The A6 Intruder

DU: 621.4 2000 0556 McGraw-Hill Series in Mechanical Engineering INTERNAL Jack P. Holman, Southern Methodist University Consulting Editor COMBUSTION Anderson: Modern Compressible Flow: With Historical Perspective ENGINE Dieter: Engineering Design: A Materials and Processing Approach Eckert and Drake: Analysis of Heat and Mass Transfer FUNDAMENTALS Heywood: Internal Combustion Engine Fundamentals Hinze: Turbulence, 2/e Hutton: Applied Mechanical Vibrations Juvinall: Engineering Considerations of Stress, Strain, and Strength John B ., Heywood Kane and Levinson: Dynamics: Theory and Applications Professor of Mechanical Engineering Kays and Crawford: Convective Heat and Mass Transfer Director, Sloan Automotive Laboratory Martin: Kinematics and Dynamics of Machines Massachusetts Institute of Technology Phelan: Dynamics of Machinery Phelan: Fundamentals of Mechanical Design, 3/e Pierce: Acoustics : An Introduction to Its Physical Principles and Applications Raven: Automatic Control Engineering, 4/e. Rosenberg and Karnopp: Introduction to Physics Schlichting: Boundary-Layer Theory, 7/e Shames: Mechanics of Fluids, 2/e Shigley: Kinematic Analysis of Mechanisms, 2/e Shigley and Mitchell: Mechanical Engineering Design, 4/e Shigley and Uicker: Theory of Machines and Mechanisms Stoecker and Jones: Refrigeration and Air Conditioning, 2/e Vanderplaats: Numerical Optimization Techniques for Engineering Design: With Applications Dauerleine an; FV/SLE4 ( Si) Änderung nur über Fachbibliothek BFV21 (Sh) McGraw-Hill, Inc. New York St. Louis San Francisco Auckland Bogotá Caracas Lisbon London .Madrid Mexico City Milan Montreal New Delhi San Juan Singapore Sydney Tokyo Toronto INTERNAL COMBUSTION ENGINE FUNDAMENTALS This book was set in Times Roman. ABOUT THE AUTHOR The editors were Anne Duffy and John M. Morriss; the designer was Joan E. O'Connor; the production supervisor was Denise L. Puryear. New drawings were done by ANCO. Project Supervision was done by Santype International Ltd. R. R. Donnelley & Sons Company was printer and binder. See acknowledgements on page xxi. Copyright C 1988 by McGraw-Hill, Inc. All rights reserved. Printed in the United States of America. Except as permitted under the United States Copyright Act of 1976, no part of this publication may be reproduced or distributed in any form or by any means, or stored in a data base or retrieval system, without the prior written permission of the publisher. 14 15 16 17 DOC/DOC 9 9 8 ISBN 0-07-028637-X Library of Congress Cataloging-in-Publication Data Heywood, John B. Internal combustion engine fundamentals. Dr. John B. Heywood received the Ph.D. degree in mechanical engineering from (McGraw-Hill series in mechanical engineering) the Massachusetts Institute of Technology in 1965. Following an additional post- Bibliography: p. Includes index. doctoral year of research at MIT, he worked as a research officer at the Central 1. Internal combustion engines. I. Title. II. Series. Electricity Generating Board's Research Laboratory in England on magneto- TJ755.H45 1988 621.43 87-15251 hydrodynamic power generation. In 1968 he joined the faculty at MIT where he is Professor of Mechanical Engineering. At MIT he is Director of the Sloan This book is printed on acid-free paper. Automotive Laboratory. He is currently Head of the Fluid and Thermal Science Division of the Mechanical Engineering Department, and the Transportation Energy Program Director in the MIT Energy Laboratory. He is faculty advisor to the MIT Sports Car Club. Professor Heywood's teaching and research interests lie in the areas of ther- modynamics, combustion, energy, power, and propulsion. During the past two decades, his research activities have centered on the operating characteristics and fuels requirements of automotive and aircraft engines. A major emphasis has been on computer models which predict the performance, efficiency, and emis- sions of spark-ignition, diesel, and gas turbine engines, and in carrying out experiments to develop and validate these models. He is also actively involved in technology assessments and policy studies related to automotive engines, auto- mobile fuel utilization, and the control of air pollution. He consults frequently in -the automotive and petroleum industries, and for the U.S. Government. His extensive research in the field of engines has been supported by the U.S. Army, Department of Energy, Environmental Protection Agency, NASA, National Science Foundation, automobile and diesel engine manufacturers, and petroleum companies. He has presented or published over a hundred papers on v vi ABOUT THE AUTHOR his research in technical conferences and journals. He has co-authored two pre- vious books: Open-Cycle MHD Power Generation published by Pergamon Press THIS BOOK IS DEDICATED TO MY FATHER, in 1969 and The Automobile and the Regulation of Its Impact on the Environment Harold Heywood: published by University of Oklahoma Press in 1975. He is a member of the American Society of Mechanical Engineers, an associ- I have followed many of the paths he took. ate fellow of the American Institute of Aeronautics and Astronautics, a fellow of the British Institution of Mechanical Engineers, and in 1982 was elected a Fellow of the U.S. Society of Automotive Engineers for his technical contributions to automotive engineering. He is a member of the editorial boards of the journals Progress in Energy and Combustion Science and the International Journal of Vehicle Design. His research publications on internal combustion engines, power generation, and gas turbine combustion have won numerous awards. He was awarded the Ayreton Premium in 1969 by the British Institution of Electrical Engineers. Pro- fessor Heywood received a Ralph R. Teetor Award as an outstanding young engineering educator from the Society of Automotive Engineers in 1971. He has twice been the recipient of an SAE Arch T. Colwell Merit Award for an outstand- ing technical publication (1973 and 1981). He received SAE's Horning Memorial Award for the best paper on engines and fuels in 1984. In 1984 he received the Sc.D. degree from Cambridge University for his published contributions to engineering research. He was selected as the 1986 American Society of Mechani- cal Engineers Freeman Scholar for a major review of "Fluid Motion within the Cylinder of Internal Combustion Engines." vii CONTENTS Preface xvii Commonly Used Symbols, Subscripts, and - Abbreviations xxiii Chapter 1 Engine Types and Their Operation T 1.1 Introduction and Historical Perspective 1 1.2 Engine Classifications 7 1.3 Engine Operating Cycles 9 1.4 Engine Components 12 1.5 Spark-Ignition Engine Operation 15 1.6 Examples of Spark-Ignition Engines 19 1.7 Compression-Ignition Engine Operation 25 1.8 Examples of Diesel Engines 31 1.9 Stratified-Charge Engines 37 Chapter 2 Engine Design and Operating Parameters 42 2.1 Important Engine Characteristics 42 2.2 Geometrical Properties of Reciprocating Engines 43 2.3 Brake Torque and Power 45 2.4 Indicated Work Per Cycle 46 2.5 Mechanical Efficiency 48 2.6 Road-Load Power 49 2.7 Mean Effective Pressure 50 2.8 Specific Fuel Consumption and Efficiency 51 2.9 Air/Fuel and Fuel/Air Ratios 53 ix X CONTENTS CONTENTS Xi 2.10 Volumetric Efficiency 53 2.11 Engine Specific Weight and Specific Volume 54 Chapter 5 Ideal Models of Engine Cycles 161 2.12 Correction Factors for Power and Volumetric Efficiency 54 5.1 Introduction 161 2.13 Specific Emissions and Emissions Index 56 5.2 Ideal Models of Engine Processes 162 2.14 Relationships between Performance Parameters 56 5.3 Thermodynamic Relations for Engine Processes 164 2.15 Engine Design and Performance Data 57 5.4 Cycle Analysis with Ideal Gas Working Fluid with c, and c, Constant 169 5.4.1 . Constant-Volume Cycle 169 Chapter 3 Thermochemistry of Fuel-Air Mixtures 62 5.4.2 Limited- and Constant-Pressure Cycles 172 5.4.3 Cycle Comparison 173 3.1 Characterization of Flames 62 5.5 Fuel-Air Cycle Analysis 177 3.2 Ideal Gas Model 64 5.5.1 SI Engine Cycle Simulation 178 3.3 Composition of Air and Fuels 64 5.5.2 CI Engine Cycle Simulation 180 3.4 Combustion Stoichiometry 68 5.5.3 Results of Cycle Calculations 181 3.5 The First Law of Thermodynamics and Combustion 72 5.6 Overexpanded Engine Cycles 183 3.5.1 Energy and Enthalpy Balances 72 5.7 Availability Analysis of Engine Processes 186 3.5.2 Enthalpies of Formation 76 5.7.1 Availability Relationships 186 3.5.3 Heating Values 78 5.7.2 Entropy Changes in Ideal Cycles 188 3.5.4 Adiabatic Combustion Processes 80 5.7.3 Availability Analysis of Ideal Cycles 189 3.5.5 Combustion Efficiency of an Internal Combustion Engine 81 5.7.4 Effect of Equivalence Ratio 192 3.6 The Second Law of Thermodynamics Applied to Combustion 83 5.8 Comparison with Real Engine Cycles 193 3.6.1 Entropy 83 3.6.2 Maximum Work from an Internal Combustion Chapter 6 Gas Exchange Processes 205 Engine and Efficiency 83 3.7 Chemically Reacting Gas Mixtures 85 6.1 Inlet and Exhaust Processes in the Four-Stroke Cycle 206 3.7.1 Chemical Equilibrium 86 6.2 Volumetric Efficiency 209 3.7.2 Chemical Reaction Rates 6.2.1 Quasi-Static Effects 209 6.2.2 Combined Quasi-Static and Dynamic Effects 212 6.2.3 Variation with Speed, and Valve Area, Lift, and Timing 216 Chapter 4 Properties of Working Fluids 100 6.3 Flow Through Valves 220 220 4.1 Introduction 100 6.3.1 Poppet Valve Geometry and Timing 4.2 Unburned Mixture Composition 102 6.3.2 Flow Rate and Discharge Coefficients 225 4.3 Gas Property Relationships 107 6.4 Residual Gas Fraction 230 4:4 A Simple Analytic Ideal Gas Model 109 6.5 Exhaust Gas Flow Rate and Temperature Variation 231 1.5 Thermodynamic Charts 112 6.6 Scavenging in Two-Stroke Cycle Engines 235 4.5.1 . Unburned Mixture Charts 112 6.6.1 Two-Stroke Engine Configurations 235 4.5.2 Burned Mixture Charts 116 6.6.2 Scavenging Parameters and Models 237 4.5.3 Relation between Unburned and Burned 6.6.3 Actual Scavenging Processes 240 Mixture Charts 123 6.7 Flow Through Ports 245 Tables of Properties and Composition 127 6.8 Supercharging and Turbocharging 248 4.6 4.7 Computer Routines for Property and Composition Calculations 130 6.8.1 . Methods of Power Boosting 248 4.7.1 Unburned Mixtures 130 6.8.2 Basic Relationships 249 4.7.2 Burned Mixtures 135 6.8.3 Compressors 255 4.8 Transport Properties 141 6.8.4 Turbines 263 4.9 Exhaust Gas Composition 145 6.8.5 Wave-Compression Devices 270 4.9.1 Species Concentration Data 145 4.9.2 Equivalence Ratio Determination from Exhaust Chapter 7 SI Engine Fuel Metering and Manifold Gas Constituents 148 Phenomena 279 4.9.3 Effects of Fuel/Air Ratio Nonuniformity 152 7.1 Spark-Ignition Engine Mixture Requirements 279 4.9.4 Combustion Inefficiency 154 7.2 Carburetors 282 xii CONTENTS CONTENTS xiii 7.2.1 Carburetor Fundamentals 282 9.6.2 Knock Fundamentals 457 7.2.2 Modern Carburetor Design 285 9.6.3 Fuel Factors 470 7.3 Fuel-Injection Systems 294 7.3.1 Multipoint Port Injection 294 7.3.2 Single-Point Throttle-Body Injection 299 Chapter 10 Combustion in Compression-Ignition Engines 491 7.4 Feedback Systems 301 10.1 Essential Features of Process 491 7.5 Flow Past Throttle Plate 304 10.2 Types of Diesel Combustion Systems 493 7.6 Flow in Intake Manifolds 308 10.2.1 Direct-Injection Systems 493. 7.6.1 Design Requirements 308 10.2.2 Indirect-Injection Systems 494 7.6.2 Air-Flow Phenomena 309 10.2.3 Comparison of Different Combustion Systems 495 7.6.3 Fuel-Flow Phenomena 314 10.3 Phenomenological Model of Compression-Ignition Engine Combustion 497 497 Chapter 8 Charge Motion within the Cylinder 10.3.1 Photographic Studies of Engine Combustion 326 10.3.2 Combustion in Direct-Injection, Multispray Systems 503 8.1 Intake Jet Flow 326 10.3.3 Application of Model to Other Combustion Systems 506 8.2 Mean Velocity and Turbulence Characteristics 330 10.4 Analysis of Cylinder Pressure Data 508 8.2.1 Definitions 330 10.4.1 Combustion Efficiency 509 8.2.2 Application to Engine Velocity Data 336 10.4.2 Direct-Injection Engines 509 8.3 Swirl 342 10.4.3 Indirect-Injection Engines 514 8.3.1 Swirl Measurement 343 10.5 Fuel Spray Behavior 517 8.3.2 Swirl Generation during Induction 345 10.5.1 Fuel Injection 517 8.3.3 Swirl Modification within the Cylinder 349 10.5.2 Overall Spray Structure 522 8.4 Squish 353 10.5.3 Atomization 525 8.5 Prechamber Engine Flows 357 10.5.4 Spray Penetration 529 8.6 Crevice Flows and Blowby 360 10.5.5 Droplet Size Distribution 532 8.7 Flows Generated by Piston-Cylinder Wall Interaction 365 10.5.6 Spray Evaporation 535 10.6 Ignition Delay 539 Chapter 9 Combustion in Spark-Ignition Engines 371 10.6.1 Definition and Discussion 539 10.6.2 541 9.1 Essential Features of Process Fuel Ignition Quality 371 542 9.2 10.6.3 Autoignition Fundamentals Thermodynamic Analysis of SI Engine Combustion 376 10.6.4 Physical Factors Affecting Delay 546 9.2.1 Burned and Unburned Mixture States 376 10.6.5 Effect of Fuel Properties 550 9.2.2 Analysis of Cylinder Pressure Data 383 10.6.6 Correlations for Ignition Delay in Engines 553 9.2.3 Combustion Process Characterization 389 Mixing-Controlled Combustion 555 9.3 10.7 Flame Structure and Speed 390 10.7.1 Background 555 9.3.1 Experimental Observations 390 10.7.2 Spray and Flame Structure 555 9.3.2 Flame Structure 395 10.7.3 Fuel-Air Mixing and Burning Rates 558 9.3.3 Laminar Burning Speeds 402 9.3.4 Flame Propagation Relations 406 9.4 Cyclic Variations in Combustion, Partial Burning, and Misfire 413 Chapter 11 Pollutant Formation and Control 567 9.4.1 Observations and Definitions 413 11.1 Nature and Extent of Problem 567 9.4.2 Causes of Cycle-by-Cycle and Cylinder-to-Cylinder 11.2 Nitrogen Oxides 572 Variations 419 11.2.1 Kinetics of NO Formation 572 9.4.3 Partial Burning, Misfire, and Engine Stability 424 11.2.2 Formation of NO2 577 9.5 Spark Ignition 427 11.2.3 NO Formation in Spark-Ignition Engines 578 9.5.1 Ignition Fundamentals 427 11.2.4 NO, Formation in Compression-Ignition Engines 586 9.5.2 Conventional Ignition Systems 437 11.3 Carbon Monoxide 592 9.5.3 Alternative Ignition Approaches 443 11.4 Unburned Hydrocarbon Emissions 596 9.6 Abnormal Combustion: Knock and Surface Ignition 450 11.4.1 Background 596 9.6.1 Description of Phenomena 450 11.4.2 Flame Quenching and Oxidation Fundamentals 599 xiv CONTENTS CONTENTS XV 11.4.3 HC Emissions from Spark-Ignition Engines 601 13.3.1 Lubricated Friction 715 11.4.4 Hydrocarbon Emission Mechanisms in Diesel Engines 620 13.3.2 Turbulent Dissipation 719 11.5 Particulate Emissions 626 13.3.3 Total Friction 719 11.5.1 Spark-Ignition Engine Particulates 626 13.4 Measurement Methods 719 11.5.2 Characteristics of Diesel Particulates 626 13.5 Engine Friction Data 722 11.5.3 Particulate Distribution within the Cylinder 631 13.5.1 SI Engines 722 11.5.4 Soot Formation Fundamentals 63,5 13.5.2 Diesel Engines 724 11.5.5 Soot Oxidation 642 13.6 Engine Friction Components 725 11.5.6 Adsorption and Condensation 646 13.6.1 Motored Engine Breakdown Tests 725 11.6 Exhaust Gas Treatment 648 13.6.2 Pumping Friction 726 11.6.1 Available Options 648 13.6.3 Piston Assembly Friction 729 11.6.2 Catalytic Converters 649 13.6.4 Crankshaft Bearing Friction 734 11.6.3 Thermal Reactors 657 13.6.5 Valve Train Friction 737 11.6.4 Particulate Traps 659 13.7 Accessory Power Requirements 739 13.8 Lubrication 740 Chapter 12 Engine Heat Transfer 668 13.8.1 Lubrication System 740 13.8.2 Lubricant Requirements 741 12.1 Importance of Heat Transfer 668 12.2 Modes of Heat Transfer 670 12.2.1 Conduction 670 Chapter 14 Modeling Real Engine Flow and Combustion 12.2.2 Convection 670 Processes 748 12.2.3 Radiation 671 12.2.4 Overall Heat-Transfer Process 671 14.1 Purpose and Classification of Models 748 12.3 14.2 Heat Transfer and Engine Energy Balance 673 Governing Equations for Open Thermodynamic System 750 12.4 Convective Heat Transfer 676 14.2.1 Conservation of Mass 750 12.4.1 Dimensional Analysis 676 14.2.2 Conservation of Energy 751 12.4.2 Correlations for Time-Averaged Heat Flux 14.3 Intake and Exhaust Flow Models 677 753 12.4.3 Correlations for Instantaneous Spatial 14.3.1 Background 753 Average Coefficients 678 14.3.2 Quasi-Steady Flow Models 753 12.4.4 Correlations for Instantaneous Local Coefficients 681 14.3.3 Filling and Emptying Methods 754 12.4.5 Intake and Exhaust System Heat Transfer 682 14.3.4 Gas Dynamic Models 756 12.5 Radiative Heat Transfer 14.4 683 Thermodynamic-Based In-Cylinder Models 762 12.5.1 Radiation from Gases 683 14.4.1 Background and Overall Model Structure 762 12.5.2 Flame Radiation 684 14.4.2 Spark-Ignition Engine Models 766 12.5.3 Prediction Formulas 688 14.4.3 Direct-Injection Engine Models 778 12.6 Measurements of Instantaneous Heat-Transfer Rates 689 14.4.4 Prechamber Engine Models 784 12.6.1 Measurement Methods 689 14.4.5 Multicylinder and Complex Engine System Models 789 12.6.2 Spark-Ignition Engine Measurements 690 14.4.6 Second Law Analysis of Engine Processes 792 12.6.3 Diesel Engine Measurements 692 14.5 Fluid-Mechanic-Based Multidimensional Models 797 12.6.4 Evaluation of Heat-Transfer Correlations 694 14.5.1 Basic Approach and Governing Equations 797 12.6.5 Boundary-Layer Behavior 14.5.2 Turbulence Models 697 800 12.7 Thermal Loading and Component Temperatures 698 14.5.3 Numerical Methodology 803 12.7.1 Component Temperature Distributions 14.5.4 Flow Field Predictions 698 807 12.7.2 Effect of Engine Variables 701 14.5.5 Fuel Spray Modeling 813 14.5.6 Combustion Modeling 816 Chapter 13 Engine Friction and Lubrication 712 13.1 Background 712 Chapter 15 Engine Operating Characteristics 823 13.2 Definitions 714 15.1 Engine Performance Parameters 823 13.3 Friction Fundamentals 715 15.2 Indicated and Brake Power and MEP 824 xvi CONTENTS 15.3 Operating Variables That Affect SI Engine Performance, Efficiency, and Emissions 827 15.3.1 Spark Timing 827 PREFACE 15.3.2 Mixture Composition 829 15.3.3 Load and Speed 839 15.3.4 Compression Ratio 841 15.4 SI Engine Combustion Chamber Design 844 15.4.1 Design Objectives and Options 844 15.4.2 Factors That Control Combustion 846 15.4.3 Factors That Control Performance 850 15.4.4 Chamber Octane Requirement 852 15.4.5 Chamber Optimization Strategy 857 15.5 Variables That Affect CI Engine Performance, Efficiency, and Emissions 858 15.5.1 Load and Speed 858 15.5.2 Fuel-Injection Parameters 863 15.5.3 Air Swirl and Bowl-in-Piston Design 866 15.6 Supercharged and Turbocharged Engine Performance 869 15.6.1 Four-Stroke Cycle SI Engines 869 15.6.2 Four-Stroke Cycle CI Engines 874 15.6.3 Two-Stroke Cycle SI Engines 881 Internal combustion engines date back to 1876 when Otto first developed the 15.6.4 Two-Stroke Cycle CI Engines 883 Engine Performance Summary 886 spark-ignition engine and 1892 when Diesel invented the compression-ignition 15.7 engine. Since that time these engines have continued to develop as our knowledge Appendixes of engine processes has increased, as new technologies became available, as demand for new types of engine arose, and as environmental constraints on A Unit Conversion Factors 899 engine use changed. Internal combustion engines, and the industries that develop B Ideal Gas Relationships 902 and manufacture them and support their use, now play a dominant role in the B.1 Ideal Gas Law 902 fields of power, propulsion, and energy. The last twenty-five years or so have seen B.2 The Mole 903 B.3 Thermodynamic Properties 903 an explosive growth in engine research and development as the issues of air pol- B.4 Mixtures of Ideal Gases 905 lution, fuel cost, and market competitiveness have become increasingly impor- C Equations for Fluid Flow through a Restriction 906 tant. An enormous technical literature on engines now exists which has yet to be C.1 Liquid Flow 907 adequately organized and summarized. C.2 Gas Flow 907 This book has been written as a text and a professional reference in response D Data on Working Fluids 911 to that need. It contains a broadly based and extensive review of the fundamental principles which govern internal combustion engine design and operation. It Index 917 attempts to provide a simplifying framework for the vast and complex mass of technical material that now exists on spark-ignition and compression-ignition engines, and at the same time to include sufficient detail to convey the real world dimensions of this pragmatic engineering field. It is the author's conviction that a sound knowledge of the relevant fundamentals in the many disciplines that con- tribute to this field, as well as an awareness of the extensive practical knowledge base which has been built up over many decades, are essential tools for engine research, development, and design. Of course, no one text can include everything about engines. The emphasis here is on the thermodynamics, combustion physics and chemistry, fluid flow, heat transfer, friction, and lubrication processes rele- vant to internal combustion engine design, performance, efficiency, emissions, and fuels requirements. xvii xviii PREFACE PREFACE xix From a fundamental point of view, how the fuel-air mixture within an inter- ating variables. These final two chapters effectively integrate the analytical under- nal combustion engine cylinder is ignited appropriately organizes the field. From standing and practical knowledge of individual engine processes together to the method of ignition-spark-ignition or compression-ignition-follows each describe overall spark-ignition and compression-ignition engine behavior. type of engine's important features: fuel requirements, method of mixture prep- Material on internal combustion engine fuels is distributed appropriately aration, combustion chamber design, details of the combustion process, method throughout the book. Each chapter is extensively illustrated and referenced, and of load control, emission formation mechanisms, and performance and efficiency includes problems for both undergraduate and graduate level courses. characteristics. While many engine processes (such as intake and exhaust flows, While this book contains much advanced material on engine design and convective heat transfer, and friction) are similar in both types of engines, this operation intended for the practitioner, each major topic is developed from its distinction is fundamental and lies behind the overall organization of the book. beginnings and the more sophisticated chapters have introductory sections to The book is arranged in four major sections. The first (Chapters 1 to 5) facilitate their use in undergraduate courses. The chapters are extensively cross- provides an introduction to, and overview of, the major characteristics of spark- referenced and indexed. Thus several arrangements of the material for a course ignition and compression-ignition engines, defines the parameters used to on engines can be followed. For example, an introductory course on internal describe engine operation, and develops the necessary thermodynamics and com- combustion engines could begin with Chapters 1 and 2, which review the differ- bustion theory required for a quantitative analysis of engine behavior. It con- ent types of engines and how their performance is characterized, and continue cludes with an integrated treatment of the various methods of analyzing idealized with the parts of Chapters 3 and 5, which introduce the key combustion concepts models of internal combustion engine cycles. The second section (Chapters 6 to 8) necessary to understand the effects of fuel/air ratio, and ideal cycle analysis. Se- focuses on engine flow phenomena. The details of the gas exchange process- lections from the introductory sections of Chapters 6, 9, 10, 11, and 15 could then intake and exhaust processes in four-stroke and scavenging in two-stroke be used to explain several of the practical and design aspects of spark-ignition cycles-and the various methods of supercharging engines -- are reviewed. Fuel and diesel engine intake and exhaust processes, combustion, emissions, and per- metering methods for spark-ignition engines and air- and fuel-flow phenomena in formance. A more advanced course would review this introductory material more intake manifolds are described. The essential features of the various types of fluid rapidly, and then move on to those sections of Chapters 4 and 5, which cover motion within the engine cylinder are then developed. These flow processes fuel-air cycle analysis, a more extensive discussion of engine breathing using addi- control the amount of air an engine will induct (and therefore its power), and tional sections of Chapter 6, and more in-depth treatment of engine combustion largely govern the rate at which the fuel-air mixture will burn during combustion. and emissions processes based on the appropriate sections of Chapters 9, 10, and The third section of the book focuses on engine combustion phenomena. 11. Material on engine heat transfer and friction selected from Chapters 12 and These chapters (9, 10, and 11) are especially important. The combustion process 13 could be included next. While Chapter 14 on modeling the thermodynamics releases the fuel's energy within the engine cylinder for eventual conversion to and fluid dynamics of real engine processes is primarily intended for the pro- useful work. What fraction of the fuel's energy is converted depends strongly on fessional scientist and engineer, material from this chapter along with selections how combustion takes place. The spark-ignition and compression-ignition engine from Chapter 15 could be used to illustrate the performance, efficiency, and emis- combustion processes (Chapters 9 and 10, respectively) therefore influence essen- sions characteristics of the different types of internal combustion engines. I have tially all aspects of engine behavior. Air pollutants are undesirable byproducts of also used much of the more sophisticated material in Chapters 6 through 15 for combustion. Our extensive knowledge of how the major pollutants form during review seminars on individual engine topics and more extensive courses for pro- these combustion processes and how such emissions can be controlled is fessional engineers, an additional important educational and reference reviewed in Chapter 11. opportunity. The last section of the book focuses on engine operating characteristics. First, Many individuals and organizations have assisted me in various ways as I the fundamentals of engine heat transfer and friction, both of which detract from have worked on this book over the past ten or so years. I am especially indebted engine performance, are developed in Chapters 12 and 13. Chapter 14 then to my colleagues in the Sloan Automotive Laboratory at M.I.T ., Professors Wai focuses on the methods available for predicting important aspects of engine K. Cheng, Ahmed F. Ghoniem, and James C. Keck, and Drs. Jack A. Ekchian, behavior based on realistic models of engine flow and combustion processes. David P. Hoult, Joe M. Rife, and Victor W. Wong, for providing a stimulating Since the various thermodynamic-based and fluid-mechanic-based models which environment in which to carry out engine research and for assuming additional have been developed over the past fifteen years or so are increasingly used in burdens as a result of my writing. Many of the Sloan Automotive Laboratory's engine research and development, a knowledge of their basic structure and capa- students have made significant contributions to this text through their research; bilities is most important. Then, Chapter 15 presents a summary of how the their names appear in the reference lists. The U.S. Department of Energy provid- operating characteristics-power, efficiency, and emissions-of spark-ignition ed support during the early stages of the text development and funded the work and compression-ignition engines depend on the major engine design and oper- on engine cycle simulation used extensively in Chapters 14 and 15. I am grateful XX PREFACE to Churchill College, Cambridge University, for a year spent as a Richard C. Mellon Visiting Fellow, 1977-78, and the Engineering Department, Cambridge University, for acting as my host while I developed the outline and earlier chap- ACKNOWLEDGMENTS ters of the book. The M.I.T. sabbatical leave fund supported my full-time writing for eight months in 1983, and the Mechanical Engineering Department at Imperial College graciously acted as host. I also want to acknowledge several individuals and organizations who have provided major inputs to this book beyond those cited in the references. Members of General Motors Research Laboratories have interacted extensively with the Sloan Automotive Laboratory over many years and provided valuable advice on engine research developments. Engineers from the Engine Research and Fluid Mechanics Departments at General Motors Research Laboratories reviewed and critiqued the final draft manuscript for me. Charles A. Amann, Head of the Engine Research Department, made especially helpful inputs on engine performance. John J. Brogan of the U.S. Department of Energy provided valuable assistance with the initial organization of this effort. My regular inter- actions over the years with the Advanced Powertrain Engineering Office and Scientific Research Laboratories of the Ford Motor Company have given me a broad exposure to the practical side of engine design and operation. A long-term The author wishes to acknowledge the following organizations and publishers relationship with Mobil Research and Development Corporation has provided for permission to reproduce figures and tables from their publications in this comparable experiences in the area of engine-fuels interactions. Many organi- text: The American Chemical Society; American Institute of Aeronautics & zations and individuals supplied specific material and illustrations for the text. I Astronautics; American Society of Mechanical Engineers; Robert Bosch GmbH, am especially grateful to those who made available the high-quality photographs CIMAC, Cambridge University Press; The Combustion Institute; Elsevier and line drawings which I have used and acknowledged. Science Publishing Company; G. T. Foulis & Co. Ltd.; General Motors Corpo- McGraw-Hill and the author would like to express their thanks to the fol- ration; Gordon & Breach Science Publishers; The Institution of Mechanical lowing reviewers for their useful comments and suggestions: Jay A. Bolt, Uni- Engineers; The Japan Society of Mechanical Engineers; M.I.T. Press; Macmil- versity of Michigan; Gary L. Borman and William L. Brown, University of lan Press Ltd.; McGraw-Hill Book Company; Mir Publishers; Mobil Oil Corpo- Wisconsin at Madison; Dwight Bushnell, Oregon State University; Jerald A. ration; Morgan-Grampian Publishers; Pergamon Journals, Inc.; Plenum Press Caton, Texas A & M University; David E. Cole, University of Michigan; Law- Corporation; The Royal Society of London; Scientific Publications Limited; rence W. Evers, Michigan Technological University; Samuel S. Lestz, Pennsylva Society of Automotive Engineers; Society of Automotive Engineers of Japan, nia State University; Willard Pulkrabek, University of Wisconsin; Robert F. Inc.; Society of Tribologists and Lubrications Engineers; Department of Mecha- Sawyer, University of California at Berkeley; Joseph E. Shepherd, Rensselaer nical Engineering, Stanford University. Polytechnic Institute; and Spencer C. Sorenson, The Technical University of Denmark. Special thanks are due to my secretaries for their faithful and thoughtful assistance with the manuscript over these many years, beyond the "call of duty"; Linda Pope typed an earlier draft of the book, and Karla Stryker was responsible for producing and coordinating subsequent drafts and the final manuscript. My wife Peggy, and sons James, Stephen, and Ben have encouraged me throughout this long and time-consuming project which took many hours away from them. Without their continuing support it would never have been finished; for their patience, and faith that it would ultimately come to fruition, I will always be grateful. John B. Heywood xxi COMMONLY USED SYMBOLS, SUBSCRIPTS, AND ABBREVIATIONS+ 1. SYMBOLS a Crank radius Sound speed Specific availability a Acceleration A Area Ac Valve curtain area Ach Cylinder head area Ae Exhaust port area AE Effective area of flow restriction Inlet port area AD Piston crown area B Cylinder bore Steady-flow availability C Specific heat Cp Specific heat at constant pressure Cs Soot concentration (mass/volume) Specific heat at constant volume C Absolute gas velocity + Nomenclature specific to a section or chapter is defined in that section or chapter. xxiii xxiv COMMONLY USED SYMBOLS, SUBSCRIPTS, AND ABBREVIATIONS COMMONLY USED SYMBOLS, SUBSCRIPTS, AND ABBREVIATIONS XXV Number of moles = Cs Swirl coefficient CD Discharge coefficient Polytropic exponent Vehicle drag coefficient Number of crank revolutions per power stroke Diameter N Crankshaft rotational speed Fuel-injection-nozzle orifice diameter Soot particle number density dn D Diameter Turbocharger shaft speed Diffusion coefficient P Cylinder pressure Pressure Da Droplet diameter Sauter mean droplet diameter P Power DSM D. Valve diameter Heat-transfer rate per unit area e Radiative emissive power Heat-transfer rate per unit mass of fluid Heat transfer Specific energy Heat-transfer rate EA Activation energy f Coefficient of friction Qch Fuel chemical energy release or gross heat release Fuel mass fraction QHV Fuel heating value Net heat release F Force Radius Gravitational acceleration r Specific Gibbs free energy Compression ratio R Gibbs free energy Connecting rod length/crank radius Gas constant h Clearance height Radius Oil film thickness Specific enthalpy R+, R- One-way reaction rates Swirl ratio hc Heat-transfer coefficient S Port open height Crank axis to piston pin distance hp Specific entropy hs Sensible specific enthalpy S Entropy H Enthalpy Moment of inertia Spray penetration Turbulent burning speed J Flux k Thermal conductivity SL Laminar flame speed Turbulent kinetic energy .Sp Piston speed Forward, backward, rate constants for ith reaction Time K Constant Temperature K. Equilibrium constant expressed in concentrations Torque u Kp Equilibrium constant expressed in partial pressures Specific internal energy Characteristic length scale Velocity - Connecting rod length Turbulence intensity Sensible specific internal energy IT Characteristic length scale of turbulent flame L Piston stroke UT Characteristic turbulent velocity U Fuel-injection-nozzle orifice length Compressor/turbine impellor tangential velocity Ln Fluid velocity Valve lift Internal energy m Mass im Mass flow rate Specific volume Velocity m, Mass of residual gas V Velocity M Mach number Molecular weight Ups Valve pseudo-flow velocity XXVi COMMONLY USED SYMBOLS, SUBSCRIPTS, AND ABBREVIATIONS COMMONLY USED SYMBOLS, SUBSCRIPTS, AND ABBREVIATIONS xxvii Usa Squish velocity Dynamic viscosity V Cylinder volume Chemical potential of species i Volume Kinematic viscosity u/p Vc Clearance volume Stoichiometric coefficient of species i Va Displaced cylinder volume Flow friction coefficient W Relative gas velocity 0 Density Soot surface oxidation rate Pa.o , Pa.i Air density at standard, inlet conditions W Work transfer Normal stress Work per cycle Standard deviation W. Pumping work Stefan-Boltzmann constant x, y, z Spatial coordinates Surface tension x Mass fraction T Characteristic time Mole fraction Induction time Xb Burned mass fraction Shear stress Residual mass fraction Vid Ignition delay time H/C ratio of fuel Fuel/air equivalence ratio Volume fraction Flow compressibility function [Eq. (C.11)] Ya Concentration of species a per unit mass Isentropic compression function [Eq. (4.15b)] Z Inlet Mach index Molar N/O ratio Angle Throttle plate open angle Thermal diffusivity k/(pc) Isentropic compression function [Eq. (4.15a)] B Angle Angular velocity Specific heat ratio cp/c. Frequency Angular momentum of charge Boundary-layer thickness 2. SUBSCRIPTS Laminar flame thickness Ahr.i Molal enthalpy of formation of species i a Air Rapid burning angle b Burned gas Flame development angle C Coolant 3 4/(4 + y): y = H/C ratio of fuel Cylinder Turbulent kinetic energy dissipation rate C Compression stroke na Availability conversion efficiency Compressor nc Combustion efficiency cr Crevice nc Compressor isentropic efficiency e Equilibrium .. nct Charging efficiency Exhaust Fuel conversion efficiency E Expansion stroke 1m Mechanical efficiency f Flame ist Scavenging efficiency Friction n Thermal conversion efficiency Fuel nT Turbine isentropic efficiency Gas Trapping efficiency i Indicated Volumetric efficiency Intake no Crank angle Species i Relative air/fuel ratio ig Gross indicated V Delivery ratio in Net indicated xxviii COMMONLY USED SYMBOLS, SUBSCRIPTS, AND ABBREVIATIONS COMMONLY USED SYMBOLS, SUBSCRIPTS, AND ABBREVIATIONS XXix - Liquid ON Fuel octane number ... L Laminar Re Reynolds number pul/u Piston sfc Specific fuel consumption Port TC, ATC, BTC Top-center crank position, after TC, before TC P Prechamber We Weber number p1u2D/o r, 0, z r, 0, z components R Reference value S Isentropic Stoichiometric T Nozzle or orifice throat Turbine Turbulent n Unburned Valve w Wall x, y, z x, y, z components 0 Reference value Stagnation value 3. NOTATION A Difference Average or mean value Value per mole [] Concentration, moles/vol Mass fraction Rate of change with time 4. ABBREVIATIONS (A/F) Air/fuel ratio BC, ABC, BBC Bottom-center crank position, after BC, before BC CN Fuel cetane number Da Damköhler number tr/tL EGR Exhaust gas recycle EI Emission index EPC, EPO Exhaust port closing, opening EVC, EVO Exhaust valve closing, opening (F/A) Fuel/air ratio (G/F) Gas/fuel ratio IPC, IPO Inlet port closing, opening IVC, IVO Inlet valve closing, opening mep Mean effective pressure Nu Nusselt number he l/k CHAPTER ENGINE TYPES AND THEIR OPERATION 1.1 INTRODUCTION AND HISTORICAL PERSPECTIVE The purpose of internal combustion engines is the production of mechanical power from the chemical energy contained in the fuel. In internal combustion engines, as distinct from external combustion engines, this energy is released by burning or oxidizing the fuel inside the engine. The fuel-air mixture before com- bustion and the burned products after combustion are the actual working fluids. The work transfers which provide the desired power output occur directly between these working fluids and the mechanical components of the engine. The internal combustion engines which are the subject of this book are spark-ignition engines (sometimes called Otto engines, or gasoline or petrol engines, though other fuels can be used) and compression-ignition or diesel engines.+ Because of their simplicity, ruggedness and high power/weight ratio, these two types of engine have found wide application in transportation (land, sea, and air) and power generation. It is the fact that combustion takes place inside the work- + The gas turbine is also, by this definition, an "internal combustion engine." Conventionally, however, the term is used for spark-ignition and compression-ignition engines. The operating prin- ciples of gas turbines are fundamentally different, and they are not discussed as separate engines in this book. - 2 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE TYPES AND THEIR OPERATION 3 producing part of these engines that makes their design and operating character- TABLE 1.1 istics fundamentally different from those of other types of engine. Comparison of Otto four-stroke cycle and Otto-Langen Practical heat engines have served mankind for over two and a half cen- engines2 turies. For the first 150 years, water, raised to steam, was interposed between the Otto and Langen Otto four-stroke combustion gases produced by burning the fuel and the work-producing piston- in-cylinder expander. It was not until the 1860s that the internal combustion Brake horsepower 2 2 4000 1250 engine became a practical reality.1 2 The early engines developed for commercial Weight, lb, approx. Piston displacement, in3 4900 310 use burned coal-gas air mixtures at atmospheric pressure-there was no com- Power strokes per min 28 80 pression before combustion. J. J. E. Lenoir (1822-1900) developed the first mar- Shaft speed, rev/min 90 160 ketable engine of this type. Gas and air were drawn into the cylinder during the Mechanical efficiency, % 68 84 first half of the piston stroke. The charge was then ignited with a spark, the Overall efficiency, % 11 14 Expansion ratio 10 2.5 pressure increased, and the burned gases then delivered power to the piston for the second half of the stroke. The cycle was completed with an exhaust stroke. Some 5000 of these engines were built between 1860 and 1865 in sizes up to six horsepower. Efficiency was at best about 5 percent. A more successful development-an atmospheric engine introduced in 1867 3. The greatest possible expansion ratio by Nicolaus A. Otto (1832-1891) and Eugen Langen (1833-1895)-used the pres- 4. The greatest possible pressure at the beginning of expansion sure rise resulting from combustion of the fuel-air charge early in the outward stroke to accelerate a free piston and rack assembly so its momentum would The first two conditions hold heat losses from the charge to a minimum. The generate a vacuum in the cylinder. Atmospheric pressure then pushed the piston third condition recognizes that the greater the expansion of the postcombustion inward, with the rack engaged through a roller clutch to the output shaft. Pro- gases, the greater the work extracted. The fourth condition recognizes that higher duction engines, of which about 5000 were built, obtained thermal efficiencies of initial pressures make greater expansion possible, and give higher pressures up to 11 percent. A slide valve controlled intake, ignition by a gas flame, and throughout the process, both resulting in greater work transfer. Although Beau exhaust. de Rochas' unpublished writings predate Otto's developments, he never reduced To overcome this engine's shortcomings of low thermal efficiency and these ideas to practice. Thus Otto, in the broader sense, was the inventor of the excessive weight, Otto proposed an engine cycle with four piston strokes: an modern internal combustion engine as we know it today. intake stroke, then a compression stroke before ignition, an expansion or power Further developments followed fast once the full impact of what Otto had stroke where work was delivered to the crankshaft, and finally an exhaust stroke. achieved became apparent. By the 1880s several engineers (e.g ., Dugald Clerk, He also proposed incorporating a stratified-charge induction system, though this 1854-1913, and James Robson, 1833-1913, in England and Karl Benz, 1844- was not achieved in practice. His prototype four-stroke engine first ran in 1876. A 1929, in Germany) had successfully developed two-stroke internal combustion comparison between the Otto engine and its atmospheric-type predecessor indi- engines where the exhaust and intake processes occur during the end of the cates the reason for its success (see Table 1.1): the enormous reduction in engine power stroke and the beginning of the compression stroke. James Atkinson weight and volume. This was the breakthrough that effectively founded the inter- (1846-1914) in England made an engine with a longer expansion than compres- nal combustion engine industry. By 1890, almost 50,000 of these engines had sion stroke, which had a high efficiency for the times but mechanical weaknesses. been sold in Europe and the United States. It was recognized that efficiency was a direct function of expansion ratio, yet In 1884, an unpublished French patent issued in 1862 to Alphonse Beau de compression ratios were limited to less than four if serious knock problems were Rochas (1815-1893) was found which described the principles of the four-stroke to be avoided with the available fuels. Substantial carburetor and ignition system cycle. This chance discovery cast doubt on the validity of Otto's own patent for developments were required, and occurred, before high-speed gasoline engines this concept, and in Germany it was declared invalid. Beau de Rochas also out- suitable for automobiles became available in the late 1880s. Stationary engine lined the conditions under which maximum efficiency in an internal combustion progress also continued. By the late 1890s, large single-cylinder engines of 1.3-m engine could be achieved. These were: bore fueled by low-energy blast furnace gas produced 600 bhp at 90 rev/min. In Britain, legal restrictions on volatile fuels turned their engine builders toward keroserie. Low compression ratio "oil" engines with heated external fuel vapor- 1. The largest possible cylinder volume with the minimum boundary surface izers and electric ignition were developed with efficiencies comparable to those of 2. The greatest possible working speed gas engines (14 to 18 percent). The Hornsby-Ackroyd engine became the most 4 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE TYPES AND THEIR OPERATION 5 popular oil engine in Britain, and was also built in large numbers in the United During the past three decades, new factors for change have become impor- States.2 tant and now significantly affect engine design and operation. These factors are, In 1892, the German engineer Rudolf Diesel (1858-1913) outlined in his first, the need to control the automotive contribution to urban air pollution and, patent a new form of internal combustion engine. His concept of initiating com- second, the need to achieve significant improvements in automotive fuel con- bustion by injecting a liquid fuel into air heated solely by compression permitted sumption. a doubling of efficiency over other internal combustion engines. Much greater The automotive air-pollution problem became apparent in the 1940s in the expansion ratios, without detonation or knock, were now possible. However, Los Angeles basin. In 1952, it was demonstrated by Prof. A. J. Haagen-Smit that even with the efforts of Diesel and the resources of M.A.N. in Ausburg combined, the smog problem there resulted from reactions between oxides of nitrogen and it took five years to develop a practical engine. hydrocarbon compounds in the presence of sunlight.8 In due course it became Engine developments, perhaps less fundamental but nonetheless important clear that the automobile was a major contributor to hydrocarbon and oxides of to the steadily widening internal combustion engine markets, have continued ever nitrogen emissions, as well as the prime cause of high carbon monoxide levels in since.2-4 One more recent major development has been the rotary internal com- urban areas. Diesel engines are a significant source of small soot or smoke par- bustion engine. Although a wide variety of experimental rotary engines have been ticles, as well as hydrocarbons and oxides of nitrogen. Table 1.2 outlines the proposed over the years,5 the first practical rotary internal combustion engine, dimensions of the problem. As a result of these developments, emission standards the Wankel, was not successfully tested until 1957. That engine, which evolved for automobiles were introduced first in California, then nationwide in the through many years of research and development, was based on the designs of United States, starting in the early 1960s. Emission standards in Japan and the German inventor Felix Wankel.6, 7 Europe, and for other engine applications, have followed. Substantial reductions Fuels have also had a major impact on engine development. The earliest in emissions from spark-ignition and diesel engines have been achieved. Both the engines used for generating mechanical power burned gas. Gasoline, and lighter use of catalysts in spark-ignition engine exhaust systems for emissions control fractions of crude oil, became available in the late 1800s and various types of and concern over the toxicity of lead antiknock additives have resulted in the carburetors were developed to vaporize the fuel and mix it with air. Before 1905 reappearance of unleaded gasoline as a major part of the automotive fuels there were few problems with gasoline; though compression ratios were low (4 or market. Also, the maximum lead content in leaded gasoline has been substan- less) to avoid knock, the highly volatile fuel made starting easy and gave good tially reduced. The emission-control requirements and these fuel developments cold weather performance. However, a serious crude oil shortage developed, and have produced significant changes in the way internal combustion engines are to meet the fivefold increase in gasoline demand between 1907 and 1915, the yield designed and operated. from crude had to be raised. Through the work of William Burton (1865-1954) Internal combustion engines are also an important source of noise. There and his associates of Standard Oil of Indiana, a thermal cracking process was are several sources of engine noise: the exhaust system, the intake system, the fan developed whereby heavier oils were heated under pressure and decomposed into used for cooling, and the engine block surface. The noise may be generated by less complex more volatile compounds. These thermally cracked gasolines satis- aerodynamic effects, may be due to forces that result from the combustion fied demand, but their higher boiling point range created cold weather starting process, or may result from mechanical excitation by rotating or reciprocating problems. Fortunately, electrically driven starters, introduced in 1912, came engine components. Vehicle noise legislation to reduce emissions to the along just in time. environment was first introduced in the early 1970s. On the farm, kerosene was the logical fuel for internal combustion engines During the 1970s the price of crude petroleum rose rapidly to several times since it was used for heat and light. Many early farm engines had heated carbu- its cost (in real terms) in 1970, and concern built up regarding the longer-term retors or vaporizers to enable them to operate with such a fuel. availability of petroleum. Pressures for substantial improvements in internal The period following World War I saw a tremendous advance in our combustion engine efficiency (in all its many applications) have become very sub- understanding of how fuels affect combustion, and especially the problem of stantial indeed. Yet emission-control requirements have made improving engine knock. The antiknock effect of tetraethyl lead was discovered at General fuel consumption more difficult, and the removal and reduction of lead in gas- Motors,4 and it became commercially available as a gasoline additive in the oline has forced spark-ignition engine compression ratios to be reduced. Much United States in 1923. In the late 1930s, Eugene Houdry found that vaporized work is being done on the use of alternative fuels to gasoline and diesel. Of the oils passed over an activated catalyst at 450 to 480ºC were converted to high- non-petroleum-based fuels, natural gas, and methanol and ethanol (methyl and quality gasoline in much higher yields than was possible with thermal cracking. ethyl alcohols) are receiving the greatest attention, while synthetic gasoline and These advances, and others, permitted fuels with better and better antiknock diesel made from shale oil or coal, and hydrogen could be longer-term pos- properties to be produced in large quantities; thus engine compression ratios sibilities. steadily increased, improving power and efficiency. It might be thought that after over a century of development, the internal 6 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE TYPES AND THEIR OPERATION 7 TABLE 1.2 present a formidable challenge to automotive engineers, they will be made pos- The automotive urban air-pollution problem sible in large part by the enormous expansion of our knowledge of engine pro- Automobile emissions Truck emissions++ cesses which the last twenty years has witnessed. Mobile source Reduction emissions Uncontrolled in new SI as % of vehicles, 1.2 ENGINE CLASSIFICATIONS vehicles, engines, Diesel, Pollutant Impact totalt g/kmt % T g/km g/km There are many different types of internal combustion engines. They can be clas- Oxides of Reactant in 40-60 2.5 75 7 sified by: 12 nitrogen photochemical (NO and NO2) smog; NO, is 1. Application. Automobile, truck, locomotive, light aircraft, marine, portable toxic power system, power generation Carbon Toxic 90 65 95 150 17 monoxide 2. Basic engine design. Reciprocating engines (in turn subdivided by arrange- (CO) ment of cylinders: e.g ., in-line, V, radial, opposed), rotary engines (Wankel Unburned Reactant in 30-5 10 90 17++ 3 and other geometries) hydrocarbons photochemical 3. Working cycle. Four-stroke cycle: naturally aspirated (admitting atmospheric (HC, many smog hydrocarbon air), supercharged (admitting precompressed fresh mixture), and turbo- compounds) charged (admitting fresh mixture compressed in a compressor driven by an Particulates Reduces 50 0.56 409 n 0.5 exhaust turbine), two-stroke cycle: crankcase scavenged, supercharged, and (soot and visibility; turbocharged absorbed some of HC 4. Valve or port design and location. Overhead (or I-head) valves, underhead (or hydrocarbon compounds compounds) mutagenic L-head) valves, rotary valves, cross-scavenged porting (inlet and exhaust ports on opposite sides of cylinder at one end), loop-scavenged porting (inlet + Depends on type of urban area and source mix. and exhaust ports on same side of cylinder at one end), through- or uniflow- $ Average values for pre-1968 automobiles which had no emission controls, determined by U.S. test procedure scavenged (inlet and exhaust ports or valves at different ends of cylinder) which simulates typical urban and highway driving. Exhaust emissions, except for HC where 55 percent are exhaust emissions, 20 percent are evaporative emissions from fuel tank and carburetor, and 25 percent are crankcase 5. Fuel. Gasoline (or petrol), fuel oil (or diesel fuel), natural gas, liquid pet- blowby gases. roleum gas, alcohols (methanol, ethanol), hydrogen, dual fuel $ Diesel engine automobiles only. Particulate emissions from spark-ignition engines are negligible. 6. Method of mixture preparation. Carburetion, fuel injection into the intake 1 Compares emissions from new spark-ignition engine automobiles with uncontrolled automobile levels in previous column. Varies from country to country. The United States, Canada, Western Europe, and Japan have standards ports or intake manifold, fuel injection into the engine cylinder with different degrees of severity. The United States, Europe, and Japan have different test procedures. Standards 7. Method of ignition. Spark ignition (in conventional engines where the mixture are strictest in the United States and Japan. is uniform and in stratified-charge engines where the mixture is non-uniform), ++ Representative average emission levels for trucks. ## With 95 percent exhaust emissions and 5 percent evaporative emissions. compression ignition (in conventional diesels, as well as ignition in gas n = negligible. engines by pilot injection of fuel oil) 8. Combustion chamber design. Open chamber (many designs: e.g ., disc, wedge, combustion engine has reached its peak and little potential for further improve- hemisphere, bowl-in-piston), divided chamber (small and large auxiliary ment remains. Such is not the case. Conventional spark-ignition and diesel chambers; many designs: e.g ., swirl chambers, prechambers) engines continue to show substantial improvements in efficiency, power, and 9. Method of load control. Throttling of fuel and air flow together so mixture degree of emission control. New materials now becoming available offer the pos- composition is essentially unchanged, control of fuel flow alone, a com- sibilities of reduced engine weight, cost, and heat losses, and of different and more bination of these efficient internal combustion engine systems. Alternative types of internal com- 10. Method of cooling. Water cooled, air cooled, uncooled (other than by natural bustion engines, such as the stratified-charge (which combines characteristics nor- convection and radiation) mally associated with either the spark-ignition or diesel) with its wider fuel tolerance, may become sufficiently attractive to reach large-scale production. The All these distinctions are important and they illustrate the breadth of engine engine development opportunities of the future are substantial. While they designs available. Because this book approaches the operating and emissions 8 ENGINE TYPES AND THEIR OPERATION INTERNAL COMBUSTION ENGINE FUNDAMENTALS TABLE 1.3 engines, the predominant type of engine used in each classification listed, and the Classification of reciprocating engines by application approximate engine power range in each type of service. Approximate Predominant type engine power 1.3 ENGINE OPERATING CYCLES Class Service range, kW D or SI Cycle Cooling Most of this book is about reciprocating engines, where the piston moves back Road vehicles Motorcycles, scooters 0.75-70 SI 2, A and forth in a cylinder and transmits power through a connecting rod and crank Small passenger cars 15-75 SI 4 A. W mechanism to the drive shaft as shown in Fig. 1-1. The steady rotation of the Large passenger cars 75-200 SI 4 W Light commercial 35-150 SI, D 4 W crank produces a cyclical piston motion. The piston comes to rest at the top- Heavy (long-distance) 120-400 D 4 W center (TC) crank position and bottom-center (BC) crank position when the commercial cylinder volume is a minimum or maximum, respectively.+ The minimum cylin- Off-road vehicles Light vehicles (factory, 1.5-15 SI 2, 4 A, W der volume is called the clearance volume V. The volume swept out by the airport, etc.) Agricultural 3-150 SI, D 2, 4 A, W Earth moving 40-750 D 2, 4 Military 40-2000 D 2, 4 A, W + These crank positions are also referred to as top-dead-center (TDC) and bottom-dead-center Railroad Rail cars 150-400 D 2, 4 W (BDC). Locomotives 400-3000 D 2, 4 W Marine Outboard 0.4-75 SI W Inboard motorcrafts 4-750 SI, D W Light naval craft 30-2200 D 2, 4 W Ships 3500-22,000 D 2, 4 W Ships' auxiliaries 75-750 D 4 W Airborne Airplanes 45-2700 SI 4 A vehicles Helicopters 45-1500 SI 4 A Vc TC Home use Lawn mowers 0.7-3 SI 2, 4 A Bore - Snow blowers 2-5 SI 2, 4 A Light tractors 2-8 SI 4 A V. Stationary Building service 7-400 D 2, 4 W Stroke Electric power 35-22,000 D 2, 4 W Gas pipeline 750-5000 SI 2, 4 W O SI = spark-ignition; D = diesel; A = air cooled; W = water cooled. BC Source: Adapted from Taylor.9 characteristics of internal combustion engines from a fundamental point of view, the method of ignition has been selected as the primary classifying feature. From the method of ignition-spark-ignition or compression-ignitiont-follow the important characteristics of the fuel used, method of mixture preparation, com- bustion chamber design, method of load control, details of the combustion process, engine emissions, and operating characteristics. Some of the other classi- fications are used as subcategories within this basic classification. The engine operating cycle-four-stroke or two-stroke-is next in importance; the principles of these two cycles are described in the following section. 270º --- 90º Table 1.3 shows the most common applications of internal combustion FIGURE 1-1 Basic geometry of the reciprocating internal com- 180º bustion engine. Ve, V, and V, indicate clearance, BC displaced, and total cylinder volumes. + In the remainder of the book, these terms will often be abbreviated by SI and CI, respectively. 10 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE TYPES AND THEIR OPERATION 11 Inlet Exhaust Inlet Exhaust Inlet Exhaust Inlet Exhaust As the piston approaches BC the exhaust valve opens to initiate the exhaust process and drop the cylinder pressure to close to the exhaust pressure. 4. An exhaust stroke, where the remaining burned gases exit the cylinder: first, because the cylinder pressure may be substantially higher than the exhaust TC pressure: then as they are swept out by the piston as it moves toward TC. As the piston approaches TC the inlet valve opens. Just after TC the exhaust O valve closes and the cycle starts again. BC Though often called the Otto cycle after its inventor, Nicolaus Otto, who built O the first engine operating on these principles in 1876, the more descriptive four- stroke nomenclature is preferred. The four-stroke cycle requires, for each engine cylinder, two crankshaft rev- olutions for each power stroke. To obtain a higher power output from a given engine size, and a simpler valve design, the two-stroke cycle was developed. The two-stroke cycle is applicable to both SI and CI engines. Figure 1-3 shows one of the simplest types of two-stroke engine designs. Ports in the cylinder liner, opened and closed by the piston motion, control the exhaust and inlet flows while the piston is close to BC. The two strokes are: (a) Intake (b) Compression (c) Expansion (d) Exhaust 1. A compression stroke, which starts by closing the inlet and exhaust ports, and FIGURE 1-2 The four-stroke operating cycle.10 then compresses the cylinder contents and draws fresh charge into the crank- casc. As the piston approaches TC, combustion is initiated. piston, the difference between the maximum or total volume V, and the clearance volume, is called the displaced or swept volume V4. The ratio of maximum volume to minimum volume is the compression ratio re . Typical values of re are 8 to 12 for SI engines and 12 to 24 for CI engines. The majority of reciprocating engines operate on what is known as the four-stroke cycle. Each cylinder requires four strokes of its piston-two revol- Exhaui) utions of the crankshaft-to complete the sequence of events which produces one Deflector ports Transfer power stroke. Both SI and CI engines use this cycle which comprises (see Fig. ports 1-2): O 1. An intake stroke, which starts with the piston at TC and ends with the piston at BC, which draws fresh mixture into the cylinder. To increase the mass inducted, the inlet valve opens shortly before the stroke starts and closes after Peed spring it ends. inlet valve 2. A compression stroke, when both valves are closed and the mixture inside the cylinder is compressed to a small fraction of its initial volume. Toward the end of the compression stroke, combustion is initiated and the cylinder pressure rises more rapidly. 3. A power stroke, or expansion stroke, which starts with the piston at TC and ends at BC as the high-temperature, high-pressure, gases push the piston down Exhaust blowdown Scavenging and force the crank to rotate. About five times as much work is done on the FIGURE 1-3 piston during the power stroke as the piston had to do during compression. The two-stroke operating cycle. A crankcase-scavenged engine is shown.10 12 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE TYPES AND THEIR OPERATION 13 2. A power or expansion stroke, similar to that in the four-stroke cycle until the Air Cleaner piston approaches BC, when first the exhaust ports and then the intake ports are uncovered (Fig. 1-3). Most of the burnt gases exit the cylinder in an exhaust blowdown process. When the inlet ports are uncovered, the fresh charge which has been compressed in the crankcase flows into the cylinder. Carburetor The piston and the ports are generally shaped to deflect the incoming charge Camshaft from flowing directly into the exhaust ports and to achieve effective scavenging of the residual gases. ... Rocker Arm Each engine cycle with one power stroke is completed in one crankshaft revolution. However, it is difficult to fill completely the displaced volume with Hydraulic fresh charge, and some of the fresh mixture flows directly out of the cylinder Adjuste during the scavenging process.+ The example shown is a cross-scavenged design; o other approaches use loop-scavenging or uniflow systems (see Sec. 6.6). Intake Valve Cam Sprocket Exhaust Valve 1.4 ENGINE COMPONENTS Piston Labeled cutaway drawings of a four-stroke SI engine and a two-stroke CI engine are shown in Figs. 1-4 and 1-5, respectively. The spark-ignition engine is a four- Connecting cylinder in-line automobile engine. The diesel is a large V eight-cylinder design Rod with a uniflow scavenging process. The function of the major components of Timing Belt - these engines and their construction materials will now be reviewed. The engine cylinders are contained in the engine block. The block has tradi- tionally been made of gray cast iron because of its good wear resistance and low cost. Passages for the cooling water are cast into the block. Heavy-duty and truck engines often use removable cylinder sleeves pressed into the block that can Timing Belt Crankshaft be replaced when worn. These are called wet liners or dry liners depending on Tensioner whether the sleeve is in direct contact with the cooling water. Aluminum is being used increasingly in smaller SI engine blocks to reduce engine weight. Iron cylin- der liners may be inserted at the casting stage, or later on in the machining and assembly process. The crankcase is often integral with the cylinder block. The crankshaft has traditionally been a steel forging; nodular cast iron crankshafts are also accepted normal practice in automotive engines. The crank- Oil Pump shaft is supported in main bearings. The maximum number of main bearings is Crankshaft Oil Pickup Sprocket one more than the number of cylinders; there may be less. The crank has eccen- FIGURE 1-4 tric portions (crank throws); the connecting rod big-end bearings attach to the Cutaway drawing of Chrysler 2.2-liter displacement four-cylinder spark-ignition engine.11 Bore 87.5 crank pin on each throw. Both main and connecting rod bearings use steel- mm, stroke 92 mm, compression ratio 8.9, maximum power 65 kW at 5000 rev/min. backed precision inserts with bronze, babbit, or aluminum as the bearing materials. The crankcase is sealed at the bottom with a pressed-steel or cast aluminum oil pan which acts as an oil reservoir for the lubricating system. Pistons are made of aluminum in small engines or cast iron in larger slower-speed engines. The piston both seals the cylinder and transmits the combustion-generated gas pressure to the crank pin via the connecting rod. The connecting rod, usually a steel or alloy forging (though sometimes aluminum in + It is primarily for this reason that two-stroke SI engines are at a disadvantage because the lost fresh small engines), is fastened to the piston by means of a steel piston pin through the charge contains fuel and air. rod upper end. The piston pin is usually hollow to reduce its weight. 14 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE TYPES AND THEIR OPERATION 15 -Water Discharge Manifold Exhaust Elbow- - Lifting Shackle Base The valves shown in Fig. 1-4 are poppet valves, the valve type normally used Top Deck Cover - Exhaust Valve Rocker Arm in four-stroke engines. Valves are made from forged alloy steel; the cooling of the Camshaft Injector Rocker Arm Exhaust Valve Bridge exhaust valve which operates at about 700ºC may be enhanced by using a hollow Overspeed Trip Shaft Exhaust Valve Spring stem partially filled with sodium which through evaporation and condensation Fuel Manifold, carries heat from the hot valve head to the cooler stem. Most modern spark- Injector Adjusting Link Exhaust Valve Injector Control Shaft ignition engines have overhead valve locations (sometimes called valve-in-head or Cylinder Head Injector Rack - I-head configurations) as shown in Fig. 1-4. This geometry leads to a compact Cylinder Test Valve -- Piston combustion chamber with minimum heat losses and flame travel time, and Thrust Washer Fuel Injector Piston Carrier improves the breathing capacity. Previous geometries such as the L head where Cylinder Head Crab Boft Piston Pin valves are to one side of the cylinder are now only used in small engines. Air Inlet Ports Crank case The valve stem moves in a valve guide, which can be an integral part of the Cylinder Liner Air Box cylinder head (or engine block for L-head engines), or may be a separate unit Blade Connecting Rod Water Inlet Jumper pressed into the head (or block). The valve seats may be cut in the head or block Oif Drain And Vent Water Inlet Manifold metal (if cast iron) or hard steel inserts may be pressed into the head or block. A Air Box Main Lube Oil Manifold Handhole Cover valve spring, attached to the valve stem with a spring washer and split keeper, Piston Cooling holds the valve closed. A valve rotator turns the valves a few degrees on opening Fork Connecting Rod - Oil Pipe to wipe the valve seat, avoid local hot spots, and prevent deposits building up in Connecting Rod Basket - Piston Cooling Oil Manifold Main Bearing "A" Frame- the valve guide. Oil Pan Main Bearing Cap - Handhole Cover A camshaft made of cast iron or forged steel with one cam per valve is used to Crankshaft Oil Pan open and close the valves. The cam surfaces are hardened to obtain adequate life. Crankshaft Counterweight Oil Level Gauge In four-stroke cycle engines, camshafts turn at one-half the crankshaft speed. Strainer Box Mechanical or hydraulic lifters or tappets slide in the block and ride on the cam. Depending on valve and camshaft location, additional members are required to transmit the tappet motion to the valve stem; e.g ., in in-head valve engines with the camshaft at the side, a push rod and rocker arm are used. A recent trend in automotive engines is to mount the camshaft over the head with the cams acting FIGURE 1-5 either directly or through a pivoted follower on the valve. Camshafts are gear, Cross-section drawing of an Electro-Motive two-stroke cycle diesel engine. This engine uses a uniflow belt, or chain driven from the crankshaft. scavenging process with inlet ports in the cylinder liner and four exhaust valves in the cylinder head. Bore 230.2 mm, stroke 254 mm, displaced volume per cylinder 10.57 liters, rated speed 750-900 An intake manifold (aluminum or cast iron) and an exhaust manifold rev/min. (Courtesy Electro-Motive Division, General Motors Corporation.) (generally of cast iron) complete the SI engine assembly. Other engine com- ponents specific to spark-ignition engines-carburetor, fuel injectors, ignition systems-are described more fully in the remaining sections in this chapter. The oscillating motion of the connecting rod exerts an oscillating force on The two-stroke cycle CI engine shown in Fig. 1-5 is of the uniflow scav- the cylinder walls via the piston skirt (the region below the piston rings). The enged design. The burned gases exhaust through four valves in the cylinder head. piston skirt is usually shaped to provide appropriate thrust surfaces. The piston These valves are controlled through cam-driven rocker arms. Fresh air is com- is fitted with rings which ride in grooves cut in the piston head to seal against gas pressed and fed to the air box by a Roots blower. The air inlet ports at the leakage and control oil flow. The upper rings are compression rings which are bottom of each cylinder liner are uncovered by the descending piston, and the forced outward against the cylinder wall and downward onto the groove face. scavenging air flows upward along the cylinder axis. The fuel injectors are The lower rings scrape the surplus oil from the cylinder wall and return it to the mounted in the cylinder head and are driven by the camshaft through rocker crankcase. The crankcase must be ventilated to remove gases which blow by the arms. Diesel fuel-injection systems are discussed in more detail in Sec. 1.7. piston rings, to prevent pressure buildup. The cylinder head (or heads in V engines) seals off the cylinders and is made 1.5 SPARK-IGNITION ENGINE OPERATION of cast iron or aluminum. It must be strong and rigid to distribute the gas forces acting on the head as uniformly as possible through the engine block. The cylin- In SI engines the air and fuel are usually mixed together in the intake system der head contains the spark plug (for an SI engine) or fuel injector (for a CI prior to entry to the engine cylinder, using a carburetor (Fig. 1-6) or fuel-injection engine), and, in overhead valve engines, parts of the valve mechanism. system (Fig. 1-7). In automobile applications, the temperature of the air entering 16 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE TYPES AND THEIR OPERATION 17 dle air bleed choke plate carburetor meters an appropriate fuel flow for the engine air flow in the following float chamber ventilation air correction jet manner. The air flow through the venturi (a converging-diverging nozzle) sets up emulsion tube full load enrichment a pressure difference between the venturi inlet and throat which is used to meter fuel discharge port an appropriate amount of fuel from the float chamber, through a series of ori- auxiliary air bleed fices, into the air flow at the venturi throat. Just downstream of the venturi is a fuel inlet auxiliary fuel jet throttle valve or plate which controls the combined air and fuel flow, and thus loat needle boost venturi the engine output. The intake flow is throttled to below atmospheric pressure by valve reducing the flow area when the power required (at any engine speed) is below idle jet accelerator the maximum which is obtained when the throttle is wide open. The intake mani- float pump fold is usually heated to promote faster evaporation of the liquid fuel and obtain main jet -ball valve more uniform fuel distribution between cylinders. part load control ~venturi Fuel injection into the intake manifold or inlet port is an increasingly common alternative to a carburetor. With port injection, fuel is injected through idle mixture control screw throttle valve auxiliary mixture control screw individual injectors from a low-pressure fuel supply system into each intake port. FIGURE 1-6 There are several different types of systems: mechanical injection using an injec- Cross section of single-barrel downdraft carburetor.12 (Courtesy Robert Bosch GmbH and SAE.) tion pump driven by the engine; mechanical, driveless, continuous injection; elec- tronically controlled, driveless, injection. Figure 1-7 shows an example of an electronically controlled system. In this system, the air flow rate is measured directly; the injection valves are actuated twice per cam shaft revolution by injec- the intake system is controlled by mixing ambient air with air heated by contact tion pulses whose duration is determined by the electronic control unit to with the exhaust manifold. The ratio of mass flow of air to mass flow of fuel must provide the desired amount of fuel per cylinder per cycle.12 An alternative be held approximately constant at about 15 to ensure reliable combustion. The approach is to use a single fuel injector located above the throttle plate in the position normally occupied by the carburetor. This approach permits electronic control of the fuel flow at reduced cost. The sequence of events which take place inside the engine cylinder is illus- Fuel-pressure regulator trated in Fig. 1-8. Several variables are plotted against crank angle through the entire four-stroke cycle. Crank angle is a useful independent variable because Relay set engine processes occupy almost constant crank angle intervals over a wide range of engine speeds. The figure shows the valve timing and volume relationship for a Injection valve Start valve Air flow typical automotive spark-ignition engine. To maintain high mixture flows at high sensor engine speeds (and hence high power outputs) the inlet valve, which opens before TC, closes substantially after BC. During intake, the inducted fuel and air mix in Electronic the cylinder with the residual burned gases remaining from the previous cycle. control After the intake valve closes, the cylinder contents are compressed to above unit TOT atmospheric pressure and temperature as the cylinder volume is reduced. Some heat transfer to the piston, cylinder head, and cylinder walls occurs but the effect Auxiliary-air device on unburned gas properties is modest. Throttle-valve switch Between 10 and 40 crank angle degrees before TC an electrical discharge across the spark plug starts the combustion process. A distributor, a rotating Fuel filter Electric fuel pump Temperature switch driven off the camshaft, interrupts the current from the battery through sensor the primary circuit of the ignition coil. The secondary winding of the ignition Thermo-time switch coil, connected to the spark plug, produces a high voltage across the plug elec- FIGURE 1-7 trodes as the magnetic field collapses. Traditionally, cam-operated breaker points Schematic drawing of L-Jetronic port electronic fuel-injection system.12 (Courtesy Robert Bosch have been used; in most automotive engines, the switching is now done elec- GmbH and SAE.) tronically. A turbulent flame develops from the spark discharge, propagates 18 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE TYPES AND THEIR OPERATION 19 300 -- Combustion 2000 brake-torque (MBT) timing,t this optimum timing is an empirical compromise psia kPa Intake Exhaust between starting combustion too early in the compression stroke (when the work 200- transfer is to the cylinder gases) and completing combustion too late in the Compression Expansion expansion stroke (and so lowering peak expansion stroke pressures). -1000 About two-thirds of the way through the expansion stroke, the exhaust 100- P IVO EVC IVC EVO IVO valve starts to open. The cylinder pressure is greater than the exhaust manifold Spark pressure and a blowdown process occurs. The burned gases flow through the TC BC TC BC valve into the exhaust port and manifold until the cylinder pressure and exhaust pressure equilibrate. The duration of this process depends on the pressure level in the cylinder. The piston then displaces the burned gases from the cylinder into the Burned 1.01 manifold during the exhaust stroke. The exhaust valve opens before the end of the expansion stroke to ensure that the blowdown process does not last too far Xb into the exhaust stroke. The actual timing is a compromise which balances V max reduced work transfer to the piston before BC against reduced work transfer to the cylinder contents after BC. Unburned The exhaust valve remains open until just after TC; the intake opens just TC BC TC BC TC before TC. The valves are opened and closed slowly to avoid noise and excessive -360º -180º 0º 180º 360º cam wear. To ensure the valves are fully open when piston velocities are at their Crank position and angle highest, the valve open periods often overlap. If the intake flow is throttled to FIGURE 1-8 below exhaust manifold pressure, then backflow of burned gases into the intake Sequence of events in four-stroke spark-ignition engine operating cycle. Cylinder pressure p (solid manifold occurs when the intake valve is first opened. line, firing cycle; dashed line, motored cycle), cylinder volume V/V.mx, and mass fraction burned x, are plotted against crank angle. 1.6 EXAMPLES OF SPARK-IGNITION ENGINES across the mixture of air, fuel, and residual gas in the cylinder, and extinguishes This section presents examples of production spark-ignition engines to illustrate at the combustion chamber wall. The duration of this burning process varies with the different types of engines in common use. engine design and operation, but is typically 40 to 60 crank angle degrees, as Small SI engines are used in many applications: in the home (e.g ., lawn shown in Fig. 1-8. As fuel-air mixture burns in the flame, the cylinder pressure in mowers, chain saws), in portable power generation, as outboard motorboat Fig. 1-8 (solid line) rises above the level due to compression alone (dashed line). engines, and in motorcycles. These are often single-cylinder engines. In the above This latter curve-called the motored cylinder pressure-is the pressure trace applications, light weight, small bulk, and low cost in relation to the power gen- obtained from a motored or nonfiring engine.| Note that due to differences in the erated are the most important characteristics; fuel consumption, engine vibration, flow pattern and mixture composition between cylinders, and within each cylin- and engine durability are less important. A single-cylinder engine gives only one der cycle-by-cycle, the development of each combustion process differs somewhat. power stroke per revolution (two-stroke cycle) or two revolutions (four-stroke As a result, the shape of the pressure versus crank angle curve in each cylinder, cycle). Hence, the torque pulses are widely spaced, and engine vibration and and cycle-by-cycle, is not exactly the same. smoothness are significant problems. There is an optimum spark timing which, for a given mass of fuel and air Multicylinder engines are invariably used in automotive practice. As rated inside the cylinder, gives maximum torque. More advanced (earlier) timing or power increases, the advantages of smaller cylinders in regard to size, weight, and retarded (later) timing than this optimum gives lower output. Called maximum improved engine balance and smoothness point toward increasing the number of + MBT timing has traditionally been defined as the minimum spark advance for best torque. Since In practice, the intake and compression processes of a firing engine and a motored engine are not the torque first increases and then decreases as spark timing is advanced, the definition used here is exactly the same due to the presence of burned gases from the previous cycle under firing conditions. more precise. 20 INTERNAL COMBUSTION ENGINE FUNDAMENTALS cylinders per engine. An upper limit on cylinder size is dictated by dynamic con- siderations: the inertial forces that are created by accelerating and decelerating the reciprocating masses of the piston and connecting rod would quickly limit the maximum speed of the engine. Thus, the displaced volume is spread out amongst several smaller cylinders. The increased frequency of power strokes with a multi- cylinder engine produces much smoother torque characteristics. Multicylinder engines can also achieve a much better state of balance than single-cylinder engines. A force must be applied to the piston to accelerate it during the first half of its travel from bottom-center or top-center. The piston then exerts a force as it decelerates during the second part of the stroke. It is desirable to cancel these inertia forces through the choice of number and arrangement of cylinders to achieve a primary balance. Note, however, that the motion of the piston is more rapid during the upper half of its stroke than during the lower half (a conse- quence of the connecting rod and crank mechanism evident from Fig. 1-1; see also Sec. 2.2). The resulting inequality in piston acceleration and deceleration produces corresponding differences in inertia forces generated. Certain com- binations of cylinder number and arrangement will balance out these secondary inertia force effects. Four-cylinder in-line engines are the most common arrangements for auto- mobile engines up to about 2.5-liter displacement. An example of this in-line arrangement was shown in Fig. 1-4. It is compact-an important consideration WW for small passenger cars. It provides two torque pulses per revolution of the Cross-section drawings of General Motors 60 degree V-6 spark-ignition engine.13 Displacement 2.8 liter, bore 89 mm, stroke 76 mm, compression ratio 8.5, crankshaft and primary inertia forces (though not secondary forces) are balanced. V engines and opposed-piston engines are occasionally used with this number of cylinders. The V arrangement, with two banks of cylinders set at 90º or a more acute angle to each other, provides a compact block and is used extensively for larger displacement engines. Figure 1-9 shows a V-6 engine, the six cylinders being arranged in two banks of three with a 60º angle between their axis. Six cylinders are usually used in the 2.5- to 4.5-liter displacement range. Six-cylinder engines provide smoother operation with three torque pulses per revolution. The in-line arrangement results in a long engine, however, giving rise to crankshaft torsional vibration and making even distribution of air and fuel to each cylinder more difficult. The V-6 arrangement is much more compact, and the example shown provides primary balance of the reciprocating components. With the V engine, however, a rocking moment is imposed on the crankshaft due to the secondary inertia forces, which results in the engine being less well balanced than the in-line version. The V-8 and V-12 arrangements are also commonly used to provide maximum power 86 kW at 4800 rev/min. compact, smooth, low-vibration, larger-displacement, spark-ignition engines. Turbochargers are used to increase the maximum power that can be obtained from a given displacement engine. The work transfer to the piston per cycle, in each cylinder, which controls the power the engine can deliver, depends FIGURE 1-9 on the amount of fuel burned per cylinder per cycle. This depends on the amount of fresh air that is inducted each cycle. Increasing the air density prior to entry into the engine thus increases the maximum power that an engine of given dis- 22 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE TYPES AND THEIR OPERATION 23 placement can deliver. Figure 1-10 shows an example of a turbocharged four- Lubricating passage Lock plate cylinder spark-ignition engine. The turbocharger, a compressor-turbine Compressed air outlet Seal ring combination, uses the energy available in the engine exhaust stream to achieve Compressor housing compression of the intake flow. The air flow passes through the compressor (2), Compressor wheel intercooler (3), carburetor (4), manifold (5), and inlet valve (6) as shown. Engine Center housing inlet pressures (or boost) of up to about 100 kPa above atmospheric pressure are typical. The exhaust flow through the valve (7) and manifold (8) drives the Turbine housing Air inlet side turbine (9) which powers the compressor. A wastegate (valve) just upstream of the Thrust spring turbine bypasses some of the exhaust gas flow when necessary to prevent the boost pressure becoming too high. The wastegate linkage (11) is controlled by a boost pressure regulator. While this turbocharged engine configuration has the Thrust bearing carburetor downstream of the compressor, some turbocharged spark-ignition engines have the carburetor upstream of the compressor so that it operates at or. below atmospheric pressure. Figure 1-11 shows a cutaway drawing of a small automotive turbocharger. The arrangements of the compressor and turbine Exhaust gas outlet side Radial bearing Turbine wheel Exhaust gas bypass passage" LI Exhaust gas inlet side FIGURE 1-11 Cutaway view of small automotive engine turbocharger. (Courtesy Nissan Motor Co ., Ltd.) 7 6 rotors connected via the central shaft and of the turbine and compressor flow passages are evident. Figure 1-12 shows a two-stroke cycle spark-ignition engine. The two-stroke cycle spark-ignition engine is used for small-engine applications where low cost and weight/power ratio are important and when the use factor is low. Examples of such applications are outboard motorboat engines, motorcycles, and chain 8 saws. All such engines are of the carburetor crankcase-compression type which is one of the simplest prime movers available. It has three moving parts per cylin- der: the piston, connecting rod, and the crank. The prime advantage of the two- stroke cycle spark-ignition engine relative to the four-stroke cycle engine is its higher power per unit displaced volume due to twice the number of power 10 strokes per crank revolution. This is offset by the lower fresh charge density achieved by the two-stroke cycle gas-exchange process and the loss of fresh mixture which goes straight through the engine during scavenging. Also, oil con- sumption is higher in two-stroke cycle engines due to the need to add oil to the E.T.A.I. France fuel to lubricate the piston ring and piston surfaces. The Wankel rotary engine is an alternative to the reciprocating engine geometry of the engines illustrated above. It is used when its compactness and FIGURE 1-10 higher engine speed (which result in high power/weight and power/volume Turbocharged four-cylinder automotive spark-ignition engine. (Courtesy Regie Nationale des Usines.) ratios), and inherent balance and smoothness, offset its higher heat transfer, and 24 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE TYPES AND THEIR OPERATION 25 Fixed timing gear EXHAUST Center housing Exhaust port Side Housing Internal timing gear Drive end Eccentric SECTION AZA Flywheel 7 Rotor Eccentric shaft - Intake port Coolant passages Side housing Induction Compression Ignition Power Exhaust FIGURE 1-13 (a) Major components of the Wankel rotary engine; (b) induction, compression, power, and exhaust processes of the four-stroke cycle for the chamber defined by rotor surface AB. (From Mobil Technical FIGURE 1-12 Bulletin, Rotary Engines, C Mobil Oil Corporation, 1971.) Cutaway drawing of two-cylinder two-stroke cycle loop-scavenged marine spark-ignition engine. Dis- placed volume 737 cm3, maximum power 41 KW at 5500 rev/min. (Courtesy Outboard Marine Corpo- ration.) thus for each eccentric (output) shaft revolution there is one power pulse. Figure 1-14 shows a cutaway drawing of a two-rotor automobile Wankel engine. The its sealing and leakage problems. Figure 1-13 shows the major mechanical parts two rotors are out of phase to provide a greater number of torque pulses per of a simple single-rotor Wankel engine and illustrates its geometry. There are two shaft revolution. Note the combustion chamber cut out in each rotor face, the rotating parts: the triangular-shaped rotor and the output shaft with its integral rotor apex, and side seals. Two spark plugs per firing chamber are often used to eccentric. The rotor revolves directly on the eccentric. The rotor has an internal obtain a faster combustion process. timing gear which meshes with the fixed timing gear on one side housing to maintain the correct phase relationship between the rotor and eccentric shaft 1.7 COMPRESSION-IGNITION ENGINE rotations. Thus the rotor rotates and orbits around the shaft axis. Breathing is OPERATION through ports in the center housing (and sometimes the side housings). The com- bustion chamber lies between the center housing and rotor surface and is sealed In compression-ignition engines, air alone is inducted into the cylinder. The fuel by seals at the apex of the rotor and around the perimeters of the rotor sides. (in most applications a light fuel oil, though heated residual fuel is used in marine Figure 1-13 also shows how the Wankel rotary geometry operates with the four- and power-generation applications) is injected directly into the engine cylinder stroke cycle. The figure shows the induction, compression, power, and exhaust just before the combustion process is required to start. Load control is achieved processes of the four-stroke cycle for the chamber defined by rotor surface AB. by varying the amount of fuel injected each cycle; the air flow at a given engine The remaining two chambers defined by the other rotor surfaces undergo exactly speed is essentially unchanged. There are a great variety of CI engine designs in the same sequence. As the rotor makes one complete rotation, during which the use in a wide range of applications-automobile, truck, locomotive, marine, eccentric shaft rotates through three revolutions, each chamber produces one power generation. Naturally aspirated engines where atmospheric air is inducted, power "stroke." Three power pulses occur, therefore, for each rotor revolution; turbocharged engines where the inlet air is compressed by an exhaust-driven ENGINE TYPES AND THEIR OPERATION 27 turbine-compressor combination, and supercharged engines where the air is com- pressed by a mechanically driven pump or blower are common. Turbocharging and supercharging increase engine output by increasing the air mass flow per unit displaced volume, thereby allowing an increase in fuel flow. These methods are used, usually in larger engines, to reduce engine size and weight for a given power output. Except in smaller engine sizes, the two-stroke cycle is competitive with the four-stroke cycle, in large part because, with the diesel cycle, only air is lost in the cylinder scavenging process. The operation of a typical four-stroke naturally aspirated CI engine is illus- trated in Fig. 1-15. The compression ratio of diesels is much higher than typical SI engine values, and is in the range 12 to 24, depending on the type of diesel engine and whether the engine is naturally aspirated or turbocharged. The valve timings used are similar to those of SI engines. Air at close-to-atmospheric pres- sure is inducted during the intake stroke and then compressed to a pressure of about 4 MPa (600 lb/in2) and temperature of about 800 K (1000ºF) during the compression stroke. At about 20º before TC, fuel injection into the engine cylin- der commences; a typical rate of injection profile is shown in Fig. 1-15b. The liquid fuel jet atomizes into drops and entrains air. The liquid fuel evaporates; fuel vapor then mixes with air to within combustible proportions. The air tem- perature and pressure are above the fuel's ignition point. Therefore after a short delay period, spontaneous ignition (autoignition) of parts of the nonuniform fuel- Cutaway drawing of two-rotor Wankel spark-ignition engine. Displacement of each working chamber 573 cm3, compression ratio 9.4. maximum air mixture initiates the combustion process, and the cylinder pressure (solid line in Fig. 1-15c) rises above the nonfiring engine level. The flame spreads rapidly through that portion of the injected fuel which has already mixed with sufficient air to burn. As the expansion process proceeds, mixing between fuel, air, and burning gases continues, accompanied by further combustion (see Fig. 1-15d). At full load, the mass of fuel injected is about 5 percent of the mass of air in the cylinder. Increasing levels of black smoke in the exhaust limit the amount of fuel that can be burned efficiently. The exhaust process is similar to that of the four- stroke SI engine. At the end of the exhaust stroke, the cycle starts again. In the two-stroke CI engine cycle, compression, fuel injection, combustion, and expansion processes are similar to the equivalent four-stroke cycle processes; it is the intake and exhaust pressure which are different. The sequence of events in a loop-scavenged two-stroke engine is illustrated in Fig. 1-16. In loop- scavenged engines both exhaust and inlet ports are at the same end of the cylin- der and are uncovered as the piston approaches BC (see Fig. 1-16a). After the exhaust ports open, the cylinder pressure falls rapidly in a blowdown process (Fig. 1-16b). The inlet ports then open, and once the cylinder pressure p falls below the inlet pressure p;, air flows into the cylinder. The burned gases, dis- placed by this fresh air, continue to flow out of the exhaust port (along with some of the fresh air). Once the ports close as the piston starts the compression stroke, FIGURE 1-14 compression, fuel-injection, fuel-air mixing, combustion and expansion processes proceed as in the four-stroke CI engine cycle. The diesel fuel-injection system consists of an injection pump, delivery pipes, and fuel injector nozzles. Several different types of injection pumps and 28 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE TYPES AND THEIR OPERATION 29 15 P IVC EVO TTTTT P EPO IPO IPC EPC BC TC BC SOI EOI TC 90º BC 270º TC BC BC Ac 160 --- .20 soc/ T TC 90º BC 270º P IVC SOI EVO (atm) 80 EOC 40} TTTTT T VITTTTTTTITT BC TC BC P Pe- Piz SOI mfp TC 90º BC 2700 TC T Crank angle BC TC BC - 180º -90º 00 90º 180º FIGURE 1-16 Crank angle Sequence of events during expansion, gas exchange, and compression processes in a loop-scavenged FIGURE 1-15 two-stroke cycle compression-ignition engine. Cylinder volume/clearance volume V/V2, cylinder pres- Sequence of events during compression, combustion, and expansion processes of a naturally aspirated sure p, exhaust port open area A ., and intake port open area A; are plotted against crank angle. compression-ignition engine operating cycle. Cylinder volume/clearance volume V/V2, rate of fuel injection m ;, cylinder pressure p (solid line, firing cycle; dashed line, motored cycle), and rate of fuel burning (or fuel chemical energy release rate) m, are plotted against crank angle. line. The injection nozzle (Fig. 1-18) has one or more holes through which the fuel sprays into the cylinder. A spring-loaded valve closes these holes until the pressure in the injection line, acting on part of the valve surface, overcomes the nozzles are used. In one common fuel pump (an in-line pump design shown in spring force and opens the valve. Injection starts shortly after the line pressure Fig. 1-17) a set of cam-driven plungers (one for each cylinder) operate in closely begins to rise. Thus, the phase of the pump camshaft relative to the engine crank- fitting barrels. Early in the stroke of the plunger, the inlet port is closed and the shaft controls the start of injection. Injection is stopped when the inlet port of the fuel trapped above the plunger is forced through a check valve into the injection pump is uncovered by a helical groove in the pump plunger, because the high 30 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE TYPES AND THEIR OPERATION 31 Annular groove Overflow line Nozzle-and- Annular groove- holder assembly Two-stage Iniet port- fuet filter Inlet port- Nozzle needle Nozzle needle Pressure Pressure chamber- Spray angle chamber Timing device Pintle nozzle closed Pintle nozzle open Multihole nozzle open Drive from A engine Adjusting screw Fuel-injection Governor pump Leak-oil connection . Supply pump Edge-type filter Spindie Pressure passage Nozzle Nozzle-holder assembly with nozzle la nozzle Helix Vertical groove ·cm suction Ba.el Pager. Maximum delivery Partial delivery Zero delivery BDC Port BDC Port BDC Supply opening opening Governor pump Timing device Fuel delivery control (lower helix) FIGURE 1-17 Diesel fuel system with in-line fuel-injection pump (type PE).12 (Courtesy Robert Bosch GmbH.) FIGURE 1-18 Details of fuel-injection nozzles, nozzle holder assembly and fuel-delivery control.12 (Courtesy Robert Bosch GmbH.) pressure above the plunger is then released (Fig. 1-18). The amount of fuel injected (which controls the load) is determined by the injection pump cam design and the position of the helical groove. Thus for a given cam design, rotating the Distributor pumps can operate at higher speed and take up less space than plunger and its helical groove varies the load. in-line pumps. They are normally used on smaller diesel engines. In-line pumps Distributor-type pumps have only one pump plunger and barrel, which are used in the mid-engine-size range. In the larger diesels, individual single- meters and distributes the fuel to all the injection nozzles. A schematic of a barrel injection pumps, close mounted to each cylinder with an external drive as distributor-type pump is shown in Fig. 1-19. The unit contains a low-pressure shown in Fig. 1-5, are normally used. fuel pump (on left), a high-pressure injection pump (on right), an overspeed gov- ernor, and an injection timer. High pressure is generated by the plunger which is 1.8 EXAMPLES OF DIESEL ENGINES made to describe a combined rotary and stroke movement by the rotating eccen- tric disc or cam plate; the rotary motion distributes the fuel to the individual A large number of diesel engine configurations and designs are in common use. injection nozzles. The very large marine and stationary power-generating diesels are two-stroke 32 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE TYPES AND THEIR OPERATION 33 Overflow valve size of diesel engine, it is often necessary to use a swirling air flow rotating about Control lever Governor 'Sliding sleeve Stop lever spring the cylinder axis, which is created by suitable design of the inlet port and valve, Overflow throttle to achieve adequate fuel-air mixing and fuel burning rates. The fuel injector, Flyweights Full-load shown left-of-center in the drawing, has a multihole nozzle, typically with three to adjusting screw five holes. The fuel jets move out radially from the center of the piston bowl into Pressure- regulating Tensioning the (swirling) air flow. The in-line fuel-injection pump is normally used with this lever Fine filter valve Starting lever type of diesel engine. Figure 1-21 shows a four-cylinder in-line overhead-valve-cam design auto- Nozzle- holder mobile diesel engine. The smallest diesels such as this operate at higher engine assem- speed than larger engines; hence the time available for burning the fuel is less and bly with the fuel-injection and combustion system must achieve faster fuel-air mixing nozzle rates. This is accomplished by using an indirect-injection type of diesel. Fuel is injected into an auxiliary combustion chamber which is separated from the main Distributor- Delivery pump valve combustion chamber above the piston by a flow restriction or nozzle. During the Regulating collar. latter stages of the compression process, air is forced through this nozzle from the Supply pump1) Drive Governor Cam Maximum effective stroke, start hub drive plate Timing device2) 1) Shown additionally turned through 90º Presupply pump 2) Shown turned through 90ª FIGURE 1-19 Diesel fuel system with distributor-type fuel-injection pump with mechanical governor.12 (Courtesy Robert Bosch GmbH.) cycle engines. Small- and medium-size engines use the four-stroke cycle. Because air capacity is an important constraint on the amount of fuel that can be burned in the diesel engine, and therefore on the engine's power, turbocharging is used extensively. All large engines are turbocharged. The majority of smaller diesels are not' turbocharged, though they can be turbocharged and many are. The details of the engine design also vary significantly over the diesel size range. In particular, different combustion chamber geometries and fuel-injection character- istics are required to deal effectively with a major diesel engine design problem- achieving sufficiently rapid fuel-air mixing rates to complete the fuel-burning process in the time available. A wide variety of inlet port geometries, cylinder head and piston shapes, and fuel-injection patterns are used to accomplish this over the diesel size range. Figure 1-20 shows a diesel engine typical of the medium-duty truck applica- tion. The design shown is a six-cylinder in-line engine. The drawing indicates that diesel engines are generally substantially heavier than spark-ignition engines because stress levels are higher due to the significantly higher pressure levels of the diesel cycle. The engine shown has a displacement of 10 liters, a compression ratio of 16.3, and is usually turbocharged. The engine has pressed-in cylinder liners to achieve better cylinder wear characteristics. This type of diesel is called a FIGURE 1-20 direct-injection diesel. The fuel is injected into a combustion chamber directly Direct-injection four-stroke cycle six-cylinder turbocharged Cummins diesel engine. Displaced volume above the piston crown. The combustion chamber shown is a "bowl-in-piston" 10 liters, bore 125 mm, stroke 136 mm, compression ratio 16.3, maximum power 168 to 246 kW at design, which puts most of the clearance volume into a compact shape. With this rated speed of 2100 rev/min. (Courtesy Cummins Engine Company, Inc.) 34 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE TYPES AND THEIR OPERATION 35 Diesel engines are turbocharged to achieve higher power/weight ratios. By increasing the density of the inlet air, a given displaced volume can induct more air. Hence more fuel can be injected and burned, and more power delivered, while avoiding excessive black smoke in the exhaust. All the larger diesels are turbo- charged; smaller diesels can be and often are. Figure 1-22 shows how a turbo- charger connects to a direct-injection diesel. All the above diesels are water cooled; some production diesels are air cooled. Figure 1-23 shows a V-8 air-cooled direct-injection naturally aspirated FIGURE 1-21 Four-cylinder naturally aspirated indirect-injection automobile Volkswagen diesel engine.14 Dis- placed volume 1.47 liters, bore 76.5 mm, stroke 80 mm, maximum power 37 kW at 5000 rev/min. cylinder into the prechamber at high velocity. Fuel is injected into this highly turbulent and often rapidly swirling flow in this auxiliary or prechamber, and very high mixing rates are achieved. Combustion starts in the prechamber, and the resulting pressure rise in the prechamber forces burning gases, fuel, and air into the main chamber. Since this outflow is also extremely vigorous, rapid mixing then occurs in the main chamber as the burning jet mixes with the remaining air and combustion is completed. A distributor-type fuel pump, which is normally used in this engine size range, driven off the camshaft at half the crankshaft speed by a toothed belt, is shown on the right of the figure. It supplies high-pressure fuel pulses to the pintle-type injector nozzles in turn. A glow plug is also shown in the auxiliary chamber; this plug is electrically heated prior to and during cold engine start-up to raise the temperature of the air charge and the fuel FIGURE 1-22 sufficiently to achieve autoignition. The compression ratio of this engine is 23. Turbocharged aftercooled direct-injection four-stroke cycle Caterpillar six-cylinder in-line heavy-duty Indirect-injection diesel engines require higher compression ratios than direct- truck diesel engine. Bore 137.2 mm, stroke 165.1 mm, rated power 200-300 KW and rated speed of injection engines to start adequately when cold. 1600-2100 rev/min depending on application. (Courtesy Caterpillar Tractor Company.) 36 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE TYPES AND THEIR OPERATION 37 FIGURE 1-24 FIGURE 1-23 Large Sulzer two-stroke turbocharged marine V-8 air-cooled direct-injection naturally aspirated diesel engine. Displacement 13.4 liter, bore 128 mm, diesel engine. Bore 840 mm, stroke 2900 mm, stroke 130 mm, compression ratio 17, maximum rated power 188 kW at rated speed of 2300 rev/min. rated power 1.9 MW per cylinder at 78 rev/min, 4 (Courtesy Klöcker-Humboldt-Deutz AG.15) to 12 cylinders. (Courtesy Sulzer Brothers Ltd.) diesel. The primary advantage compared to the water-cooled engines is lower turbine. Compressed air enters via the inlet ports and induces forced scavenging; engine weight. The fins on the cylinder block and head are necessary to increase air is supplied from the turbocharger and cooler. At part load electrically driven the external heat-transfer surface area to achieve the required heat rejection. An blowers cut in to compress the scavenge air. Because these large engines operate air blower, shown on the right of the cutaway drawing, provides forced air con- at low speed, the motion induced by the centrally injected fuel jets is sufficient to vection over the block. The blower is driven off the injection pump shaft, which mix the fuel with air and burn it in the time available. A simple open combustion in turn is driven off the camshaft. The in-line injection pump is placed between chamber shape can be used, therefore, which achieves efficient combustion even the two banks of cylinders. The injection nozzles are located at an angle to the with the low-quality heavy fuels used with these types of engines. The pistons are cylinder axis. The combustion chamber and fuel-injection characteristics are water cooled in these very large engines. The splash oil piston cooling used in similar to those of the engine in Fig. 1-22. The nozzle shown injects four fuel medium- and small-size diesels is not adequate. sprays into a reentrant bowl-in-piston combustion chamber. Diesels are also made in very large engine sizes. These large engines are used for marine propulsion and electrical power generation and operate on the 1.9 STRATIFIED-CHARGE ENGINES two-stroke cycle in contrast to the small- and medium-size diesels illustrated Since the 1920s, attempts have been made to develop a hybrid internal com- above. Figure 1-24 shows such a two-stroke cycle marine engine, available with bustion engine that combines the best features of the spark-ignition engine and from 4 to 12 cylinders, with a maximum bore of 0.6-0.9 m and stroke of 2-3 m, the diesel. The goals have been to operate such an engine at close to the optimum which operates at speeds of about 100 rev/min. These engines are normally of the compression ratio for efficiency (in the 12 to 15 range) by: (1) injecting the fuel crosshead type to reduce side forces on the cylinder. The gas exchange between directly into the combustion chamber during the compression process (and cycles is controlled by first opening the exhaust valves, and then the piston thereby avoid the knock or spontaneous ignition problem that limits convention- uncovering inlet ports in the cylinder liner. Expanding exhaust gases leave the al spark-ignition engines with their premixed charge); (2) igniting the fuel as it cylinder via the exhaust valves and manifold and pass through the turbocharger mixes with air with a spark plug to provide direct control of the ignition process 38 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE TYPES AND THEIR OPERATION 39 (and thereby avoid the fuel ignition-quality requirement of the diesel); (3) control- ling the engine power level by varying the amount of fuel injected per cycle (with the air flow unthrottled to minimize work done pumping the fresh charge into the cylinder). Such engines are often called stratified-charge engines from the need to produce in the mixing process between the fuel jet and the air in the cylinder a "stratified" fuel-air mixture, with an easily ignitable composition at the spark plug at the time of ignition. Because such engines avoid the spark-ignition engine requirement for fuels with a high antiknock quality and the diesel requirement for fuels with high ignition quality, they are usually fuel-tolerant and will operate with a wide range of liquid fuels. Many different types of stratified-charge engine have been proposed, and some have been partially or fully developed. A few have even been used in prac- tice in automotive applications. The operating principles of those that are truly fuel-tolerant or multifuel engines are illustrated in Fig. 1-25. The combustion chamber is usually a bowl-in-piston design, and a high degree of air swirl is created during intake and enhanced in the piston bowl during compression to achieve rapid fuel-air mixing. Fuel is injected into the cylinder, tangentially into the bowl, during the latter stages of compression. A long-duration spark dis- charge ignites the developing fuel-air jet as it passes the spark plug. The flame spreads downstream, and envelopes and consumes the fuel-air mixture. Mixing continues, and the final stages of combustion are completed during expansion. Most successful designs of this type of engine have used the four-stroke cycle. This concept is usually called a direct-injection stratified-charge engine. The engine can be turbocharged to increase its power density. Texaco M.A.N. FIGURE 1-26 Sectional drawing of M.A.N. high-speed multifuel four-cylinder direct-injection stratified-charge engine. Bore 94.5 mm, stroke 100 mm, displacement 2.65 liters, compression ratio 16.5, rated power 52 kW at 3800 rev/min.17 Swirl Swir A commercial multifuel engine is shown in Fig. 1-26. In this particular design, the fuel injector comes diagonally through the cylinder head from the Injector Spark plug Injector Spark , plug upper left and injects the fuel onto the hot wall of the deep spherical piston bowl. The fuel is carried around the wall of the bowl by the swirling flow, evaporated off the wall, mixed with air, and then ignited by the discharge at the spark plug Late injection which enters the chamber vertically on the right. This particular engine is air cooled, so the cylinder block and head are finned to increase surface area. Piston cup Piston cup An alternative stratified-charge engine concept, which has also been mass produced, uses a small prechamber fed during intake with an auxiliary fuel system to obtain an easily ignitable mixture around the spark plug. This concept, first proposed by Ricardo in the 1920s and extensively developed in the Soviet Union FIGURE 1-25 Two multifuel stratified-charge engines which have been used in commercial practice: the Texaco and Japan, is often called a jet-ignition or torch-ignition stratified-charge engine. Controlled Combustion System (TCCS)16 and the M.A.N .- FM System.17 Its operating principles are illustrated in Fig. 1-27 which shows a three-valve 40 ENGINE TYPES AND THEIR OPERATION 41 INTERNAL COMBUSTION ENGINE FUNDAMENTALS 1.5. The two-stroke cycle has twice as many power strokes per crank revolution as the Rich four-stroke cycle. However, two-stroke cycle engine power outputs per unit displaced Intake" volume are less than twice the power output of an equivalent four-stroke cycle engine at the same engine speed. Suggest reasons why this potential advantage of the two- Leon stroke cycle is offset in practice. Intake 1.6. Suggest reasons why multicylinder engines prove more attractive than single-cylinder Orifice engines once the total engine displaced volume exceeds a few hundred cubic centi- meters. 1.7. The Wankel rotary spark-ignition engine, while lighter and more compact than a reciprocating spark-ignition engine of equal maximum power, typically has worse effi- IN TAKE COMPRESSION COMBUSTION ciency due to significantly higher gas leakage from the combustion chamber and FIGURE 1-27 higher total heat loss from the hot combustion gases to the chamber walls. Based on Schematic of three-valve torch-ignition stratified-charge spark-ignition engine. the design details in Figs. 1-4, 1-13, and 1-14 suggest reasons for these higher losses. REFERENCES carbureted version of the concept.18 A separate carburetor and intake manifold feeds a fuel-rich mixture (which contains fuel beyond the amount that can be 1. Cummins, Jr ., C. L.: Internal Fire, Carnot Press, Lake Oswego, Oreg ., 1976. 2. Cummins, Jr ., C. L.: "Early IC and Automotive Engines," SAE paper 760604 in A History of the burned with the available air) through a separate small intake valve into the Automotive Internal Combustion Engine, SP-409, SAE Trans ., vol. 85, 1976. prechamber which contains the spark plug. At the same time, a very lean mixture 3. Hempson, J. G. G.: "The Automobile Engine 1920-1950," SAE paper 760605 in A History of the (which contains excess air beyond that required to burn the fuel completely) is fed Automotive Internal Combustion Engine, SP-409, SAE, 1976. to the main combustion chamber through the main carburetor and intake mani- 4. Agnew, W. G.: "Fifty Years of Combustion Research at General Motors," Progress in Energy and Combustion Science, vol. 4, pp. 115-156, 1978. fold. During intake the rich prechamber flow fully purges the prechamber 5. Wankel, F.: Rotary Piston Machines, Iliffe Books, London, 1965. volume. After intake valve closing, lean mixture from the main chamber is com- 6. Ansdale, R. F.: The Wankel RC Engine Design and Performance, Iliffe Books, London, 1968. pressed into the prechamber bringing the mixture at the spark plug to an easily 7. Yamamoto, K.: Rotary Engine, Toyo Kogyo Co. Ltd ., Hiroshima, 1969. ignitable, slightly rich, composition. After combustion starts in the prechamber, 8. Haagen-Smit, A. J.: "Chemistry and Physiology of Los Angeles Smog," Ind. Eng. Chem ., vol. 44, rich burning mixture issues as a jet through the orifice into the main chamber, p. 1342, 1952. entraining and igniting the lean main chamber charge. Though called a stratified- 9. Taylor, C. F.: The Internal Combustion Engine in Theory and Practice, vol. 2, table 10-1, MIT Press, Cambridge, Mass ., 1968. charge engine, this engine is really a jet-ignition concept whose primary function 10. Rogowski, A. R.: Elements of Internal Combustion Engines, McGraw-Hill, 1953. is to extend the operating limit of conventionally ignited spark-ignition engines 11. Weertman, W. L ., and Dean, J. W.: "Chrysler Corporation's New 2.2 Liter 4 Cylinder Engine," to mixtures leaner than could normally be burned. SAE paper 810007, 1981. 12. Bosch: Automotive Handbook, 1st English edition, Robert Bosch GmbH, 1976. 13. Martens, D. A.: "The General Motors 2.8 Liter 60º V-6 Engine Designed by Chevrolet," SAE paper 790697, 1979. PROBLEMS 14. Hofbauer, P ., and Sator, K.: "Advanced Automotive Power Systems-Part 2: A Diesel for a Subcompact Car," SAE paper 770113, SAE Trans ., vol. 86, 1977. 1.1. Describe the major functions of the following reciprocating engine components: 15. Garthe, H.: "The Deutz BF8L 513 Aircooled Diesel Engine for Truck and Bus Application," SAE piston, connecting rod, crankshaft, cams and camshaft, valves, intake and exhaust paper 852321, 1985. manifolds. 16. Alperstein, M ., Schafer, G. H ., and Villforth, F. J.: "Texaco's Stratified Charge Engine-Multifuel, 1.2. Indicate on an appropriate sketch the different forces that act on the piston, and the Efficient, Clean, and Practical," SAE paper 740563, 1974. direction of these forces, during the engine's expansion stroke with the piston, con- 17. Urlaub, A. G ., and Chmela, F. G.: "High-Speed, Multifuel Engine: L9204 FMV," SAE paper 740122, 1974. necting rod, and crank in the positions shown in Fig. 1-1. 18. Date, T ., and Yagi, S.: "Research and Development of the Honda CVCC Engine," SAE paper 1.3. List five important differences between the design and operating characteristics of 740605, 1974. spark-ignition and compression-ignition (diesel) engines. 1.4. Indicate the approximate crank angle at which the following events in the four-stroke and two-stroke internal combustion engine cycles occur on a line representing the full cycle (720º for the four-stroke cycle; 360" for the two-stroke cycle): bottom- and top- center crank positions, inlet and exhaust valve or port opening and closing, start of combustion process, end of combustion process, maximum cylinder pressure. ENGINE DESIGN AND OPERATING PARAMETERS 43 CHAPTER Engine performance is more precisely defined by: 2 1. The maximum power (or the maximum torque) available at each speed within the useful engine operating range 2. The range of speed and power over which engine operation is satisfactory The following performance definitions are commonly used: ENGINE DESIGN Maximum rated power. The highest power an engine is allowed to develop for short periods of operation. AND OPERATING Normal rated power. The highest power an engine is allowed to develop in PARAMETERS continuous operation. Rated speed. The crankshaft rotational speed at which rated power is devel- oped. 2.2 GEOMETRICAL PROPERTIES OF RECIPROCATING ENGINES The following parameters define the basic geometry of a reciprocating engine (see Fig. 2-1): Compression ratio re : maximum cylinder volume re = Vat Vc (2.1) 2.1 IMPORTANT ENGINE minimum cylinder volume CHARACTERISTICS where Va is the displaced or swept volume and Ve is the clearance volume. In this chapter, some basic geometrical relationships and the parameters com- Ratio of cylinder bore to piston stroke: monly used to characterize engine operation are developed. The factors impor- B tant to an engine user are: Rbs = = (2.2) 1. The engine's performance over its operating range Ratio of connecting rod length to crank radius: 2. The engine's fuel consumption within this operating range and the cost of the required fuel R == (2.3) 3. The engine's noise and air pollutant emissions within this operating range In addition, the stroke and crank radius are related by 4. The initial cost of the engine and its installation 5. The reliability and durability of the engine, its maintenance requirements, and L = 2a how these affect engine availability and operating costs Typical values of these parameters are: re = 8 to 12 for SI engines and re = 12 to 24 for CI engines; B/L = 0.8 to 1.2 for small- and medium-size engines, decreas -. These factors control total engine operating costs-usually the primary consider- ing to about 0.5 for large slow-speed CI engines; R = 3 to 4 for small- and ation of the user-and whether the engine in operation can satisfy environmental medium-size engines, increasing to 5 to 9 for large slow-speed.CI engines. regulations. This book is concerned primarily with the performance, efficiency, The cylinder volume V at any crank position 0 is and emissions characteristics of engines; the omission of the other factors listed above does not, in any way, reduce their great importance. B2 V = Vo+-A (1+a-s) (2.4) 42 44 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE DESIGN AND OPERATING PARAMETERS 45 rc 1.6 B 1.4 1.2 .... 1.0 BC 0.8 0.6- 0.4 - 0.2 R = l/a = 3.5 0 FIGURE 2-2 FIGURE 2-1 90 180 Instantaneous piston speed/mean piston speed Geometry of cylinder, piston, connecting rod, TC Crank angle, 0 BC as a function of crank angle for R = 3.5. and crankshaft where B = bore, L = stroke, I = connecting road length, a = crank radius, 0 = crank angle. more appropriate parameter than crank rotational speed for correlating engine behavior as a function of speed. For example, gas-flow velocities in the intake where s is the distance between the crank axis and the piston pin axis (Fig. 2-1), and the cylinder all scale with S ,. The instantaneous piston velocity S, is obtained and is given by from s = a cos 0 + (12 - a2 sin2 0)1/2 (2.5) Sp= ds at (2.10) The angle 0, defined as shown in Fig. 2-1, is called the crank angle. Equation (2.4) with the above definitions can be rearranged: The piston velocity is zero at the beginning of the stroke, reaches a maximum near the middle of the stroke, and decreases to zero at the end of the stroke. V . Differentiation of Eq. (2.5) and substitution gives V ; = 1 + + (re - 1)[R + 1 - cos 0 - (R2 - sin2 0)1/2] (2.6) COS 0 The combustion chamber surface area A at any crank position 0 is given by Sp = [ sin 0 1 + [ R2 - Sin 2 0) 1/2 (2.11) A = Ach + Ap + B(l + a -s) (2.7) Figure 2-2 shows how S, varies over each stroke for R = 3.5. Resistance to gas flow into the engine or stresses due to the inertia of the where Ach is the cylinder head surface area and A, is the piston crown surface moving parts limit the maximum mean piston speed to within the range 8 to 15 area. For flat-topped pistons, A, = AB2/4. Using Eq. (2.5), Eq. (2-7) can be rear- m/s (1500 to 3000 ft/min). Automobile engines operate at the higher end of this ranged: range; the lower end is typical of large marine diesel engines. A = Ach + Ap + TBL 2= [R + 1 - cos 0 - (R2 - sin2 0)1/2] (2.8) 2.3 BRAKE TORQUE AND POWER An important characteristic speed is the mean piston speed S,: Engine torque is normally measured with a dynamometer.1 The engine is S, = 2LN (2.9) clamped on a test bed and the shaft is connected to the dynamometer rotor. where N is the rotational speed of the crankshaft. Mean piston speed is often a Figure 2-3 illustrates the operating principle of a dynamometer. The rotor is 46. INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE DESIGN AND OPERATING PARAMETERS 47 2-stroke 4-stroke 4-stroke Stator Force F Rotor EVOL N + Blowdown Cylinder pressure Cylinder pressure Compression- Load Cylinder pressure Expansion cell Ignition IVO EO IO Exhaust FIGURE 2-3 B -C Schematic of principle of operation of dynamometer. B Intake IVCS TC Vol. BC TC Compression Vol. BC TC Vol. BC coupled electromagnetically, hydraulically, or by mechanical friction to a stator, (a) (b) (c) which is supported in low friction bearings. The stator is balanced with the rotor stationary. The torque exerted on the stator with the rotor turning is measured FIGURE 2-4 by balancing the stator with weights, springs, or pneumatic means. Examples of p-V diagrams for (a) a two-stroke cycle engine, (b) a four-stroke cycle engine; (c) a four-stroke cycle spark-ignition engine exhaust and intake strokes (pumping loop) at part load. Using the notation in Fig. 2-3, if the torque exerted by the engine is T: T = Fb (2.12) The power P delivered by the engine and absorbed by the dynamometer is the area enclosed on the diagram: product of torque and angular speed: P = 2ANT (2.13a) Wo.i = ppav (2.14) where N is the crankshaft rotational speed. In SI units: With two-stroke cycles (Fig. 2-4a), the application of Eq. (2.14) is straightforward. P(KW) = 2zN(rev/s)T(N . m) x 10-3 (2.13b) With the addition of inlet and exhaust strokes for the four-stroke cycle, some ambiguity is introduced as two definitions of indicated output are in common or in U.S. units: use. These will be defined as: P(hp) = (rev/min) T(lbf. ft) 5252 (2.13c) Gross indicated work per cycle We,ig . Work delivered to the piston over the compression and expansion strokes only. Note that torque is a measure of an engine's ability to do work; power is the rate Net indicated work per cycle We.in . Work delivered to the piston over the at which work is done. entire four-stroke cycle. The value of engine power measured as described above is called brake power Po. This power is the usable power delivered by the engine to the load-in this case, a "brake." In Fig. 2-4b and c, We,it is (area A + area C) and We,in is (area A + area C) - (area B + area C), which equals (area A - area B), where each of these areas is regarded as a positive quantity. Area B + area C is the work transfer between the 2.4 INDICATED WORK PER CYCLE piston and the cylinder gases during the inlet and exhaust strokes and is called the pumping work W, (see Chaps. 5 and 13). The pumping work transfer will be to Pressure data for the gas in the cylinder over the operating cycle of the engine the cylinder gases if the pressure during the intake stroke is less than the pressure can be used to calculate the work transfer from the gas to the piston. The cylin- during the exhaust stroke. This is the situation with naturally aspirated engines. der pressure and corresponding cylinder volume throughout the engine cycle can The pumping work transfer will be from the cylinder gases to the piston if the be plotted on a p-V diagram as shown in Fig. 2-4. The indicated work per cycle exhaust stroke pressure is lower than the intake pressure, which is normally the We.it (per cylinder) is obtained by integrating around the curve to obtain the case with highly loaded turbocharged engines.+ The term indicated is used because such p-V diagrams used to be generated directly with a device + With some two-stroke engine concepts there is a piston pumping work term associated with com- called an engine indicator. pressing the scavenging air in the crankcase. 48 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE DESIGN AND OPERATING PARAMETERS 49 The power per cylinder is related to the indicated work per cycle by supplied by the dynamometer to overcome all these frictional losses. The engine Pi = We. N (2.15) speed, throttle setting, oil and water temperatures, and ambient conditions are n R kept the same in the motored test as under firing conditions. The major sources where nr is the number of crank revolutions for each power stroke per cylinder. of inaccuracy with this method are that gas pressure forces on the piston and For four-stroke cycles, ng equals 2; for two-stroke cycles, n equals 1. This power rings are lower in the motored test than when the engine is firing and that the oil is the indicated power; i.e ., the rate of work transfer from the gas within the temperatures on the cylinder wall are also lower under motoring conditions. The ratio of the brake (or useful) power delivered by the engine to the cylinder to the piston. It differs from the brake power by the power absorbed in overcoming engine friction, driving engine accessories, and (in the case of gross indicated power is called the mechanical efficiency nm : indicated power) the pumping power. In discussing indicated quantities of the four-stroke cycle engine, such as Pb = 1 - Pig P 18 (2.17) work per cycle or power, the definition used for "indicated" (i.e ., gross or net) should always be explicitly stated. The gross indicated output, the definition most Since the friction power includes the power required to pump gas into and out of commonly used, will be chosen where possible in this book for the following the engine, mechanical efficiency depends on throttle position as well as engine reasons. Indicated quantities are used primarily to identify the impact of the com- design and engine speed. Typical values for a modern automotive engine at wide- pression, combustion, and expansion processes on engine performance, etc. The open or full throttle are 90 percent at speeds below about 30 to 40 rev/s (1800 to gross indicated output is, therefore, the most appropriate definition. It represents 2400 rev/min), decreasing to 75 percent at maximum rated speed. As the engine is the sum of the useful work available at the shaft and the work required to over- throttled, mechanical efficiency decreases, eventually to zero at idle operation. come all the engine losses. Furthermore, the standard engine test codes2 define procedures for measuring brake power and friction power (the friction power test provides a close approximation to the total lost power in the engine). The sum of 2.6 ROAD-LOAD POWER brake power and friction power provides an alternative way of estimating indi- cated power; the value obtained is a close approximation to the gross indicated A part-load power level useful as a reference point for testing automobile engines power. is the power required to drive a vehicle on a level road at a steady speed. Called The terms brake and indicated are used to describe other parameters such road-load power, this power overcomes the rolling resistance which arises from as mean effective pressure, specific fuel consumption, and specific emissions (see the friction of the tires and the aerodynamic drag of the vehicle. Rolling resist- the following sections) in a manner similar to that used for work per cycle and ance and drag coefficients, CR and CD, respectively, are determined empirically. An approximate formula for road-load power P, is power. P. = (CRM, 9 + EPa CDA, S2)S. (2.18a) 2.5 MECHANICAL EFFICIENCY where CR = coefficient of rolling resistance (0.012 < CR < 0.015)3 We have seen that part of the gross indicated work per cycle or power is used to My = mass of vehicle [for passenger cars: curb mass plus passenger load of expel exhaust gases and induct fresh charge. An additional portion is used to 68 kg (150 lbm); in U.S. units W, = vehicle weight in lbf] overcome the friction of the bearings, pistons, and other mechanical components g = acceleration due to gravity of the engine, and to drive the engine accessories. All of these power requirements Pa = ambient air density are grouped together and called friction power P .. + Thus: CD = drag coefficient (for cars: 0.3 < CD < 0.5)3 A, = frontal area of vehicle Pig = P; + PS (2.16) S, = vehicle speed Friction power is difficult to determine accurately. One common approach for high-speed engines is to drive or motor the engine with a dynamometer (i.e ., With the quantities in the units indicated: operate the engine without firing it) and measure the power which has to be P,(KW) = [2.73CM,(kg) + 0.0126CD A,(m2)S,(km/h)2]S,(km/h) x 10-3 (2.18b) + The various components of friction power are examined in detail in Chap. 13. or P.(hp) = CR W.(Ibf) + 0.0025CD A,(ft2)S,(mi/h)2]S,(mi/h) 375 (2.18c) 50 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE DESIGN AND OPERATING PARAMETERS 51 2.7 MEAN EFFECTIVE PRESSURE lb/in2) range, with the bmep at the maximum rated power of about 700 kPa (100 While torque is a valuable measure of a particular engine's ability to do work, it lb/in2). Turbocharged four-stroke diesel maximum bmep values are typically in depends on engine size. A more useful relative engine performance measure is the range 1000 to 1200 kPa (145 to 175 lb/in2); for turbocharged aftercooled obtained by dividing the work per cycle by the cylinder volume displaced per engines this can rise to 1400 kPa. At maximum rated power, bmep is about 850 cycle. The parameter so obtained has units of force per unit area and is called the to 950 kPa (125 to 140 lb/in2). Two-stroke cycle diesels have comparable per- mean effective pressure (mep). Since, from Eq. (2.15), formance to four-stroke cycle engines. Large low-speed two-stroke cycle engines can achieve bmep values of about 1600 kPa. Work per cycle = = " An example of how the above engine performance parameters can be used N to initiate an engine design is given below. where np is the number of crank revolutions for each power stroke per cylinder (two for four-stroke cycles; one for two-stroke cycles), then Example. A four-cylinder automotive spark-ignition engine is being designed to provide a maximum brake torque of 150 N .m (110 Ibf . ft) in the mid-speed range PAR mep = (2.19a) ( ~ 3000 rev/min). Estimate the required engine displacement, bore and stroke, and VAN the maximum brake power the engine will deliver. Equation (2.20a) relates torque and mep. Assume that 925 kPa is an appropri- For SI and U.S. units, respectively, ate value for bmep at the maximum engine torque point. Equation (2.20a) gives P(KW)ng x 103 mep(kPa) = (2.19b) V(dm3) = 6.28m Tmax (N. m) 6.28 x 2 × 150 V.(dm3)N(rev/s) bmepmax(KPa) 925 = 2 dm3 mep(lb/in2) = = P(hp)n × 396,000 (2.19c) For a four-cylinder engine, the displaced volume, bore, and stroke are related by V (in3)N(rev/min) Mean effective pressure can also be expressed in terms of torque by using V2 = 4 x - B2L Eq. (2.13): Assume B = L; this gives B = L = 86 mm. mep(kPa) =- 6.28n RT(N . m) (2.20a) The maximum rated engine speed can be estimated from an appropriate value V.(dm3) for the maximum mean piston speed, 15 m/s (see Sec. 2.2): mep(lb/in2) = 75.4nRT(Ibf . ft) or (2.20b) Spmax = 2LN max -+ Nmax = 87 rev/s (5200 rev/min) VAin3) The maximum brake mean effective pressure of good engine designs is well The maximum brake power can be estimated from the typical bmep value at maximum power, 800 kPa (116 lb/in2), using Eq. (2.19b): established, and is essentially constant over a wide range of engine sizes. Thus, the actual bmep that a particular engine develops can be compared with this Pbmax(KW) = bmep(kPa)V(dm3)N max(rev/s) 800 x 2 x 87 norm, and the effectiveness with which the engine designer has used the engine's NR X 103 2 × 103 - = 70 kw displaced volume can be assessed. Also, for design calculations, the engine dis- placement required to provide a given torque or power, at a specified speed, can be estimated by assuming appropriate values for bmep for that particular appli- 2.8 SPECIFIC FUEL CONSUMPTION cation. AND EFFICIENCY Typical values for bmep are as follows. For naturally aspirated spark- ignition engines, maximum values are in the range 850 to 1050 kPa ( ~ 125 to In engine tests, the fuel consumption is measured as a flow rate-mass flow per 150 lb/in2) at the engine speed where maximum torque is obtained (about 3000 unit time m. A more useful parameter is the specific fuel consumption (sfc)-the rev/min). At the maximum rated power, bmep values are 10 to 15 percent lower. fuel flow rate per unit power output. It measures how efficiently an engine is For turbocharged automotive spark-ignition engines the maximum bmep is in using the fuel supplied to produce work: the 1250 to 1700 kPa (180 to 250 lb/in2) range. At the maximum rated power, bmep is in the 900 to 1400 kPa (130 to 200 lb/in2) range. For naturally aspirated sfc =- four-stroke diesels, the maximum bmep is in the 700 to 900 kPa (100 to 130 : P (2.21) 52 ENGINE DESIGN AND OPERATING PARAMETERS 53 INTERNAL COMBUSTION ENGINE FUNDAMENTALS With units, or with units: 1 m (g/s) ng (2.24b) sfc(mg/J) = P (KW) (2.22a) sfc(mg/J)QHv(MJ/kg) 3600 m (g/h) sfc(g/kW . h)QHv(MJ/kg) (2.24c) or sfc(g/kW . h) == P(KW) = 608.3 sfc(lbm/hp . h) (2.22b) 2545 ns = or sfc(lbm/hp . h) =: im (Ibm/h) = 1.644 x 103 sfc(g/kW . h) (2.22c) sfc(Ibm/hp . h)QHv(Btu/lbm) (2.24d) P(hp) Typical heating values for the commercial hydrocarbon fuels used in Low values of sfc are obviously desirable. For SI engines typical best values of engines are in the range 42 to 44 MJ/kg (18,000 to 19,000 Btu/lbm). Thus, specific brake specific fuel consumption are about 75 ug/J = 270 g/kW . h = 0.47 1bm/ fuel consumption is inversely proportional to fuel conversion efficiency for normal hydrocarbon fuels. hp . h. For CI engines, best values are lower and in large engines can go below 55 Note that the fuel energy supplied to the engine per cycle is not fully re- ug/J = 200 g/kW . h = 0.32 Ibm/hp . h. The specific fuel consumption has units. A dimensionless parameter that leased as thermal energy in the combustion process because the actual com- bustion process in incomplete. When enough air is present in the cylinder to relates the desired engine output (work per cycle or power) to the necessary input oxidize the fuel completely, almost all (more than about 96 percent) of this fuel (fuel flow) would have more fundamental value. The ratio of the work produced per cycle to the amount of fuel energy supplied per cycle that can be released in energy supplied is transferred as thermal energy to the working fluid. When insuf- ficient air is present to oxidize the fuel completely, lack of oxygen prevents this the combustion process is commonly used for this purpose. It is a measure of the fuel energy supplied from being fully released. This topic is discussed in more engine's efficiency. The fuel energy supplied which can be released by combustion detail in Secs. 3.5 and 4.9.4. is given by the mass of fuel supplied to the engine per cycle times the heating value of the fuel. The heating value of a fuel, QHy, defines its energy content. It is determined in a standardized test procedure in which a known mass of fuel is 2.9 AIR/FUEL AND FUEL/AIR RATIOS fully burned with air, and the thermal energy released by the combustion process is absorbed by a calorimeter as the combustion products cool down to their In engine testing, both the air mass flow rate ma and the fuel mass flow rate m, original temperature. are normally measured. The ratio of these flow rates is useful in defining engine This measure of an engine's "efficiency," which will be called the fuel con- operating conditions: version efficiency ns,t is given by Air/fuel ratio (A/F) = ima im (2.25) (PnR/N) P ns (2.23) my QHV (m/ nR/N)OHV my 2HV Fuel/air ratio (F/A) = = ima (2.26) where my is the mass of fuel inducted per cycle. Substitution for P/m, from Eq. The normal operating range for a conventional SI engine using gasoline fuel is (2.21) gives 12 < A/F < 18 (0.056 < F/A < 0.083); for CI engines with diesel fuel, it is 1 18 < A/F < 70 (0.014 < F/A < 0.056). (2.24a) sfc QHV 2.10 VOLUMETRIC EFFICIENCY The intake system-the air filter, carburetor, and throttle plate (in a spark- + This empirically defined engine efficiency has previously been called thermal efficiency or enthalpy ignition engine), intake manifold, intake port, intake valve-restricts the amount efficiency. The term fuel conversion efficiency is preferred because it describes this quantity more of air which an engine of given displacement can induct. The parameter used to precisely, and distinguishes it clearly from other definitions of engine efficiency which will be devel- measure the effectiveness of an engine's induction process is the volumetric effi- oped in Sec. 3.6. Note that there are several different definitions of heating value (see Sec. 3.5). The ciency no. Volumetric efficiency is only used with four-stroke cycle engines which numerical values do not normally differ by more than a few percent, however. In this text, the lower have a distinct induction process. It is defined as the volume flow rate of air into heating value at constant pressure is used in evaluating the fuel conversion efficiency. 54 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE DESIGN AND OPERATING PARAMETERS 55 the intake system divided by the rate at which volume is displaced by the piston: conditions used are: 2m. (2.27a) Dry air pressure Water vapour pressure Temperature Pa,i Va N 736.6 mmHg 9.65 mmHg 29.4ºC where Pa, is the inlet air density. An alternative equivalent definition for volu- 29.00 inHg 0.38 inHg 5ºF metric efficiency is The basis for the correction factor is the equation for one-dimensional ma (2.27b) steady compressible flow through an orifice or flow restriction of effective area AF Paiva (see App. C): where ma is the mass of air inducted into the cylinder per cycle. im = - AEDO { 27 The inlet density may either be taken as atmosphere air density (in which VRT. 6 -TL( 2 ) ( 2 ) + Xxx72112 (2.30) case 1, measures the pumping performance of the entire inlet system) or may be In deriving this equation, it has been assumed that the fluid is an ideal gas with taken as the air density in the inlet manifold (in which case n, measures the gas constant R and that the ratio of specific heats (c./c, = y) is a constant; po and pumping performance of the inlet port and valve only). Typical maximum values To are the total pressure and temperature upstream of the restriction and p is the of n, for naturally aspirated engines are in the range 80 to 90 percent. The volu- pressure at the throat of the restriction. metric efficiency for diesels is somewhat higher than for SI engines. Volumetric If, in the engine, p/po is assumed constant at wide-open throttle, then for a efficiency is discussed more fully in Sec. 6.2. given intake system and engine, the mass flow rate of dry air m. varies as 2.11 ENGINE SPECIFIC WEIGHT AND (2.31) SPECIFIC VOLUME For mixtures containing the proper amount of fuel to use all the air avail- Engine weight and bulk volume for a given rated power are important in many able (and thus provide maximum power), the indicated power at full throttle P; applications. Two parameters useful for comparing these attributes from one will be proportional to ma, the dry air flow rate. Thus if engine to another are: Pi,3 = Cy Pi.m (2.32) Specific weight = engine weight (2.28) where the subscripts s and m denote values at the standard and measured condi- rated power tions, respectively, the correction factor CF is given by 1/2 Specific volume - engine volume (2.29) CF = Ps .d Pm - Po.m (2.33) rated power For these parameters to be useful in engine comparisons, a consistent definition where p, a = standard dry-air absolute pressure of what components and auxiliaries are included in the term "engine" must be Pm = measured ambient-air absolute pressure adhered to. These parameters indicate the effectiveness with which the engine Pr.m = measured ambient-water vapour partial pressure designer has used the engine materials and packaged the engine components.4 Tm = measured ambient temperature, K T; = standard ambient temperature, K The rated brake power is corrected by using Eq. (2.33) to correct the indi- 2.12 CORRECTION FACTORS FOR cated power and making the assumption that friction power is unchanged. Thus POWER AND VOLUMETRIC EFFICIENCY Pb, = C; Pi,m - Pf.m (2.34) The pressure, humidity, and temperature of the ambient air inducted into an Volumetric efficiency is proportional to ma/p. [see Eq. (2.27)]. Since pa is engine, at a given engine speed, affect the air mass flow rate and the power proportional to p/T, the correction factor for volumetric efficiency, C'F, is output. Correction factors are used to adjust measured wide-open-throttle power and volumetric efficiency values to standard atmospheric conditions to provide a 1/2 more accurate basis for comparisons between engines. Typical standard ambient no .m (2.35) 56 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE DESIGN AND OPERATING PARAMETERS 57 2.13 SPECIFIC EMISSIONS AND For four-stroke cycle engines, volumetric efficiency can be introduced: EMISSIONS INDEX Levels of emissions of oxides of nitrogen (nitric oxide, NO, and nitrogen dioxide, p = "s n. NVa QHV Pa. (F/A) 2 (2.39) NO2, usually grouped together as NO„), carbon monoxide (CO), unburned hydrocarbons (HC), and particulates are important engine operating character- For torque T: istics. The concentrations of gaseous emissions in the engine exhaust gases are T = "no Va LHV P .. (F/A) (2.40) usually measured in parts per million or percent by volume (which corresponds 4Tt to the mole fraction multiplied by 106 or by 102, respectively). Normalized indi- For mean effective pressure: cators of emissions levels are more useful, however, and two of these are in common use. Specific emissions are the mass flow rate of pollutant per unit power mep = n, no CHv Pa, (F/A) (2.41) output: The power per unit piston area, often called the specific power, is a measure of the mNO, SNO. engine designer's success in using the available piston area regardless of cylinder x (2.36a) P size. From Eq. (2.39), the specific power is mco SCO : (2.36b) Pngn. NLQHV P .. (F/A) P Ap 2 (2.42) MHC (2.36c) Mean piston speed can be introduced with Eq. (2.9) to give SHC = P P no no Se CHv Pa. (F/A) Ap (2.43) sPart =- mpart (2.36d) P Specific power is thus proportional to the product of mean effective pressure and Indicated and brake specific emissions can be defined. Units in common use are mean piston speed. ug/J, g/kW . h, and g/hp . h. These relationships illustrate the direct importance to engine performance Alternatively, emission rates can be normalized by the fuel flow rate. An of: emission index (EI) is commonly used: e.g ., 1. High fuel conversion efficiency EINO, = IMNO,(g/s) (2.37) 2. High volumetric efficiency im,(kg/s) 3. Increasing the output of a given displacement engine by increasing the inlet air with similar expressions for CO, HC, and particulates. density 4. Maximum fuel/air ratio that can be usefully burned in the engine 2.14 RELATIONSHIPS BETWEEN 5. High mean piston speed PERFORMANCE PARAMETERS The importance of the parameters defined in Secs. 2.8 to 2.10 to engine per- 2.15 ENGINE DESIGN AND formance becomes evident when power, torque, and mean effective pressure are PERFORMANCE DATA expressed in terms of these parameters. From the definitions of engine power [Eq. (2.13)], mean effective pressure [Eq. (2.19)], fuel conversion efficiency [Eq. Engine ratings usually indicate the highest power at which manufacturers expect (2.23)], fuel/air ratio [Eq. (2.26)], and volumetric efficiency [Eq. (2.27)], the fol- their products to give satisfactory economy, reliability, and durability under lowing relationships between engine performance parameters can be developed. service conditions. Maximum torque, and the speed at which it is achieved, is For power P: usually given also. Since both of these quantities depend on displaced volume, for comparative analyses between engines of different displacements in a given P = Ism. NOHv(F/A) (2.38) engine category normalized performance parameters are more useful. The follow- nR Ing measures, at the operating points indicated, have most significance:4 58 TABLE 2.1 Typical design and operating data for internal combustion engines Rated maximum Weight/ Approx. Power per power best Operating Compression Bore Stroke/ Speed, bmep, unit volume ratio, bsfc, cycle ratio m bore rev/min atm kW/dm3 kg/k W g/kW . h Spark-ignition engines: Small (e.g ., motorcycles) 28,4 6-11 0.05-0.085 1.2-0.9 4500-7500 4-10 20-60 5.5-2.5 350 Passenger cars 4S 8-10 0.07-0.1 1.1-0.9 4500-6500 7-10 20-50 4-2 270 Trucks 4S 7-9 0.09-0.13 1.2-0.7 3600-5000 6.5-7 25-30 6.5-2.5 300 Large gas engines 25,4S 8-12 0.22-0.45 1.1-1.4 300-900 6.8-12 3-7 23-35 200 Wankel engines 4S 22 0.57 dm3 per chamber 6000-8000 9.5-10.5 35-4 1.6-0.9 300 Diesel engines: Passenger cars 4S 17-23 0.075-0.1 1.2-0.9 4000-5000 5-7.5 18-22 5 - 2.5 250 Trucks (NA) 4S 16-22 0.1-0.15 1.3-0.8 2100-4000 6-9 15-22 7-4 210 Trucks (TC) 4S 14-20 0.1-0.15 1.3-0.8 2100-4000 12-18 7-3.5 200 Locomotive, 4S,2S 12-18 0.15-0.4 1.1-1.3 425-1800 7-23 5-20 6-18 190 ,industrial, marine Large engines, 2S 10-12 0.4-1 1.2-3 110-400 9-17 2-8 12-50 180 marine and stationary used . PROBLEMS has been utilized. maximum torque. maximum fuel-pump setting: 1. At maximum or normal rated point: Brake specific emissions. engine is run for long periods of time: flow and use it effectively over the full range. piston area is used, regardless of cylinder size. fuel conversion efficiencies of over 55 percent can be obtained. of success in handling higher gas pressures and thermal loading. to inertia of the parts, resistance to air flow, and/or engine friction. Brake specific fuel consumption or fuel conversion efficiency. 2.1. Explain why the brake mean effective pressure of a naturally aspirated diesel engine diesel engines, brake fuel conversion efficiencies of about 50 percent and indicated charged and supercharged engines is higher than for naturally aspirated engines. of about 8 to 15 m/s. The maximum brake mean effective pressure for turbo- engine size increases, maintaining the maximum mean piston speed in the range cycle dominates except in the smallest and largest engine sizes. The larger engines 2. At all speeds at which the engine will be used with full throttle or with diesels, their naturally aspirated maximum bmep levels are higher. As engine size normal production size range are summarized in Table 2.1.4 The four-stroke increases, due to reduced importance of heat losses and friction. For the largest 3. At all useful regimes of operation and particularly in those regimes where the increases, brake specific fuel consumption decreases and fuel conversion efficiency are turbocharged or supercharged. The maximum rated engine speed decreases as Because the maximum fuel/air ratio for spark-ignition engines is higher than for stress limited. It then reflects the product of volumetric efficiency (ability to fuel conversion efficiency. In supercharged engines bmep indicates the degree induct air), fuel/air ratio (effectiveness of air utilization in combustion), and bmep is lower at the maximum rated power for a given engine than the bmep at the is lower than that of a naturally aspirated spark-ignition engine. Explain why the Typical performance data for spark-ignition and diesel engines over the Specific weight. Indicates relative economy with which materials are Brake mean effective pressure. In naturally aspirated engines bmep is not Mean piston speed. Measures comparative success in handling loads due Power per unit piston area. Measures the effectiveness with which the Specific volume. Indicates relative effectiveness with which engine space Brake mean effective pressure. Measures ability to obtain/provide high air ENGINE DESIGN AND OPERATING PARAMETERS 59 60 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE DESIGN AND OPERATING PARAMETERS 61 2.2. Describe the impact on air flow, maximum torque, and maximum power of changing a spark-ignition engine cylinder head from 2 valves per cylinder to 4 valves (2 inlet for your design? What would be the brake torque (N . m) and the fuel flow rate (g/h) and 2 exhaust) per cylinder. at this maximum speed? Assume a maximum mean piston speed of 12 m/s is typical of good engine designs. 2.3. Calculate the mean piston speed, bmep, and specific power of the spark-ignition engines in Figs. 1-4, 1-9, and 1-12 at their maximum rated power. 2.12. The power per unit piston area P/A, (often called the specific power) is a measure of 2.4. Calculate the mean piston speed, bmep, and specific power of the diesel engines in the designer's success in using the available piston area regardless of size. Figs. 1-20, 1-21, 1-22, 1-23, and 1-24 at their maximum rated power. Briefly explain (a). Derive an expression for P/A, in terms of mean effective pressure and mean any significant differences. piston speed for two-stroke and four-stroke engine cycles. 2.5. Develop an equation for the power required to drive a vehicle at constant speed up a (b) Compute typical maximum values of P/A, for a spark-ignition engine (e.g ., Fig. hill of angle a, in terms of vehicle speed, mass, frontal area, drag coefficient, coeffi- 1-4), a turbocharged four-stroke cycle diesel engine (e.g ., Fig. 1-22), and a large cient of rolling resistance, a, and acceleration due to gravity. Calculate this power marine diesel (Fig. 1-24). Table 2-1 may be helpful. State your assumptions when the car mass is 1500 kg, the hill angle is 15 degrees, and the vehicle speed is clearly. 50 mi/h. 2.13. Several velocities, time, and length scales are useful in understanding what goes on 2.6. The spark-ignition engine in Fig. 1-4 is operating at a mean piston speed of 10 m/s. inside engines. Make estimates of the following quantities for a 1.6-liter displacement The measured air flow is 60 g/s. Calculate the volumetric efficiency based on atmo- four-cylinder spark-ignition engine, operating at wide-open throttle at 2500 rev/min. spheric conditions. (a) The mean piston speed and the maximum piston speed. 2.7. The diesel engine of Fig. 1-20 is operating with a mean piston speed of 8 m/s. Calcu- (b) The maximum charge velocity in the intake port (the port area is about 20 late the air flow if the volumetric efficiency is 0.92. If (F/A) is 0.05 what is the fuel percent of the piston area). flow rate, and the mass of fuel injected per cylinder per cycle? (c) The time occupied by one engine operating cycle, the intake process, the com- 2.8. The brake fuel conversion efficiency of a spark-ignition engine is 0.3, and varies little pression process, the combustion process, the expansion process, and the exhaust with fuel type. Calculate the brake specific fuel consumption for isooctane, gasoline, process. (Note: The word process is used here not the word stroke.) methanol, and hydrogen (relevant data are in App. D). (d) The average velocity with which the flame travels across the combustion 2.9. You are doing a preliminary design study of a turbocharged four-stroke diesel chamber. engine. The maximum rated power is limited by stress considerations to a brake (e) The length of the intake system (the intake port, the manifold runner, etc.) which mean effective pressure of 1200 kPa and maximum value of the mean piston speed of is filled by one cylinder charge just before the intake valve opens and this charge 12 m/s. enters the cylinder (i.e ., how far back from the intake valve, in centimeters, one (a) Derive an equation relating the engine inlet pressure (pressure in the inlet mani- cylinder volume extends in the intake system). fold at the turbocharger compressor exit) to the fuel/air ratio at this maximum (f) The length of exhaust system filled by one cylinder charge after it exits the cylin- rated power operating point. Other reciprocating engine parameters (e.g ., volu- der (assume an average exhaust gas temperature of 425ºC). metric efficiency, fuel conversion efficiency, bmep, etc.) appear in this equation You will have to make several appropriate geometric assumptions. The calculations also. are straightforward, and only approximate answers are required. (b) The maximum rated brake power requirement for this engine is 400 kW. Esti- 2.14. The values of mean effective pressure at rated speed, maximum mean piston speed, mate sensible values for number of cylinders, cylinder bore, stroke, and deter- and maximum specific power (engine power/total piston area) are essentially inde- mine the maximum rated speed of this preliminary engine design. pendent of cylinder size for naturally aspirated engines of a given type. If we also (c) If the pressure ratio across the compressor is 2, estimate the overall fuel/air and assume that engine weight per unit displaced volume is essentially constant, how will air/fuel ratios at the maximum rated power. Assume appropriate values for any the specific weight of an engine (engine weight/maximum rated power) at fixed total other parameters you may need. displaced volume vary with the number of cylinders? Assume the bore and stroke 2.10. In the reciprocating engine, during the power or expansion stroke, the gas pressure are equal. force acting on the piston is transmitted to the crankshaft via the connecting rod. List the forces acting on the piston during this part of the operating cycle. Show the REFERENCES direction of the forces acting on the piston on a sketch of the piston, cylinder, con- necting rod, crank arrangement. Write out the force balance for the piston (a) along 1. Obert, E.F.: Internal Combustion Engines and Air Pollution, chap. 2, Intext Educational Publishers, the cylinder axis and (b) transverse to the cylinder axis in the plane containing the New York, 1973. connecting rod. (You are not asked to manipulate or solve these equations.) 2. SAE Standard: " Engine Test Code-Spark Ignition and Diesel," SAE J816b, SAE Handbook. 3. Bosch: Automotive Handbook, 2nd English edition, Robert Bosch GmbH, Stuttgart, 1986. 2.11. You are designing a four-stroke cycle diesel engine to provide a brake power of 300 4. Taylor, C.F.: The Internal Combustion Engine in Theory and Practice, vol. II, MIT Press, Cam- kW naturally aspirated at its maximum rated speed. Based on typical values for bridge, Mass ., 1968. brake mean effective pressure and maximum mean piston speed, estimate the required engine displacement, and the bore and stroke for sensible cylinder geometry and number of engine cylinders. What is the maximum rated engine speed (rev/min) THERMOCHEMISTRY OF FUEL-AIR MIXTURES. 63 CHAPTER® injected and vaporized fuel with this hot air starts the combustion process, which spreads rapidly. Burning then proceeds as fuel and air mix to the appropriate composition for combustion to take place. Thus, fuel-air mixing plays a control- 3 ling role in the diesel combustion process. Chapters 3 and 4 focus on the thermochemistry of combustion: i.e ., the composition and thermodynamic properties of the pre- and postcombustion working fluids in engines and the energy changes associated with the combustion processes that take place inside the engine cylinder. Later chapters (9 and 10) deal THERMOCHEMISTRY with the phenomenological aspects of engine combustion: i.e ., the details of the OF FUEL-AIR physical and chemical processes by which the fuel-air mixture is converted to burned products. At this point it is useful to review briefly the key combustion MIXTURES phenomena which occur in engines to provide an appropriate background for the material which follows. More detailed information on these combustion pheno- mena can be found in texts on combustion such as those of Fristrom and Westenberg1 and Glassman.2 The combustion process is a fast exothermic gas-phase reaction (where oxygen is usually one of the reactants). A flame is a combustion reaction which can propagate subsonically through space; motion of the flame relative to the unburned gas is the important feature. Flame structure does not depend on whether the flame moves relative to the observer or remains stationary as the gas moves through it. The existence of flame motion implies that the reaction is con- fined to a zone which is small in thickness compared to the dimensions of the apparatus-in our case the engine combustion chamber. The reaction zone is 3.1 CHARACTERIZATION OF FLAMES usually called the flame front. This flame characteristic of spatial propagation is the result of the strong coupling between chemical reaction, the transport pro- Combustion of the fuel-air mixture inside the engine cylinder is one of the pro- cesses of mass diffusion and heat conduction, and fluid flow. The generation of cesses that controls engine power, efficiency, and emissions. Some background in heat and active species accelerate the chemical reaction; the supply of fresh reac- relevant combustion phenomena is therefore a necessary preliminary to under- tants, governed by the convection velocity, limits the reaction. When these pro- standing engine operation. These combustion phenomena are different for the cesses are in balance, a steady-state flame results.1 two main types of engines-spark-ignition and diesel-which are the subject of Flames are usually classified according to the following overall character- this book. In spark-ignition engines, the fuel is normally mixed with air in the istics. The first of these has to do with the composition of the reactants as they engine intake system. Following the compression of this fuel-air mixture, an elec- enter the reaction zone. If the fuel and oxidizer are essentially uniformly mixed trical discharge initiates the combustion process; a flame develops from the together, the flame is designated as premixed. If the reactants are not premixed "kernal" created by the spark discharge and propagates across the cylinder to and must mix together in the same region where reaction takes place, the flame is the combustion chamber walls. At the walls, the flame is "quenched" or extin- called a diffusion flame because the mixing must be accomplished by a diffusion guished as heat transfer and destruction of active species at the wall become the process. The second means of classification relates to the basic character of the dominant processes. An. undesirable combustion phenomenon-the gas flow through the reaction zone: whether it is laminar or turbulent. In laminar "spontaneous" ignition of a substantial mass of fuel-air mixture ahead of the (or streamlined) flow, mixing and transport are done by molecular processes. flame, before the flame can propagate through this mixture (which is called the Laminar flows only occur at low Reynolds number. The Reynolds number end-gas)-can also occur. This autoignition or self-explosion combustion (density x velocity x lengthscale/viscosity) is the ratio of inertial to viscous phenomenon is the cause of spark-ignition engine knock which, due to the high forces. In turbulent flows, mixing and transport are enhanced (usually by a sub- pressures generated, can lead to engine damage. stantial factor) by the macroscopic relative motion of eddies or lumps of fluid In the diesel engine, the fuel is injected into the cylinder into air already at which are the characteristic feature of a turbulent (high Reynolds number) flow. high pressure and temperature, near the end of the compression stroke. The A third area of classification is whether the flame is steady or unsteady. The autoignition, or self-ignition, of portions of the developing mixture of already distinguishing feature here is whether the flame structure and motion change with 62 64 INTERNAL COMBUSTION ENGINE FUNDAMENTALS THERMOCHEMISTRY OF FUEL-AIR MIXTURES 65 TABLE 3.1 time. The final characterizing feature is the initial phase of the reactants-gas, Principle constitutents of dry air liquid, or solid. Flames in engines are unsteady, an obvious consequence of the internal Molecular Mole Molar combustion engine's operating cycle. Engine flames are turbulent. Only with sub- Gas ppm by volume weight fraction ratio stantial augmentation of laminar transport processes by the turbulent convection 209,500 31.998 0.2095 1 processes can mixing and burning rates and flame-propagation rates be made fast N, 780,900 28.012 0.7905 3.773 enough to complete the engine combustion process within the time available. Ar 9,300 39.948 The conventional spark-ignition flame is thus a premixed unsteady turbu- CO, 300 44.009 Air 1,000,000 28.962 lent flame, and the fuel-air mixture through which the flame propagates is in the 1.0000 4.773 gaseous state. The diesel engine combustion process is predominantly an unsteady turbulent diffusion flame, and the fuel is initially in the liquid phase. In combustion, oxygen is the reactive component of air. It is usually suffi- Both these flames are extremely complicated because they involve the coupling of ciently accurate to regard air as consisting of 21 percent oxygen and 79 percent the complex chemical mechanism, by which fuel and oxidizer react to form pro- inert gases taken as nitrogen (often called atmospheric or apparent nitrogen). For ducts, with the turbulent convective transport process. The diesel combustion each mole of oxygen in air there are process is even more complicated than the spark-ignition combustion process, because vaporization of liquid fuel and fuel-air mixing processes are involved too. 1 - 0.2095 = 3.773 Chapters 9 and 10 contain a more detailed discussion of the spark-ignition 0.2095 engine and diesel combustion processes, respectively. This chapter reviews the moles of atmospheric nitrogen. The molecular weight of air is obtained from basic thermodynamic and chemical composition aspects of engine combustion. Table 3.1 with Eq. (B.17) as 28.962, usually approximated by 29. Because atmo- spheric nitrogen contains traces of other species, its molecular weight is slightly 3.2 IDEAL GAS MODEL different from that of pure molecular nitrogen, i.e ., The gas species that make up the working fluids in internal combustion engines MON2 = 28.962 - 0.2095 x 31.998 = 28.16 (e.g ., oxygen, nitrogen, fuel vapor, carbon dioxide, water vapor, etc.) can usually 1 - 0.2095 be treated as ideal gases. The relationships between the thermodynamic proper- In the following sections, nitrogen will refer to atmospheric nitrogen and a ties of an ideal gas and of ideal gas mixtures are reviewed in App. B. There can be. molecular weight of 28.16 will be used. An air composition of 3.773 moles of found the various forms of the ideal gas law: nitrogen per mole of oxygen will be assumed. R T = nRT The density of dry air can be obtained from Eq. (3.1) with R = 8314.3 J/ PV = mRT = MM (3.1) kmol . K and M = 28.962: where p is the pressure, V the volume, m the mass of gas, R the gas constant for p(kg/m3) = 3.483 x 10-3p(Pa) (3.2a) the gas, T the temperature, R the universal gas constant, M the molecular weight, and n the number of moles. Relations for evaluating the specific internal energy u, enthalpy h, and entropy s, specific heats at constant volume c, and constant or p(Ibm/ft3) = : 2.699p(lbf/in2) (3.2b) pressure c ,, on a per unit mass basis and on a per mole basis (where the notation T(R) u, h, 3, 2 ,, and c, is used) of an ideal gas, are developed. Also given are equations Thus, the value for the density of dry air at 1 atmosphere (1.0133 x 105 Pa, for calculating the thermodynamic properties of mixtures of ideal gases. 14.696 lbf/in2) and 25ºC (77ºF) is 1.184 kg/m3 (0.0739 lbm/ft3). Actual air normally contains water vapor, the amount depending on tem- 3.3 COMPOSITION OF AIR AND FUELS perature and degree of saturation. Typically the proportion by mass is about 1 percent, though it can rise to about 4 percent under extreme conditions. The Normally in engines, fuels are burned with air. Dry air is a mixture of gases that relative humidity compares the water vapor content of air with that required to has a representative composition by volume of 20.95 percent oxygen, 78.09 saturate. It is defined as: percent nitrogen, 0.93 percent argon, and trace amounts of carbon dioxide, neon, helium, methane, and other gases. Table 3.1 shows the relative proportions of the The ratio of the partial pressure of water vapor actually present to the saturation major constituents of dry air.3 pressure at the same temperature. 66 INTERNAL COMBUSTION ENGINE FUNDAMENTALS THERMOCHEMISTRY OF FUEL-AIR MIXTURES 67 Water vapor content is measured with a wet- and dry-bulb psychrometer. 10 This consists of two thermometers exposed to a stream of moist air. The dry-bulb Normal air -R- temperature is the temperature of the air. The bulb of the other thermometer is wetted by a wick in contact with a water reservoir. The wet-bulb temperature is L Cp - lower than the dry-bulb temperature due to evaporation of water from the wick. % Variation from normal ait It is a good approximation to assume that the wet-bulb temperature is the adia- Y batic saturation temperature. Water vapor pressure can be obtained from -4 observed wet- and dry-bulb temperatures and a psychrometric chart such as Fig. ~k × 10 -6 3-1.4 The effect of humidity on the properties of air is given in Fig. 3-2.5 -8 The fuels most commonly used in internal combustion engines (gasoline or - 10 0 0.02 0.04 0.06 0.08 petrol, and diesel fuels) are blends of many different hydrocarbon compounds Kgwater Kgair obtained by refining petroleum or crude oil. These fuels are predominantly carbon and hydrogen (typically about 86 percent carbon and 14 percent hydro- FIGURE 3-2 gen by weight) though diesel fuels can contain up to about 1 percent sulfur. Other Effect of humidity on properties of air: R is the gas constant; c, and c, are specific heats at constant fuels of interest are alcohols (which contain oxygen), gaseous fuels (natural gas volume and pressure, respectively; y = c,/c,; k is the thermal conductivity. (From Taylor.5) and liquid petroleum gas), and single hydrocarbon compounds (e.g ., methane, propane, isooctane) which are often used in engine research. Properties of the molecular structure is necessary in order to understand combustion mecha- more common internal combustion engine fuels are summarized in App. D. nisms.º The different classes are as follows: Some knowledge of the different classes of organic compounds and their Alkyl Compounds Paraffins Single-bonded open-chain saturated hydrocarbon mol- (alkanes) ecules : i.e ., no more hydrogen can be added. For the larger 600 0.05 molecules straight-chain and branched-chain configu- 590 H H. rations exist. These are called normal (n-) and iso com- 7 580 111 38 H-C-C-H - pounds, respectively. Examples: CH4, methane; C2H6, 570 H H ethane; C3H8, propane; C8H18, n-octane and isooctane. 0.04 36 6 560 There are several "isooctanes," depending on the relative CH 2n+ 2 550 34 position of the branches. By isooctane is usually meant 100% 2,2,4-trimethylpentane, indicating five carbon atoms in the 5 540 mixture enthalpy , KJ/kg ., 32 90% Pwater 0.03 straight chain (pentane) with three methyl (CH3) branches kPa 530 T adiabatic saturation (wet-bulb) and dew-point 'temperature. . 70% located respectively at C-atoms 2, 2, and 4. Radicals defi- 5.20 kg water 600 .Kgair cient in one hydrogen take the name methyl, ethyl, propyl, 510 26 0.02 etc. 1.15 1.13. 22 1.14 20 Cycloparaffins Single bond (no double bond) ring hydrocarbons. Unsatu- or napthenes rated, since ring can be broken and additional hydrogen 0.0 (cyclanes) added. Examples: C3H6, cyclopropane (three C-atom H H ring); C4H8, cyclobutane (four C-atom ring); C5H10, cyclopentane (five C-atom ring). H-C-C-H 12 16 20 24 28 32 36 40 44 48 Mixture (dry-bulb) temperature, ºC H H FIGURE 3-1 Psychrometric chart for air-water mixtures at 1 atmosphere. (From Reynolds.4) C. H2m 68 INTERNAL COMBUSTION ENGINE FUNDAMENTALS THERMOCHEMISTRY OF FUEL-AIR MIXTURES 69 Olefins Open-chain hydrocarbons containing a double bond; (alkenes) hence they are unsaturated. Examples are: C2H4, ethene If sufficient oxygen is available, a hydrocarbon fuel can be completely oxi- H (or ethylene); C3H6, propene (or propylene); C4H8, dized. The carbon in the fuel is then converted to carbon dioxide CO2 and the butene (or butylene); .... From butene upwards several hydrogen to water H2O. For example, consider the overall chemical equation for. structural isomers are possible depending on the location the complete combustion of one mole of propane C3H8: H H of the double bond in the basic carbon chain. Straight- C3H: + @02 = bCO2 + cH20 (3.3) and branched-chain structures exist. Diolefins contain two C ,, H 2n double bonds. A carbon balance between the reactants and products gives b = 3. A hydrogen balance gives 2c = 8, or c = 4. An oxygen balance gives 2b + c = 10 = 2a, or Acetylenes Open-chain unsaturated hydrocarbons containing one a = 5. Thus Eq. (3.3) becomes (alkynes) carbon-carbon triple bond. First member is acetylene, C3Hg + 502 = 3CO2 + 4H2O (3.4) H-C=C-H H-C=C-H. Additional members of the alkyne series C,H2n-2 comprise open-chain molecules, similar to higher alkenes Note that Eq. (3.4) only relates the elemental composition of the reactant and but with each double bond replaced by a triple bond. product species; it does not indicate the process by which combustion proceeds, which is much more complex. Air contains nitrogen, but when the products are at low temperatures the Aromatics nitrogen is not significantly affected by the reaction. Consider the complete com- bustion of a general hydrocarbon fuel of average molecular composition C.H. H Building block for aromatic hydrocarbons is the benzene with air. The overall complete combustion equation is (C6H6) ring structure shown. This ring structure is very CH stable and accommodates additional -CH2 groups in side HC. CH chains and not by ring expansion. Examples: C7H8, C.H, + (a + 2 )(02 + 3.773N2) = aCO2 + " H20 + 3.773( a + 2)N2 (3.5) toluene; C8H10, xylene (several structural Note that only the ratios of the numbers in front of the symbol for each chemical H arrangements); .... More complex aromatic hydrocar- bons incorporate ethyl, propyl, and heavier alkyl side species are defined by Eq. (3.5); i.e ., only the relative proportions on a molar basis C.H21 -6 chains in a variety of structural arrangements. are obtained. Thus the fuel composition could have been written CH, where y = b/a. Equation (3.5) defines the stoichiometric (or chemically correct or Alcohols theoretical) proportions of fuel and air; i.e ., there is just enough oxygen for con- version of all the fuel into completely oxidized products. The stoichiometric air/ Monohydric In these organic compounds, one hydroxyl (-OH) group fuel or fuel/air ratios (see Sec. 2.9) depend on fuel composition. From Eq. (3.5): alcohols is substituted for one hydrogen atom. Thus methane becomes methyl alcohol, CH3OH (also called methanol); H (A) -() = (1+ y/4X32 + 3.773 x 28.16) ethane becomes ethyl alcohol, C2HsOH (ethanol); etc. 12.011 + 1.008y H-C-OH 34.56(4 + y) H 12.011 + 1.008y (3.6) CH2n+1OH The molecular weights of oxygen, atmospheric nitrogen, atomic carbon, and atomic hydrogen are, respectively, 32, 28.16, 12.011, and 1.008. (A/F), depends only on y; Fig. 3-3 shows the variation in (A/F), as y varies from 1 (e.g ., benzene) 3.4 COMBUSTION STOICHIOMETRY to 4 (methane). This section develops relations between the composition of the reactants (fuel and air) of a combustible mixture and the composition of the products. Since these Example 3.1. A hydrocarbon fuel of composition 84.1 percent by mass C and 15.9 relations depend only on the conservation of mass of each chemical element in percent by mass H has a molecular weight of 114.15. Determine the number of the reactants, only the relative elemental composition of the fuel and the relative moles of air required for stoichiometric combustion and the number of moles of proportions of fuel and air are needed. products produced per mole of fuel. Calculate (A/F) ,, (F/A) ,, and the molecular weights of the reactants and the products. 70 INTERNAL COMBUSTION ENGINE FUNDAMENTALS THERMOCHEMISTRY OF FUEL-AIR MIXTURES 71 Per unit mass fuel: 17 1 + 15.14 = 16.14 Thus for stoichiometric combustion, 1 mole of fuel requires 59.66 moles of air and produces 64.16 moles of products. The stoichiometric (A/F), is 15.14 and (F/A), 16- is 0.0661. The molecular weights of the reactants Mp and products Mpare Ist- MR = - In, Mi = 60 1 60.66 (1 x 114.15 + 59.66 x 28.96) M p = = [ n, M ,- 1 64.16 (8 x 44.01 + 9 x 18.02 + 47.16 x 28.16) 14- or MR = 30.36 Mp = 28.71 13 Fuel-air mixtures with more than or less than the stoichiometric air require- FIGURE 3-3 ment can be burned. With excess air or fuel-lean combustion, the extra air 2 3 Stoichiometric air/fuel ratio for air-hydrocarbon fuel mixtures as a function of fuel molar H/C ratio. appears in the products in unchanged form. For example, the combustion of Fuel molar H/C ratio isooctane with 25 percent excess air, or 1.25 times the stoichiometric air require- ment, gives Assume a fuel composition C. H ,. The molecular weight relation gives CgH18 + 1.25 × 12.5(02 + 3.773N2) = 8CO2 + 9H20 + 3.1302 + 58.95N2 114.15 = 12.011a + 1.008b (3.7) The gravimetric analysis of the fuel gives With less than the stoichiometric air requirement, i.e ., with fuel-rich com- b. 15.9/1.008 a 84.1/12.011 = 2.25 bustion, there is insufficient oxygen to oxidize fully the fuel C and H to CO2 and H2O. The products are a mixture of CO2 and H2O with carbon monoxide CO Thus and hydrogen H2 (as well as N2). The product composition cannot be determined a = 8 b = 181 from an element balance alone and an additional assumption about the chemical composition of the product species must be made (see Secs. 4.2 and 4.9.2). The fuel is octane C8H18 . Equation (3.5) then becomes Because the composition of the combustion products is significantly differ- Fuel Air Products ent for fuel-lean and fuel-rich mixtures, and because the stoichiometric fuel/air C8H18 + 12.5(02 + 3.773N2) = 8CO2 + 9H2O + 47.16N2 ratio depends on fuel composition, the ratio of the actual fuel/air ratio to the stoichiometric ratio (or its inverse) is a more informative parameter for defining In moles: mixture composition. The fuel/air equivalence ratio o, - + 12.5(1 + 3.773) = 8 + 9 + 47.16 1 + 59.66 = 64.16 (F/A)actual (FIA), (3.8) Relative mass: 114.15 + 59.66 x 28.96 = 8 x 44.01 + 9 x 18.02 + 47.16 x 28.16 will be used throughout this text for this purpose. The inverse of o, the relative air|fuel ratio 2, 114.5 + 1727.8 = 1842.3 2 = 4-1 _ A/F )actual (A/F), (3.9) Note that for fuels which are mixtures of hydrocarbons, a and b need not be integers. is also sometimes used. 72. THERMOCHEMISTRY OF FUEL-AIR MIXTURES 73 INTERNAL COMBUSTION ENGINE FUNDAMENTALS For fuel-lean mixtures: ¢ <1,1> 1 of thermodynamics can be used to relate the end states of mixtures undergoing a combustion process; its application does not require that the details of the For stoichiometric mixtures: ¢=1=1 process be known. For fuel-rich mixtures: ¢>1,1 < 1 The first law of thermodynamics relates changes in internal energy (or When the fuel contains oxygen (e.g ., with alcohols), the procedure for deter- enthalpy) to heat and work transfer interactions. In applying the first law to a mining the overall combustion equation is the same except that fuel oxygen is system whose chemical composition changes, care must be exercised in relating included in the oxygen balance between reactants and products. For methyl the reference states at which zero internal energy or enthalpy for each species or alcohol (methanol), CH3OH, the stoichiometric combustion equation is groups of species are assigned. We are not free, when chemical reactions occur, to choose independently the zero internal energy or enthalpy reference states of CH3OH + 1.5(02 + 3.773N2) = CO2 + 2H2O + 5.66N2 (3.10) chemical substances transformed into each other by reaction. and (A/F), = 6.47. For ethyl alcohol (ethanol), C2HsOH, the stoichiometric com- Consider a system of mass m which changes its composition from reactants bustion equation is to products by chemical reaction as indicated in Fig. 3-4. Applying the first law to the system between its initial and final states gives C2H,OH + 3(02 + 3.773N2) = 2CO2 + 3H20 + 11.32N2 (3.11) QR-P - WR-P = Up - UR (3.13) and (A/F), = 9.00. If there are significant amounts of sulfur in the fuel, the appropriate oxida- Heat transfer OR-p and work transfer Wp-p due to normal force displacements tion product for determining the stoichiometric air and fuel proportions is sulfur may occur across the system boundary. The standard thermodynamic sign con- dioxide, SO2. vention for each energy transfer interaction -- positive for heat transfer to the For hydrogen fuel, the stoichiometric equation is system and positive for work transfer from the system-is used. H2 + 2(02 + 3.773N2) = H20 + 1.887N2 (3.12) We will consider a series of special processes: first, a constant volume process where the initial and final temperatures are the same, T'. Then Eq. (3.13) and the stoichiometric (A/F) ratio is 34.3. becomes Note that the composition of the products of combustion in Eqs. (3.7) and (3.10) to (3.12) may not occur in practice. At normal combustion temperatures QR-P = Up - UR = (AU)V. T. (3.14) significant dissociation of CO2 and of H2O occurs (see Sec. 3.7.1). Whether, at The internal energy of the system has changed by an amount (AU)y, T, which can low temperatures, recombination brings the product composition to that indi- be measured or calculated. Combustion processes are exothermic [i.e ., OR_p and cated by these overall chemical equations depends on the rate of cooling of the (AU)y, T, are negative]; therefore the system's internal energy decreases. If Eq. product gases. More general relationships for the composition of unburned and (3.14) is expressed per mole of fuel, then (AU)y, T, is known as the increase in burned gas mixtures are developed in Chap. 4. The stoichiometric (A/F) and (F/A) ratios of common fuels and representa- tive single hydrocarbon and other compounds are given in App. D along with QR-P other fuel data. 3.5 THE FIRST LAW OF System THERMODYNAMICS AND COMBUSTION+ 3.5.1 Energy and Enthalpy Balances WR - P In a combustion process, fuel and oxidizer react to produce products of different composition. The actual path by which this transformation takes place is under- stood only for simple fuels such as hydrogen and methane. For fuels with more Initial state Combustion process Final state complicated structure, the details are not well defined. Nonetheless, the first law Reactants Heat and work transfer Products . TR. PR. VR, UR interactions Tp, Pp, Vp, Up FIGURE 3-4 System changing from reactants to products for first law analysis. The approach used here follows that developed by Spalding and Cole." 74 INTERNAL COMBUSTION ENGINE FUNDAMENTALS THERMOCHEMISTRY OF FUEL-AIR MIXTURES 75 U or Ht Reactants, Products perature; also, the magnitude of (AU)v. T. [or (AH)p. T.] decreases with increasing temperature because c, (or c.) for the products is greater than for the reactants. The difference between (AH) ,, T. and (AU)y, z, is (AH)p. T. - (AU)V, T' = P(VP - VR) (3.17) Only if the volumes of the products and reactants in the constant pressure process are the same are (AH) ,, T, and (AU)y. ,, equal. If all the reactant and Ug or HR product species are ideal gases, then the ideal gas law Eq. (3.1) gives (AH)p. T. - (AU)V. T = R(np - n'T' (3.18) (AU)y, I, or - (AH)p. T. Note that any inert gases do not contribute to (n'p - n'R). With a hydrocarbon fuel, one of the products, H2O, can be in the gaseous FIGURE 3-5 or liquid phase. The internal energy (or enthalpy) of the products in the constant Schematic plot of internal energy (U) volume (or constant pressure) processes described above in Fig. 3-5 will depend or enthalpy (H) of reactants and pro- on the relative proportions of the water in the gaseous and liquid phases. The To T' ducts as a function of temperature. limiting cases of all vapor and all liquid are shown in Fig. 3-6a for a U-T plot. The internal energy differences between the curves is |(AU)v. T, H20 liq| - I(AU)y. T. H20 vap 1 = MH20 1456 1H20 (3.19) internal energy at constant volume, and -(AU)y T' is known as the heat of reac- tion at constant volume at temperature T'. where myzo is the mass of water in the products and u's, H2o is the internal energy Next, consider a constant pressure process where the initial and final tem- of vaporization of water at the temperature and pressure of the products. Similar peratures are the same, T'. For a constant pressure process U WR - P = PdV = P ( Vp - VR ) (3.15) JR Reactants Reactants Fuel so Eq. (3.13) becomes vap Products QR-P - P(Vp - VR) = Up - UR or QR-P = (U'p + PVP) -(U'R + PVR) = H'p - HR = (AH) ,, T. (3.16) The enthalpy of the system has changed by an amount (AH), T ., which can be measured or calculated. Again for combustion reactions, (AH), T, is a negative (AU)V. T quantity. If Eq. (3.16) is written per mole of fuel, (AH) ,, T, is called the increase in Fuel liq enthalpy at constant pressure and -(AH) ,, T, is called the heat of reaction at constant pressure at T'. These processes can be displayed, respectively, on the internal energy or enthalpy versus temperature plot shown schematically in Fig. 3-5. If U (or H) for the reactants is arbitrarily assigned a value Ux (or H) at some reference tem- perature To, then the value of (AU)y, ToCor (AH) ,, T.] fixes the relationship T T' T between U(T) or H(T), respectively, for the products and the reactants. Note that (a) (b) the slope of these lines (the specific heat at constant volume or pressure if the FIGURE 3-6 diagram is expressed per unit mass or per mole) increases with increasing tem- Schematic plots of internal energy of reactants and products as a function of temperature. (a) Effect of water in products as either vapor or liquid. (b) Effect of fuel in reactants as either vapor or liquid. THERMOCHEMISTRY OF FUEL-AIR MIXTURES 77 76 INTERNAL COMBUSTION ENGINE FUNDAMENTALS TABLE 3.2 curves and relationships apply for enthalpy: Standard enthalpies of formation I(AH)p. T ., H20 liq | - |(AH)p. T. H20 vap| = MH20 159 H20 (3.20) Species State Ang, MJ/kmol For some fuels, the reactants may contain the fuel as either liquid or vapor. Gas 0 The U-T (or H-T) line for the reactants with the fuel as liquid or as vapor will be 02 N2 Gas 0 different, as indicated in Fig. 3-6b. The vertical distance between the two reactant H1 Gas 0 curves is my usef (or m, hrgf) where the subscript f denotes fuel. C Gas 0 CO2 Gas -393.52 H,O Gas -241.83 H,O Liquid -285.84 3.5.2 Enthalpies of Formation CO Gas -- 110.54 CH. Gas -74.87 For fuels which are single hydrocarbon compounds, or where the precise fuel C3Hg Gas -103.85 composition is known, the internal energies or enthalpies of the reactants and the CH,OH Gas -201.17 Liquid -238.58 products can be related through the enthalpies of formation of the reactants and CH,OH C.H18 Gas -208.45 products. C,H18 Liquid -249.35 The enthalpy of formation Ah, of a chemical compound is the enthalpy increase associated with the reaction of forming one mole of the given compound + A: 298.15 K (25ºC) and 1 atm. from its elements, with each substance in its thermodynamic standard state at the given temperature. The standard state is the state at one atmosphere pressure and the tem- and the enthalpy of the reactants is given by perature under consideration. We will denote the standard state by the super- script º. HR = [ n.Ah ,., (3.21b) reactants Since thermodynamic calculations are made as a difference between an initial and a final state, it is necessary to select a datum state to which all other The enthalpy increase, (AH)p. To, is then obtained from the difference (H; - HR). thermodynamic states can be referred. While a number of datum states have been The internal energy increase can be obtained with Eq. (3.17). used in the literature, the most common datum is 298.15 K (25ºC) and 1 atmo- sphere. We will use this datum throughout this text. Elements at their reference Example 3.2. Calculate the enthalpy of the products and reactants, and the enthalpy state are arbitrarily assigned zero enthalpy at the datum temperature. The refer- increase and internal energy increase of the reaction, of a stoichiometric mixture of ence state of each element is its stable standard state; e.g ., for oxygen at 298.15 K, methane and oxygen at 298.15 K. the reference state is O2 gas. The stoichiometric reaction is Enthalpies of formation are tabulated as a function of temperature for all CH4 + 202 = CO2 + 2H2O commonly occurring species. For inorganic compounds, the JAN AF Thermoche- mical Tables are the primary reference source.º These tables include values of the Thus, for H2O gas, from Table 3.2 and Eq. (3.21a, b): molar specific heat at constant pressure, standard entropy, standard Gibbs free H; = - 393.52 + 2(-241.83) = - 877.18 MJ/kmol CH4 energy (called free energy in the tables), standard enthalpy, enthalpy of formation and Gibbs free energy of formation, and log10 equilibrium constant for the for- For H2O liquid: mation of each species from its elements. Some primary references for thermody- Ho = - 393.52 + 2(-285.84) = - 965.20 MJ/kmol CHA namic data on fuel compounds are Maxwell,9 Rossini et al ., 1º and Stull et al.11 Enthalpies of formation of species relevant to hydrocarbon fuel combustion are HR = - 74.87 MJ/kmol CH4 tabulated in Table 3.2. Selected values of thermodynamic properties of relevant Hence for H2O gas: species are tabulated in App. D. For a given combustion reaction, the enthalpy of the products at the stan- (AH); = - 877.18 + 74.87 = - 802.31 MJ/kmol CH4 dard state relative to the enthalpy datum is then given by and for H2O liquid: (3.21a) (AH); = - 965.20 + 74.87 = - 890.33 MJ/kmol CH4 products 78 INTERNAL COMBUSTION ENGINE FUNDAMENTALS THERMOCHEMISTRY OF FUEL-AIR MIXTURES 79 Use Eq. (3.18) to find (AU);. With H2O gas, the number of moles of reactants and Heating valuest of fuels are measured in calorimeters. For gaseous fuels, it products are equal, so is most convenient and accurate to use a continuous-flow atmosphere pressure (AU); = (AH); = - 802.3 MJ/kmol CH4 calorimeter. The entering fuel is saturated with water vapor and mixed with suffi- cient saturated air for complete combustion at the reference temperature. The For H2O liquid: mixture is burned in a burner and the combustion products cooled with water- (AU); = - 890.33 - 8.3143 x 10-3(1 - 3)298.15 MJ/kmol CH4 cooled metal tube coils to close to the inlet temperature. The heat transferred to (AU); = - 885.4 MJ/kmol CH, the cooling water is calculated from the measured water flow rate and water temperature rise. The heating value determined by this process is the higher Note that the presence of nitrogen in the mixture or oxygen in excess of the stoi- heating value at constant pressure. chiometric amount would not change any of these calculations. For liquid and solid fuels, it is more satisfactory to burn the fuel with oxygen under pressure at constant volume in a bomb calorimeter. A sample of 3.5.3 Heating Values the fuel is placed in the bomb calorimeter, which is a stainless-steel container For fuels where the precise fuel composition is not known, the enthalpy of the immersed in cooling water at the standard temperature. Sufficient water is placed reactants cannot be determined from the enthalpies of formation of the reactant in the bomb to ensure that the water produced in the combustion process will species. The heating value of the fuel is then measured directly. condense. Oxygen at 30 atmospheres is admitted to the bomb. A length of firing The heating value Ouy or calorific value of a fuel is the magnitude of the cotton is suspended into the sample from an electrically heated wire filament to heat of reaction at constant pressure or at constant volume at a standard tem- act as a source of ignition. When combustion is complete the temperature rise of perature [usually 25ºC (77ºF)] for the complete combustion of unit mass of fuel. the bomb and cooling water is measured. The heating value determined by this Thus process is the higher heating value at constant volume. The heating values of common fuels are tabulated with other fuel data in CHV, = - (AH)p, TO (3.22a) App. D. The following example illustrates how the enthalpy of a reactant mixture and QHVV = - (AU)y. TO (3.22b) relative to the enthalpy datum we have defined can be determined from the mea- sured heating value of the fuel. Complete combustion means that all carbon is converted to CO2, all hydrogen is converted to H2O, and any sulfur present is converted to SO2. The Example 3.3. Liquid kerosene fuel of the lower heating value (determined in a bomb heating value is usually expressed in joules per kilogram or joules per kilomole of calorimeter) of 43.2 MJ/kg and average molar H/C ratio of 2 is mixed with the fuel (British thermal units per pound-mass or British thermal units per pound- stoichiometric air requirement at 298.15 K. Calculate the enthalpy of the reactant mole). It is therefore unnecessary to specify how much oxidant was mixed with mixture relative to the datum of zero enthalpy for C, O2, N2, and H2 at 298.15 K. the fuel, though this must exceed the stoichiometric requirement. It is immaterial The combustion equation per mole of C can be written whether the oxidant is air or oxygen. CH2 + ¿O2 + 3.773N2) = CO2 + H2O + 5.660N2 For fuels containing hydrogen, we have shown that whether the H2O in the products is in the liquid or gaseous phase affects the value of the heat of reaction. or 14 kg fuel + 7.160 kmol 7.66 kmol The term higher heating value OHHy (or gross heating value) is used when the 207.4 kg air 221.4 kg . products H2O formed is all condensed to the liquid phase; the term lower heating value where M = 28.962 for atmospheric air. QLHV (or net heating value) is used when the H2O formed is all in the vapor The heating value given is at constant volume, - (AU);. (AH), is obtained phase. The two heating values at constant pressure are related by . from Eq. (3.18), noting that the fuel is in the liquid phase: CHHV, = @LHV, + ms (3.23) (AH); = - 43.2 + 8.3143 x 10-3 (7.66 - 7.160) x = , 298.15 14 where (MH20/m) is the ratio of mass of H2O produced to mass of fuel burned. A = - 43.2 + 0.09 = - 43.1 MJ/kg fuel similar expression with usg H120 replacing hfg H120 applies for the higher and lower heating value at constant volume. The heating value at constant pressure is the more commonly used; often the qualification "at constant pressure" is omitted. The difference between the Standard methods for measuring heating values are defined by the American Society for Testing heating values at constant pressure and constant volume is small. Materials. 80 INTERNAL COMBUSTION ENGINE FUNDAMENTALS THERMOCHEMISTRY OF FUEL-AIR MIXTURES 81 The enthalpy of the products per kilogram of mixture is found from the enthalpies of formation (with H2O vapor): U or H| Reactants Products hp =. 1(-393.52) + 1(-241.83) 221.4 -(AU)y, To or -(AR )p, To = 2.87 MJ/kg The enthalpy of the reactants per kilogram of mixture is then hp = hp - (Ah); = 2.87 + 43.1 x 14 221.4 = 5.59 MJ/kg FIGURE 3-7 3.5.4 Adiabatic Combustion Processes Adiabatic constant-volume combustion process on U-T diagram or adiabatic constant-pressure combustion We now use the relationships developed above to examine two other special To TR TP T ‘ process on H-T diagram. processes important in engine analysis: constant-volume and constant-pressure adiabatic combustion. For an adiabatic constant-volume process, Eq. (3.13) batic combustion process, the constant-volume or constant-pressure constraint becomes must also be used explicitly. Up - UR = 0 (3.24) The final temperature of the products in an adiabatic combustion process is called the adiabatic flame temperature. Examples of typical adiabatic flame tem- when Up and UR are evaluated relative to the same datum (e.g ., the enthalpies of C, O2, N2, and H2 are zero at 298.15 K, the datum used throughout this text). peratures are shown later in Fig. 3-11. Frequently, however, the tables or graphs of internal energy or enthalpy for species and reactant or product mixtures which are available give internal ener- 3.5.5 Combustion Efficiency of an Internal gies or enthalpies relative to the species or mixture value at some reference tem- Combustion Engine perature To , i.e ., U(T) - U(To) or H(T) - H(To) are tabulated. Since In practice, the exhaust gas of an internal combustion engine contains incomplete Up(To) - UR(To) = (AU)y. TO combustion products (e.g ., CO, H2, unburned hydrocarbons, soot) as well as complete combustion products (CO2 and H2O) (see Sec. 4.9). Under lean oper- it follows from Eq. (3.24) that ating conditions the amounts of incomplete combustion products are small. [UP(T) - Up(To)] - [UR(T) - UR(To)] = - (AU)y, To (3.25) Under fuel-rich operating conditions these amounts become more substantial since there is insufficient oxygen to complete combustion. Because a fraction of relates the product and reactant states. Figure 3-7 illustrates the adiabatic the fuel's chemical energy is not fully released inside the engine during the com- constant-volume combustion process on a U-T diagram. Given the initial state of bustion process, it is useful to define a combustion efficiency. The engine can be the reactants (TR, V) we can determine the final state of the products (Tp, V). analyzed as an open system which exchanges heat and work with its surrounding For an adiabatic constant-pressure combustion process, Eq. (3.13) gives environment (the atmosphere). Reactants (fuel and air) flow into the system; pro- Hp - HR = 0 ducts (exhaust gases) flow out. Consider a mass m which passes through the control volume surrounding the engine shown in Fig. 3-8; the net chemical energy which combines with Eq. (3.16) to give release due to combustion within the engine is given by [Hp(T) - Hp(To)] - [HR(T) - HR(To)] = - (AH)). To (3.26) Figure 3-7 illustrates this process also. Given the initial reactant state (TR, p) we [HR(TA) - Hp(TA)] = m [ .Ah;. - [ n.Ah ,. :) i, reactants i, products can determine the final product state (Tp, p). Note that while in Figs. 3-5, 3-6, and 3-7 we have shown U and H for the Enthalpy is the appropriate property since PR = Pp = Patm. n; is the number of moles of species i in the reactants or products per unit mass of working fluid and reactants and products to be functions of T only, in practice for the products at high temperature U and H will be functions of p and T. The analysis presented Ah? ; is the standard enthalpy of formation of species i at ambient temperature T. here is general; however, to determine the final state of the products in an adia- The amount of fuel energy supplied to the control volume around the engine which can be released by combustion is my QHy. Hence, the combustion 82 INTERNAL COMBUSTION ENGINE FUNDAMENTALS THERMOCHEMISTRY OF FUEL-AIR MIXTURES 83 Control volume bustion process remains stable. For diesel engines, which always operate lean, the combustion efficiency is normally higher-about 98 percent. Details of exhaust Fuel gas composition, on which these combustion efficiencies are based, can be found Exhaust gas Engine in Sec. 4.9. Air FIGURE 3-8 Control volume surrounding engine. 3.6 THE SECOND LAW OF THERMODYNAMICS APPLIED TO COMBUSTION efficiency-the fraction of the fuel energy supplied which is released in the com- bustion process-is given by12 3.6.1 Entropy HR(TA) - HP(T) In App. B, it is shown how the entropy of a mixture of ideal gases of known (3.27) my CHV composition can be calculated. The discussion earlier relating enthalpies (or inter- nal energies) of reactant and product mixtures applies to entropy also. The stan- Note that m and my could be replaced by the average mass flow rates m and m. dard state entropies of chemical species are tabulated in the JANAF tables Figure 3-9 shows how combustion efficiency varies with the fuel/air equiva- relative to zero entropy at 0 K. If the entropies of the elements at a datum lence ratio for internal combustion engines. For spark-ignition engines, for lean temperature are arbitrarily set equal to zero, then the values of the entropy of a equivalence ratios, the combustion efficiency is usually in the range 95 to 98 reactant mixture of given composition and of the resulting product mixture of percent. For mixtures richer than stoichiometric, lack of oxygen prevents com- given composition are both determined. plete combustion of the fuel carbon and hydrogen, and the combustion efficiency steadily decreases as the mixture becomes richer. Combustion efficiency is little affected by other engine operating and design variables, provided the engine com- 3.6.2 Maximum Work from an Internal Combustion Engine and Efficiency 100 ** An internal combustion engine can be analyzed as an open system which X × × exchanges heat and work with its surrounding environment (the atmosphere). Reactants (fuel and air) flow into the system; products (exhaust gases) flow out. XX O 90 By applying the second law of thermodynamics to a control volume surrounding the engine, as illustrated in Fig. 3-8, we can derive an expression for the maximum useful work that the engine can deliver. Consider a mass m of fluid as it passes through the control volume sur- 80|- rounding the engine. The first law gives Combustion efficiency AQ - AWy = AH 70 where AWy is the useful work transfer (i.e ., non-p dV work) to the environment and AH = Hp - HR. Since the heat transfer AQ occurs only with the atmosphere x Diesels at temperature T. , from the second law 60- Spark-ignition TA S AS 50 0 0.2 These equations combine to give 0.4 0.6 0.8 1.2 1.4 1.6 Exhaust equivalence ratio AWy < -(AH - TAAS) = - AB FIGURE 3-9 where B is the steady-flow availability function, H - TAS.13 Usually PR = PA and Variation of engine combustion efficiency with fuel/air equivalence ratio. TR = TA. The maximum work will be obtained when Pp = DA and Tp = T1. 84 INTERNAL COMBUSTION ENGINE FUNDAMENTALS THERMOCHEMISTRY OF FUEL-AIR MIXTURES 85 TABLE 3.3 which was defined as the fuel conversion efficiency in Sec. 2.8. Note that some- Enthalpies and free energies of combustion reactions times the higher heating value is used in Eq. (3.30) and sometimes the lower heating value. Whichever value is used should be explicitly stated. The normal Reaction+ Akiss, MJ/kmol Agios, MJ/kmol practice in internal combustion engine analysis is to use the lower heating value C+ 02 -+ CO2 -393.52 -394.40 at constant pressure, since the engine overall is a steady flow device and the water H2 + 102 -> H2O -240.91 -232.78 in the exhaust is always in vapor form. We will use QLHy, in Eq. (3.30) through- CH4 + 202 -> CO2 + 2H¿O -802.30 -800.76 out this text. The fuel conversion efficiency is the most commonly used definition CH,O(1) + 202 - CO2 + 2H2O -638.59 -685.35 of engine efficiency because it uses an easily measured quantity, the heating value, C3H8(g) + 502 -+ 3CO2 + 4H2O -2044.0 -2074.1 to define the usable fuel energy supplied to the engine. For hydrocarbon fuels, C6H6(!) + 4502 - 6CO2 + 3H2O -3135.2 -3175.1 C&H18(/) + 202 -> 8CO2 + 9H2O -5074.6 -5219.9 since Ahº ~ Ago, the fuel conversion efficiency and the availability conversion efficiency are closely comparable in value. + H2O (gas) in products. In practice, not all the fuel energy supplied to the engine is released by the combustion process since combustion is incomplete: the combustion efficiency Under these conditions, [Eq. (3.27)] is less than unity. It is sometimes useful to separate out the effects of incomplete combustion by defining an efficiency which relates the actual work AWy 3 -[(H - TS)p ... .. - (H - TS)RT .... ] - - (AG)TAMPA per cycle to the amount of fuel chemical energy released in the combustion or AWU max = - (AG)TA, PA (3.28) process. We will call this the thermal conversion efficiency ni: G is the Gibbs free energy, H - TS, and (AG)] > is the Gibbs free energy We We W. increase in the reaction of the fuel-air mixture to products at atmospheric tem- no = HR(TA) - HP(TA)" (AH) TA ne my PHv (3.31) perature and pressure. - (AG)T, „, will be a maximum when combustion is com- plete. Obviously the fuel conversion, thermal conversion, and combustion efficiencies A fundamental measure of the effectiveness of any practical internal com- are related by bustion engine is the ratio of the actual work delivered compared with this maximum work. This ratio will be called the availability conversion efficiency na: ng = Mc n. (3.32) AW AW It is important to understand that there is a fundamental difference between ng = = (3.29) A Wy max (AG)TA, PA availability conversion efficiency as defined by Eq. (3.29) [and the fuel conversion efficiency for internal combustion engines, Eq. (3.30), which closely approximates The property availability is the maximum useful work transfer that can be it] and the efficiency of a thermodynamic heat engine. The second law limit to obtained from a system atmosphere (or control-volume atmosphere) combination the availability conversion efficiency is unity. For a thermodynamic heat engine at a given state. This efficiency therefore defines the fraction of the availability of (which experiences heat-transfer interactions with at least two heat reservoirs) the the unburned fuel and air which, passing through the engine and interacting only efficiency is limited to a value substantially less than unity by the temperatures of with the atmosphere, is actually converted to useful work. Availability analysis of the heat reservoirs available.13 engine operation is proving valuable in identifying where the significant irrevers- ibilities or losses in availability occur. This topic is discussed more fully in Sec. 5.7. 3.7 CHEMICALLY REACTING GAS (AG)TA. PA, Or (49)TA, PA, is not an easy quantity to evaluate for practical MIXTURES fuels; it is the heating value, - (Ah)T ,, which is usually measured. Values of (Ag)298 and (Ah)29g for selected fuel combustion reactions are given in Table 3.3. The working fluids in engines are mixtures of gases. Depending on the problem For the pure hydrocarbons they are closely comparable because at 298 K, A3º < under consideration and the portion of the engine cycle in which it occurs chemi- Ahº/T. For hydrogen and methanol the differences are larger, however. Because cal reactions may: (1) be so slow that they have a negligible effect on mixture for practical fuels -(Ah)298 is measured directly as the heating value of the fuel, it composition (the mixture composition is essentially "frozen"); (2) be so rapid is standard practice to use the following definition of efficiency: that the mixture state changes and the composition remains in chemical equi- We librium; (3) be one of the rate-controlling processes that determine how the com- ng = = (3.30) position of the mixture changes with time. my 2Hv 86 INTERNAL COMBUSTION ENGINE FUNDAMENTALS THERMOCHEMISTRY OF FUEL-AIR MIXTURES 87. 3.7.1 Chemical Equilibrium Consider a reactive mixture of ideal gases. The reactant species Ma, M ., It is a good approximation for performance estimates in engines to regard the etc ., and product species M1, Mm, etc ., are related by the general reaction whose burned gases produced by the combustion of fuel and air as in chemical equi- stoichiometry is given by librium.+ By this we mean that the chemical reactions, by which individual V . M a + V , M x + . . . = v , M , + V , Mrm + ... (3.34a) species in the burned gases react together, produce and remove each species at This can be written as equal rates. No net change in species composition results. For example, if the temperature of a mass of carbon dioxide gas in a vessel E v. M , = 0 (3.34b) is increased sufficiently, some of the CO2 molecules dissociate into CO and O2 molecules. If the mixture of CO2, CO, and O2 is in equilibrium, then CO2 mol- where the vi are the stoichiometric coefficients and by convention are positive for ecules are dissociating into CO and O2 at the same rate as CO and O2 molecules the product species and negative for the reactant species. are recombining in the proportions required to satisfy the equation Let an amount on, of M. react with on, of M ,, etc ., and produce on, of M1, on, of Mm, etc. These amounts are in proportion: CO + 102 = CO2 on; = vion In combustion products of hydrocarbon fuels, the major species present at (3.35) low temperatures are N2, H2O, CO2, and O2 or CO and H2. At higher tem- The change in Gibbs free energy of a mixture of ideal gases, at constant peratures (greater than about 2200 K), these major species dissociate and react to pressure and temperature, as the composition changes is given by form additional species in significant amounts. For example, the adiabatic com- (AG)p. T = LAion, (3.36) bustion of a stoichiometric mixture of a typical hydrocarbon fuel with air pro- duces products with species mole fractions of: N2 ~ 0.7; H2O, CO2 ~ 0.1; CO, where on is the change in number of moles of species i and u is the chemical OH, O2, NO, H2 ~ 0.01; H, O ~ 0.001; and other species in lesser amounts. potential. The chemical potential, an intensive property, is defined as The second law of thermodynamics defines the criterion for chemical equi- librium as follows. Consider a system of chemically reacting substances under- (3.37) going a constant-pressure, constant-temperature process. In the absence of shear work (and electrical work, gravity, motion, capillarity), the first law gives It is equal in magnitude to the specific Gibbs free energy at a given temperature 6Q = dH and pressure. For an ideal gas, it follows from Eqs. (B.13), (B.15) and (3.37) that The second law gives À = p;(T) + RT In 24 Po (3.38) 6Q ST ds where uf equals g;, the standard specific Gibbs free energy of formation. The Combining these gives standard state pressure po is usually one atmosphere. dH - T ds < 0 Substitution in Eq. (3.36) gives, at equilibrium, Since we are considering constant-temperature processes, this equation holds for I ( Hi + RT In P. )on, = 0 finite changes: Pol AH - T AS = AG $ 0 or [ ( Fi + RT In P: )v, on = 0 Thus, reactions can only occur (at constant pressure and temperature) if G Po (=H - TS) for the products is less than G for the reactants. Hence at equilibrium We can divide by on and rearrange, to obtain (AG)p. T = 0 (3.33) [ in AGO Po RT RT = In Kp (3.39) K, is the equilibrium constant at constant pressure: This assumption is not valid late in the expansion stroke and during the exhaust process (see Sec. 4.9). Nor does it take account of pollutant formation processes (see Chap. 11). K, = In (PS)" (3.40) 88 INTERNAL COMBUSTION ENGINE FUNDAMENTALS THERMOCHEMISTRY OF FUEL-AIR MIXTURES 89 It is obtained from the Gibbs free energy of the reaction which can be calculated If the degree of dissociation in the products is a (i.e ., a fraction a of the CO2 from the Gibbs free energy of formation of each species in the reaction, as indi- formed by complete combustion is dissociated), the product composition is cated in Eq. (3.39) above. It is a function of temperature only. In the JANAF tables,8 to simplify the calculation of equilibrium constants, CO2, (1 - a) ; CO, a; values of log10 (Kp)1, the equilibrium constants of formation of one mole of each species from their elements in their standard states, are tabulated against tem- For this mixture, the number of moles of reactants n is ; the number of moles of perature. The equilibrium constant for a specific reaction is then obtained via the products np is (1 + a/2). relation The ideal gas law gives log10 (K p)reaction = E vi logi0 (Kp)i (3.41) PR V = nR RTR PpV = npRTP Thus With the JANAF table values of (K,);, the pressures in Eqs. (3.40) and (3.41) must PP 1 2500 be in atmospheres. × 1.5 = 5.555 atm/mol np 300 The effect of pressure on the equilibrium composition can be deduced from Eq. (3.40). Substitution of the mole fractions x, and mixture pressure p gives The equilibrium relation [Eq. (3.40)] gives 1 - a 1/2 a(x/2)1/2 = 27.5 PP which can be solved to give a = 0.074. If [, v = 0, changes in pressure have no effect on the composition. If [& v; > 0 The composition of the products in mole fractions is, therefore, (dissociation reactions), then the mole fractions of the dissociation products decrease as pressure increases. If _; v, < 0 (recombination reactions), the con- XCO2 1 -₫ = 0.893 np verse is true. An equilibrium constant, Kc, based on concentrations (usually expressed in Xco - = 0.071 gram moles per cubic centimeter) is also used: np K = [I [M]" (3.42) 2. For complex systems such as this, the following approach to equi- variables and an iteration procedure is generally required to obtain their solu- librium composition calculations is now more widely used. tion. Once the composition is determined, additional relations, such as those in Standardized computer methods for the calculation of complex chemical App. B which define the thermodynamic properties of gas mixtures, must then be used. equilibrium compositions have been developed. A generally available and well- documented example is the NASA program of this type.14 The approach taken is For each species, standard state enthalpies hº are obtained by combining to minimize explicitly the Gibbs free energy of the reacting mixture (at constant standard enthalpies of formation at the datum temperature (298.15 K) Ah9298 temperature and pressure) subject to the constraints of element mass conserva- with sensible enthalpies (hº - h298), i.e ., tion. The basic equations for the NASA program are the following. ho = Ah,298 + (ho - h298) (3.49) If the stoichiometric coefficients a;; are the number of kilomoles of element i per kilomole of species j, by is the number of kilomoles of element i per kilogram For the elements in their reference state, Ah9298 is zero [the elements important of mixture, and n; is the number of kilomoles of species j per kilogram of mixture, in combustion are C (solid, graphite), H2(g), O2(g), N2(g)]. element mass balance constraints are For each species, the thermodynamic quantities specific heat, enthalpy, and entropy as functions of temperature are given in the form: [ agn, - bt = 0 for i = 1, 2, ..., 1 (3.44) : = a1 + @2T + @3 72 + @4 T3. + as T4 (3.50a) The Gibbs free energy per kilogram of mixture is ho g = Lijn; RT 2 = a1 + = 2 7 + = 3 72 +- 4 + 94 T3 + == T+ + # (3.50b) (3.45) 92 significant. omitted from consideration. significant amounts), x; 3.7.2 Chemical Reaction Rates INTERNAL COMBUSTION ENGINE FUNDAMENTALS 300 to 1000 K and 1000 to 5000 K) (see Sec. 4.7). peratures occur slightly rich of stoichiometric. the program calculates and prints out the following: = a, ln T + a2 T +-3T2 +" (@ In V/2 In p), (@ In V/0 In T)p, Cp, Y ,, and a (sound speed) property data from the JANAF tables. Usually two sets of coefficients are The coefficients are obtained by least-squares matching with thermodynamic included for two adjacent temperature intervals (in the NASA program these are must be specified as an input to the calculation. In the NASA program, all allow- able species are included in the calculation, though species may be specifically 1. Thermodynamic mixture properties (obtained from the equilibrium composi- 2. Equilibrium composition. Mole fractions of each species (which are present in combustion of isooctane-air mixtures at selected temperatures and 30 atm pres- mixtures. As temperature increases, the burned-gas mixture composition becomes CO2, H2O, and O2 for lean mixtures and N2, CO2, H2O, CO, and H2 for rich sure varies with the equivalence ratio. At low temperatures, the products are N2, much more complex with dissociation products such as OH, O, and H becoming tions as a function of the equivalence ratio, obtained with the NASA program using the methodology of Sec. 3.5.4. The isooctane-air unburned mixture state pressure (where PR and HR are specified) and at constant volume (where VR and was 700 K and 10 atm. Flame temperatures for adiabatic combustion at constant tion and the appropriate gas mixture rule; see App. B). p, T, p, h, s, M, UR are specified) are shown. Flame temperatures at constant volume are higher, because the final pressure is higher and dissociation is less. Maximum flame tem- In some equilibrium programs, the species to be included in the mixture Whether a system is in chemical equilibrium depends on whether the time con- For each reactant composition and pair of thermodynamic state variables, stants of the controlling chemical reactions are short compared with time scales over which the system conditions (temperature and pressure) change. Chemical equilibrium phenomena are the flame reaction zone where the fuel is oxidized, processes in engines are often not in equilibrium. Important examples of non- controlled by the rates at which the actual chemical reactions which convert and the air-pollutant formation mechanisms. Such nonequilibrium processes are Figure 3-10 shows how the equilibrium composition of the products of Figure 3-11 shows adiabatic flame temperatures for typical engine condi- @4 +3 + - 5 T4 + a7 4 . (3.50c) N N 2 N2 2250 K 2750 K 1750 K H2O H2 O H2O O O 10- 10-1 10-1 CO CO2 Co2 CO CO NO co NO 10-2 Mole fraction Mole fraction 10-2 H2 OH TTT Mole fraction OH H2 NO o OH 10-3|- 10-3 10-3- TT CO OH H H O OH 10-45 10- 0.8 1.0 1.2 1.4 0.2 0.4 0.6 0.2 0.8 1.0 1.2 1.4 0.2 0.4 1.0 1.2 0.4 (c) ( a) ( b ) FIGURE 3-10 Mole fractions of equilibrium combustion products of isooctane-air mixtures as a function of fuel/air equivalence ratio at 30 atmospheres and (a) 1750 K; (b) 2250 K; and (c) 2750 K. 93 94 INTERNAL COMBUSTION ENGINE FUNDAMENTALS THERMOCHEMISTRY OF FUEL-AIR MIXTURES 95 3200 M is any molecule (such as N2) which takes part in the collision and carries away the excess energy. 2800- The law of mass action states that the rate at which product species are produced and the rate at which reactant species are removed is proportional to 2400 -- Tp , " the product of the concentrations of reactant species, with the concentration of each species raised to the power of its stoichiometric coefficient v. Thus, for 2000 50 reaction (3.51) above, the reaction rate R+ in the forward (+) direction, reactants Tp, P --- to products, is given by 1600 ------ 40 Pressure, atm Temperature, K R+ = -= d[M]+ d[M] = k+[M ][M] dt dt (3.53) PP. 1200 30 ----- If the reaction can also proceed in the reverse (-) direction, then the backward rate R~ is given by 800 20 d[M] - 4002 10 d[M ] = k [M ][M] dt dt (3.54) k+ and k- are the rate constants in the forward and reverse directions for this OS 0.2 0.4 0.6 0.8 1.0 1.2 reaction. The net rate of production of products or removal of reactants is, there- Fuel/air equivalence ratio fore, FIGURE 3-11 Equilibrium product temperatures for constant-volume (Tp „) and constant-pressure (Tp. „) adiabatic d[M ]+ + R+ - R- = d[M.] + -= -= d[M.] combustion of isooctane-air mixture initially at 700 K and 10 atm, as a function of fuel/air equiva- dt dt dt dt lence ratio. Pressure (pp. ») is equilibrium pressure for adiabatic constant-volume combustion. = k+[M][M ] - k-[M][M] (3.55) reactants to products occur. The rates at which chemical reactions proceed These results can be stated more generally as follows. Any reaction can be depend on the concentration of the reactants, temperature, and whether any cata- written as lyst is present. This field is called chemical kinetics and some of its basic relations will now be reviewed.2 E VR, MR. = E VP , Mp, (3.56) Most of the chemical reactions of interest in combustion are binary reac- tions, where two reactant molecules, Ma and M ,, with the capability of reacting where v; is the stoichiometric coefficient of species M;, subscripts R and P denote together collide and form two product molecules, Me and Ma; i.e ., reactants and products, respectively, and there are n reactant species and m Ma + M, = Mc + M. (3.51) product species. The forward reaction rate R+ and the reverse reaction rate R- are given by An important example of such a reaction is the rate-controlling step in the process by which the pollutant nitric oxide, NO, forms: R + = k + [ I [MR.]VR 0 + N2 = NO + N (3.57) This is a second-order reaction since the stoichiometric coefficients of the reac- R = k [[ [MP, ] tants ve and v, are each unity and sum to 2. (The only first-order reactions are decomposition processes.) Third-order reactions are important in combustion, also. Examples are the recombination reactions by which radical species such as The net rate of removal of reactant species MR, is H, O, and OH combine during the final stage of the fuel oxidation process : e.g ., d[MR.] H + H + M = H2 + M* (3.52) = VR(R+ - R-) dt (3.58a) THERMOCHEMISTRY OF FUEL-AIR MIXTURES 97 96 INTERNAL COMBUSTION ENGINE FUNDAMENTALS 3.3. The molar composition of dry exhaust gas of a propane-fueled SI engine is given below (water and the net rate of production of product species Mp, is was removed before the measurement). Calculate the equivalence ratio. d[MPd = Vp(R+ - R-) CO2 = 10.8%, O2 = 4.5%, CO = 0%, H2 = 0% (3.58b) = 3.4. Evaluate and compare the lower heating values per unit mass of stoichiometric mixture and per unit volume of stoichiometric mixture (at standard atmospheric The rate constants, k, usually follow the Arrhenius form: conditions) for methane, isooctane, methyl alcohol, and hydrogen. Assume the fuel is fully vaporized. k = A exp ( RI ) (3.59) 3.5. The measured engine fuel flow rate is 0.4 g/s, air flow rate is 5.6 g/s, and exhaust gas composition (measured dry) is CO2 = 13.0%, CO = 2.8% with O2 essentially zero. Unburned hydrocarbon emissions can be neglected. Compare the equivalence ratio where A is called the frequency or preexponential factor and may be a (moderate) calculated from the fuel and air flow with the equivalence ratio calculated from function of temperature; E, is the activation energy. The Boltzmann factor exhaust gas composition. The fuel is gasoline with a H/C ratio of 1.87. Assume a H2 exp (-E,/RT) defines the fraction of all collisions that have an energy greater concentration equal to one-third the CO concentration. than E -i.e ., sufficient energy to make the reaction take place. The functional" 3.6. The brake fuel conversion efficiency of an engine is 0.3. The mechanical efficiency is dependence of k on T and the constants in the Arrhenius form, Eq. (3.59), if that 0.8. The combustion efficiency is 0.94. The heat losses to the coolant and oil are is appropriate, are determined experimentally. 60 kW. The fuel chemical energy entering the engine per unit time, my QHy, is At equilibrium, the forward and reverse reaction rates are equal. Then, from 190 KW. What percentage of this energy becomes (a) brake work; (b) friction work; Eq. (3.55), with R+ - R- = 0: (c) heat losses; (d) exhaust chemical energy; (e) exhaust sensible energy. 3.7. An upper estimate can be made of the amount of NO formed in an engine from k* [M ][M] _ = K. considering the equilibrium of the reaction N2 + O2 = 2NO. Calculate the NO con- k - [M.][M.] centration at equilibrium at 2500 K and 30 atm. log10 K, = - 1.2 for this reaction at 2500 K. Assume N/O ratio in the combustion products is 15. N2, O2, and NO where K, is the equilibrium constant based on concentrations defined by Eq. are the only species present. (3.42). It can be related to K„, the equilibrium constant based on partial pres- 3.8. Carbon monoxide reacts with air at 1 atm and 1000 K in an exhaust gas reactor. The mole fractions of the exhaust gas-air mixture flowing into the reactor are CO, sures, by Eq. (3.43). The chemical reaction mechanisms of importance in combustion are much 3%; O2, 7%; N2, 74%; CO2, 6%; H2O, 10%. more complex than the above illustrations of rate-controlled processes. Such (a) Calculate the concentration of CO and O2 in gram moles per cm3 in the entering mixture. mechanisms usually involve both parallel and sequential interdependent reac- (b) The rate of reaction is given by tions. The methodology reviewed above still holds; however, one must sum alge- braically the forward and reverse rates of all the reactions which produce (or d[CO]/dt = - 4.3 x 1011 x [CO][O2]0.25 exp [-E/(RT)] remove) a species of interest. In such complex mechanisms it is often useful to [ ] denotes concentration in gram moles per cm3, E/R == 20,000 K. Calculate the assume that (some of) the reactive intermediate species or radicals are in steady initial reaction rate of CO, d[CO]/dt: time is in seconds. state. That is, these radicals react so quickly once they are formed that their (c) The equilibrium constant K, for the reaction CO + 1O2 = CO2 at 1000 K is concentrations do not rise but are maintained in steady state with the species 1010. Find the equilibrium CO concentration. with which they react. The net rate at which their concentration changes with (d) Determine the time to reach this equilibrium concentration of CO using the time is set equal to zero. initial reaction rate. (The actual time will be longer but this calculation indicates approximately the time required.) 3.9. The exhaust gases of a hydrogen-fueled engine contain 22.3 percent H2O, 7.44 PROBLEMS percent O2, and 70.2 percent N2 . At what equivalence ratio is it operating? 3.10. Gas is sampled at 1 atmosphere pressure from the exhaust manifold of an internal 3.1. Isooctane is supplied to a four-cylinder spark-ignition engine at 2 g/s. Calculate the combustion engine and analyzed. The mole fractions of species in the exhaust are: air flow rate for stoichiometric combustion. If the engine is operating at 1500 rev/ min, estimate the mass of fuel and air entering each cylinder per cycle. The engine H2O, 0.0468; CO2, 0.0585; O2, 0.123; N2, 0.772 displaced volume is 2.4 liters. What is the volumetric efficiency? Other species such as CO and unburned hydrocarbons can be neglected. 3.2. Calculate the exhaust gas composition of a butane-fueled spark-ignition engine oper- (a) The fuel is a synthetic fuel derived from coal containing only carbon and hydro- ating with equivalence ratio of 0.9. Assume the fuel is fully burned within the cylin- gen. What is the ratio of hydrogen atoms to carbon atoms in the fuel? der. Butane is C4H10. 98 INTERNAL COMBUSTION ENGINE FUNDAMENTALS THERMOCHEMISTRY OF FUEL-AIR MIXTURES 99 (b) Calculate the fuel/air equivalence ratio at which this engine is operating. (c) Is the internal combustion engine a conventional spark-ignition or a diesel 6. Goodger, E. M.: Hydrocarbon Fuels, Macmillan, London, 1975. engine ? Explain. 7. Spalding, D. B ., and Cole, E. H.: Engineering Thermodynamics, 3d ed ., Edward Arnold, 1973. (d) The engine has a displaced volume of 2 liters. Estimate approximately the per- 8. JAN AF Thermochemical Tables, National Bureau of Standards Publication NSRDS-NBS37, 1971. centage by which the fuel flow rate would be increased if this engine were oper- 9. Maxwell, J. B.: Data Book on Hydrocarbons, Van Nostrand, New York, 1950. ated at its maximum load at this same speed (2000 rev/min). Explain briefly what 10. Rossini, F. D ., Pitzer, K. S ., Arnelt, R. L ., Braun, R. M ., and Primentel, G. C.: Selected Values of limits the equivalence ratio at maximum load. Physical and Thermodynamic Properties of Hydrocarbons and Related Compounds, Carnegie Press, 3.11. The following are approximate values of the relative molecular mass (molecular Pittsburgh, Pa ., 1953. weights): oxygen O2, 32; nitrogen N2, 28; hydrogen H2 , 2; carbon C, 12. Determine 11. Stull, D. R ., Westrum, E. F ., and Sinke, G. C.: The Chemical Thermodynamics of Organic Com- the stoichiometric fuel/air and air/fuel ratios on a mass basis, and the lower heating pounds, John Wiley, New York, 1969. value per unit mass of stoichiometric mixture for the following fuels: 12. Matthews, R. D.: " Relationship of Brake Power to Various Energy Efficiencies and Other Engine Parameters: The Efficiency Rule," Int. J. of Vehicle Design, vol. 4, no. 5, pp. 491-500, 1983. Methane (CH4), isooctane (CH18), benzene (CH6), hydrogen (H2), methyl 13. Keenan, J. H.: Thermodynamics, John Wiley, New York, 1941 (MIT Press, Cambridge, Mass ., alcohol (CH3OH) 1970). 14. Svehla, R. A ., and McBride, B. J.: "Fortran IV Computer Program for Calculation of Thermody- Heating values for these fuels are given in App. D. namic and Transport Properties of Complex Chemical Systems," NASA Technical Note TN 3.12. Liquid petroleum gas (LPG) is used to fuel spark-ignition engines. A typical sample D-7056, NASA Lewis Research Center, 1973. of the fuel consists of 70 percent by volume propane C3Hg 5 percent by volume butane C4H10 25 percent by volume propene C3H6 The higher heating values of the fuels are: propane, 50.38 MJ/kg; butane, 49.56 MJ/kg; propylene (propene), 48.95 MJ/kg. (a) Work out the overall combustion reaction for stoichiometric combustion of 1 mole of LPG with air, and the stoichiometric F/A and A/F. (b) What are the higher and lower heating values for combustion of this fuel with excess air, per unit mass of LPG? 3.13. A spark-ignition engine is operated on isooctane fuel (C8H18). The exhaust gases are cooled, dried to remove water, and then analyzed for CO2, CO, H2, O2. Using the overall combustion reaction for a range of equivalence ratios from 0.5 to 1.5, calcu- late the mole fractions of CO2, CO, H2, and O2 in the dry exhaust gas, and plot the results as a function of equivalence ratio. Assume: (a) that all the fuel is burnt inside the engine (almost true) and that the ratio of moles CO to moles H2 in the exhaust is 3 : 1, and (b) that there is no hydrogen in the exhaust for lean mixtures. For high-power engine operation the air/fuel ratio is 14 : 1. What is the exhaust gas composition, in mole fractions, before the water is removed? REFERENCES 1. Fristrom, R. M ., and Westenberg, A. A.: Flame Structure, McGraw-Hill, 1965. 2. Glassman, I.: Combustion, Academic Press, 1977. 3. Kaye, G. W. C ., and Laby, T. H.: Tables of Physical and Chemical Constants, Longmans, London, 1973. 4. Reynolds, W. C.: Thermodynamic Properties in SI, Department of Mechanical Engineering, Stan- ford University, 1979. 5. Taylor, C. F.: The Internal Combustion Engine in Theory and Practice, vol. 1, MIT Press, Cam- bridge, Mass ., 1960. PROPERTIES OF WORKING FLUIDS 101 TABLE 4.1 CHAPTER Working fluid constituents Process Spark-ignition engine Compression-ignition engine 4 Intake Air Air Fuelt Recycled exhaustt Recycled exhaustt Residual gass Residual gass Compression Air Air PROPERTIES Fuel vapor Recycled exhaust OF WORKING Recycled exhaust Residual gas Residual gas FLUIDS Expansion Combustion products Combustion products (mixture of N2, H2O, (mixture of N2, H2O, CO2, CO, H2, O2, NO, CO2, CO, H2, O2, NO, OH, O, H, ... ) OH, O, H, ... ) Exhaust Combustion products Combustion products [mainly N2, CO2, H2O, (mainly N2, CO2, H2O, and either O2 (¢ < 1) and O2) or CO and H2 (> > 1)] Liquid and vapor in the intake; mainly vapor within the cylinder. Sometimes used to control NO, emissions (see Secs. 11.2, 15.3.2, and 15.5.1). $ Within the cylinder. ciently accurate to assume the composition is frozen. For the compression- ignition engine, the unburned mixture prior to injection contains no fuel; it consists of air and previously burned gas. The combustion products or burned mixture gases, during the combustion 4.1 INTRODUCTION process and much of the expansion process, are close to thermodynamic equi- The study of engine operation through an analysis of the processes that occur librium. The composition of such mixtures has already been discussed (Sec. 3.7.1). inside the engine has a long and productive history. The earliest attempts at this As these combustion products cool, recombination occurs as indicated in Fig. analysis used the constant-volume and constant-pressure ideal cycles as approx- 3-10. Towards the end of the expansion process, the gas composition departs imations to real engine processes (see Chap. 5). With the development of high- from the equilibrium composition; recombination can no longer occur fast speed digital computers, the simulation of engine processes has become much enough to maintain the reacting mixture in equilibrium. During the exhaust more sophisticated and accurate (see Chap. 14). All these engine simulations process, reactions are sufficiently slow so that for calculating thermodynamic (from the simplest to the most complex) require models for the composition and properties the composition can be regarded as frozen. properties of the working fluids inside the engine, as well as models for the indi- The models used for predicting the thermodynamic properties of unburned vidual processes-induction, compression, combustion, expansion, and exhaust- and burned mixtures can be grouped into the five categories listed in Table 4.2. that make up the engine operating cycle. This chapter deals with models for the The first category is only useful for illustrative purposes since the specific heats of working fluid composition, and thermodynamic and transport properties. unburned and burned mixtures are significantly different. While the specific heats The composition of the working fluid, which changes during the engine of the working fluids increase with increasing temperature in the range of interest, operating cycle, is indicated in Table 4.1. The unburned mixture for a spark- a constant-specific-heat model can be matched to the thermodynamic data over a ignition engine during intake and compression consists of air, fuel, and previously limited temperature range. This approach provides a simple analytic model which burned gases. It is, therefore, a mixture of N2, O2, CO2, H2O, CO, and H2 for can be useful when moderate accuracy of prediction will suffice. The appropri- fuel-rich mixtures, and fuel (usually vapor). The composition of the unburned ateness of frozen and equilibrium assumptions has already been discussed above. mixture does not change significantly during intake and compression. It is suffi- Approximations to thermodynamic equilibrium calculations are useful because of 100. 102 INTERNAL COMBUSTION ENGINE FUNDAMENTALS PROPERTIES OF WORKING FLUIDS 103 TABLE 4.2 Categories of models for thermodynamic properties the percent of exhaust gas recycled (%EGR) is defined as the percent of the total intake mixture which is recycled exhaust, t Unburned mixture Burned mixture EGR(%) =( MEGR × 100 1. Single ideal gas throughout operating cycle with c, (and mi (4.2) hence c,) constant 2. Ideal gas; c .,, constant Ideal gas; c .,, constant where MEGR is the mass of exhaust gas recycled, then the burned gas fraction in 3. Frozen mixture of ideal Frozen mixture of ideal the fresh mixture is gases; c ., (T) gases; c .. ((T) - MEGR + m. X6 = -= = EGR 100 (1 - xx) + x, ( 4.3) 4. Frozen mixture of ideal Approximations fitted to mc gases; c.(T) equilibrium thermodynamic properties Up to about 30 percent of the exhaust can be recycled; the burned gas fraction 5. Frozen mixture of ideal Mixture of reacting ideal during compression can, therefore, approach 30 to 40 percent. gases; c .. (T) gases in thermodynamic The composition of the burned gas fraction in the unburned mixture can be equilibrium calculated as follows. The combustion equation for a hydrocarbon fuel of average molar H/C ratio y [e.g ., Eq. (3.5)] can be written per mole O2 as Note: Subscripts i, u, and b denote species i in the gas mixture, the unburned mixture, and burned mixture properties, respectively. EpC + 2(1 - 8)PH2 + 02 + VN2 - nco2 CO2 + H20H2O + ncoCO + H2 H2 + 10202 + NN2N2 (4.4) the savings in computational time, relative to full equilibrium calculations, which where y = the molar N/O ratio (3.773 for air) can result from their use. 4 Values of thermodynamic properties of unburned and burned mixtures rele- vant to engine calculations are available from charts, tables, and algebraic 4 + y relationships developed to match tabulated data. A selection of this material is y = the molar H/C ratio of the fuel included in this chapter and App. D. The references indicate additional sources. ¢ = fuel/air equivalence ratio n; = moles of species i per mole O2 reactant 4.2 UNBURNED MIXTURE The n; are determined using the following assumptions : COMPOSITION 1. For lean and stoichiometric mixtures ( < 1) CO and H2 can be neglected. The mass of charge trapped in the cylinder (m.) is the inducted mass per cycle 2. For rich and stoichiometric mixtures ( 2 1) O2 can be neglected. (m;), plus the residual mass (m,) left over from the previous cycle. The residual 3. For rich mixtures, either (a) the water gas reaction fraction (x,) is CO2 + H2 = CO + H2O X , m me (4.1) Typical residual fractions in spark-ignition engines range from 20 percent at light + An alternative definition of percent EGR is also used based on the ratio of EGR to fresh mixture load to 7 percent at full load. In diesels, the residual fraction is smaller (a few (luel and air): percent) due to the higher compression ratio, and in naturally aspirated engines is approximately constant since the intake is unthrottled. If the inducted mixture is EGR*(%) = MEGR x 100 m . + m s ) fuel and air (or air only), then the burned gas fraction (x3) in the unburned mixture during compression equals the residual fraction. The two definitions are related by In some engines, a fraction of the engine exhaust gases is recycled to the EGR* EGR EGR EGR* intake to dilute the fresh mixture for control of NO, emissions (see Sec. 11.2). If 100 100 - EGR and 100 100 + EGR* 104 INTERNAL COMBUSTION ENGINE FUNDAMENTALS PROPERTIES OF WORKING FLUIDS 105 can be assumed to be in equilibrium with the equilibrium constant K(T): where K(T) = ^H2Onco 5 = 2 nc022H2 2 - 82 (1 - 0) (4.7b) where K(T) can be determined from a curve fit to JANAF table data:8 If we write 1.761 x 103 1.611 x 106 0.2803 x 109 In K(T) = 2.743 - 8Z T T2 + - T 3 (4.5) and 4* = (1- 2 54 (4.7c) where T is in K, or (b) K can be assumed constant over the normal engine operating range. A value of 3.5 is often assumed (see Sec. 4.9), which corre- the reactant expression (4.7a) becomes sponds to evaluating the equilibrium constant at 1740 K. @ *= C + 20*(1 - 8)H2 + 02 + 4*N2 The n; obtained from an element balance and the above assumptions are which is identical in form to the reactant expression for a hydrocarbon fuel (4.4). shown in Table 4.3. The value of c is obtained by solving the quadratic: Thus Table 4.3 can still be used to give the composition of the burned gas (K - 1)c2 - c(K[2(0 - 1) + €¢] + 2(1 - 0)} + 2Ke0( - 1) = 0 (4.6) residual fraction in the unburned mixture, except that o* replaces o and * replaces y in the expressions for ni. The mole fractions are given by Now consider the unburned mixture. The number of moles of fuel per mole El O2 in the mixture depends on the molecular weight of the fuel, M. If the no average molecular formula of the fuel is (CH,). then where n = >; n is given in the bottom line of Table 4.3. M, = a(12 + y) While Eq. (4.4) is for a fuel containing C and H only, it can readily be modified for alcohols or alcohol-hydrocarbon blends. For a fuel of molar com- The fresh fuel-air mixture (not yet diluted with EGR or residual), position CH, O2, the reactant mixture eQC + 2(1 - :)@H2 + 02 + VN2 CH, 02 +- (1+4-3(02 +VN2) then becomes can be rearranged per mole of O2 reactant as 4 (1 + 28 ) ¢( CH, )a + O2 + VN2 (DEC + 250(1 - E)H2 + 02+ (1 -3)SVN2 (4.7a) The unburned mixture (fuel, air, and a burned gas fraction), per mole O2 in the mixture, can be written: TABLE 4.3 Burned gas composition under 1700 K (1 - xx ) Ms ( 1 + 22 ) 6 ( CH , la + 02 + WN 2 n;, moles/mole O2 reactant + x(nco2 + "H20 + nco + 1H2 + no2 + "N2) Species The number of moles of each species in the unburned mixture, per mole O2, is CO2 EQ - C summarized in Table 4.4. The mole fractions of each species are obtained by H,O 2(1 - 2)¿ 2(1 - 80) + c dividing by the total number of moles of unburned mixture ny, CO 0 c H2 2(¢ - 1) - c 4(1 + 28 )@ O2 1 - 0 n = (1 -xx) + 1+ 4 + * (4.8) N, Sum: n. (1 - 8)¢ + 1 + ¥ (2-8)$ +¥ where ny is given in Table 4.3. t c defined by Eq. (4.6). The molecular weights of the (low-temperature) burned and unburned 106 INTERNAL COMBUSTION ENGINE FUNDAMENTALS PROPERTIES OF WORKING FLUIDS 107 TABLE 4.4 TABLE 4.5 Unburned mixture composition Factors for relating properties on molar and mass basis n;, moles/mole O2 reactant Quantity, per mole Equation for C3H18-air O2 in the mixture General equationt mixtures Species ¢ > 1 Moles of burned no = (1 -e)$ + 1+4, 7 = 0.36¢ + 4.773 Fuel 4(1 - x3)(1 + 28)¢/M, mixture no ny = (2 - 2)¢ + ¥, $ > 1 n = 1.36¢ + 3.773 O2 1 - X, ¢ 1 - x5 N2 n_ = 0.08¢ + 4.773 CO2 x ED x3(€¢ - c) Moles of unburned (1 - x%) [411 + 25 )4 + 1 + 4 + com +0.28×3 ¢ H2O 2x,(1 - £)¢ x,[2(1 - ed) + c] mixture n. 10>1 n_ = 0.08¢ + 4.773 CO 0 x. c +x,(1.28¢ - 1) H2 0 x5[2(¢ - 1) - c] Mass of mixturet mxp = 32 + 4d(1 + 28) + 28.16¢ 138.2 + 9.12¢ Sumt n (burned or unburned) t Given by Eq. (4.8). Mass of airt 32 + 28.16% 138.2 + Units: kg/kmol or Ibm/lb . mol. mixture can now be determined. The mass of mixture (burned or unburned) per For hydrocarbon fuels, y for air = 3.773; for fuels containing oxygen, o* and * given by Eq. (4.7c) are substi- tuted for o and w, respectively. mole O2 in the mixture, MRp , is given by MRP = 32 + 40(1 + 28) + 28.16% (4.9) The molecular weight of the burned mixture, Mb, is therefore The molecular weight of the unburned mixture, M ., is M = MRP (4.10) M = RP (4.11) Figure 4-1 gives M. and M, for a range of o and x for air, isooctane, burned gas 32 mixtures. Frequently, thermodynamic properties of unburned and burned mixtures 31 are expressed per unit mass of air in the original mixture (for burned mixture this ID = 0 is the mixture before combustion). To obtain properties in these units, we need ).1 the mass of original air, per mole O2 in the mixture, which is 30 My 0.2 0.3 (32 + 28.16 4) 0.4 29 with units of kilograms per kilomole or pound-mass per pound-mole. Table 4.5 summarizes the factors needed to relate properties expressed on a Molecular weight molar and a mass basis. 28 --- 27 4.3 GAS PROPERTY RELATIONSHIPS 26- FIGURE 4-1 The individual species in the unburned and burned gas mixtures can with suffi- Molecular weight of unburned and low- cient accuracy be modeled as ideal gases. Ideal gas relationships are reviewed in temperature burned isooctane-air 0.4 0.6 0.8 1.0 1 . 2 1.4 mixtures as a function of fuel/air equiva- App. B. The most important relationships for property determination for engine Equivalence ratio ¢ lence ratio and burned gas fraction. calculations are summarized below. 108 INTERNAL COMBUSTION ENGINE FUNDAMENTALS PROPERTIES OF WORKING FLUIDS 109 Since internal energy and enthalpy are functions of temperature only, the In these equations, the units of u and h can be on a per unit mass or molar specific heats at constant volume and constant pressure are given by basis [i.e ., joules per kilogram (British thermal units per pound-mass) or joules du per kilomole (British thermal units per pound-mole)]; similarly, s, co, cp, R, Y, aT = = C ( T ) (4.12a) and @ can be in joules per kilogram-kelvin (British thermal units per pound- mass-degree Rankine) or joules per kilomole-kelvin (British thermal units per oh dh Cp = = cp(T) pound-mole-degree Rankine). OT dT (4.12b) For gas mixtures, once the composition is known, mixture properties are and determined either on a mass or molar basis from (4.19a) u - uo = c, dT (4.13a) JTo h = [ xchi (4.19b) (4.19c) h - ho = s = [ xis; CpdT (4.13b) JTo and The entropy s(T, v) or s(T, p) is given by C. = 2 xico.i (4.20a) S - So = C + + RIn (4.20b) JTo Do (4.14a) Cp = [ x, Cp.i s - So= dT Cp F - R In P (4.14b) JTo 4.4 A SIMPLE ANALYTIC The integrals in Eqs. (4.14a, b) are functions of temperature only, and are IDEAL GAS MODEL useful in evaluating entropy changes and in following isentropic processes. If we define While the first category of model listed in Table 4.2 is too inaccurate for other than illustrative purposes, the second category-constant but different specific (4.15a) heats for the unburned and burned gas mixtures-can with careful choice of specific heat values be made much more precise. The advantages of a simple analytic model may be important for certain problems. and (4.15b) Figure 4-2 shows an internal energy versus temperature plot for a stoichio- metric mixture. It is a quantitative version of Fig. 3-5. The unburned mixture line then is for a burned gas fraction of 0.1. The fuel is isooctane. Data to construct such graphs can be obtained from charts or tables or computer programs (see Secs. 4.5 S - So = Y + R In (4.16a) to 4.7). The units for u are kilojoules per kilogram of air in the original mixture (the units of the charts in Sec. 4.5). The datum is zero enthalpy for O2, N2, H2, and C (solid) at 298 K. Note that the specific heats of the unburned and burned S - So = D - R in (4.16b) mixtures (the slopes of the lines in Fig. 4-2) are a function of temperature; at high temperatures, the internal energy of the burned mixture is a function of tem- Thus, for example, the entropy change between states (T1, P1) and (T2, P2) is perature and pressure. However, the temperature range of interest for the unburned mixture in an $2 - S1 = 02 - 01 - R in (Pz (4.17) SI engine is 400 to 900 K (700 to 1600ºR); for the burned gas mixture, the extreme end states are approximately 2800 K, 35 atm (5000ºR, 500 1b/in2 abs) For an isentropic process, and 1200 K, 2 atm (2200ºR, 30 1b/in2 abs). Linear approximations to the unburned and burned mixture curves which minimize the error in u over the In Pz = temperature (and pressure) ranges of interest are shown as dashed lines. The error PI (4.18) R in T for a given u is less than 50 K. 110 INTERNAL COMBUSTION ENGINE FUNDAMENTALS PROPERTIES OF WORKING FLUIDS 111 For a constant-pressure adiabatic combustion process, 2000 ¢ = 1.0, x = 0.1 P and it can similarly be shown that 1000 -- 10 T = YO - 1 ( M2 ) ( . T. + Ahs) (4.24) 100 To use the model, suitable values of yu, Yo, M ., (M./M.), and Ah,/R, must atm be determined. Values for M, and M, can be obtained from Eqs. (4.10) and (4.11).+ Values of yu, Yo, and Ah,/Ry, can be obtained from graphs such as Fig. 4-2 Internal energy, KJ/kg air - 1000F- (see Example 4.1 below). Values for Yu, Y, and Ah,/R, are available in the liter- ature (e.g ., Refs. 1 and 2) for a range of o and x ,. However, values used for computations should always be checked over the temperature range of interest, 2000 to ensure that the particular linear fit to u(T) used is appropriate. Example 4.1. Determine the values of yu, Yo, and Ah,/R, which correspond to the straight-line fits for u_(T) and up(T) in Fig. 4-2. -3000 -- Equations for the straight lines in Fig. 4-2 are u. (KJ/kg air) = 0.967(K) - 700 500 1000 1500 2000 2500 3000 and ub (KJ/kg air) = 1.57(K) - 4250 Temperature, K From Table 4.5, for isooctane fuel with ( = 1.0 and x) = 0.1, the number of FIGURE 4-2 moles of unburned mixture per mole O2 in the mixture is Internal energy versus temperature plot for stoichiometric unburned and burned gas mixtures: iso- octane fuel; unburned residual fraction 0.1. n_ = 0.08 x 1 + 4.773 + 0.28 x 0.1 x 1 = 4.881 The mass of air per mole O2 in the mixture is 138.2. Thus, the number of moles of unburned mixture per unit mass of air in the original mixture is The basis for this ideal gas model is 4.881 = 0.0353 hy = Cp. . Tut hs .u (4.21a, b) 138.2 U , = Co , b To + hs .b ho = Cp. o To + hs .b (4.22a, b) The molar specific heat of the unburned mixture cy. is therefore 0.96 where hy, and h. are the enthalpies of formation of unburned and burned gas - 0.0353 = 27.2 kJ/kmol . K mixture, respectively, at 0 K. Then, for a constant-volume adiabatic combustion process, Since R = 8.314 kJ/kmol . K, 27.2 + 8.314 = 1.31 27.2 or Co . .. Tu + hs u = Co, 6 Th + hs.b The number of moles of burned mixture per mole O2 is (from Table 4.5) If we solve for T; and use the relations (R$/R.) = (M/M)) and c./R = 1/(y - 1), 1) = 0.36 x 1 + 4.773 = 5.133 we obtain R. ) (4.23) where Ah , = hs .. - hs.b . + The error in ignoring the effect of dissociation on M, is small. 112 INTERNAL COMBUSTION ENGINE FUNDAMENTALS PROPERTIES OF WORKING FLUIDS 113 The number of moles of burned mixture per unit mass of air in the original mixture TABLE 4.6 is Unburned mixture composition for charts 5.133 = 0.0371 Equivalence Kilograms of mixture Moles of mixture Kilomole of mixture n. R.t 138.2 ratio ¢ (F/A) per kilogram of air . per mole of O2 per kilogram of air J/kg air . K The molar specific heat ¿ ,,, is therefore 0.4 0.0264 1.0264 4.805 + 0.112x, 0.0348 + 0.00081x3 289 0.6 0.0396 1.0396 4.821 + 0.168x, 0.0349 + 0.00122x> 290 1.5 = 40.4 kJ/kmol - K 0.8 0.0528 1.0528 4.837 + 0.224x3 Co b = 0.0350 + 0.00162x, 291 0.0371 1.0 0.0661 1.0661 4.853 + 0.280x> 0.0351 + 0.00203x 292 1.2 0.0792 1.0792 4.869 + 0.536x, 0.0352 + 0.00388x3 292 and y is 40.4 + 8.314 For x, = 0. Error in neglecting x, is usually small. Yo = - = 1.21 40.4 To find Ah,/R ., R, is given by 2. The fuel is in the vapor phase. Ry = 8.314 × 0.0353 = 0.293 KJ/kg air ~K 3. The mixture composition is homogeneous and frozen (no reactions between the fuel and air). and so 4. Each species in the mixture can be modeled as an ideal gas. Ah_ (-700) - (-4250) 2 = 1.2 × 104 K 5. The burned gas fraction is zero.+ 0.293 It proves convenient to assign zero internal energy or enthalpy to the 4.5 THERMODYNAMIC unburned mixture at 298.15 K. Internal energy and enthalpies relative to this CHARTS datum are called sensible internal energy u, or sensible enthalpy h ,. By sensible we mean changes in u or h which result from changes in temperature alone, and we One method of presenting thermodynamic properties of unburned and burned exclude changes due to chemical reaction or phase change. gas mixtures for internal combustion engine calculations is on charts. Two sets of Table 4.6 provides the basic composition data for the unburned mixture charts are in common use: those developed by Hottel et al.3 and those developed charts. Equations (4.13a, b) provide the basis for obtaining the u ,. „(T) and h ,, „(T) by Newhall and Starkman.4, 5 Both these sets of charts use U.S. units. We have curves shown in Fig. 4-3. developed a new set of charts in SI units, following the approach of Newhall and Equations (4.15) and (4.16) provide the basis for following a reversible adia- Starkman. Charts are no longer used extensively for engine cycle calculations; batic (i.e ., isentropic) compression process. Between end states 1 and 2, we obtain, computer models for the thermodynamic properties of working fluids have per kilogram of air in the mixture, replaced the charts. Nonetheless, charts are useful for illustrative purposes, and afford an easy and accurate method where a limited number of calculations are required. The charts presented below are for isooctane fuel, and the following Y(T2) = Y(T;) - n, R in (2) (4.25a) equivalence ratios: ¢ = 0.4, 0.6, 0.8, 1.0, 1.2. @(T2) = @(71) + n k ln ( P2) (4.25b) 4.5.1 Unburned Mixture Charts The thermodynamic properties of each unburned fuel-air mixture are represented where n, is the number of moles of unburned mixture per kilogram of air. Values by two charts. The first chart is designed to relate the mixture temperature, pres- sure, and volume at the beginning and at the end of the compression process; the second gives the mixture internal energy and enthalpy as functions of tem- perature. 1 This assumption introduces negligible error into calculations of the compression process for mix- The following assumptions are made: tures with normal burned gas fractions, since the major constituent of the residual is N ,. The burned gas fraction must, however, be included when the unburned mixture properties are related to burned 1. The compression process is reversible and adiabatic. mixture properties in a combustion process. 114 INTERNAL COMBUSTION ENGINE FUNDAMENTALS PROPERTIES OF WORKING FLUIDS 115 1400 = 1.0 = 0.8 Figure 4-4 then gives ₱ 0.6 T2 = 682 K 1200 „¢ = 0.4 The ideal gas law [Eq. (4.26)] gives Unburned mixture properties = 1.2 Fuel: Isooctane, C8H18 ¢ = 1.0 292 × 350 Ó = 0.8 V1= = 1.0 m3/kg air 1000 1 x 1.013 x 105 ò = 0.6 ò = 0.4 682 and P2 - P1 Ti) 350 - x 8 = 15.5 atm 800 h. 1.0 V2 = = 0.125 m 3/kg air Sensible enthalpy and internal energy, KJ/kg air 600 Note that p2 can also be obtained from Fig. 4-4 and Eq. (4.25b): In ( P2 2 - 01 _ 980 - 180 292 - = 2.74 400 P2 = 15.5 atm = 1.57 MPa The compression stroke work, assuming the process is adiabatic and using the 200 data in Fig. 4-3, is - W1-2 = 4,(72) - u,(T1) = 350 - 40 = 310 KJ/kg air 0 300 400 500 600 700 800. 900 1000 1100 1200 1300 = 1.2 1400 ៛ = 1.0 Temperature, K = 0.8 ֏ 0.6 FIGURE 4-3 = 0.4 € Sensible enthalpy and internal energy of unburned isooctane-air mixtures as function of temperature. 200 Units: KJ/kg air in mixture. Isentropic compression chart = 1.2 for unburned mixture ¢ = 1.0 Fuel: Isooctane, C8H18 = 0.8 1000 ₲ = 0.6 b = 0.4 of n, and ny R are given in Table 4.6. "I(T) and ((T) are given in Fig. 4-4. Note that u, p, and T are related by 800 $(7). @(T) = n. R in (p/po) and Y(T) = - n. R In (r/co), J/kg air . K p(Pa)(m3/kg air) = n_R(J/kg air -K)T(K) (4.26) 500 Example 4.2. The compression process in an internal combustion engine can be modeled approximately as adiabatic and reversible (i.e ., as an isentropic process). A 400 spark-ignition engine with a compression ratio of 8 operates with a stochiometric fuel vapor-air mixture which is at 350 K and 1 atm at the start of the compression stroke. Find the temperature, pressure, and volume per unit mass of air at the end of 200 the compression stroke. Calculate the compression stroke work. Given T1 = 350 K at the start of compression, find T2 at the end of compres- 0 sion using the isentropic compression chart, Fig. 4-4, and Eq. (4.25a). For T1 = 350 300 400 500 600 700 800 900 1000 K, Y = 150 J/kg air . K. From Eq. (4.25a), Temperature, K FIGURE 4-4 Y 2 ( 7 2 ) = Y , ( T ) - n R In D2 = 150 - 292 In = 757 J/kg air . K Isentropic compression functions, D and , as function of temperature for unburned isooctane-air mixtures. Units: J/kg air . K. 116 INTERNAL COMBUSTION ENGINE FUNDAMENTALS 4.5.2 Burned Mixture Charts The primary burned mixture charts are for the products of combustion at high SCE 101 = 0 25.0 20 40.0 temperatures, i.e ., for the working fluid during the expansion process. The follow- = 50 = 2800 001 = A = 4 E T = 2600 ing assumptions are made: D = 200 1. Each species in the mixture can be modeled as an ideal gas. = 4.0 T = 2400 2. The mixture is in thermodynamic equilibrium at temperatures above 1700 K; V = 2.5 00$ = d T = 2200 the mixture composition is frozen below 1700 K. 3. Datum. At the datum state of 298.15 K (25ºC or 77ºF) and 1 atm the chemical T = 2000 p = 1,000 = 1800 elements in their naturally occurring form (N2, O2, H2 as diatomic gases and C as solid graphite) are assigned zero enthalpy and entropy. -V = 1.0. - T = 1600 p = 2,200 The charts were prepared with the NASA equilibrium program described in - T = 1400 Sec. 3.7.9,10 The C/H/O/N ratio of the mixture is specified for each chart. The T = 1200 - v = 0.38 extensive properties (internal energy, enthalpy, entropy, and specific volume) are P = 4,000 9.7 all expressed per unit mass of air in the original mixture; i.e ., they correspond to the combustion of 1 kg of air with the appropriate mass of fuel. The mass basis v = 0.20 p = 6,500 for the unburned and burned mixture charts are the same. 9.5 Figures 4-5 to 4-9 are property charts for the high-temperature burned gas; - v = 0.10 9.3 each is a plot of internal energy versus entropy for a particular fuel and equiva- < p = 13,000 lence ratio. Lines of constant temperature, pressure, and specific volume are p = 20,000 Entropy s, KJ/kg air . K drawn on each chart. An illustration of the use of these charts follows. v = 0.04 8.9 9.1 Example 4.3. The expansion process in an internal combustion engine, following completion of combustion, can be modeled approximately as an adiabatic and - v = 0.02 8.7 reversible process (i.e ., isentropic). Under full-load operation, the pressure in the cylinder of a spark-ignition engine at top-center immediately following combustion is 7100 kPa. Find the gas state at the end of the expansion stroke and the expansion T = 36007- Internal energy versus entropy chart for equilibrium burned gas mixture, isooctane fuel; equivalence ratio 0.4. stroke work. The compression ratio is 8, the mixture is stoichiometric, and the T = 3400/ volume per unit mass of air at the start of expansion is 0.125 m3/kg air. T = 3200/ Locate p1 = 7100 kPa and v1 = 0.125 m3/kg air on the o = 1.0 burned gas chart (Fig. 4-8). This gives T1 = 2825 K, un = - 5 kJ/kg air, and s, = 9.33 KJ/kg air . K. The gas expands at constant entropy to 02 = 8 x 01 = 1 m3/kg air. Follow- ing a constant entropy process from state 1 on Fig. 4-8 gives Fuel: Isooctane C8 H18 Burned gas properties Volume: v, m3/kg air Equivalence ratio: 0.4 7.1 7.3 7.5 7.7 7.9 8.1 8.3 8.5 Pressure: p, kN/m2 Temperature: T, K T2 = 1840 K, P2 = 570 kPa, and U2 = - 1540 KJ/kg air The expansion stroke work, assuming the process is adiabatic, is W1-2= - (42 - 11) = 1540 - 5 = 1535 KJ/kg air As the burned gases in an engine cylinder cool during the expansion process, the composition eventually "freezes"-becomes fixed in composition- 4600 4200 3000 380 1800 600 340 2600 2200 1400 200 because the chemical reactions become extremely slow. This is usually assumed 1000 -200 FIGURE 45 -600 to occur at about 1700 K (see Sec. 4.9). The equilibrium assumption is then no Energy u, KJ/kg air longer valid. For lean and stoichiometric mixtures this distinction is not impor- tant because the mole fractions of dissociated species below this temperature are 117 1.0 = 4.0 p = 500 p = 1,000 = 200 v = 2.5 = 118 4000 10.0 p Burned gas properties = 2,200 = 101.325 3600 Pressure: p, kN/m2 v = 0.38 Volume: v, m3/kg air -- p = 4,000 v = 0.20 - p = 6,500 3200 Temperature: T, K -v = 0.10 Fuel: Isooctane C8 H18 p = 13,000 P = 20,000 2800 Equivalence ratio: 0.6 -v= 0.02 -v = 0.04 P = 50 -- v = 25.0 2400 = 40.0 T = 3600 2000 V P = 20 T = 3400/ 1600 Energy u, KJ/kg air T = 3200 1200 T = 2600 T = 3000/ 800 T = 2800/1 T = 2400 400 -T = 2200 0 T = 2000 -400 T = 1800 -800 -T = 1600 LT = 1400 __T = 1200 -12007 7.2 7.4 7.6 7.8 8.2 8.4 8.6 8.8 9.0 1.2 9.4 9 9.6 9.8 10.0 10.2' 10.4 10.6 10.8 11.0 11.2 Entropy s, KJ/kg air .K FIGURE 4-6 Internal energy versus entropy chart for equilibrium burned gas mixture, isooctane fuel; equivalence ratio 0.6. 3400 __v = 1.0 = 500 3000 = 2,200 ₭= 2.5 - v = 0.38 p = 1,000 P = 4,000 --- = 4.0 VP = 200) Burned gas properties = 0.2 -P = 6,500 2600 = 20,000 y = 0.10 Pressure: p, kN/m2 - v = 0.04 P = 13,000 = 101.325 = 0.02 Volume: v, m3/kg air -- v = 10 2200 Temperature: T, K - 1800 Fuel: Isooctane C8 H18 Equivalence ratio: 0.8 T = 3600- 25 P = 5 1400 T = 34007 40 1000 = 20 = T = 3200 Energy u, KJ/kg air 600 T = 3000/ 200 T = 2800 - T = 2400 T = 2600 -200 -600 T = 2200 - T = 2000 - 1000 T = 1800- - 1400 - T = 1600 T = 1400 - 1800, - T = 1200 7.3 7.5 7.7 7.9 8.1 8.3 8.5 8.7 8.9 9.1 9.3 9.5 9.7 9.9 10.1 10.3 10.5 10.7 10.9 11.1 11.3 Entropy s, KJ/kg air . K FIGURE 4-7 119 Internal energy versus entropy chart for equilibrium burned gas mixture, isooctane fuel; equivalence ratio 0.8. 120 2800 = 1.0 p = 4,000 P = 6,500 -v = 0.38 P = 2,200 p = 1,000 -v = 2.5 2400 = 13,000 7v = 0.10 Kv = 0.20 P = 500 -v = 4.0 = 0.04 Burned gas properties P = 200 V = 0.02 2000 Pressure: p, kN/m2 101.325 Volume: v, m3/kg air - 1600 Temperature: T, K = 10-7 Fuel: Isooctane C8 H18 T = 3600/ 1200 Equivalence ratio: 1.0 50 T = 3400/ 800 IT = 3200 Yy = 25 400 p = 20 40 Energy u, KJ/kg air T = 3000 V= 0 T = 2800 2400 = 5 -400|- T = 2600/ - 800 =T = 2200 - 1200 T = 2000 - 1600- T = 1800 -2000 +T = 1600 -T = 1400 -2400L -T = 1200 7.4 7.6 7.8 8.0 8.2 8.4 8.6 8.8 9.0 9.2 9.4 9.6 9.8 10.0 10.2 10.4 10.6 10.8 11.0 11.2 11.4 Entropy s, KJ/kg air . K FIGURE 4-8 Internal energy versus entropy chart for equilibrium burned gas mixture, isooctane fuel; equivalence ratio 1.0. 3000 = 1.0 2600! v = 0.38 p = 1,000 p = 500 tv = 2.5 Av = 4.0 P = 2,200, V = 0.20 P = 4,000 Burned gas properties P = 6,500 2200 Pressure: p, kN/m2 = 13,000 v = 0.10 - p = 200 P = 20,000 +v = 0.04 1800 Volume: v, m3/kg air -- = 0.02 Temperature: 7, K = 101.325 1400 Fuel: Isooctane C8 H18 v = 10 Equivalence ratio: 1.2 T = 3600 1000 T = 34007 600 ? = T = 3200 V = Energy u, KJ/kg air 200 -T = 3000 p = 20 y : -200 T = 2800 T = 2600/ -600 -T = 2400 - 1000 -T = 2200 - 1400 T = 2000 -T = 1800 - 1800 -T = 1600 -2200 -T = 1400 -T = 1200 7.8 8.0 8.2. 8.4 8.6 8.8 9.0 9.2 9.4 9.6 9.8 10.0 10.2 10.4 10.6 10.8 11.0 11.2 11.4 11.6 11.8 FIGURE 4-9 Entropy s, KJ/kg air . K 121 Internal energy versus entropy chart for equilibrium burned gas mixture, isooctane fuel; equivalence ratio 1.2. 122 INTERNAL COMBUSTION ENGINE FUNDAMENTALS PROPERTIES OF WORKING FLUIDS 123 1400 = 1.0 .b = 0.8 small. For rich mixtures, a frozen composition must be selected and used because "= 0.6 the mole fractions of CO2, CO, H2O, and H2 would continue to change if equi- ¢ = 0.4 1200 librium is assumed as the temperature decreases. Internal energy and enthalpy, Burned mixture properties d = 1.2 per kilogram of air in the original mixture, of the frozen burned mixture are b = 1.0 Fuel: Isooctane, C&H18 plotted against temperature in Fig. 4-10. The assumed frozen burned mixture ¢ = 0.8 1000 ₲ = 0.6 compositions are listed in Table 4.7. These are sensible internal energies and = 0.4 enthalpies, given relative to their values at 298.15 K. 800 4.5.3 Relation between Unburned Sensible enthalpy and internal energy, KJ/kg air and Burned Mixture Charts 600 We now address the questions: Given unburned mixture at T1, P1, 01, what is the state of the burned mixture following (1) constant-volume adiabatic combustion or (2) constant-pressure adiabatic combustion ? 400 The datum for internal energy and enthalpy for the unburned mixture in Fig. 4-3 is different from the datum for internal energy and enthalpy for the burned mixture. For the unburned mixture, zero internal energy and enthalpy for 200 the mixture at 298.15 K was assumed. For the burned mixture, zero enthalpy for the gaseous species O2, N2, and H2, and C (solid graphite) at 298.15 K was assumed. These data can be related through the enthalpies of formation, from 0 O2, N2, H2, and C, of each species in the unburned mixture. 300 400 500 600 700 800 900 1000 1100 1200 1300 If Ah;, is the enthalpy of formation of species i at 298.15 K, per kilomole, Temperature, K and Ahy ., is the enthalpy of formation of the unburned mixture at 298.15 K, per FIGURE 4-10 kilogram of air in the original mixture, then Sensible enthalpy and internal energy of low-temperature burned gases as function of temperature, isooctane fuel. Units: KJ/kg air in original mixture. Ahi .. = En, Ahi.: (4.27) where n; is the number of kilomoles of species i per kilogram of air. The unburned mixture enthalpy hy, with the same datum as the burned mixture enth- alpy, is therefore given by the sum of the sensible enthalpy h ,,, and Ah? .. : TABLE 4.7 Frozen burned gas composition: C8H18-air combustion h = hs ,. + Ahi .. (4.28) Similarly, the internal energy u, is given by CO2 H2O CO H2 02 N2 Sum Units Un = Us " + Au " . (4.29) 0.4 0.0521 0.0586 0.122 0.767 1.000 mole fractions 1.85 2.08 4.34 27.3 35.6 mol/kg airt Auf .. can be obtained from 0.6 0.0770 0.0866 0.0802 0.756 1.000 mole fractions 2.78 3.13 2.89 27.3 36.1 mol/kg airt (4.30) 0.8 0.101 0.113 0.0395 0.746 1.000 mole fractions 3.70 4.14 1.45 27.3 36.6 mol/kg airt Alternately, Eq. (3.18) can be used to obtain Aus.„ from Ahy:' 1.0 0.125 0.140 0.735 1.000 mole fractions 4.64 5.2 27.3 37.1 mol/kg airt Auj.„ = Ahy .. - (np - nr)RT (4.31) 1.2+ 0.0905 0.138 0.0516 0.0224 0.698 1.000 mole fractions 3.54 5.38 2.02 0.876 27.3 39.1 mol/kg airt Enthalpies and internal energies of formation of the relevant burned gas species and individual fuel compounds are given in Table 4.8 and App. D. Values K(T) in Eq. (4.6) evaluated at 1740 K; K = 3.5. of n; are obtained from Tables 4.4 and 4.7. Following the procedure used in Note mol/kg air; multiply by 10-3 for kmol/kg air. Example 4.4 below, expressions for Ah? ,, and Auf .,, in kilojoules per kilogram of 124 INTERNAL COMBUSTION ENGINE FUNDAMENTALS PROPERTIES OF WORKING FLUIDS 125 TABLE 4.8 With Ahj, from Table 4.8, Eq. (4.27) gives Standard enthalpies and internal energies of formation+ Ahy = 5.787 x 10-4 x (-224.1 x 1061 - x3) Ahy .,, MJ/kmol Aug.1, MJ/kmol + x8[4.629 × 10-3 x (- 393.5 x 106) + 5.208 × 10-3 x (-241.8 x 106)] CO2 -393.5 -393.5 Ahj. = (-129.7 - 2951x%) x 103 .J/kg air H2O (gas) -241.8 -240.6 With Aug ., from Table 4.8, Eq. (4.30) gives CO -110.5 -- 111.7 C8H18 (gas) -224.1 -204.3 Aug ., = 5.787 x 10-4 x (-204.3 x 106X1 - x) + At 298.15 K. Ah, for O2, N ,, and H, are zero by defini- + x6[4.629 x 10-3 x (-393.5 x 106) + 5.208 x 10-3 x (-240.6 x 106)] tion. Sources: JANAF tables," Rossini et al.16 Auj .. = (-118.2 - 2956x%) x 103 J/kg air Alternatively, we can determine Auf .,, from Ah? ., using Eq. (4.31). For this calcu- air can be obtained. For the charts of Figs. 4-3 and 4-5 to 4-9, these expressions lation, the "product" gas is the unburned mixture and the "reactant" gas is the are : mixture of elements from which the unburned mixture is formed. The number of gaseous moles in the unburned mixture np, per mole O2 in the original mixture, is $ = 0.4: (from Table 4.5 for ( < 1) Ahy = - 51.9 - 1181x, Auj. = - 47.3 - 1183x $ = 0.6: np = ( 1 - x 6 ) 4( 1 + 2 6 ) 2 + 1 + 4 + * [ ( 1 - 2 ) 0 + 1 + 4 ] Ahy = - 77.8 - 1771x, Aus = - 70.9 - 1774x The elemental reactant mixture from which the unburned mixture is formed is, from Eq. (4.4), ¢ = 0.8: Ahy. = - 103.8 - 2361x, Auf = - 94.6 - 2365x% (4.32) epC + 2(1 - 8)0H2 + 02 + VN2 Thus, nR, the moles of gaseous elements, is ¢ = 1.0: Auj. = - 118.2 - 2956x nr = 2( 1 - 8 ) 0 + 1 + 4 Ahy = - 129.7 - 2951x, For air, 4 = 3.773; for CH18 fuel, e = 0.64 and M, = 114. For o = 1, ¢ = 1.2: Aug ., = - 141.9 - 2769%% np - nr = - 0.64 + 0.28x, Ah9 .. = - 155.6 - 2759x5 moles/mole O2 and (np - na)RT = (-0.64 + 0.28x3) x 8.3143 x 103 x - 298.15 Example 4.4. Calculate Ah? „, the enthalpy of formation of the unburned mixture, ‘ 138.2 and Au? „, the internal energy of formation of the unburned mixture, for a or (np - nr)RT = (-11.5 + 5.0x2) x 103 J/ kg air C8H18-air mixture with o = 1.0 and burned gas fraction x3. Table 4.4 gives the moles of each species in the unburned mixture, per mole Since O2 with o = 1.0, as Auf.„ = Ahy .. - (np - np)RT CgH18, 0.08(1 - x%) CO2, 0.64x3 Aus .. = (-129.7 - 2951x%) x 103 - (-11.5 + 5.0x%) × 103 O2, 1- xb H2O, 0.72x Aus. = (-118.2 - 2956x%) x 103 J/kg air N2, 3.773 CO and H2, 0 Table 4.5 gives the mass of air per mole O2 as 138.2 kg/kmol. Thus the number of The combustion process links the unburned and burned mixture properties as follows: kilomoles of each species per kilogram of air is CgHis, 5.787 x 10-4 (1 - x%) CO2, 4.629 x 10-3x3 For an adiabatic constant-volume combustion process, O2, 7.233 x 10-3 (1 - x%) H2O, 5.208 x 10-3x3 U = u = Us + Auj .. (4.33) N2, 2.729 x 10-2 CO and H2, 0 and 126 INTERNAL COMBUSTION ENGINE FUNDAMENTALS PROPERTIES OF WORKING FLUIDS 127 Thus, given u, , and vy, the state of the burned mixture can be determined from A trial-and-error solution for , and u, along the p = 1570 kPa line on Fig. 4-8 gives the appropriate burned mixture chart. 4) = - 655 kj/kg air, T; = 2440 K, ", = 0.485 m3/kg air For an adiabatic constant-pressure combustion process, (Use the ideal gas law to estimate p, T, or v more accurately.) ho = hy = hs, + Ahi .. (4.34) Since 4.6 TABLES OF PROPERTIES AND COMPOSITION given hy, and p, 4, and v, must be found by trial and error along the specified Tables of thermodynamic properties of air are useful for analysis of motored constant-pressure line on the appropriate burned mixture chart. engine operation, diesels and compressors. Keenan, Chao, and Kaye's Gas Tables6 are the standard reference for the thermodynamic properties of air at low pressures (i.e ., at pressures substantially below the critical pressure when the ideal Example 4.5. Calculate the temperature and pressure after constant-volume adia- gas law is accurate). These gas tables are in U.S. and SI units. A set of tables for batic combustion and constant-pressure adiabatic combustion of the unburned mixture (with ¢ == 1.0 and x) = 0.08) at the state corresponding to the end of the air in SI units has been prepared by Reynolds7 following the format of the compression process examined in Example 4.2. Keenan et al. tables. A condensed table of thermodynamic properties of air, The state of the unburned mixture at the end of the compression process in derived from Reynolds, is given in App. D. It contains: Example 4.2 was h = enthalpy, KJ/kg T. = 682 K, 43.x = 350 KJ/kg air, pu = 1.57 MPa, v = 0.125 m3/kg air u = internal energy, KJ/kg For an adiabatic constant-volume combustion process [Eq. (4.33)], Y = dT, KJ/kg . K JO 4x = U = U + Auf .. O = dT, KJ/kg . K For d = 1.0, Auf .. is given by Eq. (4.32) as JO p, = relative pressure Auf.„ = - 118.2 - 2956x) = - 118.2 - 236.5 = - 355 KJ/kg air v, = relative volume Hence c, = specific heat at constant pressure, KJ/kg . K („ = specific heat at constant volume, kJ/kg . K Ub = 350 - 355 = - 5 KJ/kg air y = ratio of specific heats Also all as a function of T(K). 1) = vy = 0.125 m3/kg air D is the standard state entropy at temperature T and 1 atm pressure, rela- Locating (u ., Us) on the burned gas chart (Fig. 4-8) gives tive to the entropy at 0 K and 1 atm pressure. The entropy at pressures other than 1 atm is obtained using Eq. (4.14b). T; = 2825 K, Po = 7100 kPa The relative pressure p, is defined by For a constant-pressure combustion process [Eq. (4.34)], h = h = h + Ahi .. In p, = = (4.35) For $ = 1.0, Ahy ., is given by Eq. (4.32) as and is a function of T only. Along a given isentropic, it follows from Eq. (4.18) Ah9.„ = - 129.7 - 2951x) = - 129.7 - 236 = - 366 KJ/kg air that the ratio of actual pressures p2 and p1 corresponding to temperatures T2 and T1 is equal to the ratio of relative pressures, i.e ., At T. = 682 K, h ,.. = 465 KJ/kg air, so h) = 465 - 366 = 99 KJ/kg air P 2 P1s=const = (4.36) Since p. = P. = 1.57 MPa, the internal energy u, is given by This affords a means of determining T2, for an isentropic process, given T1 and 4x = h, - P. U. = 99 - 1.57 x 1030, KJ/kg air P2/P1 (see Example 4.6). 128 INTERNAL COMBUSTION ENGINE FUNDAMENTALS The relative volume v, is defined by RT (4.37) h of C, H2, N2, O2 zero at 0ºR keep h, > 0 Enthalpy datum h = 0 at OK The units are selected so that v, is in cubic meters per kilogram when T is in hy = 0 at 0 K Arbitrary, to kelvins and p, is in pascals. Along a given isentropic, the ratio of actual volumes 1/2 and 11 (for a fixed mass) at temperatures 72 and T1, from Eq. (4.37), is equal to the ratio of relative volumes 0.01-30 atm 1-800 atm (4.38) P range s = const Low Low Low This affords a means of determining T2 for an isentropic process, given Ti and V/2/11 (see Example 4.6). Tables giving the composition and thermodynamic properties of com- 100-6000 K 100-3600 K 200-1500 K 100-2000 K 600-5000ºR bustion products have been compiled. They are useful sources of property and T range species concentrations data in burned gas mixtures for a range of equivalence ratios, temperatures, and pressures. Summary information on four generally available sets of tables is given in Table 4.9. The most extensive set of tables of combustion product composition and thermodynamic properties is the AGARD 0.25, 0.5, 1.0 set, Properties of Air and Combustion Products with Kerosene and Hydrogen range 0.25-4 0.2-2 0.2-2 Fuels, by Banes et al.12 Note, however, that their enthalpy datum differs from the usual datum (enthalpy for O2, N2, H2, and C is zero at 298.15 K). The elements in their reference state at 298.15 K were assigned arbitrary positive values for U.S ., SI Unit enthalpy to avoid negative enthalpies for the equilibrium burned gas mixture. U.S. SI SI Example 4.6. In a diesel engine, the air conditions at the start of compression are p1 = 1 atm and T1 == 325 K. At the end of compression p2 = 60 atm. Find the tem- Air-(CH2). (CH2) ,- air (CH2) ,- air Mixture perature T2 and the compression ratio 11/12. H ,- air Air Air Air tables (see App. D), at T1 = 325 K, give P.1 = 97.13 and U21 = 960.6 Use Eq. (4.36), composition C Properties P, P and C P and C Pr2 = P2 = 60 P Tables of properties of air and combustion products P1 to give Pr2 - 5828 Tables then give Keenan, Chao, and Kaye6 T2 = 992 K and 0,2 = 48.92 General Electric11 TABLE 4.9 AGARD12 The compression ratio is given by Reynolds7 Source VI 960.6 = 19.6 V 2 48.92 129 130 from Eq. (4.14), is then functional form:13-15 FOR PROPERTY AND heat cp, is approximated by Si 4.7.1 Unburned Mixtures 4.7 COMPUTER ROUTINES RT COMPOSITION CALCULATIONS = ail+ INTERNAL COMBUSTION ENGINE FUNDAMENTALS calculate unburned and burned mixture properties. vary considerably in range of application and accuracy. The standard state enthalpy of species i is then given by = ail In T + @2 T + 2 2 H2O, O2, N2, H2, and CO, as a function of temperature. @ 2 T + 3 Cp.j = Ag1 + Arzt + Ag3t2 + Agat3 + Ars ships which model the composition and/or thermodynamic properties of required, engine process calculations are carried out on a computer. Relation- unburned and burned gas mixtures have been developed for computer use. These When large numbers of computations are being made or high accuracy is unburned mixture is frozen in composition and (2) the burned mixture is in equi- modynamic data for each species in the mixture and the assumptions that (1) the summarized here because it is consistent with the approach used throughout to librium. The approach used as the basis for representing JANAF table thermody- namic data8 in the NASA equilibrium program9.10 (see Sec. 3.7) will be Cpi = a11 + 02 T + a13 T2 + ai4 T3 + 025 T4 The standard state entropy of species i at temperature T(K) and pressure 1 atm, H from the NASA program are given in Table 4.10. Two temperature ranges are Values of the coefficients al for CO2, H2O, CO, H2, O2, N2, OH, NO, O, and perty calculations. Figure 4-11 gives values of c./R for the major species, CO2, calculations. The 1000 to 5000 K range is appropriate for burned mixture pro- given. The 300 to 1000 K range is appropriate for unburned mixture property Polynomial functions for various fuels (in the vapor phase) have been fitted to the @: 3 72 + 4 The most complete models are based on polynomial curve fits to the ther- For each species i in its standard state at temperature T(K), the specific 4 @:4 73 + 5 + 9:4 53 + 4 @:5 74 + + @15 T4 + aiq . T 016 (4.39) (4.40) (4.41) (4.42) TABLE 4.10 Coefficients for species thermodynamic properties Species T range, K @13 a15 @16 of CO. 1000-5000 0.44608( + 1) 0.30982(-2) -0.12393(-5) 0.22741(-9) -0.15526(-13) -0.48961( +5) -0.98636(0) 300-1000 0.24008( + 1) 0.87351(-2) -0.66071(-5) 0.20022(-8) 0.63274(-15) -0.48378( + 5) 0.96951(+1) H2( 1000-5000 0.27168(+1) 0.29451(-2) -0.80224(-6) 0.10227(-9) -0.48472(-14) -0.29906( + 5) 0.66306( + 1) 300-1000 0.40701(+ 1) 0.11084(-2) 0.41521(-5) -0.29637(-8) 0.80702(-12) -0.30280( + 5) 0.32270(0) CO 1000-5000 0.29841(+1) 0.14891(-2) -0.57900(-6) 0.10365(-9) -0.69354(-14) -0.14245( +5) 0.63479( + 1) 300-1000 0.37101(+1) -0.16191(-2) 0.36924(-5) -0.20320(-8) 0.23953(-12) -0.14356( + 5) 0.29555( + 1) H 2 1000-5000 0.31002( + 1) 0.51119( -3) 0.52644(-7) -0.34910(-10) 0.36945(-14) -0.87738(+3) -0.19629( + 1) 300-1000 0.30574( +1) 0.26765(-2) -0.58099(-5) 0.55210(-8) -0.18123(-11) -0.98890( +3) -0.22997( +1) O2 1000-5000 0.36220(+1) 0.73618( -3) -0.19652( -6) 0.36202(-10) -- 0.28946(-14) -0.12020( +4) 0.36151( + 1) 300-1000 0.36256( + 1) -0.18782(-2) 0.70555( -5) -0.67635( -8) 0.21556(-11) -0.10475(+4) 0.43053( + 1) 1000-5000 0.28963(+1) 0.15155(-2) -0.57235(-6) 0.99807(-10) -0.65224(-14) -0.90586( +3) 0.61615( + 1) 300-1000 0.36748( +1) -0.12082(-2) 0.23240(-5) -0.63218(-9) -0.22577(-12) -0.10612(+4) 0.23580( + 1) OH 1000-5000 0.29106(+ 1) 0.95932(-3) -0.19442(-6) 0.13757(-10) 0.14225( - 15) 0.39354( +4) 0.54423(+1) NO 1000-5000 0.31890(+ 1) 0.13382(-2) -0.52899(-6) 0.95919(-10) -0.64848(- 14) 0.98283(+4) 0.67458( + 1) O 1000-5000 0.25421( + 1) -0.27551(-4) -0.31028( -8) 0.45511(-11) -0.43681(- 15) 0.29231( +5) 0.49203( + 1) H 1000-5000 0.25(+ 1) 0.0 0.0 0.0 0.0 0.25472(+ 5) -0.46012(0) Source: NASA Equilibrium Code. 131 132 7- 6 3L 4 300 with t = T(K)/1000. 1000 INTERNAL COMBUSTION ENGINE FUNDAMENTALS N2 Temperature, K 2000 t2 CO2 02 H2 H20 Table 4.11. The units for cp,s are cal/gmol . K, and for h, are kcal/gmol. CO t3 where t = T(K)/1000. As6 is the constant for the datum of zero enthalpy for C, hydrocarbon compounds, the coefficients Af; were found by fitting Eqs. (4.42) and (4.43) to data from Rossini et al.16 Values for relevant pure fuels are given in H2, O2, and N2 at 298.15 K. For a 0 K datum, Afg is added to Ar6. For pure paraffins (including cycloparaffins). The fuel was then modeled as composed of a analysis of the fuel was performed to obtain the H/C ratio, average molecular weight, heating value, and the weight percent of aromatics, olefins, and total representative aromatic, olefin, and paraffin hydrocarbon. From atomic conser- molar fractions and average carbon numbers can be determined. Table 4.11 gives vation of hydrogen and carbon and the chemical analysis results, component of the coefficients give cp, and hy in cal/gmol . K and kcal/gmol, respectively, values for the coefficients Af1 to Afs for typical petroleum-based fuels. The units Multicomponent fuel coefficients were determined as follows.14 Chemical 3000 - + Af4 4 FIGURE 4-11 t 14 A15 + A,6 + A78 CO. (From JAN AF tables.8) Specific heat at constant pressure, c /R, as function of temperature for species CO2, H2O, 02, N2, H2, and (4.43) TABLE 4.11 Coefficients for polynomials [Eqs. (4.42) and (4.43)| for fuel enthalpy and specific heat Molecular Fuel Formula weight (A/F), (F/A), Ag4 Ags Ago A 58 Methane CH 16.04 17.23 0.0580 -0.29149 26.327 - 10.610 1.5656 0.16573 - 18.331 4.3000 Propane C3H. 44.10 15.67 0.0638 - 1.4867 74.339 -39.065 8.0543 0.01219 -27.313 8.852 Hexane C6H14 86.18 15.2 0.0656 -20.777 210.48 -164.125 52.832 0.56635 -39.836 15.611 Isooctane C&H18 114.2 15.14 0.0661 -0.55313 181.62 -97.787 20.402 -0.03095 -60.751 20.232 Methanol CH3OH 32.04 6.47 0.1546 -2.7059 44.168 -27.501 7.2193 0.20299 -48.288 5.3375 Ethanol C2HsOH 46.0 9.00 0.1111 6.990 39.741 -11.926 0 -60.214 7.6135 Gasoline C8.26H15.5 114.8 14.64 0.0683 -24.078 256.63 -201.68 64.750 0.5808 -27.562 17.792 C7.76H13.1 106.4 14.37 0.0696 -22.501 227.99 -177.26 56.048 0.4845 -17.578 15.235 Diesel C10.8H 18.7 148.6 14.4 0.0694 -9.1063 246.97 -143.74 32.329 0.0518 -50.128 23.514 Units of A ,, such that h, is in kcal/gmol and 2, , is in cal/gmol . K with t = T(K)/1000. A,6 gives enthalpy datum at 298.15 K; (A,6 + 47g) gives enthalpy datum at 0 K. 133 134 INTERNAL COMBUSTION ENGINE FUNDAMENTALS PROPERTIES OF WORKING FLUIDS 135 1.4 The thermodynamic properties of the unburned mixture can now be obtained. With the moles of each species per mole O2, n;, determined from Table 4.4, and the mass of mixture per mole O2, Mpp, determined from Table 4.5, the -0 unburned mixture properties are given by -0.4 1.3- C.p.u = = [ ncp.i (4.44a) map (4.44b) .2 Cp . . KJ/kg . K Os Sy = (4.44c) where p is in atmospheres. Figures 4-12 and 4-13, obtained with the above relations, show how cp .. and yu( = Cp,w/Cp.) vary with temperature, equivalence ratio, and burned gas frac- P = 0 tion, for a gasoline-air mixture. FIGURE 4-12 Specific heat at constant pressure of unburned gasoline, air, burned gas mixtures as function of temperature, 4.7.2 Burned Mixtures 1.0 equivalence ratio, and burned gas 400 500 800 T, K 1000 fraction. Units: KJ/kg mixture . K. The most accurate approach for burned mixture property and composition cal- culations is to use a thermodynamic equilibrium program at temperatures above about 1700 K and a frozen composition below 1700 K. The properties of each species at high and low temperatures are given by polynomial functions such as Eqs. (4.39) to (4.41) and their coefficients in Table 4.10. The NASA equilibrium program (see Sec. 3.7) is readily available for this purpose and is well docu- mented.9, 1º The following are examples of its output. o = 0 Figure 3-10 showed species concentration data for burned gases as a func- 1.35 tion of equivalence ratio at 1750, 2250, and 2750 K, at 30 atm. Figure 4-14 shows ===== = =------ the burned gas molecular weight My, and Figs. 4-15 and 4-16 give c ,, and y, as functions of equivalence ratio at 1750, 2250, and 2750 K, at 30 atm. Figures 4-17 and 4-18 show c ,., and y, as a function of temperature and pressure for selected equivalence ratios for mixtures lean and rich of stoichiometric.17 For rich mix- 0.8 tures (> > 1), for T > 2000 K, c ,., and 1 are equilibrium values. For 1200 ~ 1.3 K ST < 2000 K, "frozen" composition data are shown where the gas composi- tion is in equilibrium at the given T and p but is frozen as c, and c, are com- puted. Below about 1500 K, fixed composition data are shown corresponding to a value of 3.5 for the water-gas equilibrium constant which adequately describes exhaust gases (see Sec. 4.9). 1.25 ·0.2 Because the computational time involved in repeated use of a full equi- - 0.4 FIGURE 4-13 librium program can be substantial, simpler equilibrium programs and approx- Ratio of specific heats, Y. = c ,. w/c .... imate fits to the equilibrium thermodynamic data have been developed. The of unburned gasoline, air, burned approach usually used is to estimate the composition and/or properties of undis- gas mixtures as function of tem- perature, equivalence ratio, and sociated combustion products and then to use iterative procedures or corrections 1.2 800 T . K 1000 burned gas fraction. lo account for the effects of dissociation. 400 600 136 INTERNAL COMBUSTION ENGINE FUNDAMENTALS PROPERTIES OF WORKING FLUIDS 137 1.30 29 T = 1750 K 1.28 2250 K 2750 K 1.26 1750 K 28- 1.24 2250 K Molecular weight My 1.22 1.20- 27 2750 K 1.18 30 atm 30 atm FIGURE 4-16 T FIGURE 4-14 1.16L Ratio of specific heats, y, = C,b/cy.b, for Molecular weight of equilibrium equilibrium burned gases as a function burned gases as a function of equiv- of equivalence ratio at T = 1750, 2250, 0.6 0.8 1.0 1.2 1.4 alence ratio at T = 1750, 2250, and 0.2 0.4 0.6 0.8 1.0 1.2 1.4 and 2750 K, and 30 atm. Fuel: iso- Fuel/air equivalence ratio 2750 K, and 30 atm. Fuel: isooctane. Fuel/air equivalence ratio octane. --- A computer program for calculating properties of equilibrium combustion products, designed specifically for use in internal combustion engine applications, has been developed by Olikara and Borman and is readily available.18 The fuel 2.5 30 atm composition (C,HO,N}), fuel/air equivalence ratio, and product pressure and temperature are specified. The species included in the product mixture are: CO2, H2O, CO, H2, O2, N2, Ar, NO, OH, O, H, and N. The element balance equa- tions and equilibrium constants for seven nonredundant reactions provide the set of 11 equations required for solution of these species concentrations (see Sec. 3.7). The equilibrium constants are curve fitted from data in the JANAF tables.8 The .. 0 initial estimate of mole fractions to start the iteration procedure is the non- 2750 K dissociated composition. Once the mixture composition is determined, the ther- modynamic properties and their derivatives with respect to temperature, Cp, b, KJ/kg . K pressure, and equivalence ratio are computed. This limited set of species has been found to be sufficiently accurate for engine burned gas calculations, and is much 2250 K more rapid than the extensive NASA equilibrium program.9, 10 1.5 Several techniques for estimating the thermodynamic properties of high- temperature burned gases for engine applications have been developed. One com- 1750 K monly used approach is that developed by Krieger and Borman.19 The internal FIGURE 4-15 energy and gas constant of undissociated combustion products were first Specific heat at constant pressure of described by polynomials in gas temperature. The second step was to limit the equilibrium burned gases as a function range of T and p to values found in internal combustion engines. Then the devi- 1.0L of equivalence ratio at T = 1750, 2250, 0.2 0.4 0.6 0.8 1.0 1.2 1.4 and 2750 K, and 30 atm. Fuel: iso- ations between the equilibrium thermodynamic property data published by Fuel/air equivalence ratio octane. Units: KJ/kg mixture . K. Newhall and Starkman4.5 and the calculated nondissociated values were fitted 138 INTERNAL COMBUSTION ENGINE FUNDAMENTALS PROPERTIES OF WORKING FLUIDS 139 5000 1.4 p, atm ¢ = 1.01 === 4000 100 = 0.0 0.5 OO ------ 0.0 = 1.3- 0.5 100 10 3000 Cp. b. J/kg . K 2000- ₲ = 1.0 0.5~ p. atm 0.0 1.2- $ = 0.01 10 100 0.5 1000 ioo 1.0 100 0.0F .IF 500 1000 1500 2000 2500 3000 3500 Temperature, K 500 1000 1500 2000 2500 3000 3500 (a) Temperature, K (a) 5000 1.4 p, atm ¢ = 2.0 1.2 4000 = 1.2 ..-.-.-.--- io Frozen 100 1.2 1.5 100 1.3 $ = 2.0 3000 2.0 1.5 Fixed composition 1.2 Frozen Yb Cp. b. J/kg . K 2000 $ = 2.0 1.2- p. atm ₲ = 1.2 io 1000 100 2.0. Fixed composition Equilibrium 1.5 10 100 1.1- 2.0 10 100 Equilibrium 1.2. 500 1000 1500 2000 2500 3000 3500 Temperature, K 500 1000 1500 2000 2.500 3000 3500 (b) Temperature, K (b) FIGURE 4-17 Specific heat at constant pressure for equilibrium, frozen, and fixed composition burned gases as a FIGURE 4-18 function of temperature and pressure: (a) equivalence ratio ¢ < 1.0; (b) equivalence ratio ( > 1. Ratio of specific heats, y, = c ,. b/c ..,, for equilibrium, frozen, and fixed composition burned gases as a Units: J/kg mixture . K. Fuel: C,H2. . function of temperature and pressure: (a) equivalence ratio ¢ < 1.0; (b) equivalence ratio ¢ > 1. Fuel: C.H 2 .. by an exponential function of T, p, and o. For o < 1, a single set of equations resulted. For $ 2 1, sets of equations were developed, each set applying to a tions, the undissociated equations for thermodynamic properties are sufficiently accurate. specific value of equivalence ratio (see Ref. 19). In general, the fit for internal energy is within 22 percent over the pressure and temperature range of interest, An alternative approach for property calculations, applicable to a wide and the error over most of the range is less than 1 percent. For many applica- range of hydrocarbon and alcohol fuels, is used extensively in the author's labor- atory.20 With this method, the products of combustion of hydrocarbon (or 140 INTERNAL COMBUSTION ENGINE FUNDAMENTALS PROPERTIES OF WORKING FLUIDS 141 alcohol)-air mixtures are divided into triatomic, diatomic, and monatomic mol- 4.8 TRANSPORT PROPERTIES ecules, M3, M2, and M1, respectively. Then, if Y is the extra number of moles of diatomic molecules due to dissociation of triatomic molecules and U is the extra The processes by which mass, momentum, and energy are transferred from one number of monatomic molecules due to dissociation of diatomic molecules, the point in a system to another are called rate processes. In internal combustion engines, examples of such processes are evaporation of liquid fuel, fuel-air mixing, combustion reaction can be written as friction at a gas/solid interface, and heat transfer between gas and the walls of the :@C + 2(1 - 8)0H2 + 02 + VN2 - engine combustion chamber. In engines, most of these processes are turbulent [(2 - 8)4 - 2YJM3 + [1 - 0 + 3Y - U + V]M2 + 2UM1 and are therefore strongly influenced by the properties of the fluid flow. However, turbulent rate processes are usually characterized by correlations between dimen- for @ < 1 (4.45) sionless numbers (e.g ., Reynolds, Prandtl, Nusselt numbers, etc.), which contain [(2 - 80)0 - 2Y]M3 + [2(0 -1) + 3Y - U + V]M2 + 2UM, the fluid's transport properties of viscosity, thermal conductivity, and diffusion for ¢ > 1 coefficient as well as the flow properties. The simplest approach for computing the transport properties is based on The method is based upon a fitting of data obtained from sets of detailed chemi- the application of kinetic theory to a gas composed of hard-sphere molecules. By cal equilibrium calculations to this functional form. Two general dissociation analyzing the momentum flux in a plane Couette flow,+ it can be shown reactions: (Chapman and Cowling, Ref. 21, p. 218) that the viscosity u of a monatomic 2M3 = 3M2 and M2 = 2M1 hard-sphere gas [where u = t/(du/dx), t being the shear stress and (du/dx) the velocity gradient] is given by are then used with fitted equilibrium constants K,(T) and K2(T) to calculate the relative species concentrations. This approach has been developed to give equa- U = [5 / ( 16 2 ) ] (mkT ) 1/2 tions for enthalpy which sum the translational, rotational, and vibrational contri- d2 (4.48) butions to the specific heat, and the enthalpy of formation: where m is the mass of the gas molecule, d is the molecular diameter, and & is maph = " 2 (8N3 + 7N2 + 5N1)T + R(3N3 + N2) exp (T./T) - 1+ + maphs Boltzmann's constant, 1.381 x 10-23 J/K. For such a gas, the viscosity varies as T1/2, but will not vary with gas (4.46) pressure or density. Measurements of viscosity show it does only vary with tem- where N3, N2, and N1 are the number of moles of triatomic, diatomic, and perature, but generally not proportionally to T1/2. The measured temperature monatomic molecules respectively per mole O2 reactant, T ,, is a fitted vibrational dependence can only be explained with more sophisticated models for the inter- temperature, map is the mass of products per mole O2 reactant [Eq. (4.9)], and hy molecular potential energy than that of a hard sphere. Effectively, at higher tem- is the average specific enthalpy of formation of the products. peratures, the higher average kinetic energy of a pair of colliding molecules The molecular weight is given by requires that they approach closer to each other and experience a greater repul- sive force to be deflected in the collision. As a result, the molecules appear to be MRP - 1 + 1 - E ) $ + + + Y + U for ¢ < 1 smaller spheres as the temperature increases. (4.47) An expression for the thermal conductivity k of a monatomic hard-sphere MRP gas [k = à/(dT/dx), where q is the heat flux per unit area and dT/dx is the tem- or M . = 2 - E ) Q + 4 + Y + U for ¢ > 1 perature gradient] can be derived from an analysis of the thermal equivalent of plane Couette flow (Ref. 21, p. 235): U and Y are found using an approximate solution to the equations obtained by applying the fitted equilibrium constants to the dissociation reactions; hy is k =! [75 / ( 64 T2 ) ] ( 6 3T / m ) 1/2 obtained by fitting a correction to the undisssociated products enthalpy of forma- d2 (4.49) tion. Equations are presented for the partial derivatives of enthalpy h and density p with respect to T, p, and 6.2º These relationships have been tested for fuels with H/C ratios of 4 to 0.707, equivalence ratios 0.4 to 1.4, pressures 1 to 30 atm, and temperatures 1000 to 3000 K. The error for burned mixture temperatures + In Couette flow, the fluid is contained between two infinite plane parallel surfaces, one at rest and relevant to engine calculations is always less than +10 K. The errors in density one moving with constant velocity. In the absence of pressure gradients, the fluid velocity varies linearly across the distance between the surfaces. are less than +0.2 percent. 142 INTERNAL COMBUSTION ENGINE FUNDAMENTALS PROPERTIES OF WORKING FLUIDS 143 which has the same temperature dependence as u. Equations (4.48) and (4.49) can tivity, and Prandtl number in addition to the thermodynamic calculations be combined to give described in Secs. 3.7 and 4.7 for high-temperature equilibrium and frozen gas k = Euc . composition mixtures. The procedures used in the NASA program to compute these transport properties are based on the techniques described in Hirschfelder since, for a monatomic gas, the specific heat at constant volume is 3k/(2m). This et al.22 The NASA program has been used to compute the transport properties of simple equality is in good agreement with measurements of u and k for mon- hydrocarbon-air combustion products.17 These quantities are functions of tem- atomic gases. perature T, equivalence ratio ¢, and (except for viscosity) pressure p. Approx- The above model does not take into account the vibrational and rotational imate correlations were then fitted to the calculated data of viscosity and Prandtl energy exchange in collisions between polyatomic molecules which contribute to number. The principal advantage of these correlations is computational speed. energy transport in gases of interest in engines. Experimental measurements of k For Prandtl number (uc,/k), it was found convenient to use y, the specific heat and u show that k is less than zuc, for such polyatomic gases, where c, is the sum ratio (cp/Cu), as an independent variable. Values of y and c, then permit determi- of the translational specific heat and the specific heat due to internal degrees of nation of the thermal conductivity. freedom. It was suggested by Eucken that transport of vibrational and rotational The viscosity of hydrocarbon-air combustion products over the tem- energy was slower than that of translational energy. He proposed an empirical perature range 500 up to 4000 K, for pressures from 1 up to 100 atm, for ¢ = 0 expression up to o = 4 is shown in Fig. 4-19. The viscosity as a function of temperature of 97 - 5 hydrocarbon-air combustion products differs little from that of air. Therefore, a k == power law based on air viscosity data was used to fit the data: 4 (4.50) 4y Hais(kg/m . s) = 3.3 x 10-7 x To.7 (4.52) or Pr = Hcp k 97 - 5 where T is in kelvins. The viscosity of combustion products is almost indepen- where Pr is the Prandtl number, which is in good agreement with experimental data. 1.5 x 10-4 A similar analysis of a binary diffusion process, where one gas diffuses NASA Eq.(4.53) 0.0 through another, leads to an expression for the binary diffusion coefficient Di. 1.0 DJ is a transport property of the gas mixture composed of species i and j, defined IDDO 1. x 10-4 - - - by Fick's law of molecular diffusion which relates the fluxes of species i and j, I, 4.0 and Ixj, in the x direction to the concentration gradients, dn:/dx and dn,/dx (n is the molecular number density): 7. x 10-5 Viscosity. kg/m . s 5. x 10-5 The binary diffusion coefficient for a mixture of hard-sphere molecules is (Ref. 21, p. 245) 3 2 kT 1/2 (4.51) 3. x 10-5 Diy - 16nd2 Emy where mij is the reduced mass m; my/(m; + m;). A more rigorous treatment of gas transport properties, based on more real- 2. x 10-5L 500 700 1000 istic intermolecular potential energy models, can be found in Hirschfelder et al ., 22 1500 2000 3000 4000 Temperature, K who also present methods for computing the transport properties of mixtures of FIGURE 4-19 gases. The NASA computer program "Thermodynamic and Transport Proper- Viscosity, kg/m .s, of combustion products as a function of temperature and equivalence ratio. Equa- ties of Complex Chemical Systems"10 computes the viscosity, thermal conduc- tions shown are (4.52) and (4.53). 144 INTERNAL COMBUSTION ENGINE FUNDAMENTALS PROPERTIES OF WORKING FLUIDS 145 dent of pressure. This correlation was corrected to include the effect of the equiv- cosity of a multicomponent gas mixture is alence ratio o on the viscosity of hydrocarbon-air combustion products:. Umixt = > Hair 121 x1/4; + 1.385 [j=1. j+1 x, x(RT/PM, Dy) (4.56) Hprod = 4.53) 1 + 0.027¢ where x; and M; are the mole fraction and molecular weight of the ith species, H!; Figure 4-19 shows that the viscosity predicted using Eqs. (4.52) and (4.53) is very is the viscosity of the ith species, v is the number of species in the mixture, and Dij close to the viscosity values calculated with the NASA program. There is less is the binary diffusion coefficient for species i and j.22 than 4 percent error. The Prandtl number of hydrocarbon-air combustion products has also 4.9 EXHAUST GAS been correlated over the above ranges of temperatures, pressures, and equiva- COMPOSITION lence ratios. Since the expression for Prandtl number of a monatomic hard- While the formulas for the products of combustion used in Sec. 3.4 are useful for sphere molecule gas is a function of y, a second-order polynomial of y was used to curve-fit the calculated Prandtl number data. A good fit to the data for lean determining unburned mixture stoichiometry, they do not correspond closely to the actual burned gas composition. At high temperatures (e.g ., during combustion combustion product mixtures was the following: and the early part of the expansion stroke) the burned gas composition corre- Pr = 0.05 + 4.2(y - 1) - 6.7(y - 1)2. (4.54) sponds closely to the equilibrium composition at the local temperature, pressure, and equivalence ratio. During the expansion process, recombination reactions The values of Pr predicted with Eq. (4.54) are within 5 percent of the equilibrium simplify the burned gas composition. However, late in the expansion stroke and Pr values calculated with the NASA program. For rich mixtures the following during exhaust blowdown, the recombination reactions are unable to maintain equation is a good fit to the equilibrium values of Pr using equilibrium values of the gases in chemical equilibrium and, in the exhaust process, the composition y, for temperatures greater than 2000 K: becomes frozen. In addition, not all the fuel which enters the engine is fully burned inside the cylinder; the combustion inefficiency even when excess air is Pr = - 0.05 + 4.2(y - 1) - 6.7(y - 1)2 present is a few percent (see Fig. 3-9). Also, the contents of each cylinder are not 1 < ¢ 5 4 (4.55) 1 + 0.015 x 10-6((T)2 necessarily uniform in composition, and the amounts of fuel and air fed to each cylinder of a multicylinder engine are not exactly the same. For all these reasons, The predicted values of Pr in this case are also close to the calculated values of the composition of the engine exhaust gases cannot easily be calculated. Pr, with less than 10 percent error. Equation (4.55) is also a reasonable fit to the It is now routine to measure the composition of engine exhaust gases. This frozen valuest of Pr for rich mixtures, using frozen values of y, for the tem- is done to determine engine emissions (e.g ., CO, NO ,, unburned hydrocarbons, perature range 1200 to 2000 K. As there are no data for Pr of rich mixtures at and particulates). It is also done to determine the relative proportions of fuel and low temperatures, we suggest that where a fixed composition for the mixture is air which enter the engine so that its operating equivalence ratio can be com- appropriate (e.g ., during the exhaust process in an internal combustion engine), puted. In this section, typical engine exhaust gas composition will be reviewed, Eq. (4.55) can also be used with fixed composition values of y. and techniques for calculating the equivalence ratio from exhaust gas composi- The Prandtl number can be obtained from the above relations if y is tion will be given. known. The thermal conductivity can be obtained from the Prandtl number if values of u and c, are known. Values of y and co ., as functions of temperature, 4.9.1 Species Concentration Data pressure, and equivalence ratio are given in Figs. 4-15 to 4-18. Since the fundamental relations for viscosity and thermal conductivity are Standard instrumentation for measuring the concentrations of the major exhaust complicated, various approximate methods have been proposed for evaluating gas species has been developed.23 Normally a small fraction of the engine exhaust these transport properties for gas mixtures. A good approximation for the vis- gas stream is drawn off into a sample line. Part of this sample is fed directly to the instrument used for unburned hydrocarbon analysis, a flame ionization detec- tor (FID). The hydrocarbons present in the exhaust gas sample are burned in a small hydrogen-air flame, producing ions in an amount proportional to the number of carbon atoms burned. The FID is effectively a carbon atom counter. It In the NASA program, "frozen " means the gas composition is in equilibrium at the given T and p. is calibrated with sample gases containing known amounts of hydrocarbons. but is frozen as c ,, c ., and k are computed. Unburned hydrocarbon concentrations are normally expressed as a mole fraction 146 INTERNAL COMBUSTION ENGINE FUNDAMENTALS PROPERTIES OF WORKING FLUIDS 147 or volume fraction in parts per million (ppm) as C1. Sometimes results are 0.16 expressed as ppm propane (C3H8) or ppm hexane (C6H14); to convert these to D & L ppm Ci multiply by 3 or 6, respectively. Older measurements of unburned hydro- St, Sp 0.14 x +0 H & S carbons were often made with a nondispersive infrared (NDIR) analyzer, where MIT 4 the infrared absorption by the hydrocarbons in a sample cell was used to deter- 0.12 mine their concentration.23 Values of HC concentrations in engine exhaust gases A wk & Co2 measured by an FID are about two times the equivalent values measured by an DD NDIR analyzer (on the same carbon number basis, e.g ., C1). NDIR-obtained 0.10 concentrations are usually multiplied by 2 to obtain an estimate of actual HC concentrations. Substantial concentrations of oxygen in the exhaust gas affect the Mole fractions CO. CO2, O2. H? ).08 FID measurements. Analysis of unburned fuel-air mixtures should be done with special care.23 To prevent condensation of hydrocarbons in the sample line 0.06 (especially important in diesel exhaust gas), the sample line is often heated. NDIR analyzers are used for CO2 and CO concentration measurements. 80090 .04 Infrared absorption in a sample cell containing exhaust gas is compared to absorption in a reference cell. The detector contains the gas being measured in 0.02 two compartments separated by a diaphragm. Radiation not absorbed in the H/C = 2-2.25 sample cell is absorbed by the gas in the detector on one side of the diaphragm. 0.00 L ... Radiation not absorbed in the reference cell is absorbed by the gas in the other 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.4 1.5 half of the detector. Different amounts of absorption in the two halves of the Exhaust equivalence ratio detector result in a pressure difference being built up which is measured in terms FIGURE 4-20 of diaphragm distention. NDIR detectors are calibrated with sample gases of Spark-ignition engine exhaust gas composition data in mole fractions as a function of fuel/air equiva- known composition. Since water vapor IR absorption overlaps CO2 and CO lence ratio. Fuels: gasoline and isooctane, H/C 2 to 2.25. (From D'Alleva and Lovell,24 Stivender,25 absorption bands, the exhaust gas sample is dried with an ice bath and chemical Harrington and Shishu,26 Spindt,27 and data from the author's laboratory at MIT.) dryer before it enters the NDIR instrument. Oxygen concentrations are usually measured with paramagnetic analyzers. different on the lean and the rich side of the stoichiometric air/fuel or fuel/air Oxides of nitrogen, either the amount of nitric oxide (NO) or total oxides of ratios; thus, the fuel/air equivalence ratio o (or its inverse, the relative air/fuel nitrogen (NO + NO2, NO,), are measured with a chemiluminescent analyzer. ratio 1) is the appropriate correlating parameter. On the lean side of stoichiomet- The NO in the exhaust gas sample stream is reacted with ozone in a flow reactor. ric, as o decreases, CO2 concentrations fall, oxygen concentrations increase, and The reaction produces electronically excited NO2 molecules which emit radiation CO levels are low but not zero (~0.2 percent). On the rich side of stoichiometric, as they decay to the ground state. The amount of radiation is measured with a CO and H2 concentrations rise steadily as o increases and CO2 concentrations photomultiplier and is proportional to the amount of NO. The instrument can fall. O2 levels are low (~0.2 to 0.3 percent) but are not zero. At stoichiometric also convert any NO2 in the sample stream to NO by decomposition in a heated operation, there is typically half a percent O2 and three-quarters of a percent CO. stainless steel tube so that the total NO, (NO + NO2) concentration can be Fuel composition has only a modest effect on the magnitude of the species determined.23 Gas chromatography can be used to determine all the inorganic concentrations shown. Measurements with a wide range of liquid fuels show that species (N2, CO2, O2, CO, H2) or can be used to measure the individual hydro- CO concentrations depend only on the equivalence ratio or relative fuel/air ratio carbon compounds in the total unburned hydrocarbon mixture. Particulate emis- (see Fig. 11-20).26 A comparison of exhaust CO concentrations with gasoline, sions are measured by filtering the particles from the exhaust gas stream onto a propane (C3H8), and natural gas (predominantly methane, CH4) show that only previously weighed filter, drying the filter plus particulate, and reweighing. with the high H/C ratio of methane, and then only for CO 2 4 percent, is fuel composition significant.28 The values of CO2 concentration at a given o are SPARK-IGNITION ENGINE DATA. Dry exhaust gas composition data, as a func- slightly affected by the fuel H/C ratio. For example, for stoichiometric mixtures tion of the fuel/air equivalence ratio, for several different multi- and single- with 0.5 percent O2 and 0.75 percent CO, as the H/C ratio decreases CO2 cylinder automotive spark-ignition engines over a range of engine speeds and concentrations increase from 13.7 percent for isooctane (H/C = 2.25), to 14.2 loads are shown in Fig. 4-20. The fuel compositions (gasolines and isooctane) had to 14.5 percent for typical gasolines (H/C in range 2-1.8), to 16 for toluene H/C ratios ranging from 2.0 to 2.25. Exhaust gas composition is substantially (H/C = 1.14).29 148 INTERNAL COMBUSTION ENGINE FUNDAMENTALS PROPERTIES OF WORKING FLUIDS 149 12 0.16 Hydrocarbon fuels: 0.14- 10 approx. 85% C 0g* a 0.12 8 0.10- 0.08 Mole tractions CO. CO2. * Hydrogen, % by vol. 6 0.06 A to CO2 * 0 0.02 2 CO 0.00 0.0 0.1 exgol 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 10 12 14 16 Exhaust equivalence ratio Carbon monoxide, % by vol. FIGURE 4-22 FIGURE 4-21 Exhaust gas composition from several diesel engines in mole fractions on a dry basis as a function of Hydrogen concentration in spark-ignition engine exhaust as a function of carbon monoxide concen- fuel/air equivalence ratio.31 tration. Units: percent by volume.30 Unburned hydrocarbon exhaust concentrations vary substantially with the composition of fuel can be represented as C_H_O ,. For conventional engine design and operating conditions. Spark-ignition engine exhaust levels in a petroleum-based fuels, oxygen will be absent; for fuels containing alcohols, modern low-emission engine are typically of the order of 2000 ppm C, with oxygen will be present. The overall combustion reaction can be written as liquid hydrocarbon fuels, and about half that level with natural gas and propane Fuel + oxidizer -> products fuels. Hydrogen concentrations in engine exhaust are not routinely measured. The fuel is C,H„O,; the oxidizer is air (O2 + 3.773N2). The products are CO2, However, when the mixture is oxygen-deficient-fuel rich-hydrogen is present H2O, CO, H2, O2, NO1, N2, unburned hydrocarbons (unburned fuel and pro- with CO as an incomplete combustion product. Figure 4-21 summarizes much of ducts of partial fuel reaction), and soot particles (which are mainly solid carbon). the available data on H2 concentrations plotted as a function of CO.3º The amount of solid carbon present is usually sufficiently small (<0.5 percent of the fuel mass) for it to be omitted from the analysis. The overall combustion DIESEL EXHAUST DATA. Since diesels normally operate significantly lean of reaction can be written explicitly as stoichiometric (¢ < 0.8) and the diesel combustion process is essentially complete (combustion inefficiency is <2 percent), their exhaust gas composition is straight- C,HO, + -02 (02 + 3.773N2) = P(Xc.H. C.H, + XcoCO + xco2 CO2 forward. Figure 4-22 shows that O2 and CO2 concentrations vary linearly with the fuel/air equivalence ratio over the normal operating range. Diesel emissions + 0202 + XN2N2 + XNoNO of CO and unburned HC are low. + XNO,NO2 + XH20H20 + XH2H2) (4.57) 4.9.2 Equivalence Ratio Determination where o is the measured equivalence ratio [(F/A)actual/(F/A)stoichiometric], no2 is the number of O, molecules required for complete combustion (n + m/4 - r/2), np is from Exhaust Gas Constituents the total number of moles of exhaust products, and x; is the mole fraction of the Exhaust gas composition depends on the relative proportions of fuel and air fed ith component. to the engine, fuel composition, and completeness of combustion. These relation- There are several methods for using Eq. (4.57) to determine o, the equiva- ships can be used to determine the operating fuel/air equivalence ratio of an lence ratio, depending on the amount of information available. Normally CO2, engine from a knowledge of its exhaust gas composition. A general formula for CO, O2, NO, concentrations as mole fractions and unburned hydrocarbon (as 150 INTERNAL COMBUSTION ENGINE FUNDAMENTALS PROPERTIES OF WORKING FLUIDS 151 mole fraction or ppm C1, i.e ., xcH) are measured. The concentration of the inor- OXYGEN BALANCE AIR/FUEL AND EQUIVALENCE RATIOS. For fuels com- ganic gases are usually measured dry (i.e ., with H2O removed) or partially dry. prised of carbon and hydrogen only, when all species are measured with the same Unburned hydrocarbons may be measured wet or dry or partially dry. NO, is background moisture (wet, dry, or partially dry), the following expression based on mainly nitric oxide (NO); its concentration is usually sufficiently low (<0.5 the ratio of measured and computed oxygen-containing species to measured percent) for its effect on equivalence ratio determination to be negligible small. carbon-containing species gives the air/fuel ratio. It has been assumed that the Thus, in Eq. (4.57) there are seven unknowns which are: (, xH ,, xH120, XN2, np, a, unburned hydrocarbons have the same C/H ratio as the fuel:32 b. (There will be additional unknowns if the measurements listed above are A Maix (CO2) + (CO)/2 + (H20)/2 + (NO)/2 + (NO2) + (02) incomplete.) = 4.773 To solve for these unknowns we need seven additional equations. We can (HC) + (CO) + (CO2) obtain five equations using an atomic balance for each element and the definition (4.64) of mole fraction, as follows: where ( ) are molar concentrations (all with the same background moisture) in Carbon balance: percent, Mair = 28.96, M, = 12.01 + 1.008y where y is the H/C ratio of the fuel, (4.58) (HC) is molar percent unburned hydrocarbons as C1, and n = np(axc.Ho + &co + xco2) (H20) = 0.5y 7 (CO2) + (CO) Hydrogen balance: (CO)/[K(CO2)] + 1 (4.65) m = np(bxcaHe + 2XH20 + 2XH2) (4.59) Since nitrogen oxides collectively comprise less than 0.5 percent of the exhaust Oxygen balance: mixture, their concentrations can be omitted with negligible error. The fuel/air equivalence ratio o is obtained from the ratio of the stoichio- r + 2102 - nofxco + 2xco2 + XNo + 2x02 + XH;O) (4.60) metric air/fuel ratio [Eq. (3.6)] and Eq. (4.64) above. Nitrogen balance: CARBON BALANCE AIR/FUEL AND EQUIVALENCE RATIOS. When oxygen analysis is not available, for fuels comprised of carbon and hydrogen only, a 7.546no2 = np(2%N2 + XNO) (4.61) carbon balance air/fuel ratio may be employed:25. 32 A Mir [ 100 + (HC) - (CO)/2 + 3(H2O)/2 - (H2O)a _ y Mole fractions add up to 1: FM, (HC) + (CO) + (CO2) (4.66) &c.Ho + &co + XH2 + XH20 + XN2 + XNo + Xco2 + X02 = 1 (4.62) The symbols are as defined above. (H2O) is the molar percent water in the com- bustion products defined by Eq. (4.65) and (H2O) is the molar percent water An additional assumption is made, based on available exhaust gas composition vapor at the analyzers. data, that CO2, CO, H2O, and H2 concentrations are related by This carbon balance (A/F) is sensitive to moisture concentration at the analyzers. The use of ice bath exhaust sample chillers generally reduces the X CO XH20 = K (4.63) (H2O) term to less than 1 percent and little accuracy is then lost by neglecting it. xc02 XH2 For completely "wet" analysis (uncondensed), (H2O). = (H2O), and Eq. (4.66) is where K is a constant.+ Values of 3.824, 25 and 3.527 are commonly used for K. accurate. For partially dry exhaust gas analysis, knowledge of the dew point of The difference between these values has little effect on the computed magnitude the mixture will provide the (H2O), term by reference to steam tables. of . To complete the analysis, various assumptions are made concerning the The fuel/air equivalence ratio o is obtained from the ratio of the stoichio- composition and relative importance of the unburned hydrocarbons. The most metric air/fuel ratio [Eq. (3.6)] and Eq. (4.66) above. common approaches are summarized below. EQUIVALENCE RATIO BASED ON WET HC AND DRY INORGANIC GAS ANALYSIS. Engine exhaust gas composition is often determined by analyzing a fully dried sample stream for CO2, CO, O2, and NO ,, and a fully wet Equation (4.63) is often described as assuming a specific value for the water-gas reaction equilibrium (uncondensed) stream with an FID for unburned hydrocarbons. Equations (4.64) constant. In fact K is an empirical constant determined from exhaust gas composition data. and (4.66) are not applicable under these circumstances. The following equations 152 INTERNAL COMBUSTION ENGINE FUNDAMENTALS PROPERTIES OF WORKING FLUIDS 153 define the exhaust gas composition and equivalence ratio under these conditions. 16 The notation x; denotes the wet mole fraction of species i and * denotes the dry mole fraction of species i. Equations (4.57) to (4.62), with Eq. (4.63) to relate CO2, 14 H20 CO, H2O, and H2 concentrations and the assumption that b/a = m/n, were used 121 to derive these results. The equations apply for a fuel of composition C_H,O ,. The fuel/air equivalence ratio is given by 10 2 002 CO2 Mole fraction (wel), % np XH20 + np(1 - XH20)(xco + 2x202 + 2x02 + xNo + 2xNo,) - r : (4.67) where the wet and dry mole fractions are related by 6 x = (1 - XH20 )xt co NO and H2 n np &CHO/a + (1 - XH20Xxco + xco2) 0.7 0.8 0.9 .0 m xco + xco2 Fuel/air equivalence ratio X H20 = 2n [1 + xco/(Kxco2) + (m/2n)(xco + xco2)] (4.68) FIGURE 4-23 Wet exhaust gas species concentrations as a function of fuel/air equivalence ratio, based on the dry XH20 XEo exhaust gas composition data in Fig. 4-20 and Eqs. (4.68). X H2 Kxco2 3.773no2 _ (1 - XH20) (XNO + XNO2) cycle are exactly equal. In addition, mixing of fuel and air within each cylinder is XN2 onp 2 not necessarily completely uniform. Thus the exhaust gas composition may corre- Note that XcHy is the measured (wet) HC concentration as a mole fraction C, spond to a distribution in the fuel/air ratio in the unburned mixture about the (ppm C1 x 10-6): XCH>) = axc.H, . Figure 4-23 shows wet exhaust gas concentra- mean value. For example, if the mean fuel/air ratio is stoichiometric, extra tions, based on the MIT measured dry concentrations of CO2, CO, O2 shown in oxygen will be contributed by any cylinders operating lean of the average and Fig. 4-20, and wet HC concentration, as well as Eqs. (4.68). extra carbon monoxide by any cylinders rich of the average, so that the exhaust For lean mixtures, varying the value of K between 1.5 and 5.5 had a negligi- gas will have higher levels of O2 and CO (and a lower level of CO2) relative to an ble effect on the value of o computed from Eq. (4.67). For stoichiometric mix- engine operating with identical fuel/air ratios in each cylinder. tures, varying K from 2.5 to 4.5 varied the computed o by 2 to 3 percent. For Eltinge has developed a method for defining this nonuniformity in the fuel/ ~ 1.2, varying K from 2.5 to 4.5 varied the computed o by 3 to 4 percent. The air ratio distribution for spark-ignition engines which operate close to stoichio- error in o involved in omitting NO, is 0.2 percent for an NO, level of 1000 ppm, metric.29 A function f(x) for the fuel/air ratio distribution (x = F/A) was increasing to 1 percent for an NO, level of 5000 ppm. The sensitivity of the assumed, with standard deviation S ,. For each value of x (i.e ., F/A), complete computed o to errors in the measurements of CO2, CO, and O2 is modest within utilization of the available oxygen was assumed (i.e ., no exhaust HC) and the the normal range of o used. A 2 percent error in CO2 or CO or O2 at ( ~ l CO2, H2O, CO, H2 concentrations were related by Eq. (4.63) (with K = 3.5). The gives about a 0.1 percent error in computed o. For leaner and richer mixtures, concentrations of each species for each (F/A) were weighted by the distribution the error in 4 increases for errors in measured CO, concentration, and CO function f(x) and summed to produce the average exhaust concentration. (A cor- ( > 1) and O2 (¢ < 1) concentrations, but is still significantly less than the mea- rection to allow for the presence of unburned HC in the exhaust was also devel- surement error in fuel and air flow. oped.) Figure 4-24 shows one set of results, for a fuel H/C ratio of 1.8 (typical of gasoline), for a normal distribution in the fuel/air ratio. For each mean (F/A) and 4.9.3 Effects of Fuel/ Air Ratio maldistribution parameter S, (the standard deviation of the F/A distribution) the Nonuniformity corresponding dry concentrations of CO2, CO, and O2 are shown. This type of information can be used to define the fuel-air mixture nonuniformities in spark- Neither the masses of air inducted into the different cylinders of a multicylinder ignition engines operating relatively close to stoichiometric. For diesel engines engine per cycle nor the masses of fuel which enter the different cylinders per the variations of major exhaust gas species concentrations with fuel/air equiva- PROPERTIES OF WORKING FLUIDS 155 154 INTERNAL COMBUSTION ENGINE FUNDAMENTALS PROBLEMS Assumptions: 12.0 Fuel H/C atom ratio = 1.80 1.1. (a) Calculate the low-temperature burned gas composition resulting from the com- AIF Water gas constant = 3.5 5 . No unburned fuel bustion of 7 g/s air with 0.48 g/s ethane (C2H6). Assume K = 3.5 in Eq. (4.6). 12.4 Stoichiometric air/fuel ratio = 14.5 (b) Calculate the low-temperature burned gas composition for the combustion of 7 g/s air with 0.48 g/s ethanol (C2H6O). Assume K = 3.5 in Eq. (4.6). (Note the large difference in burned gas composition due to this difference in fuel.) - 12.8 1.2. To evaluate the accuracy of the simple analytic ideal gas model, use the results of Example 4.1 and Eqs. (4.23) (constant-volume adiabatic combustion) and (4.24) 13.2 (constant-pressure adiabatic combustion) to calculate T, for a stoichiometric 4/ isooctane-air mixture. Compare this result with that obtained using Figs. 4-3 and 13.6 CO, % 4-8. Assume the following unburned mixture conditions: T = 700 K, v = 0.125 m3/ 14.0 kg air, p = 15 atm, and x, = 0.1. 3 14.4 4.3. (a) In Fig. 4-1, why does M, decrease as x, increases? 14.8 (b) In Fig. 4-14, why does M, decrease as T increases ? 15.2 FIGURE 4-24 S. = 0.014 (c) In Fig. 4-14, why does M, decrease as ¢ increases? 13.5 %CO Computed relationship between dry 0.012 15.6 exhaust gas composition (CO2, CO, and 4.4. Show how, for an ideal gas with fixed composition, the molecular weight M is 16.0 0.010 O2), air/fuel ratio, and maldistribution related to the specific heats c, and cu, and R. Use Figs. 4-15 and 4-16 to calculate M, 14.0 16.4 0. 16.8 parameter Sx. Fuel: (CH1.8)„. The correc- for T; = 1750, 2250, and 2750 K, and ¢ = 1.0. Compare these results with values of 17.2 17.6 14.5 0.006 18.0 tion to be added to the burned gas M, obtained from Fig. 4-14 and explain any differences. 15.0 (F/A) which allows for the measured un- 4.5. EGR, exhaust gas recirculation, is often used to reduce NO, by acting as a diluent in 10.004 burned hydrocarbon concentration is 0.002 the intake mixture. 4.7 x 10-7 x HC (ppm C1) (From 2 02, % 5 Eltinge.29) (a) For low-temperature isooctane-air combustion products at o = 1.0, determine the percentage of the burned gases' average specific heat at constant pressure which comes from each component in the burned gas mixture. lence ratio are linear, so the effects of any nonuniformities are not apparent in (b) Compare the specific heat at constant pressure of isooctane-air combustion pro- this manner (see Fig. 4-22). ducts at d = 1.0 to that of air, both at 1750 K. This difference is one reason why EGR instead of leaner fuel-air mixture is used to control NO, emissions. 4.6. Explain qualitatively the causes of the trends in the curves in Fig. 4-23 as ¢ is 4.9.4 Combustion Inefficiency increased from 1.0 and decreased from 1.0. Internal combustion engine exhaust contains combustible species: CO, H2, 4.7. Compare the O2, CO2, and CO data in Fig. 4-22 with the predictions of the elemen- unburned hydrocarbons, and particulates. When their concentrations are known, tal balance in Table 4.3. Explain any differences. Assume diesel fuel is a hydrocarbon the combustion efficiency ne given by Eq. (3.27) can be calculated. The chemical with H/C ratio of 2. energy carried out of the engine in these combustibles represents the combustion 4.8. The following exhaust data were obtained from a four-stroke cycle spark-ignition engine. CO, CO2, and NO, molar concentrations are all measured fully dry; HC is inefficiency, 1 - ne: measured fully wet as ppm C1. Determine the fuel/air equivalence ratio o for the 1 - no = F Lix , CHV . (4.69) following three sets of data. Make the following assumptions: (1) the constant K in [img/(ma + ms)]QHv, Eq. (4.63) equals 3.5; (2) the fuel composition is C8H15.12; (3) the unburned hydro- carbon H/C ratio is the same as that of the fuel; and (4) NO, is entirely NO. where the x; are the mass fractions of CO, H2, HC, and particulates, respectively, (a) CO2 14.0%, CO 0.64%, O2 0.7%, NO× 3600 ppm, HC 3200 ppm. the QHv, are the lower heating values of these species, and the subscripts f and a (b) CO2 13.8%, CO 3.05%, O2 0%, NO, 1600 ppm, HC 3450 ppm. denote fuel and air. The heating values for CO (10.1. MJ/kg) and H2(120 MJ/kg) (c) CO2 12.5%, CO 0.16%, O2 4.0%, NO, 4600 ppm, HC 2100 ppm. are well defined. The composition of the unburned HC is not usually known. 4.9. The exhaust from a spark-ignition engine has the following composition (in mole However, the heating values of hydrocarbons are closely comparable, so the fuel fractions): heating value (typically 42 to 44 MJ/kg) is used. The particulates (only present in diesels) are soot with some adsorbed hydrocarbons; usually the mass fraction is CO2 = 0.12; H2O = 0.14; CO = 0.01; H2 = 0.005; low enough for their contribution to be small, and a heating value for solid N2 = 0.7247; CgH18 = 0.0003 carbon (32.8 MJ/kg) can be used. Combustion efficiency data as a function of The fuel is isooctane, C8H18; as shown above, a small fraction of the fuel escapes equivalence ratio have already been presented in Fig. 3-9. 156 INTERNAL COMBUSTION ENGINE FUNDAMENTALS PROPERTIES OF WORKING FLUIDS 157 from the cylinder unburned. The lower heating value of isooctane is 44.4 MJ/kg, of Hot Rankine cycle boiler Cooled carbon monoxide is 10.1 MJ/kg, and of hydrogen is 120 MJ/kg. The atomic weights diesel diesel of the elements are: C = 12, O = 16, H = 1, N = 14. Diesel exhaust exhaust (a) Calculate the combustion inefficiency in the engine; i.e ., the percentage of the engine 400 K~ entering fuel enthalpy which is not fully released in the combustion process and my leaves the engine in the exhaust gases (for this problem, the exhaust can be assumed to be at room temperature). (b) What fraction of this inefficiency is due to the unburned fuel emissions? Rankine ( liquid ) 4.10. A 2-liter displacement four-cylinder engine, operating at 2000 rev/min and 30 cycle fluid percent of its maximum power at that speed, has the following exhaust composition (vapor) FIGURE P4-12 (in percent by volume or mole percent): CO2, 11%; H2O, 11.5%; CO, 0.5%; H2, 0%; O2, 2%; unburned hydrocarbons with an equivalence ratio of 0.8 and an air flow of 0.5 kg/s. The gross indicated fuel (expressed as CH2, i.e ., with a molecular weight of 14), 0.5%; N2, 74.5% conversion efficiency is 45 percent, and the heat losses from the working fluid to the engine coolant and elsewhere within the engine are 280 kW. Diesel fuel has a heating The fuel is (CH2), with a heating value of 44 MJ/kg. The atomic weights of the value of 42 MJ/kg and stoichiometric fuel/air ratio of 0.067. Fuel and air enter the elements are C = 12, H = 1, O = 16, N = 14. The heating values of CO and HC are engine at ambient conditions. The mechanical efficiency of the diesel engine is 85 10 and 44 MJ/kg, respectively. percent. (a) Is the engine a diesel or spark-ignition engine? Is there enough oxygen in the (a) Calculate the rated brake power of the engine, the average sensible enthalpy of exhaust to burn the fuel completely ? Briefly explain. the exhaust gases as they leave the engine, and the average exhaust gas tem- (b) Calculate the fraction of the input fuel energy (my QHy) which exits the engine perature. unburned as (1) CO and (2) unburned HC. (b) Since the exhaust gas temperature is significantly above ambient, the advantages (c) An inventor claims a combustion efficiency of 100 percent can be achieved. What of using the diesel exhaust gas stream to heat the boiler of a Rankine cycle (see percentage improvement in engine specific fuel consumption would result? sketch) and generate additional power are to be explored. If the exhaust gases 4.11. A gasoline engine operates steadily on a mixture of isooctane and air. The air and leave the Rankine-cycle system boiler at 400 K and 30 percent of the heat trans- fuel enter the engine at 25ºC. The fuel consumption is 3.0 g/s. The output of the ferred from the exhaust gas stream in the boiler is converted to power at the engine is 50 kW. The temperature of the combustion products in the exhaust mani- Rankine-cycle power plant drive shaft, calculate the additional power obtained fold is 660 K. At this temperature, an analysis of the combustion products yields the and the brake fuel conversion efficiency of the combined cycle system (diesel plus following values (on a dry volumetric basis): Rankine cycle). CO2, 11.4%; O2, 1.6%; CO, 2.9%; N 2, 84.1% 4.13. A diesel engine has a compression ratio of 22 : 1. The conditions in the cylinder at the start of compression are p = 101.3 kPa and T = 325 K. Calculate the pressure (a) Find the composition in moles (number of moles per mole of isooctane) of the and temperature at the end of compression, assuming the compression process is reactants and the reaction products. isentropic: (b) Determine the heat-transfer rate from the working fluid as the working fluid (a) Assume the cylinder contains an ideal gas with y = 1.4 and R = 287 passes through the engine. J/kg . K. Constants for the calculations: (b) Assume the cylinder contains air which may be regarded as a semiperfect gas (use the gas tables). (c) Compare the work of compression in (a) and (b) above. Enthalpy of formation, Sensible enthalpy at 660 K, In practice, heat losses reduce the final compression temperature by 100 K. For a kJ/kmol kJ/kmol diesel engine operating at an equivalence ratio of 0.75 (full load): ....... (d) Calculate the ratio of heat loss during compression to the fuel energy added per C&H18 -259280 cycle. CO2 - 393522 15823 CO -110529 10789 4.14. While the geometric compression ratio of an engine is V/max/Vmin, the actual compres- H,O -241827 12710 sion process starts somewhere between bottom-center and when the inlet valve O2 11200 closes, and it is conditions at time of spark (for an SI engine) or fuel injection (for a N, 10749 CI engine) that determine ignition. At low engine speed, compression starts about the time when the inlet valve closes. With this assumption, for the diesel engine of Prob. 4.13, calculate the air pressure and temperature at the start of injection. The 4.12. A direct-injection four-stroke cycle diesel engine is used to provide power for inlet valve closes at 30º after BC; injection commences 15º before TC. Use the gas pumping water. The engine operates at its maximum rated power at 2000 rev/min, tables. Compare your answers with those of Prob. 4.13(b). 158 INTERNAL COMBUSTION ENGINE FUNDAMENTALS PROPERTIES OF WORKING FLUIDS 159 4.15. Use an equilibrium computer code (which calculates the composition and properties Internal energy of combustion of butane at 298 K is Au = - 2.659 MJ/gmol. of chemically reacting gas mixtures in equilibrium) to calculate the data you need for Extract from gas tables for products of combustion for 50 percent stoichiometric the following graphs: fuel: (a) Values of c ,, y, molecular weight, and gas composition (mole fractions of N2, CO2, H2O, CO, H2, O2, OH, O, H, and NO) as a function of the equivalence ratio ( = 0.2 to 1.4) for products of combustion of isooctane (CaH18) with air at T, K u, J/gmol p = 40 atm and T = 2500 K. Put all species concentrations on the same graph. Use a log scale for the composition axis. 298 6,293 (b) Unburned mixture consisting of isooctane vapor and air at 700 K and 20 atm is 2075 54,227 burned first at constant pressure and then at constant volume. 2080 54,380 (1) Calculate the enthalpy and internal energy of isooctane vapor at 700 K in cal/gmol; also calculate the volume per unit mass of mixture (cm3/g) for ¢ = 0.2, 0.4, 0.6, 0.8, 1.0, 1.2, 1.4. (2) Use these data and the equilibrium program to calculate the temperature REFERENCES attained after combustion at constant pressure, and temperature and pressure attained after combustion at constant volume. Plot these temperatures and 1. Komiyama, K ., and Heywood, J. B.: "Predicting NO, Emissions and the Effects of Exhaust Gas Recirculation in Spark-Ignition Engines," SAE paper 730475, SAE Trans ., vol. 82, 1973. pressures against the equivalence ratio o. 2. Danieli, G ., Ferguson, C ., Heywood, J ., and Keck, J. "Predicting the Emissions and Performance Characteristics of a Wankel Engine," SAE paper 740186, SAE Trans ., vol. 83, 1974. 3. Hottel, H. C ., Williams, G. C ., and Satterfield, C. N.: Thermodynamic Charts for Combustion Thermodynamic data for isooctane vapor Processes, John Wiley, 1949. See also charts in C. F. Taylor, The Internal Combustion Engine in Theory and Practice, vol. 1, MIT Press, 1960. T, K 2 ,, cal/mol . K h - h298, kcal/mol Ah ,, kcal/mol 4. Newhall, H. K ., and Starkman, E. S.: "Thermodynamic Properties of Octane and Air for Engine Performance Calculations," in Digital Calculations of Engine Cycles, Progress in Technology, vol. 298 45.14 0.00 -53.57 TP-7, pp. 38-48, SAE, 1964. 700 85.66 27.02 -62.79 5. Starkman, E. S ., and Newhall, H. K.: "Thermodynamic Properties of Methane and Air, and Propane and Air for Engine Performance Calculations," SAE paper 670466, SAE Trans ., vol. 76, 1967. 6. Keenan, J. H ., Chao, J ., and Kaye, J.: Gas Tables, 2d ed ., John Wiley, 1983. 4.16. A heavy wall bomb with a volume of 1000 cm3 contains a mixture of isooctane with 7. Reynolds, W. C.: Thermodynamic Properties in SI; Graphs, Tables, and Computational Equations the stoichiometric air requirement at p = 101.3 kPa and T = 25ºC. The mixture is for Forty Substances, Department of Mechanical Engineering, Stanford University, 1979. then ignited with a spark. Find the pressure and temperature of the equilibrium 8. JANAF Thermochemical Tables, 2d ed ., NSRDS-NB537, U.S. National Bureau of Standards, June 1971. combustion products just after combustion is complete (i.e ., before heat losses to the wall are significant). Assume the burned gases are uniform. 9. Gordon, S ., and McBride, B. J.: “Computer Program for the Calculation of Complex Chemical Equilibrium Composition, Rocket Performance, Incident and Reflected Shocks, and Chapman- 4.17. A gas engine, running on a gaseous mixture of butane, C.H10, and air has the Jouguet Detonations," NASA publication SP-273, 1971 (NTIS number N71-37775). following conditions in the cylinder prior to constant-volume adiabatic combustion: 10. Svehla, R. A ., and McBride, B. J.: "Fortran IV Computer Program for Calculation of Thermody- pressure, 6.48 x 105 N/m2; temperature, 600 K. The charge composition is air plus namic and Transport Properties of Complex Chemical Systems," NASA technical note TND- 50 percent of the stoichiometric quantity of butane fuel. Calculate the pressure and 7056, 1973 (NTIS number N73-15954). temperature at the end of combustion using the data given below. 11. Fremont, H. A ., et al.: Properties of Combustion Gases, General Electric Company, Cincinnati, Ohio, 1955. 12. Banes, B ., Mcintyre, R. W ., and Sims, J. A.: Properties of Air and Combustion Products with Kerosene and Hydrogen Fuels, vols. I-XIII, Propulsion and Energetics Panel, Advisory Group for Aerospace Research and Development (AGARD), NATO, published by Bristol Siddeley Engines For air T, K u, J/gmol Ltd ., Filton, Bristol, England, 1967. 13. Hires, S. D ., Ekchian, A ., Heywood, J. B ., Tabaczynski, R. J ., and Wall, J. C.: "Performance and 298 6,161 NO, Emissions Modelling of a Jet Ignition Prechamber Stratified Charge Engine," SAE paper 600 12,596 760161, SAE Trans ., vol. 85, 1976. 14. LoRusso, J. A.: "Combustion and Emissions Characteristics of Methanol, Methanol-Water, and For butane T, K u, J/gmol Gasoline-Methanol Blends in a Spark Ignition Engine," S. M. Thesis, Department of Mechanical Engineering, MIT, May 1976. 298 -77 15. By. A ., Kempinski, B ., and Rife, J. M.: "Knock in Spark-Ignition Engines," SAE paper 810147, 600 38,424 1981. 160 INTERNAL COMBUSTION ENGINE FUNDAMENTALS 16. Rossini, F. D ., Pitzer, K. S ., Arnett, R. L ., Braun, R. M ., and Primentel, G. C.: Selected Values of Physical and Thermodynamic Properties of Hydrocarbons and Related Compounds, Carnegie Press, Pittsburgh, Pa ., 1953. CHAPTER 17. Mansouri, S. H ., and Heywood, J. B.: "Correlation for the Viscosity and Prandtl Number of Hydrocarbon-Air Combustion Products," Combust. Sci. Technology, vol. 23, pp. 251-256, 1980. 18. Olikara, C ., and Borman, G. L.: "A Computer Program for Calculating Properties of Equilibrium 5 Combustion Products with Some Applications to I.C. Engines," SAE paper 750468, 1975. 19. Krieger, R. B ., and Borman, G. L.: "The Computation of Apparent Heat Release for Internal Combustion Engines," in Proc. Diesel Gas Power, ASME paper 66-WA/DGP-4, 1966. 20. Martin, M. K ., and Heywood, J. B.: "Approximate Relationships for the Thermodynamic Proper- ties of Hydrocarbon-Air Combustion Products," Combust. Sci. Technology, vol. 15, pp. 1-10, IDEAL 1977. 21. Chapman, S ., and Cowling, T. G.: The Mathematical Theory of Non-Uniform Gases, Cambridge MODELS OF University Press, Cambridge, 1955. ENGINE 22. Hirschfelder, J. O ., Curtiss, C. F ., and Bird, R. B.: Molecular Theory of Gases and Liquids, John Wiley, 1954. CYCLES 23. Patterson, D. J ., and Henein, N. A.: Emissions from Combustion Engines and Their Control, Ann Arbor Science Publishers, 1972. 24. D'Alleva, B. A ., and Lovell, W. G.: “Relation of Exhaust Gas Composition to Air-Fuel Ratio," SAE J ., vol. 38, no. 3, pp. 90-96, March 1936. 25. Stivender, D. L.: “Development of a Fuel-Based Mass Emission Measurement Procedure," SAE paper 710604, 1971. 26. Harrington, J. A ., and Shishu, R. C.: "A Single-Cylinder Engine Study of the Effects of Fuel Type, Fuel Stoichiometry and Hydrogen-to-Carbon Ratio on CO, NO, and HC Exhaust Emissions," SAE paper 730476, 1973. 27. Spindt, R. S ., "Air-Fuel Ratios from Exhaust Gas Analysis," SAE paper 650507, SAE Trans ., vol. 74, 1965. 28. Fleming, R. D ., and Eccleston, D. B.: "The Effect of Fuel Composition, Equivalence Ratio, and Mixture Temperature on Exhaust Emissions," SAE paper 710012, 1971. 5.1 INTRODUCTION 29. Eltinge, L.: "Fuel-Air Ratio and Distribution from Exhaust Gas Composition," SAE paper 680114, SAE Trans ., vol. 77, 1968. The operating cycle of an internal combustion engine can be broken down into a 30. Leonard, L. S.: "Fuel Distribution by Exhaust Gas Analysis," SAE paper 379A, 1961. sequence of separate processes: intake, compression, combustion, expansion, and 31. Bishop, R. P.: "Combustion Efficiency in Internal Combustion Engines," B.S. Thesis, Department exhaust. With models for each of these processes, a simulation of a complete of Mechanical Engineering, MIT, February 1985. 32. The Engine Test Code Subcommittee of the General Technical Committee, General Motors Auto- engine cycle can be built up which can be analysed to provide information on motive Engine Test Code for Four Cycle Spark Ignition Engines, 6th ed ., 1975. engine performance. Models of individual engine processes at various levels of approximation have been developed. In this chapter we consider the simplest set of models which provide useful insights into the performance and efficiency of engines. The cycles analysed are commonly called the constant-volume, constant- pressure, and limited-pressure cycles; each title describes the approximation made for the engine combustion process.+ The description of more accurate simulations of engine processes is deferred until Chap. 14. For each engine cycle, a choice of models for working fluid thermodynamic properties must be made. These models have been reviewed in Chap. 4. Ideal f These cycles are also referred to by the titles Otto cycle, Diesel cycle, and dual cycle, respectively, for historical reasons. The more descriptive titles used above are preferred because they avoid the often-made assumption that the SI or Otto engine is best approximated by the constant-volume cycle, and the CI or diesel engine is best approximated by the constant-pressure cycle. These assumptions are not necessarily correct. 161 162 INTERNAL COMBUSTION ENGINE FUNDAMENTALS IDEAL MODELS OF ENGINE CYCLES 163 engine cycle models combined with a simple ideal gas model (specific heats con- TABLE 5.1 stant throughout the engine cycle-model 1 in Table 4.2) provide analytic results Ideal models of engine processes and are useful for illustrative purposes; we will call these cycles ideal gas standard cycles. Ideal engine cycles combined with more realistic models of working fluid Process Assumptions properties (a frozen mixture of ideal gases for the unburned mixture and an equi- Compression (1-2) 1. Adiabatic and reversible (hence isentropic) librium mixture for the burned mixture-model 5 in Table 4.2) are called fuel-air Combustion (2-3) 1. Adiabatic cycles and provide more quantitative information on engine operation. 2. Combustion occurs at An internal combustion engine is not a heat engine in the thermodynamic (a) Constant volume definition of the term. It is not a closed system. The working fluid does not (b) Constant pressure (c) Part at constant volume and part at constant execute a thermodynamic cycle. The temperature changes which occur around pressure (called limited pressure) minimum and maximum cylinder volumes are not primarily a result of heat- 3. Combustion is complete (n = 1) transfer interactions. An engine can best be analyzed as an open system which Expansion (3-4) 1. Adiabatic and reversible (hence isentropic) exchanges heat and work with its surrounding environment (the atmosphere). Exhaust (4-5-6) 1. Adiabatic Reactants (fuel and air) flow into the system; products (exhaust gases) flow out. and 2. Valve events occur at top- and bottom-center intake (6-7-1) 3. No change in cylinder volume as pressure (An overall second law analysis of the engine from this point of view has already differences across open valves drop to zero been presented in Sec. 3.6.) Thus, the cycles discussed in this chapter are not 4. Inlet and exhaust pressures constant thermodynamic cycles. Rather, each is a consecutive sequence of processes 5. Velocity effects negligible through which we can follow the state of the working fluid as the engine executes a complete operating cycle. 5.2 IDEAL MODELS OF ENGINE PROCESSES P 3a 36 The sequence of processes which make up a typical SI and CI engine operating cycle has been described in Sec. 1.3. To illustrate these processes, cylinder pres- sure (p) and volume (V) data from a throttled four-stroke cycle SI engine are plotted as a p-V diagram in Fig. 5-1. The cycle can be divided into compression, 35 6 - | 1, 5 6 , 5 - juin A 30 V (a) (c) (e) 25 P 3. 20 3 Cylinder pressure/p 15 10 th 0 0.0 0.2 0.4 0.6 0.8 1.0 V (b) (d) Cylinder volume/ Vmax FIGURE 5-2 FIGURE 5-1 Pressure-volume diagrams of ideal cycles. Unthrottled operation: (a) constant-volume combustion; Pressure-volume diagram of firing spark-ignition engine. re = 8.4, 3500 rev/min, p; = 0.4 atm, pe = 1 (b) constant-pressure combustion; (c) limited-pressure combustion. (d) Throttled constant-volume atm, imep, = 2.9 atm. cycle; (e) supercharged constant-volume cycle. 164 INTERNAL COMBUSTION ENGINE FUNDAMENTALS IDEAL MODELS OF ENGINE CYCLES 165 combustion, expansion, exhaust, and intake processes. Sets of assumptions which Compression stroke: simplify each of these processes to a form convenient for analysis are given in Table 5.1. Pressure-volume diagrams for the constant-volume, constant-pressure, and $ 1 =rc 0 2 (5.4) limited-pressure cycles for unthrottled engine operation are illustrated in Fig. 5-2a to c. Throttled engine operation (pi < pe) and supercharged engine operation Since the process is adiabatic and reversible (P1 > Pe) are shown in Fig. 5-2d and e. In each cycle 1-2 is the compression process, 2-3 is the combustion process, 3-4 (or 2-4 in the constant-pressure cycle) S2 = $1 (5.5) is the expansion process, 4-5-6 is the exhaust process, and 6-7-1 is the intake process. The compression work is The most critical assumption in determining how useful these ideal cycles are as indicators of engine performance is the form assumed for the combustion Wc = U, - U2 = m(u1 - u2) (5.6) process. The real engine combustion process occupies a finite crank angle period Combustion process: (between about 20 and 70 crank angle degrees), and the spark or fuel-injection For the constant-volume cycle, timing may be retarded from its optimum advance to closer to TC. The constant- volume cycle is the limiting case of infinitely fast combustion at TC; the constant- V3 = V 2 43 - 42 = 0 pressure cycle would correspond to slow and late combustion; the limited- (5.7a, b) pressure cycle lies in between. For the constant-pressure cycle, P3 = P2 h3 - h2 = 0 (5.7c, d) 5.3 THERMODYNAMIC RELATIONS FOR ENGINE PROCESSES For the limited-pressure cycle, The overall engine operating parameters of greatest interest which can be deter- P3b = P3a mined from a thermodynamic analysis of the engine operating cycle are: (5.7e, f) and Usa - U2 = 0 h30 - 23a = 0 (5.7g, h) The indicated fuel conversion efficiency nr.i: Expansion stroke: For the constant-volume cycle, (5.1) my 2LHV S4 = $ 3 (which, since the combustion efficiency is unity, is equal to the indicated thermal V3 (5.8a, b) conversion efficiency n ., 1; see Sec. 3.6.2) and the expansion work is The indicated mean effective pressure (imep): W2 = U3 - U4 = m(u3 - u4) (5.9) imep = cl _ MCLHV NS.1 (5.2) For the constant-pressure cycle, Va We ,, the indicated work per cycle, is the sum of the compression stroke work and P3 = P2 DA = rc S4 = $ 3 (5.10a, b, c) the expansion stroke work: and the expansion work is Wei = Wc + WE (5.3) WE = U3 - U4 + P2(V3 - V2) Using the notation of Fig. 5-2 to define the endpoints of each engine process, the following relationships are obtained by applying the first and second laws of = m[(u3 - u4) + P2(V3 - 02 ) ] thermodynamics to the cylinder contents: = m[(h3 - ha ) + PAVA - P2 02] (5.11) - 166 INTERNAL COMBUSTION ENGINE FUNDAMENTALS IDEAL MODELS OF ENGINE CYCLES 167 For the limited-pressure cycle, 0 4/ V30 = c P3b = P3a S4 = $3b (5.12a, b, c) and the expansion work is TA - WE = U3b - U4 + P3(V3b - V34) = m [ ( 4 36 - U 4 ) + P 3 ( 0 36 - 0 3 2 ) ] = m[(h30 - ha) + P4 V4 - P3 V3a] (5.13) 5 The indicated fuel conversion efficiency is found by substitution into Eqs. FIGURE 5-3 Enthalpy-entropy diagram of gas state during (5.3) and (5.1): exhaust process. See text for explanation. For the constant-volume cycle: ml ( u3 - 1 4 ) - ( u2 - 41)] gas to leave have the same state-5. There is, therefore, a gradient in temperature (5.14) my 2LHV within the exhausted gas. The temperature of the first element exhausted, Ta, is slightly less than T4; the temperature of the last element exhausted is Ts. For the constant-pressure cycle: A displacement of gas out of the cylinder follows the blowdown process as the piston moves from BC to TC. If heat-transfer and kinetic energy dissipation m [ ( h3 - h4 ) - ( 112 - 11 ) + P4 V 4 - P2 02] (5.15) effects are neglected, no change in thermodynamic state of the gas occurs. In this ms QLHV displacement process, the mass of gas within the cylinder at the end of the blow- down process is further decreased by the ratio Vs/V6. For the limited-pressure cycle: The mass of residual gas m, in the cylinder at point 6 in the cycle is m [ ( h38 - 1 4 ) - ( 11 2 - 1 1 ) + P 4 V 4 - P3 U30] obtained by first determining the state of the gas (Ts, Us) at the end of the blow- (5.16) down process following an isentropic expansion from p4 to pe and then by ms QLHV reducing the cylinder volume to the clearance volume 16. The residual mass The state of the mixture at point 1 in the cycle depends on the intake fraction is thus given by mixture properties and the residual gas properties at the end of the exhaust m. stroke. x, = V 2 (5.17) When the exhaust valve opens at point 4, the cylinder pressure is above the m exhaust manifold pressure and a blowdown process occurs. In the ideal exhaust The average state of the exhausted gas can be determined by considering process model, this blowdown occurs with the piston stationary at BC. During the open system defined by the piston face, cylinder walls, and cylinder head, this blowdown process, the gas which remains inside the cylinder expands isen- shown in Fig. 5-4. Applying the first law of thermodynamics for an open system tropically. The gases escaping from the cylinder undergo an unrestrained expan- gives sion or throttling process which is irreversible. It is assumed that the kinetic energy acquired by each gas element as it is accelerated through the exhaust U6 - U4 = Pe(VA - V6) - H. (5.18a) valve is dissipated in a turbulent mixing process in the exhaust port into internal energy and flow work. Since it is also assumed that no heat transfer occurs, the where H. is the enthalpy of the mass of gas exhausted from the cylinder. The average specific exhaust enthalpy is, therefore, enthalpy of each element of gas after it leaves the cylinder remains constant. These processes are illustrated on an h-s diagram in Fig. 5-3. The gas remaining in the cylinder expands isentropically along the line 4-5. The first h = m4 U4 - mou6 + P. Ve (5.18b) m4 - m6 element of gas which leaves the cylinder at point 4 enters the exhaust manifold at state a on the pressure = pe line. An element that leaves the cylinder at an inter- which, with p = pe, defines the average exhausted-gas state. mediate state b on the expansion line 4-5 would enter the exhaust manifold at The mixture temperature at the end of the intake stroke and at the start of state c. At the end of the blowdown process the gas in the cylinder and the last the compression stroke (point 1 in Fig. 5-2) can now be determined, again using 168 INTERNAL COMBUSTION ENGINE FUNDAMENTALS IDEAL MODELS OF ENGINE CYCLES 169 The pumping mean effective pressure (pmep) is usually defined as a positive Inlet Exhaust quantity. Thus: For Pi < Pe: pmep = Pe - Pi (5.23a) For Pi > Pe: pmep = P; - Pe (5.23b) The net and gross indicated mean effective pressures are related by imep ,, = imep, - (pe - P) (5.24) System boundary FIGURE 5-4 The net indicated fuel conversion efficiency is related to the gross indicated fuel Definition of system boundary for thermodynamic conversion efficiency by analysis of ideal cycle processes. imep, (5.25) the open system in Fig. 5-4. Application of the first law between points 6 and 1 gives 5.4 CYCLE ANALYSIS WITH U1 - U6 = - p(V1 - V6) + (m1 - m6)h; (5.19a) IDEAL GAS WORKING FLUID or mi us - mo us = - P ( V1 - V6 ) + ( m1 - m6 )hi (5.19b) WITH c. AND c, CONSTANT or mj h, = m6 h6 + ( m1 - m 6 )h, + V2 ( Pi - Pe ) (5.19c) If the working fluid in these ideal cycles is assumed to be an ideal gas, with c, and c, constant throughout the engine operating cycle, the equations developed in the where h; is the specific enthalpy of the inlet mixture and p1 = Pi. previous section which describe engine performance and efficiency can be further Note that when p: < pe, part of the residual gas in the cylinder at the end of simplified. We will use the notation of Fig. 5-2. the exhaust stroke will flow into the intake system when the intake valve opens. This flow will cease when the cylinder pressure equals p ,. However, provided no heat transfer occurs, this backflow will not affect Eqs. (5.19) above, since the flow 5.4.1 Constant-Volume Cycle of residual through the intake valve is a constant enthalpy process. The compression work (Eq. 5.6) becomes In many engines, the closing of the exhaust valve and the opening of the intake valve overlap. Flow of exhausted gases from the exhaust system through Wc = mcy(T1 - T2) (5.26) the cylinder into the intake system can then occur. Equations (5.18) and (5.19) The expansion work (Eq. 5.9) becomes would have to be modified to account for valve overlap. In the four-stroke engine cycle, work is done on the piston during the WE = mcy(T3 - T4) (5.27) intake and the exhaust processes. The work done by the cylinder gases on the piston during exhaust is The denominator in Eq. (5.14), my QLHv, can be related to the temperature rise during combustion. For the working fluid model under consideration, the We = Pe ( V2 - V1) (5.20) U(T) lines for the reactants and products on a U-T diagram such as Fig. 3-5 are parallel and have equal slopes, of magnitude c ,. Hence, for a constant-volume The work done by the cylinder gases on the piston during intake is adiabatic combustion process Wi = p&(V1 - V2) (5.21) mc (T3 - T2) = my QLHV (5.28)+ The net work to the piston over the exhaust and intake strokes, the pumping work, is Wp = (p1 - PX(V1 - 12) (5.22) + Note that if insufficient air is available for complete combustion of the fuel, Eq. (5.28) must be modified. The right-hand side of the equation should then be ne my QLHy, where ne is the combustion which, for the cylinder gas system, is negative for p < Pe and positive for p; > Pe. efficiency given by Eq. (3.27). 170 INTERNAL COMBUSTION ENGINE FUNDAMENTALS IDEAL MODELS OF ENGINE CYCLES 171 Note that the heating values at constant volume and constant pressure are the The indicated mean effective pressure, using Eqs. (5.2) and (5.31), becomes same for this working fluid. For convenience we will define imep (5.32) Q =MCLHV (5.29) P1 m The dimensionless numbers re, y, and Q*/(c, T1) are sufficient to describe Q* is the specific internal energy (and enthalpy) decrease, during isothermal com- the characteristics of the constant-volume ideal gas standard cycle, relative to its bustion, per unit mass of working fluid. initial conditions p1, T1. The relation for indicated fuel conversion efficiency (Eq. 5.14) becomes It is useful to compare the imep-a measure of the effectiveness with which the displaced volume of the engine is used to produce work-and the maximum 15.1 = 3 - T4) -(T2 - T1) T4 - Ti pressure in the cycle, p3 . The ratio p3/p1 can be determined from the ideal gas T3 - T2 (5.30) T3 - T2 law applied at points 2 and 3, and the relation Since 1-2 and 3-4 are isentropic processes between the same volumes, 11 and 1/2, T3 . 0* = 1+. T2 - = r?-1 = 7 - 1 T2 (5.33) obtained from Eq. (5.28). Equations (5.32) and (5.33) then give where y = cp/c ,. Hence: imep 1 - 1/7:-1 (1 - 1)r:\ C. T1/Q* + 1/r?-1 (5.34) T4 P3 . 13 T, A high value of imep/p3 is desirable. Engine weight will increase with increasing p, to withstand the increasing stresses in components. and Eq. (5.30) can be rearranged as The indicated fuel conversion efficiency and the ratios imep/p1 and imep/p3 - for this ideal cycle model do not depend on whether the cycle is throttled or (5.31) supercharged. However, the relationships between the working fluid properties at points 1 and 6 do depend on the degree of throttling or supercharging. For throt- Values of n. for different values of y are shown in Fig. 5-5. The indicated tled engine operation, the residual gas mass fraction x, can be determined as fuel conversion efficiency increases with increasing compression ratio and follows. From Eq. (5.17), since state 5 corresponds to an isentropic expansion decreases as y decreases. from state 4 to p = pe, x, is given by x = P/P4)!!! (P/P.)"(P1/P4)Y 0.9 0.8 Since Y = 1.4 0.71 P1 _ P1 P2 P3 1 12 ,y = (1+- 0* ) -1 1.3 P4 P2 P3 P4 . 0.6 1.25 it follows that 0.5 nf. ig 1 (Pe/P)1/7 0.4 - x , = r [1 + Q*/(c, Tr?)]17 (5.35) 0.3 The residual mass fraction increases as p; decreases below pe, decreases as r. increases, and decreases as Q*/(c. T1) increases. 0.2 Through a similar analysis, the temperature of the residual gas To can be 0.1L determined: FIGURE 5-5 Ideal gas constant-volume cycle fuel con- 1 /7 4 8 12 16 20 24 . 28 version efficiency as a function of com- To De) ( 1 + (5.36) Tc pression ratio; y = c,/c .. 172 INTERNAL COMBUSTION ENGINE FUNDAMENTALS IDEAL MODELS OF ENGINE CYCLES 173 The mixture temperature at point 1 in the cycle can be related to the inlet The mean effective pressure is related to p1 and p3 via mixture temperature, Ti, with Eq. (5.19). For a working fluid with c, and c, con- imep Q* stant, this equation becomes (5.44) P1 Cp T1 = x, C. T6 + (1 - x,)c. T; - RTx ( Pe - 1) ( 5.37 ) imep 1 ( * ) P3 arcoTTI)(.Tns .. (5.45) Use of Eqs. (5.36) and (5.37) leads to the relation 1 - xx Ti 1 - 1/(vr)[P/P: + (y - 1)] (5.38) 5.4.3 Cycle Comparison Extensive results for the constant-volume cycle with y = 1.4 can be found in The above expressions are most useful if values for y and Q*/(c, T1) are chosen to Taylor.1 match real working fluid properties. Figure 5-5 has already shown the sensitivity of n : for the constant-volume cycle to the value of y chosen. In Sec. 4.4, average 5.4.2 Limited- and Constant- values of y. and y, were determined which match real working fluid properties over the compression and expansion strokes, respectively. Values for a stoichio- Pressure Cycles metric mixture appropriate to an SI engine are yy ~ 1.3, y ~ 1.2. However, The constant-pressure cycle is a limited-pressure cycle with p3 = P2. For the analysis of pressure-volume data for real engine cycles indicates that pV", where limited-pressure cycle, the compression work remains n ~ 1.3, is a good fit to the expansion stroke p-V data.2 Heat transfer from the Wc = mcx(T1 - T2) (5.39) burned gases increases the exponent above the value corresponding to yo. A value of y = 1.3 for the entire cycle is thus a reasonable compromise. The expansion work, from Eq. (5.13), becomes Q *, defined by Eq. (5.29), is the enthalpy decrease during isothermal com- W E = m [ Co ( T 38 - T 4 ) + P 3 ( 036 - 03a ) ] (5.40) bustion per unit mass of working fluid. Hence For the combustion process, Eqs. (5.7g, h) give (5.46) m32 - 3a QLHV = mc,(T32 - T2) (5.41a) A simple approximation for (m/m) is (re - 1)/re; i.e ., fresh air fills the displaced m 538 - 36 CLHV = mcp ( T36 - T3a) (5.41b) volume and the residual gas fills the clearance volume at the same density. Then, or my QLHV = m[C,(T32 - T2) + Cp (T36 - T32)] (5.41c) for isooctane fuel, for a stoichiometric mixture, (* is given by 2.92 x 106 fr. - 1)/re J/kg air. For y = 1.3 and an average molecular weight M = 29.3, c, = for a working fluid with c, and c, constant throughout the cycle. 946 J/kg . K. For T1 = 333 K, Q*/(c. T1) becomes 9.3 (rc - 1)/rc . For this value of Combining Eqs. (5.1), (5.3), and (5.39) to (5.41) and simplifying gives Q*/(c] T1) all cycles would be burning a stoichiometric mixture with an appropri- T4 - T1 ate residual gas fraction. nsx = 1 - 7 ( T 32 - 7 2 ) + y ( T 36 - 732 ) Pressure-volume diagrams for the three ideal cycles for the same compres- sion ratio and unburned mixture composition are shown in Fig. 5-6. For each Use of the isentropic relationships for the working fluid along 1-2 and 3b-4, with cycle, y = 1.3, re = 12, 0*/(c, T1) = 9.3(re - 1)/r = 8.525. Overall performance the substitutions characteristics for each of these cycles are summarized in Table 5.2. The constant- V38 volume cycle has the highest efficiency, the constant-pressure cycle the lowest a = P3 (5.42a, b) V3a efficiency. This can be seen from Eq. (5.43) where the term in square brackets is P2 equal to unity for the constant-volume cycle and greater than unity for the leads to the result limited- and constant-pressure cycles. The imep values are proportional to nf.; nsx = 1 -- 1 aBY - 1 since the mass of fuel burned per cycle is the same in all three cases. ay(B -1) + x - 1| (5.43) As the peak pressure p3 is decreased, the ratio of imep to p3 increases. This ratio is important because imep is a measure of the useful pressure on the piston, For f = 1 this result becomes the constant-volume cycle efficiency (Eq. 5.31). For and the maximum pressure chiefly affects the strength required of the engine « = 1, this result gives the constant-pressure cycle efficiency as a special case. structure. 174 INTERNAL COMBUSTION ENGINE FUNDAMENTALS IDEAL MODELS OF ENGINE CYCLES 175 0.8 130 3 0.7- Constant volume 0.6 110- Constant volume 100 - Limited pressure 67 - Constant pressure 0.5 Limited pressure 3ab 33 0.4- Constant pressure 0.3- 0.2 40 0.1- 2 0 8 12 16 20 24 28 20 FIGURE 5-7 Fuel conversion efficiency as a function of compression ratio, for constant-volume, constant-pressure, and limited-pressure ideal gas cycles. y = 1.3, 0*/(c. T1) = 9.3(r - 1)/re. For limited-pressure cycle, Py/P1 = 33, 67, 100. 0 12 VIVE FIGURE 5-6 Pressure-volume diagrams for constant-volume, limited-pressure, and constant-pressure ideal gas A more extensive comparison of the three cycles is given in Figs. 5-7 and standard cycles. r. = 12, y = 1.3, 0*/(c. T1) = 9.3(r - 1)/r = 8.525, P3J/P1 = 67. 5-8, over a range of compression ratios. For all cases y = 1.3 and 0*/(c, T1) = 9.3(rc - 1)/re. At any given re, the constant-volume cycle has the highest effi- ciency and lowest imep/p3. For a given maximum pressure p3, the constant- pressure cycle has the highest efficiency (and the highest compression ratio). For the limited-pressure cycle, at constant p3/P1, there is little improvement in effi- ciency and imep above a compression ratio of about 8 to 10 as re is increased. TABLE 5.2 Example 5.1 shows how ideal cycle equations relate residual and intake Comparison of ideal cycle results conditions with the gas state at point 1 in the cycle. An iterative procedure is imep imep required if intake conditions are specified. Pmax nr.i P1 P3 P1 Constant volume 0.525 16.3 0.128 128 Example 5.1. For y = 1.3, compression ratio re = 6, and a stoichiometric mixture Limited pressure 0.500 15.5 0.231 67 Constant pressure 0.380 11.8 0.46€ 25.3 with intake temperature 300 K, find the residual gas fraction, residual gas tem- perature, and mixture temperature at point 1 in the constant-volume cycle for y = 1.3; r = 12; Q*/(c. T;) = 8.525. Pe/P: == 1 (unthrottled operation) and 2 (throttled operation). 176 INTERNAL COMBUSTION ENGINE FUNDAMENTALS IDEAL MODELS OF ENGINE CYCLES 177 T T1 1- x, 300 1 - [1/(1.3 x 6)](P./P. + 0.3) (d) 0.7 - Y = 1.3 , A trial-and-error solution of Eqs. (a) to (d) is required. It is easiest to estimate x ,, solve for T1 from (d), evaluate Q*/(c. T1) from (a), and check the value of x, assumed with that given by (b). 0.6 - For (p/p) = 1 (unthrottled operation) the following solution is obtained: Constant pressure x, = 0.044, T1 = 344 K, c. T1 - = 8.1, in T, = 1316 K P3/P1 = 33 For (p./p.) = 2 the following solution is obtained: imer 0.4|- P 3 x, = 0.082, T1 = 391 K, c, T1 = 6.8, T, = 1580 K 0.3 - 67 0.2 H 100 5.5 FUEL-AIR CYCLE Limited pressure ANALYSIS 0.1 - Constant volume A more accurate representation of the properties of the working fluid inside the engine cylinder is to treat the unburned mixture as frozen in composition and the burned gas mixture as in equilibrium. Values for thermodynamic properties for 0 8 12 16 20 24 28 these working fluid models can be obtained with the charts for unburned and burned gas mixtures described in Sec. 4.5, or the computer codes summarized in FIGURE 5-8 Sec. 4.7. When these working fluid models are combined with the ideal engine Indicated mean effective pressure (imep) divided by maximum cycle pressure (p3) as a function of process models in Table 5.1, the resulting cycles are called fuel-air cycles.1 The compression ratio for constant-volume, constant-pressure, and limited-pressure cycles. Details same sequence of processes and assumptions are (with the notation of Fig. 5-2): as Fig. 5-7. 1-2 Reversible adiabatic compression of a mixture of air, fuel vapor, and For a stoichiometric mixture, for isooctane, residual gas without change in chemical composition. 44.38 2-3 Complete combustion (at constant volume or limited pressure or con- m 16.14 (1 - x) = 2.75(1 - x,) MJ/kg stant pressure), without heat loss, to burned gases in chemical equilibrium. 3-4 Reversible adiabatic expansion of the burned gases which remain in For y = 1.3, c. = 946 J/kg . K and chemical equilibrium. @* 2.75 x 106 2910 4-5-6 Ideal adiabatic exhaust blowdown and displacement processes with 946T, - (1 -x) = - (1 - x,) (a) C. T1 the burned gases fixed in chemical composition. 6-7-1 Ideal intake process with adiabatic mixing between residual gas and Equations (5.35), (5.36), and (5.38), for re = 6 and y = 1.3, become fresh mixture, both of which are fixed in chemical composition. (p./p)0. 769 x, 6 [1 + 0*/(c. T1 x 60.3)70.769 (b) The basic equations for each of these processes have already been presented in 10.23/ = = P) ( 1 + 10.769 (c) Sec. 5.3. The use of the charts for a complete engine cycle calculation will now be C. T1 x 60.3 illustrated. 178 INTERNAL COMBUSTION ENGINE FUNDAMENTALS IDEAL MODELS OF ENGINE CYCLES 179 5.5.1 SI Engine Cycle Simulation can be checked against the calculated values and an additional cycle computa- The mixture conditions at point 1 must be known or must be estimated. The tion carried out with the new calculated values if required. The convergence is following approximate relationships can be used for this purpose:3 rapid. The indicated fuel conversion efficiency is obtained from Eq. (5.1). The indi- (5.47) cated mean effective pressure is obtained from Eq. (5.2). The volumetric efficiency (see Sec. 2.10) for a four-stroke cycle engine is given by T1 = T,rox,( Pi) (5.48) Te(1 - x,) no - vi Padre - 1) (5.52) where T, = 1400 K and (y - 1)/y = 0.24 are appropriate average values to use for where Pa,; is the inlet air density (in kilograms per cubic meter) and v1 is the chart initial estimates. mixture specific volume (in cubic meters per kilogram of air in the original Given the equivalence ratio o and initial conditions T1 (K), p1 = p; (Pa), mixture). and v1 (m3/kg air), the state at point 2 at the end of compression through a volume ratio v1/02 = re is obtained from Eq. (4.25a) and the isentropic compres- sion chart (Fig. 4-4). The compression work Wc (J/kg air) is found from Eq. (5.6) Example 5.2. Calculate the performance characteristics of the constant-volume fuel- with the internal energy change determined from the unburned mixture chart air cycle defined by the initial conditions of Examples 4.2, 4.3, and 4.5. The compres- sion ratio is 8; the fuel is isooctane and the mixture is stoichiometric; the pressure (Fig. 4-3). and temperature inside the cylinder at the start of compression are 1 atm and 350 The use of charts to relate the state of the burned mixture to the state of the K, respectively. Use the notation of Fig. 5-2a to define the states at the beginning unburned mixture prior to combustion, for adiabatic constant-volume and and end of each process. constant-pressure combustion, has already been illustrated in Sec. 4.5.3. Example 4.2 analyzed the compression process: For the constant-volume cycle, T1 = 350 K, P1 = 101.3 kPa, 01 = 1 m3/kg air, 14,1 = 40 KJ/kg air 143 = 14,2 + Auf, J/kg air (5.49) T2 = 682 K, P2 = 1.57 MPa, 02 = 0.125 m3/kg air, 14,2 = 350 KJ/kg air where us2 is the sensible internal energy of the unburned mixture at T2 from Fig. 4-3 and Au? , is the internal energy of formation of the unburned mixture [given W1_2 = Wc = - 310 KJ/kg air by Eq. (4.32)]. Since 13 = 12, the burned gas state at point 3 can be located on Example 4.5 analyzed the constant-volume adiabatic combustion process (it the appropriate burned gas chart (Figs. 4-5 to 4-9). was assumed that the residual gas fraction was 0.08): For the constant-pressure cycle, U,3 = U 2 = 43.42 + Aus = - 5 KJ/kg air, $3 = 9.33 KJ/kg air . K h3 = h32 + Ahi .. J/kg air (5.50) 13 = 12 = 0.125 m3/kg air, T3 = 2825 K, P3 = 7100 kPa Since p3 = p2, the burned gas state at point 3 can be located (by iteration) on the Example 4.3 analyzed the expansion process, from these conditions after com- high-temperature burned gas charts, as illustrated by Example 4.5. bustion at TC, to the volume v, at BC of 1 m3/kg air: For the limited-pressure cycle, application of the first law to the mixture TA = 1840 K, between states 2 and 3b gives P4 = 570 kPa, U4 = - 1540 KJ/kg air 230 = 136 + P3 0gb = U2 + P3 02 = 432 + Aus. + P3 02 J/kg air (5.51) W3.4 = WE = 1535 KJ/kg air Since p3 for a limited-pressure cycle is given, point 3b can be located on the To check the assumed residual gas fraction, the constant entropy expansion process on the chart in Fig. 4-8 is continued from state 4 to the exhaust pressure P5 appropriate burned gas chart. of 1 atm = 101.3 kPa. This gives vs = 4.0 m3/kg air and Ts = 1320 K. The residual The expansion process 3-4 follows an isentropic line from 13 to v4 (V4 = 11) fraction from Eq. (5.17) is on the burned mixture charts. Equation (5.9) [or (5.11) or (5.13)] now gives the expansion work WE. The state of the residual gas at points 5 and 6 in the cycle is V2 0.125 x, = = 0.031 obtained by continuing this isentropic expansion from state 4 to p = pe. The Vs 4.0 residual gas temperature can be read from the equilibrium burned gas chart; the which is significantly different from the assumed value of 0.08. The combustion and residual gas fraction is obtained from Eq. (5.17). If values of T, and x, were expansion calculations are now repeated with the new residual fraction of 0.031 (the assumed at the start of the cycle calculation to determine T1, the assumed values compression process will not be changed significantly and the initial temperature is IDEAL MODELS OF ENGINE CYCLES 181 180 INTERNAL COMBUSTION ENGINE FUNDAMENTALS where my is the mass of fuel injected, urg is the latent heat of vaporization of the assumed fixed): fuel, co,f is the specific heat at constant volume of the fuel vapor, T2, is the 1453 = 350 - 118.2 - 2956 x 0.031 = 140 KJ/kg air mixture temperature (assumed uniform) after vaporization and mixing is com- With v3 = 0.125 m3/kg air, Fig. 4-8 gives plete, ma is the mass of air used, and c ,., is the specific heat at constant volume of P3 = 7270 kPa, T3 = 2890 K air. Substitution of typical values for fuel and air properties gives (72 - T2,) ~ 70 K at full load. Localized cooling in a real engine will be greater. Expand at constant entropy to 04 = 1 m3/kg air: The limited-pressure cycle is a better approximation to the diesel engine T4 = 1920 K, 44 = - 1457 KJ/kg air than the constant-pressure or constant-volume cycles. P4 = 595 kPa, Note that because nonuniformities in the fuel/air ratio exist during and W3.4 = WE = 1597 KJ/kg air after combustion in the CI engine, the burned gas charts which assume uniform Continue expansion at constant entropy to the exhaust pressure, ps = 1 atm: composition will not be as accurate an approximation to working fluid properties as they are for SI engines. vs = 4 m3/kg air, Ts = 1360 K Equation (5.17) now gives the residual fraction 5.5.3 Results of Cycle Calculations 0.125 Extensive results of constant-volume fuel-air cycle calculations are available.1, 3, 4 IS x = = 0.031 4 Efficiency is little affected by variables other than the compression ratio re and equivalence ratio o. Figures 5-9 and 5-10 show the effect of variations in these which agrees with the value assumed for the second iteration. The fuel conversion efficiency can now be calculated: two parameters on indicated fuel conversion efficiency and mean effective pres- sure. From the available results, the following conclusions can be drawn: WE + WC my 2LHV 1. The effect of increasing the compression ratio on efficiency at a constant equivalence ratio is similar to that demonstrated by the constant y constant- where my = kg fuel/kg air at state 1 = (4 x1 - x> volume cycle analysis (provided the appropriate value of y is used; see Fig. 5-19). Thus 2. As the equivalence ratio is decreased below unity (i.e ., the fuel-air mixture is 1597 - 310 made progressively leaner than stoichiometric), the efficiency increases. This ns.1 - 0.0661 x (1 - 0.031) x 44.4 x 103 = 0.45 occurs because the burned gas temperatures after combustion decrease, decreasing the burned gas specific heats and thereby increasing the effective The indicated mean effective pressure is value of y over the expansion stroke. The efficiency increases because, for a WE + Wc . 1597 - 310 given volume-expansion ratio, the burned gases expand through a larger tem- imep = = 1470 kPa Va 1 - 0.125 perature ratio prior to exhaust; therefore, per unit mass of fuel, the expansion stroke work is increased. or imep = 14.6 3. As the equivalence ratio increases above unity (i.e ., the mixture is made pro- Pi gressively richer than stoichiometric), the efficiency decreases because lack of sufficient air for complete oxidation of the fuel more than offsets the effect of decreasing burned gas temperatures which decrease the mixture's specific 5.5.2 CI Engine Cycle Simulation heats. With a diesel engine fuel-air cycle calculation, additional factors must be taken 4. The mean effective pressure, from Eq. (5.2), is proportional to the product into account. The mixture during compression is air plus a small amount of ons .. This exhibits a maximum between o ~ 1.0 and o ~ 1.1, i.e ., slightly rich residual gas. At point 2 liquid fuel is injected into the hot compressed air at of stoichiometric. For o less than the value corresponding to this maximum, temperature T2; as the fuel vaporizes and heats up, the air is cooled. For a the decreasing fuel mass per unit displaced volume more than offsets the constant-volume mixing process which is adiabatic overall, the mixture internal increasing fuel conversion efficiency. For o greater than this value, the energy is unchanged, i.e.: decreasing fuel conversion efficiency (due to decreasing combustion efficiency) my[us, + Co.(T2. - To)] + maCo,a(T2. - T2) = 0 (5.53) more than offsets the increasing fuel mass. IDEAL MODELS OF ENGINE CYCLES 183 18 T1 = 388 K 16 x, = 0.05 PI = 1.0 atm 1.6 14 +24- 20 12 1.4 16 12 8 1.2 6 imep/p, 10 P1 = 0.5 atm 1.0 Fuel/air equivalence ratio ¢ 8 0.8 6 0.6 Pi = 1.0 atm Ti = 388 K FIGURE 5-10 4 Fuel-air cycle results for indicated mean effective pressure as a function 0.4 of equivalence ratio and compres- 0.30 0.25 0.40 0.35 0.50 0.55 0.45 0.60 21 sion ratio. Fuel: octene; p, = 1 atm, 0.6 0.4 0.6 0.8 1.0 1.2 1.4 1.6 T1 = 388 K, x, = 0.05. (From Edson Fuel/air equivalence ratio o and Taylor.4) Fuel-air cycle results for indicated fuel conversion efficiency as a function of compression ratio and equivalence ratio. Fuel: octene; p, = 1 atm, T; = 388 K, 30 5. Variations in initial pressure, inlet temperature, residual gas fraction, and atmospheric moisture fraction have only a modest effect on the fuel conver- 1.0 1.2 25 10.8 sion efficiency. The effects of variations in these variables on imep are more substantial, however, because imep depends directly on the initial charge density. 20 6. Comparison of results from limited-pressure and constant-volume fuel-air cycles1 shows that placing a realistic limit on the maximum pressure reduces Compression ratio re 15 the advantages of increased compression ratio on both efficiency and imep. 10 5.6 OVEREXPANDED ENGINE CYCLES P: = 1.0 atm T = 388 K x, = 0.05 The gas pressure within the cylinder of a conventional four-stroke engine at exhaust valve opening is greater than the exhaust pressure. The available energy 0 of the cylinder gases at this point in the cycle is then dissipated in the exhaust x, = 0.05. (From Edson and Taylor.4) 0.30 0.25 0.35 0.40 0.50 0.55 blowdown process. Additional expansion within the engine cylinder would 0.65 0.60 f. : 0.45 increase the indicated work per cycle, as shown in Fig. 5-11, where expansion continues beyond point 4' (the conventional ideal cycle exhaust valve opening FIGURE 5-9 point) at VA. = r Ve to point 4 at 14 = r. V .. The exhaust stroke in this over- expanded cycle is 4-5-6. The intake stroke is 6-1. The area 14'451 has been added 182 184 INTERNAL COMBUSTION ENGINE FUNDAMENTALS IDEAL MODELS OF ENGINE CYCLES 185 maximum at maximum load. This contrasts with the ideal constant-volume cycle 3 efficiency [Eq. (5.31)], which is independent of load. The ratio rr, for complete expansion is given by FIGURE 5-11 Pressure-volume diagram for overexpanded engine rx = 1 + - 2 C. Tiry-1 (5.56) Pi. Pem .5* cycle (1234561) and Atkinson cycle (1235+61). re and r. are volumetric compression and expansion ratios, VC V respectively. The effect of overexpansion on fuel conversion efficiency is shown in Fig. 5-12 for r = 4, 8, and 16 with y = 1.3. The ratio of overexpanded cycle efficiency to the standard cycle efficiency is plotted against r. The Atkinson cycle (complete expansion) values are indicated by the transition from a continuous line to a to the conventional cycle p-V diagram area, for the same fuel input, thereby dashed line. Significant increases in efficiency can be achieved, especially at low increasing the engine's efficiency. compression ratios. Complete expansion within the cylinder to exhaust pressure pe (point 5*) is One major disadvantage of this cycle is that imep and power density called the Atkinson cycle. Unthrottled operation is shown in Fig. 5-11; throttled decrease significantly because only part of the total displaced volume is filled operating cycles can also be generated. Many crank and valve mechanisms have with fresh charge. From Eqs. (5.2), (5.29), and the relations Va = V1(re - 1)/r, and been proposed to achieve this additional expansion. For example, it can be achieved in a conventional four-stroke cycle engine by suitable choice of exhaust valve opening and intake valve closing positions relative to BC. If the crank 20 angle between exhaust valve opening and BC on the expansion stroke is less than the crank angle between BC and intake valve closing on the compression stroke, y = 1.3 = 9.3(re - 1) -18 then the actual volumetric expansion ratio is greater than the actual volumetric compression ratio (these actual ratios are both less than the nominal compression 1.6/ 16 ratio with normal valve timing). The effect of overexpansion on efficiency can be estimated from an analysis ֏14 of the ideal cycle shown in Fig. 5-11. An ideal gas working fluid with specific heats constant throughout the cycle will be assumed. The indicated work per 1.4 12 cycle for the overexpanded cycle is imep Wc.[ = m[(143 - 4 4) -( 42- 41 ) - P1 ( Va - 0 1 ) ] (5.54) -10 8 1.2 - The isentropic relations for 1-2 and 3-4 are 8 16 12 = r? -1 Is =ri-1 16 -6 8 1.0 4 With Eq. (5.33) to relate T3 and T2, the following expression for indicated fuel re = 4 conversion efficiency can be derived from Eqs. (5.1), (5.29), and (5.54): -2 15.1 = 1 - 7 1 ( rm ) - 11 (5.55) 2 r = rare where r = Te FIGURE 5-12 Indicated fuel conversion efficiency and mean effective pressure for overexpanded engine cycle as a Note that the efficiency given by Eq. (5.55) is a function of load (via Q*), and is a function of r /r . Efficiencies given relative to r. = re value, n.10 . y = 1.3, 0*/(c. T1) = 9.3(r - 1)/r .. Solid to dashed line transition marks the complete expansion point (Atkinson cycle). 186 INTERNAL COMBUSTION ENGINE FUNDAMENTALS IDEAL MODELS OF ENGINE CYCLES 187 P.V1 = mRT, it follows that imep for the overexpanded cycle is given by ductive use. It must, therefore, be subtracted from the total work to obtain the useful work transfer: imer (5.57) P1 WU.1-2 5 -[(U2 - U1) + Po(V2 - V1) - To(S2 - S1)] (5.59) Values of imep/p1 are plotted in Fig. 5-12 as a function of r(=r/rc). The substan- The maximum useful work will be obtained when the final state of the system is in thermal and mechanical equilibrium with the atmosphere.+ The availability of tial decrease from the standard constant-volume cycle values at r = 1 is clear. this system which is in communication with the atmosphere A = U+ P.V. - TOS (5.60) 5.7 AVAILABILITY ANALYSIS OF ENGINE PROCESSES is thus the property of the system-atmosphere combination which defines its capacity for useful work. The useful work such a system-atmosphere combination 5.7.1 Availability Relationships can provide, as the system changes from state 1 to state 2, is less than or equal to the change in availability: Of interest in engine performance analysis is the amount of useful work that can be extracted from the gases within the cylinder at each point in the operating WU. 1-2 5 -(A2 - A1) (5.61) cycle. The problem is that of determining the maximum possible work output (or When mass flow across the system boundary occurs, the availability associ- minimum work input) when a system (the charge within the cylinder) is taken ated with this mass flow is from one specified state to another in the presence of a specified environment (the atmosphere). The first and second laws of thermodynamics together define this B = H - TOS (5.62) maximum or minimum work, which is best expressed in terms of the property of B is usually called the steady-flow availability function. such a system-environment combination called availability or sometimes With these relations, an availability balance for the gas working-fluid exergy.6, 7 system around the engine cycle can be carried out. For any process between Consider the system-atmosphere combination shown in Fig. 5-13. In the specified end states which this system undergoes (interacting only with the absence of mass flow across the system boundary, as the system changes from atmosphere), the change in availability AA is given by state 1 to state 2, the first and second laws give AA = Ain - Aout - Adestroyed (5.63) W1-2 = - (U2 -U1)+21-2 The availability transfers in and out occur as a result of work transfers, heat @1-2 S To(S2 - S1) transfers, and mass transfers across the system boundary. The availability trans- fer associated with a work transfer is equal to the work transfer. The availability Combining these two equations gives the total work transfer: transfer dA, associated with a heat transfer 6Q occurring when the system tem- W. 1-2 < -[(U2-U2) - To(S2 - S1 )] (5.58) perature is T is given by The work done by the system against the atmosphere is not available for pro- dAQ = 801 -To) (5.64) since both an energy and entropy transfer occurs across the system boundary. Atmosphere (To, Po) The availability transfer associated with a mass transfer is given by Eq. (5.62). + The issue of chemical equilibrium with the atmosphere must also be considered. Attainment of System mout (T, p. m) chemical equilibrium with the environment requires the capacity to extract work from the partial pressure differences between the various species in the working fluid and the partial pressures of those min same species in the environment. This would require such devices as ideal semipermeable membranes FIGURE 5-13 and efficient low input pressure, high pressure ratio, expansion devices (which are not generally avail- System-atmosphere configuration for availability able for mobile power plant systems). Inclusion of these additional steps to achieve full equilibrium W analysis. beyond equality of temperature and pressure is inappropriate.8 188 INTERNAL COMBUSTION ENGINE FUNDAMENTALS IDEAL MODELS OF ENGINE CYCLES 189 Availability is destroyed by the irreversibilities that occur in any real process. The 3800 availability destroyed is given by Adestroyed = To ASirrev (5.65) Constant volume. 3000 where ASirrey is the entropy increase associated with the irreversibilities occurring Limited pressure P3/P: = 67 within the system boundary.7, 8 r. K 2000 3a 5.7.2 Entropy Changes in Ideal Cycles 4 The ideal models of engine processes examined earlier in this chapter provide 1000 Constant pressure useful illustrative examples for availability analysis. First, however, we will con- sider the variation in the entropy of the cylinder gases as they proceed through these ideal operating cycles. 300 0 For an adiabatic reversible compression process, the entropy is constant. 1.0 ( s - Si ) /CV 2.0 For the combustion process in each of the ideal gas standard cycles, the entropy FIGURE 5-14 increase can be calculated from the relations of Eq. (4.14) (with constant specific Temperature-entropy diagram for ideal gas constant-volume, constant-pressure, and limited-pressure cycles. Assumptions same as in Fig. 5-6. heats): s - So = c, In + R in (" = Cp In - R In 5.7.3 Availability Analysis of Ideal Cycles For the constant-volume cycle: An availability analysis for each process in the ideal cycle illustrates the magni- S3 - S2 = m(s3 - $2) = mc, In (5.66a) tude of the availability transfers and where the losses in availability occur.9 In general, for the system of Fig. 5-4 in communication with an atmosphere at po, To as indicated in Fig. 5-13, the change in availability between states i and j For the constant-pressure cycle: during the portion of the cycle when the valves are closed is given by S3 - S2 = m(S3 - S2) = mc, In (5.66b) . A, - A, = m(a; - a) = m[(u; - u.) + Po(v, - v.) - To(s; - s)] (5.67) The appropriate normalizing quantity for these changes in availability is the For the limited-pressure cycle: thermomechanical availability of the fuel supplied to the engine cylinder each cycle, m (-Ag298)t (see Sec. 3.6.2). However, it is more convenient to use S36 - S2 = C, In = C, In a + c , In B . (5.66c) my(-Ah298)t = my OLHy as the normalizing quantity since it can be related to the temperature rise during combustion via Eq. (5.28). As shown in Table 3.3, with a and ß defined by Eq. (5.42). these two quantities differ by only a few percent for common hydrocarbon fuels. Equation (5.67), with Eq. (5.29), then becomes Since the expansion process, after combustion is complete, is adiabatic and reversible, there is no further change in entropy, 3 to 4 (or 3b to 4). Figure 5-14 A, - A1 _ m(a, - a) shows the entropy changes that occur during each process of these three ideal engine operating cycles, calculated from the above equations, on a T-s diagram. my QLHV my QLHV 0* (5.68) The three cycles shown correspond to those of the p-V diagrams of Fig. 5-6 with T = 12, y = 1.3, and Q*/(c, T1) = 8.525. Since the combustion process was assumed to be adiabatic, the increase in entropy during combustion clearly + 49 298 is the Gibbs free energy change for the combustion reaction, per unit mass of fuel. demonstrates the irreversible nature of this process. $ 4h298 is the enthalpy change for the combustion reaction, again per unit mass of fuel. 190 INTERNAL COMBUSTION ENGINE FUNDAMENTALS IDEAL MODELS OF ENGINE CYCLES 191 The compression process is isentropic, so: 0.2 A2 - A1_ 92 - 91 ( 11 2 - 1 1 ) + Po( v2 - 0 1 ) 0 ms QLHV : [(2-1)-6-11-2)] -0.2 I [(2 - 1) - (6- 10(1-1)] (5.69) -0.4 where we have assumed po = p1. The first term in the square brackets is the -0.6- compression stroke work transfer. The second term is the work done by the atmosphere on the system, which is subtracted because it does not increase the -0.8- Constant-volume cycle FIGURE 5-15 T = 12, 0*/(c. T1) = 8.525 Availability of cylinder charge relative to avail- useful work which the system-atmosphere combination can perform. y = 1.3, To = 300 K ability at state 1 for constant-volume ideal gas During combustion, for the constant-volume cycle, the volume and internal cycle as a function of cylinder volume. Availability energy remain unchanged (Eqs. 5.7a, b). Thus 0 2 4 6 8 10 12 made dimensionless by m/ 2LHv . Assumptions as VIVO in Fig. 5-6. A3 - A2_03-02 -. To(s3 - $2) my 2LHV 0* 0* The availability of the gases inside the cylinder relative to their availability C. To at (T1, P1) over the compression and expansion strokes of the constant-volume º In Co To1. -In 1 +- C " Tir ! - (5.70) operating cycle example used in Figs. 5-6 and 5-14 is shown in Fig. 5-15. Equa- tions (5.69) and (5.71), with T2 and T4 replaced by temperatures intermediate This loss in availability results from the increase in entropy associated with the between Ti and T2 and T3 and T4, respectively, were used to compute the varia- irreversibilities of the combustion process. This lost or destroyed availability, as a tions during compression and expansion. Table 5.3 summarizes the changes in fraction of the initial availability of the fuel-air mixture, decreases as the compres- availability during each process and the availability of the cylinder gases, at the sion ratio increases (since T2 increases as the compression ratio increases, T3/T2 beginning and end of each process, relative to the datum for the atmosphere decreases for fixed heat addition) and increases as Q* decreases [e.g ., when the mixture is made leaner; see Eq. (5.46)]. The changes in availability during com- TABLE 5.3 bustion for the constant-pressure and limited-pressure cycles are more complex because there is a transfer of availability out of the system equal to the expansion Availability changes in constant-volume cycle work transfer which occurs. For the constant-volume cycle expansion stroke: Aj - At Process or state A4 - A3 9 4 - 93 ( UA - U3 ) + Po(VA - V3 ) my 2LHV my @LHV my QLHV 1.0294 1-2 0.0976 2 .. 1270 2-3 -0.1710 3 0.9560 3-4 4 -0.6237 0.3323 Fuel conversion The availability of the exhaust gas at state 4 relative to its availability at efficiency ns.i 0.526 Availability conversion (T1, P1) is given by efficiency na,i 0.511 A - A - 8 2 (2 - 1 ) (F)] (5.72) :: = 12 7 = 1.3, 0*/(c. T.) = 8.525, To = 300 K, T, = 333 my QLHV 192 INTERNAL COMBUSTION ENGINE FUNDAMENTALS IDEAL MODELS OF ENGINE CYCLES 193 (1 atm, 300 K). The availability at state 1 of the fuel, air, residual-gas mixture is (1.0286 + 0.0008)m/ QLHv. 1.0286 is the ratio (-Ag298)/(-Ah298) for isooctane (see Table 3.3). The second number, 0.0008, allows for the difference between T, 0.8 and To. Because both work-transfer processes in this ideal cycle case are reversible, the fuel conversion efficiency n ., is given by (43 - A4)/(m/ 2LHV) - (A, 0.6 - A1)/(m/ OLHv). It is, of course, equal to the value obtained for re = 12 and y = 1.3 from the formula for efficiency (Eq. 5.31), obtained previously. The avail- ability conversion efficiency is n /1.0286. Note that it is the availability 0 4 FIGURE 5-17 destroyed during combustion, plus the inability of this ideal constant-volume -- re = 36 Availability of combustion products after cycle to use the availability remaining in the gas at state 4, that decrease the 0.2 Te = 12 constant-volume combustion relative to avail- availability conversion efficiency below unity. Both these loss mechanisms ability before combustion following isentropic decrease in magnitude, relative to the fuel availability, as the compression ratio compression from ambient through specified com- 0.2 0.4 0.6 0.8 1.0 increases. This is the fundamental reason why engine indicated efficiency pression ratio as a function of equivalence ratio. Fuel/air equivalence ratio (From Flynn et al.8) increases with an increasing compression ratio. availability increases as the equivalence ratio decreases.+ The combustion loss is 5.7.4 Effect of Equivalence Ratio a stronger function of the rise in temperature and pressure which occurs than of The fuel-air cycle with its more accurate models for working fluid properties can the change in the specific heat ratio that occurs. be used to examine the effect of variations in the fuel/air equivalence ratio on the Why then does engine efficiency increase with a decreasing equivalence availability conversion efficiency. Figure 5-16 shows the temperature attained and ratio as shown in Fig. 5-9? The reason is that the expansion stroke work transfer, the entropy rise that occurs in constant-volume combustion of a fuel-air mixture as a fraction of the fuel availability, increases as the equivalence ratio decreases; of different equivalence ratios, following isentropic compression from ambient hence, the availability lost in the exhaust process, again expressed as a fraction of temperature and pressure through different volumetric compression ratios.8 The the fuel availability, decreases. The increase in the expansion stroke work as the entropy increase is the result of irreversibilities in the combustion process and cquivalence ratio decreases more than offsets the increase in the availability lost mixing of complete combustion products with excess air. The significance of these during combustion; so the availability conversion efficiency (or the fuel conver- combustion-related losses-the destruction of availability that occurs in this sion efficiency which closely approximates it) increases. process-is shown in Fig. 5-17 where the availability after constant-volume com- bustion divided by the availability of the initial fuel-air mixture is shown as a function of equivalence ratio for compression ratios of 12 and 36.8 The loss of 5.8 COMPARISON WITH REAL ENGINE CYCLES 3500 To put these ideal models of engine processes in perspective, this chapter will conclude with a brief discussion of the additional effects which are important in 3000 - real engine processes. 1.0 A comparison of a real engine p-V diagram over the compression and Initial 0.8 2500 compression expansion strokes with an equivalent fuel-air cycle analysis is shown in Fig. ratio 0.6 5-18.“ The real engine and the fuel-air cycle have the same geometric compres- Temperature, K 2000 sion ratio, fuel chemical composition and equivalence ratio, residual fraction and -16 mixture density before compression. Midway through the compression stroke, 1500 30 FIGURE 5-16 @ = 0.2 Temperature and entropy of combustion products after constant-volume combustion following isea- 1000 12 tropic compression from ambient conditions 500 through specified compression ratio as a function 1 This is consistent with the ideal gas standard cycle result (Eq. 5.70). As ¢ decreases, so does 0 0.5 1.0 1.5 2.0 of compression ratio and equivalence ratio. (From Qº/(c. T1). The factor which multiplies the natural logarithm (which increases) has a greater impact Entropy, KJ/kg.K. Flynn et al.8) than the logarithmic term (which decreases). 194 INTERNAL COMBUSTION ENGINE FUNDAMENTALS IDEAL MODELS OF ENGINE CYCLES 195 cycle will be lower. During expansion, heat transfer will cause the gas pressure in the real cycle to fall below an isentropic expansion line as the volume increases. A decrease in efficiency results from this heat loss. Vc = 68 cm3 2. Finite combustion time. In an SI engine with spark-timing adjusted for rc = 11 optimum efficiency, combustion typically starts 10 to 40 crank angle degrees 5 before TC, is half complete at about 10º after TC, and is essentially complete 30 to 40º after TC. Peak pressure occurs at about 15º after TC (see Fig. 1-8). Pressure, MPa 4- In a diesel engine, the burning process starts shortly before TC. The pressure rises rapidly to a peak some 5 to 10º after TC since the initial rate of burning 3 is fast. However, the final stages of burning are much slower, and combustion Fuel-air cycle continues until 40 to 50º after TC (see Fig. 1-15). Thus, the peak pressure in Actual cycle the engine is substantially below the fuel-air cycle peak pressure value, because 1- FIGURE 5-18 combustion continues until well after TC, when the cylinder volume is much 0 Pressure-volume diagram for actual spark-ignition greater than the clearance volume. After peak pressure, expansion stroke pres- 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 engine compared with that for equivalent fuel-air sures in the engine are higher than fuel-air cycle values in the absence of other Displaced volume, dm3 cycle. r. = 11. (From Edson and Taylor.4) loss mechanisms, because less work has been extracted from the cylinder gases. A comparison of the constant-volume and limited-pressure cycles in Fig. 5-6 demonstrates this point. the pressure in the fuel-air cycle has been made equal to the real cycle pressure.t For spark or fuel-injection timing which is retarded from the optimum The compression stroke pressures for the two cycles essentially coincide. Modest for maximum efficiency, the peak pressure in the real cycle will be lower, and differences in pressure during intake and the early part of the compression expansion stroke pressures after the peak pressure will be higher than in the process result from the pressure drop across the intake valve during the intake optimum timing cycle. process and the closing of the intake valve 40 to 60º after BC in the real engine. The expansion stroke pressures for the engine fall below the fuel-air cycle pres- 3. Exhaust blowdown loss. In the real engine operating cycle, the exhaust valve is sures for the following reasons: (1) heat transfer from the burned gases to the opened some 60º before BC to reduce the pressure during the first part of the walls; (2) finite time required to burn the charge; (3) exhaust blowdown loss due exhaust stroke in four-stroke engines and to allow time for scavenging in two- to opening the exhaust valve before BC; (4) gas flow into crevice regions and stroke engines. The gas pressure at the end of the expansion stroke is therefore leakage past the piston rings; (5) incomplete combustion of the charge. reduced below the isentropic line. A decrease in expansion-stroke work trans- fer results. These differences, in decreasing order of importance, are described below. Together, they contribute to the enclosed area on the p-V diagram for a properly 4. Crevice effects and leakage. As the cylinder pressure increases, gas flows into adjusted engine with optimum timing being about 80 percent of the enclosed crevices such as the regions between the piston, piston rings, and cylinder wall. area of an equivalent fuel-air cycle p-V diagram. The indicated fuel conversion or These crevice regions can comprise a few percent of the clearance volume. availability conversion efficiency of the actual engine is therefore about 0.8 times This flow reduces the mass in the volume above the piston crown, and this the efficiency calculated for the fuel-air cycle.1 Use is often made of this ratio to flow is cooled by heat transfer to the crevice walls. In premixed charge estimate the performance of actual engines from fuel-air cycle results. engines, some of this gas is unburned and some of it will not burn. Though much of this gas returns to the cylinder later in the expansion, a fraction, from 1. Heat transfer. Heat transfer from the unburned mixture to the cylinder walls behind and between the piston rings, flows into the crankcase. However, has a negligible effect on the p-V line for the compression process. Heat trans- leakage in a well-designed and maintained engine is small (usually less than fer from the burned gases is much more important (see Chap. 12). Due to heat one percent of the charge). All these effects reduce the cylinder pressure during transfer during combustion, the pressure at the end of combustion in the real the latter stages of compression, during combustion, and during expansion below the value that would result if crevice and leakage effects were absent. 5. Incomplete combustion. Combustion of the cylinder charge is incomplete; the exhaust gases contain combustible species. For example, in spark-ignition + Note that in the fuel-air cycle with idealized valve timing, the compression process starts imme- engines the hydrocarbon emissions from a warmed-up engine (which come diately after BC. In most engines, the charge compression starts later, close to the time that the inlet valve closes some 40 to 60º after BC. This matching process is approximate. largely from the crevice regions) are 2 to 3 percent of the fuel mass under 196 INTERNAL COMBUSTION ENGINE FUNDAMENTALS IDEAL MODELS OF ENGINE CYCLES 197 normal operating conditions; carbon monoxide and hydrogen in the exhaust 0.8 contain an additional 1 to 2 percent or more of the fuel energy, even with excess air present (see Sec. 4.9). Hence, the chemical energy of the fuel which is Y = 1.4 released in the actual engine is about 5 percent less than the chemical energy 0.7| of the fuel inducted (the combustion efficiency, see Sec. 3.5.5, is about 95 -- percent). The fuel-air cycle pressures after combustion will be higher because P = 0.4 complete combustion is assumed. In diesel engines, the combustion ineffi- 1.3 0.6 0.8 ciency is usually less, about 1 to 2 percent, so this effect is smaller. ---- 1.0 -- 1.25 SUMMARY. The effect of all these loss mechanisms on engine efficiency is best 0.5 -- ----- defined by an availability balance for the real engine cycle. A limited number of --- such calculations have been published (e.g ., Refs. 8, 10, and 11). Table 5.4 shows (12) ¢ for -- the magnitude of the loss in availability (as a fraction of the initial availability) 77. is 0.4 (13) best efficiency that occurs due to real cycle effects in a typical naturally aspirated diesel (14) engine.1º The combustion and exhaust losses are present in the ideal cycle $ = 1.0 models also (they are smaller, however9). The loss in availability due to heat 0.3 losses, flow or aerodynamic losses, and mechanical friction are real engine effects. - - Figure 5-19 shows standard and fuel-air cycle efficiencies as a function of the compression ratio compared with engine indicated efficiency data. The top 0.2 three sets of engine data are for the best efficiency air/fuel ratio. Differences in the data are in part due to different fuels [(12) isooctane; (13) gasoline; (14) propane] which affect efficiency slightly through their different composition and heating values (see Table D.4). They also result from different combustion chamber 0.1- shapes which affect the combustion rate and heat transfer. The trends in the data with increasing compression ratio and the ¢ = 0.8 fuel-air cycle curve (which corresponds approximately to the actual air/fuel ratios used) are similar. The 0 4 12 16 20 24 factor of 0.8 relating real engine and fuel-air cycle efficiencies holds roughly. At 28 Compression ratio Ic compression ratios above about 14, however, the data show that the indicated FIGURE 5-19 efficiency of actual engines is essentially constant. Increasing crevice and heat Indicated fuel conversion efficiency as a function of compression ratio for ideal gas constant-volume cycle (dashed lines, y = 1.25, 1.3, 1.4) and fuel-air cycle (solid lines, ¢ = 0.4, 0.8, 1.0). Also shown are available engine data for equivalence ratios given : best efficiency 6;12-14 ( = 1.14 TABLE 5.4 Availability losses in naturally aspi- rated diesel losses offset the calculated ideal cycle efficiency increase as the compression ratio Loss, fraction of is raised above this value. The standard ideal gas cycle analysis results, with an Loss mechanism fuel availability appropriate choice for the value of y (1.25 to 1.3), correspond closely to the fuel- air cycle analysis results. Combustion ).225 Exhaust 0.144 The ideal cycle provides a convenient but crude approximation to the real Heat transfer 0.135 engine operating cycle. It is useful for illustrating the thermodynamic aspects of Aerodynamic 0.047 engine operation. It can also provide approximate estimates of trends as major Mechanical friction 0.048 Total losses 0.599 engine parameters change. The weakest link in these ideal cycles is the modeling Availability conversion 0.40 of the combustion processes in SI and CI engines. None of the models examined efficiency (brake) in this chapter are sufficiently close to reality to provide accurate predictions of engine performance. More sophisticated models of the spark-ignition and diesel Source: Traupel.1º engine operating cycles have been developed and are the subject of Chap. 14. 198 INTERNAL COMBUSTION ENGINE FUNDAMENTALS IDEAL MODELS OF ENGINE CYCLES 199 PROBLEMS 5.6. Use a limited-pressure cycle analysis to obtain a plot of indicated fuel conversion efficiency versus p3/p1 for a compression ratio of 15 with light diesel oil as fuel. 5.1. Many diesel engines can be approximated by a limited-pressure cycle. In a limited- Assume m/m = 0.04, T1 = 45ºC. Use y = 1.3 and cy = 946 J/kg . K. pressure cycle, a fraction of the fuel is burnt at constant volume and the remaining 5.7. Explain why constant-volume combustion gives a higher indicated fuel conversion fuel is burnt at constant pressure. Use this cycle approximation with y = c./c. = 1.3 efficiency than constant-pressure combustion for the same compression ratio. to analyze the following problem: 5.8. Two engines are running at a bmep of 250 kPa. One is an SI engine with the throttle Inlet conditions: P1 = 1.0 bar, T = 289 K partially closed to maintain the correct load. The second engine is a naturally aspi- Compression ratio: 15:1 rated CI engine which requires no throttle. Mechanical friction mep for both engines Heat added during combustion: 43,000 KJ/kg of fuel is 100 kPa. If the intake manifold pressures for the SI and CI engines are 25 kPa and Overall fuel/air ratio: 0.045 kg fuel/kg air 100 kPa respectively, and both exhaust manifold pressures are 105 kPa, use an ideal cycle model to estimate and compare the gross imep of the two engines. You may (a) Half of the fuel is burnt at constant volume, then half at constant pressure. Draw neglect the pressure drop across the valves during the intake and exhaust processes. a p-V diagram and compute the fuel conversion efficiency of the cycle. 5.9. (a) Plot net imep versus p; for 20 kPa < p; < 100 kPa for a constant-volume cycle (b) Compare the efficiency and peak pressure of the cycle with the efficiency and using the following conditions: m /m = 0.06, T1 = 40ºC, c) = 946 J/kg . K, peak pressure that would be obtained if all of the fuel were burnt at constant y = 1.3, Te = 9.5, QLHy = 44 MJ/kg fuel. Assume pe = 100 kPa. pressure or at constant volume. (b) What additional information is necessary to draw a similar plot for the engine's 5.2. It is desired to increase the output of a spark-ignition engine by either (1) raising the indicated torque, and indicated power? compression ratio from 8 to 10 or (2) increasing the inlet pressure from 1.0 atm to 5.10. (a) Draw a diagram similar to those in Fig. 5-2 for a supercharged cycle with 1.5 atm. Using the constant-volume cycle as a model for engine operation, which constant-pressure combustion. procedure will give: (b) Use the ideal gas cycle with constant-pressure combustion to model an engine (a) The highest pressure of the cycle? with a compression ratio of 14 through such a supercharged cycle. Find the (b) The highest efficiency? pressure and temperature at points corresponding to 2, 3, 4, and 5 in Fig. 5-2. (c) The highest mep? Assume a pressure of 200 kPa and temperature of 325 K at point 1, and a pres- Assume y = 1.3 and (m/ 2Hv)/(mc, T1) = 9.3(re - 1)/r .. sure of 100 kPa at points 5 and 6. m,/m = 0.03 and the fuel is a light diesel oil. 5.3. When a diesel engine, originally designed to be naturally aspirated, is turbocharged (c) Calculate the gross and net indicated fuel conversion efficiency and imep for this the fuel/air equivalence ratio o at full load must be reduced to maintain the engine under these operating conditions. maximum cylinder pressure essentially constant. If the naturally aspirated engine 5.11. Use the appropriate tables and charts to carry out a constant-pressure fuel-air cycle was designed for o = 0.75 at full load, estimate the maximum permissible value of ¢ calculation for the supercharged engine described in Prob. 5.10. Assume the same for the turbocharged engine at full load if the air pressure at the engine inlet is 1.6 initial conditions at point 1, with ¢ = 0.4 and a residual gas fraction of 0.025. A atm. Assume that the engine can be modeled with the limited-pressure cycle, with single cycle calculation is sufficient. half the injected fuel burned at constant volume and half at constant pressure. The (a) Determine the pressure and temperature at points 2, 3, 4, and 5. Calculate the compression ratio is 16. The fuel heating value is 42.5 MJ/kg fuel. Assume y == compression stroke, expansion stroke, and pumping work per cycle per kg air. Cp/Cy = 1.35, that the air temperature at the start of compression is 325 K, and (b) Find the gross and net indicated fuel conversion efficiency and imep. (F/A)stoich = 0.0666. (c) Compare the calculated residual gas fraction with the assumed value of 0.025. 5.4. A spark-ignition engine is throttled when operating at part load (the inlet pressure is 5.12. One method proposed for reducing the pumping work in throttled spark-ignition reduced) while the fuel/air ratio is held essentially constant. Part-load operation of engines is early intake valve closing (EIVC). The ideal cycle p-V diagram shown the engine is modeled by the cycle shown in Fig. 5-2d; the inlet air is at pressure P;, illustrates the concept. The EIVC cycle is 1-2-3-4-5-6-7-8-1 (the conventional throt- the exhaust pressure is atmospheric pa, and the ambient temperature is T .. Derive an tled cycle is 1-2-3-4-5-6-7 *- 1). With EIVC, the inlet manifold is held at a pressure pi expression for the decrease in net indicated fuel conversion efficiency due to throt- (which is higher than the normal engine intake pressure, pf), and the inlet valve is tling from the ideal constant-volume cycle efficiency and show that it is proportional closed during the inlet stroke at 8. The trapped fresh charge and residual is then to (Pa/P1 - 1). Assume mass fuel < mass air. expanded to the normal cycle (lower) intake pressure, pr. You can assume that both 5.5. (a) Use the ideal gas cycle with constant-volume combustion to describe the oper- cycles have the same mass of gas in the cylinder, temperature, and pressure at state 1 ation of an SI engine with a compression ratio of 9. Find the pressure and tem- of the cycle. perature at points 2, 3, 4, and 5 on Fig. 5-2a. Assume a pressure of 100 kPa and a (a) On a sketch of the intake and exhaust process p-V diagram, shade in the area temperature of 320 K at point 1. Assume m /m = 0.06, cy = 946 J/kg . K, y = 1.3. that corresponds to the difference between the pumping work of the EIVC cycle QLHV for gasoline is 44 MJ/kg. and that of the normal cycle. (b) Find the indicated fuel conversion efficiency and imep for this engine under these (b) What value of p, and Veryc will give the maximum reduction in pumping work operating conditions. for the EIVC cycle. IDEAL MODELS OF ENGINE CYCLES 201 200 INTERNAL COMBUSTION ENGINE FUNDAMENTALS can be done with valve timing.) If the expansion ratio re is 12, while the compres- P 3 sion ratio and other details of the cycle remain the same as in (a), what is the indicated efficiency and mean effective pressure (based on the new, larger, dis- placed volume) of this new engine cycle? 5.15. In spark-ignition engines, exhaust gas is recycled to the intake at part load to reduce the peak burned gas temperatures and lower emissions of nitrogen oxides. (a) Calculate the reduction in burned gas temperature that occurs when, due to exhaust gas recycle, the burned gas fraction in the unburned gas mixture (x)) Per inside the cylinder is increased from 10 percent (the normal residual fraction) to 30 percent. Assume combustion occurs at top-center, at constant volume, and is P; 7 * adiabatic. Conditions at the end of compression for both cases are: T = 700 K, VEINC Vm FIGURE P5-12 p = 1000 kPa, v = 0.2 m3/kg air in the original mixture; the equivalence ratio is 1.0. The fuel can be modeled as isooctane. (b) The compression ratio is 8. The compression stroke work is 300 kj/kg air in the (c) Derive an expression for this maximum difference in pumping work between the original mixture. Find the indicated work per cycle for the compression and normal cycle and the EIVC cycle in terms of pe, pr, Ve, and Vw. You can make expansion strokes, per kilogram of air in the original mixture, for these two cases. the appropriate ideal cycle assumptions. (c) Briefly explain how you would increase the work per cycle with 30 percent 5.13. Calculate the following parameters for a constant-volume fuel-air cycle (Fig. 5-2a): burned gas fraction in the unburned mixture to the value obtained with 10 (a) The pressures and temperatures at states 1, 2, 3, 4, 5, and 6 percent burned gas fraction, with fixed engine geometry. (A qualitative answer, (b) The indicated fuel conversion efficiency only, is required here.) (c) The imep 5.16. The following cycle has been proposed for improving the operation of a four-stroke (d) The residual fraction cycle engine. Its aim is to expand the postcombustion cylinder gases to a lower (e) The volumetric efficiency pressure and temperature by extending the expansion stroke, and hence extract more Inlet pressure = 1 atm, exhaust pressure = 1 atm, inlet temperature = 300 K, work per cycle. compression ratio == 8 : 1, equivalence ratio = .= 1. The cycle consists of: (1) an intake stroke; (2) a compression stroke, where the Calculate the above parameters (points a-e) using the SI units charts. Use 44.4 inlet valve remains open (and the cylinder pressure is constant) for the first portion MJ/kg for heating value of the fuel. Hint: Start the calculations using the residual of the stroke; (3) a combustion process, which occurs rapidly close to top-center; (4) mass fraction 0.03 and the residual gas temperature 1370 K. an expansion stroke, where the exhaust valve remains closed until the end of the 5.14. The cycle 1-2-3-4-5-6-1 is a conventional constant-volume fuel-air cycle with a com- stroke; (5) an exhaust stroke, where the cylinder pressure blows down to the exhaust pression ratio of 8. The fuel is isooctane, C8H18, with a lower heating value of pressure rapidly and most of the remaining combustion products are expelled as the 44.4 MJ/kg. The gas state at 1 is T1 = 300 K, p1 = 1 atmosphere with an equivalence piston moves from the BC to the TC position. Thus, for this engine concept, the ratio of 1.0 and zero residual fraction. The specific volume at state 1 is 0.9 m3/kg air compression ratio re (ratio of cylinder volume at inlet valve closing to clearance in the mixture. The temperature at the end of compression at state 2 is 600 K. volume) is less than the expansion ratio re (ratio of cylinder volume at exhaust valve (a) Find the indicated fuel conversion efficiency and mean effective pressure of this opening to clearance volume). fuel-air cycle model of a spark-ignition engine. (a) Sketch a p-V diagram for the cylinder gases for this cycle operating unthrottled. (b) The efficiency of the cycle can be increased by increasing the expansion ratio !, (b) Using the charts in SI units developed for fuel-air cycle calculations, carry out an while maintaining the same compression ratio re (cycle 1-2-3-44-5'4-6-1). (This analysis of an appropriate ideal model for this cycle where the compression ratio re is 8 and the expansion ratio r is (1) 8; (2) 16. Assume the following: Pressure in the cylinder at inlet valve close 1 atm Mixture temperature at inlet valve close 300 K Mixture equivalence ratio = 1.0 Fuel : isooctane C8H18 Lower heating value = 44.4 MJ/kg Residual gas mass fraction at inlet valve close 0.05 Stoichiometric fuel/air ratio = 0.066 Calculate the indicated work per cycle per kg of air in the original mixture (the 4A SA standard chart units) and the indicated mean effective pressure for these two -+ V expansion ratios. Base the mean effective pressure on the volume displaced by 8V. TeVe FIGURE P5-14 202 INTERNAL COMBUSTION ENGINE FUNDAMENTALS IDEAL MODELS OF ENGINE CYCLES 203 the piston during the expansion stroke. Tabulate your answers. (Note: You are Fuel : isooctane C8H18 given the initial conditions for the cycle calculation; changing the value of r. Lower heating value = 44.4 MJ/kg requires only modest changes in the cycle calculation.) Clearance volume negligible (c) Comment briefly on the effect of increasing the ratio re/re above 1.0 with this concept on engine efficiency and specific power (power per unit engine weight). (c) Compare these values with typical values for the constant-volume fuel-air cycle. Additional calculations are not required. Explain (with thermodynamic arguments) why the two cycles have such different 5.17. In a direct-injection stratified-charge (DISC) engine fuel is injected into the engine indicated mean effective pressures and efficiencies. cylinder just before top-center (like a diesel); a spark discharge is then used to initi- (d) Explain briefly why the real Lenoir engine would have a lower efficiency than the ate the combustion process. A four-stroke cycle version of this engine has a displaced value you calculated in (b) (the actual brake fuel conversion efficiency of the volume of 2.5 liters and a compression ratio of 12. At high load, the inlet pressure is engine was about 5 percent). boosted by a compressor to above atmospheric pressure. The compressor is geared 5.19. Estimate from fuel-air cycle results the indicated fuel conversion efficiency, the indi- directly to the engine drive shaft. The exhaust pressure is 1 atm. This DISC engine is cated mean effective pressure, and the maximum indicated power (in kilowatts) at to replace an equal displacement conventional naturally aspirated spark-ignition (SI) wide-open throttle of these two.four-stroke cycle spark-ignition engines: engine, which has a compression ratio of 8. (a) Draw qualitative sketches of the appropriate constant-volume ideal cycle A six-cylinder engine with a 9.2-cm bore, 9-cm stroke, compression ratio of 7, pressure-volume diagrams for the complete operating cycles for these two operated at an equivalence ratio of 0.8 engines at maximum load. A six-cylinder engine with an 8.3-cm bore, 8-cm stroke, compression ratio of (b) Use available fuel-air results to estimate how much the DISC engine inlet pres- 10, operated at an equivalence ratio of 1.1 sure must be boosted above atmospheric pressure by the compressor to provide Assume that actual indicated engine efficiency is 0.8 times the appropriate fuel-air the same maximum gross indicated power as the naturally aspirated SI engine. cycle efficiency. The inlet manifold pressure is close to 1 atmosphere. The maximum The SI engine operates with an equivalence ratio of 1.2; the DISC engine is permitted value of the mean piston speed is 15 m/s. Briefly summarize the reasons limited by smoke emissions to a maximum equivalence ratio of 0.7. why: (c) Under these conditions, will the brake powers of these engines be the same, given (a) The efficiency of these two engines is approximately the same despite their differ- that the mechanical rubbing friction is the same? Briefly explain. ent compression ratios. (d) At part load, the SI engine operates at an equivalence ratio of 1.0 and inlet (b) The maximum power of the smaller displacement engine is approximately the pressure of 0.5 atm. At part load the DISC engine has negligible boost and same as that of the larger displacement engine. operates with an inlet pressure of 1.0 atm. Use fuel-air cycle results to determine 5.20. The constant-volume combustion fuel-air cycle model can be used to estimate the the equivalence ratio at which the DISC engine must be operated to provide the effect of changes in internal combustion engine design and operating variables on same net indicated mean effective pressure as the SI engine. What is the ratio of engine efficiency. The following table gives the major differences between a diesel and DISC engine net indicated fuel conversion efficiency to SI engine efficiency at a spark-ignition engine both operating at half maximum power. these conditions? 5.18. The earliest successful reciprocating internal combustion engine was an engine devel- oped by Lenoir in the 1860s. The operating cycle of this engine consisted of two Diesel strokes (i.e ., one crankshaft revolution). During the first half of the first stroke, as the Spark-ignition engine piston moves away from its top-center position, fuel-air mixture is drawn into the engine cylinder through the inlet valve. When half the total cylinder volume is filled with Compression ratio 16: 1 9:1 fresh mixture, the inlet valve is closed. The mixture is then ignited and burns rapidly. Fuel/air equivalence ratio 0.4 1.0 During the second half of the first stroke, power is delivered from the high-pressure Inlet manifold pressure 1 atm 0.5 atm burned gases to the piston. With the piston in its bottom-center position, the exhaust valve is opened. The second stroke, the exhaust stroke, completes the cycle as the piston returns to top-center. (a) Use the graphs of fuel-air cycle results (Figs. 5-9 and 5-10) to estimate the ratio of (a) Sketch a cylinder pressure versus cylinder volume diagram for this engine. the diesel engine brake fuel conversion efficiency to the spark-ignition engine (b) Using the charts in SI units developed for fuel-air cycle calculations, carry out a brake fuel conversion efficiency. cycle analysis and determine the indicated fuel conversion efficiency and mean (b) Estimate what percentage of the higher diesel brake fuel conversion efficiency effective pressure for the Lenoir engine. Assume the following: comes from: (1) The higher diesel compression ratio Inlet pressure = 1 atm (2) The leaner diesel equivalence ratio Inlet mixture temperature = 300 K (3) The lack of intake throttling in the diesel compared with the spark-ignition Mixture equivalence ratio = 1.0 engine 204 INTERNAL COMBUSTION ENGINE FUNDAMENTALS The values of fuel conversion efficiency and mean effective pressure given in the graphs are gross indicated values (i.e ., values obtained from | p dV over the compression and expansion strokes only). CHAPTER You may assume, if necessary, that for the real engines, the gross indicated efficiency and gross indicated mean effective pressure are 0.8 times the fuel-air cycle values. Also, the mechanical rubbing friction for each engine is 30 percent of the net indicated power or mep. 6 REFERENCES GAS 1. Taylor, C. F.: The Internal Combustion Engine in Theory and Practice, vol. 1: Thermodynamics, EXCHANGE Fluid Flow, Performance, 2d ed ., chaps. 2 and 4, 1966. PROCESSES 2. Lancaster, D. R ., Krieger, R. B ., and Lienesch, J. H.: "Measurement and Analysis of Engine Pressure Data," SAE paper 750026, SAE Trans ., vol. 84, 1975. 3. Edson, M. H.: "The Influence of Compression Ratio and Dissociation on Ideal Otto Cycle Engine Thermal Efficiency," Digital Calculations of Engine Cycles, SAE Prog. in Technology, vol 7, pp. 49-64, 1964. 4. Edson, M. H ., and Taylor, C. F.: "The Limits of Engine Performance-Comparison of Actual and Theoretical Cycles," Digital Calculations of Engine Cycles, SAE Prog. in Technology, vol. 7, pp. 65-81, 1964. 5. Keenan, J. H.: Thermodynamics, John Wiley, New York, 1941; MIT Press, Cambridge, Mass ., 1970. 6. Haywood, R. W.: "A Critical Review of the Theorems of Thermodynamic Availability, with Concise Formulations; Part 1. Availability," J. Mech. Engng Sci ., vol. 16, no. 3, pp. 160-173, 1974. 7. Haywood, R. W.: "A Critical Review of the Theorems of Thermodynamic Availability, with Concise Formulations; Part 2. Irreversibility," J. Mech. Engng Sci ., vol. 16, no. 4, pp. 258-267. 1974. 8. Flynn, R. F ., Hoag, K. L ., Kamel, M. M ., and Primus, R. J.: "A New Perspective on Diesel Engine Evaluation Based on Second Law Analysis," SAE paper 840032, SAE Trans ., vol. 93, This chapter deals with the fundamentals of the gas exchange processes-intake 1984. and exhaust in four-stroke cycle engines and scavenging in two-stroke cycle 9. Clarke, J. M.: "The Thermodynamic Cycle Requirements for Very High Rational Efficiencies," paper C53/76, Institution of Mechanical Engineers, J. Mech. Engng Sci ., 1974. engines. The purpose of the exhaust and inlet processes or of the scavenging 10. Traupel, W.: "Reciprocating Engine and Turbine in Internal Combustion Engineering," in Proc. process is to remove the burned gases at the end of the power stroke and admit CIMAC Int. Congr. on Combustion Engines, Zurich, pp. 39-54, 1957. the fresh charge for the next cycle. Equation (2.38) shows that the indicated 11. Clarke, J. M.: "Letter: Heavy Duty Diesel Fuel Economy," Mech. Engng, pp. 105-106, March power of an internal combustion engine at a given speed is proportional to the 1983. mass flow rate of air. Thus, inducting the maximum air mass at wide-open throt- 12. Caris, D. F ., and Nelson, E. E.: "A New Look at High Compression Engines," SAE Trans ., vol. 67, pp. 112-124, 1959. tle or full load and retaining that mass within the cylinder is the primary goal of 13. Kerley, R. V ., and Thurston, K, W.: "The Indicated Performance of Otto-Cycle Engines," SAE the gas exchange processes. Engine gas exchange processes are characterized by Trans ., vol. 70, pp. 5-37, 1962. overall parameters such as volumetric efficiency (for four-stroke cycles), and scav- 14. Bolt, J. A ., and Holkeboer, D. H.: "Lean Fuel-Air Mixtures for High-Compression Spark. enging efficiency and trapping efficiency (for two-stroke cycles). These overall Ignition Engines," SAE Trans ., vol. 70, p. 195, 1962. parameters depend on the design of engine subsystems such as manifolds, valves, and ports, as well as engine operating conditions. Thus, the flow through individ- ual components in the engine intake and exhaust system has been extensively studied also. Supercharging and turbocharging are used to increase air flow through engines, and hence power density. Obviously, whether the engine is natu- rally aspirated or supercharged (or turbocharged) significantly affects the gas exchange processes. The above topics are the subject of this chapter. For spark-ignition engines, the fresh charge is fuel, air, and (if used for emission control) recycled exhaust, so mixture preparation is also an important 205 206 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 207 goal of the intake process. Mixture preparation includes both achieving the appropriate mixture composition and achieving equal distribution of air, fuel, and recycled exhaust amongst the different cylinders. In diesels, only air (or air * p plus recycled exhaust) is inducted. Mixture preparation and manifold flow Po Exhaust 4 (IVC) phenomena are discussed in Chap. 7. A third goal of the gas exchange processes 3 is to set up the flow field within the engine cylinders that will give a fast-enough (IVO) combustion process for satisfactory engine operation. In-cylinder flows are the APai subject of Chap. 8. TC Po, To BC P 2 6.1 INLET AND EXHAUST PROCESSES (EVC) Intake IN THE FOUR-STROKE CYCLE ( EVO) In a spark-ignition engine, the intake system typically consists of an air filter, a 3 carburetor and throttle or fuel injector and throttle or throttle with individual Length of intake system fuel injectors in each intake port, and intake manifold. During the induction APu process, pressure losses occur as the mixture passes through or by each of these components. There is an additional pressure drop across the intake port and valve. The exhaust system typically consists of an exhaust manifold, exhaust pipe, -Va often a catalytic converter for emission control, and a muffler or silencer. Figure ( APthr (b) 6-1 illustrates the intake and exhaust gas flow processes in a conventional spark- APvalve rc BC ignition engine. These flows are pulsating. However, many aspects of these flows -P- Ap can be analysed on a quasi-steady basis, and the pressures indicated in the intake system in Fig. 6-1a represent time-averaged values for a multicylinder engine. The drop in pressure along the intake system depends on engine speed, the (a) flow resistance of the elements in the system, the cross-sectional area through which the fresh charge moves, and the charge density. Figure 6-1d shows the inlet p. Lvt and exhaust valve lifts versus crank angle. The usual practice is to extend the O O valve open phases beyond the intake and exhaust strokes to improve emptying Po O and charging of the cylinders and make the best use of the inertia of the gases in Lvexh L ving the intake and exhaust systems. The exhaust process usually begins 40 to 60º (c) P before BC. Until about BC the burned cylinder gases are discharged due to the Qw pressure difference between the cylinder and the exhaust system. After BC, the cylinder is scavenged by the piston as it moves toward TC. The terms blowdown 2 and displacement are used to denote these two phases of the exhaust process. Typically, the exhaust valve closes 15 to 30º after TC and the inlet valve opens 10 to 20º before TC. Both valves are open during an overlap period, and when BC TC BC P:/Pe < 1, backflow of exhausted gas into the cylinder and of cylinder gases into (d) the intake will usually occur. The advantage of valve overlap occurs at high FIGURE 6-1 engine speeds when the longer valve-open periods improve volumetric efficiency. Intake and exhaust processes for four-stroke cycle spark-ignition engine: (a) intake system and aver- As the piston moves past TC and the cylinder pressure falls below the intake age pressures within it; (b) valve timing and pressure-volume diagrams; (c) exhaust system; (d) cylin- pressure, gas flows from the intake into the cylinder. The intake valve remains der pressure p and valve lift L, versus crank angle 0. Solid lines are for wide-open throttle, dashed lines for part throttle; po, To, atmospheric conditions; Ap.;, = pressure losses in air cleaner; Ap = open until 50 to 70º after BC so that fresh charge may continue to flow into the valve.1 intake losses upstream of throttle; Ap,hr = losses across throttle; APvalve = losses across the intake cylinder after BC. In a diesel engine intake system, the carburetor or EFI system and the throttle plate are absent. Diesel engines are more frequently turbocharged. A 208 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 209 4 6.2 VOLUMETRIC EFFICIENCY 3 Exhaust P Volumetric efficiency is used as an overall measure of the effectiveness of a four- stroke cycle engine and its intake and exhaust systems as an air pumping device. It is defined [see Sec. 2.10, Eq. (2.27)] as 2 mg Po, To 121 Intake no =- Pa.o VaN (6.1) Pil The air density Pa,o can be evaluated at atmospheric conditions; n, is then the 3 overall volumetric efficiency. Or it can be evaluated at inlet manifold conditions; „, then measures the pumping performance of the cylinder, inlet port, and valve Pe Po alone. This discussion will cover unthrottled (wide-open throttle) engine oper- ation only. It is the air flow under these conditions that constrains maximum engine power. Lesser air flows in SI engines are obtained by restricting the intake Pi T Vc Va system flow area with the throttle valve. Volumetric efficiency is affected by the following fuel, engine design, and engine operating variables: TC BC. 1. Fuel type, fuel/air ratio, fraction of fuel vaporized in the intake system, and fuel heat of vaporization 2. Mixture temperature as influenced by heat transfer 3. Ratio of exhaust to inlet manifold pressures Per Te 4. Compression ratio 5. Engine speed 6. Intake and exhaust manifold and port design FIGURE 6-2 Intake and exhaust process for turbocharged four-stroke cycle engine. The turbocharger compressor 7. Intake and exhaust valve geometry, size, lift, and timings C raises air pressure and temperature from ambient po, To to p ,, Ti. Cylinder pressure during intake is less than p ,. During exhaust, the cylinder gases flow through the exhaust manifold to the turbo- The effects of several of the above groups of variables are essentially quasi charger turbine T. Manifold pressure p. may vary during the exhaust process and lies between cylin- steady in nature; i.e ., their impact is either independent of speed or can be der pressure and ambient.1 described adequately in terms of mean engine speed. However, many of these variables have effects that depend on the unsteady flow and pressure wave phe- nomena that accompany the time-varying nature of the gas exchange processes. similar set of diagrams illustrating the intake and exhaust processes for a turbo- charged four-stroke diesel is shown in Fig. 6-2. When the exhaust valve opens, the burned cylinder gases are fed to a turbine which drives a compressor which 6.2.1 Quasi-Static Effects compresses the air prior to entry to the cylinder. VOLUMETRIC EFFICIENCY OF AN IDEAL CYCLE. For the ideal cycles of Fig. Due to the time-varying valve open area and cylinder volume, gas inertia 5-2d and e, an expression for volumetric efficiency can be derived which is a effects, and wave propagation in the intake and exhaust systems, the pressures in function of the following variables: intake mixture pressure pi, temperature Ti, the intake, the cylinder, and the exhaust during these gas exchange processes vary and fuel/air ratio (F/A); compression ratio re; exhaust pressure pe; and molecular in a complicated way. Analytical calculation of these processes is difficult (sce weight M and y for the cycle working fluid. The overall volumetric efficiency is Secs. 7.6.2 and 14.3 for a review of available methods). In practice, these processes are often treated empirically using overall parameters such as volumetric effi- no = - ma m(1 - x,) ciency to define intake and exhaust system performance. Pa,o Va Pa,o[1 + (F/A)] (rc - 1)VI GAS EXCHANGE PROCESSES 211 210 INTERNAL COMBUSTION ENGINE FUNDAMENTALS where m is the mass in the cylinder at point 1 in the cycle. Since For conventional liquid fuels such as gasoline the effect of fuel vapor, and therefore fuel/air ratio, is small. For gaseous fuels and for methanol vapor, the R R Pi V1 = m T1 and Pa,o = Pa.0 Ma Ta,o volumetric efficiency is significantly reduced by the fuel vapor in the intake mixture. and Eq. (5.38) relates Ti to Ti, the above expression for n, can be written 1 Ma Pas Ti ) [1 + (FLA] 76 -1 7. - D[(p:) +6- 1) (6.2) 1 FRACTION FUEL VAPORIZED, HEAT OF VAPORIZATION, AND HEAT no = TRANSFER. For a constant-pressure flow with liquid fuel evaporation and with heat transfer, the steady-flow energy equation is For (Pe/P.) = 1, the term in { } is unity. [maha + (1 - xe)mghs. L + xem,hsvla =@+(maha +mghs.L)B (6.4) EFFECT OF FUEL COMPOSITION, PHASE, AND FUEL/AIR RATIO. In a spark- ignition engine, the presence of gaseous fuel (and water vapor) in the intake where xe is the mass fraction evaporated and the subscripts denote: a, air proper- system reduces the air partial pressure below the mixture pressure. For mixtures ties; f, fuel properties; L, liquid; V, vapor; B before evaporation; A after evapo- of air, water vapor, and gaseous or evaporated fuel we can write the intake mani- ration. Approximating the change in enthalpy per unit mass of each component fold pressure as the sum of each component's partial pressure: of the mixture by c, AT, and with hr, y - hy. [ = hy. Ly (the enthalpy of vaporization), Eq. (6-4) becomes P: = Pat + Ps.: + Pw.i which with the ideal gas law gives TA - TA = (0/m) - x(F/A)hs. LV P: [+(MXM.) +(MXM.)] Cp. a + (F/A)CS. L (6.5) (6.3) Since cf, L ~ 2c, . the last term in the denominator can often be neglected. The water vapor correction is usually small (<0.03). This ratio, P .. :/p;, for several If no heat transfer to the inlet mixture occurs, the mixture temperature common fuels as a function of (m/m_) is shown in Fig. 6-3. Note that (m /ma) decreases as liquid fuel is vaporized. For complete evaporation of isooctane, with only equals the engine operating fuel/air ratio if the fuel is fully vaporized. ¢ = 1.0, T - T; = - 19ºC. For methanol under the same conditions, T - TR would be - 128ºC. In practice heating occurs; also, the fuel is not necessarily fully evaporated prior to entry to the cylinder. Experimental data show that the decrease in air temperature that accompanies liquid fuel evaporation more than 1.0 C8H18 offsets the reduction in air partial pressure due to the increased amount of fuel vapor: for the same heating rate, volumetric efficiency with fuel vaporization is C3 H8 higher by a few percent.2 The ideal cycle equation for volumetric efficiency [Eq. (6.2)] shows that the 0.9 CH4 effect of gas temperature variations, measured at entry to the cylinder, is through the factor (Ta,o/T;). Engine test data indicate that a square root dependence of Pa, i - Pi volumetric efficiency on temperature ratio is closer to real engine behavior. The CH3OH square root dependence is a standard assumption in engine test data reduction 0.8 (see Sec. 2.12). EFFECT OF INLET AND EXHAUST PRESSURE RATIO AND COMPRESSION 0.7 FIGURE 6-3 RATIO. As the pressure ratio (pe/Pi) and the compression ratio are varied, the H2 Effect of fuel (vapor) on inlet air partial pressure. fraction of the cylinder volume occupied by the residual gas at the intake pressure Ratio of air inlet pressure pa, to mixture inlet pres- varies. As this, volume increases so volumetric efficiency decreases. These effects sure p; versus fuel/air equivalence ratio o for iso- on ideal-cycle volumetric efficiency are given by the { } term in Eq. (6.2). For 0.6 0.5 1.0 1.5 octane vapor, propane, methane, methanol vapor, y = 1.3 these effects are shown in Fig. 6-4. Equivalence ratio ¢ and hydrogen. 212 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 213 Y = 1.3 geometric details and 1; is the local velocity. Assuming the flow is quasi-steady, ", is related to the mean piston speed S, by 0, A; = S,A, where A; and A, are the component minimum flow area and the piston area, respectively. Hence, the total quasi-steady pressure loss due to friction is 1.0 Patm - Pc = [ Ap, = [ 5,pu] - PS3 2 5, (At) (6.6) 24 Equation (6.6) indicates the importance of large component flow areas for reducing frictional losses, and the dependence of these losses on engine speed. 16 Figure 6-5 shows an example of the pressure losses due to friction across the air cleaner, carburetor, throttle, and manifold plenum of a standard four-cylinder 00 120 io Throttle 0 0.5 FIGURE 6-4 Air cleaner 1.0 1.5 2.0 Pe Effect of exhaust to inlet pressure ratio on Atmosphere ideal-cycle volumetric efficiency. 100 Di Patm 6.2.2 Combined Quasi-Static and Pp 80- Dynamic Effects Pr When gas flows unsteadily through a system of pipes, chambers, ports, and valves, both friction, pressure, and inertial forces are present. The relative impor- Ap, mmHg 60 tance of these forces depends on gas velocity and the size and shape of these passages and their junctions. Both quasi-steady and dynamic effects are usually significant. While the effects of changes in engine speed, and intake and exhaust manifold, port and valve design are interrelated, several separate phenomena 40 which affect volumetric efficiency can be identified. Patm - P FRICTIONAL LOSSES. During the intake stroke, due to friction in each part of 20 the intake system, the pressure in the cylinder pe is less than the atmospheric pressure Patm by an amount dependent on the square of the speed. This total pressure drop is the sum of the pressure loss in each component of the intake Patm - Pp system: air filter, carburetor and throttle, manifold, inlet port, and inlet valve. 0 Each loss is a few percent, with the port and valve contributing the largest drop. 20 40 60 80 As a result, the pressure in the cylinder during the period in the intake process when the piston is moving at close to its maximum speed can be 10 to 20 percent m, g/s 1200 lower than atmospheric. For each component in the intake (and the exhaust) 2400 3600 4800 system, Bernoulli's equation gives Engine speed, rev/min Ap; = 5,pv] FIGURE 6-5 Pressure losses in the intake system of a four-stroke cycle spark-ignition engine determined under where ¿, is the resistance coefficient for that component which depends on its steady flow conditions.3 Stroke = 89 mm. Bore = 84 mm. 214 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 215 automobile engine intake system. These steady flow tests, conducted over the full speed for a four-cylinder automobile spark-ignition engine.4 At high speeds and engine speed range,3 show that the pressure loss depends on speed squared. loads the exhaust manifold operates at pressures substantially above atmo- Equivalent flow-dependent pressure losses in the exhaust system result in spheric. the exhaust port and manifold having average pressure levels that are higher than atmospheric. Figure 6-6 shows the time-averaged exhaust manifold gauge pres- RAM EFFECT. The pressure in the inlet manifold varies during each cylinder's sure as a function of inlet manifold vacuum (which varies inversely to load) and intake process due to the piston velocity variation, valve open area variation, and the unsteady gas-flow effects that result from these geometric variations. The mass of air inducted into the cylinder, and hence the volumetric efficiency, is Inlet manifold vacuum, kPa 20 30 40 50 60 70 almost entirely determined by the pressure level in the inlet port during the short 0 10 period before the inlet valve is closed.5 At higher engine speeds, the inertia of the gas in the intake system as the intake valve is closing increases the pressure in the 30H Wide-open throttle port and continues the charging process as the piston slows down around BC and starts the compression stroke. This effect becomes progressively greater as engine speed is increased. The inlet valve is closed some 40 to 60º after BC, in part to take advantage of this ram phenomenon. 25 4000 rev/min REVERSE FLOW INTO THE INTAKE. Because the inlet valve closes after the 3600 start of the compression stroke, a reverse flow of fresh charge from the cylinder back into the intake can occur as the cylinder pressure rises due to piston motion 3200 toward TC. This reverse flow is largest at the lowest engine speeds. It is an inevi- table consequence of the inlet valve closing time chosen to take advantage of the 20 2800 ram effect at high speeds. TUNING. The pulsating flow from each cylinder's exhaust process sets up pres- sure waves in the exhaust system. These pressure waves propagate at the local sound speed relative to the moving exhaust gas. The pressure waves interact with Exhaust manifold pressure, kPa the pipe junctions and ends in the exhaust manifold and pipe. These interactions cause pressure waves to be reflected back toward the engine cylinder. In multi- cylinder engines, the pressure waves set up by each cylinder, transmitted through the exhaust and reflected from the end, can interact with each other. These pres- 10+ sure waves may aid or inhibit the gas exchange processes. When they aid the 2400 process by reducing the pressure in the exhaust port toward the end of the exhaust process, the exhaust system is said to be tuned.6 The time-varying inlet flow to the cylinder causes expansion waves to be 2000 propagated back into the inlet manifold. These expansion waves can be reflected at the open end of the manifold (at the plenum) causing positive pressure waves 1600 to be propagated toward the cylinder. If the timing of these waves is appropri- ately arranged, the positive pressure wave will cause the pressure at the inlet 1200 valve at the end of the intake process to be raised above the nominal inlet pres- sure. This will increase the inducted air mass. Such an intake system is described as tuned.6 OL 90 80 70 60 50 40 30 100 Inlet manifold pressure, kPa Methods which predict the unsteady flows in the intake and exhaust systems of internal combustion engines with good accuracy have been developed. FIGURE 6-6 Exhaust manifold pressure as a function of load (defined by inlet manifold vacuum) and speed, four- These methods are complicated, however, so more detailed discussion is deferred to Chap. 14. stroke cycle four-cylinder spark-ignition engine." 216 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 217 1200 rev/min 4800 rev/min 100 1.2[ IO PI 1.2[ P1 1.0 1.0 Diesel 0.8 90 IO 0.8L 1.8 1.6 P2 80 Spark-ignition Pressure, atm abs 1.4 1.2 EO FIGURE 6-8 Volumetric efficiency versus mean piston speed 1.4 1.0 2 4 6 10 12 for a four-cylinder automobile indirect-injection Mean piston speed, m/s P 3 diesel8 and a six-cylinder spark-ignition engine.9 1.2 P3 1.6 1.0 volumetric efficiency versus mean piston speed for a four-cylinder automobile 180 360 540 720 180 360 540 720 indirect-injection diesel engine8 and a six-cylinder spark-ignition engine.9 The Crank angle, deg Crank angle, deg volumetric efficiencies of spark-ignition engines are usually lower than diesel values due to flow losses in the carburetor and throttle, intake manifold heating, FIGURE 6-7 the presence of fuel vapor, and a higher residual gas fraction. The diesel curve Instantaneous pressures in the intake and exhaust manifolds of a four-stroke cycle four-cylinder spark-ignition engine, at wide-open throttle. Locations: p1, intake manifold runner 150 mm from with its double peak shows the effect of intake system tuning. cylinder 1; p2, exhaust manifold runner 200 mm from cylinder 1; p3, exhaust manifold runner The shape of these volumetric efficiency versus mean piston speed curves 700 mm from cylinder 1. IO and EO, intake and exhaust valve open periods for that cylinder, respec- can be explained with the aid of Fig. 6-9. This shows, in schematic form, how the tively.3 Stroke = 89 mm. Bore == 84 mm. 100% Examples of the pressure variations in the inlet and exhaust systems of a Quasi-static effects four-cylinder automobile spark-ignition engine at wide-open throttle are shown Charge heating in Fig. 6-7. The complexity of the phenomena that occur is apparent. The ampli- tude of the pressure fluctuations increases substantially with increasing engine B- speed. The primary frequency in both the intake and exhaust corresponds to the Flow friction Backflow frequency of the individual cylinder intake and exhaust processes. Higher har- monics that result from pressure waves in both the intake and exhaust are clearly Volumetric efficiency important also. Tuning Choking 6.2.3 Variation with Speed, and Valve Area, Lift, and Timing Flow effects on volumetric efficiency depend on the velocity of the fresh mixture Ram effect in the intake manifold, port, and valve. Local velocities for quasi-steady flow are equal to the volume flow rate divided by the local cross-sectional area. Since the intake system and valve dimensions scale approximately with the cylinder bore, Mean piston speed mixture velocities in the intake system will scale with piston speed. Hence, volu- FIGURE 6-9 metric efficiencies as a function of speed, for different engines, should be com- Effect on volumetric efficiency of different phenomena which affect the air flow rate as a function of pared at the same mean piston speed. Figure 6-8 shows typical curves of speed. Solid line is final n, versus speed curve. 218 INTERNAL COMBUSTION ENGINE FUNDAMENTALS effect on volumetric efficiency of each of the different phenomena described in this section varies with speed. Non-speed-dependent effects (such as fuel vapor 6000 pressure) drop ny below 100 percent (curve A). Charge heating in the manifold and cylinder drops curve A to curve B. It has a greater effect at lower engine speeds due to longer gas residence times. Frictional flow losses increase as the square of engine speed, and drop curve B to curve C. At higher engine speeds, the flow into the engine during at least part of the intake process becomes choked (see Sec. 6.3.2). Once this occurs, further increases in speed do not increase the 4000 flow rate significantly so volumetric efficiency decreases sharply (curve C to D). The induction ram effect, at higher engine speeds, raises curve D to curve E. Late inlet valve closing, which allows advantage to be taken of increased charging at Engine speed, rev/min (b) higher speeds, results in a decrease in nu at low engine speeds due to backflow (curves C and D to F). Finally, intake and/or exhaust tuning can increase the volumetric efficiency (often by a substantial amount) over part of the engine 45 60 19|10 4 2000 speed range, curve F to G. An example of the effect on volumetric efficiency of tuning the intake mani- Timing in 11 8.5 fold runner is shown in Fig. 6-10. In an unsteady flow calculation of the gas Lift, mm exchange processes of a four-cylinder spark-ignition engine, the length of the intake manifold runners was increased successively by factors of 2. The 34-cm length produces a desirable "tuned" volumetric efficiency curve with increased 80 70- 60- 90 low-speed air flow and flat mid-speed characteristics. While the longest runner Volumetric efficiency, % further increases low-speed air flow, the loss in n, at high speed would be unac- ceptable.10 Further discussion of intake system tuning can be found in Sec. 7.6.2. Effect of (a) valve timing and (b) valve lift on volumetric efficiency versus speed curves. Four-cylinder 1.6-dm3 displacement spark-ignition engine at wide-open throttle, firing conditions, (A/F) = 13, MBT ignition timing. Timing numbers are: inlet valve opens (before TC) top left, closes (after BC) Figure 6-11 shows data from a four-cylinder spark-ignition engine3 which illustrates the effect of varying valve timing and valve lift on the volumetric effi- 6000 ciency versus speed curve. Earlier-than-normal inlet valve closing reduces back- flow losses at low speed and increases n .. The penalty is reduced air flow at high 30|20 speed. Later-than-normal inlet valve closing is only advantageous at very high bottom left; exhaust valve opens (before BC) bottom right, closes (after TC) top right.3 Stroke == 89 mm. 15 50 45 60 70,60 19|10 10|0 . 4000 1.0 Timing 0.91 11 8.5 (a) Engine speed, rev/min Lift, mm Volumetric efficiency 0.7- 2000 0.6 - L = 8.5 cm -- L = 17 cm - is -- L = 34 cm L = 68 cm 80 FIGURE 6-11 90 FIGURE 6-10 70- 60 Effect of intake runner length on volumetric effi- T Volumetric efficiency, % 0 2000 4000 6000 ciency versus speed for 2.3-dm3 four-cylinder Speed, rev/min spark-ignition engine.1º 219 220 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 221 speeds. Low valve lifts significantly restrict engine breathing over the mid-speed and high-speed operating ranges. Above a critical valve lift, lift is no longer a 0.20-0.22D -- | major constraint on effective valve open area (see Sec. 6.3). 0.80-0.85D 200 400 6.3 FLOW THROUGH VALVES The valve, or valve and port together, is usually the most important flow 10.88-0.93D Minimum protrusion of guide boss restriction in the intake and the exhaust system of four-stroke cycle engines. The D characteristics of flows through poppet valves will now be reviewed. Largest possible radius 150 6.3.1 Poppet Valve Geometry and Timing 0.075-0.085D (30º seat) 0.085-0.095D (45º seat) Figure 6-12 shows the main geometric parameters of a poppet valve head and 1.10-1.12D (300) seat. Figure 6-13 shows the proportions of typical inlet and exhaust valves and 1.09-1.10D (45º) ports, relative to the valve inner seat diameter D. The inlet port is generally (a) circular, or nearly so, and the cross-sectional area is no larger than is required to achieve the desired power output. For the exhaust port, the importance of good valve seat and guide cooling, with the shortest length of exposed valve stem, leads Core close to bottom of valve guide 0.23-0.25D to a different design. Although a circular cross section is still desirable, a rec- tangular or oval shape is often essential around the guide boss area. Typical valve head sizes for different shaped combustion chambers in terms of cylinder Minimum or no guide protrusion bore B are given in Table 6.1.11 Each of these chamber shapes (see Secs. 10.2 and 15.4 for a discussion of spark-ignition and diesel combustion chamber design) imposes different constraints on valve size. Larger valve sizes (or four valves com- 0.90-1.0D| pared with two) allow higher maximum air flows for a given cylinder displace- < 0.90D 70.35D ment. Typical valve timing, valve-lift profiles, and valve open areas for a four- Section Z-Z Area > 0.75 area at 'D' stroke cycle spark-ignition engine are shown in Fig. 6-14. There is no universally 200 accepted criterion for defining valve timing points. Some are based upon a spe- 0.095-0.105D D' Core close to seat 1.10-1.11D ( b) FIGURE 6-13 Stem diameter Ds Shape, proportions, and critical design areas of typical inlet (top) and exhaust (bottom) valves and Inner seat diameter D ports. 11 Seat width w cific lift criterion. For example, SAE defines valve timing events based on refer- seat angle B ence valve-lift points:13 Lift Ly 1. Hydraulic lifters. Opening and closing positions are the 0.15-mm (0.006-in) FIGURE 6-12 valve-lift points. Head diameter, Dy Parameters defining poppet valve 2 Mechanical lifters. Valve opening and closing positions are the points of geometry. 0.15-mm (0.006-in) lift plus the specified lash. 222 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 223 TABLE 6.1 IVO TC EVC Valve head diameter in terms of cylinder bore B11. 16 /12/ Approximate mean Combustion piston speed, chamber shapet Inlet Exhaust max power, m/s x = 73 ATC max = 73 BTC max Wedge or bathtub 0.43-0.46B 0.35-0.37B 15 Bowl-in-piston 0.42-0.44B 0.34-0.37B 14 Exh t = 110 In t = 106 Hemispherical 0.48-0.5B 0.41-0.43B 18 Four-valve pent-roof 0.35-0.37B 0.28-0.32B 20 + See Fig. 15-15. 48 52 Ivc EVO BC Alternatively, valve events can be defined based on angular criteria along the lift ( a ) curve.12 What is important is when significant gas flow through the valve-open area either starts or ceases. 2 3 The instantaneous valve flow area depends on valve lift and the geometric . Dp details of the valve head, seat, and stem. There are three separate stages to the flow area development as valve lift increases,14 as shown in Fig. 6-14b. For low valve lifts, the minimum flow area corresponds to a frustrum of a right circular cone where the conical face between the valve and the seat, which is perpendicu lar to the seat, defines the flow area. For this stage: 5 $ 0.123 2 0.125 4DL, 2 D'3 - D; W (b) sin B cos B >L, >0 and the minimum area is Valve lift, mm ON - 865 Am = IL, cos B( D. - 2w + 2º sin 23 (6.7) 10- 1-4-24 3 ---- 2 --- 1 --- where ß is the valve seat angle, L, is the valve lift, D, is the valve head diameter (the outer diameter of the seat), and w is the seat width (difference between the Intake inner and outer seat radii). For the second stage, the minimum area is still the slant surface of a frus- Flow area, cm2 trum of a right circular cone, but this surface is no longer perpendicular to the valve seat. The base angle of the cone increases from (90 - 8)º toward that of a -142 - 3 - -42-1- cylinder, 90º. For this stage: A Exhaust (Di Di) -W27 1/2 + w tan B > L , > - W 60 40. 20 0 -20 -40 -60 sin B cos B Camshaft angle, deg (c) and (6.8) FIGURE 6-14 Am = RDm [ ( L , - w tan B)2 + w2] 1/2 (a) Typical valve timing diagram for high-speed 2.2-dm3 four-cylinder spark-ignition engine. (b) Sche- where D, is the port diameter, D, is the valve stem diameter, and Dm, is the mean matic showing three stages of valve lift. (c) Valve-lift curve and corresponding minimum intake and exhaust valve open areas as a function of camshaft angle. Inlet and exhaust valve diameters are 3.6 seat diameter (Du - w). and 3.1 cm, respectively.12 224 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 225 Finally, when the valve lift is sufficiently large, the minimum flow area is no 5.0 longer between the valve head and seat; it is the port flow area minus the section- 4.0F al area of the valve stem. Thus, for 3.0 Valve area Am, cm2 ONAGE 2.0 -- Exhaust Intake + w tan B 3ª 1.0- then (44) max A = " (D: - D3) 0.8 / max (6.9) Lvmax $ 0.7- 0.6 L vmax Intake and exhaust valve open areas corresponding to a typical valve-lift ៛0.5- profile are plotted versus camshaft angle in Fig. 6-14c. These three different flow 0.4- regimes are indicated. The maximum valve lift is normally about 12 percent of 0.3- Exhaust Intake the cylinder bore. Inlet valve opening (IVO) typically occurs 10 to 25º BTC. Engine per- 0.1- formance is relatively insensitive to this timing point. It should occur sufficiently 180 140 100 60 20 0 20 ₣ 60 100 140 before TC so that cylinder pressure does not dip early in the intake stroke. Inlet 180 Crank angle from TC, deg valve closing (IVC) usually falls in the range 40 to 60º after BC, to provide more time for cylinder filling under conditions where cylinder pressure is below the FIGURE 6-15 intake manifold pressure at BC. IVC is one of the principal factors that deter- Rate of change of cylinder volume dV/de, valve minimum flow area A ,, and pseudo flow velocity as function of crank angle for exhaust and inlet valves of Fig. 6-14.12 mines high-speed volumetric efficiency; it also affects low-speed volumetric effi- ciency due to backflow into the intake (see Sec. 6.2.3). Exhaust valve opening (EVO) occurs 50 to 60º before BC, well before the end of the expansion stroke, so between the wrist pin and crank axis [see Fig. 2-1 and Eq. (2.5)] and Am, is the that blowdown can assist in expelling the exhaust gases. The goal here is to valve area given by Eqs. (6.7), (6.8), or (6.9). Instantaneous pseudo flow velocity reduce cylinder pressure to close to the exhaust manifold pressure as soon as profiles for the exhaust and intake strokes of a four-stroke four-cylinder engine possible after BC over the full engine speed range. Note that the timing of EVO are shown in Fig. 6-15. Note the appearance of two peaks in the pseudo flow affects the cycle efficiency since it determines the effective expansion ratio. velocity for both the exhaust and intake strokes. The broad peaks occurring at Exhaust valve closing (EVC) ends the exhaust process and determines the dura- maximum piston velocity reflect the fact that valve flow area is constant at this tion of the valve overlap period. EVC typically falls in the range 8 to 20º after point. The peaks close to TC result from the exhaust valve closing and intake TC. At idle and light load, in spark-ignition engines (which are throttled), it valve opening profiles. The peak at the end of the exhaust stroke is important therefore regulates the quantity of exhaust gases that flow back into the com- since it indicates a high pressure drop across the valve at this point, which will bustion chamber through the exhaust valve under the influence of intake mani- result in higher trapped residual mass. The magnitude of this exhaust stroke fold vacuum. At high engine speeds and loads, it regulates how much of the pseudo velocity peak depends strongly on the timing of exhaust valve closing. cylinder burned gases are exhausted. EVC timing should occur sufficiently far The pseudo velocity peak at the start of the intake stroke is much less important. after TC so that the cylinder pressure does not rise near the end of the exhaust That the pseudo velocities early in the exhaust stroke and late in the intake stroke. Late EVC favors high power at the expense of low-speed torque and idle stroke are low indicates that phenomena other than quasi-steady flow govern the combustion quality. Note from the timing diagram (Fig. 6-14a) that the points of flow rate. These are the periods when exhaust blowdown and ram and tuning maximum valve lift and maximum piston velocity (Fig. 2-2) do not coincide. effects in the intake are most important. The effect of valve geometry and timing on air flow can be illustrated con- ceptually by dividing the rate of change of cylinder volume by the instantaneous 6.3.2 Flow Rate and Discharge minimum valve flow area to obtain a pseudo flow velocity for each valve:12 Coefficients 1 dv RB2 ds Ups = - (6.10) The mass flow rate through a poppet valve is usually described by the equation Am de . 4Am de for compressible flow through a flow restriction [Eqs. (C.8) or (C.9) in App. C]. where V is the cylinder volume [Eq. (2.4)], B is the cylinder bore, s is the distance This equation is derived from a one-dimensional isentropic flow analysis, and 226 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 227 real gas flow effects are included by means of an experimentally determined dis- 0.8 Ac AE charge coefficient CD. The air flow rate is related to the upstream stagnation Flow pattern pressure po and stagnation temperature To, static pressure just downstream of b -c the flow restriction (assumed equal to the pressure at the restriction, pr), and a ).6- reference area AR characteristic of the valve design: Discharge coefficient im = CD AR PO (RT.) 1/2 (6.11) 0.4 When the flow is choked, i.e ., PT/Po < [2/(y + 1)]"(x-1), the appropriate equation 0 4 . 2 0.1 0.3 is im = - CD AR PO y y1/2/ 2 )(7+ 1 )/2 (y - 1 ) (RT0)1/2 ( 2 + 1 (6.12) (a) (b) (c) For flow into the cylinder through an intake valve, po is the intake system pres- sure pi and pr is the cylinder pressure. For flow out of the cylinder through an exhaust valve, po is the cylinder pressure and pr is the exhaust system pressure. The value of C, and the choice of reference area are linked together: their product, CD AR, is the effective flow area of the valve assembly AF. Several differ- ent reference areas have been used. These include the valve head area zD3/4,7 the port area at the valve seat zD3/4,15 the geometric minimum flow area [Eqs. (6.7), FIGURE 6-16 (6.8), and (6.9)], and the curtain area RD, L ,, 16 where L, is the valve lift. The Discharge coefficient of typical inlet poppet valve (effective flow area/valve curtain area) as a function of valve lift. Different segments correspond to flow regimes indicated.16 choice is arbitrary, though some of these choices allow easier interpretation than others. As has been shown above, the geometric minimum flow area is a complex function of valve and valve seat dimensions. The most convenient reference area steady flow operation. Also, as has been discussed in Sec. 6.2.2, the pressure up- in practice is the so-called valve curtain area: stream of the valve varies significantly during the intake process. However, it has Ac = nD. L. (6.13) been shown that over the normal engine speed range, steady flow discharge- coefficient results can be used to predict dynamic performance with reasonable since it varies linearly with valve lift and is simple to determine. precision. 14, 18 In addition to valve lift, the performance of the inlet valve assembly is influ- INLET VALVES. Figure 6-16 shows the results of steady flow tests on a typical. enced by the following factors: valve seat width, valve seat angle, rounding of the inlet valve configuration with a sharp-cornered valve seat.16 The discharge coeffi- seat corners, port design, cylinder head shape. In many engine designs the port cient based on valve curtain area is a discontinuous function of the valve-lift/ and valve assembly are used to generate a rotational motion (swirl) inside the diameter ratio. The three segments shown correspond to different flow regimes as engine cylinder during the induction process, or the cylinder head can be shaped indicated. At very low lifts, the flow remains attached to the valve head and seat, to restrict the flow through one side of the valve open area to generate swirl. giving high values for the discharge coefficient. At intermediate lifts, the flow Swirl production is discussed later, in Section 8.3. Swirl generation significantly separates from the valve head at the inner edge of the valve seat as shown. An reduces the valve (and port) flow coefficient. Changes in seat width affect the abrupt decrease in discharge coefficient occurs at this point. The discharge coeffi- Le/DD at which the shifts in flow regimes illustrated in Fig. 6-16 occur. CD cient then increases with increasing lift since the size of the separated region increases as seat width decreases. The seat angle ß affects the discharge coefficient remains approximately constant while the minimum flow area is increasing. At in the low-lift regime in Fig. 6-16. Rounding the upstream corner of the valve seat high lifts, the flow separates from the inner edge of the valve seat as well.14. reduces the tendency of the flow to break away, thus increasing CD at higher lifts. Typical maximum values of L./D, are 0.25. At low valve lifts, when the flow remains attached, increasing the Reynolds An important question is whether these steady flow data are representative number decreases the discharge coefficient. Once the flow breaks away from the of the dynamic flow behavior of the valve in an operating engine. There is some wall, there is no Reynolds number dependence of C ,. 16 evidence that the "change points" between different flow regimes shown in Fig. For well-designed ports (e.g ., Fig. 6-13) the discharge coefficient of the port 6-16 occur at slightly different valve lifts under dynamic operation than under and valve assembly need be no lower than that of the isolated valve (except when .. 228 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 229 the port is used to generate swirl). However, if the cross-sectional area of the port is not sufficient or the radius of the surface at the inside of the bend is too small, a significant reduction in CD for the assembly can result.16 At high engine speeds, unless the inlet valve is of sufficient size, the inlet flow during part of the induction process can become choked (i.e ., reach sonic velocity at the minimum valve flow area). Choking substantially reduces volu- metric efficiency. Various definitions of inlet Mach number have been used to identify the onset of choking. Taylor and coworkers7 correlated volumetric effi- ciencies measured on a range of engine and inlet valve designs with an inlet Mach index Z formed from an average gas velocity through the inlet valve: Z = ApSe C; A;a (6.14) Low lift High lift where A; is the nominal inlet valve area (zD'/4), C; is a mean valve discharge FIGURE 6-17 coefficient based on the area A;, and a is the sound speed. From the method used Flow pattern through exhaust valve at low and high lift.16 to determine C;, it is apparent that C; A; is the average effective open area of the valve (it is the average value of CDAD, Ly). Z corresponds closely, therefore, to At high lifts, Ly/D, 2 0.2, the breakaway of the flow reduces the discharge coeffi- the mean Mach number in the inlet valve throat. Taylor's correlations show that cient. (At Lu/D, = 0.25 the effective area is about 90 percent of the minimum n, decreases rapidly for Z _ 0.5. An alternative equivalent approach to this geometric area. For Lo/D, < 0.2 it is about 95 percent.16) The port design signifi- problem has been developed, based on the average flow velocity through the cantly affects Cp at higher valve lifts, as indicated by the data from four port valve during the period the valve is open.19 A mean inlet Mach number was designs in Fig. 6-18. Good designs can approach the performance of isolated defined : Mi = Di 1.0 a (6.15) where D; is the mean inlet flow velocity during the valve open period. M; is Isolated valve, sharp corners related to Z via - X- Z(no/ 100)180 (6.16) OIVC - OIVO This mean inlet Mach number correlates volumetric efficiency characteristics Discharge coefficient Cp 0.5 better than the Mach index. For a series of modern small four-cylinder engines, when M; approaches 0.5 the volumetric efficiency decreases rapidly. This is due to the flow becoming choked during part of the intake process. This relationship can be used to size the inlet valve for the desired volumetric efficiency at maximum engine speed. Also, if the inlet valve is closed too early, volumetric b efficiency will decrease gradually with increasing M;, for M < 0.5, even if the d valve open area is sufficiently large.19 0. 0.2 0.3 EXHAUST VALVES. In studies of the flow from the cylinder through an exhaust Lv poppet valve, different flow regimes at low and high lift occur, as shown in Fig. Dy 6-17. Values of CD based on the valve curtain area, for several different exhaust FIGURE 6-18 valve and port combinations, are given in Fig. 6-18. A sharp-cornered isolated Discharge coefficient as function of valve lift for several exhaust valve and port designs.16 a,2º b,15 poppet valve (i.e ., straight pipe downstream, no port) gives the best performance. c,20 d.21 230 GAS EXCHANGE PROCESSES 231 INTERNAL COMBUSTION ENGINE FUNDAMENTALS valves, however. Exhaust valves operate over a wide range of pressure ratios (1 to The residual gas mass fraction x, (or burned gas fraction if EGR is used) is 5). For pressure ratios greater than about 2 the flow will be choked, but the effect usually determined by measuring the CO2 concentration in a sample of gas of pressure ratio on discharge coefficient is small and confined to higher lifts (e.g ., extracted from the cylinder during the compression stroke. Then +5 percent at Ly/Dy = 0.3).15 XX = (co2)c (6.17) (&co2)e 6.4 RESIDUAL GAS FRACTION where the subscripts C and e denote compression and exhaust, and xco2 are mole The residual gas fraction in the cylinder during compression is determined by the fractions in the wet gas. Usually CO2 volume or mole fractions are measured in exhaust and inlet processes. Its magnitude affects volumetric efficiency and engine dry gas streams (see Sec. 4.9). A correction factor K, performance directly, and efficiency and emissions through its effect on working- fluid thermodynamic properties. The residual gas fraction is primarily a function K = (1)wer (xDary (6.18) of inlet and exhaust pressures, speed, compression ratio, valve timing, and (1)ary 1 + 0.5[y(xco, + x80) - 0.74x20] exhaust system dynamics. where y is the molar hydrogen/carbon ratio of the fuel and xto ,, xco are dry mole fractions, can be used to convert the dry mole fraction measurements. 20 T 20 Residual gas measurements in a spark-ignition engine are given in Fig. 6-19, which shows the effect of changes in speed, valve overlap, compression ratio, and 45 air/fuel ratio for a range of inlet manifold pressures for a 2-dm3, 88.5-mm bore, 15 15 32 four-cylinder engine.22 The effect of variations in spark timing was negligible. 27º Inlet pressure, speed, and valve overlap are the most important variables, though 10 -- 10- the exhaust pressure also affects the residual fraction.23 Normal settings for inlet 1000 valve opening (about 15º before TC) and exhaust valve closing (about 12º after Valve overlap 1400 1800 TC) provide sufficient overlap for good scavenging, but avoid excessive backflow 5 rev/min from the exhaust port into the cylinder. Residual gas fractions in diesel engines are substantially lower than in SI 0 0 300 400 500 600 700 300 400 500 600 700 engines because inlet and exhaust pressures are comparable in magnitude and the Manifold pressure, mmHg abs Manifold pressure, mmHg abs compression ratio is 2 to 3 times as large. Also, a substantial fraction of the Residual gas fraction, % residual gas is air. 20 20 6.5 EXHAUST GAS FLOW RATE AND 15 15 36 22 TEMPERATURE VARIATION O The exhaust gas mass flow rate and the properties of the exhaust gas vary signifi- 10 10 8.5 cantly during the exhaust process. The origin of this variation for an ideal 9.5 exhaust process is evident from Fig. 5-3. The thermodynamic state (pressure, tem- 5 50 10.5 64 perature, etc.) of the gas in the cylinder varies continually during the exhaust blowdown phase, until the cylinder pressure closely approaches the exhaust man- ifold pressure. In the real exhaust process, the exhaust valve restricts the flow out 600 700 '10 12 14 16 18 300 400 500 of the cylinder, the valve lift varies with time, and the cylinder volume changes Manifold pressure, mmHg abs Air/fuel ratio during the blowdown process, but the principles remain the same. FIGURE 6-19 Measurements have been made of the variation in mass flow rate through Residual gas fraction for 2-dm3 four-cylinder spark-ignition engine as a function of intake manifold the exhaust valve and gas temperature at the exhaust port exit during the exhaust pressure for a range of speeds, compression ratios, and valve overlaps: also as a function of air/fuel process of a spark-ignition engine.24 Figure 6-20 shows the instantaneous mass ratio for a range of volumetric efficiencies. Operating conditions, unless noted: speed = 1400 rev/min. flow rate data at three different engine speeds. The blowdown and displacement A/F = 14.5, spark timing set to give 0.95 maximum torque, compression ratio = 8.5.22 232 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 233 50 M >1-1-M<1 Data rev/min O 1200 -- 1500 me. E's 1800 A ---- 1800 40 Model -- 1200 30+ Pc 250 Mass flow rate of exhaust gas, g/s Exhaust Exhaust 20 valve valve opens closes 1200- Tc 200 10 1000 Pc, KPa BC OD T . K 100 140 180 220 260 300 340 TC 20 -150 Crank angle, deg. 800 FIGURE 6-20 Instantaneous mass flow rate of exhaust gas through the valve versus crank angle: equivalence ratio = 1.2, wide-open throttle, compression ratio = 7. Dash-dot line is one-dimensional compressible -100 flow model for blowdown and incompressible displacement model for exhaust stroke.24 600 EVC T EVO 120 220 320 420 phases of the exhaust process are evident. Simple quasi-steady models of these Crank angle phases give good agreement with the data at lower engine speeds. The blowdown FIGURE 6-21 model shown applies orifice flow equations to the flow across the exhaust valve Measured cylinder pressure pe, calculated cylinder-gas temperature Te, exhaust mass flow rate m ., using the measured cylinder pressure and estimated gas temperature for upstream and measured gas temperature at exhaust port exit T ,, for single-cylinder spark-ignition engine. stagnation conditions. Equation (C.9) is used when the pressure ratio across the Speed = 1000 rev/min, imep = 414 kPa, equivalence ratio = 1.2, spark timing = 10º BTC, r = 7.2.25 valve exceeds the critical value. Equation (C.8) is used when the pressure ratio is less than the critical value. The displacement model assumes the gas in the cylin- der is incompressible as the piston moves through the exhaust stroke. As engine der walls. The gas temperature at the port exit at the start of the exhaust flow speed increases, the crank angle duration of the blowdown phase increases. There pulse is a mixture of hotter gas which has just left the cylinder and cooler gas is evidence of dynamic effects occurring at the transition between the two phases. which left the cylinder at the end of the previous exhaust process and has been The peak mass flow rate during blowdown does not vary substantially with speed stationary in the exhaust port while the valve has been closed. The port exit since the flow is choked. The mass flow rate at the time of maximum piston speed temperature has a minimum where the transition from blowdown flow to dis- during displacement scales approximately with piston speed. As the inlet mani- placement occurs, and the gas comes momentarily to rest and loses a substantial fold pressure is reduced below the wide-open throttle value, the proportion of the fraction of its thermal energy to the exhaust port walls. charge which exits the cylinder during the blowdown phase decreases but the Figure 6-22 shows the effect of varying load and speed on exhaust port exit mass flow rate during displacement remains essentially constant. temperatures. Increasing load (A -> B-> C) increases the mass and temperature in The exhaust gas temperature varies substantially through the exhaust the blowdown pulse. Increasing speed (B -> D) raises the gas temperature process, and decreases due to heat loss as the gas flows past the exhaust valve throughout the exhaust process. These effects are the result of variations in the and through the exhaust system. relative importance of heat transfer in the cylinder and heat transfer to the Figure 6-21 shows the measured cylinder pressure, calculated cylinder gas exhaust valve and port. The time available for heat transfer, which depends on temperature and exhaust mass flow rate, and measured gas temperature at the engine speed and exhaust gas flow rate, is the most critical factor. The exhaust exhaust port exit for a single-cylinder spark-ignition engine at mid-load and low temperature variation with equivalence ratio follows from the variation in expan- speed.25 The average cylinder-gas temperature falls rapidly during blowdown, sion stroke temperatures, with maximum values at o = 1.0 and lower values for and continues to fall during the exhaust stroke due to heat transfer to the cylin- leaner and richer mixtures.24 Diesel engine exhaust temperatures are significantly 234 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 235 1000 1000 g 6.6 SCAVENGING IN TWO-STROKE EVO EVO CYCLE ENGINES 800₺ 800 6.6.1 Two-Stroke Engine Configurations 600 500 In two-stroke cycle engines, each outward stroke of the piston is a power stroke. To achieve this operating cycle, the fresh charge must be supplied to the engine cylinder at a high-enough pressure to displace the burned gases from the previous D C 1000 cycle. Raising the pressure of the intake mixture is done in a separate pump or 1000 blower or compressor. The operation of clearing the cylinder of burned gases and filling it with fresh mixture (or air)-the combined intake and exhaust process-is 800 800 called scavenging. However, air capacity is just as important as in the four-stroke cycle; usually, a greater air mass flow rate must be achieved to obtain the same 600: 600 output power. Figures 1-12, and 1-5 and 1-24 show sectioned drawings of a two-stroke spark-ignition engine and two two-stroke diesels. 120 220 320 420 120 220 320 420 The different categories of two-stroke cycle scavenging flows and the port Crank angle (and valve) arrangements that produce them are illustrated in Figs. 6-23 and 6-24. FIGURE 6-22 Scavenging arrangements are classified into: (a) cross-scavenged, (b) loop- Measured gas temperature at exhaust port exit as a function of crank angle, single-cylinder spark- scavenged, and (c) uniflow-scavenged configurations. The location and orientation ignition engine, for different loads and speeds. Curve A: imep = 267 kPa, 1000 rev/min; curve B: of the scavenging ports control the scavenging process, and the most common imep = 414 kPa, 1000 rev/min; curve C: imep = 621 kPa, 1000 rev/min; curve D: imep = 414 kPa, arrangements are indicated. Cross- and loop-scavenging systems use exhaust and 1600 rev/min. Equivalence ratio = 1.2, spark timing = 10º BTC, compression ratio = 7.2.25 inlet ports in the cylinder wall, uncovered by the piston as it approaches BC.27 The uniflow system may use inlet ports with exhaust valves in the cylinder head, lower than spark-ignition engine exhaust temperatures because of the lean oper- ating equivalence ratio and their higher expansion ratio during the power stroke. The average exhaust gas temperature is an important quantity for deter- mining the performance of turbochargers, catalytic converters, and particulate traps. The time-averaged exhaust temperature does not correspond to the average energy of the exhaust gas because the flow rate varies substantially. An enthalpy-averaged temperature EVC T. = ( EVC JEVO inc, To de ) /(J JEVO inc ; de (6.19) is the best indicator of exhaust thermal energy. Average exhaust gas temperatures are usually measured with a thermocouple. Thermocouple-averaged temperatures 00.0000 are close to time-averaged temperatures. Mass-averaged exhaust temperatures (which are close to Th, if c, variations are small) for a spark-ignition engine at the exhaust port exit are about 100 K higher than time-averaged or thermocouple- determined temperatures. Mass-average temperatures in the cylinder during the (a) (b) exhaust process are about 200 to 300 K higher than mass-averaged port tem- (c) peratures. All these temperatures increase with increasing speed, load, and spark FIGURE 6-23 retard, with speed being the variable with the largest impact.26 (a) Cross-scavenged, (b) loop-scavenged, and (c) uniflow-scavenged two-stroke cycle flow configu- rations. 236 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 237 Intake Exhaust BC P (a) (b) (c) mo Scavenging Asc Asc Pc PC Po Po Vsc 90 180 90 FIGURE 6-24 Common porting arrangements that go with (a) cross-scavenged, (b) loop-scavenged, and (c) uniflow- Exhaust scavenged configurations. blowdown LV Exhaust scavenging or inlet and exhaust ports with opposed pistons. Despite the different flow pat- terns obtained with each cylinder geometry, the general operating principles are -Air from compressor PC similar. Air in a diesel, or fuel-air mixture in a spark-ignition engine, must be (a) (b) supplied to the inlet ports at a pressure higher than the exhaust system pressure. FIGURE 6-25 Figure 6-25 illustrates the principles of the scavenging process for a uniflow Gas exchange process in two-stroke cycle uniflow-scavenged diesel engine: (a) valve and port timing engine design. Between 100 and 110º after TC, the exhaust valve opens and a and pressure-volume diagram; (b) pressure, scavenging port open area A ,, and exhaust valve lift L. blowdown discharge process commences. Initially, the pressure ratio across the as functions of crank angle.1 exhaust valve exceeds the critical value (see App. C) and the flow at the valve will be sonic: as the cylinder pressure decreases, the pressure ratio drops below the similar sequence of events for a loop-scavenged engine. Proper flow patterns for critical value. The discharge period up to the time of the scavenging port opening the fresh charge are extremely important for good scavenging and charging of the is called the blowdown (or free exhaust) period. The scavenging ports open cylinder. between 60 and 40º before BC when the cylinder pressure slightly exceeds the scavenging pump pressure. Because the burned gas flow is toward the exhaust Common methods for supercharging or pressurizing the fresh charge are shown in Fig. 6-26. In large two-stroke cycle engines, more complex com- valves, which now have a large open area, the exhaust flow continues and no binations of these approaches are often used, as shown in Fig. 1-24. backflow occurs. When the cylinder pressure falls below the inlet pressure, air enters the cylinder and the scavenging process starts. This flow continues as long as the inlet ports are open and the inlet total pressure exceeds the pressure in the 6.6.2 Scavenging Parameters and Models cylinder. As the cylinder pressure rises above the exhaust pressure, the fresh charge flowing into the cylinder displaces the burned gases: a part of the fresh The following overall parameters are used to describe the scavenging process. 13 The delivery ratio A: charge mixes with the burned gases and is expelled with them. The exhaust valves usually close after the inlet ports close. Since the flow in the cylinder is toward mass of delivered air (or mixture) per cycle the exhaust valve, additional scavenging is obtained. Figure 1-16 illustrates the reference mass (6.20) 238 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 239 The charging efficiency nch: mass of delivered air (or mixture) retained Ich = displaced volume x ambient density (6.24) O indicates how effectively the cylinder volume has been filled with fresh air (or mixture). Charging efficiency, trapping efficiency, and delivery ratio are related by nch = Anu (6.25) O When the reference mass in the definition of delivery ratio is the trapped cylinder mass my (or closely approximated by it) then nsc = Ants (6.26) FIGURE 6-26 In real scavenging processes, mixing occurs as the fresh charge displaces the Common methods of pressurizing the fresh charge in two-stroke cycle engines: left, crankcase com- burned gases and some of the fresh charge may be expelled. Two limiting ideal pression; center, roots blower; right, centrifugal compressor.7 models of this process are: (1) perfect displacement and (2) complete mixing. Perfect displacement or scavenging would occur if the burned gases were pushed compares the actual scavenging air mass (or mixture mass) to that required out by the fresh gases without any mixing. Complete mixing occurs if entering in an ideal charging process.+ The reference mass is defined as displaced fresh mixture mixes instantaneously and uniformly with the cylinder contents. volume x ambient air (or mixture) density. Ambient air (or mixture) density is For perfect displacement (with my as the reference mass in the delivery determined at atmospheric conditions or at intake conditions. This definition is ratio), useful for experimental purposes. For analytical work, it is often convenient to nsc = A and ntr = 1 for A < 1 use the trapped cylinder mass my as the reference mass. nsc = 1 and for A > 1 (6.27) The trapping efficiency nur: mass of delivered air (or mixture) retained For the complete mixing limit, consider the scavenging process as a quasi- (6.21) steady flow process. Between time t and t + dt, a mass element dmad of air is mass of delivered air (or mixture) delivered to the cylinder and is uniformly mixed throughout the cylinder volume. indicates what fraction of the air (or mixture) supplied to the cylinder is retained An equal amount of fluid, with the same proportions of air and burned gas as the in the cylinder. cylinder contents at time t, leaves the cylinder during this time interval. Thus the The scavenging efficiency nsc: mass of air delivered between t and t + dt which is retained, dm ., , is given by mass of delivered air (or mixture) retained nsc = (6.22) dmar = dmax( 1 - max mass of trapped cylinder charge indicates to what extent the residual gases in the cylinder have been replaced with Assuming my is constant, this integrates over the duration of the scavenging fresh air. process to give The purity of the charge: mar = 1 - exp (6.28) my Purity == mass of air in trapped cylinder charge (6.23 mass of trapped cylinder charge Thus, for complete mixing, with the above definitions, indicates the degree of dilution, with burned gases, of the unburned mixture in nsc = 1 -e-A the cylinder. (6.29) 1 , (1 - e-4) If scavenging is done with fuel-air mixture, as in spark-ignition engines, then mixture mass is used Figure 6-27 shows nge and nu for the perfect displacement and complete mixing instead of air mass. assumptions as a function of A, the delivery ratio. 240 INTERNAL COMBUSTION ENGINE FUNDAMENTALS Perfect displacement - Perfect mixing 1.0K- 1sc - 1sc and nur 175º 1.0 2.0 170 Delivery ratio, A FIGURE 6-27 Scavenging efficiency n, and trapping efficiency nu versus delivery ratio A for perfect displacement and complete mixing models. An additional possibility is the direct flow of fresh mixture through the cylinder into the exhaust without entraining burned gases. This is called short- 165º 175 º 195º circuiting; it is obviously undesirable since some fresh air or mixture is wasted. Photos of one fluid scavenging another in liquid analog experiments in model loop-scavenged engine cylinder. Top two rows: View perpen- There is no simple model for this process. When short-circuiting occurs, lower scavenging efficiencies result even though the volume occupied by the short- circuiting flow through the cylinder does displace an equal volume of the burned gases. Another phenomenon which reduces scavenging efficiency is the formation of pockets or dead zones in the cylinder volume where burned gases can become dicular to scavenging loop. Bottom row: Orthogonal views. Dark denser fluid displacing light less dense fluid.29 trapped and escape displacement or entrainment by the fresh scavenging flow. 160º 170º 190 These unscavenged zones are most likely to occur in regions of the cylinder that remain secluded from the main fresh mixture flow path. 6.6.3 Actual Scavenging Processes Several methods have been developed for determining what occurs in actual cylinder scavenging processes.2 Accurate measurement of scavenging efficiency is 155º difficult due to the problem of measuring the trapped air mass. Estimation of 11sc 185 1650 from indicated mean effective pressure and from gas sampling are the most reli- able methods.7 Flow visualization experiments28-30 in liquid analogs of the cylinder and flow velocity mapping techniques31 have proved useful in providing a qualitative picture of the scavenging flow field and identifying problems such as short-circuiting and dead volumes. Flow visualization studies indicate the key features of the scavenging FIGURE 6-28 process. Figure 6-28 shows a sequence of frames from a movie of one liquid 150€ 180 º 160 º scavenging another in a model of a large two-stroke cycle loop-scavenged 241 242 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 243 diesel.29 The physical variables were scaled to maintain the same values of the appropriate dimensionless numbers for the liquid analog flow and the real engine flow. The density of the liquid representing air (which is dark) was twice the density of the liquid representing burned gas (which is clear). Early in the scav. enging process, the fresh air jets penetrate into the burned gas and displace it first toward the cylinder head and then toward the exhaust ports (the schematic gives the location of the ports). During this initial phase, the outflowing gas contains no air; pure displacement of the burned gas from the cylinder is being achieved. Then short-circuiting losses start to occur, due to the damming-up or buildup of fresh air on the cylinder wall opposite the exhaust ports. The short-circuiting fluid flows directly between the scavenge ports and the exhaust ports above them. Since this damming-up of the inflowing fresh air back toward the exhaust ports continues, short-circuiting losses will also continue. While the scavenging front remains distinct as it traverses the cylinder, its turbulent character indicates that mixing between burned gas and air across the front is taking place. For both these reasons (short-circuiting and short-range mixing), the outflowing gas, once FIGURE 6-29 the "displacement" phase is over, contains an increasing amount of fresh air. Desirable air flow in loop-scavenged engine: air from the entering jets impinges on far cylinder wall and flows toward the cylinder head.31 Outflowing fluid composition measurements from this model study of a Sulzer two-stroke loop-scavenged diesel engine confirm this sequence of events. At 24 crank angle degrees after the onset of scavenging, fresh air due to short- scavenging jets enter symmetrically with sufficient velocity to fill up about half circuiting was detected in the exhaust. At the time the displacement front reached the cylinder cross section, and thereafter flow at lower velocity along the cylinder the exhaust port (65º after the onset of scavenging), loss of fresh air due to scav- wall toward the cylinder head. By proper direction of the scavenging jets it is enging amounted to 13 percent of the scavenge air flow. The actual plot of degree possible to achieve almost no outflow in the direction of the exhaust from the of purity (or ns) versus delivery ratio (A) closely followed the perfect displace- cross-hatched stagnation zone on the opposite cylinder wall. In fact, measure- ment line for A < 0.4. For A > 0.4, the shape of the actual curve was similar in ment of the velocity profile in this region is a good indicator of the effectiveness shape to the complete mixing curve. of the scavenging flow. If the flow along the cylinder wall toward the head is Engine tests confirm these results from model studies. Initially, the stable, i.e ., if its maximum velocity occurs near the wall and the velocity is near exhausted gas contains no fresh air or mixture; only burned gas is being dis- zero on the plane perpendicular to the axis of symmetry of the ports (which placed from the cylinder. However, within the cylinder both displacement and passes through the cylinder axis), the scavenging flow will follow the desired path. mixing at the interface between burned gas and fresh gas are occurring. The If there are "tongues" of scavenging flow toward the exhaust port, either in the departure from perfect scavenging behavior is evident when fresh mixture first center of the cylinder or along the walls, then significant short-circuiting will appears in the exhaust. For loop-scavenged engines this is typically when occur.31 A ~ 0.4. For uniflow scavenging this perfect scavenging phase lasts somewhat In uniflow-scavenged configurations, the inlet ports are evenly spaced longer; for cross-scavenging it is over sooner (because the short-circuiting path is around the full circumference of the cylinder and are usually directed so that the shorter). scavenging jets create a swirling flow within the cylinder (see Fig. 6-24). Results of The mixing that occurs is short-range mixing, not mixing throughout the measurements of scavenging front location in rig flow tests of a valved uniflow cylinder volume. The jets of scavenging mixture, on entering the cylinder, mix two-stroke diesel cylinder, as the inlet port angle was varied to give a wide range readily with gases in the immediate neighborhood of the jet efflux. More efficient of swirl, showed that inlet jets directed tangentially to a circle of half the cylinder scavenging-i.e ., less mixing-is obtained by reducing the size of the inlet ports radius gave the most stable scavenging front profile over a wide range of condi- while increasing their number.32 It is important that the jets from the inlet ports tions. 33 slow down significantly once they enter the cylinder. Otherwise the scavenging Though the scavenging processes in spark-ignition and diesel two-stroke front will reach the exhaust ports or valves too early. The jets are frequently engines are similar, these two types of engine operate with quite different delivery directed to impinge on each other or against the cylinder wall. Swirl in uniflow- ratios. In mixture-scavenged spark-ignition engines, any significant expulsion of scavenged systems may be used to obtain an equivalent result. fresh fluid with the burned gas results in a significant loss of fuel and causes high The most desirable loop-scavenging flow is illustrated in Fig. 6-29. The hydrocarbon emissions as well as loss of the energy expended in pumping the 244 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 245 flow which passes straight through the cylinder. In diesels the scavenging medium 1.0 is air, so only the pumping work is lost. One consequence of this is that two- stroke spark-ignition engines are usually crankcase pumped. This approach pro- vides the maximum pressure and thus also the maximum velocity in the 0.8 scavenging medium at the start of the scavenging process just after the cylinder pressure has blown down; as the crankcase pressure falls during the scavenging Perfect displacement process, the motion of the scavenging front within the cylinder also slows down. Figure 6-30 shows the delivery ratio and trapping, charging, and scavenging effi- 0.6 ciencies of two crankcase-scavenged spark-ignition engines as a function of engine speed. These quantities depend significantly on intake and exhaust port Degree of purity Complete mixing design and open period and the exhaust system configuration.34-36 For two- stroke cycle spark-ignition engines, which use crankcase pumping, delivery ratios 0.4 vary between about 0.5 and 0.8. Uniflow scavenging Figure 6-31 shows scavenging data typical of large two-stroke diesels.37 The purity (mass of air in trapped cylinder charge/mass of trapped cylinder charge) is -Loop scavenging shown as a function of the delivery ratio. The different scavenging configurations 0.2 Cross scavenging have different degrees of effectiveness, with uniflow scavenging being the most efficient. These diesel engines normally operate with delivery ratios in the range 1.2 to 1.4. 0.4 0.6 0.8 1.0 1.2 1.4 1.6 Delivery ratio A 0.9 FIGURE 6-31 7sc Purity as a function of delivery ratio A for different types of large marine two-stroke diesel engines.37 0.8 ntr 6.7 FLOW THROUGH PORTS 0.7 nr nch, 77sc The importance of the intake and exhaust ports to the proper functioning of the 0.6 two-stroke cycle scavenging process is clear from the discussion in Sec. 6.6. The crank angle at which the ports open, the size, number, geometry, and location of 0.5 ich the ports around the cylinder circumference, and the direction and velocity of the 11ch jets issuing from the ports into the cylinder all affect the scavenging flow. A 0.4 summary of the information available on flow through piston-controlled ports .8 can be found in Annand and Roe.16 Both the flow resistance of the inlet and -- --- exhaust port configurations, as well as the details of the flow pattern produced by V 3.7 the port system inside the cylinder during scavenging, are important. Figure 6-32 defines the essential geometrical characteristics of inlet ports. Rectangular ports 0.6 make best use of the cylinder wall area and give precise timing control. Ports can V be tapered, and may have axial and tangential inclination as shown. 0.5 700 2000 3000 4000 5000 6000 Figure 6-33 illustrates the flow patterns expected downstream of piston- Speed, rev/min controlled inlet ports. For small openings, the flow remains attached to the port walls. For fully open ports with sharp corners the flow detaches at the upstream FIGURE 6-30 Delivery ratio A, trapping efficiency #u, charging efficiency nen, and scavenging efficiency ne, at full corners. Both a rounded entry and converging taper to the port help prevent flow load, as functions of speed for two single-cylinder two-stroke cycle spark-ignition engines. Solid line detachment within the port. Discharge coefficients for ports have been measured 147 cm3 displacement engine.34 Dashed line is loop-scavenged 246 cm3 displacement engine.35 as a function of the open fraction of the port, the pressure ratio across the port, 246 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 247 -X- 1.0 Corner radius r 4. Height Y o Rounded entry Rectangular Skewed Cochcient of discharge $ 0.8- x === P ====== Radial convergence Thickness ? A Rounded entry, circular ports Axial convergence O Rounded entry, square ports Sharp entry O Sharp entry, circular ports Open height X Sharp entry, square ports 0.6 0.2 0.4 0.6 0.8 Tangential inclination Tangent radion 1.0 Axial inclination Port open fraction Piston FIGURE 6-34 Discharge coefficients as a function of port open fraction (uncovered height/port height) for different inlet port designs. Pressure ratio across port = 2.35.2º FIGURE 6-32 Parameters which define geometry of inlet ports.16 ratio increases. Empirical relations that predict this variation with pressure ratio have been developed.38 and port geometry and inclination (see Ref. 16 for a detailed summary). The most Tangentially inclined inlet ports are used when swirl is desired to improve appropriate reference area for evaluating the discharge coefficient is the open scavenging or when jet focusing or impingement within the cylinder off the cylin- area of the port (see Fig. 6-32). For the open height h ,, less than (Y - r) but der axis is required (see Sec. 6.6.3). The discharge coefficient decreases as the jet tangential inclination increases. The jet angle and the port angle can deviate greater than r this is significantly from each other depending on the details of the port design and the AR = Xh, - 0.43r2 (6.30) open fraction. 31 where Y is the port height, X the port width, and r the corner radius. For h, = Y, In piston-controlled exhaust ports, the angle of the jet from a thin-walled exhaust port increases as indicated in Fig. 6-35.31 In thick ports, the walls are the reference area is AR = XY - 0.8612 (6.31) 90 The effect of variations in geometry and operating conditions on the discharge coefficient CD can usually be interpreted by reference to the flow patterns illus- 70 trated in Fig. 6-33. The effects of inlet port open fraction and port geometry on CD are shown in Fig. 6-34: geometry effects are most significant at small and large open fractions.2º CD varies with pressure ratio, increasing as the pressure 0 20 40 60 80 100 Uncovered port height, % (a) (b) FIGURE 6-33 Flow pattern through piston-controlled inlet ports: (a) port axis perpendicular to wall; small openny FIGURE 6-35 and large opening with sharp and rounded entry; (b) port axis inclined.16 Angle of jet exiting exhaust port as a function of open port height.31 248 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 249 0.9 The term supercharging refers to increasing the air (or mixture) density by increasing its pressure prior to entering the engine cylinder. Three basic methods are used to accomplish this. The first is mechanical supercharging where a 5 separate pump or blower or compressor, usually driven by power taken from the ₱2.5 engine, provides the compressed air. The second method is turbocharging, where a turbocharger-a compressor and turbine on a single shaft -- is used to boost the inlet air (or mixture) density. Energy available in the engine's exhaust stream is 0.8| 2 used to drive the turbocharger turbine which drives the turbocharger compressor PE which raises the inlet fluid density prior to entry to each engine cylinder. The $1.67 third method-pressure wave supercharging-uses wave action in the intake and CD exhaust systems to compress the intake mixture. The use of intake and exhaust 1.43 manifold tuning to increase volumetric efficiency (see Sec. 6.2.2) is one example of this method of increasing air density. An example of a pressure wave super- 1.25 charging device is the Comprex, which uses the pressure available in the exhaust 0.7 gas stream to compress the inlet mixture stream by direct contact of the fluids in 1.0 narrow flow channels (see Sec. 6.8.5). Figure 6-37 shows typical arrangements of the different supercharging and turbocharging systems. The most common arrangements use a mechanical supercharger (Fig. 6-37a) or turbocharger (Fig. 6-37b). Combinations of an engine-driven compressor and a turbocharger (Fig. Port 6-37c) are used (e.g ., in large marine engines; Fig. 1-24). Two-stage turbocharging 0.6 O 0.2 0.4 0.6 0.8 1.0 (Fig. 6-37d) is one viable approach for providing very high boost pressures (4 to Port open fraction 7 atm) to obtain higher engine brake mean effective pressures. Turbocompound- ing, i.e ., use of a second turbine in the exhaust directly geared to the engine drive FIGURE 6-36 Discharge coefficient of a single rectangular exhaust port (7.6 mm wide x 12.7 mm high) in the wall of shaft (Fig. 6-37e), is an alternative method of increasing engine power (and a 51-mm bore cylinder as a function of open fraction and pressure ratio. Steady-flow rig tests at 21ºC. efficiency). Charge cooling with a heat exchanger (often called an aftercooler or Pe = cylinder pressure, pe =: exhaust system pressure.39 intercooler) after compression, prior to entry to the cylinder, can be used to increase further the air or mixture density as shown in Fig. 6-37f. usually tapered to allow the outward flow to diffuse. The pressure ratio across Supercharging is used in four-stroke cycle engines to boost the power per the exhaust ports varies substantially during the exhaust process. The pressure unit displaced volume. Some form of supercharging is necessary in two-stroke ratio has a significant effect on the exhaust port discharge coefficient, as shown in cycle engines to raise the fresh air (or mixture) pressure above the exhaust pres- Fig. 6-36. The changes in exit jet angle and separation point explain the effects of sure so that the cylinder can be scavenged effectively. With additional boost in increasing open fraction and pressure ratio. The discharge coefficient also two-stroke cycle engines, the power density is also raised. This section reviews the increases modestly with increasing gas temperature.39 operating characteristics of the blowers, compressors, turbines, and wave- compression devices used to increase inlet air or mixture density or convert exhaust-gas availability to work. The operating characteristics of supercharged 6.8 SUPERCHARGING AND and turbocharged engine systems are discussed in Chap. 15. TURBOCHARGING 6.8.1 Methods of Power Boosting 6.8.2 Basic Relationships The maximum power a given engine can deliver is limited by the amount of fud Expressions for the work required to drive a blower or compressor and the work that can be burned efficiently inside the engine cylinder. This is limited by the produced by a turbine are obtained from the first and second laws of thermody- amount of air that is introduced into each cylinder each cycle. If the inducted ai namics. The first law, in the form of the steady flow energy equation, applied to a is compressed to a higher density than ambient, prior to entry into the cylinder. control volume around the turbomachinery component is the maximum power an engine of fixed dimensions can deliver will be increased. This is the primary purpose of supercharging; Eqs. (2.39) to (2.41) show how ( 6.32) power, torque, and mean effective pressure are proportional to inlet air density. GAS EXCHANGE PROCESSES 251 where Q is the heat-transfer rate into the control volume, W is the shaft work- C T transfer rate out of the control volume, m is the mass flow, h is the specific enth- alpy, C2/2 is the specific kinetic energy, and gz is the specific potential energy C (which is not important and can be omitted). E A stagnation or total enthalpy, ho, can be defined as E ho = h + 2 (6.33) (a) (b) For an ideal gas, with constant specific heats, a stagnation or total temperature follows from Eq. (6.33): C2 To = T + 2c. (6.34) C1 A stagnation or total pressure is also defined : it is the pressure attained if the gas is isentropically brought to rest: C2 T y/ (y - 1 ) Po = P (6.35) C2 12 in Eq. (6.32) for pumps, blowers, compressors, and turbines is usually small C1 enough to be neglected. Equation (6.32) then gives the work-transfer rate as E - W= m(ho, out - ho, in) (6.36) E A component efficiency is used to relate the actual work-transfer rate to the work-transfer rate required (or produced) by an equivalent reversible adiabatic device operating between the same pressures. The second law is then used to (c) ( d) determine this reversible adiabatic work-transfer rate, which is that occurring in an isentropic process. For a compressor, the compressor isentropic efficiency nc is C T nc == reversible power requirement actual power requirement (6.37) C Figure 6-38 shows the end states of the gas passing through a compressor on an h-s diagram. Both static (p1, P2) and stagnation (Po1> Po2) constant-pressure lines are shown. The total-to-total isentropic efficiency is, from Eq. (6.37), T2 nCTT hors - hol E hoz - hos (6.38) E which, since c, is essentially constant for air, or fuel-air mixture, becomes nCTT = 7 To2s - TOI (e) To2 - TO1 (6.39) FIGURE 6-37 Since the process 01 to 02s is isentropic, Supercharging and turbocharging configurations: (a) mechanical supercharging; (b) turbocharging; (c) engine-driven compressor and turbocharger; (d) two-stage turbocharging; (e) turbocharging with turbocompounding; (f) turbocharger with intercooler. C Compressor, E Engine, / Inter-cooler, I Tozs = To, ( Po2 ) (-1)/Y Poi Turbine. 250 252 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 253 Po2 h 02 3 the blower or compressor. Thus the power required to drive the device, - Wc,D, will be 02s - WC. D = WC (6.43) 25 where nm is the blower or compressor mechanical efficiency. im Figure 6-39 shows the gas states at inlet and exit to a turbine on an h-s diagram. State 03 is the inlet stagnation state; 4 and 04 are the exit static and stagnation states, respectively. States 4s and 04s define the static and stagnation POI exit states of the equivalent reversible adiabatic turbine. The turbine isentropic P efficiency is defined as 01 actual power output FIGURE 6-38 nT = reversible power output (6.44) Enthalpy-entropy diagram for compressor. Inlet state 01, exit state 2; equivalent isentropic com- Thus, the total-to-total turbine efficiency is pressor exit state 2s. ho3 - h04 nTTT (6.45) Equation (6.39) becomes ho3 - hoas (Poz/Por)(Y - 1)Wy - 1 If the exhaust gas is modeled as an ideal gas with constant specific heats, then Eq. 1CTT (6.45) can be written (To2/Toi) - 1 (6.40) To3 - TO4 1 -(To4/To3) In deriving Eq. (6.40) it has been tacitly assumed that the kinetic energy nTTT = T Tos - TOAS 1 - (Po4/Po3)(7- 1)/Y (6.46) pressure head (po2 - P2) can be recovered. In internal combustion engine applica- tions the compressor feeds the engine via a large manifold, and much of this Note that for exhaust gas over the temperature range of interest, c, may vary kinetic energy will be dissipated. The blower or compressor should be designed significantly with temperature (see Figs. 4-10 and 4-17). for effective recovery of this kinetic energy before the exit duct. Since the kinetic energy of the gas leaving the compressor is not usually recovered, a more realistic definition of efficiency is based on exit static conditions:40 Po3 T25 -To1_ (P2/Por)(- 1)/7 - 1 nCTS - (6.41) To2 - TO1 (To2/Tos) - 1 03 This is termed the total-to-static efficiency. The basis on which the efficiency is calculated should always be clearly stated. W The work-transfer rate or power required to drive the compressor is obtained by combining Eq. (6.36), the ideal gas model, and Eq. (6.40): P04 04 - Wc = m; Cp. (Toz - To1) - miCe.a Toi ( Poz ) (7- 11/7. (6.42) P4 nCTT 045 where the subscript i denotes inlet mixture properties. If ners is used to define the 45 FIGURE 6-39 compressor performance, then p2 replaces poz in Eq. (6.42). Equation (6.42) gives Enthalpy-entropy diagram for a turbine. Inlet the thermodynamic power requirement. There will also be mechanical losses in state 03, exit state 4; equivalent isentropic S turbine exit state 4s. 254 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 255 Since the kinetic energy at the exit of a turbocharger turbine is usually For a particular device, the dimensions are fixed and the value of R is fixed. So it wasted, a total-to-static turbine isentropic efficiency, where the reversible adia- has become the convention to plot batic power output is that obtained between inlet stagnation conditions and the exit static pressure, is more realistic:40 ATO im To,in , n, To in N Po. out Po, in ho - ho4 _ To3 - To4__1 - (To4/To3) To, in Po, in ) (6.53) n ITS To3 - T45 . 1 - (P4/Po3)(7- 1)/Y (6.47) ho3 - has im To, in/Po, in is referred to as the corrected mass flow; N/ To, im is referred to as The power delivered by the turbine is given by [Eqs. (6.36) and (6.46)] the corrected speed. The disadvantage of this convention of removing D and R is that the groups of variables are no longer dimensionless, and performance plots or maps relate to a specific machine. W = me(ho - ho4) = me Cp,e(Tos - To4) = me Cp.enTTT To3 1 - (Dos ) ("e - 13/7.] Compressor characteristics are usually plotted in terms of the pressure ratio (Po2/Po1) or (P2/Po1) against the corrected mass flow (m / To1/Po1) along lines of (6.48) constant corrected speed (N/ /To1). Contours of constant efficiency are super- posed. Similar plots are used for turbines: Po3/P4 against mTo3/Po3 along lines where the subscript e denotes exhaust gas properties. If the total-to-static turbine of constant N// To3. Since these occupy a narrow region of the turbine per- efficiency (nTTs) is used in the relation for Wr, then p4 replaces Po4 in Eq. (6.48). formance map, other plots are often used (see Sec. 6.8.4). With a turbocharger, the turbine is mechanically linked to the compressor. Hence, at constant turbocharger speed, - Wc = nm WT (6.49) 6.8.3 Compressors where nm is the mechanical efficiency of the turbocharger. The mechanical losses Practical mechanical supercharging devices can be classified into: (1) sliding vane are mainly bearing friction losses. The mechanical efficiency is usually combined compressors, (2) rotary compressors, and (3) centrifugal compressors. The first with the turbine efficiency since these losses are difficult to separate out. two types are positive displacement compressors; the last type is an aerodynamic It is advantageous if the operating characteristics of blowers, compressors, compressor. Four different types of positive displacement compressors are illus- trated in Fig. 6-40. and turbines can be expressed in a manner that allows easy comparison between different designs and sizes of devices. This can be done by describing the per- In the sliding vane compressor (Fig. 6-40a), deep slots are cut into the rotor formance characteristics in terms of dimensionless numbers.4º The most impor- to accommodate thin vanes which are free to move radially. The rotor is tant dependent variables are: mass flow rate m, component isentropic efficiency ", mounted eccentrically in the housing. As the rotor rotates, the centrifugal forces and temperature difference across the device ATo. Each of these are a function of acting on the vanes force them outward against the housing, thereby dividing the the independent variables: Po, in , Po, out (Or Pour), To, in, N(speed), D(characteristic crescent-shaped space into several compartments. Ambient air is drawn through the intake port into each compartment as its volume increases to a maximum. dimension), R(gas constant), y (Cp/c,), and u(viscosity); i.e ., The trapped air is compressed as the compartment volume decreases, and is then im, n, ATo = f(Po, in, Po. out, To, in, N, D, R, Y, 4) (6.50) discharged through the outlet port. The flow capacity of the sliding vane com- pressor depends on the maximum induction volume which is determined by the By dimensional analysis, these eight independent variables can be reduced to four housing cylinder bore, rotor diameter and length, eccentricity, number of vanes, dimensionless groups: dimensions of the inlet and outlet ports. The actual flow rate and pressure rise at constant speed will be reduced by leakage. Also, heat transfer from the moving in RTO , in ATO ND Po , out in (6.51) vanes and rotor and stator surfaces will reduce compression efficiency unless Po, in D2 To . in RTo, in Po, in ' MD'Y cooling is used to remove the thermal energy generated by friction between the vanes, and the rotor and stator. The volumetric efficiency can vary between 0.6 The Reynolds number, m/(UD), has little effect on performance and y is fixed by and 0.9 depending on the size of the machine, the quality of the design, and the the gas. Therefore these variables can be omitted and Eq. (6.51) becomes . by method of lubrication and cooling employed. The displaced volume VD is given im RTo. in ATO ND Po, out (6.52) , n, 7 Po, in D2 ' ' To, in RTo, in Po, in ) VD = nel ( 27 + 8 ) (6.54) GAS EXCHANGE PROCESSES 257 where r is the rotor radius, & the eccentricity, and / the axial length of the com- pressor. The mass flow rate parameter is im Tol Istd . @ = constant x p, n. Nel(2r + 2) Po/Psta (6.55) where ne is the device volumetric efficiency, N its speed, and the subscripts i, 0 and std refer to inlet, inlet stagnation, and standard atmospheric conditions, respectively. Figure 6-41 shows the performance characteristics of a typical sliding vane compressor. The mass flow rate at constant speed depends on the pressure ratio only through its (weak) effect on volumetric efficiency. The isentro- pic efficiency is relatively low.41 (d) An alternative positive displacement supercharger is the roots blower (Fig. Screw compressor (c) Lysholm compressor 6-40b). The two rotors are connected by gears. The working principles are as follows. Air trapped in the recesses between the rotor lobes and the housing is carried toward the delivery port without significant change in volume. As these recesses open to the delivery line, since the suction side is closed, the trapped air is suddenly compressed by the backflow from the higher-pressure delivery line. This intermittent delivery produces nonuniform torque on the rotor and pressure pulses in the delivery line. Roots blowers are most suitable for small pressure ratios (about 1.2). The volumetric efficiency depends on the running clearances, - rotor length, rotational speed, and pressure ratio. In the three-lobe machines shown (two lobes are sometimes used) the volume of each recess VR is VR = 0.546R21 Positive displacement compressors: (a) sliding vane compressor ; (b) roots blower ; (c) Lysholm compressor; (d) screw compressor."" 2.2 Eccentricity 60 45 30 35 nc = 25 N = 20 rev/s 0.50 0.55 Sliding vane 1.8 -0.58 0.62 HA Pressure rallo por Roots blower 0.68 (b) (a) 1.4 Rotor . LOL 0 0.1 0.2 0.3 0.4 FIGURE 6-40 0.5 Mass flow mv Tol Tstd ( po/Psta) , kg/s FIGURE 6-41 Performance map for sliding vane compressor.41 256 258 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 259 where R is the rotor radius and / the blower length. The mass flow parameter is Collector Diffuser in Tol Tard = constant x p. n. NR21 Po/ Pstd (6.56) 3 Diffuser A performance map of a typical small roots blower is shown in Fig. 6-42. It is similar in character to that of the sliding vane compressor. At constant speed, the flow rate depends on increasing pressure ratio only through the resulting Shroud decrease in volumetric efficiency (Eq. 6.56).41 The advantage of the roots blower is that its performance range is not limited by surge and choking as is the cen- trifugal compressor (see below). Its disadvantages are its high noise level, poor efficiency, and large size.42 Lalet Screw compressors (Fig. 6-40c and d) must be precision machined to achieve close tolerances between rotating and stationary elements for satisfactory operation. They run at speeds between 3000 and 30,000 rev/min. It is usually necessary to cool the rotors internally. High values of volumetric and isentropic efficiency are claimed.41 Impeller A centrifugal compressor is primarily used to boost inlet air or mixture FIGURE 6-43 density coupled with an exhaust-driven turbine in a turbocharger. It is a single- Schematic of centrifugal compressor.40 stage radial flow device, well suited to the high mass flow rates at the relatively low pressure ratios (up to about 3.5) required by the engine. To operate efficiently The centrifugal compressor consists of a stationary inlet casing, a rotating it must rotate at high angular speed. It is therefore much better suited to direct bladed impellor, a stationary diffuser (with or without vanes), and a collector or coupling with the exhaust-driven turbine of the turbocharger than to mechanical volute to bring the compressed air leaving the diffuser to the engine intake system coupling through a gearbox to the engine for mechanical supercharging. (see Fig. 6-43). Figure 6-44 indicates, on an h-s diagram, how each component contributes to the overall pressure rise across the compressor. Air at stagnation 1.6 120 0.30 Po2 -- 140 revis P03 02 1.5- -- ho3, 202 03 -- P3 0.40 1.4- 0.50 Pressure ratio poi nc = 0.55 - 1.3 · P2 25 1 1.2|- 1011 POI I.IL no , hor. 0.02 0.04 0.06 0.08 0.10 0.12 PO Air mass flow rate, kg/s FIGURE 6-44 FIGURE 6-42 Performance map at standard inlet conditions for roots blower.42 Enthalpy-entropy diagram for flow S through centrifugal compressor. 260 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 261 state 0 is accelerated in the inlet to pressure p1 and velocity C1. The enthalpy This is often called the Euler equation. Normally in compressors the inlet flow is change 01 to 1 is C1/2. Compression in the impeller flow passages increases the axial so Co1 = 0. Thus Eq. (6.58) can be written: pressure to p2 and velocity to C2, corresponding to a stagnation state 02 if all the exit kinetic energy were recovered. The isentropic equivalent compression process Wc = U2 Cor = U2(1 - , Cz cot B2 ) (6.59) has an exit static state 2s. The diffuser, 2 to 3, converts as much as practical of the in U2 air kinetic energy at exit to the impeller (C2/2) to a pressure rise (p3 - P2) by where B2 is the backsweep angle. In the ideal case with no slip, f2 is the blade slowing down the gas in carefully shaped expanding passages. The final state, in angle, B2b. In practice, there is slip and 62 is less than B2b. Many compressors the collector, has static pressure p3, low kinetic energy C3/2, and a stagnation have radial vanes (i.e ., B2b = 90º). A recent trend is backswept vanes (B2b < 90º) pressure po3 which is less than po2 since the diffusion process is incomplete as which give higher efficiency. Since work transfer to the gas occurs only in the well as irreversible.4º impeller, the work-transfer rate given by Eq. (6.59) equals the change in stagna- The work transfer to the gas occurs in the impeller. It can be related to the tion enthalpy (ho3 - ho1) in Fig. 6-44 [see Eq. (6.36)]. change in gas angular momentum via the velocity components at the impeller The operating characteristics of the centrifugal compressor are usually entry and exit, which are shown in Fig. 6-45. Here C1 and C2 are the absolute gas described by a performance map. This shows lines of constant compressor effi- velocities, U1 and U2 are the tangential blade velocities, and w1 and w2 are the ciency nc, and constant corrected speed N/To, in, on a plot of pressure gas velocities relative to the impeller all at inlet (1) and exit (2), respectively. The ratio Po, out/Po, in against corrected mass flow mTo, in/Po. in [see Eq. (6.53)]. torque T exerted on the gas by the impeller equals the rate of change of angular Figure 6-46 indicates the form of such a map. The stable operating range in the momentum: center of the map is separated from an unstable region on the left by the surge T = m(r2 Co2 - r1 Cos) (6.57) line. When the mass flow is reduced at a constant pressure ratio, local flow reversal eventually occurs in the boundary layer. Further reductions in mass flow The rate of work transfer to the gas is given by cause the flow to reverse completely, causing a drop in pressure. This relieves the adverse pressure gradient. The flow reestablishes itself, builds up again, and the -Wc = Tw = maxr2 Co2 - 71Co,) = m(U2 Co2 - U, Co,) (6.58) process repeats. Compressors should not be operated in this unstable regime. The C2 Unstable B21 02 -- Stable B2b Surge line Choking Actual --- ---- Ideal (no slip) Pressure rano .C1 With prewhirl - Without prewhirl Rough running WI FIGURE 6-46 FIGURE 6-45 Mass flow rate MVTo Schematic of compressor operating map showing Velocity diagrams at inlet (1) and exit (2) to centrifugal compressor rotor or impeller.40 Po stable operating range.40 262 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 263 3.0 6.8.4 Turbines 2.8 - The turbocharger turbine is driven by the energy available in the engine exhaust. The ideal energy available is shown in Fig. 6-48. It consists of the blowdown 2.6 work transfer produced by expanding the gas in the cylinder at exhaust valve opening to atmospheric pressure (area abc) and (for the four-stroke cycle engine) 2.4 the work done by the piston displacing the gases remaining in the cylinder after blowdown (area cdef). 2.2 Ncor = 152,400 rev/min The reciprocating internal combustion engine is inherently an unsteady pul- sating flow device. Turbines can be designed to accept such an unsteady flow, but Pressure ratio 2.0 142,200 they operate more efficiently under steady flow conditions. In practice, two = 70%X `131,400 68% approaches for recovering a fraction of the available exhaust energy are com- 1.8 121,200 monly used: constant-pressure turbocharging and pulse turbocharging .. In ¿65% constant-pressure turbocharging, an exhaust manifold of sufficiently large volume 1.61 109,200 to damp out the mass flow and pressure pulses is used so that the flow to the turbine is essentially steady. The disadvantage of this approach is that it does not 1.4 95,400 make full use of the high kinetic energy of the gases leaving the exhaust port; the FIGURE 6-47 78,000 losses inherent in the mixing of this high-velocity gas with a large volume of 1.2H Centrifugal compressor operating map. low-velocity gas cannot be recovered. With pulse turbocharging, short small- -55,800 Lines of constant corrected speed and compressor efficiency are plotted on a cross-section pipes connect each exhaust port to the turbine so that much of the 1.0 0 . 0.05 0.10 0.15 0.20 0.25 graph of pressure ratio against corrected kinetic energy associated with the exhaust blowdown can be utilized. By suitably Corrected mass flow rate micor, kg/s mass flow.43 grouping the different cylinder exhaust ports so that the exhaust pulses are sequential and have minimum overlap, the flow unsteadiness can be held to an acceptable level. The turbine must be specifically designed for this pulsating flow to achieve adequate efficiencies. The combination of increased energy available at stable operating regime is limited on the right by choking. The velocities increase as m increases, and eventually the flow becomes sonic in the limiting area of the the turbine, with reasonable turbine efficiencies, results in the pulse system being machine. Extra mass flow through the compressor can only be obtained by more commonly used for larger diesels.4º For automotive engines, constant- pressure turbocharging is used. higher speed. When the diffuser is choked, compressor speed may rise substan- Two types of turbines are used in turbochargers: radial and axial flow tur- tially with only a limited increase in the mass flow rate.40 bines. The radial flow turbine is similar in appearance to the centrifugal compres- Figure 6-47 shows an actual turbocharger compressor performance map. la sor; however, the flow is radially inward not outward. Radial flow turbines are practice, the map variables corrected speed and mass flow rate are usually defined as44 1/2 N cor = N (6.60) Pch = charging pressure mcor = im To, in 1/2 Pref Pa = ambient pressure Tref Po, in) where Tref and Pref are standard atmospheric temperature and pressure, respec- tively. Though the details of different compressor maps vary, their general char- acteristics are similar. The high efficiency region runs parallel to the surge line (and close to it for vaneless diffusers). A wide flow range for the compressor (sc FIGURE 6-48 Fig. 6-46) is important in turbochargers used for transportation applications. Constant-volume cycle p-V diagram showing available exhaust energy. 264 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 265 Casing of the radial turbine shown in the h-s diagram of Fig. 6-50. The velocity triangles at entry and exit to the rotor, shown in Fig. 6-54, relate the work transfer from the gas to the rotor to the change in angular momentum: -Nozzles- Rotor W = WT = maxr2 C 62 + 1 3 Co3 ) 2 3 Po 01 P1 FIGURE 6-49 hoi Schematic of radial flow turbine. normally used in automotive or truck applications. Larger engines-locomotive, stationary, or marine-use axial flow turbines. A drawing of a radial flow turbine is shown in Fig. 6-49. It consists of an inlet casing or scroll, a set of inlet nozzles (often omitted with small turbines), and P2 the turbine rotor or wheel. The function of each component is evident from the MY 25 h-s diagram and velocity triangles in Fig. 6-50. The nozzles (01-2) accelerate the flow, with modest loss in stagnation pressure. The drop in stagnation enthalpy, and hence the work transfer, occurs solely in the rotor passages, 2-3: hence, the 03 .PO3 rotor is designed for minimum kinetic energy C3/2 at exit. The velocity triangles at inlet and exit relate the work transfer to the change in angular momentum via 23 the Euler equation: W = To = mar2 Co2 - 3 C63) = m(U2 Co2 - U3 C63) (6.61) 35 where T is the torque and @ the rotor angular speed. For maximum work trans- fer the exit velocity should be axial. The work-transfer rate relates to the change (a) in stagnation enthalpy via W = m ( ho2 - hos ) = m(ho1 - hos ) (6.62) W 2 C2 The turbine isentropic efficiency is given by Eqs. (6.44) to (6.47). Many different types of plots have been used to define radial flow turbine Cor characteristics. Figure 6-51 shows lines of constant corrected speed and efficiency U2 on a plot of pressure ratio versus corrected mass flow rate. As flow rate increases at a given speed, it asymptotically approaches a limit corresponding to the flow becoming choked in the stator nozzle blades or the rotor. For turbines, efficiency is usually presented on a different diagram because the operating regime in Fig- 6-51 is narrow. Figure 6-52 shows an alternative plot for a radial turbine: cor- rected mass flow rate against corrected rotor speed. On this map, the operating regime appears broader. # 3 A schematic of a turbocharger axial flow turbine is shown in Fig. 6-53. C3 Usually a single stage is sufficient to expand the exhaust gas efficiently through the pressure ratios associated with engine turbocharging. This turbine consists of FIGURE 6-50 Cos an annular flow passage, a single row of nozzles or stator blades, and a rotating (a) Enthalpy-entropy diagram for blade ring. The changes in gas state across each component are similar to those ( b) radial turbine. (b) Velocity diagrams at turbine rotor inlet (2) and exit (3). 266 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 2.8 267 0.24 2.6 0.22 2.4 0.20 70 Turbine efficiency, % /3/18 72 2.8 65 TTS = 0.70 0.18 70 2.2 2.2 Corrected mass flow rate Megr. kg/s 0.16 2.0F 0.14- 1.8 Po3 PA Total-to-static pressure ratio 1.8- 0.12 imcor = in To/922 K "Po/99.4 kPa --- 0.65- 0.10 -- Noor VT6/922 K 1.6 1.5 4000 1.3 0.00- 10 1.4- 3.500 ==== 20 30 40 50 60 70 80 90 3000 Corrected rotor speed Ncor, 103 rev/min 2500 1.2 0.40 0.50 0.52. FIGURE 6-52 1500 500 rotor speed.45 Alternative radial turbine performance map: corrected mass flow rate is plotted against corrected 1.0 1.5 2.0 2.5 . 3.0 nV TO3 or the wheel tip speed for a radial flow turbine, divided by the velocity equivalent Pos of the isentropic enthalpy drop across the turbine stage, C,; i.e ., FIGURE 6-51 Radial turbine performance map showing lines of constant corrected speed and efficiency on a plot of Blade speed ratio = " pressure ratio versus corrected mass flow rate. To3 = turbine inlet temperature (K), Po3 = turbine inlet pressure (bar), p. = turbine exit pressure (bar), m = mass flow rate (kg/s), N = speed (rev/min).40 where C3 = [ 2 (h03 - 24s ) ] 1/2 (6.64) Since the mid-radius r2 usually equals the mid-radius r3, W, = imU(Co2 + Co3) = mU(C2 sin @2 + C3 sin a3) Inlet Exit = mU(Cz, tan B2 + C1, tan B3) (6.63) (2) 3 Equation (6.62) relates the work-transfer rate to the stagnation enthalpy change 01 as in the radial turbine. Figure 6-55 shows axial turbine performance characteristics on the stan- dard dimensionless plot of pressure ratio versus corrected mass flow rate. Here the constant speed lines converge to a single choked flow limit as the mass flow is increased. In the radial turbine, the variation in centrifugal effects with speed cause a noticeable spread in the constant speed lines (Fig. 6-51). An alternative performance plot for turbines is efficiency versus blade speed Nozzles ratio. This ratio is the blade speed U (at its mean height) for an axial flow turbine FIGURE 6-53 Blade ring Schematic of single-stage axial flow turbine. 268 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 269 0.75 203 = 1.5 0 70 - 2.5 0.90 Po3 = 1.5 to 3 Cor Boost 0.80 Nozzle blades 0.70 W 2 0.60 0.60 C2 O SSL 0 .3 1.4 0.5 ).6 0.7 0.8 0.50. 0.2 0.4 0.6 0.8 1.0 (a) Axial flow (b) Radial flow FIGURE 6-56 Rotor blades Plot of turbine total-to-static efficiency versus blade speed ratio U/C, for (a) axial flow and (b) radial flow turbines. 40 W3 This method of displaying performance is useful for matching compressor and FIGURE 6-54 turbine wheel size for operation of the turbine at optimum efficiency. Figure 6-56 Velocity diagrams at entry (2) and exit (3) to axial flow turbine blade ring.40 shows such plots for an axial and radial flow turbine. The peak efficiency can occur for 0.4 < U/C, < 0.8, depending on turbine design and application.4º For a given turbocharger, the compressor and turbine characteristics are linked. Since the compressor and turbine are on a common shaft with speed N: 2.4 Turbine choking JTOS ( TO ) (6.65) For imc = imT = im (if mc[1 + (F/A)] = my, the equation is easily modified): 2.2- im Toi Toi ( Pom)( To) 1/2 Po1 Po3 (6.66) 2.0 Since the compressor and turbine powers are equal in magnitude: 2 1.8 hoz - hou = nm ( ho - ho4) (6.67) or, with an ideal gas model, 1.6- ... Cp.(To2 - Tox) = 1m Cp. I(To3 - To4) (6.68) 1.4 Equation (6.68), with Eqs. (6.40) and (6.46), gives 350 00.50% FIGURE 6-55 Axial flow turbine performance map: pressure 1.2 N =. 300 ratio is plotted against corrected mass flow rate NTO3 SPI 1-( PA ) - 1 = n c nt lm co.c L' TOI (6.69) To3 = turbine inlet temperature (K), Pos " 70 80 90 100 110 120 turbine inlet pressure (bar), p. = turbine exit pressure (bar), m = mass flow rate (kg/'s) Assuming that the turbine exit pressure p4 equals atmospheric pressure Po1, the n Tos PO3 N = speed (rev/min).40 equilibrium or steady-state running lines for constant values of To3/To1 can be 270 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 271 3.4 T 3.2- 6453 -0.65 3.0 0.68 2.8 nc = 0.70 5870 0.550.60- 2.6 0.71 2.4 3.0+ =2.75 21& 2.2 .2.5 = 5286 -2.25- 2.0 .. 8 2.0 FIGURE 6-57 3514 4097 Steady-state turbocharger operating FIGURE 6-58 1.6 lines plotted as constant To3/To1 lines Schematic of Comprex supercharger.47 a Engine, b on compressor map. Turbine charac- Cell wheel or rotor, c Belt drive, d High-pressure 1.4 103 = 1.75 teristics defined by Fig. 6-51. Po1 exhaust gas (G-HP), e High-pressure air (A-HP), f Tos 2935 P03 = P02 compressor inlet pressure (bar), Po2 = Low-pressure air (A-LP), g Low-pressure exhaust 1.2 gas (G-LP) `2351 compressor exit pressure (bar), To1 = compressor inlet temperature (K) 1.0L 6 To3 = turbine inlet temperature (K), m = mass flow rate (kg/s), N = speed (rev/min).40 between two castings by a belt driven from the crankshaft (c). There is no contact between the rotor and the casing, but the gaps are kept small to minimize leakage. The belt drive merely overcomes friction and maintains the rotor at a speed proportional to engine speed (usually 4 or 5 times faster): it provides no determined. Figure 6-57 shows an example of such a set of turbocharger charac- compression work. One casing (the air casing) contains the passage which brings teristics, plotted on a turbocharger compressor map for a radial turbine with low-pressure air (f) to one set of ports and high-pressure air (e) from another set characteristics similar to Fig. 6-51. The dash-dot-dash line is for po2 = Po3. To of ports in the rotor-side inner casing. The other casing (the gas casing) connects the right of this line, Po3 > Po2; to the left of this line Po2 > Po3 . 40 the high-pressure engine exhaust gas (d) to one set of ports at the other end of the The problem of overspeeding the turbocharger and generating very high rotor, and connects a second set of ports to the exhaust system (g). Fluid can flow cylinder pressures often requires that some of the exhaust be bypassed around the into and out of the rotor channels through these ports. The exhaust gas inlet port turbine. The bypass valve or wastegate is usually built into the turbocharger is made small enough to cause a significant pressure rise in the exhaust manifold casing. It consists of a spring-loaded valve acting in response to the inlet mani- (c.g ., 2 atm) when the engine is operated at its rated power. The pressure wave fold pressure on a controlling diaphragm. When the wastegate is open, only a process does not depend on the pressure and flow fluctuations within the mani- portion of the exhaust gases will flow through the turbine and generate power; fold caused by individual cylinder exhaust events: its operation can be explained the remainder passes directly into the exhaust system downstream of the turbine. assuming constant pressure at each set of ports. As the rotor makes one revol- ution, the ends of each channel are alternatively closed, or are open to a flow 6.8.5 Wave-Compression Devices passage. By appropriate arrangement of these passages and selection of the geometry and location of the ports, an efficient energy transfer between the Pressure wave superchargers make use of the fact that if two fluids having differ- · engine exhaust gases and the fresh charge can be realized.46 ent pressures are brought into direct contact in long narrow channels, equi- The wave-compression process in the Comprex can be explained in more lization of pressure occurs faster than mixing. One such device, the Comprex, has detail with the aid of Fig. 6-59, where the rotational motion of the channels has been developed for internal combustion engine supercharging which operates been unrolled. Consider the channel starting at the top; it is closed at both ends using this principle.46 It is shown schematically in Fig. 6-58. The working chan- and contains air at atmospheric pressure. As it opens at the upper edge of the nels of the Comprex are arranged on a rotor or cell wheel (b) which is rotated high-pressure gas (G-HP) duct, a compression or shock wave (1) propagates from 272 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 273 purged by the scavenging air flow (A-S) and filled with fresh air at atmospheric pressure. At wave (9), the cell is closed at both ends, restoring it to its initial state. $7 The speed of these pressure waves is the local sound speed and is a function -G-HP of local gas temperature only. Thus, the above process will only work properly A-HP - for a given exhaust gas temperature at a particular cell speed. The operating range is extended by the use of "pockets" as shown in Fig. 6-59. The pockets prevent the reflection of sound waves from a closed channel end which would cause a substantial change in flow velocity in the channel. These pockets, marked CP and EP on the air side and GP on the exhaust gas side, allow flow from one A-LP G-LP channel to adjacent channels via the pocket if the wave action requires it. Thus the device can be tuned for full-load medium-speed operation and still give FIGURE 6-59 acceptable performance at other loads and speeds because the pockets allow the Unrolled view of the Comprex pressure-wave particle paths to change without major losses.46 process.47 A Air, G Gas, S Scavenging, HP High Figure 6-60 shows the apparent compressor performance map of a pressure, LP Low pressure; CP, EP, GP are Comprex when connected to a small three-cylinder diesel engine. Note that the pockets. map depends on the engine to which the device is coupled because the exhaust gas expansion process and fresh air compression process occur within the same rotor. The volume flow rate is the net air: it is the total air flow into the device the right end of the channel toward the left, compressing the air through which it less the scavenging air flow. The values of isentropic efficiency [defined by Eq. passes. The compressed air behind the wave occupies less space so the high- (6.39)] are comparable to those of mechanical and aerodynamic compressors. pressure exhaust gas moves into the channel as indicated by the dotted line. This line is the boundary between the two fluids. As this wave (1) reaches the left end, the channel is opened and compressed air flows into the engine inlet duct (A-HP). 2.2 The inlet duct is shaped to provide the same mass flow at lower velocity: this 12000 2500 3500 14000 deceleration of the air produces a second compression wave (2) which propagates back into the channel. As a result the compressed air leaving the cell on the left 2.0 has a higher pressure than the driving gas on the right. As this wave (2) arrives at the right-hand side, the high-pressure gas (G-HP) channel closes. An expansion 4500 rev/min wave (3) then propagates back to the left, separating the now motionless and .8 1500 partly expanded fluid on the right from still-moving fluid on the left. When this wave (3) reaches the left-hand end, A-HP is closed and all the gases in the 1000 rev/min channel are at rest. Note that the first gas particles (dotted line) have not quite 7c = 75% Charge-air pressure fatio .6 70/ 65 ! reached the air end of the channel: a cushion of air remains to prevent break- through. The cell's contents are still at a higher pressure than the low pressure in the 1.4 exhaust gas duct. When the right-hand end of the cell reaches this duct, the cell's contents expand into the exhaust. This motion is transferred through the channel by an expansion wave (4) which propagates to the left at sonic speed. When this 1.2 FIGURE 6-60 wave reaches the left-hand end, the cell opens to the low-pressure air duct (A-LP) Full load Apparent compressor map of and fresh air is drawn into the cell. The flow to the right continues, but with Comprex connected to a 1.2-dm3 decreasing speed due to wave action (5, 6, 7, 8) and pressure losses at each end of diesel engine: charge-air pressure 1.00 the cell. When the dotted line-the interface between air and the exhaust gas- 0.01 ratio plotted versus net air 0.02 0.03 0.04 0.05 volume flow rate (total air flow reaches the right end of the cell, all the driving gas has left. The cell is then Typical net air volume flow rate, m3/s less scavenging air flow).46 274 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 275 PROBLEMS 6.1. A conventional spark-ignition engine operating with gasoline will not run smoothly (due to incomplete combustion) with an equivalence ratio leaner than about o = 0.8. It is desirable to extend the smooth operating limit of the engine to leaner equiva- lence ratios so that at part-throttle operation (with intake pressure less than 1 atmosphere) the pumping work is reduced. Leaner than normal operation can be 2 2 achieved by adding hydrogen gas (H2) to the mixture in the intake system. The addition of H2 makes the fuel-air mixture easier to burn. (a) The fuel composition with "mixed" fuel operation is H2 + C8H18-one mole of hydrogen to every mole of gasoline, which is assumed the same as isooctane. Pe 8 What is the stoichiometric air/fuel ratio for the "mixed " fuel? 5 (b) The lower heating value of H2 is 120 MJ/kg and for isooctane is 44.4 MJ/kg. pit PI What is the heating value per kilogram of fuel mixture? (c) Engine operation with isooctane and the mixed (H2 + C8H18) fuel is compared V in a particular engine at a part-load condition (brake mean effective pressure of (a) (b) 275 kPa and 1400 rev/min). You are given the following information about the FIGURE P6-3 engine operation: Fuel C&H18 H2 + C.H. 18 Equivalence ratio 0.8 0.5 load, the inlet valve is closed rapidly partway through the intake stroke at point 8. Gross indicated fuel conversion efficiency 0.35 0.4 The gas in the cylinder at inlet valve closing at 8 is then expanded isentropically to 1 with the inlet valve closed. The pressure p1 at the start of compression is the same Mechanical rubbing friction mep 138 kPa 138 kPa for both cycles. Inlet manifold pressure 46 kPa Pumping mep 55 kPa ? (a) Indicate on p-V diagrams the area that corresponds to the pumping work per cycle for cycles (a) and (b). Which area is greater? Estimate approximately the inlet manifold pressure and the pumping mean (b) Derive expressions for the pumping work per cycle W, in terms of m, c ,, y, T1, effective pressure with (H2 + C8H18) fuel. Explain your method and assumptions (p/P1), and the compression ratio re for cycles (a) and (b). Be consistent about the clearly. Note that mechanical efficiency n is defined as signs of the work transfers to and from the gas. (c) For y = 1.3, re = 8, and (p/p1) = 2 find the ratio W,(b)/W,(a), assuming the bmep bmep values of T1 and m are the same in both cases. imep. bmep + rfmep + pmep 6.4. For four-stroke cycle engines, the inlet and exhaust valve opening and closing crank 6.2. Hydrogen is a possible future fuel for spark-ignition engines. The lower heating angles are typically: IVO 15º BTC; IVC 50º ABC; EVO 55º BBC; EVC 10º ATC. value of hydrogen is 120 MJ/kg and for gasoline (C8H14) is 44 MJ/kg. The stoichio- Explain why these valve timings improve engine breathing relative to valve opening metric air/fuel ratio for hydrogen is 34.3 and for gasoline is 14.4. A disadvantage of and closing at the beginnings and ends of the intake and exhaust strokes. Are there hydrogen fuel in the SI engine is that the partial pressure of hydrogen in the H2-air additional design issues that are important? mixture reduces the engine's volumetric efficiency, which is proportional to the 6.5. Estimate approximately the pressure drop across the inlet valve about halfway partial pressure of air. Find the partial pressure of air in the intake manifold down- through the intake stroke and across the exhaust valve halfway through the exhaust stream of the hydrogen fuel-injection location at wide-open throttle when the total stroke, when the piston speed is at its maximum for a typical four-stroke cycle intake manifold pressure is 1 atmosphere; the equivalence ratio is 1.0. Then estimate spark-ignition engine with B = L = 85 mm at 2500 and 5000 rev/min at WOT. the ratio of the fuel energy per unit time entering a hydrogen-fueled engine operating Assume appropriate values for any valve and port geometric details required, and with a stoichiometric mixture to the fuel energy per unit time entering an identical for the gas composition and state. gasoline-fueled engine operating at the same speed with a stoichiometric mixture. 6.6. Using the data in Fig. 6-21, estimate the fraction of the original mass left in the (Note that the "fuel energy" per unit mass of fuel is the fuel's heating value.) cylinder: (a) at the end of the blowdown process and (b) at the end of the exhaust 6.3. Sketch (a) shows an ideal cycle p-V diagram for a conventional throttled spark- stroke. ignition engine, 1-2-3-4-5-6-7-1. The gas properties c ,, c ,, y, R throughout the cycle 6.7. Compare the engine residual gas fraction data in Fig. 6-19 with ideal cycle estimates are constant. The mass of gas in the cylinder is m. The exhaust pressure is pe. of residual gas fraction as follows. Using Eq. (5.47) plot the fuel-air cycle residual Sketch (b) shows an ideal cycle p-V diagram 1-2-3-4-5-6-8-1 for a spark- mass fraction x, against p:/pe for re = 8.5 on the same graph as the engine data in ignition engine with novel inlet valve timing. The inlet manifold is unthrottled; it has Fig. 6-19 at 1400 rev/min and 27º valve overlap. Assume T, = 1400 K and (y - 1)/ essentially the same pressure as the exhaust. To reduce the mass inducted at part y = 0.24 in Eq. (5.47). Suggest an explanation for any significant difference. .. 276 INTERNAL COMBUSTION ENGINE FUNDAMENTALS GAS EXCHANGE PROCESSES 277 6.8. One concept that would increase SI engine efficiency is early intake valve closing 11. Barnes-Moss, H. W.: "A Designers Viewpoint," in Passenger Car Engines, Conference Pro- (EIVC) where the intake valve closes before the piston reaches BC on the intake ceedings, pp. 133-147, Institution of Mechanical Engineers, London, 1975. stroke, thus limiting the amount of charge inducted into the cylinder. 12 Asmus, T. W.: " Valve Events and Engine Operation," SAE paper 820749, SAE Trans ., Vol. 91, 1982. (a) Explain why EIVC improves engine efficiency at part load. (Hint: consider what 13. SAE Recommended Practice, "Engine Terminology and Nomenclature-General," in SAE Hand- must happen to the inlet manifold pressure in order to maintain constant mass in book, J604d. the cylinder as the intake valve is closed sooner.) 14. Kastner, L. J ., Williams, T. J ., and White, J. B.: "Poppet Inlet Valve Characteristics and Their (b) This part load reduction in charge could be achieved by using late intake valve Influence on the Induction Process," Proc. Instn Mech. Engrs, vol. 178, pt. 1, no. 36, pp. 955-978, closing where the intake valve is not closed until the compression stroke has 1963-1964. pushed some of the cylinder gases back out into the intake manifold. Based on a 15. Woods, W. A ., and Khan, S. R.: "An Experimental Study of Flow through Poppet Valves," Proc. comparison of p-V diagrams, is this method inferior to EIVC? Instn Mech. Engrs, vol. 180, pt. 3N, pp. 32-41, 1965-1966. 6.9. An eight-cylinder turbocharged aftercooled four-stroke cycle diesel engine operates 16. Annand, W. J. D ., and Roe, G. E.: Gas Flow in the Internal Combustion Engine, Haessner Publi- shing, Newfoundland, N.J ., 1974. with an inlet pressure of 1.8 atmospheres at its maximum rated power at 2000 rev/ 17. Tanaka, K.: "Air Flow through Exhaust Valve of Conical Seat," Int. Congr. Appl. Mech ., vol. 1, min. B = 128 mm, L = 140 mm, n, (based on inlet manifold conditions of 1.8 atm pp. 287-295, 1931. and 325 K after the aftercooler) = 0.9. The compressor isentropic efficiency is 0.7. 18. Bicen, A. F ., and Whitelaw, J. H.: "Steady and Unsteady Air Flow through an Intake Valve of a (a) Calculate the power required to drive the turbocharger compressor. Reciprocating Engine," in Flows in Internal Combustion Engines-II, FED-vol. 20, Winter Annual (b) If the exhaust gas temperature is 650ºC and the turbocharger isentropic efficiency Meeting, ASME, 1984. is 0.65, estimate the pressure at turbine inlet. The turbine exhausts to the atmo- 19. Fukutani, I ., and Watanabe, E.: "An Analysis of the Volumetric Efficiency Characteristics of sphere. 4-Stroke Cycle Engines Using the Mean Inlet Mach Number Mim," SAE paper 790484, SAE Trans ., vol. 88, 1979. 6.10. The charging efficiency of two-stroke cycle diesel engines can be estimated from 20. Wallace, W. B.: "High-Output Medium-Speed Diesel Engine Air and Exhaust System Flow measurement of the concentration of O2 and CO2 in the burned gases within the Losses," Proc. Instn Mech. Engrs, vol. 182, pt. 3D, pp. 134-144, 1967-1968. cylinder, or in the exhaust blowdown pulse prior to any mixing with fresh air. The 21. Cole, B. N ., and Mills, B.: "The Theory of Sudden Enlargements Applied to Poppet Exhaust- engine bore = 125 mm, stroke = 150 mm, compression ratio = 15. The fuel flow rate Valve, with Special Reference to Exhaust-Pulse Scavenging," Proc. Instn Mech. Engrs, pt. 1B, pp. at 1800 rev/min is 1.6 g/s per cylinder. The conditions used to evaluate the air 364-378, 1953. density for the reference mass are 300 K and 1 atm. The molar concentrations (dry) 22. Toda, T ., Nohira, H ., and Kobashi, K.: "Evaluation of Burned Gas Ratio (BGR) as a Predomi- of CO2 and O2 in the in-cylinder burned gases are 7.2 and 10.4 percent (see Fig. nant Factor to NO„," SAE paper 760765, SAE Trans ., vol. 85, 1976. 4-22). The scavenging air flow rate is 80 g/s. Evaluate (a) the charging efficiency, (b) 23. Benson, J. D ., and Stebar, R. F.: “Effects of Charge Diluation on Nitric Oxide Emission from a Single-Cylinder Engine," SAE paper 710008, SAE Trans ., vol. 80, 1971. the delivery ratio, and (c) the trapping efficiency (assuming the trapped mass equals 24. Tabaczynski, R. J ., Heywood, J. B ., and Keck, J. C.: "Time-Resolved Measurements of Hydrocar- the reference mass). bon Mass Flow Rate in the Exhaust of a Spark-Ignition Engine," SAE paper 720112, SAE Trans ., vol. 81, 1972. 25. Caton, J. A ., and Heywood, J. B.: " An Experimental and Analytical Study of Heat Transfer in an Engine Exhaust Port," Int. J. Heat Mass Transfer, vol. 24, no. 4, pp. 581-595, 1981. REFERENCES 26. Caton, J. A.: "Comparisons of Thermocouple, Time-Averaged and Mass-Averaged Exhaust Gas 1. Khovakh, M.: Motor Vehicle Engines, English Translation, Mir Publishers, Moscow, 1976. Temperatures for a Spark-Ignited Engine," SAE paper 820050, 1982. 2. Matsuoka, S ., Tasaka, H ., and Tsuruta, J.: "The Evaporation of Fuel and Its Effect on Volu- 27. Phatak, R. G.: "A New Method of Analyzing Two-Stroke Cycle Engine Gas Flow Patterns," SAE paper 790487, SAE Trans ., vol. 88, 1979. metric Efficiency," JARI technical memorandum no. 2, pp. 17-22, 1971. 3. Takizawa, M ., Uno, T ., Oue, T ., and Yura, T.: " A Study of Gas Exchange Process Simulation of 28. Rizk, W.: "Experimental Studies of the Mixing Processes and Flow Configurations in Two-Cycle an Automotive Multi-Cylinder Internal Combustion Engine," SAE paper 820410, SAE Trans ., Engine Scavenging," Proc. Instn Mech. Engrs, vol. 172, pp. 417-437, 1958. vol. 91, 1982. 29. Dedeoglu, N.: "Scavenging Model Solves Problems in Gas Burning Engine," SAE paper 710579, SAE Trans ., vol. 80, 1971. 4. Kay, I. W.: " Manifold Fuel Film Effects in an SI Engine," SAE paper 780944, 1978. 5. Ohata, A ., and Ishida, Y.: "Dynamic Inlet Pressure and Volumetric Efficiency of Four Cycle 30. Sher, E.: "Investigating the Gas Exchange Process of a Two-Stroke Cycle Engine with a Flow Four Cylinder Engine," SAE paper 820407, SAE Trans ., vol. 91, 1982. Visualization Rig," Israel J. Technol ., vol. 20, pp. 127-136, 1982. 6. Benson, R. S ., and Whitehouse, N. D.: Internal Combustion Engines, vol. 2, Pergamon Press, 1979. 31. Jante, A.: "Scavenging and Other Problems of Two-Stroke Cycle Spark-Ignition Engines," SAE paper 680468, SAE Trans ., vol. 77, 1968. 7. Taylor, C. F.: The Internal-Combustion Engine in Theory and Practice, vol. 1, 2d ed ., revised, MIT Press, Cambridge, Mass ., 1985. 32. Kannappan, A.: "Cumulative Sampling Technique for Investigating the Scavenging Process in Two-Stroke Engine," ASME paper 74-DGP-11, 1974. 8. Hofbauer, P ., and Sator, K.: "Advanced Automotive Power Systems, Part 2: A Diesel for a Subcompact Car," SAE paper 770113, SAE Trans ., vol. 86, 1977. 33. Ohigashi, S ., Kashiwada, Y ., and Achiwa, J.: "Scavenging the 2-Stroke Diesel Engine," Bull. JSME, vol. 3, no. 9, pp. 130-136, 1960. 9. Armstrong, D. L ., and Stirrat, G. F.: "Ford's 1982 3.8L V6 Engine," SAE paper 820112, SAE Trans ., vol. 91, 1982. 34. Huber, E. W.: "Measuring the Trapping Efficiency of Internal Combustion Engines through 10. Chapman, M ., Novak, J. M ., and Stein, R. A.: " Numerical Modeling of Inlet and Exhaust Flows Continuous Exhaust Gas Analysis," SAE paper 710144, SAE Trans ., vol. 80, 1971. in Multi-Cylinder Internal Combustion Engines," in Flows in Internal Combustion Engines, 35. Blair, G. P ., and Kenny, R. G.: "Further Developments in Scavenging Analysis for Two-Cycle Engines," SAE paper 800038, SAE Trans ., vol. 89, 1980. Winter Annual Meeting, ASME, New York, 1982. 278 INTERNAL COMBUSTION ENGINE FUNDAMENTALS 36. Baudequin, F ., and Rochelle, P.: "Some Scavenging Models for Two-Stroke Engines," Proc. Insta Mech. Engrs, Automobile Division, vol. 194, no. 22, pp. 203-210, 1980. 37. Gyssler, G.: “Problems Associated with Turbocharging Large Two-Stroke Diesel Engines," Proc. CHAPTER CIMAC, paper B.16, 1965. 38. Annand, W. J. D.: "Compressible Flow through Square-Edged Orifices: An Empirical Approx- imation for Computer Calculations," J. Mech. Engng Sci ., vol. 8, p. 448, 1966. 39. Benson, R. S.: "Experiments on a Piston Controlled Port," The Engineer, vol. 210, pp. 875-880, 7 1960. 40. Watson, N ., and Janota, M. S.: Turbocharging the Internal Combustion Engine, Wiley-Interscience Publications, John Wiley, New York, 1982. 41. Bhinder, F. S.: "Supercharging Compressors-Problems and Potential of the Various Alterna- tives," SAE paper 840243, 1984. SI ENGINE 42. Bhinder, F. S.: "Some Fundamental Considerations Concerning the Pressure Charging of Small FUEL Diesel Engines," SAE paper 830145, 1983. 43. Brandstetter, W ., and Dziggel, R.: "The 4- and 5-Cylinder Turbocharged Diesel Engines for METERING Volkswagen and Audi," SAE paper 820441, SAE Trans ., vol. 91, 1982. 44. SAE Recommended Practice, "Turbocharger Nomenclature and Terminology," in SAE Hand- AND MANIFOLD book, J922. 45. Flynn, P. F.: "Turbocharging Four-Cycle Diesel Engines," SAE paper 790314, SAE Trans ., vol. PHENOMENA 88, 1979. 46. Gyarmathy, G.: "How Does the Comprex Pressure-Wave Supercharger Work?" SAE paper 830234, 1983. 47. Kollbrunner, T. A.: "Comprex Supercharging for Passenger Diesel Car Engines," SAE paper 800884, SAE Trans ., vol. 89, 1980. 7.1 SPARK-IGNITION ENGINE MIXTURE REQUIREMENTS The task of the engine induction and fuel systems is to prepare from ambient air and fuel in the tank an air-fuel mixture that satisfies the requirements of the engine over its entire operating regime. In principle, the optimum air/fuel ratio for a spark-ignition engine is that which gives the required power output with the lowest fuel consumption, consistent with smooth and reliable operation. In prac- tice, the constraints of emissions control may dictate a different air/fuel ratio, and may also require the recycling of a fraction of the exhaust gases (EGR) into the intake system. The relative proportions of fuel and air that provide the lowest fuel consumption, smooth reliable operation, and satisfy the emissions require- ments, at the required power level, depend on engine speed and load. Mixture requirements and preparation are usually discussed in terms of the air/fuel ratio or fuel/air ratio (see Sec. 2.9) and percent EGR [see Eq. (4.2)]. While the fuel metering system is designed to provide the appropriate fuel flow for the actual air flow at each speed and load, the relative proportions of fuel and air can be stated more generally in terms of the fuel/air equivalence ratio o, which is the actual fuel/air ratio normalized by dividing by the stoichiometric fuel/air ratio [Eq. 279 280 INTERNAL COMBUSTION ENGINE FUNDAMENTALS SI ENGINE FUEL METERING AND MANIFOLD PHENOMENA 281 (3.8)]. The combustion characteristics of fuel-air mixtures and the properties of combustion products, which govern engine performance, efficiency, and emis- or the amount of recycled exhaust slows down the combustion process and increases its variability from cycle to cycle. A certain minimum combustion sions, correlate best for a wide range of fuels relative to the stoichiometric mixture proportions. Where appropriate, therefore, the equivalence ratio will be repeatability or stability level is required to maintain smooth engine operation. Deterioration in combustion stability therefore limits the amount of dilution an used as the defining parameter. A typical value for the stoichiometric air/fuel engine can tolerate. As load decreases, less dilution of the fresh mixture can be ratio of gasoline is 14.6.+ Thus, for gasoline, tolerated because the internal dilution of the mixture with residual gas increases 14.6 (see Sec. 6.4). At idle conditions, the fresh mixture will not usually tolerate any A/F 7.1) EGR and may need to be stoichiometric or fuel-rich to obtain adequate com- bustion stability. The effects of equivalence ratio variations on engine combustion, emissions, Mixture composition requirements over the engine load and speed range and performance are discussed more fully in Chaps. 9, 11, and 15. A brief are illustrated schematically for the two approaches outlined above in Fig. 7-1. If summary is sufficient here. Mixture requirements are different for full-load (wide- stoichiometric operation and EGR are not required for emissions control, as load open throttle) and for part-load operation. At the former operating condition, increases the mixture is leaned out from a fuel-rich or close-to-stoichiometric complete utilization of the inducted air to obtain maximum power for a given composition at very light load. As wide-open throttle operation is approached at displaced volume is the critical issue. Where less than the maximum power at a each engine speed, the mixture is steadily enriched to rich-of-stoichiometric at the given speed is required, efficient utilization of the fuel is the critical issue. At maximum bmep point. With the stoichiometric operating conditions required for wide-open throttle, maximum power for a given volumetric efficiency is obtained three-way-catalyst-equipped engines, when EGR is used, the percentage of re- with rich-of-stoichiometric mixtures, o ~ 1.1 (see the discussion of the fuel-air cycled exhaust increases from zero at light load to a maximum at mid-load, and cycle results in Sec. 5.5.3). Mixtures that are richer still are sometimes used to then decreases to zero as wide-open throttle conditions are approached so increase volumetric efficiency by increasing the amount of charge cooling that maximum bmep can be obtained. Combinations of these strategies are possible. accompanies fuel vaporization [see Eq. (6.5)], thereby increasing the inducted air For example, lean operation at light load can be used for best efficiency, and density. At part-load (or part-throttle) operating conditions, it is advantageous to dilute the fuel-air mixture, either with excess air or with recycled exhaust gas. 1.2 .Low Mid 12 This dilution improves the fuel conversion efficiency for three reasons:1 (1) the High speed expansion stroke work for a given expansion ratio is increased as a result of the Equivalence ratto o. 1.0 14 change in thermodynamic properties of the burned gases-see Secs. 5.5.3 and 5.7.4; (2) for a given mean effective pressure, the intake pressure increases with No EGR 16 0.8 increasing dilution, so pumping work decreases-see Fig. 5-10; (3) the heat losses 18 to the walls are reduced because the burned gas temperatures are lower. In the 0 20 absence of strict engine NO, emission requirements, excess air is the obvious 100% diluent, and at part throttle engines have traditionally operated lean. When tight control of NO„, HC, and CO emissions is required, operation of the engine with 20 a stoichiometric mixture is advantageous so that a three-way catalystt can be ¢ = 1.0 used to clean up the exhaust. The appropriate diluent is then recycled exhaust EGR. 4 gases which significantly reduces NO, emissions from the engine itself. The Low High speed amount of diluent that the engine will tolerate at any given speed and load Mid depends on the details of the engine's combustion process. Increasing excess air -- 0 100% Intake mass flow rate FIGURE 7-1 Typical value only. Most gasolines have (A/F), in the range 14.4 to 14.7. (A/F), could lie between 14.1 and 15.2. Typical mixture requirements for two common operating strategies. Top diagram shows equivalence ratio variation with intake mass flow rate (percent of maximum flow at rated speed) at constant low, $ A three-way catalyst system, when operated with a close-to-stoichiometric mixture, achieves sub- mid, and high engine speeds. Bottom diagram shows recycled exhaust (EGR) schedule as a function of stantial reductions in NO ,, CO, and HC emissions simultaneously; see Sec. 11.6.2. intake flow rate, for low, mid, and high speeds for stoichiometric operation. 282 INTERNAL COMBUSTION ENGINE FUNDAMENTALS SI ENGINE FUEL METERING AND MANIFOLD PHENOMENA 283 stoichiometric mixtures (with a three-way catalyst) and/or EGR can be used at Air mid loads to control NO, emissions. In practical spark-ignition engine induction systems, the fuel and air dis- tribution between engine cylinders is not uniform (and also varies in each individ- Fuel ual cylinder on a cycle-by-cycle basis). A spread of one or more air/fuel ratios between the leanest and richest cylinders over the engine's load and speed range is not uncommon in engines with conventional carburetors. The average mixture FIGURE 7-2 composition must be chosen to avoid excessive combustion variability in the Schematic of elementary carburetor. leanest operating cylinder. Thus, as the spread in mixture nonuniformity 1 Inlet section increases, the mean equivalence ratio must be moved toward stoichiometric and 2 Venturi throat away from the equivalence ratio which gives minimum fuel consumption. 3 Float chamber 4 Pressure equalizing passage 5 Calibrated orifice 7.2 CARBURETORS 6 Fuel discharge tube 7 Throttle plate 7.2.1 Carburetor Fundamentals A carburetor has been the most common device used to control the fuel flow into rev/min), is the characteristic time of this periodic cylinder filling process. Gener- the intake manifold and distribute the fuel across the air stream. In a carburetor, ally, the characteristic times of changes in throttle setting are longer; it takes the air flows through a converging-diverging nozzle called a venturi. The pressure several engine operating cycles to reestablish steady-state engine operation after a difference set up between the carburetor inlet and the throat of the nozzle (which sudden change in throttle position.2 It is usually assumed that the flow processes depends on the air flow rate) is used to meter the appropriate fuel flow for that in the carburetor can be modeled as quasi steady. air flow. The fuel enters the air stream through the fuel discharge tube or ports in the carburetor body and is atomized and convecteur the air stream past the FLOW THROUGH THE VENTURI. Equation (C.8) in App. C relates the mass throttle plate and into the intake manifold. Fuel evaporation starts within the flow rate of a gas through a flow restriction to the upstream stagnation pressure carburetor and continues in the manifold as fuel droplets move with the air flow and temperature, and the pressure at the throat. For the carburetor venturi: and as liquid fuel flows over the throttle and along the manifold walls. A modern carburetor which meters the appropriate fuel flow into the air stream over the im = CDT ATPO (PT) 1/ 27 [ 1 complete engine operating range is a highly developed and complex device. There VRTO (po) (7.2) are many types of carburetors; they share the same basic concepts which we will now examine. where Cp, and Ar are the discharge coefficient and area of the venturi throat, Figure 7-2 shows the essential components of an elementary carburetor. respectively. If we assume the velocity at the carburetor inlet can be neglected, The air enters the intake section of the carburetor (1) from the air cleaner which Eq. (7.2) can be rearranged in terms of the pressure drop from upstream condi- removes suspended dust particles. The air then flows into the carburetor ventun tions to the venturi throat for the air stream, Apa = Po - PT, as (a converging-diverging nozzle) (2) where the air velocity increases and the pres- ma = CDT AT(20a, Ap )1/20 sure decreases. The fuel level is maintained at a constant height in the float (7.3) where chamber (3) which is connected via an air duct (4) to the carburetor intake section (1). The fuel flows through the main jet (a calibrated orifice) (5) as a result of the pressure difference between the float chamber and the venturi throat and through the fuel discharge nozzle (6) into the venturi throat where the air stream 1 - ( PT / Po ) (7.4) atomizes the liquid fuel. The fuel-air mixture flows through the diverging section and accounts for the effects of compressibility. Figure C-3 shows the value of O as of the venturi where the flow decelerates and some pressure recovery occurs. Ibe a function of pressure drop. For the normal carburetor operating range, where flow then passes the throttle valve (7) and enters the intake manifold. Apa/Po < 0.1, the effects of compressibility which reduce " below 1.0 are small. Note that the flow may be unsteady even when engine load and speed are constant, due to the periodic filling of each of the engine cylinders which draws FLOW THROUGH THE FUEL ORIFICE. Since the fuel is a liquid and therefore air through the carburetor venturi. The induction time, 1/(2N) (20 ms at 1500 essentially incompressible, the fuel flow rate through the main fuel jet is given by 284 INTERNAL COMBUSTION ENGINE FUNDAMENTALS SI ENGINE FUEL METERING AND MANIFOLD PHENOMENA 285 Eq. (C.2) in App. C as 1.0 im, = CD. A.(2p, Aps)1/2 CDT (7.5) 1.8 CD. where CD, and A, are the discharge coefficient and area of the orifice, respec- tively, Ap is the pressure difference across the orifice, and the orifice area is 0.6 assumed much less than the passage area. Usually, the fuel level in the float chamber is held below the fuel discharge nozzle, as shown in Fig. 7-2, to prevent im(A/F), fuel spillage when the engine is inclined to the horizontal (e.g ., in a vehicle on a ma slope). Thus, Ap, = Apa - Psgh 1.2 13 1.0- 14 where h is typically of order 10 mm. 16 AIF The discharge coefficient CD, in Eq. (7.5) represents the effect of all devi- 0.8 18 FIGURE 7-3 ations from the ideal one-dimensional isentropic flow. It is influenced by many 0 2 3 4 5 Performance of elementary carburetor: varia- factors of which the most important are the following: (1) fluid mass flow rate; (2) Ap, kN/m2 tion of Cpr, Cp ., , my (A/F);, ma, and equiva- orifice length/diameter ratio; (3) orifice/approach-area ratio; (4) orifice surface lence ratio o with venturi pressure drop. area; (5) orifice surface roughness; (6) orifice inlet and exit chamfers; (7) fluid specific gravity; (8) fluid viscosity; and (9) fluid surface tension. The use of the orifice Reynolds number, Re, = pVD./u, as a correlating parameter for the dis- give a stoichiometric mixture at an air flow rate corresponding to 1 kN/m2 charge coefficient accounts for effects of mass flow rate, fluid density and vis- venturi pressure drop (middle graph in Fig. 7-3). At higher air flow rates, the cosity, and length scale to a good first approximation. The discharge coefficient carburetor will deliver a fuel-rich mixture; at very high flow rates it will even- of a typical carburetor main fuel-metering system orifice increases smoothly with tually deliver an essentially constant equivalence ratio. At lower air flow rates, increasing Re ,. 3 the mixture delivered leans out rapidly. Thus, the elementary carburetor cannot provide the variation in mixture ratio which the engine requires over the com- CARBURETOR PERFORMANCE. The air/fuel ratio delivered by this carburetor plete load range at any given speed (see Fig. 7-1). The deficiencies of the elementary carburetor can be summarized as is given by follows: me = 1 Apa 1/2 Apa - Psgh ) (7.6) ims 1. At low loads the mixture becomes leaner; the engine requires the mixture to be enriched at low loads. and the equivalence ratio ¢ [=(4/F)/(A/F)] by 2. At intermediate loads, the mixture equivalence ratio increases slightly as the 6 - AND- (ED) (40) ( Br )112 ( 1 - 2/9h ) 1/2 air flow increases. The engine requires an almost constant equivalence ratio. Apa ) (7.7) 3. As the air flow approaches the maximum wide-open throttle value, the equiva- lence ratio remains essentially constant. However, the mixture equivalence where (A/F), is the stoichiometric air/fuel ratio. The terms A ., AT, Ps, and Pao ratio should increase to 1.1 or greater to provide maximum engine power. are all constant for a given carburetor, fuel, and ambient conditions. Also, except 4. The elementary carburetor cannot compensate for transient phenomena in the for very low flows, prgh < Apa. The discharge coefficients, CD, and CD ,, and intake manifold. Nor can it enrich the mixture during engine starting and vary with flow rates, however. Hence, the equivalence ratio of the mixture warm-up. delivered by an elementary carburetor is not constant. Figure 7-3 illustrates the performance of the elementary carburetor. The top 5. The elementary carburetor cannot adjust to changes in ambient air density (due primarily to changes in altitude). set of curves shows how O, Cpr, and Cp, typically vary with the venturi pressure drop.4 Note that for Apa S prgh there is no fuel flow. Once fuel starts to flow, as a consequence of these variations the fuel flow rate increases more rapidly than 7.2.2 Modern Carburetor Design the air flow rate. The carburetor delivers a mixture of increasing fuel/air equiva- lence ratio as the flow rate increases. Suppose the venturi and orifice are sized to The changes required in the elementary carburetor so that it provides the equiva- lence ratio versus air flow distribution shown in Fig. 7-1 are: SI ENGINE FUEL METERING AND MANIFOLD PHENOMENA 287 286 INTERNAL COMBUSTION ENGINE FUNDAMENTALS 1. The main metering system must be compensated to provide essentially constant at the main venturi throat, a higher vacuum Apt = pmgh, is obtained at the boost lean or stoichiometric mixtures over the 20 to 80 percent air flow range. venturi throat which can be used to obtain more precise metering of the fuel (pm Is the manometer fluid density). Best results are obtained with the boost venturi 2. An idle system must be added to meter the fuel flow at idle and light loads. exit slightly upstream (~5 mm) of the main venturi throat. Because only a frac- 3. An enrichment system must be added so the engine can provide its maximum tion of the total air flow goes through the boost venturi, the use of multiple power as wide-open throttle is approached. venturis makes it possible to obtain a high velocity air stream (up to 200 m/s) 4. An accelerator pump which injects additional fuel when the throttle is opened where the fuel is introduced at the boost venturi throat, and adequate vacuum, rapidly is required to maintain constant the equivalence ratio delivered to the and to reduce the pressure loss across the total venturi system, without increasing engine cylinder. the height of the carburetor. The fuel is better atomized in the smaller boost 5. A choke must be added to enrich the mixture during engine starting and- venturi with its higher air velocity, and since this air and fuel mixture is dis- warm-up to ensure a combustible mixture within each cylinder at the time of charged centrally into the surrounding venturi, a more homogeneous mixture results. The vacuum developed at the venturi throat of a typical double-venturi ignition. 6. Altitude compensation is required to adjust the fuel flow to changes in air system is about twice the theoretical vacuum of a single venturi of the same flow area.5 A triple-venturi system can be used to give further increases in metering density. signal. The overall discharge coefficient of a multiple-venturi carburetor is lower In addition, it is necessary to increase the magnitude of the pressure drop than a single-venturi carburetor of equal cross-sectional area. The throat area of available for controlling the fuel flow. Two common methods used to achieve this the main venturi in a multiple-venturi system is usually increased, therefore, above the single-venturi size to compensate for this. Some decrease in air stream are the following. velocity is tolerated to maintain a high discharge coefficient (and hence a high BOOST VENTURIS. The carburetor venturi should give as large a vacuum at the volumetric efficiency).6 throat as possible at maximum air flow, within the constraints of a low pressure loss across the complete venturi and diffuser. In a single venturi, as the diameter MULTIPLE-BARREL CARBURETORS. Use of carburetors with venturi systems in of the throat is decreased at a given air flow to increase the flow velocity and parallel is a common way of maintaining an adequate part-load metering signal, hence the metering signal at the throat, the pressure loss increases. A higher high volumetric efficiency at wide-open throttle, and minimum carburetor height vacuum signal at the venturi throat and higher velocities for improved atom- as engine size and maximum air flow increases. As venturi size in a single-barrel ization can be obtained without increasing the overall pressure loss through the carburetor is increased to provide a higher engine air flow at maximum power, use of multiple venturis. Figure 7-4 shows the geometry and the pressure distribu- the venturi length increases and the metering signal generated at low flows tion in a typical double-venturi system. A boost venturi is positioned upstream of decreases. Maximum wide-open throttle air flow is some 30 to 70 times the idle the throat of the larger main venturi, with its discharge at the location of air flow (the value depending on engine displacement). Two-barrel carburetors maximum velocity in the main venturi. Only a fraction of the air flows through usually consist of two single-barrel carburetors mounted in parallel. Four-barrel the boost venturi. Since the pressure at the boost venturi exit equals the pressure carburetors consist of a pair of two-barrel carburetors in parallel, with throttle plates compounded on two shafts. Air flows through the primary barrel(s) at low and intermediate engine loads. At higher loads, the throttle plate(s) on the sec- ondary barrel(s) (usually of larger cross-sectional area) start to open when the air flow exceeds about 50 percent of the maximum engine air flow. There are many different designs of complete carburetors. The operating principles of the methods most commonly used to achieve the above listed modi- fications will now be reviewed. Figure 7-5 shows a schematic of a conventional modern carburetor and the names of the various components and fuel passages. COMPENSATION OF MAIN METERING SYSTEM. Figure 7-6 shows a main fuel- metering system with air-bleed compensation. As the pressure at the venturi throat decreases, air is bled through an orifice (or series of orifices) into the main fuel well. This flow reduces the pressure difference across the main fuel-metering FIGURE 7-4 Schematic of carburetor double-venturi system. office which no longer experiences the full venturi vacuum. The mixing of bleed 288 INTERNAL COMBUSTION ENGINE FUNDAMENTALS SI ENGINE FUEL METERING AND MANIFOLD PHENOMENA 289 Fuel Air small and does not significantly affect the composition of the mixture. The air- bleed mass flow rate is given by ima, = CD. Ab[2(Po - P3)Pa]1/2 (7.8) where CD, and Ab are the discharge coefficient and the area of the air-bleed orifice. The fuel mass flow rate through the fuel orifice is given by 10 im, = CD.A.[2(P1 - P2)P,]1/2 (7.9) where P1 = Po + Psgh3 and P2 = P3 + Psg (h3 - h2) 6 14. The density of the emulsion fem in the main well and nozzle is usually approximated by 12 13 May + ms FIGURE 7-5 Pem im Jea + msles (7.10) Schematic of modern carburetor. 1 Main venturi 8 Throttle plate Since typical values are p = 730 kg/m3 and pa = 1.14 kg/m3, usually ps >> 2 Boost venturi 9 Idle air-bleed orifice Pem >> Pa. Thus, as the air-bleed flow rate increases, the height of the column of 3 Main metering spray tube or nozzle 10 Idle fuel orifice 4 Air-bleed orifice 11 Idle mixture orifice emulsion becomes less significant. However, the decrease in emulsion density due 5 Emulsion tube or well 12 Transition orifice to increasing air bleed increases the flow velocity, which results in a significant 6 Main fuel-metering orifice 13 Idle mixture adjusting screw pressure drop across the main nozzle. This pressure drop depends on nozzle 7 Float chamber 14 Idle throttle setting adjusting screw length and diameter, fuel flow rate, bleed air flow rate, relative velocity between Fuel enters the air stream from the main metering system through (3). At idle, fuel enters air at (11). fuel and bleed air, and fuel properties. It is determined empirically, and has been During transition, fuel enters at (11), (12), and (3). (Courtesy S.p.A.E. Weber.) found to correlate with Pem [as defined by Eq. (7.10)].2, 6 The pressure loss at the main discharge nozzle with two-phase flow can be several times the pressure loss air with the fuel forms an emulsion which atomizes more uniformly to smaller with single-phase flow. drop sizes on injection at the venturi throat. The schematic in Fig. 7-6 illustrates Figure 7-7 illustrates the behavior of the system shown in Fig. 7-6: ma, my, the operating principle. When the engine is not running, the fuel is at the same and the fuel/air equivalence ratio o are plotted against Ap ,. Once the bleed level in the float bowl and in the main well. With the engine running, as the system is operating (Ap) > pr gh2) the fuel flow rate is reduced below its equiva- throttle plate is opened, the air flow and the vacuum in the venturi throat lent elementary carburetor value (the A) = 0 line). As the bleed orifice area is increase. For Apy(=Po - Po) < prgh1, there is no fuel flow from the main meter. increased, in the limit of large A, and neglecting the pressure losses in the main ing system. For prgh1 < Ap, < prghz, only fuel flows through the main well and nozzle, the fuel flow rate remains constant (A) -> co). An appropriate choice of nozzle, and the system operates just like an elementary carburetor. For Ap, > bleed orifice area A, will provide the desired equivalence ratio versus pressure ps gh2, air enters the main well together with fuel. The amount of air entering the drop or air flow characteristic. well is controlled by the size of the main air-bleed orifice. The amount of air is Additional control flexibility is obtained in practice through use of a second orifice, or of a series of holes in the main well or emulsion tube as shown in Fig. 7-5. Main metering systems with controlled air bleed provide reliable and stable Po Po control of mixture composition at part throttle engine operation. They are P4 Po simple, have considerable design flexibility, and atomize the fuel effectively. In some carburetor designs, an additional compensation system consisting of a tapered rod or needle in the main metering orifice is used. The effective open area 22 of the main metering orifice, and hence the fuel flow rate, can thus be directly related to throttle position (and manifold vacuum). FIGURE 7-6 P3 Schematic of main metering system with air-bleed A wide range of two-phase flow patterns can be generated by bleeding an compensation. air flow into a liquid flow. Fundamental studies of the generation and flow of P2 Pi SI ENGINE FUEL METERING AND MANIFOLD PHENOMENA 291 290 INTERNAL COMBUSTION ENGINE FUNDAMENTALS systems are coupled, they interact and the main system behavior in this transition reyion is modified by the fuel flow through the idle system. The total combined fuel flow provides a rich (or close-to-stoichiometric) mixture at idle, a progressive ma leaning of the mixture as air flow increases, and eventually (as the main system takes over full control of the fuel flow rate) an approximately constant mixture composition. -- Ab = 0 - Pfgh2- - POWER ENRICHMENT SYSTEM. This system delivers additional fuel to enrich ing the mixture as wide-open throttle is approached so the engine can deliver its maximum power. The additional fuel is normally introduced via a submerged valve which communicates directly with the main discharge nozzle, bypassing the metering orifice. The valve, which is spring loaded, is operated either mechani- -- - Ab = 0 cally through a linkage with the throttle plate (opening as the throttle approaches 1.2+ its wide-open position) or pneumatically (using manifold vacuum). 1.0 FIGURE 7-7 0.8 Ab Metering characteristics of system with air-bleed Ab + 00 compensation: mass flow rate of air m ., mass ACCELERATOR PUMP. When the throttle plate is opened rapidly, the fuel-air Apy = Po - Pv flow rate of fuel m , and equivalence ratio o as mixture flowing into the engine cylinder leans out temporarily. The primary functions of venturi pressure drop for different reason for this is the time lag between fuel flow into the air stream at the carbu- - Pf8h2- air-bleed orifice areas Ab. retor and the fuel flow past the inlet valve (see Sec. 7.6.3). While much of the fuel flow into the cylinder is fuel vapor or small fuel droplets carried by the air two-phase mixtures in small diameter tubes with bleed holes similar to those used stream, a fraction of the fuel flows onto the manifold and port walls and forms a in carburetors have been carried out.7 For a given pipe and bleed hole size, the liquid film. The fuel which impacts on the walls evaporates more slowly than fuel type of flow pattern set up depends on the flow rates of the two phases. carried by the air stream and introduces a lag between the air/fuel ratio produced at the carburetor and the air/fuel ratio delivered to the cylinder. An accelerator IDLE SYSTEM. The idle system is required because at low air flows through the pump is used as the throttle plate is opened rapidly to supply additional fuel into carburetor insufficient vacuum is created at the venturi throat to draw fuel into the air stream at the carburetor to compensate for this leaning effect. Typically, the air stream. However, at idle and light loads, the manifold vacuum is high, fuel is supplied to the accelerator pump chamber from the float chamber via a with the pressure drop occurring across the almost-closed throttle plate. This low small hole in the bottom of the fuel bowl, past a check valve. A pump diaphragm manifold pressure at idle is exploited for the idle fuel system by connecting the and stem is actuated by a rod attached to the throttle plate lever. When the main fuel well to an orifice in the throttle body downstream of the throttle plate. throttle is opened to increase air flow, the rod-driven diaphragm will increase the Figure 7-5 shows the essential features of an idle system. The main well (5) is fuel pressure which shuts the valve and discharges fuel past a discharge check connected through one or more orifices (10), past one or more idle air-bleed valve or weight in the discharge passage, through the accelerator pump discharge orifices (or holes) (9), past an idle mixture adjusting screw (13), to the idle dis- nozzle(s), and into the air stream. A calibrated orifice controls the fuel flow. A charge port (11) in the throttle body. Emulsifying air is admitted into the idle spring connects the rod and diaphragm to extend the fuel discharge over the system [at (9) and (12)] to reduce the pressure drop across the idle port and appropriate time period and to reduce the mechanical strain on the linkage. permit larger-sized ports (which are easier to manufacture) to be used. Satisfac- tory engine operation at idle is obtained empirically by means of the idle throttle CHOKE. When a cold engine is started, especially at low ambient temperatures, position stop adjustment (14) and the idle mixture adjustment (13). As the throttle the following factors introduce additional special requirements for the complete is opened from its idle position, the idle metering system performs a transitional carburetor: function. One or more holes (12) located above the idle discharge port (11) assist as air bleeds when the throttle is at or near its idle position. As the throttle plate 1. Because the starter-cranked engine turns slowly (70 to 150 rev/min) the intake opens and the air flow increases, these additional discharge holes are exposed to manifold vacuum developed during engine start-up is low. the manifold vacuum. Additional fuel is forced out of these holes into the air stream to provide the appropriate mixture ratio. As the throttle plate is opened 2. This low manifold vacuum draws a lower-than-normal fuel flow from the car- buretor idle system. further, the main fuel metering system starts to supply fuel also. Because the two 292 INTERNAL COMBUSTION ENGINE FUNDAMENTALS SI ENGINE FUEL METERING AND MANIFOLD PHENOMENA 293 3. Because of the low manifold temperature and vacuum, fuel evaporation in the carburetor, manifold, and inlet port is much reduced. bring in the power-enrichment system at a lower air flow rate due to decreased manifold vacuum. To reduce the impact to changes in altitude on engine emis- Thus, during cranking, the mixture which reaches the engine cylinder would be sions of CO and HC, modern carburetors are altitude compensated. A number of too lean to ignite. Until normal manifold conditions are established, fuel distribu- methods can be used to compensate for changes in ambient pressure with altitude: tion is also impaired. To overcome these deficiencies and ensure prompt starts and smooth operation during engine warm-up, the carburetor must supply a fuel-rich mixture. This is obtained with a choke. Once normal manifold condi- 1. Venturi bypass method. To keep the air volume flow rate through the venturi tions are established, the choke must be excluded. The primary element of a equal to what it was at sea-level atmospheric pressure (calibration condition), typical choke system is a plate, upstream of the carburetor, which can close off a bypass circuit around the venturi for the additional volume flow is provided. the intake system. At engine start-up, the choke plate is closed to restrict the air 2. Auxiliary jet method. An auxiliary fuel metering orifice with a pressure- flow into the carburetor barrel. This causes almost full manifold vacuum within controlled tapered metering rod connects the fuel bowl to the main well in the venturi which draws a large fuel flow through the main orifice. When the parallel with the main metering orifice. engine starts, the choke is partly opened to admit the necessary air flow and 3. Fuel bowl back-suction method. As altitude increases, an aneroid bellows moves reduce the vacuum in the venturi to avoid flooding the intake with fuel. As the a tapered rod from an orifice near the venturi throat, admitting to the bowl an engine warms up, the choke is opened either manually or automatically with a increasing amount of the vacuum signal developed at the throat. thermostatic control. For normal engine operation the choke plate is fully open 4. Compensated air-bleed method. The orifices in the bleed circuits to each carbu- and does not influence carburetor performance. A manifold vacuum control is retor system are fitted with tapered metering pins actuated by a single aneroid often used to close the choke plate partially if the engine is accelerated during bellows.8 warm-up. During engine warm-up the idle speed is increased to prevent engine stalling. A fast idle cam is rotated into position by the automatic choke lever. TRANSIENT EFFECTS. The pulsating and transient nature of the flow through a carburetor during actual engine operation is illustrated by the data shown in ALTITUDE COMPENSATION. An inherent characteristic of the conventional Fig. 7-8.2 The changes in pressure with time in the intake manifold and at the float type carburetor is that it meters fuel flow in proportion to the air volume boost venturi throat of a standard two-barrel carburetor installed on a pro- flow rate. Air density changes with ambient pressure and temperature, with duction V-8 engine are shown as the throttle is opened from light load (22º) to changes due to changes in pressure with altitude being most significant. For wide-open throttle at 1000 rev/min. Note the rapid increase in boost venturi example, at 1500 m above sea level, mean atmospheric pressure is 634 mmHg or suction as the throttle is suddenly opened. This results from the sudden large 83.4 percent of the mean sea-level value. While ambient temperature variations, increase in the air flow rate and corresponding increase in air velocity within the winter to summer, can produce changes of comparable magnitude, the tem- boost venturi. Note also that the pressure fluctuations decay rapidly, and within perature of the air entering the carburetor for warmed-up engine operation is a few engine revolutions have stabilized at the periodic values associated with the controlled to within much closer tolerances by drawing an appropriate fraction new throttle angle. At wide-open throttle, the pulsating nature of the flow as each of the air from around the exhaust manifold. Equation (7.6) shows how the air/fuel ratio delivered by the main metering system will vary with inlet air conditions. The primary dependence is through the 22º Pao term; depending on what is held constant (e.g ., throttle setting or air mass 50 0 flow rate) there may be an additional, much smaller dependence through ® and WOTH Throttle angle Apa (see Ref. 5): To a good approximation, the enrichment E with increasing altitude z is given by OF cm 1H2O Pas (7.11) 20 - 1 + E = (F/A)2 1/2 (F/A)O 40- Boost venturi suction 0 FIGURE 7-8 For z = 1500 m, E = 9.5 percent; thus, a cruise equivalence ratio of 0.9 of 20 - Intake manifold vacuum Throttle angle, boost venturi suction, and intake (A/F) = 16.2 would be enriched to close to stoichiometric. 40 manifold vacuum variation with time as throttle The effects of increase in altitude on the carburetor flow curve shown in is opened from light load (22º) to wide-open Fig. 7-1 are: (1) to enrich the entire part-throttle portion of the curve and (2) 10 throttle at 1000 rev/min. Standard two-barrel Time carburetor and production V-8 engine.2 294 INTERNAL COMBUSTION ENGINE FUNDAMENTALS SI ENGINE FUEL METERING AND MANIFOLD PHENOMENA 295 cylinder draws in its charge is evident. The pressure drop across the main meter. ing jet also fluctuates. The pulsations in the venturi air flow (and hence fuel flow) Fuel- pressure due to the filling of each cylinder in turn are negligible small at small throttle regulator angles and increase to a maximum at wide-open throttle. At small throttle open- Injection valve ings, the choked flow at the throttle plate prevents the manifold pressure fluctua- Start valve tions from propagating upstream into the venturi. The effective time-averaged Temperature sensor boost venturi suction is greater for the pulsating flow case than for the steady Electronic flow case. If the ratio of the effective metering signal for a pulse cycle to that for Pressure control steady air flow at the same average mass flow is denoted as 1 + 2, where _2 is the sensor unit Throttle pulsation factor, then _ is related to the amplitude and frequency of pressure I valve waves within the intake manifold as well as the damping effect of the throttle Auxiliary air device- switch Ignition plate. An empirical equation for 12 is distributor with trigger contacts constant x (1 - M)Pm nR Temperature Fuel filter (7.12) sensor Electric Nnc/b Thermo-time switch fuel where M is the throttle plate Mach number, p. the manifold pressure, np the pump number of revolutions per power stroke, N the crank speed, and not the number FIGURE 7-9 of cylinders per barrel. The value of the constant depends on carburetor and Speed-density electronic multipoint port fuel-injection system: Bosch D-Jetronic System.9 (Courtesy Robert Bosch GmbH and SAE.) engine geometry. For pm in kilonewtons per square meter and N in revolutions per minute a typical value for the constant is 7.3.2 Thus, at wide-open throttle at 1500 rev/min, Q has a value of about 0.2. The transient behavior of the air and 270 kN/m2, 39 lb/in2, usually relative to manifold pressure to maintain a con- fuel flows in the manifold are discussed more fully in Sec. 7.6. stant fuel pressure drop across the injectors). Branch lines lead to each injector; the excess fuel returns to the tank via a second line. The inducted air flows 7.3 FUEL-INJECTION SYSTEMS through the air filter, past the throttle plate to the intake manifold. Separate runners and branches lead to each inlet port and engine cylinder. An electromag- 7.3.1 Multipoint Port Injection netically actuated fuel-injection valve (see Fig. 7-10) is located either in the intake manifold tube or the intake port of each cylinder. The major components of the The fuel-injection systems for conventional spark-ignition engines inject the fuel injector are the valve housing, the injector spindle, the magnetic plunger to which into the engine intake system. This section reviews systems where the fuel is the spindle is connected, the helical spring, and the solenoid coil. When the sole- injected into the intake port of each engine cylinder. Thus these systems require noid is not excited, the solenoid plunger of the magnetic circuit is forced, with its one injector per cylinder (plus, in some systems, one or more injectors to supple- seal, against the valve seat by the helical spring and closes the fuel passage. When ment the fuel flow during starting and warm-up). There are both mechanical the solenoid coil is excited, the plunger is attracted and lifts the spindle about injection systems and electronically controlled injection systems. The advantages of port fuel injection are increased power and torque through improved volu- metric efficiency and more uniform fuel distribution, more rapid engine response Pintle Winding to changes in throttle position, and more precise control of the equivalence ratio during cold-start and engine warm-up. Fuel injection allows the amount of fuel injected per cycle, for each cylinder, to be varied in response to inputs derived from sensors which define actual engine operating conditions. Two basic approaches have been developed; the major difference between the two is the method used to determine the air flow rate. Figure 7-9 shows a schematic of a speed-density system, where engine speed and manifold pressure and air temperature are used to calculate the engine air Valve needle Return spring flow. The electrically driven fuel pump delivers the fuel through a filter to the fuck FIGURE 7-10 line. A pressure regulator maintains the pressure in the line at a fixed value (c.g- Cross section of fuel injector.10 296 INTERNAL COMBUSTION ENGINE FUNDAMENTALS SI ENGINE FUEL METERING AND MANIFOLD PHENOMENA 297 Fuel-pressure regulator 0.15 mm so that fuel can flow through the calibrated annular passage around the Relay valve stem. The front end of the injector spindle is shaped as an atomizing pintle set with a ground top to atomize the injected fuel. The relatively narrow spray cone Injection valve of the injector, shown in the photo in Fig. 7-11, minimizes the intake manifold Start valve Air flow sensor wall wetting with liquid fuel. The mass of fuel injected per injection is controlled by varying the duration of the current pulse that excites the solenoid coil. Typical injection times for automobile applications range from about 1.5 to 10 ms.11 Electronic control The appropriate coil excitation pulse duration or width is set by the elec- unit tronic control unit (ECU). In the speed-density system, the primary inputs to the ECU are the outputs from the manifold pressure sensor, the engine speed sensor Auxiliary air device (usually integral with the distributor), and the temperature sensors installed in the Throttle-valve switch intake manifold to monitor air temperature and engine block to monitor the water-jacket temperature-the latter being used to indicate fuel-enrichment Fuel filter Electric fuel pump Temperature requirements during cold-start and warm-up. For warm-engine operation, the sensor Thermo-time switch mass of air per cylinder per cycle ma is given by FIGURE 7-12 ma = no(N)pa(Ti, Pi)Va - No VaPi Ra Ti (7.13) Electronic multipoint port fuel-injection system with air-flow meter: Bosch L-Jetronic system.9 (Courtesy Robert Bosch GmbH and SAE.) where ny is the volumetric efficiency, N is engine speed, pa is the inlet air density, and Va is the displaced volume per cylinder. The electronic control unit forms the Figure 7-12 shows an alternative EFI system (the Bosch L-Jetronic) which pulse which excites the injector solenoids. The pulse width depends primarily on uses an air-flow meter to measure air flow directly. The air-flow meter is placed the manifold pressure; it also depends on the variation in volumetric efficiency n. upstream of the throttle. The meter shown measures the force exerted on a plate with speed N and variations in air density due to variations in air temperature. as it is displaced by the flowing air stream; it provides a voltage proportional to The control unit also initiates mixture enrichment during cold-engine operation the air flow rate. An alternative air-flow measuring approach is to use a hot-wire and during accelerations that are detected by the throttle sensor. air mass flow meter.10 The advantages of direct air-flow measurement are:12 (1) automatic compensation for tolerances, combustion chamber deposit buildup, wear and changes in valve adjustments; (2) the dependence of volumetric effi- ciency on speed and exhaust backpressure is automatically accounted for; (3) less acceleration enrichment is required because the air-flow signal precedes the filling of the cylinders; (4) improved idling stability; and (5) lack of sensitivity of the system to EGR since only the fresh air flow is measured. The mass of air inducted per cycle to each cylinder, ma, varies as ma oc (7.14) Thus the primary signals for the electronic control unit are air flow and engine speed. The pulse width is inversely proportional to speed and directly pro- portional to air flow. The engine block temperature sensor, starter switch, and throttle valve switch provide input signals for the necessary adjustments for cold- start, warm-up, idling, and wide-open throttle enrichment. For cold-start enrichment, one (or more) cold-start injector valve is used to FIGURE 7-11 inject additional fuel into the intake manifold (see Figs. 7-9 and 7-12). Since short Short time-exposure photograph of liquid fuel spray from Bosch-type injector. (Courtesy Roben opening and closing times are not important, this valve can be designed to Bosch GmbH.) 298 INTERNAL COMBUSTION ENGINE FUNDAMENTALS SI ENGINE FUEL METERING AND MANIFOLD PHENOMENA 299 provide extremely fine atomization of the fuel to minimize the enrichment - required. Mechanical, air-flow-based metering, continuous injection systems are also W used. Figure 7-13 shows a schematic of the Bosch K-Jetronic system.9. 10 Air is Cylinder drawn through the air filter, flows past the air-flow sensor, past the throttle valve, number into the intake manifold, and into the individual cylinders. The fuel is sucked out 2 of the tank by a roller-cell pump and fed through the fuel accumulator and filter 0 to the fuel distributor. A primary pressure regulator in the fuel distributor main- 120 240 360 480 600 deg tains the fuel pressure constant. Excess fuel not required by the engine flows back Injection group 2 Injection group 1 to the tank. The mixture-control unit consists of the air-flow sensor and fuel distributor. It is the most important part of the system, and provides the desired D. Injection duration metering of fuel to the individual cylinders by controlling the cross section of the 2. Iniet valve 4 Ignition metering slits in the fuel distributor. Downstream of each of these metering slits is FIGURE 7-14 a differential pressure valve which for different flow rates maintains the pressure Injection pulse diagram for D-Jetronic system in six-cylinder engine.10 drop at the slits constant. Fuel-injection systems offer several options regarding the timing and loca- tion of each injection relative to the intake event.1º The K-Jetronic mechanical from about 10º at light load and low speed to about 3000 at maximum speed and injection system injects fuel continuously in front of the intake valves with the load. Thus the pulse width varies from being much less than to greater than the spray directed toward the valves. Thus about three-quarters of the fuel required duration of the intake stroke. To reduce the complexity of the electronic control for any engine cycle is stored temporarily in front of the intake valve, and one- unit, groups of injectors are often operated simultaneously. In the Bosch L- quarter enters the cylinder directly during the intake process. Jetronic system, all injectors are operated simultaneously. To achieve adequate With electronically controlled injection systems, the fuel is injected inter- mixture uniformity, given the short pulse width relative to the intake process over mittently toward the intake valves. The fuel-injection pulse width to provide the much of the engine load-speed range, fuel is injected twice per cycle; each injec- appropriate mass of fuel for each cylinder air charge varies from about 1.5 to tion contributes half the fuel quantity required by each cylinder per cycle. (This 10 ms over the engine load and speed range. In crank angle degrees this varies approach is called simultaneous double-firing.) In the speed-density system, the injectors are usually divided into groups, each group being pulsed simulta- neously. For example, for a six-cylinder engine, two groups of three injectors may Injection be used. Injection for each group is timed to occur while the inlet valves are Idle-speed valve Start valve Fuel distributor closed or just starting to open, as shown in Fig. 7-14. The other group of injec- adjusting screw Throttle Primary tors inject one crank revolution later. Sequential injection timing, where the valve pressure regulator phasing of each injection pulse relative to its intake valve lift profile is the same, Air flow is another option. Engine performance and emissions do change as the timing of sensor the start of injection relative to inlet valve opening is varied. Injection with valve Intake manifold lift at its maximum, or decreasing, is least desirable.10 Thermo- Auxiliary time air device Mixture-control switch junit 7.3.2 Single-Point Throttle-Body Injection it Warm-up regulator Single-point fuel-injection systems, where one or two electronically controlled Fuel filter injectors meter the fuel into the air flow directly above the throttle body, are also Fuel used. They provide straightforward electronic control of fuel metering at lower accumulator cost than multipoint port injection systems. However, as with carburetor systems, Electric fuel pump the problems associated with slower transport of fuel than the air from upstream of the throttle plate to the cylinder must now be overcome (see Sec. 7.6). Figure FIGURE 7-13 Mechanical multipoint port fuel-injection system: Bosch K-Jetronic system.9 (Courtesy Robert Bosch 7-15 shows a cutaway of one such system.13 Two injectors, each in a separate air-flow passage with its own throttle plate, meter the fuel in response to cali- GmbH and SAE.) brations of air flow based on intake manifold pressure, air temperature, and SI ENGINE FUEL METERING AND MANIFOLD PHENOMENA 301 300 INTERNAL COMBUSTION ENGINE FUNDAMENTALS by the pressure drop across the throttle shears and atomizes the liquid sheet. Fuel-pressure Vigorous mixing of fuel and air then occurs, especially at part throttle, and pro- regulator vides good mixture uniformity and distribution between cylinders. Injector fuel delivery scheduling is illustrated in Fig. 7-16 for an eight-cylinder engine for a throttle-body fuel-injection system.14 1.4 FEEDBACK SYSTEMS Air flow It is possible to reduce engine emissions of the three pollutants-unburned hydrocarbons, carbon monoxide, and oxides of nitrogen-with a single catalyst in the exhaust system if the engine is operated very close to the stoichiometric air/fuel ratio. Such systems (called three-way catalyst systems) are described in more detail in Sec. 11.6.2. The engine operating air/fuel ratio is maintained close to stoichiometric through the use of a sensor in the exhaust system, which pro- Electromechanical Fuel supply fuel injector- from pump vides a voltage signal dependent on the oxygen concentration in the exhaust gas stream. This signal is the input to a feedback system which controls the fuel feed Fuel return to tank FIGURE 7-15 to the intake. Cutaway drawing of a two- The sensor [called an oxygen sensor or lambda sensor-A being the symbol injector throttle-body electronic used for the relative air/fuel ratio, Eq. (3.9)] is an oxygen concentration cell with fuel-injection system.13 a solid electrolyte through which the current is carried by oxygen ions. The elec- trolyte is yttria (Y2O3)-stabilized zirconia (ZrO2) ceramic which separates two gas chambers (the exhaust manifold and the atmosphere) which have different engine speed using the speed-density EFI logic described in the previous section. oxygen partial pressures. The cell can be represented as a series of interfaces as Injectors are fired alternatively or simultaneously, depending on load and speed follows: and control logic used. Under alternative firing, each injection pulse corresponds to one cylinder filling. Smoothing of the fuel-injection pulses over time is Exhaust Metal Ceramic Metal Air achieved by proper placement of the fuel injector assembly above the throttle Por Me ZrO2, Y2O3 Me bore and plate. The walls and plate accumulate liquid fuel which flows in a sheet Por toward the annular throttle opening (see Sec. 7.6.3). The high air velocity created o2 is the oxygen partial pressure of the air (~20 kN/m2) and po, is the equi- librium oxygen partial pressure in the exhaust gases. An electrochemical reaction takes place at the metal electrodes: TBI injectors (four cylinders each) 02 + 4M2 = 202- Injector A -200411 and the oxygen ions transport the current across the cell. The open-circuit output Injector B voltage of the cell V, can be related to the oxygen partial pressures po, and po2 through the Nernst equation: Firing order -180 Crank angle, deg V. = = RT In (Box (7.15) Idle 600 rev/min Por -50 m FIGURE 7-16 1-2 ms Injector fuel delivery schedule for two- where F is the Faraday constant. Equilibrium is established in the exhaust gases Injector A - injector throttle-body injection system by the catalytic activity of the platinum metal electrodes. The oxygen partial Injector B for eight-cylinder engine with dual plane pressure in equilibrated exhaust gases decreases by many orders of magnitude as intake manifold. Each injection nozzle WOT 4400 rev/min the equivalence ratio changes from 0.99 to 1.01, as shown in Fig. 7-17a. Thus the 6.67 ms-| feeds one plane of the manifold and its Injector A à zstumm Qui four cylinders.14 sensor output voltage increases rapidly in this transition from a lean to a rich 302 INTERNAL COMBUSTION ENGINE FUNDAMENTALS SI ENGINE FUEL METERING AND MANIFOLD PHENOMENA 303 (a) (b) Positive electrical - terminal 400ºC Insulator 103 1000 500ºC Shell (negative Vent 102 800 900ºC electrical terminal) 10 750ºC - 600 Graphite seal Sensor output voltage, mV Oxygen partial pressure, Pa and contact 10-5 400 Sensor Body 900ºC 10-10 Housing TTTTTTTTTT 750ºC 200 Flute 10-15 10-20 500ºC Shield -L 0.7 0.8. 0.9 1.0 1.1 1.2 .1.3 0.7 0.8 0.9 1.0 1.1 1.2 1.3 Rich Lean Rich Lean Relative air/fuel ratio À Exhaust gases FIGURE 7-17 FIGURE 7-18 Oxygen-sensor characteristics. Variation as a function of relative air/fuel ratio and temperature of: (a) Exhaust manifold oxygen partial pressure in equilibrated combustion products; (b) sensor output voltage.15 Cross-section drawing of exhaust oxygen sensor.16 mixture at the stoichiometric point, as shown in Fig. 7-17b. Since this transition is not temperature dependent, it is well suited as a sensor signal for a feedback output varies the fuel quantity linearly in the opposite direction to the sign of the system. 15 comparator signal. There is a time lag t in the loop composed of the transport . Figure 7-18 shows a cross-section drawing of a lambda sensor, screwed into time of fuel-air mixture from the point of fuel admission in the intake system to the wall of the exhaust manifold. This location provides rapid warm-up of the the sensor location in the exhaust, and the sensor and control system time delay. sensor following engine start-up. It also gives the shortest flow time from the fuel Because of this time lag, the controller continues to influence the fuel flow rate in injector or carburetor location to the sensor-a delay time which is important in the same direction, although the reference point ¢ = 1.0 has been passed, as the operation of the feedback system. The sensor body is made of ZrO2 ceramic shown in Fig. 7-19b. Thus, oscillations in the equivalence ratio delivered to the stabilized with Y2O3 to give adequate low-temperature electrical conductivity. engine exist even under steady-state conditions of closed-loop control. This The inner and outer electrodes are 10-um thick porous platinum layers to behavior of the control system is called the limit cycle. The frequency f of this provide the required catalytic equilibration. The outer electrode which is exposed limit cycle is given by to the exhaust gases is protected against corrosion and erosion by a 100-um spinal coat and a slotted shield. Air passes to the inner electrode through holes in 1 fic = 47, (7.16) the protective sleeve. The shield, protective sleeve, and housing are made from heat- and corrosion-resistant steel alloys. Such sensors were first developed for air/fuel ratio control at close to the stoichiometric value. Use of a similar sensor to control air/fuel ratios at lean-of-stoichiometric values during part-throttle engine operation is also feasible. 1 ) - Rich -Reference level For closed-loop feedback control at close-to-stoichiometric, use is made of Lean the sensor's low-voltage output for lean mixtures and a high-voltage output for FIGURE 7-19 rich mixtures. A control voltage reference level is chosen at about the mid-point Rich Operation of electronic control unit for of the steep transition in Fig. 7-17b. In the electronic control unit the sensor 1b ) - Stoichiometric closed-loop feedback: (a) sensor signal com- signal is compared to the reference voltage in the comparator as shown in F& Lean pared with reference level; (b) controller 7-19a. The comparator output is then integrated in the integral controller whose output voltage-the integrated comparator Time - output. 12 304 INTERNAL COMBUSTION ENGINE FUNDAMENTALS SI ENGINE FUEL METERING AND MANIF. and the change in equivalence ratio peak-to-peak is 4Q = 2KIL (7.17) where K is the integrator gain (in equivalence ratio units per unit time). Depending on the details of the three-way catalyst used for cleanup of all three pollutants (CO, HC, and NO,) in the exhaust, the optimum average equiva- lence ratio may not be precisely the stoichiometric value. Furthermore, the refer- ence voltage for maximum sensor durability may not correspond exactly to the stoichiometric point or the desired catalyst mean operating point. While a small shift (~ +1 percent) in operating point from the stoichiometric can be obtained by varying the reference voltage level, larger shifts are obtained by modifying the control loop to provide a steady-state bias. One method of providing a bias- asymmetrical gain rate biasing17-uses two separate integrator circuits with dif- ferent gain rates K+ and K~ to integrate the comparator output, depending on whether the comparator output is positive (rich) or negative (lean). An alternative biasing technique incorporates an additional delay time to so that the controller (a) 20º throttle plate angle output continues decreasing (or increasing) even though the sensor signal has (b) 30º throttle plate angle switched from the high to the low level (or vice versa). By introducing this addi- tional delay only on the negative slope of the sensor signal, a net lean bias is produced. Introducing the additional delay on the positive slope of the sensor signal produces a net rich bias. 12 Note that the sensor only operates at elevated temperatures. During engine start-up and warm-up, the feedback system does not operate and conventional controls are required to obtain the appropriate fuel-air mixture for satisfactory engine operation. 7.5 FLOW PAST THROTTLE PLATE Except at or close to wide-open throttle, the throttle provides the minimum flow area in the entire intake system. Under typical road-load conditions, more than 90 percent of the total pressure loss occurs across the throttle plate. The minimum-to-maximum flow area ratio is large-typically of order 100. Throttle (c) 45º throttle plate angle (d) 60º throttle plate angle FIGURE 7-21 DSV cos Vo Photographs of flow in two-dimensional hydraulic analog of carburetor venturi, throttle plate, and manifold plenum floor at different throttle plate angles. 18 -D plate geometry and parameters are illustrated in Fig. 7-20. A throttle plate of Closed Open to angle y conventional design such as Fig. 7-20 creates a three-dimensional flow field. At part-throttle operating conditions the throttle plate angle is in the 20 to 45º range and jets issue from the "crescent moon"-shaped open areas on either side of the FIGURE 7-20 throttle plate. The jets are primarily two dimensional. Figure 7-21 shows pho- Throttle plate geometry.2 tographs taken of a two-dimensional hydraulic analog of a typical carburetor 306 INTERNAL COMBUSTION ENGINE FUNDAMENTALS SI ENGINE FUEL METERING AND MANIFOLD PHENOMENA 307 venturi and throttle plate in steady flow at different throttle angles. The path lines traced by the particles in the flow indicate the relative magnitude of the flow (C-9)]. For pressure ratios across the throttle less than the critical value (PT/Po = 0.528), the mass flow rate is given by velocity.18 The flow accelerates through the carburetor venturi (separation occurs at the corners of the entrance section); it then divides on either side of the throttle mch = - CD 4in Po ( PT) { 27 [ plate. There is a stagnation point on the upstream side of the throttle plate. The VRTO (DO ) (7.19) wake of the throttle plate contains two vortices which rotate in opposite direc- tions. The jets on either side of the wake (at part throttle) are at or near sonic where Ath is the throttle plate open area [Eq. (7.18)], po and To are the upstream velocity. There is little or no mixing between the two jets. Thus, if maldistribution pressure and temperature, pr is the pressure downstream of the throttle plate of the fuel-air mixture occurs above the throttle plate, it is not corrected imme- (assumed equal to the pressure at the minimum area: i.e ., no pressure recovery diately below the throttle plate. occurs), and Cp is the discharge coefficient (determined experimentally). For pres- In analyzing the flow through the throttle plate, the following factors sure ratios greater than the critical ratio, when the flow at the throttle plate is choked, should be considered :2, 19, 20 mith = CD Ath Po ,1/2( 2 ) (2+1)/2(7-1) 1. The throttle plate shaft is usually of sufficient size to affect the throttle open VRTO 7 + 1 (7.20) area. 2. To prevent binding in the throttle bore, the throttle plate is usually completely The relation between air flow rate, throttle angle, intake manifold pressure, closed at some nonzero angle (5, 10, or 15º). and engine speed for a two-barrel carburetor and a 4.7-dm3 (288-in3) displace- 3. The discharge coefficient of the throttle plate is less than that of a smooth ment eight-cylinder production engine is shown in Fig. 7-22. While the lines are converging-diverging nozzle, and varies with throttle angle, pressure ratio, and predictions from a quasi-steady computer simulation, the agreement with data is throttle plate Reynolds number. excellent. The figure shows that for an intake manifold pressure below the critical 4. Due to the manufacturing tolerances involved, there is usually some minimum leakage area even when the throttle plate is closed against the throttle bore. This leakage area can be significant at small throttle openings. 120 WOT 5. The measured pressure drop across the throttle depends (+10 percent) on the circumferential location of the downstream pressure tap. 6. The pressure loss across the throttle plate under the actual flow conditions 100 (which are unsteady even when the engine speed and load are constant, see Throttle angle y Fig. 7-8) may be less than under steady flow conditions. The throttle plate open area Ath, as a function of angle y for the geometry 36º in Fig. 7-20, is given by2 Air flow rate, g/s 4Ath 60 4000 rev/min (1- Cos y 3500 1+4 a 3000 1 D 2 cos 4. TI COS y ( cos 2 4 - a 2 cos 2 4 0 ) 1 1 2 2600 cos $ - sin-1 ( 7 .18 ) 2000 (a costo ) - a(1 - a2)112 + sin -" a| 40 26º 1600 cos vo Cos 1200 where a = d/D, d is the throttle shaft diameter, D is the throttle bore diameter. 210 800 and wo is the throttle plate angle when tightly closed against the throttle bore. 20 180. When y = cos"1 (a cos vo), the throttle open area reaches its maximum value (~ RD2/4 - dD). The throttle plate discharge coefficient (which varies with An FIGURE 7-22 10º Variation in air flow rate past a and minimum leakage area, must be determined experimentally. throttle, with inlet manifold pres- SH The mass flow rate through the throttle valve can be calculated from stan- 0 10 30 40 50 60 70 sure, throttle angle, and engine dard orifice equations for compressible fluid flow [see App. C, Eqs. (C-8) and Intake manifold pressure, cmHg speed. 4.7-dm3 displacement eight-cylinder engine.2 308 INTERNAL COMBUSTION ENGINE FUNDAMENTALS SI ENGINE FUEL METERING AND MANIFOLD PHENOMENA 309 value (0.528 x Patm = 53.5 kN/m2 == 40.1 cmHg) the air flow rate at a given throttle position is independent of manifold pressure and engine speed because the flow at the throttle plate is choked.2 Air-fuel mixture passage 7.6 FLOW IN INTAKE MANIFOLDS 7.6.1 Design Requirements The details of the air and fuel flow in intake manifolds are extremely complex. EGR gas passage" Section A-A Coolant The combination of pulsating flow into each cylinder, different geometry flow heat passage paths from the plenum beneath the throttle through each runner and branch of the manifold to each inlet port, liquid fuel atomization, vaporization and trans- port phenomena, and the mixing of EGR with the fresh mixture under steady- state engine operating conditions are difficult enough areas to untangle. During engine transients, when the throttle position is changed, the fact that the pro- cesses which govern the air and the fuel flow to the cylinder are substantially EGR valve different introduces additional problems. This section reviews our current under- standing of these phenomena. FIGURE 7-23 Intake manifolds consist typically of a plenum, to the inlet of which bolts Distributor Inlet manifold for four-cylinder 1.8-dm3 displace- Throttle linkage the throttle body, with individual runners feeding branches which lead to each ment spark-ignition engine.21 cylinder (or the plenum can feed the branches directly). Important design criteria are: low air flow resistance; good distribution of air and fuel between cylinders; be considered as unaffected by the fuel flow. The reverse is definitely not the case: runner and branch lengths that take advantage of ram and tuning effects; suffi- the fuel flow-liquid and vapor-depends strongly on the air flow. These two cient (but not excessive) heating to ensure adequate fuel vaporization with carbu- topics will therefore be reviewed in sequence. reted or throttle-body injected engines. Some compromises are necessary; e.g ., runner and branch sizes must be large enough to permit adequate flow without 7.6.2 . Air-Flow Phenomena allowing the air velocity to become too low to transport the fuel droplets. Some of these design choices are illustrated in Fig. 7-23 which shows an inlet manifold The air flow out of the manifold occurs in a series of pulses, one going to each and carburetor arrangement for a modern four-cylinder 1.8-dm3 engine. In this cylinder. Each pulse is approximately sinusoidal in shape. For four- and eight- design the four branches that link the plenum beneath the carburetor and throt- cylinder engines, these flow pulses sequence such that the outflow is essentially tle with the inlet ports are similar in length and geometry, to provide closely zero between pulses. For six-cylinder arrangements the pulses will overlap. When comparable flow paths. This manifold is heated by engine coolant as shown and the engine is throttled, backflow from the cylinder into the intake manifold uses an electrically heated grid beneath the carburetor to promote rapid fuel occurs during the early part of the intake process until the cylinder pressure falls evaporation.21 Exhaust gas heated stoves at the floor of the plenum are also used below the manifold pressure. Backflow can also occur early in the compression in some intake manifolds to achieve adequate fuel vaporization and distribution. stroke before the inlet valve closes, due to rising cylinder pressure. The flow at Note that with EGR, the intake manifold may contain passages to bring the the throttle will fluctuate as a consequence of the pulsed flow out of the manifold exhaust gas to the plenum or throttle body. into the cylinders. At high intake vacuum, the flow will be continuously inward at With port fuel-injection systems, the task of the inlet manifold is to control the throttle and flow pulsations will be small. When the outflow to the cylinder the air (and EGR) flow. Fuel does not have to be transported from the throttle which is undergoing its intake stroke is greater than the flow through the throt- body through the entire manifold. Larger and longer runners and branches, with tle, the cylinder will draw mixture from the rest of the intake manifold. During larger angle bends, can be used to provide equal runner lengths and take greater the portion of the intake stroke when the flow into the cylinder is lower than the advantage of ram and tuning effects. With port fuel injection it is not normally flow through the throttle, mixture will flow back into the rest of the manifold. At necessary to heat the manifold. wide-open throttle when the flow restriction at the throttle is a minimum, flow A large number of different manifold arrangements are used in practice. pulsations at the throttle location will be much more pronounced.19 Different cylinder arrangements (e.g ., four, V-six, in-line-six, etc.) provide quite The air flows to each cylinder of a multicylinder engine, even under steady different air and fuel distribution problems. Air-flow phenomena in manifolds can operating conditions, are not identical. This is due to differences in runner and 310 INTERNAL COMBUSTION ENGINE FUNDAMENTALS SI ENGINE FUEL METERING AND MANIFOLD PHENOMENA 311 branch length and other geometric details of the flow path to each cylinder. Also, cmHg deg as each cylinder's intake flow commences, air is drawn from the branch and pour Air flow runner leading to the cylinder, the plenum, and the other runners and branches -25 feeding the plenum, as well as past the throttle. This "drawing down" of other Manifold pressure parts of the intake manifold depends on the arrangement of the plenum, runners, 20 40 and branches, and the firing order of the cylinders. Thus the air flow to each Throttle angle -15 individual cylinder is affected by the details of its own branch, how that branch connects to the rest of the intake manifold, and the cylinder firing order.22 The 20 110 differences between the air flows to individual cylinders have been measured. Variations in the average air mass flow rate to each individual cylinder of up to 0.05 0.1 0.15 about 5 percent above and below the average are quite common. Larger peak-to- 0.2 peak variations (+ 15 percent) have been measured. The extent of each cylinder's Time, s difference from the average flow varied significantly as engine speed and load FIGURE 7-24 were varied. 23, 24 Throttle angle, intake manifold pressure, and air flow rate past the throttle versus time for 10º part- load throttle opening. 5-dm3 V-8 engine.25 Typical quantities that characterize manifold air flow are given in Table 7.1 for a four- and an eight-cylinder spark-ignition engine. The volume of mixture pulled into each cylinder per cycle is about the same as the volume of one direct flow path between the throttle plate and inlet valve. Thus, one stroke loads the the manifold, the pressure level in the manifold increases more slowly than would manifold, the next one pulls the charge into the cylinder. be the case if steady-state conditions prevailed at each throttle position. Thus, the An additional phenomenon becomes important when engine load is pressure difference across the throttle is larger than it would be under steady flow changed by opening or closing the throttle: the mass of air in the induction conditions and the throttle air flow overshoots its steady-state value. The air flow system volume takes, a finite time to adjust to the new engine operating condi- into each cylinder depends on the pressure in the manifold, so this lags the throt- tions. For example, as the throttle is opened the air flow into the manifold tle air flow. This transient air-flow phenomenon affects fuel metering. For increases as the throttle open area increases. However, due to the finite volume of throttle-body injection or a carburetor, fuel flow should be related to throttle air flow. For port fuel injection, fuel flow should be related to cylinder air flow. Actual results for the air flow rate and manifold pressure in response to an opening of the throttle (increase in throttle angle) are shown in Fig. 7-24. The TABLE 7.1 Parameters that characterize manifold air flow overshoot in throttle air flow and lag in manifold pressure as the throttle angle is increased are evident. Opposite effects will occur for a decrease in throttle angle. Engine geometry 1-4+ V-8+ AIR-FLOW MODELS. Several models of the flow in an intake manifold have been Typical flow-path distance between proposed.26, 27 One simple manifold model that describes many of the above throttle bore and intake valve, cm 33 30 Average intake-passage flow area, cm2 9.4 16 phenomena is the plenum or filling and emptying model. It is based on the Volume of one direct flow path from assumption that at any given time the manifold pressure is uniform. The contin- throttle bore to intake valve, cm3 300 500 uity equation for air flow into and out of the intake manifold is Range of speeds, etc. Maximum Minimum dt dma.m = ma. th - [ ma, cy1 (7.21) Crankshaft, rev/min 5000 650 Peak air velocity in where ma " is the mass of air in the manifold, and ma, th and ma, cyl are the air manifold branch, m/s 1301, 100+ 15 Peak Reynolds number mass flow rates past the throttle and into each cylinder, respectively. The flow in manifold branch 4 x 105 5 x 104 rate past the throttle is given by Eq. (7.19) or (7.20). For manifold pressures Duration of individual sufficiently low to choke the flow past the throttle plate, the flow rate is indepen- cylinder intake process, ms 6 46 dent of manifold pressure. The mass flow rate to the engine cylinders can be + 1.8-dm3 four-cylinder in-line SI engine.21 modeled at several levels of accuracy. The air flow through the valve to each $ 5.6-dm3 V eight-cylinder SI engine.25 cylinder can be computed from the valve area, discharge coefficient, and pressure 312 INTERNAL COMBUSTION ENGINE FUNDAMENTALS SI ENGINE FUEL METERING AND MANIFOLD PHENOMENA 313 difference across the valve; or a sine wave function can be assumed. In the general case, Eq. (7.21) must be combined with the first law for an open system (see Sec. 14.2.2). For calculating the manifold response to a change in load or throttle setting, simplifying assumptions can be made. A quasi-steady approx- O imation for the cylinder air flow: O A2 ma, cyl = No Pa. m VON 2 (a) is usually adequate, and the air temperature can be assumed constant.25 Then (b) using the ideal gas law for the manifold, Pm Vm = ma, m Ra Tm, Eq. (7.21) can be FIGURE 7-25 written as Helmholtz resonator models for (a) single-cylinder engine and (b) multicylinder engine.27 dpm no VaN RT, dt 2V Pm = Ma, th Vm (7.22) The Helmholtz theory for multicylinder engines treats the pipes of cylinders Both n. and ma, th have some dependence on pm [e.g ., see Eq. (6.2)]. In the absence not undergoing induction as an additional volume. The two pipes, (11, 41) and of this weak dependence, Eq. (7.22) would be a first-order equation for pm, with a (l;. A2), and two volumes, V1 and 12, in Fig. 7-25b form a vibrating system with time constant t = 2Vm/(n. V. N) ~ Vm/ Veyl, which is 2 to 4 times the intake stroke two degrees of freedom and two resonant frequencies. The following equation, duration. The smooth curves in Fig. 7-24 are predictions made with Eq. (7.22) based on an electrical analog (in which capacitors represent volumes and induc- and show good agreement with the data. The plenum model is useful for investi- tors pipes), gives the two frequencies at which the manifold shown in Fig. 7-25b would be tuned:28 gating manifold pressure variations that result from load changes. It provides no information concerning pressure variations associated with momentum effects. f + = 1 f(xB + a + 1) + [(aB + a + 1) 2 - 4aB]1/2 ) 1/2 Helmholtz resonator models for the intake system have been proposed. This type of model can predict the resonant frequencies of the combined intake and 2aBLIC1 (7.24) engine cylinder system, and hence the engine speeds at which increases in air flow due to intake tuning occur. It does not predict the magnitude of the increase in where a = L2/L1, B = C2/C1, C1 = V1, C2 = 12, 1 = (1/4)1, L2 = (1/A)2, and volumetric efficiency. The Helmholtz resonator theory analyzes what happens Vett = V1. The Helmholtz theory predicts the engine speeds at which positive tuning resonances occur with reasonable accuracy.27 during one inlet stroke, as the air in the manifold pipe is acted on by a forcing The dynamics of the flow in multicylinder intake (and exhaust) systems can function produced by the piston motion. As the piston moves downward during be modeled most completely using one-dimensional unsteady compressible flow the intake stroke, a reduced pressure occurs at the inlet valve relative to the equations. The standard method of solution of the governing equations has been pressure at the open end of the inlet pipe. A rarefraction wave travels down the the method of characteristics (see Benson29). Recently, finite difference techniques intake pipe to the open end and is reflected as a compression wave. A positive which are more efficient have been used.3º The assumptions usually made in this tuning effect occurs when the compression wave arrives at the inlet valve as the type of analysis are: valve is closing.27 A single-cylinder engine modeled as a Helmholtz resonator is shown in Fig. 7-25a. The effective resonator volume Verf is chosen to be one-half of the displaced volume plus the clearance volume; the piston velocity is then 1. The intake (or exhaust) system can be modeled as a combination of pipes, junctions, and plenums. close to its maximum and the pressure in the inlet system close to its minimum. The tuning peak occurs when the natural frequency of the cylinder volume 2. Flow in the pipes is one dimensional and no axial heat conduction occurs. coupled to the pipe is about twice the piston frequency. For a single-cylinder, fed 3. States in the engine cylinders and plenums are uniform in space. by a single pipe open to the atmosphere, the resonant tuning speed N, is given by 4. Boundary conditions are considered quasi steady. 5. Coefficients of discharge, heat transfer, pipe friction, and bend losses for steady N,(rev/min) == 955 A ) 1/2 K IVeFG ) (7.23) flow are valid for unsteady flow. 6. The gases can be modeled as ideal gases. where a is the sound speed (m/s), A the effective cross-sectional area of the inlet system (cm2), I the effective length of the inlet system (cm), K a constant equal to about 2 for most engines, and Verf = Va(re + 1)/[2(r. - 1)] (cm3).28 Sec. 14.3.4. This approach to intake and exhaust flow analysis is discussed more fully in 314 INTERNAL COMBUSTION ENGINE FUNDAMENTALS SI ENGINE FUEL METERING AND MANIFOLD PHENOMENA 315 7.6.3 Fuel-Flow Phenomena W FT TTTTT 70% 760 T 50% TRANSPORT PROCESSES. With conventional spark-ignition engine liquid-fwd 700- metering systems, the fuel enters the air stream as a liquid jet. The liquid 0- 100% 30%- atomizes into droplets. These mix with the air and also deposit on the walls of 90% the intake system components. The droplets vaporize; vaporization of the liqu? 80% fuel on the walls occurs. The flow of liquid fuel along the walls can be significant, 60% 10% 500 -- OL Manifold pressure, mmHg The transport of fuel as vapor, droplets, and liquid streams or films can all 40 % important. The fuel transport processes in the intake system are obvioush -0% extremely complex. 20- 20% 00 The details of the fuel transport process are different for multipoint. fuel injection systems than for carburetor and throttle-body injection systems. For the latter systems, fuel must be transported past the throttle plate and through the complete intake manifold. For the former systems, the liquid fuel is injected in the | 0.04 | 0.06 0.1 0.2 0.3 400 1.0 inlet port, toward the back of the intake valve. For all these practical fuel meter 0.03 0.05 1.2 1.4 1.6 1.8 2.0 ing systems, the quality of the mixture entering the engine is imperfect. The fuel Fuel/air ratio Fuel evaporated at reduced pressure Fuel evaporated at atmospheric pressure air, recycled exhaust, mixture is not homogeneous; the fuel may not be fully (a) vaporized as it enters the engine. The charge going to each cylinder is not usually (b) uniform in fuel/air ratio throughout its volume, and the distribution of fuel FIGURE 7-26 between the different engine cylinders is not exactly equal. During engine tran- i) Percentage of indolene fuel evaporated at equilibrium at 1 atmosphere pressure. (b) Effect of pressure on amount of indolene fuel evaporated.18 sients, when engine fuel and air requirements and manifold conditions change, it is obvious that the above fuel transport processes will not all vary with time in the same way. Thus, in addition to the transient non-quasi-steady air-flow phe- For carbureted and throttle-body injection systems, the fuel path is the fol- nomena described above, there are transient fuel-flow phenomena. These have to lowing. Until the throttle plate is close to fully opened, most of the fuel metered be compensated for in the fuel metering strategy. into the air stream impacts on the throttle plate and throttle-body walls. Only a Since gasoline, the standard spark-ignition engine fuel, is a mixture of a modest fraction of the fuel vaporizes upstream of the throttle. The liquid is re- large number of individual hydrocarbons it has a boiling temperature range entrained as the air flows at high velocity past the throttle plate. The fuel does rather than a single boiling point. Typically, this range is 30 to 200ºC. Individual not usually divide equally on either side of the throttle plate axis. The air under- hydrocarbons have the saturation pressure-temperature relationships of a pure goes a 90º bend in the plenum beneath the throttle; much of the fuel which has substance. The lower the molecular weight, the higher will be the saturated vapor aot evaporated is impacted on the manifold floor. Observations of fuel behavior pressure at a given temperature. The boiling point of hydrocarbons depends pri- in intake manifolds with viewing ports or transparent sections show that there is marily on their molecular weight: the vapor pressure also depends on molecular substantial liquid fuel on the walls with carburetor fuel metering systems. Figure structure. The equilibrium state of a hydrocarbon-air mixture depends therefore 7-27 shows the engine conditions under which liquid fuel was observed on the on the vapor pressure of the hydrocarbon at the given temperature, the relativ floor of the manifold plenum beneath the throttle plate and in the manifold amounts of the hydrocarbon and air, and the total pressure of the mixture. The runners, in a standard four-cylinder production engine.23 This manifold was equilibrium fraction of fuel evaporated at a given temperature and pressure can heated by engine coolant at 90ºC. The greatest amount of liquid was present at be calculated from Bridgeman charts31 and the distillation characteristics of the high engine loads and low speeds. Heating the manifold to a higher temperature fuel (defined by the ASTM distillation curve32). Figure 7-26a shows the effect of with steam at 115ºC resulted in a substantial reduction in the amount of liquid: mixture temperature on percent of indolene fuel (a specific gasoline) evaporated there was no extensive puddling on the plenum floor, liquid films or rivulets were at equilibrium at atmospheric pressure. Figure 7-26b shows the effect of reduced, observed in a zone bounded by 120 mmHg vacuum and 2500 rev/min, and there manifold pressure on the amount evaporated.18 While insufficient time is usually were no films or rivulets in the manifold runner. Depending on engine operating available in the manifold to establish equilibrium, the trends shown are indicatiw conditions, transport of fuel as a liquid film or rivulet in the manifold and vapor- of what happens in practice: lower pressures increase the relative amount of fue nation from these liquid fuel films and rivulets and subsequent transport as vaporized and charge heating is usually required to vaporize a substantial frac vapor may occur. tion of the fuel. Vaporized fuel and liquid droplets which remain suspended in the air 316 INTERNAL COMBUSTION ENGINE FUNDAMENTALS SI ENGINE FUEL METERING AND MANIFOLD PHENOMENA 317 400 100 of the liquid fuel droplets decreases rapidly (by up to about 30ºC35), and the Liquid No liquid films No liquid films or rivulets fraction of the fuel vaporized is small (in the 2 to 15 percent range35, 36). 300 films or or rivulets 300 Liquid fuel drops, due to their density being many times that of the air, will rivulets not exactly follow the air flow. Droplet impaction on the walls may occur as the Intake manifold vacuum, mmHg fow changes direction, and the greater inertia of the droplets causes them to 200 200- move across the streamlines to the outer wall. Deposition on the manifold floor 100 100 due to gravity may also occur. The equation of motion for an individual droplet Extensive Liquid films in a flowing gas stream is puddling or rivulets 0 1000 2000 3000 4000 5000 0 1000 2000 3000 4000 5000 ( &Dip , )a = mag - Hva - v)Iva - VIP, CD- (7.25) Engine speed, rev/min Engine speed, rev/min where Da is the droplet diameter, ps and p, are liquid and gas densities, va and v, (a) (b) are the droplet and gas velocities, a is the droplet acceleration, g acceleration due FIGURE 7-27 to gravity, and CD is the drag coefficient. For 6 < Re < 500 the drag coefficient of Regions of engine load and speed range where extensive pools or puddles, liquid films, or rivulets an evaporating droplet is a strong function of the Reynolds number, Re: e.g ., were observed: (a) on manifold plenum floor and (b) in manifold runner. Four-cylinder automobile CD = 27 Re ~ 0.84 engine. Manifold heated by coolant at 90ºC.23 (7.26) where Re = (p, Dalva - val/kg). Studies of droplet impaction and evaporation using the above equations stream will be transported with the air stream. However, droplet deposition on and typical manifold conditions and geometries indicate the following. 26, 35, 37 the manifold walls may occur due to gravitational settling and to inertial effects For 90º bends, drops of less than 10 um diameter are essentially carried by the as the flow goes round bends in the manifold. gas stream (<10 percent impaction); almost all droplets larger than 25 um The fuel transport processes for port fuel-injection systems are different and impact on the walls. Droplet sizes produced first in the carburetor venturi or fuel will depend significantly on the timing and duration of the injection pulse. Fuel is injector spray and then by secondary atomization as liquid fuel is entrained from injected onto the back of the inlet valve (and surrounding port wall), usually the throttle plate and throttle-body walls depend on the local gas velocity: higher while the valve is closed or only partly open. Vaporization of liquid fuel off the local relative velocities between the gas and liquid produce smaller drop sizes. valve and walls occurs, enhanced by the backflow of hot residual gases from the Approximate estimates which combine the two phenomena outlined above show cylinder (especially at part load). There is evidence that, even under fully warmed- that at low engine air flow rates, almost all of the fuel will impact first on the up engine conditions, some fuel is carried as liquid drops into the cylinder.33 throttle plate and then on the manifold floor as the flow turns 90º into the mani- fold runners. At high air flows, because the drops are smaller, a substantial frac- tion of the drops may stay entrained in the air flow. Secondary atomization at FUEL DROPLET BEHAVIOR. With carburetor and throttle-body injection the throttle at part-load operating conditions is important to the fuel transport systems, the liquid fuel atomizes as it enters the air stream. In the carburetor process: the very high air velocities at the edge of the throttle plate produce venturi this occurs as the fuel-air emulsion from the fuel jet(s) enters the high- velocity (> 100 m/s) air stream. With an injector, the velocity of the liquid jet as it droplets of order or less than 10 um diameter. However, coalescence and deposi- tion on the walls and subsequent reentrainment probably increase the mean exits the nozzle is high enough to shatter the flowing liquid, and its interaction droplet size. In the manifold, gravitational settling of large (> 100 um) droplets with the coaxial air flow further atomizes the fuel. Typical droplet-size distribu- would occur at low air flow rates, 38 but these drops are also likely to impact the tions are not well defined; size would vary over the load and speed range walls due to their inertia as the flow is turned. Droplet diameters in the 25 to 100 um range are usually assumed to be represen- Estimates of droplet evaporation rates in the manifold indicate the follow- tative: larger drops are also produced. The liquid fuel drops are accelerated by ing. With a representative residence time in the manifold of about one crank the surrounding air stream and start to vaporize. Vaporization rates have been calculated using established formulas for heat and mass transfer between a revolution (10 ms at 6000 rev/min, 100 ms at 600 rev/min), only drops of size less droplet and a surrounding flowing gas stream (see Ref. 34 for a review of methods than about 10 um will evaporate at the maximum speed; 100 um droplets will of calculating droplet vaporization rates). Calculations of fuel vaporization in not vaporize fully at any speed. Most of these large droplets impact on the walls, carburetor venturi and upstream of the throttle plate show that the temperature anyway. Drops small enough to be carried by the air stream are likely to vapor- ize in the manifold. 26 318 INTERNAL COMBUSTION ENGINE FUNDAMENTALS SI ENGINE FUEL METERING AND MANIFOLD PHENOMENA 319 FUEL-FILM BEHAVIOR. The fuel which impacts on the wall will also vaporize Manifold FIGURE 7-29 and, depending on where in the manifold deposition occurs and the local manis Schematic of fuel flow paths in the manifold when liquid film flows along the manifold runner floor. fold geometry, may be transported along the manifold as a liquid film or rivulet, F (d + v) F ( d + v) A Air If the vaporization rate off the wall is sufficiently high, then a liquid film will not F (d ) A F (v) A F Fuel build up. Any liquid film or pool on the manifold floor or walls is important f Liquid fuel film because it introduces additional fuel transport processes-deposition, liquid F (f) d Liquid fuel droplets transport, and evaporation -- which together have a much longer time constant Liquid film Fuel vapor than the air transport process. Thus changes in the air and the fuel flow into each engine cylinder, during a change in engine load, will not occur in phase with each increasing load and speed. The time constant is of order 2 seconds for a fully other unless compensation is made for the slower fuel transport. warmed-up engine; it varies with engine operating conditions and is especially Several models of the behavior of liquid-fuel wall-films have been devel- sensitive to intake manifold temperature. Such models have been used primarily oped. One approach analyzes a liquid puddle on the floor of the manifold to develop fuel metering strategies which compensate for the fuel transport lag.38 plenum.38 Metered fuel enters the puddle; fuel leaves primarily through vapor- An alternative model, for liquid film flow in the manifold runner and ization. The equation for rate of change of mass of fuel in the puddle is branch, has been developed.37 Fuel is deposited on the manifold walls and forms mf.2 a film which flows toward the cylinder due to the shear force at the gas/liquid ims. p = ims, in - ims, out = Xim ,, m -- (7.27) interface as shown in Fig. 7-29. Vaporization from the film also occurs. An I analysis of the dynamics of the fuel film leads to expressions for steady-state film where my, is the mass of fuel in the puddle, m, , is the metered fuel flow rate, velocity and thickness. As air and metered fuel flows change due to a throttle and x is the fraction of the metered flow that enters the puddle. It is assumed that position change, the characteristic time for reestablishing steady state is I/(2us), the reentrainment/evaporation rate is proportional to the mass of fuel in the where I is the manifold length and u the average film velocity. This characteristic puddle divided by the characteristic time t of the reentrainment/evaporation response time is of order 1 second for typical manifold conditions, in approx- process. The puddle behavior predicted by this model in response to a step imate agreement with values obtained from transient engine experiments. increase in engine load is shown in Fig. 7-28a. Because only part (1 - x) of the A more extensive analysis of both fuel droplet and film evaporation in a fuel flows directly with the air, as the throttle is opened rapidly a lean air/fuel complete carburetor, throttle, manifold system,35 with a multicomponent model ratio excursion is predicted. Figure 7-28b shows that this behavior (without any for gasoline based on its distillation curve, indicates the following phenomena are metering compensation) is observed in practice. Estimates of the volume of fuel in important. Secondary atomization of the liquid fuel at the throttle, which pro- the puddle (for a 5-liter V-8 engine) are of order 1000 mm3, and increase with duces the smallest droplet sizes when the throttle open angle is small, signifi- cantly increases the fraction of fuel evaporation in the manifold. Increasing inlet air temperature increases the fraction of fuel vaporized; this effect is larger at ms.m + A mj.m lower loads since secondary atomization under these conditions increases the mf . m 16.0 Metered fuel flow liquid fuel surface area significantly. Heating the wall, which heats the liquid film Tx(mf. m + 4 mf. m) 5.5- on the wall directly, provides a greater increase in fraction evaporated than does equivalent heating of the air flow upstream of the carburetor. Due to the multi- Txmf . m Film mass Air/fuel ratio 15.0 component nature of the fuel, the residual liquid fuel composition changes signifi- cantly as fuel is transported from the carburetor to the manifold exit. Of the full boiling range liquid composition at entry, all the light ends, most of the mid- Leaner Equivalence ratio 14.5 range components, but only a modest amount of the high boiling point fraction have evaporated at the manifold exit. The predicted fuel fraction evaporated Richer 14.0 -2 0 ranged from 40 to 60 percent for the conditions examined. One set of measure- Time - Time, s ments of the fraction of fuel vaporized in the manifold of a warmed-up four- (a) ( b) cylinder engine showed that 70 to 80 percent of the fuel had vaporized, confirming that under these operating conditions "most" but not necessarily FIGURE 7-28 (a) Predicted behavior of the fuel film for an uncompensated step change in engine operating condi- "all " the fuel enters the cylinder in vapor form.39 tions. (b) Observed variation in air/fuel ratio for uncompensated throttle opening at 1600 rev/mm The engine operating range where fuel puddling, fuel films, and rivulets are which increased manifold pressure from 48 to 61 cmHg.38 observed (see Fig. 7-27) can now be explained. At light load, secondary atom- 320 INTERNAL COMBUSTION ENGINE FUNDAMENTALS SI ENGINE FUEL METERING AND MANIFOLD PHENOMENA 321 ization at the throttle and the lower manifold pressure would reduce the amount mtake, the droplet number density in the clearance volume increased to a of liquid fuel impinging on the manifold plenum floor. Also, typical manifold maximum at the end of injection (the injection lasted from 45 to 153º ATC) and heating at light load substantially exceeds the heat required to vaporize the fuel then decreased due to evaporation during compression to a very small value at completely,40 and manifold floor temperatures are of order 15ºC higher than at the time of ignition. Average droplet size during intake was 10 to 20 um in diam- full load. All the above is consistent with less liquid on the floor and none in the cter; it increased during compression as the smaller drops in the distribution runners at light load, compared to what occurs at full load. At high speed, drop evaporated. At the conditions tested, some 10 to 20 percent of the fuel was in sizes produced in the carburetor are much smaller, so impingement on the walls droplet form at the end of injection. At ignition, the surviving droplets contained is much reduced. 1 negligible fraction of the fuel. During injection, the distribution of droplets The fuel flow to each cylinder per cycle is not exactly the same. There is a across the clearance volume was nonuniform. It became much more uniform with geometric variation where fuel is not divided equally among individual cylinders. time, after injection ended.33 There is also a time variation under steady-state engine conditions where the air/fuel ratio in a given cylinder varies cycle-by-cycle.41 Data on time-averaged air/fuel ratios in each cylinder of multicylinder engines show that the extent of PROBLEMS the maldistribution varies from engine to engine, and for a particular engine varies over the load and speed range. Spreads in the equivalence ratio (maximum 1.1. The equivalence ratio in a conventional spark-ignition engine varies from no load (idle) to full load, at a fixed engine speed, as shown at the top of Fig. P7-1. (By load to minimum) of about 5 percent of the mean value are typical at light load for is meant the percentage of the maximum brake torque at that speed.) Also shown is carbureted engines. Largest variations between cylinders usually occur at wide- the variation in total friction (pumping plus mechanical rubbing plus accessory open throttle. WOT spreads in the equivalence ratio of about 15 percent of the friction). Using formats similar to those shown, draw carefully proportioned qualitat- mean appear to be typical, again for carbureted engines, while spreads as high as ive graphs of the following parameters versus load (0 to 100 percent): 20 to 30 percent are not uncommon at particular speeds for some engines.23.40 Combustion efficiency, nc Time variations are less well defined; the limited data available suggest they Gross indicated fuel conversion efficiency, nf.ig could be of comparable magnitude.41 Gross indicated mean effective pressure, imep, With multipoint port fuel-injection systems, the fuel transport processes are Brake mean effective pressure, bmep substantially different and are not well understood. Air-flow phenomena are com- Mechanical efficiency, 1m parable to those with carbureted or throttle-body injection systems. However, manifold design can be optimized for air flow alone since fuel transport from the Indicate clearly where the maximum occurs if there is one, and where the value is throttle through the manifold is no longer a design constraint. Because the manu- zero or unity or some other obvious value, if appropriate. Provide a brief justification for the shape of the curves you draw. facture and operation of individual fuel injectors are not identical, there is still some variation in fuel mass injected cylinder-to-cylinder and cycle-to-cycle. Since 1.2 individual cylinder air flows depend on the design of the manifold, whereas the 1.0 amount of fuel injected does not, uniform air distribution is especially important 0.8 THAT with port injection systems. The fuel vaporization and transport processes will 100% depend on the duration of injection and the timing of injection pulse(s) relative Total fmep to the intake valve-lift profile. Some of the injected fuel will impinge on the port 0 walls, valve stem, and backside of the valve, especially when injection toward a 100 % FIGURE P7-1 closed valve occurs. Backflow of hot residual gases at part-load operation will - have a substantial effect on fuel vaporization. Compensation for fuel lag during 1.2. The four-cylinder spark-ignition engine shown in the figure uses an oxygen sensor in transient engine operation is still required; sudden throttle openings are accom- the exhaust system to determine whether the exhaust gas composition is lean or rich panied by a "lean spike" in the mixture delivered to the engine, comparable to of the stoichiometric point, and a throttle-body injection system with feedback to though smaller than that shown in Fig. 7-28 for a throttle-body fuel-injection maintain engine operation close to stoichiometric. However, since there is a time delay between a change in the fuel/air ratio at the injector location and the detection system. Thus wall wetting, evaporation off the wall, and liquid flow along the of that change by the sensor (corresponding to the flow time between the injector wall are all likely to be important with port fuel-injection systems also. and the sensor), the control system shown results in oscillations in fuel/air ratio With port fuel-injection systems, liquid fuel enters the cylinder and droplets about the stoichiometric point. are present during intake and compression. Limited measurements have been (a) Estimate the average flow time between the injector and the sensor at an engine made of the distribution, size, and number density of these fuel droplets. During speed of 2000 rev/min. 322 INTERNAL COMBUSTION ENGINE FUNDAMENTALS SI ENGINE FUEL METERING AND MANIFOLD PHENOMENA 323 (b) The sensor and control unit provide a voltage V of + V, volts when the fuel/air equivalence ratio o is less than one and a voltage of - V, volts when o is greater 7.4. Port fuel-injection systems are replacing carburetors in automobile spark-ignition engines. List the major advantages and any disadvantages of fuel metering with port than one. The feedback injection system provides a fuel/air ratio (F/A) given by fuel injection relative to carburetion. (1 + CVt) 75. With multipoint port fuel injection and single-point injection systems, the fuel flow rate is controlled by the injection pulse duration. If each injector operates contin- uously at the maximum rated power point (wide-open throttle, A/F = 12, 5500 where t is the time (in seconds) after the voltage signal last changed sign, rev/min) of an automobile spark-ignition engine, estimate approximately the injec- (F/A)=0 is the fuel/air ratio at the injector at t = 0, and C is a constant. Develop tion pulse duration (in crank angle degrees) for the same engine at idle. Idle condi- carefully proportioned quantitative sketches of the variation in the fuel/air ratio tions are: 700 rev/min, 0.3 atm inlet manifold pressure, stoichiometric-mixture. at the injector and at the exhaust sensor, with time, showing the phase relation between the two curves. Explain briefly how you developed these graphs. 7.6. The fuel-air cycle results indicate that the maximum imep is obtained with gasoline- (c) Find the value of the constant C, in volts-1-seconds" (the feedback system air mixtures at equivalence ratios of about 1.0. In practice, the maximum wide-open gain), such that (F/A) variations about the stoichiometric value do not exceed throttle power of a spark-ignition engine at a given speed is obtained with an air/fuel ratio of about 12. The vaporization of the additional gasoline lowers the temperature + 10 percent for V, = 1 V. of the inlet air and the richer mixture has a lower ratio of specific heats y_ during compression. Estimate approximately the change in mixture temperature due to Signal Control vaporization of the additional fuel used to decrease A/F from 14.6 (an equivalence unit ratio of 1.0) to 12.2 in the intake system, and the combined effect of vaporization and 0.4 m- -0.3 m- lower yy on the unburned mixture temperature at WOT when the cylinder pressure is at its peak of 40 atm. (The principal effect of the richer mixture is its impact on Air Engine knock.) Intake manifold, Sensor 7.7. (a) Plot dimensionless throttle plate open area 4/,/(RD2) as a function of throttle Fuel 3 cm diam Exhaust manifold, plate angle v. Assume vo = 10º, D (throttle bore diameter) = 57 mm, d (throttle Va = 500 cm3 3 cm diam shaft diameter) = 10.4 mm. What is the throttle plate area? Feedback per cylinder (b) Estimate the average velocity of the air flowing through the throttle plate open injector FIGURE P7-2 area for v = 26º at 3000 rev/min and v = 36º at 2000 rev/min. Use the relation- 7.3. In many spark-ignition engines, liquid fuel is added to the inlet air upstream of the ship between w, engine speed, and inlet manifold pressure given in Fig. 7-22. Assume a discharge coefficient CD = 0.8. inlet manifold above the throttle. The inlet manifold is heated to ensure that under steady-state conditions the fuel is vaporized before the mixture enters the cylinder. (c) For the throttle of part (a), estimate and plot the total force on the throttle plate (a) At normal wide-open throttle operating conditions, in a four-stroke cycle and shaft, and the force parallel and perpendicular to the throttle bore axis (i.e ., 1.6-dm3 displacement four-cylinder engine, at 2500 rev/min, the temperature of in the mean flow direction and normal to that direction) as a function of throttle the air entering the carburetor is 40ºC. The heat of vaporization of the fuel is angle at 2000 rev/min. Again use Fig. 7-22 for the relationship between y and inlet manifold pressure. 350 KJ/kg and the rate of heat transfer to the intake mixture is 1.4 kW. Estimate the temperature of the inlet mixture as it passes through the inlet valve, assuming 7.8. For the engine and intake manifold shown in Fig. 7-23, estimate the ratio of the that the fuel is fully vaporized. The volumetric efficiency is 0.85. The air density is intake manifold runner cross-sectional area to (TB2/4), the ratio of the length of the 1.06 kg/m3 and c, for air is 1 kj/kg . K. You may neglect the effects of the heat flow path from the intake manifold entrance to the inlet valve seat to the bore, the capacity of the liquid and vapor fuel. ratio of the volume of each inlet port to each cylinder's displaced volume, and the (b) With port electronic fuel-injection systems, the fuel is injected directly into the ratio of the volume of each intake manifold runner to each cylinder's displaced intake port. The intake manifold is no longer heated. However, the fuel is only volume. The cylinder bore is 89 mm. partly vaporized prior to entering the cylinder. Estimate the mixture.temperature as it passes through the inlet valve with the EFI system, assuming that the air temperature entering the intake manifold is still 40ºC and 50 percent of the fuel is REFERENCES vaporized. (c) Estimate the ratio of the maximum indicated power obtained at these conditions 1. Nakajima, Y ., Sugihara, K ., Takagi, Y ., and Muranaka, S.: " Effects of Exhaust Gas Recirculation with this engine with a carburetor, to the maximum power obtained with port on Fuel Consumption," in Proceedings of Institution of Mechanical Engineers, Automobile Divi- fuel injection. Assume that the inlet valve is the dominant restriction in the flow sion, vol. 195, no. 30, pp. 369-376, 1981. into the engine and that the pressure ratio across the inlet valve is the same for 2. Harrington, D. L ., and Bolt, J. A.: " Analysis and Digital Simulation of Carburetor Metering," SAE paper 700082, SAE Trans ., vol. 79, 1970. both carbureted and port-injection fueled engines. The intake mixture pressure 3. Bolt, J. A ., Derezinski, S. J ., and Harrington, D. L.: "Influence of Fuel Properties on Metering in and equivalence ratio remain the same in both these cases. Carburetors," SAE paper 710207, SAE Trans ., vol. 80, 1971. 324 INTERNAL COMBUSTION ENGINE FUNDAMENTALS SI ENGINE FUEL METERING AND MANIFOLD PHENOMENA 325 4. Khovakh, M.: Motor Vehicle Engines (English translation), Mir Publishers, Moscow, 1976. 5. Bolt, J. A ., and Boerma, M. J.: " Influence of Air Pressure and Temperature on Carburetor Meter- 11. Bridgeman, O. C.: "Equilibrium Volatility of Motor Fuels from the Standpoint of Their Use in ing," SAE paper 660119, 1966. Internal Combustion Engines," National Bureau of Standards research paper 694, 1934. 6. Shinoda, K ., Koide, H ., and Yii, A.: " Analysis and Experiments on Carburetor Metering at the 12 ASTM Standard Method: " Distillation of Petroleum Products," ANSI/ASTM D86 (1P 123/68). Transition Region to the Main System," SAE paper 710206, SAE Trans ., vol. 80, 1971. 11. Peters, B. D.: "Laser-Video Imaging and Measurement of Fuel Droplets in a Spark-Ignition 7. Oya, T.: "Upward Liquid Flow in a Small Tube into which Air Streams," Bull. JSME, vol. 14. Engine," in Proceedings of Conference on Combustion in Engineering, Oxford, Apr. 11-14, 1983, no. 78, pp. 1320-1329, 1971. Institution of Mechanical Engineers, 1983. 8. Wrausmann, R. C ., and Smith, R. J.: "An Approach to Altitude Compensation of the Carbu- . Sirignano, W. A.: "Fuel Droplet Vaporization and Spray Combustion Theory," Prog. Energy and retor," SAE paper 760286, 1976. Combust. Sci ., vol. 9, pp. 291-322, 1983. 9. Bosch, Automotive Handbook, 1st English ed ., Robert Bosch GmbH, Stuttgart, 1978. 1. Boam, D. J ., and Finlay, I. C.: "A Computer Model of Fuel Evaporation in the Intake System of 10. Glöckler, O ., Knapp, H ., and Manger, H.: " Present Status and Future Development of Gasoline a Carbureted Petrol Engine," Conference on Fuel Economy and Emissions of Lean Burn Engines, Fuel Injection Systems for Passenger Cars," SAE paper 800467, 1980. London, June 12-14, 1979, paper C89/79, Institution of Mechanical Engineers, 1979. 11. Greiner, M ., Romann, P ., and Steinbrenner, U.: "BOSCH Fuel Injectors-New Developments," 16. Yun, H. J ., and Lo, R. S.: "Theoretical Studies of Fuel Droplet Evaporation and Transportation SAE paper 870124, 1987. in a Carburetor Venturi," SAE paper 760289, 1976. 12. Gorille, I ., Rittmannsberger, N ., and Werner, P.: "Bosch Electronic Fuel Injection with Closed 37. Servati, H. B ., and Yuen, W. W.: "Deposition of Fuel Droplets in Horizontal Intake Manifolds Loop Control," SAE paper 750368, SAE Trans ., vol. 84, 1975. and the Behavior of Fuel Film Flow on Its Walls," SAE paper 840239, SAE Trans ., vol. 93, 1984. 13. Czadzeck, G. H.: " Ford's 1980 Central Fuel Injection System," SAE paper 790742, 1979. 38. Hires, S. D ., and Overington, M. T.: "Transient Mixture Strength Excursions -- An Investigation 14. Bowler, L. L.: "Throttle Body Fuel Injection (TBI)-An Integrated Engine Control System," SAE of Their Causes and the Development of a Constant Mixture Strength Fueling Strategy," SAE paper 800164, SAE Trans ., vol. 89, 1980. paper 810495, SAE Trans ., vol. 90, 1981. 15. Hamann, E ., Manger, H ., and Steinke, L.: "Lambda-Sensor with Y2O3-Stabilized ZrO2-Ceramic 39. Collins, M. H.: "A Technique to Characterize Quantitatively the Air/Fuel Mixture in the Inlet for Application in Automotive Emission Control Systems," SAE paper 770401, SAE Trans ., vol Manifold of a Gasoline Engine," SAE paper 690515, SAE Trans ., vol. 78, 1969. 86, 1977. 4). Blackmore, D. R ., and Thomas, A.: Fuel Economy of the Gasoline Engine, John Wiley, 1977. 16. Seiter, R. E ., and Clark, R. J.: "Ford Three-Way Catalyst and Feedback Fuel Control System," 41. Yu, H. T. C.: "Fuel Distribution Studies-A New Look at an Old Problem," SAE Trans ., vol. 71, SAE paper 780203, SAE Trans ., vol. 87, 1978. pp. 596-613, 1963. 17. Camp, J ., and Rachel, T.: "Closed-Loop Electronic Fuel and Air Control of Internal Combustion --- Engines," SAE paper 750369, 1975. 18. Liimatta, D. R ., Hurt, R. F ., Deller, R. W ., and Hull, W. L.: " Effects of Mixture Distribution oo Exhaust Emissions as Indicated by Engine Data and the Hydraulic Analogy," SAE paper 710618. SAE Trans ., vol. 80, 1971. 19. Benson, R. S ., Baruah, P. C ., and Sierens, I. R.: "Steady and Non-steady Flow in a Simple- Carburetor," in Proceedings of Institution of Mechanical Engineers, vol. 188, no. 53/74, pp. 537- 548, 1974. 20. Woods, W. A ., and Goh, G. K.: "Compressible Flow through a Butterfly Throttle Valve in a Pipe," in Proceedings of Institution of Mechanical Engineers, vol. 193, no. 10, pp. 237-244, 1979. 21. Walker, J. W.: "The GM 1.8 Liter L-4 Gasoline Engine Designed by Chevrolet," SAE paper 820111, SAE Trans ., vol. 91, 1982. 22. Chapman, M.: "Two Dimensional Numerical Simulation of Inlet Manifold Flow in a Four Cylinder Internal Combustion Engine," SAE paper 790244, 1979. 23. Kay, I. W.: "Manifold Fuel Film Effects in an SI Engine," SAE paper 780944, 1978. 24. Brandstetter, W. R ., and Carr, M. J.: "Measurement of Air Distribution in a Multicylinder Engine by Means of a Mass Flow Probe," SAE paper 730494, 1973. 25. Aquino, C. F.: "Transient A/F Control Characteristics of the 5 Liter Central Fuel Injection Engine," SAE 810494, SAE Trans ., vol. 90, 1981. 26. Trayser, D. A ., Creswick, F. A ., Giesike, J. A ., Hazard, H. R ., Weller, A. E ., and Locklin, D. W.: " A Study of the Influence of Fuel Atomization, Vaporization, and Mixing Processes on Pollutant Emissions from Motor-Vehicle Powerplants," Battelle Memorial Institute, Columbus, Ohio, 1969. 27. Tabaczysnki, R. J.: " Effects of Inlet and Exhaust System Design on Engine Performance," SAE paper 821577, 1982. 28. Engelman, H. W.: "Design of a Tuned Intake Manifold," ASME paper 73-WA/DGP-2, 1973. 29. Benson, R. S.: in J. H. Horlock and D. E. Winterbone (eds.), The Thermodynamics and G Dynamics of Internal Combustion Engines, vol. 1, Clarendon Press, Oxford, 1982. 30. Chapman, M ., Novak, J. M ., and Stein, R. A.: "Numerical Modeling of Inlet and Exhaust Flows in Multi-cylinder Internal Combustion Engines," in Flows in Internal Combustion Engines, ASMB Winter Annual Meeting, Nov. 14-19, 1982, ASME, New York. CHARGE MOTION WITHIN THE CYLINDER 327 CHAPTER Mean -o- rms 8 CHARGE MOTION 30 WITHIN cooog THE 100-0-000 r, mm 20 CYLINDER 10 0 2 4 6 10 20 30 40 z, mm 070 z, mm V 2 SOIINE FIGURE 8-1 Radial mean velocity D, and root mean square (rms) velocity fluctuations v', at the valve exit plane, and axial mean velocity o, and rms velocity fluctuation v', 15 mm below the cylinder head, at 36º ATC in model engine operated at 200 rev/min. Valve lift = 6 mm. Velocities normalized by mean piston speed. 1 Gas motion within the engine cylinder is one of the major factors that controls the combustion process in spark-ignition engines and the fuel-air mixing and combustion processes in diesel engines. It also has a significant impact on heat seat and lip, producing shear layers with large velocity gradients which generate transfer. Both the bulk gas motion and the turbulence characteristics of the flow turbulence. This separation of the jet sets up recirculation regions beneath the are important. The initial in-cylinder flow pattern is set up by the intake process. valve head and in the corner between the cylinder wall and cylinder head. It may then be substantially modified during compression. This chapter reviews The motion of the intake jet within the cylinder is shown in the schlieren the important features of gas motion within the cylinder set up by flows into and photographs in Fig. 8-2 taken in a transparent engine. This engine had a square out of the cylinder through valves or ports, and by the motion of the piston. cross-section cylinder made up of two quartz walls and two steel walls, to permit easy optical access. The schlieren technique makes regions with density gradients in the flow show up as lighter or darker regions on the film.2 The engine was 8.1 INTAKE JET FLOW throttled to one-half an atmosphere intake pressure, so the jet starts after the The engine intake process governs many important aspects of the flow within the intake stroke has commenced, at 35º ATC, following backflow of residual into cylinder. In four-stroke cycle engines, the inlet valve is the minimum area for the the intake manifold. The front of the intake jet can be seen propagating from the flow (see Sec. 6.3) so gas velocities at the valve are the highest velocities set up valve to the cylinder wall at several times the mean piston speed. Once the jet during the intake process. The gas issues from the valve opening into the cylinder reaches the wall (0 > 41º ATC), the wall deflects the major portion of the jet as a conical jet and the radial and axial velocities in the jet are about 10 times the downward toward the piston; however, a substantial fraction flows upward mean piston speed. Figure 8-1 shows the radial and axial velocity components toward the cylinder head. The highly turbulent nature of the jet is evident. close to the valve exit, measured during the intake process, in a motored model The interaction of the intake jet with the wall produces large-scale rotating engine with transparent walls and single valve located on the cylinder axis, using flow patterns within the cylinder volume. These are easiest to visualize where the laser doppler anemometry (see next section).1 The jet separates from the valve engine geometry has been simplified so the flow is axisymmetric. The photograph 326 CHARGE MOTION WITHIN THE CYLINDER 329 328 INTERNAL COMBUSTION ENGINE FUNDAMENTALS of the intake generated flow in a thin illuminated plane through the cylinder axis. The streaks are records of the paths of tracer particles in the flow during the period the camera shutter is open. The bulk of the cylinder as the piston moves down is filled with a large ring vortex, whose center moves downward and remains about halfway between the piston and the head. The upper corner of the cylinder contains a smaller vortex, rotating in the opposite direction. These vor- tices persist until about the end of the intake stroke, when they became unstable and break up.3 With inlet valve location and inlet port geometry more typical of normal ( b) 350 (c) 36º (a) 30º engine practice, the intake generated flow is more complex. However, the pre- sence of large-scale rotating flow patterns can still be discerned. Figure 8-4a shows the effect of off-axis valve location (with the flow into the valve still paral- lel to the cylinder and valve axis). During the first half of the inlet stroke, at least, a flow pattern similar in character to that in Fig. 8-3 is evident. The vortices are now displaced to one side, however, and the planes of their axes of rotation are no longer perpendicular to the cylinder axis but are tipped at an angle to it. The vortices become unstable and break up earlier in the intake stroke than was the case with the axisymmetric flow.3 With an offset valve and a normal inlet port (d) 390 ( e) 41º (f) 700 configuration which turns the flow through 50 to 70º (see Fig. 6-13), photographs FIGURE 8-2 Sequence of schlieren photographs of intake jet as it develops during intake stroke. Numbers are crank angle degrees after TC.2 in Fig. 8-3 of a water analog of an engine intake flow was taken in a transparent model of an engine cylinder and piston. The valve is located in the center of the cylinder head, and the flow into the valve is along the cylinder axis. The experi- mental parameters have been scaled so that the appropriate dimensionless numbers which govern the flow, the Reynolds and Strouhal numbers, were main- tained equal to typical engine values. The photograph shows the major features (a) (b) FIGURE 8-3 FIGURE 8-4 Large-scale rotating flow pattern Sketches from: (a) streak photographs of in-cylinder intake generated flow in water analog of intake set up within the cylinder by the intake jet. Photograph of streak process in model engine with offset inlet valve, at 90º ATC;3 (b) streak photographs of flow in diam- lines in water flow into model ctral plane; 30 mm below cylinder head, with intake port and valve geometry shown, with steady water flow into cylinder. Valve lift = 4 mm.4 engine with axisymmetric valve." 330 INTERNAL COMBUSTION ENGINE FUNDAMENTALS CHARGE MOTION WITHIN THE CYLINDER 331 of the flow pattern in a diametral plane show an additional large-scale rotation. Figure 8-4b shows the flow pattern observed in a water-flow model of the cylin- domness. Statistical methods are therefore used to define such a flow field. The der in a plane 30 mm (one-third of the bore) from the cylinder head, with quantities normally used are: the mean velocity, the fluctuating velocity about standard inlet port design. The direction of flow with this vortex pair is toward the mean, and several length and time scales. In a steady turbulent flow situation, the left across the center of the cylinder. This flow pattern occurs because the the instantaneous local fluid velocity U (in a specific direction) is written: cylinder wall closest to the valve impedes the flow out of the valve and forces the U(t) = U + u(t) flow on either side of the plane passing through the valve and cylinder axes to (8.1) circulate around the cylinder in opposite directions. The upper vortex follows the For steady flow, the mean velocity U is the time average of U(t): flow direction of the port and becomes larger still as the valve lift increases. The 1 'to+z details of this aspect of the intake flow depend on the port design, valve stem U = lim - U(t) dt (8.2) orientation, and the valve lift.4 With suitable port and/or cylinder head design, it [ ~ 00 1 Jto is possible to develop a single vortex flow within the bulk of the cylinder. The The fluctuating velocity component u is defined by its root mean square value, production and characteristics of such "swirling" flows are reviewed in Sec. 8.3. the turbulence intensity, u': In summary, the jet-like character of the intake flow, interacting with the cylinder walls and moving piston, creates large-scale rotating flow patterns within to + + u' = lim 1/2 the cylinder. The details of these flows are strongly dependent on the inlet port, u2 dt (8.3a) + 00 Jto valve, and cylinder head geometry. These flows appear to become unstable, either Alternatively, during the intake or the compression process, and break down into three- dimensional turbulent motions. Recirculating flows of this type are usually sensi- tive to small variations in the flow: hence there are probably substantial u' = lim ( U2 - 02) dc 1 - 00 (8.3b) Jzo cycle-by-cycle flow variations.5 since the time average of (uU) is zero. In engines, the application of these turbulence concepts is complicated by 8.2 MEAN VELOCITY AND the fact that the flow pattern changes during the engine cycle. Also, while the TURBULENCE CHARACTERISTICS overall features of the flow repeat each cycle, the details do not because the mean flow can vary significantly from one engine cycle to the next. There are both 8.2.1 Definitions cycle-to-cycle variations in the mean or bulk flow at any point in the cycle, as The flow processes in the engine cylinder are turbulent. In turbulent flows, the well as turbulent fluctuations about that specific cycle's mean flow. rates of transfer and mixing are several times greater than the rates due to molec- One approach used in quasi-periodic flows such as that which occurs in the ular diffusion. This turbulent "diffusion" results from the local fluctuations in the engine cylinder is ensemble-averaging or phase-averaging. Usually, velocity mea- flow field. It leads to increased rates of momentum and heat and mass transfer, surements are made over many engine cycles, and over a range of crank angles. and is essential to the satisfactory operation of spark-ignition and diesel engines. The instantaneous velocity at a specific crank angle position 0 in a particular Turbulent flows are always dissipative. Viscous shear stresses perform deforma- cycle i can be written as tion work on the fluid which increases its internal energy at the expense of its turbulence kinetic energy. So energy is required to generate turbulence: if no U(0, i) = Ù(0, i) + u(0, i) (8.4) energy is supplied, turbulence decays. A common source of energy for turbulent The ensemble- or phase-averaged velocity, U(0), is defined as the average of velocity fluctuations is shear in the mean flow. Turbulence is rotational and is values at a specific phase or crank angle in the basic cycle. Figure 8-5 shows this characterized by high fluctuating vorticity: these vorticity fluctuations can only approach applied schematically to the velocity variation during a two-stroke persist if the velocity fluctuations are three dimensional.6. engine cycle, with small and large cycle-to-cycle variations. The ensemble- The character of a turbulent flow depends on its environment. In the engine averaged velocity is the average over a large number of measurements taken at cylinder, the flow involves a complicated combination of turbulent shear layers, the same crank angle (two such points are indicated by dots): recirculating regions, and boundary layers. The flow is unsteady and may exhibe substantial cycle-to-cycle fluctuations. Both large-scale and small-scale turbulent Ü BA(O) == motions are important factors governing the overall behavior of the flow.5 (8.5) An important characteristic of a turbulent flow is its irregularity or ran- where Ne is the number of cycles for which data are available. By repeating this CHARGE MOTION WITHIN THE CYLINDER 333 332 INTERNAL COMBUSTION ENGINE FUNDAMENTALS (a) Low cycle-to-cycle variation Instantaneous Velocity IK Time Ensemble average - FIGURE 8-6 (b) Large cycle-to-cycle variation Schematic of jet created by flow through the intake valve indicating its turbulent structure. 5, 6 Instantaneous Individual cycle mean Velocity Figure 8-5 illustrates this breakdown of the instantaneous velocity into an ensemble-averaged component, an individual-cycle mean velocity, and a com- ponent which randomly fluctuates in time at a particular point in space in a single cycle. This last component is the conventional definition of the turbulent Time velocity fluctuation. Whether this differs significantly from the fluctuations about Ensemble average the ensemble-averaged velocity depends on whether the cycle-to-cycle fluctua- tions are small or large. The figure indicates these two extremes.+ In turbulent flows, a number of length scales exist that characterize different FIGURE &-5 Schematic of velocity variation with crank angle at a fixed location in the cylinder during two con- aspects of the flow behavior. The largest eddies in the flow are limited in size by secutive cycles of an engine. Dots indicate measurements of instantaneous velocity at the same crank the system boundaries. The smallest scales of the turbulent motion are limited by angle. Ensemble- or phase-averaged velocity obtained by averaging over a large number of such molecular diffusion. The important length scales are illustrated by the schematic measurements shown as solid smooth line. Top graph: low cycle-to-cycle flow variations. Here the of the jet issuing into the cylinder from the intake valve in Fig. 8-6. The eddies individual-cycle mean velocity and ensemble-averaged velocity are closely comparable. Bottom graph: large cycle-to-cycle variations. Here the individual-cycle mean velocity (dotted line) is different responsible for most of the turbulence production during intake are the large from the ensemble-averaged mean by U. The turbulent fluctuation u is then defined in relation to the eddies in the conical inlet jet flow. These are roughly equal in size to the local jet individual-cycle mean.5 thickness. This scale is called the integral scale, l,: it is a measure of the largest scale structure of the flow field. Velocity measurements made at two points separated by a distance x significantly less than I, will correlate with each other; process at many crank angle locations, the ensemble-averaged velocity profile with x >> l, no correlation will exist. The integral length scale is, therefore, over the complete cycle is obtained. defined as the integral of the autocorrelation coefficient of the fluctuating velocity The ensemble-averaged mean velocity is only a function of crank angle at two adjacent points in the flow with respect to the variable distance between since the cyclic variation has been averaged out. The difference between the mean velocity in a particular cycle and the ensemble-averaged mean velocity over many cycles is defined as the cycle-by-cycle variation in mean velocity: There is considerable debate as to whether the fluctuating components of the velocity U(0, i) defined 0(0, i) = 0(0, 1) - UEA(0) (8.6) by Eq. (8.7) (cycle fluctuations in the mean velocity and fluctuations in time about the individual cycle mean) are physically distinct phenomena. The high-frequency fluctuations in velocity are often defined Thus the instantaneous velocity, given by Eq. (8.4), can be split into three com- as "turbulence." The low-frequency fluctuations are generally attributed to the variations in the mean ponents:7 Sow between individual cycles, a phenomenon that is well established. Whether this distinction is (8.7) valid has yet to be resolved. U(0, 1) = UEA(0) + 0(0, 1) + u(0, 1) 334 INTERNAL COMBUSTION ENGINE FUNDAMENTALS CHARGE MOTION WITHIN THE CYLINDER 335 to be isotropic (have no preferred direction) than are the large eddies, and have a structure like that of other turbulent flows. The dissipation of turbulence energy takes place in the smallest structures. At this smallest scale of the turbulent motion, called the Kolmogorov scale, molecular viscosity acts to dissipate small- 4 ( 1 0 ) u ( to + x ) ( x )u'(x + x ) scale kinetic energy into heat. If & is the energy dissipation rate per unit mass and , the kinematic viscosity, Kolmogorov length and time scales are defined by FIGURE 8-7 Spatial velocity autocorrelation R, as a function of ( 4 3) 1/4 1/2 RI = - x, defining the integral length scale l, and the 1 x = TK = (8.11) IM micro length scale IM. The Kolmogorov length scale indicates the size of the smallest eddies. The Kolmogorov time scale characterizes the momentum-diffusion of these smallest the points, as shown in Fig. 8-7: i.e ., structures. A third scale is useful in characterizing a turbulent flow. It is called the 8 Rx dx (8.8a) microscale (or Taylor microscale). The micro length scale IM is defined by relating Jo the fluctuating strain rate of the turbulent flow field to the turbulence intensity: where ou u' (8.12) 1 u ( xo )u( xo + x ) R, = (8.8b) N_- 1 1=1 u'(xo)u'(xo + x) It can be determined from the curvature of the spatial correlation curve at the This technique for determining the integral scale requires simultaneous measure- origin, as shown in Fig. 8-7.5. 6 More commonly, the micro time scale ty is deter- mined from the temporal autocorrelation function of Eq. (8.9): ments at two points. Due to the difficulty of applying such a technique in engines, most efforts to determine length scales have first employed correlations to deter- 2 mine the integral time scale, t1. The integral time scale of turbulence is defined as TM = " (22 R,/at2)Lo a correlation between two velocities at a fixed point in space, but separated in For turbulence which is homogeneous (has no spatial gradients) and is isotropic time: (has no preferred direction), the microscales IM and ty are related by = R, dt (8.9a) (8.13) JO These different scales are related as follows. The turbulent kinetic energy where per unit mass in the large-scale eddies is proportional to u'2. Large eddies lose a " u(to)u(to + t) substantial fraction of this energy in one "turnover" time Ij/u'. In an equilibrium 1- (8.9b R: - N - 1 1-1 u'(to)u'(to + t) situation the rate of energy supply equals the rate of dissipation: y'3 and N, is the number of measurements. Under conditions where the turbulence pattern is convecteurast the observation point without significant distortion and the turbulence is relatively weak, the integral length and time scales are related Thus, by 3/4 1, = Ût1 (8.10) V = Re- 3/4 (8.14) In flows where the large-scale structures are convecteur, is a measure of the time where Rer is the turbulent Reynolds number, u'll/v. it takes a large eddy to pass a point. In flows without mean motion, the integral Within the restrictions of homogeneous and isotropic turbulence, an energy time scale is an indication of the lifetime of an eddy.5, 8 budget can be used to relate I, and IM:6 Superposed on this large-scale flow is a range of eddies of smaller and Au'3 smaller size, fed by the continual breakdown of larger eddies. Since the smalla 15 vu'2 eddies respond more rapidly to changes in local flow pattern, they are more likely 11 336 INTERNAL COMBUSTION ENGINE FUNDAMENTALS CHARGE MOTION WITHIN THE CYLINDER 337 where A is a constant of order 1. Thus, velocity fluctuation is 15 1/2 Re - 1/ 2 (8.15) "'F . EA (D ) = 14 (8.18) These restrictions are not usually satisfied within the engine cylinder during intake. They are approximately satisfied at the end of compression. where Ui, j = UI,J - UEA (8.19) 8.2.2 Application to Engine Velocity Data As has already been explained, this definition of fluctuation intensity [the As has been explained above, it is necessary to analyze velocity data on an indi- ensemble-averaged rms velocity fluctuation, Eq. (8.18)] includes cyclic variations vidual cycle basis as well as using ensemble-averaging techniques. The basic defi- in the mean flow as well as the turbulent fluctuations about each cycle's mean nitions for obtaining velocities which characterize the flow will now be developed. flow.7 It is necessary to determine the mean and fluctuating velocities on an The ensemble-averaged velocity UEA has already been defined by Eq. (8.5). The individual-cycle basis to characterize the flow field more completely. The critical ensemble-averaged fluctuation intensity UF, EA is given by part of this process is defining the mean velocity at a specific crank angle (or within a small window centered about that crank angle) in each cycle. Several L'F , BA (0) = [ [u(o, )]2( 1/2 =N [V(O , ) 2 - U E A ( 0 ) 2 ] ( 8.16 ) methods have been used to determine this individual-cycle mean velocity (e.g ., moving window, low-pass filtering, data smoothing, conditional sampling; see Ref. 7 for a summary). A high data rate is required. It includes all fluctuations about the ensemble-averaged mean velocity. In this individual-cycle velocity analysis the individual-cycle time-averaged Use of Eqs. (8.5) and (8.16) requires values for U and u at each specific or mean velocity U(0, i) is first determined.7.12 The ensemble average of this crank angle under consideration. While some measurement techniques (e.g ., hot- mean velocity wire anemometry) provide this, the preferred velocity measurement method (laser doppler anemometry) provides an intermittent signal. With laser doppler ane- mometry (LDA), interference fringes are produced within the small volume UFA(0) = + 20 ( 0 + 2, (8.20) created by the intersection of two laser beams within the flow field. When a small particle passes through this volume, it scatters light at a frequency proportional is identical to the ensemble-averaged value given by Eq. (8.17). The root mean to the particle velocity. By seeding the flow with particles small enough to be square fluctuation in individual-cycle mean velocity can then be determined from carried without slip by the flow and collecting this scattered light, the flow veloc- ity is determined.9 A signal is only produced when a particle moves through the measurement volume; thus one collects data as velocity crank angle pairs. It is URMS ( ) = 0(0 8.1 ) - DEA(O) ]]12 (8.21) necessary, therefore, to perform the ensemble-averaging over a finite crank angle window 40 around the specific crank angle of interest, 0. The ensemble-averaged This indicates the magnitude of the cyclic fluctuations in the mean motion. The velocity equation becomes instantaneous velocity fluctuation from the mean velocity, within a specified window 40 at a particular crank angle 0, is obtained from Eq. (8.4). This instan- (8.17) taneous velocity fluctuation is ensemble-averaged, because it varies substantially cycle-by-cycle and because the amount of data is usually insufficient to give reli- able individual-cycle results: where N; is the number of velocity measurements recorded in the window during the ith cycle, Ne is the number of cycles, and N, is the total number of measure- ments.+ The corresponding equation for the ensemble-averaged root mean square (8.22) This quantity is the ensemble-averaged turbulence intensity. Several different techniques have been used to measure gas velocities within + This need to ensemble-average over a finite crank angle window introduces an error called crank the engine cylinder (see Refs. 13 and 14 for brief reviews and references). The angle broadening, due to change in the mean velocity across the window. This error depends on the technique which provides most complete and accurate data is laser doppler ane- velocity gradient, and can be made negligible small by suitable choice of window size.9-11 mometry.9 Sample results obtained with this technique will now be reviewed to 338 INTERNAL COMBUSTION ENGINE FUNDAMENTALS CHARGE MOTION WITHIN THE CYLINDER 339 illustrate the major features of the in-cylinder gas motion. The available resuha, must be interpreted with caution since they have been obtained in special engine 6 Smoothed ensemble where the geometry and flow have been modified to make the experiments and Smoothed ensemble Cycle by cycle b Cycle by cycle c their interpretation easier. Also, the flow within the cylinder is three dimensional 4 in nature. It takes measurements at many points within the flow field and the use Turbulence or rms velocity fluctuation, m/s of a flow visualization technique to characterize the flow adequately. Figure 8-8 shows ensemble-averaged velocities throughout the engine cycle 2 at two measurement locations in a special L-head engine designed to generate a swirling flow within the cylinder. The engine was motored at 300 rev/min, giving a mean piston speed of 0.76 m/s. Figure 8-8b shows the mean velocity in the path 180 360 540 720 180 360 540 72 of the swirling intake flow within the clearance volume, in the swirl direction Intake Compression Expansion Exhaust Intake Compression Expansion Exhaust Crank angle, deg High velocities are generated during the intake process, rising to a maximum and Crank angle, deg then decreasing in response to the piston motion (see Fig. 2-2). During the com- (a) ( b ) pression stroke, the velocity continues to decrease but at a much slower rate. This NGURE &-9 is a motored engine cycle. A comparison of intake and compression velocities Emsemble-averaged rms velocity fluctuation and ensemble-averaged individual-cycle turbulence inten- with an equivalent firing cycle showed close agreement.15 The expansion and uty as a function of crank angle: (a) at location b in Fig. 8-8a; (b) at location c in Fig. 8-8a.11 exhaust stroke velocities are not typical of firing engine behavior, however.+ Figure 8-8c shows the mean velocity in the clearance volume in the same direc- br tion but on the cylinder axis. At this location, positive and negative flow veloc- Intake ities were measured. Since this location is out of the path of the intake generated - c+- fow, velocities during the intake stroke are much lower. The nonhomogeneous character of this particular ensemble-mean flow is evident. Exhaust Figure 8-9 shows the ensemble-averaged rms velocity fluctuation (which includes contributions from cycle-by-cycle variations in the mean flow and turbulence) and the ensemble-averaged individual-cycle turbulence intensity at these same two locations and directions. The difference between the two curves in (a) cach graph is an indication of the cycle-by-cycle variation in the mean flow [see 20 Eq. (8.7)]. During the intake process, within the directed intake flow pattern, the 40 cycle-by-cycle variation in the mean flow is small in comparison to the high b turbulence levels created by the intake flow. Outside this directed flow region, 30 10F again during intake, this cycle-by-cycle contribution is more significant relative to the turbulence. During compression, the cycle-by-cycle mean flow variation is Mean velocity, m/s Mean velocity, m/s 20 comparable in magnitude to the ensemble-averaged turbulence intensity. It is therefore highly significant. Two important questions regarding the turbulence in the engine cylinder 0 -10L 180 360 540 720 -5 are whether it is homogeneous (i.e ., uniform) and whether it is isotropic (i.e ., 0 180 360 540 720 0 Intake Compression Exhaust Expansion independent of direction). The data already presented in Figs. 8-8 and 8-9 show Intake Compression Expansion Exhaust Crank angle, deg Crank angle, deg that during intake the flow is far from homogeneous. Nor is it isotropic.11 (b) (c) FIGURE 8-8 Ensemble-averaged velocities throughout the engine cycle in motored four-stroke L-head engine: rev/min, mean piston speed 0.76 m/s. (a) Engine schematic showing measurement locations and velo The increase in velocity when the exhaust valve opens is due to the flow of gas into the cylinder Acause, due primarily to heat losses, the cylinder pressure is then below 1 atm. ity directions; (b) velocity at b in intake flow path; (c) velocity at c on cylinder axis.11 340 INTERNAL COMBUSTION ENGINE FUNDAMENTALS CHARGE MOTION WITHIN THE CYLINDER 341 However, it is the character of the turbulence at the end of the compression Swirl process that is most important: that is what controls the fuel-air mixing and Ported O x UT. EA OX burning rates. Only limited data are available which relate to these questions With open disc-shaped combustion chambers, measurements at different locs tions in the clearance volume around TC at the end of compression show the turbulence intensity to be relatively homogeneous. In the absence of an intak generated swirling flow, the turbulence intensity was also essentially isotropic near TC.16 These specific results support the more general conclusion that the inlet boundary conditions play the dominant role in the generation of the meant flow and turbulence fields during the intake stroke. However, in the absence of swirl, this intake generated flow structure has almost disappeared by the time the compression process commences. The turbulence levels follow this trend in the mean flow, with the rapid decay process lasting until intake valve closing. Later in the compression process the turbulence becomes essentially homogeneous.17 4 6 00 When a swirling flow is generated during intake, an almost solid-body 10 Mean piston speed, m/s rotating flow develops which remains stable for much longer than the inlet jet FIGURE 8-11 generated rotating flows illustrated in Fig. 8-3. With simple disc-shaped cham- Individual-cycle turbulence intensity u'r. EA (OX) and ensemble-averaged rms fluctuation velocity bers, the turbulence still appears to become almost homogeneous at the end of (remaining symbols) at TC at the end of compression, for a number of different flow configurations compression. With swirl and bowl-in-piston geometry chambers, however, the and chamber geometries as a function of mean piston speed.16 Two data sets for two-stroke ported flow is more complex (see Sec. 8.3). mgines. Four data sets with intake generated swirl. The flow through the intake valve or port is responsible for many features of the in-cylinder motion. The flow velocity through the valve is proportional to the piston speed [see Eq. (6.10) for pseudo valve flow velocity, and Eq. 2.10)]. It appropriateness of this velocity scaling.+ Other results support this conclusion, though in the absence of an ordered mean motion such as swirl when the would be expected therefore that in-cylinder flow velocities at different engine speeds would scale with mean piston speed [Eq. (2.11)]. Figure 8-10 shows ensemble-averaged mean velocities at the end of compression are low, this scaling ensemble-averaged mean and rms velocity fluctuations, normalized by the mean for the mean velocity does not always hold.16 Figure 8-11 shows a compilation of piston speed through the cycle at three different engine speeds. The measurement ensemble-averaged rms fluctuation velocity or ensemble-averaged individual- cycle turbulence intensity results at TC at the end of compression, from 13 differ- location is in the path of the intake generated swirling flow (point b in Fig. 8-80). ent flow configurations and combustion chamber geometries. Two of these sets of All the curves have approximately the same shape and magnitude, indicating the results are from two-stroke cycle ported configurations. The measured fluctuating velocities or turbulence intensities are plotted against mean piston speed. The linear relationship holds well. There is a substantial variation in the proportion- 50 ,400 ality constant, in part because in most of these studies (identified in the figure) the 40 12 800, rev/min ensemble-averaged rms fluctuation velocity was the quantity measured. Since this 30 800 rev/min includes the cycle-by-cycle fluctuation in the mean velocity, it is larger (by up to a ₡270 Sks 20 270 8 factor of 2) than the average turbulence intensity u'T, EA . 400 U'F, EA 800/1. 400 270 A consensus conclusion is emerging from these studies that the turbulence intensity at top-center, with open combustion chambers in the absence of swirl, 270 has a maximum value equal to about half the mean piston speed :16. 18 O 180 360 540 720 180 360 540 UT, EA(TC) ~ 0.55, (8.23) Intake Compression Expansion Exhaust Intake Compression Expansion Exhaust Crank angle, deg Crank angle, deg FIGURE 8-10 Ensemble-averaged mean and rms velocity fluctuations, normalized by mean piston speed, through Note that because of the valve and combustion chamber of this particular engine, the ratio of U to out the engine cycle for three engine speeds. Location b in Fig. 8-8a.11 , is higher than is typical of normal engine geometries . 342 INTERNAL COMBUSTION ENGINE FUNDAMENTALS CHARGE MOTION WITHIN THE CYLINDER 343 The available data show that the turbulence intensity at TC with swirl is usually and the injected fuel. Swirl is also used to-speed up the combustion process in higher than without swirl16 (see the four data sets with swirl in Fig. 8-11). Some spark-ignition engines. In two-stroke engines it is used to improve scavenging. In data, however, indicate that the rms fluctuation intensity with swirl may be some designs of prechamber engines, organized rotation about the prechamber lower.18 The ensemble-averaged cyclic variation in individual-cycle mean velocity axis is also called swirl. In prechamber engines where swirl within the precom- at the end of compression also scales with mean piston speed. This quantity can bustion chamber is important, the flow into the prechamber during the compres- be comparable in magnitude to the turbulence intensity. It usually decreases sion process creates the rotating flow. Prechamber flows are discussed in Sec. 8.5. when a swirling flow is generated within the cylinder during the intake process. 11, 16 During the compression stroke, and also during combustion while the 8.3.1 Swirl Measurement cylinder pressure continues to rise, the unburned mixture is compressed. Turbu- lent flow properties are changed significantly by the large and rapidly imposed The nature of the swirling flow in an actual operating engine is extremely difficult distortions that result from this compression. Such distortions, in the absence of to determine. Accordingly, steady flow tests are often used to characterize the dissipation, would conserve the angular momentum of the flow: rapid compres- swirl. Air is blown steadily through the inlet port and valve assembly in the sion would lead to an increase in vorticity and turbulence intensity. There is cylinder head into an appropriately located equivalent of the cylinder. A common evidence that, with certain types of in-cylinder flow pattern, an increase in turbu- technique for characterizing the swirl within the cylinder has been to use a light lence intensity resulting from piston motion and combustion does occur toward paddle wheel, pivoted on the cylinder centerline (with low friction bearings), the end of the compression process. The compression of large-scale rotating flows mounted between 1 and 1.5 bore diameters down the cylinder. The paddle wheel can cause this increase due to the increasing angular velocity required to con- diameter is close to the cylinder bore. The rotation rate of the paddle wheel is serve angular momentum resulting in a growth in turbulence generation by used as a measure of the air swirl. Since this rotation rate depends on the loca- shear.19 tion of the wheel and its design, and the details of the swirling flow, this tech- Limited results are available which characterize the turbulence time and nique is being superseded by the impulse swirl meter shown in Fig. 8-12. A honeycomb flow straightener replaces the paddle wheel: it measures the total length scales in automobile engine flows. During the intake process, the integral length scale is of the order of the intake jet diameter, which is of the order of the torque exerted by the swirling flow. This torque equals the flux of angular valve lift (<10 mm in automobile-size engines). During compression the flow relaxes to the shape of the combusion chamber. The integral time scale at the end of compression decreases with increasing engine speed. It is of order 1 ms at engine speeds of about 1000 rev/min. The integral length scale at the end of Alr compression is believed to scale with the clearance height and varies little with engine speed. It decreases as the piston approaches TC to about 2 mm (0.2 x clearance height). The micro time scale at the end of compression is of order 0.1 ms at 1000 rev/min, and decreases as engine speed increases (again in automobile-size engine cylinders). Micro length scales are of order 1 mm at the end of compression and vary little with engine speed. Kolmogorov length scales at the end of compression are of order 10-2 mm. 8, 20, 21 8.3 SWIRL Swirl is usually defined as organized rotation of the charge about the cylinder axis. Swirl is created by bringing the intake flow into the cylinder with an initial angular momentum. While some decay in swirl due to friction occurs during the engine cycle, intake generated swirl usually persists through the compression, combustion, and expansion processes. In engine designs with bowl-in-piston combustion chambers, the rotational motion set up during intake is substantially modified during compression. Swirl is used in diesels and some stratified-charge FIGURE 8-12 Schematic of steady-flow impulse torque swirl engine concepts to promote more rapid mixing between the inducted air charge Restraining torque meter. 22 344 INTERNAL COMBUSTION ENGINE FUNDAMENTALS CHARGE MOTION WITHIN THE CYLINDER 345 momentum through the plane coinciding with the flow-straightener upstream the momentum produced under corresponding conditions of flow and valve lift face. For each of these approaches, a swirl coefficient is defined which essentially remains in the cylinder. Steady-state impulse torque-meter flow rig data can be used to estimate engine swirl in the following manner.23. Assuming that the port compares the flow's angular momentum with its axial momentum. For the and valve retain the same characteristics under the transient conditions of the paddle wheel, the swirl coefficient C, is defined by engine as on the steady-flow rig, the equivalent solid-body angular velocity @, at C = " B the end of the intake process is given by ( 8.24) Do 2 00 where @, is the paddle wheel angular velocity (=2zN ,, where N, is the rotation- al speed) and the bore B has been used as the characteristic dimension. The where 91 and 02 are crank angles at the start and end of the intake process and characteristic velocity, vo, is derived from the pressure drop across the valve the torque T and mass flow rate m are evaluated at the valve lift corresponding using an incompressible flow equation: to the local crank angle. Using Eq. (8.27) for T, Eq. (6.11) for m, assuming vo and [2( 00 - P 2 1 1 /2 , are constant throughout the intake process, and introducing volumetric effi- P (8.25) ciency n, based on intake manifold conditions via Eq. (2.27), it can be shown that or a compressible flow equation: R $ = - Rn. BL (4. CD)C. de| (4. Ca) del? (8.29) J 27 Do| 1 -(Pc)" (7 - 1 1/ x 7 7 1 /2 8.26) where A, CD is the effective valve open area at each crank angle. Note that the ((y - 1) po L crank angle in Eq. (8.29) should be in radians. Except for its (weak) dependence where the subscripts 0 and c refer to upstream stagnation and cylinder values, on ny, Eq. (8.29) gives R, independent of operating conditions directly from rig respectively. The difference between Eqs. (8.25) and (8.26) is usually small. With test results and engine geometry. the impulse torque meter, characteristic velocity and length scales must also be The relationship between steady-flow rig tests (which are extensively used introduced. Several swirl parameters have been defined.22, 23 The simplest is because of their simplicity) and actual engine swirl patterns is not fully under- stood. Steady-flow tests adequately describe the swirl generating characteristics of 8T. (8.27) the intake port and valve (at fixed valve lift) and are used extensively for this mvo B purpose. However, the swirling flow set up in the cylinder during intake can change significantly during compression. where T is the torque and m the air mass flow rate. The velocity vo, defined by Eq. (8.25) or Eq. (8.26), and the bore have again been used to obtain a dimension- less coefficient. Note that for solid-body rotation of the fluid within the cylinder 8.3.2 Swirl Generation during Induction at the paddle wheel speed @ ,, Eqs. (8.24) and (8.27) give identical swirl coeffi. Two general approaches are used to create swirl during the induction process. In cients. In practice, because the swirling flow is not solid-body rotation and one, the flow is discharged into the cylinder tangentially toward the cylinder wall, because the paddle wheel usually lags the flow due to slip, the impulse torque where it is deflected sideways and downward in a swirling motion. In the other, meter gives higher swirl coefficients.23 When swirl measurements are made in an the swirl is largely generated within the inlet port: the flow is forced to rotate operating engine, a swirl ratio is normally used to define the swirl. It is defined as about the valve axis before it enters the cylinder. The former type of motion is the angular velocity of a solid-body rotating flow w ,, which has equal angular achieved by forcing the flow distribution around the circumference of the inlet momentum to the actual flow, divided by the crankshaft angular rotational valve to be nonuniform, so that the inlet flow has a substantial net angular speed: momentum about the cylinder axis. The directed port and deflector wall port in Ws (8.28) Fig. 8-13 are two common ways of achieving this result. The directed port brings R, = ; 2n N the flow toward the valve opening in the desired tangential direction. Its passage Is straight, which due to other cylinder head requirements restricts the flow area During the induction stroke in an engine the flow and the valve open area, and results in a relatively low discharge coefficient. The deflector wall port uses and consequently the angular momentum flux into the cylinder, vary with crank the port inner side wall to force the flow preferentially through the outer periph- angle. Whereas in rig tests the flow and valve open area are fixed and the angular cry of the valve opening, in a tangential direction. Since only one wall is used to momentum passes down the cylinder continuously, in the engine intake process obtain a directional effect, the port areas are less restrictive. 346 INTERNAL COMBUSTION ENGINE FUNDAMENTALS CHARGE MOTION WITHIN THE CYLINDER 347 (a) (c) Shrouded Masked FIGURE 8-14 (b) (d) Shrouded inlet valve and masked cylinder head approaches for producing net in-cylinder angular momentum. FIGURE 8-13 Different types of swirl-generating inlet ports: (a) deflector wall; (b) directed; (c) shallow ramp helical; (d) steep ramp helical.24 ratios for these ports calculated from this rig data using Eqs. (8.27) and (8.29) are: 2.5 for the directed port, 2.9 for the shallow ramp helical, and 2.6 for the steep ramp helical. Vane swirl-meter swirl ratios were about 30 percent less. These Flow rotation about the cylinder axis can also be generated by masking off impulse-swirl-meter derived engine swirl ratios are within about 20 percent of the or shrouding part of the peripheral inlet valve open area, as shown in Fig. 8-14. solid-body rotation rate which has equal angular momentum to that of the cylin- Use is often made of a mask or shroud on the valve in research engines because der charge determined from tangential velocity measurements made within the changes can readily be made. In production engines, the added cost and weight, cylinder of an operating engine with the same port, at the end of the induction process 23 problems of distortion, the need to prevent valve rotation, and reduced volu- metric efficiency make masking the valve an unattractive approach. The more practical alternative of building a mask on the cylinder head around part of the 0.8 inlet valve periphery is used in production spark-ignition engines to generate 0.7 Plain directed swirl. It can easily be incorporated in the cylinder head casting process. Shallow ramp helical. The second broad approach is to generate swirl within the port, about the 3 0.6F Steep ramp helical valve axis, prior to the flow entering the cylinder. Two examples of such helical 0.5- ports are shown in Fig. 8-13. Usually, with helical ports, a higher flow discharge coefficient at equivalent levels of swirl is obtained, since the whole periphery of Kig swirl coefficient. 0.4 the valve open area can be fully utilized. A higher volumetric efficiency results. 0.3 Also, helical ports are less sensitive to position displacements, such as can occur in casting, since the swirl generated depends mainly on the port geometry above Maximum valve lift 0.2 the valve and not the position of the port relative to the cylinder axis. Figure 8-15 compares steady-state swirl-rig measurements of examples of 0.1 the ports in Fig. 8-13. The rig swirl number increases with increasing valve lift, FIGURE &-15 reflecting the increasing impact of the port shape and decreasing impact of the 0.1 0.2 0.3 Steady-state torque meter swirl measurements of flow restriction between the valve head and seat. Helical ports normally impart Valve lift Valve diameter directed, shallow ramp, and steep ramp helical ports more angular momentum at medium lifts than do directed ports.23, 25 The swirl as a function of inlet valve lift/diameter ratio.23 348 INTERNAL COMBUSTION ENGINE FUNDAMENTALS CHARGE MOTION WITHIN THE CYLINDER 349 Directed and deflector wall ports, and masked valve or head designs produce a tangential flow into the cylinder by increasing the flow resistance the full valve open area. The radial and axial velocities are essentially uniform through that part of the valve open area where flow is not desired. A highly around the valve periphery. The swirl velocity about the valve axis (anticlockwise nonuniform flow through the valve periphery results and the flow into the cylin- when viewed from above) for this helical port is relatively uniform and is about der has a substantial ug velocity component in the same direction about the cylin- half the magnitude of the radial and axial velocities. der axis. In contrast, helical ports produce the swirl in the port upstream of the The swirling air flow within the cylinder of an operating engine is not valve, and the velocity components v ,, and v, through the valve opening, and u. uniform. The velocities generated at the valve at each point in the induction about the valve axis are approximately uniform around the valve open area. process depend on the valve open area and piston velocity. The velocities are Figure 8-16 shows velocity data measured at the valve exit plane in steady-flow highest during the first half of the intake process as indicated in Fig. 6-15. Thus, rig tests with examples of these two types of port. The valve and cylinder wall the swirl velocities generated during this portion of the induction stroke are locations are shown. In Fig. 8-16a, the deflector wall of the tangentially oriented higher than the swirl generated during the latter half of the stroke: there is swirl port effectively prevents any significant flow around half the valve periphery. In stratification. Also, the flow pattern close to the cylinder head during induction is contrast, in Fig. 8-16b with the helical port, the air flows into the cylinder around comparatively disorganized, and not usually close to a solid-body rotation. It consists of a system of vortices, created by the high-velocity tangential or spiral- ing intake jet. Further down the cylinder, the flow pattern is closer to solid-body (a) (b) rotation with the swirl velocity increasing with increasing radius.23. 24 This more ordered flow directly above the piston produces higher swirl velocities in that region of the cylinder. As the piston velocity decreases during intake, the swirl pattern redistributes, with swirl speeds close to the piston decreasing and swirl speeds in the center of the cylinder increasing.27 Note that the axis of rotation of 13 the in-cylinder gases may not exactly coincide with the cylinder axis. 10 8.3.3 Swirl Modification within the Cylinder ,8 The angular momentum of the air which enters the cylinder at each crank angle Swirl during induction decays throughout the rest of the intake process and during the - Axial compression process due to friction at the walls and turbulent dissipation within Radial the fluid. Typically one-quarter to one-third of the initial moment of momentum = 50 m/s about the cylinder axis will be lost by top-center at the end of compression. However, swirl velocities in the charge can be substantially increased during compression by suitable design of the combustion chamber. In many designs of 3 direct-injection diesel, air swirl is used to obtain much more rapid mixing between the fuel injected into the cylinder and the air than would occur in the 10 absence of swirl. The tangential velocity of the swirling air flow set up inside the cylinder during induction is substantially increased by forcing most of the air into a compact bowl-in-piston combustion chamber, usually centered on the cylinder axis, as the piston approaches its top-center position. Neglecting the effects of friction, angular momentum is conserved, and as the moment of inertia of the air is decreased its angular velocity must increase. However, the total angular momentum of the charge within the cylinder does decay due to friction at the chamber walls. The angular momentum of the FIGURE 8-16 cylinder charge I, changes with time according to the moment of momentum Swirl, axial, and radial velocities measured 2 mm from cylinder head around the valve circumference conservation equation: for (a) tangential deflector-wall port and (b) helical port; magnitude of velocity is given by the distance along a radial line (from valve axis), from valve outline to the respective curve scaled by the reference length (examples of radial velocity indicated by two arrows); valve lift = 12.8 mm. 26, 27 dt di c = J, - T, (8.30) 350 INTERNAL COMBUSTION ENGINE FUNDAMENTALS CHARGE MOTION WITHIN THE CYLINDER 351 where J, is the flux of angular momentum into the cylinder and Ty is the torque ungential velocity 19 at the wall varies with radius, the shear stress should be due to wall friction. At each point in the intake process J, is given by evaluated at each radius and integrated over the surface: e.g ., 29 J = JAO' pru.v . dA . (r) = C1tp[ve(r)]2 Re- 0.2 (8.34) (8.31) with Re = PUT)r where dA, is an element of the valve open area, as defined in Fig. 8-17. While the angular momentum entering the cylinder during the intake process is where C1 is an empirical constant (~0.055). An alternative approximate Tot = prus . dA , dt approach is to evaluate these components of the wall shear stress at the mean JEivo radius. 28 Next, consider the effects on swirl of radially inward displacement of the air the actual angular momentum within the cylinder at the end of induction will be less, due to wall friction during the intake process. Friction continues through the charge during compression. The most common example of this phenomenon occurs with the bowl-in-piston combustion chamber design of medium- and high- compression process so the total charge angular momentum at the end of com- speed direct-injection diesels (see Sec. 10.2.1). However, in spark-ignition engines pression is further reduced. There is friction on the cylinder wall, cylinder head, and piston crown where swirl is used to increase the burning rate, the shape of the combustion chamber close to top-center can also force radially inward motion of the charge. (including any combustion chamber within the crown). This friction can be esti- For a given swirling in-cylinder flow at the end of induction and neglecting the mated with sufficient accuracy using friction formulas developed for flow over a effects of friction, as the moment of inertia of the air about the cylinder axis is flat plate, with suitable definition of characteristic length and velocity scales. Fric- decreased the air's angular velocity must increase to conserve angular momen- tion on the cylinder wall can be estimated from the wall shear stress: tum. For example, for solid-body rotation of the cylinder air charge of mass me, the initial angular momentum Ie, ; and solid-body rotation @, ; are related at ( 8.32) bottom-center by where w, is the equivalent solid-body swirl. The friction factor CF is given by the flat plate formula: where I, is the moment of inertia of the charge about the cylinder axis. For a CF = 0.0371(Reg) - 0.2 (8.33) disc-shaped combustion chamber, I = m. B2/8 and is constant. For a bowl-in- piston combustion chamber, where 1 is an empirical constant introduced to allow for differences between the flat plate and cylinder wall (1 ~ 1.5)28 and Reg is the equivalent of the flat plate I =- m. B2 [(z/hg) + (DB/ B)ª] Reynolds number [Reg = p(Bw,/2)(1B)/u]. Friction on the cylindrical walls of a 8 [(z/hB) + (DB/B)2] (8.35) piston cup or bowl can be obtained from the above expressions with DB, the where DB and h, are the diameter and depth of the bowl, respectively, and z is the bowl diameter, replacing the bore. distance of the piston crown from the cylinder head. At TC crank position, z ~ 0 Friction on the cylinder head, piston crown, and piston bowl floor can be and I ~ me Da/8. At the end of induction, I ~me B2/8. Thus, in the absence of estimated from expressions similar to Eqs. (8.32) and (8.33). However, since the friction w, would increase by (B/DB)2, usually a factor of about 4. In an operating engine with this bowl-in-piston chamber design, the observed increase in swirl in the bowl is less; it is usually about a factor of 2 to 3.23. 25 This is because of wall friction, dissipation in the fluid due to turbulence and velocity gradients, and the fact that a fraction of the fluid remains in the clearance height above the piston crown. The loss in angular momentum due to these effects will vary with geometric details, initial swirl flow pattern, and engine day speed. Swirl velocity distributions in the cylinder at the end of induction show the V1 langential velocity increasing with radius, except close to the cylinder wall where FIGURE 8-17 Definition of symbols in equation for angular friction causes the velocity to decrease. While the velocity distribution is not that momentum flux into the cylinder [Eq. (8.31)]. of a solid-body rotation, depending on port design and operating conditions it is 352 INTERNAL COMBUSTION ENGINE FUNDAMENTALS CHARGE MOTION WITHIN THE CYLINDER 353 often close to solid-body rotation.23, 25 Departures from the solid-body velocity 110 mm O 20 Lo0 distribution are greater at higher engine speeds, suggesting that the flow pattern 66 mm -10º in the cylinder at this point in the cycle is still developing with time.23, 30 o o O In the O O 10 absence of radially inward gas displacement during compression, the flow pattern Ve, m/s -20º O 18 mm O continues to develop toward a solid-body distribution throughout the compres sion stroke.25. Swirl ratios of 3 to 5 at top-center can be achieved with the porta 20 shown in Fig. 8-13, with flat-topped pistons (i.e ., in the absence of any swirl o C amplification during compression). 23, 25 O 20 -20º O With combustion chambers where the chamber radius is less than the cylin- o oo 00 der bore, such as the bowl in piston, the tangential velocity distribution with - Ve, m/s -- 10- 10 O - 50 Doo radius will change during compression. Even if the solid-body rotation assump- --- -300 Liner Bow! tion is reasonable at the end of induction, the profile will distort as gas moves into the piston bowl. Neglecting the effects of friction, the angular momentum of 0 20 40 50 20 40 60 each fluid element will remain constant as it moves radially inward. Thus the Radius, mm Radius, mm increase in tangential velocity of cach fluid element as it moves radially inward is FIGURE &-18 proportional to the change in the reciprocal of its radius. Measurements of the Velocity measurements as a function of radius across the combustion chamber of a firing, bowl-in- swirl velocity distribution within the cylinder of bowl-in-piston engine designs piston, direct-injection diesel engine. Schematic shows the chamber geometry. Solid lines are calcu- support this description. The rate of displacement of gas into the bowl depends lations based on the assumption of constant angular momentum for fluid elements as they move on the bowl volume VB, cylinder volume V, and piston speed Sp, at that particu- radially inward.31 lar piston position: dmz - the effects of wall friction (enhanced by the higher gas velocities in the bowl). dt Sometimes the bowl axis is offset from the cylinder axis and some additional loss The gas velocity into the bowl will therefore increase rapidly toward the end of in swirl amplification results.25 the compression stroke and reach a maximum just before TC (see Sec. 8.4 where The effect of swirl generation during induction on velocity fluctuations in this radial "squish" motion is discussed more fully). Thus, there is a rapid the combustion chamber at the end of compression has been examined.32 The increase in v, in the bowl as the crank angle approaches TC. The lower layers of turbulence intensity with swirl was higher than without swirl (with the same chamber geometry). Integral scales of the turbulence were smaller with swirl than the bowl rotate slower than the upper layers because that gas entered the bowl earlier in the compression process. 23, 25 without. Cyclic fluctuations in the mean velocity are, apparently, reduced by Velocity measurements illustrating the development of this radial distribu- swirl. Also, some studies show that the ensemble-averaged fluctuation intensity goes down when swirl is introduced.18 There is evidence that swirl makes the tion in tangential velocity are shown in Fig. 8-18. These measurements were turbulence intensity more homogeneous.30 made by analysing the motion of burning carbon particles in the cylinder of an operating diesel engine from movies of the combustion process. The figure shows the engine geometry and the data compared with a model based on gas displace- 8.4 SQUISH ment and conservation of angular momentum in each element of the charge as it is displaced inward. Different swirl velocity profiles exist within and outside the Squish is the name given to the radially inward or transverse gas motion that bowl as the piston approaches TC. Swirl velocities within the bowl increase a occurs toward the end of the compression stroke when a portion of the piston TC is approached, roughly as predicted by the ideal model. Outside the bowl, the face and cylinder head approach each other closely. Figure 8-19 shows how gas is swirl velocity decreases with increasing radius due to the combined effects of thereby displaced into the combustion chamber. Figure 8-19a shows a typical friction and inward gas displacement as the clearance height decreases. wedge-shaped SI engine combustion chamber and Fig. 8-19b shows a bowl-in- Swirl ratios in bowl-in-piston engine designs of up to about 15 can be piston diesel combustion chamber. The amount of squish is often defined by the achieved with DE ~ 0.5B, at top-center. Amplification factors relative to fiat- percentage squish area: i.e ., the percentage of the piston area, TB2/4, which closely topped piston swirl are typically about 2.5 to 3, some 30 percent lower than the approaches the cylinder head (the shaded areas in Fig. 8-19). Squish-generated ideal factor of (B/DB)2 given by Eq. (8.35) as z -+ 0. This difference is due to the gas motion results from using a compact combustion chamber geometry. mass remaining within the clearance height which does not enter the bowl, and A theoretical squish velocity can be calculated from the instantaneous dis- 354 INTERNAL COMBUSTION ENGINE FUNDAMENTALS CHARGE MOTION WITHIN THE CYLINDER 355 DB I B (a) (b) (a) (b) FIGURE 8-19 Schematics of how piston motion generates squish: (a) wedge-shaped SI engine combustion chamber; FIGURE 8-20 (b) bowl-in-piston direct-injection diesel combustion chamber. (a) Schematic of axisymmetric bowl-in-piston chamber for Eq. (8.36). (b) Schematic of wedge chamber with transverse squish for Eq. (8.37). placement of gas across the inner edge of the squish region (across the dashed lines in the drawings in Fig. 8-20a and b), required to satisfy mass conservation. Ignoring the effects of gas dynamics (nonuniform pressure), friction, leakage past The theoretical squish velocity for a bowl-in-piston engine normalized by the piston rings, and heat transfer, expressions for the squish velocity are: the mean piston speed S, is shown in Fig. 8-21 for different ratios of Dp/B and clearance heights c. The maximum squish velocity occurs at about 10º before TC. 1. Bowl-in-piston chamber (Fig. 8-20a):33 After TC, Usq is negative; a reverse squish motion occurs as gas flows out of the bowl into the clearance height region. Under motored conditions this is equal to VB (8.36) the forward motion. A c 2 + VB These models omit the effects of gas inertia, friction, gas leakage past the piston rings, and heat transfer. Gas inertia and friction effects have been shown to where V, is the volume of the piston bowl, A, is the cross-sectional area of the be small. The effects of gas leakage past the piston rings and of heat transfer are cylinder (TB2/4), S, is the instantaneous piston speed [Eq. (2.11)], and z is the more significant. The squish velocity decrement Avg due to leakage is proportion- distance between the piston crown top and the cylinder head (z = c + Z. al to the mean piston speed and a dimensionless leakage number: where Z = 1 + a - s; see Fig. 2-1). 2. Simple wedge chamber (Fig. 8-20b):34 NL = AEL VYRTINC NVa (8.38) Usq A, Z + c ) (8.37) S. b(Z + c) C + Z ) where AE, L is the effective leakage area and Tivc is the temperature of the cylin- der gases at inlet valve closing. Leakage was modeled as a choked flow through where A, is the squish area, b is the width of the squish region, and C is the effective leakage area. Values of Av1/1,q are shown in Fig. 8-22. The effect of Z/(re - 1) evaluated at the end of induction. leakage on Usq is small for normal gas leakage rates. A decrement on squish 356 INTERNAL COMBUSTION ENGINE FUNDAMENTALS CHARGE MOTION WITHIN THE CYLINDER 357 o DB = 0.45, c = 1.0 mm) = 0.011 B 0.55, 1.0 Measured 0.027 X 0.55, 2.5 30- 6 No losses DE = 0.35 --- Includes AVL, AVH 0.45 -B- 20+ 4- Flies Squash vekaity , mis 0.55 Vsq 10- 0.65 DB 0 -60 -50 -40 -30 -20 -10 TO -30 -20 -10 TO Crank angle, deg Crank angle, deg FIGURE 8-21 FIGURE 8-23 Theoretical squish velocity divided by mean piston speed for bowl-in-piston chambers, for different Comparison of measured squish velocities in bowl-in-piston combustion chambers, with different D /B and c/L (clearance height/stroke). B/L = 0.914, V /V = 0.056, connecting rod length/crank bowl diameter/bore ratios and clearance heights, to calculated ideal squish velocity (solid lines) and radius == 3.76.35 calculations corrected for leakage and heat transfer (dashed lines). Bore = 85 mm, stroke = 93 mm, 1500 rev/min.35 velocity due to heat transfer, Avg, has also been derived, using standard engine heat-transfer correlations (see Sec. 12.4). Values of AvH/vsq are also shown in 0.06 Fig. 8-22. Again the effects are small in the region of maximum squish, though DB 0.45 B NL = 0.015 they become more important as the squish velocity decreases from its maximum 0.55 = 0.011 0.04 value as the piston approaches TC. Y = 1.4 0.010 Velocity measurements in engines provide good support for the above theory. The ideal theory adequately predicts the dependence on engine speed.36 0.02 | With appropriate corrections for leakage and heat-transfer effects, the above 0.005 theory predicts the effects of the bowl diameter/bore ratio and clearance height on squish velocity (see Fig. 8-23). The change in direction of the radial motion as the piston moves through TC has been demonstrated under motored engine con- 0.15 ditions. Under firing conditions, the combustion generated gas expansion in the C = 0.011 open portion of the combustion chamber substantially increases the magnitude of 0.027 the reverse squish motion after TC.37 ).10F Y = 1.35 1500 rev/min AVH R = 0 - 0.45 8.5 PRECHAMBER ENGINE FLOWS 0.05 Small high-speed diesel engines use an auxiliary combustion chamber, or pre- chamber, to achieve adequate fuel-air mixing rates. The prechamber is connected 0 -30 .20 -10 TC to the main combustion chamber above the piston via a nozzle, passageway, or Crank angle, deg one or more orifices. Flow of air through this restriction into the prechamber FIGURE 8-22 during the compression process sets up high velocities in the prechamber at the Values of squish velocity decrement due to leakage Av, and heat transfer Avg, normalized by the time the fuel-injection process commences. This results in the required high fuel- ideal squish velocity, as a function of crank angle.35 air mixing rates. Figures 1-21 and 10-2 show examples of these prechamber or 358 INTERNAL COMBUSTION ENGINE FUNDAMENTALS CHARGE MOTION WITHIN THE CYLINDER 359 indirect-injection diesels. The two most common designs of auxiliary chamber 30 are: the swirl chamber (Fig. 10-2a), where the flow through the passageway enters Tc = 23, R = 3.5 the chamber tangentially producing rapid rotation within the chamber, and the 25 Vp = 0.5 prechamber (Fig. 10-2b) with one or more connecting orifices designed to AT 20 produce a highly turbulent flow but no ordered motion within the chamber. Aux- - = 0.008 xB2/4 iliary chambers are sometimes used in spark-ignition engines. The torch-ignition (mN) ) 3 15 les three-valve stratified-charge engine (Fig. 1-27) is one such concept. The precham- ber is used to create a rich mixture in the vicinity of the spark plug to promote 10 rapid flame development. An alternative concept uses the prechamber around the spark plug to generate turbulence to enhance the early stages of combustion, but has no mixture stratification. The most critical phase of flow into the prechamber occurs towards the end 180 150 120 90 50 30 BC rc of compression. While this flow is driven by a pressure difference between the Crank angle, deg main chamber above the piston and the auxiliary chamber, this pressure differ- FIGURE &-24 ence is small, and the mass flow rate and velocity at the nozzle, orifice, or pas- Velocity and mass flow rate at the prechamber nozzle throat, during compression, for a typical small sageway can be estimated using a simple gas displacement model. Assuming that swirl-prechamber automotive diesel. the gas density throughout the cylinder is uniform (an adequate assumption toward the end of compression-the most critical period), the mass in the pre- chamber mp is given by me(Vp/V), where mc is the cylinder mass, V the cylinder chamber diesel usually starts just before TC, and the pressure in the prechamber volume, and Vp the prechamber volume. The mass flow rate through the throat of then rises significantly above the main chamber pressure. The outflow from the the restriction is, therefore, prechamber is then governed by the development of the combustion process, and the above simple gas displacement model no longer describes the flow. This com- dmp 2 = me Ve dV. bustion generated prechamber gas motion is discussed in Sec. 14.4.4. dt V2 dt (8.39) In prechamber stratified-charge engines, the flow of gas into the precham- Using the relations dV/dt = - (TB2/4)S, where S, is the instantaneous piston ber during compression is critical to the creation of an appropriate mixture in the prechamber at the crank angle when the mixture is ignited. In the concept shown speed, V/a = B2L/4, and S, = 2NL, Eq. (8.39) can be written as in Fig. 1-27, a very rich fuel-air mixture is fed directly to the prechamber during im = 2(r )(V) (S) (*) intake via the prechamber intake valve, while a lean mixture is fed to the main (8.40) chamber via the main intake valve. During compression, the flow into the pre- chamber reduces the prechamber equivalence ratio to a close-to-stoichiometric where Ve is the clearance volume, Sp/S, is given by Eq. (2.11), and V/Ve is given by value at the time of ignition. Figure 8-25 shows a gas displacement calculation of Eq. (2.6). The gas velocity at the throat ur can be obtained from m via the rela- this process and relevant data; the prechamber equivalence ratio, initially greater tion pvT AT = m, the density p = mc/V, and Eq. (8.40): (8.41) 4.0 40 where AT is the effective cross-sectional area of the throat. The variation of m/(m. N) and vr/S, with crank angle during the compression process for values of 3.0 30 Te, Vp/Ve, and AT/(B2/4) typical of a swirl prechamber diesel are shown in Fig. 8-24. The velocity reaches its peak value about 20º before TC: very high gas 2.0 Mass flow rate through orifice, ug/deg CA Equivalence ratio 20 velocities, an order of magnitude or more larger than the mean piston speed, can be achieved depending on the relative effective throat area. Note that as the 1.0 10 FIGURE 8-25 piston approaches TC, first the nozzle velocity and then the mass flow rate Effect of gas flow into the prechamber during com- · Measured O pression on the prechamber equivalence ratio in a decrease to zero. After TC, in the absence of combustion, an equivalent flow in O BC 30 60 90. 120 150 TC three-valve prechamber stratified-charge engine. the reverse direction out of prechamber would occur. Combustion in the pre- Crank angle, deg Calculations based on gas displacement model.38 360 INTERNAL COMBUSTION ENGINE FUNDAMENTALS CHARGE MOTION WITHIN THE CYLINDER 361 The largest crevices are the volumes between the piston, piston rings, and cylinder wall. Some gas flows out of these regions into the crankcase; it is called blowby. Other crevice volumes in production engines are the threads around the spark plug, the space around the plug center electrode, the gap around the fuel injector, crevices between the intake and exhaust valve heads and cylinder head, and the head gasket cutout. Table 8.1 shows the size and relative importance of 128º BTC 86º BTC 51º BTC Scale: ~ 0.5 m/s Scale: - 1 m/s Scale: - 1.5 m/s these crevice regions in one cylinder of a production V-6 spark-ignition engine determined from measurements of cold-engine components. Total crevice volume is a few percent of the clearance volume, and the piston and ring crevices are the dominant contributors. When the engine is warmed up, dimensions and crevice volumes will change. The important crevice processes occurring during the engine cycle are the following. As the cylinder pressure rises during compression, unburned mixture or air is forced into each crevice region. Since these volumes are thin they have a large surface/volume ratio; the gas flowing into the crevice cools by heat transfer 24º BTC 1.5º BTC to close to the wall temperature. During combustion while the pressure continues Scale: - 2 m/s Scale: ~ 1.5 m/s to rise, unburned mixture or air, depending on engine type, continues to flow into FIGURE 8-26 these crevice volumes. After flame arrival at the crevice entrance, burned gases Calculations of developing flow field in (two-dimensional) swirl prechamber during compression will flow into each crevice until the cylinder pressure starts to decrease. Once the process. Lines are instantaneous flow streamlines, analogous to streak photographs of flow field.41 crevice gas pressure is higher than the cylinder pressure, gas flows back from each crevice into the cylinder. The volumes between the piston, piston rings, and cylinder wall are shown than 3, is leaned out to unity as mass flows through the orifice into the precham- schematically in Fig. 8-27. These crevices consist of a series of volumes ber (whose volume is 8.75 percent of the clearance volume).38 Charts for estimat- (numbered 1 to 5) connected by flow restrictions such as the ring side clearance ing the final equivalence ratio, based on gas displacement, for this prechamber and ring gap. The geometry changes as each ring moves up and down in its ring concept are available.39 groove, sealing either at the top or bottom ring surface. The gas flow, pressure The velocity field set up inside the prechamber during compression is distribution, and ring motion are therefore coupled. Figure 8-28 illustrates this strongly dependent on the details of the nozzle and prechamber geometry. Velo- behavior: pressure distributions, ring motion, and mass flow of gas into and out cities vary linearly with mean piston speed.4º In swirl prechambers, the nozzle flow sets up a vortex within the chamber. Figure 8-26 shows calculations of this developing flow field; instantaneous flow streamlines have been drawn in, with TABLE 8.1 the length of the streamlines indicating how the particles of fluid move relative to V-6 engine crevice data+42 each other.41 The velocities increase with increasing crank angle as the compres- cm3 % sion process proceeds, and reach a maximum at about 20º before TC. Then, as the piston approaches TC and the flow through the passageway decreases to Displaced volume per cylinder 632 zero, the vortex in the swirl chamber expands to fill the entire chamber and mean Clearance volume per cylinder 89 100 velocities decay. Very high swirl rates can be achieved just before TC: local swirl Volume above first ring 0.93 1.05 ratios of up to 60 at intermediate radii and up to 20 at the outer radius have been Volume behind first ring 0.47 0.52. measured. These high swirl rates produce large centrifugal accelerations. Volume between rings 0.68 0.77 Volume behind second ring 0.47 0.52 Total ring crevice volume 2.55 2.9 Spark plug thread crevice 0.25 0.28 8.6 CREVICE FLOWS AND BLOWBY Head gasket crevice 0.3 0.34 The engine combustion chamber is connected to several small volumes usually Total crevice volume 3.1 3.5 called crevices because of their narrow entrances. Gas flows into and out of these + Determined for cold engine. volumes during the engine operating cycle as the cylinder pressure changes. 362 INTERNAL COMBUSTION ENGINE FUNDAMENTALS CHARGE MOTION WITHIN THE CYLINDER 363 40 10 Combustion chamber 8 a Plane a 30 b 1, 2 6 c Top land crevice Mass in, % d 20 4 Pressure, atm b Ring side clearance 2 b 2 -2, 3, 4 0 8 Region behind rings - 180 -60 60 180 300 420 540 a Crank angle, deg 6 Top ring gap, 8 4 (a) Mass out, % 4 5 2 b Oil ring Groove upper surface Top ring Relative position of the rings Groove lower surface Mass trapped, % Groove upper surface .4 FIGURE 8-27 Groove lower surface Second ring Schematic of piston and ring assembly in automotive spark-ignition engine. 2 -180 -60 60 180 300 420 540 0 -120 0 120 240 360 480 of the regions defined by planes a, b, c, d, and through the ring gap g are plotted Crank angle, deg Crank angle, deg versus crank angle through compression and expansion. These results come from (b) (c) an analysis of these regions as volumes connected by passageways, with a pre- FIGURE 8-28 scribed cylinder pressure versus crank angle profile coupled with a dynamic (a) Pressures in the combustion chamber (1), in region behind top ring (2), in region between rings (3), model for ring motion, and assuming that the gas temperature equals the wall and behind second ring (4); (b) relative position of top and second rings; (c) percentage of total cylinder mass that flows into and out of the different crevice regions across planes a, b, c, and d and temperature.42 During compression and combustion, the rings are forced to the through the ring gap g in Fig. 8-27, and the percentage of mass trapped beneath these planes, as a groove lower surfaces and mass flows into all the volumes in this total crevice function of crank angle. Automotive spark-ignition engine at wide-open throttle and 2000 rev/min.42 region. The pressure above and behind the first ring is essentially the same as the cylinder pressure; there is a substantial pressure drop across each ring, however. the top ring gap later in the expansion process when the pressure difference Once the cylinder pressure starts to decrease (after 15º ATC) gas flows out of across the ring changes sign have been observed. Figure 8-29 shows these flows regions 1 and 2 in Fig. 8-27 into the cylinder, but continues to flow into regions 3, with explanatory schematics. 4, and 5 until the pressure in the cylinder falls below the pressure beneath the top Blowby is defined as the gas that flows from the combustion chamber past ring. The top ring then shifts to seal with the upper grove surface and gas flows out of regions 2, 3, and 4 (which now have the same pressure), both into the the piston rings and into the crankcase. It is forced through any leakage paths afforded by the piston-bore-ring assembly in response to combustion chamber cylinder and as blowby into the crackcase. Some 5 to 10 percent of the total pressure. If there is good contact between the compression rings and the bore, cylinder charge is trapped in these regions at the time of peak cylinder pressure. and the rings and the bottom of the grooves, then the only leakage path of Most of this gas returns to the cylinder; about 1 percent goes to the crankcase as blowby. The gas flow back into the cylinder continues throughout the expansion consequence is the ring gap. Blowby of gases from the cylinder to the crankcase removes gas from these crevice regions and thereby prevents some of the crevice process. In spark-ignition engines this phenomenon is a major contributor to gases from returning to the cylinder. Crankcase blowby gases used to be vented unburned hydrocarbon emissions (see Sec. 11.4.3). In all engines it results in a directly to the atmosphere and constituted a significant source of HC emissions. loss of power and efficiency. The crankcase is now vented to the engine intake system and the blowby gases There is substantial experimental evidence to support the above description are recycled. Blowby at a given speed and load is controlled primarily by the of flow in the piston ring crevice region. In a special square-cross-section flow greatest flow resistance in the flow path between the cylinder and the crankcase. visualization engine, both the low-velocity gas expansion out of the volume This is the smallest of the compression ring ring-gap areas. Figure 8-30 shows above the first ring after the time of peak pressure and the jet-like flows through how measured blowby flow rates increase linearly with the smallest gap area.43 CHARGE MOTION WITHIN THE CYLINDER INTERNAL COMBUSTION ENGINE FUNDAMENTALS 365 364 Calculations of blowby based on the model described earlier are in good agree- ment. 42 Extrapolation back to the zero gap area gives nearly zero blowby. Note, however, that if the bore finish is rough, or if the rings do not contact the bore all around, or if the compression rings lift off the bottom of the groove, this linear 40º ATC relationship may no longer hold. 8.7 FLOWS GENERATED BY PISTON-CYLINDER WALL INTERACTION Expanding flow out of Jet through inner inner piston top-land piston ring gap Because a boundary layer exists on the cylinder wall, the motion of the piston generates unusual flow patterns in the corner formed by the cylinder wall and the piston face. When the piston is moving away from top-center a sink-type flow occurs. When the piston moves toward top-center a vortex flow is generated. Figure 8-31 shows schematics of these flows (in a coordinate frame with the piston face at rest). The vortex flow has been studied because of its effect on gas motion at the time of ignition and because it has been suggested as a mechanism 55º ATC for removing hydrocarbons off the cylinder wall during the exhaust stroke (see Sec. 11.4.3). The vortex flow has been studied in cylinders with water as the fluid over the range of Reynolds numbers typical of engine operation.44,45 Laminar, tran- sition, and turbulent flow regimes have been identified. It has been shown that a (b) (a) quasi-steady flow assumption is valid and that FIGURE 8-29 Schlieren photographs of the flow out of the piston-cylinder wall crevices during the expansion Ar = ( UM L) stroke. A production piston was inserted into the square cross-section piston of the visualization engine. Gas flows at low velocity out of the crevice entrance all around the production piston circum- where Ay is the vortex area (area inside the dashed line in Fig. 8-31), L is the ference once the cylinder pressure starts decreasing early in the expansion stroke. Gas flows out of the ring gap as a jet once the pressure above the ring falls below the pressure beneath the ring.42 stroke, uw is the wall velocity in piston stationary coordinates (Vw = S, in the engine), v is the kinematic viscosity, and (vw L/v) is a Reynolds number. - Shear area Vortex area 1200 rev/min 0 , - Experimental Pin = 0.6 atm (Wentworth) · Calculations Boundary layer Entrained area 0.3H Boundary layer 0.21 Blowby flow, cfm PA 0.1 DO +Production range- (a) (b) FIGURE 8-30 O 2 3 5 6 7 8 9 10 x 10-4 FIGURE 8-31 Measured blowby for one cylinder of an in2 automobile spark-ignition engine as a func- Schematics of the flow pattern set up in the piston face-cylinder wall corner, in piston-stationary coordinates, due to the boundary layer on the cylinder wall. Piston crown on left; cylinder wall at o 1 2 3 5 6 × 10-3 tion of the smallest ring gap area, compared with blowby calculations based on flow bottom. (a) Sink flow set up during intake and expansion; (b) vortex flow set up during compression cm2 and exhaust.44 Arrow shows cylinder wall velocity relative to piston. Smaller ring gap area model described in text.42. 43 366 INTERNAL COMBUSTION ENGINE FUNDAMENTALS CHARGE MOTION WITHIN THE CYLINDER 367 For the laminar flow regime, a good assumption is that Ay is proportional to the shear area in the vortex (shown cross-hatched), which equals the boundary-layer area; this can be estimated from boundary-layer theory. In the turbulent flow regime, an entrainment theory was used, which assumed that the rate of change of vortex area was proportional to the product of the exposed perimeter of the vortex and the velocity difference between the vortex and the stationary fluid (~vw). The relevant relationships are: - 1/2 20º BTC For (vw L/v) < 2 × 104: (8.42a) 60º BTC For (vw L/v) 2 2 x 104: 72 = 0.006 (8.42b) Figure 8-32 shows these two theories correlated against hydraulic analog data. These theories are for constant values of v. During compression, v decreases substantially as the gas temperature and pressure increase (v decreases by a factor of 4 for a compression ratio of 8). This will decrease the size of the vortex until the turbulent regime is reached. During the exhaust stroke following blowdown, v FIGURE 8-33 will remain approximately constant as the pressure and temperature do not Schlieren photographs of in-cylinder flow during later stages of exhaust stroke. Growing vortex in the change significantly. Typical parameter values at 1500 rev/min are: S, = 5 m/s, piston face-cylinder wall corner and turbulent outflow toward the valve are apparent at 60º BTC. At L = 0.1 m; average values of v are 1.2 x 10-5 and 1.4 x 10-4 m2/s for compres- 20º BTC, the vortex has grown to of order 0.2B diameter.42 sion and exhaust stroke, respectively. Hence a Reynolds number for the compres- sion stroke is 4 x 104, Ay/L2 ~ 0.006, and the vortex diameter dy ~ 0.09L. For This vortex flow has been observed in an operating engine. Figure 8-33 the exhaust stroke, the Reynolds number is 4 x 103, Ay/L2 ~ 0.015, and dy ~ shows schlieren photographs taken during the exhaust stroke in a special square- 0.14L. Thus the vortex dimensions at the end of the upward stroke of the piston cross-section flow visualization spark-ignition engine. The accompanying sche- are comparable to the engine clearance height. matic identifies the vortex structure which is visible in the photo because the cool boundary-layer gas is being scraped off the cylinder wall by the upward-moving piston and "rolled up." The vortex diameter as the piston approaches TC is about 20 percent of the bore. 0.06 PROBLEMS Stable Transition Turbulent 8.1. (a) Estimate the ratio of the maximum gas velocity in the center of the hollow cone ODD inlet jet to the mean piston speed from the data in Fig. 8-1. (b) Compare this ratio with the ratio of inlet valve pseudo flow velocity determined 21 0.01 from Fig. 6-15 to the mean piston speed at the same crank angle. The engine is that of Fig. 1-4. TTTTTT (c) Are the engine velocity data in (a) consistent with the velocity calculated from the simple piston displacement model of (b)? Explain. 8.2. Given the relationship between turbulence intensity and mean piston speed [Eq. (8.23)] and that the turbulence integral scale is ~ 0.2 x clearance height, use Eqs. 0.002 LLLLL 104 (8.14) and (8.15) to estimate the following quantities for a spark-ignition engine with 103 105 Reynolds number " bore = stroke = 86 mm, r = 9, at 1000 and 5000 rev/min and wide-open throttle: (a) Mean and maximum piston speed, maximum gas velocity through the inlet valve FIGURE 8-32 (see Prob. 8.1) Ratio of area of vortex in piston face-cylinder wall corner to square of stroke, as a function of (b) Turbulence intensity, integral length scale, micro length scale, and Kolmogorov Reynolds number based on piston velocity, for piston moving toward the cylinder head.44 length scale, all at TC 368 INTERNAL COMBUSTION ENGINE FUNDAMENTALS CHARGE MOTION WITHIN THE CYLINDER 369 8 .3. The swirl ratio at the end of induction at 2000 rev/min in a direct-injection diesel 7. Rask, R. B.: "Laser Doppler Anemometer Measurements of Mean Velocity and Turbulence in engine of bore = stroke = 100 mm is 4.0. What is the average tangential velocity Internal Combustion Engines," ICALEO '84 Conference Proceedings, vols. 45 and 47, Inspection, (evaluated at the inlet valve-axis radial location) required to give this swirl ratio? Measurement and Control and Laser Diagnostics and Photochemistry, Laser Institute of America, What is the ratio of this velocity to the mean piston speed and to the mean flow Boston, November 1984. velocity through the inlet valve estimated from the average valve open area and 8. Tabaczynski, R. J.: "Turbulence and Turbulent Combustion in Spark-Ignition Engines," Prog. Energy Combust. Sci ., vol. 2, pp. 143-165, 1976. open time? 9. Witze, P. O.: "A Critical Comparison of Hot-Wire Anemometry and Laser Doppler Velocimetry 8.4. (a) Derive a relationship for the depth (or height) hp of a disc-shaped bowl-in-piston for I.C. Engine Applications," SAE paper 800132, SAE Trans ., vol. 89, 1980. direct-injection diesel engine combustion chamber in terms of compression ratio 10. Witze, P. O ., Martin, J. K ., and Borgnakke, C.: "Conditionally-Sampled Velocity and Turbulence re, bore B, stroke L ,. bowl diameter DB, and top-center cylinder-head to piston- Measurements in a Spark Ignition Engine," Combust. Sci. Technol ., vol. 36, pp. 301-317, 1984. crown clearance c. For B = L = 100 mm, re = 16, DE = 0.5B, c = 1 mm find the 11. Rask, R. B.: "Comparison of Window, Smoothed-Ensemble, and Cycle-by-Cycle Data Reduction fraction of the air charge within the bowl at TC. Techniques for Laser Doppler Anemometer Measurements of In-Cylinder Velocity," in T. Morel, (b) If the swirl ratio at the end of induction at 2500 rev/min is 3 find the swirl ratio R. P. Lohmann, and J. M. Rackley (eds.), Fluid Mechanics of Combustion Systems, pp. 11-20, and average angular velocity in the bowl-in-piston chamber of dimensions given ASME, New York, 1981. above. Assume the swirling flow is always a solid-body rotation. Compare the 12. Liou, T-M ., and Santavicca, D. A.: "Cycle Resolved LDV Measurements in a Motored IC tangential velocity at the bowl edge with the mean piston speed. Neglect any Engine," ASME Trans ., J. Fluids Engng, vol. 107, pp. 232-240, 1985. 13. Amann, C. A.: "Classical Combustion Diagnostics for Engine Research," SAE paper 850395, in friction effects. Engine Combustion Analysis: New Approaches, P-156, SAE, 1985. (c) What would the swirl ratio be if the top-center clearance height was zero? 14. Dyer, T. M.: "New Experimental Techniques for In-Cylinder Engine Studies," SAE paper 850396, 8.5. Using Eq. (8.37) and Fig. 8-20b plot the squish velocity divided by the mean piston in Engine Combustion Analysis: New Approaches, P-156, SAE, 1985. speed at 10º BTC (the approximate location of the maximum) as a function of squish 15. Rask, R. B.: "Laser Doppler Anemometer Measurements in an Internal Combustion Engine," area expressed as a percentage of the cylinder cross section, A,/(zB2/4) x 100, from SAE paper 790094, SAE Trans ., vol. 88, 1979. 50 tò 0 percent. re = 10, c/B = 0.01, B/L = 1, R = 1/a = 3.5. 16. Liou, T .- M ., Hall, M ., Santavicca, D. A ., and Bracco, F. V.:" Laser Dopper Velocimetry Measure- ments in Valved and Ported Engines," SAE paper 840375, SAE Trans ., vol. 93, 1984. 8.6. Figure 8-24 shows the velocity at the prechamber nozzle throat during compression 17. Arcoumanis, C ., and Whitelaw, J. H.: "Fluid Mechanics of Internal Combustion Engines: A for dimensions typical of a small swirl chamber indirect-injection diesel. Assuming Review," Proc. Instn Mech. Engrs ., vol. 201, pp. 57-74, 1987. that the swirl chamber shape is a disc of height equal to the diameter, that the nozzle 18. Bopp, S ., Vafidis, C ., and Whitelaw, J. H.: "The Effect of Engine Speed on the TDC Flowfield in throat is at 0.8 x prechamber radius, and that the flow enters the prechamber a Motored Reciprocating Engine," SAE paper 860023, 1986. tangentially, estimate the swirl ratio based on the total angular momentum about 19. Wong, V. W ., and Hoult, D. P.: "Rapid Distortion Theory Applied to Turbulent Combustion," the swirl chamber axis in the prechamber at top-center. Assume B = L; neglect fric- SAE paper 790357, SAE Trans ., vol. 88, 1979. 20. Fraser, R. A ., Felton, P. G ., and Bracco, F. V.: "Preliminary Turbulence Length Scale Measure- tion. ments in a Motored IC Engine," SAE paper 860021, 1986. 8.7. The total crevice volume in an automobile spark-ignition engine is about 3 percent 21. Ikegami, M ., Shioji, M ., and Nishimoto, K.: "Turbulence Intensity and Spatial Integral Scale of the clearance volume. If the gas in these crevice regions is close to the wall tem- during Compression and Expansion Strokes in a Four-cycle Reciprocating Engine," SAE paper perature (450 K) and at the cylinder pressure, estimate the fraction of the cylinder 870372, 1987. mass within these crevice regions at these crank angles: inlet valve closing (50º ABC), 22. Uzkan, T ., Borgnakke, C ., and Morel, T.: “Characterization of Flow Produced by a High-Swirl spark discharge (30º BTC), maximum cylinder pressure (15º ATC), exhaust valve Inlet Port," SAE paper 830266, 1983. opening (60º BBC), TC of the exhaust stroke. Use the information in Fig. 1-8 for 23. Monaghan, M. L ., and Pettifer, H. F.: "Air Motion and Its Effects on Diesel Performance and your input data, and assume the inlet pressure is 0.67 atm. Emissions," SAE paper 810255, in Diesel Combustion and Emissions, pt. 2, SP-484, SAE Trans ., vol. 90, 1981. REFERENCES 24. Tindal, M. J ., Williams, T. J ., and Aldoory, M.: "The Effect of Inlet Port Design on Cylinder Gas 1. Bicen, A. F ., Vafidis, C ., and Whitelaw, J. H.: "Steady and Unsteady Airflow through the Intake Motion in Direct Injection Diesel Engines," in Flows in Internal Combustion Engines, pp. 101-111, Valve of a Reciprocating Engine," ASME Trans ., J. Fluids Engng, vol. 107, pp. 413-420, 1985. ASME, New York, 1982. 2. Namazian, M ., Hansen, S. P ., Lyford-Pike, E. J ., Sanchez-Barsse, J ., Heywood, J. B ., and Rife, J.: 25. Brandl, F ., Reverencic, I ., Cartellieri, W ., and Dent, J. C.: "Turbulent Air Flow in the Com- "Schlieren Visualization of the Flow and Density Fields in the Cylinder of a Spark-Ignition bustion Bowl of a D.I. Diesel Engine and Its Effect on Engine Performance," SAE paper 790040, Engine," SAE paper 800044, SAE Trans ., vol. 89, 1980. SAE Trans ., vol. 88, 1979. 3. Ekchian, A ., and Hoult, D. P.: "Flow Visualization Study of the Intake Process of an Internal 26. Brandstätter, W ., Johns, R. J. R ., and Wigley, G.: "Calculation of Flow Produced by a Tangential Combustion Engine," SAE paper 790095, SAE Trans ., vol. 88, 1979. Inlet Port," in International Symposium on Flows in Internal Combustion Engines-III, FED vol. 4. Hirotomi, T ., Nagayama, I ., Kobayashi, S ., and Yamamasu, M.: "Study of Induction Swirl in a 28, pp. 135-148, ASME, New York, 1985. Spark Ignition Engine," SAE paper 810496, SAE Trans ., vol. 90, 1981. 27. Brandstätter, W ., Johns, R. J. R ., and Wigley, G.: "The Effect of Inlet Port Geometry on In- 5. Reynolds, W. C.: " Modeling of Fluid Motions in Engines -- An Introductory Overview," in J. N. Cylinder Flow Structure," SAE paper 850499, 1985. Mattavi and C. A. Amann (eds.), Combustion Modelling in Reciprocating Engines, pp. 69-124, 28. Davis, G. C ., and Kent, J. C.: "Comparison of Model Calculations and Experimental Measure- Plenum Press, 1980. ments of the Bulk Cylinder Flow Processes in a Motored PROCO Engine," SAE paper 790290, 6. Tennekes, H ., and Lumley, J. L.: A First Course in Turbulence, MIT Press, 1972. 1979. 370 INTERNAL COMBUSTION ENGINE FUNDAMENTALS 29. Borgnakke, C ., Davis, G. C ., and Tabaczynski, R. J.: "Predictions of In-Cylinder Swirl Velocity and Turbulence Intensity for an Open Chamber Cup in Piston Engine," SAE paper 810224, SAE Trans ., vol. 90, 1981. CHAPTER 30. Arcoumanis, C ., Bicen, A. F ., and Whitelaw, J. H.: "Squish and Swirl-Squish Interaction in Motored Model Engines," Trans. ASME, J. Fluids Engng, vol. 105, pp. 105-112, 1983. 31. Ikegami, M ., Mitsuda, T ., Kawatchi, K ., and Fujikawa, T.: "Air Motion and Combustion in. Direct Injection Diesel Engines," JARI technical memorandum no. 2, pp. 231-245, 1971. 9 32. Liou, T .- M ., and Santavicca, D. A.: "Cycle Resolved Turbulence Measurements in a Ported Engine With and Without Swirl," SAE paper 830419, SAE Trans ., vol. 92, 1983. 33. Fitzgeorge, D ., and Allison, J. L.: "Air Swirl in a Road-Vehicle Diesel Engine," Proc. Instn Meck Engrs (A.D.), no. 4, pp. 151-168, 1962-1963. 34. Lichty, L. C.: Combustion Engine Processes, McGraw-Hill, 1967. COMBUSTION IN 35. Shimamoto, Y ., and Akiyama, K.: "A Study of Squish in Open Combustion Chambers of a Diesel SPARK-IGNITION Engine," Bull. JSME, vol. 13, no. 63, pp. 1096-1103, 1970. 36. Dent, J. C ., and Derham, J. A.: "Air Motion in a Four-Stroke Direct Injection Diesel Engine," ENGINES Proc. Instn Mech. Engrs, vol. 188, 21/74, pp. 269-280, 1974. 37. Asanuma, T ., and Obokata, T.: "Gas Velocity Measurements of a Motored and Firing Engine by Laser Anemometry," SAE paper 790096, SAE Trans ., vol. 88, 1979. 38. Asanuma, T ., Babu, M. K. G ., and Yagi, S.: "Simulation of Thermodynamic Cycle of Three-Valve Stratified Charge Engine," SAE paper 780319, SAE Trans ., vol. 87, 1978. 39. Hires, S. D ., Ekchian, A ., Heywood, J. B ., Tabaczynski, R. J ., and Wall, J. C.: " Performance and NO, Emissions Modelling of a Jet Ignition Prechamber Stratified Charge Engine," SAE paper 760161, SAE Trans ., vol. 85, 1976. 40. Zimmerman, D. R.: “Laser Anemometer Measurements of the Air Motion in the Prechamber of an Automotive Diesel Engine," SAE paper 830452, 1983. 41. Meintjes, K ., and Alkidas, A. C.: "An Experimental and Computational Investigation of the Flow in Diesel Prechambers," SAE paper 820275, SAE Trans ., vol. 91, in Diesel Engine Combustion, Emissions, and Particulates, P-107, SAE, 1982. 42. Namazian, M ., and Heywood, J. B.: "Flow in the Piston-Cylinder-Ring Crevices of a Spark- Ignition Engine: Effect on Hydrocarbon Emissions, Efficiency and Power," SAE paper 820088. SAE Trans ., vol. 91, 1982. 43. Wentworth, J. T.: "Piston and Ring Variables Affect Exhaust Hydrocarbon Emissions," SAE paper 680109, SAE Trans ., vol. 77, 1968. 44. Tabaczynski, R. J ., Hoult, D. P ., and Keck, J. C.: "High Reynolds Number Flow in a Moving 9.1 ESSENTIAL FEATURES OF PROCESS Corner," J. Fluid Mech ., vol. 42, pp. 249-255, 1970. 45. Daneshyar, H. F ., Fuller, D. E ., and Deckker, B. E. L.: " Vortex Motion Induced by the Piston of In a conventional spark-ignition engine the fuel and air are mixed together in the an Internal Combustion Engine," Int. J. Mech. Sci ., vol. 15, pp. 381-390, 1973. intake system, inducted through the intake valve into the cylinder, where mixing with residual gas takes place, and then compressed. Under normal operating conditions, combustion is initiated towards the end of the compression stroke at the spark plug by an electric discharge. Following inflammation, a turbulent flame develops, propagates through this essentially premixed fuel, air, burned gas mixture until it reaches the combustion chamber walls, and then extinguishes. Photographs of this process taken in operating engines illustrate its essential fea- tures. Figure 9-1 (color plate) shows a sequence of frames from a high-speed color movie of the combustion process in a special single-cylinder engine with a glass piston crown.1 The spark discharge is at -30º. The flame first becomes visible in the photos at about -24º. The flame, approximately circular in outline in this FIGURE 9-1 (On color plate opposite p. 498) Color photographs from high-speed movie of spark-ignition engine combustion process, taken through glass piston crown. Ignition timing 30º BTC, light load, 1430 rev/min, (A/F) = 19.1 371 372 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 373 view through the piston, then propagates outward from the spark plug location. The blue light from the flame is emitted most strongly from the front. The irregu- 2 lar shape of the turbulent flame front is apparent. At TC the flame diameter is 30 about two-thirds of the cylinder bore. The flame reaches the cylinder wall farthest from the spark plug about 15º ATC, but combustion continues around parts of 5 the chamber periphery for another 10º. At about 10º ATC, additional radiation- initially white, turning to pinky-orange-centered at the spark plug location is evident. This afterglow comes from the gases behind the flame which burned p. atm 20H earlier in the combustion process, as these are compressed to the highest tem- peratures attained within the cylinder (at about 15º ATC) while the rest of the charge burns. 2, 3 Additional features of the combustion process are evident from the data in Fig. 9-2, taken from several consecutive cycles of an operating spark-ignition 10/ Motored engine. The cylinder pressure, fraction of the charge mass which has burned (determined from the pressure data, see Sec. 9.2), and fraction of the cylinder volume enflamed by the front (determined from photographs like Fig. 9-1) are -20 Spark 20 shown, all as a function of crank angle.4 Following spark discharge, there is a period during which the energy release from the developing flame is too small for 1.0[ the pressure rise due to combustion to be discerned. As the flame continues to grow and propagate across the combustion chamber, the pressure then steadily 0.8 -- 2 rises above the value it would have in the absence of combustion. The pressure reaches a maximum after TC but before the cylinder charge is fully burned, and 0.6- then decreases as the cylinder volume continues to increase during the remainder of the expansion stroke. 0.4 The flame development and subsequent propagation obviously vary, cycle- by-cycle, since the shape of the pressure, volume fraction enflamed, and mass 0.2- fraction burned curves for each cycle differ significantly. This is because flame growth depends on local mixture motion and composition. These quantities vary -20 in successive cycles in any given cylinder and may vary cylinder-to-cylinder. 20 40 60 Especially significant are mixture motion and composition in the vicinity of the 1.0 spark plug at the time of spark discharge since these govern the early stages of flame development. Cycle-by-cycle and cylinder-to-cylinder variations in com- 0.8- bustion are important because the extreme cycles limit the operating regime of the engine (see Sec. 9.4.1). 0.6 Note that the volume fraction enflamed curves rise more steeply than the mass fraction burned curves. In large part, this is because the density of the 0.4 unburned mixture ahead of the flame is about four times the density of the burned gases behind the flame. Also, there is some unburned mixture behind the 0.2 visible front to the flame: even when the entire combustion chamber is fully enflamed, some 25 percent of the mass has still to burn. From this description it 0 is plausible to divide the combustion process into four distinct phases: (1) spark -20 0 20 40 60 ignition; (2) early flame development; (3) flame propagation; and (4) flame termi- Crank angle, deg FIGURE 9-2 nation. Our understanding of each of these phases will be developed in the Cylinder pressure, mass fraction burned, and volume fraction enflamed for five consecutive cycles in a remainder of this chapter. rev/min, ¢ = 0.98.4 spark-ignition engine as a function of crank angle. Ignition timing 30º BTC, wide-open throttle, 1044 The combustion event must be properly located relative to top-center to 374 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 375 obtain the maximum power or torque. The combined duration of the flame Empirical rules for relating the mass burning profile and maximum cylinder development and propagation process is typically between 30 and 90 crank angle pressure to crank angle at MBT timing are often used. For example, with degrees. Combustion starts before the end of the compression stroke, continues optimum spark timing: (1) the maximum pressure occurs at about 16º after TC; through the early part of the expansion stroke, and ends after the point in the (2) half the charge is burned at about 10º after TC. In practice, the spark is often cycle at which the peak cylinder pressure occurs. The pressure versus crank angle retarded to give a 1 or 2 percent reduction in brake torque from the maximum curves shown in Fig. 9-3a allow us to understand why engine torque (at given value, to permit a more precise definition of timing relative to the optimum. engine speed and intake manifold conditions) varies as spark timing is varied So far we have described normal combustion in which the spark-ignited relative to TC. If the start of the combustion process is progressively advanced flame moves steadily across the combustion chamber until the charge is fully before TC, the compression stroke work transfer (which is from the piston to the consumed. However, several factors-e.g ., fuel composition, certain engine design cylinder gases) increases. If the end of the combustion process is progressively and operating parameters, and combustion chamber deposits-may prevent this delayed by retarding the spark timing, the peak cylinder pressure occurs later in normal combustion process from occurring. Two types of abnormal combustion the expansion stroke and is reduced in magnitude. These changes reduce the have been identified: knock and surface ignition. expansion stroke work transfer from the cylinder gases to the piston. The Knock is the most important abnormal combustion phenomenon. Its name optimum timing which gives the maximum brake torque-called maximum brake comes from the noise that results from the autoignition of a portion of the fuel, torque, or MBT, timing-occurs when the magnitudes of these two opposing air, residual gas mixture ahead of the advancing flame. As the flame propagates trends just offset each other. Timing which is advanced or retarded from this across the combustion chamber, the unburned mixture ahead of the flame- optimum gives lower torque. The optimum spark setting will depend on the rate called the end gas-is compressed, causing its pressure, temperature, and density of flame development and propagation, the length of the flame travel path across to increase. Some of the end-gas fuel-air mixture may undergo chemical reactions the combustion chamber, and the details of the flame termination process after it prior to normal combustion. The products of these reactions may then autoig- reaches the wall. These depend on engine design and operating conditions, and nite: i.e ., spontaneously and rapidly release a large part or all of their chemical the properties of the fuel, air, burned gas mixture. Figure 9-3b shows the effect of energy. When this happens, the end gas burns very rapidly, releasing its energy at variations in spark timing on brake torque for a typical spark-ignition engine. a rate 5 to 25 times that characteristic of normal combustion. This causes high- The maximum is quite flat. frequency pressure oscillations inside the cylinder that produce the sharp metallic noise called knock. The presence or absence of knock reflects the outcome of a race between the advancing flame front and the precombustion reactions in the unburned end 1.0 Spark advance = 50 deg gas. Knock will not occur if the flame front consumes the end gas before these 3 reactions have time to cause the fuel-air mixture to autoignite. Knock will occur if the precombustion reactions produce autoignition before the flame front arrives. 30 The other important abnormal combustion phenomenon is surface ignition. N 0.9 Relative torque Surface ignition is ignition of the fuel-air charge by overheated valves or spark Cylinder pressure, MPa MBT plugs, by glowing combustion-chamber deposits, or by any other hot spot in the 10 engine combustion chamber: it is ignition by any source other than normal spark ignition. It may occur before the spark plug ignites the charge (preignition) or ignition after normal ignition (postignition). It may produce a single flame or many flames. Uncontrolled combustion is most evident and its effects most severe when 20 30 10 0 10 it results from preignition. However, even when surface ignition occurs after the -60 -40 -20 TC 20 40 60 80 100 Spark advance, deg spark plug fires (postignition), the spark discharge no longer has complete Crank angle, deg (a) (b) control of the combustion process. Surface ignition may result in knock. Knock which occurs following normal FIGURE 9-3 spark ignition is called spark knock to distinguish it from knock which has been (a) Cylinder pressure versus crank angle for overadvanced spark timing (50º), MBT timing (30º), and retarded timing (10º). (b) Effect of spark advance on brake torque at constant speed and (A/F), al preceded by surface ignition. Abnormal combustion phenomena are reviewed in more detail in Sec. 9.6. wide-open throttle. MBT is maximum brake torque timing.5 376 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 377 9.2 THERMODYNAMIC ANALYSIS OF SI ENGINE COMBUSTION tions in the burned and unburned gas are then determined by conservation of mass : 9.2.1 Burned and Unburned Mixture States Because combustion occurs through a flame propagation process, the changes in m = 0% dx + 0 dx (9.1) state and the motion of the unburned and burned gas are much more complex than the ideal cycle analysis in Chapter 5 suggests. The gas pressure, temperature, and conservation of energy: and density change as a result of changes in volume due to piston motion. During combustion, the cylinder pressure increases due to the release of the fuel's Uo - W - Q chemical energy. As each element of fuel-air mixture burns, its density decreases m = 4, dx + u , dx Jo Jxo 9.2) by about a factor of four. This combustion-produced gas expansion compresses the unburned mixture ahead of the flame and displaces it toward the combustion where V is the cylinder volume, m is the mass of the cylinder contents, u is the chamber walls. The combustion-produced gas expansion also compresses those specific volume, x, is the mass fraction burned, U, is the internal energy of the parts of the charge which have already burned, and displaces them back toward cylinder contents at some reference point 00, u is the specific internal energy, W the spark plug. During the combustion process, the unburned gas elements move is the work done on the piston, and Q is the heat transfer to the walls. The away from the spark plug; following combustion, individual gas elements move subscripts u and b denote unburned and burned gas properties, respectively. The back toward the spark plug. Further, elements of the unburned mixture which work and heat transfers are burn at different times have different pressures and temperatures just prior to combustion, and therefore end up at different states after combustion. The ther- modynamic state and composition of the burned gas is, therefore, non-uniform. A w =pav' e - [ (BGON (9.3) first law analysis of the spark-ignition engine combustion process enables us to quantify these gas states. where Q is the instantaneous heat-transfer rate to the chamber walls. Consider the schematic of the engine cylinder while combustion is in To proceed further, models for the thermodynamic properties of the burned progress, shown in Fig. 9-4. Work transfer occurs between the cylinder gases and and unburned gases are required. Several categories of models are described in the piston (to the gas before TC; to the piston after TC). Heat transfer occurs to Chap. 4. Accurate calculations of the state of the cylinder gases require an equi- the chamber walls, primarily from the burned gases. At the temperatures and librium model (or good approximation to it) for the burned gas and an ideal gas pressures typical of spark-ignition engines it is a reasonable approximation to mixture model (of frozen composition) for the unburned gas (see Table 4.2). assume that the volume of the reaction zone where combustion is actually However, useful illustrative results can be obtained by assuming that the burned occurring is a negligible fraction of the chamber volume even though the thick- and unburned gases are different ideal gases, each with constant specific heats;6 i.e ., ness of-the turbulent flame may not be negligible compared with the chamber dimensions (see Sec. 9.3.2). With normal engine operation, at any point in time or crank angle, the pressure throughout the cylinder is close to uniform. The condi- Us = Co , o To + hs .b (9.4) PUR = Ru Tu (9.5) Combining Eqs. (9.1) to (9.5) gives m PV = x,R. T. + (1 - x,)R. T. (9.6) FIGURE 9-4 and U. - W-Q A = = x 6 ( Co . b Tp + h3 .b ) + ( 1 - x 6 X Co. , Tu + 25.1) Schematic of flame in the engine cylinder during m (9.7) .. .. .. combustion: unburned gas (U) to left of flame, where burned gas to right. A denotes adiabatic burned-gas core, BL denotes thermal boundary layer in burned BL gas, W is work-transfer rate to piston, Q is heat- transfer rate to chamber walls. X o Jo 1 T dx 378 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 379 are the mean temperatures of the burned and unburned gases. Equations (9.6) 2.5 ¢ = 1.1 and (9.7) may now be solved to obtain 2.0- 1.0 - PV - Po Vo + (76 - 1)(W + Q) + (Vb - Y)mc ... (Tu - To) P 1.51 (9.8) 40.8 m [ ( Yo - 1 Xh su - 25.6 ) + ( Yo - Y) Co .. [2] P. M. 10.6 x b 1.0} 10.4 and = RUT + PV - mR, ₸ 0.5 10.2 Ro mRE Xb (9.9) 3000 FIGURE 9-5 Cylinder pressure, mass fraction burned, If we now assume the unburned gas is initially uniform and undergoes isen- and gas temperatures as functions of tropic compression, then 2000 T, K crank angle during combustion. T ., is To . Tb. ! unburned gas temperature, T, is burned (y= - 1)/Yu gas temperature, the subscripts e and / -1000 To = (9.10) denote early and late burning gas ele- ments, and T, is the mean burned gas temperature.9 (Reprinted with per- -40 -20 0 20 40 60 mission. Copyright 1973, American This equation, with Eqs. (9.8) and (9.9) enables determination of both x, and T, Crank angle, deg Chemical Society.) from the thermodynamic properties of the burned and unburned gases, and known values of p, V, m, and Q. Alternatively, if x, is known then p can be determined. Mass fraction burned and cylinder gas pressure are uniquely related. where Ti(x, Xb) is the temperature of the element which burned at the pressure While Eq. (9.9) defines a mean burned gas temperature, the burned gas is p(x) when the pressure is p(x)), and not uniform. Mixture which burns early in the combustion process is further compressed after combustion as the remainder of the charge is burned. Mixture T ( x ) = 11. 4 - 25 . 0 + C Q . I . ( x 2 ) (9.12) which burns late in the combustion process is compressed prior to combustion Cp,b and, therefore, ends up at a different final state. A temperature gradient exists is the temperature resulting from isenthalpic combustion of the unburned gas at across the burned gas with the earlier burning portions at the higher tem- T.(x'), p(xi). An example of the temperature distribution computed with this perature.7, 8 Two limiting models bracket what occurs in practice: (1) a fully model is shown in Fig. 9-5. A mixture element that burns right at the start of the mixed model, where it is assumed that each element of mixture which burns combustion process reaches, in the absence of mixing, a peak temperature after mixes instantaneously with the already burned gases (which therefore have a combustion about 400 K higher than an element that burns toward the end of uniform temperature), and (2) an unmixed model, where it is assumed that no the combustion process. The mean burned gas temperature is closer to the lower mixing occurs between gas elements which burn at different times. of these temperatures. These two models approximate respectively to situations In the fully mixed model the burned gas is uniform, T; = T;, and the equa- where the time scale that characterizes the turbulent mixing process in the tions given above fully define the state of the cylinder contents. In the unmixed burned gases is (1) much less than the overall burning time (for the fully mixed model, the assumption is made that no mixing occurs between gas elements that model) or (2) much longer than the overall burning time (for the unmixed model). burn at different times, and each burned gas element is therefore isentropically The real situation lies in between. compressed (and eventually expanded) after combustion.+ Thus: Measurements of burned gas temperatures have been made in engines using T. ( x's , Xx ). [P(x) ](6-17/76 spectroscopic techniques through quartz windows in the cylinder head. Examples (9.11) of measured temperatures are shown in Fig. 9-6. The solid lines marked A, B, and Ti(x') P (x'%)_ C are the burned gas temperatures measured by Rassweiler and Withrow7 using the sodium line reversal technique in an L-head engine, for the spark plug end (A), the middle (B), and the opposite end (C) of the chamber, respectively. Curves labeled W2 and W3 were measured by Lavoie8 through two different windows, W2 and W3 (with W2 closer to the spark), again in an L-head engine. Each set of + This model applies to burned gas regions of the chamber away from the walls. Heat transfer to the experimental temperatures shows a temperature gradient across the burned gas walls results in a thermal boundary layer on the walls which grows with time. The gas in the bound- comparable to that predicted, and the two sets have similar shapes. ary layer is not isentropically compressed and expanded. 380 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 381 2800 fame with a constant burning velocity propagates outward from the center of a 2700 Early burned spherical container. Applying this gas motion model to an engine, it can be con- cluded that a window in the cylinder head initially views earlier burned gas (of higher temperature and entropy) and that as more of the charge burns, the 2600 window views later burned gas of progressively lower entropy. The experimental curves fit this description: they cross the constant entropy lines toward lower 2500 entropy. Note that the gradient in temperature persists well into the expansion T, K stroke, indicating that the "unmixed" model is closer to reality than the "fully 2400 As = 0 Late burned mixed " model. More accurate calculations relating the mass fraction burned, gas pressure, ¢ = 1.1 2300- and gas temperature distribution are often required. Note that the accuracy of Window W2 Window W3 FIGURE 9-6 such calculations depends on the accuracy with which the time-varying heat loss A Near spark Burned gas temperatures measured using spec- to the chamber walls can be estimated (see Sec. 12.4.3) and whether flows into 2200 B Middle troscopic techniques through windows in the and out of crevice regions are significant (see Sec. 8.6), as well as the accuracy of C Opposite spark cylinder head, as a function of cylinder pressure. the models used to describe the thermodynamic properties of the gases. Appro- 21001 Temperatures measured closer to spark plug 1.5 2 3 have higher values. Dashed lines show isentropic priate more accurate models for the thermodynamic properties are: an equi- p, MPa behavior.7, 8 librium model for the burned gas, and specific heat models which vary with temperature for each of the components of the unburned mixture (see Secs. 4.1 and 4.7). In the absence of significant crevice effects, Eqs. (9.1) and (9.2) can be In the unmixed model, the temperature of each burned gas element follows written as a different isentropic line as it is first compressed as p increases to Pmar and then expanded as the pressure falls after Pmax. The measured temperature curves in V - = 06 X6 + 0 (1 - xx) (9.13) Fig. 9-6 do not follow the calculated isentropes because of gas motion past the m observation ports. As has already been mentioned, the expansion of a gas element which occurs during combustion compresses the gas ahead of the flame Uo - W -2 - 4 x6 + u (1 - xx) m . (9.14) and moves it away from the spark plug. At the same time, previously burned gas is compressed and moved back toward the spark plug. Defining this motion in an where engine requires sophisticated flow models, because the combustion chamber 1 "Xb shape is rarely symmetrical, the spark plug is not usually centrally located, and DO Un dx and 1 0 dx often there is a bulk gas motion at the time combustion is initiated. However, the Xb JO 1 - xx J xo gas motion in a spherical or cylindrical combustion bomb with central ignition which can readily be computed illustrates the features of the combustion-induced and similar definitions hold for us and uy. For a given equivalence ratio, fuel and motion in an engine. Figure 9-7 shows calculated particle trajectories for a stoi- burned gas fraction: chiometric methane-air mixture, initially at ambient conditions, as a laminar hu = h ( T ) ho = ho ( Tb, P ) (9.15a, b) 1.0 -78 (9.16a, b) 0.8 4 and us = ho - pix (9.17a, b) 0.6 Burned gas Normalized time 2 To simplify the calculations, it is convenient to assume that, for the burned gas, Pressure 0.4 Flame fronty FIGURE 9-7 Particle trajectories in unburned and burned gas " = up(Tb, p) and D = vp(T), p). This corresponds to the fully mixed assumption 0.2 -- Unburned gas as flame propagates outward at constant velocity described above. The effect of neglecting the temperature distribution in the cal- from the center of a spherical combustion bomb. culation of mass fraction burned is small. In addition, the heat losses from the OL JI 0 0.2 0.4 0.6 0.8 1.0 Stoichiometric methane-air mixture initially at 1 unburned gas can usually be neglected; the unburned gas is then compressed Normalized radius atm and 300 K. isentropically. T, is specified for some initial state of the unburned gas (where 382 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 383 X = 0) by Po, Vo, M ., and the mass of charge m. Then, since for any isentropie 3000 process Pmax ap =. v - (@h/ap)I (8h/T), (9.1 :300 - T, can be determined. Equations (9.13) to (9.18) constitute a set of nine equations for the niney unknowns Du, Do, y, ub, hu, , , , ando . One convenient soluto method is to eliminate x, from Eqs. (9.13) and (9.14) to obtain Spark (V/m) - Uu ------- ( U / m ) - u y = f ( T . , T . ) = 0 (9.19) 1400 Do - Du 0.95 0.8 where U = U. - W - Q. T ., can be determined from Eq. (9.18). Equation (9.19) 0.2 0.4 b = 0.05 can then be solved using an appropriate iterative technique for To, and x, can be FIGURE 9-9 2200 obtained from Eq. (9.13). An alternative formulation based on the rate of chant Calculated temperature distribution in of pressure dp/de and equations for dT./de, dT /de, dm./de, and dV,/de can be the adiabatic core of the burned gas zone for the unmixed model assuming found in Ref. 10. Some examples of mass fraction burned curves obtained from measured pressure data, with gasoline and methanol fuels, are shown in Fig. 9-2 10 thermodynamic equilibrium. ¢ = 1.0. 15 20 Dashed line is temperature of each With accurate pressure versus crank angle records, values of final mass fraction Pressure, atm element just after it burns. burned should be close to but lower than unity, usually in the range 0.93 to 0.98: the difference from unity is the combustion inefficiency for lean mixtures (see Fin divided into an adiabatic core and a boundary layer that grows in thickness with 3-9) and incomplete oxygen utilization for rich mixtures (see Fig. 4-20). time. In the adiabatic core, in the absence of mixing between gas elements that More accurate burned gas temperature calculations need to account for the burn at different times, burned gas is compressed and then expanded isentropi- presence of a thermal boundary layer (of order 1 mm thick) around the com- cally. The burned gas temperature distribution can be calculated as follows. bustion chamber walls (see Sec. 12.6.5). The burned gas region in Fig. 9-4 can bj Given the pressure versus crank angle data, the unburned mixture state can be determined using Eq. (9.18) above. Each small element of unburned mixture 1.2 1.2, burns in a constant-enthalpy constant-pressure process. So the burned state of an Indolene Methanol clement of unburned charge, which burns at p = p;, can be obtained from the relation 1.0L 1.0 ho( Tb, ;, Pi) = hu(Tu ., Pi) ).8 o = 1.13 0.8- 0 = 1.12 ¢ = 1.01 ¢ = 0.97 After combustion, this element which burned at p = p; is compressed and ¢ = 0.94 ¢ = 0.77. expanded along the isentropic: 0.6 ¢ = 0.80 -$ = 0.69 -¢ = 0.5% Mass fraction burned 0.6- $ = 0.72 Mass fraction burned -$ = 0.68 L-$ = 0.57 Sa(Tb, P) = So(Tb.i, Pi) 0.4- 0.4- An example of the temperature distribution computed in this manner for this unmixed model in the burned gas adiabatic core is shown in Fig. 9-9. The 0.2H 0.2- clement ignited by the spark is compressed to the highest peak temperature at Pent. The temperature difference across the bulk of the charge (0.05 < x] < 0.95) ts about 200 K. 0 20 40 60 80 100 0 20 40 60 80 Degrees after spark Degrees after spark 9.22 Analysis of Cylinder Pressure Data (a) (b) FIGURE 9-8 Cylinder pressure changes with crank angle as a result of cylinder volume change, Mass fraction burned curves determined from measured cylinder pressure data using two-zone combustion, heat transfer to the chamber walls, flow into and out of crevice bustion model: (a) gasoline; (b) methanol. ¢ = fuel/air equivalence ratio.11 "sions, and leakage. The first two of these effects are the largest. The effect of 384 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 385 volume change on the pressure can readily be accounted for; thus, combustion bustion can be located approximately in similar fashion; the expansion stroke rate information can be obtained from accurate pressure data provided models following combustion is essentially linear with slope 1.33. Since both the com- for the remaining phenomena can be developed at an appropriate level of pression of the unburned mixture prior to combustion and the expansion of the approximation. The previous section has developed the fundamental basis for burned gases following the end of combustion are close to adiabatic isentropic such calculations. processes (for which pV" = constant; y = c./cp), the observed behavior is as Cylinder pressure is usually measured with piezoelectric pressure trans- expected. More extensive studies12, 13 show that the compression and expansion ducers. This type of transducer contains a quartz crystal. One end of the crystal is processes are well fitted by a polytropic relation: exposed through a diaphragm to the cylinder pressure; as the cylinder pressure pV" = constant (9.20) increases, the crystal is compressed and generates an electric charge which is proportional to the pressure. A charge amplifier is then used to produce an The exponent n for the compression and expansion processes is 1.3 (+0.05) for output voltage proportional to this charge. Accurate cylinder pressure versus conventional fuels. It is comparable to the average value of y, for the unburned crank angle data can be obtained with these systems provided the following steps mixture over the compression process, but is larger than y, for the burned gas are carried out: (1) the correct reference pressure is used to convert the measured mixture during expansion due to heat loss to the combustion chamber walls (see pressure signals to absolute pressures; (2) the pressure versus crank angle (or Figs. 4-13 and 4-16). volume) phasing is accurate to within about 0.2º; (3) the clearance volume is Log p-log V plots such as Fig. 9-10 approximately define the start and end estimated with sufficient accuracy; (4) transducer temperature variations (which of combustion, but do not provide a mass fraction burned profile. One well- can change the transducer calibration factor) due to the variation in wall heat established technique for estimating the mass fraction burned profile from the flux during the engine cycle are held to a minimum. Log p versus log V plots can pressure and volume data is that developed by Rassweiler and Withrow.2 They be used to check the quality of cylinder pressure data. The first three of the above correlated cylinder pressure data with flame photographs, and showed how Eq. requirements can be validated using log p-log V diagrams for a motored engine. (9.20) could be used to account for the effect of cylinder volume change on the If the effects of thermal cycling are significant, the expansion stroke on the log p- pressure during combustion. Assuming that the unburned gas filling the volume log V plot for a firing engine shows excessive curvature.12 V. ahead of the flame at any crank angle during combustion has been compressed Figure 9-10 shows pressure-volume data from a firing spark-ignition engine polytropically by the advancing flame front, then the volume Vu, o it occupied at on both a linear p-V and a log p-log V diagram. On the log p-log V diagram the time of spark is compression process is a straight line of slope 1.3. The start of combustion can be 1 / m identified by the departure of the curve from the straight line. The end of com- Vu. o = V.( 2) (9.21) Similarly, the burned gas behind the flame filling the volume 1, would, at the end 3000 104 of combustion, fill a volume Vo, s given by Vb. s = V. ( 2 ) 1 / 2 (9.22) 2000 103 The mass fraction burned x, is equal to 1 - (Vu, o/Vo) and to Vo, y/V ,, where Vo Pressure, kPa Pressure, kPa amd V, are the total cylinder volumes at time of spark and at the end of com- 1000 102 bustion, respectively. Since V = V. + V ,, Eqs. (9.21) and (9.22) then give: T p 'INV - P. " Vo (9.23) TT PH"V, - P. "Vo ol 101 0.0 0.4 0.8 0.1 0.2 0.4 0.6 0.8 1.0 This method is widely used, though it contains several approximations. Fraction of maximum volume Fraction of maximum volume Heat-transfer effects are included only to the extent that the polytropic exponent (a) (b) n in Eq. (9.22) differs appropriately from y. The pressure rise due to combustion is proportional to the amount of fuel chemical energy released rather than the mass FIGURE 9-10 (a) Pressure-volume diagram; (b) log p-log (V/V ... ) plot. 1500 rev/min, MBT timing, imep = 513 kPa, of mixture burned. Also, the polytropic exponent n is not constant during com- ¢ = 0.8, re = 8.72, propane fuel.12 bustion. Selecting an appropriate value for n (whether n is assumed to be con- 386 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 387 stant or to vary through the combustion process) is the major difficulty where applying this pressure data analysis procedure. The effects of heat transfer, crevices, and leakage can be explicitly incorpos dm, > 0 when flow is out of the cylinder into the crevice rated into cylinder pressure data analysis by using a "heat release" approach dm -, < 0 when flow is from the crevice to the cylinder based on the first law of thermodynamics. A major advantage of such an h' is evaluated at cylinder conditions when dm- > 0 and at crevice conditions when dm- < 0 approach is that the pressure changes can be related directly to the amount of fuel chemical energy released by combustion, while retaining the simplicity of Use of the ideal gas law (neglecting the change in gas constant R) with Eq. (9.25) treating the combustion chamber contents as a single zone. Figure 9-11 shows then gives the appropriate open-system boundary for the combustion chamber.14 The first" law for this open system is 8Qch = (Ce)V dp + (C+ +1)pdv + (h - utc, T) dma + 80m (9.26) 80ch = dUs + 8Que + SW + Eh, dm; (9.24) This equation can be used in several ways. When the heat or energy release The change in sensible energy of the charge dU, is separated from that due to term, 0Och, is combined with the heat-transfer and crevice terms, the com- bination is termed net heat release-the combustion energy release less heat lost change in composition: the term 60ch represents the "chemical energy" released by combustion. The work is piston work and equal to p dV. 60ht is heat transfer to the walls. It is equal to the first two terms on the right-hand side of Eq. (9.26) to the chamber walls. The mass flux term represents flow across the system which, together, represent the sensible energy change and work transfer to the boundary. In the absence of fuel injection, it represents flow into and out of piston. While heat losses during combustion are a small fraction of the fuel crevice regions (see Sec. 8.6). energy (10 to 15 percent), the distributions of heat release and heat transfer with The accuracy with which this energy balance can be made depends on how crank angle are different; heat transfer becomes more important as the com- bustion process ends and average gas temperatures peak. The net heat-release adequately each term in the above equation can be quantified. Assuming that U. profile obtained from integrating the first two terms on the right-hand side of Eq. is given by mu(T), where T is the mean charge temperature and m is the mass (9.26), normalized to give unity at its maximum value, is often interpreted as the within the system boundary, then burned mass fraction (or, more correctly, the energy-release fraction) versus crank dU, = mc,(T) dT + u(T) dm angle profile. Use of Eq. (9.26) requires a value for c./R [=1/(y - 1)]. The ratio of specific Note that this mean temperature determined from the ideal gas law is close to the heats y for both unburned and burned gases decreases with increasing tem- mass-averaged cylinder temperature during combustion since the molecular perature and varies with composition (see Figs. 4-13, 4-16, and 4-18). As the mean weights of the burned and unburned gases are essentially identical. Crevice effects charge temperature increases during compression and combustion and then can usually be modeled adequately by flow into and out of a single volume at the decreases during expansion, y should vary. An approximate approach, modeling cylinder pressure, with the gas in the crevice at a substantially lower temperature. "T) with a linear function of temperature fitted to the appropriate curves in Figs. Leakage to the crankcase can usually be neglected. Then Eq. (9.24), on substitut- 4-13, 4-16, and 4-18 and with y constant during combustion, has been shown to ing for dU, and dm; (=dm- = - dm), becomes give adequate results. 15 The convective heat-transfer rate to the combustion chamber walls can be 80ch = mc, dT + (h' - u)dmex + PdV + 80ht (9.25) calculated from the relation dem = Ah (T - Tw) dt hinjdmf Open system boundary where A is the chamber surface area, T is the mean gas temperature, Tw, is the mean wall temperature, and he is the heat-transfer coefficient (averaged over the chamber surface area). he can be estimated from engine heat-transfer correlations (see Sec. 12.4.3). Since crevice effects are usually small, a sufficiently accurate model for their overall effect is to consider a single aggregate crevice volume i'dmer HE FIGURE 9-11 where the gas is at the same pressure as the combustion chamber, but at a differ- Open system boundary for combustion ent temperature. Since these crevice regions are narrow, an appropriate assump- OW chamber for heat-release analysis. ton is that the crevice gas is at the wall temperature. Inserting this crevice model 388 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 389 into Eq. (9.26), with y(T) = a + bT, gives the chemical energy- or gross heat- region must be developed. Due to this complexity, crevice models are usually release rate: omitted despite the fact that their impact can be significant. dech _ Y dv 1 -+ dp de y - 1 y - 1 9.2.3 Combustion Process Characterization T dp LTW + Tu (Y - 1) + BTW do (9.27) The mass fraction burned profiles as a function of crank angle in each individual +1 cycle shown in Fig. 9-2 and the chemical energy- or gross heat-release curve in Fig. 9-12 have a characteristic S-shape. The rate at which fuel-air mixture burns An example of the use of Eq. (9.27) to analyze an experimental pressure increases from a low value immediately following the spark discharge to a versus crank angle curve for a conventional spark-ignition engine is shown in maximum about halfway through the burning process and then decreases to Fig. 9-12. The integrated heat release is plotted against crank angle. The lowest close to zero as the combustion process ends. It proves convenient to use these curve shown is the net heat release. The addition of heat transfer and crevice mass fraction burned or energy-release fraction curves to characterize different models gives the chemical energy release. The curve at the top of the figure is the stages of the spark-ignition engine combustion process by their duration in crank mass of fuel within the combustion chamber times its lower heating value. It angles, thereby defining the fraction of the engine cycle that they occupy. The decreases slightly as Pmax is approached due to flow into crevices. The difference flame development process, from the spark discharge which initiates the com- between the final value of 2ch and (my QLHv) should equal the combustion ineffi- bustion process to the point where a small but measurable fraction of the charge ciency (which is a few percent of my QLHv). The combustion inefficiency can be has burned, is one such stage. It is influenced primarily by the mixture state, determined from the exhaust gas composition (see Sec. 4.9.4). Inaccuracies in the composition, and motion in the vicinity of the spark plug (see Sec. 9.3). The stage cylinder pressure data and the heat-release calculation will also contribute to this during which the major portion of the charge burns as the flame propagates to difference. An important advantage of a heat-release analysis that relates the the chamber walls is next. This stage is obviously influenced by conditions pressure changes to the amount of fuel chemical energy within the cylinder is that throughout the combustion chamber. The final stage, where the remainder of the this error can be determined. In the example in Fig. 9-12, the measured com- charge burns to completion, cannot as easily be quantified because energy-release bustion inefficiency was close to the amount shown. rates are comparable to other energy-transfer processes that are occurring. Two-zone models (one zone representing the unburned mixture ahead of The following definitions are most commonly used to characterize the the flame and one the burned mixture behind the flame) are used to calculate the energy-release aspects of combustion: mass fraction burned profile from measured cylinder pressure data.1º Figure 9-8 shows results from such an analysis, using the methodology described in Sec. Flame-development angle 402. The crank angle interval between the spark discharge 9.2.1. The advantage of a two-zone analysis is that the thermodynamic properties and the time when a small but significant fraction of the cylinder mass has burned of the cylinder contents can be quantified more accurately. The disadvantages are or fuel chemical energy has been released. Usually this fraction is 10 percent, though that the unburned and the burned zone heat-transfer areas must both now be other fractions such as 1 and 5 percent have been used.+ estimated, and a model for the composition of the gas flowing into the crevice Rapid-burning angle 40 ,. The crank angle interval required to burn the bulk of the charge. It is defined as the interval between the end of the flame-development stage Inefficiency (usually mass fraction burned or energy-release fraction of 10 percent) and the end 1000 of the flame-propagation process (usually mass fraction burned or energy-release my2LHV Crevice /1999o Qch fraction of 90 percent).+ 800 Heat transfer Overall burning angle 40 .. The duration of the overall burning process. It is the sum 600 of Al and Alb. 400 200 f This angle is sometimes called the ignition delay. Since the flame starts to propagate outward imme- 2500 rev/min diately following the spark discharge there is no delay, and the terminology used here is preferred (see 0.7 atm FIGURE 9-12 Sec. 9.3). Results of heat-release analysis showing -200 An alternative definition for A0, uses the maximum burning rate to define an angle or time charac- - 150 -100 -50 0 50 100 150 the effects of heat transfer, crevices, and teristic of the bulk charge burning process4 (see Fig. 9-13). 40, and A, are usually closely compara- Degrees ATC combustion inefficiency.14 ble. 390 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 391 dame can be obtained from photographs taken with techniques that are sensitive 100 to density changes in the flow field, such as schlieren and shadowgraph. With these techniques, a parallel light beam is passed through the combustion chamber. Portions of the beam which pass through regions where density gra- Percent dients normal to the beam exist are deflected, due to the refractive index gra- FIGURE 9-13 dients that result from the density gradients. In the schlieren technique, the beam Definition of flame-development angle, A0 ,. is focused on a knife edge; the deflected parts of the beam are displaced relative 0 -+ rapid-burning angle, Al ,, on mass fraction burne to the knife edge and produce lighter or darker regions when subsequently refo- 0,0% 090% versus crank angle curve. Us cused onto film. With the shadowgraph technique, the parallel beam emerging from the combustion chamber is photographed directly; deflected parts of the Figure 9-13 illustrates these definitions on a mass fraction burned, or frac- beam produce lighter and darker regions on the film. With these techniques, details of flame structure can be discerned. tion of fuel energy released, versus crank angle plot. While the selection of the 10 and 90 percent points is arbitrary, such a choice avoids the difficulties involved in Figure 9-14 shows a set of photographs from one engine cycle, from a high- determining accurately the shape of the curve at the start and end of combustion speed schlieren movie taken in a special visualization spark-ignition engine oper- These angles can be converted to times (in seconds) by dividing by 6N (with N in" ating at 1400 rev/min and 0.5 atm inlet pressure. Also shown are the cylinder revolutions per minute). pressure versus crank angle data, and the mass fraction burned profile calculated A functional form often used to represent the mass fraction burned versus from the pressure data using the method of Rassweiler and Withrow2 (see Sec. crank angle curve is the Wiebe function: 9.2.2). This engine had a square-cross-section cylinder with two quartz walls to permit easy optical access, but otherwise operated normally.17 Visualization of X% = 1 - exp -( 00 ) +x7 (9.28) where 0 is the crank angle, 00 is the start of combustion, A0 is the total com- bustion duration (x) = 0 to x) = 1), and a and m are adjustable parameters. Varying a and m changes the shape of the curve significantly. Actual mass frac tion burned curves have been fitted with a = 5 and m = 2.16 b 9.3 FLAME STRUCTURE AND SPEED 9.3.1 Experimental Observations d The combustion process in the spark-ignition engine takes place in a turbulent flow field. This flow field is produced by the high shear flows set up during the intake process and modified during compression, as described in Chap. 8. The importance of the turbulence to the engine combustion process was recognized 10 long ago through experiments where the intake event, and the turbulence it gen- f 72:0 -0.8 Cylinder pressure, atm erates, was eliminated and the rate of flame propagation decreased substantially. LA -0.6 Understanding the structure of this engine flame as it develops from the spark Mass fraction burned Spark 0.4 discharge and the speed at which it propagates across the combustion chamber, and how that structure and speed depend on charge motion, charge composition, 0.2 Jo.0 and chamber geometry, are critical to engine optimization. This section reviews - 60 TT -40 -20 0 20 . 40 60 experimental evidence that describes the essential features of the flame develop? Crank angle, deg ment and propagation processes. FIGURE 9-14 Direct flame photographs such as those in Fig. 9-1 indicate the location and Sequence of movie frames from one engine cycle in a square-cross-section cylinder, single-cylinder, shape of the actual reaction zone which radiates in the blue region of the visible engine with two glass walls, and corresponding pressure and mass fraction burned curves. 1400 rev/ spectrum. An irregular front is apparent. Further insight into the structure of the min, 0.5 atm inlet pressure. 92 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 393 the flame is especially important during the early stages of flame development when the pressure rise due to combustion is too small to be detected. These photographs show how the flame "ball," roughly spherical in shape, No swirl grows steadily from the time of spark discharge. The effect of turbulence is already visible in the convoluted flame surface in Fig. 9-14a. The volume enflamed behind the front continues to grow in a roughly spherical manner, except where intersected by the chamber walls, as seen in Fig. 9-14b and c. The mass fraction burned and the associated pressure rise due to combustion become significant by the time the flame front has traversed two-thirds to three-quarters of the field of view. Note that the fraction of the cylinder filled with enflamed Normal swirl charge is less than is suggested by the photos because the front of the flame is approximately spherical and the cylinder has a square cross section. Maximum cylinder pressure occurs close to the time the flame makes contact with the far wall, as seen in Fig. 9-14e. Finally, the unburned mixture ahead of and within the front burns out and the density gradients associated with the flame reaction zone disappear, clearing the field of view as shown in Fig. 9-14e and f. A useful relationship between the mass fraction burned, x,(=m/m), and the High swirl volume fraction occupied by the burned gas, y,(= V/,/V), can be obtained from the identities m = my + mb V = 11 + 16 and the ideal gas law: (9.29) No swirl While the density ratio (pu/Pt) does depend on the equivalence ratio, burned gas fraction in the unburned mixture, gas temperature, and pressure, its value is close to 4 for most spark-ignition engine operating conditions. Thus, the plot of x, against y has a universal form,10 as shown in Fig. 9-15. This curve is an impor- tant aid in interpreting flame geometry information. No swirl 1.0 10º 15º 20º 25º 0.8 Crank angle, degrees from ignition FIGURE 9-16 0.6 Laser shadowgraph photographs of engine combustion process taken in single-cylinder engine with transparent cylinder head. From top to bottom: side plug without swirl; side plug with normal swirl; Burned mass fraction, Xb side plug with high swirl; central plug without swirl; two plugs without swirl. 18 0.4 x, = 0.2 x, = The above-described features of the developing and propagating flame are 0.2 common to almost all engine geometries and operating conditions. Figure 9-16 FIGURE 9-15 Relation between mass fraction burned x, and vol- shows shadowgraph photographs of the flame at fixed crank angle intervals after 0.2 0.4 0.6 0.8 1.0 ume fraction burned y. x, is residual mass frac- Ignition, taken through a transparent cylinder head with different geometric and Burned volume fraction, yb tion.4 How configurations.18 The approximately spherical development of the flame 394 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 395 from the vicinity of the spark plug, except where it intercepts the chamber walls is evident for side and center ignition with one plug, and for ignition with two It is also well established that unburned mixture composition and state plugs in the absence of any intake generated swirl. With normal levels of swirl affect the burning rate. Reducing the inlet pressure (and maintaining the ratio of the flame center is convecteurh the swirling flow, but the flame front as it exhaust to inlet pressure fixed to hold the residual gas fraction constant) increases grows is still approximately spherical in shape. Only with unusually high levels of both the flame development and rapid burning angles.19 The fuel/air equivalence swirl and aerodynamic stabilization of the flame at the spark plug location does ratio affects the burning rate. Both flame development and burning angles show a the flame become stretched out and distorted by the flow in a major way.18 minimum for slightly rich mixtures (( ~ 1.2) and increase significantly as the At any given flame radius, the geometry of the combustion chamber and mixture becomes substantially leaner than stoichiometric.19, 20 The burned gas the spark plug location govern the flame front surface area-the area of the fraction in the unburned mixture, due to the residual gas fraction and any recy- approximately spherical surface corresponding to the leading edge of the flame cled exhaust gases, affects the burning rate: increasing the burned gas fraction contained by the piston, cylinder head, and cylinder wall. The larger this surface slows down both flame development and propagation.20 Fuel composition area, the greater the mass of fresh charge that can cross this surface and enter the changes can be significant also. While mixtures of isooctane or conventional gas- flame zone. The photos in Fig. 9-16 illustrate the importance of flame area. The olines with air and burned gases (at identical conditions) have closely comparable center plug location gives approximately twice the flame area of the side plug burning rates, propane, methane, methanol, and ethanol mixtures exhibit modest geometry at a given flame radius, and burns about twice as fast (the fraction of differences in burning rate and hydrogen-air mixtures substantial differences. The the cylinder volume enflamed is about twice the size, at a fixed crank angle inter- basic combustion chemistry of the fuel, air, burned gas mixture influences the val after spark). The arrangement with two spark plugs at opposite sides of the combustion process. However, the relative importance of combustion chemistry chamber is not significantly different in enflamed volume from the single center effects depends on combustion chamber design and burn rate. Faster burning plug because, once the flame fronts are intersected by the cylinder wall, the flame engines (which have higher turbulence) are less sensitive to changes in mixture front areas are comparable. composition, pressure, and temperature than are slower burning engines (which Mixture burning rate is strongly influenced by engine speed. It is well estab- have lower turbulence). The effects of chamber geometry, gas motion, and gas composition and state are interrelated.21 lished that the duration of combustion in crank angle degrees only increases slowly with increasing engine speed.19 Figure 9-17 shows how the interval between the spark discharge and 10 percent mass fraction burned, the flame 9.3.2 Flame Structure development angle A04, and the interval between the spark and 90 percent mass fraction burned, the overall burning angle A0. + 40, (see Sec. 9.2.3), vary with Laminar flames in premixed fuel, air, residual gas mixtures are characterized by a engine speed.20 Both intervals increase by a factor of about 1.6 for a factor of 4 laminar flame speed SL and a laminar flame thickness ô1 (see Sec. 9.3.3). The increase in engine speed; i.e ., the burning rate throughout the combustion process laminar flame speed is the velocity at which the flame propagates into quiescent increases almost, though not quite, as rapidly as engine speed. Additionally, at a premixed unburned mixture ahead of the flame. There are several ways to define given engine speed, increasing in-cylinder gas velocities (e.g ., with intake gen- the thickness of a laminar flame.22 Given the molecular diffusivity DI (see Sec. erated swirl) increases the burning rate: the flame size for the swirling flows in 4.8), dimensional arguments give the most commonly used definition: 6, == Fig. 9-16 is larger than for the quiescent case with the same plug location at the Dy/SL. Turbulent flames are also characterized by the root mean square velocity crank angle intervals after spark shown. Increasing engine speed and introducing fluctuation, the turbulence intensity u' [Eq. (8.3)], and the various length scales of swirl both increase the levels of turbulence in the engine cylinder at the time of the turbulent flow ahead of the flame. The integral length scale /, [Eq. (8.8)] is a combustion (see Sec. 8.2.2). Increased turbulence increases the rate of develop- measure of the size of the large energy-containing structures of the flow. The ment and propagation of the turbulent premixed engine flame. Kolmogorov scale Ix [Eq. (8.11)] defines the smallest structures of the flow where small-scale kinetic energy is dissipated via molecular viscosity. Several dimensionless parameters are used to characterize turbulent pre- 60 90% burn mixed flames. The dimensionless parameter used to define the turbulence is the turbulent Reynolds number, Rer = u'l /v. For homogeneous and isotropic (no 40 preferred direction) turbulence, the integral and Kolmogorov scales are related Combustion duration, deg FIGURE 9-17 20 10% burn by Eq. (8.14): 1x/l, = Ref 3/4. A characteristic turbulent eddy turnover time tr can Effect of engine speed on flame-development be defined as angle (0 to 10 percent burned) and overall 1000 2000 3000 4000 burning angle (0 to 90 percent burned). o = 1.A Engine speed, rev/min intake pressure 0.54 atm, spark 30º BTC.20 396 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 397 A characteristic chemical reaction time is the residence time in a laminar flame: Values of Da and Rer for a typical spark-ignition engine (the cross-hatched region in Fig. 9-18) lie predominantly in the reaction sheet flame regime. Engine operation at high speed (the lower right boundary) and low load (the lower left The ratio of the characteristic eddy turnover time to the laminar burning time is boundary) gives values of Da and Rer which fall below the Ix/6L = 1 line. This is called the Damköhler number: largely due to the low values of laminar flame speed that result from the high amounts of residual gas and EGR under these conditions (see Sec. 9.3.4 and Fig. IT S Da = = L (9.30) 6-19). Whether the flame structure under these conditions is significantly different is not known. Observations of engine flames to date, described below, lie above It is an inverse measure of the influence of the turbulent flow on the chemical the lx/81 = 1 line, within the reaction sheet regime. One would expect, then, the processes occurring in the flame. Other ratios are of interest. The ratio 6 /1x is a structure of the flame in a spark-ignition engine, once developed, to be that of a measure of the stretch or local distortion to which a laminar flame is subjected by thin reaction sheet wrinkled and convoluted by the turbulent flow. the turbulent flow. Unless 11/61 >> 1 the concept of a localized flame region has Detailed observations have been made of flame structure from ignition to little significance. The ratio u'/SL is a measure of the relative strength of the fame extinguishing at the far cylinder wall. A flame develops from the spark turbulence. 22 discharge which causes ignition as follows. In the initial breakdown phase of Different regimes of turbulent flames are apparent in the plot of Damkoh- ignition, a cylindrical discharge between the spark plug electrodes is estab- ler number versus turbulent Reynolds number in Fig. 9-18.22 It has been assumed lished.23 As electrical energy is fed into the discharge, the arc expands and exo- that D, ~ v and that the relationships for homogeneous isotropic turbulence are thermic chemical reactions capable of sustaining a propagating flame develop. valid. Two regimes-distributed reactions and reaction sheets-are normally Figure 9-19 shows how this development of a flame kernel occurs, with a set of identified. In the distributed reaction regime, chemical reactions proceed in dis- shadowgraph photographs taken at 40-us intervals of the spark plug electrode tributed reaction zones and thin-sheet flames do not occur. A sufficient condition gap in one cylinder of a 2-liter conventional engine. The first photograph is for this regime is l, < oL. In the reaction sheet regime, propagating reaction between 20 and 50 us after the spark breakdown occurred. The complete fronts are wrinkled and convoluted by the turbulence. A sufficient condition for sequence (~200 us) corresponds to 1.3 crank angle degrees. The outer boundary the existence of reaction sheets is Ix >> 61. For Rer > 1, there is a region in Fig. of this developing flame kernel is approximately spherical and is smooth with 9-18 where I, > 61 > lx: the characteristics of flames in this regime are unclear. modest irregularities, corresponding to a thin reaction zone with high- temperature gases inside. As this developing sheetlike flame grows it interacts with the turbulent flow field in the vicinity of the spark plug: the flame outer 108 surface becomes increasingly convoluted and the flame center can be convecte Weak turbulence 10-2 away from the plug in a direction and with a velocity that can vary substantially cycle-by-cycle, as seen in Fig. 9-20.4, 17 106 S = 1 The structure of the flame continues to develop as it propagates across the chamber. Evidence, largely from schlieren photographs and studies of flame 102 TL Reaction structure with laser diagnostics, shows that early in the burning process the flame 104 sheets Engine regime Damköhler number Da = ; 104 1 102 = - FIGURE 9-18 10-2- 10-2 104 Different turbulent flame regimes on plot Damköhler number versus turbulence Reynolds 2 3 Distributed number, u' is turbulence intensity; S, is lamine FIGURE 9-19 reactions 10- flame speed; l1, lx, and ôr are integral scale, Kof Shadowgraph photographs of spark-generated kernel between the spark plug electrodes. First pho- 1 102 10 106 108 mogorov scale, and laminar flame thickne tograph on left, 20 to 50 us after breakdown; 40 us between photos. Stoichiometric mixture, 1100 Turbulent Reynolds number Rer (From Abraham et al.22) rev/min. (Courtesy A. Douad, Institut Francais du Petrole.) 398 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 399 FIGURE 9-20 Schlieren photographs of developing flame, 5º after spark discharge, showing different convection processes in two different cycles. Spark plug wires 0.8 mm diameter. 1400 rev/min; 0.5 atm inlet pressure; propane fuel; o = 0.9; spark advance 45º BTC.17 is a thin, moderately wrinkled but simply connected, front or reaction sheet between unburned and burned gas. The thickness of the front is about 0.1 mm which is comparable to the thickness of a laminar flame under the prevailing conditions. The scale of the wrinkles is typically about 2 mm at engine speeds of 1000 to 2000 rev/min. As the flame propagates across the chamber, the thickness of the reaction sheet front remains roughly constant, the flame front becomes (a) more convoluted, and the scale of the wrinkles tends to decrease with time.24 (b) Further evidence that the thin reaction sheet front becomes highly wrinkled FIGURE 9-22 and convoluted by the turbulent flow field into a thick turbulent flame "brush" Enlarged schlieren photographs of (a) flame front and (b) flame back in square-cross-section cylinder is provided by the schlieren photographs in Fig. 9-21. These show a flame propa. engine with two glass side walls. 1400 rev/min, 0.5 atm inlet pressure, propane fuel, ( = 0.9.17 gating across the combustion chamber of a square-cross-section single-cylinder engine with a special transparent piston crown containing a 13-mm wide channel to isolate a small section of the flame.25 The flame on the left is for a propane-air burned mixture exists-is apparent. It was 4 to 5 mm for propane and 1.5 mm for hydrogen. The difference is due to the substantially higher laminar flame mixture; that on the right is hydrogen-air. The energy density per unit volume of mixture and the flow field are comparable for each fuel. The effective thickness of speed for the hydrogen, air, burned gas mixture (see Sec. 9.3.4) which increases this turbulent flame-the average distance between the region ahead of the flame, the Damköhler number and shifts the flame toward the weak-turbulence (i.e ., less wrinkled) flame regime in Fig. 9-18. where only unburned mixture exists, and the region behind the flame, where only Additional insight into the structure of the developed engine flame can be obtained by enlarging photographs of the leading and trailing edges of the flame, obtained with the schlieren or shadowgraph technique. Figure 9-22a shows the front of the flame 40º after the spark, when it has propagated about halfway across the chamber. It shows the irregular but smoothly curved surfaces which comprise the leading edge of the flame. Figure 9-22b shows the back of the flame 70º after the spark, when the front of the flame has just reached the wall of the (a) (b) combustion chamber farthest from the spark plug. It shows large clear regions of burned gas behind the flame and smaller clear regions connected by a lacelike FIGURE 9-21 Schlieren photographs of flame in square-cross-section cylinder engine, with narrow channel in piston crown (at bottom of pictures) which permits observation of 13-mm wide section of flame. (a) Prop fuel, spark timing 36º BTC, photograph at 14º ATC, flame thickness ~ 4.6 mm. (b) Hydrogen spark timing 3º BTC, photograph at 10º ATC, flame thickness ~ 1.5 mm. Stoichiometric mixtur * These photographs were selected from a large number to give the minimum flame thickness corre- 1380 rev/min, 0.5 atm inlet pressure.25 ponding to the flame front perpendicular to the channel length. 400 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 401 Intensity ~ density (a) (b) (6) (d) Resolution ->+ Unburned - Laser beam gas FIGURE 9-23 Upper picture: oscillogram of output of optical Height, mm multichannel analyzer showing the intensity of light scattered from a narrow laser beam as a funo- Burned gas tion of distance through the flame. Intensity is a (e) () (8) (h) Laminar flame measure of gas density. Lower picture: schematic of flame structure corresponding to this signal. 600 FIGURE 9-24 30 29 28 27 rev/min, 1 atm inlet pressure, propane fuel, Shadowgraph photographs, with light beam normal to flame front, of 10-mm diameter section of Flame radius, mm $ = 1.0.24 name. Photographs, arranged in order of increasing turbulent Reynolds number, suggest this is an appropriate scaling parameter (see Table 9.1).27 structure within which the remaining regions of unburned mixture are being con- sumed.17 The analogy with a crumpled sheet of paper is appropriate. developed turbulent flames in spark-ignition engines, under normal operating Laser scattering experiments, where "snapshots" of the density profile conditions, are highly wrinkled and probably multiply-connected thin reaction along a laser beam passed through the flame were obtained using Rayleigh scat- sheets. The overall thickness of the turbulent flame "brush," front to back, is of tered light from the gas molecules, provide explicit evidence of this structure.26 order 1 cm. The thickness of the thin reaction sheet is comparable to estimates of Figure 9-23 shows an oscillogram of the output of the optical analyzer. Each dot the laminar flame thickness under the prevailing unburned mixture conditions represents the intensity of the scattered light which is a measure of the gas which are of order 0.1 mm. The scale of the wrinkles is of order 1 mm.24 Direct density. The signal on the left corresponds to unburned gas. The flame is propa- evidence to date is limited to the low to mid engine speed and low to high engine gating from right to left. The oscillogram shows a thin transition zone of width 0.25 mm between unburned and burned gas, followed at a distance of 1.5 mm by an "island" of unburned gas. The fraction of oscillograms showing such TABLE 9.1 "islands" varied from 0 at 300 rev/min to 20 percent at 1800 rev/min. An inter- Parameters for shadowgraph photographs in Fig. 9-24 pretation of this signal consistent with the available photographic evidence is shown underneath. 24 Engine Turbulence The above results suggest that increasing engine speed, which increases speed, intensity, Photograph rev/min turbulence levels in the unburned charge, increasingly convolutes and probably m/s Valve REM Rer multiply-connects the thin reaction sheet flame front. Enlarged schlieren pho- 300 0.44 106 229 tographs of a 9-mm diameter section of the developed engine flame, viewed 600 0.88 S 157 503 normal to the flame surface, indicate that increasing turbulence intensity and 300 1.07 US 173 611 900 decreasing turbulence scales result in increasingly finely wrinkled flame struc- 1.33 193 760 1200 -. 1.80 S tures. Figure 9-24 shows a set of such photographs, arranged in order of increas- 224 1024 600 1.95 US 234 1117 ing turbulent Reynolds number. Increases in turbulence intensity, which was 900 2.90 US 285 1658 measured, were achieved by increases in engine speed and by modifying the inlet 1200 4.0 US 333 2263 valve. Relevant parameters for each photograph are given in Table 9.1. Note: Valve: S, shrouded; US, unshrouded (produced higher turbulence due to The above theoretical discussion and experimental evidence indicates that kas ordered flow). Reg = ly W/v; Ref = 1, N'/v (see Sec. 8.2.1).27 402 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 403 load ranges. Whether the structure becomes significantly different at high engine treated as negligible thin. Laminar burning velocities for methane, propane, iso- speed is not known. Models of this turbulent flame development and propaga- octane, methanol, gasoline, and hydrogen-premixed with air-at pressures, tem- tion process are reviewed in Chap. 14. peratures, and equivalence ratios which occur in engines have been measured using this technique.29-33 Also, the effect of a burned gas diluent on laminar 9.3.3 Laminar Burning Speeds burning velocity with gasoline-air mixtures has been determined.32 Correlations derived from these data are the most accurate means available for estimating An important intrinsic property of a combustible fuel, air, burned gas mixture is laminar burning velocities for mixtures and conditions relevant to spark-ignition its laminar burning velocity. This burning velocity is defined as the velocity, rela- engines. tive to and normal to the flame front, with which unburned gas moves into the The effect of the mixture fuel/air equivalence ratio on laminar burning front and is transformed to products under laminar flow conditions. Some details velocity for several hydrocarbon fuels and methanol is shown in Fig. 9-25. The of flame structure help explain the significance of this quantity. A flame is the burning velocity peaks slightly rich of stoichiometric for all the fuels shown. The result of a self-sustaining chemical reaction occurring within a region of space values for isooctane and gasoline are closely comparable. Data at higher pres- called the flame front where unburned mixture is heated and converted into pro- sures and temperatures have been fitted to a power law of the form: ducts. The flame front consists of two regions: a preheat zone and a reaction zone. In the preheat zone, the temperature of the unburned mixture is raised (9.33) mainly by heat conduction from the reaction zone: no significant reaction or energy release occurs and the temperature gradient is concave upward (02T/ where To = 298 K and po = 1 atm are the reference temperature and pressure, @x2 > 0). Upon reaching a critical temperature, exothermic chemical reaction and SL, o, a, and ß are constants for a given fuel, equivalence ratio, and burned begins. The release of chemical energy as heat results in a zone where the tem- gas diluent fraction. For propane, isooctane, and methanol, these constants can perature gradient is concave downward (02T/2x2 < 0). The region between the be represented by temperature where exothermic chemical reaction begins and the hot boundary at the downstream equilibrium burned gas temperature is called the reaction zone. a = 2.18 - 0.8(¢ - 1) (9.34a) The thicknesses of the preheat and reaction zones can be calculated for one- B = - 0.16 + 0.22(0 - 1) (9.34b) dimensional flames from conservation equations of mass and energy. The thick- and ness of the preheat zone OL, ph is SL, O = Bm + B (Q - Q m)2 (9.35) 4.6k OL, ph (9.31) where om is the equivalence ratio at which SL, o is a maximum with value Bm. C . P . SL where k and c, are the mean thermal conductivity and specific heat at constant 50 pressure in the preheat zone and St is the laminar burning velocity.28 Thus, the factors which govern the laminar burning velocity of a specific unburned 45- mixture the velocity at which this flame structure propagates relative to the Methanol () unburned gas ahead of it-are the temperature and species concentration gra- 40 Propane ( 4) dients within the flame and the mixture transport and thermodynamic properties. Laminar burning velocities at pressures and temperatures typical of 35 Gasoline (0) unburned mixture in engines are usually measured in spherical closed vessels by Burning velocity, cm/s propagating a laminar flame radially outward from the vessel center. The laminar 30 burning velocity is then given by Isooctane dmp/dt 25- ....... ... . SL = (9.32) Methane (v) A, Pu O 20H where the mass burning rate is determined from the rate of pressure rise in the FIGURE 9-25 vessel and A, is the flame area. Because the laminar flame thickness [e.g ., given Laminar burning velocity for several fuels as func- 15 L by Eq. (9.31)] under engine conditions is of order 0.2 mm26 and is therefore much 0.6 tion of equivalence ratio, at 1 atm and 300 K. 0.8 1.0 1.2 1.4 1.6 Lines are least-squares polynomial fits to less than characteristic vessel dimensions, in applying Eq. (9.32) the flame can be Fuel/air equivalence ratio ¢ data, 29, 30 404 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 405 1.0, TABLE 9.2 Pi, atm ₾ 0.7 2 Parameters ¢_, B ., and B. for Eq. A 0.8 2 (9.35) 0.8 1.0 0 1.0 2 B ., 1.2 1.2 2 Fuel cm/s cm/s Ref. 0.6 Methanol 1.11 36.9 -140.5 30 SL(*b) SL(O) Propane 1.08 34.2 -138.7 30. Isooctane 1.13 26.3 -84.7 30 0.4 Gasoline 1.21 30.5 -54.9 32 Note: Values of St o given by Eq. (9.35) are obtained 0.2 from least-squares fits of Eq. (9.33) to data over the range p = 1-8 atm, T. = 300-700 K. They do not correspond exactly to the laminar flame speed data at 1 atm and FIGURE 9-26 298 K in Fig. 9-25. OF 0.1 Effect of burned gas mole fraction %, in unburned 0.2 0.3 Mole fraction diluent &b mixture on laminar burning velocity. Fuel: gas- oline.32 Values of dm, Bm, and Bo are given in Table 9.2.30 For gasoline (a reference gasoline with average molecular weight of 107 and an H/C ratio of 1.69) addi- tional data were available and were correlated by32 The presence of burned gas in the unburned cylinder charge due to residual a = 2.4 - 0.27163.51 (9.36a) gases and any recycled exhaust gases causes a substantial reduction in the laminar burning velocity. Any burned gas in the unburned mixture reduces the B. = - 0.357 + 0.1402.77 (9.36b) heating value per unit mass of mixture and, thus, reduces the adiabatic flame temperature. It acts as a diluent. The effect of increasing burned gas or diluent For methane, simple equations such as (9.34a, b) do not adequately correlate the fraction on laminar flame speed is shown in Fig. 9-26. The diluent used was a data over the range of p and Ty, relevant to engines. However, laminar burning mixture of CO2 and N2, chosen to match the heat capacity of actual gasoline-air velocity data from a spherical constant-volume bomb experiment have been combustion products.+ The proportional reduction in laminar burning velocity is obtained along an unburned gas isentropic path, as the pressure in the bomb essentially independent of the unburned mixture equivalence ratio, pressure, and rises during combustion. Variation in laminar burning velocity along such temperature over the range of interest in engines. The data in Fig. 9-26 are corre- unburned gas isentropes does correlate with a power law: lated by the relation: (Du )‘ SL.S = SL. (Puo): (9.37) SL(x) = SL(X6 = 0)(1 - 2.06x9.77) (9.38) Values for SL, o and e from the literature are summarized in Table 9.3. where × is the mole fraction of burned gas diluent. Other studies corroborate the magnitude of this burned gas effect.32 Note that for equal heat capacity added to the unburned mixture, burned gases have a much larger effect on laminar burning velocity than does excess air. TABLE 9.3 Parameters for methane-air laminar For example, the laminar burning velocity of a stoichiometric mixture as it is burning velocity correlation [Eq. (9.37)] leaned to d = 0.8 is reduced by 23 percent. The excess air required has a heat capacity of about 0.2 times that of the combustion products of the undiluted Stout mixture. Adding the same heat capacity by adding stoichiometric burned gases atm cm/s Ref. (which requires a burned gas mole fraction of 0.175) reduces the laminar burning 1.0 0.5 49 0.51 31 1.0 1.0 35 0.2 31 0.8-1.2 1-8 0.17-0.19 33 + At 298 K initial temperature. The water in actual residual and exhaust gas was omitted. A mixture of 80 percent N2 and 20 percent CO2, by volume, was used. $ See Fig. 9-25. 406 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 407 velocity by 55 percent.32 Proper allowance for the burned gas fraction in estimat Top ing laminar burning velocities for spark-ignition engines is most important. Cylinder wall The above correlations define the laminar burning velocity as a function of unburned mixture thermodynamic properties and composition, only. It has been assumed that flame thickness and curvature effects are negligible.28 Our interest in laminar burning velocity is twofold: first, it is used to define the characteristic chemical reaction time of the mixture in Eq. (9.30); second, a presumed conse- O Side quence of the wrinkled thin-reaction-sheet turbulent-flame structure is that, Flame front locally, the sheet propagates at the laminar burning velocity. The above correla- tions adequately characterize a quiescent burning process. However, laminar 0- TC flame propagation can be influenced by the local flow field in the unburned gas. If the flame thickness is less than the Kolmogorov scale, the primary effect is one of straining which affects both the flame area (usually referred to as flame stretch- ing for an area increase) and the local (laminar) burning velocity. While this problem is not yet well understood, it is known that straining can affect the laminar burning velocity and can cause flame extinction. The laminar burning ,Flame center velocity decreases with increasing strain rate, and the Lewis number of the Angle after spark Flame! unburned mixture has a significant influence on this rate of decrease. The Lewis "front number is the ratio of diffusivities of heat and mass. For stoichiometric mixtures Mean it is close to one; it increases above about unity as the unburned fuel-air mixture particle FIGURE 9-27 is leaned out. Thus the local flow field may have a discernable effect on the local path burning velocity of the thin laminarlike reaction-sheet flame, especially for lean OL Schematic of spherical flame front in engine com- -R Spark 0 R bustion chamber identifying parameters which or dilute mixtures. Distance define flame geometry. (From Beretta et al.4) 9.3.4 Flame Propagation Relations spherical surface within the combustion chamber which would contain all the burned gas behind it; i.e ., If the heat-release or mass burning rate analysis of Sec. 9.2.2 is coupled with an analysis of flame geometry data, substantial additional insight into the behavior V6 ( Tb , Tc , Qc, 2c) = VB(P, 0) (9.39) of spark-ignition engine flames is obtained. Flame photographs (such as those in Figs. 9-1, 9-14, and 9-16 and Refs. 4 and 35) effectively define the position of the The spherical burning area A, is the area of this spherical surface; i.e ., front or leading edge of the turbulent engine flame. The "shadow" of the enflamed zone, under normal engine conditions, is close to circle: only in the Or b (9.40) presence of very high swirl does substantial distortion of the flame shape occur.18 Thus, to a good approximation, the surface which defines the leading edge of the The laminar burning area AL is the surface area the flame would have if it burned turbulent flame (ahead of which only unburned mixture exists) is a portion of the at the laminar flame speed, i.e ., surface of a sphere. Figure 9-27 indicates the geometrical parameters which define dmy/dt this flame surface: rc , de , Zc , the coordinates of the flame center; ry, the radius of AL == PUSL (9.41) the best-fit circle to the flame front silhouette; and the geometry of the com- bustion chamber walls. The flame is initiated at the spark plug; however, it may where Sy is the laminar flame speed in the unburned mixture ahead of the flame move away from the plug during the early stages of its development as shown. (see Sec. 9.3.3). We define the flame front area As as the spherical surface of radius ry coinciding Several velocities can be defined. The mean expansion speed of the front us with the leading edge of the flame contained within the combustion chamber, and is given by the enflamed volume V, as the volume within the chamber behind this flame front. The thermodynamic analysis of cylinder pressure data allows us to define dAs/dt additional geometrical parameters. The burned gas radius r, is the radius of the us Ls (9.42) 408 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 409 where As is the "shadow" area enclosed by the "best-fit" circle through the T T leading edge of the flame and 60 60 @As 50 Ls = = 50 40- 40 is the arc length within the chamber of this "best-fit" circle. The mean expansion speed of the burned gas u, is Flame radius rf, mm Flame radius rf, mm 30 30 - 0V/6/at (9.43) 20 20 Ab 10 10 This derivative is taken with the piston position fixed since only burned volume changes due to combustion are of interest. The burning speed S, is defined by - 30 -20 -10 0 10 20 10 20 30 40 50 60 dm,/dt Crank angle 0, deg 9.44) Burned gas radius ry, mm P. A. (a) (b) The mean gas speed just ahead of the flame front u, is 1g = 1b - Sb (9.45) 16 Note that combining Eqs. (9.41) and (9.44) gives the relation 14 1.0 S. Ar = SL AL (9.46) 12 0.8 5 43 Uf ub Also, it follows from Eqs. (9.29), (9.43), and (9.44) that 10 0.6 yf(rc = 0.5) 8 Speed, m/s Pu (1 - yb) + yb == Pu/Pb 235 = [(P /Pb) - 1]x, + 1 (9.47) 0.4 6 54 Sb 4 Po 2 Up 5 4 S 3 43 642 3 5 ? As x, and y -> 0, ub/S, approaches the expansion ratio pu/pt. As x, and y, - + 1, 4 52 43 ub/St approaches unity. 2 SL (x, = 0) -SL (x, = 0.2) The variation of the above quantities during the engine combustion Piston Near = AL al A Rh O process, coupled with the photographs and discussion in Sec. 9.3.2, provide sub- face wall b 45 5 stantial insight into the flame development and propagation process. Figure 9-28 10 15 2 2 2 525 52 2 54 2 shows results from an analysis of cylinder pressure data and the corresponding 8 AL Ab 4.3 g 32ª 5 5 flame front location information (determined from high-speed movies through a 2 43 6 5 window in the piston) of several individual engine operating cycles. The com- £5 43 f(Vc = 0.5) 4 bustion chamber was a typical wedge design with a bore of 102 mm and a com- 2 pression ratio of 7.86. The flame radius initially grows at a rate that increases with time and exhibits substantial cycle-by-cycle variation in its early develop- 10 20 30 40 50 60 70 10 20 30 40 50 ment (Fig. 9-28a). Later (r _ 30 mm) the growth rate, which approximates the Flame radius rf, mm Flame radius re, mm expansion speed up, reaches an essentially constant value. The flame radius r is (c) (d) initially equal to the burned gas radius r); it increases above r, as the flame grows FIGURE 9-28 and becomes increasingly distorted by the turbulent flow field (Fig. 9-28b). Even- Variation of flame geometry and velocity parameters during four individual combustion cycles at tually ry - r$ goes to an essentially constant value of about 6 mm for r, _ 30 1044 rev/min, ¢ = 0.98, 1 atm inlet pressure: (a) flame radius r, versus crank angle; (b) flame radius r, mm. This difference, r - r$, is approximately half the thickness of the turbulent versus burned gas radius r,; (c) normalized enflamed volume y , burned volume y ,, normalized flame front area a ,, and laminar area a, versus flame radius; (d) front expansion speed up, burning speed flame brush. So, and laminar flame speed S, versus flame radius. (From Beretta et al.4) 410 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 411 3.5 Normalized enflamed and burned volumes, and flame front area and laminar burning area, are shown in Fig. 9-28c. Volumes are normalized by the 3.0 cylinder volume, and areas by zRh, where h is the average clearance height and R UBAY the cylinder radius. Discontinuities occur in the flame area a, at the points where 2.5 the flame front contacts first the piston face and then the near cylinder wall. The Ignition 2.0 laminar area AL is initially close to the flame area A, and then increases rapidly as the flame grows beyond 10 mm in radius. During the rapid burning com- Urms 41.5 rms fluctuation Unns and uf, m/s Mean velocity UEA, m/s bustion phase (ys _ 0.2) the value of ys is significantly greater than yb . During 27 -1.0 this phase, the laminar area exceeds the flame area by almost an order of magni- T tude. These observations indicate the existence of substantial pockets of -0.5 unburned mixture behind the leading edge of the flame.4 0.0 The ratio of the volume of the unburned mixture within the turbulent flame zone (V, - V) to the reaction-sheet area within the flame zone (AL - Af) defines -1L 342 346 350 354 358 -0.5 362 a characteristic length Crank angle, deg Vs - Vo (9.48) FIGURE 9-29 AL - As Laser doppler anemometer measurements of ensemble-averaged mean velocity UFA [Eq. (8.20)], rms which can be thought of as the scale of the pockets of unburned mixture within fluctuation in individual-cycle mean velocity U.m. [Eq. (8.21)] and turbulence intensity u'f [Eq. (8.22)], close to the cylinder axis, from before ignition to after flame arrival. Disc-shaped chamber, spark plug the flame. For the data set of Fig. 9-28, IT is approximately constant and of order in cylinder wall, measurement at x/B = 0.57, B = 76 mm, 300 rev/min. S, = 0.83 m/s.36 1 mm.24 These flame geometry results would be expected from the previous pho- tographic observations of how the flame grows from a small approximately because the mean flow varies cycle-by-cycle, the turbulence is not homogeneous, spherical smooth-surfaced kernal shortly after ignition to a highly wrinkled and the flame motion and shape show substantial cyclic variations. The results in reaction-sheet turbulent flame of substantial overall thickness. Initially, the Fig. 9-29 were taken in a special single-cylinder engine with a disc-shaped com- amount of unburned gas within the enflamed volume is small. During the rapid bustion chamber where the spark plug was located in the cylinder liner. Shown burning phase of the combustion process, however, a significant fraction (some 25 are the ensemble-averaged mean velocity, cyclic variation in mean velocity, and percent; see Fig. 9-2) of the gas entrained into the flame zone is unburned. the turbulence intensity, normal to the front, during the major portion of the The front expansion speed us, burning speed S ,, and laminar flame speed combustion process, close to the chamber center. Sz are shown in Fig. 9-28d. The expansion speed increases as the flame develops The mean velocity normal to the front increases steadily from shortly after to a maximum value that is several times the mean piston speed of 3.1 m/s and is ignition, as the combustion-produced gas expansion displaces unburned mixture comparable to the mean flow velocity through the inlet valve of 18 m/s.24 The toward the wall. It peaks as the flame arrives at the measurement location. The burning speed increases steadily from a value close to the laminar flame speed at cyclic variation in mean velocity and the turbulence intensity normal to the front early times to almost an order of magnitude greater than SL during the rapid remain essentially constant until a few degrees before the flame arrival. These two burning phase. During this rapid burning phase, since (r - r$) is approximately quantities are comparable in magnitude; thus the turbulence intensity is lower constant, the flame front expansion speed and the mean burned gas expansion than the rms fluctuation velocity (in this case by about a factor of 2) (see Sec. speed are essentially equal. The difference between us ~ us and S, is the unburned 8.2.2). Whether the increase in turbulence as the flame approaches is due to rapid gas speed ug just ahead of the flame front. Note that the ratio us/S, (~ub/St) distortion resulting from the compression of the unburned mixture which occurs decreases monotonically from a value equal to the expansion ratio (p./pt) at during combustion or is the result of inadequate resolution of cycle-to-cycle flow spark to unity as the flame approaches the far wall, as required by Eq. (9.47). variations is unclear. Rapidly imposed distortions of a turbulent flow field, such The effect of flame propagation on the flow field in the unburned mixture as those imposed by combustion-produced gas expansion, would lead to an ahead of the flame is important because it is the turbulence just ahead of the increase in vorticity and turbulence intensity. Other studies, e.g ., Ref. 37, indicate flame that determines the local burning velocity. Measurements of mean veloc- there is little or no increase in turbulence intensity ahead of the flame. ities, rms fluctuation velocities, and turbulence intensities have been made using The variation of burning speed with engine speed has also been carefully laser doppler anemometry (see Sec. 8.2.2) at different locations within engine examined in a study where flame position was determined from high-speed combustion chambers (e.g ., Refs. 36 and 37). Such data are difficult to interpret movies, mass burning rates from cylinder pressure, and turbulence information 412 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 413 from experiments at equivalent motored conditions.55 The turbulence quantity 9.4 CYCLIC VARIATIONS IN obtained during motoring experiments was the ensemble-averaged root mean COMBUSTION, PARTIAL BURNING, square velocity fluctuation defined by Eq. (8.18). Values of Sp/SL and ur/SL were AND MISFIRE determined at two points in the combustion process: at a flame radius of 30 mm (the end of the flame development process) and at mass fraction burned equal to 9.4.1 Observations and Definitions 0.5 (halfway through the rapid burning phase). To correct the motored turbu- lence data for the higher pressure levels corresponding to engine firing condi- Observation of cylinder pressure versus time measurements from a spark-ignition tions, a simple rapid distortion model (see Sec. 14.4.2) based on conservation of engine, for successive operating cycles, shows that substantial variations on a angular momentum in turbulent eddies was used. A linear correlation between S, cycle-by-cycle basis exist. Since the pressure development is uniquely related to and u's results, as shown in Fig. 9-30, for the rapid burning combustion phase. the combustion process, substantial variations in the combustion process on a Note that as u'p/SL goes to zero, Sp/SL approaches a value close to unity. cycle-by-cycle basis are occurring. In addition to these variations in each individ- Once the flame front reaches the far cylinder wall (see Fig. 9-27) the front ual cylinder, there can be significant differences in the combustion process and can no longer propagate: however, combustion continues behind the front until pressure development between the cylinders in a multicylinder engine. Cyclic all the unburned mixture entrained into the enflamed region is consumed. This variations in the combustion process are caused by variations in mixture motion final burning or termination phase of the combustion process can be approx- within the cylinder at the time of spark cycle-by-cycle, variations in the amounts imated by an exponential decay in the mass burning rate with a characteristic of air and fuel fed to the cylinder each cycle, and variations in the mixing of fresh time constant t, of order 1 ms. Since these "islands" or "pockets" of unburned mixture and residual gases within the cylinder each cycle, especially in the vicin- mixture behind the leading edge of the flame have a characteristic scale lr based ity of the spark plug. Variations between cylinders are caused by differences in on the laminar flame area [Eqs. (9.41) and (9.48)], it follows that these same phenomena, cylinder-to-cylinder. Cycle-by-cycle variations in the combustion process are important for two l = T, SL (9.49) reasons. First, since the optimum spark timing is set for the "average" cycle, In summary, the above flame data analysis procedures show that the faster-than-average cycles have effectively overadvanced spark-timing and relationships between r and ro, V, and V ,, Af and AL, us, S, and SL are dis- slower-than-average cycles have retarded timing, so losses in power and efficiency tinctly different in the three phases of combustion: (1) the development phase, result. Second, it is the extremes of the cyclic variations that limit engine oper- where a highly wrinkled reaction-sheet "thick "-overall turbulent flame evolves ation. The fastest burning cycles with their overadvanced spark timing are most from the essentially spherical flame kernal established by the spark discharge; (2) likely to knock. Thus, the fastest burning cycles determine the engine's fuel the rapid-burning phase, where this thick "developed" turbulent flame propa- octane requirement and limit its compression ratio (see Sec. 9.6.3). The slowest gates across the combustion chamber to the far wall, during which most of the burning cycles, which are retarded relative to optimum timing, are most likely to mass is burned; and (3) the termination phase after the flame front has reached burn incompletely. Thus these cycles set the practical lean operating limit of the the far wall and propagation of the front is no longer possible, when the remain- engine or limit the amount of exhaust gas recycle (used for NO emissions control) ing unburned mixture within the flame burns up. The burning velocity, in the which the engine will tolerate. Due to cycle-by-cycle variations, the spark timing rapid-burning phase of the combustion process, scales with turbulence intensity, and average air/fuel ratio must always be compromises, which are not necessarily which in turn scales with engine speed. the optimum for the average cylinder combustion process. Variations in cylinder pressure have been shown to correlate with variations in brake torque which directly relate to vehicle driveability. 35 An example of the cycle-by-cycle variations in cylinder pressure and the 30 variations in mixture burning rate that cause them are shown in Fig. 9-31. Pres- 25 sure and gross heat-release rate [calculated from the cylinder pressure using Eq. (9.27)] for several successive cycles at a mid-load, mid-speed point are shown as a 20 function of crank angle. The maximum heat-release rate and the duration of the 15 FIGURE 9-30 heat release or burning process vary by a factor of two from the slowest to the Variation of burning speed with turbulence 10 -- fastest burning cycle shown. The peak cylinder pressure varies accordingly. The intensity. The ensemble-averaged rms velocity fluctuation was measured during motoring faster burning cycles have substantially higher values of maximum pressure than engine operation. The ratio p/p. (firing pressure/ do the slower burning cycles; with the faster burning cycles peak pressure occurs S 10 15 20 motoring pressure) corrects for the effect of addi- closer to TC. 0.75 tional compression on the turbulence intensity. The heat-release rate data in Fig. 9-31 show that there are cycle-by-cycle Range of engine speeds and spark timings.35 variations in the early stages of flame development (from zero to a few percent of 414 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 415 40 the point of efficiency, hydrocarbon emissions, torque variations, and roughness. 35 The partial-burn and misfire regimes are discussed in Sec. 9.4.3. Various measures of cycle-by-cycle combustion variability are used. It can 30 be defined in terms of variations in the cylinder pressure between different cycles, 25 or in terms of variations in the details of the burning process which cause the differences in pressure. The following quantities have been used: Pressure, atm 20 15 1. Pressure-related parameters. The maximum cylinder pressure Pmax; the crank angle at which this maximum pressure occurs 0pm.; the maximum rate of 10 pressure rise (dp/d0)mar; the crank angle at which (dp/d0)mm occurs; the indi- cated mean effective pressure [which equals , p dV/V4, see Eqs. (2.14), (2.15), and (2.19)]. 0 - 150 - 100 -50 0 50 100 150 2. Burn-rate-related parameters. The maximum heat-release rate (net or gross, see Sec. 9.2.2); the maximum mass burning rate; the flame development angle A0. 70 and the rapid burning angle, Al, (see Sec. 9.2.3). 3. Flame front position parameters. Flame radius, flame front area, enflamed or 60 burned volume, all at given times; flame arrival time at given locations. 50 Pressure-related quantities are easiest to determine; however, the relation 40 between variations in combustion rate and variations in cylinder pressure is dech , J/deg 30 complex.38 Equation (9.26) defines the factors that govern this relationship. Because the rate of change of pressure is substantially affected by the rate of de 20 change of cylinder volume as well as rate of burning, changes in the phasing of 10 the combustion process relative to TC (e.g ., which result from changes in flame development angle) as well as changes in the shape and magnitude of the heat- 0 release rate profile affect the pressure. Figure 9-32 illustrates how the magnitude - 10 of the maximum cylinder pressure pmar and the crank angle at which it occurs - 150 - 100 -50 0 50 100 150 Oom vary as the crank angle at which combustion effectively starts (e.g ., 0 at Degrees ATC which 1 percent of the cylinder mass has burned) and the burning rate are varied. FIGURE 9-31 Curve CABE shows how Pmax and 0pmar vary for a fixed fast-burning heat-release Measured cylinder pressure and calculated gross heat-release rate for ten cycles in single-cylinder profile (the duration of the heat-release process and its maximum value are held spark-ignition engine operating at 1500 rev/min, ( = 1.0, Pintet = 0.7 atm, MBT timing 25º BTC.14 the total heat release) and in the major portion of the combustion process-the Fast burn rapid-burning phase-indicated by the variations in the maximum burning rate. As the mixture becomes leaner with excess air or more dilute with a higher burned gas fraction from residual gases or exhaust gas recycle, the magnitude of Pmax cycle-by-cycle combustion variations increases. Eventually, some cycles become sufficiently slow burning that combustion is not completed by the time the 50% mass -2 ... burned lines B' exhaust opens: a regime where partial burning occurs in a fraction of the cycles is D' Retard E FIGURE 9-32 encountered. For even leaner or more dilute mixtures, the misfire limit is reached. E' Slow burn-2___ Schematic of variation in maximum cylinder pressure At this point, the mixture in a fraction of the cycles fails to ignite. While spark- and crank angle at which it occurs, in individual cycles. CABE typical of fast heat-release process; ignition engines will continue to operate with a small percentage of the cycles in TC C'A'B'D'E' typical of slow heat-release process. (From the partial-burn or misfire regimes, such operation is obviously undesirable from OP max Matekunas.38) 416 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 417 constant), as the phasing of this combustion process relative to TC is varied. A must be done with care. The location of maximum pressure 0pm ., also depends on corresponds to MBT timing where the start of combustion is phased to give relative phasing of combustion and on the burn rate profile. In addition, for maximum brake torque, B corresponds to retarded timing, and C to over- slow-burning chambers and retarded timing (around E') variations produce little advanced timing. C'A'B'D'E' is a similar curve for a slow-burning heat-release change in Opmax . However, for fast-burning chambers, with MBT or only slightly profile. A' corresponds to MBT timing, and B' and D' to increasingly retarded retarded timing, the location of peak pressure depends essentially on the phasing timing. Note that with a sufficiently slow-burning heat-release profile, beyond D', of each combustion process relative to its MBT phasing, and is independent of pmax decreases as the burn process is increasingly retarded. This occurs when the charging variations. For these reasons, 0pmax is a useful measure of variability in rate of increase of pressure due to combustion becomes so low that it is more combustion event phasing. 38 than offset by the pressure decrease due to volume increase: eventually for One important measure of cyclic variability, derived from pressure data, is extremely slow and late burning, the maximum pressure approaches the motored the coefficient of variation in indicated mean effective pressure. It is the standard pressure at TC. The dashed lines show the constant start-of-combustion timing deviation in imep divided by the mean imep, and is usually expressed in percent: relative to MBT for each burn rate curve. Note that 0, pmax for constant relative timing varies little as the heat-release profile or burn rate varies. 38 COVimep Timer x 100 =im (9.50) We can now explain the effects of variations in the heat-release profile (both imep in the development stage of the burning process, which effectively changes the It defines the cyclic variability in indicated work per cycle, and it has been found location of the start of combustion, and in the rate of burning throughout the that vehicle driveability problems usually result when COVime, exceeds about 10 process) on pmax and 0,mar, when the spark timing occurs at a fixed crank angle. percent. For a fixed burning rate profile (duration of burn and maximum burning rate) as Figure 9-33 illustrates the relationships between Pmar, 0pmax, and imep for the start of combustion is delayed to be closer to TC, Pmax decreases and 0, 120 cycles of an engine cylinder at fixed operating conditions and three different pmax initially increases (A to B or A' to B'). This is the effect of a change in relative spark timings.38 The MBT timing data show a spread in imep at a fixed value of timing or phase of the burning process due to a slower initial rate of flame devel- Opmax . This imep data band is relatively flat and is centered around 0pmax ~ 160; opment with fixed spark timing. If, in addition to the flame development being only at later values of 0pmax does imep fall off significantly. The vertical spread in slower, the heat-release rate throughout the burning process is lower, then that imep around Opmax = 16º is due to variations in the amount of fuel entering the combustion process is even more retarded from the optimum and Pmar decreases cylinder each cycle; normal variations in the burn profile under these conditions, and 0pm increases further, to their values at D'. The effect of a faster initial flame which effectively change the phasing of the combustion process, produce only development and faster burning rate, with fixed spark time, is the opposite. The modest reductions in imep. For early 0pmax (the extreme upper left of Fig. 9-33a), magnitudes of the changes in Pmar and 0pmar depend, obviously, on the extent of the variations in Pmaz are also due mainly to these fuel-charging variations, cycle- the cyclic variations; they also depend on whether the average burn process is by-cycle; these are the fastest burning cycles with the most advanced phasing. As fast or slow. For fast-burning engines, a larger fraction of the heat release occurs near TC when the chamber volume is changing relatively slowly. Thus pressure variations are mainly due to combustion variations. With slow-burning engines, where a significant fraction of the energy release occurs well after TC, the effect of 1600 + MBT 360 volume change also becomes significant and augments the effect of combustion + MBT 1400 . MBT - 10º 340 . MBT - 10º variations. For large variations and a slow average burning process, Pmar can fall . MBT - 20º · MBT - 20º below E', and 0,max then decreases. A fast-burning combustion process signifi- 1200 3201 cantly reduces the impact of cyclic combustion variations on engine per- Maximum pressure, kPa Fast burn imep, kPa 300 1000 formance. 39 Slow burn 280- We can now evaluate the various measures of combustion variability. The 800 260 maximum pressure variation has been shown to depend on both changes in 600 240 phasing and burning rate. The magnitude of this variation depends on whether 10 15 20 25 30 35 10 15 20 25 30 35 the combustion chamber is faster or slower burning, on average. It also depends @pen, deg ATC Opww, deg ATC on whether the burning process is substantially retarded relative to MBT. It (a) (b) depends, too, on cyclic cylinder fuel and air charging variations. Thus the inter- FIGURE 9-33 pretation of variations in pmx [or in the maximum rate of pressure rise (a) Individual-cycle maximum pressure versus crank angle at which p ... occurs. (b) Individual-cycle (dp/d0)[x] in terms of variations in the rate and phasing of the burning process indicated mean effective pressure versus 0 38 pmax 418 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 419 pmax Increases, the dispersion increases as cyclic variations in phasing and 9.4.2 Causes of Cycle-by-Cycle and burning rate have increasing impact. Cylinder-to-Cylinder Variations An important issue is whether variations in the early stages of flame devel- opment and variations in subsequent portions of the burning process are inde- Cycle-by-cycle combustion variations are evident from the beginning of the com- pendent of each other or are correlated. Plots of early flame development angle bustion process. Analysis of flame photographs from many engine cycles taken in (spark to 1 percent mass burned) against the burning angle (1 to 90 percent special research engines with windows in the combustion chamber has shown burned) from individual cycles for several different combustion chambers indicate that dispersion in the fraction of the combustion chamber volume inflamed is the following. There is a trend with increasing flame development angle for the present from the start of combustion (e.g ., see Refs. 3 and 23). Dispersion in burning angle to increase (or the burning rate to decrease); however, there is burning rate is also evident throughout the combustion process (see Figs. 9-2 and much scatter about this trend (for a given value of flame development angle 9-31). Three factors have been found to influence this dispersion:45 different cycles show a substantial range in burning angle), and the quantitative aspects of the trend depend on operating conditions and on combustion chamber 1. The variation in gas motion in the cylinder during combustion, cycle-by-cycle design. In addition, as the mean rapid burning angle increases (due to changing 2. The variation in the amounts of fuel, air, and recycled exhaust gas supplied to operating conditions or a slower-burning chamber design) the mean flame devel- a given cylinder each cycle opment angle, the cyclic variation in the flame development angle, and the cyclic 3. Variations in mixture composition within the cylinder each cycle -especially variation in the rapid burning angle all increase.4º This topic is discussed more near the spark plug-due to variations in mixing between air, fuel, recycled fully in the following section. exhaust gas, and residual gas The shapes of the frequency distributions in individual-cycle pressure data (e.g ., in Pmax, (dp/d0)max , 8pmax) and in burn rate data such as A04, 40 ,, (dQ/d0)m. The relative importance of these factors is not yet fully defined, and depends depend on whether the combustion process is fast and "robust" (e.g ., with close- on engine design and operating variables. The variation in the velocity field to-stoichiometric mixtures at higher loads at optimum timing-well away from within the engine cylinder throughout the cycle, and from one cycle to the next, the lean operating limit of the engine) or slower and less repeatable, closer to the has been reviewed in Sec. 8.2.2. Toward the end of the compression stroke, the lean or dilute-mixture operating limit. Under robust combustion conditions these ensemble-averaged rms velocity fluctuation is of comparable magnitude to the distributions are close to normal distributions.41 43 When the combustion mean piston speed, and may be larger than the mean flow velocity if there is no process is much slower, the cyclic variability becomes large and the distribution strongly directed local mean flow pattern (see Figs. 8-8 and 8-9). This ensemble- becomes skewed toward the slower burning cycles which have low imep (due to averaged velocity fluctuation combines both cycle-by-cycle variation in the mean the substantial retard of these slower cycles). When partial burning and then flow and the turbulent velocity fluctuations. During compression, these two com- misfire occur, the low-pressure tail of the distribution approaches the motored ponents are of comparable magnitude (see Figs. 8-9 and 9-29). While this data pressure value at TC.44 Examples of the frequency distributions of imep in these base is limited, it indicates that substantial variations in the mean flow exist, two combustion variability regimes are shown later in Fig. 9-36. cycle-by-cycle, both in the vicinity of the spark plug and throughout the com- Cylinder pressure data are often averaged over many cycles to obtain the bustion chamber. Velocity variations contribute in a major way to variations in mean cylinder pressure at each crank angle. The primary use of this average the initial motion of the flame center as it grows from the kernel established by pressure versus crank angle data is in calculating the average indicated mean the spark, and in the initial rate of growth of the flame; they can also affect the effective pressure (which is a linear function of p). Since combustion parameters burning rate once the flame has developed to fill a substantial fraction of the are not linearly related to the cylinder pressure [see Eq. (9-27)], analysis of the combustion chamber. Variations in gas motion near the spark plug convect the average pressure data will not necessarily yield accurate values of average com- flame in its early stages in different directions and at different velocities, cycle-by- bustion parameters. The error will be most significant when the combustion cycle. This affects the flame's interaction with the cylinder walls, changing the variability is largest. It is best to determine mean combustion parameters by flame area development with time. Variations in the turbulent velocity fluctua- averaging their values obtained from a substantial number of individual cycle tions near the spark plug will result in variations in the rate at which the small analysis results. The number of cycles which must be averaged to obtain the initially laminarlike flame kernel develops into a turbulent flame. Variations in desired accuracy depends on the extent of the combustion variability. For the mean flow throughout the chamber will produce differences in flame front example, while 40 to 100 cycles may define imep to within a few percent when shape; also, they may produce differences in turbulence which affect the propaga- combustion is highly repeatable, several hundred cycles of data may be required tion velocity of the front (see Fig. 9-30 for the relation between mean burning when cyclic combustion variations are large.12 speed and turbulence intensity). 420 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 421 It is well known that, on a time-averaged basis, the fuel, air, and recycled 0r exhaust gas flows into each cylinder of a multicylinder engine are not identical. 5 These flow rate differences are typically a few percent (see Secs. 7.6.2 and 7.6.3). It 10- is also known that the flow patterns within the different cylinders are not neces- sarily identical due to differences between the individual intake manifold runner and port geometries in many production engines. All these factors contribute to 20 cylinder-to-cylinder variations in the combustion process: there can be significant Cycle number CO2 concentration, % differences in the mean burn rate parameters as well as in the cyclic variations in 30- these parameters.41 Also, the limited data available on the variation in mixture composition within each cylinder for each cycle indicates that cyclic charging 40 variations in individual cylinders are comparable in magnitude to cylinder-to- 3- cylinder differences (i.e ., of order +5 percent46). Whether the amount of residual gas left in the cylinder varies significantly, cycle-by-cycle, is not known. At higher 50 16 17 19 20 9 10 loads, where the combustion process is more repeatable (and always completed 11 relatively early in the expansion stroke) and the residual gas fraction smaller (see HC concentration, 104 ppm C ( a ) (b) Sec. 6.4), variations in the total amount of residual are not expected to be signifi- cant. At light loads (particularly at idle), where combustion variability is much FIGURE 9-34 higher and partial-burning cycles may occur, and especially with high valve (a) Air/fuel ratio in 50 consecutive cycles, in vicinity of spark plug, measured just after ignition with a overlap engine designs, variations in the residual gas mass and its composition rapid-acting sampling valve located in the plug center electrode. Engine operated at 1400 rev/min, MBT timing, imep = 314 kPa.47 (b) CO2 and unburned HC concentrations in gas sampled in individ- may become important. ual cycles from the vicinity of the spark plug just prior to ignition. Engine operated at 1200 rev/min, In addition, mixing of fuel, air, recycled exhaust, and residual is not com- ¢ = 0.98, Pinlet = 0.5 atm, gasoline fuel.48 plete: nonuniformities in composition exist within the cylinder at the start of combustion. Composition variations, cycle-by-cycle, in the vicinity of the spark plug electrode gap will affect the early stages of flame development, especially as gas were removed in turn as contributors to cycle-by-cycle variations (by com- the flame grows through the laminarlike burning phase following the creation of paring premixed propane operation with conventional carbureted operation with a small flame kernel by the spark discharge (see Sec. 9.5.1). Figure 9-34 indicates gasoline, and by removing residual gas by purging with nonfiring cycles), showed the extent of these composition nonuniformities. The available data comes from that the three contributing factors to cyclic combustion variations-velocity experiments where a small rapid-acting sampling valve located in the spark plug variations, fuel/air ratio variations, and residual gas mixing variations-are of center electrode was used to extract gas from the vicinity of the electrode gap, comparable importance at road-load conditions.45 close to the start of combustion, for individual cycle composition analysis. Figure An explanation for cycle-by-cycle variations can be developed from the 9-34a shows the cycle-by-cycle air/fuel ratio fluctuations in the burned gases description of the turbulent flame propagation process in Sec. 9.3. Conditions in -- sampled from one cylinder of a four-cylinder gasoline-fueled carbureted engine the vicinity of the spark plug will influence the initial stages of the flame propaga- just after combustion has started. The standard deviation was typically 2 to 6 tion process-establishing a stable kernel and its development into a turbulent percent of the mean (A/F).46-48 Figure 9-34b shows the relationship between flame. During the developed flame propagation phase, the average conditions in total hydrocarbon and CO2 concentrations in unburned mixture, sampled just the bulk gas within the combustion chamber will be the determining factors since before spark discharge. The CO2 concentration is a measure of the burned gas the flame front spans the chamber, effectively averaging out local non- fraction in the sampled unburned mixture; hence on average it correlates uniformities. By conditions are meant the turbulent velocity fluctuations and inversely with the total hydrocarbon concentration. However, there is substantial length scales in the flow, proportions of fuel, air, and burned gas in the mixture, and the mixture state. fluctuation in CO2 concentration about the mean value, at a given fuel fraction, indicating significant fluctuations, cycle-by-cycle, in the mixing of fresh mixture Using the turbulent combustion model described in Sec. 14.4.2 (which is with residual gas. Nonuniformities in EGR distribution between cylinders and based on the description of the flame development and propagation process in EGR mixing within the cylinder would also increase the variations in burned gas Sec. 9.3), the flame development angle 404 (the time to burn a few large eddies fraction locally at the spark gap, cycle-by-cycle.48 and establish a developed turbulent flame) can be expressed as20 Experiments in a multicylinder production SI engine, where the fuel/air ratio nonuniformities and the nonuniform mixing of fresh mixture with residual 40%= c(4)(M ) (9.51) 422 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 423 The turbulent flow field influences 40, through l1, u', and IM, the integral scale, cycles being retarded and faster developing and/or burning cycles being over- the turbulence intensity, and the microscale, respectively. The mixture composi- advanced. The curve of torque versus combustion timing (relative to optimum tion influences A04 through the laminar flame speed S1. There is therefore a timing), Fig. 9-3b, is almost independent of the burning rate; i.e ., a given magni- variability in the flame development period, since all of these quantities can vary tude retard (of say 10º) relative to optimum timing gives almost the same in the vicinity of the spark plug on a cycle-by-cycle basis. reduction in torque for a very fast burn as it does for a very slow burn. This is The pressure development during the rapid-burning developed turbulent because the burning process, for optimum timing, is centered at about 10º ATC flame propagation phase, when the flame spans the combustion chamber, independent of the burn rate, and retard or advance shifts this "center" by equal depends on the average rate of burning in the flame. Thus, variations in the amounts for all burn rates.16 One of the major advantages of fast-burn engines is turbulent flow field and mixture composition across the gas entering the flame now apparent. The magnitude of the variations in the flame development process front are averaged out and are not important. However, variation in chamber- and subsequent flame propagation rate are decreased as the burning rate is average quantities are significant. The combustion model in Sec. 14.4.2 leads to increased (see Fig. 9-35): the ratio of standard deviation in 40, and 40, to the the following expression for the maximum burning rate: mean values remains approximately constant. Thus, these smaller combustion Cm,(h*/BXp#/p.)10/9[(u'*St)/h.]2/3 variations in fast-burn engines, which correspond to modest retard and over- (9.52) advance in nonaverage burn rate cycles, have little effect on torque. In contrast, dt max v*1/3 the larger combustion variations of slow-burning engines result in significant Here, m is the mass of fuel in the chamber, h is the instantaneous (mean) clear- cyclic torque variations. ance height, B is the bore, p the density, ü' the average turbulence intensity across the flame front, and v the kinematic viscosity; the * denotes the value at the time of the maximum burning rate and the subscript i the value at spark; C is a 60 Data bounds constant depending on engine design. It can be seen from Eq. (9.52) that cycle-by- 50 cycle variations in the maximum burning rate can result from variations in the 40 overall flow pattern within the combustion chamber (which vary u'*) and from 40d, deg variations in the amount of fuel (m ) that enters the cylinder each cycle.¡ Also, it 20- can be seen that variations in the flame development process will result in varia- tions in the maximum burning rate because the crank angle at which the 10- maximum burning rate occurs is shifted, and all the starred parameters in Eq. 0 20 30 40 50 60 70 80 90 100 110 (9.52) will have different values. From the discussion of flame development and structure in Sec. 9.3, and Eqs. (9.51) and (9.52) above, we would expect that mixture conditions and motion leading to slower flame development rates (longer flame development angles, A0)-lower turbulence intensities and more dilute mixtures-would also give º49; deg 4- lower burning rates (longer rapid burning angles, A0,). Data from many different engines and a wide range of operating cases show that this is the case, on 2 average, though there is substantial variation about the mean trend. Figure 9-35 OL shows these trends; it also shows that the standard deviation of 40, and the 20 30 40 50 60 70 80 90 100 110 standard deviation of A0, for a given chamber and operating condition generally 25 increase as the average burning process becomes slower.4º 20 One final factor of importance is how variations in flame development and burning rate affect engine torque. With fixed spark timing, such variations in the 15 FIGURE 9-35 Das, deg combustion process cycle-by-cycle result in slower developing and/or burning Variation in mean value of flame development angle 10 40, (spark to 1 percent mass burned) and standard deviations of flame development angle and rapid burning angle A0, (1 to 90 percent mass burned) 0 with mean rapid burning angle 40 ,. Range of com- Variations in the total amount of air, recycled exhaust gas, and residual in the chamber could also, 20 30 40 50 60 70 80 90 100 110 bustion chamber geometries and engine operating for some operating regimes, be significant. 40 ,, deg conditions.40 424 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 425 9.4.3 Partial Burning, Misfire, and Engine Stability 20 As the unburned mixture in a spark-ignition engine is leaned out with excess air COVimeps % 10 or is diluted with increasing amounts of burned residual gas and exhaust gas Unstable zone recycle, the flame development period, the duration of the rapid burning phase, 80 and the cycle-by-cycle fluctuations in the combustion process all increase. Even- EGR, % ₲ 6000 tually a point is reached where engine operation becomes rough and unstable, 50 HC, ppm C 000 and hydrocarbon emissions increase rapidly. The point at which these phenom- 0 20 ena occur effectively defines the engine's stable operating limit.+ These phenomena 2000- . 28 Frequency, % 40 result from the lengthening of all stages of the combustion process as the 10} unburned mixture is diluted. With increasing dilution, first a fraction of the cycles 20 Normal burn Slow burns so slowly that combustion is only just completed prior to exhaust valve burn Frequency, % 5 Partial burn opening. Then as burning lengthens further, in some cycles there is insufficient Misfire time to complete combustion within the cylinder; also, flame extinguishment 100 200 300 100 500 10 20 30 before the exhaust valve opens and before the flame has propagated across the imep, kPa EGR rate, % chamber may start to occur in some cycles. Finally, misfiring cycles where the (a) (b) mixture never ignites may start to occur. The proportion of partial burning or FIGURE 9-36 nonburning cycles increases rapidly if the mixture is made even more lean or (a) Frequency distributions in indicated mean effective pressure at different EGR rates; 0 percent gave dilute, and the point is soon reached where the engine will not run at all. excellent engine stability, 20 percent acceptable stability, and 28 percent poor stability. (b) Coefficient The impact on engine stability of increasing combustion variability, due to of variation in imep, HC emissions, and percentage of normal, slow, partial-burn, and misfire cycles. increased exhaust gas recycle at part-load, is shown in Fig. 9-36. Figure 9-36a Engine conditions: 1400 rev/min, ¢ = 1.0, MBT timing, imep = 324 kPa.49 shows the distributions of individual-cycle indicated mean effective pressure values for 0, 20, and 28 percent EGR. Without EGR at these conditions, the Partial-burn-limited spark timing or the partial-burn limit. The spark timing spread in imep is narrow. Increasing EGR widens the distribution significantly (retarded from MBT) at which incomplete flame propagation occurs at a given and cycles with low imep, and eventually zero imep, occur. Figure 9-36b shows mixture composition in a given small percent of the cycles (again, this frequency is how the coefficient of variation of imep and hydrocarbon emissions increase as selected arbitrarily for experimental convenience). EGR is increased. Slow burn, then partial burn, and then misfire cycles occur with increasing frequency. In the slow-burn cycles, combustion is complete but Lean misfire limit at MBT spark. The leanest mixture stoichiometry at which the ends after 80º ATC and the indicated mean effective pressure is low (between 85 engine could be stabilized to operate at MBT spark timing with a misfire frequency and 46 percent of the mean value). Imep in partial-burn cycles was less than 46 below a specified value (again, this frequency, usually a percent or less of the cycles, percent of the mean. In misfiring cycles, imep < 0. Empirically, it has been found is selected arbitrarily for convenience). A dilute misfire limit, the maximum amount that COVimep [see Eq. (9.50)] is about 10 percent at the engine's stable operating of exhaust gas recycle that can be absorbed at a given stoichiometry for stable limit, which here occurs just before the onset of partial-burning cycles49. engine operation, can be similarly defined. An explanation of combustion phenomenon at the engine stable operating limit has been developed by Quader.5º It involves the following terms: Engine experiments have defined the locations of the ignition limit line and the partial-burn limit line; they are shown qualitatively in Fig. 9-37. At a given Ignition-limited spark timing or the ignition limit. The spark timing [advanced from spark timing, on this spark timing versus equivalence ratio plot, progressive maximum brake torque (MBT) timing] at which misfire (i.e ., failure of flame leaning of the mixture fed to the engine will lead to the onset of misfire or to the initiation) first occurs at a given mixture composition, in a given small but arbitrary onset of partial burning, depending on the location of the lines and the spark fraction of cycles (e.g ., 0.5 to 1 percent). timing selected. The individual figures show the possible interactions of the maximum brake torque (MBT) timing line-leaner mixtures require greater advance-with the ignition limit and partial-burn limit lines. At MBT timing, the partial-burn limit may or may not be reached prior to misfire or the ignition + This limit has often been called the lean operating limit. Since what limits engine operation in practice is excessive torque fluctuations, cycle-by-cycle, and high hydrocarbon emissions, resulting limit. It will depend on the engine and ignition system design and operating from the use of mixtures made overly dilute with either air or burned gases (or with both), stable conditions. For spark timings retarded relative to MBT, partial burning and not operating limit is a more appropriate term. failure of flame initiation is the primary cause of unstable engine operation. 426 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 427 agation following successful ignition is usually the factor which limits engine Ignition Ignition, limit limit limit operation with dilute mixtures. Experiments have shown that a limited interval, -Lean limit -Lean limit of order 80 crank angle degrees (depending on engine geometry and spark plug Spark advance location), is available during the engine cycle when conditions are favorable for MBT MBT Partial MBT Partial complete flame propagation.49. 51 Outside of this interval, the mixture pressure Partial burn limit burn limit burn limit and temperature are too low, and the turbulence intensity is too low to sustain a sufficiently rapid rate of combustion. Thus, it is factors which increase the flame Lean Rich Lean Rich Lean Rich development and propagation rates which primarily extend the partial-burn Fuel/air equivalence ratio limit. FIGURE 9-37 Schematics of three possible combinations of ignition limit, partial-burn limit, and MBT timing 9.5 SPARK IGNITION curves as function of fuel/air equivalence ratio. (From Quader.50) In spark-ignition engines, the electrical discharge produced between the spark plug electrodes by the ignition system starts the combustion process close to the An example of these limiting combustion regimes for lean engine operation end of the compression stroke. The high-temperature plasma kernel created by is shown in Fig. 9-38. With MBT timing, as the mixture is leaned out (at constant the spark develops into a self-sustaining and propagating flame front-a thin air flow rate), complete combustion in all cycles changes to partial burning in reaction sheet where the exothermic combustion chemical reactions occur. The some cycles which changes to no ignition in some cycles at the ignition limit. In function of the ignition system is to initiate this flame propagation process, in a the partial-burning regime, the most common type of incomplete-combustion repeatable manner cycle-by-cycle, over the full load and speed range of the engine cycle was a slow-burning cycle which required more time to complete burning at the appropriate point in the engine cycle. Shadowgraph and schlieren pho- than was available: flame extinguishment during expansion was much less tographs of the kernel created by the discharge between the plug electrodes, the common. Engine performance measurements showed that the engine stability growth of that kernal, and its transition to a propagating flame have already limit-evidenced by minimum fuel consumption and onset of rapid increase in been presented in Figs. 9-19 and 9-20, and described in the accompanying text. A HC emissions-occurred at o = 0.65, just before the partial-burn limit line where spark can arc from one electrode to another when a sufficiently high voltage is some slow-burning cycles occur but combustion is still complete in all cycles.44 applied. Ignition systems commonly used to provide this spark are: battery igni- From the above it is clear that flame initiation is a necessary but not suffi- tion systems where the high voltage is obtained with an ignition coil (coil ignition cient condition for complete combustion. Too-slow flame development and prop- systems); battery systems where the spark energy is stored in a capacitor and transferred as a high-voltage pulse to the spark plug by means of a special trans- former (capacitive-discharge ignition systems); and magneto ignition systems where the magneto-a rotating magnet or armature-generates the current used 100 to produce a high-voltage pulse. Ignition This section reviews our basic understanding of electrical discharges in 80- limit Partial burn limit inflammable gas mixtures relevant to engine ignition (Sec. 9.5.1), the major design 60 Complete burns in all cycles and operating characteristics of conventional engine ignition systems (Sec. 9.5.2), Spark timing, deg BTC and, briefly, some alternative approaches to generating a propagating flame (Sec. 9.5.3). 40 MBT spark timing Partial 20 burn regime: 9.5.1 Ignition Fundamentals A spark can arc from one plug electrode to the other only if a sufficiently high 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 voltage is applied. In a typical spark discharge, the electrical potential across the Equivalence ratio electrode gap is increased until breakdown of the intervening mixture occurs. FIGURE 9-38 Ionizing streamers then propagate from one electrode to the other. The imped- Actual limiting combustion regimes for lean-operating engine. 1200 rev/min, volumetric ance of the gap decreases drastically when a streamer reaches the opposite elec- efficiency ~ 60 percent, methane fuel, 40 mJ spark energy, 2.5 ms spark duration.44 trode, and the current through the gap increases rapidly. This stage of the 428 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 429 discharge is called the breakdown phase. It is followed by the arc phase, where the (<100 V), though the current can be as high as the external circuit permits. In thin cylindrical plasma expands largely due to heat conduction and diffusion and, contrast to the breakdown phase where the gas in the channel is fully dissociated with inflammable mixtures, the exothermic reactions which lead to a propagating and ionized, in the arc phase the degree of dissociation may still be high at the flame develop. This may be followed by a glow discharge phase where, depending center of the discharge, but the degree of ionization is much lower (about 1 on the details of the ignition system, the energy storage device (e.g ., the ignition percent). Voltage drops at the cathode and anode electrodes are a significant coil) will dump its energy into the discharge circuit. 52, 53 fraction of the arc voltage, and the energy deposited in these electrode sheath Figure 9-39 shows the behavior of the discharge voltage and current as a regions, which is conducted away by the metal electrodes, is a substantial fraction function of time for a conventional coil ignition system. Typical values are of the total arc energy (see Table 9.4 below). The arc requires a hot cathode spot, shown; actual values depend on the details of the electrical components. The so evaporation of the cathode material occurs. The arc increases in size due pri- breakdown phase is characterized by a high-voltage (~ 10 kV), high-peak current marily to heat conduction and mass diffusion. Due to these energy transfers the (~200 A), and an extremely short duration (~10 ns). A narrow (~40 um gas temperature in the arc is limited to about 6000 K: the temperature and diameter) cylindrical ionized gas channel is established very early. The energy degree of dissociation decrease rapidly with increasing distance from the arc axis. supplied is transferred almost without loss to this plasma column. The tem- Currents less than 200 mA, a large electrode voltage drop at the cathode (300 to perature and pressure in the column rise very rapidly to values up to about 500 V), a cold cathode, and less than 0.01 percent ionization are typical for the 60,000 K and a few hundred atmospheres, respectively. A strong shock or blast glow discharge. Energy losses are higher than in the arc phase, and peak equi- wave propagates outward, the channel expands, and, as a result, the plasma tem- librium gas temperatures are about 3000 K.52 perature and pressure fall. Some 30 percent of the plasma energy is carried away About 0.2 mJ of energy is required to ignite a quiescent stoichiometric fuel- by the shock wave; however, most of this is regained since spherical blast waves air mixture at normal engine conditions by means of a spark. For substantially transfer most of their energy to the gas within a small (~2 mm diameter) sphere leaner and richer mixtures, and where the mixture flows past the electrodes, an into which the breakdown plasma soon expands. 52, 53 order of magnitude greater energy (~3 mJ) may be required.54 Conventional A breakdown phase always precedes arc and glow discharges: it creates the ignition systems deliver 30 to 50 mJ of electrical energy to the spark. Due to the electrically conductive path between the electrodes. The arc phase voltage is low physical characteristics of the discharge modes discussed above, only a fraction of the energy supplied to the spark gap is transmitted to the gas mixture. The energy balance for the breakdown, arc, and glow phases of the discharge is given in Table 9.4. Radiation losses are small throughout. The end of the breakdown phase occurs when a hot cathode spot develops, turning the discharge into an Predischarge phase Transition region Breakdown phase Transition region Glow discharge arc; heat losses to the electrodes then become substantial. The breakdown phase Arc phase 105 reaches the highest power level (~ 1 MW), but the energy supplied is small (0.3 to 1 mJ). The glow discharge has the lowest power level (~10 W) but the highest 104 15 kV 50 V 500 V mJ 30 mJ energy (30 to 100 mJ), due to its long discharge time. The arc phase lies between. 103 1 mJ The proportions of the electrical energy supplied, which can be transferred Voltage, V 102 to the plasma in these three phases of the discharge, are shown in Fig. 9-40.53, 55 The different transfer capabilities for breakdown arc and glow discharges arise 101 100 0-9 10- -6 2 - 3T 10- TABLE 9.4 Energy distribution for breakdown, arc, and glow 103 dischargest 102 Breakdown, % Arc, % Glow, % 101 Radiation loss <1 Current, A 100 FIGURE 9-39 Heat loss to electrodes 5 45 70 Schematic of voltage and current variation with Total losses 6 50 70 10-1 time for conventional coil spark-ignition system. Plasma energy 94 50 30 10-2 LLL Typical values for energy and voltage in the three 10-9 10-6 10- phases-breakdown, arc, and glow discharge-are + Typical values, under idealized conditions with small electrodes. Time, s given. 52 Source : Maly and Vogel . 52 430 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 431 100 primarily from the differences in heat losses to the electrodes, as explained above. Breakdown These losses increase with increasing supplied energy. In arc and glow discharges, 50 - - - Arc increases in either discharge time or discharge current (or in both) always lead to --- - Glow substantial decreases in energy-transfer efficiency. If the glow discharge current is 20 increased above about 100 mA, the discharge changes to the arc mode and heavy electrode erosion will result. Thus there are practical limits to the arc and glow 10- discharge currents; also, the time available for ignition in the engine limits increases in discharge time. 5- The initial expansion velocities in the discharge are much higher than those Transferred plasma energy, mJ in self-propagating flames. Figure 9-41 shows the expansion velocities and diam- 2 eters of the volumes activated by a 3-mJ, 100-us discharge from a capacitor- discharge ignition system as a function of time after spark onset.52 The curves shown are (1) for the shock wave following a discharge in air at 1 atm; (2) the 100 % 0.5 plasma in air at 1 atm; (3) the electrical and chemical plasma following a dis- charge in a stoichiometric methane-air mixture at 1 atm. The initially strong 25% shock wave attenuates rapidly to the local sound speed. Up to times of order 0.2 microseconds, the effects of fuel combustion chemistry are small. The change in slope of the velocity curves for the plasmas at about 100 us indicates the tran- 0.1 0.2 1 2 5 10 20 50 100 200 sition from an expansion caused by the initial high pressure in the breakdown Supplied electrical energy, mJ - discharge to expansion resulting from heat conduction and diffusion. FIGURE 9-40 The temperature distributions within the three different types of discharge Energy transferred to the spark kernal as a function of supplied electrical energy for breakdown, arc, provide additional insight. During the breakdown discharge, on a time scale of and glow discharges.55 nanoseconds, the temperature rises to 60,000 K. Increasing the breakdown energy does not produce higher kernel temperatures; instead the channel diam- eter increases, producing a larger plasma volume. The kernel temperatures then decrease to the order of 10,000 K on a microsecond time scale as the plasma 104 expands behind the shock wave. Arc and glow discharges, because their power d1 inputs are much lower, do not increase the kernel temperatures; rather, they 103 10 extend the cooling period on a microsecond and millisecond time scale, respec- tively. 52 - 8 The characters of the temperature profiles that each of these three types of 102- a3/ Diameter, mm discharge create in air, with essentially the same total electrical energy input (30 V 3 a to 33 mJ), are indicated in Fig. 9-42. The radial profiles in the undisturbed mid- Velocity, m/s plane of the arc are shown. The expansion-wave-induced expansion of the plasma 101 V2 a2 behind the shock with the breakdown discharge produces a larger plasma, earlier, with a steep temperature front. Thus it creates favorable conditions for trans- ferring heat and radicals to the surrounding unburned mixture. In addition cold 2 gas flows into the central region of the plasma, due to the boundary layers which the rapidly expanding flow sets up on the electrodes, effectively insulating the 10-1 T 0 10- 10-7 10-6 10 - 10-4 10-3 hottest plasma region from the cold electrode surfaces.53 The arc and glow dis- charges, each preceded by a much lower energy initial breakdown process, show Time, s a much slower expansion rate and the more gradual temperature profile produc- FIGURE 9-41 ed by heat conduction and diffusion. Diameters (d) and expansion velocities (D) of volumes activated by a capacitive-discharge ignition Chemical reactions can be observed spectroscopically a few nanoseconds system : 3 mJ electrical energy, 100 us duration. Subscripts denote: 1, shock in air at 1 atm; 2, plasma in air at 1 atm; 3, electrical and chemical plasma in stoichiometric methane-air mixture at 1 atm.52 after spark onset. They are initiated by the very high radical density in the break- 432 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 433 5000 rior of the plasma still consists of a fully dissociated reacted gas mixture with most of its energy stored in radicals. An indication of the gas composition across the steep temperature profile at the plasma interface can be obtained from Table 4000 -- 9.5, which shows the equilibrium composition of C8H18-air combustion products over the relevant temperature range. Note, however, that the gas will not be in TB 3000 TG FIGURE 9-42 equilibrium. The different radicals have different diffusivities, with the hydrogen Radial temperature profiles at selected times atom some five times that of other species. Thus the H radical will diffuse furthest Temperature, K 2 ms after spark onset for ignition systems with differ- into the as-yet unreacted mixture. On the high-temperature side of the inflamma- 2000| 500 us 50 us ent electrical energies and discharge times in air tion zone, the large number of radical particles transfer their energy to the at 1 atm. Te, breakdown discharge, 30 mJ mixture molecules within a few collisions. On the low-temperature side of the energy, 60 ns duration; T1, capacitive discharge, 1000 3 mJ energy, 100 us duration with superimposed zone, above-equilibrium concentrations of combustion-initiating radicals (O and current arc of 2 A, 30 mJ energy, for 230 us; Te, H) build up. In addition, a high heat flux into the region occurs, from the plasma capacitive-discharge system, 3 mJ energy, 100 us core, by conduction down the steep temperature gradient. As a consequence of duration with superimposed constant-current these conditions, reactions will occur and energy will be released more rapidly 00 3 glow discharge of 60 mA, 30 mJ energy, 770 us than in a normal flame.53 Radius, mm duration. 52 Figure 9-43a shows the size of the activated volume as a function of time for several types of discharge, both in a stoichiometric mixture and in air. It can down plasma where all the heavy particles are present as highly excited atoms be seen from the air curves for the capacitive discharge (CDI) and breakdown and ions. Since the kernel temperatures are much too high to allow the species discharge that, after about 10 us, the plasma ceases to be the energy source for present in normal combustion products to exist, combustion reactions take place continued growth of the activated volume. At this time, the plasma temperature at the outer plasma surface where the conditions are ideal for rapid chemical at the interface has fallen to a value comparable to flame temperatures. With activity (temperatures of one thousand to a few thousand kelvins). The chemical combustible mixtures, molecules such as OH, CH, C2, CO, etc ., appear, indicat- energy released is added to the plasma energy and becomes evident when the ing that combustion reactions are now occurring. Figure 9-43a shows that for plasma velocity falls below about 100 m/s (see Fig. 9-41). At this point, the inte- t 2 20 us, the volume activated by the discharge with the combustible mixture grows much faster than the volume activated in air. Continuing the supply of electrical energy in the arc and glow discharge does produce a higher expansion TABLE 9.5 rate in both the combustible mixture and in air, due to additional heat conduc- Equilibrium composition of stoichiometric tion and diffusion to the interface, but the onset of inflammation is not signifi- isooctane-air combustion products cantly affected. The radial temperature profiles across the discharge at selected times, shown in Fig. 9-43b, illustrate these points. This, therefore, is the critical Temperature, Kt point in the inflammation process: at some 20 us after onset of the discharge the Species 2000 3000 4000 5000 flame reactions must be proceeding sufficiently rapidly to be self-sustaining; i.e ., chemical energy release must more than offset heat losses across the front to the CO 2.4(-3) 6.1(-2) 9.4(-2) 9.0(-2) surrounding unburned mixture via diffusion and conduction.53 CO 2 1.2(-1) 5.7(-2) 5.0(-3) 3.5( -4) Thus the characteristics of the breakdown phase of the discharge have the H 2(-5) 9.7(-3) 1.2( -1) 1.9(-1) H2 6.1(-4) 1.5(-2) 2.4(-2) 3.5(-3) greatest impact on inflammation. The size of the activated volume a given time H2O 1.4(-1) 1.0(-1) 1.1(-2) 1.2(-4) interval after spark initiation, the temperature difference across the kernel inter- N n 1(-5) 6.6(-4) 1.2(-2) face, and the velocity of the interface are all substantially increased by increasing NH n n 2(-5) 6(-5) the breakdown phase energy (see Fig. 9-43). Additional energy input during the NO 5.7( -4) 1.5(-2) 2.9(-2) 1.7(-2) N 2 7.3( -1) 6.9(-1) 5.7(-1) 5.1(-1) arc and glow discharge phases has a more modest effect on these critical kernel O 1(-5) 8.6(-3) 9.8(-2) 1.6(-1) properties. This is graphically illustrated by Fig. 9-44 where the same energy OH 4.5(-4) 2.2(-2) 3.3(-2) 5.6(-3) input into breakdown, arc, and glow discharge modes produces substantially dif- 1.1(-3) 2.3(-2) 1.7(-2) 2.2(-3) ferent ignition limits. t At 4 atm pressure: mole fractions, 9.0( -2) = 9.0 x 10-2. Several models of the plasma-unburned mixture interface have been devel- n = <5 x 10-6 oped in attempts to quantify the complex phenomena described above (e.g ., Refs. 434 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 435 100 1000- TT 80 :3 CH4-air, p = 1 atm, 0 = 1 T = Air, p = 1 atm 60 -v = 1 m/s = 8 m/s 100 FIGURE 9-44 Probability of inflammation, % Probability of inflammation of stoichiometric 40 Glow methane-air mixture, at 300 K, 4 atm, as function Arc- of relative air/fuel ratio 1 (=1/¢) for different igni- TT tion discharges with equal total electrical energy Activated volume, mm3 10- 20 Breakdown (30-33 mJ). Breakdown: 30 mJ, 60 ns duration. Arc: CDI, 3 mJ, 100 us duration; plus 2 A, 30 mJ, 230 us duration arc. Glow: CDI, 3 mJ, 100 us 0.8 1.0 1.2 1.4 1.6 duration; plus 60 mA, 30 mJ, 770 us duration a, c, d FIGURE 9-43 Relative air/fuel ratio glow discharge. v = mixture velocity.52 1 (a) Size of discharge-activated volume as function of time for several types of discharge in air growth (up to 10 to 100 us) is not greatly affected by the mixture equivalence and in stoichiometric methane-air ratio, the inflammation process and the thickness and rate of propagation of the mixture at 1 atm. a: CDI, 3 mJ 0.1L 1 10 100 1000 energy, 100 us duration; b: break- resulting flame are strongly affected. The lean side of the minimum is of more down discharge, 30 mJ energy, practical interest than the rich side. Because the chemical energy density of the Time after spark onset, us 20 ns duration; c: CDI plus arc mixture and flame temperature decrease as the mixture is leaned out, the flame (a) discharge, 1.5 A, 40 V, 500 us speed decreases and the flame becomes thicker. Thus more time is available for duration; d: CDI + glow dis- heat losses from the inflammation zone, less energy is available to offset these 6000 charge, 30 mA, 500 V, 2 ms dura- tion. (b) Temperature profiles for losses, and the rate of energy transfer into the zone decreases. The consequence is 5000 different discharge modes at dif- that, as the mixture is leaned out, the approximately spherical discharge-created 230 us 150 ps plasma must grow to a larger size before inflammation will occur. Substantially, 4000 ferent times in stoichiometric CH4-air, p = 4 atm, o = 1 methane-air mixture at 300 K, more energy must therefore be supplied to the discharge.53 Temperature, K 3000 4 atm. a: CDI, 3 mJ energy, In engines, the mixture is not quiescent: mean and fluctuating velocities in d b 100 us duration; b: breakdown 2000 discharge, 20 mJ energy, 80 ns the range 1 to 10 m/s exist in the clearance volume at TC (see Sec. 8.2.2). On the 770 us duration; c: CDI, 3 mJ energy, time scale of the breakdown discharge phase (10 ns), this fluid motion is not 1000 770 MS ,770 us 100 us duration, plus 2 A, 30 mJ important. In the arc and glow discharge phases, however, the arc is convecte energy, 230 us duration arc; d: 2 3 CDI, 3 mJ energy, 100 us dura- Radius, mm tion, plus 60 mA, 30 mJ energy, 80 (b) 770 us duration glow discharge.53 Velocity, m/s 15 56 and 57). They are based on the requirement that the energy release due to chemical reaction in the plasma front (and production of radical species) exceed the losses due to conduction and diffusion to the unburned gas ahead of the Minimum ignition energy, mj 40- front. While these models are incomplete, they do provide a theoretical basis for the well-known fact that the minimum energy required to ignite a premixed fuel- Stagnant air mixture depends strongly on mixture composition. Figure 9-45 shows a 20 mixtures typical set of results on the minimum ignition energy as a function of the equiva- lence ratio under quiescent conditions.58 The curve shows a minimum for slightly FIGURE 9-45 Effect of mixture equivalence ratio and flow veloc- rich-of-stoichiometric mixtures; the minimum energy required for successful igni- 1.0 1.5 2.0 2.5 ity on minimum ignition energy for propane-air tion increases rapidly as the mixture is leaned out. While the initial plasma kernel Equivalence ratio mixtures at 0.17 atm. 58 436 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 437 4. The time over which the ignition energy can be used effectively for inflamma- tion decreases as the initial flame velocity increases. Ignition energy supplied after inflammation has occurred will have only a modest impact on flame propagation. 15 m/s 35 m/s 9.5.2 Conventional Ignition Systems 0 m/s FIGURE 9-46 The ignition system must provide sufficient voltage across the spark plug elec- Photographs of single glow discharge (30 mJ, 0.77-1.5 ms) in air at 2 atm flowing perpendicular to trodes to set up the discharge and supply sufficient energy to the discharge to axis of electrodes. Below 25 m/s almost no multiple discharges; above 25 m/s only multiple dis- ignite the combustible mixture adjacent to the plug electrodes under all operating charges. 1.2 mm electrode gap.56 conditions. It must create this spark at the appropriate time during the compres- sion stroke. Usually spark timing is set to give maximum brake torque for the by the flow and lengthened accordingly, as illustrated in Fig. 9-46. For velocities specific operating condition, though this maximum torque may be constrained by below 15 m/s a steady increase in discharge channel length occurs. For higher emission control or knock control requirements. For a given engine design, this velocities, an increasing number of reignitions occur, so the discharge energy is optimum spark timing varies as engine speed, inlet manifold pressure, and distributed into many separate channels. As the channel lengthens, the ratio of mixture composition vary. Thus, in most applications, and especially the automo- total discharge voltage to anode plus cathode voltage drop increases substan- tive applications, the system must have means for automatically changing the tially, and the relative importance of heat losses to the electrodes decreases. Thus spark timing as engine speed and load vary. more energy is transferred to the gas-the energy-transfer efficiency increases. With an equivalence ratio best suited for ignition and with homogeneous However, as the channel is lengthened, the energy transferred is spread over a mixture distribution, spark energies of order 1 mJ and durations of a few micro- larger volume. Depending on flow velocity, and mixture and discharge condi- seconds would suffice to initiate the combustion process. In practice, circum- tions, increasing velocity may increase or decrease the minimum ignition energy stances are less ideal. The air, fuel, and recycled exhaust are not uniformly or the lean ignition limit for a specific ignition system. Both the mean flow veloc- distributed between cylinders; the mixture of air, fuel, recycled exhaust gas, and ity and turbulence levels are important.53-58 With a conventional coil ignition residual gas within each cylinder is not homogeneous. Also, the pressure, tem- system, increasing mean flow velocity up to the point where reignitions start to perature, and density of the mixture between the spark plug electrodes at the time occur extends the lean limit. With breakdown discharge systems, the lean ignition the spark is needed affect the voltage required to produce a spark. These vary limit decreases as flow velocity increases.53 With capacitive-discharge systems at significantly over the load and speed range of an engine. The spark energy and low flow velocities, the flow has little impact: at high flow velocities the minimum duration, therefore, has to be sufficient to initiate combustion under the most ignition energy increases. 59 unfavorable conditions expected in the vicinity of the spark plug over the com- Maly53 has summarized these fundamental aspects of spark-discharge plete engine operating range. Usually if the spark energy exceeds 50 mJ and the ignited flames as follows: duration is longer than 0.5 ms reliable ignition is obtained. In addition to the spark requirements determined by mixture quality, pres- 1. Of the total electrical energy supplied to the spark, only that fraction con- sure, temperature, and density, there are others determined by the state of the tained within the outer surface layer of the plasma (of thickness of the order of plugs. The erosion of the plug electrodes over extended mileage increases the gap the inflammation zone) is available for initiating the flame propagation width and requires a higher breakdown voltage. Also, spark plug fouling due to process. The energy density and the temperature gradient in this layer depend deposit buildup on the spark plug insulator can result in side-tracking of the on the discharge mode. Highest energy densities and temperature gradients spark. When compounds formed by the burning of fuel, lubricating oil, and their are achieved if the ignition energy is supplied in the shortest time interval. additives are deposited on the spark plug insulator, these deposits provide an 2. A minimum radius of the spark plasma is required for inflammation of the alternative path for the spark current. If the resistance of the spark plug deposits fuel-air mixture to occur. This radius increases rapidly as the mixture is leaned is sufficiently low, the loss of electrical energy through the deposits may prevent out (or diluted); it decreases with increasing pressure and increasing plasma the voltage from rising to that required to break down the gas. The influence of expansion velocity. side-tracking on spark generation decreases with lower source impedance of the 3. After inflammation, burning rates are proportional to flame surface area. Thus high-voltage supply, and therefore with a higher available energy. discharges and plasma geometries that produce the largest inflammation zone The fundamental requirements of the high-voltage ignition source can be surface area, most rapidly, are advantageous. summarized as: (1) a high ignition voltage to break down the gap between the 438 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 439 plug electrodes; (2) a low source impedance or steep voltage rise; (3) a high Breaker energy storage capacity to create a spark kernel of sufficient size; (4) sufficient points Breaker -open duration of the voltage pulse to ensure ignition. There are several commonly used Primary current . points Time concepts that partly or fully satisfy these requirements. 54, 60 close Secondary voltage (no spark plug) Available1 COIL IGNITION SYSTEMS. Breaker-operated inductive ignition systems have voltage Secondary voltage been used in automotive engines for many years. While they are being replaced (with spark plug) Required FIGURE 9-48 IL with more sophisticated systems (such as transistorized coil ignition systems), voltage Spark occurs Current and voltage waveforms for breaker igni- they provide a useful introduction to ignition system design and operation. -Spark duration tion system. 61 Figure 9-47 shows the circuit of a typical breaker ignition system. The system includes a battery, switch, resistor, coil, distributor, spark plugs, and the neces- sary wiring. The circuit functions as follows. If the breaker point is closed when this induced voltage will have a damped sinusoidal waveform, as shown in the the ignition is switched on, current flows from the battery, through the resistor, center trace. The peak value of this voltage is the maximum voltage that can be primary winding of the ignition coil, contacts, and back to the battery through produced by the system and is called the available voltage Va of the system. The ground. This current sets up a magnetic field within the iron core of the coil. maximum energy transferred to the secondary system is given by When ignition is required, the breaker points are opened by the action of the distributor cam, interrupting the primary current flow. The resulting decay of Es, max = = C. v. magnetic flux in the coil induces a voltage in both the primary and secondary windings. The voltage induced in the secondary winding is routed by the dis- where C, is the total capacitance of the secondary circuit. Hence, the available tributor to the correct spark plug to produce the ignition spark. voltage of the system is given by The current and voltage waveforms are shown in Fig. 9-48. The primary current for any given time of contact closure t is given by V. - (25%.max) (9.54) 1, = Vo R (1 - e - REILp ) (9.53) If all the energy stored in the primary circuit of the coil, 11, 13, is transferred to where I, is the primary current, Vo is the supply voltage, R is the total primary the secondary, circuit resistance, and L, is the primary circuit inductance. The primary current 1/2 requires time to build up. At low speeds the time of contact closure is sufficient (9.55) for the primary current to reach the maximum permitted by the circuit resis- tance; at high speeds the primary current may not reach its maximum. Thus, When the coil is connected to a spark plug, the secondary voltage will rise only at higher engine speeds does the term e-Ri/Lp become significant. When the to the breakdown potential of the spark plug, and a discharge between the plug points open the primary current falls to zero and a voltage of order 15 kV is electrodes will occur. This alters the waveform as shown in the bottom trace of induced in the secondary winding. If the coil is not connected to a spark plug, Fig. 9-48. After the spark occurs, the voltage is reduced to a lower value until all the energy is dissipated and the arc goes out. The value of this voltage which Ignition caused breakdown to occur is called the required voltage of the spark plug. The coil Distributor interval during which the spark occurs is called the spark duration. The available voltage of the ignition system must always exceed the required voltage of the spark plug to ensure breakdown. The spark must then possess sufficient energy and duration to initiate combustion under all conditions of operation. The major limitations of the breaker-operated induction-coil system are the decrease in available voltage as engine speed increases due to limitations in the current switching capability of the breaker system, and the decreasing time avail- Ignition Spark FIGURE 9-47 able to build up the primary coil stored energy. Also, because of the high source condenser plugs Schematic of conventional coil ignition system. impedance (about 500 k_2) the system is sensitive to side-tracking across the H Contact breaker (Courtesy Robert Bosch GmbH and SAE.54) spark plug insulator. A further disadvantage is that due to their high current 440 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 441 load, the breaker points are subject to electrical wear in addition to mechanical spark plugs as in the conventional breaker system. The control module contains wear, which results in short maintenance intervals. The life of the breaker points timing circuits which then close the primary circuit so that buildup of primary is dependent on the current they are required to switch. Acceptable life is circuit current can occur. There are many types of pulse generators that could obtained with I, ~ 4 A; increased currents cause a rapid reduction in breaker trigger the electronic circuit of the ignition system.60 A magnetic pulse generator, point life and system reliability. where a gear-shaped iron rotor driven by the distributor shaft rotates past the stationary pole piece of the pickup, is usually used. The number of teeth on the TRANSISTORIZED COIL IGNITION (TCI) SYSTEMS. In automotive applica- rotor is the same as the number of cylinders. A magnetic field is provided by a tions, the need for much reduced ignition system maintenance, extended spark permanent magnet. As each rotor tooth passes the pole piece it first increases and plug life, improved ignition of lean and dilute mixtures, and increased reliability then decreases the magnetic field strength y linked with the pickup coil, produc- and life has led to the use of coil ignition systems which provide a higher output ing a voltage signal proportional to du/dt. The electronic module switches off the voltage and which use electronic triggering to maintain the required timing coil current to produce the spark as the rotor tooth passes through alignment without wear or adjustment (see Refs. 54 and 61). These are called transistorized and the pickup coil voltage abruptly reverses and passes through zero. The coil ignition (TCI) or high-energy electronic-ignition systems. The higher output increasing portion of the voltage waveform, after this voltage reversal, is used by voltage is required because spark plugs are now set to wider gaps (e.g ., about the electronic module to establish the point at which the primary coil current is 1 mm) to extend the ability to ignite the fuel mixture over a wider range of engine switched on for the next ignition pulse. operation, and because during the extended mileage between spark plug replace- ment electrode erosion further increases the gap. In automotive applications an CAPACITIVE-DISCHARGE IGNITION (CDI) SYSTEMS. With this type of system available ignition voltage of 35 kV is now usually provided. In addition to higher (shown schematically in Fig. 9-50) a capacitor, rather than an induction coil, is voltage, longer spark duration (about 2 ms) has been found to extend the engine used to store the ignition energy. The capacitance and charging voltage of the operating conditions over which satisfactory ignition is achieved. capacitor determine the amount of stored energy. The ignition transformer steps Most of the solid-state ignition systems now in use operate on the same up the primary voltage, generated at the time of spark by the discharge of the basic principle. Figure 9-49 shows the block circuit diagram of a transistorized capacitor through the thyristor, to the high voltage required at the spark plug. coil ignition system. The distributor points and cam assembly of the conventional The CDI trigger box contains the capacitor, thyristor power switch, charging ignition system are replaced by a magnetic pulse generating system which detects device (to convert battery voltage to the charging voltage of 300 to 500 V by the distributor shaft position and sends electrical pulses to an electronic control means of pulses via a voltage transformer), pulse shaping unit, and control unit. module. The module switches off the flow of current to the coil primary windings, The principal advantage of CDI is its insensitivity to electrical shunts in the inducing the high voltage in the secondary windings which is distributed to the high-voltage ignition circuit that result from spark plug fouling. Because of the fast capacitive discharge, the spark is strong but short (0.1 to 0.3 ms). This can lead to ignition failure at operating conditions where the mixture is very lean or dilute. 54 Induction-type pulse Regulator generator Ignition coil D Dwell-angle C control system ND Pulse- Driver stage Distributor shaping !Charging Switching Idevice circuit Thyristor Ignition output stage witch Pulse- Control shaping unit circuit Trigger Spark box plugs Ignition transformer -4 To induction-type FIGURE 9-49 pulse generator FIGURE 9-50 Schematic of transistorized coil ignition system with induction pulse generator. (Courtesy Robert Schematic of capacitive-discharge system. (Courtesy Bosch GmbH and SAE.54) To distributor‘ Robert Bosch GmbH and SAE.54) 442 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 443 components: an insulator, electrodes, and a shell. The insulated material must K IA - have adequate thermal shock resistance, tensile and compressive strength, and impact strength. It must also have low porosity to limit absorption of com- bustion gases and high resistivity to prevent leakage of high-voltage charge at S both ambient temperatures and normal operating temperatures. Alumina is FIGURE 9-51 usually used as the insulator material. Schematic of breaker-triggered magneto system with The electrodes are normally made of high-nickel alloys to withstand the Pole wheel ignition armature. (Courtesy Robert Bosch GmbH and high ignition voltage, high temperatures, and corrosive gases with minimum SAE.54) erosion. The center-electrode surface temperatures can average 650-700ºC under normal operating conditions. A wide range of electrode geometries are available. MAGNETO IGNITION. With this type of ignition system, a magneto supplies the The location of the special conductive seal within the shell affects the heat ignition voltage for the spark discharge independent of a battery or generator. rating of the spark plug. For a "hot" plug, an insulator with a long conical nose Magneto ignition is commonly used in small four-stroke and two-stroke engines. is used; for a "cold" plug a short-nosed insulator is used. The length of the heat Figure 9-51 illustrates the system and its operation. A time-varying magnetic flux conduction path from the insulator nose to the shell is changed in this way to is set up in the ignition armature (IA) as the rotating permanent magnets on vary and control the temperature of the exposed part of the insulator. The spark the pole wheel generate a current in the closed primary winding W1. This plug insulator tip temperature increases with increasing speed. It is desirable to primary current generates an additional flux 1, giving a resultant flux OR = 00 have the tip hotter than about 350ºC to prevent fouling at low speed. High-speed + 01. To generate the ignition voltage, the primary current flow is interrupted high-load tip temperatures must be kept below about 950ºC to prevent preigni- and the flux collapses rapidly from R to Do, producing a high-voltage pulse in tion. Normally, the gap between the center and ground electrodes is 0.7 to the winding which is connected to the spark plug electrode. The current can be 0.9 mm. For extremely dilute mixtures this is usually increased to 1.2 mm. interrupted with contact breakers (breaker-triggered magneto) or with a tran- Magneto ignition systems use smaller gaps (~0.5 mm). High-compression-ratio sistor (semiconductor magneto). Since the flux generated by the rotating pole racing engines use smaller gaps (0.3 to 0.4 mm).54 wheel depends on engine speed, the magnitude of the ignition voltage varies with speed, 54 9.5.3 Alternative Ignition Approaches SPARK PLUG DESIGN. The function of the spark plug is to provide an electrode A large number of methods for initiating combustion in spark-ignition engines gap across which the high-voltage discharge occurs which ignites the compressed with electrical discharges, in addition to those described in the previous section, mixture of fuel vapor and air in the combustion chamber. In addition, it must have been proposed and examined. These include different designs of spark plug, provide a gas-tight conducting path from the high-voltage wire to the electrode use of more than one plug, use of higher power, higher energy, or longer-duration gap. Figure 9-52 shows a typical spark plug design. There are three principal discharges, and ignition systems that initiate the main combustion process with a high-temperature reacting jet-plasma-jet and flame-jet ignition systems.62 Con- ventional ignition systems normally ignite the unburned fuel, air, burned gas mixture within the cylinder and perform satisfactorily under conditions away Terminal nut from the lean or dilute engine stable operating limit. Thus, these alternative igni- tion approaches have the goal of extending the engine stability limit (and/or of Leakage-current barriers reducing the cyclic combustion variability near the stability limit), usually by achieving a faster initial burning rate than can be obtained with conventional Insulator . Terminal stud systems. This section describes the more interesting of these alternative ignition approaches. Special ALTERNATIVE SPARK-DISCHARGE APPROACHES. There are many different conductive seal Captive gasket designs of spark plugs. These use different geometry electrodes, gap widths, and Internal seal gap arrangements. The effects of the major plug electrode design features on the FIGURE 9-52 Center electrode Cutaway drawing of conventional spark plug. engine's stable lean operating limit are illustrated in Fig. 9-53.63 Ignition system Ground electrode .... Insulator nose (Courtesy Robert Bosch GmbH and SAE.54) effects are important when misfire due to the quenching effect of the spark plug 444 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 445 20 20 (b) 1.5 1.0 development angle, and rapid-burning angle, and imep and torque fluctuations 0.5 have defined the effects of both increasing the number of ignition sites from 1 to 1.0 18 18 12 and of changing their geometric location.64 These results confirm that increas- Lean limit 0.75 ing the number of simultaneously developing flame kernels increases the initial Center-electrode diameter, mm mixture burning rate, as anticipated. It also extends the lean stable operating Lean limit 16 16 0.5 limit and reduces cyclic combustion variability under conditions where slow and Gap width, mm 2.5 occasional partial-burning cycles would occur with fewer spark plug gaps. 14 - 14 Many studies have examined the effects of higher-energy discharges on 0 10 20 30 40 50 engine operation near the lean operating limit. It is useful to differentiate between 2.9 20 12-1 13.0 higher current discharges and longer duration discharges: most high-energy C) 0.5 1.0 1.5 (conventional-type) ignition systems have both these features. The results of these 2.0 2.5 8.0 Spark gap width, mm studies show that away from the lean or dilute stable operating limit, increasing 18 3.5 (a) the discharge current or duration has no significant effect on engine operating Lean limit characteristics. The higher current does, as would be expected, result in a larger 16 flame kernel during the inflammation process and thereby modestly reduces the Spark gap projection, mm spark advance required for maximum brake torque with a given combustion chamber and set of operating conditions. Figure 9-54 shows these trends for rich, 14 stoichiometric, and slightly lean mixtures. The figure also shows that both higher 10 20 30 40 50 Spark timing, deg BTC FIGURE 9-53 ', ms Is, mA Effect of spark plug electrode diameter, plug gap width, and projection of gap into chamber on 25 440 air/fuel ratio at engine's lean stable operating limit. Baseline conditions: 30 mJ spark energy, 3.5 mm 100 projection, center electrode diameter 2.5 mm, gap width 0.75 mm, 40º BTC spark timing. 1600 rev/ 5 25 Fuel consumption, g/kW.h 400 min, intake pressure 300 mmHg.63 100 360 electrodes determines the stable operating limit.+ Thus, smaller spark plug center-electrode diameters (Fig. 9-53a), larger electrode gap widths (Fig. 9-53b), 320 and higher electrode temperatures (obtained by projecting the gap further into 10 the combustion chamber; Fig. 9-53c), all extend the lean stability limit to leaner (or more dilute) mixtures for the more advanced spark timings and smaller gap 8 HC emission, g/kW.h widths. For spark timing closer to TC and for larger gap widths, these data show the lean stability limit to be much less sensitive (or not sensitive at all) to plug 6- geometry or spark energy. Multigap plugs, designed to produce a series of dis- charges which together form a long arc, have also been used to generate a larger initial flame kernel and thereby extend the lean limit. Use of more than one plug, at separate locations in the combustion 60 chamber and fired simultaneously, is also common.49 The advantages are Spark timing, deg BTC twofold. First, the effective flame area in the early stages of flame development is 40 increased substantially (e.g ., by almost a factor of two for two widely spaced plugs). Second, the variations in flow velocity and mixture composition in the 20 vicinity of the (multiple) plugs produce less variability in the initial mixture 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 burning rate than occurs with a single plug. Studies of heat-release rates, flame Relative air/fuel ratio À FIGURE 9-54 Effect of higher spark currents I, and longer spark durations t, on fuel consumption, HC emissions, and MBT spark timing, as a function of relative air/fuel ratio 1 (=1/o). 2000 rev/min, bmep = 3 atm, t See Sec. 9.4.3 for a detailed discussion of the stable operating limit. 2.8-dm3 six-cylinder engine.65 446 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 447 80F Mass fraction burned ATC trodes is substantially increased above values used in normal ignition systems by 60 allowing a capacitor to discharge at a relatively low voltage and high current MBT 95 % 40- through the spark generated in a conventional manner with a high-voltage low- 1.0 current ignition system. Stored energies of about 1 J are typically used, and this TCI BD 0.8 energy is discharged in some 20 us. This high-power discharge creates a high- Crank angle, deg TC MBT timing temperature plasma so rapidly that the pressure in the cavity increases substan- 0.6H 5% rCI BD tially, causing a supersonic jet of plasma to flow from the cavity into the main 20 Standard deviation of imep, bar 0.4- combustion chamber. The plasma enters the combustion chamber as a turbulent 40 jet, preceded by a hemispherical blast wave. The gas dynamic effects of the blast BTC 0.2- 60 Spark discharge wave are dissipated by the time combustion starts, which is typically of order 0.9 1.0 1.1 0 1.2 1.4 1.3 1.4 1.5 0.9 1.0 1.5 1 ms after the discharge commences. Ignition in the main combustion chamber 1.1 1.2 1.3 1.4 Relative air/fuel ratio À Relative air/fuel ratio À takes place in the turbulent jet; the flame starts out as a turbulent flame in con- (b) trast to the flame with conventional ignition systems, which is initially lami- (a) narlike. The penetration of the jet depends on its initial momentum; it thus FIGURE 9-55 (a) MBT spark timing and location of 5 and 95 percent mass fraction burned points for conventional depends on the amount of energy deposited, cavity size, and orifice area. If the transistorized coil ignition (TCI) system (43 mJ energy, 2 ms duration) and breakdown system (BD) cavity is filled prior to ignition with a hydrocarbon (or a mixture of (43 mJ energy, ~10 ns duration) as function of relative air/fuel ratio À (=1/o). (b) Standard deviation hydrocarbons) the ignition capabilities are enhanced due to the large increase in in indicated mean effective pressure as a function of relative air/fuel ratio 1 for TCI and BD hydrogen atoms created in the plasma.62 systems. 66 The effects of plasma-jet ignitors on engine combustion are similar to those of breakdown ignition systems: the flame development period is significantly shortened, and the engine's lean stable operating limit is extended. In addition, currents and longer duration discharges do extend the lean engine stability limit the phenomenon of misfire, which is the failure to initiate combustion in a frac- [and also the dilute (with EGR) stability limit]. Note the HC emissions data, tion of the engine's operating cycles, no longer occurs.68 The lean operating limit which indicate that longer discharges have a greater impact than higher currents of the engine is, however, normally controlled by flame extinguishment. on extending the misfire limit. The increase in fuel consumption as the lean oper- ating limit is approached is due to the rapidly increasing cycle-by-cycle com- FLAME-JET IGNITION. With this type of system, ignition occurs in a precham- bustion variability. ber cavity which is physically separated from the main chamber above the piston The discussion of discharge fundamentals in Sec. 9.5.1 showed that depos- and is connected to it via one or more orifices or nozzles. As the flame develops iting energy into the discharge during the initial short breakdown phase resulted in this cavity the pressure of the gases in the prechamber rises, forcing gas out in faster flame kernel growth than did depositing the same energy at slower rates. into the main chamber through the orifice (or orifices) as one or more turbulent Ignition systems of this type have been used. In such systems, a capacitor is burning jets. The jet or jets penetrate into the main chamber, igniting the connected in parallel with the spark plug electrodes, and a low-impedance dis- unburned mixture in the main chamber, thereby initiating the primary com- charge path allows the energy stored in the capacitor to be discharged into the bustion process. Ignition within the cavity is usually achieved with a convention- gap very rapidly. The anticipated effects on the engine's combustion process are al spark discharge. The function of the prechamber or cavity is to transform the observed. Away from the lean engine stability limit, the primary impact is a initial flame around the spark plug electrodes into one or more flame jets in the reduction in the flame development period due to the more rapid initial flame main chamber, which have a substantial surface area that can ignite extremely kernel growth. Thus MBT timing is less advanced with these breakdown systems lean or dilute mixtures in a repeatable manner. Many different systems for than with conventional systems (see Fig. 9-55a). The lean limit can also be achieving this goal have been developed; some of these have been used in pro- extended, and acceptable engine stability obtained (i.e ., tolerable cycle-by-cycle duction spark-ignition engines. Examples of the three major types of flame-jet combustion variations) for leaner or more dilute engine operation, as shown in ignition systems are shown in Fig. 9-56. Fig. 9-55b.66, 67 Figure 9-56a shows an example of the simplest type of flame-jet ignition concept (often called a torch cell). The cavity has no separate valve so is PLASMA-JET IGNITION. In a plasma-jet ignitor, the spark discharge is confined unscavenged; nor is there any prechamber fuel metering system. The function of within a cavity that surrounds the plug electrodes, which connects with the com- the prechamber cavity is to increase the initial growth rate of the flame imme- bustion chamber via an orifice. The electrical energy supplied to the plug elec- diately following spark discharge by having this flame growth take place in a 448 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 449 Spark plug Figure 9-56b and c shows two prechamber stratified-charge engine flame-jet ignition concepts. Here the mixture in the prechamber cavity is enriched by addi- tion of fuel so that it ends up being slightly rich-of-stoichiometric at the time of Turbulence generating pot spark discharge. The initial inflammation process in the cavity then occurs more Intake port. Orifice rapidly and more repeatably. The operating principle of these stratified-charge Main combustion systems is described briefly in Sec. 1.9. Figure 9-56b shows a system where the chamber prechamber is unscavenged and fuel is injected directly into the prechamber cavity (in addition to the main fuel-injection process which occurs into the fresh charge in the intake system) to richen the mixture (which is lean overall) at the time of spark to an easily ignitable rich-of-stoichiometric composition. With this approach, the prechamber volume is usually 20 to 25 percent of the clearance (a) volume. Figure 9-56c shows a prechamber stratified-charge flame-jet ignition system where the prechamber is scavenged between each combustion event. With Auxiliary intake Auxiliary intake this approach, a separate small intake valve feeds very rich mixture into the valve passage prechamber during the intake process, while the main fuel metering system feeds lean mixture to the main intake valve. During intake the prechamber is com- Spark plug Spark plug pletely scavenged by the rich intake stream. During compression the lean mixture Prechamber flowing from the main chamber to the prechamber brings the prechamber Injection nozzle pp mixture equivalence ratio to slightly rich-of-stoichiometric at the time of spark Passage Main intake discharge.71 Main chamber valve The number and size of the orifices connecting the prechamber and the main chamber have a significant effect on the development of the main chamber burning process. Two different approaches are shown in the flame development stage in Fig. 9-57. Figure 9-57a shows the jets produced when the prechamber has more than one small nozzle which direct the burning prechamber mixture deep into the main chamber charge. A fast burning of the lean main-chamber (b) (c) charge results. With this approach, prechamber volumes of 2 to 3 percent of the FIGURE 9-56 clearance volume and nozzle area/prechamber volume ratios of 0.03 to 0.04 cm - 1 Flame-jet ignition concepts: (a) turbulence-generating torch cell;69 (b) prechamber stratified-charge are used. Figure 9-57b shows the approach used by Honda in their CVCC engine. engine with auxiliary fuel injector with no prechamber scavenging;7º (c) prechamber stratified-charge engine with prechamber inlet valve and auxiliary carburetor.71 more turbulent region than the main combustion chamber: the flame jet or jets which then emerge from the cavity produce a large initial flame surface area in the main chamber to start the bulk charge combustion process. The prechamber system shown was called a turbulent generating pot.69 Another approach is to incorporate the cavity into the spark plug. Systems with prechamber volumes varying from 20 percent of the clearance volume to less than 1 percent have been developed. The flow pattern produced within the prechamber by the flow into the cavity during compression and the location of the spark plug electrodes within the cavity and of the nozzle or orifice are critical design issues. A major problem (a) (b) with these systems is that the prechamber is never completely scavenged by fresh FIGURE 9-57 mixture between cycles, so the burned gas fraction in the unburned mixture Two different approaches to prechamber orifice design with the prechamber stratified-charge carbu- within the prechamber is always substantially higher than the burned gas fraction reted and scavenged engine: (a) one or more small orifice(s) for deep jet penetration and faster in the unburned main chamber mixture.72 burning process; (b) large orifice for lower-velocity jet and slower burn.73 150 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 451 A larger prechamber (5 to 12 percent) and larger orifice (orifice area/prechamber volume ratio of 0.04 to 0.2 cm ~1) gives a lower velocity jet which penetrates the main chamber charge more slowly, resulting in a slower burn. Normal combustion Abnormal combustion All these concepts extend the engine's lean stable operating limit, relative to A combustion process in which a flame equivalent conventional engines, by several air/fuel ratios. For example, the A combustion process which is initiated solely by a timed spark and in which the front may be started by hot combustion- unscavenged cavity without auxiliary fueling can operate satisfactorily at part- flame front moves completely across the chamber surfaces either prior to or after load with air/fuel ratios of 18 (equivalence ratio o ~ 0.8, relative air/fuel ratio combustion chamber in a uniform manner at spark ignition, or a process in which some a normal velocity. part or all of the charge may be consumed 1 ~ 1.25). The prechamber stratified-charge flame-jet ignition concepts can at extremely high rates. operate much leaner than this; however, the best combination of fuel consump- tion and emissions characteristics is obtained with o ~0.9 -0.75 1 ~ 1.1 - 1.3.70,71 One performance penalty associated with all these flame-jet ignition concepts is the additional heat losses to the prechamber walls due to Surface ignition increased surface area and flow velocities. The stratified-charge prechamber con- Spark knock* hot spots-combustion-chamber deposits cepts also suffer an efficiency penalty, relative to equivalent operation with A knock which is recurrent and repeatable Surface ignition is ignition of the fuel-air in terms of audibility. It is controllable by charge by any hot surface other than the uniform air/fuel ratios, due to the presence of fuel-rich regions during the com- the spark advance; advancing the spark spark discharge prior to the arrival of the bustion process. increases the knock intensity and retarding normal flame front. It may occur before the the spark reduces the intensity. spark ignites the charge (preignition) or after normal ignition (postignition). 9.6 ABNORMAL COMBUSTION: KNOCK AND SURFACE IGNITION 9.6.1 Description of Phenomena Run-on Abnormal combustion reveals itself it many ways. Of the various abnormal com- Continuation of engine firing bustion processes which are important in practice, the two major phenomena are after the electrical ignition is Knocking* surface ignition shut off. knock and surface ignition. These abnormal combustion phenomena are of Nonknocking surface Knock which has been concern because: (1) when severe, they can cause major engine damage; and (2) ignition preceded by surface even if not severe, they are regarded as an objectionable source of noise by the ignition. It is not controlla- Surface ignition which does engine or vehicle operator. Knock is the name given to the noise which is trans- ble by spark advance. Runaway surface ignition not result in knock. mitted through the engine structure when essentially spontaneous ignition of a Surface ignition which occurs earlier and earlier in portion of the end-gas-the fuel, air, residual gas, mixture ahead of the propagat- the cycle. It can lead to ing flame-occurs. When this abnormal combustion process takes place, there is serious overheating and an extremely rapid release of much of the chemical energy in the end-gas, causing structural damage to the Wild ping engine. Rumble very high local pressures and the propagation of pressure waves of substantial Knocking surface ignition A low-pitched thudding amplitude across the combustion chamber. Surface ignition is ignition of the fuel- characterized by one or noise accompanied by air mixture by a hot spot on the combustion chamber walls such as an overheat- more erratic sharp cracks. It engine roughness. Probably is probably the result of caused by the high rates of ed valve or spark plug, or glowing combustion chamber deposit: i.e ., by any early surface ignition from pressure rise associated with means other than the normal spark discharge. It can occur before the occurrence deposit particles. very early ignition or multiple surface ignition. of the spark (preignition) or after (postignition). Following surface ignition, a turb- ulent flame develops at each surface-ignition location and starts to propagate *Knock: The noise associated with autoignition of a portion of the fuel-air mixture ahead of the advancing flame front. across the chamber in an analogous manner to what occurs with normal spark Autoignition is the spontaneous ignition and the resulting very rapid reaction of a portion or all of the fuel-air mixture. ignition. FIGURE 9-58 Because the spontaneous ignition phenomenon that causes knock is gov- Definition of combustion phenomena-normal and abnormal (knock and surface ignition)-in a erned by the temperature and pressure history of the end gas, and therefore by spark-ignition engine. (Courtesy Coordinating Research Council.) the phasing and rate of development of the flame, various combinations of these two phenomena-surface ignition and knock-can occur. These have been cate- gorized as indicated in Fig. 9-58. When autoignition occurs repeatedly, during 452 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 453 otherwise normal combustion events, the phenomena is called spark-knock. Higher heat rejection causes higher temperature components which, in turn, can Repeatedly here means occurring more than occasionally: the knock phenome- advance the preignition point even further until critical components can fail. The non varies substantially cycle-by-cycle, and between the cylinders of a multi- parts which can cause preignition are those least well cooled and where deposits cylinder engine, and does not necessarily occur every cycle (see below). build up and provide additional thermal insulation: primary examples are spark Spark-knock is controllable by the spark advance: advancing the spark increases plugs, exhaust valves, metal asperities such as edges of head cavities or piston the knock severity or intensity and retarding the spark decreases the knock. Since bowls. Under normal conditions, using suitable heat-range spark plugs, preigni- surface ignition usually causes a more rapid rise in end-gas pressure and tem- tion is usually initiated by an exhaust valve covered with deposits coming from perature than occurs with normal spark ignition (because the flame either starts the fuel and from the lubricant which penetrates into the combustion chamber. propagating sooner, or propagates from more than one source), knock is a likely Colder running exhaust valves and reduced oil consumption usually alleviate this outcome following the occurrence of surface ignition. To identify whether or not problem: locating the exhaust valve between the spark plug and the end-gas surface ignition causes knock, the terms knocking surface ignition and non- region avoids contact with both the hottest burned gas near the spark plug and knocking surface ignition are used. Knocking surface ignition usually originates the end-gas. Engine design features that minimize the likelihood of preignition from preignition caused by glowing combustion chamber deposits: the severity of are: appropriate heat-range spark plug, removal of asperities, radiused metal knock generally increases the earlier that preignition occurs. Knocking surface edges, well-cooled exhaust valves with sodium-cooled valves as an extreme ignition cannot normally be controlled by retarding the spark timing, since the option. 75, 76 spark-ignited flame is not the cause of knock. Nonknocking surface ignition is Knock primarily occurs under wide-open-throttle operating conditions. It is usually associated with surface ignition that occurs late in the operating cycle. thus a direct constraint on engine performance. It also constrains engine effi- The other abnormal combustion phenomena in Fig. 9-58, while less ciency, since by effectively limiting the temperature and pressure of the end-gas, it common, have the following identifying names. Wild ping is a variation of knock- limits the engine compression ratio. The occurrence and severity of knock depend ing surface ignition which produces sharp cracking sounds in bursts. It is thought on the knock resistance of the fuel and on the antiknock characteristics of the to result from early ignition of the fuel-air mixture in the combustion chamber by engine. The ability of a fuel to resist knock is measured by its octane number: glowing loose deposit particles. It disappears when the particles are exhausted higher octane numbers indicate greater resistance to knock (see Sec. 9.6.3). Gas- and reappears when fresh particles break loose from the chamber surfaces. oline octane ratings can be improved by refining processes, such as catalytic Rumble is a relatively stable low-frequency noise (600 to 1200 Hz) phenomenon cracking and reforming, which convert low-octane hydrocarbons to high-octane associated with deposit-caused surface ignition in high-compression-ratio hydrocarbons. Also, antiknock additives such as alcohols, lead alkyls, or an engines. This type of surface ignition produces very high rates of pressure rise organomanganese compound can be used. The octane-number requirement of an following ignition. Rumble and knock can occur together. Run-on occurs when engine depends on how its design and the conditions under which it is operated the fuel-air mixture within the cylinder continues to ignite after the ignition affect the temperature and pressure of the end-gas ahead of the flame and the system has been switched off. During run-on, the engine usually emits knocklike time required to burn the cylinder charge. An engine's tendency to knock, as noises. Run-on is probably caused by compression ignition of the fuel-air defined by its octane requirement-the octane rating of the fuel required to avoid mixture, rather than surface ignition. Runaway surface ignition is surface ignition knock-is increased by factors that produce higher temperatures and pressures that occurs earlier and earlier in the cycle. It is usually caused by overheated or lengthen the burning time. Thus knock is a constraint that depends on both spark plugs or valves or other combustion chamber surfaces. It is the most the quality of available fuels and on the ability of the engine designer to achieve destructive type of surface ignition and can lead to serious overheating and struc- the desired normal combustion behavior while holding the engine's propensity to tural damage to the engine.74 knock at a minimum. 74 After some additional description of surface-ignition phenomena, the The pressure variation in the cylinder during knocking combustion indi- remainder of Sec. 9.6 will focus on knock. This is because surface ignition is a cates in more detail what actually occurs. Figure 9-59 shows the cylinder pressure problem that can be solved by appropriate attention to engine design, and fuel variation in three individual engine cycles, for normal combustion, light knock, and lubricant quality. In contrast, knock is an inherent constraint on engine and heavy knock, respectively.77 When knock occurs, high-frequency pressure performance and efficiency since it limits the maximum compression ratio that fluctuations are observed whose amplitude decays with time. Figures 9-59a and b can be used with any given fuel. have the same operating conditions and spark advance. About one-third of the Of all the engine surface-ignition phenomena in Fig. 9-58, preignition is cycles in this engine at these conditions had no trace of knock and had normal, potentially the most damaging. Any process that advances the start of com- smoothly varying, cylinder pressure records as in Fig. 9-59a. Knock of varying bustion from the timing that gives maximum torque will cause higher heat rejec- severity occurred in the remaining cycles. With light or trace knock, knock tion because of the increasing burned gas pressures and temperatures that result. occurs late in the burning process and the amplitude of the pressure fluctuations 454 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 455 sity of knock because it depends on the amount of end-gas which ignites sponta- neously and rapidly, and because engine damage due to knock results from the high gas pressures (and temperatures) in the end-gas region. Use of this measure Pressure of knock severity or intensity shows there is substantial variation in the extent of knock, cycle-by-cycle. Figure 9-60 shows the knock intensity in one hundred con- secutive cycles in a given cylinder of a multicylinder engine operating at fixed conditions for knocking operation. The intensity varies randomly from essentially no knock to heavy knock.79 Cylinder-to-cylinder variations are also substantial due to variations in compression ratio, mixture composition and conditions, burn 20 TC 20 40 OCA 20 TC 20 40 ºCA -20 TC 20 40 .CA rate, and combustion chamber cooling. One or more cylinders may not knock at all while others may be knocking heavily.80 (a) Normal combustion, (b) Slight knock, (c) Intense knock, spark 28ºBTC spark 28ºBTC spark 32ºBTC Since the knock phenomenon produces a nonuniform state in the cylinder, and since the details of the knock process in each cycle and in each cylinder are FIGURE 9-59 different, a fundamental definition of knock intensity or severity is extremely diffi- Cylinder pressure versus crank angle traces of cycles with (a) normal combustion, (b) light knock, and cult. The ASTM-CFR method for rating fuel octane quality (see Sec. 9.6.3) by the (c) heavy knock. 4000 rev/min, wide-open throttle, 381-cm3 displacement single-cylinder engine.77 severity of knocking combustion uses the time derivative of pressure during the cycle. Cylinder pressure is determined with a pressure transducer. The low- is small (Fig. 9-59b). With heavy knock, illustrated here with more advanced frequency component of pressure change due to normal combustion is filtered spark timing and by selecting an especially high intensity knocking cycle, knock out and the rate of pressure rise is averaged over many cycles during the pressure occurs closer to top-center earlier in the combustion process and the initial fluctuations following knock. This approach obviously provides only an average amplitude of the pressure fluctuation is much larger. These pressure fluctuations relative measure of knock intensity. The maximum rate of pressure rise has been produce the sharp metallic noise called "knock." They are the result of the essen- used to quantify knock severity. An accelerometer mounted on the engine can tially spontaneous release of much of the end-gas fuel's chemical energy. This give indications of relative knock severity provided that it is mounted in the same produces a substantial local increase in gas pressure and temperature, thereby location for all tests. The most precise measure of knock severity is the maximum causing a shock wave to propagate away from the end-gas region across the amplitude of the pressure oscillations that occur with knocking combustion. The combustion chamber. This shock wave, the expansion wave that accompanies it, cylinder pressure signal (from a high-frequency response pressure transducer) is and the reflection of these waves by the chamber walls create the oscillatory filtered with a band-pass filter so that only the component of the pressure signal pressure versus time records shown in Fig. 9-59b and c. Note that once knock that corresponds to the fluctuations occurring after knock remains. The filter is occurs, the pressure distribution across the combustion chamber is no longer set for the first transverse mode of gas vibration in the cylinder (in the 3 to uniform: transducers located at different points in the chamber will record differ- 10 KHz range, depending on bore and chamber geometry). The maximum ampli- ent pressure levels at a given time until the wave propagation phenomena tude of pressure oscillation gives a good indication of the severity of knock.81 described above have been damped out.78 The knock intensities in individual cycles shown in Fig. 9-60 were determined in Many methods of knock detection and characterization have been used. The human ear is a surprisingly sensitive knock detector and is routinely used in determining the octane requirement of an engine-the required fuel quality the 400 engine must have to avoid knock. Knock detectors used for knock control systems normally respond to the vibration-driven acceleration of parts of the 300 engine block caused by knocking combustion pressure waves. A high-intensity flash is observed when knock occurs; this is accompanied by a sharp increase in Knock intensity, kPa 200 ionization. Optical probes and ionization detectors have therefore been used. The spark plug can serve as an ionization detector. For more detailed studies of 100 knock in engines, the piezoelectric pressure transducer is the most useful moni- FIGURE 9-60 toring device. Often the transducer signal is filtered so that the pressure fluctua- Knock intensity (maximum amplitude of band- tions caused by knock are isolated.75 pass-filtered pressure signal) in one hundred indi- 20 40 60 80 100 vidual consecutive cycles. One cylinder of V-8 The amplitude of the pressure fluctuation is a useful measure of the inten- Firing cycle number engine, 2400 rev/min, wide-open throttle.79 456 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES . 457 this manner.79 Note that because the pressure fluctuations are the consequence of a wave propagation phenomenon, the location of the pressure transducer in rela- tion to the location of the knocking end-gas and the shape of the combustion chamber will affect the magnitude of the maximum recorded pressure-fluctuation amplitude. The impact of knock depends on its intensity and duration. Trace knock has no significant effect on engine performance or durability. Heavy knock can lead to extensive engine damage. In automobile applications, a distinction is .. ..... usually made between "acceleration knock" and "constant-speed knock." Accel- (a) eration knock is primarily an annoyance, and due to its short duration is unlikely (b) to cause damage. Constant-speed knock, however, can lead to two types of engine damage. It is especially a problem at high engine speeds where it is masked by other engine noises and is not easily detected. Heavy knock at con- stant speed can easily lead to: 1. Preignition, if significant deposits are present on critical combustion chamber components. This could lead to runaway preignition.+ 2. Runaway knock-spark-knock occurring earlier and earlier, and therefore more and more intensely. This soon leads to severe engine damage. 3. Gradual erosion of regions of the combustion chamber, even if runaway knock (c) (d) does not occur. FIGURE 9-61 Examples of component damage from abnormal engine combustion. (a) Piston holing by preigni- The engine can be damaged by knock in different ways: piston ring stick- tion;83 (b) piston crown erosion after 10 hours of high-speed knocking;82 (c) cylinder head gasket ing; breakage of the piston rings and lands; failure of the cylinder head gasket; splitting failure due to heavy knock;83 (d) erosion of aluminum cylinder head along the top of the cylinder head erosion; piston crown and top land erosion; piston melting and cylinder liner due to heavy knock. 83 holing. Examples of component damage due to preignition and knock are shown in Fig. 9-61.82-84 The mechanisms that cause this damage are thought to be the following. range. These high local pressures are combined with the higher-than-normal local Preignition damage is largely thermal as evidenced by fusion of spark plugs or surface temperatures which occur with the higher knocking heat fluxes and pistons. When knock is very heavy, substantial additional heat is transferred to weaken the material. Pitting and erosion due to fatigue with these excessive mechanical stresses, and breakage of rings and lands, can then occur 78, 82-84 the combustion chamber walls and rapid overheating of the cylinder head and piston results. Under these conditions, knock is not stable: the overheating increases the engine's octane requirement which in turn increases the intensity of 9.6.2 Knock Fundamentals knock. It becomes heavier and heavier, and the uncontrolled running away of this phenomena can lead to engine failure in minutes. This damage is due to As yet, there is no complete fundamental explanation of the knock phenomenon overheating of the engine: the piston and rings seize in the bore. The damage due over the full range of engine conditions at which it occurs. It is generally agreed to heavy knock over extended periods-erosion of piston crowns and (aluminum) that knock originates in the extremely rapid release of much of the energy con- cylinder heads in the end-gas region-is due primarily to the high gas pressures tained in the end-gas ahead of the propagating turbulent flame, resulting in high in this region. Extremely high pressure pulses of up to 180 atm due to heavy local pressures. The nonuniform nature of this pressure distribution causes pres- knock can occur locally in the end-gas region, in the 5 to 10 kHz frequency sure waves or shock waves to propagate across the chamber, which may cause the chamber to resonate at its natural frequency. Two theories have been advanced to explain the origin of knock: the autoignition theory and the detona- tion theory. The former holds that when the fuel-air mixture in the end-gas + Note that heavy knock can also remove deposits from the combustion chamber walls, thereby region is compressed to sufficiently high pressures and temperatures, the fuel oxi- decreasing the octane requirement of the engine. dation process-starting with the preflame chemistry and ending with rapid 458 INTERNAL COMBUSTION ENGINE FUNDAMENTALS energy release-can occur spontaneously in parts or all of the end-gas region. The latter theory postulates that, under knocking conditions, the advancing flame front accelerates to sonic velocity and consumes the end-gas at a rate much faster than would occur with normal flame speeds.75 These theories attempt to describe what causes the rapid release of chemical energy in the end-gas which No. 9 (46.00) No. 9 (22.9º) creates very high pressures, locally, in the end-gas region. The engine phenome- 40 non "knock" includes also the propagation of strong pressure waves across the chamber, chamber resonance, and transmission of sound through the engine 20 structure. The detonation theory has led many to call knock "detonation." 1 2 3 4 5 6789 No. 8 (43.6º) However, the more general term "knock" is preferred, since this engine pheno- Crank angle, deg ATC No. 8 (20.50) menon includes more than the end-gas energy release, and there is much less (knocking cycle) evidence to support the detonation theory than the autoignition theory as the initiating process. Most recent evidence indicates that knock originates with the 20 spontaneous or autoignition of one or more local regions within the end-gas. Cylinder pressure, MPa No. 7 (41.2º) No. 7 (18.1º) Additional regions (some adjacent to already ignited regions and some separated from these regions) then ignite until the end-gas is essentially fully reacted. This sequence of processes can occur extremely rapidly. Thus, the autoignition theory is most widely accepted. No. 6 (38.8º) Photographic studies of knocking combustion have been an important No. 6 (15.7º) Observation window for photography source of insight into the fundamentals of the phenomenon over the past fifty years.+ Figure 9-62 shows two sets of schlieren photographs, one from a cycle with normal combustion and the other from a cycle with knock.88 These pho- tographs were taken in an overhead valve engine with a disc-shaped combustion No. 5 (36.4º) No. 5 (13.30) chamber, with a window which permits observation of the chamber opposite to the spark plug. A reflecting mirror on the piston crown permits use of the sch- lieren technique which identifies regions where changes in gas density exist. Oper- Exhaust valve Intake valve Spark plug ating conditions, except for spark advance, were the same for both cycles. In the No. 4 (34.00) normal combustion sequence, the turbulent flame front moves steadily through No. 4 (10.90 the end-gas as combustion goes to completion. The cylinder pressure varies smoothly throughout this process. When the spark is advanced by 15º, the end-gas temperature and pressure are increased significantly and knock occurs. 60 In this sequence of photographs (b), the initial flame propagation process No. 3 (31. 60) No. 3 (8.50) (photographs 1 to 3) is like that of the normal combustion sequence: then almost 40 172 334 5 6789 Crank angle, deg ATC (normal cycle) No. 2 (29.2º) + See Refs. 85 to 87 for early high-speed photographic studies. See Refs. 88 to 91 for some recent No. 2 (6.1º) 20 studies. 2t Cylinder pressure, MPa FIGURE 9-62 No. 1 (26.8º ATC) No. 1 (3.7º ATC) Schlieren photographs from high-speed movies of (a) normal flame propagation through the end gas and (b) knocking combustion (autoignition occurs in photograph 4), with corresponding cylinder (a) pressure versus crank angle traces. Disc-shaped combustion chamber with details of window shown in (b) insert. 1200 rev/min, 80 percent volumetric efficiency, (A/F) = 12.5; spark timing: (a) 10º BTC, (b) 25º BTC.88 150 460 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 461 the entire region ahead of the flame appears dark (photograph 4). Between pho- nition sites varied with engine operating conditions, the majority of occurrences tographs 3 and 4 substantial changes in the density and temperature throughout in this study were in the vicinity of the cylinder wall. The rate of spread of the most of the end-gas region have occurred. Examination of the cylinder pressure autoignited end-gas region also varied significantly. Under heavy knocking con- trace shows that this corresponds closely to the time when the pressure recorded ditions the entire end-gas region became ignited very rapidly and high-amplitude by the pressure transducer rises rapidly. Immediately after this, pressure oscil- pressure oscillations occurred. Under trace knock conditions, autoignition could lations (at 6 to 8 kHz) are detected. In photograph 5 the flame is no longer occur; yet the spread of the autoignited region could be sufficiently slow for no visible: the end-gas expansion has pushed the flame back out of the field of view. pressure oscillations to be detected. Subsequent photographs are alternatively lighter and darker, indicative of chang- When the above-described end-gas ignition process occurs rapidly, the gas ing local density fields as pressure waves propagate back and forth across the pressure in the end-gas region rises substantially due to the rapid release of the chamber. end-gas fuel's chemical energy. The erosion damage that knock can produce, due More extensive studies of this type, which relate photographs from high- to stress-induced material fatigue as described in the previous section, indicates speed movies of the combustion process to the cylinder pressure development, the location of this high-pressure region. With the chamber geometry typical of indicate that the location or locations where ignition of one or more portions of most engines where the flame propagates toward the cylinder wall, the damage is the end-gas first occur and the subsequent rate with which the ignition process confined to the thin crescent-shaped region on the opposite side of the chamber develops throughout the rest of the end-gas vary substantially cycle-by-cycle and to the spark plug, where one expects the end-gas to be located. A shock wave with the intensity of the knocking process. Figure 9-63 shows five shadowgraph propagates from the outer edge of this high-pressure end-gas region across the photographs from a knocking engine cycle in a research engine similar to that chamber at supersonic velocity, and an expansion wave propagates into the high- shown in Fig. 9-62.89 The photographs are 33 us apart; the total sequence shown pressure region toward the near wall. The presence of such a shock wave has lasts 1º. The first photograph shows the flame front prior to onset of knock. The been observed photographically.90 The shock wave and expansion wave reflect second photograph shows the onset of autoignition with the appearance of dark off the walls of the chamber, eventually producing standing waves. Usually these regions near the wall (two identified by arrows) where substantial density gra- standing waves are due to transverse gas vibration and are of substantial ampli- dients resulting from local energy release exist. The third, fourth, and fifth pho- tude. The amplitude of the pressure oscillations builds up as the standing waves tographs show the spread of these ignited regions with time through the are established, and then decays as the gas motion is damped out. The frequency remaining end-gas. The exact location where autoignition occurred was identified of the pressure oscillations (normally in the 5 to 10 kHz range) decreases with with a photodigitizing system which ranked regions of the photograph by their time as the initially finite-amplitude supersonic pressure waves decay to small- brightness or darkness. The digitized version of the second photograph is also amplitude sound waves.8º Thus the pressure signal detected with a transducer shown in Fig. 9-63. Additional smaller regions of autoignition are evident adjac- during a knocking combustion cycle will depend on the details of the end-gas ent to the wall and in the vicinity of the flame front. While the location of autoig- ignition process, the combustion chamber geometry, and the location of the transducer in relation to the end-gas region. The pressure variation across the cylinder bore, due to knock, can be illus- trated by the following example. Consider the disc-shaped combustion chamber with the spark plug located in the cylinder wall shown in Fig. 9-64a. Figure 9-64b shows the pressure and temperature distribution across the combustion chamber due to the normal flame propagation process, at the time rapid ignition of the Flame, end-gas occurs. If the end-gas ignites completely and instantaneously, its pressure and temperature will suddenly rise, as shown. A shock wave will now propagate Autoignition sites to the right and an expansion wave to the left, as shown in the distance-time diagram in Fig. 9-64c. These waves reflect off the walls and interact. The pres- sures at the cylinder wall, in the end-gas region and on the opposite side of the Cylinder wall chamber at the spark plug, develop as shown in Fig. 9-64d. Figure 9-65 shows 18.83 19.07 19.31 19.5 19.79 two simultaneously recorded pressure traces with heavy knock from two trans- FIGURE 9-63 ducers at the top of the cylinder liner, one located near the spark plug and one in Five shadowgraph photographs of knocking combustion cycle identifying location of autoignition the end-gas region. In the end-gas region the pressure rises extremely rapidly sites (arrows). Crank angle of each photo indicated: 33 us between frames. Operating conditions and when knock occurs, to a value considerably higher than that recorded on the engine details as in Fig. 9-62. Photodigitized picture of second photograph showing additional details opposite side of the chamber where the pressure rises more gradually. Standing of the autoignition sites in the end-gas region on right.89 ....... 462 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 463 End gas Flame Burned gas 120 7- Spark plug Position 1 80 (a) 0.3 150 40 120 - Shock Cylinder pressure, atm Time, ms p - 100 Pi Expansion wave FIGURE 9-65 50 --- 80 Simultaneously recorded pressure Position 2 traces on opposite sides of spark- -1 0 0 R ignition engine combustion (c) -40 chamber in heavily knocking cycle. 3000 Position 1: at top of cylinder liner closest to spark plug. Position 2: at top of cylinder liner farthest 2500 T 180 o 200 from spark plug. 2000 rev/min, 220 240 260 280 T, K wide-open throttle, (A/F) = 12.3, Pressure Crank angle, deg spark advance 20º BTC.78 000 0 O R Autoignition leased by the reaction as heat is larger than the heat lost to the surroundings; as Time a result the temperature of the mixture increases, thereby rapidly accelerating, (b) (d) due to their exponential temperature dependence, the rates of the reactions FIGURE 9-64 involved. The state at which such spontaneous ignition occurs is called the self- Illustration of how cylinder pressure distribution develops following knock. (a) Schematic of disc- ignition temperature and the resulting self-accelerating event where the pressure shaped combustion chamber at time of knock. (b) Pressure and temperature across the diameter of a and temperature increases rapidly is termed a thermal explosion. disc-shaped engine combustion chamber, before and after end-gas autoignition (assumed to occur very rapidly, at constant end-gas density). (c) Schematic of shock and expansion wave pattern follow- In complex reacting systems such as exist in combustion, the "reaction" is ing end-gas autoignition on distance-time plot. (d) Pressure variation with time at cylinder wall in the not a single- or even a few-step process; the actual chemical mechanism consists end-gas region, and at the opposite side of the cylinder. of a large number of simultaneous, interdependent reactions or chain reactions. In such chains there is an initiating reaction where highly reactive intermediate species or radicals are produced from stable molecules (fuel and oxygen). This waves are then set up and the amplitudes of the oscillations decay as the waves step is followed by propagation reactions where radicals react with the reactant are damped out. 78 molecules to form products and other radicals to continue the chain. The process The fundamental theories of knock are based on models for the autoigni- ends with termination reactions where the chain propagating radicals are tion of the fuel-air mixture in the end-gas. Autoignition is the term used for a removed. Some propagating reactions produce two reactive radical molecules for rapid combustion reaction which is not initiated by any external ignition source. each radical consumed. These are called chain-branching reactions. When, due to Often in the basic combustion literature this phenomenon is called an explosion. chain-branching, the number of radicals increases sufficiently rapidly, the reac- Before discussing the relevant theories of hydrocarbon oxidation, the necessary tion rate becomes extremely fast and a chain-branching explosion occurs. While terminology will be defined and illustrated with the autoignition behavior of the the terms thermal and chain-branching explosions have been introduced sepa- much simpler hydrogen-oxygen system. For the hydrocarbons commonly found rately, in many situations the self-accelerations in temperature and radicals occur in practical fuels, the chemical reaction schemes by which the fuel molecules are simultaneously and the two phenomena must be combined.92, 93 broken down and react to form products are extremely complicated, and are as The oxidation of hydrogen at high pressures and temperatures provides a yet imperfectly understood. good illustration of these phenomena. For stoichiometric hydrogen-oxygen mix- The autoignition of a gaseous fuel-air mixture occurs when the energy re- tures, no reaction occurs below 400ºC unless the mixture is ignited by an external 464 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 465 source such as a spark; above 600ºC explosion occurs spontaneously at all or multiple cool flames (slightly exothermic reactions); two-stage ignition (cool pressures.+ flame followed by a hot flame); single-stage ignition (hot flame). Slow reactions The initiating steps proceed primarily through hydrogen peroxide (H2O2) are a low-pressure, low-temperature (<<200ºC) phenomenon not normally to form the hydroxyl radical (OH) at lower temperatures, or through dissociation occurring in engines. At 300 to 400ºC one or more combustion waves often of H2 at higher temperatures to form the hydrogen atom radical H. The basic appear, accompanied by faint blue light emission; the reaction is quenched, radical-producing chain sequence is composed of three reactions: however, when only a small fraction of the reactants have reacted and the tem- (R1) H + 02 = O + OH perature rise is only tens of degrees. These are called cool flames. Depending on conditions and the fuel, a cool flame may be followed by a "hot flame" or high- (R2) O + H2 = H + OH (9.56) temperature explosion where the reaction accelerates rapidly after ignition. This (R3) H2 + OH = H2O + H is termed two-stage ignition. As the temperature of the mixture increases, a tran- sition from two-stage to single-stage ignition occurs. While all hydrocarbons The first two reactions are chain-branching; two radicals are produced for each exhibit induction intervals which are followed by a very rapid reaction rate, some one consumed. The third reaction is necessary to complete the chain sequence: hydrocarbon compounds do not exhibit the cool flame or two-stage ignition starting with one hydrogen atom, the sequence R1 then R2 and R3 produces two behavior. H; starting with OH, the sequence R3, then R1, then R2 produces two OH. Since Figure 9-66 shows these ignition limits for isooctane, methane, and benzene. all three reactions are required, the multiplication factor is less than 2 but greater For isooctane, ignition in the low-temperature regions is by a two-stage process: than 1. Repeating this sequence over and over again rapidly builds up high con- there is a first time interval before the cool flame appears and then a second time centrations of radicals from low initial levels. interval from the appearance of the cool flame to the hot flame combustion However, these three reactions do not correspond to the overall stoichiom- process. Ignition in the high-temperature region is by a continuous one-stage etry process. The cool flame phenomena vary enormously with hydrocarbon struc- H2 + 102 = H20 ture. Normal paraffins give strong cool flames, branched-chain paraffins are more and other reactions must become important. In flames, the chain-branching resistant. Olefins give even lower luminosity cool flames with longer induction process ceases when the reverse of reactions R1-R3 become significant. A quasi periods. Methane shows only the high-temperature ignition limit, as indicated in equilibrium is established; while the overall process has proceeded a considerable Fig. 9-66. Benzene, also, does not exhibit the cool flame phenomenon and other way toward completion, a substantial amount of the available energy is still con- aromatics give hardly detectable luminosity.75 It is thought that some com- tained in the high radical concentrations. Over a longer time scale, this energy is released through three-body recombination reactions (the principal chain termi- 800 Isooctane nating reactions): Methane Benzene (R4) H + H + M = H2 + M (R5) O + 0 + M = O2 + M (9.57) 500 (R6) H + OH + M = H2O + M Ignition temperature, C -- + ---- M refers to any available third-body species, required in these recombination High-temperature region reactions to remove the excess energy.92 400 With this introductory background let us now turn to the autoignition of hydrocarbon-air mixtures. The process by which a hydrocarbon is oxidized can Cool flame r Low-temperature region exhibit four different types of behavior, or a sequential combination of them, depending on the pressure and temperature of the mixture: slow reactions; single 200 2 6 8 Pressure, atm FIGURE 9-66 + In between, three separate explosion limits-pressure-temperature boundaries for specific mixture Ignition diagrams for isooctane, methane, and benzene. Two-stage ignition occurs in the low- ratios of fuel and oxidizer that separate regions of slow and fast reaction-exist; these are not, temperature region; the first stage may be a cool flame. Single-stage ignition occurs in the high- however, of interest to us here. temperature region.94 466 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 467 organic radical (formed by abstracting a hydrogen atom from a fuel hydrocarbon molecule). However, at higher temperatures, ROOH is no longer the major C. product of the chain propagation process: instead, it is hydrogen peroxide, B , Ap H2O2. While H2O2 is relatively stable at lower temperatures, above 500ºC it Pressure - T2 decomposes into two OH radicals. An outline of the basic hydrocarbon (RH) oxidation process due to Semenov is as follows:93 (R1 RH + O2 -+ R + HO2 Chain initiation (R2) R + O2 + RÓ2 FIGURE 9-67 R + O2 Pressure records of the autoignition of isooctane-air (R3) > olefin + HO2 E mixture (( = 0.9) in a rapid compression machine. (R4) RO2 + RH ROOH + R Chain propagation B A -> B -+ C is the compression process. Top: two- Pressure total - stage ignition (at D, then at E) at postcompression (R5) RO > R'CHO + R"O (9.59) pressure of 1.86 MPa and temperature of 686 K. Bottom: single-stage ignition at E at post- (R6) HO2 + RH -+ H2O2 + R compression conditions of 2.12 MPa and 787 K. ( R7 ) ROOH RO + OH Vertical scale: 690 kPa/division. Horizontal scale: Degenerate branching Time 1 ms/division.95 (R8) R'CHO + O2 R'CO + HO2 (R9) RÒ2 destruction Chain termination pounds knock by a low-temperature two-stage ignition mechanism, some via a The dot denotes an active radical; each dash denotes the number of free bonds high-temperature single-stage ignition mechanism, and for some fuels both on the organic radical R. mechanisms may play a role. Reaction R1 is slow and explains the induction period in hydrocarbon com- Examples of these two mechanisms in rapid-compression machine experi- bustion. R2 is fast and of near-zero activation energy. R3 leads to olefins known ments, where a homogeneous isooctane-air mixture was compressed to different to occur in the oxidation of saturated hydrocarbons. R4 and R5 yield the main final conditions in a piston-cylinder apparatus and allowed to autoignite, are intermediates. The degenerate branching comes about from the delay in decom- shown in Fig. 9-67. A -> B- C is the piston-motion-produced compression. The position of the reactive species in R7 and R8. As one radical is used up to form top trace shows a well-defined cool flame at D, preceding hot ignition at E. The the reactants in R7 and R8, the multiple radicals do not appear until these reac- lower trace, at a higher temperature, shows a single-stage ignition process.95 tants decompose.93 A more extensive discussion of hydrocarbon oxidation Many of the above phenomena-long induction periods, initial slow mechanisms can be found in Benson.96 increase in reaction rate, two-stage ignition process-cannot be explained with The following evidence indicates the relevance of the above mechanism to simple mechanisms like the hydrogen oxidation process reviewed above. knock in engines. End-gas sampling studies have identified products of slow com- Although chain-branching reactions are taking place, the radical generation bustion reactions of isooctane; these principally include olefins, cyclic ethers, process must be more complex. Explanations of the long induction periods are aldehydes (R'CHO), and ketones (R"CO).97 Such studies have shown increasing based on the formation of unstable but long-lived intermediates. These interme- concentrations of peroxides (predominantly H2O2 with traces of organic diates can then either react to form stable molecules or to form active radicals, peroxides) with isoparaffinic fuels which show two-stage ignition behavior. the dominance and rate of either of these paths depending on the temperature: Higher temperature, single-stage ignition fuels such as benzene and toluene gave i.e. 93 no detectable peroxide. Aldehydes and ketones have been measured in significant Stable molecules and increasing concentrations in motored engines where the peak cycle tem- A - M * (9.58) perature was steadily increased. In motored engines, the occurrence of cool II Radicals flames, the two-stage nature of the autoignition ignition process at intermediate compression temperatures, and the transition to a single-stage ignition process at This is called a degenerate-branching mechanism. Hydroperoxides are an impor- very high compression ratios (i.e ., motored-engine gas temperatures with a peak tant metastable intermediate produced in the chain propagation process in the much higher than normal to simulate end-gas conditions in a firing engine) have low-temperature ignition process. They have the form ROOH, where R is an also been demonstrated. 75, 94 468 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 469 Two types of models of this autoignition process have been developed and piston (which run at higher temperatures than the water-cooled wall regions) used: (1) empirical induction-time correlations; (2) chemical mechanisms which could be at higher temperatures than adiabatically compressed mixture, due to embody many or all of the features of the "full" hydrocarbon oxidation process substantial heat transfer to the unburned mixture. Thus, the end-gas temperature given in Eq. (9.59). is not uniform and the distribution of temperature is extremely complex. Often, Induction-time correlations are derived by matching an Arrhenius function the mean unburned mixture temperature is used to characterize its state. Alterna- to measured data on induction or autoignition times, for given fuel-air mixtures, tively, the core temperature corresponding to adiabatic compression of mixture over the relevant mixture pressure and temperature ranges. It is then assumed from conditions at the start of compression is used. In the absence of substantial that autoignition occurs when heating by the exhaust valve and piston, the core temperature is a better repre- sentation of the maximum unburned mixture at any point in the cycle. dt = 1 (9.60) While more complex and complete chemical models of the autoignition process are being developed for simple paraffinic hydrocarbon fuel compounds where t is the induction time at the instantaneous temperature and pressure for (e.g ., Refs. 91 and 101), no detailed models are yet available for use with real the mixture, t is the elapsed time from the start of the end-gas compression blended fuels in engines. However, a generalized kinetic model for hydrocarbon process (t = 0), and t; is the time of autoignition. This equation can be derived by oxidation based on a degenerate branched-chain mechanism, known as the Shell assuming that the overall rate of production of the critical species in the induc- model, has been developed and tested with some success. The model uses generic tion period chemistry, for a given mixture, depends only on the gas state and that chemical entities representative of a variety of individual species which undergo a the concentration of the critical species required to initiate autoignition is fixed set of generalized reactions. This is justified by the broadly similar (though (i.e ., independent of the gas state).98 complex) ignition behavior of a variety of different fuel molecules and the similar A number of empirical relations for induction time for individual hydrocar- kinetics exhibited by the organic radicals of the same type in the hydrocarbon bons and blended fuels have been developed from fundamental or engine studies oxidation process [Eq. (9.59)].95. 102 of autoignition (see Ref. 99). These relations have the form The Shell model is based on a generic eight-step degenerate chain- branching reaction scheme. The scheme involves the fuel (RH), oxygen, radicals t = Ap " exp (9.61) formed from the fuel (R), products (P), intermediate product (Q), and degenerate- branching agent (B). The rate constants are either fixed at values consistent with where A, n, and B are fitted parameters that depend on the fuel. The ability of the literature or fitted so that measured induction times (such as those illustrated these types of equations to predict the onset of knock with sufficient accuracy is in Fig. 9-67) are adequately predicted.102 An example of results obtained with unclear. The most extensively tested correlation is that proposed by Douaud and this scheme is shown in Fig. 9-68. It shows the calculated pressure, temperature, Eyzat:100 and species concentrations in the end-gas region of an operating spark-ignition ON 3.402 t = 17.68 p - 1.7 exp 3800 100 T (9.62) 10-4 2300| 100- e where t is in milliseconds, p is absolute pressure in atmospheres, and T is in WWT. kelvin. ON is the appropriate octane number of the fuel (see Sec. 9.6.3). If the 1900- 80 - 10-7 temperature and pressure time history of the end-gas during an individual cycle are known, Eqs. (9.60) and (9.62) together can be used to determine whether 1.500 60 [R] -10-9 Concentration, mole/cm3 Pressure, bar Temperature, K autoignition occurs before the normally propagating flame consumes the end-gas. -[R] 1100 40 [B]+ An important question with any model of the end-gas autoignition process 10-11 is characterizing the end-gas temperature. During intake, the combustion p 700 20 10-13 chamber walls are hotter than the entering gases: thus, heat is transferred from the walls to the fresh mixture. During compression, the mixture temperature rises 300 OL 10-15 to levels substantially above the wall temperature. A thermal boundary layer will 145 155 165 175 185 195 205 215 build up adjacent to the wall, as heat is now transferred to the wall. Unburned Crank angle, deg ABC mixture away from the wall will be compressed essentially adiabatically (see Sec. FIGURE 9-68 12.6.5). In addition, any unburned mixture which for some portion of time during Pressure, temperature, and composition in the end-gas, before, during, and after autoignition, predict- intake and compression has been in close proximity to the exhaust valve and ed by the Shell model in spark-ignition engine combustion process. B, Q, R defined in text.103 470 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 471 engine, leading up to knock at 209º ABC. In this example, the autoignition model 8. Lengthening the side chain attached to the basic ring structure increases the has been incorporated in a multidimensional model of the flow and flame propa- knocking tendency in both groups of fuels, whereas branching of the side gation processes within the combustion chamber (see Sec. 14.5).103 chain decreases the knocking tendency. Figure 9-69 identifies the magnitude of these trends on a plot of the critical 9.6.3 Fuel Factors compression ratio against the number of carbon atoms in the molecule. The strong dependence of knocking tendency on fuel molecular size and structure is The tendency to knock depends on engine design and operating variables which apparent. influence end-gas temperature, pressure, and the time spent at high values of Practical fuels are blends of a large number of individual hydrocarbon com- these two properties before flame arrival. Thus, for example, the tendency to pounds from all the hydrocarbon series of classes: alkanes (paraffins), cyclanes knock is decreased through reductions in the end-gas temperature that follow (napthenes), alkenes (olefins), and aromatics (see Sec. 3.3). A practical measure from decreasing the inlet air temperature and retarding the spark from MBT of a fuel's resistance to knock is obviously required.107, 108 This property is timing. However, knock is a phenomenon that is governed by both engine and defined by the fuel's octane number. It determines whether or not a fuel will knock fuel factors; its presence or absence in an engine depends primarily on the anti- in a given engine under given operating conditions: the higher the octane knock quality of the fuel. number, the higher the resistance to knock. Octane number is not a single-valued Individual hydrocarbon compounds vary enormously in their ability to quantity, and may vary considerably depending on engine design, operating con- resist knock, depending on their molecular size and structure. Their tendency to ditions during test, ambient weather conditions during test, mechanical condition knock has been measured by the critical compression ratio of an engine: i.e ., the of engine, and type of oil and fuel used in past operation. The octane number compression ratio at which, under specified operating conditions, the specific fuel (ON) scale is based on two hydrocarbons which define the ends of the scale. By compound will exhibit incipienteock. Knocking tendency is related to molecu- definition, normal heptane (n-C7H16) has a value of zero and isooctane (C8H18: lar structuret as follows:105, 106 2,2,4-trimethylpentane) has an octane number of 100. These hydrocarbons were chosen because of the great difference in their ability to resist knock and the fact Paraffins that isooctane had a higher resistance to knock than any of the gasolines avail- 1. Increasing the length of the carbon chain increases the knocking tendency. able at the time the scale was established. Blends of these two hydrocarbons 2. Compacting the carbon atoms by incorporating side chains (thereby short- define the knock resistance of intermediate octane numbers: e.g ., a blend of 10 ening the length of the basic chain) decreases the tendency to knock. percent n-heptane and 90 percent isooctane has an octane number of 90. A fuel's 3. Adding methyl groups (CH3) to the side of the basic carbon chain, in the octane number is determined by measuring what blend of these two hydrocar- second from the end or center position, decreases the knocking tendency. bons matches the fuel's knock resistance. Several octane rating methods for fuels have been developed. Two of Olefins these-the research method (ASTM D-2699)+ and the motor method (ASTM 4. The introduction of one double bond has little antiknock effect; two or three D-2700)-are carried out in a standardized single-cylinder engine. In the motor double bonds generally result in appreciably less knocking tendency. method, the engine operating conditions are more severe; i.e ., the conditions are 5. Exceptions to this rule are acetylene (C2H2), ethylene (C2H4), and propylene more likely to produce knock. In addition, road octane rating methods have been (C3H6), which knock much more readily than the corresponding saturated developed to define the antiknock quality of fuels in cars operated on the road or hydrocarbons. on chassis dynamometers. The engine used in the ASTM research and motor Napthenes and aromatics methods is the single-cylinder engine developed under the auspices of the Co- 6. Napthenes have significantly greater knocking tendency than have the corre- operative Fuel Research Committee in 1931-the CFR engine.+ This test engine is a robust four-stroke overhead-valve engine with an 82.6-mm (3.25-in) bore and sponding size aromatics. 7. Introducing one double bond has little antiknock effect; two and three double 114.3-mm (4.5-in) stroke. The compression ratio can be varied from 3 to 30 while bonds generally reduce knocking tendency appreciably. + ASTM denotes American Society for Testing and Materials; the letter and number defines the specific testing code. See Sec. 3.3 for a review of hydrocarbon structure and its nomenclature. A more extensive dis- * The Cooperative Fuel Research Committee is now the Coordinating Research Council, Inc. cussion is given by Goodger.104 472 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 473 TABLE 9.6 Operating conditions for research and motor methods Research method Motor method Inlet temperature 52ºC (125ºF) 149ºC (300ºF) Inlet pressure Atmospheric 14- Humidity 0.0036-0.0072 kg/kg dry air Coolant temperature 100ºC (212ºF) Engine speed 600 rev/min 900 rev/min Aromatics Spark advance 13º BTC 19-26º BTC (constant) (varies with compression ratio) Air/fuel ratio Adjusted for maximum knock 12- -C 1 c c the engine is operating, with a mechanism which raises or lowers the cylinder and - n-paraffins C-C+c-c cylinder head assembly relative to the crankcase. A special valve mechanism 10 C maintains a constant tappet clearance with vertical adjustment of the head. The 6 engine is equipped with multiple-bowl carburetors so two reference fuels (usually 8 blends of n-heptane and isooctane) and the fuel being rated can be placed in separate bowls. By means of a selector valve, the engine can be operated on any Critical compression ratio €-C c-C of the three fuels. The engine operating conditions of the research and motor methods are summarized in Table 9.6. The test conditions are chosen to represent B the engine operating range where knock is most severe. With the fuel under test, 2) Isooctane the fuel/air ratio is adjusted for maximum knock. The compression ratio is then adjusted to produce knock of a standardized intensity, as measured with a mag- -0 -@-C-C netostriction knock detector. The level of knock obtained with the test fuel is 4) bracketed by two blends of the reference fuels not more than two octane numbers 6 C-C=C-C apart (with one knocking more and one less than the test fuel). The octane C-c =C number of the gasoline is then obtained by interpolation between the knock- C-C&C-C (3) meter scale readings for the two reference fuels and their octane numbers. For fuels below 100 ON, the primary reference fuels are blends of isooctane and n- 3 C-C-C-C heptane; the percent by volume of isooctane in the blend is the octane number. 4 For fuels above 100 ON, the antiknock quality of the fuel is determined in terms ---- of isooctane plus milliliters of the antiknock additive, tetraethyl lead, per U.S. C-C- C-C C gallon.+ C=C C-C-C+C-C-C The octane ratings of several individual hydrocarbon compounds and n-heptane common blended fuels are summarized in App. D, Table D.4. Practically all fuels C-C-C-€-C-C-C exhibit a difference between their research and motor octane numbers. The motor 3 4 5. 6 7 9 10 method of determining ON uses more severe operating conditions than the Number of carbon atoms FIGURE 9-69 Critical compression ratio (for incipienteock at 600 rev/min and 450 K coolant temperature) as a function of number of carbon atoms in hydrocarbon molecule, illustrating the effects of changes in t The octane number of the fuel is calculated from molecular structure. (Developed from Lovell.105) ON = 100 + 28.28T/ [1.0 + 0.736T + (1.0 + 1.472T - 0.03521672)1/2], where T is milliliters of tetraethyl lead per U.S. gallon. Tetraethyl lead, (C2H3)4Pb, contains 64.06 weight percent lead; 1 ml of TEL contains 1.06 grams of lead. COMBUSTION IN SPARK-IGNITION ENGINES 475 474 INTERNAL COMBUSTION ENGINE FUNDAMENTALS research method (higher inlet mixture temperature, more advanced timing). Thus, Modern gasolines contain a number of chemical additives designed to the motor octane number (MON) is usually lower than the research octane improve fuel quality. These additives are used to raise the octane number of the number (RON). The numerical difference between these octane numbers is called fuel, control surface ignition, reduce spark plug fouling, resist gum formation, prevent rust, reduce carburetor icing, remove carburetor or injector deposits, the fuel sensitivity: minimize deposits in the intake system, and prevent valve sticking. The octane Fuel sensitivity = RON - MON (9.63) number of hydrocarbon fuels can be increased by antiknock agents. Their use Fuel sensitivity varies with the source of crude petroleum and refining pro- generally allows an increase in antiknock quality to be achieved at less expense cesses used. The primary reference fuels themselves (mixtures of isooctane and than modifying the fuel's hydrocarbon composition by refinery processing. The n-heptane), by definition, have the same octane numbers by both the research most effective antiknock agents are lead alkyls. Tetraethyl lead (TEL), (C2H5)4 and motor methods. Since the primary reference fuels are paraffins, we would Pb, was first introduced in 1923. Tetramethyl lead (TML), (CH3)4Pb, was intro- expect other paraffins to have little or no sensitivity. In contrast, olefins and duced in 1960. Since TML boils in the mid-range of a gasoline (110ºC), whereas aromatics have high sensitivity. In general, therefore, straight-run gasolines con- TEL boils at the high end (200ºC), the introduction of TML permitted better taining high percentages of saturated hydrocarbons have low sensitivity, while distribution of octane amongst the cylinders of an engine. In 1959 a manganese cracked or reformed gasolines containing large percentages of unsaturated hydro- antiknock compound (methylcyclopentadienyl manganese tricarbonyl), now carbons have high sensitivity. Fuels having high sensitivity generally, but not known as MMT, was introduced as a supplementary antiknock agent for TEL. It always, have lower road octane ratings (i.e ., octane ratings determined in cars in is also an antiknock agent in its own right. on-the-road use) than do low-sensitivity fuels of the same research octane About 1970, in the United States, low-lead and unleaded gasolines were number. Regular grade unleaded gasoline typically has a RON of at least 91 and introduced. Two factors influenced the reduction in the use of lead alkyls: a MON of about 83, giving a sensitivity of 8. concern about the toxicological aspects of lead in the urban environment and the Research and motor octane number fuel ratings are made in a single- use of catalytic devices for emission control that are poisoned by lead. Unleaded cylinder engine run at constant speed, wide-open throttle and fixed spark timing. and reduced-lead-content gasolines are now required in the United States. Japan These methods do not always predict how a fuel will behave in an automobile has almost completely converted to unleaded fuel. In Europe, requirements which engine operated under a variety of speed, load, and weather conditions. Several reduce the lead content of gasolines and introduce unleaded fuel were implement- methods of rating a gasoline in actual vehicles, either on the road or on chassis ed in the late 1980s. MMT is sometimes used as an antiknock additive in dynamometers which duplicate outdoor road conditions, have, therefore, been unleaded gasoline. However, its role as a deposit in plugging exhaust catalytic developed. These methods determine the fuel's road octane number.107 The road converters limits its use to low concentrations. The expanding use of unleaded ratings of current fuels usually lie between the motor and research ratings. Road fuels has increased interest in other methods of boosting the octane rating of octane number can be related to motor and research ratings with equations of gasolines. The use of oxygenates-alcohols and ethers-as gasoline extenders, the form which due to their excellent antiknock quality increase the fuel's octane rating, is Road ON = a (RON) + b (MON) + c becoming more common. A brief review of the mechanism by which these addi- tives and compounds improve the knock resistance of gasolines follows. where a, b, and c are experimentally derived constants. Recent studies show Lead is the most effective antiknock element known. In the form of lead a & b ~ 0.5 gives good agreement. An antiknock index which is the mean of the alkyls, it is stable and fuel-soluble. The precise mechanism by which lead alkyls research and motor octane numbers is now used in the United States to charac- control knock is not fully known. It is generally agreed that the alkyls decompose terize antiknock quality: before they exert their antiknock action. The decomposed material-lead oxide, RON + MON Antiknock index = 9.64) PbO-either as a vapor or as a fog (a dispersion of fine particles), inhibits the 2 preflame chain-branching reaction which leads to autoignition of the fuel-air charge, thus slowing the reaction rate. However, lead has little effect on two-stage Refiners and automobile manufacturers are interested in the octane number ignition until after the cool flame.109 Commercially available tetraalkyl lead anti- requirement of engines or vehicles on the road. The octane number requirement knock fluids are based on TEL, TML, physical mixtures of TEL and TML, and (usually abbreviated to OR) of an engine or vehicle-engine combination is defined mixed ethyl-methyl compounds produced by reacting TML and TEL. This range as the minimum fuel octane number that will resist knock throughout the of compounds offers volatilities between the extremes of TEL and TML. The engine's operating speed and load range. The octane number requirement of a individual alkyls vary in antiknock behavior as a function of fuel composition single engine or vehicle does not usually provide adequate information for that and combustion conditions. The average effect of various amounts of TEL in a particular model; every model has a range of requirements due to production large number of regular gasolines is summarized in Fig. 9-70: the effectiveness of tolerances and variations in engine and vehicle condition. 476 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 477 g Pb per U.S. gallon 2 3 CH3OH), ethanol (C2H,OH), tertiary butyl alcohol (TBA) (C4H,OH), and methyl tertiary butyl ether (MTBE). Table 9.7 lists the antiknock characteristics of these compounds and their physical and chemical characteristics relative to gasoline. The blending value of antiknock index [Eq. (9.64)] given in the table is not necessarily the same as the ORNTTTTTIO Increase in RON and MON compound's antiknock index [(R + M)/2 from App. D, Table D.4] when used FIGURE 9-70 alone as a fuel, and depends on the gasoline composition with which the com- Gasoline octane number increase resulting from pound is blended. MTBE-gasoline blends have good water stability, and MTBE 0.2 0.4 0.6 0.8 1.0 1.2 use of antiknock additive tetraethyl lead. Varies has little effect on vapor pressure and material compatibility. TBA is moderately g Pb per liter with fuel composition: average values shown. susceptible to water extraction and loss. Ethanol is technically feasible as a high octane supplement or substitute for gasoline; economically it is less attractive than methanol. each successive increment added steadily decreases. The addition of about 0.8 g Methanol, because it can be made from natural gas, coal, or cellulose lead per liter (3 g Pb per U.S. gallon, the maximum economic limit to lead materials, has near- and long-term potential. Its high octane quality (130 RON, concentration) provides an average gain of about 10 octane numbers in modern 95 MON), when used in low-concentration (~5 percent) methanol-cosolvent- gasolines, though effectiveness varies with chemical composition of the fuel. TML gasoline blends, can help offset the octane loss from lead alkyl phase-out. Prob- offers a greater octane number gain than TEL in many gasolines, particularly in lems with these blends include poor solubility in gasoline in the presence of highly aromatic fuels with a low sulfur content. One of its major values, however, water; toxicity; an energy content about half that of gasoline; high latent heat of is in overcoming engine knock that results from fuel component segregation in vaporization and oxygen content which contribute to poor driveability; incom- the intake manifold; gasoline fractions of different volatility separate in the intake patibility with many commonly used metals and elastomers; blending effects on manifold of a multicylinder engine and the heavier fractions lag behind (see Sec. gasoline volatility which may force the displacement of large volumes of butane. 7.6.3). When a gasoline containing a lead alkyl is burned in a spark-ignition Some of these problems can be partially reduced by using cosolvents such as engine, it produces nonvolatile combustion products. These deposit on the walls TBA or isobutanol. Use of methanol as a neat fuel in specially designed engines of the combustion chamber and on the spark plug, causing lead-fouling of the permits advantage to be taken of its high octane rating via high compression spark plug electrodes and tracking across the plug insulator, and hot corrosion of ratios. Problems include its energy content of one-half that of gasoline; engine the exhaust valve. Commercial antiknock fluids, therefore, contain scavenging starting problems which require starting aids such as 5 to 10 percent isopentane agents-combinations of ethylene dibromide and ethylene dichloride which at temperatures below 10ºC or intake system heaters; toxicity; extensive engine transfer the lead oxides which would otherwise deposit into volatile lead-bromine modifications. 110 compounds which are largely exhausted with the combustion gases. The octane requirement of an engine-vehicle combination usually increases Low concentrations of methylcyclopentadienyl manganese tricarbonyl during use, primarily due to the buildup of combustion chamber deposits within (MMT) act as an octane promoter; it is most effective in highly paraffinic gas- the engine cylinder. While these deposits increase the engine's compression ratio olines. It is sometimes used as a supplement to TEL. MMT-TEL antiknock fluids modestly, their largest effect is to increase the temperature of the outer surface of are tailored to optimize octane cost effectiveness and the average fluid contains about 0.03 g Mn/g Pb. The octane gains vary significantly with fuel composition. TABLE 9.7 In unleaded fuels MMT is sometimes used in low concentrations to provide from Oxygenate properties 110 0.5 to 1 octane number gain. On a weight of metal basis, MMT is about twice as effective as TEL in terms of research octane number gain and about equally Methanol Ethanol TBA MTBE Gasoline effective in terms of motor octane number gain.109 The use of oxygenates (oxygen containing organic compounds) as extenders Typical (R + M)/2 blending value 112 110 98 105 87-93 Weight percent oxygen 50 35 22 18 0 or substitutes for gasoline is increasing.11º In some cases, this is because the Stoichiometric (A/F) 6.5 9.0 11.2 11.7 14.5 oxygenate can be produced from nonpetroleum sources (e.g ., biomass, coal) and Specific gravity 0.796 0.794 0.791 0.746 0.74 thus may offer strategic or economic benefits. In other cases, the good antiknock Lower heating value, MJ/kg 20.0 26.8 32.5 35.2 44.0 blending characteristics of oxygenates can aid in meeting octane quality demands Latent heat of vaporization, MJ/kg 1.16 0.84 0.57 0.34 0.35 where increasingly stringent regulations limit lead alkyl use. Several oxygenates Boiling temperature, ºC 65 78 83 55 27-227 have been used as automotive fuels; those of major interest are methanol TBA, tertiary butyl alcohol; MTBE, methyl teriary butyl ether. 478 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 479 the combustion chamber. This increases heat transfer to the fresh mixture during (c) Estimate the ratios of the maximum indicated mean effective pressure, at the induction and decreases heat transfer from the unburned charge during compres. same conditions as in (b), of the natural gas and methanol fueled engines to the sion. End-gas temperatures are therefore higher, thus increasing the likelihood of maximum imep of the gasoline engine. knock. As the combustion chamber deposits stabilize (over 15,000 to 25,000 km (d) If the methane engine burns 33 percent faster than the gasoline engine (i.e ., its of driving), the engine's octane requirement typically increases by about 5 octane combustion event takes three-quarters of the time), sketch a carefully drawn numbers; the increase can vary from between 1 to over 13 octane numbers. qualitative cylinder pressure versus crank angle curve for the two engines, from Essentially all of the octane requirement increase results from the buildup halfway through the compression stroke to halfway through the expansion stroke. Put both curves on the same graph. Conditions are as in (b). Show the of deposits on the combustion chamber walls: when the deposits are completely motored and firing pressure curves for each engine. The spark timing should be removed from the engine the octane requirement returns to close to its original set for maximum brake torque for each engine. Show the location of spark value. The volume of deposits for leaded and unleaded fuels which build up timing, the location of maximum cylinder pressure, and the approximate location inside the engine cylinder are similar in magnitude, and are in the range 0.3 to of the end of combustion. 1 cm3 per cylinder. The compression ratio increase associated with this volume of deposits is small (0.1 compression ratio) and contributes therefore on the order of 10 percent of the octane requirement increase. The primary effect of the deposits Fuel properties is thought to be changes in heat transfer between the end-gas and the combustion ¢ at chamber walls, as explained above. In an experiment where the deposits were Research lean Stoichiometric Heating removed from various regions of the combustion chamber in sequence and the octane operating air/fuel value, octane requirement after each removal was determined, it was found that the Fuel Formula number limit ratio MJ/kg deposits on the cylinder head around the end-gas region were the cause of about Natural CHA 120 0.7 17.2 50 0.78 two-thirds of the total ORI observed. Though the volume of the deposits for gas leaded and unleaded fuels are comparable, the composition and density are sub- Methanol CH3OH 105 0.8 6.4 Gasoline (CH2). 95 0.9 20 0.75 stantially different. For unleaded fuels the major element is carbon (one-third to 14.9 0.85 one-half by mass). The deposit density with leaded fuels is 2 to 5 g/cm3; with ¢ = fuel/air equivalence ratio. unleaded fuels it is a factor of 5 lower. 111 9.2. The figure shows the flame propagating radially outward from the center of a disc- PROBLEMS shaped combustion chamber. Combustion in such a device has many features in common with spark-ignition engine combustion. The chamber diameter is 10 cm 9.1. The table gives relevant properties of three different spark-ignition engine fuels. The and the height is 1.5 cm. For this problem, the flame can be thought of as a thin design and operating parameters of a four-cylinder 1.6-liter displaced volume engine cylindrical sheet. Its radius increases approximately linearly with time. The volume are to be optimized for each fuel over the engine's operating load and speed range. of the chamber is constant. The fuel is a typical hydrocarbon fuel; the mixture is You may assume that for each engine-fuel combination, the gross indicated fuel stoichiometric; the initial temperature is room temperature. conversion efficiency and imep at any operating condition are given by 0.8 times the (a) Sketch qualitative but carefully proportioned graphs of the following quantities fuel-air cycle efficiency at those conditions. Also, assume that for every five research versus time from the start of combustion to the end of combustion: octane number increase above 95 (the gasoline value) in fuel antiknock rating, the (1) The ratio of actual pressure in the chamber to the initial pressure compression ratio can be increased by one unit. For gasoline, the engine compres- (2) The ratio of average density of the gas ahead of the flame to the initial density sion ratio is 9, so if the octane number of the fuel is 100, a compression ratio of 10 (3) The ratio of the average density of the gas behind the flame to the initial can be used. density (a) At part-throttle operation-at an intake pressure of 0.5 atm and a speed of 2500 (b) On a qualitative but carefully proportioned graph of r/Ro versus time show how rev/min -- estimate the gross indicated fuel conversion efficiency and specific fuel the radial positions of gas elements, initially at r/Ro = 0, 0.5, and 1.0 before com- consumption for each engine-fuel combination. Each engine-fuel combination bustion, change during the combustion process as the flame propagates radially operates at the lean limit given. outward from the center (r = 0) to the outer wall (r = Ro). (b) At the appropriate equivalence ratio for maximum power, with 1 atm inlet pres- Note: Accurate numerical calculations are not required to answer this question. You sure at the same speed (2500 rev/min), the volumetric efficiencies are as shown. will be graded on the amount of physical insight your diagrams and the supporting Explain these volumetric efficiency values. Each fuel is fully vaporized in the inlet brief explanations of your logic communicate. You should write down any equations manifold and the inlet mixture temperature is held constant for all fuels. Mani- or approximate numerical values for relevant quantities that help explain your fold and valve design remains the same. reasoning. COMBUSTION IN SPARK-IGNITION ENGINES 481 480 INTERNAL COMBUSTION ENGINE FUNDAMENTALS Disc-shaped chamber RO Flame Compact chambers FIGURE P9-3 (d) The efficiency of the compact-chamber engine at part-throttle conditions is Gas element- higher than that of the conventional engine. Briefly explain why this is the case at radius r and estimate approximately the ratio of the two engine efficiencies. FIGURE P9-2 9.4. In a spark-ignition engine, a turbulent flame propagates through the uniform pre- mixed fuel-air mixture within the cylinder and extinguishes at the combustion chamber walls. 9.3. Highly compact bowl-in-piston or bowl-in-head combustion chambers permit SI (a) Draw a carefully proportioned qualitative graph of cylinder pressure p and mass engine operation at higher-than-normal compression ratios and with leaner mix- fraction burned x, as a function of crank angle 0 for -90º < 0 < 90º for a tures. A vigorous swirling flow in the combustion chamber just prior to combustion typical SI engine at wide-open throttle with the spark timing adjusted for is achieved by these chamber designs. The table compares the characteristics of a maximum brake torque. Mark in the crank angles of spark discharge, and of the compact-chamber spark-ignition engine (Fig. P9-3) with a more conventional spark- flame development period (0 to 10 percent burned) and end of combustion, on ignition engine which has an open chamber (see Fig. 9-4). both p and x, versus 0 curves relative to the top-center crank position. (b) Estimate approximately the fraction of the cylinder volume occupied by burned gases when the mass fraction burned is 0.5 (i.e ., halfway through the burning Compact chamber Conventional process). 1.5 liter 1.5 liter (c) A simple model for this turbulent flame is shown on the left in Fig. P9-4. The rate of burning of the charge dmb/dt is given by Compression ratio 14: 1 9:1 Fuel octane requirement 97 RON 97 RON Maximum bmep 980 kPa 965 kPa amb = A ,Pu ST dt Air/fuel ratio at WOT 17:1 13:1 MBT timing at 2000 rev/min 5º BTC 25º BTC where As is the area of the flame front (which can be approximated by a portion and WOT of a cylinder whose axis is at the spark plug position), p, is the unburned mixture Air/fuel ratio at cruise 22 : 1 16:1 conditions Spark Flame For gasoline, the stoichiometric air/fuel ratio is 14.6. RON = research octane A B number, WOT = wide-open throttle, MBT = maximum brake torque. Use the data provided, and any additional quantitative information you have *Spark plug or can generate easily, to answer the following questions: Spark (a) Explain whether the combustion process rate in the compact-chamber engine B Radius r will be faster, slower, or about the same rate as the conventional engine com- + bustion process. (b) The compact-chamber engine has a higher compression ratio than the conven- tional engine, yet it has about the same maximum bmep at WOT in the mid- Burned speed range. Explain quantitatively how the differences in compression ratio, Unburned air/fuel ratio, and MBT spark timing at WOT between the two engines influence maximum bmep and result in negligible total change. (c) Both engines have been designed to have the same fuel octane requirement, yet the compact-chamber engine has a much higher compression ratio. Explain A which features of the engine allow it to operate without knock at a 14 : 1 com- C pression ratio while the conventional engine can only operate without knock at FIGURE P9-4 9 :1. 482 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 483 density, and ST is the turbulent flame speed (the speed at which the front moves higher-compression-ratio engine using "X" will be more efficient, have about the relative to the unburned mixture ahead of it). The rate of mass burning is influ same efficiency, or be less efficient than the faster-burning lower-compression- enced therefore by combustion chamber geometry (through A, ) as well as those ratio gasoline engine. factors that influence ST (turbulent intensity, fuel/air ratio, residual gas fraction 9.6. Knock in spark-ignition engines is an abnormal combustion condition. Almost and EGR). Compare combustion chambers A and B shown in Fig. P9-4. Sketch everyone who rides a motorcycle or drives a car experiences this phenomenon at the approximate location of the flame front when 50 percent of the mass has been some time and usually changing into a lower gear will take the engine away from burned. (A careful qualitative sketch is sufficient; however, provide a quantitative this condition. Use your experience of what changes in other variables do, and justification for your sketch.) Sketch the mass fraction burned versus crank angle consult this and other textbooks to complete a table with the dependent variables curves for these two combustion chambers on the same graph, each with its shown at the top of the columns. spark timing set for maximum brake torque. You may assume the value of ST is The independent variables are: speed, compression ratio, chamber surface/ the same for A and B. volume ratio, spark plug distance from cylinder axis, percent EGR, inlet mixture (d) Compare combustion chambers A and C in Fig. P9-4 which have the same flame temperature, inlet mixture pressure, (F/A), wall temperature, air swirl, squish motion, travel distance but have different chamber shapes. Which chamber has fuel octane number. In these columns show the corresponding influence on the (1) the faster rate of mass burning during the first half of the combustion dependent variables by a " + " for an increase and a " - " for a decrease. Show the process; effect of increase in engine system independent variables on: cylinder pressure and (2) the faster rate of mass burning during the second half of the combustion temperature, flame speed, total burn time, autoignition induction period, tendency to process; knock. Provide in the extreme right-hand column brief comments to explain your (3) the more advanced timing for maximum brake torque? answers. Explain your answers. 9.5. A new synthetic fuel with chemical formula (CH2O) ,, is being developed from coal Dependent variables for automotive use. You are making an evaluation of what changes in the spark- Effect of increase ignition engines you produce might be required if gasoline is replaced by this fuel in engine "X." First make the following calculations: independent Cylinder Total Tendency (a) What is the stoichiometric air/fuel ratio for "X"? How does this compare with variable Flame burn Induction to the stoichiometric air/fuel ratio for gasoline? Pressure Temperature speed time period knock Explanation (b) Given that in a constant-volume calorimeter experiment to determine the heating Speed, value of "X" the combustion of 50 g of fuel with excess air at standard condi- rev/min tions resulted in a temperature rise of 1.25ºC of the water and calorimeter (of combined heat capacity 650 kJ/K), what is the heating value of a stoichiometric Etc. mixture of "X" and air, per kilogram of mixture? How does this value compare with a stoichiometric gasoline-air mixture? 9.7. The attached graph gives the pressure-crank angle curve for a spark-ignition engine Then, running at o = 1.0. The mass fraction burned x, is also shown. Estimate the tem- (c) Compare approximately the specific fuel consumption of a spark-ignition engine operated on stoichiometric gasoline and "X" mixtures. What do you conclude about the relative size of the fuel systems required to provide equal power? @ = 1 2500 (d) "x"-air mixtures take twice as long to burn as gasoline-air mixtures (the crank angle between the spark and end of combustion is twice as large). Sketch care- 2000 fully drawn pressure-time curves over the entire engine four-stroke cycle for the 1.0 two mixtures, for the same displacement and compression ratio engines, for the 0.8- 1500 same imep (for throttled engine operation) for stoichiometric mixtures, with p, kPa optimum spark timing, indicating the relative crank angle location of spark, peak X6 0.6- af 1000 cylinder pressure, and end of combustion, and the relative values of intake pres- 0.4 sure, peak cylinder pressure, and pressure at the exhaust valve opening, for the 0.2- -500 two mixtures at the same equivalence ratio. (You do not need to make calcu- lations to sketch these graphs.) 0.0 10 (e) Though "X"-air mixtures are slower burning than gasoline-air mixtures, the engine compression ratio can be increased to 14 : 1 from the 8 : 1 values typical 40 -20 0 20 40 60 for gasoline. Using typical fuel-air cycle efficiencies and the relation between the Crank angle, deg ATC fuel-air cycle and real cycle efficiency, determine whether the slower-burning FIGURE P9-7 484 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 485 perature of the reactants at a number of crank angles and plot a graph of Tu versus the spark timing is varied from very advanced (e.g ., 60º before TC) to close to TC. 0. Assume the reactants in the cylinder are at 333 K and 1 atm pressure at the start What is the "best" spark timing? Explain how it varies with engine speed and load. of compression. It is necessary to make simplifications in order to do this and you 9.13. (a) Explain the causes of the observed variation in cylinder pressure versus crank should explain clearly what other assumptions you make; they should be compatible angle and imep in spark-ignition engines, cycle-by-cycle. with Prob. 9.8 below. (b) What impacts do these cylinder pressure variations have on engine operation? 9.8. If the combustion takes place progressively through a large number of very small 9.14. (a) Describe briefly what occurs when a spark-ignition engine "knocks." zones of gas and there is no mixing between the zones, determine: (b) SI engine knock is primarily a problem at wide-open throttle and lower engine (a) the temperature at -30º, just after combustion, of the zone which burns at speeds. Explain why this is the case. - 300; (c) With a knock sensor, the normal knock control strategy is to retard the spark (b) the temperature at 0º, just after combustion, of the zone which burns at 0º; timing when knock is detected, until knock no longer occurs. Explain why this (c) the temperature at + 30º, just after combustion of the zone which burns at +300; strategy is effective and is preferred over other possible approaches (e.g ., throt- (d) the temperature of the products in these three zones at + 30º. tling the inlet, adding EGR). Plot your results versus crank angle to show whether there is a spatial distribution (d) In a knocking engine, the crank angle at which autoignition occurs and the mag- of temperature in the cylinder after combustion. nitude of the pressure oscillations which result vary substantially, cycle-by-cycle. Note: Each small unburned gas element burns at essentially constant pressure and is Suggest reasons why this happens. subsequently compressed and/or expanded. Use charts (Chap. 4) or an equilibrium 9.15. (a) The electrical energy stored in a typical ignition system coil is 50 mJ. Almost all computer code. The unburned gas state is given by Prob. 9.7. this energy is transferred from the coil during the glow discharge phase. If the 9.9. An approximate way to calculate the pressure in the end-gas just after knock occurs glow discharge lasts for 2 ms, use the data in Fig. 9-39 to estimate the glow is to assume that all the end-gas (the unburned gas ahead of the flame) burns instan- discharge voltage and current. taneously at constant volume. We assume that the inertia of the burned gases pre- (b) Only a fraction of this energy is transferred to the fuel-air mixture between the vents significant gas motion while the end-gas autoignites. spark plug electrodes. Estimate the energy transferred during the breakdown and For the pressure data in Fig. P9-7, assume autoignition occurs at 10 crank glow phases of the discharge, using the data in Fig. 9-40. angle degrees after the top-center position. Determine the maximum pressure (c) Overall about one-tenth of the coil energy is transferred to the fuel-air mixture. reached in the end-gas after knock occurs. From the mass fraction burned and What fraction of the cylinder contents' chemical energy (m/ 2LHv) does this corre- approximate average burned gas conditions at this time, estimate the volume spond to at a typical part-load condition (p; = 0.5 atm, ( = 1.0)? Assume occupied by the end-gas as a fraction of the cylinder volume just before autoignition 500 cm3 per cylinder displaced volume. If the average burned gas temperature occurs. Use the charts (Chap. 4) or an equilibrium computer code. within the flame kernal just after spark is 3500 K and the cylinder pressure is 9.10. At spark timing (30º BTC) in an automobile spark-ignition engine (with 6 atm, what radius of kernal has fuel chemical energy equal to the electrical bore = stroke = 85 mm and re = 9) at 2000 rev/min, operating on gasoline, the energy transferred to the kernal? cylinder pressure is 7.5 atm and the mixture temperature is 650 K. The fuel-air mixture is stoichiometric with a residual gas fraction of 8 percent. The rapid burning angle A0, (10 to 90 percent mass burned) is 35º. Estimate (a) the mean piston speed; REFERENCES (b) the average flame travel speed based on A0, (the spark plug is located 15 mm from the cylinder axis); (c) the turbulence intensity at TC [see Eq. (8.23)]; (d) the 1. Nakamura, H ., Ohinouye, T ., Hori, K. Kiyota, Y ., Nakagami, T ., Akishino, K ., and Tsukamoto, laminar flame speed at spark; (e) the turbulent burning speed S, at TC using Fig. Y.: "Development of a New Combustion System (MCA-JET) in Gasoline Engine," SAE paper 9-30 (appropriate firing and motored pressures are given in Fig. 9-2a); (f) the mean 780007, SAE Trans ., vol. 87, 1978. expansion speed of the burned gases up at TC. Discuss briefly the relative magni- 2. Rassweiler, G. M ., and Withrow, L.: "Motion Pictures of Engine Flames Correlated with Pres- sure Cards," SAE Trans ., vol. 83, pp. 185-204, 1938. Reissued as SAE paper 800131, 1980. tudes of these velocities. 3. Nakanishi, K ., Hirano, T ., Inoue, T ., and Ohigashi, S.: "The Effects of Charge Dilution on 9.11. The following combustion chamber design changes increase the mass burning rate in Combustion and Its Improvement-Flame Photograph Study," SAE paper 750054, SAE Trans ., a spark-ignition engine at fixed compression ratio, bore, speed, and inlet mixture vol. 84, 1975. conditions. Explain how each change affects the burning rate. 4. Beretta, G. P ., Rashidi, M ., and Keck, J. C.: "Turbulent Flame Propagation and Combustion in (a) Reducing the amount of EGR. Spark Ignition Engines," Combust. Flame, vol. 52, pp. 217-245, 1983. (b) Using two spark plugs per cylinder instead of one. 5. Amann, C. A.: "Cylinder-Pressure Measurement and Its Use in Engine Research," SAE paper 852067, 1985. (c) Generating swirl within the cylinder using a mask on the cylinder head (see Fig. 6. Lavoie, G. A ., Heywood, J. B ., and Keck, J. C.: "Experimental and Theoretical Investigation of 8-14). Nitric Oxide Formation in Internal Combustion Engines," Combust. Sci. Technol ., vol. 1, pp. (d) Using a combustion chamber with higher clearance height near the spark plug 313-326, 1970. and a more central plug location. 7. Rassweiler, G. M ., and Withrow, L.: "Flame Temperatures Vary with Knock and Combustion- 9.12. Explain why SI engine torque varies, at fixed speed and inlet mixture conditions, as Chamber Position," SAE Trans ., vol. 36, pp. 125-133, 1935. 486 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 487 8. Lavoie, G. A.: "Spectroscopic Measurement of Nitric Oxide in Spark-Ignition Engines," Hydrogen-Hydrocarbon Gas Mixtures," Combust. Flame, vol. 58, pp. 13-22, 1984. Combust. Flame, vol. 15, pp. 97-108, 1970. 32. Rhodes, D. B ., and Keck, J. C.: "Laminar Burning Speed Measurements of Indolene-Air- 9. Heywood, J. B ., and Keck, J. C.: "Formation of Hydrocarbons and Oxides of Nitrogen in Diluent Mixtures at High Pressures and Temperature," SAE paper 850047, 1985. Automobile Engines," Env. Sci. Technol ., vol. 7, pp. 216-223, 1973. 33. Metghalchi, M.: "Laminar Burning Velocity of Isooctane-Air, Methane-Air, and Methanol-Air 10. Kreiger, R. B ., and Borman, G. L.: "The Computation of Apparent Heat Release for Internal Mixtures at High Temperature and Pressure," M.S. Thesis, Department of Mechanical Engin- Combustion Engines," ASME paper 66-WA/DGP-4, 1966. eering, MIT, 1976. 11. LoRusso, J. A ., and Tabaczynski, R. J.: "Combustion and Emissions Characteristics of Meth- 34. Daneshyar, H ., Mendes-Lopes, J. M. C ., Ludford, G. S. S ., and Tromans, P. S.: "The Influence anol, Methanol-Water, and Gasoline-Methanol Blends in a Spark-Ignition Engine," SAE paper of Straining on a Premixed Flame and Its Relevance to Combustion in SI Engines," paper 769019, Proceedings of Eleventh Intersociety Energy Conversion Engineering Conference, Lake C50/83, in Combustion in Engineering, vol. 1, pp. 191-199, Proceedings of an International Con- Tahoe, Nev ., pp. 122-132, Sept. 12-17, 1976. ference of the Institution of Mechanical Engineers, Mechanical Engineering Publications Ltd ., 12. Lancaster, D. R ., Kreiger, R. B ., and Lienesch, J. H.: "Measurement and Analysis of Engine London, 1983. Pressure Data," SAE paper 750026, SAE Trans ., vol. 84, 1975. 35. Groff, E. G ., and Matekunas, F. A.: "The Nature of Turbulent Flame Propagation in a Homo- 13. Caris, D. F ., and Nelson, E. E.: "A New Look at High Compression Engines," SAE Trans ., vol. geneous Spark-Ignited Engine," SAE paper 800133, SAE Trans ., vol. 89, 1980. 67, pp. 112-124, 1959. 36. Witze, P. O ., and Mendes-Lopes, J. M. C.: "Direct Measurement of the Turbulent Burning 14. Gatowski, J. A ., Balles, E. N ., Chun, K. M ., Nelson, F. E ., Ekchian, J. A ., and Heywood, J. B.: Velocity in a Homogeneous-Charge Engine," SAE paper 851531, 1986. "Heat Release Analysis of Engine Pressure Data," SAE paper 841359, SAE Trans ., vol. 93, 1984. 37. Hall, M. J ., Bracco, F. V ., and Santavicca, D. A.: "Cycle-Resolved Velocity and Turbulence 15. Chun, K. M ., and Heywood, J. B.: " Estimating Heat-Release and Mass-of-Mixture Burned from Measurements in an IC Engine with Combustion," SAE paper 860320, 1986. Spark-Ignition Engine Pressure Data," Combust. Sci. Technol ., vol. 54, pp. 133-144, 1987. 38. Matekunas, F. A.: "Modes and Measures of Cyclic Combustion Variability," SAE paper 16. Heywood, J. B ., Higgins, J. M ., Watts, P. A ., and Tabaczynski, R. J.: "Development and Use of a 830337, SAE Trans ., vol. 92, 1983. Cycle Simulation to Predict SI Engine Efficiency and NO, Emissions," SAE paper 790291, 1979. 39. Young, M. B.: "Cyclic Dispersion in the Homogeneous-Charge Spark-Ignition Engine-A Lit- 17. Gatowski, J. A ., Heywood, J. B ., and Deleplace, C.: "Flame Photographs in a Spark-Ignition erature Survey," SAE paper 810020, SAE Trans ., vol. 90, 1981. Engine," Combust. Flame, vol. 56, pp. 71-81, 1984. 40. Young, M. B.: "Cyclic Dispersion-Some Quantitative Cause and Effect Relationships," SAE 18. Witze, P. O.: "The Effect of Spark Location on Combustion in a Variable-Swirl Engine," SAE paper 800459, 1980. paper 820044, SAE Trans ., vol. 91, 1982. 41. Patterson, D. J.: "Cylinder Pressure Variations Present a Fundamental Combustion Problem," 19. Taylor, C. F.: The Internal Combustion Engine in Theory and Practice, vol. II, Combustion, Fuels, SAE paper 660129, SAE Trans ., vol. 75, 1966. Materials, Design, MIT Press, 1968. 42. Fisher, R. V ., and Macey, J. P.: "Digital Data Acquisition with Emphasis on Measuring Pres- 20. Hires, S. D ., Tabaczynski, R. J ., and Novak, J. M.: "The Prediction of Ignition Delay and sure Synchronously with Crank Angle," SAE paper 750028, 1975. Combustion Intervals for a Homogeneous Charge, Spark Ignition Engine," SAE paper 780232, 43. Douaud, A ., and Eyzat, P.: "DIGITAP-An On-Line Acquisition and Processing System for SAE Trans ., vol. 87, 1978. Instantaneous Engine Data-Applications," SAE paper 770218, 1977. 21. Heywood, J. B.: "Combustion Chamber Design for Optimum Spark-Ignition Engine Per- 44. Peters, B. D.: "Mass Burning Rates in a Spark Ignition Engine Operating in the Partial-Burn formance," Int. J. Vehicle Design, vol. 5, no. 3, pp. 336-357, 1984. Regime," paper C92/79, Conference Proceedings on Fuel Economy and Emissions of Lean Burn 22. Abraham, J ., Williams, F. A ., and Bracco, F. V.: " A Discussion of Turbulent Flame Structure in Engines, Institution of Mechanical Engineers, London, June 12-14, 1979. Premixed Charges," SAE paper 850345, in Engine Combustion Analysis: New Approaches, P-156, 45. Hansel, J. G.: "Lean Automotive Engine Operation-Hydrocarbon Exhaust Emissions and 1985. Combustion Characteristics," SAE 710164, SAE Trans ., vol. 80, 1971. 23. Namazian, M ., Hansen, S. P ., Lyford-Pike, E. J ., Sanchez-Barsse, J ., Heywood, J. B ., and Rife, J.: 46. Yu, H. T. C.: "Fuel Distribution Studies-A New Look at an Old Problem," SAE Trans ., vol. "Schlieren Visualization of the Flow and Density Fields in the Cylinder of a Spark-Ignition 71, pp. 596-613, 1963. Engine," SAE paper 800044, SAE Trans ., vol. 89, 1980. 47. Nakagawa, Y ., Nakai, M ., and Hamai, K.: " A Study of the Relationship between Cycle-to-Cycle 24. Keck, J. C.: "Turbulent Flame Structure and Speed in Spark-Ignition Engines," Proceedings of Variations of Combustion and Heat Release Delay in a Spark-Ignited Engine," Bull. JSME, vol. the Nineteenth International Symposium on Combustion, The Combustion Institute, pp. 1451- 25, no. 199, pp. 54-60, 1982. 1466, 1982. 48. Matsui, K ., Tanaka, T ., and Ohigashi, S.: "Measurement of Local Mixture Strength at Spark 25. Heywood, J. B ., and Vilchis, F. R.: "Comparison of Flame Development in a Spark-Ignition Gap of S.I. Engines," SAE paper 790483, SAE Trans ., vol. 88, 1979. Engine Fueled with Propane and Hydrogen," Combust. Sci. Technol ., vol. 38, pp. 313-324, 1984. 49. Kuroda, H ., Nakajima, Y ., Sugihara, K ., Takagi, Y ., and Muranaka, S.: "The Fast Burn with 26. Smith, J. R.: “Turbulent Flame Structure in a Homogeneous-Charge Engine," SAE paper Heavy EGR, New Approach for Low NO, and Improved Fuel Economy," SAE 780006, SAE 820043, SAE Trans ., vol. 91, 1982. Trans ., vol. 87, 1978. 27. Smith, J. R.: "The Influence of Turbulence on Flame Structure in an Engine," in T. Uzkan (ed.), 50. Quader, A. A.: "What Limits Lean Operation in Spark Ignition Engines-Flame Initiation or Flows in Internal Combustion Engines, pp. 67-72, ASME, New York, 1982. Propagation ?," SAE paper 760760, SAE Trans ., vol. 85, 1976. 28. Rallis, C. J ., and Garforth, A. M.: "The Determination of Laminar Burning Velocity," Prog. 51. Quader, A. A.: "Lean Combustion and the Misfire Limit in Spark Ignition Engines," SAE paper Energy Combust. Sci ., vol. 6, pp. 303-330, 1980. 741055, SAE Trans ., vol. 83, 1974. 29. Metghalchi, M ., and Keck, J. C.: "Laminar Burning Velocity of Propane-Air Mixtures at High 52. Maly, R ., and Vogel, M.: "Ignition and Propagation of Flame Fronts in Lean CH ,- Air Mix- Temperature and Pressure," Combust. Flame, vol. 38, pp. 143-154, 1980. tures by the Three Modes of the Ignition Spark," Proceedings of Seventeenth International Sym- 30. Metghalchi, M ., and Keck, J. C.: “Burning Velocities of Mixtures of Air with Methanol, Iso- posium on Combustion, pp. 821-831, The Combustion Institute, 1976. octane, and Indolene at High Pressure and Temperature," Combust. Flame, vol. 48, pp. 191-210, 53. Maly, R.: "Spark Ignition: Its Physics and Effect on the Internal Combustion Process," in J. C. ............. 1982. Hilliard and G. S. Springer (eds.), Fuel Economy in Road Vehicles Powered by Spark Ignition 31. Milton, B. E ., and Keck, J. C.: "Laminar Burning Velocities in Stoichiometric Hydrogen and Engines, chap. 3, Plenum Press, New York, 1984. 488 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN SPARK-IGNITION ENGINES 489 54. Bosch: Automotive Handbook, Robert Bosch GmbH, Stuttgart, West Germany, 1976. 78. Lee, W ., and Schaefer, H. J.: "Analysis of Local Pressures, Surface Temperatures and Engine 55. Maly, R ., Saggau, B ., Wagner, E ., and Ziegler, G.: "Prospects of Ignition Enhancement," SAE Damages under Knock Conditions," SAE paper 830508, SAE Trans ., vol. 92, 1983. paper 830478, SAE Trans ., vol. 92, 1983. 79. Leppard, W. R.: "Individual-Cylinder Knock Occurrence and Intensity in Multicylinder 56. Maly, R.: "Ignition Model for Spark Discharges and the Early Phase of Flame Front Growth," Engines," SAE paper 820074, 1982. Proceedings of Eighteenth International Symposium on Combustion, pp. 1747-1754, The Com- 80. Nakajima, Y ., Onoda, M ., Nagai, T ., and Yoneda, K.: "Consideration for Evaluating Knock bustion Institute, 1981. Intensity," JSAE Rev ., vol. 9, pp. 27-35, 1982. 57. Refael, S ., and Sher, E.: " A Theoretical Study of the Ignition of a Reactive Medium by Means of 81. Benson, G ., Fletcher, E. A ., Murphy, T. E ., and Scherrer, H. C.: "Knock (Detonation) Control an Electrical Discharge," Combust. Flame, vol. 59, pp. 17-30, 1985. by Engine Combustion Chamber Shape," SAE paper 830509, 1983. 58. Ballal, D. R ., and Lefebvre, A. H.: "Influence of Flow Parameters on Minimum Ignition Energy 82. Cornetti, G. M ., DeCristofaro, F ., and Gozzelino, R.: "Engine Failure and High Speed Knock," and Quenching Distance," Proceedings of Fifteenth International Symposium on Combustion, pp. SAE paper 770147. 1473-1481, The Combustion Institute, 1974. 83. Renault, F.: "A New Technique to Detect and Control Knock Damage," SAE paper 820073, 59. Kono, M ., Linuma, K ., Kumagai, S ., and Sakai, T.: "Spark Discharge Characteristics and Ignit- 1982. ing Ability of Capacitor Discharge Ignition Systems," Combust. Sci. Technol ., vol. 19, pp. 13-18, 84. The Relationship between Knock and Engine Damage, Tentative Code of Practice, CEC Report 1978. M-07-T-83, Co-ordinating European Council (CEC), London, England, 1984. 60. Rittmannsberger, N.: “Developments in High-Voltage Generation and Ignition Control," paper 85. Withrow, L ., and Rassweiler, G. M.: "Slow-Motion Shows Knocking and Non-Knocking C106/72, in the Automobile Division Conference on Automotive Electric Equipment, Institution Explosions," SAE J ., vol. 39, no. 2, pp. 297-303, 312, 1936. of Mechanical Engineers, London, 1972. 86. Miller, C. D ., Olsen, H. L ., Logan, W. O ., and Osterstrom, G. E.: NACA report 857, 1946. 61. Huntzinger, G. O ., and Rigsby, G. E.: "HEI-A New Ignition System Through New Tech- 87. Male, T.: "Photography at 500,000 Frames per Second of Combustion and Detonation in a nology," SAE paper 750346, SAE Trans ., vol. 84, 1975. Reciprocating Engine," Proceedings of Third Symposium on Combustion, Flame and Explosion 62. Dale, J. D ., and Oppenheim, A. K.: "Enhanced Ignition for I.C. Engines with Premixed Gases," Phenomena, p. 721, Williams and Wilkins, Combustion Institute, Pittsburgh, 1949. SAE paper 810146, SAE Trans ., vol. 90, 1981. 88. Nakajima, Y ., Nagai, T ., Iijima, T ., Yokoyama, J ., and Nakamura, K.: " Analysis of Combustion 63. Tanuma, T ., Sasaki, K ., Kaneko, T ., and Kawasaki, H.: "Ignition, Combustion, and Exhaust Patterns Effective in Improving Anti-Knock Performance of a Spark-Ignition Engine," JSAE Emissions of Lean Mixtures in Automotive Spark Ignition Engines," SAE paper 710159, SAE Rev ., vol. 13, pp. 9-17, 1984. Trans ., vol. 80, 1971. 89. Nakagawa, Y ., Takagi, Y ., Itoh, T ., and Iijima, T.: "Laser Shadowgraphic Analysis of Knocking 64. Nakamura, N ., Baika, T ., and Shibata, Y.: "Multipoint Spark Ignition for Lean Combustion," in S.I. Engine," SAE paper 845001; also in XX FISITA Congress Proceedings, vol. P-143, 1984. SAE paper 852092, 1985. 90. Hayashi, T ., Taki, M ., Kojima, S ., and Kondo, T.: "Photographic Observation of Knock with a 65. Schwartz, H.: "Ignition Systems for Lean Burn Engines," paper C95/79, Proceedings of Con- Rapid Compression and Expansion Machine," SAE paper 841336, SAE Trans ., vol. 93, 1984. ference on Fuel Economy and Emissions of Lean Burn Engines, pp. 87-96, Institution of Mechani- 91. Smith, J. R ., Green, R. M ., Westbrook, C. K ., and Pitz, W. J.: " An Experimental and Modeling cal Engineers, London, June 1979. Study of Engine Knock," Proceedings of Twentieth International Symposium on Combustion, pp. 66. Ziegler, G. F. W ., Wagner, E. P ., Saggau, B ., and Maly, R. R.: "Influence of a Breakdown 91-100, The Combustion Institute, 1984. Ignition System on Performance and Emission Characteristics," SAE paper 840992, SAE Trans ., 92. Barnard, J. A.: Flame and Combustion, 2d ed ., Chapman and Hall, 1985. vol. 93, 1984. 93. Glassman, I.: Combustion, Academic Press, 1977. 67. Anderson, R. W ., and Asik, J. R.: "Lean Air-Fuel Ignition System Comparison in a Fast-Burn 94. Downes, D.: "Chemical and Physical Studies of Engine Knock," in Six Lectures on the Basic Engine," SAE paper 850076, 1985. Combustion Processes, pp. 127-155, Ethyl Corporation, Detroit, Mich ., 1954. 68. Edwards, C. F ., Oppenheim, A. K ., and Dale, J. D.: "A Comparative Study of Plasma Ignition 95. Halstead, M. P ., Kirsch, L. J ., Prothero, A ., and Quinn, C. P.: "A Mathematical Model for Systems," SAE paper 830479, 1983. Hydrocarbon Autoignition at High Pressures," Proc. R. Soc ., ser. A, vol. 346, pp. 515-538, 1975. 69. Noguchi, M ., Sanda, S ., and Nakamura, N.: "Development of Toyota Lean Burn Engine," SAE 96. Benson, S. W.: "The Kinetics and Thermochemistry of Chemical Oxidation with Application to paper 760757, SAE Trans ., vol. 85, 1976. Combustion and Flames," Prog. Energy Combust. Sci ., vol. 7, pp. 125-134, 1981. 70. Brandstetter, W. R ., Decker, G ., and Reichel, K.: "The Water-Cooled Volkswagen PCI- 97. Alperstein, M ., and Bradow, R. L.: "Investigations into the Composition of End Gases from Stratified Charge Engine," SAE paper 750869, SAE Trans ., vol. 84, 1975. Otto Cycle Engines," SAE paper 660410, SAE Trans ., vol. 75, 1966. 71. Date, T ., and Yagi, S.: "Research and Development of the Honda CVCC Engine," SAE paper 98. Livengood, J. C ., and Wu, P. C.: "Correlation of Autoignition Phenomenon in Internal Com- 740605, 1974. bustion Engines and Rapid Compression Machines," Proceedings of Fifth International Sympo- 72. Konishi, M ., Nakamura, N ., Oono, E ., Baika, T ., and Sanda, S.: "Effects of a Prechamber on sium on Combustion, p. 347, Reinhold, 1955. NO, Formation Process in the SI Engine," SAE paper 790389, 1979. 99. By, A ., Kempinski, B ., and Rife, J. M.: "Knock in Spark Ignition Engines," SAE paper 810147, 73. Turkish, M. C.: "Prechamber and Valve Gear Design for 3-Valve Stratified Charge Engines," 1981. SAE paper 751004, SAE Trans ., vol. 84, 1975. 100. Douaud, A. M ., and Eyzat, P.: "Four-Octane-Number Method for Predicting the Anti-Knock 74. Ethyl Corporation: "Engine Combustion Noises," Ethyl technical note PCDTN-MS 117768 Behavior of Fuels and Engines," SAE paper 780080, SAE Trans ., vol. 87, 1978. Rev. 774. 101. Cox, R. A ., and Cole, J. A.: "Chemical Aspects of the Autoignition of Hydrocarbon-Air Mix- 75. Wheeler, R. W.: "Abnormal Combustion Effects on Economy," in J. C. Hilliard and G. S. tures," Combust. Flame, vol. 60, pp. 109-123, 1985. Springer (eds.), Fuel Economy in Road Vehicles Powered by Spark-Ignition Engines, chap. 6, pp. 102. Halstead, M. P ., Kirsch, L. J ., and Quinn, C. P.: "Autoignition of Hydrocarbon Fuels at High 225-276, Plenum Press, 1984. Temperatures and Pressures-Fitting of a Mathematical Model," Combust. Flame, vol. 30, pp. 76. Guibet, J. C ., and Duval, A.: "New Aspects of Preignition in European Automotive Engines," 45-60, 1977. SAE paper 720114, SAE Trans ., vol. 81, 1972. 103. Schapertons, H ., and Lee, W.: "Multidimensional Modelling of Knocking Combustion in SI 77. Douaud, A ., and Eyzat, P.: "DIGITAP-An On-Line Acquisition and Processing System for Engines," SAE paper 850502, 1985. Instantaneous Engine Data-Applications," SAE paper 770218, 1977. 104. Goodger, E. M.: Hydrocarbon Fuels, Macmillan, London, 1975. 490 INTERNAL COMBUSTION ENGINE FUNDAMENTALS 105. Lovell, W. G.: "Knocking Characteristics of Hydrocarbons," Ind. Engng. Chem ., vol. 40, pp. 2388-2438, 1948. CHAPTER 106. Lichty, L. C.: Combustion Engine Processes, McGraw-Hill, 1967. 107. Ethyl Corporation: "Determining Road Octane Numbers," Ethyl technical note PCDTN SP-347 (113) Rev. 573. 108. Benson, J.: "What Good are Octanes," in Chemtech, pp. 16-22, American Chemical Society, 10 January 1976. 109. Barusch, M. R ., Macpherson, J. B ., and Amberg, G. H.: " Additives, Engine Fuel," in J. J. McKetta and W. A. Cunningham (eds.), Encyclopedia of Chemical Processing and Design, vol. 2, pp. 1-77, Marcel Dekker, New York and Basel, 1977. 110. McCabe, L. J ., Fitch, F. B ., and Lowther, H. V.: "Future Trends in Automotive Fuels and COMBUSTION IN Engine Oils," SAE paper 830935, in Proceedings of Second International Pacific Conference on COMPRESSION-IGNITION Automotive Engineering, pp. 678-697, Tokyo, Japan, Nov. 7-10, 1983. 111. Benson, J. D.: "Some Factors Which Affect Octane Requirement Increase," SAE paper 750933, ENGINES SAE Trans ., vol. 84, 1975. 10.1 ESSENTIAL FEATURES OF PROCESS The essential features of the compression-ignition or diesel engine combustion process can be summarized as follows. Fuel is injected by the fuel-injection system into the engine cylinder toward the end of the compression stroke, just before the desired start of combustion. Figures 1-17, 1-18, and 1-19 illustrate the major components of common diesel fuel-injection systems. The liquid fuel, usually injected at high velocity as one or more jets through small orifices or nozzles in the injector tip, atomizes into small drops and penetrates into the combustion chamber. The fuel vaporizes and mixes with the high-temperature high-pressure cylinder air. Since the air temperature and pressure are above the fuel's ignition point, spontaneous ignition of portions of the already-mixed fuel and air occurs after a delay period of a few crank angle degrees. The cylinder pressure increases as combustion of the fuel-air mixture occurs. The consequent compression of the unburned portion of the charge shortens the delay before ignition for the fuel and air which has mixed to within combustible limits, which then burns rapidly. It also reduces the evaporation time of the remaining liquid fuel. Injection continues until the desired amount of fuel has entered the cylinder. Atomization, vaporization, fuel-air mixing, and combustion continue until essen- tially all the fuel has passed through each process. In addition, mixing of the air remaining in the cylinder with burning and already burned gases continues throughout the combustion and expansion processes. 492 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 493 It will be clear from this summary that the compression-ignition com- foregoing discussion indicates (and more detailed analysis will confirm) that bustion process is extremely complex. The details of the process depend on the mixing rates control the fuel burning rate. Commercial diesel engines are made characteristics of the fuel, on the design of the engine's combustion chamber and with a very large range of cylinder sizes, with cylinder bores varying from about fuel-injection system, and on the engine's operating conditions. It is an unsteady, 70 to 900 mm. The mean piston speed at maximum rated power is approximately heterogeneous, three-dimensional combustion process. While an adequate con- constant over this size range (see Sec. 2.2), so the maximum rated engine speed ceptual understanding of diesel engine combustion has been developed, to date will be inversely proportional to the stroke [see Eq. (2.9)]. For a fixed crank an ability to describe many of the critical individual processes in a quantitative angle interval for combustion (of order 40 to 50º to maintain high fuel conversion manner is lacking. efficiency), the time available for combustion will, therefore, scale with the stroke. Some important consequences of this combustion process on engine oper- Thus, at the small end of the diesel engine size range, the mixing between the ation are the following: injected fuel and the air must take place on a time scale some 10 times shorter than in engines at the large end of this range. It would be expected, therefore, 1. Since injection commences just before combustion starts, there is no knock that the design of the engine combustion chamber (including the inlet port and limit as in the spark-ignition engine resulting from spontaneous ignition of the valve) and the fuel-injection system would have to change substantially over this premixed fuel and air in the end-gas. Hence a higher engine compression ratio size range to provide the fuel and air motion inside the cylinder required to can be used in the compression-ignition engine, improving its fuel conversion achieve the desired fuel-air mixing rate. As engine size decreases, more vigorous efficiency relative to the SI engine. air motion is required while less fuel jet penetration is necessary. It is this logic, 2. Since injection timing is used to control combustion timing, the delay period primarily, that leads to the different diesel combustion chamber designs and between the start of injection and start of combustion must be kept short (and fuel injection systems found in practice over the large size range of commercial reproducible). A short delay is also needed to hold the maximum cylinder gas diesel engines. pressure below the maximum the engine can tolerate. Thus, the spontaneous ignition characteristics of the fuel-air mixture must be held within a specified 10.2 TYPES OF DIESEL COMBUSTION range. This is done by requiring that diesel fuel have a cetane number (a SYSTEMS measure of the ease of ignition of that fuel in a typical diesel environment; see Sec. 10.6.2) above a certain value. Diesel engines are divided into two basic categories according to their com- 3. Since engine torque is varied by varying the amount of fuel injected per cycle bustion chamber design: (1) direct-injection (DI) engines, which have a single open with the engine air flow essentially unchanged, the engine can be operated combustion chamber into which fuel is injected directly; (2) indirect-injection unthrottled. Thus, pumping work requirements are low, improving part-load (IDI) engines, where the chamber is divided into two regions and the fuel is mechanical efficiency relative to the SI engine. injected into the "prechamber" which is connected to the main chamber (situated above the piston crown) via a nozzle, or one or more orifices. IDI 4. As the amount of fuel injected per cycle is increased, problems with air uti- lization during combustion lead to the formation of excessive amounts of soot engine designs are only used in the smallest engine sizes. Within each category there are several different chamber geometry, air-flow, and fuel-injection arrange- which cannot be burned up prior to exhaust. This excessive soot or black ments. smoke in the exhaust constrains the fuel/air ratio at maximum engine power to values 20 percent (or more) lean of stoichiometric. Hence, the maximum indicated mean effective pressure (in a naturally aspirated engine) is lower than 10.2.1 Direct-Injection Systems values for an equivalent spark-ignition engine. In the largest-size engines, where mixing rate requirements are least stringent, 5. Because the diesel always operates with lean fuel/air ratios (and at part load quiescent direct-injection systems of the type shown in Fig. 10-1a are used. The with very lean fuel/air ratios), the effective value of y (=c,/c,) over the expan- momentum and energy of the injected fuel jets are sufficient to achieve adequate sion process is higher than in a spark-ignition engine. This gives a higher fuel fuel distribution and rates of mixing with the air. Additional organized air conversion efficiency than the spark-ignition engine, for a given expansion motion is not required. The combustion chamber shape is usually a shallow bowl ratio (see Sec. 5.5.3). in the crown of the piston, and a central multihole injector is used. As engine size decreases, increasing amounts of air swirl are used to achieve The major problem in diesel combustion chamber design is achieving suffi- faster fuel-air mixing rates. Air swirl is generated by suitable design of the inlet ciently rapid mixing between the injected fuel and the air in the cylinder to com- port (see Sec. 8.3); the swirl rate can be increased as the piston approaches TC by plete combustion in the appropriate crank angle interval close to top-center. The forcing the air toward the cylinder axis, into a bowl-in-piston type of combustion 494 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 495 Fuel jets Fuel jets Fuel jet Air swirl- + (a) (b) FIGURE 10-2 Air swirl Two common types of small indirect-injection diesel engine combustion system: (a) swirl prechamber; (b) turbulent prechamber. (a) (b) (c) ing passage and chamber are shaped so that this flow within the auxiliary FIGURE 10-1 chamber rotates rapidly. Common types of direct-injection compression-ignition or diesel engine combustion systems: (a) quiescent chamber with multihole nozzle typical of larger engines; (b) bowl-in-piston chamber with Fuel is usually injected into the auxiliary chamber at lower injection-system swirl and multihole nozzle; (c) bowl-in-piston chamber with swirl and single-hole nozzle. (b) and (c) pressure than is typical of DI systems through a pintle nozzle as a single spray, as used in medium to small DI engine size range. shown in Fig. 1-18. Combustion starts in the auxiliary chamber; the pressure rise associated with combustion forces fluid back into the main chamber where the jet chamber. Figure 10-1b and c shows the two types of DI engine with swirl in issuing from the nozzle entrains and mixes with the main chamber air. The glow common use. Figure 10-1b shows a DI engine with swirl, with a centrally located plug shown on the right of the prechamber in Fig. 10-2 is a cold-starting aid. The multihole injector nozzle. Here the design goal is to hold the amount of liquid plug is heated prior to starting the engine to ensure ignition of fuel early in the fuel which impinges on the piston cup walls to a minimum. Figure 10-1c shows engine cranking process. the M.A.N. "M system" with its single-hole fuel-injection nozzle, oriented so that most of the fuel is deposited on the piston bowl walls. These two types of designs 10.2.3 Comparison of Different Combustion are used in medium-size (10- to 15-cm bore) diesels and, with increased swirl, in Systems small-size (8- to 10-cm bore) diesels. The number of different combustion chamber types proposed and tried since the 10.2.2 Indirect-Injection Systems beginnings of diesel engine development is substantial. Over the years, however, through the process of evolution and the increased understanding of the physical Inlet generated air swirl, despite amplification in the piston cup, has not provided and chemical processes involved, only a few designs based on a sound principle sufficiently high fuel-air mixing rates for small high-speed diesels such as those have survived. The important characteristics of those chambers now most com- used in automobiles. Indirect-injection or divided-chamber engine systems have monly used are summarized in Table 10.1. The numbers for dimensions and been used instead, where the vigorous charge motion required during fuel injec- operating characteristics are typical ranges for each different type of diesel engine tion is generated during the compression stroke. Two broad classes of IDI and combustion system. systems can be defined: (1) swirl chamber systems and (2) prechamber systems, as The largest, slowest speed, engines for power generation or marine applica- illustrated in Fig. 10-2a and b, respectively. During compression, air is forced tions use open quiescent chambers which are essentially disc shaped; the motion from the main chamber above the piston into the auxiliary chamber, through the of the fuel jets is responsible for distributing and mixing the fuel. These are nozzle or orifice (or set of orifices). Thus, toward the end of compression, a vigor- usually two-stroke cycle engines. In the next size range, in large truck and loco- ous flow in the auxiliary chamber is set up; in swirl chamber systems the connect- motive engines, a quiescent chamber consisting of a shallow dish or bowl in the 496 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 497 TABLE 10.1 Characteristics of Common Diesel Combustion Systems auxiliary chamber during the latter part of the compression stroke, using a nozzle or connecting passage that enters the auxiliary chamber tangentially, or they Direct injection Indirect injection generate intense turbulence in the prechamber through use of several small ori- fices and obstructions to the flow within the prechamber. The most common Medium High swirl High swirl Swir Pre- wirl design of swirl chamber is the Ricardo Comet design shown in Fig. 10-2a. An System Quiescent "M" multispray chamber chamber alternative IDI engine to the two types listed in Table 10-1 is the air cell system. Size Largest Medium Medium- Medium- Smallest Smallest In that system the fuel is injected into the main chamber and not the auxiliary smaller small "air cell." The auxiliary chamber acts as a turbulence generator as gas flows into Cycle 2-/ 4-stroke 4-stroke 4-stroke 4-stroke 4-stroke and out of the cell. 4-stroke Turbocharged/ TC/S TC/NA TC/NA NA/TC NA/TC NA/TC supercharged/ naturally 10.3 PHENOMENOLOGICAL MODEL OF aspirated COMPRESSION-IGNITION ENGINE Maximum speed, 120-2100 1800-3500 2500-5000 3500 -- 4300 3600-4800 4500 COMBUSTION rev/min Bore, mm 900-150 150-100 130-80 100-80 95-70 95-70 Studies of photographs of diesel engine combustion, combined with analyses of Stroke/bore 3.5-1.2 1.3-1.0 1.2-0.9 1.1-0.9 1.1-0.9 1.1-0. engine cylinder pressure data, have led to a widely accepted descriptive model of Compression 12-15 15-16 16-18 16-22 20-24 22-24 ratio the compression-ignition engine combustion process. The concept of heat-release Chamber Open or Bowl-in- Deep bowl- Deep bowl- Swirl pre- Single/ rate is important to understanding this model. It is defined as the rate at which shallow piston In-piston in-piston chamber multi- the chemical energy of the fuel is released by the combustion process. It can be dish orifice calculated from cylinder pressure versus crank angle data, as the energy release pre- required to create the measured pressure, using the techniques described in Sec. chamber Air-flow Quiescent Medium High swirl Highest Very high Very turbu- 10.4. The combustion model defines four separate phases of diesel combustion, pattern swirl swirl swirl lent in pre- each phase being controlled by different physical or chemical processes. Although in pre- chamber the relative importance of each phase does depend on the combustion system chamber used, and engine operating conditions, these four phases are common to all diesel Number of Multi Multi Single Multi Single Single engines. nozzle holes Injection Very high High Medium High Lowest Lowest pressure 10.3.1 Photographic Studies of Engine Combustion High-speed photography at several thousand frames per second has been used extensively to study diesel engine combustion. Some of these studies have been piston crown is often used. The air utilization in these engines is low, but they are carried out in combustion chambers very close to those used in practice, under invariably supercharged or turbocharged to obtain high power density. normal engine operating conditions (e.g ., Refs. 1 and 2). Sequences of individual In the DI category, as engine size decreases and maximum speed rises, swirl frames from movies provide valuable information on the nature of the com- is used increasingly to obtain high-enough fuel-air mixing rates. The swirl is gen- bustion process in the different types of diesel engines. Figure 10-3 shows four erated by suitably shaped inlet ports, and is amplified during compression by combustion chamber geometries that have been studied photographically. These forcing most of the air toward the cylinder axis into the deep bowl-in-piston are: (a) a quiescent chamber typical of diesel engines in the 3 to 20 dm3/cylinder combustion chamber. In about the same size range, an alternative system to the displacement used for industrial, marine, and rail traction applications (only the multihole nozzle swirl system is the M.A.N. "M" system (or wall-wetting system), burning of a single fuel spray of the multispray combustion system could be where most of the fuel from the single-hole pintle nozzle is placed on the wall of studied2); (b) a smaller high-speed DI engine with swirl and four fuel jets centrally the spherical bowl in the piston crown. injected; (c) an M.A.N. "M" DI system; and (d) a Ricardo Comet V swirl In the smallest engine sizes, the IDI engine has traditionally been used to chamber IDI system.1 obtain the vigorous air motion required for high fuel-air mixing rates. There are The combustion sequences were recorded on color film and show the fol- several different geometries in use. These either generate substantial swirl in the lowing features: 498 INTERNAL COMBUSTION ENGINE FUNDAMENTALS -30º -200 -10º TO 10º Exhaust valve Spark plug 20º Inlet valve 30º FIGURE 9-1 (a) (b) FIGURE 10-3 Four diesel combustion chambers used to obtain photographs of the compression-ignition com- bustion process shown in Fig. 10-4 on color plate: (a) quiescent DI chamber; (b) multihole nozzle DI chamber with swirl: on p. 499; (c) M.A.N. "M" DI chamber; (d) Ricardo Comet IDI swirl chamber.1,2 Fuel spray(s). The fuel droplets reflect light from spot lamps and define the extent of the liquid fuel spray prior to complete vaporization. Premixed flame. These regions are of too low a luminosity to be recorded with the exposure level used. The addition of a copper additive dope to the fuel gives these normally blue flames a green color bright enough to render them (a) (b) visible. Diffusion flame. The burning high-temperature carbon particles in this flame provide more than adequate luminosity and appear as yellow-white. As the flame cools, the radiation from the particles changes color through orange to red. - 1 mm Over-rich mixture. The appearance of a brown region, usually surrounded 1 mm by a white diffusion flame, indicates an excessively rich mixture region where substantial soot particle production has occurred. Where this fuel-rich soot-laden cloud contacts unburned air there is a hot white diffusion flame. Table 10.2 summarizes the characteristics of these different regions, discernable in the photographs shown in Fig. 10-4 on the color plate. (c) FIGURE 10-5 (d) COMBUSTION IN COMPRESSION-IGNITION ENGINES 499 in 38º TC (d) 10.5 O 50 in -2 (c) (c) (d) in 300 TC -30 - 70 Figure 10-4a shows a sequence of photographs from one combustion event of the single spray, burning under conditions typical of a large quiescent DI engine. The fuel spray is shown penetrating into the chamber. Ignition occurs at (b) -8º in the fuel-air mixture left behind on the edge of the spray not far from the injector. The flame then spreads rapidly (-7º) along the outside of the spray to the spray tip. Here some of the fuel, which has had a long residence time in the chamber, burns with a blue-green low-luminosity flame (colored green by the copper fuel additive). The flame engulfing the remainder of the spray is brilliant white-yellow from the burning of the soot particles which have been formed in 31º 20 11º -1º -70 -100 FIGURE 10-4 (On Color Plate, facing this page) Sequence of photographs from high-speed movies taken in special visualization diesel engines shown in Fig. 10-3: (a) combustion of single spray burning under large DI engine conditions; (b) combustion of four sprays in DI engine with counterclockwise swirl; (c) combustion of single spray in M.A.N. (a) "M" DI diesel; (d) combustion in prechamber (on left) and main chamber (on right) in Ricardo Comet IDI swirl chamber diesel. 1250 rev/min, imep = 827 kPa (120 1b/in2)1. 2 (Courtesy Ricardo Consulting Engineers.) FIGURE 10-4 500 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 501 TABLE 10.2 Interpretation of diesel engine combustion color photographs1 formed by the richer mixture. Flame propagation back to the injector follows extremely rapidly and at TC the bowl is filled with flame. At 5º ATC the flame Color Interpretation spreads out over the piston crown toward the cylinder wall due to combustion- Background; the gas (air in early produced gas expansion and the reverse squish flow (see Sec. 8.4). The brown Grey stages, combustion products later) regions (13º) are soot-laden fuel-rich mixture originating from the fuel which is transparent and not glowing impinges on the wall. The last frame (30º ATC) shows the gradual diminution of Green Early in combustion process; low the soot-particle-laden regions as they mix with the excess air and burn up. The luminosity " premixed "-type flame, last dull-red flame visible on the film is at about 75º ATC, well into the expansion rendered visible by copper added to stroke. fuel. Later; burned gas above about 1800ºC Figure 10-4c shows the combustion sequence for the M.A.N. "M"-type DI White, and yellow-white Carbon particle burnup in diffusion engine. In the version of the system used for these experiments, the fuel was flame, 2000-2500ºC injected through a two-hole nozzle which produces a main jet directed tangen- Yellow, orange-red Carbon burnup in diffusion flame tially onto the walls of the spherical cup in the piston crown, and an auxiliary at lower temperatures; last visible spray which mixes a small fraction of the fuel directly with the swirling air flow. in film at 1000ºC More recent "M" systems use a pintle nozzle with a single variable orifice.3 At Brown Soot clouds from very fuel-rich -5º the fuel spray is about halfway round the bowl. Ignition has just occurred of mixture regions. Where these meet fuel adjacent to the wall which has mixed sufficiently with air to burn. The flame air (grey) there is always a white fringe of hot flame spreads rapidly (-2º, 1º) to envelop the fuel spray, and is convecteuround the cup by the high swirl air flow. By shortly after TC the flame has filled the bowl and is spreading out over the piston crown. A soot cloud is seen near the top right of the picture at 5º ATC which spreads out around the circumference of the enflamed region. There is always a rim of flame between the soot cloud and the the fuel-rich spray core. At this stage (-1º), about 60 percent of the fuel has been cylinder liner as excess air is mixed into the flame zone (10.5º). The flame is of the injected. The remainder is injected into this enflamed region, producing a very carbon-burning type throughout; little premixed green flame is seen even at the fuel-rich zone apparent as the dark brown cloud (11º). This soot cloud moves to beginning of the combustion process. the outer region of the chamber (11º to 20º); white-yellow flame activity con- Figure 10-4d shows the combustion sequence in a swirl chamber IDI engine tinues near the injector, probably due to combustion of ligaments of fuel which of the Ricardo Comet V design. The swirl chamber (on the left) is seen in the view issued from the injector nozzle as the injector needle was seating. Combustion of the lower drawing of Fig. 10-3d (with the connecting passageway entering the continues well into the expansion stroke (31ºC). swirl chamber tangentially at the bottom left to produce clockwise swirl). The This sequence shows that fuel distribution is always highly nonuniform main chamber is seen in the plan view of the upper drawing of Fig. 10-3d. Two during the combustion process in this type of DI engine. Also the air which is sprays emerge from the Pintaux nozzle after the start of injection at -11º. The between the individual fuel sprays of the quiescent open-chamber diesel mixes smaller auxiliary spray which is radial is sharply deflected by the high swirl. with each burning spray relatively slowly, contributing to the poor air utilization Frame 1 shows how the main spray follows the contour of the chamber; the with this type of combustion chamber. auxiliary spray has evaporated and can no longer be seen. The first flame occurs Figure 10-4b shows a combustion sequence from the DI engine with swirl at -1º in the vaporized fuel from the auxiliary spray and is a green premixed (the chamber shown in Fig. 10-3b). The inner circle corresponds to the deep bowl flame. The flame then spreads to the main spray (TC), becoming a yellow-white in the piston crown, the outer circle to the cylinder liner. The fuel sprays (of carbon-particle-burning flame with a green fringe. At 4º ATC the swirl chamber which two are visible without obstruction from the injector) first appear at - 13º. appears full of carbon-burning flame, which is being blown down the throat and At -7º they have reached the wall of the bowl; the tips of the sprays have been into the recesses in the piston crown by the combustion generated pressure rise in deflected slightly by the anticlockwise swirl. The frame at -3º shows the first the prechamber. The flame jet impinges on the piston recesses entraining the air ignition. Bright luminous flame zones are visible, one on each spray. Out by the in the main chamber, leaving green patches where all carbon is burned out (4º, bowl walls, where fuel vapor has been blown around by the swirl, larger greenish 11º, 15º). A brown soot cloud is emerging from the throat. By 15º ATC this soot burning regions indicating the presence of premixed flame can be seen. The fuel cloud has spread around the cylinder, with a bright yellow-white flame at its downstream of each spray is next to ignite, burning yellow-white due to the soot periphery. This soot then finds excess air and burns up, while the yellow-white 502 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 503 flame becomes yellow and then orange-red as the gases cool on expansion. By 10.3.2 Combustion in Direct-Injection, 38º ATC most of the flame is burnt out. Multispray Systems Magnified color photographs of the flame around a single fuel spray under Figure 10-6 shows typical data for cylinder pressure (p), fuel-injector needle-lift, conditions typical of a direct-injection diesel engine, shown in Fig. 10-5 on the and fuel pressure in the nozzle gallery through the compression and expansion color plate, provide additional insight into the compression-ignition and flame- strokes of a direct-injection diesel. The engine had central fuel injection through a development processes.4 These photographs were obtained in a rapid compres- four-hole nozzle into a disc-shaped bowl-in-piston combustion chamber. The rate sion machine: this device is a cylinder-piston apparatus in which air is rapidly of fuel injection can be obtained from the fuel-line pressure, cylinder pressure, compressed by moving the piston to temperatures and pressures similar to those nozzle geometry, and needle-lift profiles by considering the injector as one or in the diesel engine combustion chamber at the time of injection. A single fuel more flow restrictions; it is similar in phasing and comparable in shape to the spray was then injected into the disc-shaped combustion chamber. The air flow needle-lift profile. There is a delay of 9º between the start of injection and start of prior to compression was forced to swirl around the cylinder axis and much of combustion [identified by the change in slope of the p(0) curve]. The pressure that swirl remains after compression. rises rapidly for a few crank angle degrees, then more slowly to a peak value Figure 10-5a shows a portion of the liquid fuel spray (which appears black about 5º after TC. Injection continues after the start of combustion. A rate-of- due to back lighting) and the rapidly developing flame 0.4 ms after ignition heat-release diagramt from the same study, corresponding to this rate of fuel occurs. Ignition commences in the fuel vapor-air mixture region, set up by the jet injection and cylinder pressure data, is shown in Fig. 10-7. The general shape of motion and swirling air flow, away from the liquid core of the spray. In this the rate-of-heat-release curve is typical of this type of DI engine over its load and region the smaller fuel droplets have evaporated in the hot air atmosphere that speed range. The heat-release-rate diagram shows negligible heat release until surrounds them and mixed with sufficient air for combustion to occur. Notice toward the end of compression when a slight loss of heat during the delay period that the fuel vapor concentration must be nonuniform; combustion apparently (which is due to heat transfer to the walls and to fuel vaporization and heating) is occurs around small "lumps" of mixture of the appropriate composition and temperature. Figure 10-5b shows the same flame at a later time, 3.2 ms after ignition. The flame now surrounds most of the liquid spray core. Its irregular boundary reflects the turbulent character of the fuel spray and its color variation indicates that the temperature and composition in the flame region are not uniform. Figure 10-5c shows a portion of this main flame region enlarged to show its - P internal structure. A highly convoluted flame region is evident, which has a similar appearance to a gaseous turbulent diffusion flame. The major portion of IN the diesel engine flame has this character, indicative of the burning of fuel vapor- air pockets or lumps or eddies of the appropriate composition. Only at the end of the combustion process is there visible evidence of individual fuel droplets - PI burning with an envelope flame. Figure 10-5d shows the same region of the com- bustion chamber as Fig. 10-5c, but at the end of the burning process well after injection has been completed. A few large droplets are seen burning with individ- FIGURE 10-6 ual droplet flames. It is presumed that such large drops were formed at the end of -80 -60 -40 -20 TC 20 40 60 80 Cylinder pressure p, injector needle lift ly, and injection-system fuel-line pressure p ,, as functions of the injection process as the injector nozzle was closing. Crank angle, deg crank angle for small DI diesel engine.5 . ..... .- ....... FIGURE 10-5 (On Color Plate, facing page 498) t The heat-release rate plotted here is the net heat-release rate (see Sec. 10.4). It is the sum of the Photographs from high-speed movie of single fuel spray injected into a swirling air flow in a rapid- change of sensible internal energy of the cylinder gases and the work done on the piston. It differs compression machine. (a) Spray and flame 0.4 ms after ignition; scale on right in millimeters. from the rate of fuel energy released by combustion by the heat transferred to the combustion (b) Flame surrounding spray 3.2 ms after ignition. (c) Magnified photograph of main portion of flame. chamber walls. The heat loss to the walls is 10 to 25 percent of the fuel heating value in smaller (d) Individual droplet burning late in combustion process after injection completed. Air temperature engines; it is less in larger engine sizes. This net heat release can be used as an indicator of actual heat ~ 500ºC. 50 mg fuel injected." (Courtesy Professor M. Ogasawara, Osaka University.) release when the heat loss is small. COMBUSTION IN COMPRESSION-IGNITION ENGINES 505 504 INTERNAL COMBUSTION ENGINE FUNDAMENTALS evident. During the combustion process the burning proceeds in three distin- guishable stages. In the first stage, the rate of burning is generally very high and lasts for only a few crank angle degrees. It corresponds to the period of rapid Rate of injection cylinder pressure rise. The second stage corresponds to a period of gradually decreasing heat-release rate (though it initially may rise to a second, lower, peak 1|2|3 4 5 6 7 8 9 as in Fig. 10-7). This is the main heat-release period and lasts about 40º. Nor- mally about 80 percent of the total fuel energy is released in the first two periods. The third stage corresponds to the "tail" of the heat-release diagram in which a small but distinguishable rate of heat release persists throughout much of the Ignition Injection expansion stroke. The heat release during this period usually amounts to about Rate of burning 20 percent of the total fuel energy. 4 5 From studies of rate-of-injection and heat-release diagrams such as those in Fig. 10-7, over a range of engine loads, speeds, and injection timings, Lyn6 devel- oped the following summary observations. First, the total burning period is much longer than the injection period. Second, the absolute burning rate increases pro- 10 TO 10 20 30 40 50 60 portionally with increasing engine speed; thus on a crank angle basis, the burning interval remains essentially constant. Third, the magnitude of the initial Crank angle, deg peak of the burning-rate diagram depends on the ignition delay period, being FIGURE 10-8 higher for longer delays. These considerations, coupled with engine combustion Schematic of relationship between rate of fuel injection and rate of fuel burning or energy release.6 photographic studies, lead to the following model for diesel combustion. Figure 10-8 shows schematically the rate-of-injection and rate-of-burning mixture (the shaded region in Fig. 10-8) is then added to the mixture which diagrams, where the injected fuel as it enters the combustion chamber has been becomes ready for burning after the end of the delay period, producing the high divided into a number of elements. The first element which enters mixes with air initial rate of burning as shown. Such a heat-release profile is generally observed and becomes "ready for burning" (i.e ., mixes to within combustible limits), as with this type of naturally aspirated DI diesel engine. Photographs (such as those shown conceptually by the lowest triangle along the abscissa in the rate-of- in Fig. 10-4a and b) show that up to the heat-release-rate peak, flame regions of burning figure. While some of this fuel mixes rapidly with air, part of it will mix low green luminosity are apparent because the burning is predominantly of the much more slowly. The second and subsequent elements will mix with air in a premixed part of the spray. After the peak, as the amount of premixed mixture similar manner, and the total "ready-for-burning" diagram, enclosed by the available for burning decreases and the amount of fresh mixture mixed to be dashed line, is obtained. The total area of this diagram is equal to that of the " ready for burning" increases, the spray burns essentially as a turbulent diffusion rate-of-injection diagram. Ignition does not occur until after the delay period is flame with high yellow-white or orange luminosity due to the presence of carbon over, however. At the ignition point, some of the fuel already injected has mixed particles. with enough air to be within the combustible limits. That "premixed" fuel-air To summarize, the following stages of the overall compression-ignition diesel combustion process can be defined. They are identified on the typical heat- release-rate diagram for a DI engine in Fig. 10-9. 150- -10 125- 8 Ignition delay (ab). The period between the start of fuel injection into the combustion chamber and the start of combustion [determined from the change 100H 6 Pressure, MPa in slope on the p-0 diagram, or from a heat-release analysis of the p(0) data, or mg and On, KJ/kg air 75- P from a luminosity detector]. 4 FIGURE 10-7 50- Cylinder pressure p, rate of fuel Premixed or rapid combustion phase (bc). In this phase, combustion of the On 2 injection my, and net heat- fuel which has mixed with air to within the flammability limits during the igni- 25 release rate Q ,, calculated from p tion delay period occurs rapidly in a few crank angle degrees. When this burning OL for small DI diesel engine, mixture is added to the fuel which becomes ready for burning and burns during -40 -20 TO 20 40 60 80 100 1000 rev/min, normal injection this phase, the high heat-release rates characteristic of this phase result. Crank angle, deg timing, bmep = 620 kPa.5 506 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 507 rates after ignition are relatively high, mixing must speed up. This occurs due to the centrifugal forces set up in the swirling flow. Initially, the fuel is placed near the wall, and mixing is inhibited by the effect of the high centrifugal forces on the Ignition Premixed combustion phase fuel vapor which is of higher density than the air and so tends to remain near the delay Rate of heat release period wall. Once ignition occurs, the hot burning mixture expands, decreases in density, and is then moved rapidly toward the center of the chamber. This strong radial Mixing-controlled combustion phase mixing is the rate-determining process. An additional delaying mechanism exists Late if significant fuel is deposited on the wall. At compression air temperatures, the combustion heat transferred to the fuel film on the wall from the gases in the cylinder is too C SOI EOI phase d e small to account for the observed burning rates. Only after combustion starts will 160 a 170 b 1 180 190 200 210 the gas temperature and heat-transfer rates be high enough to evaporate the fuel Crank angle, deg off the wall at an adequate rate. In the swirl chamber IDI engine, where the air in the main chamber is not FIGURE 10-9 Typical DI engine heat-release-rate diagram identifying different diesel combustion phases. immediately available for mixing, again the rate-determining processes are differ- ent.6 There is no initial spike on the rate-of-heat-release curve as was the case with DI engines. The small size of the chamber, together with the high swirl rate Mixing-controlled combustion phase (cd). Once the fuel and air which pre- generated just before injection, results in considerable fuel impingement on the mixed during the ignition delay have been consumed, the burning rate (or heat- walls. This and the fact that the ignition delay is usually shorter with the IDI release rate) is controlled by the rate at which mixture becomes available for engine due to the higher compression ratio used account for the low initial burning. While several processes are involved-liquid fuel atomization, vapor- burning rate. ization, mixing of fuel vapor with air, preflame chemical reactions-the rate of Based on the above discussion Lyn6 proposed three basic injection, mixing, burning is controlled in this phase primarily by the fuel vapor-air mixing process. and burning patterns important in diesel engines: The heat-release rate may or may not reach a second (usually lower) peak in this phase; it decreases as this phase progresses. A. Fuel injection across the chamber with substantial momentum. Mixing pro- Late combustion phase (de). Heat release continues at a lower rate well into ceeds immediately as fuel enters the chamber and is little affected by com- the expansion stroke. There are several reasons for this. A small fraction of the bustion. fuel may not yet have burned. A fraction of the fuel energy is present in soot and B. Fuel deposition on the combustion chamber walls. Negligible mixing during fuel-rich combustion products and can still be released. The cylinder charge is the delay period due to limited evaporation. After ignition, evaporation nonuniform and mixing during this period promotes more complete combustion becomes rapid and its rate is controlled by access of hot gases to the surface, and less-dissociated product gases. The kinetics of the final burnout processes radial mixing being induced by differential centrifugal forces. Burning is there- become slower as the temperature of the cylinder gases fall during expansion. fore delayed by the ignition lag. C. Fuel distributed near the wall: mixing proceeds during the delay, but at a rate 10.3.3 Application of Model to Other smaller than in mechanism A. After ignition, mixing is accelerated by the Combustion Systems same mechanism as in mechanism B. In the M.A.N. "M" DI engine system, and in IDI systems, the shapes of the heat-release-rate curve are different from those of the quiescent or moderate swirl Figure 10-10 shows, schematically, the construction of the burning-rate or DI shown in Figs. 10-7 and 10-9. With the "M" system, the initial heat-release heat-release-rate diagrams (from the same injection-rate diagram) for the DI "spike" is much less pronounced (in spite of the fact that a large fraction of the diesel combustion system with a central multihole nozzle, for the "M"-type DI fuel is injected during the delay period) though the total burning period is about diesel, and for the swirl chamber IDI. For the DI engine with multihole nozzle, the same. Lynº has suggested that the lower initial burning rate is due to the fact mechanism A is predominant. For the DI engine with fuel sprayed tangentially to that the smaller number of nozzle holes (one or two instead of about four or the wall, mechanisms B and C prevail; the delayed mixing prevents excessively more) and the directing of the main spray tangentially to the wall substantially high initial burning rates. For the IDI swirl chamber engine, the shorter ignition reduce the free mixing surface area of the fuel jets. However, since the burning delay together with mixing process C during the delay period produces a gradual increase in burning rate, as shown in Fig. 10-10c. COMBUSTION IN COMPRESSION-IGNITION ENGINES 509 508 INTERNAL COMBUSTION ENGINE FUNDAMENTALS Injection the relative importance of these crevices. Thus crevices increase heat transfer Injection Mechanism C Injection Mechanism C Delayed mixing by and contain a nonnegligible fraction of the cylinder charge at conditions that mechanisms B and C Delayed mixing by are different from the rest of the combustion chamber. Immediate mixing mechanism C by mechanism A Due to difficulties in dealing with these problems, both sophisticated methods of analysis and more simple methods give only approximate answers. Time Time Time Ignition Long Short delay (a) delay (b) delay (c) FIGURE 10-10 10.4.1 Combustion Efficiency Schematic injection-rate and burning-rate diagrams in three different types of naturally aspirated diesel combustion system: (a) DI engine with central multihole nozzle; (b) DI "M"-type engine with In both heat-release and fuel mass burned estimations, an important factor is the fuel injected on wall; (c) IDI swirl chamber engine. Mechanisms A, B, and C defined in text.6 completeness of combustion. Air utilization in diesels is limited by the onset of black smoke in the exhaust. The smoke is soot particles which are mainly carbon. 10.4 ANALYSIS OF CYLINDER While smoke and other incomplete combustion products such as unburned PRESSURE DATA hydrocarbons and carbon monoxide represent a combustion inefficiency, the magnitude of that inefficiency is small. At full load conditions, if only 0.5 percent Cylinder pressure versus crank angle data over the compression and expansion of the fuel supplied is present in the exhaust as black smoke, the result would be strokes of the engine operating cycle can be used to obtain quantitative informa- unacceptable. Hydrocarbon emissions are the order of or less than 1 percent of tion on the progress of combustion. Suitable methods of analysis which yield the the fuel. The fuel energy corresponding to the exhausted carbon monoxide is rate of release of the fuel's chemical energy (often called heat release), or rate of about 0.5 percent. Thus, the combustion inefficiency [Eq. (4.69)] is usually less fuel burning, through the diesel engine combustion process will now be described. than 2 percent; the combustion efficiency is usually greater than about 98 percent The methods of analysis are similar to those described in Sec. 9.2.2 for spark- (see Fig. 3-9). While these emissions are important in terms of their air-pollution ignition engines and start with the first law of thermodynamics for an open impact (see Chap. 11), from the point of view of energy conversion it is a good system which is quasi static (i.e ., uniform in pressure and temperature). The first approximation to regard combustion and heat release as essentially complete. law for such a system (see Fig. 9-11) is do av dt - P dt - + [mih, = d at (10.1) 10.4.2 Direct-Injection Engines where dQ/dt is the heat-transfer rate across the system boundary into the system, For this type of engine, the cylinder contents are a single open system. The only p(dV/dt) is the rate of work transfer done by the system due to system boundary mass flows across the system boundary (while the intake and exhaust valves are displacement, m; is the mass flow rate into the system across the system boundary closed) are the fuel and the crevice flow. An approach which incorporates the at location i (flow out of the system would be negative), h, is the enthalpy of flux i crevice flow has been described in Sec. 9.2.2; crevice flow effects will be omitted entering or leaving the system, and U is the energy of the material contained here. Equation (10.1) therefore becomes inside the system boundary. dv du The following problems make the application of this equation to diesel dt - P dt - + mgh s = dt (10.2) combustion difficult: Two common methods are used to obtain combustion information from 1. Fuel is injected into the cylinder. Liquid fuel is added to the cylinder which pressure data using Eq. (10.2). In both approaches, the cylinder contents are vaporizes and mixes with air to produce a fuel/air ratio distribution which is assumed to be at a uniform temperature at each instant in time during the com- nonuniform and varies with time. The process is not quasi static. bustion process. One method yields fuel energy- or heat-release rate; the other 2. The composition of the burned gases (also nonuniform) is not known. method yields a fuel mass burning rate. The term apparent is often used to 3. The accuracy of available correlations for predicting heat transfer in diesel describe these quantities since both are approximations to the real quantities which cannot be determined exactly. engines is not well defined (see Chap. 12). 4. Crevice regions (such as the volumes between the piston, rings, and cylinder HEAT-RELEASE ANALYSIS. If U and hy in Eq. (10.2) are taken to be the sensible wall) constitute a few percent of the clearance volume. The gas in these regions internal energy of the cylinder contents and the sensible enthalpy of the injected is cooled to close to the wall temperature, increasing its density and, therefore, 510 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 511 100 fuel, respectively,+ then dQ/dt becomes the difference between the chemical energy or heat released by combustion of the fuel (a positive quantity) and the Heat transfer 80 heat transfer from the system (in engines, the heat transfer is from the system and Qch by thermodynamic convention is a negative quantity). Since h ,, f ~ 0, Eq. (10.2) 60 en Crevices becomes Heat release, % fuel energy do. doch dQnt dV dU du, 40 + (10.3) dt dt dt = p dt dt 20 - FIGURE 10-11 The apparent net heat-release rate, dQ,/dt, which is the difference between the Fuel vaporization Gross and net heat-release profile during com- apparent gross heat-release rate dQ /dt and the heat-transfer rate to the walls 0 bustion, for a turbocharged DI diesel engine in mid- dOm/dt, equals the rate at which work is done on the piston plus the rate of load, mid-speed range, showing relative magnitude -50 0 50 100 change of sensible internal energy of the cylinder contents. 150 of heat transfer, crevice, and fuel vaporization and If we further assume that the contents of the cylinder can be modeled as an Crank angle, deg ATC heatup effects. ideal gas, then Eq. (10.3) becomes den = p- dv dT (10.4) and (3) the effects of the crevice regions. These additional phenomena must be dt dt - + mc, dt dealt with at an equivalent level of accuracy for more complex heat-release From the ideal gas law, pV = mRT, with R assumed constant, it follows that models to be worth while. For many engineering applications, Eq. (10.6) is ade- quate for diesel engine combustion analysis. dp + dv dT (10.5) Additional insight can be obtained by incorporating a model for the largest P of the effects omitted from Eq. (10.6), the heat transfer dOn/dt (see Chap. 12); we thereby obtain a close approximation to the gross heat-release rate. The integral Equation (10.5) can be used to eliminate T from Eq. (10.4) to give of the gross heat-release rate over the complete combustion process should then den - dt 1 + = P - dv equal (to within a few percent only, since the analysis is not exact) the mass of at at fuel injected my times the fuel lower heating value QLHv : i.e ., de .. dv 1 dp or Qch = tend dech dt = m, LHV P 10.6) (10.7) dt y - 1 dt 7 - 1 dt Jt start dt Here y is the ratio of specific heats, c,/c ,. An appropriate range for y for diesel Of course, Eqs. (10.1) to (10.4), (10.6), and (10.7) also hold with crank angle 0 as heat-release analysis is 1.3 to 1.35; Eq. (10.6) is often used with a constant value the independent variable instead of time t. of y within this range. More specifically, we would expect y for diesel engine Figure 10-11 illustrates the relative magnitude of gross and net heat release, heat-release analysis to have values appropriate to air at end-of-compression- heat transfer, crevice effects, and heat of vaporization and heating up of the fuel stroke temperatures prior to combustion (~1.35) and to burned gases at the for a turbocharged DI diesel engine operating in the mid-load, mid-speed range. overall equivalence ratio following combustion (~1.26-1.3). The appropriate The net heat release is the gross heat release due to combustion, less the heat values for y during combustion which will give most accurate heat-release infor- transfer to the walls, crevice effects, and the effect of fuel vaporization and heatup mation are not well defined.7.8 (which was omitted above by neglecting the mass addition term in dU/dt). This More complete methods of heat-release analysis based on Eq. (10.2) have last term is sufficiently small to be neglected. The enthalpy of vaporization of been proposed and used. These use more sophisticated models for the gas proper- diesel fuel is less than 1 percent of its heating value; the energy change associated ties before, during, and after combustion, and for heat transfer and crevice with heating up fuel vapor from injection temperature to typical compression air effects.8 However, it is also necessary to deal with the additional issues of: (1) temperatures is about 3 percent of the fuel heating value. The heat transfer inte- mixture nonuniformity (fuel/air ratio nonuniformity and burned and unburned grated over the duration of the combustion period is 10 to 25 percent of the total gas nonuniformities); (2) accuracy of any heat-transfer model used (see Chap. 12); heat released. FUEL MASS BURNING RATE ANALYSIS. If the internal energies of the fuel, air, 1 That is, U = U, = U(T) - U(298 K) and h, = h. , = h(T) - h, (298 K); see Sec. 5.5 for definition. and burned gases in Eq. (10.1) are evaluated relative to a consistent datum (such 512 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 513 as that described in Sec. 4.5.2), then this equation can be used to obtain an appar- -8 ent fuel mass burning rate from cylinder pressure versus crank angle data. (With 3 such a species energy datum the "heat release" is properly accounted for in the internal energy and enthalpy terms.) Following Krieger and Borman,9 Eq. (10.2) 2 6 can be written as Fuel mass burning rate, mg/deg dv dm 5 Pressure, MPa d (mu) = - P at + de dt ths dt (10.8) - 4 FIGURE 10-12 dt Cylinder pressure p and fuel mass Here Q is the heat transfer to the gas within the combustion chamber (that is, 3 burning rate calculated from p, as 0 Q = - (h), m is the mass within the combustion chamber, and dm/dt has been a function of crank angle, using 12 the Krieger and Borman method9 substituted for ms. 160 TC 200 220 240 260 for DI diesel engine at 3200 rev/ Since the properties of the gases in the cylinder during combustion Crank angle, deg min and full load. (assumed to be uniform and in chemical equilibrium at the pressure p and average temperature T) are in general a function of p, T, and the equivalence ratio o, Equation (10.12) can be solved numerically for m(t) given mo, do, p(t), and appro- u = u(T, p, o) and R = R(T, p, $) priate models for the working fluid properties (see Sec. 4.7) and for the heat- transfer term dQ/dt (see Chap. 12). Therefore Figure 10-12 shows cylinder pressure data for an open chamber DI diesel du au dT ou dp du do and fuel mass burning rate dm/de calculated from that data using the above + + of dt (10.9a) method. The heat-transfer model of Annand was used (see Sec. 12.4.3). The result dt ÔT dt Op dt obtained is an apparent fuel mass burning rate. It is best interpreted, after multi- dr aR dT OR dp OR do (10.9b) plying by the heating value of the fuel, as the fuel chemical-energy or heat- dt + OT dt Op dt + ao dt release rate. The actual fuel burning rate is unknown because not all the fuel "burns" with sufficient air available locally to produce products of complete Also, combustion. About 60 percent of the fuel has burned in the first one-third of the m 1 + (F/A). total combustion period. The integral of the fuel mass burning rate over the Q = @0 + mo (10.10) (F/A), combustion process should equal the total fuel mass burned; in this case it is 3 percent less than the total fuel mass injected. Note that chemical energy con- do 1 + (F/A)o dm and (10.11) tinues to be released well into the expansion process. The accuracy of this type of dt (F/A), mo dt calculation then decreases, however, since errors in estimating heat transfer sig- nificantly affect the apparent fuel burning rate. (F/A) is the fuel/air ratio; the subscript 0 denotes the initial value prior to fuel Krieger and Borman also carried out sensitivity analyses for the critical injection and the subscript s denotes the stoichiometric value. It then follows that assumptions and variables. They found that the effect of dissociation of the 1 dm _ _ (RT/VXdV/dt) - (du/@pdp/dt) + (1/m)(dQ/dt) - CB (10.12 ) product gases was negligible. This permits a substantial simplification of Eq. m dt u - hy + D(ou/0d) - C[1 + (D/RX(OR/06)] (10.12). With no dissociation, u = u(T, o), and R = R/M can be taken as con- stant, since the molecular weight M changes little. Then where dm [1 + (c/R)]p(dV/dt) + (c/R)V(dp/dt) - (dQ/dt) B == 1 dp 1 @R dp 1 dv dt p dt R op dt V dt hy + (cu/RXpV/m) - u - D(ou/ao) (10.13) where D, as before, is [1 + (F/A)o]m/[(F/A), mo]. Given the uncertainties inherent T(Ou/OT) c = in the heat-transfer model and the neglect of nonuniformities and crevices, Eq. 1 + (T/RX(OR/OT) (10.13) represents an adequate level of sophistication. D = [1 + (F/A)0]m The other sensitivity variations studied by Krieger and Borman were: shift- ing of the phasing of the pressure data 2º forward and 2º backward; translating (F/A)3 mo the pressure data +34 kPa (5 lb/in2); changing the heat transfer +50 percent; COMBUSTION IN COMPRESSION-IGNITION ENGINES 515 514 INTERNAL COMBUSTION ENGINE FUNDAMENTALS P2, 12, 12, m2 Here dm/dt is the mass flow rate between the chambers with positive flow from the prechamber to the main chamber. If dm/dt > 0, h2.1 = h2; if dm/dt < 0, Prechamber - h2.1 = h1. If we define U, and U2 as sensible internal energies and hy as the Main chamber sensible enthalpy of the fuel, then dQ1/dt and dQ2/dt represent the net heat- release rates-the difference between the combustion energy-release rates and the dm rates of heat transfer to the walls. If we use an ideal gas model for the working fluid in each chamber, with cp, Cp, and M constant, the relation p1 V1 = m1RT1 and P2 V2 = m2 RT2 can be used to eliminate m and T from the dU/dt terms and, with the fact that h ,, y = 0, can be used to write Eqs. (10.14) and (10.15) as FIGURE 10-13 de1 Y Schematic defining variables in main chamber P1 dv1 + 1 am dt 7 - 1 dt y - 1 - V1 at dp1 - Cp T2.1 dt (10.16) (subscript 1) and prechamber (subscript 2) for IDI engine heat-release analysis. dQ2 1 dm = V2 dp2 + C. 12,1 dt y - 1 dt dt (10.17) increasing the initial mass 5 percent. The initial mass change had a negligible When Eqs. (10.16) and (10.17) are added together, the term representing the en- effect on the fuel burning rate calculations. The heat-transfer changes of +50 percent changed the mass of fuel burned by about +5 percent. The change in thalpy flux between the two chambers cancels out, and the following equation for total net heat-release results: phasing of the pressure data was more significant. It needs to be stressed that accurate (in magnitude and phasing) pressure data are a most important require- de de1 de2 1 1 - + - + api dp2 ment for useful heat-release or fuel mass burning rate analysis. dt dt dt -101 dt y - V1 V2 (10.18) dt dt 10.4.3 Indirect-Injection Engines The comments made in the previous section regarding the interpretation of the net heat release (it is the gross heat release due to combustion less the heat In IDI diesel engines, the pressures in each of the two chambers, main and aux- iliary, are not the same during the combustion process. Since combustion starts transfer to the walls, and other smaller energy transfers due to crevices, fuel vaporization, and heatup) also hold here. in the auxiliary or prechamber, the fuel energy release in the prechamber causes the pressure there to rise above the main chamber pressure. Depending on com- In practice, Eq. (10.18) is difficult to use since it requires experimental data for both the main and auxiliary chamber pressures throughout the combustion bustion chamber design and operating conditions, the prechamber pressure rises to be 0.5 to 5 atm above that in the main chamber. This pressure difference process. Access for two pressure transducers through the cylinder head is not causes a flow of fuel, air, and burning and burned gases into the main chamber, often available; even when access can be achieved, the task of obtaining pressure where additional energy release now occurs. The analysis of the DI diesel in the data from two different transducers under the demanding thermal loading condi- tions found in IDI diesels, of sufficient accuracy such that the difference between previous section was based on uniform pressure throughout the combustion chamber. For IDI engines the effect of the pressure difference between the cham- the pressures (of order 0.5 to 5 atm) at pressure levels of 60 to 80 atm can be interpreted, requires extreme diligence in technique.10,11 Figure 10-14a and b bers must usually be included. Figure 10-13 shows an IDI combustion chamber divided at the nozzle into shows apparent net heat-release rate profiles for an IDI diesel obtained using Eq. two open systems. Applying the first law [Eq. (10.1)] to the main chamber yields (10.18) with y = 1.35.11 Curves of dQ/dt and dQ/de are shown at three different speeds and essentially constant fuel mass injected per cycle. While the absolute dV 1 + h2, 1 dm dU 1 heat-release rates increase with increasing speed, the relative rates are essentially de1 - P1 dt (10.14) dt dt dt independent of speed, indicating that combustion rates, which depend on fuel-air mixing rates, scale approximately with engine speed. and to the auxiliary chamber yields Equation (10.18) (or its equivalent) has been used assuming p2 = p1 and dm dm, dU 2 using either main chamber or auxiliary chamber pressure data alone. The error dQ2 - h2,1 dt (10.15) dt tns dt dt associated with this approximation can be estimated as follows. If we write p2 = 516 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 517 300 rev/min mg fuel 2500 16.5 rev/min mg fuel 1.2|- 1.2|- 250 1500 15.0 25 2500 16.5 1035 15.8 1.500 15.0 200 20 1035 15.8 0.8 .8 150- 15- Q, kJ 2, kJ Heat-release rate, kw Heat-release rate, J/deg 100- 10H 0.4 0.4 50 -20 0 0 20 40 -40 20 40 -40 0 40 80 120 -40 40 80 120 Ic TC -50 -5- 0.12 0.06 Crank angle, deg Crank angle, deg ( a ) (b) 1320 rev/min 2800 rev/min FIGURE 10-14 0.08 0.04 - Calculated net heat-release-rate profiles for IDI diesel engine at constant load (0.29 < < 0.32). (a) Q, kJ/deg Q, kJ/deg Heat-release rate in kilowatts. (b) Heat-release rate in joules per degree.11 0.04. 0.02 - 2 P1 + Ap then Eq. (10.18) becomes 2 0 dQ dv1 V1 + V2 dp1 +- V2 d(Ap) 0 (10.19) -40 0 40 80 120 y - 1 P1 -40 0 dt 40 80 dt y - 1 dt y -1 dt 120 Crank angle, deg Crank angle, deg If the last term is omitted, Eq. (10.19) is identical to Eq. (10.6) derived for the DI (a) (b) diesel. Since the term V(dp1/dt)/(y - 1) is much larger than the first term on the FIGURE 10-15 right-hand side of Eq. (10.19) during the early stages of the combustion process, Calculated gross heat-release rates in IDI swirl-chamber diesel engine at full load. 1 Prechamber heat the error involved in omitting the last term is given to a good approximation by release. 2 Main chamber heat release. Top figures: integrated heat release. Bottom figures: heat- release rate. (a) 1320 rev/min; (b) 2800 rev/min.10 [V2/(V1 + V2)]d(Ap)/dp1. In the initial stages of combustion this error can be quite large (of order 0.25 based on data in Ref. 10 close to TC). Later in the com- 10-15. For this particular engine, at low engine speeds two-thirds of the heat bustion process it becomes negligible (of order a few percent after 20º ATC).10 release occurs in the prechamber; at higher engine speeds about two-thirds of the Thus, neglecting Ap will lead to errors in predicting the initial heat-release rate. heat release occurs in the main chamber. The magnitude of the error will depend on the design of the combustion chamber and on engine speed and load (with more restricted passageways, higher loads 10.5 FUEL SPRAY BEHAVIOR and speeds, giving higher values of Ap and, therefore, greater error). Later in the combustion process the error is much less, so integrated heat-release data derived 10.5.1 Fuel Injection ignoring Ap will show a smaller error. A model analogous to the above, but using the approach of Krieger and The fuel is introduced into the cylinder of a diesel engine through a nozzle with a Borman9 (see Sec. 10.4.2), for the IDI diesel has been developed and used by large pressure differential across the nozzle orifice. The cylinder pressure at injec- Watson and Kamel.1º The energy conservation equation for an open system tion is typically in the range 50 to 100 atm. Fuel injection pressures in the range developed in Sec. 14.2.2, with energy and enthalpy modeled using a consistent 200 to 1700 atm are used depending on the engine size and type of combustion datum (see Sec. 4.5.2), with appropriate models for convective and radiation heat system employed. These large pressure differences across the injector nozzle are transfer and for gas properties, was applied to the main chamber and also to the required so that the injected liquid fuel jet will enter the chamber at sufficiently prechamber. These equations were solved using accurately measured main high velocity to (1) atomize into small-sized droplets to enable rapid evaporation chamber and prechamber pressure data to determine the apparent rate of heat and (2) traverse the combustion chamber in the time available and fully utilize the air charge. release (here, the rate of fuel burning multiplied by the fuel heating value) in the main chamber and prechambers through the combustion process. The engine was Examples of common diesel fuel-injection systems were described briefly in a Ricardo Comet swirl chamber IDI design. Some results are shown in Fig. Sec. 1.7 and illustrated in Figs. 1-17 to 1-19. (See also Refs. 12 and 13 for more 518 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 519 extensive descriptions of diesel fuel-injection systems.) The task of the fuel. Fuel intake injection system is to meter the appropriate quantity of fuel for the given engine Fuel-injection tubing speed and load to each cylinder, each cycle, and inject that fuel at the appropriate Delivery-valve holder time in the cycle at the desired rate with the spray configuration required for the Fuel filter particular combustion chamber employed. It is important that injection begin Delivery valve and end cleanly, and avoid any secondary injections. Nozzle holder assembly Barrel To accomplish this task, fuel is usually drawn from the fuel tank by a with nozzle supply pump, and forced through a filter to the injection pump. The injection Control rack Sof pump sends fuel under pressure to the nozzle pipes which carry fuel to the injec- Plunger tor nozzles located in each cylinder head. Excess fuel goes back to the fuel tank Figures 1-17 and 1-19 show two common versions of fuel systems used with Plunger return spring multicylinder engines in the 20 to 100 KW per cylinder brake power range which Control sleeve operate with injection pressures between about 300 and 1200 atm. Plunger control arm In-line injection pumps (Fig. 1-17) are used in engines in the 40 to 100 kw Guide sleeve per cylinder maximum power range. They contain a plunger and barrel assembly Fuel-injection pump for each engine cylinder. Each plunger is raised by a cam on the pump camshaft Retainer and is forced back by the plunger return spring. The plunger stroke is fixed. The FIGURE 10-16 plunger fits sufficiently accurately within the barrel to seal without additional Fuel-injection system with single-barrel pump. Left: system layout. Right: section through fuel- sealing elements, even at high pressures and low speeds. The amount of fuel injection pump. (Courtesy Robert Bosch GmbH and SAE.14) delivered is altered by varying the effective plunger stroke. This is achieved by means of a control rod or rack, which moves in the pump housing and rotates through the fuel-injection pump.14 Such pumps are driven by an auxiliary cam the plunger via a ring gear or linkage lever on the control sleeve. The plunger on the engine camshaft. Also used extensively on larger engines are unit injectors chamber above the plunger is always connected with the chamber below the where the pump and injector nozzle are combined into a single unit. An example plunger helix by a vertical groove or bore in the plunger. Delivery ceases when of a unit injector and its driving mechanism used on a large two-stroke cycle the plunger helix exposes the intake port (port opening), thus connecting the diesel engine is shown in Fig. 10-17. Fuel, supplied to the injector through a plunger chamber with the suction gallery. When this takes place depends on the fuel-distributing manifold, enters the cavity or plunger chamber ahead of the rotational position of the plunger. In the case of a lower helix, delivery always plunger through a metering orifice. When fuel is to be injected, the cam via the starts (port closing) at the same time, but ends sooner or later depending on the rocker arm pushes down the plunger, closing the metering orifice and compres- rotational position of the plunger. With a plunger with an upper helix, port sing the fuel, causing it to flow through check valves and discharge into the closing (start of delivery) not port opening is controlled by the helix and is varied cylinder through the injector nozzles or orifices. The amount of fuel injected is by rotating the plunger. Figure 1-18 illustrates how the plunger helix controls controlled by the rack, which controls the spill of fuel into the fuel drain manifold fuel delivery.14 by rotating the plunger with its helical relief section via the gear. Distributor-type fuel-injection pumps (such as that illustrated in Fig. 1-19) The most important part of the injection system is the nozzle. Examples of are normally used in multicylinder engines with less than 30 kW per cylinder different nozzle types and a nozzle holder assembly are shown in Fig. 1-18. The maximum power with injection pressures up to 750 atm. These pumps have only nozzles shown are fluid-controlled needle valves where the needle is forced one plunger and barrel. The pump plunger is made to describe a combined rotary against the valve seat by a spring. The pressure of the fuel in the pressure and stroke movement by the rotating cam plate. The fuel is accurately metered to chamber above the nozzle aperture opens the nozzle by the axial force it exerts each injection nozzle in turn by this plunger which simultaneously acts as the on the conical surface of the nozzle needle. Needle valves are used to prevent distributor. Such units are more compact and cheaper than in-line pumps but dribble from the nozzles when injection is not occurring. It is important to keep cannot achieve such high injection pressures. The distributor-type fuel-injection the volume of fuel left between the needle and nozzle orifices (the sac volume) as pump is combined with the automatic timing device, governor, and supply pump small as possible to prevent any fuel flowing into the cylinder after injection is to form a single unit. over, to control hydrocarbon emissions (see Sec. 11.4.4). Multihole nozzles are Single-barrel injection pumps are used on small one- and two-cylinder used with most direct-injection systems; the M.A.N. "M" system uses a single- diesel engines, as well as large engines with outputs of more than 100 kW per hole nozzle. Pintle nozzles, where the needle projects into and through the nozzle cylinder. Figure 10-16 shows the layout of the injection system and a section hole, are used in indirect-injection engine systems. The shape of the pin on the 520 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 521 Terminal Solenoid stator Armature Motion- Plunger Plunger Motion booggg Body Fuel Gear manifold- Rack Pin Retainer Nut - Bushing Spill deflector Poppet control valve Plunger Body forging To nozzle To nozzle Bushing Mechanical injector Electronic unit injector Spray tip FIGURE 10-18 Electronically controlled unit fuel-injection system.16 FIGURE 10-17 sophisticated hydraulic models: Hiroyasu17 provides an extensive reference list of Unit fuel injector and its driving mechanism, typically used in large diesel engines.15 such models. However, approximate estimates of the injection rate through the injector nozzles can be made as follows. If the pressure upstream of the injector nozzle(s) can be estimated or measured, and assuming the flow through each end of the nozzle needle controls the spray pattern and fuel-delivery character- nozzle is quasi steady, incompressible, and one dimensional, the mass flow rate of istics. Auxiliary nozzle holes are sometimes used to produce an auxiliary smaller fuel injected through the nozzle my is given by spray to aid ignition and starting. Open nozzle orifices, without a needle, are also used. im, = CDAnV2p, Ap (10.20) The technology for electronic control of injection is now available. In an electronic injector, such as that shown in Fig. 10-18, a solenoid operated control where A, is the nozzle minimum area, CD the discharge coefficient, p, the fuel valve performs the injection timing and metering functions in a fashion analo- density, and Ap the pressure drop across the nozzle. If the pressure drop across gous to the ports and helices of the mechanical injector. Solenoid valve closure the nozzle and the nozzle open area are essentially constant during the injection initiates pressurization and injection, and opening causes injection pressure decay period, the mass of fuel injected is then and end of injection. Duration of valve closure determines the quantity of fuel injected. The unit shown uses camshaft/rocker arm driven plungers to generate my = CDA 2p, Ap 40 the injection pressure, and employs needle-valve nozzles of conventional design. 360N (10.21) Increased flexibility in fuel metering and timing and simpler injector mechanical where 40 is the nozzle open period in crank angle degrees and N is engine speed. design are important advantages.16 Equations (10.20) and (10.21) illustrate the dependence of injected amounts of fuel Accurate predictions of fuel behavior within the injection system require on injection system and engine parameters. COMBUSTION IN COMPRESSION-IGNITION ENGINES 523 522 INTERNAL COMBUSTION ENGINE FUNDAMENTALS Spray angle Break up length 1004 2 3 3. 3.5 ms FIGURE 10-20 Shadowgraph and back-illuminated pho- tographs of evaporating spray injected into Droplet size distribution ?! Spray tip penetration nitrogen at 3.4 MPa and 670 K in rapid- mm compression machine. Times in milliseconds are after start of injection: injection duration 50 3.3 ms. Top (shadowgraph) photographs show full vapor and liquid region. Bottom (back-illuminated) photographs only show 2 3 3.3 3.5 ms liquid-containing core.19 FIGURE 10-19 Schematic of diesel fuel spray defining its major parameters.18 tographic techniques, back lighting and shadowgraph,t have been used to distinguish the liquid-containing core of the jet and the extent of the fuel vapor 10.5.2 Overall Spray Structure region of the spray which surrounds the liquid core. The region of the jet closest to the nozzle (until injection ceases at 3.3 ms) contains liquid drops and liga- The fuel is introduced into the combustion chamber of a diesel engine through ments; the major region of the spray is a substantial vapor cloud around this one or more nozzles or orifices with a large pressure differential between the fuel narrow core which contains liquid fuel. supply line and the cylinder. Different designs of nozzle are used (e.g ., single- Different spray configurations are used in the different diesel combustion orifice, multiorifice, throttle, or pintle; see Fig. 1-18), depending on the needs of systems described earlier in this chapter. The simplest configuration involves the combustion system employed. Standard diesel injectors usually operate with multiple sprays injected into quiescent air in the largest-size diesels (Fig. 10-1a). fuel-injection pressures between 200 and 1700 atm. At time of injection, the air in Figures 10-19 and 10-20 illustrate the essential features of each spray under these the cylinder has a pressure of 50 to 100 atm, temperature about 1000 K, and circumstances until interactions with the wall occur. Each liquid fuel jet atomizes density between 15 and 25 kg/m3. Nozzle diameters cover the range 0.2 to 1 mm into drops and ligaments at the exit from the nozzle orifice (or shortly thereafter). diameter, with length/diameter ratios from 2 to 8. Typical distillate diesel fuel The spray entrains air, spreads out, and slows down as the mass flow in the spray properties are: relative specific gravity of 0.8, viscosity between 3 and 10 kg/m .s. increases. The droplets on the outer edge of the spray evaporate first, creating a and surface tension about 3 x 10-2 N/m (at 300 K). Figure 10-19 illustrates the fuel vapor-air mixture sheath around the liquid-containing core. The highest structure of a typical DI engine fuel spray. As the liquid jet leaves the nozzle it velocities are on the jet axis. The equivalence ratio is highest on the centerline becomes turbulent and spreads out as it entrains and mixes with the surrounding (and fuel-rich along most of the jet), decreasing to zero (unmixed air) at the spray air. The initial jet velocity is greater than 102 m/s. The outer surface of the jet boundary. Once the sprays have penetrated to the outer regions of the com- breaks up into drops of order 10 um diameter, close to the nozzle exit. The liquid bustion chamber, they interact with the chamber walls. The spray is then forced column leaving the nozzle disintegrates within the cylinder over a finite length to flow tangentially along the wall. Eventually the sprays from multihole nozzles called the breakup length into drops of different sizes. As one moves away from the nozzle, the mass of air within the spray increases, the spray diverges, its width increases, and the velocity decreases. The fuel drops evaporate as this air- entrainment process proceeds. The tip of the spray penetrates further into the 1 The back lighting identifies regions where sufficient liquid fuel (as ligaments or drops) is present to combustion chamber as injection proceeds, but at a decreasing rate. Figure 10-20 attenuate the light. The shadowgraph technique responds to density gradients in the test section, so it shows photographs of a diesel fuel spray injected into quiescent air in a rapid- identifies regions where fuel vapor exists. compression machine which simulates diesel conditions.19 Two different pho- 524 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 525 3 12/ cm 10% FIGURE 10-21 Sketches of outer vapor boundary of diesel fuel spray -2 Combustion from 12 successive frames of rapid-compression- chamber wall- machine high-speed shadowgraph movie showing FIGURE 10-23 - 3 L interaction of vaporizing spray with cylindrical wall of Schlieren photographs of vaporizing sprays injected into swirling air flow in transparent prechamber 2 3 5 combustion chamber. Injection pressure 60 MPa. of special IDI diesel.21 Left: high sensitivity, showing boundaries of the vapor regions of spray. Right: Time between frames 0.14 ms.20 low sensitivity, showing liquid-containing core (dark) in relation to vapor regions (mottled). Radius, cm prechamber with high clockwise swirl. The photograph on the left, with high interact with one another. Figure 10-21 shows diesel fuel sprays interacting with sensitivity, shows the outer boundary of the fuel vapor region of the spray; the the cylindrical outer wall of a disc-shaped combustion chamber in a rapid- low-sensitivity photograph on the right locates the liquid phase regions of the compression machine, under typical diesel-injection conditions. The cylinder wall spray.21 The interaction between the swirl and both liquid and vapor spray causes the spray to split with about half flowing circumferentially in either direc- regions is evident, as is the spray interaction with the chamber wall. tion. Adjacent sprays then interact forcing the flow radially inward toward the Other spray flow patterns are used. The spray may enter the swirling air chamber axis.20 flow tangentially as in the M.A.N. "M" system shown in Fig. 10-1c. The spray Most of the other combustion systems in Figs. 10-1 and 10-2 use air swirl to then interacts immediately with the combustion chamber walls. increase fuel-air mixing rates. A schematic of the spray pattern which results To couple the spray-development process with the ignition phase of the when a fuel jet is injected radially outward into a swirling flow is shown in Fig. combustion, it is important to know which regions of the spray contain the fuel 10-22. Because there is now relative motion in both radial and tangential direc- injected at the beginning of the injection process. These regions of the sprays are tions between the initial jet and the air, the structure of the jet is more complex. likely to autoignite first. Each spray develops as follows. At the start of injection As the spray entrains air and slows down it becomes increasingly bent toward the the liquid fuel enters the quiescent air charge, atomizes, moves outward from the swirl direction; for the same injection conditions it will penetrate less with swirl nozzle, and slows down rapidly as air is entrained into the spray and accelerated. than without swirl. An important feature of the spray is the large vapor- This start-up process forms a vortex or "puff" at the head of the spray. The containing region downstream of the liquid-containing core. Figure 10-23 shows injected fuel which follows encounters less resistance; thus drops from that fuel schlieren photographs of four fuel jets injected on the axis of an IDI diesel engine overtake the drops from first-injected fuel, forcing them outward toward the per- iphery of the spray. At the tip of the unsteady spray the drops meet the highest aerodynamic resistance and slow down, but the spray continues to penetrate the Nozzle hole air charge because droplets retarded at the tip are continually replaced by new Core Upstream edge higher-momentum later-injected drops.22 Accordingly, droplets in the periphery of the spray and behind the tip of the spray come from the earliest injected fuel.23 As Figs. 10-20 and 10-23 indicate, these drops evaporate quickly. 10.5.3 Atomization FIGURE 10-22 Schematic of fuel spray injected radially outward Under diesel engine injection conditions, the fuel jet usually forms a cone-shaped from the chamber axis into swirling air flow. Shape spray at the nozzle exit. This type of behavior is classified as the atomization Air swirl of equivalence ratio (o) distribution within jet is breakup regime, and it produces droplets with sizes very much less than the Downstream edge indicated. nozzle exit diameter. This behavior is different from other modes of liquid jet 526 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 527 breakup. At low jet velocity, in the Rayleigh regime, breakup is due to the mm/2 unstable growth of surface waves caused by surface tension and results in drops O larger than the jet diameter. As jet velocity is increased, forces due to the relative motion of the jet and the surrounding air augment the surface tension force, and 0.5 lead to drop sizes of the order of the jet diameter. This is called the first wind- 1.0 induced breakup regime. A further increase in jet velocity results in breakup 1.5 characterized by divergence of the jet spray after an intact or undisturbed length 20 downstream of the nozzle. In this second wind-induced breakup regime, the unstable growth of short-wavelength waves induced by the relative motion between the liquid and surrounding air produces droplets whose average size is much less than the jet diameter. Further increases in jet velocity lead to breakup in the atomization regime, where the breakup of the outer surface of the jet occurs at, or before, the nozzle exit plane and results in droplets whose average diameter is much smaller than the nozzle diameter. Aerodynamic interactions at the liquid/gas interface appear to be one major component of the atomization mechanism in this regime. 22, 24 A sequence of very short time exposure photographs of the emergence of a liquid jet from a nozzle of 0.34 mm diameter and L,/d, = 4 into high-pressure nitrogen at ambient temperature is shown in Fig. 10-24. The figure shows how the spray tip penetrates and the spray spreads during the early part of its travel.25 Data such as these were used to examine the dependence of the spray mm/2 development on gas and liquid density, liquid viscosity, and nozzle geometry.24 26 The effects of the most significant variables, gas/liquid density ratio and nozzle geometry, on initial jet spreading angle are shown in Fig. 10-25. -1.0 For a given geometry (cylindrical hole and length/diameter = 4), the initial jet - 2.0 spreading or spray angle increases with increasing gas/liquid density ratio as 3.0 shown in Fig. 10-25a. Typical density ratios for diesel injection conditions are between 15 x 10-3 and 30 x 10-3. Of several different nozzle geometry param- -4.0 eters examined, the length/diameter ratio proved to be the most significant (see Steady state Fig. 10-25b). FIGURE 10-24 For jets in the atomization regime, the spray angle 0 was found to follow Photographs showing initial emergence and steady state (bottom right) of high-pressure liquid spray. Time between frames 2.1 us. Liquid: water. Gas: nitrogen at 1380 kPa. Ap across nozzle 11 MPa. the relationship Nozzle diameter 0.34 mm.25 tan 1 = 1223 2 A 6 (10.22) symbols) to atomization regime breakup (open symbols). The growth of aero- where pe and p, are gas and liquid densities and A is a constant for a given nozzle dynamic surface waves is known to be responsible for jet breakup in the second geometry.+ The data in Fig. 10-25a are fitted by A = 4.9. This behavior is in wind-induced breakup regime. Such a mechanism can explain the observed data accord with the theory that aerodynamic interactions are largely responsible for trends in the atomization regime, if an additional mechanism is invoked to jet breakup. Note that the data in Fig. 10-25b show a continuous trend as the jet explain nozzle geometry effects. One possible additional mechanism is liquid breakup regime makes a transition from second wind-induced breakup (solid cavitation. A criterion for the onset of jet atomization at the nozzle exit plane was developed. For (p/p.)(Re[/We)2 > 1 (which is true for distillate fuel injection applications) the design criterion is + An empirical equation for A is A = 3.0 + 0.28 (L,/d,), where L,/d ,, is the length/diameter ratio of the 1/2 < k nozzle. 25 (10.23) 528 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 529 where k is an empirical constant depending on nozzle geometry in the range 6 to 12 (k = 18.3// A). Jet breakup trends can be summarized as follows. The initial jet divergence 10 0 angles increase with increasing gas density. Divergence begins progressively closer to the nozzle as gas density increases until it reaches the nozzle exit. Jet 20 divergence angles increase with decreasing fuel viscosity; divergence begins at the nozzle exit once the liquid viscosity is below a certain level. Nozzle design affects the onset of the jet atomization regime. Jet divergence angles decrease with increasing nozzle length. For the same length, rounded inlet nozzles produce less 0, deg divergent jets than sharp-edged inlet nozzles. The initial jet divergence angle and 10- intact spray length are quasi steady with respect to changes in operating condi- Ln = 4 tions which occur on time scales longer than about 20 us.25 Note that while all these results were obtained under conditions where evaporation was not occurring, the initial spray-development processes are not significantly affected by evaporation (see Sec. 10.5.6). 0 10 20 30 40 50 60 70 80 90 × 103 10.5.4 Spray Penetration (a) The speed and extent to which the fuel spray penetrates across the combustion chamber has an important influence on air utilization and fuel-air mixing rates. 24 O Divergence at nozzle exit Intact before diverging In some engine designs, where the walls are hot and high air swirl is present, fuel Marginal " = 0.5 (0) impingement on the walls is desired. However, in multispray DI diesel com- bustion systems, overpenetration gives impingement of liquid fuel on cool sur- 20 : (0) faces which, especially with little or no air swirl, lowers mixing rates and increases O emissions of unburned and partially burned species. Yet underpenetration results 16| Eq. (10.23) in poor air utilization since the air on the periphery of the chamber does not then contact the fuel. Thus, the penetration of liquid fuel sprays under conditions Eq. (10.22) 2.1 (RI ) (0) typical of those found in diesel engines has been extensively studied. 0, deg 12 Many correlations based on experimental data and turbulent gas jet theory 10 ( v ) have been proposed for fuel spray penetration.17 These predict the penetration S 49 (0) of the fuel spray tip across the combustion chamber for injection into quiescent 00 air, as occurs in larger DI engines, as a function of time. An evaluation of these correlations27 indicated that the formula developed by Dent,28 based on a gas jet mixing model for the spray, best predicts the data:+ 85 (A) T UL S = 3.07(Ap) 1/4 (td,) 1/2/ 294) 1/4 10 20 30 40 50 60 Pa. (10.24) 8 X 103 PI where Ap is the pressure drop across the nozzle, t is time after the start of injec- (b) tion, and d ,, is the nozzle diameter. All quantities are expressed in SI units: t in FIGURE 10-25 (a) Initial spray angle of atomizing jets versus density ratio (p /p, = gas density/liquid density) for fixed nozzle geometry shown. Various fluids and gases at liquid pressures of 3.4-92 MPa. Nozzle diameters d, = 0.254, 0.343, and 0.61 mm.22 (b) Initial spray angle versus density ratio for a wide range of nozzle geometries. L,/d, = nozzle length/diameter ratio (RI = rounded inlet geometry). Solid symbols indicate jets which break up and diverge downstream of nozzle exit. Open symbols indicate For nozzles where 2 < L,/d, < 4, and for t > 0.5 ms. At exceptionally high chamber densities (p > 100 atm) Eq. (10.24) overpredicts penetration. jet breakup at nozzle exit.25 530 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 531 300 300 Measurement po = 10 MPa Measurement pq = 2 MPa 4.1 ms-3.2 ms o PQ = 1 MPa O Po = 15 MPa 2.0 ms- 100- o Pq = 2 MPa Pq = 1 MPa 100- o Po = 10 MPa Po = 15 MPa 1.2 ms -1.2 ms 2.0 ms 1.0 ms A Pa = 3 MPa A Po = 7 MPa 3.0 ms- 2.0 ms 0.5 ms 50 50 0.7 ms -0.7 ms 1.0 ms 30- 30 ~0.5 ms 2.0 ms Swirl DDDD DD DDD 3.0 ms Swirl Spray tip penetration, mm 10 Pa = 2 MPa 10H Spray tip penetration, mm 33.2 ms -- 5 Do Pa = 3 MPa 5 Po = 10 MPa N3 = 2100 rev/min Po = 7 MPa Ns = 4600 rev/min (a) W W 60 Ns = 0 rev/min a D O 14 4 0.5 N. = 2100 rev/min 0.5 D 0.3 0.3 A 0.03 0.05 0.1 0.3 0.5 3 5 10 0.03 0.05 0.1 0.3 0.5 1 3 5 10 40 D N. = 3000 rev/min Time, ms Time, ms N. = 4600 rev/min Spray tip penetration, mm FIGURE 10-26 N, = 7500 rev/min Spray tip penetration as function of time at various ambient pressures (p.) and injection pressures (po). Measurement Fuel jets injected into quiescent air at room temperature.29 20 Po = 20 MPa ( N = 0 rev/min Pa = 1.1 MPa O 2100 3000 seconds, S and d ,, in meters, Ap in pascals, p, in kilograms per cubic meter, and T. 40 0 4600 in kelvins. 7500 More detailed studies have examined the spray tip location as a function of 0 0.5 1.0 1.5 2.0 time, following start of a diesel injection process in high-pressure bombs. Data Time, ms taken by Hiroyasu et al ., 29 shown in Fig. 10-26, illustrate the sensitivity of the (b) spray tip position as a function of time to ambient gas state and injection pres- FIGURE 10-27 sure for fuel jets injected into quiescent air at room temperature. These data show (a) Measured outer boundary of sprays injected into swirling air flow. (b) Spray tip penetration as a that the initial spray tip penetration increases linearly with time t (i.e ., the spray function of time for different swirl rates. Solid lines show Eq. (10.27).29 tip velocity is constant) and, following jet breakup, then increase as t. Injection pressure has a more significant effect on the initial motion before breakup; ambient gas density has its major impact on the motion after breakup. Hiroyasu diameter (meters), and t is time (seconds). The results of Reitz and Bracco25 indi- et al. correlated their data for spray tip penetration S(m) versus time as cate that the breakup or intact length depends on nozzle geometry details in 2Ap 1/2 addition to the diameter (see Fig. 10-25b). Note that under high injection pres- t < break : S = 0.39| sures and nozzle geometries with short length/diameter ratios, the intact or (10.25) breakup length becomes very short; breakup can occur at the nozzle exit plane. Ap 1/4 t > tbreak : S = 2.95 (d ,, t)1/2 The effect of combustion air swirl on spray penetration is shown in Fig. 10-27. Figure 10-27a shows how the spray shape and location changes as swirl is increased; Fig. 10-27b shows how spray tip penetration varies with time and swirl where rate.29 These authors related their data on spray tip penetration with swirl, S ,, 29p1 dn through a correlation factor to the equivalent penetration, S, without swirl given t break = (Pg Ap)1/2 (10.26) by Eq. (10.25): and Ap is the pressure drop across the nozzle (pascals), p, and p, are the liquid >> = ( 1 + "R. NS) - 1 and gas densities, respectively (in kilograms per cubic meter), d, is the nozzle 30v , (10.27) 532 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 533 where R, is the swirl ratio which equals the swirl rate in revolutions per minute 15 x 10-3- divided by the engine speed N (revolutions per minute), and 1; is the initial fuel jet velocity (meters per second). The curves in Fig. 10-27b correspond to Eg. (10.27). Swirl both reduces the penetration of the spray and spreads out the spray more rapidly. 10- 10.5.5 Droplet Size Distribution Frequency volume ()(20 ) . ((un) -1 Previous sections in Sec. 10.5 have discussed the overall characteristics of the diesel engine fuel spray-its spreading rate and penetration into the combustion chamber. While the distribution of fuel via the spray trajectory throughout the 5- 74.5 mm combustion chamber is important, atomization of the liquid fuel into a large 18.5 mm number of small drops is also necessary to create a large surface area across 54.5 mm which liquid fuel can evaporate. Here we review how the drop size distribution in 37.5 mm the fuel spray depends on injection parameters and the air and fuel properties. Since the measurement of droplet characteristics in an operating diesel engine is 0 40 80 120 160 200 extremely difficult, most results have come from studies of fuel injection into Droplet diameter Da, um constant-volume chambers filled with high-pressure quiescent air at room tem- FIGURE 10-28 perature. Droplet size distribution in diesel fuel spray injected through throttling pintle nozzle into quiescent During the injection period, the injection conditions such as injection pres- room-temperature air at 11 atm. Nozzle opening pressure 9.9 MPa. Pump speed 500 rev/min. Drop- sure, nozzle orifice area, and injection rate may vary. Consequently, the droplet lets are sampled well downstream of injector at given radial distances from spray axis.32 size distribution at a given location in the spray may also change with time during the injection period. In addition, since the details of the atomization which will change the droplet size distribution and mean diameter. The down- process are different in the spray core and at the spray edge, and the trajectories stream drop size in the solid-cone sprays used in diesel-injection systems is mark- of individual drops depend on their size, initial velocity, and location within the edly influenced by both drop coalescence and breakup. Eventually a balance is spray, the drop size distribution will vary with position within the spray.29 None reached as coalescence decreases (due to the expansion of the spray) and breakup of these variations has yet been adequately quantified. ceases (due to the reduced relative velocity between the drops and the entrained The aerodynamic theory of jet breakup in the atomization regime sum- gas). 31 marized in Sec. 10.5.3 (which is based on work by G. I. Taylor) leads to the Measurements of droplet size distributions within a simulated diesel spray prediction that the initial average drop diameter Da is proportional to the length indicate how size varies with location. Figure 10-28 shows the variation in drop of the most unstable surface waves:22 size distribution with radial distance from the spray axis, at a fixed axial location. Da = C 270 The drop sizes were measured with a liquid immersion technique where a sample (10.28) of drops is collected in a small cell filled with an immiscible liquid. Size distribu- tions can be expressed in terms of: where o is the liquid-fuel surface tension, p, is the gas density, v, is the relative velocity between the liquid and gas (taken as the mean injection velocity v,), C is 1. The incremental number of drops An within the size range Da - ADa/2 < a constant of order unity, and 1* is the dimensionless wavelength of the fastest Da < Da + AD 2/2 growing wave. 1* is a function of the dimensionless number (p/p.)(Re,/We,)2, 2. The incremental volume AV of drops in this size range where the jet Reynolds and Weber numbers are given by Re, = p1 v; d,/u, and Wej = puj d,/o and d, is the nozzle orifice diameter. A* goes to 3/2 as this 3. The cumulative number of drops n less than a given size Da number increases above unity. Near the edge of the spray close to the nozzle, this 4. The cumulative volume V of drops less than a given size Da equation predicts observed drop size trends with respect to injection velocity, fuel properties, nozzle L/d, and nozzle diameter, though measured mean drop sizes Since the drops are spherical: are larger by factors of 2 to 3.30 However, within the dense early region of the dn 6 dv spray, secondary atomization phenomena-coalescence and breakup-occur dDa nDi dDa (10.29) 534 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 535 100 75 75 = 50 80 -- dn = 0.3 mm dr = 0.4 mm Um = 4 ,1, 10 0.3 dn 3 0.1 60 50 -- 50 Cumulative volume V, % Pa O 0.1 MPa Sauter mean diameter DSM, um 40 × 1.1 MPa Sauter mean diameter DSM, um ₲ = 8 A 2.1 MPa 25 20 v 3.0 MPa 25 0 5.0 MPa 0 0.5 1.0 1.5 2.0 2.5 0 20 40 60 80 100 O 20 40 60 80 100 Droplet diameter Da Injection pressure, MPa Injection pressure, MPa Dm (a) (b) FIGURE 10-29 FIGURE 10-30 Normalized drop-size cumulative frequency distribution in spray injected into ambient-temperature Effect of fuel-injection pressure and nozzle geometry and size on Sauter mean drop diameter. (a) air for air pressures from 0.1 to 5 MPa. Throttling pintle nozzle with nozzle opening pressure of Effect of nozzle length/diameter ratio L,/d, and injection pressure. (b) Effect of nozzle diameter d, and 9.9 MPa. Median drop diameter D. = 1.224D `SM . 32 injection pressure. 18 An empirical expression for the Sauter mean diameter Dsm (in micrometers) for The distributions shown in the figure are frequency distributions of drop typical diesel fuel properties given by Hiroyasu and Kadota32 is volume.32 The peak in the distribution shifts to larger drop diameters as the DSM = A(Ap) -0.13500.121 /0.131 (10.32) radial position decreases: on average, the drops are smaller at the periphery of the spray. where Ap is the mean pressure drop across the nozzle in megapascals, pa is the air To characterize the spray, expressions for drop size distribution and mean density in kilograms per cubic meter, and V, is the amount of fuel delivered per diameter are desirable. An appropriate and commonly used mean diameter is the cycle per cylinder in cubic millimeters per stroke. A is a constant which equals Sauter mean diameter : 25.1 for pintle nozzles, 23.9 for hole nozzles, and 22.4 for throttling pintle nozzles. Other expressions for predicting Dsm can be found in Ref. 17. Dsm - ( Da an)/ ([ D'? an) The effects of injection pressure, nozzle geometry and size, air conditions, (10.30) and fuel properties on Sauter mean diameter in sprays obtained with diesel fuel- injection nozzles have been extensively studied. Various immersion, photo- where dn is the number of drops with diameter Da in the range Da - dDa/2 < graphic, and optical techniques for making such measurements have been used.17 Da < Da + dDa/2. The integration is usually carried out by summing over an Some of the major effects are illustrated in Figs. 10-30 and 10-31 which show appropriate number of drop size groups. The Sauter mean diameter is the diam- eter of the droplet that has the same surface/volume ratio as that of the total 150 300 v = 0.7 - 1.4 x 10-6 m2/s dr = 0.3 mm spray. y = - 50 x 10-6 m2/s Various expressions for the distribution of drop sizes in liquid sprays have 24 - 27 × 10-6 Heavy oil been proposed. One proposed by Hiroyasu and Kadota32 based on the chi- 100 61 x 10-6 200 -- G = 33 x 10-3 N/m Glycerine-alcohol-water solution square statistical distribution fits the available experimental data. Figure 10-29 Sauter mean diameter DSM, um Sauter mean diameter DSM, um 0 = 51 × 10-3 - shows how the chi-square distribution with a degree of freedom equal to 8 fits Glycerine-water solution 50- 100- 0 = 66 x 10-3 well to experimental measurements of the type shown in Fig. 10-28. Here Dm, is the median drop diameter which for this chi-square curve is 1.224Dsm. The non- in = 0.3 mm dimensional expression for drop size distribution in sprays injected through hole 0 20 40 60 80 100 0 20 40 60 80 100 nozzles, pintle nozzles, and throttling pintle nozzles given by the chi-square dis- Injection pressure, MPa Injection pressure, MPa tribution is (a) (b) FIGURE 10-31 dv - 13.5(De) exp [ -3(2) ]= (Bs) (10.31) Effect of (a) liquid viscosity v and (b) liquid surface tension o on Sauter mean drop diameter as a function of injection pressure. Air conditions: 3 MPa and ambient temperature.17 536 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 537 average Sauter mean diameters determined optically from studies of steady diese fuel sprays in a pressurized vessel. Figure 10-30 shows that nozzle size affects the mean drop size in the expected direction. Nozzle length/diameter ratio is also shown to be important: an L,/d ,, = 4 gives the minimum mean drop size at low and intermediate injection pressures. This L,/d, also corresponds to the minimum value of spray breakup length and to the maximum spray cone angle. Fuel vis- cosity and surface tension also affect mean drop size as shown in Fig. 10-31, with the effects being most significant at lower injection pressures. 10.5.6 Spray Evaporation The injected liquid fuel, atomized into small drops near the nozzle exit to form a 10 mm spray, must evaporate before it can mix with air and burn. Figure 10-20 showed the basic structure of an evaporating diesel spray under conditions typical of a large direct-injection engine. Back illumination showed that a core exists along the axis of the spray where the liquid fuel ligaments or drops are sufficiently dense to attenuate the light beam. Once the start-up phase of the injection process is over, the length of this core remains essentially constant until injection ends. This core is surrounded by a much larger vapor-containing spray region which continues to penetrate deeper into the combustion chamber: the core extends only partway to the spray tip. Additional insight into the physical struc- ture of evaporating sprays can be obtained from the schlieren photographs taken mm just after the end of injection in a prechamber engine with air swirl, shown in Fig. 10-32. The lowest magnification picture (A) shows the overall structure of the spray. The only liquid-containing region evident is that part of the core nearest 1 mm the nozzle which shows black on the left of the photograph. The spreading vapor Nozzle tip in region of the spray, carried around the chamber by the swirling air flow, appears mottled due to local turbulent vapor concentration and temperature fluctuations. The dark region within the spray vapor region is due to soot formed where the fuel vapor concentration is sufficiently high. It is probable that, after the breakup length, the dense black liquid core of the spray is composed of individual droplets Droplets with but the concentration is so high along the optical path that the light beam is fully B vapor trails extinguished. However, the last part of the core close to the nozzle tip (B) dis- Background due to perses sufficiently for individual features to be resolved. The small black dots are air turbulence liquid fuel drops in the size range 20 to 100 um. Fuel drop vapor trails can be near window observed in the highest magnification photo (C) corresponding to various stages of evaporation. These range from drops showing little surrounding vapor to FIGURE 10-32 vapor trails with little liquid remaining at the head. The vapor trails show Shadowgraph photographs at three magnifications taken just after the end of injection of diesel-fuel random orientations relative to the spray axis, presumably due to local air turbu- spray into swirling air flow in prechamber of special diesel engine. Nozzle hole diameter = 0.25 mm.21 lence. The process of droplet evaporation under normal engine operating condi- tions appears to be rapid relative to the total combustion period.21 Let us examine the drop evaporation process in more detail. Consider a 1. Deceleration of the drop due to aerodynamic drag liquid drop at close to ambient temperature injected into air at typical end-of- compression engine conditions. Three phenomena will determine the history of 2. Heat transfer to the drop from the air the drop under these conditions: 3. Mass transfer of vaporized fuel away from the drop 538 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 539 1.4 Mass mfv Ta = 1200 K ma 1200 1.2 = 1200 K 1000 1000 K Te -1.0 1000 K Diameter 800 800 K 0.8 mfy Evaporation ma Equilibrium temperature, K 600 K rate 500 800 K - 0.6 Temperature FIGURE 10-33 Schematic of variation of mass, diameter, tem- 4001 n-dodecane 0.4 Q from air perature, evaporation rate, heat-transfer rate from --- n-heptane Q to liquid air, and heat-transfer rate to liquid drop core as 200 0.2 function of time during evaporation process of 600 K 0 individual drop in diesel environment at the time 0.01 0.05 0.1 0.5 1.0 5.0 10.0 Time of injection. mg ma FIGURE 10-34 Adiabatic-saturation conditions for equilibrium mixtures formed by injecting n-dodecane and n- As the droplet temperature increases due to heat transfer, the fuel vapor pressure heptane, initially liquid at 300 K, into air at initial temperature T. between 600 and 1200 K and initial increases and the evaporation rate increases. As the mass transfer rate of vapor density 6.5 kg/m3. Equilibrium mixture temperature (T) and ratio of fuel vapor mass (my) to air mass away from the drop increases, so the fraction of the heat transferred to the drop (m.) shown as function of ratio of total fuel mass m, to ma. Fuel vapor only present to left of peaks in surface which is available to increase further the drop temperature decreases. As mry/m_ curves: liquid fuel also present to right of peaks.35 the drop velocity decreases, the convective heat-transfer coefficient between the air and the drop decreases. The combination of these factors gives the behavior lets and the air within the combustion chamber. Various phenomenological shown in Fig. 10-33 where drop mass, temperature, velocity, vaporization rate, models and computational fluid dynamic models have been developed for this and heat-transfer rate from the air are shown schematically as a function of time purpose (see Secs. 14.4.3 and 14.5.5). In the most sophisticated of these, the spray following injection.33 Analysis of individual fuel drops 25 um in diameter, is assumed to be composed of discrete computational particles each of which injected into air at typical diesel conditions, indicates that evaporation times are represents a group of droplets of similar size, temperature, etc. The distribution usually less than 1 ms.34 functions in droplet size, velocity, temperature, etc ., produced by the fuel injector Such an analysis is relevant to drops that are widely separated (e.g ., at the are statistically sampled and the resulting discrete particles are followed along edge of the spray). In the spray core, where drop number densities are high, the lagrangian trajectories as they interact and exchange mass, momentum, and evaporation process has a significant effect on the temperature and fuel-vapor energy with the surrounding gas. Drops interact directly with each other via concentration in the air within the spray. As fuel vaporizes, the local air tem- collisions and indirectly via evaporation by modifying the ambient vapor concen- perature will decrease and the local fuel vapor pressure will increase. Eventually tration and gas temperature. Studies with such models indicate that, under thermodynamic equilibrium would pertain: this is usually called adiabatic satura- normal diesel engine conditions, 70 to 95 percent of the injected fuel is in the tion.33 Calculated thermodynamic equilibrium temperatures for diesel spray con- vapor phase at the start of combustion. Evaporation is more than 90 percent ditions are plotted in Fig. 10-34 as a function of the fuel/air mass ratio for complete after 1 ms. However, only 10 to 35 percent of the vaporized fuel has n-dodecane and n-heptane. The initial liquid fuel temperature was 300 K. The mixed to within flammability limits in a typical medium-speed DI diesel engine. ratio of fuel vapor to air mass at these equilibrium conditions is also shown. To Thus combustion is largely mixing-limited, rather than evaporation-limited.36 Of the left of the peaks in the mrv/ma curves, only fuel vapor is present. To the right course, under cold-starting conditions, evaporation becomes a major constraint. of these peaks, liquid fuel is also present because the vapor phase is saturated.35 Liquid fuel vaporization causes substantial reductions in gas temperature. While this equilibrium situation may not be reached within the spray, these results are 10.6 IGNITION DELAY useful for understanding the temperature distribution within an evaporating 10.6.1 Definition and Discussion spray. To quantify accurately the fuel vaporization rate within a diesel fuel spray The ignition delay in a diesel engine was defined as the time (or crank angle) requires the solution of the coupled conservation equations for the liquid drop- interval between the start of injection and the start of combustion. The start of 540 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 541 injection is usually taken as the time when the injector needle lifts off its seat perature and pressure, as well as the physical processes described above which (determined by a needle-lift indicator). The start of combustion is more difficult govern the distribution of fuel throughout the air charge. to determine precisely. It is best identified from the change in slope of the heat- Since the ignition characteristics of the fuel affect the ignition delay, this release rate, determined from cylinder pressure data using the techniques property of a fuel is very important in determining diesel engine operating char- described in Sec. 10.4, which occurs at ignition. Depending on the character of acteristics such as fuel conversion efficiency, smoothness of operation, misfire, the combustion process, the pressure data alone may indicate when pressure smoke emissions, noise, and ease of starting. The ignition quality of a fuel is change due to combustion first occurs; in DI engines under normal conditions defined by its cetane number. Cetane number is determined by comparing the ignition is well defined, but in IDI engines the ignition point is harder to identify. ignition delay of the fuel with that of primary reference fuel mixtures in a stan- Flame luminosity detectors are also used to determine the first appearance of the dardized engine test (see below). For low cetane fuels with too long an ignition flame. Experience has shown that under normal conditions, the point of appear- delay, most of the fuel is injected before ignition occurs, which results in very ance of the flame is later than the point of pressure rise and results in greater rapid burning rates once combustion starts with high rates of pressure rise and uncertainty or error in determining the ignition point. high peak pressures. Under extreme conditions, when autoignition of most of the Both physical and chemical processes must take place before a significant injected fuel occurs, this produces an audible knocking sound, often referred to as fraction of the chemical energy of the injected liquid fuel is released. The physical "diesel knock." For fuels with very low cetane numbers, with an exceptionally processes are: the atomization of the liquid fuel jet; the vaporization of the fuel long delay, ignition may occur sufficiently late in the expansion process for the droplets; the mixing of fuel vapor with air. The chemical processes are the pre- burning process to be quenched, resulting in incomplete combustion, reduced combustion reactions of the fuel, air, residual gas mixture which lead to autoigni- power output, and poor fuel conversion efficiency. For higher cetane number tion. These processes are affected by engine design and operating variables, and fuels, with shorter ignition delays, ignition occurs before most of the fuel is fuel characteristics, as follows. injected. The rates of heat release and pressure rise are then controlled primarily Good atomization requires high fuel-injection pressure, small injector hole by the rate of injection and fuel-air mixing, and smoother engine operation diameter, optimum fuel viscosity, and high cylinder air pressure at the time of results. injection (see Sec. 10.5.3). The rate of vaporization of the fuel droplets depends on the size of the droplets, their distribution, and their velocity, the pressure and temperature inside the chamber, and the volatility of the fuel. The rate of fuel-air 10.6.2 Fuel Ignition Quality mixing is controlled largely by injector and combustion chamber design. Some The ignition quality of a diesel fuel is defined by its cetane number. The method combustion chamber and piston head shapes are designed to amplify swirl and used to determine the ignition quality in terms of cetane number is analogous to create turbulence in the air charge during compression. Some engine designs use that used for determining the antiknock quality of gasoline in terms of octane a prechamber or swirl chamber to create the vigorous air motion necessary for number. The cetane number scale is defined by blends of two pure hydrocarbon rapid fuel-air mixing (see Sec. 10.2). Also, injector design features such as the reference fuels. Cetane (n-hexadecane, C16H34), a hydrocarbon with high ignition number and spatial arrangement of the injector holes determine the fuel spray quality, represents the top of the scale with a cetane number of 100. An isocetane, pattern. The details of each nozzle hole affect the spray cone angle. The penetra- heptamethylnonane (HMN), which has a very low ignition quality, represents the tion of the spray depends on the size of the fuel droplets, the injection pressure, bottom of the scale with a cetane number of 15.+ Thus, cetane number (CN) is the air density, and the air-flow characteristics. The arrangement of the sprays, given by the spray cone angle, the extent of spray penetration, and the air flow all affect the rate of air entrainment into the spray. These physical aspects of fuel-injection CN = percent n-cetane + 0.15 x percent HMN (10.33) and fuel-spray behavior are reviewed in Sec. 10.5. The engine used in cetane number determination is a standardized single- The chemical component of the ignition delay is controlled by the precom- cylinder, variable compression ratio engine with special loading and accessory bustion reactions of the fuel. A fundamental discussion of autoignition or sponta- equipment and instrumentation. The engine, the operating conditions, and the neous hydrocarbon oxidation in premixed fuel-air mixtures is given in Sec. 9.6.2. test procedure are specified by ASTM Method D613.37 The operating require- Since the diesel engine combustion process is heterogeneous, its spontaneous ments include: engine speed-900 rev/min; coolant temperature-100ºC; intake ignition process is even more complex. Though ignition occurs in vapor phase air temperature 65.6ºC (150ºF); injection timing-13º BTC; injection regions, oxidation reactions can proceed in the liquid phase as well between the fuel molecules and the oxygen dissolved in the fuel droplets. Also, cracking of large hydrocarbon molecules to smaller molecules is occurring. These chemical t In the original procedure a-methylnapthalene (C11H10) with a cetane number of zero represented processes depend on the composition of the fuel and the cylinder charge tem- the bottom of the scale. Heptamethylnonane, a more stable compound, has replaced it. 542 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 543 pressure-10.3 MPa (1500 1b/in2). With the engine operating under these condi. 1000 105 tions, on the fuel whose cetane number is to be determined, the compression ratio T T is varied until combustion starts at TC: i.e ., an ignition delay period of 13º 300 p = 1 atm O $ = 0.3 (2.4 ms at 900 rev/min) is produced. The above procedure is then repeated using ¢ = 0.5 reference fuel blends. Each time a reference fuel is tried, the compression ratio is 100 6 atm ¢ = 0.7 ¢ = 1.0 adjusted to give the same 13º ignition delay. When the compression ratio - 11 atm 104 required by the actual fuel is bracketed by the values required by two reference 30 21 atm blends differing by less than five cetane numbers, the cetane number of the fuel is Ignition delay Tid, ms determined by interpolation between the compression ratios required by the two -31 atm 10 Tidp2, ms (atm)2 reference blends. Because of the expense of the cetane number test, many correlations which 3 .03L predict ignition quality based on the physical properties of diesel fuels have been developed.38. 39 A calculated cetane index (CCI) is often used to estimate ignition Light oil quality of diesel fuels (ASTM D97640). It is based on API gravity and the mid- boiling point (temperature 50 percent evaporated). It is applicable to straight-run No. 2 diesel 0.3 WILL 102 fuels, catalytically cracked stocks, and blends of the two. Its use is suitable for 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 x 10-3 0.4 0. 1.2 1.6 2.0 x 10-3 most diesel fuels and gives numbers that correspond quite closely to cetane 1 , K-1 Temperature number. A diesel index is also used. It is based on the fact that ignition quality is Temperature , K-1 (a) linked to hydrocarbon composition: n-paraffins have high ignition quality, and (b) FIGURE 10-35 aromatic and napthenic compounds have low ignition quality. The aniline point (a) Ignition delay as function of reciprocal air temperature for light oil spray injected into constant- (ASTM D61141-the lowest temperature at which equal volumes of the fuel and volume combustion bomb. Injection pressure 9.8 MPa (100 atm). Air pressures indicated.42 (b) Igni- aniline become just miscible) is used, together with the API gravity, to give the tion delay x (pressure)2 measured in steady-flow reactor for No. 2 diesel fuel as function of reciprocal diesel index: temperature. Fuel/air equivalence ratio o varied from 0.3 to 1.0.43 Diesel index = aniline point (ºF) x 24 API gravityt 100 (10.34) tion into constant-temperature and pressure environments have shown that the The diesel index depends on the fact that aromatic hydrocarbons mix completely temperature and pressure of the air are the most important variables for a given with aniline at comparatively low temperatures, whereas paraffins require con- fuel composition. Ignition delay data from these experiments have usually been siderably higher temperatures before they are completely miscible. Similarly, a correlated by equations of the form: high API gravity denotes low specific gravity and high paraffinicity, and, again, good ignition quality. The diesel index usually gives values slightly above the Tid = Ap" exp EA RT (10.35) cetane number. It provides a reasonable indication of ignition quality in many (but not all) cases. where Tid is the ignition delay (the time between the start of injection and the start of detectable heat release), E, is an apparent activation energy for the fuel autoignition process, R is the universal gas constant, and A and n are constants 10.6.3 Autoignition Fundamentals dependent on the fuel (and, to some extent, the injection and air-flow Basic studies in constant-volume bombs, in steady-flow reactors, and in rapid- characteristics). compression machines have been used to study the autoignition characteristics of Figure 10-35a shows ignition delay data obtained by injecting liquid fuel fuel-air mixtures under controlled conditions. In some of these studies the fuel sprays into a high-pressure heated constant-volume bomb.42 Figure 10-35b and air were premixed; in some, fuel injection was used. Studies with fuel injec- shows ignition delay data from a steady-flow high-pressure reactor where vapor- ized fuel was mixed rapidly with the heated air stream.43 The match between the form of Eq. (10.35) and the data is clear. Figure 10-35b also shows an equivalence ratio dependence of the ignition delay. Representative values for A, n, and E, for API gravity is based on specific gravity and is calculated from: API gravity, deg = (141.5/specific Eq. (10.35), taken from these and other studies, are given in Table 10.3. Ignition gravity at 60ºF) - 131.5. delay times calculated with these formulas for various diesel engines are given in 544 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 545 TABLE 10.3 TABLE 10.4 Constants for Arrhenius equation for ignition delay:44 Calculated ignition delay times44 Tid(ms) = Ap"" exp [E,/(RT)] Conditions Conditions Parameters Spadaccini and Investigator Apparatus Fuel p, atm T, K A E JR, K TeVelde43 Speed, P, Stringer Hiroyasu Spadaccini Steady No. 2 diesel 10-30 650-900 2 2.43 x 10 - 9 20,926 Engine rev/min atm No. 1 No. 2 et al.45 Wolfer4€ et al.29 and TeVelde43 flow No. 1 IDI Diesel Spadaccini Steady No. 2 diesel 10-30 650-900 1 4.00 x 10-10 20,08 1. Low swirl 600 45.6 690 17.3 38.2 6.26 3.94 9.60 and TeVelde43 flow 1200 49.3 747 1.47 3.83 3.22 2.15 3.90 1800 No. 2 52.5 758 0.86 2.44 2.76 1.82 3.13 Stringer Steady Diesel 30-60 770-980 0.757 0.0405 5,473 2. High swirl 600 45.2 674 36.3 76.9 7.60 4.68 12.5 et al.45 flow 45-50 cetane 1200 48.4 721 4.18 10.3 4.25 2.75 5.67 number 1800 51.8 744 1.47 4.07 3.19 2.08 3.82 Wolfer46 Constant- Fuel with 8-4 590-782 1.19 0.44 4,650 DI Diesel volume cetane number 1. Low compression 42.8 781 0.57 1.37 2.60 1.92 bomb >50 2.39 ratio Hiroyasu Constant- Kerosene 1-30 673-973 1.23 0.027€ 7,280 2. High compression 1500 58.8 975 0.0015 0.0060 0.508 et al.29 volume 0.407 0.322 ratio bomb Table 10.4. Air pressures and temperatures at TC piston position were estimated affect the ignition delay. While the work of Spadaccini and TeVelde probably describes the chemical ignition delay more accurately, since great care was taken from measured cylinder pressure data. Measured ignition delay times in these to obtain a uniform mixture and flow, the experiments in constant-volume bombs types of engines are: 0.6 to 3 ms for low-compression-ratio DI diesel engines over a wide range of operating conditions; 0.4 to 1 ms for high-compression-ratio and with diesel-type fuel injectors are more relevant to the diesel engine compression- ignition combustion process because they include the appropriate physical and turbocharged DI diesel engines; 0.6 to 1.5 ms for IDI diesel engines.44 The variation in the calculated delay times can be attributed to several chemical processes. The available engine ignition delay data suggest that for delays less than about 1 ms, the rate of decrease in delay with increasing tem- factors: perature becomes much less than that indicated by the data in Fig. 10-35. This is 1. In some cases the correlations are being extrapolated outside their original due to the increasing relative importance of physical processes relative to chemi- cal processes during the delay period. Thus relations of the form of Eq. (10.35) range of operating conditions. should be used with caution. 2. The methods used to detect the start of combustion, and hence the duration of In general, tia is a function of mixture temperature, pressure, equivalence the delay, are not identical. ratio, and fuel properties (though no accepted form for the variation with equiva- 3. The experimental apparatus and the method of fuel-air mixture preparation lence ratio is yet established). In the above referenced studies, the fuel was are different. injected into a uniform air environment where the pressure and temperature only changed due to the cooling effect of the fuel-vaporization and fuel-heating pro- The third factor is probably the most significant. As has been explained, the cesses. In an engine, pressure and temperature change during the delay period phenomenon of autoignition of a fuel spray consists of sequences of physical and due to the compression resulting from piston motion. To account for the effect of chemical processes of substantial complexity. The relative importance of each changing conditions on the delay the following empirical integral relation is usu- process depends on the ambient conditions, on fuel properties, and on how the ally used: fuel-air mixture is prepared. For example, fuel evaporation times are significant in cold engines, but not under fully warmed-up conditions. Thus, an equation of dt = 1 (10.36) the simple form of Eq. (10.35) can only fit the data over a limited range of condi- Jist tions. The correlations of Spadaccini and TeVelde43 have much higher activation where to; is the time of start of injection, tid is the ignition delay period, and t is energies. Normally, lower values of E /R imply that physical processes such as the ignition delay at the conditions pertaining at time t. Whether the variation in vaporization and mixing are important and relevant to chemical processes. Thus, conditions is significant depends on the amount of injection advance before TC fuel preparation, mixture inhomogeneity, heat loss, and nonuniform flow patterns that is used and the length of the delay. 546 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 547 10.6.4 Physical Factors Affecting Delay 3.0 The physical factors that affect the development of the fuel spray and the air O charge state (its pressure, temperature, and velocity) will influence the ignition delay. These quantities depend on the design of the fuel-injection system and 2.5 combustion chamber, and the engine operating conditions. The injection system variables affecting the fuel-spray development are injection timing, quantity, velocity, rate, drop size, and spray form or type. The relevant charge conditions o CN = 30 O depend on the combustion system employed, the details of the combustion 2.0 CN = 39 Ignition delay, ms o chamber design, inlet air pressure and temperature, compression ratio, injection timing, the residual gas conditions, coolant and oil temperature, and engine o O speed. Data on these interactions are available for different types of diesel O engines. The trends observed with the different diesel combustion systems are 1.5 * * generally similar, though some of the details are different. In this section the ignition delay trends during normal (fully warmed-up) engine operation are con- * CN = 52 sidered. The dependence of the ignition delay on engine design and operating 1.0 variables during engine starting and warm-up is also very important, and may be FIGURE 10-36 different from fully warmed-up behavior due to lower air temperature and pres- Ignition delays measured in a small four-stroke cycle 0 100 200 300 400 500 600 DI diesel engine with r == 16.5 as a function of load at sure. bmep, kPa 1980 rev/min. Fuel cetane numbers 30, 39, and 52.48 INJECTION TIMING. At normal engine conditions (low to medium speed, fully warmed engine) the minimum delay occurs with the start of injection at about 10 Experiments by Lyn and Valdmanis47 have shown that none of these factors has to 15º BTC.47 The increase in the delay with earlier or later injection timing a significant effect on the delay. At normal operating conditions, increasing injec- occurs because the air temperature and pressure change significantly close to TC. tion pressure produces only modest decreases in the delay. Doubling the nozzle If injection starts earlier, the initial air temperature and pressure are lower so the hole diameter at constant injection pressure to increase the fuel flow rate (by a delay will increase. If injection starts later (closer to TC) the temperature and factor of about 4) and increase the drop size (by about 30 percent) had no signifi- pressure are initially slightly higher but then decrease as the delay proceeds. The cant effect on the delay. Studies of different nozzle hole geometries showed that most favorable conditions for ignition lie in between. the length/diameter ratio of the nozzle was not significant; nor did changes in nozzle type (multihole, pintle, pintaux) cause any substantial variation in delay at INJECTION QUANTITY OR LOAD. Figure 10-36 shows the effect of increasing normal engine conditions. injection quantity or engine load on ignition delay. The delay decreases approx- imately linearly with increasing load for this DI engine. As load is increased, the INTAKE AIR TEMPERATURE AND PRESSURE. Figure 10-35 showed values of residual gas temperature and the wall temperature increase. This results in higher ignition delay for diesel fuels plotted against the reciprocal of charge temperature charge temperature (and also, to a lesser extent, charge pressure) at injection, thus for several charge pressures at the time of injection. The intake air temperature shortening the ignition delay. When adjustment is made for this increasing tem- and pressure will affect the delay via their effect on charge conditions during the perature, it is found that increasing the quantity of fuel injected has no significant delay period. Figure 10-37 shows the effects of inlet air pressure and temperature effect on the delay period under normal operating conditions. Under engine start- as a function of engine load. The fundamental ignition data available show a ing conditions, however, the delay increases due to the larger drop in mixture strong dependence of ignition delay on charge temperatures below about 1000 K temperature associated with evaporating and heating the increased amount of at the time of injection. Above about 1000 K, the data suggest that the charge fuel.47 This latter result should be expected since it is the first part of the injected temperature is no longer so significant. Through this temperature range there is fuel which ignites first; subsequent injected fuel (above the minimum required to an effect of pressure at the time of injection on delay: the higher the pressure the maintain the fuel-air mixture within the flammability limits during the delay) shorter the delay, with the effect decreasing as charge temperatures increase and does not influence the delay. delay decreases. Since air temperature and pressure during the delay period are such important variables, other engine variables that affect the relation between DROP SIZE, INJECTION VELOCITY, AND RATE. These quantities are deter- the inlet air state and the charge state at the time of injection will influence the mined by injection pressure, injector nozzle hole size, nozzle type, and geometry. delay. Thus, an increase in the compression ratio will decrease the ignition delay, 548 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 549 300 300 With wall CN = 50 --- Without wall With wall Without wall Ti 2 100 L = 100 mm 100 Ta = 440ºC P; (gauge) o 25ºC Tw = Tu L = 100 mm o Naturally · 66ºC 50 50 aspirated 200ºC 2 103 kPa CN = 34 30- 30 280ºC Ignition delay, ms Ignition delay, ms 2 Ignition delay, ms Ignition delay, ms 340ºC 10- 470ºC - 520ºC 10 TW = 570ºC CN = 50 Ta = 600ºC 3- w 530ºC 0 400 800 1200 1600 400 800 1200 1600 440ºC bmep, kPa bmep, kPa TT T (a) (b) 0.1 0.3 0.5 3 5 10 0. 0.3 0.5 1 3 5 10 FIGURE 10-37 Ambient pressure, MPa Ambient pressure, MPa (a) (b) Effect of inlet air pressure and temperature on ignition delay over load range of small DI diesel at 1980 rev/min. (a) Engine naturally aspirated and with 1 atm boost; inlet air temperature T; = 25ºC; FIGURE 10-38 50 cetane number fuel. (b) Engine naturally aspirated; T; = 25 and 66ºC; 34 and 50 cetane number Effect of spray impingement on wall 100 mm from nozzle on ignition delay from combustion bomb fuel. 48 studies. (a) Effect of air temperature as a function of air pressure; Tw = T.t. (b) Effect of wall tem- perature at 440ºC air temperature.29 and injection timing will affect the delay (as was discussed earlier), largely due to pressures and temperatures studied, but has no significant effect at the high pres- the changes in charge temperature and pressure at the time of injection. sures and temperatures more typical of normal diesel operation. Engine experi- ments where the delay was measured while the jet impingement process was ENGINE SPEED. Increases in engine speed at constant load result in a slight varied showed analogous trends. The jet impingement angle (the angle between decrease in ignition delay when measured in milliseconds; in terms of crank angle the fuel jet axis and the wall) was varied from almost zero (jet and wall close to degrees, the delay increases almost linearly.48 A change in engine speed changes parallel) to perpendicular. The delay showed a tendency to become longer as the the temperature/time and pressure/time relationships. Also, as speed increases, impingement angle decreased. The most important result is not so much the injection pressure increases. The peak compression temperature increases with modest change in delay but the difference in the initial rate of burning that results increasing speed due to smaller heat loss during the compression stroke.47 from the differences in fuel evaporation and fuel-air mixing rates. COMBUSTION CHAMBER WALL EFFECTS. The impingement of the spray on SWIRL RATE. Changes in swirl rate change the fuel evaporation and fuel-air the combustion chamber wall obviously affects the fuel evaporation and mixing mixing processes. They also affect wall heat transfer during compression and, processes. Impingement of the fuel jet on the wall occurs, to some extent, in hence, the charge temperature at injection. Only limited engine studies of the almost all of the smaller, higher speed engines. With the "M" system, this effect of swirl rate variations on ignition delay have been made. At normal oper- impingement is desired to obtain a smooth pressure rise. The ignition delay with ating engine speeds, the effect of swirl rate changes on the delay are small. Under the "M" system is longer than in conventional DI engine designs.47 Engine and engine starting conditions (low engine speeds and compression temperatures) the combustion bomb experiments have been carried out to examine the effect of effect is much more important,47 due presumably to the higher rates of evapo- wall impingement on the ignition delay. Figure 10-38 shows the effect of jet wall ration and mixing obtained with swirl. impingement on ignition delay measured in a constant-volume combustion bomb, for a range of air pressures and temperatures, and wall temperatures.29 OXYGEN CONCENTRATION. The oxygen concentration in the charge into The wall was perpendicular to the spray and was placed 100 mm from the nozzle which the fuel is injected would be expected to influence the delay. The oxygen tip. The data shows that the presence of the wall reduces the delay at the lower concentration is changed, for example, when exhaust gas is recycled to the intake 550 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 551 25 35 72 . Base fuel: 25% n-cetane + 75% i-octane 20- A n-paraffins 30 Isoalkanes 5-20% by volume O Cycloalkanes in base fuel -1.5 × Aromatics 15- Ignition delay, ms Ignition delay, deg 25 1 10 Ignition delay, deg o p varied, foz constant FIGURE 10-39 . fo2 varied, p constant -0.5 Effect of oxygen density in gas on ignition delay in 20 n-pentane single-cylinder DI engine of 1.3-dm3 displacement with r = 15 at 1800 rev/min. Oxygen density changed by recycling exhaust gas at constant inlet 2,7-dimethyloctane T 0 Base fuel 0 0.5 1 1.5 density and by varying inlet pressure from 0.5 to 15 n-hexyl benzene 50% 3,4-dimethyldecane Relative oxygen density 3 atm with air.49 AA 50% 3,3-diethyloctane 10, 30 35 40 50 55 for the control of oxides of nitrogen emissions (see Chap. 11). Results of a study carried out in a single-cylinder DI engine operated at a constant air/fuel ratio Observed cetane number (30 : 1), manifold temperature, injection timing, and speed (1800 rev/min), where FIGURE 10-40 the oxygen concentration was varied by recirculating known amounts of cooled Effect of type of hydrocarbon structure on ignition quality of fuels in DI diesel combustion process at constant compression ratio and engine operating conditions.5º exhaust, are shown in Fig. 10-39.49 Oxygen density is normalized by the natu- rally aspirated no-recirculation test value. As oxygen concentration is decreased, ignition delay increases. ment increases. Isoalkanes, depending on the degree of branching, degrade igni- tion quality (unless the branching is concentrated at one end of the molecule, when these types of isoalkanes improve ignition quality). Cycloalkanes and aro- 10.6.5 Effect of Fuel Properties matics generally reduce the cetane number, unless they have a long n-alkane Since both physical and chemical processes take place during the ignition delay, chain attached to the ring. The cetane number of a fuel (a measure of its ability to the effects of changes in the physical and chemical properties of fuels on the delay autoignite) generally varies inversely with its octane number (a measure of its period have been studied. The chemical characteristics of the fuel are much the ability to resist autoignition; see Fig. 9-69 for the effect of hydrocarbon structure more important. The ignition quality of the fuel, defined by its cetane number, on knock). The cetane number of commercial diesel fuel is normally in the range will obviously affect the delay. The dependence of cetane number on fuel molecu- 40 to 55. lar structure is as follows. Straight-chain paraffinic compounds (normal alkanes) Engine ignition delay data with diesel fuels of different cetane number, at have the highest ignition quality, which improves as the chain length increases. various constant loads and speeds, shown in Fig. 10-41, demonstrate similar Aromatic compounds have poor ignition quality as do the alcohols (hence, the trends. Within the normal diesel fuel cetane number range of 40 to 55, an approx- difficulties associated with using methanol and ethanol, possible alternative fuels, imately linear variation is evident. However, decreasing fuel cetane number below in compression-ignition engines). Figure 10-40 illustrates these effects. A base fuel about 38 may result in a more rapid increase in ignition delay. was blended with pure paraffinic (normal, iso -, and cycloalkanes), aromatic, and Cetane number is controlled by the source of crude oil, by the refining olefinic hydrocarbons of various carbon numbers, by up to 20 percent by volume. process, and by additives or ignition accelerators. Just as it is possible to reduce The base fuel, a blend of 25 percent n-hexadecane and 75 percent isooctane, had the tendency to knock or autoignite in spark-ignition engine fuels by adding a cetane number of 38.3. The figure shows that the resulting ignition delays corre- antiknock agents, so there are additives that improve the ignition quality of late well as a function of cetane number at constant compression ratio and engine compression-ignition engine fuels. Generally, substances that increase the ten- operating conditions. Addition of normal alkanes (excluding n-pentane and lower dency to knock enhance ignition, and vice versa. Ignition-accelerating additives carbon number alkanes) improve the ignition quality. As the chain length of the include organic peroxides, nitrates, nitrites, and various sulfur compounds. The added paraffin gets longer (higher carbon number) the cetane number improve- most important of these commercially are the alkyl nitrates (isopropyl nitrate, 552 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 553 10, 2.0 75ºF inlet air o Engine 1 - 15ºF inlet air Engine 2 Engine 3 1.5 1000 rev/min, 1/4 load 2000 rev/min, 81 X- 1/2 load -- 3600 rev/min, full load DDDD ----- 1.0 DND Predicted delay, ms Ignition delay, deg 2000 rev/min, 1/2 load ====== 2000 rev/min, full load 0.5 2000 rev/min, full load FIGURE 10-42 T Comparison of engine ignition delays 0.5 1000 rev/min, 1/4 load 1.0 1.5 predicted with Eq. (10.37) with corre- 600 rev/min, idle Measured delay, ms sponding measured values. 54 3 35 40 45 50 55 Cetane number FIGURE 10-41 Effect of fuel cetane number on ignition delay over the load and speed range of 6.2-dm3 eight-cylinder 10.6.6 Correlations for Ignition Delay in Engines IDI swirl-chamber diesel engine. Top curve indicates typical fit between data and least-squares Many correlations have been proposed for predicting ignition delay as a function straight line over this cetane number range.51 of engine and air charge variables. These usually have the form of Eq. (10.35) and have been based on data from more fundamental experiments in combustion primary amyl nitrates, primary hexyl nitrates, octyl nitrate). 52 Typically, about bombs and flow reactors. An important factor in assessing the appropriateness of 0.5 percent of these additives by volume in a distillate fuel gives about a 10 cetane any correlation is how it is to be used to predict the magnitude of the delay. If an number increase in a fuel's ignition quality, though their effectiveness may equation for predicting the complete delay process (including all the physical and depend on the composition of the base fuel. The incremental effect of increasing chemical processes from injection to combustion) is required, then the data show amounts of ignition-accelerating additives on cetane number decreases.48 that such a simple form for the equation is unlikely to be adequate for the full Usually, the ignition delay obtained with cetane improved blends are found to be range of engine conditions (see Table 10.4). However, if a model for the autoigni- equivalent to those obtained with natural diesel fuels of the same cetane number. tion process of a premixed fuel-air mixture during the delay period is required, Two potential practical uses for ignition accelerators are in upgrading the igni- for use in conjunction with models for the physical processes of fuel evaporation tion characteristics of poorer quality diesel fuel and (in much larger amounts) and fuel-air mixing, then correlations such as those listed in Table 10.3 may be making possible the use of alcohols in compression-ignition engines. sufficiently accurate. The physical characteristics of diesel fuel do not significantly affect the igni- An empirical formula, developed by Hardenberg and Hase53 for predicting tion delay in fully or partially warmed-up engines. Tests with fuels of different the duration of the ignition delay period in DI engines, has been shown to give front-end volatility (over the cetane number range 38 to 53), and with substan- good agreement with experimental data over a wide range of engine conditions tially different front-end ignition quality for the same average cetane number, (see Fig. 10-42).54 This formula gives the ignition delay (in crank angle degrees) in showed no discernible differences. Fuel viscosity variations over a factor of 2.5 terms of charge temperature T (kelvins) and pressure p (bars) during the delay were also tested and showed no significant effect.48 Thus, in a warmed-up engine, (taken as TC conditions) as variations in fuel atomization, spray penetration, and vaporization rate over rea- sonable ranges do not appear to influence the duration of the delay period signifi- cantly (see also Sec. 10.5.6 on fuel vaporization). Tia(CA) = (0.36 + 0.2252) exp| Ex RT - 17,190)(P - 12.4) 21 .2 0.637 000 ] (10.37) 554 COMBUSTION IN COMPRESSION-IGNITION ENGINES 555 INTERNAL COMBUSTION ENGINE FUNDAMENTALS T .. 2 10.7 MIXING-CONTROLLED COMBUSTION n n 10.7.1 Background 1.2 1.11 Earlier sections of this chapter have developed our current understanding of the 1.0 individual processes which together make up the total injection-mixing-burning 900 sequence-atomization, vaporization, fuel spray development, air entrainment, 20 Cold starting ignition, and combustion. While the phenomenological model developed by Lynº 800 500- T; = 273 K 16 provides satisfactory logical links between these processes, quantitative links are Te = 20 ,16ª still lacking. Especially difficult to quantify are the relations between fuel spray 700 Tc = 12 behavior, flame structure, and fuel burning rate-the area of focus of this section. 12 600 TTC, K TTC, K 400 - The color photographs of the compression-ignition combustion process in differ- ent types of diesel engines in Figs. 10-4 and 10-5 (see color plate between 498 and 500 Warm engine 499) show how the flame immediately following ignition spreads rapidly and T; = 300 K L = 130 mm envelops the spray. Depending on the spray configuration, the visible flame may 400 300 then fill almost the entire combustion chamber. The flame and spray geometries are closely related. Mixing processes are also critical during the ignition delay 300 0 500 1000 1500 2000 2500 0 50 100 150 200 period: while the duration of the delay period is not influenced in a major way by Engine speed, rev/min the rates of spray processes which together control "mixing," the amount of fuel mixed with air to within combustible limits during the delay (which affects the FIGURE 10-43 Exponent n for polytropic model of compression process in Eq. (10.39) and corresponding end-of- rate of pressure rise once ignition has occurred) obviously is directly related to compression air temperature at TC. Warm and cold DI diesel engine with 130 mm stroke.53 mixing rates. Thus substantial observational evidence supports the mixing- controlled character of diesel engine combustion. However, while it is well accepted that diesel combustion is normally con- trolled by the fuel-air mixing rate, fundamentally based quantitatively accurate where S, is the mean piston speed (meters per second) and R is the universal gas models for this coupled mixing and combustion process are not yet available. constant (8.3143 J/mol . K). E, (joules per mole) is the apparent activation The difficulties are twofold. First, the spray geometry in real diesel combustion energy, and is given by systems is extremely complex. Second, the phenomena which must be described (and especially the unsteady turbulent diffusion diesel engine flame) are inade- 618,840 EA = - CN + 25 (10.38) quately understood. Current capabilities for modeling these phenomena are reviewed in Chap. 14. Thermodynamic-based models of the diesel combustion process with atomization, vaporization, and spray development described by where CN is the fuel cetane number. The apparent activation energy decreases empirical or turbulent-jet-based submodels have been developed and used to with increasing fuel cetane number. The delay in milliseconds is given by predict burning rates. These are described in Secs. 14.4.3 and 14.4.4, and show reasonable but not precise agreement with data. Fluid-mechanic-based models of Tid(ms) == _ Tia(CA) air flow, fuel spray behavior, and combustion are under active development (see 0.006N Sec. 14.5). While realistic air-flow predictions are now feasible, predictions of spray behavior are less well developed and combustion-rate predictions are still where N, engine speed, is in revolutions per minute. Values for T and p can be exploratory. estimated using a polytropic model for the compression process: In the following sections, the evidence linking spray characteristics to flame TTC = Tir?-1 PIC = Pire (10.39a,b) structure and burning rates is summarized. where n is the polytropic exponent (see Sec. 9.2.2), re is the compression ratio, and 10.7.2 Spray and Flame Structure the subscript i denotes intake manifold conditions. Values of the polytropic expo- nent are given in Fig. 10-43 for a direct-injection diesel under warm and cold The structure of each fuel spray is that of a narrow liquid-containing core engine operating conditions, 49,53 (densely filled with drops of order 20 um in diameter) surrounded by a much 556 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 557 12- 12 13 3ºBTC 3ºBTC A. 1.17 ms after SOI B. 0.13 ms after A C. 0.65 ms after A 12º ATC 12º ATC (a) CO (b) CO2 FIGURE 10-45 D. 2.6 ms after A E. 3.1 ms after A F. 5.6 ms after A Contours of constant CO and CO2 concentration in a plane along the spray axis calculated from gas composition data obtained with a rapid-acting sampling valve in a large 30.5-cm bore DI quiescent- chamber diesel engine with r = 12.85 operated at 500 rev/min. Injection starts 17º BTC, ignition occurs 8º BTC, injection ends 5º ATC.60 Air or fuel-air mixture Liquid fuel Luminous flame Combustion products FIGURE 10-44 Tracings of outer boundary of liquid fuel spray and flame from high-speed movies of diesel com- bustion taken in a rapid-compression machine, looking down on piston through transparent head. later-injected fuel as explained in Sec. 10.5.2) and the injector nozzle. Experiments First occurrence of luminous flame at A, 1.17 ms after start of injection. End of injection at D.57 where air/fuel ratio contours for a gaseous fuel jet injected into a swirling air flow in a rig simulated the fuel-air mixing process in open-chamber diesels,58 under conditions chosen to match a set of diesel combustion rapid-compression- larger gaseous-jet region containing fuel vapor (see Fig. 10-20). The fuel concen- machine experiments where the autoignition sites and subsequent flame develop- tration in the core is extremely high: local fuel/air equivalence ratios near the ment were recorded on movies,59 showed that autoignition occurred in a nozzle of order 10 have been measured during the injection period. Fuel concen- concentration band between the equivalence ratios of 1 and 1.5. Subsequent trations within the spray decrease with increasing radial and axial position at any flame development, along mixture contours close to stoichiometric, occurs given time, and with time at a fixed location once injection has ended.55 The fuel rapidly, as indicated in Fig. 10-44. Initially, this is thought to be due to sponta- distribution within the spray is controlled largely by turbulent-jet mixing pro- neous ignition of regions close to the first ignition site due to the temperature rise cesses. Fuel vapor concentration contours determined from interferometric associated with the strong pressure wave which emanates from each ignition site studies of unsteady vaporizing diesel-like sprays, presented by Lakshminarayan due to local rapid chemical energy release. Also, spontaneous ignition at addi- and Dent, 56 confirm this gaseous turbulent-jet-like structure of the spray, with its tional sites on the same spray, well separated from the original ignition location, central liquid-containing core which evaporates relatively quickly once fuel injec- can occur. Turbulent mixing provides another flame-spreading mechanism. From tion ends. this point flame development is rapid, and the gas expansion which occurs on The location of the autoignition sites and subsequent spreading of the burning deforms the original spray form. These processes take place in each fuel enflamed region in relation to the fuel distribution in the spray provides further spray in a closely comparable though not necessarily identical manner. Com- evidence of the mixing-controlled character of combustion. Figure 10-44 shows bustion movies such as those in Figs. 10-4 and 10-44 show that flame rapidly how this process occurs under conditions typical of direct-injection quiescent- envelops each spray once spontaneous ignition occurs. chamber diesel engines. It shows tracings of the liquid fuel spray and flame Gas-sampling data indicate that the burned gases within the flame- boundaries taken from high-speed movies of the injection and combustion pro- enveloped spray are only partially reacted and may be fuel-rich. Figure 10-45 cesses with central injection of five fuel jets into a disc-shaped chamber in a shows CO and CO, concentration contours determined from rapid-acting rapid-compression machine.57 These and other similar studies show that autoig- sample valve measurements from the combustion chamber of a large quiescent- nition first occurs toward the edge of the spray, between the spray tip (which may chamber two-stroke cycle diesel engine.60 The contour maps shown correspond by then have interacted with the combustion chamber walls, and which contains to the centerline of one of the five injected fuel sprays. Injection commenced at 558 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 559 17º BTC and ended about 5º ATC; ignition occurred at 8º BTC. The contours at 3º BTC show high CO concentrations in the burned gases which now occupy most of the spray region, indicating locally very fuel-rich conditions. Later, at 12º ATC, fuel injection has ceased, this rich core has moved outward to the piston- 3.5 . (61) Quiescent chamber, spray centerline bowl wall, and combustion within the expanded spray region is much more com- Combustion plete. This oxidation of CO, as air is entrained into the spray region, mixes, and starts at 3.0 o (62) Quiescent chamber, top of spray 5ºBTC burns, releases substantial additional chemical energy. * (61) Edge of deep bowl, spray centerline The role of air swirl in promoting much more rapid fuel-air mixing in 2.5- medium-size and smaller diesel engines is evident from similar gas-sampling "(63) Bowl, - 22º injection studies in engines with these different combustion systems. The variation of gas species and unburned hydrocarbon concentrations within critical regions of the 2.0 = (63) Bowl, - 27º injection combustion chamber have been mapped out by a number of investigators.61-63 Fuel/air equivalence ratio These data show that during the early stages of injection and combustion, the 1.5 boundaries of the individual sprays can be identified as they are convecte around the combustion chamber bowl by the swirl. The fuel distribution within 1.07 the combustion chamber is highly nonuniform. However, within each spray, suffi- cient air has mixed into the spray to bring the peak fuel/air equivalence ratios within the spray, in the outer regions of the chamber, to close to stoichiometric 0.5 values.63 This substantially different character of the spray with swirl is clear O from the data in Fig. 10-46. Figure 10-46a shows equivalence ratio values deter- 15 -10 0 10 20 30 40 50 mined from gas sampling, versus crank angle, from several studies with different Crank angle, deg designs of combustion chamber. While the local values obviously depend on the (a) relation of the sample valve location to spray position at any given crank angle, the much lower values of equivalence ratio with swirl relative to quiescent cham- bers, during injection and the early stages of combustion, clearly indicate that 0.7 swirl enhances mixing rates substantially. As combustion ends, these data indi- 0.6 cate relatively uniform fuel distribution within the combustion chamber, at least Quiescent chamber 0.5- (61) on a gross geometric scale. However, early in the combustion process the high Moles CO per mole CH1.75 0.4- CO levels, found in all these combustion systems as shown in Fig. 10-46b, indi- cate that the burned gases are only partially reacted. With quiescent chambers 0.3 Bowl (61) this is largely due to lack of oxygen. With swirl, however, substantial oxygen is 0.2Bowl (63) present. Whether the high CO with swirl is due to kinetic limitations or to 0.1 smaller-scale mixture nonhomogeneities is unclear. OL -25-20-10 0 10 20 30 40 50 60 Crank angle, deg 10.7.3 Fuel-Air Mixing and Burning Rates (b) The model of diesel combustion obtained from heat-release analyses of cylinder FIGURE 10-46 pressure data identifies two main stages of combustion (see Fig. 10-9). The first is Time and space-resolved gas-composition data obtained from rapid-acting sampling valves from the premixed-combustion phase, when the fuel which has mixed sufficiently with within the combustion chambers of quiescent and high-swirl bowl-in-piston DI diesel engines. (a) air to form an ignitable mixture during the delay period is consumed. The second Local fuel/air equivalence ratios on spray centerline and periphery with quiescent chamber, edge of deep bowl with swirl, and within a shallow bowl with swirl, three-quarters of the way out to the bowl is the mixing-controlled combustion phase, where rates of burning (at least in wall, for two injection timings (-22º and -27º). (b) CO concentration on spray centerline with naturally aspirated engines) are lower. Experimental evidence from heat-release quiescent chamber, edge of deep bowl, and within shallow bowl with swirl.61-63 analysis indicates that the majority of the fuel (usually more than 75 percent) burns during the second mixing-controlled phase. Such evidence forms the basis for the heat-release models used in diesel engine cycle simulations. For example, 560 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 561 2500 T the fraction of the fuel f which burns in the premixed phase has been correlated 90ºC by Watson et al. (see Sec. 14.4.3) by the relation Initial air temperature 2000 100ºC .10ºC acho 128ºC B =1 - (10.40) 1500 Tid Net heat release rate, KJ/s 1000 where o is the fuel/air equivalence ratio, tid the ignition delay (in milliseconds), and a ~ 0.9, b ~ 0.35, and c ~ 0.4 are constants depending on engine design. 500 149ºC Equation (10.40) shows the expected trends for the premixed fraction, with 0 changes in the overall equivalence ratio o (increasing injection duration as load is increased) and changes in the ignition delay. 500 0 That the fuel-burning or heat-release rate is predominantly mixing con- 2 4 6 10 trolled is supported by the following types of evidence. Estimates of the rate at Time after start of injection, ms which fuel-air mixture with composition within the combustible limits is produc- ed in diesel sprays under typical engine conditions, based on a variety of 3500 turbulent-jet models of the spray (e.g ., see Refs. 29, 36, and 59 and also Sec. 3000 14.4.3), show that mixing rates and burning rates are comparable in magnitude. 2500 Estimates of characteristic times for the turbulent-jet mixing processes in diesel Injection pressure combustion chambers show these to be comparable to the duration of the heat- 2000 120 MPa Net heat release rate, kI/s release process, and much longer than characteristic times for evaporation and 500 90 MPa the combustion chemical kinetics. 36, 44 1000- 60 MPa Then, measured diesel-combustion heat-release profiles show trends with 500 engine design and operating parameter changes that correspond to fuel-air mixing being the primary controlling factor. Examples of heat-release profiles measured in rapid-compression-machine studies of diesel combustion, shown in -500| 0 2 4 6 8 Fig. 10-47, illustrate this clearly. The rapid-compression machine had a disc- Time after start of injection, ms shaped chamber of 10 cm diameter with a 3.1 cm clearance height at the end of a compression process through a volume ratio of 15.4; a five-hole centrally located fuel-injector nozzle was used. Figure 10-47a shows the heat-release profiles for 2000 different initial air temperatures which produce different ignition delays. Longer Swirl 1600 delays allow more fuel to mix to within combustible limits during the delay, so 6000 rev/min 4000 rev/min the peak premixed heat-release rate increases. However, the mixing-controlled- 1200 No swirl phase heat-release-rate magnitudes are essentially the same because the spray- Net heat release rate, kJ/s mixing processes are little affected by these changes in air temperature. Figure 800 10-47b and c shows that heat-release rates throughout the combustion process are increased by increased fuel-injection rate (achieved by increasing the fuel- 400 injection pressure) and by swirl. Both these changes increase the fuel-air mixing rates within the fuel spray and therefore increase the heat-release rate during the mixing-controlled combustion phase. 0 2 4 6 8 Diesel engine heat-release rate trends, as design and operating variables are Time after start of injection, ms changed, can be related to mixing rates in analogous fashion. Table 10.5 sum- marizes the trends that have been investigated. The directional effects of changes FIGURE 10-47 in engine parameters on the ignition delay period and the fuel-air mixing rate are Net heat-release rates, as a function of time after start of injection, calculated from cylinder pressure all consistent with the measured changes in premixed and mixing-controlled data from rapid-compression-machine studies of DI diesel combustion. (a) Effect of varying initial air heat-release rates. The controlling role of fuel-air mixing in the diesel engine fuel temperature: 4000 rev/min swirl, injection pressure 60 MPa. (b) Effect of varying injection pressure: spray on combustion is clear. no swirl. (c) Effect of varying swirl : injection pressure 60 MPa.20 562 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 563 TABLE 10.5 Effects of engine design and operating variables on heat- 10.6. Estimate the following quantities for a typical direct-injection diesel fuel spray. The injection pressure is 500 atm; the cylinder pressure during injection is 50 atm. release rates (a) Assuming that the flow through the nozzle orifice is incompressible and quasi Effect on steady, estimate the liquid fuel velocity at the orifice exit. At this velocity, how long would the fuel take to reach the cylinder wall? The bore is 125 mm. Reference Parameter varied Tid (b) Each nozzle orifice diameter d ,, is 0.34 mm and L,/d ,, = 4. Determine the spray angle and plot spray tip penetration versus time. 5, 64 Injection rate 1 (c) Use Eq. (10.32) to estimate the initial average drop size assuming that the injec- 65 Turbocharger boost 1 tion process in (a) above continues for 1 millisecond and the injector nozzle has 66 Compression ratio | four orifices. 66 Number of injector holes 1 T 67, 68 Injection advance 1 10.7. Diesel fuel is injected as a liquid at room temperature into air at 50 atm and 800 K, 67, 68 Swirl 1 - T close to TC at the end of compression. If the overall equivalence ratio is 0.7, esti- 67 Intake-air temperature ! ** mate the reduction in average air temperature which would occur when the fuel is 68, 69 Injection pressure 1 T 1 fully vaporized and uniformly mixed. Assume such mixing takes place at constant -- 11, 69 Speed 1 volume prior to any combustion. Tid, ignition delay; m. = (dm/dt) -, fuel-air mixing rate; 0, = (dQ/dt) ,, heat-release 10.8. Using Eq. (10.37) estimate the ignition delay in milliseconds and crank angle rate during premixed-combustion phase; O .. = (dQ/dt) -, heat-release rate during degrees for these operating conditions in Table 10.4: low swirl IDI diesel 600 and mixing-controlled-combustion phase. 1 increase; | decrease; * minor effect. 1800 rev/min; high swirl IDI diesel 1800 rev/min; DI diesel low and high compres- Source: From Plee and Ahmad.44 sion ratio. The fuel cetane number is 45; stroke = 0.1 m. Discuss whether the pre- dicted trends with speed, swirl, and compression ratio are consistent with Sec. PROBLEMS 10.6.4. 10.9. The compression ratio of truck diesel engines must be set at about 18 so that the 10.1. Describe the sequence of processes which must occur before the liquid fuel in the engine will start when cold. Using Eqs. (10.37) to (10.39) develop a graph of tia (in injection system in a direct-injection compression-ignition engine is fully burned. degrees) as a function of compression ratio for re = 12 to 20. Assume p = 1 atm, 10.2. Small high-swirl direct-injection CI engines have fuel conversion efficiencies which T; = 255 K, n = 1.13, speed = 100 rev/min, bore = stroke = 120 mm, fuel cetane are about 10 percent higher than values typical of equivalent indirect-injection number = 45. If the ignition delay must be less than 20º CA for satisfactory start- engines. (IDI engines are used because they achieve higher bmep.) What ing, what compression ratio is required ? combustion-system-related differences contribute to this higher efficiency? 10.10. Equation (10.40) predicts the fraction 6 of the fuel injected into a direct-injection 10.3. In a diesel engine, because the fuel distribution is nonuniform the burned gas tem- diesel engine which burns in the premixed phase. Plot ß as a function of tid for perature is nonuniform. Consider small fuel-air mixture elements initially at 1000 K = 0.4. Show that for turbocharged DI diesel engines where tia is 0.4 to 1 ms, the and 6.5 MPa at top-center with a range of equivalence ratios. Each element burns premixed combustion phase is much less important than it normally is for naturally at essentially constant pressure. Calculate (using the charts in Chap. 4, or an appro- aspirated engines where Tid is between 0.7 and 3 ms. priate chemical equilibrium thermodynamic computer code) the burned gas tem- perature for mixture equivalence ratios of 0.4, 0.6, 0.8, 1.0, 1.2. Assume the fuel is REFERENCES isooctane. 1. Alcock, J. F ., and Scott, W. M.: "Some More Light on Diesel Combustion," Proc. Auto. Div ., 10.4. The levels of combustible species in the exhaust of a direct-injection diesel engine Instn Mech. Engrs, No. 5, pp. 179-191, 1962-1963. are: HC, 0.8 g/kW . h; CO, 3 g/kW . h; particulates, 0.7 g/kW . h. If the specific fuel 2. Scott, W. M.: "Understanding Diesel Combustion through the Use of High Speed Moving Pic- consumption is 210 g/kW . h calculate the combustion efficiency. tures in Color," SAE paper 690002, SAE Trans ., vol. 78, 1969. 3. Neitz, A ., and D'Alfonso, N.: "The M.A.N. Combustion System with Controlled Direct Injection 10.5. Consider the naturally aspirated direct-injection diesel engine in Fig. 1-23 operating for Passenger Car Diesel Engines," SAE paper 810479, 1979. at 2300 rev/min and an equivalence ratio of 0.7. Estimate the following: 4. Ogasawara, M ., Tokunaga, Y ., Horio, K ., Uryu, M ., and Hirofumi, N.: "Photographic Study on (a) Mass of air in each cylinder per cycle, mass, and volume (as liquid) of diesel fuel the Intermittent Spray Combustion by a Rapid Compression Machine" (in Japanese), Internal injected per cylinder per cycle. Combustion Engine, vol. 15, 1976. (b) Estimate the average drop size [e.g ., use Eq. (10.32)]. The cylinder pressure at 5. Austen, A. E. W ., and Lyn, W .- T.: "Relation between Fuel Injection and Heat Release in a time of injection (close to TC) is 50 atm; the fuel injection pressure is 500 atm. Direct-Injection Engine and the Nature of the Combustion Processes," Proc. Instn Mech. Engrs, (c) Assuming all fuel droplets are the same size as the average drop, how many No. 1, pp. 47-62, 1960-1961. drops are produced per injection? If these drops are uniformly distributed 6. Lyn, W .- T.: "Study of Burning Rate and Nature of Combustion in Diesel Engines," in Pro- throughout the air charge at TC, what is the approximate distance between ceedings of Ninth International Symposium on Combustion, pp. 1069-1082, The Combustion Insti- tute, 1962. drops? (Neither of these assumptions is correct; however, the calculations illus- 7. Cheng, W ., and Gentry, R.: "Effects on Charge Non-Uniformity on Diesel Heat Release trate the difficulty of the fuel-air mixing process.) Analysis," SAE paper 861568, 1986. 564 INTERNAL COMBUSTION ENGINE FUNDAMENTALS COMBUSTION IN COMPRESSION-IGNITION ENGINES 565 8. Gatowski, J. A ., Balles, E. N ., Chun, K. M ., Nelson, F. E ., Ekchian, J. A ., and Heywood, J. B 34. Borman, G. L ., and Johnson, J. H.: "Unsteady Vaporization Histories and Trajectories of Fuel "Heat Release Analysis of Engine Pressure Data," SAE paper 841359, SAE Trans ., vol. 93, 1984. Drops Injected into Swirling Air," in Burning a Wide Range of Fuels in Diesel Engines, SAE 9. Krieger, R. B ., and Borman, G. L.: "The Computation of Apparent Heat Release for Internal Progress in Technology, vol. 11, pp. 13-29, SAE, 1967; also SAE paper 598C, 1962. Combustion Engines," ASME paper 66-WA/DGP-4, in Proc. Diesel Gas Power, ASME, 1966. 35. Kamimoto, T ., and Matsuoka, S.: "Prediction of Spray Evaporation in Reciprocating Engines," 10. Watson, N ., and Kamel, M.: "Thermodynamic Efficiency Evaluation of an Indirect Injection SAE paper 770413, SAE Trans ., vol. 86, 1977. Diesel Engine," SAE paper 790039, SAE Trans ., vol. 88, 1979. 36. Kuo, T ., Yu, R. C ., and Shahed, S. M.: "A Numerical Study of the Transient Evaporating Spray 11. Kort, R. T ., Mansouri, S. H ., Heywood, J. B ., and Ekchian, J. A.: "Divided-Chamber Diesel Mixing Process in the Diesel Environment," SAE paper 831735, SAE Trans ., vol. 92, 1983. Engine, Part II: Experimental Validation of a Predictive Cycle-Simulation and Heat Release 37. ASTM Method D613, Cetane Test Procedure. Analysis," SAE paper 820274, SAE Trans ., vol. 91, 1982. 38. Henein, N. A ., and Fragoulis, A. N.: "Correlation between Physical Properties and Autoignition 12. Obert, E. F.: Internal Combustion Engines and Air Pollution, Intext Educational Publishers, 1973 Parameters of Alternate Fuels," SAE paper 850266, 1985. edition. 39. Gulder, O. L ., Glavincevski, B ., and Burton, G. F.: "Ignition Quality Rating Methods for Diesel 13. Weathers, Jr ., T ., and Hunter, C.: Diesel Engines for Automobiles and Small Trucks, Reston Publi- Fuels-A Critical Appraisal," SAE paper 852080, 1985. shing Company, a Prentice-Hall Company, Reston, Va. 1981. 40. ASTM D976, Calculated Cetane Index. 14. Bosch: Automotive Handbook, Ist English ed ., Robert Bosch GmbH, 1976. 41. ASTM D611. 15. Williams, Jr ., H. A.: "The GM/EMD Model 710G Series Engine," in Marine Engine Development, 42. Igura, S ., Kadota, T ., and Hiroyasu, H.: "Spontaneous Ignition Delay of Fuel Sprays in High SP-625, SAE, 1985. Also ASME paper 85-GGP-24, 1985. Pressure Gaseous Environments," Trans. Japan Soc. Mech. Engrs, vol. 41, no. 345, pp. 24-31, 16. Hames, R. J ., Straub, R. D ., and Amann, R. W.: "DDEC Detroit Diesel Electronic Control," SAE 1975. paper 850542, 1985. 43. Spadaccini, L. J ., and TeVelde, J. A.: "Autoignition Characteristics of Aircraft-Type Fuels," 17. Hiroyasu, H.: "Diesel Engine Combustion and Its Modeling," in Diagnostics and Modeling of Combust. Flame, vol. 46, pp. 283-300, 1982. Combustion in Reciprocating Engines, pp. 53-75, COMODIA 85, Proceedings of Symposium, 44. Plee, S. L ., and Ahmad, T.: "Relative Roles of Premixed and Diffusion Burning in Diesel Com- Tokyo, Sept. 4 6, 1985. bustion," SAE paper 831733, SAE Trans ., vol. 92, 1983. 18. Arai, M ., Tabata, M ., and Hiroyasu, H.: "Disintegrating Process and Spray Characterization of 45. Stringer, F. W ., Clarke, A. E ., and Clarke, J. S.: "The Spontaneous Ignition of Hydrocarbon Fuel Jet Injected by a Diesel Nozzle," SAE paper 840275, SAE Trans ., vol. 93, 1984. Fuels in a Flowing System," Proc. Instn Mech. Engrs, vol. 184, pt. 3J, 1969-1970. 19. Kamimoto, T ., Kobayashi, H ., and Matsuoka, S.: " A Big Size Rapid Compression Machine for 46. Wolfer, H. H.: "Ignition Lag in Diesel Engines," VDI-Forschungsheft 392, 1938; Translated by Fundamental Studies of Diesel Combustion," SAE paper 811004, SAE Trans ., vol. 90, 1981. Royal Aircraft Establishment, Farnborough Library No. 358, UDC 621-436.047, August 1959. 20. Balles, E.: "Fuel-Air Mixing and Diesel Combustion in a Rapid Compression Machine," Ph.D. 47. Lyn, W .- T ., and Valdmanis, E.: "Effects of Physical Factors on Ignition Delay," SAE paper Thesis, Department of Mechanical Engineering, MIT, June 1987. 680102, 1968. 21. Browne, K. R ., Partridge, I. M ., and Greeves, G.: "Fuel Property Effects on Fuel/Air Mixing in 48. Wong, C. L ., and Steere, D. E.: "The Effects of Diesel Fuel Properties and Engine Operating an Experimental Diesel Engine," SAE paper 860223, 1986. Conditions on Ignition Delay," SAE paper 821231, SAE Trans ., vol. 91, 1982. 22. Bracco, F. V.: "Modeling of Engine Sprays," SAE paper 850394, 1985. 49. Andree, A ., and Pachernegg, S. J.: “Ignition Conditions in Diesel Engines," SAE paper 690253, 23. Kuo, T ., and Bracco, F. V.: "Computations of Drop Sizes in Pulsating Sprays and of Liquid-Core SAE Trans ., vol. 78, 1969. Length in Vaporizing Sprays," SAE paper 820133, SAE Trans ., vol. 91, 1982. 50. Glavincevski, B ., Gülder, O. L ., and Gardner, L.: “Cetane Number Estimation of Diesel Fuels 24. Reitz, R. D ., and Bracco, F. V.: "Mechanism of Atomization of a Liquid Jet," Phys. Fluid, vol. 25, from Carbon Type Structural Composition," SAE paper 841341, 1984. no. 10, pp. 1730-1742, 1982. 51. Olree, R ., and Lenane, D.: "Diesel Combustion Cetane Number Effects," SAE paper 840108, SAE 25. Reitz, R. D ., and Bracco, F. V.: "On the Dependence of Spray Angle and Other Spray Param- Trans ., vol. 93, 1984. eters on Nozzle Design and Operating Conditions," SAE paper 790494, 1979. 52. Schaefer, A. J ., and Hardenberg, H. O.: "Ignition Improvers for Ethanol Fuels," SAE paper 26. Wu, K .- J ., Su, C .- C ., Steinberger, R. L ., Santavicca, D. A ., and Bracco, F. V.: "Measurements of 810249, SAE Trans ., vol. 90, 1981. the Spray Angle of Atomizing Jets," J. Fluids Engng, vol. 105, pp. 406-413, 1983. 53. Hardenberg, H. O ., and Hase, F. W.: " An Empirical Formula for Computing the Pressure Rise 27. Hay, N ., and Jones, P. L.: "Comparison of the Various Correlations for Spray Penetration," SAE Delay of a Fuel from its Cetane Number and from the Relevant Parameters of Direct-Injection paper 720776, 1972. Diesel Engines," SAE paper 790493, SAE Trans ., vol. 88, 1979. 28. Dent, J. C.: "Basis for the Comparison of Various Experimental Methods for Studying Spray 54. Dent, J. C ., and Mehta, P. S.: "Phenomenological Combustion Model for a Quiescent Chamber Penetration," SAE paper 710571, SAE Trans ., vol. 80, 1971. Diesel Engine," SAE paper 811235, SAE Trans ., vol. 90, 1981. 29. Hiroyasu, H ., Kadota, T ., and Arai, M.: "Supplementary Comments: Fuel Spray Character- 55. Chang, Y. J ., Kobayashi, H ., Matsuzawa, K ., and Kamimoto, T.: "A Photographic Study of Soot ization in Diesel Engines," in James N. Mattavi and Charles A. Amann (eds.), Combustion Model- Formation and Combustion in a Diesel Flame with a Rapid Compression Machine," in Diagnos- ing in Reciprocating Engines, pp. 369-408, Plenum Press, 1980. tics and Modeling of Combustion in Reciprocating Engines, pp. 149-157, COMODIA 85, Pro- 30. Wu, K .- J ., Reitz, R. D ., and Bracco, F. V.: "Measurements of Drop Size at the Spray Edge near ceedings of Symposium, Tokyo, Sept. 4-6, 1985. the Nozzle in Atomizing Liquid Jets," Phys. Fluids, vol. 29, no. 4, pp. 941-951, 1986. 56. Lakshminarayan, P. A ., and Dent, J. C.: "Interferometric Studies of Vaporising and Combustion 31. Reitz, R. D ., and Diwakar, R.: "Effect of Drop Breakup on Fuel Sprays," SAE paper 860469, Sprays," SAE paper 830244, SAE Trans ., vol. 92, 1983. 1986. 57. Colella, K. J ., Balles, E. N ., Ekchian, J. A ., Cheng, W. K ., and Heywood, J. B.: " A Rapid Com- 32. Hiroyasu, H ., and Kadota, T.: "Fuel Droplet Size Distribution in Diesel Combustion Chamber, pression Machine Study of the Influence of Charge Temperature on Diesel Combustion," SAE SAE paper 740715, SAE Trans ., vol. 83, 1974. paper 870587, 1987. 33. El Wakil, M. M ., Myers, P. S ., and Uyehara, O. A.: "Fuel Vaporization and Ignition Lag in 58. Morris, C. J ., and Dent, J. C.: "The Simulation of Air Fuel Mixing in High Swirl Open Chamber Diesel Combustion," in Burning a Wide Range of Fuels in Diesel Engines, SAE Progress Diesel Engines," Proc. Instn Mech. Engrs, vol. 190, no. 47/76, pp. 503-513, 1976. Technol ., vol. 11, pp. 30-44, SAE, 1967. 59. Rife, J ., and Heywood, J. B.: " Photographic and Performance Studies of Diesel Combustion with 566 INTERNAL COMBUSTION ENGINE FUNDAMENTALS a Rapid Compression Machine," SAE paper 740948, SAE Trans ., vol. 83, 1974. 60. Whitehouse, N. D ., Cough, E ., and Jeje, A. B.: "The Study of Combustion in a Quiescent Com- bustion Chamber Diesel Engine," ASME paper 82-HT-35, 1982. CHAPTER 61. Nightingale, D. R.: " A Fundamental Investigation into the Problem of NO Formation in Diese Engines," SAE paper 750848, SAE Trans ., vol. 84, 1975. 62. Bennethum, J. E ., Mattavi, J. N ., and Toepel, R. R.: "Diesel Combustion Chamber Sampling Hardware, Procedures, and Data Interpretation," SAE paper 750849, SAE Trans ., vol. 84, 1975. 11 63. Rhee, K. T ., Myers, P. S ., and Uyehara, O. A.: "Time- and Space-Resolved Species Determi- nation in Diesel Combustion Using Continuous Flow Gas Sampling," SAE paper 780226, SAE Trans ., vol. 87, 1978. 64. Shipinsky, J ., Uyehara, O. A ., and Myers, P. S.: "Experimental Correlation between Rate-of- POLLUTANT Injection and Rate-of-Heat-Release in a Diesel Engine," ASME paper 68-DGP-11, 1968. 65. Grigg, H. C ., and Syed, M. H.: "The Problem of Predicting Rate of Heat Release in Diesel FORMATION Engine," Proc. Instn Mech. Engrs, vol. 184, pt. 3J, 1969-1970. 66. Whitehouse, N. D ., Clough, E ., and Way, J. B.: "The Effect of Changes in Design and Operating AND Conditions on Heat Release in Direct-Injection Diesel Engines," SAE paper 740085, 1974. CONTROL 67. Meguerdichian, M ., and Watson, N.: "Prediction of Mixture Formation and Heat Release in Diesel Engines," SAE paper 780225, 1978. 68. Kamimoto, T ., Aoyagi, Y ., Matsui, Y ., and Matsuoka, S.: "The Effects of Some Engine Variables on Measured Rates of Air Entrainment and Heat Release in a DI Diesel Engine," SAE paper 8000253, SAE Trans ., vol. 89, 1980. 69. Dent, J. C ., Mehta, P. S ., and Swan, J.: "A Predictive Model for Automotive DI Diesel Engine Performance and Smoke Emissions," paper C126/82, presented at the International Conference on Diesel Engines for Passenger Cars and Light Duty Vehicles, Institution of Mechanical Engi- neers, London, England, Oct. 5-7, 1982. 11.1 NATURE AND EXTENT OF PROBLEM Spark-ignition and diesel engines are a major source of urban air pollution. The spark-ignition engine exhaust gases contain oxides of nitrogen (nitric oxide, NO, and small amounts of nitrogen dioxide, NO2-collectively known as NO,), carbon monoxide (CO), and organic compounds which are unburned or partially burned hydrocarbons (HC). The relative amounts depend on engine design and operating conditions but are of order: NO„, 500 to 1000 ppm or 20 g/kg fuel; CO, 1 to 2 percent or 200 g/kg fuel; and HC, 3000 ppm (as C1) or 25 g/kg fuel. Piston blowby gases, and fuel evaporation and release to the atmosphere through vents in the fuel tank and carburetor after engine shut-down, are also sources of unburned hydrocarbons. However, in most modern engines these nonexhaust sources are effectively controlled by returning the blowby gases from the crank- case to the engine intake system and by venting the fuel tank and carburetor float bowl through a vapor-absorbing carbon cannister which is purged by some of the engine intake air during normal engine operation. In diesel engine exhaust, concentrations of NO, are comparable to those from SI engines. Diesel hydro- carbon emissions are significant though exhaust concentrations are lower by about a factor of 5 than typical SI engine levels. The hydrocarbons in the exhaust may also condense to form white smoke during engine starting and warm-up. 567 568 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 569 Specific hydrocarbon compounds in the exhaust gases are the source of diesel Deposits absorb HC NO forms in CO present at odor. Diesel engines are an important source of particulate emissions; between high-temperature high temperature about 0.2 and 0.5 percent of the fuel mass is emitted as small (~0.1 um diameter) burned gas and if fuel rich particles which consist primarily of soot with some additional absorbed hydro- carbon material. Diesel engines are not a significant source of carbon monoxide. Use of alcohol fuels in either of these engines substantially increases alde- hyde emissions. While these are not yet subject to regulation, aldehydes would be End gas a significant pollutant if these fuels were to be used in quantities comparable to source of gasoline and diesel. Currently used fuels, gasoline and diesel, contain sulfur: gas- Oil layers Unburned Flame HC if absorb HC mixture combustion oline in small amounts (<600 ppm by weight S), diesel fuel in larger amounts forced into incomplete (<0.5 percent). The sulphur is oxidized (or burned) to produce sulfur dioxide, crevices SO2, of which a fraction can be oxidized to sulfur trioxide, SO3, which combines with water to form a sulfuric acid aerosol. In general, the concentrations of these pollutants in internal combustion engine exhaust differ from values calculated assuming chemical equilibrium. Thus the detailed chemical mechanisms by which these pollutants form and the kinetics of these processes are important in determining emission levels. For some pollutant species, e.g ., carbon monoxide, organic compounds, and particulates, (a) Compression the formation and destruction reactions are intimately coupled with the primary (b) Combustion fuel combustion process. Thus an understanding of the formation of these species As burned gases cool, Deposits desorb HC requires knowledge of the combustion chemistry. For nitrogen oxides and sulfur first NO chemistry, then CO chemistry freezes oxides, the formation and destruction processes are not part of the fuel com- bustion process. However, the reactions which produce these species take place in an environment created by the combustion reactions, so the two processes are Entrainment of HC from still intimately linked. A summary of the mechanisms by which these pollutants wall into form in internal combustion engines provides an introduction to this chapter. In bulk gas subsequent sections, the details of the basic formation mechanisms of each pol- Oil layers lutant and the application of these mechanisms to the combustion process in desorb HC both spark-ignition and compression-ignition engines will be developed. The processes by which pollutants form within the cylinder of a convention- Outflow of al spark-ignition engine are illustrated qualitatively in Fig. 11-1. The schematic HC from shows the combustion chamber during four different phases of the engine oper- crevices; Piston some HC ating cycle: compression, combustion, expansion, and exhaust. Nitric oxide (NO) scrapes burns HC off forms throughout the high-temperature burned gases behind the flame through walls chemical reactions involving nitrogen and oxygen atoms and molecules, which do not attain chemical equilibrium. The higher the burned gas temperature, the higher the rate of formation of NO. As the burned gases cool during the expan- (c) Expansion (d) Exhaust sion stroke the reactions involving NO freeze, and leave NO concentrations far FIGURE 11-1 in excess of levels corresponding to equilibrium at exhaust conditions. Carbon Summary of HC, CO, and NO pollutant formation mechanisms in a spark-ignition engine. monoxide also forms during the combustion process. With rich fuel-air mixtures, there is insufficient oxygen to burn fully all the carbon in the fuel to CO2; also, in the high-temperature products, even with lean mixtures, dissociation ensures During compression and combustion, the increasing cylinder pressure forces there are significant CO levels. Later, in the expansion stroke, the CO oxidation some of the gas in the cylinder into crevices, or narrow volumes, connected to the process also freezes as the burned gas temperature falls. combustion chamber: the volumes between the piston, rings, and cylinder wall The unburned hydrocarbon emissions have several different sources. are the largest of these. Most of this gas is unburned fuel-air mixture; much of it 570 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 571 Air/fuel ratio escapes the primary combustion process because the entrance to these crevices is 20 17 15 14 too narrow for the flame to enter. This gas, which leaves these crevices later in T- 13 12 Stoichiometric the expansion and exhaust processes, is one source of unburned hydrocarbon Lean Rich emissions. Another possible source is the combustion chamber walls. A quench layer containing unburned and partially burned fuel-air mixture is left at the wall when the flame is extinguished as it approaches the wall. While it has been shown that the unburned HC in these thin (<0.1 mm) layers burn up rapidly when the NO combustion chamber walls are clean, it has also been shown that the porous deposits on the walls of engines in actual operation do increase engine HC emis- NO, CO, and HC concentrations (not to scale) sions. A third source of unburned hydrocarbons is believed to be any engine oil left in a thin film on the cylinder wall, piston and perhaps on the cylinder head. These oil layers can absorb and desorb fuel hydrocarbon components, before and after combustion, respectively, thus permitting a fraction of the fuel to escape the primary combustion process unburned. A final source of HC in engines is incom- HO plete combustion due to bulk quenching of the flame in that fraction of the engine cycles where combustion is especially slow (see Sec. 9.4.3). Such conditions CO are most likely to occur during transient engine operation when the air/fuel ratio, FIGURE 11-2 spark timing, and the fraction of the exhaust recycled for emission control may Variation of HC, CO, and NO concentration in not be properly matched. 0.7 0.8 0.9 1.0 1.1 1.2 1.3 the exhaust of a conventional spark-ignition The unburned hydrocarbons exit the cylinder by being entrained in the Fuel/air equivalence ratio engine with fuel/air equivalence ratio. bulk-gas flow during blowdown and at the end of the exhaust stroke as the piston pushes gas scraped off the wall out of the exhaust valve. Substantial oxida- tion of the hydrocarbons which escape the primary combustion process by any of to reduce emissions of all three pollutants, over all engine operating modes, and the above processes can occur during expansion and exhaust. The amount of achieve acceptable average levels. oxidation depends on the temperature and oxygen concentration time histories of In the diesel engine, the fuel is injected into the cylinder just before com- these HC as they mix with the bulk gases. bustion starts, so throughout most of the critical parts of the cycle the fuel dis- One of the most important variables in determining spark-ignition engine tribution is nonuniform. The pollutant formation processes are strongly emissions is the fuel/air equivalence ratio, o. Figure 11-2 shows qualitatively how dependent on the fuel distribution and how that distribution changes with time NO, CO, and HC exhaust emissions vary with this parameter. The spark-ignition due to mixing. Figure 11-3 illustrates how various parts of the fuel jet and the engine has normally been operated close to stoichiometric, or slightly fuel-rich, to flame affect the formation of NO, unburned HC, and soot (or particulates) during ensure smooth and reliable operation. Figure 11-2 shows that leaner mixtures the "premixed " and "mixing-controlled " phases of diesel combustion in a direct- give lower emissions until the combustion quality becomes poor (and eventually injection engine with swirl. Nitric oxide forms in the high-temperature burned misfire occurs), when HC emissions rise sharply and engine operation becomes gas regions as before, but temperature and fuel/air ratio distributions within the erratic. The shapes of these curves indicate the complexities of emission control. burned gases are now nonuniform and formation rates are highest in the close- In a cold engine, when fuel vaporization is slow, the fuel flow is increased to to-stoichiometric regions. Soot forms in the rich unburned-fuel-containing core of provide an easily combustible fuel-rich mixture in the cylinder. Thus, until the the fuel sprays, within the flame region, where the fuel vapor is heated by mixing engine warms up and this enrichment is removed, CO and HC emissions are with hot burned gases. Soot then oxidizes in the flame zone when it contacts high. At part-load conditions, lean mixtures could be used which would produce unburned oxygen, giving rise to the yellow luminous character of the flame. lower HC and CO emissions (at least until the combustion quality deteriorates) Hydrocarbons and aldehydes originate in regions where the flame quenches both and moderate NO emissions. Use of recycled exhaust to dilute the engine intake on the walls and where excessive dilution with air prevents the combustion mixture lowers the NO levels, but also deteriorates combustion quality. Exhaust process from either starting or going to completion. Fuel that vaporizes from the gas recirculation (EGR) is used with stoichiometric mixtures in many engine nozzle sac volume during the later stages of combustion is also a source of HC. control systems. Note that the highest power levels are obtained from the engine Combustion generated noise is controlled by the early part of the combustion with slightly rich-of-stoichiometric mixtures and no recycled exhaust to dilute the process, the initial rapid heat release immediately following the ignition-delay incoming charge. As we will see, several emission control techniques are required period. 572 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 573 Lean flame-out Initial rapid TABLE 11.1 region: HC combustion: noise Rate constants for NO formation mechanism1 Burned gas: NO Rate constant, Temperature Uncertainty, Fuel jet Reaction cm3/mol . s range, K factor of or % mixing with air: rich mixture (1) O+ N2 -+ NO + N 7.6 × 1013 exp [ -38,000/T] 2000-5000 2 (-1) N + NO -+ N2 + 0 1.6 x 1013 300-5000 + 20% at 300 K 2 at 2000-5000 K (2) N +O2 -+ NO + 0 6.4 x 109 T exp [-3150/7] 300-3000 Premixed +30% 300-1500 K 2 at 3000 K (-2) 0 + NO - 02 + N 1.5 x 109 T exp [-19,500/T] 1000-3000 + 30% at 1000 K White/yellow flame: 2 at 3000 K soot oxidation (3) N + OH -+ NO + H 4.1 x 1013 Burned gas: NO 300-2500 +80% Rich zones (-3) H + NO - OH + N 2.0 x 1014 exp [-23,650/T] 2200-4500 2 Flame quench in fuel jet: on walls: HC soot formation FIGURE 11-3 Summary of pollutant formation The forward and reverse rate constants (kit and k; , respectively) for these reac- mechanisms in a direct-injection tions have been measured in numerous experimental studies. Recommended Fuel vapor from nozzle compression-ignition engine values for these rate constants taken from a critical review of this published data sac volume during "premixed " and "mixing- are given in Table 11.1. Note that the equilibrium constant for each reaction, Kc,i Mixing controlled controlled " combustion phases. (see Sec. 3.7.2), is related to the forward and reverse rate constants by Kc,i = kj/kt. The rate of formation of NO via reactions (11.1) to (11.3) is given by [see 11.2 NITROGEN OXIDES Eqs. (3.55) and (3.58)] 11.2.1 Kinetics of NO Formation d[NO] dt = k [O][N2] + k2 [N][O2] + k; [N][OH] While nitric oxide (NO) and nitrogen dioxide (NO2) are usually grouped together as NO, emissions, nitric oxide is the predominant oxide of nitrogen produced - ki[NO][N] - k2 [NO][O] - k; [NO][H] (11.4) inside the engine cylinder. The principal source of NO is the oxidation of atmo- spheric (molecular) nitrogen. However, if the fuel contains significant nitrogen, where [ ] denote species concentrations in moles per cubic centimeter when k; the oxidation of the fuel nitrogen-containing compounds is an additional source have the values given in Table 11.1. The forward rate constant for reaction (11.1) of NO. Gasolines contain negligible amounts of nitrogen; although diesel fuels and the reverse rate constants for reactions (11.2) and (11.3) have large activation contain more nitrogen, current levels are not significant. energies which results in a strong temperature dependence of NO formation rates. The mechanism of NO formation from atmospheric nitrogen has been studied extensively.1 It is generally accepted that in combustion of near- A similiar relation to (11.4) can be written for d[N]/dt: stoichiometric fuel-air mixtures the principal reactions governing the formation d[N] of NO from molecular nitrogen (and its destruction) aret dt ! = kt [ O][N2 ] - k [N][2] - k [N][OH] O + N2 = NO + N (11.1) - ki[NO][N] + kz[NO][O] + k, [NO][H] (11.5) N + 02 = NO + 0 (11.2) Since [N] is much less than the concentrations of other species of interest N + OH = NO + H (11.3) (~10-8 mole fraction), the steady-state approximation is appropriate: d[N]/dt is set equal to zero and Eq. (11.5) used to eliminate [N]. The NO formation rate then becomes d[NO] 1 - [NO]2/(K[O2][N2]) + This is often called the extended Zeldovich mechanism. Zeldovich1 was the first to suggest the dt = 2k+[O][N2] (11.6) importance of reactions (11.1) and (11.2). Lavoie et al.2 added reaction (11.3) to the mechanism; it 1 + kj[NO]/(k*[O2] + kj[OH]) does contribute significantly. where K = (kt/kjkj/kg). 574 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 575 TABLE 11.2 10 Typical values of R1, R1/R2, and R1/(R2 + R3)t Adiabatic flame Equivalence temperature ratio R $ RIR2 R,/(R2 + R3) 0.8 5.8 × 10-5 1.2 0.33 1.0 2.8 x 10-5 2.5 0.26 10-1 1.2 7.6 × 10-6 9.1 0.14 - at a = 0, s- 0.6 + At 10 atm pressure and 2600 K. + Units gmol/cm3. s. ).8 d(NO) 1.0 ¢ = 1.2 FIGURE 11-4 NO forms in both the flame front and the postflame gases. In engines, 10-3 Initial NO formation rate, mass fraction per however, combustion occurs at high pressure so the flame reaction zone is second (for [NO]/[NO]< < 1), as a function of extremely thin (~0.1 mm) and residence time within this zone is short. Also, the temperature for different equivalence ratios (o) and 10 15 atm pressure. Dashed line shows adiabatic cylinder pressure rises during most of the combustion process, so burned gases 2000 2100 2200 2300 2400 2500 2600 flame temperature for kerosene combustion with produced early in the combustion process are compressed to a higher tem- Temperature, K 700 K, 15 atm air.3 perature than they reached immediately after combustion. Thus, NO formation in the postflame gases almost always dominates any flame-front-produced NO. It is, therefore, appropriate to assume that the combustion and NO formation pro- cesses are decoupled and to approximate the concentrations of O, O2, OH, H, where Kp(o) is the equilibrium constant for the reaction and N2 by their equilibrium values at the local pressure and equilibrium tem- ¿02 = 0 perature. To introduce this equilibrium assumption it is convenient to use the nota- and is given by tion R1 = k+[O] [N2]e = ki[NO] [N]e, where [ ]e denotes equilibrium con- Kp(o) = 3.6 x 103 exp -31,090 centration, for the one-way equilibrium rate for reaction (11.1), with similiar atm 1/2 (11.10) definitions for R2 = kt[N][O2]e = k2 [NO] [O], and R3 = ks[N].[OH]. T = kg [NO] [H]e. Substituting [O]e, [O2]e, [OH], [H]e, and [N2]e for [O], The initial NO formation rate may then be written [combining Eqs. (11.8), (11.9), [O2], [OH], [H], and [N2] in Eq. (11.6) yields and (11.10) with k+ from Table 11.1] as d[NO] 2R1{1 - ([NO]/[NO])2) (11.7) d[NO] 6 x 1016 dt 1 + ([NO]/[NO])R1/(R2 + R3) dt T1/2 exp -69,090) T [o 2]/2[N 2Je mol/cm3 . s (11.11) Typical values of R1, R1/R2 and R1/(R2 + R3) are given in Table 11.2. The differ- The strong dependence of d[NO]/dt on temperature in the exponential term is ence between R1/R2 and R1/(R2 + R3) indicates the relative importance of evident. High temperatures and high oxygen concentrations result in high NO adding reaction (11.3) to the mechanism. formation rates. Figure 11-4 shows the NO formation rate as a function of gas The strong temperature dependence of the NO formation rate can be temperature and fuel/air equivalence ratio in postflame gases. Also shown is the demonstrated by considering the initial value of d[NO]/dt when [NO]/[NO]< < adiabatic flame temperature attained by a fuel-air mixture initially at 700 K at a 1. Then, from Eq. (11.7), constant pressure of 15 atm. For adiabatic constant-pressure combustion (an d[NO] appropriate model for each element of fuel that burns in an engine), this initial = 2R1 = 2kt[O] [N2]e (11.8) NO formation rate peaks at the stoichiometric composition, and decreases at rapidly as the mixture becomes leaner or richer. The equilibrium oxygen atom concentration is given by A characteristic time for the NO formation process, TNo, can be defined by [0] = KO,[O2]1/2 (RT)1/2 (11.9) UNO = 1 d[NO] [NO]. dt (11.12) 576 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 577 [NO]e can be obtained from the equilibrium constant TABLE 11.3 Typical nitrogen content of distillate fuels1 KNO = 20.3 x exp (-21,650/T) Fraction Average nitrogen, wt % for the reaction Range, wt % Crude 0.65 O2 + N2 = 2NO Heavy distillates 1.40 0.60-2.15 as [NO]. = (KNo[O2][N2]e)1/2. Equations (11.11) and (11.12) can be combined Light distillates 0.07 0-0.60 to give 8 x 10-16T exp (58,300/T) TNO = p1/2 (11.13) close to zero, indicating that at these high pressures there is negligible NO pro- duction within the flame front itself. where INo is in seconds, T in kelvins, and p in atmospheres. Use has been made Fuel nitrogen is also a source of NO via a different and yet to be fully of the fact that EN, ~ 0.71. For engine combustion conditions, tNo is usually com- explained mechanism. Table 11.3 shows the typical nitrogen content of parable to or longer than the times characteristic of changes in engine conditions petroleum-derived fuels. During distillation, the fuel nitrogen is concentrated in so the formation process is kinetically controlled. However, for close-to- the higher boiling fractions. In distillate fuels, the nitrogen can exist as amines stoichiometric conditions at the maximum pressures and burned gas tem- and ring compounds (e.g ., pyridine, quinoline, and carbazoles). During com- peratures, tNo is of the same order as typical combustion times (1 ms) and bustion these compounds are likely to undergo some thermal decomposition equilibrium NO concentrations may be attained. prior to entering the combustion zone. The precursors to NO formation will Evidence that this formation model is valid under conditions typical of therefore be low molecular weight nitrogen-containing compounds such as NH3, those found in engines is provided by high-pressure combustion bomb studies. HCN, and CN. The detailed information on the kinetics of NO formation from Newhall and Shahed4 have measured the NO production, using the q-band these compounds is limited. Oxidation to NO is usually rapid, occurring on a absorption technique, behind hydrogen-air and propane-air planar flames propa- time scale comparable to that of the combustion reactions. The NO yield gating axially in a cylindrical bomb. Some results are compared with predictions (amount of fuel nitrogen converted to NO) is sensitive to the fuel/air equivalence made with this kinetic scheme (coupled with an "unmixed" combustion calcu- ratio. Relatively high NO yields (approaching 100 percent) are obtained for lean lation to determine local pressure and temperature; see Sec. 9.2.1) in Fig. 11-5. and stoichiometric mixtures; relatively low yields are found for rich mixtures. NO The agreement is excellent. Note that the NO concentration rises smoothly from yields are only weakly dependent on temperature, in contrast to the strong tem- perature dependence of NO formed from atmospheric nitrogen.1 5 x 10-7 0 ¢ = 0.7 11.2.2 Formation of NO2 ¢ = 0.9 0 $ = 1.0 Chemical equilibrium considerations indicate that for burned gases at typical - Theoretical flame temperatures, NO2/NO ratios should be negligible small. While experimen- tal data show this is true for spark-ignition engines, in diesels NO2 can be 10 to 30 percent of the total exhaust oxides of nitrogen emissions.5 A plausible mecha- 5 X 10-8¢ .P nism for the persistence of NO2 is the following.º NO formed in the flame zone can be rapidly converted to NO, via reactions such as NO concentration, moles/cm3 o NO + HO2 -> NO2 + OH O (11.14) o O O O o Subsequently, conversion of this NO2 to NO occurs via 25 x 10-91 NO2 + 0 - NO + 02 FIGURE 11-5 (11.15) Measured and calculated rate-limited NO concen- unless the NO2 formed in the flame is quenched by mixing with cooler fluid. This trations behind flame in high-pressure cylindrical explanation is consistent with the highest NO2/NO ratio occurring at light load 10-90 5 10 15 bomb experiments with H ,- air mixture. O in diesels, when cooler regions which could quench the conversion back to NO Time, ms ¢ = equivalence ratio.4 are widespread.5 578 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 579 6000- T T SI engine 30 1680 3 1 80 (a) Spark 5000j- 70 2000 Diesel NO 2 60 0.5 *b 4000 20- p P. MPa 50 percent 3000 NO2, ppm NO, ppm NO2 1300 40 NO2 NOT OL 0 2000- 30 10- 1000 20 -10 O 10 20 30 40 50 60 70 80 90 20 2800 1000- rev/min 3000 Spark % = 0 (b) 18 JO 11 12 13 14 15 16 17 19 100 200 300 ADO Ti, Tu, K Air/fuel ratio bmep, kPa 1ª 2000 Xb (a) (b) FIGURE 11-6 1000 (a) NO and NO, concentrations in SI engine exhaust as function of air/fuel ratio, 1500 rev/min, wide-open throttle; (b) NO, as percent of total NO, in diesel exhaust as function of load and speed.5 20 -10 20 30 40 50 60 70 80 .90 10,000 Figure 11-6 shows examples of NO and NO2 emissions data from a spark- (c) ignition and a diesel engine. The maximum value for the ratio (NO2/NO) for the SI engine is 2 percent, at an equivalence ratio of about 0.85. For the diesel this Equilibrium _ __ Spark ratio is higher, and is highest at light load and depends on engine speed. 5000 NO, ppm It is customary to measure total oxides of nitrogen emissions, NO plus NO2, with a chemiluminescence analyzer and call the combination NO ,. It is always important to check carefully whether specific emissions data for NO, are Rate-controlled OL given in terms of mass of NO or mass of NO2, which have molecular weights of -20 -10 0 10 10 20 30 40 50 60 70 80 90 Crank angle, deg 30 and 46, respectively. FIGURE 11-7 Illustration of SI engine NO formation model: (a) measured cylinder pressure p and calculated mass 11.2.3 NO Formation in Spark-Ignition Engines fraction burned x,; (b) calculated temperature of unburned gas T, and burned gas T, in early- and late-burning elements; (c) calculated NO concentrations in early- and late-burning elements for rate- In conventional spark-ignition engines the fuel and air (and any recycled exhaust) controlled model and at equilibrium.7 are mixed together in the engine intake system, and vigorous mixing with the residual gas within the cylinder occurs during the intake process. Thus the fuel/ air ratio and the amount of diluent (residual gas plus any recycled exhaust) is approximately uniform throughout the charge within the cylinder during com- temperature distribution which develops in the burned gases due to the passage of the flame across the combustion chamber has been discussed in Sec. 9.2.1. bustion.+ Since the composition is essentially uniform, the nature of the NO for- Mixture which burns early is compressed to higher temperatures after com- mation process within the cylinder can be understood by coupling the kinetic bustion, as the cylinder pressure continues to rise; mixture which burns later is mechanism developed in Sec. 11.2.1 with the burned gas temperature distribution and pressure in the cylinder during the combustion and expansion processes. The compressed primarily as unburned mixture and ends up after combustion at a lower burned gas temperature. Figure 11-7a and b shows measured cylinder pres- sure data from an operating engine, with estimates of the mass fraction burned (x)) and the temperatures of a gas element which burned just after spark dis- + It is well known that the mixture composition within the cylinder is not completely uniform and charge and a gas element which burned at the end of the burning process. The varies from one cycle to the next. Both these factors contribute to cycle-by-cycle combustion varia- model used to estimate these temperatures assumed no mixing between mixture tions. For the present discussion, the assumption of mixture uniformity is adequate. elements which burn at different times. This assumption is more realistic than the 580 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 581 alternative idealization that the burned gases mix rapidly and are thus uniform (see Sec. 9.2.1). If the NO formation kinetic model [Eq. (11.7)] is used to calculate . Experiment W2 o Experiment W3 NO concentrations in these burned gas elements, using the equilibrium concen- -- Kinetic solutions trations of the species O, O2, N2, OH, and H corresponding to the average 10-24 fuel/air equivalence ratio and burned gas fraction of the mixture and these pres- sure and temperature profiles, the rate-limited concentration profiles in Fig. 11-7c TTT are obtained. Also shown are the NO concentrations that would correspond to chemical equilibrium at these conditions. The rate-controlled concentrations rise NO, mole fraction from the residual gas NO concentration, lagging the equilibrium levels, then cross the equilibrium levels and "freeze" well above the equilibrium values corre- @ 10 -3- FIGURE 11-8 sponding to exhaust conditions. Depending on details of engine operating condi- Spectroscopically measured NO concentrations tions, the rate-limited concentrations may or may not come close to equilibrium through two windows W3 and W2 in special - T @ = 0.9 levels at peak cylinder pressure and gas temperature. Also, the amount of decom- L-head SI engine (W2 is closer to spark than W3). The asterisks mark estimated initial conditions position from peak NO levels which occurs during expansion depends on engine and flame arrival times. The dashed lines are cal- conditions as well as whether the mixture element burned early or late.7 10-4 culated rate-limited concentrations for parts of Once the NO chemistry has frozen during the early part of the expansion 20 0 20 40 60 charge burning at these flame arrival times with stroke, integration over all elements will give the final average NO concentration Crank angle, deg zero initial NO concentration.1º in the cylinder which equals the exhaust concentration. Thus, if {NO} is the local mass fraction of NO, then the average exhaust NO mass fraction is given by time of arrival of the flame at each window. The observed NO mole fractions rise smoothly from these initial values and then freeze about one-third of the way (NO) = (NO), dx, (11.16) through the expansion process. NO levels observed at window W2, closest to the Jo spark plug, were substantially higher than those observed at window W3. The where {NO}, is the final frozen NO mass fraction in the element of charge which dashed lines show calculated NO concentrations obtained using the NO forma- burned when the mass fraction burned was x ,. Note that {NO} = [NO]MNo/P, tion kinetic model with an "unmixed" thermodynamic analysis for elements that where MNo = 30, the molecular weight of NO. The average exhaust concentra- burned at the time of flame arrival at each window. Since the calculated values tion of NO as a mole fraction is given by started from zero NO concentration at the flame front (and not the diluted &NOV = {NO} Mexh residual gas NO level indicated by the star), the calculations initially fall below (11.17) M NO the data. However, the difference between the two measurement locations and the frozen levels are predicted with reasonable accuracy. Thus, the rate-limited for- and the exhaust concentration in ppm is Noay x 106. The earlier burning frac- mation process, freezing of NO chemistry during expansion, and the existence of tions of the charge contribute much more to the exhausted NO than do later NO concentration gradients across the combustion chamber have all been burning fractions of the charge: frozen NO concentrations in these early-burning observed. elements can be an order of magnitude higher than concentrations in late- The most important engine variables that affect NO emissions are the fuel/ burning elements. In the absence of vigorous bulk gas motion, the highest NO air equivalence ratio, the burned gas fraction of the in-cylinder unburned mixture, concentrations occur nearest the spark plug. and spark timing. The burned gas fraction depends on the amount of diluent Substantial experimental evidence supports this description of NO forma- such as recycled exhaust gas (EGR) used for NO, emissions control, as well as tion in spark-ignition engines. The NO concentration gradient across the burned the residual gas fraction. Fuel properties will affect burned gas conditions; the gas in the engine cylinder, due to the temperature gradient, has been demon- effect of normal variations in gasoline properties is modest, however. The effect of strated using gas sampling techniques8, 9 and using measurements of the chemilu- variations in these parameters can be explained with the NO formation mecha- minescent radiation from the reaction NO + O -> NO2 + hv to determine the nism described above: changes in the time history of temperature and oxygen local NO concentration. Figure 11-8 shows NO concentration data as a function concentration in the burned gases during the combustion process and early part of crank angle, taken by Lavoie1º through two different windows in the cylinder of the expansion stroke are the important factors.11 head of a specially constructed L-head engine where each window was a different distance from the spark plug. The stars indicate the estimated initial NO concen- EQUIVALENCE RATIO. Figure 11-9 shows the effect of variations in the fuel/air tration that results from mixing of the residual gas with the fresh charge, at the equivalence ratio on NO emissions. Maximum burned gas temperatures occur at 582 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 583 Fuel/air equivalence ratio 3000 1.05 1.0 0.9 0.8 0.75 3000 2000 2000 ៛ 150 16X NO, ppm NO, ppm 17 A 1000 1000 FIGURE 11-9 O, 0 14 16 18 20 Variation of exhaust NO concentration with A/F and FIGURE 11-10 fuel/air equivalence ratio. Spark-ignition engine, 1600 Variation of exhaust NO concentration with per- rev/min, no = 50 percent, MBT timing.12 0L 0 cent recycled exhaust gas (EGR). Spark-ignition 10 20 engine, 1600 rev/min, ny = 50 percent, MBT tim- EGR, % ing. 12 ~ 1.1; however, at this equivalence ratio oxygen concentrations are low. As the mixture is enriched, burned gas temperatures fall. As the mixture is leaned out, increasing oxygen concentration initially offsets the falling gas temperatures and NO emissions peak at o ~ 0.9. Detailed predictions of NO concentrations in the burned gases suggest that the concentration versus time histories under fuel-lean conditions are different in character from those for fuel-rich conditions. In lean heat capacity of the cylinder charge, per unit mass of fuel. Figure 11-11 shows the mixtures NO concentrations freeze early in the expansion process and little NO effect of different diluent gases added to the engine intake flow, in a single- decomposition occurs. In rich mixtures, substantial NO decomposition occurs cylinder engine operated at constant speed, fuel flow, and air flow.13 The data in from the peak concentrations present when the cylinder pressure is a maximum. Fig. 11-11a show that equal volume percentages of the different diluents produce Thus in lean mixtures, gas conditions at the time of peak pressure are especially different reductions in NO emissions. The same data when plotted against diluent significant.7 heat capacity (diluent mass flow rate x specific heat, c,) collapse to a single BURNED GAS FRACTION. The unburned mixture in the cylinder contains fuel vapor, air, and burned gases. The burned gases are residual gas from the previous cycle and any exhaust gas recycled to the intake for NO, emissions control. The 100 100 residual gas fraction is influenced by load, valve timing (especially the extent of N2 valve overlap), and, to a lesser degree, by speed, air/fuel ratio, and compression 80 He 80 - ratio as described in Sec. 6.4. The burned gases act as a diluent in the unburned CO2 Exhaust gas ------ mixture; the absolute temperature reached after combustion varies inversely with 60 Ar 60- Reduction in mass NO, % the burned gas mass fraction. Hence increasing the burned gas fraction reduces O Reduction in mass NO, % H,O NO emissions levels. However, it also reduces the combustion rate and, therefore, 40 40 O @ CO2 makes stable combustion more difficult to achieve (see Secs. 9.3 and 9.4). ------------ O . H2O A N2 Figure 11-10 shows the effect of increasing the burned gas fraction by recy- 20- 20|- He cling exhaust gases to the intake system just below the throttle plate. Substantial + o Ar reductions in NO concentrations are achieved with 15 to 25 percent EGR, which + Exhaust gas 0 10 20 30 0.2 is about the maximum amount of EGR the engine will tolerate under normal 0.4 0.6 0.8 1.0 Diluent in intake mixture, % vol part-throttle conditions. Of course, increasing the EGR at fixed engine load and Diluent heat capacity mc ,, J/K . s (a) (b) speed increases the inlet manifold pressure, while fuel flow and air flow remain FIGURE 11-11 approximately constant. The primary effect of the burned gas diluent in the unburned mixture on the (@) Percentage reduction in mass NO emissions with various diluents. (b) Correlation of NO reduction with diluent heat capacity. Spark-ignition engine operated at 1600 rev/min, constant brake load NO formation process is that it reduces flame temperatures by increasing the (intake pressure ~0.5 atm), with MBT spark timing. 13 584 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 585 curve.+ A similiar study where the burned gas fraction in the unburned charge will correlate NO emissions.+ Figure 11-12 shows the correlation of specific NO was varied by changing the valve overlap, compression ratio, and EGR, separ- emissions, from a four-cylinder engine, over a wide range of engine operating ately, showed that, under more realistic engine operating conditions, it is the heat conditions with the air/fuel ratio and gas/fuel ratio. Lines of constant air/fuel capacity of the total diluent mass in the in-cylinder mixture that is important. ratio and volumetric efficiency are shown; the direction of increasing dilution Whether the diluent mass is changed by varying the valve overlap, EGR, or even with residual gas and EGR at constant air/fuel ratio is to the right. Excessive the compression ratio is not important.14 dilution results in poor combustion quality, partial burning, and, eventually, misfire (see Sec. 9.4.3). Lowest NO emissions consistent with good fuel consump- EXCESS AIR AND EGR. Because of the above, it is possible to correlate the influ- tion (avoiding the use of rich mixtures) are obtained with a stoichiometric ence of engine operating variables (such as air/fuel ratio, engine speed, and load) mixture, with as much dilution as the engine will tolerate without excessive dete- and design variables (such as valve timing and compression ratio) on NO emis- rioration in combustion quality.15 sions with two parameters which define the in-cylinder mixture composition: the Comparisons between predictions made with the NO formation model fuel/air equivalence ratio (often the air/fuel ratio is used instead) and the gas/fuel (described at the beginning of this section) and experimental data show good ratio. The gas/fuel ratio (G/F) is given by agreement with normal amounts of dilution.16 With extreme dilution, at NO levels of about 100 ppm or less, the NO formed within the flame reaction zone G total mass in cylinder # ( 1 + 1 - X6 ) (11.18) cannot, apparently, be neglected. Within the flame, the concentrations of radicals F fuel mass in cylinder such as O, OH, and H can be substantially in excess of equilibrium levels, where x, is the burned gas fraction [Eq. (4.3)]. These together define the relative resulting in much higher formation rates within the flame than in the postflame proportions of fuel, air, and burned gases in the in-cylinder mixture, and hence gases. It is believed that the mechanism [reactions (11.1) to (11.3)] and the forma- tion rate equation (11.6) are valid. However, neglecting flame-front-formed NO is no longer an appropriate assumption.17 SPARK TIMING. Spark timing significantly affects NO emission levels. Advanc- 15 16 ing the timing so that combustion occurs earlier in the cycle increases the peak 0 64 cylinder pressure (because more fuel is burned before TC and the peak pressure 4 50 12.5 moves closer to TC where the cylinder volume is smaller); retarding the timing 15 0 36 17 22 decreases the peak cylinder pressure (because more of the fuel burns after TC). 14.56 Higher peak cylinder pressures result in higher peak burned gas temperatures, 10 and hence higher NO formation rates. For lower peak cylinder pressures, lower 14 NO formation rates result. Figure 11-13 shows typical NO emission data for a 18 spark-ignition engine as a function of spark timing. NO emission levels steadily : 7.5 decrease as spark timing is retarded from MBT timing and moved closer to TC. Indicated specific NO, emissions, g/kW.h Since exact determination of MBT timing is difficult (and not critical for fuel consumption and power where the variation with timing around MBT is 5 13 modest), there is always considerable uncertainty in NO emissions at MBT timing. Often, therefore, an alternative reference timing is used, where spark is FIGURE 11-12 retarded from MBT timing to the point where torque is decreased by 1 or 2 2.51 12 Correlation between gas/fuel ratio (G/F) and indi- percent from the maximum value. Great care with spark timing is necessary to cated specific NO, emissions at various air/fuel obtain accurate NO emissions measurements under MBT-timing operating con- ratios (A/F) and volumetric efficiencies (n.). Spark- ditions. ignition engine operated at 1400 rev/min with spark 10 14 18 22 timing retarded to give 0.95 of maximum brake Gas/fuel ratio torque.15 f Some of the scatter in Fig. 11-11 is due to the fact that the residual gas fraction is slightly different t Spark timing also affects NO emissions, as discussed next. The above discussion relates to engines run with timing at MBT or with torque at a fixed percentage of (and close to) the maximum. for each diluent. 586 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 587 3000 2000 15 -S+ TC NO, ppm 20º 16 30º Piston 1000 17 4 00 2400 FIGURE 11-13 -A 3 6 Variation of exhaust NO concentration with spark 2200 0 40 20 0 Pressure, MPa P retard. 1600 rev/min, no = 50 percent; left-hand end Equivalence ratio 2 4 -2000 Flame temperature, K Spark timing, deg BTC of curve corresponds to MBT timing for each A/F.12 2 injection 1800 11.2.4 NO, Formation in Compression-Ignition NO A 0 0 Engines - 800 1.5 15 A- The kinetic mechanisms for NO and NO2 formation described in Secs. 11.2.1 and 20 600 Sool -0- 11.2.2 and the assumptions made regarding equilibration of species in the 1.0- 10 CO2,CO,CH4,02 mole fraction, % co - - A NO, ppm C-O-H system apply to diesels as well as to spark-ignition engines. The criti- C6H 14, soot mole fraction, % 400 cal difference, of course, is that injection of fuel into the cylinder occurs just 0.5 5 -200 before combustion starts, and that nonuniform burned gas temperature and com- CHA position result from this nonuniform fuel distribution during combustion. The 0 0 fuel-air mixing and combustion processes are extremely complex. During the -20 TC 20 40 70 "premixed " or uncontrolled diesel combustion phase immediately following the Crank angle, deg Exhaust FIGURE 11-14 ignition delay, fuel-air mixture with a spread in composition about stoichiometric Concentrations of soot, NO, and other combustion product species measured at outer edge of bowl- burns due to spontaneous ignition and flame propagation. During the mixing in-piston combustion chamber (location S) of quiescent DI diesel with rapid sampling valve. Cylinder controlled combustion phase, the burning mixture is likely to be closer to stoi- gas pressure p, mean temperature T, and local equivalence ratio ¢ shown. Bore = 95 mm, chiometric (the flame structure is that of a turbulent, though unsteady, diffusion stroke = 110 mm, r = 14.6. Four-hole nozzle with hole diameter = 0.2 mm.18 flame). However, throughout the combustion process mixing between already burned gases, air, and lean and rich unburned fuel vapor-air mixture occurs, changing the composition of any gas elements that burned at a particular equiva- is especially important since it is compressed to a higher temperature, increasing lence ratio. In addition to these composition (and hence temperature) changes the NO formation rate, as combustion proceeds and cylinder pressure increases. due to mixing, temperature changes due to compression and expansion occur as After the time of peak pressure, burned gas temperatures decrease as the cylinder the cylinder pressure rises and falls. gases expand. The decreasing temperature due to expansion and due to mixing of The discussion in Sec. 11.2.1 showed that the critical equivalence ratio for high-temperature gas with air or cooler burned gas freezes the NO chemistry. NO formation in high-temperature high-pressure burned gases typical of engines This second effect (which occurs only in the diesel) means that freezing occurs is close to stoichiometric. Figure 11-4 is relevant here: it shows the initial NO more rapidly in the diesel than in the spark-ignition engine, and much less decomposition of the NO occurs. formation rate in combustion products formed by burning a mixture of a typical hydrocarbon fuel with air (initially at 700 K, at a constant pressure of 15 atm). The above description is supported by the NO concentration data obtained NO formation rates are within a factor of 2 of the maximum value for 0.85 from experiments where gas was sampled from within the cylinder of normally ¢ <1.1. operating diesel engines with special gas-sampling valves and analyzed. Figure The critical time period is when burned gas temperatures are at a 11-14 shows time histories of major species concentrations, through the com- maximum: i.e ., between the start of combustion and shortly after the occurrence bustion process, determined with a rapid-acting sampling valve (1 ms open time) of peak cylinder pressure. Mixture which burns early in the combustion process in a quiescent direct-injection diesel engine. Concentrations at different positions in the combustion chamber were obtained; the sample valve location for the Fig. 588 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 589 4000 1.2 NO, NO 3000 -- 1.0F rev/min NO, NO3, ppm 0.8 8.50 2000 1000 1200 (NO) exh NO 0.6- 1000 0.4 FIGURE 11-16 FIGURE 11-15 OL Exhaust NO, and NO concentrations as a function 0.2- Ratio of cylinder-average NO concentration at 0.3 0.4 0.5 0.6 0.7 of overall equivalence ratio or engine load. DI given crank angle (determined from cylinder- Equivalence ratio TC diesel, 1000 rev/min, injection timing at 27º BTC.19 0 dumping experiments) to exhaust NO concen- -20 -10 0 10 20 30 40 tration. DI diesel, equivalence ratio = 0.6, Crank angle, deg injection timing at 27º BTC.19 amount of fuel injected decreases proportionally as the overall equivalence ratio is decreased, much of the fuel still burns close to stoichiometric. Thus NO emis- 11-14 data is shown. Local NO concentrations rise from the residual gas value sions should be roughly proportional to the mass of fuel injected (provided following the start of combustion, to a peak at the point where the local burned burned gas pressures and temperatures do not change greatly). gas equivalence ratio changes from rich to lean (where the CO2 concentration Similar gas-sampling studies have been done with indirect-injection diesel has its maximum value). As the local burned gas equivalence ratio becomes engines. Modeling studies suggest that most of the NO forms within the pre- leaner due to mixing with excess air, NO concentrations decrease since formation chamber and is then transported into the main chamber where the reactions becomes much slower as dilution occurs. At the time of peak NO concentrations freeze as rapid mixing with air occurs. However, the prechamber, except at light within the bowl (15º ATC), most of the bowl region was filled with flame. The load, operates rich overall so some additional NO can form as the rich com- total amount of NO within the cylinder of this type of direct-injection diesel bustion products are diluted through the stoichiometric composition.2º Figure during the NO formation process has also been measured.19 At a predetermined 11-17 shows local NO concentrations and equivalence ratios as a function of time in one cycle, once steady-state warmed-up engine operation had been crank angle determined with a rapid-acting sampling valve at different locations achieved, the contents of the cylinder were dumped into an evacuated tank by rapidly cutting open a diaphragm which had previously sealed off the tank system. Figure 11-15 shows how the ratio of the average cylinder NO concentra- 2400 tion divided by the exhaust concentration varies during the combustion process. Distance of sample valve from wall NO concentrations reach a maximum shortly after the time of peak pressure. 2000 2 mm --- 10 mm There is a modest amount of NO decomposition. Variations in engine speed have -- 15 mm 18.5 mm little effect on the shape of this curve. The 20 crank angle degrees after the start of 1600 1.5 combustion is the critical time period. NO, ppm Results from similar cylinder-dumping experiments where injection timing 1200 1.0- and load (defined by the overall equivalence ratio) were varied also showed that 800- almost all of the NO forms within the 20º following the start of combustion. As Equivalence ratio d injection timing is retarded, so the combustion process is retarded; NO forma- ).5 1 400 tion occurs later, and concentrations are lower since peak temperatures are lower. The effect of the overall equivalence ratio on NO, concentrations is shown 0 0 10 TC 10 20 30 40 50 in Fig. 11-16. At high load, with higher peak pressures, and hence temperatures, -5 TC 10 20 Ignition 30 40 and larger regions of close-to-stoichiometric burned gas, NO levels increase. Both Crank angle, deg Ignition Crank angle, deg (a) (b) NO and NO, concentrations were measured; NO2 is 10 to 20 percent of total FIGURE 11-17 NO ,. Though NO levels decrease with a decreasing overall equivalence ratio, (a) NO concentrations measured with rapid sampling valve and (b) calculated equivalence ratios at they do so much less rapidly than do spark-ignition engine NO emissions (see different distances from the wall in swirl chamber of IDI diesel engine, as function of crank angle. Fig. 11-9) due to the nonuniform fuel distribution in the diesel. Though the Engine speed = 1000 rev/min, injection at 13º BTC, ignition at 5º BTC.21 590 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 591 within the prechamber of a Comet swirl chamber IDI engine.21 The gas mixture 10 10 rapidly becomes stoichiometric or fuel-rich. Composition nonuniformities across = - 34, 300 the prechamber are substantial. Peak NO concentrations, as expected, corre- = = - 36,700 DI engine spond approximately to locally stoichiometric regions. Because the mixture remains fuel-rich in the prechamber as the burned gases expand (after the time of peak pressure which occurs between 6 and 10º ATC), substantial NO decomposi- tion within the prechamber can occur. However, by this time much of the gas EINO,/EINOr, std EINO,/E NOx, std (and NO) in the prechamber has been transferred to the main chamber where freezing of the NO chemistry will occur. Cylinder-gas dumping experiments, 0.1 0.1 where both main chamber and prechamber gases were dumped and quenched, 02 addition N2 addition IDI engine confirm this description. Cylinder average NO concentrations, determined by o 0.72 dm3 rapidly opening a diaphram which separated the engine cylinder from an evac- A 0.52 dm3 -. Present study uated tank at predetermined points in the cycle of an otherwise normally oper- O Yu and Shahed ated IDI engine, rise rapidly once combustion starts, until the NO chemistry is 0.01 - L 0.01l 3.4 3.6 3.8 4.0 4.2 4.4 4.6 3.4 3.6 3.8 4.0 4.2 4.4 4.6 4.8 effectively frozen at about 15º ATC. Little net NO decomposition occurs.22 Heat- 104 T. , 104/K 104 release-rate diagrams obtained from pressure data analysis for the same IDI , 104/K engine indicate that combustion is only about one-half complete at the time the (a) (b) NO formation process ceases. FIGURE 11-19 Diluents added to the intake air (such as recycled exhaust) are effective at Correlation of NO, emissions index EINo, for a wide range of operating conditions with reciprocal of reducing the NO formation rate, and therefore NO, exhaust emissions. As with stoichiometric mixture flame temperature for: (a) DI engines; (b) IDI engines. Flame temperatures spark-ignition engines, the effect is primarily one of reducing the burned gas tem- varied by addition of different diluents and oxygen.25, 26 Values of EINo, normalized with value at perature for a given mass of fuel and oxygen burned. Figure 11-18 shows the standard conditions. effect of dilution of the intake air with N2, CO2, and exhaust gas on NO, exhaust levels.23 The heat capacity of CO2 (per mole) at the temperatures rele- vant to diesel combustion is about twice that of N2. That of exhaust gas is capacity increases as the concentrations of CO2 and H2O are substantially slightly higher than that of N2. Therefore these data show that the effect is pri- higher. Similar studies in an indirect-injection engine show comparable trends. marily one of reduced burned gas temperatures. Note that the composition of the Addition of diluents [exhaust gas (EGR) and nitrogen] reduce peak flame tem- exhaust gas of a diesel varies with load. At idle, there is little CO2 and H2O, and peratures and NO, emissions; also, addition of oxygen (which corresponds to a the composition does not differ much from that of air. At high load the heat reduction in diluent fraction) increases flame temperatures and therefore increases NO, emissions. 24 Confirmation that NO forms in the close-to-stoichiometric burned gas regions and the magnitude of the stoichiometric burned gas temperature controls 600 NO, emissions is given by the following. Plee et al.25, 26 have shown that the 500 effects of changes in intake gas composition (with EGR, nitrogen, argon, and oxygen addition) and temperature on NO, emissions can be correlated by 400 E EINO, = constant x exp RT (11.19) NOx, ppm 300 T,(kelvin) is the stoichiometric adiabatic flame temperature (evaluated at a suit- 200 FIGURE 11-18 able reference point: fuel-air mixture at top-center pressure and air temperature) Effect of reduction in oxygen concentration by dif- and E is an overall activation energy. Figure 11-19 shows EINo, for a range of 100 CO2 ferent diluents (exhaust gas, CO2, N2) on NO, Exhaust -N2 emissions in DI diesel. Bore = 140 mm, intake air compositions and temperatures, and two DI and two IDI engines for gas stroke == 152 mm, r = 14.3. Speed = 1300 rev/ several loads and speeds, normalized by the engine NO, level obtained for stan- 02 21 20 19 18 17 16 min, fuel rate = 142 mm3/stroke, injection timing dard air, plotted on a log scale against the reciprocal of the stoichiometric adia- Oxygen concentration, vol % at 4º BTC.23 batic flame at TC conditions. A single value of E/R correlates the data over two 592 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 593 orders of magnitude. There is, of course, some scatter since the model used is The levels of CO observed in spark-ignition engine exhaust gases are lower overly simple, and load, speed, and other engine design and operating parameters than the maximum values measured within the combustion chamber, but are affect the process. The overriding importance of the burned gas temperature significantly higher than equilibrium values for the exhaust conditions. Thus the of close-to-stoichiometric mixture is clear, however. processes which govern CO exhaust levels are kinetically controlled. In premixed hydrocarbon-air flames, the CO concentration increases rapidly in the flame zone 11.3 CARBON MONOXIDE to a maximum value, which is larger than the equilibrium value for adiabatic combustion of the fuel-air mixture. CO formation is one of the principal reaction Carbon monoxide (CO) emissions from internal combustion engines are con- steps in the hydrocarbon combustion mechanism, which may be summarized by1 trolled primarily by the fuel/air equivalence ratio. Figure 11-20 shows CO levels RH - R -+ RO2 -+ RCHO -+ RCO -> CO (11.20) in the exhaust of a conventional spark-ignition engine for several different fuel compositions.27 When the data are plotted against the relative air/fuel ratio or where R stands for the hydrocarbon radical. The CO formed in the combustion the equivalence ratio, they are correlated by a single curve. For fuel-rich mixtures process via this path is then oxidized to CO, at a slower rate. The principal CO CO concentrations in the exhaust increase steadily with increasing equivalence oxidation reaction in hydrocarbon-air flames is ratio, as the amount of excess fuel increases. For fuel-lean mixtures, CO concen- CO + OH = CO2 + H (11.21) trations in the exhaust vary little with equivalence ratio and are of order 10-3 The rate constant for this reaction is1 mole fraction. Since spark-ignition engines often operate close to stoichiometric at part T load and fuel rich at full load (see Sec. 7.1), CO emissions are significant and must kto = 6.76 x 1010 exp 1102) cm3/gmol (11.22) be controlled. Diesels, however, always operate well on the lean side of stoichio- It is generally assumed that in the postflame combustion products in a spark- metric; CO emissions from diesels are low enough to be unimportant, therefore, ignition engine, at conditions close to peak cycle temperatures (2800 K) and pres- and will not be discussed further. sures (15 to 40 atm), the carbon-oxygen-hydrogen system is equilibrated. Thus CO concentrations in the immediate postflame burned gases are close to equi- 8 librium. However, as the burned gases cool during the expansion and exhaust 8 strokes, depending on the temperature and cooling rate, the CO oxidation process [reaction (11.21)] may not remain locally equilibrated. ECONOMIAWN- Newhall carried out a series of kinetic calculations for an engine expansion stroke assuming the burned gas at the time of peak cylinder pressure was uniform 6 and in equilibrium.28 Of the reactions important to CO chemistry, only three- A = AIF (A/F) stoich 10 body radical-recombination reactions such as 11 H + H + M = H2 + M (11.23) CO, vol % CO, vol % H + OH + M = H2O + M (11.24) 3- H + 02 + M = HO2 + M (11.25) ." 80 were found to be rate controlling. The bimolecular exchange reactions and the 2 2 CO oxidation reaction (11.21) were sufficiently fast to be continuously equili- brated. Only during the later stages of the expansion stroke was the CO concen- 1 tration predicted to depart from equilibrium, as shown in Fig. 11-21. Using this technique to predict average CO levels at the end of expansion over a range of 12 OL 14 17 0. 0.9 1.0 1.1 1.2 13 15 16 0.8 equivalence ratios (rich to lean), Newhall obtained a good match to experimental Relative air/fuel ratio À data (see Fig. 11-22). The kinetically controlled aspects of the CO emissions (b) (a) mechanism have thus been confirmed. These calculations showed that a partial equilibrium amongst the bimolecu- FIGURE 11-20 Variation of SI engine CO emissions with eleven fuels of different H/C ratio: (a) with air/fuel ratio; (b) lar exchange reactions occurred a posteriori. Analyses based explicitly on this with relative air/fuel ratio 2.27 partial equilibrium assumption (which are considerably simpler) have been 594 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 595 reactions were assumed to be equilibrated. The studies using this simplified kinetic model have confirmed that at peak cylinder pressures and temperatures, 0- equilibration times for CO are faster than times characteristic of changes in burnt Kinetic gas conditions due to compression or expansion. Thus the CO concentration -- rapidly equilibrates in the burnt gases just downstream of the reaction zone fol- lowing combustion of the hydrocarbon fuel. However, it has already been Carbon monoxide, mole fraction pointed out in Sec. 9.2.1 that the burnt gases are not uniform in temperature. Equilibrium Also, the blowdown of cylinder pressure to the exhaust manifold level during the exhaust process and the decrease in gas temperature that accompanies it occupies 10-3L a substantial portion of the cycle-about 60 crank angle degrees. Thus, the FIGURE 11-21 temperature- and pressure-time profiles of parts of the charge at different loca- Results of kinetic calculations of CO con- tions throughout the cylinder differ, depending on when these parts of the charge centrations during expansion stroke following burn and when they exit the cylinder through the exhaust valve and enter the 2 3 5 TC combustion in SI engine; stoichiometric exhaust manifold. Time, ms mixture. 28 The results of an idealized calculation which illustrate these effects are shown in Fig. 11-23. The CO mole fractions in different elements or parts of the carried out. 29, 30 The appropriate three-body atom and radical recombination burnt gas mixture are plotted versus crank angle; x, is the fraction of the total reactions [e.g ., (11.23) to (11.25)] were treated as the rate-limiting constraint on charge which had burned when each element shown burned; z is the mass frac- the total number of particles or moles per unit volume of burnt gases, i.e ., tion which had left the cylinder at the time each element left the cylinder. The partial equilibrium calculations show the burned gases are close to equilibrium until about 60 crank angle degrees after top-center. During the exhaust blow- at dn = [ (Ri - Rt ) (11.26) down process after the exhaust valve opens, gas which leaves the cylinder early V is the volume of the elemental system considered, n is the total number of 3 × 10-1 moles, Ri and Ri are the forward and backward rates for reaction i, and k - z = 0.01 represents the number of three-body recombination reactions included. All other Mb = 0.05 z = 0.50 z = 0.99 CO --- Equilibrium CO 10-2 14 ₣% = 0.50 c = 7 Experimental. ¢ = 1.0 measurements 12+ N = 3000 rev/min ----- - - Calculation , end of expansion 10 Equilibrium at initial point CO, mole fraction = 0.01 - 81 FIGURE 11-23 Carbon monoxide, mole % 10- CO CO concentrations in selected 6 Equilibrium at exhaust --- - elements of SI engine cylinder z = 0.50 charge, which burn at different Exhaust times and which exit the cylinder 4 valve z = 0.01 2 = 0.99 at different times. x, is mass frac- opens tion burned when element FIGURE 11-22 3 × 10-4 Predicted CO concentration at end of expansion rc burned. z is fraction of gas which BC ºATC 30 60 90 120 210 has already left cylinder during stroke, compared with measured exhaust concentra- 150 180 exhaust process prior to element tions, as function of air/fuel ratio. Equilibrium levels TT leaving cylinder. Speed = 3000 8 10 12 14 16 18 at TC combustion and exhaust conditions also 0 5 10 rev/min, rc = 7, equivalence Air/fuel ratio shown. 28 Time, ms ratio = 1.0.30 596 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 597 (z < 1) cools more rapidly than gas which leaves late (z ~ 1). In these calow TABLE 11.4 lations, mixing between gas elements which burn at different times was neglected Hydrocarbon composition of spark-ignition engine exhaust It can be seen that a CO gradient exists across the burned gases and that the Co (by class)33 concentration in the exhaust gases is unlikely to be uniform. Experiments with Carbon, percent of total HC single-cylinder engines support these conclusions that CO is in equilibrium during the combustion process but deviates from equilibrium late in the expan- Paraffins Olefins Acetylene Aromatics sion stroke (e.g ., see Refs. 10 and 31). Without catalyst 33 27 32 Conclusions from these detailed studies are as follows. The measured With catalyst 57 15 2 26 average exhaust CO concentrations for fuel-rich mixtures are close to equilibrium concentrations in the burned gases during the expansion process. For close-to- stoichiometric mixtures, the partial equilibrium CO predictions are in agreement carbon concentration expressed in parts per million carbon atoms (C1).+ While with measurements and are orders of magnitude above CO equilibrium concen- total hydrocarbon emission is a useful measure of combustion inefficiency, it is trations corresponding to exhaust conditions. For fuel-lean mixtures, measured not necessarily a significant index of pollutant emissions. Engine exhaust gases CO emissions are substantially higher than predictions from any of the models contain a wide variety of hydrocarbon compounds. Table 11.4 shows the average based on kinetically controlled bulk gas phenomena. One possible explanation of breakdown by class of the hydrocarbons in spark-ignition engine exhaust gases, this lean-mixture discrepancy is that only partial oxidation to CO may occur of both with and without a catalytic converter, with gasoline fuel. Some of these some of the unburned hydrocarbons emerging during expansion and exhaust hydrocarbons are nearly inert physiologically and are virtually unreactive from from crevices in the combustion chamber and from any oil layers or deposits on the standpoint of photochemical smog. Others are highly reactive in the smog- the chamber walls. producing chemistry. (Some hydrocarbons are known carcinogens; see Sec. While many questions about details of the CO oxidation mechanisms 11.5.2). Based on their potential for oxidant formation in the photochemical smog remain, as a practical matter exhaust emissions are determined by the fuel/air chemistry, hydrocarbon compounds are divided into nonreactive and reactive equivalence ratio. The degree of control achieved within the engine to date has categories. Table 11.5 shows one reactivity scale which has been used to estimate come from improving mixture uniformity and leaning-out the intake mixture. In the overall reactivity of exhaust gas hydrocarbon mixtures. Other scales are used multicylinder engines, because CO increases rapidly as the inlet mixture becomes for the same purpose.34 Scales that assign high values for reactivity to the olefins richer than stoichiometric, cylinder-to-cylinder variations in equivalence ratio (like Table 11.5), which react most rapidly in the photochemical smog reaction, about the mean value are important; nonuniform distribution can significantly probably best approximate smog-formation potential near the sources of hydro- increase average emissions. Thus improved cylinder-to-cylinder fuel/air ratio dis- carbon pollution. The simplest scale, which divides the HC into two classes tribution has become essential. Also, because it is necessary to enrich the fuel-air methane and nonmethane hydrocarbons probably best approximates the end mixture when the engine is cold, CO emissions during engine warm-up are much result for all HC emissions. All hydrocarbons except methane react, given enough higher than emissions in the fully warmed-up state. Further, in transient engine time. More detailed breakdowns of the composition of spark-ignition and diesel operation during acceleration and deceleration, control of fuel metering has had engine exhaust HC are available in the literature.33, 35 to be improved. Additional reductions in CO beyond what can be achieved in the Fuel composition can significantly influence the composition and magni- engine are possible with exhaust treatment devices, which are reviewed in Sec. tude of the organic emissions. Fuels containing high proportions of aromatics 11.6. Oxidation of CO in the exhaust system without use of special exhaust treat- and olefins produce relatively higher concentrations of reactive hydrocarbons. ment devices does not occur to any significant degree because the exhaust gas However, many of the organic compounds found in the exhaust are not present temperature is too low (Fig. 11-23 shows that the CO oxidation reactions effec- tively freeze as the gas passes through the exhaust valve). 11.4 UNBURNED HYDROCARBON + This is because the standard detection instrument, a flame ionization detector (FID), is effectively a EMISSIONS carbon atom counter: e.g ., one propane molecule generates three times the response generated by one methane molecule. Some data in the literature are presented as ppm propane (C3), or ppm hexane 11.4.1 Background (C6); to convert to ppm C1 multiply by 3 or by 6, respectively. Older measurements of hydrocarbon emissions were made with nondispersive infrared (NDIR) detectors which had different sensitivities Hydrocarbons, or more appropriately organic emissions, are the consequence of for the different hydrocarbon compounds. For gasoline-fueled engines, HC emissions determined by incomplete combustion of the hydrocarbon fuel. The level of unburned hydrocar FID analyzers are about twice the levels determined by NDIR analyzers,32 though this scaling is not bons (HC) in the exhaust gases is generally specified in terms of the total hydro- exact. 598 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 599 TABLE 11.5 11.4.2 Flame Quenching and Oxidation Reactivity of classes of hydrocarbons Fundamentals Hydrocarbons Relative reactivityt Flame quenching or extinction occurs at the walls of engine combustion cham- bers. The cool walls of the chamber act as a sink for heat and the active radical C1-C4 paraffins Acetylene 0 species generated within the flame. Quenching of the flame occurs under several Benzene different geometrical configurations: the flame may be propagating normal to, parallel to, or at an angle to the wall; the flame may quench at the entrance to a C4 and higher molecular weight crevice-a thin volume with a narrow entrance to the combustion chamber such paraffins as the region between the piston crown and the cylinder wall. When the flame Monoalkyl benzenes 2 Ortho- and para-dialkyl benzenes quenches, it leaves a layer or volume of unburned mixture ahead of the flame. Cyclic paraffins (Whether this results in unburned hydrocarbon emissions depends on the extent to which these quench region hydrocarbons can subsequently be oxidized.) Ethylene Flame-quenching processes are analyzed by relating the heat release within Meta-dialkyl benzenes 5 Aldehydes the flame to the heat loss to the walls under conditions where quenching just occurs. This ratio, a Peclet number (Pe), is approximately constant for any given 1-Olefins (except ethylene) geometrical configuration. The simplest configuration for study is the two-plate Diolefins 10 quench process, where the minimum spacing between two parallel plates through Tri- and tetraalkyl benzenes which a flame will propagate is determined. The Peclet number for this two-plate Internally bonded olefins 30 configuration is given by: Internally bonded olefins with Pe, = PuSLCp.Ts - T.)_ P. SLCp.sdg2 k,(T, - T)/dq2 ks (11.27) substitution at the double bond 100 Cycloolefins which is approximately constant over a wide range of conditions. p, SL, cp, T, + General Motors Reactivity Scale (0-100). Based on the NO, for- and k are the density, laminar flame speed, specific heat at constant pressure, gas mation rate for the hydrocarbon relative to the NO, formation rate temperature, and thermal conductivity, respectively, with the subscripts u and f for 2,3-dimethyl-2-benzene.34 referring to unburned and flame conditions. de2 is the two-plate quench distance. The wall temperature and unburned gas temperature are assumed to be equal; this assumption is also appropriate in the engine context since there is ample in the fuel, indicating that significant pyrolyses and synthesis occur during the time during the compression stroke for a thermal boundary layer to build up to a combustion process. thickness of at least the quench distance. Oxygenates are present in engine exhaust, and are known to participate in Lavoie36 has developed empirical correlations for two-plate quench- the formation of photochemical smog. Some oxygenates are also irritants and distance data for propane-air mixtures: only limited data for liquid hydrocarbon odorants. The oxygenates are generally categorized as carbonyls, phenols, and fuels such as isooctane are available. The data in the pressure range 3 to 40 atm other noncarbonyls. The carbonyls of interest are low molecular weight alde- are well fitted by hydes and aliphatic ketones. The volatile aldehydes are eye and respiratory tract 9.5 0.26 min ( 1 , 1/02 ) irritants. Formaldehyde is a major component (<20 percent of total carbonyls). Pe2 (11.28) Carbonyls account for about 10 percent of the HC emissions from diesel pas- senger car engines, but only a few percent of spark-ignition engine HC emissions. where p is the pressure in atmospheres and o is the fuel/air equivalence ratio. The Phenols are odorants and irritants: levels are much lower than aldehyde levels. two-plate quench distance de2 is then obtained from Eq. (11.27) and Prandtl Other noncarbonyls include methanol, ethanol, nitromethane, methyl formate. number and viscosity relations for the flame conditions (see Sec. 4.8 or Ref. 36). Whether these are significant with conventional hydrocarbon fuels is unclear.35 Thus the minimum size crevice region into which a flame will propagate can be Use of alcohol fuels increases oxygenate emissions. Both methanol and aldehyde determined. emissions increase substantially above gasoline-fueled levels with methanol-fueled For the process of a flame front quenching on a single wall, there are many spark-ignition engines. possible geometries. The simplest is where the flame front is parallel to the wall 600 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 601 and approaches it head on. The one-wall quench distance d.1, defined as the from the hot reaction zone to the wall and heat released in the reaction zone by position of closest approach of the reaction zone to the wall, scales with flam the flame reactions. The second step is the postquench diffusion and oxidation properties in a similar way to the two-plate quench distance. Thus, a one-wall occurring on a time scale of one or a few milliseconds after quenching. The diffu- Peclet number relation can be formed: sion and oxidation process ultimately reduces the mass of wall quench hydrocar- bons to several orders of magnitude below its value at the time of quenching. Per = DuSL Cpu dal ~ 8 Closed-vessel combustion experiments have also been used to show that oil ku layers on the walls of the bomb cause an increase in residual unburned HC levels where the subscript u denotes properties evaluated at unburned gas conditions. after combustion is complete. The additional HC that result in experiments with Using the wall temperature as representative of the unburned gas tem- oil films present are primarily (>95 percent) fuel molecules, and are directly pro- perature (because the thermal boundary-layer thickness is greater than typical portional to the amount of oil placed on the walls of the reactor and the solu- quench distances), Lavoie showed that bility of the specific fuel in the oil. These results show that absorption of fuel in dal _ Per = 0.2 the oil occurs prior to ignition. This dissolved fuel is then desorbed into the dq2 (11.30) burned gases well after combustion is complete. Thus fuel absorption into and Pe2 desorption from any oil layers is a potentially important engine HC mecha- is a reasonable fit to the single-wall quench data. Typical two-wall quench dis- mism. 39 tances for spark-ignition engine conditions are 0.2 to 1 mm; these distances rep- resent the minimum crevice size the flame will enter. Single-wall quench distances are, therefore, in the range 0.04 to 0.2 mm. 11.4.3 HC Emissions from Spark-Ignition While a fraction of the fuel hydrocarbons can escape the primary com- Engines bustion process unburned or only partially reacted, oxidation of some of these Unburned hydrocarbon levels in the exhaust of a spark-ignition engine under hydrocarbons can occur during the expansion and exhaust processes. Hydrocar- normal operating conditions are typically in the range 1000 to 3000 ppm C1. This bon oxidation rates have been determined in a number of different studies and corresponds to between about 1 and 24 percent of the fuel flow into the engine; several different empirical correlations of the data in the form of overall reaction the engine combustion efficiency is high. As indicated in Fig. 11-2, HC emissions rate equations have been proposed. A reasonable fit to the experimental data on rise rapidly as the mixture becomes substantially richer than stoichiometric. unburned HC burnup is the rate expression:36 When combustion quality deteriorates, e.g ., with very lean mixtures, HC emis- d[HC] 2 sions can rise rapidly due to incomplete combustion or misfire in a fraction of the = - 6.7 x 1015 exp - 18,735 &HC XO2 (11.31) T engine's operating cycles. As outlined in Sec. 11.1, there are several mechanisms dt that contribute to total HC emissions. Also, any HC escaping the primary com- where [ ] denotes concentration in moles per cubic centimeter, Hc and xo, are bustion process may oxidize in the expansion and exhaust processes. While a the mole fractions of HC and O2, respectively, t is in seconds, T in kelvins, and complete description of the HC emissions process cannot yet be given, there are the density term (p/RT) has units of moles per cubic centimeter. The spread in sufficient fundamental data available to indicate which mechanisms are likely to the data about this equation is substantial, however. be most important, and thus how major engine variables influence HC emission Studies of combustion of premixed fuel-air mixtures at high pressure in levels. closed vessels or bombs have been useful in identifying the mechanisms by which Four possible HC emissions formation mechanisms for spark-ignition hydrocarbons escape complete combustion. The residual unburned hydrocarbons engines (where the fuel-air mixture is essentially premixed) have been proposed: left in the bomb following a combustion experiment have been shown to come (1) flame quenching at the combustion chamber walls, leaving a layer of primarily from crevices in the bomb walls. Unburned HC levels were proportion- unburned fuel-air mixture adjacent to the wall; (2) the filling of crevice volumes al to total crevice volume, and decreased to very low values (~10 ppm C) as all with unburned mixture which, since the flame quenches at the crevice entrance, the crevices were filled with solid material. Thus wall quench hydrocarbons escapes the primary combustion process; (3) absorption of fuel vapor into oil apparently diffuse into the burned gases and oxidize following the quenching layers on the cylinder wall during intake and compression, followed by desorp- event. 37 Analytical studies of the flame quenching process, and postquench diffu- tion of fuel vapor into the cylinder during expansion and exhaust; (4) incomplete sion and oxidation with kinetic models of the hydrocarbon oxidation process, are combustion in a fraction of the engine's operating cycles (either partial burning in agreement with these bomb data.38 Flame quenching can be thought of as a or complete misfire), occurring when combustion quality is poor (e.g ., during two-stage process. The first step is the extinction of the flame at a short distance engine transients when A/F, EGR, and spark timing may not be adequately from the cold wall, determined by a balance between thermal conduction of heat controlled). In addition, as deposits build up on the combustion chamber walls, 602 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 603 7000F unburned HC with the bulk cylinder gases occurs during expansion and/or the 100 exhaust blowdown process. Then, the final stages of piston motion during the 6000 exhaust stroke push most of the remaining fraction of the cylinder mass with its high HC concentration into the exhaust. This would be expected to leave a high 180 5000- concentration of HC in the residual gas in the cylinder. Experiments conducted Mass flow rate in which the valve mechanism of a single-cylinder engine was arranged to disen- Exhaust 4000 Exhaust Concentrations valve 60 gage during operation and trap residual gases in the cylinder confirm this. For Hydrocarbon concentration, ppm C valve at exhaust valve closes Hydrocarbon mass flow rate, mg/s one set of typical engine operating conditions, approximately one-third of the opens 3000 hydrocarbons left unburned in an engine combustion event was retained in the cylinder and recycled.42 2000F 20 FLAME QUENCHING AT THE WALLS. The existence of quench layers on the 1000F cold combustion chamber walls of a spark-ignition engine was shown photo- graphically by Daniel.43 Photographs of the flame region immediately after flame arrival at the wall through a window in the cylinder head showed a thin non- 100 140 180 220 260 300 340 3.80 radiating layer adjacent to the wall. The quench layer thicknesses measured were Crank angle, deg in the range 0.05 to 0.4 mm (thinnest at high load), in rough agreement with FIGURE 11-24 Variation in HC concentration and HC mass flow rate at the exhaust valve during the exhaust predictions based on experiments in combustion bombs. However, more recent process. SI engine operating at 1200 rev/min and o = 1.2, unthrottled.40 work in bombs and engines indicates that diffusion of hydrocarbons from the quench layer into the burned gases and its subsequent oxidation occur on a time scale of a few milliseconds, at least with smooth clean combustion chamber walls. HC emissions increase. Whether the deposits constitute an additional mechanism The constant-volume combustion bomb data which suggested this conclusion or merely modify one of the above mechanisms is unclear. and the kinetic calculations which support this explanation of why quench layers All these processes (except misfire) result in unburned hydrocarbons close are not significant with smooth clean walls have already been described in Sec. to the combustion chamber walls, and not in the bulk of the cylinder gases. Thus, 11.4.2. The following evidence shows these conclusions are also valid in an the distribution of HC in the exhaust gases would not be expected to be uniform. engine. Experiments have been done to determine the unburned HC concentration dis- A special rapid-acting poppet valve was used in a single-cylinder engine to tribution in the exhaust port during the exhaust process to provide insight into sample the gases from a torus-shaped region, of height of order 0.25 mm and the details of the formation mechanisms. Gas concentrations were measured with diameter about 6 mm, adjacent to the wall over a 1-ms period. Sampling was a rapid-acting sampling valve placed at the exhaust port exit. Figure 11-24 shows repeated every cycle to provide a steady stream of sampled gases for analysis. results from these time-resolved HC concentration measurements. HC concentra- Figure 11-25 shows the variation in concentrations of HC species through the tions vary significantly during the exhaust process. Gas which remains in the combustion, expansion, and exhaust processes. The fuel was propane (C3H8). The exhaust port between exhaust pulses has a high HC concentration, so purging fuel concentration drops rapidly to a low value when the flame arrives at the techniques where air or nitrogen was bled into the exhaust port were used to valve; at the same time, intermediate hydrocarbon product concentrations rise displace this high HC gas while the exhaust valve was closed. The high HC and then fall sharply to values below 1 ppm. Beginning at 60º ATC, all HC concentration in the blowdown exhaust gases is clearly discernible, as is the rapid concentrations rise and vary somewhat during the remainder of the cycle in a rise in HC concentration toward the end of the exhaust stroke. The cylinder-exit way that depends strongly on engine operating conditions. The observed rapid HC concentrations were then multiplied by the instantaneous exhaust gas mass rise in partial oxidation products immediately after flame arrival is consistent flow rate to obtain the instantaneous HC mass emission rate from the cylinder with the flame quenching short of the wall. The presence of CH2O and CH3CHO throughout the exhaust process, also shown in Fig. 11-24. The unburned HC are. in significant quantities indicates that low-temperature oxidation processes are exhausted in two peaks of approximately equal mass: the first of these coincides occurring. However, since all HC product concentrations fall rapidly within 2 ms with the exhaust blowdown mass flow pulse (which removes the majority of the of flame arrival to very low values, the unburned HC in the quench layer diffuse mass from the cylinder); the second occurs toward the end of the exhaust stroke into the bulk burned gases and oxidize. The increase in HC concentrations later where HC concentrations are very high and the mass flow rate is relatively low.40 in the cycle results from the sampling of hydrocarbons from sources other than Other experiments have confirmed these observations.41 Clearly, mixing of quench layers.44 604 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 605 Flame EVO EVC flame arrives at each crevice, it can either propagate into the crevice and fully or partially burn the fuel and air within the crevice or it can quench at the crevice 104. entrance. Whether the flame quenches depends on crevice entrance geometry, the composition of the unburned mixture, and its thermodynamic state as described in Sec. 11.4.2. After flame arrival and quenching, burned gases will flow into each 103 Exhaust crevice until the cylinder pressure starts to decrease. Once the crevice gas pres- concentration sure is higher than the cylinder pressure, gas flows back from each crevice into Mole fractions, ppm 102 C3H8 FIGURE 11-25 the cylinder. C2H4 Concentrations (mole fractions) of selected The most important of these crevices, the volumes between the piston, pis- C3H6 hydrocarbons adjacent to combustion ton rings, and cylinder wall, is shown schematically in Fig. 8-27. This crevice 10 CH3CHO C2H2 chamber wall, as a function of crank angle consists of a series of volumes, connected by flow restrictions such as the ring side 5 CH4 TTTT during combustion, expansion, and exhaust processes. Mass sampled with rapid-acting clearance and ring gap whose geometry changes as the ring moves up and down valve held constant at 7.6 x 10-6 g per pulse. in the ring groove sealing either at the top or bottom ring surface. The gas flow, Total exhaust HC = 400 ppm C. Engine pressure distribution, and ring motion are therefore coupled, and their behavior speed = 1250 rev/min, imep = 380 kPa, during the compression and expansion strokes has already been discussed in Sec. 100 200 300 equivalence ratio = 0.9, MBT spark timing, 8.6. During compression and combustion, mass flows into the volumes in this Crank angle, deg ATC no EGR.44 total crevice region. Once the cylinder pressure starts to decrease (after about 15º ATC) gas flows out of the top of these crevice regions in Fig. 8-27 into the cylin- der at low velocity adjacent to the cylinder wall. The important result is that the Though quench layers on clean smooth combustion chamber walls are not fraction of the total cylinder charge (5 to 10 percent) trapped in these regions at a significant source of unburned hydrocarbons, it has been shown that wall the time of peak cylinder pressure escapes the primary combustion process. Most surface finish does affect exhaust HC levels. Comparisons have been made of this gas flows back into the cylinder during the expansion process. Depending between the standard "rough" as-cast cylinder head surfaces and the same cylin- on spark plug location in relation to the position of the top ring gap, well above der heads when smoothed. The average exhaust HC concentration decreased by 50 percent of this gas can be unburned fuel-air mixture. Its potential contribution 103 ppm C, or 14 percent; the smoothed surface area was 32 percent of the total to unburned HC emissions is obvious. combustion chamber surface area.45 Buildup of deposits on the combustion There is substantial evidence to support the above description of crevice chamber surfaces also affect HC emission levels, as will be discussed later. HC phenomena and the piston ring crevice region in particular. Visualization studies in a special engine have identified the spark plug crevice outflow, low- CREVICE HC MECHANISM. The crevices in the combustion chamber walls- velocity gas expansion out of the volume above the first ring after the time of small volumes with narrow entrances-into which the flame is unable to pen- peak pressure, and the jet-like flows through the top ring gap later in the expan- etrate have been shown to be a major source of unburned HC. The largest of sion process when the pressure difference across the ring changes sign.46 Gas these crevice regions is the volumes between the piston, piston rings, and cylinder sampling from the volume above the top ring, using a rapid-acting sample valve wall. Other crevice volumes in production engines are the threads around the mounted in the piston crown, has shown that the gas composition in this region spark plug, the space around the plug center electrode, crevices around the intake corresponds to unburned fuel-air mixture until flame arrival at the crevice and exhaust valve heads, and the head gasket crevice. Table 8.1 shows the size entrance closest to the sampling valve location. Next, product gases enter the and relative importance of these crevice regions in one cylinder of a production crevice as the cylinder pressure continues to rise. Then, during expansion as gas V-6 engine determined from measurements of cold-engine components. Total flows out of this region, the composition of the gas sampled reverts back toward crevice volume is a few percent of the clearance volume and the piston and ring that of the unburned mixture which enters the crevice region earlier.47 pack crevices are the dominant contributor. Direct evidence that the piston and ring crevice regions are a major contrib- The important crevice processes occurring during the engine cycle are the utor to exhaust HC emissions comes from experiments where the volume of this following. As the cylinder pressure rises during compression, unburned mixture is crevice region was substantially changed. Wentworth48 almost completely elimi- forced into the crevice regions. Since these volumes are thin they have a large nated this crevice by moving the top piston ring as close to the crown of the surface/volume ratio; the gas flowing into each crevice cools by heat transfer to piston as possible, and sealing this ring at top and bottom in its groove with O close to the wall temperature. During combustion, while the pressure continues rings. Tests of this sealed ring-orifice design in a production engine showed to rise, unburned mixture continues to flow into the crevice volumes. When the reductions of between 47 and 74 percent from baseline HC levels over a range of 606 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 607 6000r 1250 HC, ppm C| Production pistons 1000- Top-land volume O and rings 4000} O Clean Cylinder wall 20 h 750 Exhaust HC, ppm C, O -O 000 500 Piston 250 0.1 0.2 0.3 0.4 FIGURE 11-27 Clearance o, mm Effect of increasing crankcase blowby on exhaust 0 hydrocarbon emissions. Production pistons and (a) (b) 0 0.05 0.1 0.15 0.2 rings. SI engine at 1200 rev/min, intake manifold FIGURE 11-26 Blowby flow, dm3/s pressure 0.6 atm, A/F = 14.2.51 (a) Piston top-land crevice volume. (b) Effect of increasing top-land clearance on exhaust hydrocarbon emissions. Unthrottled spark-ignition engine, re == 6, 885 rev/min, A/F = 13, MBT timing.49 piston-bore-ring assembly in response to combustion chamber pressure. Blowby speeds and loads. Haskell and Legate,49 in experiments in a single-cylinder CFR of gases from the cylinder to the crankcase removes gas from this crevice region engine, steadily increased the piston top-land clearance (see Fig. 11-26a) and mea- and thereby prevents some of the crevice gases from returning to the cylinder. sured the effect on exhaust HC emissions. Figure 11-26b shows the results: HC Crankcase blowby gases used to be vented directly to the atmosphere and consti- emissions increase as the top-land clearance increases until the clearance equals tuted a significant source of HC emissions. The crankcase is now vented to the about 0.18 mm, when emissions drop to the zero clearance level. This clearance engine intake system, the blowby gases are recycled, and this source of HC emis- (0.18 mm) is close to the two-plate quench distance estimated from Eq. (11.27). sions is now fully controlled. Blowby at a given speed and load is controlled For piston top-land clearances above this value, the flame can enter the crevice primarily by the greatest flow resistance in the flow path between the cylinder and burn up much of the crevice HC. and the crankcase. This is the smallest of the compression ring ring-gap areas. The relative importance of the different crevices in the combustion chamber Figure 8-30 shows how blowby increases linearly with the smallest gap area. walls has been examined by using the cylinder head and piston of a four-cylinder Figure 11-27 shows how exhaust HC levels decrease as blowby increases and production engine to form two constant-volume reactors or combustion more crevice HC flows to the crankcase. Crankcase blowby gases represent a bombs.50 The cylinder head was sealed with a steel plate across the head gasket direct performance loss. They are a smaller efficiency loss because crankcase plane to make one reactor; the piston and ring pack and cylinder wall, again gases are now recycled to the engine intake system. sealed with a plate at the head gasket plane, formed the second reactor. Each The location of the ring gap in relation to the spark plug also affects HC reactor was filled with a propane-air mixture and combustion initiated with a emission levels. Experiments have shown that HC emissions are highest when the spark discharge across a spark plug; following combustion the burned gases were top ring gap is farthest from the spark plug; the gas flowing into the crevice exhausted, sampled, and analyzed. The crevices were sequentially filled with directly above the gap is then unburned mixture for the longest possible time. epoxy or viton rubber, and after filling each crevice, the exhaust HC emission With the top ring gap closest to the spark plug, HC exhaust levels are lowest level determined. It was found that the ring pack crevices produced approx- because burned gas reaches the gap location at the earliest time in the com- imately 80 percent of the total scaled HC emissions, the head gasket crevice bustion process. The difference, highest to lowest, was between 9 and 42 percent about 13 percent, and the spark plug threads 5 percent. All other HC sources in of the average level for any set of operating conditions, and in most cases was these reactors produced less than 2 percent of the total HC. While these numbers above 20 percent.51 cannot be applied directly to an operating engine (the crevice filling and empty- The fate of these crevice HC when they flow back into the cylinder during ing rates in the bomb experiments are substantially different from these rates in expansion and exhaust is not well understood. Both jet-like flows (e.g ., that from an engine), they do underline the importance of the ring pack crevice region. the ring gap) and low-velocity creeping flows (e.g ., that from the piston top-land Blowby is the gas that flows from the combustion chamber, past the piston crevice) have been observed (see Fig. 8-29). While the former could mix rapidly and into the crankcase. It is forced through any leakage paths afforded by the with the high-temperature bulk burned gases, the latter will enter the cool gases 608 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 609 TABLE 11.6 Some of the desorbed fuel vapor will mix with the high-temperature combustion Amount of gas flowing into and out of crevice regionst products and oxidize. However, desorbed vapor that remains in the cool bound- % mass ppm C ary layer or mixes with the cooler bulk gases late in the cycle may escape full oxidation and contribute to unburned HC emissions. Total gas in all crevice regions 8.2 Experiments, where measured amounts of oil were placed on the piston Total gas back to combustion chamber 7.0 crown, confirm that oil layers on the combustion chamber surface increase Unburned back to combustion chamber 3.7-7.0¢ 5000-9400 Unburned to blowby 0.5-1.2¢ exhaust HC emissions. The exhaust HC levels increased in proportion to the Total unburned escaping primary combustion 4.2-8.2¢ amount of oil added when the engine was fueled with isooctane. Addition of 0.6 cm3 of oil produced an increase of 1000 ppm C in exhaust HC concentration. + For V-6 engine operating at 2000 rev/min and wide-open throttle. Fuel and fuel oxidation species, not oil oxidation products, were responsible for t Depends on spark plug and ring gap location. most of this increase. Similar experiments performed with propane fuel showed no increase in exhaust HC emissions when oil was added to the cylinder. The in the cylinder wall boundary layer and mix and (probably) burn much more increase in exhaust HC is proportional to the solubility of the fuel in the oil. The slowly. Hydrocarbon transport and oxidation processes are discussed more fully exhaust HC levels decreased steadily back to the normal engine HC level before below. oil addition, over a period of several minutes. At higher coolant temperatures, the Table 11.6 presents a summary of estimates of the total mass of gas and increase in HC on oil addition is less, and HC concentrations decreased back to mass of unburned mixture in the piston, ring, and cylinder wall crevice region for the normal level more quickly. Increasing oil temperature would decrease vis- a typical spark-ignition engine.46 When compared to exhaust HC levels, it is cosity, increasing the rate of drainage into the sump. It also changes the solubility clear that these crevices are a major source of unburned hydrocarbons. . and diffusion rate of the fuel in the oil. 52 At the outer surface of the oil layer, the concentration of fuel vapor dis- ABSORPTION AND DESORPTION IN ENGINE OIL. The presence of lubricating solved in the oil is given by Henry's law for dilute solutions in equilibrium: oil in the fuel or on the walls of the combustion chamber is known to result in an increase in exhaust hydrocarbon levels. In experiments where exhaust HC con- .PL H (11.32) centrations rose irregularly with time, with engine operating conditions nomin- ally constant, it was shown that oil was present on the piston top during these where xs is the mole fraction of fuel vapor in the oil, ps is the partial pressure of high emission periods. When engine oil was added to the fuel, HC emissions fuel vapor in the gas, and H is Henry's constant. If the oil layer is sufficiently thin, increased, the amount of additional HC in the exhaust increasing with the and hence diffusion sufficiently rapid, Eq. (11.32) can be used to estimate the mole increasing amount of oil added. The increase in exhaust HC was primarily unre- fraction of the fuel dissolved in the oil. Since p = ny RT/V (where ny, is the acted fuel (isooctane) and not oil or oil-derived compounds.51 The increase in number of moles of fuel in the cylinder, T is the temperature, and V the cylinder HC can be substantial: exhaust HC levels from a clean engine can double or volume) and x = ns.0/(ns. + no) = nr.o/n, for no >> ns. (where n ., is the number triple when operated on a fuel containing 5 percent lubricating oil over a period of moles of fuel dissolved in the oil and n, is the number of moles of oil),53 then of order 10 minutes. (With deposits from leaded-fuel operation present on the n. RT combustion chamber walls, however, a much smaller increase in exhaust HC was nfc HV (11.33) observed.) It has been proposed that fuel vapor absorption into and desorption from oil layers on the walls of the combustion chamber could explain these phe- Diffusion is sufficiently rapid for Eq. (11.33) to be valid if the diffusion time con- nomena.49 stant ta is much less than characteristic engine times : i.e ., The absorption and desorption mechanism would work as follows. The fuel 82 vapor concentration within the cylinder is close to the inlet manifold concentra- tion during intake and compression. Thus, for about one crankshaft revolution, any oil film on the walls will absorb fuel vapor. During the latter part of com- where ô is the oil layer thickness, D is the diffusion coefficient for fuel vapor in the pression, the fuel vapor pressure is increasing so, by Henry's law, absorption will oil, and N is engine speed. D for a hydrocarbon through a motor oil is of order continue even if the oil was saturated during intake. During combustion the fuel 10 6 cm2/s at 300 K and of order 10-5 cm2/s at 400 K. Oil film thicknesses on vapor concentration in the bulk gases goes essentially to zero so the absorbed the cylinder wall vary during the operating cycle between about 1 and fuel vapor will desorb from the liquid oil film into the gaseous combustion pro- 10 um.54, 55 Thus diffusion times for engine conditions are 10~1 to 10-3 s; for ducts. Desorption could continue throughout the expansion and exhaust strokes. the thinnest oil layers approximate equilibration would be achieved. A theoretical 610 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 611 operating conditions. As the cylinder pressure falls during the expansion stroke, the temperature of the unburned mixture ahead of the flame decreases. This slows the burning rate [the laminar flame speed decreases so the burning rate in Eq. O-ring piece (9.52) decreases]. If the pressure and temperature fall too rapidly, the flame can .. 8 Gap seal be extinguished. This type of bulk quench has been observed in spark-ignition Orifice Production rings engines; it results in very high HC concentrations for that particular cycle. 1.6 Cylinder wall Engine conditions where bulk quenching is most likely to occur are at idle and 1.4 Top compression ring light load where engine speed is low and the residual gas fraction is high, with 1.2 high dilution with excessive EGR or overly lean mixtures, and with substantially retarded combustion. Even if steady-state engine calibrations of A/F, EGR, and 1.0¢ spark-timing are such that bulk quenching does not occur, under transient engine Exhaust HC emission rate, mg/s 0.8 O-rings Inter-ring crevice operation these variables may not be appropriately set to avoid bulk quenching Piston in some engine cycles due to the different dynamic characteristics of the engine 0.6 subsystems which control these variables. 0.4 Sealed-ring designs The existence of zones of stable and unstable engine operation with lean or Second compression ring dilute mixtures has already been discussed (see Sec. 9.4.3). Detailed engine com- 0.2 bustion studies have shown that, as mixture composition becomes more dilute 0.27 0.4 0.6 0.8 1.2 1.4 (e.g ., by increasing EGR) and unburned gas temperature and pressure during Oil consumption rate, mg/s Sealed-ring orifice ring design combustion become lower, combustion quality (or variability) and engine stabil- ity deteriorate. The standard deviation in a parameter such as indicated mean FIGURE 11-28 Correlation between exhaust hydrocarbon emissions and oil consumption rate. Production piston effective pressure (which depends for its magnitude on the proper timing of the rings and sealed ring-orifice ring designs. SI engine at 1600 rev/min, imep == 422 kPa, equivalence start of combustion and on the duration of the combustion process) increases due ratio = 0.9, r = 8.0, intake pressure = 54 kPa, MBT spark timing.57 first to an increase in the number of slower burning cycles, then as conditions worsen to the occurrence of partial burning cycles, and finally to some misfiring study of this problem-the one-dimensional cyclic absorption and desorption of cycles. Figure 9-36 showed how unburned hydrocarbon emissions from a spark- a dilute amount of gas in a thin (constant thickness) isothermal liquid layer where ignition engine rise as the EGR rate is increased at constant load and speed, and diffusion effects are important-has been carried out. It suggests that oil layers combustion quality (defined by the ratio of standard deviation in imep to the on the cylinder wall could be a significant contributor to HC emission levels. 56 average imep) deteriorates. Initially the increase in HC is modest and is caused Correlations between engine oil consumption and exhaust HC emissions by changes in the other HC emission mechanisms described above. However, provide a perspective on the relative importance of oil absorption/desorption and when partial burning cycles are detected, HC emissions rise more rapidly due to crevice mechanisms. Wentworth measured oil consumption and HC emissions in incomplete combustion of the fuel in the cylinder in these cycles. When misfiring a spark-ignition engine for a range of piston ring designs.57 Some of these designs cycles-no combustion-occur the rise in HC becomes more rapid still. were of the sealed ring-orifice type which effectively eliminates all the crevices The relative importance of bulk gas quenching in a fraction of the engine's between the piston, piston rings, and cylinder, and prevents any significant gas operating cycles due to inadequate combustion quality as a source of HC, com- flow into or out of the ring region. HC emissions increase with increasing oil pared with the other sources described in this section, has yet to be established. consumption for both production ring designs and the sealed ring-orifice designs, However, one obvious technique for reducing its importance, burning the as shown in Fig. 11-28. Extrapolation to zero oil consumption from normal con- mixture faster so that combustion is completed before conditions conducive to sumption levels shows a reduction in exhaust HC levels, but this decrease is slow and partial burning exist in the cylinder, does reduce engine exhaust HC significantly less than the difference in emission levels between the production emissions. Figure 11-29 shows a comparison of HC emissions from a moderate and the sealed ring-orifice designs which effectively remove the major crevice burn rate engine with HC emissions with a faster burn rate [i.e ., with improved region. The production piston used had a chamfered top land. The HC emissions combustion quality-lower coefficient of variation in imep, COVimep, Eq. (9.50)], for a normal piston top-land design would probably be higher. achieved by the use of two spark plugs instead of one.58 The exhaust measure- ments show lower HC emissions when significant amounts of EGR are used for POOR COMBUSTION QUALITY. Flame extinction in the bulk gas, before all of NO, control for the faster, and hence less variable, combustion process. Such the flame front reaches the wall, is a source of HC emissions under certain engine evidence suggests that occasional partial burning cycles may occur, even under 612 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 613 30 80 150- EV IV EV IV EV 60 20 100 'imep> % Operating limits NO3, 8/h Moderate burn El 40+ HC, g/h COV. 10 Fast burn 50 20 10 20 30 40 10 20 30 40 10 20 30 40 EGR rate, % Boundary Vortex FIGURE 11-29 Expanded layer Effect of increasing burn rate on tolerance to recycled exhaust gas (EGR) and HC and NO, emissions ring crevice levels. COVimen defined by Eq. (9.50). SI engine at 1400 rev/min, 324 kPa imep, equivalence material - Vortex - ratio = 1.0, MBT timing. 58 Piston Boundary conditions where combustion appears to be "normal," and that this mechanism layer is important in practice. Ring crevice EFFECT OF DEPOSITS. Deposit buildup on the combustion chamber walls (which occurs in vehicles over several thousands of miles) is known to increase (a) (b) (c) unburned HC emissions. With leaded gasoline operation, the increase in HC FIGURE 11-30 emissions varies between about 7 and 20 percent. The removal of the deposits Schematic of flow processes by which ring crevice HC and HC desorbed from cylinder wall oil film results typically in a reduction in HC emissions to close to clean engine levels. exit the cylinder: (a) exhaust blowdown process; (b) during exhaust stroke; (c) end of exhaust stroke.60 With unleaded gasoline, while the deposit composition is completely different (carbonaceous rather than lead oxide), the increase in HC emissions with accu- mulated mileage is comparable. Soft sooty deposits, such as those which accumu- sion and exhaust strokes can transport unburned HC into the bulk gases, most of late after running the engine on a rich mixture, also cause an increase in HC the HC will remain near the wall. Two mechanisms by which gas near the cylin- emissions. Again, when the deposits were removed the emission rate fell about 25 der wall exits the cylinder have been demonstrated. One is entrainment in the percent to the original level.59 Studies with simulated deposits (pieces of metal- vigorous gas flow out of the cylinder which occurs during the exhaust blowdown foam sheet 0.6 mm thick) attached to the cylinder head and piston also showed process. The other is the vortex generated in the piston crown-cylinder wall corner during the exhaust stroke. increases in HC emissions. The increase varied between about 10 and 100 ppm C/cm2 of simulated deposit area. The effect for a given area of deposit Figure 11-30 illustrates these flow processes. In Fig. 11-30a the engine cylin- varied with deposit location. Locations close to the exhaust valve, where the flow der is shown as the exhaust valve opens during the blowdown process. At this direction during the exhaust process would be expected to be directly into the time the unburned HC from the ring crevice regions, laid along the wall during exhaust port, showed the highest increase in emissions.45 expansion (and possibly HC from the oil film on the cylinder wall), is expanding It is believed that absorption and desorption of hydrocarbons by these into the cylinder as the cylinder pressure falls. Some of this material will be surface deposits is the mechanism that leads to an increase in emissions. Deposits entrained by the bulk gases in the rapid motion which occurs during exhaust can also build up in the piston ring crevice regions. A reduction in volume of blowdown (see Sec. 6.5). The rapid thinning of the thermal boundary-layer these crevice regions would decrease HC emissions (and such a decrease has been regions on the combustion chamber walls during blowdown, which would result observed). However, changes in piston-cylinder wall clearance due to deposits from entrainment of the denser hydrocarbon-containing gas adjacent to the wall, can affect the flame-quenching process and could increase emissions.49 has been observed in schlieren movies taken in a transparent engine.46 This process, plus entrainment of any HC from the spark plug and head gasket crev- HYDROCARBON TRANSPORT MECHANISMS. All of the above mechanisms ices, would contribute to unburned HC in the blowdown gases which contain (except misfire) result in high hydrocarbon concentrations adjacent to the com- about half the total HC emissions (see Fig. 11-24). During the exhaust stroke this bustion chamber walls. While any jet-type flows out of crevices during the expan- bulk gas entrainment process will continue, exhausting additional unburned HC, as shown in Fig. 11-30b. 614 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 615 The second mechanism starts at the beginning of the exhaust stroke in the piston crown-cylinder wall corner. The piston motion during the exhaust stroke Unburned HC sources: Crevices scrapes the boundary-layer gases off the cylinder wall (which contain the remain- Wall phenomena der of the piston and ring crevice hydrocarbons), rolls them up into a vortex, and Bulk quench pushes them toward the top of the cylinder. This piston crown-cylinder wall corner flow is discussed in Sec. 8.7, and has been observed in transparent engines as well as in water-flow engine analog studies. At the end of the exhaust stroke, the height of this vortex is comparable to the engine clearance height. As shown in Fig. 11-30c, a recirculation flow is likely to build up in the upper corner of the Oxidation products Post-combustion cylinder away from the exhaust valve, causing the vortex to detach from the wall in-cylinder mixing and partly sweep out of the cylinder. In the corner nearest the valve, the flow is deflected around the valve, also tending to pull part of the vortex out of the Residual gas chamber. In this way it is possible for a large part of the vortex, which now THCL contains a substantial fraction of the unburned HC originally located adjacent to Oxidation products the cylinder wall, to leave the cylinder at the end of the exhaust stroke. This Exhaust port vortex flow is thought to be the mechanism that leads to the high HC concentra- tions measured at the end of the exhaust process, which contributes the other half of the exhausted HC mass (see Fig. 11-24), and to be responsible for the HC concentrations measured in the residual gases being much higher than average Exhaust pipe exhaust HC levels.42 This study showed that at close to wide-open-throttle con- ditions, only about two-thirds of the HC which fail to oxidize inside the cylinder FIGURE 11-31 were exhausted, though 95 percent of the gas within the cylinder flows out Schematic of complete SI engine HC formation and oxidation mechanism within the cylinder and exhaust system. 62 through the exhaust valve. The residual gas HC concentration was about 11 times the average exhaust level. At part-throttle conditions, where the residual gas fraction is higher, it has been estimated that only about half of the unreacted HC in the cylinder will enter the exhaust.61 ducts. While the relative proportion of fuel compared to reaction product hydro- carbon compounds varies substantially with engine operating conditions, an HYDROCARBON OXIDATION. Unburned hydrocarbons which escape the average value for passenger car vehicle exhausts is that fuel compounds comprise primary engine combustion process by the mechanisms described above must 40 percent of the total HC. Though partially reacted HC are produced in the then survive the expansion and exhaust process without oxidizing if they are to flame-quenching process, these are closest to the high-temperature burned gases appear in the exhaust. Since the formation mechanisms produce unburned HC at and are likely to mix and burn rapidly. That such a large fraction of the exhaust temperatures close to the wall temperature, mixing with bulk burned gas must HC are reaction products indicates that substantial postformation HC reactions take place first to raise the HC temperature to the point where reaction can are occurring. There is direct evidence that HC oxidation in the exhaust system occur. The sequence of processes which links the source mechanisms to hydrocar- occurs.64 Since in-cylinder gas temperatures are higher, it is likely that mixing of bons at the exhaust exit is illustrated in Fig. 11-31; it involves mixing and oxida- unburned HC with the bulk cylinder gases limits the amount of oxidation rather tion in both the cylinder and the exhaust system. than the reaction kinetics directly. There is considerable evidence that substantial oxidation does occur. The Overall empirically based expressions for the rate of oxidation of hydrocar- oxidation of unburned HC in the quench layers formed on the combustion bons of the form of Eq. (11.31) have been developed and used to examine in- chamber walls on a time scale of order 1 ms after the flame is extinguished has cylinder and exhaust burnup. A characteristic time tuc for this burnup process can be defined: already been discussed. Because the quench layers are thin, diffusion of HC into the bulk burned gas is rapid. Because the burned gases are still at a high tem- 1 1 d[HC] perature, oxidation then occurs quickly. Measurements of in-cylinder HC con- centrations by gas sampling prior to exhaust valve opening show levels about 1.5 THC [HC] dt (11.34) to 2 times the average exhaust level.44.63 The exhaust unburned HC are a Using an expression similar to Eq. (11.31) for d[HC]/dt, Weiss and Keck63 have mixture of fuel hydrocarbon compounds and pyrolyses and partial oxidation pro- shown that any HC mixing with the burned gases in the cylinder prior to exhaust 616 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 617 3000 blowdown will oxidize. The in-cylinder gas temperature prior to blowdown gen- erally exceeds 1250 K; the characteristic reaction time THc is then less than 1 ms. 2000 = 700ºC During blowdown the temperature falls rapidly to values typically less than CO, ppm 10001 75000 1000 K; THc is then greater than about 50 ms. An experimental study of HC 800ºC 85008 exiting from a simulated crevice volume has shown that complete HC oxidation 770ºC only occurs when the cylinder gas temperature is above 1400 K.65 Thus a large 1500 fraction of the HC leaving crevice regions or oil layers during the exhaust process can be expected to survive with little further oxidation. Gas-sampling data show 1000 little decrease in in-cylinder HC concentrations during the exhaust stroke, thus HC, ppm C ₡ = 700ºC 500 FIGURE 11-32 supporting this conclusion.44, 63 Overall, probably about half of the unburned 850º Effect of exhaust gas temperature on HC and CO HC formed by the source mechanisms described above will oxidize within the burnup in the exhaust. SI engine at 1600 rev/min, engine cylinder (the exact amount cannot yet be predicted with any accuracy; it 50 100 150 200 250 engine air flow = 7.7 dm3/s, lean mixture with 3% is likely to depend on engine design and operating conditions61). Residence time, ms O2 and 13% (CO + CO2) in exhaust.66 As shown schematically in Fig. 11-31, oxidation of HC in the exhaust system can occur. Often this is enhanced by air addition into the port region to ensure that adequate oxygen for burnup is available. However, since the gas tem- gas stream. HC oxidation starts immediately (for T _ 600ºC), the rate of oxida- perature steadily decreases as the exhaust gases flow through the exhaust port tion increasing rapidly with increasing temperature. Under fuel-lean conditions, and manifold, the potential for HC burnup rapidly diminishes. To oxidize the incomplete HC oxidation can result in an increase in CO levels. CO oxidation hydrocarbons in the gas phase, a residence time of order 50 ms or longer at commences later, when the gas temperature rises above the entering value due to temperatures in excess of 600ºC are required. To oxidize carbon monoxide tem- heat released by the already occurring HC oxidation. The further heat released peratures in excess of 700ºC are required. Average exhaust gas temperatures at by CO oxidation accelerates the CO burnup process. These data underline the the cylinder exit (at the exhaust valve plane) are about 800ºC; average gas tem- importance of the exhaust port heat-transfer and mixing processes. Both mixing peratures at the exhaust port exit are about 600ºC.+ Figure 6-21 shows an between the hotter blowdown gases (with their lower HC concentration) and the example of the measured cylinder pressure, measured gas temperature at the cooler end-of-exhaust gases (with their higher HC concentration) and mixing exhaust port exit, and estimated mass flow rate into the port and gas temperature between burned exhaust gas and secondary air are important. in the cylinder, during the exhaust process at a part-throttle operating condition. Engine experiments where the exhaust gas reactions were quenched by Port residence time and gas temperatures vary significantly through the process. timed injection of cold carbon dioxide at selected locations within the exhaust Precise values of these variables obviously depend on engine operating condi- port have shown that significant reductions in HC concentration in the port can tions. It is apparent that only in the exhaust port and upstream end of the mani- occur. Parallel modeling studies of the HC burnup process (based on instantane- fold can any significant gas-phase HC oxidation occur. ous mass flow rate, estimated exhaust gas temperature, and an overall hydrocar- The importance of exhaust gas temperature to exhaust system emissions bon reaction-rate expression), which predicted closely comparable magnitudes burnup is illustrated by the results shown in Fig. 11-32.66 The exhaust system of and trends, indicated that gas temperature and port residence time are the critical a four-cylinder engine was modified by installing a section of heated and insu- variables. The percent of unburned HC exiting the cylinder which reacted in the lated pipe to maintain the exhaust gas temperature constant in the absence of exhaust system (with most of the reaction occurring in the port) varied between a any HC or CO burnup. The exhaust temperature entering this test section was few and 40 percent. Engine operating conditions that gave highest exhaust tem- varied by adjusting the engine operating conditions. The figure shows CO and peratures (stoichiometric operation, higher speeds, retarded spark timing, lower HC concentrations as functions of residence time in the exhaust test section (or compression ratio) and longest residence times (lighter load) gave relatively effectively as a function of distance from the engine). Te is the entering gas tem- higher percent reductions. Air injection at the exhaust valve-stem base, phased to perature. The exhaust composition was fuel lean with 3 percent O2 in the burned coincide with the exhaust process, showed that for stoichiometric and slightly rich conditions secondary air flow rates up to 30 percent of the exhaust flow substantially increased the degree of burnup. The timing of the secondary air flow relative to the exhaust flow and the location of the air injection point in the port are known to be critical.64 + Note that there is a significant variation in the temperature of the exhaust gases throughout the exhaust process. The gas exhausted first is about 100 K hotter than the gas exhausted at the end of Reductions in exhaust port heat losses through the use of larger port cross- the process (see Sec. 6.5). sectional areas (to reduce flow velocity and surface area per unit volume), inser- 618 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 619 tion of port liners to provide higher port wall temperatures, and attention to port design details to minimize hot exhaust gas impingement on the walls are known TABLE 11.7 to increase the degree of reaction occurring in the port. Critical factors and engine variables in HC emissions mechanisms 1. Formation of HC 2. In-cylinder mixing and oxidation (a) Crevices (a) Mixing rate with bulk gas SUMMARY. It will be apparent from the above that the HC emissions formation (1) Crevice volume (1) Speed process in spark-ignition engines is extremely complex and that there are several (2) Crevice location (2) Swirl ratio paths by which a small but important amount of the fuel escapes combustion. It (relative to spark plug) (3) Combustion chamber shape is appropriate here to summarize the overall structure of the spark-ignition (3) Load (b) Bulk gas temperature during (4) Crevice wall temperature engine hydrocarbon emission problem and identify the key factors and engine expansion and exhaust (5) Mixture composition+ (1) Speed variables that influence the different parts of that problem. Table 11.7 provides (b) Oil layers (2) Spark timingt such a summary. The total process is divided into four sequential steps: (1) the (1) Oil consumption (3) Mixture composition+ formation of unburned hydrocarbon emissions; (2) the oxidation of a fraction of (2) Wall temperature (4) Compression ratio these HC emissions within the cylinder, following mixing with the bulk gases; (3) (3) Speed (5) Heat losses to walls the flow of a fraction of the unoxidized HC from the cylinder into the exhaust; (4) (c) Incomplete combustion (c) Bulk gas oxygen concentration (1) Burn rate and variability the oxidation in the exhaust system of a fraction of the HC that exit the cylinder. (1) Equivalence ratio (2) Mixture composition+ (d) Wall temperature The detailed processes and the design and operating variables that influence each (3) Load (1) Important if HC source of these steps in a significant way are listed. (4) Spark timingt near wall The four separate formation mechanisms identified in step 1 have substan- (d) Combustion chamber walls (2) For crevices : importance tial, though as yet incomplete, evidence behind them. They are listed in the most (1) Deposits depends on geometry (2) Wall roughness likely order of importance. Each has been extensively described in this section. It 3. Fraction HC flowing is through each of these mechanisms that fuel or fuel-air mixture escapes the 4. Oxidation in exhaust system out of cylinder primary combustion process. That fuel must then survive the expansion and (a) Residual fraction (a) Exhaust gas temperature exhaust processes and pass through the exhaust system without oxidation if it is (1) Load (1) Speed to end up in the atmosphere as HC emissions. The rate of mixing of these (2) Exhaust pressure (2) Spark timingt unburned HC with the hot bulk cylinder gases, the temperature and composition (3) Valve overlap (3) Mixture compositiont (4) Compression ratio of the gases with which these HC mix, and the subsequent temperature-time and (4) Compression ratio (5) Speed (5) Secondary air flow composition-time histories of the mixture will govern the amount of in-cylinder (b) In-cylinder flow during (6) Heat losses in cylinder oxidation that occurs. The distribution of these HC around the combustion exhaust stroke and exhaust chamber is nonuniform (and changes with time); they are concentrated close to (1) Valve overlap (b) Oxygen concentration the walls of the chamber. The fraction of these HC that will exit the chamber (2) Exhaust valve size and (1) Equivalence ratio location during the exhaust process will depend on the details of the in-cylinder flow (2) Secondary air flow (3) Combustion chamber shape and addition point patterns that take them through the exhaust valve. Overall, the magnitude of the (4) Compression ratio (c) Residence time residual fraction will be one major factor; the residual gas is known to be much (5) Speed (1) Speed richer in HC than the average exhaust. In particular, the flow patterns in the (2) Load cylinder toward the end of the exhaust stroke as the gas scraped off the cylinder (3) Volume of critical wall by the piston moves toward the exhaust valve will be important. Finally, a exhaust system component (d) Exhaust reactors fraction of the unburned HC which leave the cylinder through the exhaust valve (1) Oxidation catalyst will burn up within the exhaust system. Gas-phase oxidation in the exhaust ports (2) Three-way catalyst and hotter parts of the exhaust manifold is significant. The amount depends on (3) Thermal reactor the gas temperature, composition, and residence time. If catalysts or a thermal t Fuel/air equivalence ratio and burned gas fraction (residual plus recycled exhaust gas). reactor are included in the exhaust system, very substantial additional reduction + Relative to MBT timing. in HC emission levels can occur. These devices and their operating characteristics $ Of at least as great an importance as engine details if present in total emission control system. See Sec. 11.6. are described in Sec. 11.6. 620 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 621 11.4.4 Hydrocarbon Emission Mechanisms in The complex heterogeneous nature of diesel combustion, where fuel evapo- Diesel Engines ration, fuel-air and burned-unburned gas mixing, and combustion can occur simultaneously, has been discussed extensively in Chap. 10. As a result of this BACKGROUND. Diesel fuel contains hydrocarbon compounds with higher boiling points, and hence higher molecular weights, than gasoline. Also, substan- complexity, there are many processes that could contribute to diesel engine tial pyrolyses of fuel compounds occurs within the fuel sprays during the diesel hydrocarbon emissions. In Chap. 10 the diesel's compression-ignition combustion combustion process. Thus, the composition of the unburned and partially oxi- process was divided into four stages: (1) the ignition delay which is the time dized hydrocarbons in the diesel exhaust is much more complex than in the between the start of injection and ignition; (2) the premixed or rapid combustion spark-ignition engine and extends over a larger molecular size range. Gaseous phase, during which the fuel that has mixed to within combustible limits during hydrocarbon emissions from diesels are measured using a hot particulate filter (at the delay period burns; (3) the mixing controlled combustion phase, during which 190ºC) and a heated flame ionization detector. Thus the HC constituents vary the rate of burning depends on the rate of fuel-air mixing to within the combusti- from methane to the heaviest hydrocarbons which remain in the vapor phase in ble limits; (4) the late combustion phase where heat release continues at a low the heated sampling line (which is also maintained at about 190ºC). Any hydro- rate governed by the mixing of residual combustibles with excess oxygen and the carbons heavier than this are therefore condensed and, with the solid-phase soot, kinetics of the oxidation process. There are two primary paths by which fuel can escape this normal combustion process unburned: the fuel-air mixture can are filtered from the exhaust gas stream upstream of the detector. The particulate emission measurement procedure measures a portion of total engine hydrocar- become too lean to autoignite or to support a propagating flame at the condi- bon emissions also. Particulates are collected by filtering from a diluted exhaust tions prevailing inside the combustion chamber, or, during the primary com- gas stream at a temperature of 52ºC or less. Those hydrocarbons that condense bustion process, the fuel-air mixture may be too rich to ignite or support a flame. at or below this temperature are absorbed onto the soot. They are the extractable This fuel can then be consumed only by slower thermal oxidation reactions later fraction of the particulate: i.e ., that fraction which can be removed by a powerful in the expansion process after mixing with additional air. Thus, hydrocarbons solvent, typically between about 15 and 45 percent of the total particulate mass. remain unconsumed due to incomplete mixing or to quenching of the oxidation process.+ This section discusses gaseous hydrocarbon emissions; particulate emissions- soot and extractable material-are discussed in Sec. 11.5. Figure 11-33 shows schematically how these processes can produce incom- plete combustion products. Fuel injected during the ignition delay (Fig. 11-33a) will mix with air to produce a wide range of equivalence ratios. Some of this fuel Fue Air Fuel will have mixed rapidly to equivalence ratios lower than the lean limit of com- Slow mixing or bustion (locally overlean mixture), some will be within the combustible range, Fuel-air Pyrolyses lack of oxygen mixture and some will have mixed more slowly and be too rich to burn (locally overrich mixture). The overlean mixture will not autoignite or support a propagating Products of flame at conditions prevailing inside the combustion chamber (though some of Locally Locally Combustible pyrolyses overlean overrich Locally this mixture may burn later if it mixes with high-temperature burned products mixture mixture mixture Combustible overrich early in the expansion stroke). In the "premixed" combustible mixture, ignition mixture mixture Bulk occurs where the local conditions are most favorable for autoignition. Unless Slow reaction, quenching quenched by thermal boundary layers or rapid mixing with air, subsequent no ignition or Slow reaction, flame propagation autoignition or flame fronts propagating from the ignition sites consume the no ignition or Ignition and Bulk combustible mixture. Complete combustion of overrich mixture depends on flame propagation flammation Flammation Quenching further mixing with air or lean already-burned gases within the time available before rapid expansion and cooling occurs. Of all these possible mechanisms, the - Products of Products of Products of Products of overlean mixture path is believed to be the most important.23 incomplete complete complete incomplete For the fuel injected after the ignition delay period is over (Fig. 11-33b), combustion combustion combustion combustion rapid oxidation of fuel or the products of fuel pyrolyses, as these mix with air, (a) (b) FIGURE 11-33 Schematic representation of diesel hydrocarbon formation mechanisms: (a) for fuel injected during + Note that under normal engine operating conditions, the combustion inefficiency is less than 2 delay period; (b) for fuel injected while combustion is occurring.23 percent; see Sec. 4.9.4 and Fig. 3-9. 622 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 623 results in complete combustion. Slow mixing of fuel and pyrolyses products with 2400 air, resulting in overrich mixture or quenching of the combustion reactions, can 2000 result in incomplete combustion products, pyrolyses products, and unburned fuel being present in the exhaust.23 1800- Hydrocarbon emission levels from diesels vary widely with operating condi- Exhaust HC, ppm C 1200 tions, and different HC formation mechanisms are likely to be most important at different operating modes. Engine idling and light-load operation produce signifi- 800 cantly higher hydrocarbon emissions than full-load operation. However, when FIGURE 11-35 400 the engine is overfueled, HC emissions increase very substantially. As will be Correlation of exhaust HC concentration with dura- explained more fully below, overmixing (overleaning) is an important source of OL tion of ignition delay for DI diesel engine. Various 4 8 12 16 20 24 HC, especially under light-load operation. Undermixing, resulting in overrich fuels, engine loads, injection timings, boost pres- Ignition delay, deg sures, at 2800 rev/min.67 mixture during the combustion period, is the mechanism by which some of the fuel remaining in the injector nozzle sac volume escapes combustion, and is also the cause of very high HC emissions during overfueling. Wall temperatures affect HC emissions, suggesting that wall quenching is important, and under especially where the fuel which has spent most time within the combustible limits is located. adverse conditions very high cyclic variability in the combustion process can However, the fuel close to the spray boundary has already mixed beyond the lean limit of combustion and will not autoignite or sustain a fast reaction front. This cause an increase in HC due to partial burning and misfiring cycles. mixture can only oxidize by relatively slow thermal-oxidation reactions which OVERLEANING. As soon as fuel injection into the cylinder commences, a dis- will be incomplete. Within this region, unburned fuel, fuel decomposition pro- tribution in the fuel/air equivalence ratio across the fuel sprays develops. The ducts, and partial oxidation products (aldehydes and other oxygenates) will exist; amount of fuel that is mixed leaner than the lean combustion limit (¢1 ~ 0.3) some of these will escape the cylinder without being burned. The magnitude of increases rapidly with time.23 Figure 11-34 illustrates this equivalence ratio dis- the unburned HC from these overlean regions will depend on the amount of fuel tribution in the fuel spray at the time of ignition. In a swirling flow, ignition injected during the ignition delay, the mixing rate with air during this period, and occurs in the slightly lean-of-stoichiometric region downstream of the spray core the extent to which prevailing cylinder conditions are conducive to autoignition. A correlation of unburned HC emissions with the length of the ignition delay would be expected. The data in Fig. 11-35 from a direct-injection naturally aspi- rated engine show that a good correlation between these variables exists. As the Air swirl delay period increases beyond its minimum value (due to changes in engine oper- ating conditions), HC emissions increase at an increasing rate.67 Thus, overlean- ing of fuel injected during the ignition delay period is a significant source of > > 1, rich hydrocarbon emissions, especially under conditions where the ignition delay is long. ¢ = 0 UNDERMIXING. Two sources of fuel which enter the cylinder during combustion Injection nozzle and which result in HC emissions due to slow or under mixing with air have been identified. One is fuel that leaves the injector nozzle at low velocity, often late in the combustion process. The most important source here is the nozzle sac volume, though secondary injections can increase HC emissions if the Fuel jet boundary problem is severe. The second source is the excess fuel that enters the cylinder under overfueling conditions. Ignition location Overmixed HC At the end of the fuel-injection process, the injector sac volume (the small volume left in the tip of the injector after the needle seats) is left filled with fuel. FIGURE 11-34 Schematic of diesel engine fuel spray showing equivalence ratio (o) contours at time of ignition. As the combustion and expansion processes proceed, this fuel is heated and ¢1 = equivalence ratio at lean combustion limit (~0.3). Shaded region contains fuel mixed leaner vaporizes, and enters the cylinder at low velocity through the nozzle holes. This than ¢ ,. 67 fuel vapor (and perhaps large drops of fuel also) will mix relatively slowly with air 624 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 625 Standard sac, volume = 1.35 mm3 500 600 Valve covers orifice 500 100 Exhaust HC, ppm C 400 Exhaust HC, ppm C Reduced sac, volume = 0.6 mm3 3001- 200 200- FIGURE 11-37 100 0.2 0.4 0.6 0.81.0 ¢ Approximate volume of nozzle holes Effect of overfueling on exhaust HC concentration. 25 50 15 100 125 DI diesel engine, speed = 1700 rev/min, injection 0.2 0.2 0.4 0.6 0.8 1.0 2 1.4 Load, % 0 timing at full-load 15º BTC.67 Nozzle sac volume, mm3 FIGURE 11-36 similar trend exists for IDI engines.67 This mechanism is not significant under Effect of nozzle sac volume on exhaust HC concentration, DI diesel engine, at minimum ignition delay. V. = 1 dm3/cylinder, 1700-2800 rev/min. 67 normal operating conditions, but can contribute HC emissions under acceler- ation conditions if overfueling occurs. However, it produces less HC than does overleaning at light load and idle.23 and may escape the primary combustion process. Figure 11-36 shows HC emis- QUENCHING AND MISFIRE. Hydrocarbon emissions have been shown to be sions at the minimum ignition delay for a direct-injection diesel engine as a func- sensitive to oil and coolant temperature: when these temperatures were increased tion of sac volume, along with drawings of some of the injector nozzles used. The from 40 to 90ºC in a DI diesel, HC emissions decreased by 30 percent. Since correlation between HC emissions (under conditions when the overleaning ignition delay was maintained constant, overmixing phenomena should remain mechanism is least significant) and sac volume is striking. The extrapolation to approximately constant. Thus, wall quenching of the flame may also be a signifi- zero HC emissions suggests that the fuel in the nozzle holes also contributes. Not cant source of HC, depending on the degree of spray impingement on the com- all the fuel in the sac volume is exhausted as unburned hydrocarbons. For bustion chamber walls. example, in Fig. 11-36 a volume of 1 mm3 gives 350 ppm C1 while 1 mm3 of fuel While cycle-by-cycle variation in the combustion process in diesel engines is would give 1660 ppm C1. The sac volume may not be fully filled with fuel. Also, generally much less than in spark-ignition engines, it can become significant the higher-boiling-point fractions of the fuel may remain in the nozzle. Significant under adverse conditions such as low compression temperatures and pressures oxidation may also occur. In indirect-injection engines, similar trends have been and retarded injection timings. Substantial variations, cycle-by-cycle, in HC emis- observed, but the HC emission levels at short ignition delay conditions are sub- sions are thought to result. In the limit, if misfire (no combustion) occurs in a stantially lower. The sac volume in current production nozzles helps to equalize fraction of the operating cycles, then engine HC emissions rise as the percentage the fuel pressures immediately upstream of the nozzle orifices. A small sac volume of misfires increases. However, complete misfires in a well-designed and ade- makes this equalization less complete and exhaust smoke deteriorates. The con- quately controlled engine are unlikely to occur over the normal operating tribution of sac and hole volumes to exhaust HC can be reduced to below 0.75 g/ range, 23 kW . h for a 1 dm3 per cylinder displacement DI engine.67 In DI engines, exhaust smoke limits the full-load equivalence ratio to about SUMMARY. There are two major causes of HC emissions in diesel engines under 0.7. Under transient conditions as the engine goes through an acceleration normal operating conditions: (1) fuel mixed to leaner than the lean combustion process, overfueling can occur. Even though the overall equivalence ratio may limit during the delay period; (2) undermixing of fuel which leaves the fuel injec- remain lean, locally overrich conditions may exist through the expansion stroke tor nozzle at low velocity, late in the combustion process. At light load and idle, and into the exhaust process. Figure 11-37 shows the effect of increasing the overmixing is especially important, particularly in engines of relatively small amount of fuel injected at constant speed, with the injection timing adjusted to cylinder size at high speed. In IDI engines, the contribution from fuel in the keep the ignition delay at its minimum value (when HC emissions from overlean- nozzle sac volume is less important than with DI engines. However, other sources ing are lowest). HC emissions are unaffected by an increasing equivalence ratio of low velocity and late fuel injection such as secondary injection can be signifi- until a critical value of about 0.9 is reached when levels increase dramatically. A cant. 626 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 627 11.5 PARTICULATE EMISSIONS peratures above 500ºC, the individual particles are principally clusters of many small spheres or spherules of carbon (with a small amount of hydrogen) with 11.5.1 Spark-Ignition Engine Particulates individual spherule diameters of about 15 to 30 nm. As temperatures decrease There are three classes of spark-ignition engine particulate emissions: lead, below 500ºC, the particles become coated with adsorbed and condensed high organic particulates (including soot), and sulfates. molecular weight organic compounds which include: unburned hydrocarbons, Significant sulfate emissions can occur with oxidation-catalyst equipped oxygenated hydrocarbons (ketones, esters, ethers, organic acids), and polynuclear engines. Unleaded gasoline contains 150 to 600 ppm by weight sulfur, which is aromatic hydrocarbons. The condensed material also includes inorganic species oxidized within the engine cylinder to sulfur dioxide, SO2 . This SO2 can be oxi- such as sulfur dioxide, nitrogen dioxide, and sulfuric acid (sulfates). dized by the exhaust catalyst to SO3 which combines with water at ambient The objective of most particulate measurement techniques is to determine temperatures to form a sulfuric acid aerosol. Levels of sulfate emissions depend the amount of particulate being emitted to the atmosphere. Techniques for on the fuel sulfur content, the operating conditions of the engine, and the details particulate measurement and characterization range from simple smoke meter of the catalyst system used. Typical average automobile sulfate emission rates are opacity readings to analyses using dilution tunnels. Most techniques require 20 mg/km or less. 68 lengthy sample-collection periods because the emission rate of individual species For automobile engines operated with regular and premium leaded gas- is usually low. The physical conditions under which particulate measurements are olines (which contain about 0.15 g Pb/liter or dm3) the particulate emission rates made are critical because the emitted species are unstable and may be altered are typically 100 to 150 mg/km. This particulate is dominated by lead com- through loss to surfaces, change in size distribution (through collisions), and pounds: 25 to 60 percent of the emitted mass is lead.69 The particulate emission chemical interactions among other species in the exhaust at any time during the rates are considerably higher when the engine is cold, following start-up. The measurement process (including sampling, storage, or examination). The most exhaust temperature has a significant effect on emission levels. The particle size basic information is normally obtained on a mass basis: for example, grams per distribution with leaded fuel is about 80 percent by mass below 2 um diameter kilometer for a vehicle, grams per kilowatt-hour for an engine, grams per kilo- and about 40 percent below 0.2 um diameter. Most of these particles are pre- gram of fuel or milligrams per cubic meter of exhaust (at standard conditions). sumed to form and grow in the exhaust system due to vapor phase condensation Smoke meters measure the relative quantity of light that passes through the enhanced by coagulation. Some of the particles are emitted directly, without set- exhaust or the relative reflectance of particulate collected on filter paper. They do tling. Some of the particles either form or are deposited on the walls where not measure mass directly. They are used to determine visible smoke emissions agglomeration may occur. Many of these are removed when the exhaust flow rate and provide an approximate indication of mass emission levels. Visible smoke is suddenly increased, and these particles together with rust and scale account for from heavy-duty diesels at high load is regulated. In the standard mass emission the increase in mass and size of particles emitted during acceleration. Only a measurement procedure, dilution tunnels are used to simulate the physical and fraction (between 10 and 50 percent) of the lead consumed in the fuel is chemical processes the particulate emissions undergo in the atmosphere. In the exhausted, the remainder being deposited within the engine and exhaust system. dilution tunnel, the raw exhaust gases are diluted with ambient air to a tem- Use of unleaded gasoline reduces particulate emissions to about 20 mg/km perature of 52ºC or less, and a sample stream from the diluted exhaust is filtered in automobiles without catalysts. This particulate is primarily soluble to remove the particulate material. (condensed) organic material. Soot emissions (black smoke) can result from com- bustion of overly rich mixtures. In properly adjusted spark-ignition engines, soot in the exhaust is not a significant problem. PARTICULATE COMPOSITION AND STRUCTURE. The structure of diesel particulate material is apparent from the photomicrographs shown in Fig. 11-38 of particulates collected from the exhaust of an IDI diesel engine. The samples 11.5.2 Characteristics of Diesel Particulates are seen to consist of collections of primary particles (spherules) agglomerated MEASUREMENT TECHNIQUES. Diesel particulates consist principally of com- into aggregates (hereafter called particles). Individual particles range in appear- bustion generated carbonaceous material (soot) on which some organic com- ance from clusters of spherules to chains of spherules. Clusters may contain as pounds have become absorbed. Most particulate material results from many as 4000 spherules. Occasional liquid hydrocarbon and sulfate droplets have incomplete combustion of fuel hydrocarbons; some is contributed by the lubri- been identified. The spherules are combustion generated soot particles which cating oil. The emission rates are typically 0.2 to 0.6 g/km for light-duty diesels in vary in diameter between 10 and 80 nm, although most are in the 15 to 30 nm an automobile. In larger direct-injection engines, particulate emission rates are range. Figure 11-39 shows a typical distribution of spherule size (solid line) deter- 0.5 to 1.5 g/brake kW . h. The composition of the particulate material depends on mined by sizing and counting images in the photomicrographs. The number- the conditions in the engine exhaust and particulate collection system. At tem- mean diameter (=> N;d;/N) is 28 nm. The volume contribution of these 628 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 629 TABLE 11.8 Chemical composition of particular matter70 0.5 um 0.5 um Idle 48 km/h Extractable composition C23H2904.7No.21 C24H3002.6No.18 H/C 1.26 1.63 Dry soot composition CH0.2700.22No.01 CHO.2100.15No.01 H/C 0.27 0.21 spherules is shown as the dashed curve in Fig. 11-39. The volume-mean diameter, ( N.d)/N)1/3, is 31 nm. Determination of the particle size distribution with a similar technique involves assigning a single dimension to a complex and irregular aggregate, and introduces uncertainties arising from only having two-dimensional images of par- ticles available. Other approaches based on inertial impactors and electrical aerosol analysers have been used. Some of the data suggest that the particle size distribution is bimodal. The smaller-size range is thought to be liquid hydrocar- bon drops and/or individual spherules characterized by number-mean diameters of 10 to 20 nm; the larger-size range is thought to be the particles of agglomer- ated spherules characterized by number-mean diameters of 100 to 150 nm. However, other particulate samples have not shown a bimodal distribution: volume-mean diameters ranged from 50 to 220 nm with no notable trend with either speed or load.70 The exhaust particulate is usually partitioned with an extraction solvent into a soluble fraction and a dry-soot fraction. Two commonly used solvents are dichloromethane and a benzene-ethanol mixture. Typically 15 to 30 mass percent FIGURE 11-38 is extractable, though the range of observations is much larger (~10 to 90 Photomicrographs of diesel particulates: cluster (upper left), chain (upper right), and collection from percent). Thermogravimetric analysis (weighing the sample as it is heated) pro- filter (bottom).70 duces comparable results. Typical average chemical compositions of the two par- ticulate fractions are given in Table 11.8. Dry soot has a much lower H/C ratio than the extractable material. Although most of the particulate emissions are formed through incomplete combustion of fuel hydrocarbons, engine oil may also contribute significantly. The number-average molecular weight of the extractable material shown in Table 11.8 ranged from about 360 to 400 for a variety of engine conditions. This fell between the average molecular weight of the fuel (199) Volume contribution Number of spherules and that of the lubricating oil (443 when fresh and 489 when aged).70 Radioactive tracer studies in a light-duty IDI diesel have shown that the oil was the origin of between 2 to 25 percent by mass of the total particulate and 16 to 80 percent of the extractable organic portion, the greatest percentages being measured at the FIGURE 11-39 highest engine speed studied (3000 rev/min). All of the oil contribution appeared 20 30 40 50 60 Typical distributions of spherule diameter and in the extractable material. The contributions from the different individual com- 10 Spherule diameter, nm volume. 70 pounds in the fuel have also been studied. All the compounds tested-paraffins, 630 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 631 tallites per spherule. The crystallites are arranged with their planes more or less parallel to the particle surface. This structure of unordered layers is called turbo- static. The spherules, 10 to 50 nm in diameter, are fused together to form par- ticles as shown in Fig. 11-40. A single spherule contains 105 to 106 carbon atoms. 70, 73 A surface area of about 200 m2/g has been measured for diesel soot. A smooth-surfaced 30-nm diameter sphere with a density of 2 g/cm3 would have a surface/mass ratio of 100 m2/g, so the measured value is about twice the super- ficial area. Approximating a particle of agglomerated spherules by a single sphere of 200 nm diameter gives a surface/mass ratio of 15 m2/g.7º These data and esti- mates of superficial area per unit mass indicate that diesel soot has low porosity. SOLUBLE FRACTION COMPONENTS. The extractable organic fraction of diesel particulate emissions includes compounds that may cause health and environmental hazards. Thus chemical and biological characterization of the soluble organic fraction are important. Both soxhlet and sonification methods are FIGURE 11-40 used to extract the organic fraction from particulate samples. Because the partic- Lattice-imaging micrograph of a diesel particulate.72 ulates are mixtures of polar and nonpolar components, full extraction requires different solvents; any one solvent is a compromise. Methylene chloride is the most commonly used extractant, however. Since a complex mixture of organic olefins, and aromatics-contributed to the particulate emissions; as a group, aro- compounds is associated with diesel particulates, a preliminary fractionation matics were the greatest contributors. Eighty percent of the carbon-14 used to tag scheme is used to group similar types of compounds before final separation and individual fuel compounds was found in the insoluble fraction and 20 percent in identification. The scheme most frequently used results in seven fractions gener- the soluble particulate fraction.71 ally labeled as: basics, acidics, paraffins, aromatics, transitionals, oxygenates, and In addition to the elements listed in Table 11.8, trace amounts of sulfur, ether insolubles. Table 11.9 indicates the types of components in each fraction zinc, phosphorus, calcium, iron, silicon, and chromium have been found in parti- and the approximate proportions. The biological activity of the soluble organic culates. Sulfur and traces of calcium, iron, silicon, and chromium are found in fraction and its subfractions is most commonly assessed with the Ames diesel fuel; zinc, phosphorus, and calcium compounds are frequently used in Salmonella/microsomal test. With this test, a quantitative dose-response curve lubricating oil additives.7º showing the mutagenicity of a sample compound is obtained. The Ames test uses A lattice image of a diesel particle is shown in Fig. 11-40; it suggests a a mutant strain of Salmonella typhimurium that is incapable of producing histi- concentric lamellate structure arranged around the center of each spherule. This dine. Mutagenicity is defined as the ability of a tested compound to revert- arrangement of concentric lamellas is similar to the structure of carbon black. back-mutate-this bacterium to its wild state, where it regains its ability to This is not surprising; the environment in which diesel soot is produced is similar produce histidine.35 to that in which oil furnace blacks are made. The carbon atoms are bonded together in hexagonal face-centered arrays in planes, commonly referred to as platelets. As illustrated in Fig. 11-41, the mean layer spacing is 0.355 nm (only 11.5.3 Particulate Distribution within the slightly larger than graphite). Platelets are arranged in layers to form crystallites. Cylinder Typically, there are 2 to 5 platelets per crystallite, and on the order of 103 crys- Measurements have been made of the particulate distribution within the com- bustion chamber of operating diesel engines. The results provide valuable infor- mation on the particulate formation and oxidation processes and how these relate to the fuel distribution and heat-release development within the com- 1.2 nm bustion chamber. Techniques used to obtain particulate concentration data 0.355 nm FIGURE 11-41 include: use of rapid-acting poppet or needle valves which draw a small gas Platelet Crystallite Particle Substructure of carbon particle.73 sample from the cylinder at a specific location and time for analysis (e.g ., Refs. 21 632 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 633 TABLE 11-9 125 mm Components of the soluble organic fraction35 75 mm Percent of 4. D. PC Fraction Components of fraction total 400- Acidic Aromatic or aliphatic 3-15 TC Acidic functional groups 15.100. Phenolic and carboxylic acids - 25 0- - Position Basic Aromatic or aliphatic <1-2 300- o-0 Centerline A 00 +4.5º A Basic functional groups +00 B Amines 4-4 ^ +18º C Paraffin Aliphatics, normal and branched 34-65 +36º D Numerous isomers From unburned fuel and/or lubricant Soot concentration, g/m3 200 2.0 Aromatic From unburned fuel, partial combustion, and 3-14 1.5 recombination of combustion products; from 1.0 lubricants 0.5 Single ring compounds 100-91 0 Polynuclear aromatics TC 20 40 60 80 100 Oxygenated Polar functional groups but not acidic or basic 7-15 Aldehydes, ketones, or alcohols Sector boundary Injection duration Aromatic phenols and quinones Transitional Aliphatic and aromatic 1-6 180 Carbonyl functional groups -20 TC 20 40 60 80 100 190/ Ketones, aldehydes, esters, ethers Crank angle, deg ATC pray, Insoluble Aliphatic and aromatic 6-25 4.5º Hydroxyl and carbonyl groups 20º cone angle- High molecular weight organic species Inorganic compounds FIGURE 11-42 Glass fibers from filters Particulate concentrations, in g/m3 at standard temperature and pressure, in various regions of the fuel spray as a function of crank angle in quiescent DI diesel engine, measured with rapid sampling valve. Different sample valve locations in combustion chamber and spray indicated on left. Cylinder bore = 30.5 cm, stroke = 38.1 cm, re = 12.9, engine speed == 500 rev/min, bmep = 827 kPa.74 and 74), optical absorption techniques (e.g ., Refs. 75 and 76), and cylinder dumping where the cylinder contents are rapidly emptied into an evacuated tank at a preset time in the cycle (e.g ., Ref. 77). Both DI and IDI engines have been studied. Of course, concentration data taken at specific locations in the cylinder injection diesel engine which illustrates these points.74 The particulate during the engine cycle are not necessarily representative of the cylinder contents concentrations on the fuel spray axis close to the injector orifice are remarkably in general; nor do they represent the time history of a given mass of gas. The fuel high (~200 to 400 g/m3 at standard temperature and pressure). This corresponds distribution, mixing, and heat-release patterns in the cylinder are highly nonuni- to a large fraction of the fuel carbon in the extremely rich fuel vapor core being form during the soot-formation process, and the details of gas motion in the sampled as particulate (as soot and condensed HC species). Such high particulate vicinity of the sampling location as the piston changes position are usually fractions of the local fuel carbon (~50 percent) have also been found in the very unknown. fuel rich cores of high-pressure liquid-fueled turbulent diffusion flames. Pyrolyses In direct-injection diesel engines, the highest particulate concentrations are of the fuel is therefore an important source of soot. These very high local soot found in the core region of each fuel spray where local average equivalence ratios concentrations decrease rapidly once fuel injection ceases and the rich core mixes are very rich (see Secs. 10.5.6 and 10.7.2). Soot concentrations rise rapidly soon to leaner equivalence ratios. Soot concentrations in the spray close to the piston after combustion starts. Figure 11-42 shows a set of sample-valve soot- bowl outer radius and at the cylinder wall rise later, are an order of magnitude concentration data from a large (30.5-cm bore, 38.1-cm stroke), quiescent, direct- less, and decay more slowly. Away from the fuel spray core, soot concentrations 634 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 635 2 mm 10 mm 2 15 mm W Soot, g/m3 ignition FIGURE 11-43 N, 109 part./cm3 21 Particulate concentrations at various distances from wall of prechamber in swirl-chamber IDI 1 Exhaust 10 TC 10 20 30 40 50 diesel engine, measured with rapid sampling valve, OL TC 10 20 Crank angle, deg Engine speed = 1000 rev/min, injection at 12º 30 40 50 110 1 120 Injection BTC, ignition at 5º BTC.21 EVO 1.0 FIGURE 11-44 0.8 0.6 Cylinder-average particle-number density N and Fy, cm3/m3 0.4 particle-volume fraction Fy, as a function of crank Exhaust 0.2 -- O angle in IDI diesel engine determined from 0 cylinder-dumping experiments. 1000 rev/min, decrease rapidly with increasing distance from the centerline. A useful compari- TC 10 20 30 40 50 110 $ 120 EVO ¢ = 0.32, injection starts at 3.5º BTC. Gas son with these soot concentrations is the fuel concentration in a stoichiometric Crank angle, deg volumes at standard temperature and pressure.77 mixture, about 75 g fuel/m3. Approximate estimates of the mean soot concentra- tion inside the cylinder through the combustion process suggests that almost all (over 90 percent) of the soot formed is oxidized prior to exhaust. Similar results 11.5.4 Soot Formation Fundamentals have been obtained in a small direct-injection engine with swirl.78, 79 Peak soot The soot particles, whose characteristics have been described in the above two concentrations in the outer regions of the fuel spray were comparable (~ 10 g/ sections, form primarily from the carbon in the diesel fuel. Thus, the formation m3). Measurements were not made in the spray core near the injector orifice; process starts with a fuel molecule containing 12 to 22 carbon atoms and an H/C however, based on the equivalence ratio results in Fig. 10-46, soot concentrations ratio of about 2, and ends up with particles typically a few hundred nanometers would be expected to be lower due to the more rapid mixing with air that occurs in diameter, composed of spherules 20 to 30 nm in diameter each containing with swirl. some 105 carbon atoms and having an H/C ratio of about 0.1. Most of the Similar data are available from sampling in the prechamber of an IDI swirl information available on the fundamentals of soot formation in combustion chamber engine.21 Figure 11-43 shows soot concentrations 2, 10, and 15 mm comes from studies in simple premixed and diffusion flames, stirred reactors, from the wall of the prechamber. Equivalence ratio distributions from this study shock tubes, and constant-volume combustion bombs. A recent review" sum- have already been shown in Fig. 11-17. Concentrations peak 5 to 10º ATC at marizes the extensive literature available from such studies. Also, the production levels ~2 g/m3; these are substantially lower than DI engine peak soot concen- of carbon black requires a high yield of soot from pyrolyses of a hydrocarbon trations (presumably due to the more rapid mixing of fuel and air in the IDI feedstock, and the literature from that field has much to contribute (see Ref. 81). engine). Concentrations in the prechamber at these locations then decrease sub- However, the characteristics of diesel combustion which make it unsuitable for stantially. more fundamental studies-the high gas temperatures and pressures, complex A better indication of average concentrations within the cylinder is given by fuel composition, dominance of turbulent mixing, the unsteady nature of the total cylinder sampling experiments. Measurements of the total number of soot process, and the three-dimensional geometry-also make it difficult to interpret particles and soot volume fraction through the combustion process have been fundamental ideas regarding soot formation in the diesel context. There is much made in an IDI passenger car diesel. The contents of the engine cylinder, at a about the soot formation process in diesel engines, therefore, that is poorly and preselected point in the cycle, were rapidly expelled through a blowdown port, incompletely understood. diluted, and collected in a sample bag. Figure 11-44 shows one set of results. Soot formation takes place in the diesel combustion environment at tem- Particles first appear shortly after the start of combustion (4 to 5º ATC). The peratures between about 1000 and 2800 K, at pressures of 50 to 100 atm, and number density rises to a maximum at 20º ATC and then falls rapidly as a result with sufficient air overall to burn fully all the fuel. The time available for the of particle coagulation and, possibly, oxidation. The exhaust particulate number formation of solid soot particles from a fraction of the fuel is in the order of density is less than one-tenth of the peak value. The volume fraction soot data milliseconds. The resulting aerosol-dispersed solid-phase particles in a gas-can (soot mass concentration is proportional to volume fraction) show a much flatter be characterized by the total amount of condensed phase (often expressed as the maximum earlier in the combustion process and a decrease (due to oxidation) soot volume fraction, Fy, the volume of soot/total volume), the number of soot from 20 to 40º ATC to about one-third of the peak value. Oxidation apparently particles per unit volume (N), and the size of the particles (e.g ., average diameter ceases at about 40º ATC at these conditions. d). Fy, N, and d are mutually dependent [e.g ., for spherical particles Fy = 636 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 637 (Tt/6)Nd3], and any two of these variables characterize the system. It is most growth nor purely PAH growth would lead to soot particles which have xH in convenient to consider N and Fy as the independent variables since they each the range 0.1 to 0.2. What is required is condensation of species with the right relate to the "almost-separate" stages of soot particle generation (the source of hydrogen content, or condensation of species with higher hydrogen content N) and soot particle growth (the source of Fy). followed by dehydrogenation, or a combination of both these processes. Obvi- These stages can be summarized as follows:8º ously some polyacetylenes and some PAH can satisfy these requirements, as can saturated platelets (e.g ., C27H27; see Sec. 11.5.2). Surface growth reactions 1. Particle formation, where the first condensed phase material arises from the lead to an increase in the amount of soot (Fy) but the number of particles (N) fuel molecules via their oxidation and/or pyrolyses products. These products remains unchanged. The opposite is true for growth by coagulation, where the typically include various unsaturated hydrocarbons, particularly acetylene and particles collide and coalesce, which decreases N with Fy constant. Once its higher analogues (C2 ,, H2), and polycyclic aromatic hydrocarbons (PAH). surface growth stops, continued aggregation of particles into chains and clus- These two types of molecules are considered the most likely precursors of soot ters can occur. in flames. The condensation reactions of gas-phase species such as these lead to the appearance of the first recognizable soot particles (often called nuclei). These stages of particle generation and growth constitute the soot forma- These first particles are very small (d < 2 nm) and the formation of large tion process. At each stage in the process oxidation can occur where soot or soot numbers of them involve negligible soot loading in the region of their forma- precursors are burned in the presence of oxidizing species to form gaseous pro- ducts such as CO and CO2. The eventual emission of soot from the engine will tion. depend on the balance between these processes of formation and burnout. The 2. Particle growth, which includes both surface growth, coagulation, and aggre- gation. Surface growth, by which the bulk of the solid-phase material is gener- emitted soot is then subject to a further mass addition process as the exhaust gases cool and are diluted with air. Adsorption into the soot particle surface and ated, involves the attachment of gas-phase species to the surface of particles and their incorporation into the particulate phase. Figure 11-45, where the log condensation to form new particles of hydrocarbon species in the exhaust gases occurs in the exhaust system and in the dilution tunnel which simulates what of the molecular weight of a species is plotted against its hydrogen mole frac- happens in the atmosphere. Figure 11-46 illustrates the relationship between tion xH, illustrates some important points about this process. Starting with a these processes.70 Although they are illustrated as discrete processes, there is fuel molecule of xy _ 0.5 it is apparent that neither purely polyacetylene chain some overlap, and they may occur concurrently in a given elemental mixture region within the diesel combustion chamber. Of course, due also to the non- homogeneous nature of the mixture and the duration of fuel injection and its overlap with combustion, at any given time different processes are in progress in 7 different regions or packets of fluid. The fundamentals of each of these processes will now be reviewed. Soot O O - Dehydrogenation 5 Nucleation Oxidation log (molar mass) Polycyclic aromatics Paraffins Surface growth - Dehydrogenation Oxidation Cylinder C24H16 Time 27H27O Polyacetylenes 2- C6H6 C24H12 C8H2 Agglomeration Dehydrogenation - Oxidation .C2H4 C2H2 C4H2 C2H6 1L 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 CH Hydrocarbons - Adsorption and Dilution tunnel condensation FIGURE 11-46 Processes leading to net production of FIGURE 11-45 Paths to soot formation on plot of species molecular weight M versus hydrogen mole fraction xH.80 diesel particulates. 70 638 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 639 SOOT PARTICLE FORMATION. Empirically, it has been found useful to define Aromatics the composition of the fuel-oxidizer mixture at the onset of soot formation in Condensation reactions Direct (fast) Soot flames by the carbon/oxygen ratio. Equilibrium considerations indicate that soot formation should occur when, in Fragmentation reactions C. H, + yO2 - 2, CO += H2 + (m - 2y)Cs (11.35) CHx Indirect (slow) Soot m becomes larger than 2y: i.e ., the C/O ratio exceeds unity. The corresponding fuel/air equivalence ratio is given by FIGURE 11-47 Fragmentation reactions $ = 2(6 )(1 + 5) (11.36) Aliphatics Mechanistic model for formation of soot from aromatic and aliphatic hydrocarbon compounds. 70 where ô = n/(4m); ¢ is 3 for (C/O) = 1, with n/m = 2. The experimentally observed critical C/O ratios are less than unity, however, varying with fuel com- position and details of the experimental setup from about 0.5 to 0.8. The critical C/O ratio for soot formation increases with increasing temperature but is only soot nuclei. Aliphatic molecules can only follow this latter less-direct route. weakly dependent on pressure. Beyond the carbon formation limit, the yield of Experimental measurements in flames suggest that polyunsaturated hydrocarbon soot increases rapidly with increasing C/O ratio and is strongly enhanced by compounds are involved in nucleation, and acetylenes and polyacetylenes have increasing pressure.80 been detected that decrease in concentration as the mass of carbon formed It is obvious that soot formation is a nonequilibrium process. Yet despite increases. Such observations fit the indirect path in Fig. 11-47. Results of studies decades of study, the precise details of the chemistry leading to the establishment of pyrolyses of benzene between 1300 and 1700 K support a physical conden- of soot nuclei still elude investigators. Several different theories have been sation mechanism for the low-temperature path. This mechanism begins with the advanced to explain the pyrolyses process-the extensive decomposition and transformation of the initial hydrocarbon into macromolecules by a gas-phase atomic rearrangement of the fuel molecules-that culminates in nucleation. reaction. The partial pressure of these macromolecules grows until supersatu- Reviews of these theories can be found in Refs. 73, 80, and 81. Often-cited mecha- ration becomes sufficient to force their condensation into liquid microdroplets. nisms are thermal cracking that results in fragmentation of fuel molecules into These become nuclei, and subsequently formed gaseous macromolecules then smaller ones, condensation reactions and polymerization that result in larger contribute to nuclei growth.70 molecules, and dehydrogenation that lowers the H/C ratio of the hydrocarbons destined to become soot. Three different paths to the production of soot appear SOOT PARTICLE GROWTH. Nucleation produces a large number of very small to exist, depending on the formation temperature. At the lowest temperatures particles with an insignificant soot loading. The bulk of the solid-phase material (<1700 K) only aromatics or highly unsaturated aliphatic compounds of high is generated by surface growth, which involves the gas-phase deposition of hydro- molecular weight are very effective in forming solid carbon through pyrolyses. At carbon intermediates on the surfaces of the spherules that develop from the intermediate temperatures typical of diffusion flames (21800 K), all normally nuclei. A qualitative description of the changes that occur as a function of time in used hydrocarbon fuels produce soot if burned at sufficiently rich stoichiometry, a premixed flame during nucleation and surface growth is shown in Fig. 11-48. but appear to do so by following a different path. At very high temperatures, The soot fraction Fy, in units of soot volume per unit volume of gas, is related to above the range of interest for diesel combustion, a third nucleation process the number density N and the volume-mean diameter of the soot particles by seems likely that involves carbon vapor.7º A simple mechanistic model for nucleation in the low and intermediate tem- Fy = " Nd3 (11.37) perature ranges which has considerable experimental support for its basic fea- tures has been advanced by Graham et al.82 It is illustrated in Fig. 11-47. At low d is the actual diameter of the spherules, or the diameter of a sphere of equivalent temperatures, an aromatic hydrocarbon can produce soot via a relatively fast volume to an agglomerated particle. The rate of change of particle number direct route that involves condensation of the aromatic rings into a graphitelike density with time t can be written structure. Above about 1800 K, however, a slower, less-direct route is favored that entails ring breakup into smaller hydrocarbon fragments. These fragments dN then polymerize to form larger unsaturated molecules that ultimately produce dt = N, - N. (11.38) ........... 640 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 641 sion and surface growth rates have diminished, the resulting particles resemble a cluster in which the original spherules retain much of their individual identity. Fv After surface growth essentially ceases, continued coalescence of the soot particles N results in the formation of chainlike structures of discrete spherules. This suggests electrostatic forces are significant. Positive charge measured on these particle chains is claimed to be the cause of their chainlike structure.70, 73 This latter Soot volume fraction Fy , particle size d, particle number density N, hydrogen/carbon ratio coalescence once surface growth ceases is termed aggregation. It has been shown experimentally that during coagulation the rate of decrease of particle number density was proportional to the product of a coagu- lation coefficient and the square of the number density: FIGURE 11-48 dN dt = KN2 Variation in soot volume fraction Fy, particle (11.39) 0 number density N, particle size d, and soot 0 Time - hydrogen/carbon ratio with time in a flame.70 This is the Smoluchowski equation for coagulation of a liquid colloid. Based on brownian motion, this equation is applicable when the Knudsen number (ratio of mean free path to particle diameter) exceeds 10. K depends on such factors as where N ,, is the rate at which fresh nuclei appear and Na is the rate of agglomer- particle size and shape, size distribution, and the temperature, pressure, and ation of spherules or particles that collide and stick. At the peak of the N curve, density of the gas. Equation (11.39) has been used to predict coagulation rates in N, = Na. To the left of the peak, N, > N ., the particle diameter remains essen- low-pressure sooting flames.73,80 It has also been modified so that it applies tially constant at the minimum detectable diameter and the (small) rise in soot where the particle size and mean free path are comparable by using a more volume is dominated by nucleation. To the right of the peak in the N curve, complex expression for K (see Ref. 83). These studies show that under conditions Na > N ,. The number of agglomerating collisions is high because of the high approximating those in engine flames, the fraction of the initial number density number density; at the same time nucleation ends because there is enough dis- No remaining at time t is given approximately by persed surface area for gaseous deposition of hydrocarbon intermediates so the probability of generating new nuclei falls to zero. With nucleation halted slightly N No ~ ( KN t ) - 1 (11.40) to the right of the N curve peak, all the subsequent increase in soot volume fraction (the majority) stems from surface growth. To the right of the N curve Thus as t increases, N/No decreases rapidly. Although these coagulation calcu- peak, the number density falls in the case illustrated by three orders of magni- lations are simplistic (in that many of the assumptions made are not strictly valid tude. This is the result of agglomeration, which is responsible for a portion of the since soot particles are not initially distributed homogeneously in the combustion increase in particle diameter. Agglomeration does not contribute to the rise in space, they are not monodisperse, and surface growth and oxidation may be soot volume fraction, Fy. Surface growth that takes place on nuclei and on taking place during agglomeration), an overall conclusion is that the rate of coag- spherules is responsible for forming the concentric shells (somewhat distorted and ulation of spherules and particles to larger particles is very sensitive to number warped) that constitute the outer portions of spherules and which are distinct density. Thus the number of particles decreases rapidly with advancing crank from the less-organized spherule center (see Figs. 11-40 and 11-41). Surface angle in the diesel engine during the early part of the expansion process (see Fig. growth on agglomerated particles may partly fill in the crevices at the junctures 11-44) and agglomeration is essentially complete well before the exhaust valve of adjoining spherules to provide the nodular structure evident in Fig. 11-40.70 opens. Once particles have formed, interparticle collisions can lead to agglomer- Throughout the soot formation process in a flame, the H/C ratio of the ation, thereby decreasing the number of particles and increasing their size. Three hydrocarbons formed in the pyrolyses and nucleation process and of the soot types of agglomeration have been identified in soot formation. During the early particles continually decreases. The H/C ratio decreases from a value of about 2, stages of particle growth, collision of two spherical particles may result in their typical of common fuels, to of order 1 in the youngest soot particles that can be coagulation into a single spheroid. This is easy to visualize in hydrocarbon pyrol- sampled, and then to 0.2 to 0.3 once surface growth has ceased in the fully ysis where the beginnings of a soot particle may have the viscosity of a tarry agglomerated soot.80 The latter stages of this process are indicated in Fig. 11-48. liquid.7º Also, when the individual particles are small, rapid surface growth will The addition of mass to the soot particles occurs by reaction with gas-phase quickly restore the original spherical shape.73 This process occurs up to diam- molecules. The reacting gas-phase hydrocarbons appear to be principally acety- eters of about 10 nm. On the other hand, if spherules have solidified before colli- lenes, with larger polymers adding faster than the smaller. Small polyacetylenes 642 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 643 undergo further polymerization in the gas phase, presumably by the same mecha- TABLE 11.10 nism leading to nucleation. As a result of preferential addition of the larger poly- Rate constants for Nagle and Strickland- mers, the H/C ratio of the particles decreases toward its steady-state value. Thus Constable soot oxidation mechanism84 most of the polyacetylenes added must be of very high molecular weight or dehy- Rate constant Units drogenation must also take place. 73, 80 k = 20 exp (- 15,100/T) g/cm2 . s . atm kg = 4.46 x 10-3 exp (-7640/T) g/cm2 . s . atm 11.5.5 Soot Oxidation kg = 1.51 x 105 exp (-48,800/T) g/cm2 · s kz = 21.3 exp (2060/T) atm~1 In the overall soot formation process, shown schematically in Fig. 11-46, oxida- tion of soot at the precursor, nuclei, and particle stages can occur. The engine cylinder soot-concentration data reviewed in Sec. 11.5.3 indicate that a large frac- tion of the soot formed is oxidized within the cylinder before the exhaust process dation rate w (g C/cm2 . s): commences. In the discussion of diesel combustion movies in Sec. 10.3.1, dark W brown regions were observed in the color photographs (see color plate, Fig. 12 KAPO2 x + KB Po2(1 - x) (11.41) 10-4); these were interpreted as soot particle clouds, and were seen to be sur- 1 + kz Por rounded by a diffusion flame which appeared white from the luminosity of the where x is the fraction of the surface occupied by type A sites and is given by high-temperature soot particles consumed in this flame. As air mixed with this soot-rich region, the white flame eradicated the dark soot clouds as the particles x = ( 1 +- KT ) - 1 were burned up. Po2 KB / (11.42) In general, the rate of heterogeneous reactions such as the oxidation of soot The empirical rate constants determined by Nagle and Strickland-Constable for depends on the diffusion of reactants to and products from the surface as well as this model are listed in Table 11.10. According to this mechanism, the reaction is the kinetics of the reaction. For particles less than about 1 um diameter, diffu- first order at low oxygen partial pressures, but approaches zero order at higher sional resistance is minimal. The soot oxidation process in the diesel cylinder is pressures. At a given oxygen pressure, the rate initially increases exponentially kinetically controlled, therefore, since particle sizes are smaller than this limit. with temperature (equivalent activation energy is ka/kz or 34,100 cal/mol). There are many species in or near the flame that could oxidize soot: examples are Beyond a certain temperature the rate decreases as the thermal rearrangement O2, O, OH, CO2, and H2O. Recent reviews of soot formation70, 73,80 have con- favors formation of less reactive B sites. When, at sufficiently high temperature, cluded that at high oxygen partial pressures, soot oxidation can be correlated the surface is completely covered with B sites, the rate is first order in oxygen with a semiempirical formula based on pyrographite oxidation studies. For fuel- partial pressure and increases again with temperature.80 rich and close-to-stoichiometric combustion products, however, oxidation by OH Park and Appleton84 have compared this formula with oxidation rate data has been shown to be more important than O2 attack, at least at atmospheric obtained from pyrographite samples, carbon black particles, and with the avail- pressure. able flame soot oxidation data. Figure 11-49 shows both the soot oxidation rate It is argued on the basis of structural similarities that the rates of oxidation predicted by Eqs. (11.41) and (11.42) as a function of temperature and oxygen of soot and of pyrographites should be the same. This is a significant simplifica- partial pressure, and the above-mentioned data. The formula correlates the data tion. It has proved difficult to follow the oxidation of soot aerosols in flames, and shown to within a factor of 2. Under diesel engine conditions, the O2 partial if care is taken to avoid diffusional resistance, studies of bulk samples of pyro- pressure can be high ( ~several atmospheres), as can the temperatures of close-to- graphite can then be used as a basis for understanding soot oxidation. The semi- stoichiometric mixtures (< 2800 K). empirical formula of Nagle and Strickland-Constable has been shown84 to Equations (11.41) and (11.42) have been used to estimate the amount of soot correlate pyrographite oxidation for oxygen partial pressures po, < 1 atm and that can be oxidized in a typical IDI diesel engine. It was assumed that soot was temperatures between 1100 and 2500 K. This formula is based on the concept present in stoichiometric combustion products at selected times in the cycle and that there are two types of sites on the carbon surface available for O2 attack. that mixing with air leaned out the burned gas mixture at different rates until the For the more reactive type A sites, the oxidation rate is controlled by the fraction overall fuel/air equivalence ratio was reached. The surface recession rate during of sites not covered by surface oxides (and therefore is of mixed order, between 0 this process was computed. Figure 11-50 shows sample results at an engine speed and 1 in po2). Type B sites are less reactive, and react at a rate which is first order of 1600 rev/min and an overall cylinder equivalence ratio of 0.58. Fast, interme- in po2. A thermal rearrangement of A sites into B sites is also allowed (with rate diate, and slow mixing occurred in 30, 70, and 140º, respectively. The surface constant kr). A steady-state analysis of this mechanism gives a surface mass oxi- recession rate rises to a maximum as po, increases and then decreases as the 644 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 645 0.06 0.06 T, K Early-burning elements 4000 3000 2000 1600 1400 1200 Intermediate mixing rate = 0.5 atm 0.05 0.05 Po Fast mixing 0.15 Nagle and Strickland- 0.05 Constable formula 0.04 Early ).04 2 O O Slow mixing .03 Carbon recession rate, mg/mm- . s 1.03 Carbon recession rate, mg/mm2 . s -3 0.02 ).02 Late 0.01 0.01 H -4 - Specific soot oxidation rate log10 w, g/cm2 . s 0 30 60 90 120 0 30 60. 90 120 Po2, atm F, nm Crank angle, deg Crank angle, deg (a) (b) -5 o 0.5 + 0.1 4.5 Present 0 0.15 + 0.03 18 FIGURE 11-50 experiment < 0.05 + 0.01 4.5 Soot particle burnup rate in diesel combustion environment: (a) in early- and late-burned fuel-air 0.04 - 0.37 Fenimore and Jones elements with intermediate mixing rate; (b) for fast and slow mixing for early-burning elements.83 -6- ~10-4 ///2 0.04 - 0.1 Lee, Thring, and Beer more important. For the late mixing element shown (mixing lean of stoichiomet- - 7 ric at 40º ATC), the total carbon mass oxidized is only 40 percent of that for the 3 4 5 6 8 9 × 10-4 early mixing calculation. This is due primarily to the decreasing gas temperatures Reciprocal temperature, K-1 as the expansion stroke proceeds, and not the longer time available for burnup.83 10- For a spherical particle, the mass burning rate w (g/cm2.s) can be con- verted to a surface recession rate using Nagle and Strickland-Constable dr 10-2 semiempirical formula - w dt p wc Por where p is the density (~2 g/cm3). The integrated values of w(t) when divided by Specific soot oxidation rate log10 w, g/cm2 . s p then give the maximum radius of a soot particle that can be burned up. Inte- .0-31 T = 2500 $ 100 K grated values of 0.1 ug/mm2 (estimated for TC start of burnup) correspond to a O Soot radius F = 4.5 nm radius of about 50 nm or diameter of 100 nm. Individual spherule diameters are A Soot radius F = 18 nm about 30 nm, so soot which mixes with air early in the expansion stroke is likely 0-4 1 10 100 to be fully burned. Thus the soot present in the exhaust would be expected to 0.01 0.1 Oxygen partial pressure po ,, atm come from regions which mix with air too late for the oxidation rate to be suffi- FIGURE 11-49 cient for particle burnup. Specific soot oxidation rate measurements and predictions as a function of temperature and oxygen Agglomeration will have an indirect influence on the amount of soot oxi- partial pressure. 84 dized through its effect on surface area. In the limiting case of a spherical cluster, n monodisperse spherules (10 < n < 100) can be imagined as compacted into a falling gas temperature more than offsets the increasing oxygen concentration. single solid sphere of equal volume. Alternatively, the same n spherules can be While the shape of the recession rate versus time curves depends on the mixing imagined compacted into a cylinder of diameter equal to that of the original rate, the total amount of carbon burned (the area under each curve in Fig. spherules. Since oxidative attack is essentially an exterior surface phenomenon, 11-50b) is about the same (0.1 ug/mm2). However, the point in the cycle at which the surface/volume ratio is the appropriate measure of the effect of particle shape the soot-containing burned gas mixture passes through stoichiometric is much on soot mass burnup rate. It can be shown that the surface/volume ratios for the 646 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 647 single sphere, cylinder, and individual spherule are in the ratio n 1/3, 3, and 1, respectively. Thus agglomeration will decrease the relative oxidation rate. In the limit spherical clusters are less desirable than a chain; the larger the cluster the bigger the relative reduction in surface area. However, the densely packed spher- Extractable ule limit does not appear to be approached in practice. A specific surface area, of fraction about 200 m2/g for diesel soot, has been measured.85 A smooth-surfaced 30-nm Particulates, mg/m3 at STP diameter spherule with a 2-g/cm3 density has a surface/mass ratio of 100 m2/g; the measured value is about twice this value, indicating low porosity and an Nonextractable agglomerate structure which is loosely rather than densely packed.83 fraction Equation (11.41) shows a maximum recession rate in combustion products corresponding to a fuel/air equivalence ratio of about 0.9. Recent evidence shows FIGURE 11-52 that in an atmospheric pressure environment with rich and close-to- Typical effect of dilution ratio on particulate mass 10 20 stoichiometric combustion products where O2 mole fractions are low, oxidation 5 50 100 emission and its partitioning between extractable Dilution ratio and nonextractable fractions. 70 by OH radical attack is much more significant than oxidation by O or O2. The OH radical may be important in oxidizing soot in the flame zone under close-to- stoichiometric conditions. where the total sample is partitioned into extractable and nonextractable frac- tions. The nonextractable fraction is the carbonaceous soot generated during combustion and is not affected by the dilution process. With no dilution (dilution 11.5.6 Adsorption and Condensation ratio of unity) the difference between the total and nonextractable mass is small; The final process in the particulate formation sequence illustrated in Fig. 11-46 is the bulk of the extractable fraction is acquired after the exhaust gas is mixed with adsorption and condensation of hydrocarbons. This occurs primarily after the dilution air. Extensive studies of this dilution process have shown that both cylinder gases have been exhausted from the engine, as these exhaust gases are adsorption and condensation occur. Adsorption involves the adherence of mol- diluted with air. In the standard particulate mass emission measurement process ecules of unburned hydrocarbons to the surfaces of the soot particles by chemical this occurs in a dilution tunnel which simulates approximately the actual atmo- or physical (van der Waals) forces. This depends on the fraction of the available spheric dilution process. A diluted exhaust gas sample is filtered to remove the particle surface area occupied by hydrocarbons and on the partial pressure of the particulate. After equilibrating the collection filter at controlled conditions to gaseous hydrocarbons that drives the adsorption process. As the dilution ratio remove water vapor, the particulate mass is obtained by weighing. In the pre- increases from unity, the effect of decreasing temperature on the number of active scribed EPA procedure, the filter temperature must not exceed 52ºC. For a given sites dominates and, as shown in Fig. 11-52, the extractable fraction increases. At exhaust gas temperature, the filter (and sample) temperature depends on the dilu- high dilution ratios, the sample temperature becomes insensitive to the dilution tion ratio, as shown in Fig. 11-51. ratio (see Fig. 11-51) but the decreasing hydrocarbon partial pressure causes the The effect of the dilution ratio (and the dependent sample temperature) on extractable mass to fall again. Condensation will occur whenever the vapor pres- collected particulate mass is shown in Fig. 11-52 for a standard dilution tunnel, sure of the gaseous hydrocarbon exceeds its saturated vapor pressure. Increasing dilution decreases hydrocarbon concentrations and hence vapor pressure. However, the associated reduction in temperature does reduce the saturation pressure. High exhaust concentrations of hydrocarbons are the conditions where condensation is likely to be most significant, and the hydrocarbons most likely to 600 condense are those of low volatility. Sources of low-volatility hydrocarbons are 500H - T = 600 K the high-boiling-point end of the fuel, unburned hydrocarbons that have been pyrolyzed but not consumed in the combustion process, and the lubricating oil.7º T = 400 K EPA maximum Experiments with a passenger car IDI diesel, where the oil was tagged with Sample temperature, K 400- temperature a radioactive tracer, have shown that the oil can contribute from 2 to 25 percent 300 -- of the total particulate mass, with the greatest contribution occurring at high Ambient temperature FIGURE 11-51 speed. On average, over half of the extractable mass was traceable to the oil. All Effect of exhaust gas dilution ratio on the temperature 200 0 2 4 6 8 10 of the collected particulate sample as a function of the material traceable to the oil was found in the extractable fraction, indicating Dilution ratio engine exhaust temperature T ,. 70 that the oil did not participate in the combustion process. However, the oil is not 648 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 649 always a significant contributer: in another engine, fuel was the dominant source method. The accumulation of mass within the trap and the increase in exhaust of extractable material. 70, 71 manifold pressure during trap operation are major development problems. Diesel particulates, once trapped, can be burned up either by initiating oxidation within 11.6 EXHAUST GAS TREATMENT the trap with an external heat source or by using a trap which contains cata- 11.6.1 Available Options lytically active material. The operation of particulate traps is reviewed briefly in Sec. 11.6.4. Our discussion so far has focused on engine emissions. Further reductions in emissions can be obtained by removing pollutants from the exhaust gases in the engine exhaust system. Devices developed to achieve this result include catalytic converters (oxidizing catalysts for HC and CO, reducing catalysts for NO ,, and 11.6.2 Catalytic Converters three-way catalysts for all three pollutants), thermal reactors (for HC and CO), The catalytic converters used in spark-ignition engines consist of an active cata- and traps or filters for particulates. lytic material in a specially designed metal casing which directs the exhaust gas The temperature of exhaust gas in a spark-ignition engine can vary from flow through the catalyst bed. The active material employed for CO and HC 300 to 400ºC during idle to about 900ºC at high-power operation. The most oxidation or NO reduction (normally noble metals, though base metals oxides common range is 400 to 600ºC. Spark-ignition engines usually operate at fuel/air can be used) must be distributed over a large surface area so that the mass- equivalence ratios between about 0.9 and 1.2 (see Sec. 7.1). The exhaust gas may transfer characteristics between the gas phase and the active catalyst surface are therefore contain modest amounts of oxygen (when lean) or more substantial sufficient to allow close to 100 percent conversion with high catalytic activity. amounts of CO (when rich). In contrast, diesel engines, where load is controlled The two configurations commonly used are shown in Fig. 11-53. One system by the amount of fuel injected, always operate lean. The exhaust gas therefore employs a ceramic honeycomb structure or monolith held in a metal can in the contains substantial oxygen and is at a lower temperature (200 to 500ºC). exhaust stream. The active (noble metal) catalyst material is impregnated into a Removal of gaseous pollutants from the exhaust gases after they leave the engine highly porous alumina washcoat about 20 um thick that is applied to the pas- cylinder can be either thermal or catalytic. In order to oxidize the hydrocarbons sageway walls. The typical monolith has square-cross-section passageways with in the gas phase without a catalyst, a residence time of order or greater than inside dimensions of ~1 mm separated by thin (0.15 to 0.3 mm) porous walls. 50 ms and temperatures in excess of 600ºC are required. To oxidize CO, tem- The number of passageways per square centimeter varies between about 30 and peratures in excess of 700ºC are required. Temperatures high enough for some 60. The washcoat, 5 to 15 percent of the weight of the monolith, has a surface homogeneous thermal oxidation can be obtained by spark retard (with some loss area of 100 to 200 m2/g. The other converter design uses a bed of spherical in efficiency) and insulation of the exhaust ports and manifold. The residence ceramic pellets to provide a large surface area in contact with the flow. With time can be increased by increasing the exhaust manifold volume to form a pellet catalysts, the noble metal catalyst is impregnated into the highly porous thermal reactor (see Sec. 11.6.3). However, this approach has limited application. surface of the spherical alumina pellets (typically 3 mm diameter) to a depth of Catalytic oxidation of CO and hydrocarbons in the exhaust can be about 250 um. The pellet material is chosen to have good crush and abrasion achieved at temperatures as low as 250ºC. Thus effective removal of these pol- resistance after exposure to temperatures of order 1000ºC. The gas flow is lutants occurs over a much wider range of exhaust temperatures than can be directed down through the bed as shown to provide a large flow area and low achieved with thermal oxidation. The only satisfactory method known for the pressure drop. The gas flow is turbulent which results in high mass-transfer rates; removal of NO from exhaust gas involves catalytic processes. Removal of NO by in the monolith catalyst passageways, it is laminar. catalytic oxidation to NO2 requires temperatures <400ºC (from equilibrium considerations) and subsequent removal of the NO2 produced. Catalytic reaction OXIDATION CATALYSTS. The function of an oxidation catalyst is to oxidize CO of NO with added ammonia NH3 is not practical because of the transient varia- and hydrocarbons to CO, and water in an exhaust gas stream which typically tions in NO produced in the engine. Reduction of NO by CO, hydrocarbons, or contains ~ 12 percent CO2 and H2O, 100 to 2000 ppm NO, ~20 ppm SO2, 1 to H2 in the exhaust to produce N2 is the preferred catalytic process. It is only 5 percent O2, 0.2 to 5 percent CO, and 1000 to 6000 ppm C1 HC, often with feasible in spark-ignition engine exhausts. Use of catalysts in spark-ignition small amounts of lead and phosphorus. About half the hydrocarbons emitted by engines for CO, HC, and NO removal has become widespread. Catalysts are the SI engine are unburned fuel compounds. The saturated hydrocarbons (which discussed in Sec. 11.6.2. comprise some 20 to 30 percent) are the most difficult to oxidize. The ease of Particulates in the exhaust gas stream can be removed by a trap. Due to the oxidation increases with increasing molecular weight. Sufficient oxygen must be small particle size involved, some type of filter is the most effective trapping present to oxidize the CO and HC. This may be supplied by the engine itself POLLUTANT FORMATION AND CONTROL 651 650 INTERNAL COMBUSTION ENGINE FUNDAMENTALS running lean of stoichiometric or with a pump that introduces air into the exhaust ports just downstream of the valve. Venturi air addition into the exhaust port using the pressure pulsations generated by the exhaust process can also be used to add the required air. Because of their high intrinsic activity, noble metals are most suitable as the catalytic material. They show higher specific activity for HC oxidation, are more thermally resistant to loss of low-temperature activity, and are much less deacti- vated by the sulfur in the fuel than base metal oxides. A mixture of platinum (Pt) Metal mesh and palladium (Pd) is most commonly used. For the oxidation of CO, olefins, Seal Shell and methane, the specific activity of Pd is higher than that of Pt. For the oxida- Catalyst tion of aromatic compounds, Pt and Pd have similar activity. For oxidation of paraffin hydrocarbons (with molecular size greater than C3), Pt is more active than Pd. Pure noble metals sinter rapidly in the 500 to 900ºC temperature range experienced by exhaust catalysts. Since catalytic behavior is manifested exclu- sively by surface atoms, the noble metals are dispersed as finely as possible on an (a) inert support such as y-A2O3 which prevents particle-to-particle metal contact and suppresses sintering. The particle size of the noble metal particles in a fresh catalyst is less than 50 nm. This can increase to ~100 nm when the catalyst is Converter shell - Insulation Outer wrap exposed to the high temperatures of the exhaust in vehicle operation. Typical noble metal concentrations in a commercial honeycomb catalyst are between 1 Retainer and 2 g/dm3 of honeycomb volume, with Pt/Pd = 2 on a weight basis. As a rough rule of thumb, the ceramic honeycomb volume required is about half the engine displaced volume. This gives a space velocity through the converter (volume flow rate of exhaust divided by converter volume) over the normal engine operating range of 5 to 30 per second.68 The conversion efficiency of a catalyst is the ratio of the rate of mass removal in the catalyst of the particular constituent of interest to the mass flow rate of that constituent into the catalyst : e.g ., for HC, neat =- MHC. in - MHC, out = 1 - MHC , out (11.43) Fill plug- Catalyst Insulation MHC , in MHC, in The variation of conversion efficiency of a typical oxidizing catalytic converter with temperature is shown in Fig. 11-54. At high enough temperatures, the steady-state conversion efficiencies of a new oxidation catalyst are typically 98 to 99 percent for CO and 95 percent or above for HC. However, the catalyst is ineffective until its temperature has risen above 250 to 300ºC. The term light-off temperature is often used to describe the temperature at which the catalyst Inlet gas Catalytic pellets - Outlet gas becomes more than 50 percent effective. (b) The above numbers apply to fresh noble metal oxidation catalysts; as cata- FIGURE 11-53 lysts spend time in service their effectiveness deteriorates. Catalysis involves the Catalytic converters for spark-ignition engine emission control: (a) monolith design; (b) pelletized adsorption of the reactants onto surface sites of high activity, followed by chemi- design.62 cal reaction, then desorption of the products. Catalyst degradation involves both the deactivation of these sites by catalyst poisons and a reduction in the effective area of these sites through sintering. Poisoning affects both the warm-up and 652 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 653 100 80 80 100 CO HC 60 60 80 Catalyst conversion efficiency, % (methane excluded) HC conversion at 500ºC, % HC conversion at 500ºC, % 40 40 60 40 FIGURE 11-55 20 20 HC conversion efficiency as a function FIGURE 11-54 20 of lead concentration on catalyst. Conversion efficiency for CO and HC as a Total HC conversion on left; non- 200 300 400 500 function of temperature for typical oxidizing O OC catalytic converter.62 0.01 0.1 1.0 10 methane HC conversion on right. Temperature, Lead on catalyst, weight % 0.001-0.013 g Pb/dm3 in fuel. 68 steady-state performance of the catalyst. When poisoning occurs, catalytic activ- sure, is inhibited by CO, olefins, and NO, and increases as the O2 partial pressure ity is impeded through prolonged contact with interfering elements that either is decreased to near-stoichiometric values.68 physically block the active sites or interact chemically with the active material. It will be apparent from the above that two extremely important consider- The lead in fuel antiknock agents and the phosphorus in oil additives are the ations for successful use of catalysts for automotive applications are the test pro- most important poisons. Though lead antiknock agents are not added to the cedure that is used to measure emissions and the methods used to determine if gasoline used with catalyst-equipped vehicles, this "unleaded" fuel can be con- the catalyst has the required durability. The U.S. Federal Test Procedure requires taminated with small amounts (~10 mg Pb/dm3) from the fuel distribution that the vehicle under test be at a temperature of 16 to 30ºC for 12 hours prior to system. Between 10 and 30 percent of the lead in the fuel ends up on the catalyst. the test and that emissions are measured from the time the ignition key is turned Its effect on catalyst conversion efficiency depends on the amount of lead on the on until the test has ended. In spark-ignition engines the mixture fed into the catalyst, as shown in Fig. 11-55. Lead depresses the catalytic oxidation of HC to engine during start-up is enriched substantially (carburetors have a choke to a greater extent than oxidation of CO. The oxidation activity of saturated hydro- accomplish this; additional fuel is injected with port or manifold fuel injection). carbons is particularly depressed. The extent of the poisoning that results from The rationale is that if sufficient fuel is added to the inlet air, enough will evapo- traces of critical elements in the fuel and oil depends on which elements are rate to start the engine. However, until the rest of the fuel is consumed, the engine present and the amounts absorbed, as well as the composition of the catalyst and then runs rich and emits high concentrations of CO and HC. The catalyst is cold its operating conditions (especially its temperature).68 Sintering is promoted by at this time, and until it warms up, these emissions will pass through without exposure of the catalyst to high operating temperatures. It involves the migration reaction. It is important that the catalyst be brought to its light-off temperature and agglomeration of sites, thus decreasing their active surface area. Sintering as quickly as possible (preferably in less than 60 s) and that mixture enrichment slows warm-up but has minimal effect on the steady-state conversion efficiency. during start-up be held to a minimum. Thus catalysts should have low thermal The oxidation kinetics of CO over Pt and Pd noble metal catalysts can be inertia for rapid warm-up and low light-off temperatures for CO and HC, so they described by become effective quickly. The closer they are placed to the engine the faster they will reach light-off. However, they will then experience higher temperatures when d[Co] KIPco Po2 (11.44) fully warmed up and so be more susceptible to thermal degradation. While it is dt (1 + K2 Pco + K3 PHC)2(1 + K4 PNO) not too difficult to prepare catalysts that are highly effective when fresh, it is much more difficult to maintain effectiveness over extended mileage (50,000 miles) where K1 to K4 and n are constants at any given temperature, and pco, Po2, PHC, in which the catalyst is exposed to high temperatures and catalyst poisons. These and PNo are the partial pressures of carbon monoxide, oxygen, hydrocarbons, and can degrade both cold-start and warmed-up performance. Also, catalyst dura- nitric oxide, respectively. A similar relationship can be written for the olefinic and bility is affected by engine durability. Any engine malfunction that will expose the aromatic HC oxidation rate (these being the most reactive hydrocarbons). These catalyst to excessive amounts of unburned fuel (such as ignition failure, misfire relationships incorporate the fact that the rates of CO and HC oxidation are with too lean a mixture, or excessively rich operation) will severely overheat the inhibited by high CO and reactive HC concentrations, and that NO concentra- catalyst. tions in the range 0 to 1000 ppm strongly inhibit oxidation also. The oxidation Oxidation-catalyst-equipped vehicles may emit sulfuric acid aerosol. rate of paraffin hydrocarbons varies with the first power of the HC partial pres- Unleaded gasoline contains 150 to 600 ppm by weight of S, which leaves the 654 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 655 100 TABLE 11.11 Possible NO reactions under Thermodynamic equilibrium reducing conditions68 80 | 1. NO + CO - IN2 + CO2 2. 2NO + 5CO + 3H2O -+ 2NH3 + 5CO2 3. 2NO + CO -+ N20 + CO2 60- 4. NO + H2 - 4N2 + H2O 5. 2NO + 5H2 - 2NH3 + 2H2O 6. 2NO + H2 - N2O + H2O SO2 conversion, % 40 Reactions 3 and 6 occur at 200ºC, which is below that usually found in auto exhausts. 20 FIGURE 11-56 A Pt:Pd 2/1 SO, conversion to SO3 as a function of temperature formation under rich operation in the first bed must be small in this two-bed o Pt:Rh 10/1 On Pd catalyst with 5% O2 concentration and no reducing gases system because the second (oxidation) catalyst readily oxidizes NH3 back to NO. 0 present. Space velocity (volume flow per unit Reduction of NO by CO or H2 can be accomplished by base metal catalysts (e.g ., :200 400 600 800 volume) ~ 10 s 1. Results for Pt-Pd, Pt-Rh, and Pd CuO, NiO) in the temperature range 350 to 600ºC. However, these catalyst Temperature, ºC catalysts.68 materials are deactivated by sulfur and have shown limited thermal stability when used in vehicle exhausts. Alumina-supported noble metal catalysts reduce NO with CO-H2 mixtures. Their NO-reduction activity is in the order combustion chamber as SO2. This SO2 can be oxidized by the catalyst to SO3 Ru > Rh > Pd > Pt. Ruthenium (Ru) and rhodium (Rh) produce considerably which combines with water at ambient conditions to form an H2SO4 aerosol. less NH3 than Pd or Pt under slightly rich conditions. While these properties The SO3 can be chemisorbed on the alumina catalyst surface; when large pellet make ruthenium a desirable NO catalyst, it forms volatile oxides under oxidizing beds are used, considerable storage of SO3 at temperatures <500ºC can occur. conditions which results in loss of ruthenium from the alumina support.68 At higher catalyst temperatures, this stored SO3 is emitted as an SO3-SO2 mixture. SO3 production can be controlled by lowering or raising the catalyst THREE-WAY CATALYSTS. If an engine is operated at all times with an air/fuel temperature. Figure 11-56 shows that at low temperatures SO3 production is ratio at or close to stoichiometric, then both NO reduction and CO and HC kinetically limited; at high temperatures SO3 production is thermodynamically oxidation can be done in a single catalyst bed. The catalyst effectively brings the limited. Palladium and rhodium produce less SO3 than Pt and have comparable exhaust gas composition to a near-equilibrium state at these exhaust conditions; HC and CO catalytic activity. By decreasing oxygen concentrations leaving the i.e ., a composition of CO2, H2O, and N2. Enough reducing gases will be present catalyst to ~1 percent, SO3 production can substantially reduced.68 to reduce NO and enough O2 to oxidize the CO and hydrocarbons. Such a catalyst is called a three-way catalyst since it removes all three pollutants simulta- NO CATALYSIS. NO is removed by reduction using the CO, hydrocarbons, and neously. Figure 11-57 shows the conversion efficiency for NO, CO, and HC as a H2 in the exhaust. The reactions are shown in Table 11.11. No catalyst is avail- function of the air/fuel ratio. There is a narrow range of air/fuel ratios near stoi- able for the decomposition of NO to O2 and N2 (thermodynamically favored at chiometric in which high conversion efficiencies for all three pollutants are exhaust temperatures) which is sufficiently active for use in engine exhausts. NO achieved. The width of this window is narrow, about 0.1 air/fuel ratios (7 x 10-3 reduction can be carried out under rich conditions where there is an excess of in equivalence ratio units) for catalyst with high mileage use, and depends on reducing species over oxidizing species. The catalyst used under these conditions catalyst formulation and engine operating conditions. is referred to as an NO reduction catalyst. Such a system requires a follow-up This window is sufficiently narrow to be beyond the control capabilities of oxidation catalyst, together with addition of air from an air pump before the an ordinary carburetor, though it can sometimes be achieved with sophisticated oxidation catalyst, to remove the remaining CO and hydrocarbons. Such a carburetors and fuel-injection systems. Thus closed-loop control of equivalence two-bed system can remove all three pollutants (NO, CO, and HC) from the ratio has been introduced. An oxygen sensor in the exhaust is used to indicate exhaust. However, the rich operation necessary for NO reduction results in a fuel whether the engine is operating on the rich or lean side of stoichiometric, and consumption penalty and constrains the performance of the NO catalyst since a provide a signal for adjusting the fuel system to achieve the desired air-fuel fraction of the NO removed is converted to ammonia NH3 rather than N2 . NH3 mixture (see Sec. 7.4). Holding the equivalence ratio precisely on the chosen near- 656 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 657 100 regime, the three-way catalyst consumes all the oxygen that is present in the exhaust, and as a consequence removes an equivalent amount of CO, H2, and NO, hydrocarbons; it is thought that the H2 is removed first. In addition, the water- 80- gas shift reaction HC - CO + H2O = H2 + CO2 60 CO. and the steam-reforming reaction Catalyst efficiency, % 80% efficiency 40 -air/fuel ratio Hydrocarbon + H2O -> CO, CO2, H2 window can consume CO and HC. The exhaust contains an H2/CO ratio of about 3 (see 20- Sec. 4.9.1), where the equilibrium ratio at 500ºC is about 4. Considerable CO Stoichiometric FIGURE 11-57 removal can be expected if the water-gas shift equilibrium is approached. Plati- air/fuel ratio Conversion efficiency for NO, CO, num is active in promoting this equilibrium. For large molecular weight paraffin and HC for a three-way catalyst as 14.3 14.4 14.5 14.6 14.7 14.8 14.9 hydrocarbons, and for olefins and aromatic hydrocarbons, the equilibrium for the Rich Lean a function of exhaust gas air/fuel steam-reforming reactions lies to the right. This reaction can therefore lead to Air/fuel ratio ratio. 68 considerable hydrocarbon removal. Rhodium is particularly active in the steam- reforming reaction; platinum is also active.68 The conversions of NO, CO, and hydrocarbons in a three-way catalyst stoichiometric value is not a practical expectation of such a feedback system, and operated with cyclical variations in equivalence ratio are larger than estimates the equivalence ratio oscillates around the set point in an approximately periodic based on summation of steady-state values during the cycle. At least part of the manner as the fuel flow is varied. Experimental data show that there is a con- improved performance is thought to be due to the ability of the catalyst to siderable widening of the air/fuel ratio window where all three pollutants are undergo reduction-oxidation reactions. Such a catalyst component is usually effectively removed, with cyclic variation of the fuel flow. The maximum conver- referred to as an oxygen-storage component. In its oxidized state it can provide sion in the middle of the window is reduced, however, from its value when there oxygen for CO and hydrocarbon oxidation in a rich exhaust gas environment, are no fluctuations. The effect of fluctuations depends on the frequency; fre- and in the process be reduced. When the exhaust cycles to lean conditions, this quencies of about 0.5 to 1 hertz are most effective and the usable window (at reduced component can react with O2 or NO (which removes NO directly or lower conversion efficiencies) can be broadened to about 1 air/fuel ratio. Some of indirectly by reducing the O2 concentration). The oxidized component can then the benefits of fluctuations in equivalence or air/fuel ratios are available even oxidize CO and HC in the next rich cycle, etc. Components such as ReO2 or without any deliberate attempt to produce such variations with closed-loop feed- CeO2 which exhibit this "redox" behavior can be included in three-way catalyst back. Open-loop systems exhibit variations in the air/fuel ratio during normal formulations. Commercial three-way catalysts contain platinum and rhodium vehicle operation. (the ratio Pt/Rh varying substantially in the range 2 to 17) with some A2O3, Because of these cyclic variations in exhaust gas composition about a set NiO, and CeO2 . Alumina is the preferred support material.68 point close to stoichiometric, it is desirable that the catalyst be able to reduce NO when a slight excess of oxygen is present (on the lean side) and remove CO and HC when there is a slight deficiency of oxygen (on the rich side). Rhodium is 11.6.3 Thermal Reactors the principal ingredient used in commercial catalysts to remove NO. It is very active for NO reduction, is much less inhibited by CO and sulfur compounds, In Secs. 11.3 and 11.4.2 it was explained that oxidation of CO and HC occurred and produces less NH3 than Pt. To remove NO under slightly lean-of- during the expansion and exhaust processes in the cylinder of a conventional stoichiometric conditions, the catalyst must react the CO, H2, or HC with NO spark-ignition engine and, under certain circumstances, in the exhaust system. rather than with O2, as the exhaust gas passes through the catalyst bed. Oxidation after passage through the exhaust port can be enhanced with a thermal Rhodium shows some NO reduction activity slightly lean of stoichiometric. On reactor-an enlarged exhaust manifold that bolts directly onto the cylinder head. the rich side, the three-way catalyst window is determined by hydrocarbon and Its function is to promote rapid mixing of the hot exhaust gases with any second- CO removal. Platinum is most commonly used for HC and CO oxidation; it has ary air injected into the exhaust port (required with fuel-rich engine operation to good activity under stoichiometric and slightly lean conditions. When sufficient produce a net oxidizing atmosphere), to remove nonuniformities in temperature rhodium is present, the participation of Pt in NO removal is minimal. In the rich and composition in the exhaust gases, and to retain the gases at a high enough 658 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 659 Exhaust gas Outer shell tial reductions in CO emissions are difficult to achieve. For very lean operation, HC burnup becomes marginal. A practical limitation to reactor effectiveness with fuel-rich engine oper- ation is mixing of secondary air and engine exhaust gases in the exhaust port and the reactor core. The secondary air flow with a conventional air pump is effec- tively shut off by the exhaust blowdown process, and virtually no oxidation occurs in the exhaust port because the air and exhaust gases are segregated. Heat shield Core Mixing in the reactor itself is promoted by suitably arranging the reactor inlet To exhaust system and exit ports and by using baffles. In systems with conventional secondary air pumps, maximum reductions in CO and HC occur with 10 to 20 percent excess Air air in the mixture. However, even with very high reactor core gas temperatures, 100 percent HC and CO oxidation is not achieved due to incomplete mixing. Improved control of secondary air flow has been shown to increase significantly FIGURE 11-58 CO emissions burnup. Schematic of exhaust thermal reactor for HC and CO oxidation. 11.6.4 Particulate Traps An exhaust treatment technology that substantially reduces diesel engine particu- late emissions is the trap oxidizer. A temperature-tolerant filter or trap removes temperature for sufficient time to oxidize much of the HC and CO which exits the particulate material from the exhaust gas; the filter is then "cleaned off" by the cylinder. An example of a thermal reactor design is shown in Fig. 11-58. oxidizing the accumulated particulates. This technology is difficult to implement The temperature levels typically required for bulk gas oxidation of HC and because: (1) the filter, even when clean, increases the pressure in the exhaust CO in a reactor are about 600 and 700ºC, respectively. Note that they are con- system; (2) this pressure increase steadily rises as the filter collects particulate siderably higher than those required for equivalent conversion in a catalytic con- matter; (3) under normal diesel engine operating conditions the collected particu- verter and that higher temperatures are required for CO oxidation than for HC late matter will not ignite and oxidize; (4) once ignition of the particulate occurs, oxidation. The exhaust gas temperature in the manifold of a conventional engine the burnup process must be carefully controlled to prevent excessively high tem- is not sufficient to achieve any substantial reduction in engine exhaust port emis- peratures and trap damage or destruction. Trap oxidizers have been put into sions. To achieve greater reductions, the reactor must be designed to reduce heat production for light-duty automobile diesel engines. Their use with heavy-duty losses and increase residence time. In addition, to achieve rapid warm-up after diesel engines poses more difficult problems due to higher particulate loading and engine start, a low thermal inertia reactor is desirable. Typically, a thin steel liner lower exhaust temperatures. acts as the core of the reactor inside a cast-iron outer casing; with suitably Types of particulate filters include: ceramic monoliths, alumina-coated wire arranged flow paths, this construction holds heat losses to a minimum by ther- mesh, ceramic foam, ceramic fiber mat, woven silica-fiber rope wound on a mally isolating the core. porous tube. Each of these has different inherent pressure loss and filtering effi- The effectiveness of the reactor depends on its operating temperature, the ciency. Regeneration of the trap by burning up the filtered particulate material availability of excess oxygen mixed throughout the reacting gases, and the can be accomplished by raising its temperature to the ignition point while pro- reactor volume. The operating temperature depends on the reactor inlet gas tem- viding oxygen-containing exhaust gas to support combustion and carry away the perature, heat losses, and the amount of HC, CO, and H2 burned up in the heat released. Diesel particulate matter ignites at about 500 to 600ºC. This is reactor. This latter factor is important: 1.5 percent CO removal results in a above the normal temperature of diesel exhaust so either the exhaust gas flowing 220 K temperature rise. As a consequence, reactors with fuel-rich cylinder through the trap during regeneration must be heated (positive regeneration) or exhaust gas and secondary air give greater fractional reductions in HC and CO ignition must be made to occur at a lower temperature with catalytic materials emissions than reactors with fuel-lean cylinder exhaust (which do not require any on the trap or added to the fuel (catalytic regeneration). Catalytic coatings on the secondary air). As has already been explained, a higher core gas temperature is trap reduce the ignition temperature by up to 200ºC. required to burn up the same fraction of CO which enters the reactor as of HC Figure 11-59 shows a ceramic-coated trap oxidizer mounted on the exhaust which enters. For lean engine exhaust gas, where the reactor core gas tem- system of a turbocharged IDI diesel engine. The trap is a ceramic honeycomb peratures are a hundred degrees K lower than under fuel-rich operation, substan- with half the cells closed at the inlet end and the other half of the cells closed at 660 POLLUTANT FORMATION AND CONTROL 661 INTERNAL COMBUSTION ENGINE FUNDAMENTALS 11.5. A three-way catalytic converter is used with the spark-ignition engine in Prob. 11.3. For 10 percent of the driving time, the catalyst is cold and ineffective, and does not reduce the engine's emissions. For 90 percent of the time, the catalyst is hot and has conversion efficiencies as given in Fig. 11-57. Estimate the average vehicle emissions of NO„, HC, and CO in grams per mile. 11.6. Figure 15-11 shows the variation in NO and HC emissions as concentrations (ppm) in the exhaust of a spark-ignition engine as a function of speed and load. Convert these data to graphs of indicated specific NO and HC emissions (g/kW . h) versus speed and imep. Assume nu (based on atmospheric air density) = imep (kPa) × 10-3. FIGURE 11-59 11.7. Use the data in Fig. 11-44 to estimate: Catalytic ceramic-monolith partic- ulate trap oxidizer mounted on (a) The exhaust particulate emissions as a fraction of the maximum particulate exhaust of turbocharged automo- loading during the cycle. bile diesel engine.86 (b) The maximum measured soot loading and the exhaust soot loading as fractions of the fuel carbon. (c) The equivalent sphere size of each soot particle at the number density peak (22º the exit end. Thus the particulate laden exhaust is forced to flow through the ATC) and in the exhaust. porous ceramic cell walls. The outside of the honeycomb is insulated and the trap Assume a particulate density of 2 g/cm3. Note that the gas volumes in Fig. 11-44 is mounted close to the engine to maintain as high a trap temperature as possible. are determined at standard temperature and pressure. The pressure drop across the unloaded trap increases from 0.02 atm at 1000 11.8. Explain the following emissions trends. Highest marks will be given for succinct rev/min to 0.15 atm at the maximum engine speed of 4500 rev/min. As the trap summaries of the important technical issues. loads up, the pressure drop increases, requiring more fuel to be injected to com- (a) Nitric oxide (NO) emissions from diesels and spark-ignition engines as the pensate for the loss in power. This leads to higher exhaust temperature which equivalence ratio is varied show significantly different behavior (see Figs. 11-9 and 11-16). Redraw these graphs on the same plot and explain the different eventually results in catalytic ignition of the particulate. The particulate oxida- trends for these two types of engines as o decreases on the lean side of stoichio- tion rate depends on the trap temperature. With suitable trap location and metric. design, the regeneration process is largely self-regulating. The particulate emis- (b) Recirculation of a fraction of the exhaust gases to the intake is used to control sions from the engine are reduced by 70 percent or more.86 engine nitric oxide emissions at part load. Exhaust gas recycle is usually more effective with spark-ignition engines than with diesels, as shown in Fig. P11-8. Explain why these trends are different. PROBLEMS 11.1. Figure 11-2 shows concentrations of NO, CO, and HC in a spark-ignition engine 1.0 exhaust as a function of fuel/air equivalence ratio. Assume the concentration scale Diesel is parts per million. Explain the trends shown as the mixture is first made richer and then leaner than stoichiometric. overall = 0.5 11.2. NO Figure 11-2 is for a spark-ignition engine. Construct a similar qualitative graph of (NO) no EGR NO, CO, and HC concentrations versus equivalence ratio for a direct-injection four-stroke cycle diesel engine. Spark-ignition 11.3. 11.3. A spark-ignition engine driving a car uses, on average, 120 grams of gasoline per OL mile traveled. The average emissions from the engine (upstream of the catalyst) are 10 20 30 1.5, 2, and 20 grams per mile of NO, (as NO2), HC, and CO, respectively. The EGR, % FIGURE P11-8 engine operates with a stoichiometric gasoline-air mixture. Find the average con- centrations in parts per million of NO„, HC (as ppm C1), and CO in the engine exhaust. (c) Brake specific particulate emissions from diesels are a major problem. Particu- 11.4. Calculate the average combustion inefficiency corresponding to the spark-ignition late emissions from conventional spark-ignition engines are negligible. Briefly engine emissions levels given in Prob. 11.3. Include any hydrogen you estimate explain why the particulate emission levels from these two types of engines are so different in magnitude. would be present in the exhaust stream. 662 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 663 (d) Diesels have low carbon monoxide (CO) emissions. Spark-ignition engine CO (b) Estimate approximately the time taken to reach equilibrium NO levels at emissions when averaged over a typical urban automobile trip (cold engine ¢ = 1, 2750 K and 3000 K, 5.5 MPa. start, warm-up, cruise, idle, acceleration, etc.) are substantial and require a cata- (c) If the stoichiometric mixture inducted into the engine reaches 3000 K and lyst for effective control. Explain this difference in average CO emissions 5.5 MPa after combustion, in the absence of any exhaust gas recirculation, cal- (upstream of any catalyst) from these two types of engines. culate the percentage of the exhaust that must be recycled to the intake (at the 11.9. The following questions refer to an engine with these geometric and operating char- initial intake temperature) to reduce the NO formation rate by a factor of 4 acteristics (see Fig. 11-26a): ( = 1.0; compression ratio = 8 : 1; bore = 100 mm; (assume the final pressure 5.5 MPa stays the same; of course, the final tem- stroke = 100 mm; piston diameter above top ring = 99.4 mm; distance from piston perature decreases as the exhaust gas is recycled). crown top to top ring = 9.52 mm; volumetric efficiency = 0.8; temperature in cylin- der at the start of compression = 333 K; pressure in cylinder at start of p = 5.5 MPa ¢ = 1.0, p = 5.5 MPa compression = 1 atm; mixture temperature before entering cylinder = 30ºC; brake Mole fraction Mole fraction specific fuel consumption = 300 g/kW . h. A substantial fraction of spark-ignition engine hydrocarbon emissions comes T(K) O N2 0 N2 NO from the crevice between the piston crown and cylinder wall. Gas is forced into this crevice as the cylinder pressure increases and flows out of this crevice as the cylin- 0.9 3000 2.1 x 10-3 0.73 2500 6 × 10-5 0.73 der pressure decreases. The gas in the crevice can be assumed to be at the wall 1.0 3000 1.5 x 10-3 0.73 2750 5 × 10-4 0.73 4 x 10-3 temperature, 400 K. The gas pushed into the crevice ahead of the flame is unburned 1.1 3000 1 × 10-3 0.73 3000 1.5 x 10-3 0.73 8 × 10-3 mixture; the gas pushed in behind the flame is burned mixture. About two-thirds of the crevice gas is unburned. The maximum cylinder pressure is 3 MPa. (a) Calculate the mass fraction of the cylinder gas which is in the crevice between the piston and cylinder wall and above the first piston ring, at the time of peak REFERENCES pressure. 1. Bowman, C. T.: "Kinetics of Pollutant Formation and Destruction in Combustion," Prog. (b) Assuming that half of the unburned fuel in this region is oxidized within the Energy Combust. Sci ., vol. 1, pp. 33-45, 1975. cylinder and a further one-third is oxidized in the exhaust port, calculate the 2. Lavoie, G. A ., Heywood, J. B ., and Keck, J. C.: "Experimental and Theoretical Investigation of engine HC emissions from this source in parts per million (ppm C1) by volume. Nitric Oxide Formation in Internal Combustion Engines," Combust. Sci. Technol ., vol. 1, pp. (c) Calculate the ratio of brake specific hydrocarbon emissions to brake specific 313-326, 1970. fuel consumption. 3. Heywood, J. B ., Fay, J. A ., and Linden, L. H.: " Jet Aircraft Air Pollutant Production and Disper- (d) Calculate the brake specific hydrocarbon emissions in grams of HC per sion," AIAA J ., vol. 9, no. 5, pp. 841-850, 1971. kilowatt-hour. 4. Newhall, H. K ., and Shahed, S. M.: "Kinetics of Nitric Oxide Formation in High-Pressure Flames," in Proceedings of Thirteenth International Symposium on Combustion, pp. 381-390, The 11.10. Nitric oxide, NO, forms via reactions (11.1) to (11.3). Reaction (11.1) is "slow" and Combustion Institute, 1971. reactions (11.2) and (11.3) are "fast," so the initial rate of formation of NO is given 5. Hilliard, J. C ., and Wheeler, R. W.: "Nitrogen Dioxide in Engine Exhaust," SAE paper 790691, by Eq. (11.8): SAE Trans ., vol. 88, 1979. d[NO] 6. Merryman, E. L ., and Levy, A.: "Nitrogen Oxide Formation in Flames: The Roles of NO, and Fuel Nitrogen," in Proceedings of Fifteenth International Symposium on Combustion, p. 1073, The dt = 2k+[N2]e[O]e Combustion Institute, 1975. where [ ] denote concentrations in gram-moles per cubic centimeter, k+ is the rate 7. Komiyama, K ., and Heywood, J. B.: "Predicting NO, Emissions and Effects of Exhaust Gas constant for reaction (11.1), and the factor of 2 enters because the N atom formed Recirculation in Spark-Ignition Engines," SAE paper 730475, SAE Trans ., vol. 82, 1973. 8. Alperstein, M ., and Bradow, R. L.: "Exhaust Emissions Related to Engine Combustion Reac- in (11.1) immediately reacts via (11.2) or (11.3) to give an additional NO molecule: tions," SAE paper 660781, SAE Trans ., vol. 75, 1966. -38,000 9. Starkman, E. S ., Stewart, H. E ., and Zvonow, V. A.: "Investigation into Formation and Modifi- kt = 7.6 x 1013 exp cm3/gmol . s cation of Exhaust Emission Precursors," SAE paper 690020, 1969. T 10. Lavoie, G. A.: "Spectroscopic Measurement of Nitric Oxide in Spark-Ignition Engines," where T is in kelvin. Combust. Flame, vol. 15, pp. 97-108, 1970. 11. Blumberg, P ., and Kummer, J. K.: " Prediction of NO Formation in Spark-Ignition Engines-An Using the equilibrium composition data provided for mole fraction atomic Analysis of Methods of Control," Combust. Sci. Technol ., vol. 4, pp. 73-96, 1971. oxygen (O), molecular nitrogen (N2), and nitric oxide (NO): 12. Sakai, Y ., Miyazaki, H ., and Mukai, K.: "The Effect of Combustion Chamber Shape on Nitrogen (a) Plot the formation rate of NO as a function of the equivalence ratio at 3000 K Oxides," SAE paper 730154, 1973. and 5.5 MPa, and as a function of temperature for a stoichiometric mixture at 13. Quader, A. A.: " Why Intake Charge Dilution Decreases Nitric Oxide Emission from Spark Igni- 5.5 MPa. tion Engines," SAE paper 710009, SAE Trans ., vol. 80, 1971. 664 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 665 14. Benson, J. D ., and Stebar, R. F.: "Effects of Charge Dilution on Nitric Oxide Emission from Speed, Quench Distance and Global Reaction Rates," SAE paper 780229, SAE Trans ., vol. 87, Single-Cylinder Engine," SAE paper 710008, SAE Trans ., vol. 80, 1971. 1978. 15. Toda, T ., Nohira, H ., and Kobashi, K.: "Evaluation of Burned Gas Ratio (BGR) as a Predomi- 37. Adamczyk, A. A ., Kaiser, E. W ., Cavolowsky, J. A ., and Lavoie, G. A.: " An Experimental Study nant Factor to NO ,, " SAE paper 760765, SAE Trans ., vol. 85, 1976. of Hydrocarbon Emissions from Closed Vessel Explosions," in Proceedings of Eighteenth Interna- 16. Lavoie, G. A ., and Blumberg, P. N.: " A Fundamental Model for Predicting Fuel Consumption, tional Symposium on Combustion, pp. 1695-1702, The Combustion Institute, 1981. NO, and HC Emissions of the Conventional Spark-Ignited Engine," Combust. Sci. Technol ., vol. 38. Westbrook, C. K ., Adamczyk, A. A ., and Lavoie, G. A.: "A Number Study of Laminar Wall 21, pp. 225-258, 1980. Quenching," Combust. Flame, vol. 40, pp. 81-91, 1981. 17. Lavoie, G. A ., and Blumberg, P. N.: " Measurements of NO Emissions from a Stratified Charge 39. Kaiser, E. W ., Adamczyk, A. A ., and Lavoie, G. A.: "The Effect of Oil Layers on the Hydrocar- Engine: Comparison of Theory and Experiment," Combust. Sci. Technol ., vol. 8, p. 25, 1973. bon Emissions Generated During Closed Vessel Combustion," in Proceedings of Eighteenth Inter- 18. Aoyagi, Y ., Kamimoto, T ., Matsui, Y ., and Matsuoka, S.: " A Gas Sampling Study on the Forma- national Symposium on Combustion, pp. 1881-1890, The Combustion Institute, 1981. tion Processes of Soot and NO in a DI Diesel Engine," SAE paper 800254, SAE Trans ., vol. 89, 40. Tabaczynski, R. J ., Heywood, J. B ., and Keck, J. C.: "Time-Resolved Measurements of Hydrocar- 1980. bon Mass Flow Rate in the Exhaust of a Spark-Ignition Engine," SAE paper 720112, SAE Trans ., 19. Vioculescu, I. A ., and Borman, G. L.: "An Experimental Study of Diesel Engine Cylinder- vol. 81, 1972. Averaged NO, Histories," SAE paper 780228, SAE Trans ., vol. 87, 1978. 41. Ekchian, A ., Heywood, J. B ., and Rife, J. M.: "Time Resolved Measurements of the Exhaust from 20. Mansouri, S. H ., Heywood, J. B ., and Radhakrishnan, K.: " Divided-Chamber Diesel Engine, Part a Jet Ignition Prechamber Stratified Charge Engine," SAE paper 770043, SAE Trans ., vol. 86, I: Cycle-Simulation Which Predicts Performance and Emissions," SAE paper 820273, SAE 1977. Trans ., vol. 91, 1982. 42. Daniel, W. A ., and Wentworth, J. T.: "Exhaust Gas Hydrocarbons-Genesis and Exodus," SAE 21. Duggal, V. K ., Priede, T ., and Khan, I. M.: "A Study of Pollutant Formation within the Com- paper 486B, March 1962; also SAE Technical Progress Series, vol. 6, p. 192, 1964. bustion Space of a Diesel Engine," SAE paper 780227, SAE Trans ., vol. 87, 1978. 43. Daniel, W. A.: "Flame Quenching at the Walls of an Internal Combustion Engine," in Pro- 22. Liu, X ., and Kittelson, D. B.: "Total Cylinder Sampling from a Diesel Engine (Part II)," SAE ceedings of Sixth International Symposium on Combustion, p. 886, Reinhold, New York, 1957. paper 820360, 1982. 44. LoRusso, J. A ., Lavoie, G. A ., and Kaiser, E. W.: " An Electrohydraulic Gas Sampling Valve with 23. Yu, R. C ., and Shahed, S. M.: "Effects of Injection Timing and Exhaust Gas Recirculation on Application to Hydrocarbon Emissions Studies," SAE paper 800045, SAE Trans ., vol. 89, 1980. Emissions from a D.I. Diesel Engine," SAE paper 811234, SAE Trans ., vol. 90, 1981. 45. Wentworth, J. T.: "More on Origins of Exhaust Hydrocarbons-Effects of Zero Oil Consump- 24. Plee, S. L ., Myers, J. P ., and Ahmed, T.: "Flame Temperature Correlation for the Effects of tion, Deposit Location, and Surface Roughness," SAE paper 720939, SAE Trans ., vol. 81, 1972. Exhaust Gas Recirculation on Diesel Particulate and NO, Emissions," SAE paper 811195, SAE 46. Namazian, M ., and Heywood, J. B.: "Flow in the Piston-Cylinder-Ring Crevices of a Spark- Trans ., vol. 90, 1981. Ignition Engine: Effect on Hydrocarbon Emissions, Efficiency and Power," SAE paper 820088, 25. Plee, S. L ., Ahmad, T ., and Myers, J. P.: "Diesel NO, Emissions-A Simple Correlation Tech- SAE Trans ., vol. 91, 1982. nique for Intake Air Effects," in Proceedings of Nineteenth International Symposium on Com- 47. Furuhama, S ., and Tateishi, Y.: "Gases in Piston Top-Land Space of Gasoline Engine," Trans. bustion, pp. 1495-1502, The Combustion Institute, Pittsburgh, 1983. SAEJ, no. 4, 1972. 26. Ahmad, T ., and Plee, S. L.: " Application of Flame Temperature Correlations to Emissions from a 48. Wentworth, J. T.: "The Piston Crevice Volume Effect on Exhaust Hydrocarbon Emission," Direct-Injection Diesel Engine," SAE paper 831734, SAE Trans ., vol. 92, 1983. Combust. Sci. Technol ., vol. 4, pp. 97-100, 1971. 27. Harrington, J. A ., and Shishu, R. C.: " A Single-Cylinder Engine Study of the Effects of Fuel Type, 49. Haskell, W. W ., and Legate, C. E.: "Exhaust Hydrocarbon Emissions from Gasoline Engines- Fuel Stoichiometry, and Hydrogen-to-Carbon Ratio and CO, NO, and HC Exhaust Emissions," Surface Phenomena," SAE paper 720255, 1972. SAE paper 730476, 1973. 50. Adamczyk, A. A ., Kaiser, E. W ., and Lavoie, G. A.: "A Combustion Bomb Study of the Hydro- 28. Newhall, H. K.: "Kinetics of Engine-Generated Nitrogen Oxides and Carbon Monoxide," in carbon Emissions from Engine Crevices," Combust. Sci. Technol ., vol. 33, pp. 261-277, 1983. Proceedings of Twelfth International Symposium on Combustion, pp. 603-613, Mono of Maryland, 51. Wentworth, J. T.: "Piston and Ring Variables Affect Exhaust Hydrocarbon Emissions," SAE 1968. paper 680109, SAE Trans ., vol. 77, 1968. 29. Keck, J. C ., and Gillespie, D.: “Rate-Controlled Partial-Equilibrium Method for Treating React- 52. Kaiser, E. W ., LoRusso, J. A ., Lavoie, G. A ., and Adamczyk, A. A.: "The Effect of Oil Layers on ing Gas Mixtures," Combust. Flame, vol. 17, pp. 237-241, 1971. the Hydrocarbon Emissions from Spark-Ignited Engines," Combust. Sci. Technol ., vol. 28, pp. 30. Delichatsios, M. M.: "The Kinetics of CO Emissions from an Internal Combustion Engine," S.M. 69-73, 1982. Thesis, Department of Mechanical Engineering, MIT, June 1972. 53. Kaiser, E. W ., Adamczyk, A. A ., and Lavoie, G. A.: "The Effect of Oil Layers on the Hydrocar- 31. Johnson, G. L ., Caretto, L. S ., and Starkman, E. S.: "The Kinetics of CO Oxidation in Recipro- bon Emissions Generated during Closed Vessel Combustion," in Proceedings of Eighteenth Inter- cating Engines," paper presented at the Western States Section, The Combustion Institute, Spring national Symposium on Combustion, pp. 1881-1890, The Combustion Institute, 1981. Meeting, April 1970. 54. McGeehan, J. A.: " A Literature Review of the Effects of Piston and Ring Friction and Lubricat- 32. Jackson, M. W.: "Analysis for Exhaust Gas Hydrocarbons-Nondispersive Infrared Versus ing Oil Viscosity on Fuel Economy," SAE paper 780673, SAE Trans ., vol. 87, 1978. Flame-Ionization," J. Air Pollution Control Ass ., vol. 16, p. 697-702, 1966. 55. Shin, K ., Tateishi, Y ., and Furuhama, S.: "Measurement of Oil-Film-Thickness between Piston 33. Jackson, M. W.: "Effect of Catalytic Emission Control on Exhaust Hydrocarbon Composition Ring and Cylinder," SAE paper 830068, SAE Trans ., vol. 92, 1983. and Reactivity," SAE paper 780624, SAE Trans ., vol. 87, 1978. 56. Carrier, G. F ., Fendell, F. E ., and Feldman, P. S.: "Cyclic Absorption/Desorption of Gas in a 34. Patterson, D. J ., and Henein, N. A.: Emissions from Combustion Engines and Their Control, Ann Liquid Wall Film," Combust. Sci. Technol ., vol. 25, pp. 9-19, 1981. Arbor Science Publishers, Ann Arbor, Michigan, 1972. 57. Wentworth, J. T.: "Effects of Top Compression Ring Profile on Oil Consumption and Blowby 35. "Diesel Technology, Impacts of Diesel-Powered Light-Duty Vehicles," report of the Technology with the Sealed Ring-Orifice Design," SAE paper 820089, 1982. Panel of the Diesel Impacts Study Committee, National Research Council, National Academy 58. Kuroda, H ., Nakajima, Y ., Sugihara, K ., Takagi, Y ., and Muranaka, S.: "The Fast Burn with Press, Washington, D.C ., 1982. Heavy EGR, New Approach for Low NO, and Improved Fuel Economy," SAE paper 780006, 36. Lavoie, G. A.: "Correlations of Combustion Data for S.I. Engine Calculations -- Laminar Flame SAE Trans ., vol. 87, 1978. 666 INTERNAL COMBUSTION ENGINE FUNDAMENTALS POLLUTANT FORMATION AND CONTROL 667 59. Jackson, M. W ., Wiese, W. M ., and Wentworth, J. T.: "The Influence of Air-Fuel Ratio, Spark Thrower (eds.), Chemistry and Physics of Carbon, vol. 14, pp. 168-294, Marcel Dekker, New York, Timing, and Combustion Chamber Deposits on Exhaust Hydrocarbon Emissions," SAE paper 1978. 486A, in Vehicle Emissions, vol. TP-6, SAE, 1962. 82. Graham, S. C ., Homer, J. B ., and Rosenfeld, J. L. J.: "The Formation and Coagulation of Soot 60. Lavoie, G. A ., Lorusso, J. A ., and Adamczyk, A. A.: "Hydrocarbon Emissions Modeling for Aerosols Generated by the Pyrolyses of Aromatic Hydrocarbons," Proc. R. Soc. Lond ., vol. A344, Spark Ignition Engines," in J. N. Mattavi and C. A. Amann (eds.), Combustion Modeling in pp. 259-285, 1975. Reciprocating Engines, pp. 409-445, Plenum Press, 1978. 83. Amann, C. A ., Stivender, D. L ., Plee, S. L ., and MacDonald, J. S.: "Some Rudiments of Diesel 61. Daniel, W. A.: "Why Engine Variables Affect Exhaust Hydrocarbon Emission," SAE paper Particulate Emissions," SAE paper 800251, SAE Trans ., vol. 89, 1980. 700108, SAE Trans ., vol. 79, 1970. 84. Park, C ., and Appleton, J. P.: "Shock-Tube Measurements of Soot Oxidation Rates," Combust. 62. Amann, C. A.: "Control of the Homogeneous-Charge Passenger-Car Engine-Defining the Flame, vol. 20, pp. 369-379, 1973. Problem," SAE paper 801440, 1980. 85. Otto, K ., Sieg, M. H ., Zinbo, M ., and Bartosiewicz, L.: "The Oxidation of Soot Deposits from 63. Weiss, P ., and Keck, J. C.: “Fast Sampling Valve Measurements of Hydrocarbons in the Cylinder Diesel Engines," SAE paper 800336, SAE Trans ., vol. 89, 1980. of a CFR Engine," SAE paper 810149, SAE Trans ., vol. 90, 1981. 86. Abthoff, J ., Schuster, H ., Langer, H ., and Loose, G.: “The Regenerable Trap Oxidizer-An Emis- 64. Caton, J. A ., Heywood, J. B ., and Mendillo, J. V.: "Hydrocarbon Oxidation in a Spark Ignition sion Control Technique for Diesel Engines," SAE paper 850015, 1985. Engine Exhaust Port," Combust. Sci. Technol ., vol. 37, nos. 3 and 4, pp. 153-169, 1984. 65. Green, R. M ., Smith, J. R ., and Medina, S. C.: "Optical Measurement of Hydrocarbons Emitted from a Simulated Crevice Volume in an Engine," SAE paper 840378, SAE Trans ., vol. 93, 1984. 66. Nakagawa, Y ., Etoh, Y ., and Maruyama, R.: " A Fundamental Analysis of HC and CO Oxidation Reaction in the Exhaust System," JSAE Rev ., no. 1, pp. 98-106, 1978. 67. Greeves, G ., Khan, I. M ., Wang, C. H. T ., and Fenne, I.: "Origins of Hydrocarbon Emissions from Diesel Engines," SAE paper 770259, SAE Trans ., vol. 86, 1977. 68. Kummer, J. T.: "Catalysts for Automobile Emission Control," Prog. Energy Combust. Sci ., vol. 6, pp. 177-199, 1981. 69. Cadle, S. H ., Nebel, G. J ., and Williams, R. L.: "Measurements of Unregulated Emissions from General Motors' Light-Duty Vehicles," SAE paper 790694, SAE Trans ., vol. 88, 1979. 70. Amann, C. A ., and Siegla, D. C.: "Diesel Particulates-~What They Are and Why," Aerosol Sci. Technol ., vol. 1, pp. 73-101, 1982. 71. Mayer, W. J ., Lechman, D. C ., and Hilden, D. L.: "The Contribution of Engine Oil to Diesel Exhaust Particulate Emissions," SAE paper 800256, SAE Trans ., vol. 89, 1980. 72. Lahaye, J ., and Prado, G.: "Morphology and Internal Structure of Soot and Carbon Blacks," in D. C. Siegla and G. W. Smith (eds.), Particulate Carbon Formation during Combustion, pp. 33-55, Plenum Press, 1981. 73. Smith, O. I.: “Fundamentals of Soot Formation in Flames with Application to Diesel Engine Particulate Emissions," Prog. Energy Combust. Sci ., vol. 7, pp. 275-291, 1981. 74. Whitehouse, N. D ., Clough, E ., and Uhunmwangho, S. O.: "The Development of Some Gaseous Products during Diesel Engine Combustion," SAE paper 800028, 1980. 75. Greeves, G ., and Meehan, J. O.: “Measurement of Instantaneous Soot Concentration in a Diesel Combustion Chamber," paper C88/75, in Combustion in Engines, pp. 73-82, Institution of Mechanical Engineers, 1975. 76. Chang, Y. J ., Kobayashi, H ., Matsuzawa, K ., and Kamimoto, T.: " A Photographic Study of Soot Formation and Combustion in a Diesel Flame with a Rapid Compression Machine," Proceedings of International Symposium on Diagnostics and Modeling of Combustion in Reciprocating Engines, COMODIA 85, pp. 149-157, Tokyo, Japan, September 4-6, 1985. 77. Du, C. J ., and Kittelson, D. B.: "Total Cylinder Sampling from a Diesel Engine: Part III- Particle Measurements," SAE paper 830243, SAE Trans ., vol. 92, 1983. 78. Aoyagi, Y ., Kamimoto, T ., Matsui, Y ., and Matsuoka, S.: " A Gas Sampling Study on the Forma- tion Processes of Soot and NO in a DI Diesel Engine," SAE paper 800254, SAE Trans ., vol. 89, 1980. 79. Matsui, Y ., Kamimoto, T ., and Matsuoka, S.: "Formation and Oxidation Processes of Soot Particles in a D.I. Diesel Engine -- An Experimental Study Via the Two-Color Method," SAE paper 820464, SAE Trans ., vol. 91, 1982. 80. Haynes, B. S ., and Wagner, H. G.: "Soot Formation," Prog. Energy Combust. Sci ., vol. 7, pp. 229-273, 1981. 81. Lahaye, J ., and Prado, G.: "Mechanisms of Carbon Black Formation," in P. L. Walker and P. A. ENGINE HEAT TRANSFER 669 CHAPTER and reduce the work per cycle transferred to the piston. Thus specific power and efficiency are affected by the magnitude of engine heat transfer. Heat transfer 12 between the unburned charge and the chamber walls in spark-ignition engines affects the onset of knock which, by limiting the compression ratio, also influ- ences power and efficiency. Most critical is heat transfer from the hot exhaust valve and piston to mixture in the end-gas region. Changes in gas temperature due to the heat-transfer impact on emission formation processes, both within the ENGINE engine's cylinder and in the exhaust system where afterburning of CO and HC occurs. The exhaust temperature also governs the power that can be obtained HEAT from exhaust energy recovery devices such as a turbocharger turbine. Friction is TRANSFER both affected by engine heat transfer and contributes to the coolant load. The cylinder liner temperature governs the piston and ring lubricating oil film tem- perature, and hence its viscosity. Piston and liner distortion due to temperature nonuniformities have a significant impact on the piston component of engine friction. Some of the mechanical energy dissipated due to friction must be rejected to the atmosphere by the cooling system. The fan and water pump power requirements are determined by the magnitude of the heat rejected. The impor- tance of engine heat transfer is clear. To examine heat transfer more fully, it is helpful to divide the engine into its subsystems. The intake system consists of intake manifold and inlet ports and valves. Heat transfer to the inflowing charge reduces volumetric efficiency (see Sec. 6.2.1). However, in spark-ignition engines, the intake mixture is heated, with carbureted and single-point injected engines, to aid in vaporizing the fuel (see Sec. 12.1 IMPORTANCE OF HEAT TRANSFER 7.6.3). Within the engine cylinder, the temperature of the charge relative to the wall temperature and the flow field vary enormously throughout the cycle. Both The peak burned gas temperature in the cylinder of an internal combustion engine is of order 2500 K. Maximum metal temperatures for the inside of the of these variables have a major influence on heat transfer. During the intake process, the incoming charge is usually cooler than the walls and the flow veloc- combustion chamber space are limited to much lower values by a number of considerations, and cooling for the cylinder head, cylinder, and piston must ities are high. During compression the charge temperature rises above the wall therefore be provided. These conditions lead to heat fluxes to the chamber walls temperature, and gas velocities decrease (see Sec. 8.2.2). Heat transfer is now from that can reach as high as 10 MW/m2 during the combustion period. However, the cylinder gases to the chamber walls. During combustion gas temperatures increase substantially and the gas expansion which occurs on combustion pro- during other parts of the operating cycle, the heat flux is essentially zero. The flux varies substantially with location: regions of the chamber that are contacted by duces increased gas motion. This is the period when heat-transfer rates to the walls are highest. Also, as the cylinder pressure rises, a small fraction of the cylin- rapidly moving high-temperature burned gases generally experience the highest fluxes. In regions of high heat flux, thermal stresses must be kept below levels der charge is forced into crevice regions, resulting in additional heat transfer (see Sec. 8.6). During expansion, gas temperatures decrease so heat-transfer rates that would cause fatigue cracking (so temperatures must be less than about 400ºC for cast iron and 300ºC for aluminum alloys). The gas-side surface of the decrease. When the exhaust valve opens, however, the blowdown process (Sec. 6.5) produces high velocities within the cylinder, and past the exhaust valve and cylinder wall must be kept below about 180ºC to prevent deterioration of the lubricating oil film. Spark plug and valves must be kept cool to avoid knock and in the exhaust port. Substantial heat transfer from the exhausting gases to the preignition problems which result from overheated spark plug electrodes or valve, port, and (to a lesser extent) manifold occurs during the exhaust process. An example of how the heat-transfer rate to the total combustion chamber walls exhaust valves. Solving these engine heat-transfer problems is obviously a major varies throughout the four-stroke operating cycle of a spark-ignition engine is design task. shown in Fig. 14-9. The heat-transfer rate was estimated from the cylinder pres- Heat transfer affects engine performance, efficiency, and emissions. For a sure, unburned and burned gas temperatures, combustion chamber surface area, given mass of fuel within the cylinder, higher heat transfer to the combustion and wall temperature, assuming gas velocities scaled with mean piston speed. An chamber walls will lower the average combustion gas temperature and pressure, ability to predict the magnitude of the heat transfer between the working fluid, 668 670 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE HEAT TRANSFER 671 the walls of the intake system, combustion chamber, and exhaust system, and to. where he is called the heat-transfer coefficient. For many flow geometries (such as the coolant is of obvious importance to the engine designer. flow through pipes or over a plate), he is given by relations of the form (PUL )" ( COM) (12.3) 12.2 MODES OF HEAT TRANSFER where L and v are a characteristic length and velocity. The terms in brackets The following modes of heat transfer are important. from left to right are the Nusselt, Reynolds, and Prandtl dimensionless numbers, respectively. For gases, the Prandtl number (c, u/k) varies little and is about 0.7 (see Sec. 4.8). 12.2.1 Conduction When boiling occurs at the surface (i.e ., vapor is formed in the liquid), as Heat is transferred by molecular motion, through solids and through fluids at may be the case in high heat flux areas on the coolant side in water-cooled rest, due to a temperature gradient. The heat transfer by conduction, per unit engines, then different relationships for he must be used. area per unit time, q, in a steady situation is given by Fourier's law: ¿ = - kVT 12.2.3 Radiation (12.1) where k is the thermal conductivity. For a steady one-dimensional temperature Heat exchange by radiation occurs through the emission and absorption of elec- tromagnetic waves. The wavelengths at which energy is transformed into thermal variation energy are the visible range (0.4 to 0.7 um) and the infrared (0.7 to 40 um). Heat 110 dT transfer by radiation occurs from the high-temperature combustion gases and the A = - k dx flame region to the combustion chamber walls (although the magnitude of this radiation heat transfer relative to convective heat transfer is only significant in Heat is transferred by conduction through the cylinder head, cylinder walls, and diesel engines). Heat transfer by radiation to the environment occurs from all the piston; through the piston rings to the cylinder wall; through the engine block hot external surfaces of the engine. and manifolds. The theory of radiant heat transfer starts from the concept of a "black body," i.e ., a body that has a surface that emits or absorbs equally well radiation of all wavelengths and that reflects none of the radiation falling on it. The heat 12.2.2 Convection flux from one plane black body at temperature Ti to another at temperature T2 Heat is transferred through fluids in motion and between a fluid and solid surface parallel to it across a space containing no absorbing material is given by in relative motion. When the motion is produced by forces other than gravity, the à = (T1 - T4) (12.4) term forced convection is used. In engines the fluid motions are turbulent (see Chap. 8). where o is the Stefan-Boltzmann constant 5.67 x 10-8 W/m2 . K4. Real surfaces Heat is transferred by forced convection between the in-cylinder gases and are not "black" but reflect radiation to an extent which depends on wavelength. the cylinder head, valves, cylinder walls, and piston during induction, compres- Gases are far from this black-body idealization. They absorb and emit radiation sion, expansion, and exhaust processes. Heat is transferred by forced convection almost exclusively within certain wavelength bands characteristic of each species. from the cylinder walls and head to the coolant (which may be liquid or gas), and These departures from black-body behavior are usually dealt with by applying a from the piston to the lubricant or other piston coolant. Substantial convective multiplying factor (an emissivity, &) to Eq. (12.4). Similarly, a "shape factor" is heat transfer occurs to the exhaust valve, exhaust port, and exhaust manifold applied to account for the fact that the angle of incidence of the radiation usually during the exhaust process. Heat transfer by convection in the inlet system is varies over any actual surface. These factors can be calculated for simple cases. used to raise the temperature of the incoming charge. Heat is also transferred from the engine to the environment by convection. 12.2.4 Overall Heat-Transfer Process In steady-flow forced-convection heat-transfer problems, the heat flux à transferred to a solid surface at temperature Tw from a flowing fluid stream at Figure 12-1 shows, schematically, the overall heat-transfer process from the gases within the cylinder through the combustion chamber wall to the coolant flow. temperature T is determined from the relation The heat flux into the wall has in general both a convective and a radiation à = he(T - Tw) (12.2) component. The heat flux is conducted through the wall and then convecteurom 672 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE HEAT TRANSFER 673 T 1.6 T 9cv + 4R 9CN qCv 1.4 (a) 400-600 cm3/cyl. Gas Coolant 1.2 1w. 8 Tw. C Tc brake power Coolant heat flow rate FIGURE 12-1 (b) 800-1600 cm3/cyl. 1.0 Schematic of temperature distribution and heat flow Distance, x across the combustion chamber wall. -(c) 2-5 dm3/cyl. 0.8 FIGURE 12-2 Ratio of coolant heat flow rate to brake power the wall to the coolant. A schematic temperature profile, and mean gas and as a function of engine speed. Different size and 0.6 coolant temperatures, T, and Te , are shown. types of engines: (a) small automotive diesels; (b) larger automotive diesels; (c) various diesels; In internal combustion engines, throughout each engine operating cycle, the 1000 2000 3000 4000 (d) spark-ignition engines. (Developed from heat transfer takes place under conditions of varying gas pressure and tem- Engine speed, rev/min Howarth.2) perature, and with local velocities which vary more or less rapidly depending on intake port and combustion chamber configuration (see Chap. 8). In addition, the surface area of the combustion chamber varies through the cycle. The heat flux 12.3 HEAT TRANSFER AND ENGINE into the containing walls changes continuously from a small negative value ENERGY BALANCE during the intake process to a positive value of order several megawatts per square meter early in the expansion process. The flux variation lags behind the Figure 12-2 shows the magnitude of the heat-rejection rate to the coolant relative change in gas temperature. This lag between heat flux and driving temperature to the brake power for a range of engine types and sizes at maximum power. This difference is clearly perceptible1 but the precision of measurements to date suffice ratio decreases with increasing engine speed and with increasing engine size. The only for a rough estimate of its magnitude. Generally, investigators have con- smaller diesel engine designs use higher gas velocities to achieve the desired fuel- cluded that the assumption that the heat-transfer process is quasi steady is suffi- air mixing rates and have less favorable surface/volume ratios (see Sec. 10.2). ciently accurate for most calculation purposes. However, gas temperature and gas An overall first law energy balance for an engine provides useful informa- velocities vary significantly across the combustion chamber. The heat flux dis- tion on the disposition of the initial fuel energy. For a control volume which tribution over the combustion chamber walls is, therefore, nonuniform. surrounds the engine (see Fig. 3-8), the steady-flow energy-conservation equation is For a steady one-dimensional heat flow through a wall as indicated in Fig. 12-1, the following equations relate the heat flux q = Q/A and the temperatures img hy + maha = P, + Qcool + Omise + (mg + ma)he (12.8) indicated : where P, is the brake power, Ocool is the heat-transfer rate to the cooling medium, Gas side: q = @cv + @R = hcg (T, - Tw.g) + GET# - Tag) (12.5) Omise is the heat rejected to the oil (if separately cooled) plus convection and radiation from the engine's external surface. It proves convenient to divide the where e is the emissivity. The radiation term is generally negligible for SI engines. exhaust enthalpy he into a sensible part he ,, = he(T) - h (298 K), plus the exhaust reference state enthalpy (see Sec. 4.5). Then Eq. (12.8) can be written: Wall: q = qCN = K(Tw. - Tw.d (12.6) tw Ps + Qcool + Omisc + Hele + mhes = ing QLHV (12.9) where He.te represents the exhaust enthalpy loss due to incomplete combustion. Coolant side: q = @cv = hec (Tw,c - Tc) (12.7) Typical values of each of these terms relative to the fuel flow x heating value are If he, and he, are known, the temperatures T ., Tw,q, Tw,c , and Te can be related given in Table 12.1. to each other. The energy balance within an engine is more complicated and is illustrated 674 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE HEAT TRANSFER 675 TABLE 12.1 100 Radiation, incomplete combustion Energy balance for automotive engines at maximum power 80 Exhaust enthalpy Ocool (percentage of fuel heating value) 60 SI engine 25-28 17-26 3-10 2-5 34-45 Diesel 34-38 16-35 2-6 1-2 22-35 Percent fuel heating value Coolant load Sources: From Khovakh, 3 Sitkei, " and Burke et al.5 40 in the energy flow diagram in Fig. 12-3. The indicated power is the sum of the brake power and the friction power. A substantial part of the friction power 20 Brake power FIGURE 12-4 (about half) is dissipated between the piston and piston rings and cylinder wall 100 1500 1800 1500 Brake power, coolant load, sensible exhaust and is transferred as thermal energy to the cooling medium. The remainder of the 2200 rev/min 128 160 202 245 bmep, kPa enthalpy, and miscellaneous energy transfers as friction power is dissipated in the bearings, valve mechanism, or drives auxiliary percent of fuel flow x heating value for spark- 30 40 devices, and is transferred as thermal energy to the oil or surrounding 50 60 70 80 90 ignition engine at road-load operating condi- tions. 6 environment (in _misc). The enthalpy initially in the exhaust gases can be sub- Vehicle speed, km/h divided into the following components: a sensible enthalpy (60 percent), an exhaust kinetic energy (7 percent), an incomplete combustion term (20 percent), and a heat transfer to the exhaust system (12 percent) (part of which is radiated to the environment and the remainder ends up in the cooling medium).+ Thus the heat carried away by the coolant medium consists of heat transferred to the combustion chamber walls from the gases in the cylinder, heat transferred to the ew Lcool exhaust valve and port in the exhaust process, and a substantial fraction of the friction work. At part-load, a greater fraction of the fuel heating value is absorbed into the .2c.e coolant. Figure 12-4 shows data for a six-cylinder SI engine operated at road- load over a range of vehicle speeds. At low speeds and loads, the coolant heat- He,s.a transfer rate is 2 to 3 times the brake power. Although the heat losses are such a substantial part of the fuel energy input, He,ic elimination of heat losses would only allow a fraction of the heat transferred to the combustion chamber walls to be converted to useful work. The remainder myQLHV Pif Qe.r would leave the engine as sensible exhaust enthalpy. Consider this example for an Ee, k automotive high-speed naturally aspirated CI engine with a compression ratio of 15. The indicated efficiency is 45 percent, and 25 percent of the fuel energy is carried away by the cooling water. Of this 25 percent, about 2 percent is due to Qmis friction. Of the remaining 23 percent, about 8 percent is heat loss during com- bustion, 6 percent heat loss during expansion, and 9 percent heat loss during exhaust. Of the 8 percent lost during combustion about half (or 4 percent of the FIGURE 12-3 Energy flow diagram for IC engine. (m, QLHy) = fuel flow rate x lower heating value, (w = fuel energy) could be converted into useful work on the piston (see Fig. 5-9). Of heat-transfer rate to combustion chamber wall, H. = exhaust gas enthalpy flux, P, = brake power, the 6 percent heat loss during expansion, about one-third (or 2 percent) could Py = total friction power, P, = indicated power, Por = piston friction power, (.) = heat-rejection rate to coolant, Que = heat-transfer rate to coolant in exhaust ports, H ... = exhaust sensible enthalpy flux entering atmosphere, Hoje = exhaust chemical enthalpy flux due to incomplete com- bustion, Qe, = heat flux radiated from exhaust system, Et = exhaust kinetic energy flux, Omisc = sum of remaining energy fluxes and transfers. + The percentages are approximate. 576 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE HEAT TRANSFER 677 have been utilized. Thus, of the 25 percent lost to the cooling system, only about 6 percent could have been converted to useful work on the piston, which would The first three groups are the familiar Nusselt, Reynolds, and Prandtl numbers, increase the indicated efficiency of the engine from 45 to 51 percent. respectively. The next has the nature of a Mach number since c, T is proportion- For a spark-ignition engine, the conversion to useful work will be lower. al to the square of the sound speed. For Mach numbers much less than 1, the because the compression ratio is lower. However, as shown in Fig. 12-4, the heat Mach number dependence is known to be small and can be omitted. It is usual to losses at part-load (an important operating regime for automobile use) are a take for v the mean piston speed S, = 2LN. Then, by introducing the bore/stroke substantially larger fraction of the fuel heating value. Studies with computer ratio B/L, the term NB/v is eliminated. z/B is a function of the compression ratio simulations of the SI engine operating cycle indicate that at typical part-load Te, the ratio of connecting rod to crank radius R = 1/a, and 0. Thus conditions a proportional reduction in combustion-chamber-wall heat losses of 10 percent results in a proportional increase (improvement) in brake fuel conver- K F ( h. B PS B Cpu B Ich sion efficiency of about 3 percent.7 M ' k ' L' pc, NT, Je, R, Vi. ..., Vm, 0, 41, ..., un) = 0 (12.10) The dimensionless groups may be varied (but not reduced in number) by com- 12.4 CONVECTIVE HEAT TRANSFER bination. While Eq. (12.10) reveals nothing about the functional form of the 12.4.1 Dimensional Analysis relationship between the groups, it does provide a basis for evaluating the corre- lations which have been proposed. While the overall time-averaged heat transfer to the coolant medium is adequate Many formulas for calculating instantaneous engine heat-transfer coeffi- for some design purposes, the instantaneous heat flux during the engine cycle is a cients have been proposed (see Ref. 9 for a review). Only those with a functional necessary input for realistic cycle calculations (see Sec. 14.4) and provides the form which fits Eq. (12.10) will be summarized here. The basis of these correla- fundamental input for obtaining the heat flux distribution to various parts of an tions is the assumption that the Nusselt, Reynolds, and Prandtl number relation- operating engine. Equations (12.5), (12.6), and (12.7) provide the framework for ship follows that found for turbulent flow in pipes or over flat plates: calculating the heat flux à, based on the assumption that at each point in the cycle the heat-transfer process is quasi steady. For example, neglecting radiation, Nu = a Rem Pr" (12.11) if T ,, Tw ,, and he, can be calculated at each point in the cycle, q(0) is obtained. Alternatively, if Te, he,c , Tg, and he ., are known, q, Tw.(, and Tw, can be com- Distinctions should be made between correlations intended to predict the puted. time-averaged heat flux to the combustion chamber walls, the instantaneous spa- Dimensional analysis can be used to develop the functional form of tially averaged heat flux to the chamber walls (which is required for engine per- relationships which govern the gas-side heat-transfer coefficient. The engine con- formance analysis), and the instantaneous local heat fluxes (which are not uniform vective heat-transfer process can be characterized geometrically by a length over the combustion chamber and may be required for thermal stress dimension-say the bore B-and a number of length ratios y1, 12, 13, etc. (of calculations). In using these heat-transfer correlations, the critical choices to be which one will be the axial cylinder length z divided by the bore z/B), which made are (1) the velocity to be used in the Reynolds number; (2) the gas tem- define the cylinder and combustion chamber geometry. The flow pattern, simi- perature at which the gas properties in Eq. (12.11) are evaluated; and (3) the gas larly, may be characterized by one chosen velocity u and a set of velocity ratios temperature used in the convective heat-transfer equation (12.2). 11, 12, 13, etc. The gas properties of importance are the thermal conductivity k, The most widely used correlations and the basis of their derivation will now the dynamic viscosity u, the specific heat c ,, and the density p. If there is com- be summarized. Because the experimental data for evaluating these correlations bustion, then the chemical energy release rate per unit volume deh may be impor- in CI engines includes both convective and radiative heat fluxes, comparison of tant. The engine speed N and relative position in the cycle denoted by crank these correlations with data is deferred to Sec. 12.6. angle 0 introduce the cyclical nature of the process. Thus f(hc, B, 2, 71, 72. ..., Vm, 0, 41, Uz, ..., Un, K, U, Cp, P, ich, N, 0) = 0 Applying dimensional analysis, with mass, length, time, and temperature as the 12.4.2 Correlations for Time-Averaged Heat Flux independent dimensions, reduces the variables to four-fewer nondimensional Taylor and Toong1º have correlated overall heat-transfer data from 19 different groups: engines. It was assumed that coolant and wall temperatures varied little between F L. B PUB COM CT NB 9ch designs and that the effects of geometrical differences were small. Thus, at a given k ' H ' k ' U2 ' U ' PC N T ' B ' X 10 . . . , Y m x 14 1 ; . . . , Un, 0 ) = 0 fuel/air ratio, the convective part of the heat flux should correlate with Reynolds number. To allow for variations in fuel/air ratio, Taylor and Toong defined an 678 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE HEAT TRANSFER 679 average effective gas temperature To ., such that 103 8 8 00 SI engines, water-cooled (10) Ah (T - Todde = 0 o CI engines, water-cooled (8) CI engine, air-cooled (1) over the engine cycle. Te, is the temperature at which the wall would stabilize if no heat was removed from outside. T .. was obtained by extrapolating average heat-transfer data plotted versus gas-side combustion chamber surface tem- perature back to the zero heat-transfer axis. The Nusselt number, defined as 10 OB Ko * Bk (T3, a - T) 3.5 Nu = ke, 10-8 W/m . K (ZB2/4XT ., a - T)k nBK (To - TC) (12.12) W 7 2.5 6 plotted against Reynolds number, defined as Nu = 48, 10-> N . s/m2 1g, @ 700 2 4m 600 MB Re = - (12.13) SI engines 500 , K Ho(ZB2/4) Diesel 400 Te 300 where m is the charge mass flow rate, is shown in Fig. 12-5. Taylor and Toong 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 proposed a power law of 0.75. Annand8 suggests three separate lines for the three 103 Equivalence ratio ¢ different types of engines covered, with slope 0.7. The diesel line is about 25 percent higher than the spark-ignition engine line (which corresponds in part to the radiative heat flux component present in diesels). The air-cooled engine line is 103 2 6 8 104 2 105 2 lower than the liquid-cooled line, presumably because surface temperatures are Re = 4 m B higher. The average gas temperature values developed by Taylor and Toong are FIGURE 12-5 shown in the insert in Fig. 12-5. Overall engine heat-transfer correlation: gas-side Nusselt number versus Reynolds number for differ- ent types of IC engines. See text for definition of symbols. Insert gives effective gas temperature (wall 12.4.3 Correlations for Instantaneous Spatial temperature for adiabatic operation), gas viscosity u, and thermal conductivity k. Lines have slope 0.7,8,10 Average Coefficients Annand8 developed the following convective heat-transfer correlation to match effect of chemical energy release is omitted. While only data from cylinder head previously published experimental data on instantaneous heat fluxes to selected thermocouple locations were used as a basis for this correlation, it has often been cylinder head locations: used to estimate instantaneous spatial average heat fluxes for the entire com- n . B ) = a ( PS B )6 bustion chamber. (12.14) Woschni11 assumed a correlation of the form Nu = 0.035 Rem The value of a varied with intensity of charge motion and engine design. With (12.16) normal combustion, 0.35 < a < 0.8 with b == 0.7, and a increases with increasing With the cylinder bore B taken as the characteristic length, with w as a local intensity of charge motion. Gas properties are evaluated at the cylinder-average average gas velocity in the cylinder, and assuming k oc 70.75, u oc 70.62, and charge temperature ₸,: p = pRT, the above correlation can be written T, PVM (12.15) h = CBm-1pm wmTo.75-1.62m (12.17) mk During intake, compression, and exhaust, Woschni argued that the average The same temperature is used in Eq. (12.2) to obtain the convective heat flux. gas velocity should be proportional to the mean piston speed. During com- Note that in developing this correlation, the effects of differences in bustion and expansion, he attempted to account directly for the gas velocities geometry and flow pattern between engines [the ratios y1, ..., ym and u1, ..., ". induced by the change in density that results from combustion (~10 m/s), which in Eq. (12.10)] have been incorporated in the proportionality constant a, and the are comparable to mean piston speeds. Thus a term proportional to the pressure 680 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE HEAT TRANSFER 681 12.4.4 Correlations for Instantaneous Local rise due to combustion (p - pm) was added (p„ is the motored cylinder pressure). Coefficients The coefficients relating the local average gas velocity w to the mean piston speed and (p - pm) were determined by fitting the correlation, integrated over the LeFeuvre et al.15 and Dent and Sulaiman16 have proposed the use of the flat- engine cycle, to time-averaged measurements of heat transfer to the coolant for a plate forced convection heat-transfer correlation formula wide range of engine operating conditions for a direct-injection four-valve diesel without swirl. T in Eq. (12.17) is the mean cylinder gas temperature defined by = 0.036( PUL)0.8( UC2 ) 0.333 (12.20) Eq. (12.15); the same temperature is used to obtain the heat flux from the heat- K transfer coefficient hc. Thus this correlation represents spatially averaged com- where I is the length of the plate and u the flow velocity over the plate. This bustion chamber heat fluxes. formula has been applied to DI diesel engines with swirl, with / and v evaluated The average cylinder gas velocity w (meters per second) determined for a at a radius r as four-stroke, water-cooled, four-valve direct-injection CI engine without swirl was expressed as follows: 1 = 2tr v = rw w = C15 , + C 2 VAI ( p - Pm )| @ being the solid-body angular velocity of the charge. The heat flux at any radius (12.18) r (with Pr = 0.73) is then given by 0.8 where Va is the displaced volume, p is the instantaneous cylinder pressure, p ,, V ,, q(r) = 0.023 [Tg(r) - Tw(r) ] (12.21) T, are the working-fluid pressure, volume, and temperature at some reference state (say inlet valve closing or start of combustion), and pm is the motored This equation can be used if the swirl variation with crank angle is known (see cylinder pressure at the same crank angle as p. Sec. 8.3) and an appropriate local gas temperature can be determined. It would For the gas exchange period: C1 = 6.18, C2 = 0 not be consistent to use the cylinder average gas temperature given by Eq. (12.15) For the compression period: C1 = 2.28, C2 = 0 because during combustion substantial temperature nonuniformities exist For the combustion and expansion period: C1 = 2.28, C2 = 3.24 × 10-3 between burned gases and air or mixture which has yet to burn or mix with already burned gas. Subsequent studies in higher-speed engines with swirl indicated higher heat transfer than these velocities predicted. For engines with swirl, cylinder averaged An alternative approach is zonal modeling, where the combustion chamber is divided into a relatively small number of zones each with its own temperature, gas velocities were given by Eq. (12.18) with: heat-transfer coefficient, and heat-transfer surface area history. This approach has For the gas exchange period: C1 = 6.18 + 0.417 been applied to spark-ignition engines (e.g ., Ref. 17), where the division of the Sp in-cylinder gases during combustion into a higher-temperature burned gas region For the rest of cycle: Us behind the flame and lower-temperature unburned gas region ahead of the flame C = 2.28 + 0.308 S. is clear (see Fig. 9-4). The heat transfer to the combustion chamber surfaces in p where v, = Bw,/2 and @, is the rotation speed of the paddle wheel used to contact with the unburned and burned gas zones [analagous to Eq. (12.2)] is given by measure the swirl velocity (see Sec. 8.3.1).12 Spark-ignition engine tests showed that the above velocities gave acceptable predictions for this type of engine Ou = Au,w he, w(Th - T.) Q. = AB, who, 6 (Tb - Tw) (12.21a,b) also, 13 Woschni's correlation, with the exponent in Eq. (12.17) equal to 0.8, can be respectively. Since he depends on local gas properties and velocities, he, y and he, b are not necessarily the same. Examples of how the burned gas wetted areas on summarized as: the piston, cylinder head, and liner vary during the combustion process are given h (W/m2. K) = 3.26B(m) -0-2p(kPa)0.8T(K) -0.55w(m/s)0.8 (12.19) in Fig. 14-8. Since the burned gas temperature T; is much larger than the unburned gas temperature, the heat flux from the burned gas zone dominates. with w defined above. Hohenberg14 examined Woschni's formula and made changes to give better One useful development of this two-zone approach is the division of the predictions of time-averaged heat fluxes measured with probes in a direct- burned gas zone into an adiabatic core and a thermal boundary layer. The injection diesel engine with swirl. The modifications include use of a length based advantages are: (1) this corresponds more closely to the actual temperature dis- tribution (see Sec. 12.6.5); (2) a model for the boundary-layer flow provides a on instantaneous cylinder volume instead of bore, changes in the effective gas more fundamental basis for evaluating the heat-transfer coefficient. The local heat velocity, and in the exponent of the temperature term. 682 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE HEAT TRANSFER 683 flux is then given by18 was developed. Here Rep = D, Dp/v, where D, is the time-averaged exhaust port ke(Tac - Tw) gas velocity and D, is the port diameter. For straight sections of exhaust pipe q = " (12.22) downstream of the port, an empirical correlation based on measurements of average heat-transfer rates to the pipe has been derived:21 where ke is the effective thermal conductivity in the boundary layer, Tac is the adiabatic core temperature, and ò the boundary-layer thickness. In the laminar Nu = 0.0483 Re0.783 (12.25) regime o would grow as t1/2; in the turbulent regime o would grow as to.8. Both The Reynolds number is based on pipe diameter and velocity. Heat-transfer cor- growth regimes are observed. relations for steady developing turbulent flow in pipes predict values about half Zonal models have also been used to describe DI diesel engine heat trans- that given by Eq. (12.25). fer.19 A bowl-in-piston chamber was divided into three flow regions and two gas-temperature zones during combustion. An effective velocity 12.5 RADIATIVE HEAT TRANSFER w = (U2 + U3 + 2k) 1/2 There are two sources of radiative heat transfer within the cylinder: the high- was used to obtain the heat-transfer coefficient, where Ux and U, are the two temperature burned gases and the soot particles in the diesel engine flame. In a velocity components parallel to the surface outside the boundary layer and k is spark-ignition engine, the flame propagates across the combustion chamber from the turbulent kinetic energy. Zonal models would be expected to be more accu- the point of ignition through previously mixed fuel and air. Although the flame rate than global models. However, only limited validation has been carried out. front is slightly luminous (see color plate, Fig. 9-1), all the chemical intermediaries in the reaction process are gaseous. Combustion is essentially complete early in the expansion stroke. In the compression-ignition engine (and in fuel-injected 12.4.5 Intake and Exhaust System Heat Transfer stratified-charge engines), most of the fuel burns in a turbulent diffusion flame as fuel and air mix together. There can be many ignition locations, and the flame Convective heat transfer in the intake and exhaust systems is driven by much conforms to the shape of the fuel spray until dispersed by air motion (see color higher flow velocities than in-cylinder heat transfer. Intake system heat transfer is plate, Figs. 10-4 and 10-5). The flame is highly luminous, and soot particles usually described by steady, turbulent pipe flow correlations.9 With liquid fuel (which are mostly carbon) are formed at an intermediate step in the combustion present in the intake, the heat-transfer phenomena become especially complicated process. (see Sec. 7.6). Exhaust flow heat-transfer rates are the largest in the entire cycle The radiation from soot particles in the diesel engine flame is about five due to the very high gas velocities developed during the exhaust blowdown times the radiation from the gaseous combustion products. Radiative heat trans- process and the high gas temperature (see Sec. 6.5). Exhaust system heat transfer fer in conventional spark-ignition engines is small in comparison with convective is important since it affects emissions burnup in the exhaust system, catalyst, or heat transfer. However, radiative heat transfer in diesel engines is not negligible; particulate trap, it influences turbocharger performance, and it contributes sig- it contributes 20 to 35 percent of the total heat transfer and a higher fraction of nificantly to the engine cooling requirements. the maximum heat-transfer rate. The highest heat-transfer rates occur during blowdown, to the exhaust valve and port. Detailed exhaust port convective heat-transfer correlations have been developed and tested. These are based on Nusselt-Reynolds number correla- 12.5.1 Radiation from Gases tions. For the valve open period, relations of the form Gases absorb and emit radiation in narrow wavelength bands rather than in a Nu = K Ref (12.23) continuous spectrum as do solid surfaces. The simpler gas molecules such as H2, O2, and N2 are essentially transparent to radiation. Of the gases important in have been proposed and evaluated.2º For L./D, < 0.2, the flow exits the valve as combustion, CO, CO2, and H2O emit sufficient energy to warrant consideration. a jet, and Re; = v; DD/v, where D, is the valve diameter, v, the velocity of the In gases, emission and absorption will occur throughout the gas volume. These exhaust gases through the valve opening, and v the kinematic velocity. For L_/DD, 2 0.2, the port is the limiting area and a pipe flow model with Re = v, Dp/v processes will be governed by the number of molecules along the radiation path. For each species, this will be proportional to the product of the species partial is more appropriate. v, is the velocity in the port and D, is the port diameter. For pressure p; and the path length /. In addition the radiative capacity depends on the valve closed period, the correlation gas temperature T2. Thus the emissivity of the gas & can be expressed as Nu = 0.022 Reg.8 (12.24) &g = f(T ,, pil, ..., Pnl) (12.26) 684 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE HEAT TRANSFER 685 1.0 The mean path length for a volume V with surface area A is given with sufficient 1.0 10º ATC accuracy by 20º ATC - 30º ATC. 0.8 0.8 1 = 0.9 x- 4V A (12.27) 0.6 4V/A is the mean path length for a hemispherical enclosure. Monochromatic emissivity Monochromatic emissive power, MW/m2 . um Standard methods have been developed for estimating & for mixtures of CO2 and H2O by Hottel and others (see Ref. 22). Charts based on experimental 0.4+ 0.4 data give &co, and 6H2o as functions of p, I and Te. Correction factors are applied for total pressures above one atmosphere and for the overlapping of spectral FIGURE 12-6 bands of CO2 and H2O. Estimates for engine combustion gases at peak condi- 0.2 0.2 Variation in monochromatic emissive power and tions give & ~ 0.1 and peak heat fluxes due to gas radiation of order 0.2 MW/ emissivity, with wavelength, at three different crank angles. DI diesel engine with 114 mm bore, m2. This amounts to ~5 percent of the peak convective heat transfer. Since gas 1995 rev/min, overall equivalence ratio 0.46. Radi- radiation is proportional to T4, this radiative flux falls off more rapidly from 2 3 5 ation from piston bowl measured through cutout peak values than convective heat flux and, when integrated over the cycle, can be Wavelength, um in piston crown.23 neglected. 12.5.2 Flame Radiation matic emissivity. Equation (12.28), combined with Planck's equation for black- Flame radiation is a much more complex process because the detailed geometry body monochromatic emissive power and chemical composition of the radiating region are not well known. Since the 2zKI radiation from the optically transparent or nonluminous flames of spark-ignition eb. 2 - 15 ( ex2/2TR - 1) (12.29) engines is small, we will deal only with luminous nontransparent flames where the radiation comes from incandescent soot particles and has a continuous spec- where K1 = 0.59548 x 10-16 W . m2 and K2 = 1.43879 cm . K, defined an appar- trum. Because the particle size distribution, number density and temperature, and ent radiation temperature TR and optical thickness kl for the radiating medium. flame geometry in a diesel engine are not well defined, flame emissivities cannot An apparent grey-body emissivity &, was also calculated from the standard equa- tion be calculated from first principles. Direct measurements of flame emissivities are required. A number of measurements of the magnitude and spectral distribution of radiation from a diesel engine combustion chamber have been made (see Ref. 9 for a summary). The most extensive of these by Flynn et al.23 in a direct-injection Jeb . adi engine used a monochromator to measure intensity of radiation at seven wave- Figure 12-7 shows sample results for four equivalence ratios at an engine speed of lengths. The viewing path cut through the piston crown into the central region of 2000 rev/min. The radiation flux has approximately the same shape and time the bowl-in-piston combustion chamber. Fuel was injected through a five-hole span as the net heat-release rate curve (which was determined from the cylinder nozzle and some air swirl was provided. At any given crank angle, the distribu- pressure curve). During the period of maximum radiation, the apparent emis- tion of energy over the seven wavelengths was used to reconstruct the complete sivity is 0.8 to 0.9; it then decreases as the expansion process proceeds.23 In energy spectrum and to calculate the apparent radiation temperature and optical previous experiments on the same engine, instantaneous total heat fluxes had thickness. been measured at various locations on the cylinder head.24 A comparison of The energy distribution was skewed from that of a grey-body model (for radiant and total heat fluxes (both peak and average) showed that the radiation which emissivity is independent of wavelength), and the monochromatic emis- heat flux can be a substantial fraction of the peak heat flux. The average radiant sivity was well fitted by the equation flux is about 20 percent of the average total flux: the percentage varies signifi- cantly with load. These conclusions are supported by other experimental data 82 = 1 - exp (12.28) summarized in the next section. The radiation or apparent flame temperatures measured in diesels by used to describe the emissivity variation from clouds of small particles. Figure several investigators show consistent results (see Fig. 12-8). Also included in the 12-6 shows sample results for the monochromatic emissive power and monochro- figure during the combustion and expansion process are typical values of: (1) the 1.5 Equivalence ratio ENGINE HEAT TRANSFER 687 0.23 0.46 50 TTTTTTT 1.0 0.51 Unstabilized pressure 0.75 jet oil flame Radiant flux, MW/m2 o Stabilized pressure jet oil flame 0.5 O Candle flame Diesel exhaust 10 ₺ 30 60 90 120 150 180 Values for - > 10 2500 2000 Calculated for c. m-/8 propane flame, Radiant temperature, K 1500} acetylene flame 1000H 500+ 0 30 60 90 120 150 180 0.5- Calculated for FIGURE 12-9 amorphous soot 100 r at 2250 K (28) Measured and calculated values of k2/c, as a function of wavelength 1. ka 80- is monochromatic absorption coeffi- 0.2 cient, c, is the soot concentration. 60 0.5 (From Field et al.28 and Greeves and Heat release rate, ml/deg Wavelength À, um Meehan.29) 40 20 temperature of any air not yet mixed with fuel or burned gases; (2) the average 30 90 120 150 180 temperature of the cylinder contents; and (3) the maximum possible flame tem- Crank angle, deg perature [corresponding to combustion of a slightly rich mixture (( = 1.1) with air at the temperature shown]. The measured radiation temperatures fall between FIGURE 12-7 Radiant heat flux, apparent radiant temperature and net heat-release rate as function of crank angle the maximum flame temperatures and the bulk temperature. Such a model has for DI diesel engine at four different loads. Engine and measurement details as in Fig. 12-6; 2000 been proposed, fits the available data, and has been used.27 Zonal models have rev/min.23 been proposed (e.g ., the burned gas region is stoichiometric16,19) to define an appropriate flame temperature. The emissivity of an incandescent soot-burning flame can be calculated 3000 from a knowledge of the monochromatic absorption coefficient, which is given by 2500 Flame k2 = 36m n 2 K 2000 FIGURE 12-8 Ps 2[(n'2 + n'2x2)2 + 4(n'2 - n2x2 + 1)] Ps = 367 3 f( 2, T) (12.30) Apparent radiation temperatures measured in Temperature, K 1500- diesel engines, compared to calculated maximum where n is the refractive index, k is the absorption term in the complex refractive adiabatic flame, cylinder-mean and air tem- 1000 Mean gas index, c, is the soot concentration in kilograms per cubic meter, and p, is the peratures, during the combustion period. Adia- density of the soot particles (~2g/m3). The absorption coefficient dependson Air batic flame temperature is temperature attained by 500 burning air at air temperature with fuel for equiva- wavelength and temperature, is independent of particle size, and depends only on TC lence ratio of 1.1. (Data from Dent and Sulaiman, 16 the soot mass loading. Figure 12-9 shows several estimates of k1/c, as a function L 340 360 380 400 420 440 460 Flynn et al ., 23 Lyn,25 Kamimoto et al.;26 calculated of 1. The strong dependence on 1 shows that clouds of soot particles are mark- Crank angle, deg curves from Assanis and Heywood.27) edly not grey. There is a considerable spread in the different estimates shown. 686 688 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE HEAT TRANSFER 689 This could be the result of different soot compositions (the C/H ratio affects the optical properties) and temperatures. To find a mean value for the absorption lation gave a reasonable match to Flynn's data, an example of which was shown in Fig. 12-7. coefficient of a wide spectrum of radiation, the expression It has been proposed that the apparent absorptivity should be proportional IKa eb, 2 da to pressure.4, 53 The assumption is then made that the proportionality constant K = feb , 2 di would be a unique function of the equivalence ratio and crank angle. Although ka is dependent on p for gas radiation, it is not clear that the same proportionality must be evaluated. For example, for radiation from a black body at temperature should apply to soot radiation. 1800 K, the mean absorption coefficient is ~1300c, per meter, where c, is in kilograms per cubic meter. At higher black-body temperatures, the value of k, would be higher. Annand3º has applied this approach to a diesel engine. The apparent flame 12.6 MEASUREMENTS OF emissivity was related to the apparent mean absorptivity by INSTANTANEOUS HEAT-TRANSFER RATES 84 = 1 - exp (-ka !) (12.31) 12.6.1 Measurement Methods For Flynn's data, the peak emissivity is 0.8 which gives kal = 1.6. Since Values of instantaneous heat flux into the combustion chamber walls have been 1 ~ 0.07 m, this gives ka ~ 22 m-1 and , ~ 16 g/m3 (~1 g/m3 at NTP), which is obtained from measurements of the instantaneous surface temperature. The tem- a soot loading comparable with values measured in diesel engines during com- perature variation at the wall is a result of the time-varying boundary condition bustion (see Figs. 11-42 to 11-44). at the gas/wall interface. It is damped out within a small distance (~1 mm) from the wall surface, so measurements must be made at the surface. Various types of 12.5.3 Prediction Formulas thermocouple or thermistor have been used.9 One-dimensional unsteady heat conduction into the wall is then assumed: Well-accepted prediction formulas for radiant heat flux in an engine are not available. Annand has proposed a radiation term of the form 1 0 at pc ax (12.34) àR = Bo(Ta - T4) (12.32) A sinusoidal variation with time of heat flux into a semi-infinite solid can be where o is the Stefan-Boltzmann constant, T, is the mean gas temperature, and shown to produce a sinusoidal variation of surface temperature of the same fre- Tw is the wall temperature. This term was coupled with a convective heat flux quency displaced in phase by 90º. The surface temperature Tw is expressed as a term to give a correlation for predicting total heat flux. In a first evaluation, when Fourier series: coupled with Eq. (12.14), f = 0.6 was proposed.8 In a later study with a modified convective heat-transfer correlation,31 6 ~ 1.6 was proposed. (Note that since the temperature used is the average gas temperature and not the apparent flame Tw = Tm + [[A, cos (not) + B, sin (not)] (12.35) temperature, ß is not an emissivity.) The limited evaluation of this approach shows that f = 0.6 gave approximately correct magnitude for aR for one engine where Tu is the time-averaged value of Tw, A ,, and B ,, are Fourier coefficients, n is and was too low for another.32 ß = 1.6 gave radiant heat fluxes higher than a harmonic number, and w is the angular frequency (radians per second). The experimental data.33 boundary conditions are T = Tw(t) at x = 0 and T = T; (constant) at x = l. The Flynn et al.23 developed an empirical expression for instantaneous radiant solution of Eq. (12.35) is1 heat flux to fit their data, of the form dR = 2GR b(a + 1)( (0 - 0) T(x, t) = Tm - (Tm - 1) + + [ exp ( -. x)F.(x, t) (12.36) 360 exp 10 - 0,10 + 1- (12.33) 360 where where 0, is the crank angle at the start of the radiation pulse. Correlations for aR. a, b, and 0, in terms of engine speed, manifold pressure, crank angle at the start of F. = A ,, cos (not - (, x) + B ,, sin (not - (,x) fuel injection, and the equivalence ratio were obtained and presented. This corre- and ¢, = (nœ/2x)1/2, where a is the thermal diffusivity of the wall material k/(pc). 690 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE HEAT TRANSFER 691 390 1.25 flux occurs depends on the flame arrival time at the measurement location. Thus Intake Compression Expansion Exhaust Intake Compression Expansion Exhaust the heat fluxes determined from surface temperature data averaged over many 1.00 - 388 cycles show their rapid rise later, the further the distance from the spark plug 0.75- location (Fig. 12-11a). The individual cycle data in Fig. 12-11b show how varia- 386 tions in the flame arrival time from one cycle to the next essentially shift the 0.50 Surface temperature, K Surface heat flux, MW/m2 rising portion of the heat flux profile in time. Note that due to this cyclic varia- 384 0.25 - tion, the average profile shows a less rapid rise in heat flux than do individual Ignition Ignition cycles. 382 0.00 Such measurements show that increasing engine speed and increasing IC TC 380 -0.25 engine load increase the surface heat flux. Retarding timing delays the rise in heat 0 180 360 540 720 180 360 540 720 Crank angle, deg Crank angle, deg FIGURE 12-10 Flame arrival Surface temperature measured with thermocouple in cylinder head, and surface heat flux calculated A ------ HT1 from surface temperature, as a function of crank angle. Spark-ignition engine operated at part-load.34 3.0H o --- HT2 0 -.-.- HT3 -- HT4 The heat flux components at each frequency that caused that variation can be calculated via Fourier's law [Eq. (12.1)], and summed to give the total fluctua- tion of heat flux with time: w = (Tm -T) + k [ Q.[(A, + B.) cos (not) - (A, - B,) sin (not)] (12.37) Surface heat flux, MW/m2 = 1 Alternative approaches for solving Eq. (12.34) are through use of an electrical 1.0- analogy to heat flow and by numerical methods. The latter become necessary if Woschni wall material properties depend significantly on temperature, as do combustion chamber deposits and some insulating ceramic materials. Several measurements Pmax of this type in spark-ignition and diesel engines have been made. A summary of -3 these measurements can be found in Ref. 9. 340 350 360 370 380 390 400 410 Radiant heat fluxes are determined by a variety of techniques: e.g ., photo- Crank angle, deg detector and infrared monochromator; thermocouple shielded by a sapphire (a) window; pyroelectric thermal detector. FIGURE 12-11 21 (a) Variation of surface heat flux with crank angle at four temp- 12.6.2 Spark-Ignition Engine Measurements erature measurement locations in 1.5 the cylinder head of a spark- Figure 12-10 shows the surface temperature variation with crank angle, and the ignition engine. Each curve is an heat flux variation calculated from it, on the cylinder head of a spark-ignition Average average over many cycles. Dis- engine at a part-load low-speed operating condition. The swing in surface tem- tances from on-axis spark plug Surface heat flux, MW/m2 1 are: HT1, 18.7 mm; HT2, 27.5 mm; perature at this point (about halfway from the on-the-cylinder-axis spark plug to HT3, 37.3 mm; HT4, 46.3 mm. the cylinder wall) is 7 K. The heat flux rises rapidly when the flame arrives at the Bore = 104.7 mm, 2000 rev/min, measurement location, has its maximum at about the time of peak cylinder pres- 0.5 part-load n. = 40 percent, A/F = 18, sure when gas temperatures peak (see Section 9.2.1), and then decays to relatively MBT timing. Solid curve shows heat low levels by 60º ATC as expansion cools the burned gases. Peak heat fluxes on flux predicted by Woschni's correla- the cylinder head of 1.5 to 3 MW/m2 were measured over the normal engine tion.34 (b) Heat flux histories for five 340 360 380 400 420 consecutive individual cycles and speed and load range.34, 35 440 198-cycle average at location HT1. The heat flux profile varies significantly with location and from one cycle to Crank angle, deg 1500 rev/min, A/F = 18, n = 40 the next. Figure 12-11 illustrates both these effects. When the rapid rise in heat (b) percent, MBT timing.35 692 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE HEAT TRANSFER 693 flux and reduces the peak value. Maximum heat fluxes occur with close-to- TC1 stoichiometric mixtures.34, 35 All these trends would be expected from the varia- 4.0r tions in burned gas temperature that result from these changes in engine TC! AV EV operation. TC2 + 3.0 TC3 A TC4 12.6.3 Diesel Engine Measurements TC8 2.0 Injector tip Measurements of instantaneous heat fluxes in diesel engines show similar fea- tures. The heat flux distribution is usually highly nonuniform. There may also be TC2 significant variations between the heat flux profiles from individual cycles. Figure 1.0 Top view of cylinder head 12-12 shows surface heat fluxes at two locations in a medium-swirl DI diesel, one over the piston bowl (higher heat flux) and the other over the piston squish area, Heat flux, MW/m2 TC6 TC3 in relation to the heat-release profile. The heat flux increases rapidly once com- TC7 TC4 bustion starts, reaches a maximum at close to the time of maximum cylinder -0.5 pressure, and decreases to a low value by 40 to 60º ATC. Peak heat fluxes to the 0 90 180 270 360 BC TC BC TC8 primary combustion chamber surfaces (the piston bowl and head directly above the bowl) are of order 10 MW/m2. 0.5 TC5 In smaller diesel engines with swirl, the mean heat flux to the piston within the piston bowl is usually higher than the mean heat fluxes to the cylinder head and the annular squish portion of the piston crown (by about a factor of 0.5L 360 450 540 630 720 two). 16, 31 This would be expected since the piston bowl is the zone where most BC TC BC of the combustion takes place and gas velocities are highest. There are, in addi- Crank angle, deg Cylinder sleeve, section A-A tion, substantial variations in heat flux at different locations within the piston FIGURE 12-13 bowl, on the head, and on the annular region of the piston crown. Measured surface heat fluxes at different locations in cylinder head and liner of naturally aspirated In contrast to the above results obtained on smaller high-speed (~ 10-cm four-stroke cycle DI diesel engine. Bore = stroke = 114 mm, 2000 rev/min, overall equivalence ratio = 0.45.15 bore) diesels with swirl with deep bowl-in-piston combustion chambers of diam- eter about half the piston diameter, results from tests on a medium-speed 30-cm bore quiescent shallow-bowl piston direct-injection supercharged diesel showed a percent of the flux to the primary combustion chamber surfaces. Again this much more uniform heat flux distribution over the combustion chamber walls.37 would be expected, since the combustion gases do not contact the lower parts of Heat fluxes to the cylinder liner are much lower still (an order of magnitude the cylinder wall until later in the expansion stroke when their temperature is less than the peak flux to the combustion chamber surface) and are also nonuni- much below the peak value. form. Figure 12-13 compares heat fluxes to the cylinder head with three locations Figure 12-14 shows examples of radiant heat flux measured above the along the liner. Even at the top of the liner, the peak heat flux is only 15 to 20 piston bowl of a medium-swirl DI diesel engine as a function of engine speed and load.16 The limited radiant heat flux data available exhibit the following trends. 8 The rapid increase in radiant heat flux following combustion is delayed relative 50% heat TCITC2 release to the start of pressure rise due to combustion (by about 5º); this delay increases 6 End of with increasing speed. The peak radiant flux remains approximately constant premixed -ICI with increasing equivalence ratios up to o ~ 0.5. Further increases in the equiva- 4 Ignition - TC2 FIGURE 12-12 lence ratio produce a drop in level of radiant flux. The time-averaged radiant Surface heat flux, MW/m2 2 90% heat release! Measured surface heat flux at two locations in the heat flux increased approximately linearly with increasing manifold pressure; cylinder head of a medium-swirl DI diesel engine. however, peak radiative flux levels remained essentially unchanged. Peak and TC1 above the piston bowl, TC2 above the piston 0 time-averaged values of the radiant heat flux decreased as injection timing was squish area as shown. Percentages of heat release 70% heat release retarded. are indicated. Bore = stroke = 114 mm, r = 16, -24 - 40 -20 0 20 40 60 80 2000 rev/min, overall equivalence ratio = 0.5, intake In diesel engines, the relative importance of radiant heat transfer (as a per- Crank angle, deg pressure = 1.5 atm.36 centage of the total heat flux) depends on the location on the combustion 694 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE HEAT TRANSFER 695 T 600 Engine speed 1500 1750 1050 rev/min rev/min rev/min 3 450 Load 0% 40% 80 % 300 FIGURE 12-14 Radiant heat flux, kW/m2 150| Measured radiant heat flux to Annand cylinder head above the piston Flat plate bowl in a high-swirl DI diesel 0 Heat flux, MW/m2 engine when load and speed are Woschni varied. Solid curve: 80 percent -150 -45 TC +45 -45 TC+45 -45 TC +45 load. Dotted curve: 40 percent load. Dashed curve: no load.16 Measured Crank angle, deg chamber surface, crank angle, engine load, engine size, and engine design. The time-averaged radiant heat transfer increases as a proportion of the total heat transfer, with increasing load, as indicated in Fig. 12-15.9 At high load, the total -45 TC +45 -45 TC +45 -45 TC +45 radiant heat flux is between 25 and 40 percent of the total time-averaged heat Crank angle, deg flux. FIGURE 12-16 Comparison of measured mean heat fluxes on the cylinder head at 1050 rev/min in a fired high-swirl DI diesel engine with various prediction equations. Annand, Eq. (12.14), Woschni, Eq. (12.19), flat 12.6.4 · Evaluation of Heat-Transfer Correlations plate, Eq. (12.20) using measured gas motion.16 The convective (or combined convective plus radiative) heat flux correlations have been compared with instantaneous engine heat-transfer measurements. One difficulty in this evaluation is the determination of spatially averaged More extensive comparisons have been made of predictions with data for combustion-chamber heat fluxes from the experimental data for comparison with diesel engines. Annand's correlations, Eqs. (12.14) and (12.32) with a = 0.06, correlations intended to predict the mean chamber heat flux as a function of b = 0.85, and f (for the combustion phase only) = 0.57, gave reasonable agree- crank angle. The area-averaged instantaneous heat flux prediction using ment with instantaneous cylinder head heat fluxes, and overall time-averaged Woschni's equation (12.19) for the spark-ignition engine conditions shown in Fig. heat fluxes for a medium-speed quiescent DI engine design.37 In a small high- 12-11a is comparable in magnitude to, though lower than, the measured heat speed diesel with swirl, values of a = 0.13, b = 0.7, and ß = 1.6 gave an approx- imate fit to estimates of the instantaneous area-mean heat flux31 and fluxes to the cylinder head. time-averaged heat flux to the piston.38 Comparisons of the Annand and Woschni correlations generally show that the Annand correlation predicts higher heat fluxes at the same crank angle.11, 24 The most careful comparison of all three correlations summarized in Secs. 12.4.3 and 12.4.4 with experimental data Oguri, 600 rev/min engine B . has been made by Dent and Sulaiman in a small high-speed diesel engine with 0.3 swirl.16 The mean experimental heat flux was estimated from a number of ther- -0- 4 - - 4_ mocouple measurements located at different points around the combustion Ebersole chamber surface. Figure 12-16 shows the comparison. The Woschni correlation Oradiat Quota falls below the others at light load because the combustion-induced velocity term Sitkei is smaller. Expansion stroke heat fluxes are underpredicted by all three correla- tions. Given the uncertainty in converting the measurements to an average heat 0.1- Oguri, 900 rev/min flux value, the agreement is reasonably good. engine A Dent examined additional modifications to the flat plate formula [Eq. FIGURE 12-15 40 Radiant heat flux as fraction of total heat flux over (12.20)], which was based on a cylinder-mean gas temperature. During com- 20 60 80 100 .................... . Percent max imep the load range of several different diesel engines.9 bustion a two-zone model is more appropriate. Assuming an equivalence ratio 696 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE HEAT TRANSFER 697 Spark Prediction using bulk & ## # # Exhaust mean temperature Intake 21 Combustion Prediction using 2-zone temperature and swirl increase 4- Heat flux, MW/m2 * Measured (mean) T TC FIGURE 12-17 1380 rev/min 3 Boundary layer thickness, mm Measured heat fluxes on cylinder head in fired Pi = 0.7 atm high-swirl DI diesel engine at 1050 rev/min and 40 percent load compared with predictions based on bulk mean gas temperature and using tem- 2 -90 -45 TC 45 90 135 peratures based on two-zone (air and burned gas) Crank angle, deg model. 16 TC 90 BC 90 TC 90 BC -90 TC for the burned gas (Dent assumed stoichiometric), a combustion zone tem- Crank angle, deg perature can be determined from the relation FIGURE 12-18 Thermal boundary-layer thickness, at the top of the cylinder wall in the clearance volume, determined T = mTe - m. To from schlieren photographic measurements in a special visualization square-piston spark-ignition engine. 39 mb where ma is the mass of air, Ta is the air temperature, and m, is the burned gas mass, my = m - m .. In addition, the observed swirl enhancement which occurs 12.6.5 Boundary-Layer Behavior due to the combustion was included by multiplying the swirl velocity used in the Measurements of thermal boundary-layer thickness in an operating spark- heat-transfer correlation by the square root of the ratio (density of air in the ignition engine have been made using schlieren photography in a special flow- motored case)/(density of burned gas in fired case). The combination of both visualization engine. Figure 12-18 shows one set of measurements on the cylinder effects (see Fig. 12-17) improves the shape of the predictions by broadening and wall in the clearance volume opposite the spark plug. The boundary-layer thick- lowering the peak. ness decreases during intake and increases steadily during compression and Each of the convective-heat-transfer correlations described has limited expansion to about 2 mm. It stops growing and becomes unstable during the experimental support. However, under engine design and operating conditions exhaust process, separating from the cylinder wall and becoming entrained into different from those under which they were derived, the predictions should be the bulk gas leaving the cylinder. The thickness of the thermal boundary layer viewed with caution. Woschni's correlation is the correlation used most exten- varies substantially at different locations throughout the chamber. While the sively for predicting spatially averaged instantaneous convective heat fluxes. trends with crank angle were similar, the layers on the cylinder head and piston However, the empirical constants which relate the mean piston speed and crown were substantially (up to 2 to 3 times) thicker during compression and combustion-induced motion to the velocity used in the Reynolds number, deter- expansion in the simple disc-shaped chamber studied.39 This different behavior mined by Woschni, will not necessarily apply to all the different types of engine. probably results because there is no bulk flow adjacent to the head and piston If local velocities are known, the flat-plate-based correlations provide the best crown. available approach. Annand's correlation has the advantage of being the simplest Estimates of thermal boundary-layer thickness in spark-ignition engines, correlation to use. Since the radiation component in diesel engine heat transfer is based on convective-heat-transfer correlations and thermal energy conservation normally 20 to 40 percent of the total an approximate estimate of its value may for the growing layer give thicknesses comparable to these measurements. Note suffice. that a substantial fraction of the cylinder mass is contained within the thermal 698 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE HEAT TRANSFER 699 boundary layer. For example, for an average thickness of 3 mm at 90º ATC Upper 5 Between inlet and during expansion, the volume of the boundary layer is 20 percent of the com- Lower 6 exhaust valves bustion chamber volume for typical engine dimensions. Since the average density Aluminum cylinder head in the boundary layer is about twice that in the bulk gases, some 30 to 40 percent Under Upper 7 (9 )Lower 240- --- Cast iron cylinder head valve 11 1 4 of the cylinder mass would be contained within this boundary layer. seat Lower 8 10 Upper 220- 200 12.7 THERMAL LOADING AND Metal temperature, ºC 180 COMPONENT TEMPERATURES 160 The heat flux to the combustion chamber walls varies with engine design and 140 operating conditions. Also, the heat flux to the various parts of the combustion chamber is not the same. As a result of this nonuniform heat flux and the differ- 120 ent thermal impedances between locations on the combustion chamber surface 1001 2 3 4 5 6 17 8 19 10 11 and the cooling fluid, the temperature distribution within engine components is Location of thermocouples nonuniform. This section reviews the variation in temperature and heat flux in FIGURE 12-20 the components that comprise the combustion chamber. Variation of cylinder head temperature with measurement location in spark-ignition engine operating at 2000 rev/min, wide-open throttle with coolant water at 95ºC and 2 atm.41 12.7.1 Component Temperature Distributions bowl. In IDI diesel engines, maximum piston temperatures occur where the pre- Figures 12-19 to 12-22 show illustrative examples of measured temperature dis- chamber jet impinges on the piston crown. tributions in various engine components. Normally, the heat flux is highest in the Figure 12-20 shows the temperatures at various locations on a four-cylinder center of the cylinder head, in the exhaust valve seat region, and to the center of SI engine cylinder head. The maximum temperatures occur where the heat flux is the piston. It is lowest to the cylinder walls. Cast-iron pistons run about 40 to high and access for cooling is difficult. Such locations are the bridge between the 80ºC hotter than aluminum pistons. With flat-topped pistons (typical of spark- valves and the region between the exhaust valves of adjacent cylinders. Figure ignition engines) the center of the crown is hottest and the outer edge cooler by 12-21 shows how the average heat flux and temperature vary along the length of 20 to 50ºC. Diesel engine piston crown surface temperatures are about 50ºC a DI diesel engine liner. Because the lower regions of the liner are only exposed higher than SI engine equivalent temperatures. As shown in Fig. 12-19, the to combustion products for part of the cycle after significant gas expansion has maximum piston temperatures with DI diesel engine pistons are at the lip of the occurred, the heat flux and temperature decrease significantly with distance from -260- kW/m2 200 100 0 150ºC. -240- 140 -220+ 130 9GL 120 180 110 160 100 95 120 120 FIGURE 12-21 90 FIGURE 12-19 Temperature and heat flux distribution in the Isothermal contours (solid lines) and heat flow paths (dashed lines) determined from measured tem- cylinder liner of a high-speed DI diesel engine at perature distribution in piston of high-speed DI diesel engine. Bore 125 mm, stroke 110 mm, re = 17, 1500 rev/min and bmep = 11 bar. àr is heat flux 3000 rev/min, and full load.40 into the liner; L is heat flux from the gas to the LA BC liner. Difference is friction-generated heat flux.42 700 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE HEAT TRANSFER 701 800ºC Ring belt zone Crown internal zone 528 700-1 A\ Pin boss zone Skirt zone 681% 649 1500 rev/min, no load 3000 rev/min, full load 600 80 500 659 `649 Ring lands Ring lands 682 60 Oil ring Oil ring Taper-faced second ring Taper-faced second ring Percentage 40 FIGURE 12-22 500 600 700 800ºC Temperature distribution in one of the four exhaust Plain top ring Plain top ring valves of a two-stroke-cycle uniflow DDA 4-53 DI diesel engine. Bore == 98 mm, stroke = 114 mm.43 20 the cylinder head. Note that the heat generated by friction between the piston OL and the liner, the difference between deL (the gas to liner heat flux) and ar (the total heat flux into the liner), is a significant fraction of the liner thermal loading. FIGURE 12-24 The exhaust valve is cooled through the stem and the guide, and the valve Heat outflow from various zones of piston as percentage of heat flow in from combustion chamber. High-speed DI diesel engine, 125 mm bore, 110 stroke, r = 17.40 seat. In small-size valves the greater part of the heat transfer occurs through the stem; with large-size valves, the valve seat carries the higher thermal load. 14.4). These define the time-averaged heat flux into the piston. Heat-transfer coef- Temperature distributions in engine components can be calculated from a ficients for the different surfaces of the piston (underside of dome, ring-land areas, knowledge of the heat fluxes across the component surface using finite element ring regions, skirt outer and inner surfaces, wrist pin bearings, etc.) were esti- analysis techniques. For steady-state engine operation, the depth within a com- mated. The actual piston shape was approximated with a three-dimensional grid ponent to which the unsteady temperature fluctuations (caused by the variations for one quadrant of the piston. A standard finite element analysis of the heat flow in heat flux during the cycle) penetrate is small, so a quasi-steady solution is through the piston yields the temperature distribution within the piston. The satisfactory. Results from such calculations for a spark-ignition engine piston thermal stresses can therefore be calculated and added to the mechanical stress illustrate the method.44 A mean heat-transfer coefficient from the combustion field to determine the total stress distribution. This can be used to define the chamber gases to the piston crown and a mean chamber gas temperature were potential fatigue regions in the actual piston design. Figure 12-23 shows the tem- defined (using the input from a cycle simulation of the type described in Sec. perature distribution calculated with this approach, compared with measure- ments (indicated by dots). The agreement is acceptable, except in the piston skirt Piston pin plane Thrust plane where the heat-transfer rate between the skirt and cylinder liner has been over- estimated. 280 268 252 256 274 272 256 252 Detailed measurements of the temperature distribution in the piston allow 268 250 242 the relative amounts of heat which flow out of the different piston surfaces to be 221 estimated. Figure 12-24 shows examples of such estimates for a DI diesel engine -220 232 at no-load and full-load. About 70 percent of the heat flows out through the ring 205 228 226 177 193 zone, and much smaller amounts through the pin boss zone, underside of the 221 176 (196) crown and skirt. In larger diesel engines and highly loaded diesel engines, one or 240 224 ) more cooling channels are incorporated into the piston crown. This reduces the 228 . 201 172 (198) FIGURE 12-23 heat flow out through the ring area significantly.45 201.8 Measured (dots) and calculated temperature (ºC) 201.9 distributions in piston pin and thrust planes of 210 201 12.7.2 Effect of Engine Variables 162 C (196) the piston of a four-cylinder 2.5-dm3 spark- (207) ignition engine at 4600 rev/min and wide-open The following variables affect the magnitude of the heat flux to the different sur- throttle.44 faces of the engine combustion chamber and the temperature distribution in the 702 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE HEAT TRANSFER 703 components that comprise the chamber: engine speed; engine load; overall 32 equivalence ratio; compression ratio; spark or injection timing; swirl and squish 30 motion; mixture inlet temperature; coolant temperature and composition; wall material; wall deposits. Of these variables, speed and load have the greatest effect. 28- Equation (12.19), derived from the Nusselt-Reynolds number relation 20º 26 60º h = constant x B-0.2p0.87-0.55 y0.8 100º Heat transfer, % 24 and the relation for heat-transfer rate per unit area [Eq. (12.2)] 22 1400 rev/min à = h.(T - Tw) bmep = 325 kPa 20 FIGURE 12-26 MBT are useful for predicting trends as engine operating and design variables change. 5.7-dm3 engine Predicted average heat-transfer rate (as percent of fuel 18 flow rate x Q_y) to combustion chamber walls of an The effect of the above variables on engine and component heat flux will LL eight-cylinder 5.7-dm3 spark-ignition engine as a func- now be summarized. The comments which follow apply primarily to spark- 0.6 0.7 0.8 0.9 1.0 1.1 1.2 tion of equivalence ratio and burn rate (40; == ignition engines. In compression-ignition engines, the distribution of heat flux Equivalence ratio o combustion duration).46 and temperature varies greatly with the size of cylinder and form of the com- bustion chamber. engine and several diesel engine designs, with appropriate values of n, can be found in Refs. 47 to 49. While this correlation is not dimensionless and does not SPEED, LOAD, AND EQUIVALENCE RATIO. Predictions of spark-ignition satisfy Eq. (12.19), it provides a convenient method for reducing the experimental engine heat transfer as a function of speed and load are shown in Fig. 12-25. The data. The heat flux to the cylinder head and liner for a spark-ignition engine were cycle heat transfer is expressed as a percent of the fuel's chemical energy (mass of well correlated by Eq. (12.38) with n = 0.6. The flux distribution over the cylinder fuel X QLHy). The heat transfer to the total combustion chamber surface head at a fuel flow rate per unit piston area of 0.195 kg/s . m2 for several different (excluding the exhaust port) was calculated using a thermodynamic-based cycle DI diesel engines were comparable in magnitude. The effect of speed at wide- simulation (see Sec. 14.4). The relative importance of heat losses per cycle open throttle on component temperatures for a spark-ignition engine can be decreases as speed and load increase: the average heat transfer per unit time, found in Ref. 47. Exhaust valve, piston crown and top ring groove, and nozzle however, increases as speed and load increase. throat temperatures for a Comet prechamber diesel as a function of fuel flow rate Since speed and load affect p, T, and w in Eq. (12.19), simpler correlations can be found in Ref. 49. have been developed to predict component heat fluxes from experimental data. The peak heat flux in an SI engine occurs at the mixture equivalence ratio Time-averaged heat fluxes at several combustion chamber locations, determined for maximum power o ~ 1.1, and decreases as is leaned out or enriched from from measurements of the temperature gradient in the chamber walls, have been this value.5º The major effect is through the gas temperature in Eqs. (12.2) and fitted with the empirical expression (12.19). However, as a fraction of the fuel's chemical energy, the heat transfer per cycle is a maximum at o = 1.0 and decreases for richer and leaner mixtures, as à = constant x (12.38) shown by the thermodynamic-based cycle-simulation predictions in Fig. 12-26. In CI engines, the air/fuel ratio variation is incorporated directly in the load varia- with n between 0.5 and 0.75 (the value of n depending on engine type and loca- tion effects. tion within the combustion chamber). Results for a modern four-cylinder SI COMPRESSION RATIO. Increasing the compression ratio in an SI engine decreases the total heat flux to the coolant until re ~ 10; thereafter heat flux 50 increases slightly as re increases.5º The magnitude of the change is modest; e.g ., a 40- A B. 10 percent decrease in the maximum heat flux (at the valve bridge) occurs for an 30 C increase in re from 7.1 to 9.4.47 Several gas properties change with increasing Heat transfer, % bmep, kPa FIGURE 12-25 compression ratio (at fixed throttle setting): cylinder gas pressures and peak 20 A 163 Predicted average heat-transfer rate (as percent of 10- ฿ 325 fuel flow rate x QLy) to combustion chamber burned gas temperatures increase; gas motion increases; combustion is faster; the € 655 walls of an eight-cylinder 5.7-dm3 spark-ignition surface/volume ratio close to TC increases; the gas temperature late in the expan- OL 500 700 1000 2000 3000 4000 engine as a function of speed and load. Stoichio- sion stroke and during the exhaust stroke is reduced. Measured mean exhaust Speed, rev/min metric operation; MBT timing.46 temperatures confirm the last point, which probably dominates the trend at lower 704 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE HEAT TRANSFER 705 T 800 900 1400 rev/min 32 bmep = 325 kPa 48, = 20º Exhaust valve Spark plug 60º ¢ = 1.0 3001 700 800 1000 5.7-dm3 engine Piston Heat transfer, % 28 2801 340 320 12 FIGURE 12-27 260| 1320 300 24 15 Predicted average heat-transfer rate (as percent- age of fuel flow rate x QLHy) to combustion 300 280- 17 chamber walls of an eight-cylinder 5.7-dm3 Metal temperature, ºC 20 20 10: 5 MBT 5 spark-ignition engine as a function of spark 11 timing and burn rate (40) == combustion 210- 260€ Spark timing, deg duration).46 6 190- 240 - -150 2 compression ratios. As the compression ratio increases further, the other factors 10 (which all increase heat transfer) become important. 170] 160 {130 The effect of changes in compression ratio on component temperatures depends on location. Generally, head and exhaust valve temperatures decrease 18 40 110 70 80 100 110 120 with increasing compression ratio, due to lower expansion and exhaust stroke Cylinder liner Coolant outlet temperature, ºC temperatures. The piston and spark plug electrode temperatures increase, at con- -120 19 stant throttle setting, due to the higher peak combustion temperatures at higher Thermocouple locations 11 Behind top ring 2, 3, and 6 Cylinder head 12 In crown under spark plug compression ratios. If knock occurs (see Sec. 9.6), increases in heat flux and com- 70 80 90 100 110 120 7 Exhaust valve seat 15 Under exhaust valve 10-17 Piston ponent temperatures result; see below. 17 Under inlet valve Coolant outlet temperature, ºC 10 Top of skirt 18 and 19 Cylinder liner FIGURE 12-28 SPARK TIMING. Retarding the spark timing in an SI engine decreases the heat Effect of coolant temperature on cylinder head, liner, exhaust valve, valve seat, piston, and spark plug flux as shown in Fig. 12-27. A similar trend in CI engines with retarding injection metal temperatures. Spark-ignition engine at 5520 rev/min and wide-open throttle. r = 8.5.47 timing would be expected. The burned gas temperatures are decreased as timing is retarded because combustion occurs later when the cylinder volume is larger. perature in a spark-ignition engine. The exhaust valve and spark plug tem- Temperature trends vary with component. Piston and spark plug electrode tem- peratures are unchanged. The smaller response of the metal temperatures to peratures change most with timing variations; exhaust valve temperature coolant temperature change occurs at higher heat flux locations (such as the increases as timing is retarded due to higher exhausting gas temperatures.47 valve bridge), and indicates that heat transfer to the coolant has entered the nucleate-boiling regime in that region. The response is greater where heat fluxes SWIRL AND SQUISH. Increased gas velocities, due to swirl or squish motion, will are lower (e.g ., the cylinder liner), indicating that there heat transfer to the result in higher heat fluxes. Equation (12.19) indicates that the effect on local heat coolant is largely by forced convection. When nucleate boiling occurs (i.e ., when flux, relative to quiescent engine designs, should be proportional to (local gas steam bubbles are formed in the liquid at the metal surface, although the bulk velocity)º·8. There is no direct evidence to support this correlation but there is temperature of the coolant is below the saturation temperature), the metal tem- evidence that use of a shrouded value to increase gas velocities within the cylin- perature is almost independent of coolant temperature and velocity. Addition of der increases the total heat transfer.50 antifreeze (ethylene glycol) to coolant water changes the thermodynamic proper- ties of the coolant. INLET TEMPERATURE. The heat flux increases linearly with increasing inlet temperature; the gas temperatures throughout the cycle are increased. An WALL MATERIAL. While the common metallic component materials of cast iron increase of 100 K gives a 13 percent increase in heat flux. 51 and aluminum have substantially different thermal properties, they both operate with combustion chamber surface temperatures (200 to 400ºC) that are low rela- COOLANT TEMPERATURE AND COMPOSITION. Increasing liquid coolant tive to burned gas temperatures. There is substantial interest in using materials temperature increases the temperature of components directly cooled by the that could operate at much higher temperatures so that the heat losses from the liquid coolant. Figure 12-28 shows the result of a 50-K rise in coolant tem- working fluid would be reduced. Ceramic materials, such as silicon nitride and 706 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE HEAT TRANSFER 707 TABLE 12.2 2000 Thermal properties of wall materials - Ceramic 1600- --- Metal Peak Temperature, K 1200- Mean gas Thermal Specific Thermal Skin temperature ILLILLA FIGURE 12-29 800 conductivity Density heat diffusivity depth swing, iston surface Mean gas temperature and piston surface tem- Material k, W/m . K p, kg/m3 c, J/kg . K a, mª/s kpc 6, mm K 400 perature profiles predicted by turbocompounded DI diesel engine simulation for water-cooled metal 200 Cast iron 54 7.2 × 103 400 480 1.57 x 10~5 1.8 x 108 600 2.8 18 800 combustion chamber walls and for partly insulated Aluminum 155 2.75 × 103 915 6.2 x 10-5 3.9 x 108 5.4 12 Crank angle, deg engine with ceramic walls. 27 Reaction- 5-10 2.5 × 103 710 2.8 × 10-6 1.3 x 107 12 70 bonded silicon of this thermal energy diffuses through the wall, during intake and compression nitride much of it is transferred back to the now low-temperature cylinder contents. The Sprayed 1.2 5.2 x 103 732 3.2 x 10-7 4.6 x 106 0.39 95 depth of penetration of the thermal wave into the material, the skin depth 6, is zirconia proportional to va/w, where a = k/(pc) is the thermal diffusivity and @ the fre- quency of the wave (proportional to engine speed). Values of a and 6 are given in Table 12.2 for an engine speed of 1900 rev/min: a/d = 1.4, a constant. The magnitude of the temperature fluctuation (important because it is a source of zirconia, have lower thermal conductivity than cast iron, would operate at higher fluctuating thermal stress) is proportional to (opc) 1: this varies as (kpc)- 1/2. temperatures, and thereby insulate the engine. The thermal properties of some of Estimated peak temperature swings for the materials in Table 12.2 are tabulated. these materials are listed in Table 12.2. With these thermally insulating materials it is possible to reduce the heat transfer through the wall by a substantial KNOCK. Knock in an SI engine is the spontaneous ignition of the unburned amount. "end-gas" ahead of the flame as the flame propagates across the combustion This approach is most feasible for diesel engines where there is the possi- chamber. It results in an increase in gas pressure and temperature above the bility of eliminating the conventional engine coolant system and improving normal combustion levels (see Sec. 9.6). Knock results in increased local heat engine efficiency. Since the coolant-side heat transfer is essentially steady during fluxes to regions of the piston, the cylinder head, and liner in contact with the each cycle, a high enough thermal resistance in the wall material can bring the end-gas. Increases to between twice and three times the normal heat flux in the net heat transfer close to zero. However, there is still substantial heat transfer end-gas region have been measured.13, 52 It is thought that the primary knock between the working fluid in the cylinder and the combustion chamber walls. damage to the piston crown in this region is due to the combination of extremely Figure 12-29 illustrates these heat-transfer processes by comparing the mean gas high local pressures and higher material temperatures. temperature to the piston surface temperature for metal and ceramic combustion chamber wall materials. The results come from a thermodynamic simulation of a PROBLEMS turbocompounded diesel engine system operating cycle. From Eq. (12.2) the heat transfer is from the gas when T. > Tw and to the gas when T, < Tw. With the 12.1. If radiation in the combustion chamber is negligible, Eqs. (12.5), (12.6), and (12.7) can ceramic material at about 800 K surface temperature, the net heat transfer is be combined into the following overall equation approximating the time-averaged much reduced compared with the metal case. However, there is substantial heat heat transfer from the engine: transfer to the gas from the ceramic walls during intake (which reduces volu- q = hc .. ( To - T ) metric efficiency) and compression (which increases compression stroke work), and still substantial heat transfer from the gas during combustion and expansion. Derive the expression for he.o . The heat transfer from the hot walls to the incoming charge makes ther- 12.2. Given that the average heat flux through a particular zone in a cast iron liner 1 cm mally insulating materials unattractive for spark-ignition engines. Such heat thick is 0.2 MW/m2, the coolant temperature is 85ºC, and the coolant side heat- transfer would increase the unburned mixture temperature leading to earlier transfer coefficient is 7500 W/m2 . K, find the average surface temperature on the combustion chamber and coolant sides of the liner at that zone. onset of knock (see Sec. 9.6). The variation in ceramic surface temperature in Fig. 12-29 indicates the 12.3. Figure 12-1 gives a schematic of the temperature profile from the gas inside the inherently unsteady nature of the heat-transfer interaction with the wall. During combustion chamber out to the coolant. Draw an equivalent figure showing schematically the temperature profiles at the following points in the engine cycle: (a) combustion and expansion, the thermal energy transferred from the gas to the intake; (b) just prior to combustion; (c) just after combustion; (d) during the exhaust wall is stored in a thin layer of wall material adjacent to the surface. While some stroke. Your sketch should be carefully proportioned. 708 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE HEAT TRANSFER 709 12.4. Using dimensional analysis, compare the relative heat losses of two geometrically (c) Estimate the fraction of the fuel energy that is transferred to the cylinder walls similar SI engines (same bore/stroke ratio, same connecting rod/stroke ratio) oper- during compression and expansion. ating at the same imep and power. Engine A has twice the displacement per cylinder Assume for the gas that the viscosity u = 7 x 10-5 kg/m . s, the thermal conductivity of engine B. Assume that the wall temperature and the gas temperature for both k = 1.5 x 10-1 J/m . s . K, the molecular weight = 28, and the Prandtl number is 0.8. engines are the same. Assume that the combustion chamber is disc-shaped with B = 102 mm, L = 88 mm, 12.5. (a) Using Woschni's correlation, evaluate the percentage increase in heat transfer and re = 9. (The calculations required for this problem are straightforward; do not expected from an engine with a mean piston speed of 10 m/s when the swirl ratio attempt anything elaborate.) is raised from 0 to 5. Do your comparison for the intake process only. The engine bore is 0.15 m and the engine speed is 2000 rev/min. REFERENCES (b) Explain how both the generation of swirl and the change in heat transfer that results affect the volumetric efficiency of an engine. 1. Overbye, V. D ., Bennethum, J. E ., Uyehara, O. A ., and Myers, P. S.: " Unsteady Heat Transfer in Engines," SAE paper 201C, SAE Trans ., vol. 69, pp. 461-494, 1961. 12.6. (a) Explain how you would estimate the thermal boundary-layer thickness on the 2. Howarth, M. H.: The Design of High Speed Diesel Engines, chap. 5, Constable, London, 1966. combustion chamber wall of an internal combustion engine. 3. Khovakh, M. (ed.): Motor Vehicle Engines, chap. 12, Mir Publishers, Moscow, 1971. (b) Using representative data, make a rough estimate of the thickness of the thermal 4. Sitkei, G.: Heat Transfer and Thermal Loading in Internal Combustion Engines, Akademiai Kiado, boundary layer in the combustion chamber of an SI engine just after the com- Budapest, 1974. pletion of combustion and the fraction of the cylinder mass contained within the 5. Burke, C. E ., Nagler, L. H ., Campbell, E. C ., Zierer, W. E ., Welch, H. L ., Lundstrom, L. C ., boundary layer. B ~ 100 mm. Transport properties given in Sec. 4.8. Kosier, T. D ., and McConnell, W. A.: "Where Does All the Power Go," SAE Trans ., vol. 65, pp. 12.7. (a) Using the analysis found in Sec. 12.6.1, calculate the depth below the surface 713-738, 1957. where the amplitude of the temperature oscillations has attenuated to 1 percent 6. Ament, F ., Patterson, D. J ., and Mueller, A.: "Heat Balance Provides Insight into Modern Engine Fuel Utilization," SAE paper 770221, 1977. of the amplitude at the surface. The wall material is aluminum and the four- 7. Novak, J. M ., and Blumberg, P. N.: "Parametric Simulation of Significant Design and Operating stroke cycle engine is operating at 2500 rev/min. For this estimate, consider only Alternatives Affecting the Fuel Economy and Emissions of Spark-Ignited Engines," SAE paper the temperature oscillations which have a frequency equal to the engine firing 780943, SAE Trans ., vol. 87, 1978. frequency (@ = 2RN/nx). 8. Annand, W. J. D.: "Heat Transfer in the Cylinders of Reciprocating Internal Combustion (b) Repeat the calculation for the engine operating at 5000 rev/min. Engines," Proc. Instn Mech. Engrs, vol. 177, no. 36, pp. 973-990, 1963. (c) What is the dependence of the penetration depth on the amplitude of surface 9. Borman, G ., and Nishiwaki, K.: "A Review of Internal Combustion Engine Heat Transfer," Prog. temperature fluctuations ? Energy Combust. Sci ., vol. 13, pp. 1-46, 1987. 10. Taylor, C. F ., and Toong, T. Y.: "Heat Transfer in Internal Combustion Engines," ASME paper 12.8. The instantaneous heat-transfer rate O from the cylinder gases to the combustion 57-HT-17, 1957. chamber walls in a spark-ignition engine may be estimated approximately from the 11. Woschni, G.: “ Universally Applicable Equation for the Instantaneous Heat Transfer Coefficient equation in the Internal Combustion Engine," SAE paper 670931, SAE Trans ., vol. 76, 1967. Q - h,A(T, - T.) 12. Sihling, K ., and Woschni, G.: “ Experimental Investigation of the Instantaneous Heat Transfer in the Cylinder of a High Speed Diesel Engine," SAE paper 790833, 1979. where he is the heat-transfer coefficient, A is the surface area, T. is the average 13. Woschni, G ., and Fieger, J.: "Experimental Investigation of the Heat Transfer at Normal and temperature of the gas in the cylinder, and Tw, is the average wall temperature. The Knocking Combustion in Spark Ignition Engines," MTZ, vol. 43, pp. 63-67, 1982. heat-transfer coefficient can be obtained from the Nusselt, Reynolds, and Prandtl 14. Hohenberg, G. F.: "Advanced Approaches for Heat Transfer Calculations," SAE paper 790825, SAE Trans ., vol. 88, 1979. number relationship: 15. LeFeuvre, T ., Myers, P. S ., and Uyehara, O. A.: "Experimental Instantaneous Heat Fluxes in a Nu = C(Re)"(Pr)" Diesel Engine and Their Correlation," SAE paper 690464, SAE Trans ., vol. 78, 1969. 16. Dent, J. C ., and Sulaiman, S. J.: "Convective and Radiative Heat Transfer in a High Swirl Direct where C = 0.4, m = 0.75, n = 0.4. The characteristic velocity and length scale used in Injection Diesel Engine," SAE paper 770407, SAE Trans ., vol. 86, 1977. this relation are the mean piston speed and the cylinder bore. 17. Krieger, R. B ., and Borman, G. L.: “The Computation of Apparent Heat Release for Internal Assuming appropriate values for the engine geometry and operating conditions Combustion Engines," ASME paper 66-WA/DGP-4, in Proceedings of Diesel Gas Power, ASME, at wide-open throttle with the wall temperature at 400 K, at an engine speed of 2500 1966. rev/min, and using the cylinder pressure versus crank angle curve of Fig. 14-9, calcu- 18. Borgnakke, C ., Arpaci, V. S ., and Tabaczynski, R. J.: "A Model for the Instantaneous Heat Transfer and Turbulence in a Spark Ignition Engine," SAE paper 800287, 1980. late the following: 19. Morel, T ., and Keribar, R.: "A Model for Predicting Spatially and Time Resolved Convective (a) The average temperature of the gas in the cylinder at 0 = - 180º, - 90º, 0º, 20º, Heat Transfer in Bowl-in-Piston Combustion Chambers," SAE paper 850204, 1985. 40º, 90º, 150º. 20. Caton, J. A ., and Heywood, J. B.: "An Experimental and Analytical Study of Heat Transfer in an (b) The instantaneous heat-transfer coefficient he and heat-transfer rate _ from the Engine Exhaust Port," Int. J. Heat Mass Transfer, vol. 24, no. 4, pp. 581-595, 1981. gas to the combustion chamber walls of one cylinder at these crank angles. Plot 21. Malchow, G. L ., Sorenson, S. C ., and Buckius, R. O.: "Heat Transfer in the Straight Section of an these results versus 0. Exhaust Port of a Spark Ignition Engine," SAE paper 790309, 1979. 710 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE HEAT TRANSFER 711 22. Hottel, H. C ., and Sarofim, A. F.: Radiative Transfer, McGraw-Hill, 1967. 46. Watts, P. A ., and Heywood, J. B.: "Simulation Studies of the Effects of Turbocharging and 23. Flynn, P ., Mizuszwa, M ., Uyehara, O. A ., and Myers, P. S.: "Experimental Determination of Reduced Heat Transfer on Spark-Ignition Engine Operation," SAE paper 800289, 1980. Instantaneous Potential Radiant Heat Transfer within an Operating Diesel Engine," SAE paper 47. French, C. C. J ., and Atkins, K. A.: "Thermal Loading of a Petrol Engine," Proc. Instn Mech. 720022, SAE Trans ., vol. 81, 1972. Engrs, vol. 187, 49/73, pp. 561-573, 1973. 24. LeFeuvre, T ., Myers, P. S ., and Uyehara, O. A.: " Experimental Instantaneous Heat Fluxes in a 48. Brock, E. K ., and Glasspoole, A. J.: "The Thermal Loading of Cylinder Heads and Pistons on Diesel Engine and Their Correlation," SAE paper 690464, SAE Trans ., vol. 78, 1969. Medium-Speed Oil Engines," in Proceedings of Symposium on Thermal Loading of Diesel Engines, 25. Lyn, W. T.: "Diesel Combustion Study by Infra Red Emission Spectroscopy," J. Instn Petroleum, vol. 179, part 3C, pp. 3-18, Institution of Mechanical Engineers, 1964-1965. vol. 43, 1957. 49. French, C. C. J ., and Hartles, E. R.: "Engine Temperatures and Heat Flows under High Load 26. Kamimoto, T ., Matsuoka, S ., Matsui, Y ., and Aoyagi, Y.: "The Measurement of Flame Tem- Conditions," in Proceedings of Symposium on Thermal Loading of Diesel Engines, vol. 179, part perature and Thermodynamic Analysis of the Combustion Process in a Direct Injection Diesel 3C, pp. 126-135, Institution of Mechanical Engineers, 1964-1965. Engine," paper C96/75, in Proceedings of Conference on Combustion in Engines, Cranfield, Institu- 50. Kerley, R. V ., and Thurston, K. W.: "The Indicated Performance of Otto-Cycle Engines," SAE tion of Mechanical Engineers, 1975. Trans ., vol. 70, pp. 5-30, 1962. 27. Assanis, D. N ., and Heywood, J. B.: "Development and Use of Computer Simulation of the 51. Taylor, C. F.: The Internal Combustion Engine in Theory and Practice, vol. 1, chap. 8, MIT Press, Turbocompounded Diesel System for Engine Performance and Components Heat Transfer 1966. Studies," SAE paper 860329, 1986. 52. Lee, W ., and Schaefer, H. J.: "Analysis of Local Pressures, Surface Temperatures and Engine 28. Field, M. A ., Gill, D. W ., Morgan, B. B ., and Hawksley, P. G. W.: " Combustion of Pulverised Damages Under Knock Conditions," SAE paper 830508, SAE Trans ., vol. 92, 1983. Coal, British Coal Utility Research Association, Leatherhead, U.K ., 1967. 29. Greeves, G ., and Meehan, J. O.: “Measurements of Instantaneous Soot Concentration in a Diesel Combustion Chamber," paper C88/75, in Proceedings of Conference on Combustion in Engines, Cranfield, Institution of Mechanical Engineers, 1975. 30. Annand, W. J. D.: "Heat Transfer from Flames in Internal Combustion Engines," in N. Afgan and J. M. Beer (eds.), Heat Transfer from Flames, chap. 24, John Wiley, 1974. 31. Annand, W. J. D ., and Ma, T. H.: "Instantaneous Heat Transfer Rates to the Cylinder Head Surface of a Small Compression-Ignition Engine," Proc. Instn Mech. Engrs, vol. 185, no. 72/71, pp. 976-987, 1970-1971. 32. Oguri, T ., and Inaba, S.: "Radiant Heat Transfer in Diesel Engines," SAE paper 720023, SAE Trans ., vol. 81, 1972. 33. Kunitomo, T ., Matsuoka, K ., and Oguri, T.: "Prediction of Radiative Heat Flux in a Diesel Engine," SAE paper 750786, SAE Trans ., vol. 84, 1975. 34. Alkidas, A. C.: "Heat Transfer Characteristics of a Spark-Ignition Engine," Trans. ASME, J. Heat Transfer, vol. 102, pp. 189-193, 1980. 35. Alkidas, A. C ., and Myers, J. P.: "Transient Heat-Flux Measurements in the Combustion Chamber of a Spark-Ignition Engine," Trans. ASME, J. Heat Transfer, vol. 104, pp. 62-67, 1982. 36. Van Gerpen, J. H ., Huang, C .- W ., and Borman, G. L.: "The Effects of Swirl and Injection Param- eters on Diesel Combustion and Heat Transfer," SAE paper 850265, 1965. 37. Whitehouse, N. D.: "Heat Transfer in a Quiescent Chamber Diesel Engine," Proc. Instn Mech. Engrs, vol. 185, no. 72/71, pp. 963-975, 1970-1971. 38. Ramchandani, M ., and Whitehouse, N. D.: "Heat Transfer in a Piston of a Four Stroke Diesel Engine," SAE paper 760007, 1976. 39. Lyford-Pike, E. J ., and Heywood, J. B.: "Thermal Boundary Layer Thickness in the Cylinder of a Spark-Ignition Engine," Int. J. Heat Mass Transfer, vol. 27, no. 10, pp. 1873-1878, 1984. 40. Furuhama, S ., and Suzuki, H.: "Temperature Distribution of Piston Rings and Piston in High Speed Diesel Engine," Bull. JSME, vol. 22, no. 174, pp. 1788-1795, 1979. 41. Finlay, I. C ., Harris, D ., Boam, D. J ., and Parks, B. I.: "Factors Influencing Combustion Chamber Wall Temperatures in a Liquid-Cooled, Automotive, Spark-Ignition Engine," Proc. Instn Mech. Engrs, vol. 199, no. D3, pp. 207-214, 1985. 42. Woschni, G.: “Prediction of Thermal Loading of Supercharged Diesel Engines," SAE paper 790821, 1979. 43. Worthen, R. P ., and Raven, D. G.: " Measurements of Valve Temperatures and Strain in a Firing Engine," SAE paper 860356, 1986. 44. Li, C-H.: "Piston Thermal Deformation and Friction Considerations," SAE paper 820086, 1982. 45. Woschni, G ., and Fieger, J.: "Determination of Local Heat Transfer Coefficients at the Piston of a High Speed Diesel Engine by Evaluation of Measured Temperature Distribution," SAE paper 790834, SAE Trans ., vol. 88, 1979. ENGINE FRICTION AND LUBRICATION 713 Pumping CHAPTER Piston-crank assembly 2.0r Other auxiliaries 3600 rev/min 13 Fuel-injection pump 4 bar bmep Camshaft 1800 rev/min 1.5- 2 bar bmep 1800 rev/min 6 bar bmep ENGINE FRICTION Mean effective pressure, bar 1.0 AND LUBRICATION 0.5 FIGURE 13-1 Comparison of major categories of friction losses: fric- tion mean effective pressure at different loads and speeds for 1.6-liter four-cylinder overhead-cam auto- 8888 3888885 motive spark-ignition (SI) and compression-ignition SI CI SI CI SI CI (CI) engines.1 compression and expansion strokes) and the usable work delivered to the drive shaft, is expended as follows: 1. To draw the fresh mixture through the intake system and into the cylinder, and to expel the burned gases from the cylinder and out of the exhaust system. This is usually called the pumping work. 13.1 BACKGROUND 2. To overcome the resistance to relative motion of all the moving parts of the engine. This includes the friction between the piston rings, piston skirt, and Not all the work transferred to the piston from the gases contained inside the cylinder wall; friction in the wrist pin, big end, crankshaft, and camshaft bear- cylinder-the indicated work-is available at the drive shaft for actual use. That ings; friction in the valve mechanism; friction in the gears, or pulleys and portion of the work transferred which is not available is usually termed friction belts, which drive the camshaft and engine accessories. work. It is dissipated in a variety of ways within the engine and engine auxiliaries. 3. To drive the engine accessories. These can include: the fan, the water pump, The friction work or power is a sufficiently large fraction of the indicated work or the oil pump, the fuel pump, the generator, a secondary air pump for emission power-varying between about 10 percent at full load and 100 percent at idle or control, a power-steering pump, and an air conditioner. no-load -- for the topic to be of great practical importance in engine design. Fric- tion losses affect the maximum brake torque and minimum brake specific fuel consumption directly; often the difference between a good engine design and an All this work is eventually dissipated as heat; the term friction work or power is average engine is the difference in their frictional losses. A large part of the fric- therefore appropriate. Figure 13-1 indicates the relative importance of these com- tion losses appear as heat in the coolant and oil which must be removed in the ponents in typical four-cylinder automotive SI and diesel engines at different radiator and oil cooler system. Thus, friction losses influence the size of the loads and speeds. The magnitudes of the friction from the major items in 1, 2, and coolant systems. A knowledge of friction power is required to relate the com- 3 above are shown for an SI and a CI engine. The absolute value of the total bustion characteristics of an engine-which influence the indicated power-and friction work varies with load, and increases as speed increases. The pumping the useful output-the brake power. work for SI engines is larger than for equivalent CI engines and becomes compa- The friction work, defined as the difference between the work delivered to rable to rubbing friction at light loads as the engine is increasingly throttled. The the piston while the working fluid is contained within the cylinder (i.e ., during the piston and crank assembly contributes the largest friction component. 717 714 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE FRICTION AND LUBRICATION 715 13.2 DEFINITIONS Gross indicated mean effective pressure, imep ,. The work delivered to the piston over the compression and expansion strokes, per cycle per unit displaced volume. The following terminology will be used in discussing engine friction. Net indicated mean effective pressure, imep ,,. The work delivered to the piston over Pumping work W ,. The net work per cycle done by the piston on the in-cylinder the entire four strokes of the cycle, per unit displaced volume. gases during the inlet and exhaust strokes. W, is only defined for four-stroke cycle engines. It is the area (B + C) in Fig. 2-4.+ From the above definitions it follows that Rubbing friction work Wer. The work per cycle dissipated in overcoming the friction imep, = imep, + pmep (13.2a) due to relative motion of adjacent components within the engine. This includes all tfmep = pmep + rfmep + amep (13.2b) the items listed in 2 above. bmep = imep, - tfmep (13.2c) Accessory work Wa. The work per cycle required to drive the engine accessories; e.g ., pumps, fan, generator, etc. Normally, only those accessories essential to engine bmep = imep ,, - rfmep - amep (13.2d) operation are included. [Note that all the quantities in Eqs. (13.2a to d) are positive, except for pmep Total friction work Ww. The total friction work is the sum of these three com- when pi > pe.] ponents, i.e ., That two different definitions of indicated output are in common use Wat = Wp + We + Wa (13.1) follows from two different approaches to determining friction work or power. In the standard engine test code procedures2 friction power is measured in a hot It is convenient to discuss the difference between indicated and brake motoring test: the engine is motored with water and oil temperatures held at the output in terms of mean effective pressure, mep, the work per cycle per unit dis- firing engine values, with the throttle setting unchanged from its firing engine placed volume: position (in an SI engine). This measures (approximately) the sum of pumping, rubbing friction, and auxiliary power. The sum of brake power, and friction mep We Va power determined in this way, is the gross indicated power. Alternatively, when an accurate record of cylinder pressure throughout the whole cycle is available, and power. Power and mep are related by pumping power can be determined directly: the sum of rubbing friction and accessory power is then the difference between the net indicated power- P = mep x V4 x 4 determined from f p dV over the whole cycle-and the brake power. NR For the reasons explained in Sec. 2.4, the gross indicated output is preferred where np (the number of revolutions per cycle) = 1 or 2 for the two-stroke or and used in this text. The distinction is most important for SI engines at part four-stroke cycle, respectively. Hence, from W ,, Wf, Wa, and Wif we can define load where the pumping power and rubbing friction power are comparable in pumping mean effective pressure and power (pmep and Pp), rubbing friction magnitude. For unthrottled engines at low speeds, the distinction becomes less mean effective pressure and power (rfmep and Prf), accessory mean effective pres- important (Fig. 13-1 shows the relative importance of pumping work under both sure and power (amep and Pa), and total friction mean effective pressure and these conditions). power (tfmep and P.f ), respectively. Brake mean effective pressure and power (bmep and Pb), indicated mean 13.3 FRICTION FUNDAMENTALS effective pressure and power (imep and P.), and mechanical efficiency have already been defined in Secs. 2.3, 2.4, 2.5, and 2.7. Note that for four-stroke cycle The friction losses outlined in Sec. 13.1 can be classified into two groups, depend- engines, two definitions of indicated output are in common use. These have been ing on the type of dissipation which occurs. One type is friction between two designated as: metal surfaces in relative motion, with a lubricant in between. The other type is turbulent dissipation. + This definition gives W, > 0 for naturally aspirated engines. For supercharged and turbocharged 13.3.1 Lubricated Friction engines at high load, where p, is usually greater than pe, this definition gives W, < 0. For such engines the sign convention for pumping work is often changed in order to maintain W, as a positive quan- A primary problem in understanding friction between lubricated surfaces in engines is the wide variation in the magnitude of the forces involved. Thus tity. 716 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE FRICTION AND LUBRICATION 717 Oil Journal Bearing 0.1- Load , J = af, + (1 - @)fL Pad Hydrodynamic lubrication Load Oil 2.01|- Mixed lubrication Coefficient of friction Boundary lubrication Supporting oil film Bearing FIGURE 13-3 Supporting oil film Stribeck diagram for journal 0.001- - 4 FIGURE 13-2 bearing: coefficient of friction f Hydrodynamic friction (1 - a)fz Schematic of a lubricated journal and a slider bearing. versus dimensionless duty param- eter uN/a, where u is the lubricant Solid friction ofs dynamic viscosity, N is rotational speed of shaft, o is the loading various regimes of lubrication can occur. Figure 13-2 shows the operating condi- UN/o force per unit area. tions of two common geometries for lubricated parts: a journal and a slider bearing. The different regimes of lubricated friction can be illustrated by means of the Stribeck diagram shown in Fig. 13-3, where the coefficient of friction f (tangential force/normal force) for a journal bearing is plotted against a dimen- chemical ones, which govern the ability of lubricant (or additive) molecules to sionless duty parameter uN/o, where u is the dynamic viscosity of the lubricant, attach themselves to the solid surfaces. Figure 13-4 shows two surfaces under N is the rotational speed of the shaft, and o is the loading force per unit area. For boundary lubrication conditions. Due to the surface asperities, the real contact sliding surfaces the dimensionless duty parameter becomes uU/(ab), where U is area is much less than the apparent contact area. The real contact area A, is the relative velocity of the two surfaces and b is the width of the sliding pad in the equal to the normal load F ,, divided by the yield stress of the material o.: direction of motion. The coefficient of friction can be expressed as A, = f = afs + (1 -a) fz (13.3) The force required to cause tangential motion is the product of the real contact where f; is the metal-to-metal coefficient of dry friction, fz is the hydrodynamic area and the shear strength of the material tm: coefficient of friction, and a is the metal-to-metal contact constant, varying between 0 and 1. As a -> 1, f -+f, and the friction is called boundary, i.e ., close to F . = A, Um solid friction. The lubricating film is reduced to one or a few molecular layers and cannot prevent metal-to-metal contact between surface asperities. As a -+ 0, f -+fL and the friction is called hydrodynamic or viscous or thick film. The lubricant film is sufficiently thick to separate completely the surfaces in relative motion. In Apparent area of contact between these regimes, there is a mixed or partial lubrication regime where the transition from boundary to hydrodynamic lubrication occurs. While Fig. 13-3 Real area of contact applies to journal bearings, this discussion holds for any pair of engine parts in relative motion with lubricant in between. Under boundary lubrication conditions, the friction between two surfaces in relative motion is determined by surface properties as well as by lubricant properties. The important surface properties are roughness, hardness, elasticity, d films FIGURE 13-4 plasticity, shearing strength, thermal conductivity, and wettability with respect to Schematic of two surfaces in relative motion under the lubricant. The important lubricant properties are mainly surface ones of boundary lubrication conditions.3 718 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE FRICTION AND LUBRICATION 719 Thus the coefficient of friction f is 13.3.2 Turbulent Dissipation ‘1 f= F, (13.4) Part of the total friction work is spent in pumping fluids through flow n restrictions. The cylinder gases, cooling water, and oil are pumped through the For dissimilar materials, the properties of the weaker material dominate the fric- engine; the fan pumps air over the engine block. This work is eventually dissi- tion behavior. Since, as shown in Fig. 13-4, the surfaces are covered by oxide pated in turbulent mixing processes. The pressure difference required to pump films and adsorbed lubricant films, the shear strength of the material in Eq. (13.4) these fluids around their flow paths is proportional to pu2, where v is a represen- is effectively the shear strength of the surface film.3 Under boundary lubrication tative fluid velocity. The proportionality constant essentially depends only on conditions, the coefficient of friction is essentially independent of speed. Bound- flow-path geometry. Hence the friction forces associated with fluid pumping will ary lubrication occurs between engine parts during starting and stopping be proportional to N2 (or S', if the piston motion forces the flow). (bearings, pistons, and rings), and during normal running at the piston ring/ cylinder interface at top and bottom center crank positions, between heavily 13.3.3 Total Friction loaded parts, and between slow moving parts such as valve stems and rocker The work per cycle for each component i of the total friction is given by inte- arms, and crankshaft timing gears and chains.4 grating the friction force Ff, : times its displacement dx around the cycle: Hydrodynamic lubrication conditions occur when the shape and relative motion of the sliding surfaces form a liquid film in which there is sufficient pres- sure to keep the surfaces separated. Resistance to motion results from the shear Ws : =| F5 . 1( 0 ) dx forces within the liquid film, and not from the interaction between surface irregu- larities, as was the case under boundary lubrication. The shear stress t in a liquid The friction force components are either independent of speed (boundary film between two surfaces in relative motion is given by friction), proportional to speed (hydrodynamic friction) or to speed squared (turbulent dissipation), or some combination of these. It follows that the total friction work per cycle (and thus the friction mean effective pressure) for a given engine geometry engine will vary with speed according to where u is the fluid viscosity and (dv/dy) the velocity gradient across the film. We (or tfmep) = C1 + C2 N + C3 N2 (13.5) Hence, the friction coefficient (shear stress/normal load stress) in this regime will be proportional to viscosity x speed + loading: i.e ., a straight line on the Stri- Some of the components of hydrodynamic lubrication friction and turbulent dis- beck diagram. Full hydrodynamic lubrication or viscous friction is independent sipation will be dependent on mean piston speed rather than crankshaft rotation- of the material or roughness of the parts, and the only property of the lubricant al speed N. Examples are piston skirt and ring friction, and the pressure losses involved is its viscosity. Hydrodynamic lubrication is present between two con- associated with gas flow through the inlet and exhaust valves. For conventional verging surfaces, moving at relatively high speed in relation to each other and engine geometries, crankshaft rotational speed is usually used to scale the total withstanding a limited load, each time an oil film can be formed. This type of friction data rather than mean piston speed,5,6 though more detailed models must include both these variables. lubrication is encountered in engine bearings, between piston skirt and cylinder liner, and between piston rings and liners for high sliding velocities. Hydrodynamic lubrication breaks down when the thickness of the fluid film 13.4 MEASUREMENT METHODS becomes about the same as the height of the surface asperities. To the viscous A true measurement of friction in a firing engine can only be obtained by sub- friction is then added metal-to-metal solid friction at the peaks of the asperities. tracting the brake power from the indicated power determined from accurate Both hydrodynamic and boundary conditions coexist. The surface texture con- measurements of cylinder pressure throughout the cycle. However, this method is trols this transition from hydrodynamic to mixed lubrication: rougher surfaces not easy to use on multicylinder engines, both because of cylinder-to-cylinder make the transition at lower loads.3 Abrupt load or speed variations or mechani- differences in indicated power and due to the difficulties in obtaining sufficiently cal vibration may cause this transition to occur. This phenomenon occurs in accurate pressure data. As a result, friction is often measured in a motored connecting rod and crankshaft bearings where periodic metal-to-metal contact engine. Friction in a firing and a motored engine are different for the reasons results from sudden breaks in the oil film. The contact area between rings and outlined below. First, the common measurement methods will be described. cylinders is a zone where, due to sudden variations in speed, load, and tem- perature, lubrication is of the mixed type. Intermittent metal-to-metal contacts 1. Measurement of fmep from imep. The gross indicated mean effective pressure is occur as the result of breaks in the oil film. obtained from f p dV over the compression and expansion strokes for a four- 720 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE FRICTION AND LUBRICATION 721 stroke engine, and over the whole cycle for a two-stroke engine. This requires 12 T accurate and in-phase pressure and volume data. Accurate pressure versus 10 crank angle data must be obtained from each cylinder with a pressure trans- ducer and crank angle indicator. Volume versus crank angle values can be calculated. Great care must be exercised if accurate imep data are to be Fuel flow, g/s obtained.7 Both imep, and pmep are obtained from the p-V data. By subtrac- Motored fmep ting the brake mean effective pressure, the combined rubbing friction plus 4 auxiliary requirements, rfmep + amep, are obtained. 2- 2. Direct motoring tests. Direct motoring of the engine, under conditions as close FIGURE 13-5 as possible to firing, is another method used for estimating friction losses. 300 200 100 0 100 200 300 400 500 600 700 Willans line method for determining Engine temperatures should be maintained as close to normal operating tem- -fmep, kPa -- bmep, kPa friction mean effective pressure.5 peratures as possible. This can be done either by heating the water and oil flows or by conducting a "grab" motoring test where the engine is switched rapidly from firing to motored operation. The power required to motor the behind the ring. The resulting boundary friction in this region makes friction engine includes the pumping power. In tests on SI engines at part-load, the in the firing engine higher. Overall, the net effect of lower piston and cylinder throttle setting is left unchanged. "Motoring" tests on a progressively dis- temperatures during motoring is unclear. assembled engine can be used to identify the contribution that each major 3. In motored operation, the exhaust blowdown phase is missing and the gases component of the engine makes to the total friction losses. discharged later in the exhaust stroke have a higher density than under firing 3. Willans line. An approximate equivalent of the direct motoring test for diesel conditions. These effects can result in different pumping work. engines is the Willans line method. A plot of fuel consumption versus brake 4. When motoring, net work is done during the compression and expansion output obtained from engine tests at a fixed speed is extrapolated back to zero strokes because of heat loss from the gas to the walls, and because of gas loss fuel consumption. An example is shown in Fig. 13-5. Generally, the plot has a through blowby. This work is not part of the true total friction work in a slight curve, making accurate extrapolation difficult. Agreement with a firing engine and should not be deducted from the indicated work of the firing motored test result is shown. engine to obtain the brake work; heat losses and blowby are additional energy 4. Morse test. In the Morse test, individual cylinders in a multicylinder engine transfers to the indicated work, friction work, and brake work. are cut out from firing, and the reduction in brake torque is determined while maintaining the same engine speed. The remaining cylinders drive the cylinder Figure 13-6 shows pumping mep, rubbing friction plus auxiliary mep, and cut out. Care must be taken to determine that the action of cutting out one total friction mep for an SI engine over the entire range of throttle positions for cylinder does not significantly disturb the fuel or mixture flow to the others. firing and motoring tests. Firing test data come from imep and bmep measure- ments. The engine was a special four-cylinder, in-line, overhead-valve, 3.26-dm3 Only the first of these four methods has the potential for measuring the true displacement tractor SI engine of rugged design and 12 : 1 compression ratio. The friction of an operating engine. The last three methods measure the power pmep values are closely comparable; the rubbing friction mep values diverge requirements to motor the engine. The motoring losses are different from the firing losses for the following reasons: 200 1. Only the compression pressure and not the firing pressure acts on the piston, tfmep piston rings, and bearings. The lower gas loadings during motoring lower the 150 -A- rubbing friction. fmep, kPa 100 2. Piston and cylinder bore temperatures are lower in motored operation. This rfmep -- A FIGURE 13-6 results in greater viscosity of the lubricant and therefore increased viscous Total friction mean effective pressure (tfmep), friction. In addition, piston-cylinder clearances are greater during motoring 50 . Firing Applep rubbing friction mep (rfmep), and pumping mep CA --- operation which tends to make friction lower. However, in firing operation, A Motoring (pmep) as a function of load for four-cylinder 3.26- the lubrication of the top ring near the top of the stroke is inadequate to dm3 spark-ignition engine with bore = 95.3 mm, 200 400 600 800 1000 stroke = 114 mm, and re = 12, operated at 1600 rev/ maintain normal hydrodynamic lubrication with the higher gas pressures bmep, kPa min. Motoring and firing conditions.8 722 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE FRICTION AND LUBRICATION 723 250 . Engine-motored, complete B, mm L, mm Va, cm3 200 Engine-motored, manifolds removed 3.5 . 80 O Engine-motored, valves removed and camshaft deactivated 73 1468 0.912 0 87.3 -- Engine-fired 82.6 1977 0.945 150 3.0 - 4 58 80 845 1.38 . 90 78 1998 0.865 fmep, kPa 2.5 j- 4 76 71 1288 0.935 100 fmep, atm 2.0 Eq. (13.6) 1.5 OL 800 1000 1200 1400 1600 1800 2000 2200 Speed, rev/min FIGURE 13-7 0.5 Rubbing friction and auxiliary mep for six-cylinder diesel engine under motored and fired conditions. Effect of removing manifolds, valves, and camshaft drive under motored conditions also shown.9 500 1000 2000 3000 4000 5000 6000 Engine speed, rev/min significantly as load increases. However, the firing friction is not necessarily FIGURE 13-8 higher than the motoring test values. Figure 13-7 shows rubbing plus auxiliary Friction mean effective pressure under motored conditions at wide-open throttle for several four- cylinder spark-ignition engines.6 mep for a six-cylinder diesel. The firing data are slightly lower than the motored data for this case.9 1.0 13.5 ENGINE FRICTION DATA 1m 0.8H 13.5.1 SI Engines Figure 13-8 shows total motored friction mep for several four-stroke cycle four- 0.6 cylinder SI engines between 845 and 2000 cm3 displacement, at wide-open throt- pmep tle, as a function of engine speed.6 The data are well correlated by an equation of 0.4 tfme the form of (13.5): tfmep(bar) = 0.97 + 0.15 N 0.2 1000 + 0.05 N (13.6) FIGURE 13-9 Mechanical efficiency n, and ratio of pumping mep to 20 where N is in revolutions per minute. Mean piston speed did not provide as good 40 60 80 100 total friction mep as a function of load for a typical a correlation as rotational speed for this friction data. The importance of avoid- percent load spark-ignition engine at fixed speed.3 ing high engine speeds in the interests of good mechanical efficiency are evident. Under normal automobile engine operating conditions, a reduction in total fric- 200 tion mean effective pressure by about 10 kPa results in about a 2 percent -O-Tc = 12 improvement in fuel consumption.1º --- rc = 7 150H Figure 13-9 shows how mechanical efficiency and the relative importance of rfmep pumping work vary over the load range idle to wide-open throttle under mid- pmep and rfmep, kPa 100 speed operating conditions. The effect of compression ratio on rubbing friction and pumping losses, as a function of load at 1600 rev/min, is shown in Fig. 13-10.8 At the same bmep, both 50 pmep FIGURE 13-10 friction and pumping mep are higher at a higher compression ratio. Friction is -o Pumping mep (pmep) and rubbing friction mep higher because peak cylinder pressures are higher. Pumping is higher at fixed On (rimep) as a function of load for r = 12 and 7, 200 400 600 800 1000 four-cylinder SI engine with B = 95.3 mm and bmep because the engine is throttled more because the efficiency is higher. bmep, kPa L = 114 mm. 1600 rev/min.8 724 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE FRICTION AND LUBRICATION 725 250, 80 Swirl-chamber 70 · Motoring 200 engines A Firing 60 DI engines Motoring mep, kPa 150 FIGURE 13-11 Motored total friction mean effective pressure as a 50 0.4 32 100 -- function of speed for several DI diesels (bores in range 100 to 137 mm) and IDI swirl-chamber pmep, kPa 40 50 diesels (bores in range 100 to 121 mm). Correla- 500 1000 1500 2000 2500 30- tions for r = 15 and L = 142 mm (DI engine) and Speed, rev/min Te = 16 and L = 142 mm (IDI engine).5 20 10 13.5.2 Diesel Engines FIGURE 13-13 0L Pumping mean effective pressure as a function of 2.5 Figure 13-11 shows total friction as determined from motoring tests for both 5 7.5 10 12.5 mean piston speed for several naturally aspirated direct-injection and swirl-chamber indirect-injection four- and six-cylinder CI Mean piston speed, m/s diesel engines.5 engines in the 10 to 14 cm bore range. The higher compression ratio IDI engines lie in the upper half of the scatter band. Correlations for a typical engine of each of the speed.5 This extra heat loss is not part of the difference between indicated type are shown, of the form: and brake output in a firing engine, as noted previously. Motoring mep (kPa) = C1 + 48 - N Pumping mean effective pressure data for a series of naturally aspirated 1000 + 0.453 (13.7 diesels under both firing and motored conditions is shown in Fig. 13-13. The solid line is the term 0.4S?, with S, in meters per second; this is the last term in where N is in revolutions per minute and S, in meters per second. For the direct- the overall motored engine friction correlation [Eq. (13.7)]. injection engine C1 = 75 kPa; for the large swirl chamber IDI engine C1 = 110 kPa. Mean piston speed was found to give a better correlation for the last term in Eq. (13.5) which is mainly pumping mep. Figure 13-12 shows similar results for 13.6 ENGINE FRICTION COMPONENTS small swirl-chamber IDI engines. The same correlation, Eq. (13.7) with C1 == 144 In this section, a more detailed analysis of the major components of engine fric- kPa, is a good fit to the data. tion is presented and, where possible, equations for predicting or scaling the dif- Friction mep increases as engine size decreases. Also, the motoring friction ferent components are developed. loss for the swirl-chamber engines is higher than for direct-injection engines, pri- marily because of heat transfer to the prechamber throat and not due to extra pumping losses which are small. Comparative motoring tests show the increase in 13.6.1 Motored Engine Breakdown Tests motored fmep to be about 27 kPa and essentially independent of speed. This is Motored engine tests, where the engine is disassembled or broken down in stages, typical of a heat-loss effect, whereas a pumping loss would increase as the square can be used to determine the friction associated with each major engine assembly. While this test procedure does not duplicate the combustion forces of actual engine operation, such tests are useful for assessing the relative importance of 500 individual friction components. Figure 13-14 shows results of breakdown tests on a spark-ignition engine and DI diesel engines. These tests show the large contri- 400- Eq. (13.7) bution from the piston assembly (piston, rings, rod, including compression Motoring mep, kPa loading effects), with the valve train, crankshaft bearings, and water and oil 300 pumps all making significant contributions to the total. An approximate break- Data FIGURE 13-12 200 down of rubbing and accessory friction is: piston assembly 50 percent; valve Motored total friction mean effective pressure as a 150 function of speed for smaller IDI swirl-chamber train 25 percent; crankshaft bearings 10 percent; accessories 15 percent. Their 1000 2000 3000 4000 diesel engines (bores in range 73 to 93 mm). Correla- relative importance varies over the speed range, however. In the sections that Engine speed, rev/min tion for r = 21 and L = 95.3 mm.5 follow, total engine friction will be discussed under the following headings: 726 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE FRICTION AND LUBRICATION 727 200 250 & (Pe - Pi) work ZValve flow work 150 Water pump and 200 Compression alternator at stroke no charge Oil pump Exhaust Pressure, kPa 150 stroke Valve train Motored fmep, kPa 100 De 100 E Pistons, rings, pins, and rods 50- (without valves) FIGURE 13-15 50 6 Pumping loop diagram for spark-ignition engine Crankshaft under firing conditions, showing throttling work and seals V max V/pe - Pi) and valve flow work.11 1000 2000 3000 4000 5000 Engine speed, rev/min (a) over the inlet and exhaust strokes. In Fig. 13-15, the firing pumping loop is com- pared with the inlet and exhaust manifold pressures, p; and pe. The work VApi - Pe) measures the effect of restrictions outside the cylinder, in the inlet and 250 | exhaust systems: air filter, carburetor, throttle valve, intake manifold (on the inlet side); exhaust manifold and tail pipe, catalytic converter, and muffler (on the 200- Total motoring loss Pumping exhaust side). The other area, shown as valve flow work, corresponds mainly to pressure losses in the inlet and exhaust valves, and to a lessor extent in the inlet and exhaust ports. As load is reduced in an SI engine, the throttle restriction is :50 Compression, gas load, and increased, the VA(pe - p1) term-called throttling work-will increase, and the Motored fmep, kPa valve gear valve flow work will decrease. The increase in throttling work is much more 100 - rapid than the decrease in valve flow work. Both throttling work and valve flow Motoring without head Rings work increase as speed increases at constant load. Piston and The manifold pressures in naturally aspirated engines can be related to 50- FIGURE 13-14 big end Motored friction mean effective imep through a set of equations developed by Bishop:11, 12 Auxiliaries Crankshaft pressure versus engine speed for 0 engine breakdown tests. (a) Four- 800 1200 1600 2000 cylinder spark-ignition engine.1º (b) imepc = 12.9pa( Pi.a - 0.1) (13.8) Engine speed, rev/min Average results for several four- and Pa (b) six-cylinder DI diesel engines.5 where pi, a is the absolute inlet manifold pressure and pa is the atmospheric pres- sure. (All pressures are in kilopascals.) For SI engines, pumping friction, piston assembly friction, valve train friction, crankshaft bearing friction, and (in Sec. 13.7) accessory power requirements. Pi,o = Pa - imepc - 10 12.9 (13.9) 13.6.2 Pumping Friction (13.10) Engine pumping mep data for SI and CI engines, as a function of speed and load, were given in Sec. 13.5. A more detailed breakdown of pumping work is devel- For diesel engines (naturally aspirated), oped here. Figure 13-15 shows the pumping loop for a firing four-stroke cycle spark-ignition engine. The pumping work per cycle (see Fig. 2-4) is the j p dv Pi ,, = 0 and imepc [in Eq. (13.8)] = 972 kPa (13.11) 728 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE FRICTION AND LUBRICATION 729 800 bmep at p; = 96 kPa 800 bmep at p; = 96 kPa Ring gap Piston crown 700 700 -Molybdenum-filled 600 Groove Compression rings 10% 600- clearance Upper compression ring 500 500 20% 400 bmep, kPa bmep, kPa 400 36% 30% -Ring belt 300 300 Side LRoad 32% clearance -Oil ring Lower compression ring 200 40% load 200 100 100_Road./ 50% load 4% 24% Segment 8 % 16% - Skirt On 0 1000 2000 3000 4000 5000 6000 1000 2000 3000 4000 5000 6000 Chrome-plated Engine speed, rev/min Engine speed, rev/min Expander (a) (b) FIGURE 13-16 Relative importance of (a) throttling friction mep and (b) valve pumping friction mep, for spark- Oil ring assembly ignition engine, as percent of total friction mep on engine load versus speed map.12 FIGURE 13-17 Construction and nomenclature of typical piston and ring assembly.10 Here pi, , and pe, q are the intake and exhaust manifold gauge pressures (both are positive numbers), pa is the atmospheric pressure, and p'e, , is the exhaust gauge 13.6.3 Piston Assembly Friction pressure (all in kilopascals) at 4000 rev/min and full load. The construction and nomenclature of a typical piston and ring assembly is The throttling mep for firing engine operation is then given by shown in Fig. 13-17. The piston skirt is a load-bearing surface which keeps the piston properly aligned within the cylinder bore. The piston lands and skirt carry mep(throttling) = Pi ., + Pe. . (13.12) the side load which is present when the connecting rod is at an angle to the The valve-pumping mep was correlated by cylinder axis. The rings control the lubrication between these surfaces and the liner. Two types of rings-compression and oil rings-perform the following mep (valve pump) = 8.96( mep. o.s N 1.7 2.98 1.28 tasks: (1) seal the clearance between the piston and cylinder to retain gas pressure 1124) 1000 (13.13) and minimize blowby; (2) meter adequate lubricant to the cylinder surface to sustain high thrust and gas force loads at high surface speed and at the same time where control oil consumption to acceptable limits; and (3) control piston temperatures by assisting in heat transfer to the cylinder walls and coolant. Automobile Niv no Div F = " iv m -1 engines normally use three rings, though two-ring designs exist. Larger diesel Va engines may use four rings. Many designs of compression ring are employed,13 the differences between and niv is the number of inlet valves per cylinder, ne the number of cylinders, Div them being in the cross-sectional shapes (and hence relative flexibility) and in the inlet valve head diameter, and Va the displaced volume. For diesel engines, in their use of wear-resistant surface treatments. Top compression rings are usually Eq. (13.13), imepe = 1124 kPa. made of cast iron. The axial profiles are chosen to facilitate hydrodynamic lubri- Figure 13-16 shows the relative importance of the throttling and valve cation. Common shapes are a rectangular cross section with inner and outer pumping losses as a percentage of the total friction mep over the speed and load edges chamfered to prevent sticking in the groove, or with a barrel-shaped range of a typical SI engine. The curves are obtained with the equations given working surface which can accommodate the rotation of the piston which occurs above for a six-cylinder, 9 : 1 compression ratio, 3.3-liter (202 in3) displacement with short piston skirts. Wear-resistant coatings (either a hard chromium-plated engine. The trends of increasing importance of valve pumping with increasing overlay or a molybdenum-filled inlay) are usually applied to the outer ring speed and increasing importance of throttling losses with decreasing load are surface. The second compression ring serves principally to reduce the pressure evident. drop across the top ring. Since the operating environment is less arduous, the 730 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE FRICTION AND LUBRICATION 731 second ring can be made more flexible to give better oil control. The objective is Combustion chamber to compensate for the torsional deflection of the ring under load so that top-edge contact with the cylinder liner is avoided. Top-edge contact tends to pump oil Lubricant toward the combustion chamber, detracting from the performance of the oil Cylinder Piston Pc control ring. Bottom-edge contact provides an oil scraping action on the down- liner' stroke. The oil control ring meters and distributes the oil directed onto the cylin- Po der liner by the crankshaft system, returning excess oil to the crankcase sump. It Pc must exert sufficient pressure against the cylinder, possess suitably shaped wiping Pc U Air BB edges (usually two thin steel rings), and provide adequate oil drainage. Slotted or P composite rings are normally used.14 Piston ring Pir The tension in all the piston rings holds them out against the cylinder wall A FIGURE 13-18 and hence contributes to friction. The gas pressure behind the compression rings Schematic of pressure distribution in the lubri- increases this radial force. The gas pressures behind the second ring are substan- Pir cating oil film and around a compression ring tially lower than behind the first ring. The gas pressures behind the rings are a during expansion stroke. Pressure profile in the Crankcase function of speed and load. An approximate rule for estimating ring friction is oil film indicated by horizontal shading.3 that each compression ring contributes about 7 kPa (1 1b/in2) mep.5 Oil rings, due to their substantially higher ring tension, operate under boundary lubrica- tion; they contribute about twice the friction of each compression ring.15 Navier-Stokes equation for the liquid film motion reduces to a Reynolds equa- The piston assembly is the dominant source of engine rubbing friction. The tion of the form: components that contribute to friction are: compression rings, oil control rings, piston skirt, and piston pin. The forces acting on the piston assembly include: 2 x ax on + 12 # 2 = 6UH 2x (13.14) static ring tension (which depends on ring design and materials); the gas pressure forces (which depend on engine load); the inertia forces (which are related to where h is the local film thickness, u the liquid viscosity, and U the relative component mass and engine speed). The major design factors which influence velocity between the two surfaces. This equation, along with the appropriate piston assembly friction are the following: ring width, ring face profile, ring force balances on the ring, can then be solved for the coupled film and ring tension, ring gap (which governs inter-ring gas pressure), liner temperature, ring- behavior (e.g ., see Ref. 15). land width and clearances, skirt geometry, skirt-bore clearance.3. Measured oil film thicknesses in an operating direct-injection diesel engine Piston assembly friction is dominated by the ring friction. The forces acting are shown in Fig. 13-19. A capacitance technique with electrodes embedded in on a typical compression ring, lubricated by a thin oil film, are shown in Fig. the top compression ring was used to make the measurements.16 At top-center 13-18. The analysis of this hydrodynamic contact is complex because the forces during combustion, the thickness is a minimum (~1 um); it then increases as gas acting on the ring vary with time and slight changes in ring face geometry can loading on the rings decreases and piston velocity increases during the expansion have large effects on the computed results. Cylinder pressure pe normally acts on stroke to a value an order of magnitude higher. Higher engine load results in the top and back of the ring. The inter-ring gas pressure pir (which depends on higher gas loading on the rings. It also results in higher lubricant temperatures cylinder pressure and the geometry of the lands, ring grooves, and ring, especially and lower viscosity, which reduce the film thickness during intake, compression, the ring gap), acts on the oil film and bottom part of the ring. The character of and exhaust. This large change in film thickness over one cycle is the reason the the gas flow into and out of the inter-ring regions and its effect on ring motion ring friction regime changes from boundary lubrication to thick-film hydrody- were discussed in Sec. 8.6. Late in the expansion stroke, pressure reversals can namic lubrication. When the oil film thickness drops below about 1 um, asperity occur which may cause the ring to move to the upper surface of the groove or to contact will begin.+ flutter in between. Ring tension acts to force the ring against the liner. The pres- An analysis of the side thrust between the piston and cylinder wall helps sure in the lubricating oil film is generated as shown by the surface A-B in Fig. explain piston design trends. A force balance on the crank/connecting rod mecha- 13-18 as the ring moves downward. It is believed that the film cavitates between nism of Fig. 2-1 leads to the following. An axial force balance relates the piston B and C so the pressure decreases to a low value and then increases to pe. When the direction of motion is reversed, C-B becomes the pressure-generating surface. Models for the ring and oil film behavior have been developed. For the practical case where the oil film thickness h is much less than the ring width, the The critical film thickness depends on both the cylinder liner and ring surface finish. 732 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE FRICTION AND LUBRICATION 733 10 No load (10 = 10.5 mm2/s) Half load (10 = 9.5 mm2/s) 00 200- F, 6 Oil film thickness, um 4 F3 2 Full load (Vo = 8.5 mm2/s) 0 -360 180 -90 0 90 80 270 360 Full load -100+ Crank angle, deg Half load No load -200- F1 Gas pressure 80 + Full load Fr (no load) 540º 00 180º 360º 5400 Half load -300- 60 F; (full load) Crank angle, deg -No load Gas pressure, atm FIGURE 13-19 40 Measured oil film thickness between top ring and cylinder liner of a DI diesel engine, operated at Motoring F1 Rings alone without pressure 1300 rev/min. Bore = 139.7 mm, stroke == 152.4 mm. vo is estimated oil viscosity. Reprinted by per- F2 Rings alone with compression pressure mission of the Society of Tribologists and Lubrication Engineers (STLE), formerly the American F3 Ring belt body alone Rings alone Firing Society of Lubrication Engineers (ASLE).16 -270 - 180 -90 90 180 270 mass m and acceleration to the net axial force: Crank angle, deg FIGURE 13-20 m dt aDe = - F, cos 0 + .B2 4 PFF , (13.15) Measured frictional force on cylinder liner of 137 mm bore and 135 mm stroke single-cylinder DI diesel engine. 1200 rev/min, coolant temperature 80ºC, cylinder liner inside temperature 97ºC.18 where o is the angle between the cylinder axis and connecting rod, and p is the cylinder gauge pressure. A transverse force balance gives the ring contact area, results in lighter loading (force/area) and promotes hydro- F, = F, sin $ = ( -m- ds LB2 dynamic lubrication. Piston skirt areas have been reduced substantially in recent dt + 4 PFF, tan & (13.16) years to reduce piston mass (which reduces side thrust) and contact area. An additional reduction in side thrust, leading to reduced skirt friction, has been Here F, is the force in the connecting rod (positive when in compression) and F, achieved with the use of an offset wrist-pin. By offsetting the pin axis by 1 to is the friction force on the piston assembly (- when piston is moving toward the 2 mm without changing its vertical location, the crank angle at which the piston crank; + when piston moves away from the crank). dS,/dt is the piston acceler- traverses the bore and "slaps" the other side of the cylinder is advanced so it ation obtained by differentiating the equation for piston velocity [Eq. (2.11)]: occurs before combustion has increased the cylinder gas pressure significantly.17 Direct measurements of the friction force associated with the piston dSe d's 2 = 12NS, cos 0 + R2 cos 20 + sin 4 0 (13.17) assembly have been made. The most common technique involves the use of a dt (R2 - sin? @)3/2 special engine where the axial force on the cylinder liner is measured directly with a load transducer (e.g ., Ref 18). Figure 13-20 shows the friction forces measured in The side thrust F, given by Eq. (13.16) is transmitted to the liner via the such an engine (a DI diesel engine) through the engine's operating cycle. Friction rings and piston skirt. It changes direction as the piston passes through top- and forces are highest just before and after top-center at the end of the compression bottom-center positions. Since the friction force changes sign at these locations stroke. The high values at the start of the expansion stroke under firing condi- and the gas pressure during expansion is greater than during compression, the tions are caused by the piston slap impulse and the high side-thrust force as well side thrust during expansion is greater. as the combustion gas pressure loading on the rings. The piston skirt carries part of this side thrust so it contributes to piston Bishop12 has developed correlations for piston and ring friction in the fol- assembly friction. The large contact area between the skirt and liner, relative to lowing categories: boundary condition friction (primarily between the rings and 734 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE FRICTION AND LUBRICATION 735 the cylinder wall due to ring tension, and gas pressure behind the compression rings) and viscous ring and piston friction. He argued that boundary condition friction was primarily due to breakdown of the oil film between the rings and -Bearing cylinder wall over part of the piston travel. Assuming that the transition to boundary lubrication occurred at a critical speed, he showed that fmep due to boundary friction was proportional to stroke/bore2, i.e.: (fmep)boundary oc loading x 2 (13.18) Journa The ring loading has two components. The component due to ring tension is essentially constant. The component due to gas pressure behind the rings will N Load depend on load. Bishop assumed it to be proportional to inlet manifold pressure. The viscous piston friction-friction between the piston and rings and cylinder hm wall under hydrodynamic lubrication conditions-was correlated by FIGURE 13-21 (fmep)nydrodyn CAP. off .Db Schematic of hydrodynamically lubricated LB 2 (13.19) journal bearing.3 where Ap, eff is the effective area of the piston skirt in contact with the cylinder liner. The relative importance of the boundary lubrication piston and ring friction, and viscous piston and ring friction over the load and speed range, is as follows. The viscous friction component increases in importance with increasing speed. The boundary lubrication friction component increases with increasing load as the cylinder gas pressures increase. 30,000 N 13.6.4 Crankshaft Bearing Friction Crankshaft friction contributions come from journal bearings (connecting rod, Location of minimum main and accessory or balance shaft bearings) and their associated seals. A sche- oil film thickness matic of a journal bearing operating under hydrodynamic lubrication is shown in Fig. 13-21. Large loads can be carried by journal bearings with low energy losses -Load vector under normal operating conditions, due to the complete separation of the two surfaces in relative motion by the lubricant film. Loads on crankshaft journal 1270% 10 904 bearings vary in magnitude and direction because they result primarily from the inertial loads of the piston/connecting rod mechanism and the cylinder gas loads [see Eq. (13.15)]. Typical loads and the resulting journal eccentricity diagram for € = 1.0 a connecting rod bearing are shown in Fig. 13-22. From the journal eccentricity 180 diagram the minimum oil film thickness is determined. This quantity, the Inside surface minimum separation distance between the journal and bearing surfaces, is a criti- of bearing cal bearing design parameter. If the film thickness is too low, asperities will break Polar load diagram for Eccentricity of journal in through the oil film and substantially increase the friction and wear. Journal connecting rod bearing connecting rod bearing bearings are usually designed to provide minimum film thicknesses of about FIGURE 13-22 2 um. Typical engine journal bearing load and eccentricity diagrams.3 736 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE FRICTION AND LUBRICATION 737 The friction force F, in the bearing is given approximately by the product Type 1: Type II: Type III: OHC, direct-acting of the bearing area, the oil viscosity, and the mean velocity gradient in the oil: OHC, end pivot rocker OHC, center pivot rocker F , ~ ( n D , La ) u .D . N) _ 12 HDE L , N (13.20) where D, and L, are the bearing diameter and length, h is the mean radial clear- ance, and N is the shaft rotational speed. A more sophisticated analysis of the friction in a hydrodynamically lubricated bearing yields the relation19 F =- R MD? L, N hew (1 - 82)1/2h + - sin ₲ (13.21) Type IV: Type V: OHC, center pivot rocker with lifter push rod where & is the eccentricity ratio (h - h,)/h and h, is the minimum clearance. The first term closely matches the approximation given in Eq. (13.20). The factor 1/(1 - 82)1/2 and the second term correct for the offset of the journal center rela- tive to the bearing center: W is the bearing load and o the attitude angle. To first order, with hydrodynamic lubrication the friction power does not depend signifi- cantly on the bearing load. If o is the loading per unit projected area of the bearing [W/(L) Du)], then the coefficient of friction f is given by FL XHD? L, N 12D; UN f = W' OL , D . h (13.22) FIGURE 13-23 Different valve train configurations.21 For a given bearing, or series of geometrically similar bearings, the friction coeffi- cient is proportional to uN/G. However, at low values of uN/o the hydrodynamic pressure in an actual bearing will be insufficient to support the shaft load and the The front and rear main bearing seals2º also contribute to crankshaft oil film becomes incomplete. The friction coefficient increases rapidly as mixed assembly friction. At 1500 rev/min they are responsible for about 20 percent of lubrication then occurs. the friction attributable to the crankshaft.1º Bishop12 summed the friction power loss in all crankshaft and con rod journal bearings and divided by the displaced volume per unit time to obtain the following correlation for bearing friction mep (in kilopascals): 13.6.5 Valve Train Friction fmep (bearings) = 41.4(2) (1000) (13.23) The valve train carries high loads over the entire speed range of the engine. Loads acting on the valve train at lower speeds are due primarily to the spring where forces, while at higher speeds the inertia forces of the component masses domi- nate. Valve train designs can be classified by type of configuration, as indicated in K = Pmb Lmb + Do Lib/m + D2, Les B3 (13.24) Fig. 13-23. Large valves and high rated speeds generally increase spring and inertia loads and friction. Friction differences between these systems are difficult to quantify. For example, measurements of valve train friction mean effective In Eq. (13.24), Dmb is the main bearing diameter, Lmb the total main bearing pressure for several of these valve train types showed significant variations ( +30 length - number of cylinders, Drt the rod bearing diameter, L, the rod bearing percent): see Fig. 13-24a.10,21 However, when the data were adjusted to a length, m the number of pistons per rod bearing, Da, the accessory shaft bearing common spring load, Fig. 13-24b, the low-speed friction mep values converged diameter, La the total length of all accessory shaft bearings : number of cylin- and the high-speed fmep differences were reduced.1º ders, and all dimensions are in millimeters. The similarity between engines is such The total valve train friction can be broken down by critical contact that K ~ 0.14 for spark-ignition engines and K ~ 0.29 for diesel engines. regions: camshaft journal bearings, rocker arm/fulcrum and cam/tappet interface. 738 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE FRICTION AND LUBRICATION 739 40 Needle-bearing rocker arm fulcrums 35 30 25 Type IV Type II Type IV Low-tension, low-mass Roller followers (tappets) Valve train drive torque, N.m 300000d Valve train drive mep, kPa 20 Type III valve spring and retainers ooOO Type II 15 Type III 10+ Camshaft Type V of 1000 2000 3000 4000 5000 op L 1000 2000 3000 4000 5000 Engine speed, rev/min Engine speed, rev/min FIGURE 13-25 (a) (b) Low friction valve train. 22 FIGURE 13-24 (a) Total valve train friction mean effective pressure as a function of speed for four engines with different valve configurations (see Fig. 13-23). (b) Valve train friction torque for three of these engines 13.7 ACCESSORY POWER after adjusting to common valve spring load.1º REQUIREMENTS The coolant water pump and oil pump are built-in accessories, essential to engine operation, and are normally considered part of the basic engine.2 A fully The shape of the valve train mep versus speed curve indicates that the predomi- equipped engine usually includes additional accessories-a fan and generator; nant regime of lubrication in the valve train at lower engine speeds is boundary in automobile use it often includes a power-steering pump, an air conditioner, lubrication. The cam/lifter interface usually contributes the largest friction loss and an air pump for emission control. The power delivered by the fully equipped due to the very high loads and small contact areas.22 engine (the net power) is lower than the power delivered by the basic engine due Effective methods of reducing valve train friction are: (1) spring load and to the power requirements of these additional accessories. valve mass reduction; (2) use of tappet roller cam followers; (3) use of rocker arm The friction mean effective pressures associated with driving the water fulcrum needle bearings. One such low-friction valve train design is shown in pump and alternator, and oil pump are shown in Fig. 13-14a. Together they Fig. 13-25.22 The roller cam-followers provide the largest benefit especially at comprise about 20 percent of the total (motored) engine friction. The water pump lower speeds: reductions of order 50 percent in valve train friction can be is typically less than about 7 kPa at 1500 rev/min;1º the oil pump 4 to 10 kPa at achieved. this speed;1º the alternator requires 7 to 10 kPa.23 These numbers vary signifi- Bishop12 developed a correlation for valve train friction from design data cantly with component design details. The generator power depends on the elec- on valve spring loads and valve weights, and experimental data from dynamom- trical load to be met and the generator blower design. A requirement of about eter tests of push rod engines. He shows that two-thirds of the peak is indicated for average generator power.23 The power requirements for a fan, generator, and power-steering pump fmep (valve train) = C[1 - 0.133(N/1000) ]n;, Di,75 B2L (13.25) typical of a 5.7-liter engine are shown in Fig. 13-26. The fan requirements are the largest and with a direct drive increase with the cube of the speed. Alternative couplings such as a viscous drive reduce the fan speed at high engine speed and where niy is the number of inlet valves per cylinder, Diy is the inlet valve head thereby reduce its power significantly. The power-steering pump is only required diameter, and B and L are bore and stroke. This relation does not include cam- to provide high pressures intermittently. Here only the fluid pumping losses are shaft bearing friction, which is included in Eq. (13.23). The functional form of Eq. charged against the engine. (13.25) is an acceptable fit to more modern engine data. Bishop's value for C Air-conditioning is standard on a majority of U.S. cars; additional power is (1.2 x 104 with fmep in kilopascals, N in revolutions per minute, and dimensions required for the air-conditioning compressor. Also, since the compressed refriger- in millimeters) gives valve train fmep values (which exclude camshaft bearing ant is condensed in a second radiator, a larger-than-standard fan is required to losses) about two-thirds the total valve train friction of current production pull additional air through the combined radiator systems. An air pump which engines. This is consistent with the data in Fig. 13-24. pumps air into the engine exhaust ports may be part of an SI engine emission 740 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE FRICTION AND LUBRICATION 741 14 1 Sump 2 Suction pipe 18 3 Lube oil pump 23 17 12- 4 Oil pressure control valve 5 Pressure pipe 10- Solid fan drive. 6 Bypass pipe or alternative 16 7 Cooling coil or, alternatively: 8 Block-type oil cooler 8H Viscous 9 Oil filter -15 Power, kW fan drive 10 Safety valve 6 11 Main oil gallery 12 Main bearing -20 13 Big end bearing 4 14 Camshaft bearing Power steering 15 Tappet (with timing groove pump Generator to pulse-lubricate rocker arm) 8 22 FIGURE 13-26 16 Push rod (hollow, used as 10 Power requirements for engine fan, generator, and rocker arm oil feed pipe) 11 0 1000 2000 3000 4000 5000 6000 power-steering pump typical of 5.7-liter eight- 17 Rocker arm bearing 3 18 Metering plug (to control 21 2 Engine speed, rev/min cylinder engine.23, 24 valve lubrication) 13 12 14 19 Push rod duct (used as cylinder-head-to-crankcase oil return pipe) 21 Piston cooling nozzle control system (see Sec. 11.6). Its power requirements (~1 kW at normal engine 20 Splash hole to lubricate 22 Oil pressure gauge adaptor speeds) must then be added to the accessory friction requirements. timing gears 23 Oil pressure gauge FIGURE 13-27 13.8 LUBRICATION Lubrication system layout for air-cooled DI diesel engine. (Courtesy Klöckner-Humboldt-Deutz AG.) The lubricant and the lubricating system perform the following functions:25 through the tappets and pushrods. For cooling pistons and lubricating cylinders, 1. Reduce the frictional resistance of the engine to a minimum to ensure oil is thrown against the underside of the piston through nozzles connected to the maximum mechanical efficiency. main bearings. Spring-loaded ball valves incorporated in the nozzles interrupt the 2. Protect the engine against wear. jet cooling at low engine speeds to insure that the oil pressure remains above a 3. Contribute to cooling the piston and regions of the engine where friction work safe level. The gears of the main timing train are splash-lubricated. The oil is is dissipated. returned from the injection pump and rocker chamber cover to the sump. 4. Remove all injurious impurities from lubricated regions. 5. Hold gas and oil leakage (especially in the ring region) at an acceptable minimum level. 13.8.2 Lubricant Requirements Table 13.1 lists the qualities required of engine oils to perform the main lubrica- tion system functions. These qualities can be summarized under the following 13.8.1 Lubrication System headings. 25 The principle moving parts of an engine are positively lubricated by introducing a supply of oil from a pressurized system. An example of a lubrication system (for OXIDATION STABILITY. The temperature of the oil and engine parts it con- an air-cooled diesel engine) is shown in Fig. 13-27. The oil pump draws oil from tacts, the presence of oxygen, the nature of the metal surfaces and debris, and the the engine sump and delivers it through a control valve to the oil cooler. The oil products of the fuel combustion, all influence the oxidation of the hydrocarbon then passes through the filter to the main oil gallery. From the main oil gallery it components in lubricating oil. High temperatures are the primary factor, and the is branched to the main, the big end, and the camshaft bearings. Oil is also top piston ring groove and the crankcase are the critical regions. The tem- ducted to the injection pump. Through a passage in the camshaft bearing the oil perature of the top ring groove can easily reach 250ºC. The lubricating oil when flows to the tappet bridges. As the oil passages of tappets and tappet bridges line subject to these conditions must not, through oxidation, contribute to deposit up during tappet motion, rocker arms and valve stems are pulse-lubricated formation, even after long periods of running. These deposits would eventually TABLE 13.1 ENGINE FRICTION AND LUBRICATION 743 Functions and qualities required of engine oils Main functions lead to ring sticking which results in excessive blowby. At high temperatures, required Where and when Qualities required deposits are related to the oxidized fraction of the oil. The oil temperature in the crankcase is 120 to 130ºC, or higher. Oil main- Reduce During cold-starting Low enough viscosity to provide good pumping tained at this temperature should neither form any acid products capable of frictional and avoid undue cranking resistance resistance Between con-rod/ Minimum viscosity without risk of metal-to-metal attacking the bearing alloys nor form insoluble products which form deposits. crankshaft bearings, contact under the varying conditions of temperature, Good-quality mineral oils cannot withstand these temperatures, so antioxidant and journals speed, and load and anticorrosive additives are used to control these problems. While anti- Between pistons, rings, Sufficiently high viscosity at high temperatures; oxidants help to reduce deposit formation, detergent/dispersant additives are and cylinders good lubrication property outside the hydrodynamic required to maintain any insoluble materials formed through oxidation in sus- condition, especially at top-center Antiseizure properties, especially during the pension. run-in period Protect During shut-down or Must protect metallic surfaces against corrosive DETERGENCY/DISPERSION. Except for deposits formed in the combustion against when running at low action of fuel decomposition products (water, SO2, chamber, deposits in the oil are controlled by its detergency. The amount of corrosion temperature HBr, HCI, etc.) deposits formed depends on the fuel used, the quality of combustion, the tem- and wear Must resist degradation (resist oxidation, have good perature of the lubricating oil and coolant, and on the effectiveness of gas sealing thermal stability) In normal running Must counteract action of fuel and lubricant at the piston rings. The detergency property is given to straight mineral oils by decomposition products at high temperatures, additives; the function of the detergent additive is to reduce the amount of depos- especially on non-ferrous metals its formed and make their removal easier. By intervention in the friction mechanism must At low temperatures, deposits are mainly byproducts of fuel combustion, reduce the consequences of unavoidable metal-to- and the detergency function is to keep them in suspension or solution in the oil. metal contact Must resist deposit formations which would affect At high temperatures, deposits come from the oxidized fractions of the oil. The lubrication (detergency or dispersive action) detergency function here is both to keep these products in suspension and to Must contribute to the elimination of dust and other inhibit the reactions that lead to the formation of varnishes and lacquers. In contaminants (dispersive action) addition, in diesel engines, the detergency helps in neutralizing the acidic reaction products from the sulfur compounds in the fuel. Assist In the ring zone, Must have sufficient viscosity at high temperatures sealing especially at TC and low volatility Must limit ring and liner wear WEAR REDUCTION. Wear is due to the individual and combined effects of cor- Must not contribute to formation of deposits in ring grooves and must prevent such formation rosion, adhesion (i.e ., metal-to-metal contact), and abrasion. Corrosive attack by acidic products of combustion is one of the chief causes Contribute Chiefly of pistons, Must have good thermal stability and oxidation of cylinder and ring wear. The effect is worst at low cylinder wall temperatures. to cooling rings, and con-rod resistance In diesel engines, the sulfur in the fuel increases the corrosive wear. Corrosive bearings Must have low volatility Viscosity must not be too high wear is effectively prevented by the use of detergent oils which neutralize the corrosive acids as they form, and by designing the cooling system to give appro- Facilitate During oil drains to Must be able to maintain in fine suspension all priate metal temperatures. the elimination eliminate atmospheric solid material (dispersivity) whatever the Adhesive wear affects certain parts of the engine. In the upper cylinder, of undesirable dust, soot from diesel temperature and physical and chemical conditions products engines, Pb salts, (water) metal-to-metal contact between piston, rings, and cylinder walls takes place each wear debris, organic Must be able to solubilize certain organic compounds, time the engine is started (most significant during cold-starts) because there is products from burned particularly heavy oxidation products insufficient oil in the top portion of the engine. Oils with antiwear additives and fuel and lubricants, low viscosity at low temperatures provide a partial remedy. Adhesive wear also and other contaminants occurs on components such as cams, tappets, drive gears, rocker arm ends, and which promote valve stems. deposits or accelerate wear Abrasion results from the presence of atmospheric dust, and metallic debris from corrosive and adhesive wear, in the lubricating oil. Efficient air filtration is Source: From Schilling.25 therefore most important (see Ref. 26 for a discussion of air filters). Elimination of 742 744 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE FRICTION AND LUBRICATION 745 abrasive particle impurities from the oil system by filtration and periodic oil 1,000,000 change is essential. 100,000 a 20,000- SAE 5000 VISCOSITY. For low resistance to cranking and ease of starting, and rapid dis- 2000 NOW tribution of the oil while the engine is cold, a low oil viscosity at low ambient 1000 400 5W B temperatures is required. When the engine (and oil) is fully warmed up, viscosity 200 in the proper range is important for adequate sealing of the piston, acceptable oil 100 C Kinematic viscosity, centistokes or mm2/s consumption, and low friction losses. The viscosity of the oil at both low and 50 A 30 normal engine temperatures (a spread of some 200 K) is, therefore, important. 20 15 The viscosity of lubricating oils decreases with increasing temperature. The pour 10 A = SAE 10W point, viscosity, and viscosity index are used to characterize the behavior of a 8 B = SAE 50 6 lubricating oil for these aspects of engine operation. C = SAE 10W/30 The pour point is determined by cooling a sample of oil in a test jar until, 4 when the jar is rotated from the vertical to the horizontal, no perceptible move- 3- 0ºF ment of the oil will occur within 5 s; 5ºF above this temperature is the pour 100ºF 210ºF FIGURE 13-28 Viscosity versus temperature curves point. -34 -18 0 20 40 60 80 100 a b illustrating SAE lubricating oil clas- The viscosity of lubricating oils is determined by measuring the time Temperature, ºC sification.25 required for a specified volume of oil to flow through a capillary tube or orifice, contained in a constant temperature water bath. The kinematic viscosity, v (v = u/p), is determined by this method. Use of a Saybolt tube with an orifice of they are based on viscosity at 99ºC (210ºF). Multigrade oils (for example, specified diameter is the standard U.S. measurement practice. The viscosity is 10W-40) satisfy service requirements at low as well as high temperatures in terms then given by the time t (in seconds) required to flow 60 cm3 of oil, and is of the SAE classification. The first number indicates the viscosity at - 18ºC; the expressed as Saybolt universal seconds, SUS. Approximate conversion to centi- second number at 99ºC. Examples are shown in Fig. 13-28. Multigrade oils have stokes (1 centistoke = 10-6 m2/s) can be obtained via a higher viscosity index than single-grade oils, which make them more attractive for engine use. b v = at - - PROBLEMS where for 115 > t > 34 s, a = 0.224 and b = 185; for 215 > t > 115 s, a = 0.223 and b = 155; and for t > 215 s, a = 0.2158 and b = 0.27 13.1. (a) Show how friction mean effective pressure for a four-stroke cycle engine can be The viscosity of lubricating oils decreases with increasing temperature. obtained from the brake power P ,, engine speed N, displaced volume Va, and Since engine oils must operate over a range of temperatures, a measure of the Sp dV over the compression and expansion strokes ( == We.i.). rate of decrease is important. The viscosity index, an empirical number indicating (b) How is pumping mean effective pressure related to , p dV over the compression the effect of temperature changes on viscosity, is used for this purpose;28 a low and expansion strokes and f p dV over the full four-stroke cycle? viscosity index indicates a relatively large change of viscosity with temperature. (c) Find the brake power, total friction power, total friction imep, and pumping imep To increase the viscosity index, lubricating oils incorporate additives called for a four-stroke cycle SI engine operating at 1800 rev/min with a measured brake torque of 32 N . m, a gross imep of 933 kPa, and a net imep of 922 kPa. "viscosity-index improvers." These are high molecular weight compounds V. = 0.496 dm3. (molecular weight ~ 103 to 104) whose primary function is to reduce the viscosity 13.2. Three categories of friction are described in Sec. 13.3: boundary friction, hydrody- variation with temperature. namic (or fully lubricated) friction, and turbulent dissipation. By means of Eq. (13.6), The lubricating oil classification used most extensively is the SAE classi- estimate the relative proportion of total friction work per cycle in each category for a fication.29 It depends solely on the viscosity of the oil. The seven different classi- four-cylinder automobile spark-ignition engine operating at 3000 rev/min. fication numbers 5W, 10W, 20W, 20, 30, 40, and 50 correspond to viscosity 13.3. For four-stroke cycle naturally aspirated multicylinder spark-ignition and diesel ranges; increasing numbers correspond to increasing viscosity, as shown in automobile engines at full load and one-third full load, at mid speed (2000 rev/min), Fig. 13-28. SAE numbers followed by W (abbreviation for winter) refer to oils for give approximate estimates of the percentages of total friction mep in these three use in cold climates, and viscosity is determined at - 18ºC (0ºF). SAE numbers categories: pumping mep, rubbing friction mep, and accessory friction mep. State without W are applied to engine oils commonly used under warmer conditions; explicitly how you develop these estimates and what you include as accessories. 746 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE FRICTION AND LUBRICATION 747 13.4. All of the friction measurement procedures except the difference between brake and 3. Rosenberg, R. C.: "General Friction Considerations for Engine Design," SAE paper 821576, 1982. gross indicated power or mep measured directly assume that motored engine friction 4. Schilling, A.: Automobile Engine Lubrication, Scientific Publication, 1972. and firing engine friction are closely comparable. This is not an accurate assumption 5. Millington, B. W ., and Hartles, E. R.: "Frictional Losses in Diesel Engines," paper 680590, SAE for the pumping component. Summarize the differences between the gas exchange Trans ., vol. 77, 1968. processes under motoring and firing conditions for a spark-ignition engine at a fixed 6. Barnes-Moss, H. W.: "A Designer's Viewpoint," in Passenger Car Engines, Conference Pro- part-load throttle setting that will result in the pumping work being significantly ceedings, pp. 133-147, Institution of Mechanical Engineers, London, 1975. different under these two conditions. 7. Lancaster, D. R ., Krieger, R. B ., and Lienesch, J. H.: "Measurement and Analysis of Engine 13.5. On separate accurately proportioned sketches of the piston, cylinder, connecting rod, Pressure Data," paper 750026, SAE Trans ., vol. 84, 1975. and crank mechanism (similar to Fig. 2-1), during the intake stroke (at 120º ATC), 8. Gish, R. E ., Mccullough, J. D ., Retzloff, J. B ., and Mueller, H. T.: "Determination of True Engine compression stroke (at 60º BTC), expansion stroke (at 60º ATC), and exhaust stroke Friction," SAE Trans ., vol. 66, pp. 649-661, 1958. 9. Brown, W. L.: "The Caterpillar IMEP Meter and Engine Friction," paper 730150, SAE Trans ., (at 120º BTC), draw an arrow for each of the forces acting on the piston (pressure vol. 82, 1973. forces, force from connecting rod, friction force, inertia force). Mark clearly the posi- 10. Kovach, J. T ., Tsakiris, E. A ., and Wong, L. T.: "Engine Friction Reduction for Improved Fuel tive direction of each force. Express each force in terms of cylinder pressure pe, Economy," SAE paper 820085, 1982. crankcase pressure pec , friction force Ff, piston area A„, effective piston (and part of 11. Cleveland, A. E ., and Bishop, I. N.: "Several Possible Paths to Improved Part-Load Economy of connecting rod) mass m ,, piston acceleration a, connecting rod force Fer. Spark-Ignition Engines," SAE paper 150A, 1960. 13.6. (a) For the DI diesel engine for the friction force data in Fig. 13-20, estimate the 12. Bishop, I. N.: "Effect of Design Variables on Friction and Economy," SAE Trans ., vol. 73, pp. maximum pressure force on the piston (under full-load conditions) and the 334-358, 1965. approximate magnitude of the inertia force [mass of piston plus part of the con- 13. Piston Rings, Mobil Technical Bulletin. 14. Nunney, M. J.: The Automotive Engine, Newnes-Butterworths, London, 1974. necting rod (7 kg) x S, x (N/4)]. Compare these forces with the piston friction 15. Furuhama, S ., Takiguchi, M ., and Tomizawa, K.: " Effect of Piston and Piston Ring Designs on force at time of peak pressure. the Piston Friction Forces in Diesel Engines," SAE paper 810977, SAE Trans ., vol. 90, 1981. (b) Figure 13-6 shows the variation in friction force acting on the piston of a DI 16. Furuhama, S ., Ashi, C ., and Hiruma, M.: " Measurement of Piston Ring Oil Film Thickness in an diesel engine under no-load and full-load firing conditions. Carefully sketch the Operating Engine," ASLE preprint 82-LC-6C-1, 1982. shape (indicating direction and rough magnitude) of the cylinder pressure force 17. McGeehan, J. A.: " A Literature Review of the Effects of Piston and Ring Friction and Lubricat- on the piston, the piston velocity, and the piston acceleration, as functions of ing Oil Viscosity on Fuel Economy," SAE paper 780673, SAE Trans ., vol. 87, 1978. crank angle on the same graph as these friction forces. Use these graphs to 18. Furuhama, S ., and Takiguchi, M.: "Measurement of Piston Frictional Force in Actual Operating explain the variation of piston friction force throughout the four strokes of the Diesel Engine," SAE paper 790855, SAE Trans ., vol. 88, 1979. cycle. 19. Cameron, A.: The Principles of Lubrication, Wiley, New York, 1966. 20. McGeehan, J. A.: " A Survey of the Mechanical Design Factors Affecting Engine Oil Consump- 13.7. (a) Show by dimensional analysis of the variables that govern the friction in a tion," SAE paper 790864, SAE Trans ., vol. 88, 1979. journal bearing (friction force F, oil viscosity u, bearing diameter Db, length L ,, 21. Armstrong, W. B ., and Buuck, B. A.: "Valve Gear Energy Consumption: Effect of Design and mean clearance h, shaft rotational speed N) that Operational Parameters," SAE paper 810787, 1981. 22. Staron, J. T ., and Willermet, P. A.: " An Analysis of Valve Train Friction in Terms of Lubrication FS Principles," SAE paper 830165, SAE Trans ., vol. 92, 1983. HD ? N 23. Burke, C. E ., Nagler, L. H ., Campbell, E. C ., Lundstrom, L. C ., Zierer, W. E ., Welch, H. L ., Kosier, T. D ., and McConnell, W. A.: " Where Does All the Power Go," SAE Trans ., vol. 65, pp. What additional physical assumptions are then required to obtain an equation of 713-737, 1957. the form of (13.20)? 24. Dean, J. W ., and Casebeer, H. M.: “Chrysler 340 Cu In. V-8 Engine Produces 275 HP at 5000 (b) Under what conditions can Eq. (13.23), an empirically developed relation for RPM," SAE paper 680019, 1968. engine bearing fmep, be obtained from Eq. (13.20)? 25. Schilling, A.: Motor Oils and Engine Lubrication, Scientific Publications, 1968. 13.8. Explain whether each of the following components of engine friction would be 26. Annand, W. J ., and Roe, G. E.: Gas Flow in the Internal Combustion Engine, Haessner Publishing, expected to depend on (1) crankshaft rotational speed N, (2) mean piston speed S ,, 1974. (3) or both of these variables. Crankshaft journal bearings, connecting rod bearings, 27. ASTM Standards, Part 17, Petroleum Products. 28. ASTM D2270-64. valve train, piston rings, piston skirt, water pump, fan, valve flow loss (resistance to 29. SAE J300a. flow through the inlet and exhaust valves). REFERENCES 1. Ball, W. F ., Jackson, N. S ., Pilley, A. D ., and Porter, B. C.: "The Friction of a 1.6 Litre Automo- tive Engine-Gasoline and Diesel," SAE paper 860418, 1986. 2. SAE Test Code J816b. MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 749 trends and tradeoffs, and, if the model is sufficiently accurate, to optimize CHAPTER design and control; 4. Providing a rational basis for design innovation. 14 Each of these contributions is valuable. Whether a model is ready to pass from one stage to the next depends on the accuracy with which it represents the actual process, the extent to which it has been tested and validated, and the time and effort required to use the model for extensive sets of calculations and to MODELING interpret the results. REAL This chapter reviews the types of models and their primary components that are being developed and used to describe engine operating and emissions ENGINE characteristics. These models describe the thermodynamic, fluid-flow, heat- FLOW AND transfer, combustion, and pollutant-formation phenomena that govern these per- formance aspects of engines. Many of the building blocks for these models have COMBUSTION been described in the previous chapters. The purpose here is to show how fluid PROCESSES dynamics, heat-transfer, thermodynamics, and kinetics fundamentals can be com- bined at various levels of sophistication and complexity to predict, with varying degrees of completeness, internal combustion engine combustion and emissions processes, and hence engine operating characteristics. For the processes that govern engine performance and emissions, two basic types of models have been developed. These can be categorized as thermodynamic or fluid dynamic in nature, depending on whether the equations which give the model its predominant structure are based on energy conservation or on a full analysis of the fluid motion. Other labels given to thermodynamic energy- conservation-based models are: zero-dimensional (since in the absence of any flow modeling, geometric features of the fluid motion cannot be predicted), phenomenological (since additional detail beyond the energy conservation equa- 14.1 PURPOSE AND CLASSIFICATION tions is added for each phenomenon in turn), and quasi-dimensional (where spe- OF MODELS cific geometric features, e.g ., the spark-ignition engine flame or the diesel fuel spray shapes, are added to the basic thermodynamic approach). Fluid-dynamic- In engineering, modeling a process has come to mean developing and using the based models are often called multidimensional models due to their inherent appropriate combination of assumptions and equations that permit critical fea- ability to provide detailed geometric information on the flow field based on solu- tures of the process to be analyzed. The modeling of engine processes continues tion of the governing flow equations. to develop as our basic understanding of the physics and chemistry of the phe- Some general observations about models of engine processes provide a nomena of interest steadily expands and as the capability of computers to solve context for the details that follow. The processes themselves are extremely complex equations continues to increase. Modeling activities can make major complex. While much is known about these processes, they are not adequately contributions to engine engineering at different levels of generality or detail, cor- understood at a fundamental level. At present, it is not possible to construct responding to different stages of model development, by: models that predict engine operation from the basic governing equations alone. Thus the objectives of any model development effort should be clearly defined, 1. Developing a more complete understanding of the process under study from and the structure and detailed content of the model should be appropriate to the discipline of formulating the model; these objectives. It is impractical to construct models that attempt to describe all 2. Identifying key controlling variables to provide guidelines for more rational important aspects of engine operation: more limited objectives are appropriate. and therefore less costly experimental development efforts; Due to this complexity of engine processes and our inadequate understand- 3. Predicting engine behavior over a wide range of design and operating vari- ing at a fundamental level, most engine models are incomplete. Empirical rela- ables to screen concepts prior to major hardware programs, to determine tions and ad hoc approximations are often needed to bridge gaps in our 748 750 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 751 understanding of critical phenomena. Hence, since models will continue to fuel elements in the combustion products) in the open system : develop greater completeness and generality, the emphasis in this chapter is on the basic relationships used in engine process models rather than the current im , = = (mf ) = Ems.s = Em, f; 14.2) status of these models. Finally, an important issue in any overall engine model is balance in com- plexity and detail amongst the process submodels. A model is no more accurate Differentiation of Eq. (14.2) leads to an equation for the rate of change of fuel than its weakest link. Thus critical phenomena should be described at compara- fraction: ble levels of sophistication. (14.3) 14.2 GOVERNING EQUATIONS FOR OPEN THERMODYNAMIC SYSTEM The fuel/air equivalence ratio o is related to f via o = f/[(F/A),(1 - f)]. Hence the rate of change of equivalence ratio of the material in the open system is It is often required to model a region of the engine as an open thermodynamic system. Examples are the cylinder volume and the intake and exhaust manifolds 1 (or portions of these volumes). Such a model is appropriate when the gas inside (F/A), (1 - f)2 (14.4) the open system boundary can be assumed uniform in composition and state at each point in time, and when that state and composition vary with time due to heat transfer, work transfer and mass flow across the boundary, and boundary displacement. Such an open system is illustrated in Fig. 14-1. The important 14.2.2 Conservation of Energy equations are mass and energy conservation. These equations for an open system, The first law of thermodynamics for the open system in Fig. 14-1 can be written: with time or crank angle as the independent variable, are the building blocks for thermodynamic-based models. E= Qw - W+ Emjh, (14.5) 14.2.1 Conservation of Mass w is the total heat-transfer rate into the system, across the system boundary, and equals the sum of the heat-transfer rates across each part of the boundary, The rate of change of the total mass m of an open system is equal to the sum of E Ow,1. W is the work-transfer rate out of the system across the boundary; the mass flows into and out of the system: where the piston is displaced, the work-transfer rate equals pV. Because all ener- gies and enthalpies are expressed relative to the same datum (see Sec. 4.5.3), it is (14.1) not necessary to include the heat released by combustion in Eq. (14-5); this is already accounted for in the energy and enthalpy terms. Mass flows into the system are taken as positive; mass flows out are taken as The goal is to define the rate of change of state of the open system in terms negative. For conservation of the fuel chemical elements, it is convenient to use of Tand p. Two approaches are commonly used, depending on whether the ther- the fuel fraction f, which is defined as m,/m, where m, denotes the mass of fuel (or modynamic property routines provide values for internal energy u or enthalpy h. Thus È in Eq. (14.5) can be expressed as ew E = - (mu) or E = " ( mb ) - - di (PV) (14.6a, b) mihi im3h3 It is assumed that the system can be characterized by T, p, and ¢; thus u = u(T, p, o) h = h( T , p, 4) P = P(T, P, $) (14.7) and the rate of change of u, h, and p can be written in the form nah W FIGURE 14-1 a = ad (14.8) Open thermodynamic system. 752 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 753 where a is u, h, or p. Using the ideal gas law in its two forms, p = pRT and of the combustion chamber is treated as one system. For the two-zone model pV = mRT, and Eq. (14.8) for p, an equation for p can be derived: used for spark-ignition engine simulations, the unburned mixture zone and the P V burned mixture zone are each treated as separate open systems, with volumes V, P == ap/ap V POT pad m ) (14.9) and V ,, respectively, where V. + V = V. If a thermal boundary-layer region is included (see Sec. 12.6.5) an additional open system must be defined. Returning now to the energy conservation equation, expressing E in terms of u or h, and ù or h in terms of partial derivatives with respect to T, p, and , 14.3 INTAKE AND EXHAUST FLOW and substituting for p with Eq. (14.9), one can obtain equations for T: MODELS V OR $ Ou Ou CD Ou 14.3.1 Background Dap 0 R) 8 / 2+ + + DT op) (14.10) The behavior of the intake and exhaust systems are important because these where systems govern the air flow into the engine's cylinders. Inducting the maximum air flow at full load at any given speed and retaining that mass within the B = - RT - - + - ( en + Im, h , - mu ) engine's cylinders is a primary design goal. The higher the air flow, the larger the amount of fuel that can be burned and the greater the power produced. The T OR important parameters are volumetric efficiency (for four-stroke cycle engines) or C=1+ ROT scavenging and trapping efficiencies (for two-stroke cycle engines), along with equal air flows to each engine cylinder (see Secs. 6.2, 6.6, and 7.6.2). D = 1 - POR The objectives of any manifold model have an important bearing on its R ap complexity and structure. If the goal is to provide the input or boundary condi- (see Ref. 1, for example). From Ref. 2, tions to a detailed model of in-cylinder processes, then sophisticated intake and exhaust system models are not necessarily required. If the manifold flows are the ( 14.11 ) primary focus, then models that adequately describe the unsteady gas-flow phe- nomena which occur are normally required. Then simple models for the in- where cylinder phenomena usually suffice to connect the intake and exhaust processes. The valves and ports, which together provide the major restriction to the intake A' = + @p/@T (1 oh and exhaust flow, largely decouple the manifolds from the cylinders. ap/Op P ap Three types of models for calculating details of intake and exhaust flows 1 - p(@h/ap) have been developed and used: B' = ap/Op 1. Quasi-steady models for flows through the restrictions which the valve and oh @plak (1. ah port (and other components) provide C' = 26 + 2p/app ap 2. Filling and emptying models, which account for the finite volume of critical manifold components Equations (14.1), (14.3), (14.4), (14.9), and (14.10) or (14.11) can now be solved to obtain the state of the open system as a function of time. V is obtained from Eq. 3. Gas dynamic models which describe the spatial variations in flow and pressure (2.6), and the thermodynamic properties and their derivatives from the models throughout the manifolds described in Chap. 4. Often, for specific applications, the above equations can be simplified sub- Each of these types of models can be useful for analyzing engine behavior. stantially. For the intake and exhaust systems (or sections of these systems such The appropriate choice depends on objectives, and the time and effort available. as the manifold or plenum, etc.), Vis zero and effects of dissociation (the terms Each will now be reviewed. Ou/Op, 0h/Op, and OR/Op) can usually be neglected. For the cylinder during com- pression, dissociation can usually be neglected, also. Application of these equa- 14.3.2 Quasi-Steady Flow Models tions during combustion must be related to the combustion model used. For the Here the manifolds are considered as a series of interconnected components, single-zone model often used in diesel engine simulations (see Sec. 10.4) the whole which each constitute a significant flow restriction: e.g ., air cleaner, throttle, port, 754 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 755 and valve for the intake system. The flow restriction each of these components Experiment represents is defined by their geometry and discharge coefficient, usually deter- EPO IN IPO IPC! (compact EPC manifold) mined empirically under steady-state conditions. The gas flow rate through each 1.5 component is computed using steady one-dimensional flow equations [see App. Experiment FIGURE 14-2 (long pipe) C, Eqs. (C.8) and (C.9)]: the actual flow is assumed to be quasi steady. These Comparison of intake and Pi = 1.18 Pe exhaust manifold pressures, P; components are connected by the gas flow passing through them and the pres- BC Prediction and p ., predicted by filling and sure ratios across them; mass accumulation between components is neglected. emptying model, with experimen- Pressure, atm abs 90 180 Quasi-steady models are often used to calculate the flow into and out of the 270 360 tal data. Single-cylinder two- cylinder through the inlet and exhaust valves (see Secs. 6.3 and 6.5 and Fig. 6-20). Experiment stroke loop-scavenged direct- EPO! IPO IPC EPC (compact .5 injection diesel engine. Different If the pressure variation with time upstream of the valve is known or is small, as manifold) ratios of exhaust system volume usually occurs with large plenums and short manifold pipe lengths, such methods = 5 --- Experiment V, to displaced volume V ., and are accurate enough to be useful. This approach has been used extensively with Va (long pipe) exhaust manifold shapes.4 engine cycle simulations which predict engine performance characteristics from a 1.0 Pe BC Prediction EPO Exhaust port opens thermodynamics-based analysis, to calculate the mass flow rates into and out of IPO Inlet port opens 90 the cylinder (see Sec. 14.4). Such methods are not able to predict the variation of 180 270 360 IPC Inlet port closes Crank angle, deg EPC Exhaust port closes volumetric efficiency with engine speed, however, because many of the phe- nomena which govern this variation are omitted from this modelling approach (see Sec. 6.2 and Fig. 6-9). fold region corresponding to each volume analyzed: however, they cannot 14.3.3 Filling and Emptying Methods describe the spatial variation of pressure (and other gas properties) due to unsteady gas dynamics in the manifolds. In "filling and emptying" models, the manifolds (or sections of manifolds) are A simple application of a filling and emptying model to the intake manifold represented by finite volumes where the mass of gas can increase or decrease with of a spark-ignition engine was described in Sec. 7.6.2. The manifold was analyzed time. Such models can range from treating the whole intake or exhaust system as as a single control volume with the throttle plate controlling mass flow into the a single volume to dividing these systems into many sections, with flow manifold and the engine cylinders controlling mass flow out. An equation for the restrictions such as the air cleaner, throttle valve, or inlet valve at the beginning, rate of change of manifold pressure [Eq. (7.22)] was derived and used to explain in between volumes, or at the end. Each volume is then treated as a control how the air flow past the throttle varied as the throttle open angle was increased, volume (an open system of fixed volume) which contains gas at a uniform state. as would occur at the start of a vehicle acceleration at part-throttle conditions The mass and energy conservation equations developed in Sec. 14.2 [Eqs. (14.1), (see Fig. 7-24). (14.3), (14.9), and (14.10) or (14.11)], coupled with information on the mass flow A second example will illustrate the conditions under which filling and rates into and out of each volume [e.g ., determined by the equations for flow emptying models give sufficiently accurate predictions to be useful.4 It concerns a through a restriction, Eqs. (C.8) and (C.9)] are used to define the gas state in each single-cylinder two-stroke cycle loop-scavenged direct-injection diesel engine. The control volume. For intake and exhaust flows these equations can be simplified engine was modeled as three open systems (the intake system, the cylinder, the since the volumes are fixed (V = 0), gas composition can be assumed frozen exhaust system) connected by flow restrictions. Various exhaust manifold (Ou/Op, 0h/Op, and @R/Op are then zero), unless backflow occurs or recycled volumes and shapes were examined, using nozzles at the manifold exit to simu- exhaust is used for emission control changes in fuel fraction are not significant, late the exhaust-driven turbine. The in-cylinder models were calibrated to match and for intake systems it may be acceptable to omit heat transfer to the walls the measured engine performance. Figure 14-2 shows the predicted and measured (w). Such methods characterize the contents of the manifold (or a region thereof) pressure variation at the exhaust system exit for two exhaust system volumes (Ve). with a single gas temperature, pressure, and composition. These vary periodically With the compact manifold the measured and predicted pressures were in good with time as each cylinder in turn draws on the intake system and discharges to agreement. With the larger exhaust system shown in the figure (Ve/V4 = 5.2) and the exhaust system. Also, under transient conditions when engine load and/or the compact manifold, good agreement is again obtained. Only with the larger speed change with time, manifold conditions will vary until the new engine volume and long pipe exhaust system is there evidence in the measured pressure steady-state conditions are established. Watson and Janota3 discuss the applica- variation of substantial unsteady gas dynamic effects. For small manifolds, and tion of filling and emptying models to manifolds in more detail. Such models can manifolds that are compact in shape, filling and emptying models can be a useful characterize these time-varying phenomena, spatially averaged over each mani- predictive tool. 756 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 757 14.3.4 Gas Dynamic Models dimensional. Mass conservation requires that the rate of change of mass within the control volume equals the net flow into the control volume: i.e ., Many induction and exhaust system design variables determine overall per- formance. These variables include the length and cross-sectional area of both primary and secondary runners, the volume and location of the plenums or junc- 2 (PA dx) = PAU -| PAU + + 2 x ( P AU ) ax (14.12) tions which join the various runners, the entrance or exit angles of the runners at a junction, the number of engine cylinders and their dimensions, intake and Retaining only first-order quantities, this equation simplifies to exhaust port and valve design, and valve lift and timing (see Secs. 6.2, 6.3, 6.7, and ap + (OU) + PUda 7.6). Most of this geometric detail is beyond the level which can be incorporated at ax = = 0 A dx (14.13) into the models discussed above. Coupled with the pulsating nature of the flow into and out of each cylinder, these details create significant gas dynamic effects The momentum conservation equation states that the net pressure forces on intake and exhaust flows which require a more complete modeling approach. plus the wall shear force acting on the control volume surface equal the rate of Gas dynamic models have been in use for a number of years to study change of momentum within the control volume plus the net flow of momentum engine gas exchange processes. These models use the mass, momentum, and out of the control volume. The net forces and momentum changes are given by: energy conservation equations for the unsteady compressible flow in the intake Pressure forces: and exhaust. Normally, the one-dimensional unsteady flow equations are used.+ These models often use a thermodynamic analysis of the in-cylinder processes to dA dx = - A - OP dx link the intake and exhaust flows. In the past, the method of characteristics was P A - P + P dx A + an dx + P dx ax used to solve the gas dynamic equations. Finite difference techniques are used in more recent intake and exhaust flow models. The basic equations and assump- Shear forces: tions of these models will now be reviewed. 5, 6 PU2 - TW RD dx = - E nD dx 2 UNSTEADY FLOW EQUATIONS. Consider the flow through the control volume within a straight duct shown in Fig. 14-3. It is assumed that the area change over where D is the equivalent diameter (4A/zt)1/2 and g is the friction coefficient given the length dx of the control volume is small so the flow is essentially one- by tw/(EpU2). The rate of change of momentum within the control volume is a af (UpA dx) + Two- and three-dimensional effects can be important and can be modeled with multidimensional flow models described in Sec. 14.5. and the net efflux of momentum across the control volume surface is ( p + Pax ( U +OU ax dx ) ( A + dA dx ) - PURA = 2 ( PU 2 A ) dx Control volume Combining these terms into the momentum equation yields TW U + ou dx - A OP dx - EPUZ ax ID dx = at (PUA dx) + > (OU2A) dx (14.14) etc. P This can be rearranged and combined with the mass conservation equation (14.13) to give A OU 1 Op + U au U2 dA d +- + 25 == 0 at ax p ox · D (14.15) FIGURE 14-3 ENERGY CONSERVATION. The first law of thermodynamics for a control Control volume for unsteady one-dimensional flow volume states that the energy within the control volume changes due to heat and analysis. shear work transfers across the control volume surface and due to a net efflux of 758 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 759 stagnation enthalpy resulting from flow across the control volume surface. The These one-dimensional unsteady flow equations have been used for a stagnation enthalpy ho is number of years to study the flow in the intake and exhaust systems of spark- U2 ignition and diesel engines, both naturally aspirated and turbocharged. Two U2 P ho = h + =u + + types of methods have been used to solve these equations: (1) the method of 2 2 characteristics and (2) finite difference procedures. The characteristic methods where u is the specific internal energy of the fluid (often approximated by c. T). have a numerical accuracy that is first order in space and time, and require a The shear work transfer across the control volume surface is zero. large number of computational points if resolution of short-wavelength varia- The heat-transfer rate _w, is given by tions is important. Finite difference techniques can be made higher order and prove to be more efficient:7, 8 this approach is now preferred. Methods for treat- 80w = apA dx ing the boundary conditions will also be described. where à is the heat transfer per unit mass of fluid per unit time into the control volume. METHOD OF CHARACTERISTICS. The method of characteristics is a well- The rate of change of energy within the control volume is established mathematical technique for solving hyperbolic partial differential equations. With this technique, the partial differential equations are transformed at Of [ (ex ) (*+ 2)] into ordinary differential equations that apply along so-called characteristic lines. Pressure waves are the physical phenomenon of practical interest in the unsteady The net efflux of stagnation enthalpy is intake flow, and these propagate relative to the flowing gas at the local sound Dx QUAL 4 + B + 2) ax speed. In this particular application, the one-dimensional unsteady flow equa- tions, (14.13) and (14.15), are rearranged so that they contain only the local fluid velocity U and local sound speed a. Hence, the equation for energy conservation becomes Since the absolute velocity of small amplitude sound waves is U + a in the at oAdxu+ + 2) | + 2x 6Us(4 + + 5) ax - upA dx = 0 direction of flow and U - a opposite to the flow direction, the lines of slope U t a are the position characteristics of the propagating pressure waves which define the position x of the pressure wave at time t. Compatability conditions (14.16) accompanying the position characteristics relate U to a. The compatability Additional simplifications are possible. Expanding Eq. (14.16) and using the mass relationships are expressed in terms of variables (called Riemann invariants) and momentum conservation equations yields which are constant along the position characteristics for constant-area homentro- pic flow, though they vary if these restrictions do not apply. Thus, the solution of ou U 3 P O(UA) the mass and momentum conservation equations for this one-dimensional at ax D PA @x (14.17) unsteady flow is reduced to the solution of a set of ordinary differential equa- If u can be represented by c, T and R/c, = y - 1 is constant, Eq. (14.17) can be tions. The equations are usually solved numerically using a rectangular grid in the rearranged and simplified to give x and t directions. The intake or exhaust system is divided into individual pipe OP UP - a2(08 + 000 ) -(7- Up(a + 280 ) = 0 (14.18) sections which are connected at junctions. A mesh is assigned to each section of at pipe between junctions. From the initial values of the variables at each mesh point at time t = 0, the values of the Riemann variables at each mesh point at where the sound speed a for an ideal gas is given by subsequent time steps are then determined. Gas pressure, density, and tem- perature can then be calculated from the energy conservation equation and the a2 = ap = Y (14.19) ideal gas law. Additional details of the method are given by Benson et al.5, 6 If friction and heat-transfer effects are small enough to be neglected, Eqs. (14.15) and (14.18) can be considerably simplified. In the absence of these effects FINITE DIFFERENCE METHODS. Finite difference methods for solving the the flow is isentropic; it has uniform entropy which is constant with time and is one-dimensional unsteady flow equations in intake and exhaust manifolds are often called homentropic flow.º If the duct area can be neglected then the contin- proving more efficient and flexible than the method of characteristics. The con- uity equation, (14.13), can be simplified also. servation equations, (14.13), (14.14), and (14.16), can be rearranged and written in 760 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 761 TABLE 14.1 Boundary conditions for unsteady one-dimensional finite element analysis9 n + 2- Pipe ends Time Out-flow Mass n + 1, j P . U , A 1 = P 2 U 2 A2 n + Energy Di = C . T 2 + = 2 At FIGURE 14-4 W Ax n Mesh in time-distance plane for application Isobaric P2 = P3 j - 1 j + 1 of one-step Lax-Wendroff method to intake Mass Distance or exhaust pipe. In-flow PIU1A1 = P2 U2A2 1,2 Energy Cp T3 = Cp T2 + 2 2 I 3 matrix form as Isentropic P2/02 = P3/P'S Pipe junctions -PU dA dx Mas , ap. V - at = = EP. U.A. p OU at PU U2 dA axl PU2 + p 25U |U = -₱ A dx - p Energy ac = [ lo , U . A . c . T. + 2 ) pu PU + pUu D 4h (T - Tw) 1 dA A dx ( 2PU3 + = ", Up ) Pressure P1 - AP1 = P2 + AP2 = P3 + AP3 = ... Dp Ap/P: = C(Ua)2 (14.20) The fluid viscous shear is small relative to friction at the wall in the momentum These finite difference solution methods usually require the introduction of equation, and heat conduction and viscous dissipation prove negligible relative to some form of dissipation or damping to prevent instabilities and large non- convective heat transfer at the wall in the energy conservation equation. These physical oscillations from occurring with nonlinear problems with large gradients equations have the vector form: (e.g ., a shock wave in the exhaust system). Amplification of the physical viscosity ÔF ÔG and the addition of artificial viscosity, damping, and smoothing terms to Eq. + = H (14.21) (14.22) are frequently used techniques.8, 9 at ax The boundary conditions at pipe ends and junctions are obtained from the where G and H are functions of F only. Several finite difference methods have appropriate conservation equations and pressure relations, as illustrated in Table been used to solve Eq. (14.21) (see Refs. 7, 8, and 9). The one-step Lax-Wendroff 14.1. Out-flows and in-flows obviously conserve mass and energy. For the flow method will be illustrated.8 Equation (14.21) can be developed into a Taylor out through a restriction, there is no pressure recovery downstream: for flow in series with respect to time, and the time and space derivatives approximated by through a restriction, the flow upstream of the restriction is isentropic. For pipe central differences around the mesh point, shown in Fig. 14-4, as junctions, the conservation equations are applied to the control volume con- tained within the dashed line in the sketch in the table. The pressure boundary F;+1 = F; - 1 At 2 Ax - (Gj+1 - Gj-1) + AtH; conditions are most easily estimated by modifying the simple constant-pressure assumption with pressure losses at each pipe exit or entry, calculated from experi- mentally determined loss coefficients (see Fig. 6-5).9 + = (As) [(G"+1 + GTXGj+1 - G;) -(G" + G" -, XG; - G;-)] (14.22) Calculations of intake and exhaust flows using these techniques predict the variations in intake and exhaust manifold pressure with crank angle (as shown, where G' = 0G/OF. This equation is first-order accurate, unless H is small. For for example, in Fig. 6-7), in single and multicylinder engines, with acceptable stability in the integration process, the time step and mesh size must satisfy the accuracy.7.9 Measured volumetric efficiency variations with engine speed, mani- requirement that fold design, and valve dimensions and timing are adequately predicted also. Figure 14-5a shows the instantaneous exhaust and intake mass flow rates for C = (U|+a) <1 (14.23) cylinder number 1 of a four-cylinder spark-ignition engine at wide-open throttle at 1500 rev/min. Note how gas dynamic effects distort the exhaust flow. Note where C is the Courant number. also the "reverse" flows into the cylinder past the exhaust valve and out of the 762 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 763 1.0 Thermodynamic analysis Phenomenological 60 of cylinder contents process models 40 0.9- O 1. Cylinder and mi Intak valve geometry 20 0.8- o 2. Thermodynamic Volumetric efficiency 0.7- Compression properties Mass flow rate, g/s -20- o Experiment Model 3. Flow rates -40- 0.6 - --- Plenum me Combustion 4. Heat transfer -60 0.5 5. Transport -80- IC properties 0 180 360 540 720 Expansion 0 2000 4000 6000 6. Combustion Crank angle, deg Speed, rev/min rate ( a ) (b) FIGURE 14-6 Exhaust 7. Emissions Logic structure of thermodynamic-based simu- FIGURE 14-5 mechanisms lations of internal combustion engine operating (a) Predicted mass flow rate through the exhaust valve m, and through the intake valve m, in cylinder cycle. 1, four-cylinder four-stroke-cycle spark-ignition engine at wide-open throttle and 1500 rev/min. Flows into cylinder are positive; flows out are negative. (b) Predicted and measured volumetric efficiency at wide-open throttle for four-cylinder spark-ignition engine. Solid line: one-dimensional unsteady flow model. Dashed line: quasi-steady flow calculation based on infinite plenums for manifolds.7 The starting point for these cycle simulations is the first law of thermody- namics for an open system, developed in Sec. 14.2. This is applied to the cylinder cylinder past the intake valve at the end of the exhaust process, and the larger volume for the intake, compression, combustion, expansion, and exhaust pro- reverse flow at the end of the intake process at this low engine speed. Figure cesses that in sequence make up the engine's operating cycle. The structure of this 14-5b shows the volumetric efficiency for this engine based on these predicted type of engine simulation is indicated in Fig. 14-6. Then, during each process, mass flow rates, as a function of speed. Experimental values and values predicted submodels are used to describe geometric features of the cylinder and valves or with quasi-steady flow equations and infinite plenums for manifolds are also ports, the thermodynamic properties of the unburned and burned gases, the mass shown. These results clearly demonstrate the important role that intake and and energy transfers across the system boundary, and the combustion process. exhaust system gas dynamics play in determining both the engine speed at which During intake and compression, the cylinder volume is modeled as a single peak breathing efficiency occurs and the air charging characteristics over the full open system. Application of the conservation equations in the form of Eqs. (14.1), engine speed range.7 (14.3), and (14.10) or (14.11) for the intake and then the compression process gives2 14.4 THERMODYNAMIC-BASED IN-CYLINDER MODELS Intake: 14.4.1 Background and Overall Model Structure m = mi - me (14.24) If the mass transfer into and out of the cylinder during intake and exhaust, the m f = = (fi -f) - Le (fe -f) m (14.25) heat transfer between the in-cylinder gases and the cylinder head, piston, and cylinder liner, and the rate of charge burning (or energy release from the fuel) are all known, the energy and mass conservation equations permit the cylinder pres- V - B $ + BM (" 1 ( im , h , - mehe - @ w) ( 14.26 ) sure and the work transfer to the piston to be calculated. Engine models of this type have been developed and used extensively to predict engine operating char- where m is the mass of gas in the cylinder, m; and me are the mass flow rates acteristics (indicated power, mean effective pressure, specific fuel consumption, through the inlet valve and the exhaust valve, and f is the fuel fraction m /m. The etc.) and to define the gas state for emission calculations. These models effectively subscripts i and e denote properties of the flow through the intake and exhaust follow the changing thermodynamic and chemical state of the working fluid valves, respectively. The thermodynamic properties for these flows are the values through the engine's intake, compression, combustion, expansion, and exhaust upstream of the valves and therefore depend on whether the flow is into or out of processes; they are often called engine cycle simulations. the cylinder. 764 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 765 Compression: Exhaust : m = 0 f= 0 (14.27a, b) im = - me f= - = " (fe -f) (14.29a, b) V Bm , (14.28) me (1 - 2 ) - VB + 11-mche - @ w ) ( 14.30 ) The pressure is then determined from Eq. (14.9). where he, the enthalpy of the flow through the exhaust valve, is the cylinder During intake and compression, the working fluid composition is frozen. average enthalpy for flow out of the cylinder and the exhaust system gas enthalpy The composition and thermodynamic properties can be determined using the if reverse flow occurs. models described in Secs. 4.2 and 4.7. Mass flows across open valves are usually The engine operating cycle should end with the working fluid at the same calculated using one-dimensional compressible flow equations for flow through a state that it started out. For the first calculations of the sequence of processes in restriction (see App. C and Secs. 6.3.2 and 14.3.2) or filling and emptying models Fig. 14-6, property values defining the initial state of the fluid in the cylinder were (Sec. 14.3.3). The more accurate unsteady gas dynamic intake (and exhaust) flow assumed. If the values of these properties at the end of the first cycle differ from models described in Sec. 14.3.4 are sometimes used to calculate the mass flow the assumed values, the cycle calculation is repeated with the appropriate new into the engine cylinder in complete engine cycle simulations when the variation initial values until the discrepancy is sufficiently small. Convergence with these in engine flow rate with speed is especially important:1º the disadvantage is much cycle simulations occurs within a few iterations. increased computing time. Heat transfer during intake and compression is calcu- The working fluid state is now defined throughout the operating cycle. The lated using one of the Nusselt-Reynolds number relations for turbulent convec- work transfer to the piston per cycle tive heat transfer described in Sec. 12.4.5. The transport properties, viscosity, and thermal conductivity used in these correlations can be obtained from relations W = Qpdv (14.31) such as Eqs. (4.52) to (4.55). During combustion which starts with the spark discharge in spark-ignition can now be obtained. From We, the masses of fuel and air inducted, m, and ma, engines and with spontaneous ignition of the developing fuel-air jets in and engine speed N, all the engine indicated performance parameters can be cal- compression-ignition engines, the actual processes to be modeled become much culated: power, torque, mean effective pressure, specific fuel consumption, fuel- more complex. Many approaches to predicting the burning or chemical energy conversion efficiency; as well as volumetric efficiency, residual gas fraction, total release rate have been used successfully to meet different simulation objectives. heat transfer, etc. With a friction model, the indicated quantities can be converted The simplest approach has been to use a one-zone model where a single ther- to brake quantities. modynamic system represents the entire combustion chamber contents and the The more sophisticated of these thermodynamic-based engine cycle simula- energy release rate is defined by empirically based functions specified as part of tions define the working fluid state throughout the cycle in sufficient detail for the simulation input. At the other extreme, quasi-geometric models of turbulent useful predictions of engine emissions to be made. The discussion in Chap. 11 of premixed flames are used with a two-zone analysis of the combustion chamber emission-formation mechanisms indicates that our understanding of how some of contents-an unburned and a burned gas region-in more sophisticated simula- these pollutants form (e.g ., NO ,, CO) is reasonably complete, and can be tions of spark-ignition engines. In compression-ignition engines, multiple-zone modeled accurately. The formation processes of the other pollutants (unburned models of the developing fuel-air jets have been used to provide more detailed hydrocarbons and particulates) are not adequately understood, though modeling predictions of the combustion process and nonuniform cylinder composition and activities are continuing to contribute to that understanding. The key features of state. These combustion models will be reviewed in the following sections (14.4.2 models for predicting engine emissions were discussed in Chap. 11. and 14.4.3) and the appropriate conservation equations for the combustion Cycle simulations and combustion models which have been developed for process will be developed there. In diesels, radiation heat transfer becomes impor- spark-ignition engines, where the fuel, air, residual gas mixture is essentially uni- tant during the combustion process (see Sec. 12.5). formly mixed, are discussed in Sec. 14.4.2. Compression-ignition engine simula- The expansion process is either treated as a continuation of the combustion tions and combustion models are then discussed in Sec. 14.4.3. The special process or, once combustion is over, can use the form of the mass, fuel, and features required for prechamber engine models are reviewed in Sec. 14,4.4. energy conservation equations which hold during compression [Eqs. (14.27) and Finally, thermodynamic-based models for more complex engine systems- (14.28)]. The exhaust process conservation equations for a one-zone open-system multicylinder, turbocharged, and turbocompounded engines-are discussed in model of the cylinder contents are2 Sec. 14.4.5. 766 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 767 14.4.2 Spark-Ignition Engine Models Disc Bowl-in-piston These models have usually followed the conceptual structure indicated in Fig. 0.6 14-6. Our focus here is on the combustion submodels that have been developed and used successfully. Features of the spark-ignition engine combustion process Bowl-in-piston, center ignition that permit major simplifying assumptions for thermodynamic modeling are: (1) 0.5 the fuel, air, residual gas charge is essentially uniformly premixed; (2) the volume occupied by the reaction zone where the fuel-air oxidation process actually 0.4 --- Disc, center occurs is normally small compared with the clearance volume-the flame is a ignition Flame area Bore2 thin reaction sheet even though it becomes highly wrinkled and convoluted by the turbulent flow as it develops (see Sec. 9.3); thus (3) for thermodynamic 0.3- analysis, the contents of the combustion chamber during combustion can be Bowl-in-piston, side ignition analyzed as two zones-an unburned and a burned zone. 0.2- Useful combustion chamber design information can be generated with Disc, side ignition simple geometric models of the flame. In the absence of strong swirl, the surface 0.1- which defines the leading edge of the flame can be approximated by a portion of the surface of a sphere. Thus the mean burned gas front can also be approx- imated by a sphere. Then, for a given combustion chamber shape and assumed IC 0.2 0.4 0.6 0.8 1.0 flame center location (e.g ., the spark plug), the spherical burning area A] [see Eq. Flame radius (9.40)], the burned gas volume V, [see Eq. (9.39)], and the combustion chamber Bore (a) surface "wetted" by the burned gases can be calculated for a given flame radius rb and piston position (defined by crank angle) from purely geometric consider- ations.+ The practical importance of such "model" calculations is that (1) the Side mass burning rate for a given burning speed S, (which depends on local turbu- Bowl-in-piston 0.8|- Center lence and mixture composition) is proportional to the spherical burning area A, as given by Eq. (9.44); (2) the heat transfer occurs largely between the burned Disc - Side 0.6 -*- Center gases and the walls and is proportional to the chamber surface area wetted by the burned gases Ab.» [see Eq. (12.21)]. Using the fact that the density ratio across Piston at TC the flame pu/p, is approximately constant and equal to 4, the unburned and Flame area Bore2 0.4 burned gas volumes V1 and 1/, can be related to the unburned and burned mass fractions (1 - x;) and x ,, respectively. 0.2 FIGURE 14-7 Examples of the results of such flame geometry calculations are shown in Calculated spark-ignition engine Figs. 14-7 and 14-8.11 Figure 14-7a shows spherical flame areas A, as a function spherical flame surface area: (a) as a of flame radius r, for two different chambers and two plug locations and the TC function of flame radius for different 0.02 0.04 0.06 0.08 0.10 0.12 combustion chamber shapes and piston position. The much larger flame area and shorter flame travel length of the Enflamned volume spark plug locations and (b) as a central plug location are obvious. Such area data can be plotted as a function of Bore3 function of enflamed volume. Piston burned gas volume 1;, as shown in Fig. 14-7b, so that comparisons of Ab(r)) for (b) in top center position.11 different chambers at the same mass fraction burned can be made. The advantage of a more compact chamber with higher central clearance height is apparent. Figure 14-8 shows that burned-gas-wetted wall area on the cylinder head, cylin- der wall, and piston as a function of flame radius and crank angle for an open chamber with central ignition. The cylinder head and piston are the dominant areas early in the expansion stroke when the burned gas temperatures and heat fluxes are highest. + Note that the center of this sphere may be convecteur from the spark plug location, especially Mass fraction burned versus crank angle profiles determined from a first if some swirl is present. However, only strong swirling and squish flows produce major distortions to law analysis of cylinder pressure data, as shown in Figs. 9-2, 9-5, and 9-8, have an the flame surface shape. essentially universal dimensionless shape, as indicated in Fig. 9-13. Much useful 768 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 769 4000 Open chamber, center ignition 3000- 700 Pressure, kPa 71.0 1.2|- 2000- 1000- Xb p -0.5 ¢ 3000F 0 1.0 la Cylinder head 0-70º 2000- Temperature, K 42º 0.8- 2000 Wetted area Bore2 50 $ 40- NO -0.04 0.6- 1600- Piston Cylinder wall 30 NO, ppm 1200- 0.03 Heat transfer rate ( ,, S"1 20- 800- 0.02 51 0.4 42º! 1 70º 0º 10- 400 0.01 FIGURE 14-8 OL OL -180 0 180 0.2 Calculated burned-gas-wetted wall TC area as a function of radius based Crank angle, deg 14º on spherical flame model of an 0.2 0.8 1.0 open-chamber SI engine with FIGURE 14-9 0.4 0.6 Flame radius center plug location, for piston Cylinder pressure p, mass fraction burned x ,, unburned and burned gas temperatures (Tu = Bore locations of 0º, 42º, and 70º.11 unburned, T. = adiabatic burned core, T, = mean burned gas temperatures), heat-transfer rate Qw (normalized by fuel flow rate x heating value), thermal boundary-layer thickness of, and mean nitric oxide concentration in the burned gases, through a four-stroke engine operating cycle, predicted by thermodynamic-based cycle simulation. 5.7-dm3 displacement eight-cylinder engine operating at wide- analysis has been done with engine simulations where this universal combustion open throttle, 2500 rev/min, with equivalence ratio = 1.1. Gross indicated mean effective pressure is profile has been used as a calculation input. The S-shaped mass fraction burned 918 kPa and specific fuel consumption is 254 g/kW . h.13 profile is often represented by the Wiebe function: Xx = 1 - exp -a(-00 ) 2+ 17 40 (14.32) from heat loss. The open-system conservation equations, (14-1) and (14.10) or (14.11), are now applied to the core and boundary-layer region separately. The where 0 is the crank angle, 00 is the start of combustion, A0 is the total com- boundary-layer region covers that portion of the combustion chamber wall bustion duration (x) == 0 to x) ~ 1), and a and m are adjustable parameters which wetted by the burned gases, as shown in Fig. 9-4, and is of thickness of, which fix the shape of the curve. Actual mass fraction burned curves have been fitted increases with time. The temperature of the boundary-layer zone (assumed with a = 5 and m = 2.12 uniform) is usually taken to be the mean of the wall temperature and burned gas The conservation equations for an open system [Eqs. (14.1) and (14.10) or core temperature. Equation (14.10) or Eq. (14.11) is used to relate the enthalpy (14.11)] are now applied to the unburned gas zone ahead of the flame and to the flux due to the mass flow across the inner edge of the boundary layer (which has burned zone behind the flame, in turn (see Fig. 9-4). For premixed engines, fand an enthalpy equal to the core gas enthalpy), the heat transfer to the wall, the o are zero. During combustion, m and m; in Eq. (14.10) or Eq. (14.11) are the changing energy within the boundary-layer system due to its increasing mass and mass flow rate across the flame sheet. This is -m, for the unburned zone system changing state, and the work transfer due to its changing volume. and +m, for the burned zone system; m, is given by mx ,, with x, obtained by An example of predictions of cylinder pressure, unburned and burned gas differentiating Eq. (14.32). temperatures, heat-transfer rate, and boundary-layer thickness, based on an To calculate the effect of heat transfer on the burned gas state more accu- assumed 50º total burn duration for a 5.7-dm3 eight-cylinder engine at wide-open rately, the burned gas zone in Fig. 9-4 can be modeled in two parts: an adiabatic throttle and 2500 rev/min is shown in Fig. 14-9.13 Appropriately based predic- core and a boundary-layer region. The intent here is to account for the fact that tions of overall engine performance parameters made with this type of thermody- heat loss to the walls primarily cools the burned gas adjacent to the wall, and namic model agree well with engine data. Figure 14-10 shows predictions of only indirectly affects the core gas through the change in pressure that results indicated specific fuel consumption and exhaust gas temperature as a function of 770 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 771 350F MBT timing 31% 300 420 isfc, g/kW .h 250 400 26 200 Experiment: MBT _ MBT + 10% Calculated 380 21 050 360- 1000 MBT + 10% FIGURE 14-10 bsfc, g/kW .h -- 4- Predicted and measured indicated spe- 11.5 cific fuel consumption and exhaust tem- 340 Exhaust temperature, K 950 perature as a function of the fuel/air FIGURE 14-11 MBT 900 equivalence ratio for a spark-ignition 320 1.3 Predicted brake specific fuel con- engine operated at 1250 rev/min and imep of 379 kPa. MBT: maximum Ecycle sumption as a function of heat transfer per cycle to the com- 850F brake torque timing. MBT + 10%: 300- bustion chamber walls (as 0.7 0.8 0.9 1.0 1.1 1.2 combustion timing retarded to give 10 percent of the fuel's heating Equivalence ratio ¢ percent fuel consumption penalty.14 280 value) and total burn duration [40 in Eq. (14.32)]. 1250 rev/min, 20 40 60 80 100 262 kPa bmep, fuel/air equiva- lence ratio = 0.91, maximum the fuel/air equivalence ratio at fixed load and speed. The isfc predictions and Total burn duration, deg brake torque spark timing. 15 data agree well (except for very lean mixtures with retarded timing where cycle- by-cycle combustion variations are sufficiently large so predictions based on the average cycle lose accuracy); the predicted curves for exhaust temperature show these models assume that the overall flame shape approximates a portion of a the same trends as the experimental data. However, they are higher due to under- sphere centered at or near the spark plug. Empirical flame models have difficulty estimation of the heat losses during the exhaust process.14 appropriately describing the three phases of the combustion process-flame The output from such thermodynamic-based cycle simulations has replaced development, rapid burning, and termination-with sufficient generality to be the fuel-air cycle as a predictor of effects of major variables on engine per- widely useful. One such model, based on the experimental data shown in Fig. formance and efficiency. An instructive example of the value of such predictions is 9-30, has been used successfully to evaluate different combustion chambers. 16, 17 shown in Fig. 14-11, where fuel consumption at constant equivalence ratio, load, The burning speed S, [defined by Eq. (9.44)] is related empirically to the laminar and speed has been computed as a function of total burn duration and heat loss flame speed S_ (see Sec. 9.3.3), the local rms velocity fluctuation u'f [see Eq. (8.22)] to the chamber walls: increasing burn duration and heat loss both worsen fuel under motored engine conditions, the firing and motored cylinder pressure at the consumption.15 Such data can be used to assess the efficiency improvements that same crank angle, and spark advance. While a good fit to the data in Fig. 9-30 should result from reduced heat transfer (e.g ., reduced chamber surface area) and for engine flames during their turbulent rapid-burning phase was obtained, increased burn rate. Obviously the dependence of burn rate on engine design and during the flame development period a correction factor was required to fit the data. operating parameters has not been modeled; the burn rate profile was a calcu- lation input. Such models are most useful either (1) when the burn rate profile is Spark-ignition engine combustion models with a more fundamental frame- not critical to the problem under study or (2) when predictions for a range of work have been developed and used. Based on coupled analysis of flame front assumed burn rate profiles provide the required information. location and cylinder pressure data, Keck and coworkers18-20 have derived the following burning law: So far we have discussed engine cycle simulations where details of the com- bustion process have been specified as input. The same thermodynamic-based simulation structure can be used in conjunction with a combustion model which dmb = pu A, St + ". dt To (14.33) predicts the rate of fuel burning. Various combustion models have been proposed and used for this purpose. Some of these are empirically based; some are based du - on the highly wrinkled, thin reaction-sheet flame model described in Sec. 9.3. All dt = P& Agur(1 - e /kb) - M TD (14.34) 772 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 773 2. Initially, as t-+ 0, Unburned S. - SL (14.36b) UT SL 3. Quasi-steady state, du/dt ~ 0, (14.36c) 4. Final burning stage after the flame front reaches the wall, t _ tw (when As - 0), - IT Burned mp = e-( -tw )/ 2b (14.36d) im, (tm) FIGURE 14-12 To apply Eqs. (14.33) and (14.34), the quantities ur and ty (or IT = 16 SL) Schematic of turbulent premixed spark-ignition engine flame, illustrating the physical basis for must be evaluated. Two approaches have been taken: (1) use of empirical correla- burning law of Eqs. (14.33) to (14.35). The approximately spherical front of the "thick" turbulent tions for these variables, derived from engine flame data (such as that described flame (dashed line) diffuses outward at the laminar flame speed S1. Fresh mixture also crosses this in Sec. 9.3.4); (2) use of more fundamental models to predict these quantities. front at a characteristic velocity ur due to turbulent convection. Schematic on left shows detailed Keck has derived the following correlations for ur and IT, based on the flame structure: 6 ,, is a reaction-sheet thickness, IT is characteristic scale of wrinkles in the sheet. application of Eqs. (14.33) and (14.34) to several sets of engine combustion data: u = 0.08u, Pu 1/2 where (14.37) u = me - mo = P.(Vs - Vb) = Puli(AL - Ag) (14.35) 12 = 0.8Li Pt 3/4 (14.38) is a parametric mass (interpreted as the mass entrained within the flame region that has yet to burn), ur a characteristic speed, and t = IT/SL is a characteristic burning time. IT, AL, Vs, Af, V, are defined in Sec. 9.3.4. ur was found to be proportional to p. (at time of spark) and to correlate well Figure 14-12 illustrates the physical basis for this model. The first term in with mean inlet gas speed it = n.(Ap/A¡y)S ,, where no is volumetric efficiency, A, Eq. (14.33) represents the laminar (diffusive) propagation forward of the approx- is piston area, Ai, is the maximum open area of the inlet valve, S, is mean piston imately spherical front of the "thick" turbulent flame; the second term represents speed. IT appears to scale with valve lift, L;; it decreases with increasing density the burning of mixture already entrained within this flame front. In Eq. (14.34), at a rate proportional to p. 3/4. While ur and IT are not constant during the combustion process, their variation is modest. 18 which describes the rate of change of unburned mixture mass u within the flame A quantitative comparison of predicted and measured flame radius as a zone, the first term represents the turbulent convection of unburned mixture across the spherical front of the flame and the second term represents the mass function of time is shown in Fig. 14-13 for hydrogen and propane fuel-air mix- tures which exhibit widely different behavior:18 the figure indicates both the rate of burning of entrained but not yet burned mixture which is contained within the "wrinkles" and "islands," which the distorting and stretching of the behavior and validity of the model. Predicted burned gas expansion speeds ub thin reaction sheet by the turbulent flow produces. This has been called an [see Eq. (9.43)] are shown in Fig. 14-13a as a function of burned gas radius; the parameters ur and Ir were chosen to fit the propane data. Figure 14-13b shows "entrainment" or "eddy-burning" model for the above reasons. The exponential term in brackets in Eq. (14.34) allows for the fact that the flame sheet initially is that the measured flame front radii, r, are in good agreement with the predicted flame and burned gas radii, r and ro, for these two fuels. The initial expansion spherical and laminarlike: it requires a time of about t, to develop into a turbu- speed of hydrogen is about 10 times that of propane. Since r ~ r, for early times, lent flame. The behavior of Eqs. (14.33) and (14.34) in four important limits is: S, ~ SL and this ratio is expected. As r, become large, (r - r)) -> Urt ,, which is several times smaller for hydrogen mixtures than for propane mixtures. 1. For a quiescent mixture, up -+ 0 or IT -+ 00, An adaption of this approach developed by Tabaczynski and coworkers21, 22 is based on the following model of turbulent flame propagation. (14.36a) The vorticity in the turbulent flow field is concentrated in vortex sheets which are 774 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 775 25 Thus, in Eqs. (14.33) and (14.34), up and ty are given by (a) ¢ = 1.0 x, = 0.2 H2 l = 1.5 mm UT ~u' and (14.39) SL 20 where IM, the microscale, is determined from the integral scale and the turbulent ACA Pb Reynolds number via Eq. (8.15), assuming that the turbulence is homogeneous and isentropic. The task therefore becomes one of evaluating u' and I1. 15 One approach used is to relate the turbulence intensity at the start of the Expansion speed up, m/s combustion process to the mean intake flow velocity through the valve: e.g ., 23 C3Hg O SuTtSL CS B2 UT Liv Div (14.40) Far wall - where S, is the mean piston speed, B the bore, and Li, and Diy the lift and diameter of the inlet valve. It is assumed that the integral length scale at the start PuSL of combustion, 11.0, is proportional to a characteristic flow dimension, usually the 20 40 60 80 clearance height h. Then, during combustion, the unburned portion of the charge is assumed to undergo isentropic compression sufficiently rapidly that the Burned gas radius ro, mm angular momentum of the "eddies" is conserved and the length scale follows the eddy size, i.e ., a simple rapid distortion process occurs : (b) Far wall P 1/3 1/3 - - - u' = u'o 1 = 12,0 18 (14.41) 60 This model predicts an increase in turbulence intensity and decrease in length scale with compression, which is only partly confirmed by experiment. Radius, mm 40 A more sophisticated approach is to describe the dynamic behavior of the rev/min = 1380 turbulence with one or more rate equations for the key turbulence parameters: k Calculated the turbulent kinetic energy and & the dissipation rate of k. Turbulence is gener- 20) ated, diffused, and dissipated by the flow field, so the rate of change of turbulent o . Experiment kinetic energy k can be written: dak 20 40 60 80 at = P& + D - PE (14.42) Degrees after spark where the term P, represents the volumetric production of turbulence and the FIGURE 14-13 (a) Calculated burned gas expansion speed u, for stoichiometric hydrogen-air and propane-air mix- diffusion term D can be modeled as a gradient diffusion with an effective turbu- tures as a function of burned gas radius r ,. (b) Comparison of experimentally measured (points) and lent viscosity which dominates the laminar diffusion process. In this application, calculated (dashed curve) flame radii r for these mixtures as a function of crank angle. Also shown Eq. (14.42) is integrated over the combustion chamber (or a region of the (solid curve) is the burned gas radius r ,. 18 chamber) to provide spatially averaged turbulence predictions. Then the diffusion terms become boundary fluxes: e.g ., the transport of kinetic energy across the of a size comparable to the Kolmogorov scale Ix [see Eq. (8-11)]. These vortex combustion chamber boundary due to flow through the inlet or exhaust valve. sheets are assumed to have a characteristic spacing which is of the order of the The dissipation rate & is related to the integral length scale via Taylor microscale ly, which is a function of the integral length scale !, and the CDK3/2 turbulent Reynolds number as indicated by Eq. (8.15). From these turbulence CD = 0.09 (14.43) assumptions it is argued that ignition sites propagate along the vortex sheets with a velocity u' + SL, where u' is the local turbulence intensity. The propagation of /, can be taken as proportional to the clearance height (l, ~ 0.22 h), or an addi- the reaction front between the vortex sheets is assumed to be a laminar process. tional rate equation for a second turbulence parameter, the dissipation rate e, can 776 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 777 20 1.0 0.8- 15 0.6F O Normal lift Mass fraction burned Central spark, º 1/4 normal lift 0.4- medium swirl Data _ Wall spark, medium swirl Wall spark 10 0.2} Turbulence intensity, m/s "zero" "swirl FIGURE 14-16 Model Comparisons of predicted and measured mass 0 10 20 30 40 50 60 70 fraction burned versus crank angle profiles for same Crank angle after spark, deg swirl levels and plug locations as Fig. 14-15.28 5 FIGURE 14-14 Predicted turbulence intensity u' as a function of TC B=B=8-8 OL crank angle and valve lift in engine operating at -400 -200 0 200 400 1500 rev/min, 414 kPa imep, with a compression reduced maximum valve-lift profile (one-quarter normal) are shown. The high Crank angle, deg ratio of 10.27 levels of turbulence generated during the first half of the intake process decay substantially before the latter stages of the compression stroke produce some amplification. Reduced valve lift produces higher levels of turbulence intensity at be used. In the more complete of these k - & turbulence models,24 the & equation combustion, as is well known.29 Figure 14-15 shows the predicted turbulence is similar to the k equation with production, diffusion, and dissipation terms. behavior during combustion for a disc-shaped chamber for different swirl levels These k - e turbulence models are discussed more fully in Sec. 14.5.2. and plug locations. Swirl is shown to increase the turbulence intensity. Compari- The application of this turbulence model to the spark-ignition engine com- son of predicted and measured mass fraction burned profiles versus crank angle bustion chamber becomes complicated and the reader is referred to references for for different swirl levels and plug locations are shown in Fig. 14-16. The large the details. 25-28 Considerable success with predicting trends in mass burning rate flame area effects (shown here in the two limiting plug locations: side wall and has been achieved with this type of model. Design variables examined include: center) and significant though lesser effect of swirl are correctly modeled. Such swirl, squish, valve lift, bore/stroke ratio. The advantage of such models is that models are useful for relating changes in spark-ignition engine design and oper- they are straightforward computationally so that extensive parametric sets of cal- ating variables to changes in engine performance, via predictions of changes in culations are feasible. The major disadvantage is the ad hoc nature of the turbu- flame development and propagation. lence and flame models which involve plausible but arbitrary assumptions. The above type of combustion model has been used to obtain explicit rela- Sample predictions are shown in Figs. 14-14 and 14-15.27, 28 Figure 14-14 shows tions for the flame development and rapid burning angles as functions of engine the variation in turbulence intensity u' in an engine with a disc-shaped com- design and operating variables.3º The equation for the mass burning rate, (14.33), bustion chamber, throughout the operating cycle. A normal valve-lift profile and was effectively integrated over the relevant portion of the total combustion process; the turbulent characteristic velocity was assumed proportional to S ,, the mean piston speed. The flame development angle was found to vary as 3.0 A02 = C(Spv) 1/3 2 2/3 (14.44) 2.8|- 2.6|- where v is the kinematic viscosity (v = u/p) and h is the clearance height at igni- tion. C is a constant which depends on engine geometry and is determined by Turbulence intensity, m/s 2.4- matching Eq. (14.44) with engine data. The rapid burn angle (here taken as the crank angle between x, = 0.01 and 1.0) is given by 2.2- Central spark, high swirl Wall spark, high swirl 2.0H FIGURE 14-15 ( Sp v # ) 1/3 / kg ) 2/3 Central spark, "zero" swirl Predicted turbulence intensity during combustion SE (14.45) for high and "zero" swirl levels for central and 0 20 40 60 80 cylinder wall spark plug locations. Same engine and where C is a constant which depends on engine geometry, B is the bore, the Crank angle, deg operating conditions as in Fig. 14-14.28 subscript i denotes the value at ignition, and the superscript * denotes the value 778 MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 779 INTERNAL COMBUSTION ENGINE FUNDAMENTALS at cylinder conditions where x = 0.5. These expressions show reasonable agree- One extensively used model of this type developed by Watson et al.35 is ment with observed trends in A0, and A0;. especially appropriate for use in total diesel system simulations where the com- bustion process details are not the primary focus. It is based on Lyn's description of compression-ignition combustion-a rapid premixed burning phase followed 14.4.3 Direct-Injection Engine Models by a slower mixing controlled burning phase. The fraction of the injected fuel In direct-injection compression-ignition and stratified-charge engines, the liquid that burns in each of these phases is empirically linked to the duration of the fuel is injected into the cylinder as one or several jets just prior to ignition. In ignition delay. One algebraic function is used to describe the premixed heat- large direct-injection compression-ignition engines, the air flow is essentially release phase and a second function to describe the mixing-controlled heat- quiescent. However, in medium and smaller size DI engines, the air flow is release phase. These two functions are weighted with a phase proportionality usually swirling about the cylinder axis at up to 10 times the crankshaft rotation- factor, ß, which is largely a function of the ignition delay. Thus: al speed; this air-flow pattern increases the rate of entrainment of air into the fuel m j . b ( 2 ) jet to increase the fuel-air mixing rate. Thus modeling of the ignition and com- = Bf1 + (1 - B) f 2 (14.46) bustion processes for direct-injection types of engines is much more complex than for premixed-charge spark-ignition engines. The unsteady liquid-fuel jet where my , is the mass of fuel burned, m, , is the total fuel mass injected per cycle phenomena-atomization, liquid jet and droplet motion, fuel vaporization, air per cylinder, and t' is time from ignition non-dimensionalized by total time entrainment, fuel-air mixing, and the ignition chemistry-all play a role in the allowed for combustion [=(t - tign)/Atcomb].t The premixed-burning function is heat-release process (see Chap. 10). It is not yet possible to model all these pheno- f1 = 1 - (1 -1)K2 (14.47) mena from a fundamental basis, even with the most sophisticated fluid-dynamic- based codes now available (see Sec. 14.5), since many of these processes are not and the mixing-controlled function is yet adequately understood. However, models at various levels of detail and f2 = 1 - exp (-K3t'k4) (14.48) empiricism have been developed and have proven useful in direct-injection diesel where K1, K2, K3, and K4 are empirical coefficients. The proportionality factor and stratified-charge engine analysis. This section reviews the important features B is given by of single-zone heat-release models and phenomenological jet-based combustion models. Their relative simplicity and modest computer time requirements make B = 1-4 (14.49) them especially useful for diesel cycle simulation and more complex engine system studies. where o is the overall fuel/air equivalence ratio and a, b, and c are empirical Single-zone models assume that the cylinder contents can be adequately constants. described by property values representing the average state, and use one or more Correlation with data from a typical turbocharged truck engine gave the algebraic formulas to define the heat-release rate. The functional forms of these following values for K1 to K4: formulas are chosen to match experimentally observed heat-release profiles (see Sec. 10.4.2). Coefficients in these formulas, which may vary with engine design K1 = 2 + 1.25 x 10-8(Zia N)2.4 details and operating conditions, are determined empirically by fitting with data. K2 = 5000 The phenomenological description of diesel combustion developed by Lyn (see Sec. 10.3) comprises three primary phases: the ignition delay period, the premixed 14.2 (14.50) fuel-burning phase, and the mixing-controlled fuel-burning phase. Ignition delay K3 - 60.644 correlations are reviewed in Sec. 10.6.6. Here models for the second and third K4 = 0.79K9.25 phases, when the major heat release occurs, are summarized (see Ref. 31 for a more extensive review). The attraction of the one-zone heat-release approach is where Tid, the ignition delay, is in milliseconds and N, engine speed, is in revo- its simplicity: however, since it cannot fully describe the complex phenomena lutions per minute. It also gave these ranges for a, b, c:35 which comprise the compression-ignition engine combustion process, substantial 0.8 < a < 0.95; 0.25 < b < 0.45; 0.25 < < < 0.5 empirical input must be used. Several one-zone heat-release models have been proposed and used (e.g ., Refs. 32 to 34). These use simple equations to describe the rate of release of the fuel's energy, sometimes modeled on the presumed con- trolling physical or chemical process and always calibrated by comparison with t The combustion duration at At comt is an arbitrary period within which combustion must be com- data. pleted. A value of 125º was used above. 780 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 781 Such single-zone heat-release models are useful because of their simplicity. Air swirl They obviously cannot relate engine design and operating variables explicitly to Air zone Drag force the details of the combustion process. Experience indicates that those models = C.V 2b dme, B, with only one function are not usually able to fit experimentally determined heat- 4 (7, 6 dt dmmp release profiles with sufficient accuracy. All single-zone heat-release models Ve (r) dt should be checked against experimentally derived heat-release profiles, and recal- ibrated if necessary, before being used for predictions. amy Rich core Many thermodynamic-based direct-injection engine simulations incorpo- Air entrainment Fuel Bj rate an explicit model for each fuel spray which attempts to describe how the spray develops with time. The spray starts out as a liquid fuel jet which then vaporizes, entrains air and (later) burned gases. Mixture preparation can be limited by the availability of either fuel vapor or air, the former limited by (a) (b) droplet evaporation and the latter by air entrainment. While there is evidence in FIGURE 14-17 the literature to support both of these phenomena as rate-limiting, more recent (a) Schematic of one-dimensional quasi-steady fuel spray model used to define spray centerline trajec- tory and width as radially outward-moving spray interacts with swirling air flow. (b) Schematic of studies36 show that most (70 to 95 percent) of the injected fuel is in the vapor multizone model for fuel spray which, based on empirically calculated spray motion and assumed phase at the start of combustion, whereas only 10 to 35 percent of the vaporized concentration distributions in the spray, successfully evolves discrete combustion zones (each contain- fuel is mixed to within the combustion limits (equivalence ratio between 0.3 and ing a fixed fraction of the fuel) as fuel is injected, vaporized, and mixed with air. (dm,/dt) = rate of fuel 3). This suggests that the combustion process in typical heavy-duty direct- injection into rich core; (dmm/dt) = rate of preparation of mixture for burning; (dme, g/dt) == rate of injection compression-ignition engines is mixing controlled rather than vapor- entrainment of air into zone B ,. 37. ization controlled. While spray geometry is an essential aspect of the fuel-air mixing process, it axisymmetric turbulent jets42 are often assumed to apply. Although the fuel spray may not be necessary to model the precise details of the actual configuration. For is initially pure liquid, the liquid fuel drops soon become a small fraction of the the purpose of heat-release and emission analysis, it suffices in many phenomeno- jet volume due to vaporization and air entrainment. Downstream of the initial logical models to calculate the evolution of the fuel mass, composition, volume, liquid breakup region, the velocity of the small drops relative to the vaporized and temperature of critical regions of the spray based on a generic spray fuel and air is small, so the spray acts as a gas jet. Adding a combustion model to geometry. Alternate approaches attempt to provide a detailed structure for the this quasi-steady gaseous jet model for fuel-air mixing is an additional major fuel spray to improve the modeling of air entrainment, effects of swirl/spray inter- step. action, and heat transfer. The more commonly used approaches are illustrated in A comparison between this type of gas jet model and an experimental Fig. 14-17. engine spray is shown in Fig. 14-18. A single fuel jet was injected into a disc- The schematic in Fig. 14-17a illustrates the simplest approach: it is shaped chamber in the location shown, and schlieren photography used to assumed that the growth and motion of the spray or jet within the chamber can observe the spray trajectory. Good agreement was obtained for the spray center- be analyzed as a quasi-steady one-dimensional turbulent gaseous jet.38-40 The line: note the significant effect of swirl. Reasonable agreement was also obtained intent here is to describe the position of the jet within the combustion chamber between predicted and measured spray boundaries. and the overall jet size as a function of time. Entrainment of air into the jet is Figure 14-17b shows a multizone model for each fuel spray which has been assumed to take place at each point along the jet surface at a rate proportional to used extensively for engine performance and emissions studies in quiescent DI the velocity difference between the jet and surrounding air at that point. Two diesels. 37, 43 The spray is modeled as a gas jet, with penetration, trajectory and empirical entrainment coefficients41 are used for the proportionality constants for spreading rate determined from empirical equations based on axisymmetric tur- the relative motion in the jet axial and transverse directions. Conservation equa- bulent jet data. These equations describe the approximate spray geometry. The tions for fuel mass and total mass, and momentum (in two or three orthogonal fuel-air distribution within the spray is determined by using a normal distribution directions) are used to determine the jet trajectory and size. The jet slows down across the spray cross section and a hyperbolic profile along the axis of the spray. due to air entrainment. Deflection of the jet results from the entrainment of air Progressively evolving, discrete combustion zones, each containing a fixed frac- with a momentum component normal to the jet axis, and from drag forces due to tion of the total fuel mass, are then superimposed on the geometrically defined the normal component of air flow past the jet. This approach does not define the fuel-air distribution. Outer zones are diluted with air and inner zones are added velocity and concentration profiles across the jet: it only calculates the mean as fuel vaporizes and mixes, as injection and combustion proceeds. The model values at any jet axial position. Experimentally determined radial profiles for implicitly assumes that combustion does not affect the mixing rates. With careful 782 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 783 50r E(I, L) y-penetration O 40 O o o O O o 30 y-penetration, mm 00 20 oo - Trajectory 10 No swirl 0 T 1 2 3 4 5 6 5 6 Injection Crank angle, deg 50 No swirl ----- 40 O Ignition (A) (B) O 30 --- Combustion x-penetration, mm Controlled by fuel Controlled by air 20 evaporation rate entrainment rate 10 x-penetration 123456 7 8 ( A) (B) Crank angle, deg - FIGURE 14-19 FIGURE 14-18 Valve open Schematic of spray model with Spray trajectory and width calculated using one-dimensional quasi-steady spray model of type illus- Complete Incomplete many small packages, each with trated in Fig. 14-17a, compared with experimental data taken in special visualization direct-injection combustion combustion the same fuel mass, and of the stratified-charge engine.39 processes that occur within each Liquid fuel Air package, developed and used by 88% Vaporized fuel Products Hiroyasu et al.44 adjustment of calibrating constants, this model describes engine performance variations with reasonable accuracy as major design and operating variables E (3, 4) E (1, 4) change. E (1, 1) More detailed geometric models of the fuel-air mixing and combustion pro- ( a ) L Rk cesses in engine sprays have been developed (e.g ., Ref. 44). The intent is to follow E (1, 1) the spray development in a swirling air flow and the spray interaction with the E (3, 2) No swirl combustion chamber wall. Figure 14-19 illustrates the approach. The liquid fuel E (1, 4) which enters the chamber through the injector nozzle is divided into many small (b) equal mass "elements." The spray motion is defined by an experimentally based correlation. Air entrainment is calculated from momentum conservation and the E (1, 1) With weak swirl spray velocity decrease predicted by this correlation. The processes which occur E (5, 2 within each element are also illustrated in Fig, 14-19. The fuel drops evaporate and fuel vapor mixes with entrained air. When ignition occurs combustible E (1 , 4) mixture prepared before ignition burns rapidly: it is assumed to burn at the With strong swirl stoichiometric composition. The continuation of the burning process then depends on the composition of the element: it may be limited by either the rate of (c) (d) production of fuel vapor by evaporation or the availability of air by the rate of entrainment (paths A and B in Fig. 14-19). FIGURE 14-20 The growth of the spray is determined from the air entrainment into each Method used with model of Fig. 14-19 to compute spray and flame configuration: (a) prior to element and the combustion-produced expansion of each element, as indicated in impingement of spray on wall-shaded elements indicate combustion; (b) and (c) show spray behavior following impingement on the cylindrical bowl wall of the DI diesel combustion chamber; (d) shows Fig. 14-20. When impingement on the wall occurs, the spray is assumed to spread effect of swirl on spray and flame configuration.44 784 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 785 along the wall with a constant thickness as shown in Fig. 14-20. When the per- Specifications of engine and operating conditions iphery of the spray reaches that of a neighboring spray the sideways growth of the spray is then prevented and the thickness of the elements along the wall increases. Swirl effects are calculated from tangential momentum considerations. Fuel-injection simulator Motoring Each annular cone ring element is shifted sideways by the swirl as indicated in Spray distribution and penetration Fig. 14-20. The heat-release rate in the combustion chamber is obtained by summing up the heat release in each element. Nitric oxide and soot formation calculations Air entrainment are based on the time histories of temperature, vaporized fuel, air and com- Evaporation of liquid fuel bustion products in each element. The overall structure of this particular com- plete diesel engine performance and emissions model is indicated in Fig. 14-21 : it Ignition delay is typical of the type of compression-ignition engine simulation used to study Heat release engine performance and emissions. Figure 14-22 shows an example of the output from the above model. The injection rate diagram, the assumed Sauter mean Pressure and temperature histories drop size of the spray, and the air swirl determine the spray development which leads to the heat-release rate predictions. This determines the cylinder pressure Equilibrium products Fuel consumption profile. Predicted engine performance results show reasonable but not precise agreement with experimental data. That is not surprising given the complexity of Flame configuration the phenomena being modeled. A review of these types of jet models is given by FIGURE 14-21 Hiroyasu.46 Soot formation and oxidation Structure of thermodynamic- based DI diesel simulation for NO formation by Zeldovich mechanism predicting engine performance 14.4.4 Prechamber Engine Models and emissions. Simulation incor- porates spray model of type illus- Small high-speed compression-ignition engines use an auxiliary combustion Emissions trated in Figs. 14-19 and 14-20.45 chamber, or prechamber, to achieve adequate fuel-air mixing rates. The precham- ber is connected to the main combustion chamber above the piston via a nozzle, passageway, or one or more orifices (see Secs. 1.8, 8.5, and 10.2.2). Auxiliary 10 0.8 chambers are sometimes used in spark-ignition engines, also. The plasma and flame-jet ignition systems described in Sec. 9.5.3 enclose the spark plug in a cavity ite 19 , KJ/deg or small prechamber which connects to the main chamber via one or more ori- de fices. The function of the prechamber is to increase the initial growth rate of the Pressure p, MPa flame. Combustion in the main chamber is initiated by one or more flame jets 5 0.4 emanating from the prechamber created by the ignition process and subsequent Heat release rate " energy release within the prechamber. If the mixture within the prechamber is richer than in the main chamber (due to fuel injection or a separate prechamber intake valve see Sec. 1.9) these are called stratified-charge engines. The additional phenomena which these prechambers introduce beyond 1200 rev/min 0.8H mf = 90 mg/stroke those already present in conventional chamber engines are: (1) gas flows through the nozzle or orifice between the main chamber and prechamber due to piston de › mg/deg Injection rate 0.4 motion; (2) gas flows between these chambers due to the combustion-generated dm pressure rise; (3) heat is transferred to the nozzle or passageway walls due to -40 0 40 80 120 160 these flows. The first of these phenomena results in nonuniform composition and Crank angle 0, deg temperature distributions between the main and prechamber due to gas displace- FIGURE 14-22 ment primarily during compression, and determines the nature of the flow field Fuel-injection rate, heat-release rate profile, and cylinder pressure predicted with thermodynamic- within the prechamber toward the end of compression just prior to combustion. based DI diesel simulation with spray and combustion model of type shown in Fig. 14-21.45 786 787 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES The second phenomena controls the rate of energy release in the main chamber. Heat transfer to the passageway and chamber walls is affected by the flows The heat losses in the passageway and to the additional chamber surface area of between the chambers: high velocities within the passageway result in high heat- the prechamber designs relative to conventional open chambers result in transfer rates to the passageway walls, and the vigorous flows set by the passage- decreased engine performance and efficiency. Thus the prechamber concept adds way exit flow entering the prechamber or the main chamber increase additional complexity to the engine processes that must be modeled to predict heat-transfer rates to the walls of these chambers. The standard engine heat- engine behavior. transfer correlations which relate the heat-transfer coefficient to mean flow field The following variables are important to prechamber engine performance variables via Nusselt-Reynolds number relationships (see Sec. 12.4) are normally and emissions characteristics, in addition to the design and operating variables used to describe these heat-transfer processes. The length scales are chosen to which govern single-chamber engine behavior: prechamber geometry-size, match the prechamber or main chamber or passageway dimensions. The charac- shape, flow area and shape of connecting passageway(s); prechamber location in teristic velocities in these relationships are equated with velocities which are rep- relation to main chamber geometry; geometry and timing of any auxiliary pre- resentative of the flow in each of these regions at the relevant time in the engine chamber valve; fuel metering strategy in prechamber compression-ignition or operating cycle. 50, 51 stratified-charge engine. Thermodynamic-based models have been developed and The utility of the more sophisticated of these prechamber engine per- used to examine the overall impact of these variables (see Ref. 47). Computational formance and emissions models is illustrated by the sample results shown in Fig. fluid dynamic models (see Sec. 14.5 and Fig. 8-26) have also been used to examine 14-23. This simulation of the indirect-injection compression-ignition engine's flow specific prechamber engine flow and combustion processes. and combustion processes describes, through the use of stochastic mixing models, Useful predictions of fuel, air, and residual gas distributions and the corre- the development of the fuel/air ratio distribution and fuel-energy release distribu- sponding temperature within the prechamber and main chamber can be obtained tion, and hence the development of the gas pressure and gas temperature dis- with simple gas displacement models. Only during combustion is the pressure tribution, within the prechamber and main chambers of the engine. With the difference across the nozzle or orifice sufficiently large in magnitude for its model- (nonuniform) gas composition and state defined, the models for NO formation ing to be essential; the assumption of uniform pressure during compression, the described in Sec. 11.2.1 was used to predict NO, emissions. The approaches used critical process for determining conditions just prior to combustion, introduces to describe the evolution of the prechamber, main chamber, and passageway con- little error into calculations of the flows between the chambers. Section 8.5 tents are summarized in Fig. 14-23a. develops the appropriate equations for these piston-motion driven gas displace- The cylinder contents were divided up into a large number of elements. ments. Use of the conservation equations for an open system, for total mass, fuel Pairs of elements are selected at random to undergo "turbulent mixing" inter- mass, residual gas, and energy given in Sec. 14.2, for the main chamber and the actions at a frequency related to the turbulence in each region. Rate processes- prechamber, then give the mean composition and temperature variation in each evaporation, ignition, NO formation, etc .- proceed within each element between chamber as a function of time due to this flow. Figure 8-25 illustrates the mean these mixing interactions. Figure 14-23b shows sample results. At about TC, after composition variation in the prechamber that results during the compression some of the injected fuel has evaporated and the ignition delay is over, com- stroke of a three-valve stratified-charge engine. bustion starts in the prechamber and the prechamber pressure p, rises above the During combustion, the pressure difference across the connecting passage- main chamber pressure pm. This forces air, fuel, and burned gases to flow from way or orifice is the driving force for the flow between chambers. Since com- the prechamber into the main chamber; fuel and rich products can now mix with bustion starts in the prechamber, the initial flow is into the main chamber; later, air and burn in the main chamber. NO starts to form in each mass element, once as the heat release in the main chamber becomes dominant, the flow may reverse it burns, at a rate dependent on each element's composition and state. Most of direction and be into the prechamber. In thermodynamic-based models, the the NO forms within the prechamber and then flows into the main chamber as equations for one-dimensional quasi-steady ideal gas flow through a restriction the expansion process proceeds. The attractive feature of this type of emission given in App. C are used to relate these flows to the pressure difference between calculation is that the kinetically controlled NO formation calculations are the two chambers. Open-system conservation equations are again used to calcu- based directly on local gas composition and temperature in a manner that late mean properties in each chamber. approximately simulates the mean and turbulent nonuniformities in these vari- Combustion models used are either empirically based [e.g ., using specified ables. Predictions of engine operation and emissions showed good agreement heat-release or mass burning rates such as Eq. (14.32)48] or are developed from with data.52 direct-injection compression-ignition engine models with spray evaporation, fuel- Fluid-dynamic-based models have been used to study fluid flow, com- air mixing, and ignition delay processes explicitly included.49 Because of the com- bustion, and pressure wave phenomena in prechamber engines. Section 14.5 plexity of these processes in the prechamber engine geometry, substantial reviews this type of engine model. Additional details of these applications can be simplifying assumptions and empiricism must be used. found in Refs. 53 and 54. 788 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 789 Injection: add liquid fuel elements 14.4.5 Multicylinder and Complex Engine Prechamber partially stirred reactor System Models The models discussed in the previous parts of Sec. 14.4 focus on the processes Elements transferred through passageway: occurring within each cylinder of an internal combustion engine. Most engines flow restriction with are multicylinder engines and the individual cylinders interact via the intake and heat transfer exhaust manifolds. Also, many engine systems are more complex: internal com- bustion engines can be supercharged, turbocharged, or turbocompounded, and the manifolds then connect to the atmosphere via compressors or turbines (see Fig. 6-37 and Sec. 6.8). Thermodynamic-based simulations of the relevant engine processes, constructed from the types of model components already described, prove extremely useful for examining the behavior of these more complex engine systems. By describing the mass and energy flows between individual components Main chamber: and cylinders of such systems throughout the engine's operating cycle, the total partially stirred reactor system preformance can be predicted. Such models have been used to examine Partially stirred reactors contain many equal mass elements. These elements may be air steady-state engine operation at constant load and speed (where time-varying (plus residual), liquid fuel, unburned mixture conditions in the manifolds due to individual cylinder filling and emptying events (fuel vapor, air, burned gas), and burned mixture affect multicylinder engine behavior), and how the total system responds to (a) changes in load and speed during engine transients. The block diagram of a turbocharged and turbocompounded diesel engine system in Fig. 14-24 illustrates the interactions between the system components. By describing the mass and energy flows between components and the heat and 80 1900 70- 1500 rev/min Tp work transfers within each component, total system behavior can be studied. In Pp 1700 60- ¢ = 0.53 Pm 1500 such engine simulations, the reciprocator cylinders, the intake manifold, and the 50 1300 various sections of the exhaust system are treated as connected open systems. 40 1100 Temperature, K Pressure, atm 30 900 The flows into and out of these volumes are usually analyzed using the quasi- 20 700 FIGURE 14-23 steady emptying and filling approach described in Sec. 14.3.3, using the open- 10 TC 500 (a) Schematic of IDI diesel system conservation equations of Sec. 14.2. The reciprocator cycle is treated as a 0 engine illustrating how stochastic sequence of processes within each cylinder: intake, compression, combustion 100 100 mixing models are applied to 80 Injection 80 prechamber, main chamber, and (including expansion), and exhaust. These are modeled using the approaches Evaporation Burned, total passageway to simulate turbulent described previously in Secs. 14.4.1 to 14.4.4. Heat transfer has, of course, an 60+ Pre 60 mixing processes and pressure- important effect on the in-cylinder processes. It also is important in the exhaust Fuel mass fraction, % burned 40} Main 40 driven flows. (b) Example of system since the performance of the turbocharger turbine and of any com- 20H 20 simulation predictions through pounded turbine depends on the gas state at the turbine inlet. The performance IC the engine's operating cycle. 0 0 of the turbomachinery components is normally defined by maps that interrelate Shown are prechamber and main Mass weighted NO ,, ppm injected, evaporated, and 800 200 chamber pressures, prechamber efficiency, pressure ratio, mass flow rate, and shaft speed for each component (see Total and main chamber average gas Secs. 6.8.2 to 6.8.4). Special provisions are usually required in the logic of the 600- 150 temperatures; fuel mass injected, turbomachinery map interpolation routines to avoid problems with the compres- --- Pre NO ,, ppm 400 Main evaporated, and burned in pre- 100 sor surge and turbine choking operating limits of these devices. NOx, ppm Mass weighted NO, chamber and main chamber, When the reciprocator is coupled with turbomachinery its manifolds no 200 - 50 average NO concentration in each chamber (and total) in ppm longer connect directly with the atmosphere: matching procedures are required -20 20 by volume and mass weighted to ensure that the pressure levels and mass flow rates of the compressor and 40 60 80 ppm (mass in chamber x NO turbines match with those of the engine. The following matching process is Crank angle, deg concentration in chamber/total typical of those used for turbocharged engines (one compressor and one turbine (b) mass in cylinder). 52 only). At a given time, the values of the variables describing the state of the 790 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 791 - Latm rotor speed according to the turbocharger dynamics equation Parum .@ w + Bw2 Wc + W = ITCw dt (14.51) Turbocharger where ITc is the rotational inertia of the turbocharger, @ is angular velocity, and B is the rotational damping. The values of the other state variables for the next 8 9 time step are determined from the solution of the mass and energy conservation 2 Ap equations for each open system, with the compressor and turbine mass flows Inter Wastegate taken from the output of the turbomachinery map interpolation routines. cooler Fuel system This approach can be used to establish the steady-state engine operating 3 characteristics from an assumed initial set of state variables. (Of course, due to the pulsating nature of the flows into and out of the cylinders, these state vari- 4 6 ables will vary in a periodic fashion throughout the engine cycle at a fixed engine load and speed.) This approach can also be used to follow transient engine Intake Multicylinder Exhaust manifold diesel engine manifold 10 behavior as load or speed is varied from such a steady-state condition.35 The additional inputs required are the fuel pump delivery characteristics as a function of fuel pump rack position and speed, with the latter evaluated from an appropri- ate model for dynamic behavior of the governor.55 From the brake torque of the Engine engine (determined by subtracting friction torque from the indicated torque), the friction torque required by the load TL, the inertia of the engine and load IF and IL, the dynamic response of the engine and load to changing fuel rate or engine speed Compound turbine can be obtained from dw (14.52) 11 Patm An example of the output from this type of engine model is shown in Fig. 14-25. FIGURE 14-24 Block diagram of turbocharged turbocompounded diesel engine system. The response of a turbocharged DI diesel engine to an increase in load from 0 to 95 percent of full load is shown. The predictions come from a model of the type shown in Fig. 14-24, and engine details correspond to the experimental configu- ration.55 The simulation follows the data through the engine transient with rea- various system components are known (from integration of the system governing sonable accuracy. Note that with the assumed constant governor setting, during equations over the previous time step). These include the intake and exhaust this transient the equivalence ratio of the trapped mixture rises to close to stoi- manifold pressures and the turbocharger rotor speed. The compressor inlet pres- chiometric because the increase in air flow lags the increase in fuel flow. This sure is atmospheric pressure less the intake air-filter pressure drop. The turbine would result in excessive smoke emissions. Such models prove extremely useful exit pressure is atmospheric plus the muffler pressure drop. By relating the com- for exploring the effect of changes in engine system design on transient pressor discharge pressure to the intake manifold pressure and the turbine inlet response. 56 pressure to the exhaust manifold pressure (through suitable pressure drops) the For two-staged turbocharged or turbocompounded systems the engine- pressure ratio across each machine is determined. Hence, the compressor and turbocharger matching process is more complicated. The division of the pressure turbine maps can be entered using the calculated pressure ratios, and the rotor ratio between the exhaust manifold and atmosphere between the two turbines in speed (same for both turbomachines) as inputs. The output from the map inter- Fig. 14-24 is not known a priori. Nor, with two compressors, is the intake pres- polation routines determines the mass flow rate and efficiency of each component sure ratio distribution known. Iterative procedures based on an assumed mass for the next time step. From these the power required to drive the compressor flow rate are used to determine the pressure level between the two turbines such (-Wc) and to drive the turbine (Wr) are determined from Eqs. (6.42) and (6.48), that mass flow and pressure continuity through the exhaust system is satisfied respectively. Any excess power (or power deficiency) will result in a change of (e.g ., Ref. 2). 792 MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 793 INTERNAL COMBUSTION ENGINE FUNDAMENTALS TABLE 14.2 1.500 200F Available energy equations for various processes 1000- 160 kPa bmep, Measured KN/m2 500 Predicted 120 Mechanism Equation Air flow Boost pressure, 0 -120 80 -1.0 Work transfer 2000 0.2 0.75 dAw = dW Heat transfer 1750 Number of cycles 60 -0.5 dAQ = dQ(1 - To/T) 0.15 Trapped kg/s rev/min Gas transfer 0.25 dA. = dm.[(h - ho) - To(s - so)] 1500 0.14 O Liquid fuel transfert dA, = dm (1.03382LHV) O 10- Control volume storage dAcv = d(mcv[(u - up) - To(s - So) + Po(v - vo)]} 60₸ press, MPa Max cyl 7.51 Fuel rack Turbo speed, Engine speed, The availability of the fuel is 1.0338 times its lower heating value; see Sec. 5.7. 40- 5 1100 position 103 rev/min Maximum 900 bustion engine processes has already been developed in Sec. 5.7. The change in Turbine inlet Maximum fueling temp, K 700 - Minimum availability of any system undergoing any process where work, heat, and mass 500 transfers across the system boundary occur (see Fig. 5.13) can be written: 6 8 O 2 Time, s Time, s AA = Ain - Aout - Adestroyed (14.53) FIGURE 14-25 where Ain and Aout represent the availability transfers into and out of the system Predicted ( --- ) and measured (-) response of a turbocharged direct-injection diesel engine to an across the boundary. Since availability is not a conserved quantity, this equation increase in load.55 can only be used to solve for the availability destruction term, Adestroyed . Table 14.2 summarizes the equations for the availability change of the system and the 14.4.6 Second Law Analysis of Engine Processes availability transfers associated with work, heat and mass transfer across the The first-law-based methods for evaluating power plant performances do not system boundary, developed in Sec. 5.7. explicitly identify those processes within the engine system that cause unre- This availability balance is applied to the internal combustion engine oper- coverable degradation of the thermodynamic state of the working fluid. However, ating cycle as follows. A first-law-based cycle analysis of the type described above second-law-based analysis methods do provide the capability to identify and in this section (14.4) is used to define the variation in working fluid thermody- quantify this unrecoverable state degradation. Thus, cause and effect relation- namic state, and the work, heat, and mass transfers that occur in each of the ships which relate these losses to individual engine processes can be determined. processes that make up the total engine cycle. Integration of the availability The first law analysis approaches summarized in this section (14.4) are based on balance over the duration of each process then defines the magnitude of the the fact that energy is conserved in every device and process. Thus, they take availability destruction that occurs during that process. account of the conversion of energy from one form to another: e.g ., chemical, To illustrate this procedure, consider the operating cycle of a 10-liter six- thermal, mechanical. Although energy is conserved, second law analysis indicates cylinder turbocharged and aftercooled direct-injection four-stroke cycle that various forms of energy have differing levels of ability to do useful mechani- compression-ignition engine, operating at its rated power and speed of 224 KW cal work. This ability to perform useful mechanical work is defined as availability. and 2100 rev/min. The variations in temperature, energy, and entropy are deter- The availability of a system at a given state is defined as the amount of mined with a first-law-based analysis. Figure 14-26 shows the T-s diagram for the useful work that could be obtained from the combination of the system and its working fluid as it goes through the sequence of processes from air inlet from the surrounding atmosphere, as the system goes through reversible processes to atmosphere (state 1) to exhaust gas exit to the atmosphere (state 10).57 The equilibrate with the atmosphere. It is a property of the system and the incoming air is compressed (with some irreversibility) in the turbocompressor to environment with which the system interacts, and its value depends on both the state 2 and cooled with an aftercooler to state 3. The air at state 3 is drawn into state of the system and the properties of the atmosphere. Availability is not a the cylinder and mixed (irreversibly) with residual gases until, at the end of the conserved property; availability is destroyed by irreversibilities in any process the intake, the cylinder gases are represented by state 4. That mixture is subsequently system undergoes. When availability destruction occurs, the potential for the compressed (with modest heat loss) to state 5. Fuel addition commences close to system to do useful mechanical work is permanently decreased. Thus to make a state 5; subsequent burning increases the combustion chamber pressure and tem- proper evaluation of the processes occurring within an engine system both energy perature along the line 5-6. At 6 the heat release, heat transfer, and volume and availability must be considered concurrently. change rate are such that the maximum cylinder pressure is reached (a few The basis for an availability analysis of realistic models of internal com- degrees after TC). From 6 to 7 combustion continues to completion, the burned 794 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 795 2000 TABLE 14.3 Comparison of first and second law analysis for six- 1600- 6 cylinder 14-liter naturally aspirated and turbocharged diesel engine at 2100 rev/min58 1200- FIGURE 14-26 T Temperature, K T-s diagram for the working fluid as it Naturally aspirated Turbocharged 800- goes through the sequence of processes - from air inlet to exhaust in turbo- First law, Second law, charged aftercooled DI compression- % fuel % fuel 400- ignition engine. The 10-liter six-cylinder energy availability engine is operated at its rated power (224 kW) and speed (2100 rev/min). The Indicated workt 40.3 39.1 43.9 OL -0. 0.0 0.1 0.2 0.3 0.4 text relates processes to numbered end Combustion loss 15.9 19.2 Entropy, kcal/kg - K states. 57 Cylinder heat transfer 25.1 21.4 17.6 Internal valve throttling 0.7 0.7 Exhaust valve throttling 2.5 2.3 Loss in compressor 1.4 gases continue to expand, doing work on the piston and losing heat to the walls. Loss in turbine 0.8 At state 7 the exhaust valve opens initiating a rapid pressure equilibration with Exhaust to ambient 34.6 20.4 14.1 the exhaust manifold to a pressure corresponding to point 8. Gases are expelled Total 100.0 100.0 100.0 from the cylinder to the exhaust manifold. After the intake valve opens, cylinder Brake power, kw 185 185 220 residual gases are mixed with incoming air at state 3 to yield gases at state 4 to + Note that the indicated work for the second law balance is a lower percentage complete the in-cylinder cycle. The exhaust gases that have been expelled from than for the first law. This occurs because the availability of the fuel is 1.0317 the cylinder experience additional thermodynamic losses and can be represented times the fuel's heating value. by state 9. These gases then pass through the turbocharger turbine to state 10 to provide the work to drive the compressor. A first law and second law analysis of a naturally aspirated diesel engine are The quantity referred to as combustion loss in Table 14.3 is determined compared in Table 14.3. Also shown is a second law analysis of a turbocharged from an availability balance for the combustion chamber over the duration of the version of the naturally aspirated diesel. These results illustrate the value of defin- combustion period. The "availability destroyed" term in Eq. (14.53) then rep- ing the losses in availability that occur in each process. resents the deviation of the actual combustion process from a completely Consider the first and second law analysis results for the naturally aspirated reversible process. The second law analysis shows that the availability loss associ- engine. While 25.1 percent of the fuel energy leaves the combustion chamber in ated with the combustion irreversibilities is 15.9 percent of the fuel's availability. the form of heat transfer, the availability transfer corresponds to 21.4 percent of This loss depends on the overall equivalence ratio at which the engine is oper- the fuel's availability. It is this latter number that indicates the maximum amount ating, as indicated in Fig. 5-17. Combustion of leaner air/fuel ratios would give a of the heat transfer that can be converted to work. The table shows that 34.6 higher fractional availability loss due to mixing of the fuel combustion products percent of the fuel energy is carried out of the engine in the exhaust gases. with increased amounts of excess air and the lower bulk temperature. However, the second law analysis shows that the exhaust contains only 20.4 Overall, the most important point emerging from this comparison is that percent of the available energy of the fuel. The ratio of these quantities shows the work-producing potential of the heat loss to the combustion chamber walls that only about 60 percent of the exhaust energy can be converted to work using and the exhaust mass flow out of the engine is not as large as the magnitude of ideal thermodynamic devices.| The exhaust gas leaves the system in a high- the energy transferred: some of these energy transfers, even with ideal thermody- temperature, ambient pressure state and therefore has high entropy (relative to namic work-producing devices, must ultimately be rejected to the environment as the po, To reference state). This, via the gas-transfer equation in Table 14.2, heat. reduces the available energy of the exhaust gas stream. A comparison of the second and third columns in Table 14.3, both obtained with a combined first and second law analysis, illustrates how turbocharging improves the performance of a naturally aspirated engine. The brake fuel conver- sion efficiency of the turbocharged engine is considerably improved-from 33.9 Of course, real thermodynamic devices will produce less work than ideal devices. to 39.2 percent. The table indicates that through turbocharging, the availability transfers associated with the heat loss and exhaust gas flow are reduced from 41.8 796 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 797 100 to 31.7 percent (a difference of 10.1 percentage points), while the combustion and Total heat transfer added turbomachinery availability losses increase from 15.9 to 21.4 percent (a 80 difference of 5.5 percentage points). By turbocharging, advantage has been taken Combustion losses of the following changes. While the leaner air/fuel ratio operation of the turbo- 60 ---- Exhaust to ambient Available energy, % fuel charged engine increases the combustion availability losses due to the use of a -------- 40 ........................................... greater portion of the chemical energy of the fuel to mix with and heat excess air, Mechanical friction Fluid flow losses the lower burned gas temperature this produces results in reduced heat losses and 20 Brake work lower cylinder exhaust temperature. In addition, the turbocharger transfers avail- FIGURE 14-28 able energy from the cylinder exhaust to the inlet air. The reduced heat loss and Distribution of available energy into 1300 1500 1700 1900 2100 major categories for the engine of Fig. lower final exhaust availability level give a substantial performance improve- Engine speed, rev/min 14-26 as a function of engine speed.57 ment. 58 To interpret the second law analysis results, one must remember that the desired output is brake work and increases in this quantity (for a given fuel flow) represent improved performance. All other availability terms represent losses or little over the load range. The combustion loss increases from 21.8 to 32.5 percent undesirable transfers from the system; decreasing these terms constitutes an as load is decreased due to an increasingly lean operation of the engine. Fluid improvement. These undesirable available energy transfer and destruction terms friction losses, as a percentage, increase slightly as load increases due to larger fall into five categories: (1) heat transfer, (2) combustion, (3) fluid flow, (4) exhaust mass flow rates. Since friction is approximately constant in absolute magnitude, to ambient, (5) mechanical friction. The available energy flows identified as heat its relative importance increases drastically as the brake output goes to zero. transfer represent the summation of all availability transfers that occur due to Exhaust flow available energy decreases from 12.2 to 8 percent as load is heat transfers. The most significant of these are the in-cylinder and aftercooler decreased from 100 to 0 percent.57 heat rejection. The combustion loss represents the amount of available energy The effect of varying engine speed (at full load) is shown in Fig. 14-28. The destroyed due to irreversibilities occurring in releasing the chemical potential of availability associated with heat transfers changes over the speed range shown the fuel as thermal energy and mixing the combustion products with any excess from 15.6 to 21 percent : more time during each cycle is available for heat transfer air. The fluid flow losses include the available energy destroyed within the at lower speeds. Fluid flow and friction losses decrease with decreasing speed. working fluid in the compressor, aftercooler, intake valve, exhaust valve, exhaust Other availability losses remain essentially constant as a percentage of the fuel's manifold, and turbine due to fluid shear and throttling. The availability availability.57 destroyed due to fluid shear and mechanical rubbing, exterior to the working fluid, are contained in the mechanical friction category. The effect of variations in engine load and speed on these five categories of losses or transfers will now be 14.5 FLUID-MECHANIC-BASED described. MULTIDIMENSIONAL MODELS Figure 14-27 shows the availability transfers or losses in each of these cate- gories for a turbocharged six-cylinder 10-liter displacement direct-injection diesel 14.5.1 Basic Approach and Governing Equations engine, expressed as a percentage of the fuel availability, as a function of engine load. The percentage of fuel availability associated with the heat transfers varies The prediction of the details of the flow field within engines, and the heat-transfer and combustion processes that depend on those flow fields, by numerical solution of the governing conservation equations has become a realizable goal. Such 100 Total heat transfer methods have been under development for more than a decade, during which 80 time they have steadily improved their ability to analyze the flow field in realistic Combustion losses engine geometries. While the overall dynamic characteristics of intake and ---- 60 Exhaust to ambient exhaust flows can usefully be studied with one-dimensional unsteady fluid Available energy, % fuel -------- ------------------ --------- 40 dynamic computer calculations (see Sec. 14.3.4), flows within the cylinder and in Mechanical Fluid flow losses 20 losses intake and exhaust ports are usually inherently unsteady and three dimensional. Brake work FIGURE 14-27 Recent increases in computing power, coupled with encouraging results with two- Distribution of available energy into dimensional calculations, indicate that useful three-dimensional calculations are 20 40 60 80 100 major categories for the engine of Fig. Load, % max 14-26 as a function of engine load.57 now feasible. However, they still do not have the capability to predict accurately 798 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 799 all the features of real engine processes of interest. Gas-flow patterns can be pre- The first term on the right gives the source terms, the second term the diffusive dicted best; predictions of fuel spray behavior are less complete, and combustion transport. The D/Dt operator provides the convective transport terms and is calculations present considerable difficulties. These computational, fluid dynamic, engine process analysis codes solve the Df _ @(pf) , (ou; f) Dt Ox ; (14.55) partial differential equations for conservation of mass, momentum, energy, and species concentrations. To apply a digital computer to the solution of a contin- Here, p is the density, u, the ith velocity component, e the internal energy per unit uum problem (such as the flow field inside the cylinder), the continuum must be mass, and Y, the concentration of species a per unit mass. represented by a finite number of discrete elements. The most common method of In the IC engine context, the thermal energy source term Q involves a discretization is to divide the region of interest into a number of small zones or viscous term and source terms arising from chemical reaction of the fuel. Both Q cells. These cells form a grid or mesh which serves as a framework for construc- and the species source term, Sa, will depend upon the chemical rate equations, ting finite volume approximations to the governing partial differential equations. which must be known to close the problem. Note that diffusion of the various The time variable is similarly discretized into a sequence of small time intervals species contributes to the diffusive flow of internal energy, q;, in addition to called time steps, and the transient solution is "marched out" in time: the solu- conductive heat diffusion. tion at time t,+1 is calculated from the known solution at time t ,. Three- The fact that turbulent flows exhibit important spatial and temporal varia- dimensional formulations of the finite difference equations are required for most tions over a range of scales (dictated at the upper end by chamber dimensions practical engine calculations; two-dimensional (or axisymmetric) formulations and at the lower end by viscous dissipative processes, see Sec. 8.2.1) makes direct can be useful, however, under simpler flow situations, and have been more exten- numerical solution of these governing equations impractical for flows of engine sively used to date due to their simpler models and computer codes and require- complexity. Recourse must therefore be made to some form of averaging or filter- ment for less computer time and storage capacity. ing which removes the need for direct calculation of the small-scale motions. Two The principal components of these multidimensional engine flow models approaches have been developed for dealing with this turbulence modeling are the following:59 problem: full-field modeling (FFM), sometimes called statistical flux modeling; and large-eddy simulation (LES) or subgrid-scale simulation. In FFM, one works 1. The mathematical models or equations used to describe the flow processes. with the partial differential equations describing suitably averaged quantities, Especially important is the turbulence model, which describes the small-scale using the same equations everywhere in the flow. For periodic engine flows, time features of the flow which are not accessible to direct calculation. averaging must be replaced by ensemble or phase averaging (see Sec. 8.2.1). The variables include the velocity field, thermodynamic state variables, and various 2. The discretization procedures used to transform the differential equations of mean turbulence parameters such as the turbulent kinetic energy, the turbulent the mathematical model into algebraic relations between discrete values of stress tensor, etc. In FFM, models are needed for various averages of the turbu- velocity, pressure, temperature, etc ., located on a computing mesh which lence quantities. These models must include the contributions of all scales of (ideally) conforms to the geometry of the combustion chamber with its moving turbulent motion. 59, 60 valves and piston. Large-eddy simulation (LES) is an approach in which one actually calcu- 3. The solution algorithm whose function is to solve the algebraic equations. lates the large-scale three-dimensional time-dependent turbulence structure in a 4. The computer codes which translate the numerical algorithm into computer single realization of the flow. Thus, only the small-scale turbulence need be language and also provide easy interfaces for the input and output of informa- modeled. Since the small-scale turbulence structure is more isotropic than the tion. large-scale structure and responds rapidly to changes in the large-scale flow field, modeling of the statistical fluxes associated with the small-scale motions is a The basic equations for all existing in-cylinder flow calculation methods are the simpler task than that faced in FFM where the large-scale turbulence must be differential equations expressing the conservation laws of mass, momentum (the included. Navier-Stokes equations-a set of three), energy, and species concentrations. An important difference between FFM and LES is their definition of These equations, in the above order, may be written: "turbulence." In FFM the turbulence is the deviation of the flow at any instant from the average over many cycles of the flow at the same point in space and 0 oscillation phase [i.e ., the fluctuation velocity defined by Eq. (8.16) or (8.18)]. D tij Thus, FFM "turbulence" contains some contribution from cycle-by-cycle flow Dt = (14.54) e Ox; variations. LES defines turbulence in terms of variations about a local average; hence in LES turbulence is related to events in the current cycle.60 800 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 801 14.5.2 Turbulence Models perature, {0} will be strongly influenced by temperature fluctuations. These In full-field modeling (FFM), equations for the averaged variables are formed issues are discussed more fully in Secs. 14.5.5 and 14.5.6. from Eqs. (14.54). With periodic engine flows, phase or ensemble averaging must The momentum equations contain terms, - puqu', which represent turbu- be used (see Secs. 8.2.1 and 8.2.2). Since the flow during the engine cycle is com- lent stresses (and are often called the Reynolds stresses). These terms must be pressed and expanded, mass-weighted averaging (called Favre averaging) can be modeled with additional equations before the set of equations, (14.59), is used to make the averaged compressible-flow equations look almost exactly like "closed" and can be solved. The most widely used turbulence model or equation the averaged equations for incompressible flows. The combined ensemble-Favre set is the k-e model.60-63 This assumes a newtonian relationship between the averaging approach works as follows.60 turbulent stresses and mean strain rates, and computes the (fictitious) turbulent We denote the phase-averaging process by { }, i.e.: viscosity appearing in this relationship from the local turbulent kinetic energy k (= ui ui/2) and its dissipation rate &. An equation governing k can be developed by {p(x, t)} = lim [ p(x, t + nt) (14.56) multiplying the u equation in Eq. (14.54) by ut, subtracting from this the equa- N + 00 N n= 1 tion formed by multiplying the u equation in Eq. (14.59) by u;, and phase- where t is the cycle period. We also write {} = p, and decompose o into averaging the result. The equation so obtained is p = p +p'. The mass-weighted phase-averaged quantities (indicated by an DK overbar) are defined by Dt = P(P - 8) - ax; (14.61a) P(x, t)f(x, t) = lim [ p(x, t + nt)f(x, t+ nt) (14.57) where P is the rate of turbulence production per unit mass N + 00 n = 1 where all flow variables (except density and pressure) have been decomposed as P = - uil'; ax; (14.61b) f =f+ f. Note that {p'} is zero, {f } = f, the mass-weighted phase average of f is zero, but {f'} is not zero. With these definitions: and Jk represents diffusive transport. {pf} = Pf In the most commonly used two-equation k-e model, all the unknown tur- bulence quantities are modeled in terms of the turbulent velocity scale k1/2 and {pf} = 0 the turbulence length scale k3/2/8 obtained from the definition of the energy dissi- {ofg) = P(g + fg) pation rate, via {ofgh} = PUgh + jgh + gph + hig + igh) (14.58) k3/2 E OC - (14.62) Phase-averaging Eq. (14.54), one obtains60 O 0 The rationale is that the rate of energy dissipation is controlled by the rate at which the large eddies feed energy to the smaller dissipative scales which in turn u pulu'; = ax ; ax (14.59) adjust to handle this energy.60 pe'u'j A turbulent viscosity ur is defined: Co pk2 where UT (14.63) a (pu, f) (14.60) Dt ôt - (pf ) +- ax ; where Co is a model constant. The turbulent stress terms appearing in Eqs. (14.59) and (14.61) are then modeled in a quasi-newtonian manner : The terms on the left-hand side in Eq. (14.59) involve only the solution variables p, ui, e, and Y'a, and hence require no modeling. However, all of the terms on the puju, = 3pkoy + 3MTV.uby - ZurSy (14.64) right, particularly the last terms that represent turbulent transport, involve turbu- where Sij is the strain rate of the u, field: lence fluctuation quantities and must be modeled in terms of the solution vari- ables. The source terms {0} and {S.} present special difficulties to the engine modeler. Due to the exponential dependence of the heat release Q on tem- Su = ! ( oui + axi (14.65) 802 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 803 The viscous-stress terms in the momentum equations are evaluated using a new- tains terms for production and decay of q and for its convection and diffusion. In tonian constitutive relation. The turbulent-diffusion terms in the various trans- the KIVA engine code,67 this equation has the form: port equations are modeled using the turbulent diffusivity. The diffusing flux of a quantity o is given by at (pa) + V . (oqu) = - 3pgV . a + o : Vu+ V. (uvq) - CpL-1q3/2 + W3 (14.66) (14.69) of @x ; where ou is a turbulent Prandtl number for o. where o is the turbulent stress tensor, u the turbulent viscosity, C a constant of The model is completed with a transport equation for a. An exact equation order unity, and L a characteristic length on the order of twice the mesh spacing. can be developed by suitable manipulation of the Navier-Stokes equations. All & We is a source term representing the production of turbulence by the motion of equation models are of the form60 fuel droplets in situations where fuel sprays are important. The physical meaning of the terms in Eq. (14.69) are as follows. The term De Dt W - (14.67) V . (pqu) is the convection of the turbulence by the resolved (large-scale) velocity axi field. The term - 3pqV . u is a compressibility term that is the turbulent analog of p dV work. The term a : Vu represents the production of turbulence by shear where W is the source term and H; is the diffusive flux of pe which is modeled in the resolved velocity field; V . (uVq) is the self-diffusion of the turbulence with similarly to the other diffusion terms. The appropriate form of W is the subject of diffusivity u/p. The term -CpL-1q3/2 represents the decay of turbulent energy much debate. For an incompressible flow, W can be modeled adequately by into thermal energy. This term appears with opposite sign as a source term in the thermal internal energy equation in place of o : Vu, which can be thought of as w =( - C2+ C1 2 ) DE2 K (14.68) the rate at which kinetic energy of the resolved motions is dissipated by the turbulence. Before it is dissipated, the kinetic energy of the resolved velocity field C1 and C2 are constants: the C2 term produces the proper behavior of homoge- is first converted into subgrid scale turbulent energy q, which is then converted neous isotropic turbulence and the C1 term modifies this behavior for homoge- into heat by the decay term CpL-1q3/2,67 neous shear. However, for a flow with compression and expansion, an additional Under most circumstances, the velocity and temperature boundary layers in term in Eq. (14.68) is needed to account for changes in & produced by dilation. an engine cylinder will be too thin to be resolved explicitly with a computing Several forms for this additional term have been proposed60,62 (for example, mesh that is practical on present-day computers. However, these layers are C3 p&V . i) and compared.63 The goal is to construct a W that predicts the appro- important because they determine the wall shear and heat flux which are essential priate physical behavior under the relevant engine conditions. While different boundary conditions for the numerical simulation, and are of practical impor- choices for modeling these terms do affect the results (especially the behavior of tance (see Secs. 8.3 and 12.6.5). Special submodels for these boundary layers, the turbulence length scale during the cycle62), the predictions of mean flow and referred to as wall functions, are used to connect the wall shear stresses, heat turbulence intensity do not differ very significantly.63 fluxes, wall temperatures, etc ., to conditions at the outer edge of the boundary One other FFM that has been applied to engines is the Reynolds stress layer. This removes the need to place grid points within the layer. Since the model (RSM) which, in its most general form, comprises seven simultaneous boundary layers are usually turbulent, the logarithmic "turbulent law of the partial differential equations for the six stress components and the dissipation wall" is commonly used. Key assumptions made are: that the finite difference rate &. This obviously imposes a much greater computing burden compared with mesh point nearest the wall lies in the law-of-the-wall region and that the law-of- the two-equation k-e model. The limited results available64 indicate that RSM the-wall relation for steady flow past a plane wall is valid under engine cylinder predictions of the flow field are closer to corresponding measured data than k-8 conditions. While these may not be valid assumptions, it is not yet feasible to model predictions.65 resolve the flow details within the boundary layer.68 The large-eddy simulation (LES) approach to turbulence modeling66 has also been applied to engines. Since here one calculates the large-scale three- 14.5.3 Numerical Methodology dimensional time-dependent flow structure directly, only the turbulence smaller in scale than the grid size need be modeled. Hence these are often referred to as The three important numerical features of multidimensional methods are: the subgrid-scale models. A new dependent variable q, which represents the kinetic computational grid arrangement, which defines the number and positions of the energy per unit mass of the turbulent length scales that are too small to resolve in locations at which the flow parameters are to be calculated; the discretization the mesh, is introduced. This variable satisfies a transport equation which con- practices used to transform the differential equations of the mathematical model 804 MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 805 INTERNAL COMBUSTION ENGINE FUNDAMENTALS into algebraic equations; and the solution algorithms employed to obtain the flow parameters from the discrete equations. 59, 65 COMPUTING MESH. The requirements of the computing mesh are: 1. It adequately fits the topography of the combustion chamber and/or inlet port, including the moving components. 2. It allows control of local resolution to obtain the maximum accuracy with a given number of grid points. 3. It has the property that each interior grid point is connected to the same number of neighboring points. (a) The first requirement obviously follows from the need to simulate the effects of changes in engine geometry. The second requirement stems from the fact that NET computing time increases at least linearly with the number of mesh points. Thus it is desirable that the mesh allow concentration of grid points in regions where steep gradients exist such as jets and boundary layers. The third requirement comes from the need for the mesh to be topologically rectangular in some trans- formed space so that highly efficient equation solvers for such mesh systems can be utilized. Early engine models used a grid defined by the coordinate surfaces of a cylindrical-polar frame. Such an approach is adequate provided the combustion chamber walls also coincide with coordinate surfaces. This only occurs for a restricted number of practical chambers (e.g ., disc and centered cylindrical bowl- in-piston shapes); even for these, the inlet and exhaust valve circumferences Plan view Elevation view would in general cut across the grid (see Fig. 14-29a). While procedures have (b) been devised for modifying the difference equations for such grids to allow for noncoincident boundaries, the preferred approach is to employ some form of flexible "body-fitting" coordinate frame/grid whose surfaces can be shaped to the chamber geometry, as illustrated in Fig. 14-29b, which shows a diesel engine combustion chamber fitted by a mesh which is orthogonal-curvilinear in the bowl. This enables the bowl shape to be accurately represented and the boundary layers on its surfaces to be resolved in greater detail. The region between the piston crown and cylinder head surfaces is fitted with a bipolar system which expands and contracts axially to accommodate the piston motion. The orthog- onality constraint of this mesh limits its usefulness: the generation of orthogonal meshes for general three-dimensional geometries is cumbersome and the resulting mesh often far from optimal. These problems are largely surmounted by arbitrary " nonorthogonal lagrangian-eulerian meshes like that used in KIVA,67 illustrated in Fig. 14-29c. This has the additional advantage that the mesh points (c) in the swept volume are not constrained to move axially: their motion can be FIGURE 14-29 arbitrarily prescribed. 59 Different types of computing mesh arrangements for engine combustion chambers. (a) Cylindrical polar mesh: dashed line shows valve head circumference.65 (b) Body-fitted orthogonal curvilinear DISCRETIZATION PRACTICES. These multidimensional engine flow models are mesh fitted to DI diesel combustion chamber bowl.69 (c) Arbitrary nonorthogonal lagrangian-eulerian time-marching programs that solve finite difference approximations to the gov- (ALE) mesh for offset diesel combustion bowl.67 806 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 807 erning differential equations. The individual cells formed by the mesh or grid needed to solve the implicit (simultaneous) system of equations at each time step. serve as the spatial framework for constructing these algebraic finite difference This solution is usually performed by iterative techniques. equations. The time variable is similarly discretized into a sequence of small time The computing time requirements of these two approaches scale with the intervals called time steps: the solution at time t,+1 is calculated from the known number of equations n and the number of mesh points m, as follows. For explicit solution at time t ,. The spatial differencing is made conservative wherever pos- methods, computing time scales as nm, but the time step is limited by the stability sible. The procedure used is to difference the basic equations in integral form, condition as summarized above. For implicit methods, computing time scales as with the volume of a typical cell used as the control volume and the divergence nom and At is only limited by accuracy considerations. terms transformed into surface integrals using the divergence theorem.67 One procedure used, a semi-implicit method, is the acoustic subcycling The discretized equations for any dependent variable o are of the general method. All terms in the governing equations that are not associated with sound form: waves are explicitly advanced with a larger time step At similar to that used with implicit methods. The terms associated with acoustic waves (the compression (14.70) terms in the continuity and energy equations and the pressure gradient in the momentum equation) are explicitly advanced using a smaller time step ot that where the A's are coefficients expressing the combined influences of convection satisfies the sound-speed stability criterion [Eq. (14.23)], and of which the main and diffusion, S. , V, is the source integral over the cell volume V ,, the subscript time step is an integral multiple. While this method works well in many IC p denotes a typical node point in the mesh, the summation __ is over its (six) engine applications where the Mach number is not unduly low, it is unsuitable nearest neighbors, and the superscripts i + 1 and i denote "new" and "old" for very low Mach number flows since the number of subcycles (At/ôt) tends to values, at times t + ot and t, respectively, where ôt is the size of the time step.69 infinity as the Mach number tends to zero. For values of At/ôt greater than 50 an Until recently all methods involved similar spatial approximations to calcu- implicit scheme becomes more efficient. Pressure gradient scaling can be used to late convective and diffusive transport, using a blend of first-order upwind differ- extend the method to lower Mach numbers. The Mach number is artificially encing for the former and second-order central differencing for the latter. increased to a larger value (but still small in an absolute sense) by multiplying the Unfortunately, all discretization practices introduce inaccuracies of some kind, pressure gradient in the momentum equation by a time-dependent scaling factor and the standard first-order upwind scheme produces spatial diffusion errors 1/a(t)2, where a(t) > 1. This reduces the effective sound speed by the factor a. This which act in the same way as real diffusion to "smooth" the solutions. The does not significantly affect the accuracy of the solution because the pressure magnitude of the numerical diffusion reduces as the mesh density is increased, gradient in low Mach number flows is effectively determined by the flow field and but even with as many as 50 mesh points in each coordinate direction, the effect not vice versa. Coupling pressure gradient scaling with acoustic subcycling is not eliminated. A recent development has been the introduction of "higher reduces the number of subcycles by a.67 order" spatial approximations which, in the past, had a tendency to produce The implicit equations that result from forward differencing consist of spurious extrema. This problem has been overcome by the use of "flux blending simultaneous sets for all variables and thus require more elaborate methods of techniques. First-order upwind and higher-order approximations are blended in solution. However, they contain no intrinsic stability constraints. Fully iterative appropriate proportions to eliminate the overshoots of the latter. Even with these solution algorithms for solution of these equation sets are being replaced with schemes, however, true mesh-independent solutions could not be achieved with more efficient simultaneous linear equation solvers.65 densities of up to 50 nodes in each coordinate direction; so there is still a need for further improvement. 65 14.5.4 Flow Field Predictions SOLUTION ALGORITHMS. Numerical calculations of compressible flows are To illustrate the potential for multidimensional modeling of IC engine flows, inefficient at low Mach numbers because of the wide disparity between the time examples of the output from such calculations will now be reviewed. A large scales associated with convection and with the propagation of sound waves. amount of information on many fluid flow and state variables is generated with While all methods use first-order temporal discretization and are therefore of each calculation, and the processing, organization, and presentation of this infor- comparable accuracy, they differ in whether forward or backward differencing is mation are tasks of comparable scope to its generation! Flow field results are employed in the transport equations leading to implicit or explicit discrete equa- usually presented in terms of the gas velocity vectors at each grid point of the tions, respectively. In explicit schemes, this inefficiency occurs because the time mesh in appropriately selected planes. Arrows are usually used to indicate the steps needed to satisfy the sound-speed stability condition are much smaller than direction and magnitude (by length) of each vector. Examples of such plots-of those needed to satisfy the convective stability condition alone. In implicit the flow pattern in the cylinder during the intake process-are shown in Fig. 14- schemes, the inefficiency manifests itself in the additional computational labor 30.70 The flow field is shown 60º ATC during the intake stroke. A helical intake 808 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 809 O D FIGURE 14-30 Computed velocity field within the cylinder at 60º ATC during the intake stroke. Top left: plane through cylinder and inlet valve axes. Bottom left: orthogonal plane through valve axis. Right: circumferential-radial plane halfway between piston and cylinder head. Reference vector arrow corre- sponds to velocity of 132 m/s. Letters denote centers of toroidal flow structures.70 (a) (b) FIGURE 14-31 Comparison of (a) measured and predicted axial velocity profiles and (b) measured and predicted port is used to general swirl, and the flow through the valve curtain area (see Sec. turbulence intensity profiles at 68º ATC during the intake stroke. Data: line with points. Predictions: 6.3.2)-the inlet boundary condition for the calculation-was determined by line without points. Each interval on the scale on cylinder axes corresponds to 2 times the mean measurement. The calculation used a curvilinear, axially expanding and contract- piston speed, 71 ing grid with about 16,000 mesh points of the type shown in Fig. 14-29b. It employed a fully iterative solution algorithm with standard upwind differencing and the k-& turbulence model. Shown in Fig. 14-30 are the plane through the valve and cylinder axis (top left), the perpendicular plane through the valve axis (bottom left), and a circumferential radial plane halfway between the cylinder head and the piston (right). The major features of the conical jet flow through the inlet valve into the -0-0-0 cylinder are apparent (see Sec. 8.1). However, the off-cylinder-axis valve and the ppooooo Oo swirl generated by the helical port produce substantial additional complexity. The letters on the figures show regions of local recirculation. Regions A and B correspond to the rotating flow structures observed in simpler geometries (see to koo Fig. 8-3): however, regions CF indicate that the swirling motion is far from solid- body rotation. 70 Figures 14-31 and 14-32 show comparisons of three-dimensional predic- tions of in-cylinder flow fields with data. The computational and experimental geometries have been matched, as have the inlet flow velocities through the valve open area and engine speed. Figure 14-31 shows predicted and measured mean (a) 72º ATC (b) 166º ATC flow velocities and turbulence intensities within the cylinder, with a conventional Measurement Prediction inlet port and valve configuration, at 68º ATC during the intake stroke.71 The experimental values come from LDA measurements (see Sec. 8.2.2). The general features of the mean flow are reproduced by the model with reasonable accuracy, FIGURE 14-32 Comparison between measured and predicted swirl velocities and turbulence intensities at 72 and though some details such as the flow along the cylinder toward the head in the 166º ATC during the intake stroke. Engine equipped with helical port.59 MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 811 810 INTERNAL COMBUSTION ENGINE FUNDAMENTALS - 45 m/s symmetry plane are not predicted. The approximate magnitude of the turbulence intensity levels are predicted, but the values within the conical intake jet are underestimated. Figure 14-32 shows in-cylinder swirl velocity predictions and measurements in an engine with a disc-shaped chamber and helical intake port, ........ .. . .. . during the intake and compression strokes. Again the major features of the ......... . .. . experimental profiles are predicted adequately, though differences in detail are ... significant. 59 ---- Comparative multidimensional modeling studies of different turbulence models,65 differencing schemes, 59, 65 and number of grid points59 indicate the following. Differences in the form and coefficients of the dilation term in the k-8 turbulence model have only modest effects on flow field predictions. Higher-order turbulence models might provide improved accuracy.65 Both mesh refinement, more finely spaced grid points, and use of higher-order differencing schemes have been shown to improve significantly the accuracy of the predictions, often of course with substantial increases in computing requirements.59 Examples of predictions of other types of engine flow processes are the following. Squish flows into bowl-in-piston combustion chambers have been extensively analyzed. Figure 14-33a shows the flow field into and within an off- (a) axis bowl in piston at 20º BTC of the compression stroke. The strong radially inward squish flow at the bowl lip is apparent. However, the bowl-axis offset produces a stronger flow where the squish region is greatest in extent and results in a net flow across the bowl center plane and a complex flow pattern within the 0.9 0.1 bowl. Turbulence intensity results are often displayed on contour plots. Figure 0.7 14-33b shows the turbulence intensity distribution within the bowl at TC after - 0.5 compression. The correspondence between high-velocity regions generated by the 0.3 squish flow (Fig. 14-33a) and higher turbulence intensities is apparent. A substan- tial variation in intensity throughout the bowl is predicted. Assimilation of detailed three-dimensional velocity data from individual two-dimensional planar vector maps is cumbersome: computer-generated three-dimensional perspective 0.1 0.9 views of the velocity field are proving valuable.72 0.3- An alternative method of displaying multidimensional model results, espe- 0.7 cially from three-dimensional calculations, is through particle traces. Infinitesimal 0.5 particles are placed at key locations in the flow field at a given crank angle (e.g ., (b) at the start of the process of interest) and their trajectories are computed from the velocity field as a function of time through the process. Figure 14-34 shows the FIGURE 14-33 traces of four particles, initially located near the center of the entrance to a helical (a) Predicted velocity flow field within the offset bowl of DI diesel chamber in two orthogonal planes through the bowl center, at 20º BTC toward end of compression. Reference vector = 45 m/s. (b) inlet port at 30º ATC, as they traverse the port.73 The particle traces illustrate Predicted relative turbulence intensity u'/S, within the bowl in the same two planes at TC at the end the mechanism by which a helical port generates swirl. A second example of of compression. Numbered contours show fraction of maximum value.72 particle traces (Fig. 14-35) within the cylinder during the intake stroke indicates the complexity of swirling flows with realistic port and valve geometries.74 The figure shows the paths traced out by six particles, initially evenly spaced around Multidimensional models also provide local composition information. the valve curtain area at TC at the start of the intake process, during the intake Studies have been done of two-stroke cycle scavenging flows (e.g ., Ref. 75) and of stroke with a tangentially directed inlet port. While all the particles follow a the mixing between fresh mixture and residual gases in four-stroke cycle engines helical path within the cylinder, the steepness of these paths varies substantially (e.g ., Ref. 76). Figure 14-36 shows how the mixing between fresh fuel and air, and depending on the initial location of each particle. 812 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 813 1.0 y - 2 < = 2 0.8 a = 7 0.6 x = 7 Concentration 0.41 0.2 OL TC -120 -60 BC 60 120 TC Intake stroke Compression stroke Crank angle, deg FIGURE 14-36 FIGURE 14-34 Computed concentration distribution of fresh fuel-air mixture and residual gas within the cylinder Computed trajectories of gas particles moving through a helical inlet port during the intake process. Particles initially located near center of port at 30º ATC.73 during the intake and compression stroke of a spark-ignition engine. Concentration expressed as fresh mixture mass/total mixture mass. z = 2 is near the cylinder head, z = 7 near the piston; y = 2 near the cylinder axis, y = 7 near the cylinder liner; x = 7 along the radius passing beneath the inlet valve. 2000 rev/min and wide-open throttle. 76 residual gases, proceeds during the intake and compression strokes of a spark- ignition engine four-stroke cycle. Concentrations (defined as fresh mixture mass/ total mixture mass) at different locations within the cylinder are plotted against nonuniformity. At part load with its higher residual fraction, one would expect crank angle (z = 2 is near the head, z = 7 near the piston; y = 2 is near the these differences to be larger.76 cylinder axis, y = 7 near the cylinder liner). A relatively long time is required for the fresh and residual gases to mix and at 30º BTC there is still several percent 14.5.5 Fuel Spray Modeling The physical behavior of liquid fuel sprays when injected into the engine cylinder, as occurs in compression-ignition (or stratified-charge) engines, has already been described in Sec. 10.5. Here the current status of models for such spray behavior which are used with multidimensional models of gas motion within the cylinder are reviewed. Fuel-injected internal combustion engines present a particularly difficult problem for numerical simulation. The fuel spray produces an inhomoge- neous fuel-air mixture: the spray interacts with and strongly affects the flow pat- terns and temperature distribution within the cylinder. The fuel is injected as liquid, it atomizes into a large number of small droplets with a wide spectrum of sizes, the droplets disperse and vaporize as the spray moves through the sur- rounding air, droplet coalescence and separation can occur, gaseous mixing of fuel vapor and air then takes place, followed, finally, by combustion. Models which explicitly treat the two-phase structure of this spray describe the spray behavior in terms of differential conservation equations for mass, momentum, and energy. FIGURE 14-35 Two such classes of model exist, usually called the continuum droplet Computed trajectories traced out during the intake stroke by six gas particles initially evenly spaced model (CDM) and the discrete droplet model (DDM). Both approaches average around the valve curtain area at TC at the start of over flow processes occurring on a scale comparable to the droplet size, and thus the intake process, with a tangentially directed inlet require independent modeling of the interactions occurring at the gas droplet port. Cylinder shown with piston at BC, at the end interface: typically this is done with correlations for droplet drag and heat and of the intake stroke.74 mass transfer. The CDM attempts to represent the motion of all droplets via an 814 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 815 eulerian partial differential spray probability equation containing, in its most can be added to Eq. (14.75) so that it describes the heat-up phase where the general case, eight independent variables: time, three spatial coordinates, droplet droplet temperature increases from its initial value to T .. 79 The heat and mass radius and the three components of the droplet velocity vector. This approach exchange rates are calculated from experimentally based correlations for droplet imposes enormous computational requirements. The DDM uses a statistical Nusselt and Sherwood numbers as functions of Reynolds, Prandtl, and Schmidt approach; a representative sample of individual droplets, each droplet being a numbers. 77, 80 member of a group of identical non-interacting droplets termed a "parcel," is Account must now be taken of the two-way nature of the coupling between tracked in a lagrangian fashion from its origin at the injector by solving ordinary the gas and the liquid. The gas velocity, density, temperature, and fuel vapor differential equations of motion which have time as the independent variable. concentration required for solving the droplet equations are taken from the This latter type of model is used in engine spray analysis.77,78 Droplet values prevailing in the grid cell in which the droplet parcel resides. At the same parcels are introduced continuously throughout the fuel-injection process, with time, a field of "sources" is assembled for the interphase mass, momentum, and specified initial conditions of position, size, velocity, number of droplets in the energy transfer, and these are subsequently fed back into the gas-phase solution parcel prescribed at the "zone of atomization" according to an assumed or preserving conservation between phases.81 The gas-phase mass, momentum, and known size distribution, initial spray angle, fuel-injection rate, and fuel tem- energy conservation equations require additional terms to account for the dis- perature at the nozzle exit. The values of these parameters are chosen to rep- placement effect of the particles, the density change associated with mixing with resent statistically all such values within the spray. They are then tracked in a the fuel vapor, the drag of the droplets, the initial momentum difference and lagrangian fashion through the computational mesh used for solving the gas- enthalpy difference between evaporated fuel at the drop surface and the sur- phase partial differential conservation equations. rounding gas, and heat transfer to the droplet.79 The equations describing the behavior of individual droplets are79 The above treatment is limited to "thin sprays" where the droplets are sufficiently far apart for interparticle interactions to be unimportant. This d - dt Xk = UK (14.71) assumption is not valid in the immediate vicinity of the injector or in narrow cone sprays. In such "thick sprays" interparticle interactions-collisions which a can result in coalescence or in reseparation of droplets-are important. di ( mx Ux ) = -- mx Vp + FD,x ( u - Ux ) (14.72) The most complete models of atomized fuel sprays represent the spray by a Px Monte Carlo-based discrete-particle technique.67, 80 The spray is described by a ahx = 9x + ( ho - hx ) - dmk droplet distribution function-a droplet number density in a phase space of dt dt (14.73) droplet position, velocity, radius, and temperature. The development of this dis- tribution function is determined by the so-called spray equation.8º The distribu- where xx is the position vector for droplet k and ux its velocity, mx is the droplet tion function is statistically sampled and the resulting discrete particles are mass and px the droplet density, u is the gas velocity, hx the droplet specific followed as they locally interact and exchange mass, momentum, and energy with enthalpy, dx the heat-transfer rate from the gas to the droplet, and hy the specific the gas, using the above lagrangian droplet equations. Each discrete droplet rep- enthalpy of fuel vapor. FD, is the droplet drag function: resents a group or parcel of droplets. Droplet collisions are described by appro- priate terms in the spray equation. FD,k = ark PCplu -uxl (14.74) Figure 14-37 shows the type of results such spray models can generate. The where rx is the droplet radius, u and p the gas viscosity and density, and CD is the calculation involves a direct-injection stratified-charge engine with an offset drag coefficient. FD. is the sum of the Stokes drag and the form drag, and in the bowl-in-piston combustion chamber and a tilted injector. Injection of a single laminar limit where CD = 24/Re with Re = 2rx p |u - ux |/u it goes to 6ark u. hollow-cone fuel spray commences at 52º BTC. Figure 14-37a and b shows the An equation for the evaporation rate completes this set: it is usually fuel spray at 39º BTC at the end of injection and later, at 28º BTC, just before assumed that the droplet is in thermal equilibrium at its wet-bulb temperature, combustion. Of the 2000 computational particles injected (each representing T .. Then a balance between heat transfer to the droplet and the latent heat of some number of identical physical droplets), 1218 remain at 39º BTC and 773 at vaporization carried away by the fuel vapor exists: 28º BTC. Evaporation and coalescence have caused these decreases. Figure 14-37c shows the gas velocity vectors at the end of injection: the flow field set up (h, - hx)= dmk by momentum exchange with the injected fuel spray can be seen, and the highest dt (14.75) velocities exist in the spray region. Figure 14-37d shows the equivalence ratio contours at 28º BTC just prior to ignition. The highly nonhomogeneous fuel While a large portion of the droplet lifetime is spent in this equilibrium, terms vapor distribution within the bowl is evident. 67,82 816 MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 817 INTERNAL COMBUSTION ENGINE FUNDAMENTALS E, are constants (usually a and b are taken to be unity, or to be 0.5). Values for the preexperimental factor A and activation energy E, are obtained by matching to experimentally determined rates of burning. While this approach "works" in the sense that, when calibrated, its predic- 11 1 1 tions can show reasonable agreement with data, it has three major problems. The first is the presumption that the complex hydrocarbon fuel oxidation process can (a) (c) be adequately represented by a single (or limited number of) overall reaction(s). The fact that it is usually necessary to adjust the constants in Eq. (14.76) as engine design and operating parameters change is one indication of this problem. Second, Eq. (14.76) uses local average values of concentrations and temperatures to calculate the local reaction rate, whereas the instantaneous local values will actually determine the reaction rate. These two rates will only be equal if the reaction time scale is much longer than that of the turbulent fluctuations, which (b) (d) is not the case in engine combustion. Third, the implied strong dependence of FIGURE 14-37 burning rate on chemical kinetics is at variance with the known experimental Predictions of single hollow-cone fuel spray behavior in direct-injection stratified-charge engine. Injec- evidence on engine combustion (see Secs. 9.3.2 and 10.3). The effects of turbulence tion commences at 52º BTC with 2000 computational particles. (a) Location of 1218 remaining spray on the burning rate, apart from the augmentation of the thermal and mass diffu- particles at 39º BTC at the end of injection. (b) Location of 773 spray particles at 28º BTC, just before sivities, are not represented by equations of the form of (14.76).83, 84 combustion. (c) Gas velocity vectors at the end of injection at 39º BTC. (d) Fuel/air equivalence ratio contours just prior to ignition at 28º BTC. The L contour is o = 0.5, the contour interval Ad is 0.5.82 An alternative, equally straightforward, approach assumes that turbulent mixing is the rate-controlling process: the kinetics are sufficiently fast for chem- 14.5.6 Combustion Modeling istry modeling to be unimportant. Thus reactions proceed instantly to com- pletion once mixing occurs at a molecular level in the smaller-scale eddies of the In numerical calculations of reacting flows, computer time and storage con- turbulent flow; the rate-controlling process is then the communication between straints place severe restrictions on the complexity of the reaction mechanisms and decay of the large-scale eddies. Thus the reaction rate is inversely pro- that can be incorporated. While it is feasible to include detailed chemical mecha- portional to the turbulent mixing time tr (= l[/u') which is equated to k/8. nisms for combustion of hydrocarbon-air mixtures in one-dimensional calcu- Whether fuel or oxygen concentration is limiting, and the need for sufficient hot lations, it becomes increasingly impractical to attempt such complexity in two- products to ensure flame spreading are also incorporated. For extremely lean (or and three-dimensional studies. Accordingly, engine calculations have been forced rich) mixtures, the reaction may become kinetically controlled. A choice between to use greatly simplified reaction schemes. In addition, detailed reaction schemes Eq. (14.76) and the above mixing-controlled model can be made depending on are only available for the simpler hydrocarbon fuels (e.g ., methane, propane, whether the ratio of a chemical reaction time to the turbulent mixing time is butane): for higher hydrocarbon compounds and practical fuels which are blends greater or less than unity. 83, 84 of a large number of hydrocarbons, the detailed mechanisms have yet to be An example of a two-dimensional calculation of flame propagation in a defined. Accordingly, multidimensional engine calculations have used highly sim- premixed-charge spark-ignition engine illustrates the type of results which have plified chemical kinetic schemes, with one or at most a few reactions, to represent the combustion process. While such schemes can be calibrated with experimental data to give acceptable results over a limited range of engine conditions, they lack an adequate fundamental basis. The most common practice has been to assume the combustion process, fuel + oxidizer - products, is governed by a single rate equation of an Arrhenius form: .10º 5º 15º 30º FIGURE 14-38 R = Ap2xj xox exp ( - RT (14.76) Isotherms and velocity vectors during the combustion process in premixed spark-ignition engine predicted by two-dimensional computational fluid dynamic code. Points show ionization probe loca- where Rf is the rate of disappearance of unburned fuel, x, and xor are unburned tions in the cylinder head in corresponding experiment: open symbols are before flame arrival; filled fuel and oxidizer mass fractions, R is the universal gas constant, and a, b, A, and symbols are after flame arrival. Crank angle values are relative to TC = 0º.85 818 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 819 been generated to date. Figure 14-3885 shows computed constant-temperature 17. Mattavi, J. N.: "The Attributes of Fast Burning Rates in Engines," SAE paper 800920, SAE lines and velocity vectors, looking down on the piston, as the flame develops Trans ., vol. 89, in SP-467, The Piston Engine-Meeting the Challenge of the 80s, 1980. from the spark. The points show ionization probe locations: open symbols 18. Keck, J. C.: "Turbulent Flame Structure and Speed in Spark-Ignition Engines," Proceedings of denote prior to and closed symbols after flame arrival. A combustion model of Nineteenth Symposium (International) on Combustion, pp. 1451-1466, The Combustion Institute, the form of Eq. (14.76) was used and results in a thick "turbulent" flame with an 1982. 19. Beretta, G. P ., Rashidi, M ., and Keck, J. C.: "Turbulent Flame Propagation and Combustion in approximately cylindrical front surface. Flame front propagation speeds are ade- Spark Ignition Engines," Combust. Flame, vol. 52, pp. 217-245, 1983. quately predicted; the flame is not modeled in sufficient detail to describe its 20. Keck, J. C ., Heywood, J. B ., and Noske, G.: "Early Flame Development and Burning Rates in actual structure (see Sec. 9.3.2). Practical use is now being made of these com- Spark-Ignition Engines," SAE paper 870164, 1987. bustion codes for both spark-ignition engine (e.g ., Refs. 83, 86, and 87) and 21. Tabaczynski, R. J ., Ferguson, C. R ., and Radhakrishnan, K.: "A Turbulent Entrainment Model compression-ignition engine studies. 84, 88 for Spark-Ignition Engine Combustion," SAE paper 770647, SAE Trans ., vol. 86, 1977. 22. Tabaczynski, R. J ., Trinker, F. H ., and Shannon, B. A. S.: "Further Refinement and Validation of a Turbulent Flame Propagation Model for Spark-Ignition Engines," Combust. Flame, vol. 39, pp. 111-121, 1980. References 23. Borgnakke, C.: "Flame Propagation and Heat-Transfer Effects in Spark-Ignition Engines," in J. 1. Streit, E. E ., and Borman, G. L.: "Mathematical Simulation of a Large Turbocharged Two- C. Hilliard and G. S. Springer (eds.), Fuel Economy in Road Vehicles Powered by Spark Ignition Stroke Diesel Engine," SAE paper 710176, SAE Trans ., vol. 80, 1971. Engines, chap. 5, pp. 183-224, Plenum Press, 1984. 2. Assanis, D. N ., and Heywood, J. B.: "Development and Use of Computer Simulation of the 24. Launder, B. E ., and Spalding, D. B.: Lectures in Mathematical Models of Turbulence, Academic Turbocompounded Diesel System for Engine Performance and Component Heat Transfer Press, 1972. Studies," SAE paper 860329, 1986. 25. Borgnakke, C ., Arpaci, V. S ., and Tabaczynski, R. J.: "A Model for the Instantaneous Heat 3. Watson, N ., and Janota, M. S.: Turbocharging the Internal Combustion Engine, John Wiley, New Transfer and Turbulence in a Spark Ignition Engine," SAE paper 800287, 1980. York, 1982. 26. Borgnakke, C ., Davis, G. C ., and Tabaczynski, R. J.: " Predictions of In-Cylinder Swirl Velocity 4. Janota, M. S ., Hallam, A. J ., Brock, E. K ., and Dexter, S. G.: "The Prediction of Diesel Engine and Turbulence Intensity for an Open Chamber Cup in Piston Engine," SAE paper 810224, SAE Performance and Combustion Chamber Component Temperatures Using Digital Computers," Trans ., vol. 90, 1981. Proc. Instn Mech. Engrs, vol. 182, pt. 3L, pp. 58-70, 1967-1968. 27. Davis, G. C ., Tabaczynski, R. J ., and Belaire, R. C.: "The Effect of Intake Valve Lift on Turbu- 5. Benson, R. S ., Garg, R. D ., and Woollatt, D.: "A Numerical Solution of Unsteady Flow Prob- lence Intensity and Burnrate in S.I. Engines," SAE paper 840030, SAE Trans ., vol. 93, 1984. lems," Int. J. Mech. Sci ., vol. 6, pp. 117-144, 1964. 28. Davis, G. C ., Mikulec, A ., Kent, J. C ., and Tabaczynski, R. J.: "Modeling the Effect of Swirl on 6. Benson, R. S.: In J. H. Horlock and D. E. Winterbone (eds.), The Thermodynamics and Gas Turbulence Intensity and Burn Rate in S.I. Engines and Comparison with Experiment," SAE Dynamics of Internal Combustion Engines, vol. 1, Clarendon Press, Oxford, 1982. paper 860325, 1986. 7. Chapman, M ., Novak, J. M ., and Stein, R. A.: "Numerical Modeling of Inlet and Exhaust Flows 29. Stivender, D. L.: "Sonic Throttling Intake Valves Allow Spark-Ignition Engine to Operate with in Multi-Cylinder Internal Combustion Engines," in T. Uzkan (ed.), Flows in Internal Combustion Extremely Lean Mixtures," SAE paper 680399, SAE Trans ., vol. 77, 1968. Engines, ASME, 1982. 30. Hires, S. D ., Tabaczynski, R. J ., and Novak, J. M.: "The Prediction of Ignition Delay and Com- 8. Bulaty, T ., and Niessner, H.: "Calculation of 1-D Unsteady Flows in Pipe Systems of I.C. bustion Intervals for a Homogeneous Charge, Spark Ignition Engine," SAE paper 780232, SAE Engines," ASME paper ASME-WA7, 1984. Trans ., vol. 87, 1978. 9. Takizawa, M ., Uno, T ., Oue, T ., and Yura, T.: " A Study of Gas Exchange Process Simulation of 31. Primus, R. J ., and Wong, V. W.: "Performance and Combustion Modeling of Heterogeneous an Automotive Multi-Cylinder Internal Combustion Engine," SAE paper 820410, SAE Trans ., Charge Engines," SAE paper 850343, 1985. vol. 91, 1982. 32. Shipinski, J ., Uyehara, O. A ., and Myers, P. S.: " Experimental Correlation between Rate of Injec- 10. Baruah, P. C ., Benson, R. S ., and Balouch, S. K.: " Performance and Emission Predictions of a tion and Rate of Heat Release in Diesel Engine," ASME paper 68-DGP-11, 1968. Multi-Cylinder Spark Ignition Engine with Exhaust Gas Recirculation," SAE paper 780663, 1978. 33. Whitehouse, N. D ., and Way, R. J. B.: " Simple Method for the Calculation of Heat Release Rates 11. Poulos, S. G ., and Heywood, J. B.: "The Effect of Chamber Geometry on Spark-Ignition Engine in Diesel Engines Based on the Fuel Injection Rate," SAE paper 710134, 1971. Combustion," SAE paper 830334, SAE Trans ., vol. 92, 1983. 34. Woschni, G ., and Anisits, F.: "Experimental Investigation and Mathematical Presentation of 12. Heywood, J. B ., Higgins, J. M ., Watts, P. A ., and Tabaczynski, R. J.: " Development and Use of a Rate of Heat Release in Diesel Engines Dependent Upon Engine Operating Conditions," SAE Cycle Simulation to Predict SI Engine Efficiency and NO, Emissions," SAE paper 790291, 1979. paper 740086, 1974. 13. Watts, P. A ., and Heywood, J. B.: "Simulation Studies of the Effects of Turbocharging and 35. Watson, N ., Pilley, A. D ., and Marzouk, M.: " A Combustion Correlation for Diesel Engine Reduced Heat Transfer on Spark-Ignition Engine Operation, SAE paper 800289, 1980. Simulation," SAE paper 800029, 1980. 14. Lavoie, G. A ., and Blumberg, P. N.: "A Fundamental Model for Predicting Fuel Consumption, 36. Kuo, T ., Yu, R. C ., and Shahed, S. M.: " A Numerical Study of the Transient Evaporating Spray NO, and HC Emissions of the Conventional Spark-Ignited Engine," Combust. Sci. and Technol ., Mixing Process in the Diesel Environment," SAE paper 831735, SAE Trans ., vol. 92, 1983. vol. 21, pp. 225-258, 1980. 37. Chiu, W. S ., Shahed, S. M ., and Lyn, W. T.: " A Transient Spray Mixing Model for Diesel Com- 15. Novak, J. M ., and Blumberg, P. N.: " Parametric Simulation of Significant Design and Operating bustion," SAE paper 760128, SAE Trans ., vol. 85, 1976. Alternatives Affecting the Fuel Economy and Emissions of Spark-Ignited Engines," SAE paper 38. Rife, J. M ., and Heywood, J. B.: "Photographic and Performance Studies of Diesel Combustion 780943, SAE Trans ., vol. 87, 1978. with a Rapid Compression Machine," SAE paper 740948, SAE Trans ., vol. 83, 1974. 16. Groff, E. G ., and Matekunas, F. A.: "The Nature of Turbulent Flame Propagation in a Homoge- 39. Sinnamon, J. F ., Lancaster, D. R ., and Steiner, J. C.: "An Experimental and Analytical Study of neous Spark-Ignited Engine," SAE paper 800133, SAE Trans ., vol. 89, 1980. Engine Fuel Spray Trajectories," SAE paper 800135, SAE Trans ., vol. 89, 1980. 40. Kobayashi, H ., Yagita, M ., Kaminimoto, T ., and Matsuoka, S.: " Prediction of Transient Diesel 820 INTERNAL COMBUSTION ENGINE FUNDAMENTALS MODELING REAL ENGINE FLOW AND COMBUSTION PROCESSES 821 Sprays in Swirling Flows Via a Modified 2-D Model," SAE paper 860332, 1986. Mattavi and C. A. Amann (eds.), Combustion Modeling in Reciprocating Engines, pp. 41-68, 41. Ricou, F. P ., and Spalding, D. B.: "Measurements of Entrainment by Axisymmetric Turbulent Plenum Press, 1980. Jets," J. Fluid Mech ., vol. 9, pp. 21-32, 1961. 61. El Tahry, S. H.: "k-e Equation for Compressible Reciprocating Engine Flows," J. Energy, vol. 7, 42. Abramovich, G. M.: The Theory of Turbulent Jets, MIT Press, Cambridge, Mass ., 1963. no. 4, pp. 345-353, 1983. 43. Shahed, S. M ., Flynn, P. F ., and Lyn, W. T.: "A Model for the Formation of Emissions in a 62. Morel, T ., and Mansour, N. N.: " Modeling of Turbulence in Internal Combustion Engines," SAE Direct-Injection Diesel Engine," in J. N. Mattavi and C. A. Amann (eds.), Combustion Modeling in paper 820040, 1982. Reciprocating Engines, pp. 345-368, Plenum Press, 1980. 63. Ahmadi-Befrui, B ., Gosman, A. D ., and Watkins, A. P.: "Prediction of In-Cylinder Flow and 44. Hiroyasu, H ., Kadota, T ., and Arai, M.: "Development and Use of a Spray Combustion Model- Turbulence with Three Versions of k-& Turbulence Model and Comparison with Data," in T. ing to Predict Diesel Engine Efficiency and Pollutant Emission," paper 214-12, Bull. JSME, vol. Uzkan (ed.), Flows in Internal Combustion Engines-II, FED-vol. 20, p. 27, ASME, New York, 1984. 26, no. 214, pp. 569-575, 1983. 45. Hiroyasu, H ., Kadota, T ., and Arai, M.: "Development and Use of a Spray Combustion Model- 64. El Tahry, S. H.: "Application of a Reynolds Stress Model to Engine Flow Calculations," in T. ing to Predict Diesel Engine Efficiency and Pollutant Emissions (Part 2. Computational Pro- Uzkan (ed.), Flows in Internal Combustion Engines-II, FED-vol. 20, pp. 39-46, ASME, New cedure and Parametric Study), “ paper 214-13, Bull. JSME, vol. 26, no. 214, pp. 576-583, 1983. York, 1984. 46. Hiroyasu, H.: "Diesel Engine Combustion and Its Modeling," in Proceedings of International 65. Gosman, A. D.: "Multidimensional Modeling of Cold Flows and Turbulence in Reciprocating Symposium on Diagnostics and Modeling of Combustion in Reciprocating Engines, COMODIA 85, Engines," SAE paper 850344, 1985. pp. 53-75, Tokyo, September 4-6, 1985. 66. Ferzieger, J. H.: "Large Eddy Simulations of Turbulent Flows," AIAA paper 76-347, 1976. 47. Blumberg, P. N ., Lavoie, G. A ., and Tabaczynski, R. J.: "Phenomenological Models for Recipro- 67. Amsden, A. A ., Butler, T. D ., O'Rourke, P. J ., and Ramshaw, J. D.: "KIVA -- A Comprehensive cating Internal Combustion Engines," Prog. Energy Combust. Sci ., vol. 5, pp. 123-167, 1979. Model for 2-D and 3-D Engine Simulations," SAE paper 850554, 1985. 48. Hires, S. D ., Ekchian, A ., Heywood, J. B ., Tabaczynski, R. J ., and Wall, J. C.: "Performance and 68. Butler, T. D ., Cloutman, L. D ., Dukowicz, J. K ., and Ramshaw, J. D.: "Multidimensional NO, Emissions Modeling of Jet Ignition Prechamber Stratified Charge Engine," SAE paper Numerical Simulation of Reactive Flow in Internal Combustion Engines," in Prog. Energy 760161, SAE Trans ., vol. 85, 1976. Combust. Sci ., vol. 7, pp. 293-315, 1981. 49. Hiroyasu, H ., Yoshimatsu, A ., and Arai, M.: "Mathematical Model for Predicting the Rate of 69. Gosman, A. D ., Tsui, Y. Y ., and Watkins, A. P.: "Calculation of Three Dimensional Air Motion Heat Release and Exhaust Emissions in IDI Diesel Engines," paper C102/82, Proceedings of Con- in Model Engines," SAE paper 840229, SAE Trans ., vol. 93, 1984. ference on Diesel Engines for Passenger Cars and Light Duty Vehicles, Institution of Mechanical 70. Brandstatter, W ., Johns, R. J. R ., and Wigley, G.: "The Effect of Inlet Port Geometry on In- Engineers, London, 1982. Cylinder Flow Structure," SAE paper 850499, 1985. 50. Watson, N ., and Kamel, M.: "Thermodynamic Efficiency Evaluation of an Indirect Injection 71. Gosman, A. D ., Tsui, Y. Y ., and Watkins, A. P.: "Calculation of Unsteady Three-Dimensional Diesel Engine," SAE paper 790039, SAE Trans ., vol. 88, 1979. Flow in a Model Motored Reciprocating Engine and Comparison with Experiment," presented at 51. Mansouri, S. H ., Heywood, J. B ., and Radhakrishnan, K.: "Divided-Chamber Diesel Engine, Part Fifth International Turbulent Shear Flow Meeting, Cornell University, August 1985. I: A Cycle-Simulation which Predicts Performance and Emissions," SAE paper 820273, SAE 72. Schapertons, H ., and Thiele, F.: "Three-Dimensional Computations for Flowfields in DI Piston Trans ., vol. 91, 1982. Bowls," SAE paper 860463, 1986. 52. Mansouri, S. H ., Heywood, J. B ., and Ekchian, J. A ., "Studies of NO ,, and Soot Emissions from 73. Isshiki, Y ., Shimamoto, Y ., and Wakisaka, T.: "Numerical Prediction of Effect of Intake Port an IDI Diesel using an Engine Cycle Simulation," paper C120/82, in Diesel Engines for Passenger Configurations on the Induction Swirl Intensity by Three-Dimensional Gas Flow Analysis," in Cars and Light Duty Vehicles, Institution of Mechanical Engineers Conference Publication Proceedings of International Symposium on Diagnostics and Modeling of Combustion in Recipro- 1982-8, pp. 215-227, 1982. cating Engines, COMODIA 85, pp. 295-304, Tokyo, September 4 6, 1985. 53. Syed, S. A ., and Bracco, F. V.: "Further Comparisons of Computed and Measured Divided- 74. Wakisaka, T ., Shimamoto, Y ., and Isshiki, Y.: "Three-Dimensional Numerical Analysis of In- Chamber Engine Combustion," SAE paper 790247, 1979. Cylinder Flows in Reciprocating Engines," SAE paper 860464, 1986. 54. Meintjes, K ., and Alkidas, A. C.: " An Experimental and Computational Investigation of the Flow 75. Diwakar, R.: "Multidimensional Modeling of the Gas Exchange Processes in a Uniflow- in Diesel Prechambers," SAE paper 820275, SAE Trans ., vol. 91, 1982. Scavenged Two-Stroke Diesel Engine," in T. Uzkan, W. G. Tiederman, and J. M. Novak (eds.), 55. Watson, N ., and Marzouk, M.: "A Non-Linear Digital Simulation of Turbocharged Diesel International Symposium on Flows in Internal Combustion Engines-III, FED-vol. 28, pp. 125- Engines under Transient Conditions," SAE paper 770123, SAE Trans ., vol. 86, 1977. 134, ASME, New York, 1985. 56. Marzouk, M ., and Watson, N.: "Load Acceptance of Turbocharged Diesel Engines," paper 76. Yamada, T ., Inoue, T ., Yoshimatsu, A ., Hiramatsu, T ., and Konishi, M.: "In-Cylinder Gas C54/78, Proceedings of Conference on Turbocharging and Turbochargers, Institution of Mechani- Motion of Multivalve Engine-Three Dimensional Numerical Simulation," SAE paper 860465, cal Engineers, London, 1978. 1986. 57. Primus, R. J ., and Flynn, P. F.: "Diagnosing the Real Performance Impact of Diesel Engine 77. Gosman, A. D ., and Johns, R. J. R.: "Computer Analysis of Fuel-Air Mixing in Direct-Injection Design Parameter Variation (A Primer in the Use of Second Law Analysis)," in Proceedings of Engines," SAE paper 800091, SAE Trans ., vol. 89, 1980. International Symposium on Diagnostics and Modeling of Combustion in Reciprocating Engines, 78. Watkins, A. P ., Gosman, A. D ., and Tabrizi, B. S.: "Calculation of Three Dimensional Spray COMODIA 85, pp. 529-538, Tokyo, September 4-6, 1985. Motion in Engines," SAE paper 860468, 1986. 58. Primus, R. J ., Hoag, K. L ., Flynn, P. F ., and Brands, M. C.: " An Appraisal of Advanced Engine 79. Butler, T. D ., Cloutman, L. D ., Dukowicz, J. K ., and Ramshaw, J. D.: "Toward a Comprehensive Concepts Using Second Law Analysis Techniques," SAE paper 841287, SAE Trans ., vol. 93, 1984. Model for Combustion in a Direct-Injection Stratified-Charge Engine," in J. N. Mattavi and C. A. 59. Gosman, A. D.: "Computer Modeling of Flow and Heat Transfer in Engines, Progress and Amann (eds.), Combustion Modelling in Reciprocating Engines, pp. 231-260, Plenum Press, 1980. Prospects," in Proceedings of International Symposium on Diagnostics and Modeling of Combustion 80. Bracco, F. V.: " Modeling of Engine Sprays," SAE paper 850394, 1985. in Reciprocating Engines, COMODIA 85, pp. 15-26, Tokyo, September 4-6, 1985. 81. Cartellieri, W ., and Johns, R. J. R.: Multidimensional Modeling of Engine Processes: Progress 60. Reynolds, W. C.: "Modeling of Fluid Motions in Engines-An Introductory Overview," in J. N. and Prospects," paper presented at the Fifteenth CIMAC-Congress, Paris, June 1, 1983. 822 INTERNAL COMBUSTION ENGINE FUNDAMENTALS 82. Amsden, A. A ., Ramshaw, J. D ., O'Rourke, P. J ., and Dukowicz, J. K.: "KIVA: A Computer Program for Two- and Three-Dimensional Fluid Flows with Chemical Reactions and Fuel CHAPTER Sprays," report LA-10245-MS, Los Alamos National Laboratory, Los Alamos, New Mexico, February 1985. 83. Ahmadi-Befrui, B ., Gosman, A. D ., Lockwood, F. C ., and Watkins, A. P.: " Multidimensional Calculation of Combustion in an Idealized Homogeneous Charge Engine: A Progress Report," 15 SAE paper 810151, SAE Trans ., vol. 90, 1981. 84. Gosman, A. D ., and Harvey, P. S.: "Computer Analysis of Fuel-Air Mixing and Combustion in an Axisymmetric D.I. Diesel," SAE paper 820036, SAE Trans ., vol. 91, 1982. 85. Basso, A ., and Rinolfi, R.: "Two-Dimensional Computations of Engine Combustion: Compari- sons of Measurements and Predictions," SAE paper 820519, 1982. ENGINE 86. Basso, A.: "Optimization of Combustion Chamber Design for Spark Ignition Engines," SAE paper 840231, 1984. OPERATING 87. Schapertons, H ., and Lee, W.: "Multidimensional Modeling of Knocking Combustion in SI Engines," SAE paper 850502, 1985. CHARACTERISTICS 88. Cheng, W. K ., and Theobald, M. A.: "A Numerical Study of Diesel Ignition," paper 87-FE-2, presented at the ASME Energy-Sources Technology Conference, Dallas, February 1987. This chapter reviews the operating characteristics of the common types of spark- ignition and compression-ignition engines. The performance, efficiency, and emis- sions of these engines, and the effect of changes in major design and operating variables, are related to the more fundamental material on engine combustion, thermodynamics, fluid flow, heat transfer, and friction developed in earlier chap- ters. The intent is to provide data on, and an explanation of, actual engine oper- ating characteristics. 15.1 ENGINE PERFORMANCE PARAMETERS The practical engine performance parameters of interest are power, torque, and specific fuel consumption. Power and torque depend on an engine's displaced volume. In Chap. 2 a set of normalized or dimensionless performance and emis- sions parameters were defined to eliminate the effects of engine size. Power, torque, and fuel consumption were expressed in terms of these parameters (Sec. 2.14) and the significance of these parameters over an engine's load and speed range was discussed (Sec. 2.15). Using these normalized parameters, the effect of engine size can be made explicit. The power P can be expressed as: P = mep A, Sp/4 (four-stroke cycle) (15.1) P = mep A,Sp/2 (two-stroke cycle) 823 824 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 825 The torque T is given by 160- T = mep Va/(47) (four-stroke cycle) P 1100 (15.2) 140- T = mep Va/(27) (two-stroke cycle) imep -1000 Thus for well-designed engines, where the maximum values of mean effective 120 bmep pressure and piston speed are either flow limited (in naturally aspirated engines) =900 100 mep, kPa or stress limited (in turbocharged engines), power is proportional to piston area 800 and torque to displaced volume. Mean effective pressure can be expressed as Power P, KW 80 700 mep = n , no CHv Pad 'A) (15.3) 60 300 for four-stroke cycle engines [Eq. (2.41)], and as 40 fmep -200 mep = n , no ACHv Pad A) 20 (15.4) 100 1000 2000 3000 4000 for two-stroke cycle engines [Eqs. (2.19), (2.38), and (6.25)]. The importance of 5000 high fuel conversion efficiency, breathing capacity, and inlet air density is clear. Engine speed, rev/min Specific fuel consumption is related to fuel conversion efficiency by Eq. (2.24): 1 0.9 -- 350 sfc = - (15.5) 7 m ns 2HV 0.8- 300 These parameters have both brake and indicated values (see Secs. 2.3, 2.4, and bsfc sfc, g/kW.h 2.5). The difference between these two quantities is the engine's friction (and 0.7- 250 pumping) requirements and their ratio is the mechanical efficiency 1m. isfc The relative importance of these parameters varies over an engine's oper- 0.6- ating speed and load range. The maximum or normal rated brake power (see Sec. 200 2.1) and the quantities such as bmep derived from it (see Sec. 2.7) define an 0 1000 2000 3000 4000 5000 engine's full potential. The maximum brake torque (and bmep derived from it), Engine speed, rev/min over the full speed range, indicates the ability of the designer to obtain a high air FIGURE 15-1 flow through the engine over the full speed range and use that air effectively. Gross indicated, brake, and friction power (P ,, P ., P,), indicated, brake, and friction mean effective Then over the whole operating range, and most especially those parts of that pressure, indicated and brake specific fuel consumption, and mechanical efficiency for 3.8-dm3 six- range where the engine will operate for long periods of time, engine fuel con- cylinder automotive spark-ignition engine at wide-open throttle. Bore = 96.8 mm, stroke = 86 mm, sumption and efficiency, and engine emissions are important. Since the operating Te = 8.6.1 and emissions characteristics of spark-ignition and compression-ignition engines are substantially different, each engine type is dealt with separately. power was obtained by adding the friction power to the brake power; it is the average rate of work transfer from the gases in the engine cylinders to the pistons 15.2 INDICATED AND BRAKE POWER during the compression and expansion strokes of the engine cycle (see Sec. 2.4). AND MEP The indicated mean effective pressure shows a maximum in the engine's mid- speed range, just below 3000 rev/min. The shape of the indicated power curve The wide-open-throttle operating characteristics of a production spark-ignition follows from the imep curve. Since the full-load indicated specific fuel consump- automotive engine are shown in Fig. 15-1. The power shown is the gross power tion (and hence indicated fuel conversion efficiency) varies little over the full for the basic engine; this includes only the built-in engine accessories.2 The speed range, this variation of full-load imep and power with speed is primarily maximum net power for the fully equipped engine with the complete intake and due to the variation in volumetric efficiency, n, [see Eq. (15.3)]. Since friction exhaust system and full cooling system is about 14 percent lower. The indicated mean effective pressure increases almost linearly with increasing speed, friction 826 1100 1000 (a) (b) 200 imep 1000 imep -900 180|- 4900 mep, kPa -800 bmep - - bmep, kPa - - 160K - =800 70 - bmep 700 - - - 140|- 700 60- =600 --- Power P, KW - 120 50 - 500 Pi 100 Power P, kw 40- - 80 30- 60 bsfc 20 240 - isfc sfc, g/kW.h 10 - _=160 0 1000 1500 2000 2500 2750 1000 2000 3000 4000 5000 Speed, rev/min Speed, rev/min FIGURE 15-2 Gross indicated and brake power (P ,, P1), mean effective pressure (imep, bmep), and specific fuel consumption (isfc, bsfc) for: (a) 8.4-dm3 six-cylinder naturally aspirated direct-injection diesel engine: bore = 115 mm, stroke = 135 mm, re = 16;3 (b) 1.8-dm3 four-cylinder naturally aspirated indirect-injection swirl-chamber diesel engine: bore = 84 mm, stroke = 82 mm, r = 22.4 13-9 and 13-10). increasing speed. 15.3.1 Spark Timing have modest additional impacts. EFFICIENCY, AND EMISSIONS 15.3 OPERATING VARIABLES THAT AFFECT SI ENGINE PERFORMANCE, the inlet manifold pressure. The effect of these variables will now be reviewed. combustion starts too late, the peak cylinder pressure is reduced and the expan- relative to top-center affected the pressure development in the SI engine cylinder. Figure 9-3 and the accompanying text explain how variations in spark timing If combustion starts too early in the cycle, the work transfer from the piston to the gases in the cylinder at the end of the compression stroke is too large; if engine heat transfer per cycle and decreasing air-flow rate, as speed increases, ciency, and emissions at any given load and speed are: spark timing, fuel/air or The major operating variables that affect spark-ignition engine performance, effi- air/fuel ratio relative to the stoichiometric ratio, and fraction of the exhaust gases that are recycled for NO, emission control. Load is, of course, varied by varying cycle) have a similar shape to the full-load characteristics in Fig. 15-2. The part-load torque and bmep characteristics (at fixed amount of fuel injected per increases as the engine is throttled, decreasing mechanical efficiency (see Figs. speed since the intake system of the diesel can have larger flow areas than the pressure for naturally aspirated DI and IDI compression-ignition engines. Except increase in friction mep with speed (see Figs. 13-7, 13-11, and 13-12). Decreasing decrease in torque and bmep with increasing engine speed is due primarily to the however, at higher speeds torque and mean effective pressure decrease more intake of SI engines with their intake-system fuel transport requirements. The lower and lower speeds as the throttle open area is reduced, increasingly limiting percent from 0.31 to 0.34 over the speed range 1000 to 4000 rev/min. This is at high engine speeds, brake torque and mep vary only modestly with engine the air flow (see Fig. 7-22). The pumping component of total friction also rapidly with increasing speed than at full load. The throttle chokes the flow at primarily due to the decreasing importance of heat transfer per cycle with min. Thus bmep peaks at a lower speed than imep. The brake power shows a decrease in P ,. The indicated fuel conversion efficiency increases by about 10 increasing speed from a maximum of about 0.9 at low speed to 0.7 at 5000 rev/ maximum at about 4300 rev/min; increases in speed above this value result in a power will increase more rapidly. Hence mechanical efficiency decreases with Figure 15-2 shows full-load indicated and brake power and mean effective At part load at fixed throttle position, these parameters behave similarly; ENGINE OPERATING CHARACTERISTICS 827 828 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 829 sion stroke work transfer from the gas to the piston decreases. There exists a torque, the effect of spark timing variations on fuel consumption at constant particular spark timing which gives maximum engine torque at fixed speed, and engine load can be evaluated. Figure 15-3b shows results obtained with a com- mixture composition and flow rate. It is referred to as MBT-maximum brake puter simulation of the engine operating cycle.6, 7 The curves for several different torque-timing. This timing also gives maximum brake power and minimum part-load operating conditions and burn durations (from fast to slow) have been brake specific fuel consumption. Figure 15-3a shows the effect of spark advance normalized and fall essentially on top of each other. Five degrees of retard in variations on wide-open-throttle brake torque at selected speeds between 1200 spark timing have only a modest effect on fuel consumption; for 10 to 20º retard, and 4200 rev/min for a production eight-cylinder engine. At each speed, as spark the impact is much more significant. is advanced from an initially retarded setting, torque rises to a maximum and Spark timing affects peak cylinder pressure and therefore peak unburned then decreases. MBT timing depends on speed; as speed increases the spark must and burned gas temperatures (see Sec. 9.2.1). Retarding spark timing from the be advanced to maintain optimum timing because the duration of the com- optimum reduces these variables. Retarded timing is sometimes used therefore for bustion process in crank angle degrees increases. Optimum spark timing also NO, emission control (see Fig. 11-13 and accompanying text) and to avoid depends on load. As load and intake manifold pressure are decreased, the spark knock (see Sec. 9.6.1). The exhaust temperature is also affected by spark timing. timing must be further advanced to maintain optimum engine performance. Retarding timing from MBT increases exhaust temperature; both engine effi- The maximum in each brake torque curve in Fig. 15-3a is quite flat. Thus ciency and heat loss to the combustion chamber walls (see Fig. 12-27) are accurate determination of MBT timing is difficult, but is important because NO decreased. Retarded timing is sometimes used to reduce hydrocarbon emissions and HC emissions vary significantly with spark timing. In practice, to permit a by increasing the fraction oxidized during expansion and exhaust due to the more precise definition of spark timing, the spark is often retarded to give a 1 or higher burned gas temperatures that result (see Sec. 11.4.3). Retarded timing may 2 percent reduction in torque from the maximum value. be used at engine idle to bring the ignition point closer to TC where conditions In Fig. 15-3a the mixture composition and flow rate were held constant at for avoiding misfire are more favorable. each engine speed. If the mixture flow rate is adjusted to maintain constant brake 15.3.2 Mixture Composition --- 1% loss line BL spark advance The unburned mixture in the engine cylinder consists of fuel (normally 440 2600 rev/min vaporized), air, and burned gases. The burned gas fraction is the residual gas plus 20005 any recycled exhaust used for NO control. Mixture composition during com- bustion is most critical, since this determines the development of the combustion 120 20º 48, process which governs the engine's operating characteristics. While substantial 1600___ 3600 1.1 1000. 10% EGR efforts are made to produce a uniform mixture within the cylinder, some nonuni- formities remain (see Sec. 9.4.2). In a given cylinder, cycle-by-cycle variations in 400 average charge composition exist. Also, within each cylinder in a given engine Brake torque, N.m $ = 0.8 1200 ____ cycle, the fuel, air, EGR, and residual gas are not completely mixed, and com- bsfc (MBT) bsfc position nonuniformities across the charge may be significant.| These together 380|- 4000 produce variations in composition at the spark plug location (the critical region 1.0- since the early stages of flame development influence the rest of the combustion ¢ = 1.0, 4%, = 60º, 0% EGR 4200 except where noted process) which can be of order + 10 percent peak-to-peak (see Fig. 9-34). In addi- 360- T tion, in multicylinder engines, the average air, fuel, and EGR flow rates to each -20 10 -10 -5 MBT +5 cylinder are not identical. Typical cylinder-to-cylinder variations have standard TC 10 20 30 40 50 - Retard |Advance- deviations of +5 percent of the mean for air flow rate and fuel flow rate (giving a Spark advance, deg BTC Spark advance (a) (b) FIGURE 15-3 (a) Variation in brake torque with spark advance, eight-cylinder automotive spark-ignition engine at + This aspect of mixture nonuniformity is least well defined. Mixing of the fresh mixture (fuel, air, and wide-open throttle, at engine speeds from 1200 to 4200 rev/min. 1 percent torque loss from MBT and EGR) with residual gas is likely to be incomplete (see Fig. 14-36), especially at light load when the spark advance for borderline knock are shown.5 (b) Predicted variation in brake specific fuel con- residual gas fraction is highest. With intake-port fuel-injection systems, there is evidence of incomplete sumption (normalized by MBT value) with spark retard at several different part-load engine condi- fuel-air mixing due to the fact that the air flow and fuel flow processes are not in phase.9 When the tions. 6, 7 engine is cold, fuel distribution within the cylinder is known to be nonuniform. 830 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 831 +7 percent variation in the air/fuel ratio) for steady-state engine operation. EGR tially burned. This increases the temperature and the number of moles of the cylinder-to-cylinder flow rates may have higher variability. Under unsteady burned gases in the cylinder. These effects increase the pressure to give increased engine operating conditions all these variations can be higher. power and mep. Fuel conversion efficiency decreases approximately as 1/o, as the It is necessary to consider the effect of mixture composition changes on mixture is richened above stoichiometric (¢ > 1) due to the decreasing com- engine operating and emissions characteristics in two regimes: (1) wide-open bustion efficiency associated with the richening mixture. throttle (WOT) or full load and (2) part throttle or load. At WOT, the engine air For mixtures lean of stoichiometric, the theoretical fuel conversion effi- flow is the maximum that the engine will induct.+ Fuel flow can be varied, but air ciency increases linearly as o decreases below 1.0. Combustion of mixtures leaner flow is set by engine design variables and speed. At part throttle, air flow, fuel than stoichiometric produces products at lower temperature, and with less disso- flow, and EGR flow can be varied. Evaluation of mixture composition changes at ciation of the triatomic molecules CO2 and H2O. Thus the fraction of the chemi- part load should be done at fixed (brake) load and speed, i.e ., under conditions cal energy of the fuel which is released as sensible energy near TC is greater; where the engine provides the desired torque level at the specified speed. To hence a greater fraction of the fuel's energy is transferred as work to the piston maintain torque (or load or bmep) constant as mixture composition is varied during expansion, and the fraction of the fuel's available energy rejected to the normally requires changes in throttle setting (and if EGR is varied, changes in exhaust system decreases (see Sec. 5.7). There is a discontinuity in the fuel conver- EGR flow-control valve setting). This distinction between part-load comparisons sion efficiency and imep curves at the stoichiometric point; the burned gas com- at specified torque or bmep, rather than at constant throttle settings (which gives position is substantially different on the rich and the lean sides of o = 1. essentially constant air flow), is important because the pumping work component Figure 15-4 shows gross indicated specific fuel consumption data for a six- of engine friction will vary at constant engine load as mixture composition cylinder spark-ignition engine at wide-open throttle and 1200 rev/min,9 and changes. At constant throttle setting and speed, the pumping work remains essen- values of gross indicated mean effective pressure and fuel conversion efficiency tially unchanged. derived from the isfc data. In these engine tests, the fuel-air mixture was prepared in two different ways: (1) with the normal carburetor and (2) with a heated vapor- AIR/FUEL OR EQUIVALENCE RATIO CHANGES. Mixture composition effects izing tank to ensure intake-mixture uniformity. Shapes of the practical efficiency are usually discussed in terms of the air/fuel ratio (or fuel/air ratio) because in curves and the theoretical curves in Fig. 5-9 differ. Cylinder-to-cylinder air/fuel engine tests, the air and fuel flow rates to the engine can be measured directly and ratio maldistribution prevents the carbureted engine operating leaner than because the fuel metering system is designed to provide the appropriate fuel flow ¢ ~ 0.85 (A/F ~ 17) without misfire under these conditions. While use of a fuel for the actual air flow at each speed and load. However, the relative proportions vaporizing and mixing tank essentially removes this maldistribution and extends of fuel and air can be stated more generally in terms of the fuel/air equivalence the lean misfire limit, n ; does not continue to increase as o decreases. The ratio o [the actual fuel/air ratio normalized by dividing by the stoichiometric reasons for this are that cycle-to-cycle pressure fluctuations and the total dura- fuel/air ratio, see Eq. (3.8)] or the relative air/fuel ratio A [see Eq. (3.9)]. The combustion characteristics of fuel-air mixtures and the properties of combustion 1100 products, which govern engine performance, efficiency, and emissions, correlate best for a wide range of fuels relative to the stoichiometric mixture proportions. 1000- imep Where appropriate, therefore, the equivalence ratio will be used as the defining 900 38 parameter. Equation (7.1) converts the air/fuel ratio with gasoline to the equiva- imep, kPa 800 lence ratio. -36 The theoretical basis for understanding the effect of changes in the equiva- 700 nf. i 34 lence ratio is the fuel-air cycle results in Figs. 5-9 and 5-10, where the indicated 320 32 nf. is percent fuel conversion efficiency and mean effective pressure are shown as a function of the fuel/air equivalence ratio, o. The mean effective pressure peaks slightly rich of stoichiometric, between o = 1 and 1.1. Due to dissociation at the high tem- · Vapor tank FIGURE 15-4 280 o Carburetor -28 Effect of the fuel/air equivalence ratio varia- isfc, g/kW.h peratures following combustion, molecular oxygen is present in the burned gases under stoichiometric conditions, so some additional fuel can be added and par- 2601- isfc tions on indicated mean effective pressure, 26 specific fuel consumption, and fuel conver- 240- sion efficiency of six-cylinder spark-ignition 220 engine at wide-open throttle and 1200 rev/ 0.6 0.8 1.0 1.2 1.4 min. Data for standard carbureted engine, and engine equipped with vapor tank which + EGR is normally zero at WOT, since maximum torque is usually desired. Fuel/air equivalence ratio extends the lean operating limit, are shown.9 832 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 833 until the lengthening burn duration and larger cycle-by-cycle variations cause 500 bmep = 250 kPa bsfc to increase. For the slower-burning conventional chamber, this deterioration 2400 rev/min 150 in combustion starts to occur almost immediately on the lean side of stoichiomet- ric, and fuel consumption worsens for o < 0.9. 400 Thus the equivalence ratio for optimum fuel consumption at a given load bsfc, g/kW .h depends on the details of chamber design (including compression ratio) and 350- mixture preparation quality. It also varies for a given chamber over the part- Compact, high re, chamber throttle load and speed range. For lighter loads and lower speeds it is closer to 300 stoichiometric since the residual gas fraction is higher and combustion quality is 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 poorer with greater dilution and at lower speeds. Fuel/air equivalence ratio At part load, as the air/fuel ratio is varied at constant brake load, the (a) pumping work varies, and this also contributes to the brake specific fuel con- 430 sumption and efficiency variation with equivalence ratio. Figure 15-6 shows the gross and net indicated fuel conversion efficiencies and brake efficiency as a func- 410 bmep = 325 kPa tion of equivalence ratio at a part-throttle constant load and speed point 390- 1400 rev/min FIGURE 15-5 MBT timing Effect of combustion chamber design (325 kPa bmep and 1400 rev/min), calculated using a thermodynamic-based com- 370 and burn rate on spark-ignition engine puter simulation of the engine's operating cycle. The difference between the net brake specific fuel consumption. (a) and gross indicated curves illustrates the magnitude of the effect of the pumping 350 bsfc, g/kW . h 1.6-dm3 four-cylinder engine with con- work changes. Part-throttle comparisons of different operating conditions should 330 ventional combustion chamber and 40; = 100º be done at constant brake load (torque or bmep) and speed: the task the engine 1.5-dm3 four-cylinder engine with 310 600 compact fast-burning high-compression- is required to perform is then the same. At constant bmep and speed, the mecha- 290 ratio chamber beneath the exhaust valve nical rubbing friction is essentially fixed; thus net imep is constant (and gross 200 with re == 13, both at bmep of 250 kPa imep will vary if the pumping mep varies). 270- and 2400 rev/min.10 (b) Predictions from Note that all the engine data show a smooth transition between the rich and 0.6 0.7 0.8 0.9 1.0 1.1 1.2 thermodynamic-based computer simula- lean characteristics at the stoichiometric point, whereas the calculated sfc and tion of engine cycle for 5.7-dm3 eight- Fuel/air equivalence ratio cylinder engine at bmep of 325 kPa and (b) 1400 rev/min with MBT spark timing.6 40 tion of the burning process increase as the mixture becomes leaner: both these factors degrade engine efficiency. Since the spark advance is set for the average 35P nf , is cycle, increasing cycle-to-cycle dispersion produces increasing imep (and hence nr. :) losses in "nonaverage" cycles due to nonoptimum timing. The lengthening burn duration directly decreases efficiency, even in the absence of cyclic varia- nf , in tions. 30- Fuel conversion efficiency, percent Engine fuel consumption and efficiency well lean of stoichiometric depend strongly on the engine combustion chamber design. Figure 15-5 shows two sets of engine bsfc data, for a conventional combustion chamber and a compact high- nf , b 25- compression-ratio chamber, at constant load and speed (250 kPa bmep and 2400 rev/min) as a function of equivalence ratio. Also shown are bsfc results obtained from a thermodynamic-based computer cycle simulation of the spark- FIGURE 15-6 ignition engine operating cycle (at 325 kPa bmep and 1400 rev/min).6 Though 20 Gross and net indicated, and brake, fuel conversion the load and speed are different, the behavior of the data and predictions for rich efficiencies predicted by thermodynamic-based cycle mixtures, o > 1, are comparable. On the lean side of stoichiometric, however, fuel 0L '0.6 simulation at constant part-load bmep (325 kPa) 0.8 1.0 .2 consumption depends on the combustion characteristics of the chamber. The and speed (1400 rev/min) for a fixed burn duration Fuel/air equivalence ratio (0-100 percent, 60º CA).6 faster-burning compact high-compression-ratio chamber shows decreasing bsfc 834 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 835 efficiency characteristics show a discontinuity in slope. The difference is due to in quantifying this variation. However, time-averaged thermocouple measure- cylinder-to-cylinder and cycle-by-cycle mixture composition variations and to ments from specific locations in the exhaust system can provide useful informa- cycle-by-cycle cylinder pressure variations which exist (though to a lesser extent) tion on trends. Figure 14-10 shows examples of predictions of the even in the absence of these mixture variations. Averaging over these variations enthalpy-averaged exhaust gas temperature at the exhaust port exit as a function smooths out the theoretical discontinuity in slope at o = 1.0. of equivalence ratio compared with time-averaged measurements. The enthalpy- The equivalence ratio requirements of a spark-ignition engine over the full averaged temperature is defined by Eq. (6.19). These are typically 50 to 100 K load and speed range can now be explained from the point of view of per- higher than time-averaged measurements. The exhaust temperature peaks at the formance and efficiency. However, since emissions depend on o also, emission stoichiometric point and decreases as the mixture is richened and leaned on control requirements may dictate a different engine calibration, as will be dis- either side. cussed later. The mixture requirements in the induction system are usually dis- The fuel/air equivalence ratio is an important parameter controlling spark- cussed in relation to steady and transient engine operation. Steady operation ignition engine emissions. The critical factors affecting emissions, that are govern- includes operation at a given speed and load over several engine cycles with a ed by the equivalence ratio, are the oxygen concentration and the temperature of warmed-up engine. Transient operation includes engine starting, engine warm-up the burned gases. Excess oxygen is available in the burned gases lean of stoichio- to steady-state temperatures, and changing rapidly from one engine load and metric. The maximum burned gas temperatures occur slightly rich of stoichio- speed to another. The mixture requirements of the engine as defined by the com- metric at the start of the expansion stroke, and at the stoichiometric composition position of the combustible mixture at the time of ignition, while they vary some- at the end of expansion and during the exhaust process. Figure 11-2 illustrates what with speed and load, are essentially the same for all these operating modes.+ the general trends in emissions with equivalence ratio which have already been However, the methods used to prepare the mixture prior to entry to the cylinder discussed. must be modified in the transient modes when liquid fuels are used, to allow for Figure 15-7 shows the effect of variations in fuel/air equivalence ratio on variations in the liquid fuel flow and fuel evaporation rate in the intake manifold NO, and HC emissions and fuel consumption when a special fuel vapor gener- as the air flow varies and as the manifold and inlet port pressure and temperature ator was used to produce a uniform fuel-air mixture. As explained in Sec. 11.2.3, change. The transient fuel metering requirements for adequate mixture prep- the formation rate of NO depends on the gas temperature and oxygen concentra- aration are discussed in Chap. 7. tion. While maximum burned gas temperatures occur at o ~ 1.1, at this equiva- At all load points at a given speed, the ideal equivalence ratio is that which gives minimum brake specific fuel consumption at the required load. However, once wide-open-throttle air flow has been reached, increases in power can only be Fuel/air equivalence ratio obtained by increasing the fuel flow rate. The equivalence ratio requirements for ).8 0.75 0.7 0.65 optimum-efficiency steady-state engine operation can be summarized on a plot of 16- equivalence ratio versus percent of maximum air flow at any given speed. A typical plot was shown in Fig. 7-1. For part-throttle operation, unless dictated otherwise by emission control requirements, the equivalence ratio is set close to the equivalence ratio for minimum fuel consumption consistent with avoiding 12 300 partial burning or misfire in one or more cylinders. At very light load the best NO, bsfc mixture is richer to compensate for slower flame speeds at lower mixture Fuel density and increased residual fraction. As wide-open throttle is approached, the Brake specific emissions, g/kW . h mixture is richened to obtain maximum power. 85 290 The exhaust gas temperature varies with the equivalence ratio. The exhaust bsfc, g/kW . h gas temperature also varies continuously as the gas leaves the engine cylinder and flows through the exhaust port and the manifold and pipe (see Sec. 6.5), so FIGURE 15-7 an appropriate definition of an average exhaust gas temperature should be used 4 HC 280 Variation of brake specific HC and NO, emissions and fuel consump- tion with (A/F) and fuel/air equiv- alence ratio. 5.7-dm3 eight-cylinder T spark-ignition engine at 385 kPa + Except during start-up and cold engine operation, when a substantial part of the fuel within the 18 20 21 22 23 bmep and 1400 rev/min with uni- cylinder can be in the liquid phase. Air/fuel ratio form vaporized fuel-air mixture.11 836 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 837 lence ratio oxygen concentrations are low. As the mixture is leaned out, recycled to the intake. As described in Sec. 6.4, the residual gas fraction is influ- increasing oxygen concentration initially offsets the falling gas temperatures and enced by load and valve timing (especially the extent of valve overlap) and, to a NO emissions peak at o ~ 0.9. Then, decreasing temperatures dominate and NO lesser degree, by the air/fuel ratio and compression ratio. The total burned gas emissions decrease to low levels. mass fraction is given by Eq. (4.3). Since the burned gases dilute the unburned Figure 15-7 also shows the effect of variations in equivalence ratio for lean mixture, the absolute temperature reached after combustion varies inversely with mixtures on unburned hydrocarbon emissions. For rich mixtures, Fig. 11-2 the burned gas mass fraction. Hence increasing the burned gas fraction reduces shows that emissions are high. This is primarily due to the lack of oxygen for the rate of formation of NO emissions. afterburning of any unburned hydrocarbons that escape the primary combustion Figure 11-10 shows the effect on NO emissions of increasing the burned gas process, within the cylinder and the exhaust system. HC emissions decrease as the fraction by recycling exhaust gases to the intake system. Substantial reductions in stoichiometric point is approached: increasing oxygen concentration and NO concentrations are achieved with 10 to 25 percent EGR. However, EGR also increasing expansion and exhaust stroke temperatures result in increasing HC reduces the combustion rate which makes stable combustion more difficult to burnup. For moderately lean mixtures, HC emission levels vary little with equiv- achieve (see Sec. 9.4.3 and Fig. 9-36). The amount of EGR a particular com- alence ratio. Decreasing fuel concentration and increasing oxygen concentration bustion chamber design will tolerate depends on its combustion characteristics, essentially offset the effect of decreasing bulk gas temperatures. As the lean oper- the speed and load, and the equivalence ratio. EGR percentages in the 15 to 30 ating limit of the engine is approached, combustion quality deteriorates signifi- range are about the maximum amount of EGR a spark-ignition engine will toler- cantly and HC emissions start to rise again due to the occurrence of occasional ate under normal part-throttle conditions. Faster-burning engines will tolerate partial-burning cycles. For still leaner mixtures, HC emissions rise more rapidly more EGR than slower-burning engines. Because of the decrease in burn rate and due to the increasing frequency of partial-burning cycles, and even the occurrence increase in cycle-by-cycle combustion variations, hydrocarbon emissions increase of completely misfiring cycles (see Sec. 9.4.3). The equivalence ratio at which with increasing EGR, as shown in Fig. 11-29. At first the increase in HC is partial-burning and misfiring cycles just start to appear depends on details of the modest and is due primarily to decreased HC burnup due to lower expansion engine combustion and fuel preparation systems, as well as the load and speed and exhaust stroke temperatures. The HC increase becomes more rapid as slow point. combustion, partial burning, and even misfire, in turn, occur with increasing fre- The effect of equivalence ratio variations on CO emissions has already been quency. EGR has no significant effect on engine CO emissions. explained in Sec. 11.3 (see Fig. 11-20). For rich mixtures, CO levels are high The effect of exhaust gas recycle on engine performance and efficiency, for because complete oxidation of the fuel carbon to CO2 is not possible due to mixtures with ¢ < 1.0, is similar to the addition of excess air. Both EGR and insufficient oxygen. For lean mixtures, CO levels are approximately constant at a excess air dilute the unburned mixture. In practice since EGR is only used at low level of about 0.5 percent or less. part-throttle conditions, o < 1.0 is the region of interest. Because three-way cata- Figure 15-7 indicates that if an engine can be designed and operated so that lysts are now used where NO, emission constraints are severe, greatest attention its stable operating limit under the appropriate part-load conditions is sufficiently has focused on dilution with EGR at o ~1.0. Figure 15-8 shows the effect of lean, excellent fuel consumption and substantial control of engine NO, HC, and increasing EGR on bsfc and enthalpy-mean exhaust temperature [defined by Eq. CO emissions can be achieved. Such an approach requires good control of (6.19)] at constant bmep, predicted using a thermodynamic-based computer mixture preparation and a fast-burning combustion chamber design (see Sec. simulation of the engine's operating cycle. Predictions made for different burn 15.4.1). However, this lean-engine approach is not compatible with the three-way durations are shown, at MBT timing for a stoichiometric mixture. At constant catalyst system (see Sec. 11.6.2) which, with close-to-stoichiometric mixtures, burn duration, bsfc and exhaust temperature decrease with increasing EGR. Only achieves substantial additional reductions in NO, HC, and CO emissions. for very long combustion processes is the burn rate especially significant. This improvement in fuel consumption with increasing EGR is due to three factors: (1) EXHAUST GAS RECYCLE. Exhaust gas recycle (EGR) is the principal technique reduced pumping work as EGR is increased at constant brake load (fuel and air used for control of SI engine NO, emissions (see Sec. 11.2.3). A fraction of the flows remain almost constant; hence intake pressure increases); (2) reduced heat exhaust gases are recycled through a control valve from the exhaust to the engine loss to the walls because the burned gas temperature is decreased significantly; intake system. The recycled exhaust gas is usually mixed with the fresh fuel-air and (3) a reduction in the degree of dissociation in the high-temperature burned mixture just below the throttle valve. EGR acts, at part load, as an additional gases which allows more of the fuel's chemical energy to be converted to sensible diluent in the unburned gas mixture, thereby reducing the peak burned gas tem- energy near TC. The first two of these are comparable in magnitude and each is peratures and NO formation rates. Note that it is the total burned gas fraction in about twice as important as the third. 12 the unburned mixture in the cylinder that acts as a diluent. These burned gases Figure 15-9 shows experimental bsfc versus EGR data for two combustion are comprised of both residual gas from the previous cycle and exhaust gas chambers: a combustion chamber with a moderate burning rate and a faster- 838 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 839 340 burning chamber with open geometry and with induction-generated swirl. ¢ = 1.0, MBT timing bmep = 325 kPa Though addition of EGR lengthens both the flame development and propagation 330} 1400 rev/min processes (as indicated by the increasing MBT spark advance requirement with increasing EGR), the faster-burning chamber follows the anticipated pattern of 40, = 100º 320 significant bsfc reductions until, at about 20 percent EGR, the combustion quality deteriorates. For the slower-burning combustion chamber, the tolerance bsfc, g/kW . h 80º to dilution with EGR is much less. 310 600 400 300} 15.3.3 Load and Speed 200 One common way to present the operating characteristics of an internal com- 0 8 12 16 20 bustion engine over its full load and speed range is to plot brake specific fuel EGR, % consumption contours on a graph of brake mean effective pressure versus engine speed. Operation of the engine coupled to a dynamometer on a test stand, over 1400- ¢ = 1.0, MBT timing its load and speed range, generates the torque and fuel flow-rate data from which bmep = 325 kPa such a performance map is derived. Equation (2.20) relates bmep to torque, and 1400 rev/min 1300 bsfc values are obtained from Eq. (2.22) at each operating point. Figure 15-10 shows an example of such a performance map for a four-cylinder spark-ignition engine. The upper envelope of the map is the wide-open-throttle performance Exhaust temperature, K 200 40) = 100º FIGURE 15-8 curve. Points below this curve define the part-load operating characteristics. 80º While details differ from one engine to another, the overall shapes of these maps .60º Effect of recycled exhaust on brake 1100- specific fuel consumption and exhaust for spark-ignition engines are remarkably similar. When mean piston speed S, is temperature at constant bmep and used instead of crankshaft speed for the abscissa, the quantitative similarity of 200 speed, stoichiometric mixture, and such maps over a wide range of engine sizes is more apparent. 1000 various burn durations (0-100 8 12 16 20 percent). Predictions from thermo- Maximum bmep occurs in the mid-speed range; the minimum bsfc island is EGR, % dynamic-based cycle simulation.6 located at a slightly lower speed and at part load. These map characteristics can be understood in terms of variations in volumetric efficiency nu, gross indicated 60 fuel conversion efficiency ni, and mechanical efficiency nm as A/F, EGR (if used), and the importance of heat losses and friction change, via Eqs. (15.3) and (15.5). 10 Spark advance, ºBTC Mean piston speed, m/s 1000 6 12 14 20 300 335 g/kW . h 475 Moderate burn rate 500 290 335 bmep, kPa 425 TT FIGURE 15-9 275 bsfc, g/kW .h Brake specific fuel consumption and MBT 305 -O- Fast burn rate spark advance as a function of percent recy- 365 375 - cled exhaust, for four-cylinder spark-ignition 200- 520 FIGURE 15-10 engine with a moderate burn rate com- 610 Performance map for 2-dm3 four- bustion chamber and a fast burn rate com- cylinder fast-burn spark-ignition 0 TO T 20 30 40 bustion chamber. 1400 rev/min, 324 kPa 0 1000 2000 3000 4000 5000 engine showing contours of constant EGR, % bmep, equivalence ratio 1.0.12 Engine speed, rev/min bsfc in grams per kilowatt-hour.13 840 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 841 The maximum bmep curve reflects the variation with speed of n ,, the decrease of than at higher loads.º The residual gas fraction decreases as speed increases, this nm as S, increases, and the increase of n. as S, increases due to decreasing effect being greater at lower inlet manifold pressures (lighter loads) (see Fig. 6-19). importance of heat transfer per cycle. The bsfc contours have the following expla- Also, the relative importance of heat transfer per cycle is less as speed increases nation. Starting at the minimum bsfc point, increasing speed at constant load (see Fig. 12-25), which would also be expected to increase NO concentration. increases bsfc due primarily to the increasing friction mep at higher speeds (which With increasing load (at constant speed), NO concentrations also increase. Again, decreases nm). While ns, it increases as speed increases, friction increases dominate. as inlet manifold pressure and load increase, the residual gas fraction decreases Decreasing speed at constant load increases bsfc due primarily to the increasing (Fig. 6-19); also, the relative importance of heat transfer per cycle decreases with importance of heat transfer per cycle (which decreases ni). Friction decreases, increasing load (Fig. 12-25). increasing nm, but this is secondary. Any mixture enrichment required to main- The hydrocarbon concentration trends with speed and load changes are the tain a sufficiently repeatable combustion process at low engine speeds (see Fig. opposite of the NO concentration trends. As indicated in Table 11.7, speed and 7-1) contributes too. Increasing load at constant speed from the minimum bsfc load are likely to affect several of the HC formation mechanisms, the in-cylinder point increases bsfc due to the mixture enrichment required to increase torque as mixing of unburned hydrocarbons which escape combustion with the bulk gases, the engine becomes increasingly air-flow limited. Decreasing load at constant and the fraction of the in-cylinder HC which escape into the exhaust. However, speed increases bsfc due to the increased magnitude of friction (due to increased not enough is yet known about the details of these processes to make these pumping work), the increased relative importance of friction, and increasing dependencies explicit. If oxygen is available, oxidation of unburned hydrocarbons importance of heat transfer (which decreases nf.i?). both within the cylinder and in the exhaust system will be significantly enhanced The effects of speed and load variations on NO and HC emissions are by increases in speed since the expansion stroke and exhaust process gas tem- shown in Fig. 15-11.14 NO concentrations increase moderately with increasing peratures increase substantially, due to the reduced significance of heat transfer speed at constant load. At lower loads, the proportional increase in NO is greater per cycle with increasing speed. This more than offsets the reduced residence time in the cylinder and in the exhaust. Measurements of the percent HC reacted in the exhaust port as a function of engine speed show the same proportional 3000 reduction in the exhaust emissions data in Fig. 15-11.15 The rationale for the variation with load is less clear. As load increases at constant speed, expansion 2000 - 3000 and exhaust stroke temperatures increase, and the in-cylinder oxidation rate, if oxygen is available, will increase. However, as the exhaust gas flow rate increases, 2500 the residence time in critical sections of the exhaust system decreases and a 1500 p reduction in exhaust port HC oxidation occurs.16 The net trend is for HC con- ---- ---- 2000 HC, ppm C1 centration to decrease modestly as load is increased. O O O 1000 -- -- HC, ppm C1 1500 - 0 P 15.3.4 Compression Ratio 500 --- 4000 1000|- The ideal cycle analysis of Chap. 5 showed that indicated fuel conversion effi- 500- 4000 ciency increased continuously with compression ratio according to Eq. (5.31). O -- 3000 With y = 1.3, this relation also matches closely the fuel-air cycle predictions with 0 3000 o ~ 1.0. However, in an actual engine other processes which influence engine 2000 NO, ppm performance and efficiency vary with changes in compression ratio: namely, com- 2000 NO, ppm bustion rate and stability, heat transfer, and friction. Over the load and speed 1000 1000 range, the relative impact that these processes have on power and efficiency varies also. Hence, the applicability of Eq. (5.31) is open to question. Also, while 30 JO the geometric compression ratio (ratio of maximum to minimum cylinder 1200 1600 2000 300 400 500 600 700 imep, kPa volume) is well defined, the actual compression and expansion processes in Speed, rev/min (b) engines depend on valve timing details and the importance of flow through the (a) valves while they are opening or closing (which depends on engine speed). Of FIGURE 15-11 course, our ability to increase the compression ratio is limited by the octane Variation in spark-ignition engine HC and NO, emissions with (a) engine speed at 379 kPa imep and (b) load (or imep) at 1250 rev/min. Equivalence ratio = 0.9, MBT spark timing, r. = 7.14 quality of available fuels and knock (see Sec. 9.6.1). 842 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 843 Only a few studies have examined the effect of compression ratio on spark- 1.3- ignition engine performance and efficiency over a wide range of compression Fuel-air cycle ratios. Figure 15-12 shows results obtained at wide-open throttle at 2000 rev/min with a series of eight-cylinder 5.3-dm3 displacement engines, from the most exten- sive of these studies.17 Gross-indicated and brake fuel conversion efficiencies and mean effective pressures are shown. Indicated mep was obtained by adding 1.2H motoring friction mep to brake mep. The mep data were obtained with (A/F) and nf. : (CN) spark timing adjusted to give maximum torque; for the efficiency data, (4/F) and nf (rc) if (r = 8) D. : (KT) nf, b (CN) spark timing were adjusted to give maximum efficiency. The mechanical efficiency remained essentially constant at 0.89 over the full compression ratio range. The 1.1 --- volumetric efficiency was also constant at 0.825. Both nig and mep show a maximum at a compression ratio of about 17; for higher compression ratios efficiency and mep decrease slightly. This trend was explained as being due to increasing surface/volume ratio and slower combustion, and is also due to the FIGURE 15-13 1.0- Wide-open throttle increasing importance of crevice volumes: at the higher compression ratios Relative fuel conversion efficiency studied the combustion chamber height became very small. improvement with increasing com- To assess more broadly the effect of compression ratio variations on fuel 00 10 12 14 16 18 20 22 pression ratio, spark-ignition 24 engines at wide-open throttle: conversion efficiency, several data sets have been normalized and compared in Compression ratio rc CN,17 KT.18 Fig. 15-13 which shows the ratio of fuel conversion efficiency at the given com- pression ratio divided by the efficiency at re = 8, for wide-open-throttle engine A similar comparison of the effect of compression ratio increases on effi- operation. The agreement for re < 14 is good. Over the compression ratio range ciency at part load is shown in Fig. 15-14.19 The figure shows brake fuel conver- that is accessible to SI engines with available fuels (r< < 12), fuel conversion effi- sion efficiency data from engines of different cylinder volume. Both the ciency increases by about 3 percent per unit of compression ratio increase. Note, compression ratio for maximum efficiency and the maximum efficiency depend of course, that engine power increases by about the same amount. on cylinder size. The wide-open-throttle and road-load data (top two curves17) confirm that the increase in efficiency with an increase in the compression ratio at 1300 part load apparently depends on the details of engine operation to a significant 0.46 degree also. For the important compression ratio range of 9 to 11, the relative 1200 20 0.42|- 7). is - imep 15 WOT (CN) mep, kPa 0.38 =1100 664 cm3 10 - RL (CN) 7f. b- Improvement in nb, % bmep - (TO) 497 cm3 FIGURE 15-12 5 443 cm3 0.34- Effect of compression ratio on indi- 1000 309 cm3 cated mean effective pressure and fuel FIGURE 15-14 conversion efficiency. 5.3-dm3 eight- 248 cm3 Relative brake fuel conversion effi- 2000 rev/min WOT cylinder spark-ignition engine at 2000 ciency improvement with increasing MBT timing rev/min and wide-open throttle. Va, cm3/cylinder 0.3- compression ratio of spark-ignition Equivalence ratio and spark timing engines of different displaced volume adjusted for maximum torque for mep -5 6 10 12 14 16 18 20 per cylinder at part throttle (except 10 12 14 16 18 20 900 data; adjusted for minimum fuel con- top curve at WOT).19 RL road load. Compression ratio sumption for efficiency data.17 Compression ratio rc CN, 17 TO.10 844 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 845 efficiency improvement is between 1 and 3 percent per unit of compression ratio (a) increase, depending on cylinder size and operating conditions. The exhaust temperature decreases as compression ratio and efficiency increase until the compression ratio corresponding to maximum efficiency is reached. It has also been shown that heat losses to the combustion chamber walls, as a fraction of the fuel's chemical energy, also decrease as the compression ratio and efficiency both increase.17 The effect of compression ratio changes on NO emissions is small. Some studies show a modest increase in specific NO emissions as the compression ratio increases at constant load and speed; other studies show a slight decrease. Increasing the compression ratio increases exhaust hydrocarbon emissions. Several trends could contribute: increased importance of crevice volumes at high (b) (d) re; lower gas temperatures during the latter part of the expansion stroke, thus producing less HC oxidation in the cylinder; decreasing residual gas fraction, E thus increasing the fraction of in-cylinder HC exhausted; lower exhaust tem- peratures, hence less oxidation in the exhaust system. 15.4 SI ENGINE COMBUSTION CHAMBER DESIGN 15.4.1 Design Objectives and Options There has always been extensive debate over the optimum SI engine combustion chamber design. There are a large number of options for cylinder head and FIGURE 15-15 piston crown shape, spark plug location, size and number of valves, and intake Examples of common spark-ignition engine combustion chamber shapes: (a) bathtub and wedge: (b) port design.2º Debate revolves around issues such as chamber compactness, bowl-in-piston; (c) four-valve pent roof; (d) hemispherical.21 surface/volume ratio, flame travel length, and use of swirl and squish types of mixture motion. Figure 15-15 shows examples of several common types of com- bustion chamber shapes. Over the past few years a consensus has developed Many methods for producing a "fast burn" have been proposed. These which favors faster-burning combustion-chamber designs. A chamber design include ways of making the combustion chamber shape more compact, moving where the fuel burning process takes place faster, i.e ., occupies a shorter crank the spark plug to a more central location within the chamber, using two plugs, angle interval at a given engine speed, produces a more robust and repeatable and increasing in-cylinder gas motion by creating swirl during the induction combustion pattern that provides emission control and efficiency gains simulta- process or during the latter stages of compression. neously. A faster-burning chamber with its shorter burn time permits operation A faster combustion process relative to more moderate burn rate engines with substantially higher amounts of EGR, or with very lean mixtures, within the does result in a direct engine efficiency gain, other factors being equal. The mag- normal constraints of engine smoothness and response. Thus greater emissions nitude of this direct gain is relatively modest. Experimental studies of the effect of control within the engine can be achieved, and at part load at this higher level of an increase in burn rate from moderate to fast at constant engine load, speed, dilution a faster-burning chamber shows an improvement in fuel consumption and mixture composition show that this effect is a few percent at most.23 Com- due to the reduced pumping work, reduced heat transfer (due to lower burned puter simulations of the engine operating cycle confirm these experimental obser- gas temperatures), and reduced amount of dissociation in the burned gases.22 vations: while a decrease in total burn duration from 100 to 60º (slow to The major combustion chamber design objectives which relate to engine moderate burn) does result in a 4 percent decrease in bsfc, a decrease in burn performance and emissions are: (1) a fast combustion process, with low cycle-by- duration from 60 to 20º gives only a further 1.5 percent bsfc decrease.6 cycle variability, over the full engine operating range; (2) a high volumetric effi- Of greater importance is the fact that the faster burn process is more robust ciency at wide-open throttle; (3) minimum heat loss to the combustion chamber and results in the engine being able to operate satisfactorily with much more walls; (4) a low fuel octane requirement. EGR, or much leaner, without a large deterioration in combustion quality. Faster 846 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 847 burning chamber designs exhibit much less cycle-by-cycle variability. This ability Geometry primarily affects combustion through the flame front surface to operate with greater dilution at part load while maintaining a short burn area. It has a lesser effect on combustion development through its influence on duration and low cycle-by-cycle variability, permits much greater control of NO in-cylinder motion. Geometric calculations (see Sec. 14.4.2), based solely on the within the engine with 20 or more percent EGR without any substantial increase assumption that the front surface of the flame can be modeled as a portion of a in HC emissions (see Fig. 11-29), or permits very lean operation. In both cases the sphere centered at the spark plug, provide data on flame front area and the efficiency gain relative to moderate burn rate engines, which must operate with less volume behind the flame front surface (the enflamed volume), contained within dilution, is sizeable.24 the combustion chamber at the appropriate flame radii and piston positions. High volumetric efficiency is required to obtain the highest possible power Flame area varies significantly from one chamber shape to another for a density. The shape of the cylinder head affects the size of valves that can be given enflamed volume. In the example shown in Fig. 14-7, the bowl-in-piston incorporated into the design. Effective valve open area, which depends on valve chamber gives flame surface areas 30 to 45 percent larger than those for the disc diameter and lift, directly affects volumetric efficiency. Swirl is used in many chamber under equivalent conditions around top-center. Hemispherical and open modern chamber designs to speed up the burning process and achieve greater or clamshell chambers showed gains of about 30 percent relative to the equiva- combustion stability. Induction-generated swirl appears to be a particularly lent disc configuration. For a given chamber shape, flame area depends even stable in-cylinder flow. Swirl results in higher turbulence inside the chamber more significantly on plug location. Figure 14-7 shows that shifting the plug from during combustion, thus increasing the rate of flame development and propaga- a side to a center location for the bowl-in-piston chamber increased the peak tion. Generating swirl during the intake process decreases volumetric efficiency. flame area by 150 percent. For hemispherical and open chambers, the increases Heat transfer to the combustion chamber walls has a significant impact on for a similar shift in plug location were 75 and 90 percent, respectively.25 engine efficiency. It is affected by cylinder head and piston crown surface area, by Maps of flame area as a function of radius at different crank angle locations the magnitude of in-cylinder gas velocities during combustion and expansion, by indicate the following pattern. For chamber geometries with side ignition, as the gas temperatures and the wall temperatures. The heat-transfer implications of flame radius increases, the flame area first rises slowly, then remains approx- a combustion chamber should be included in the design process. imately constant, and then decreases slowly to zero. In contrast, chambers with Knock effectively limits the maximum compression ratio that can be used central ignition show, as flame radius increases, a rise in flame area to a peak in any combustion chamber; it therefore has a direct impact on efficiency. Knock during the major part of the flame travel followed by a rapid decrease as the is affected by all the factors discussed above. It is the hardest of all the constraints flame encounters the chamber walls. Moving the plug location toward the center to incorporate into the design process because of its obvious complexity. of the chamber produces a larger increase in flame front area than does making Knowledge of the fundamentals of spark-ignition engine combustion, in- the chamber shape more compact (though this has a positive impact too). cylinder gas motion, and heat transfer has developed to the point where a ration- The effect of chamber geometry on burn rate has been examined using al procedure for evaluating these factors for optimum combustion chamber thermodynamic-based engine cycle simulations with various types of combustion development and design can be defined. The next two sections develop such a model (e.g ., the type developed by Keck and coworkers, see Sec. 14.4.2). Figure procedure. 15-16 shows results from one such study.25 The combustion characteristics of ten different chamber geometries were compared at fixed part-load engine operating conditions. The flame development and propagation phases were separated into 0 15.4.2 Factors That Control Combustion to 10 and 10 to 90 percent mass fraction burned times. These were then normal- Our understanding of the structure of the spark-ignition engine flame as it ized by the equivalent burn times of the slowest burning chamber-the disc with develops and propagates across the combustion chamber (see Secs. 9.3 and 9.4) side ignition. Chamber geometry has the greatest impact on the 10 to 90 percent allows us to relate the physical and chemical factors that control this process to burn time; its effect on 0 to 10 percent time is significant but substantially the relevant engine design and operating parameters. The following factors affect smaller. Total burn times can be reduced by between 20 to 30 percent by opti- the flame development and propagation processes: mizing spark plug location-comparing worst to best location for each chamber shape. Comparing worst and best chamber shapes, total burn time with fixed 1. Geometry. Combustion chamber shape and spark plug location. plug location can be reduced by about 10 percent. Increased turbulence in the unburned mixture at the time of combustion 2. Flow field characteristics. Mean velocity, turbulence intensity, and character- increases the burning rate. Turbulence is usually increased by generating swirl istic turbulence length scale in the unburned mixture during combustion. during the induction process (see Sec. 8.3.2 and below). Cycle simulation 3. Unburned mixture composition and state. Fuel, equivalence ratio, burned gas studies25 indicate that both the duration of the early stage of the burning process fraction, mixture pressure and temperature. and of the main stage decrease when the turbulent velocity at the start of com- 848 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 849 Disc Open of laminar flame speed on temperature, pressure, fuel/air equivalence ratio, and burned gas fraction (residual gas and EGR): see Sec. 9.3.3 and Eqs. (14.33) to (14.35). Table 15.1 compares the burn durations for a stoichiometric mixture, a lean mixture with o = 0.8, and a stoichiometric mixture with 20 percent EGR. Hemi Bowl-in-piston The values of the laminar flame speed at the time of spark are also given (conditions at spark as well as composition are different in each case). The longer burn durations of the more dilute mixtures are clear. Note that EGR as a diluent has a much more deleterious effect on combustion than does air at these approx- Spark plug location imately equal levels of dilution. All the above-described factors-flame geometry, fluid motion, and mixture composition-can vary cycle-by-cycle, and therefore contribute to combustion variability (see Sec. 9.4). Cyclic differences in gas motion in the vicinity of the Disc, side spark plug result in differences in motion of the flame kernel during its early Hemi, side stages of development. Differences in turbulence result in differences in the rates Bowl, side at which the initially smooth surface of the flame kernel becomes wrinkled and Open, side convoluted by the flow. Different initial flame center motions change the geo- Hemi, 1/3 metrical interaction of the flame front with the combustion chamber walls later in Chamber geometries Hemi, center 10-10% burn the flame propagation process. Differences in the amount of fuel, air, and EGR Disc, center which enter each cylinder cycle-by-cycle, the nonuniformity in composition of the Open, top center 0-90% burn entering charge, and any incomplete mixing of that entering charge with the residual gases in the cylinder also contribute to combustion variability. These Bowl, center composition nonuniformities lead to differences in the early stages of flame devel- Open, center 0-90% burn opment. The variations in the amounts of fuel, air, and EGR that enter each cylinder cycle-by-cycle and in the uniformity of that mixture are factors within 0 0.2 0.4 0.6 0.8 1.0 Burn angle ratio the direct control of the engine designer. A fast combustion process reduces cyclic combustion variability for the fol- FIGURE 15-16 lowing reasons. With a faster burn, optimum spark timing is closer to top-center: Comparison of burn angles (0-10 percent burned, 10-90 percent burned, 0-90 percent burned; see Fig. 9-13) for ten different spark-ignition engine combustion chamber geometries and spark plug mixture temperature and pressure at the time of spark are higher, so the laminar locations. Burn angles are normalized by angles for slowest burning chamber: disc with side plug.26 flame speed at the start of combustion is greater. This, combined with the higher turbulence of most fast-burn concepts, results in faster flame kernel development. bustion is increased. The faster combustion process comes primarily from the higher turbulence intensity; however, the decreased characteristic turbulence TABLE 15.1 scale that accompanies the increased turbulence is also significant since it results Effect of excess air and recycled exhaust on burn dura- in a shorter characteristic burning time [see Eq. (14.39) and the accompanying tion text]. It is important to note that the fuel conversion efficiency of higher- turbulence chambers at the same operating conditions can be lower than for Burn durations, degree EGR Sz at 0 ., normal chambers, despite the faster burn rates, due to the higher heat transfer % degree 0-10% 10-90% cm/s that accompanies the higher in-cylinder velocities. For example, predictions based on the combustion model defined by Eqs. (14.33) to (14.35), where the 1.0 340 22 17 75 characteristic mixture speed ur was increased by a factor of two, showed that the 0.8 0 336 26 21 52 1.0 20 324 31 28 23 0 to 10 percent and 10 to 90 percent burn durations decreased by about one- third. However, the indicated fuel conversion efficiency decreased by about 6 400 cm3 per cylinder displaced volume, 80 mm bore, 8.5 compression ratio, percent due to the predicted 15 percent increase in heat transfer.25 disc chamber, center plug location. 1500 rev/min, stoichiometric operation, 0, = spark timing (MBT), inlet pressure 0.5 atm, inlet temperature 350 K, Mixture composition and state affect the burn rate through the dependence S, = laminar flame speed. 26 850 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 851 More rapid initial flame growth results in less variation in flame center motion swirl, are an attractive option. However, the gains in volumetric efficiency are during the critical flame-development phase. The resulting geometric variations offset by a higher cost due to the additional complexity in port and manifold of in the flame front/chamber wall interaction are therefore reduced; this decreases the double passage and the individual throttle valves required in each port for the variations in burn rate that result from these geometric variations. Also, the flow control. faster burning process ends earlier in the expansion stroke. Thus the problem of Swirl can be intensified during compression with bowl-in-piston combustion occasional slow burning cycles, partial burning cycles, and eventually misfire, chambers by decreasing the moment of inertia of the in-cylinder charge as the which occurs with dilute mixtures under normal burning conditions due to quen- piston moves toward top-center, and thereby increasing its angular velocity (see ching of the combustion process as gas temperatures fall during expansion, is Sec. 8.3.3). An advantage here is that the swirl level generated during induction is largely avoided (see Sec. 9.4.3). less than would be required without the compression-produced radially inward motion of the charge. This approach can be used with combustion chamber designs that are axisymmetric and compact. Swirl can also be generated by squish 15.4.3 Factors That Control Performance motion toward the end of compression with a suitable design of chamber. The VOLUMETRIC EFFICIENCY. Combustion chamber shape affects volumetric effi- advantage of this approach is that there is no induction-stroke swirl-generating ciency through its constraints on maximum valve size and through the degree of volumetric efficiency penalty. However, the cylinder head geometries proposed to swirl (if any) that the chamber and port designs produce to achieve the desired date for either intensifying or generating swirl have vertical valve stems, and combustion characteristics. To obtain maximum performance and to reduce hence have smaller valve sizes which in themselves restrict air flow. Also, the pumping losses, the size of the valve heads should be as large as practical; the cylinder head geometry required to generate swirl during compression has a valve sizes that can be accommodated depend on cylinder head layout. Table 6.1 larger surface area than more open chamber designs and, therefore, has signifi- lists the typical maximum valve sizes that can be accommodated into several cantly higher heat losses. common chamber shapes (see Fig. 15-15). The approximate mean piston speed at The impact of conventional radially inward squish motion (see Sec. 8.4) on maximum power is a measure of the maximum air flow that each engine design in-cylinder turbulence, and hence combustion, is unclear. Chambers with signifi- can pump. Note that of the two-valve configurations, the designs with inclined cant squish are also more compact; for this reason alone they would be faster valve stems permit substantially greater maximum air flow. The four-valve pent- burning. roof design, which also has inclined valve stems, is the best of those listed since it accommodates the largest valve and port areas (there are other four-valve head HEAT TRANSFER. The convective engine heat transfer to the combustion designs which are comparable). chamber walls is described by equations of the form of (12.2): e.g ., Eq. (12.21). Swirl can be generated during the intake process through suitable port, The heat-transfer coefficient is usually correlated by expressions of the form of valve, and head design. It requires either that the flow through the intake valve Eq. (12.3), which relate the Nusselt, Reynolds, and Prandtl numbers (see Sec. be directed tangentially into the cylinder so that gas flows through one side of the 12.4). Thus combustion chamber surface area, and especially the surface area in valve opening preferentially (e.g ., through the use of masks to restrict flow at the contact with the burned gases, is important. Gas velocity is also important; it mask location or through the use of a tangentially directed port or a flow deflec- influences the heat-transfer rate through the Reynolds number. Various charac- tor in the port just upstream of the valve), or requires the use of a helical intake teristic velocities have been used in the Reynolds number to scale heat transfer: port that imparts an angular velocity to the flow before it enters the cylinder. In mean piston speed, mean in-cylinder gas velocity, turbulence intensity, either either case the inlet flow enters the cylinder with higher velocity than it would individually or in combination. Both of these variables, area and velocity, are have in the absence of swirl; hence the pressure drop across the valve is affected by combustion chamber design. increased, and maximum air flow through the cylinder is reduced. Well-designed Studies of engine performance using thermodynamic-based simulations of helical swirl-generating ports (see Sec. 8.3.2) appear to be the best way to create the engine's operating cycle (see Sec. 14.4) provide data that indicate the impor- swirl. However, geometric and production constraints often prevent the incorpo- tance of changes in heat transfer. At part-throttle operating conditions, such ration of such ports into the cylinder head design, and other swirl-generating simulation calculations show that a 10 percent change in combustion chamber methods must be used. The engine maximum-power penalty associated with gen- heat losses results in a change of between 2 and 5 percent in brake specific fuel erating significant swirl is of order 5 to 10 per cent. consumption; an average fuel consumption change of about one-third the magni- Since swirl is only required at part-throttle operation when enhancement of tude of the heat-transfer change (and of opposite sign) is an appropriate rule of the burn rate is most critical and is not usually required at full throttle when the thumb.25, 27 At wide-open throttle, the effect on mean effective pressure is compa- flow restriction penalty is most significant, induction systems with a separate rable: a 10 percent change in heat transfer results in about a 3 percent change in passage for the part-throttle air flow, where only this separate passage generates bmep. 852 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 853 This impact of heat transfer on engine efficiency and performance under- 60 Spark lines the importance of combustion chamber details that affect heat transfer. For 50 advance for Torque loss the chamber shapes shown in Fig. 15-16, the total heat losses as a fraction of the borderline 1% 40 knock fuel's energy, at fixed engine speed and intake conditions, were also calculated. MBT Both chamber shape and spark plug details affect heat losses since together these 30 2% govern the surface area of the hot burned gases in contact with the walls. The 7% open and hemispherical chambers had least heat transfer. Geometries such as the Spark advance, deg 20 10% --- bowl-in-piston, which obviously have a higher surface area, had about 10 percent 15% 10 Retard higher heat transfer. The effect of shifting the plug from a side to center location FIGURE 15-17 depended on chamber shape. Open and bowl-in-piston chambers showed little Automatic spark Relation between spark advance, speed, change; the hemispherical chamber showed a 4 percent reduction. Given a advance and torque loss, for spark-ignition general chamber shape choice, the details of the actual design are important also; engine at wide-open throttle, showing it is easy to add substantial surface area with piston cutouts, plug bosses, and knock limit for specific gasoline and 0 2000 4000 cylinder head masking or squish regions which will deteriorate chamber per- typical spark-advance schedule that Engine speed, rev/min avoids knock problems. 28 formance to a measurable degree. Higher in-cylinder velocities affect heat-transfer rates through the Reynolds number term in the heat-transfer coefficient correlation. Swirl- and squish- stant specified percentage torque reductions. The upper solid line traces the spark generated flows increase in-cylinder gas velocities and will, therefore, increase advance for borderline knock with a particular commercial gasoline. To avoid heat-transfer rates. knock with this fuel, the spark advance must be set to lose one percent of engine torque at 800 rev/min, with the torque loss diminishing to zero at 1200 rev/min. Above that speed this particular fuel allows operation at MBT timing without 15.4.4 Chamber Octane Requirement knocking. The lower solid curve represents a typical spark-advance schedule at Knock limits an engine's compression ratio, and hence its performance and effi- WOT. It lies below the borderline knock advance (and results in a significant ciency. The more fundamental aspects of knock were reviewed in Sec. 9.6. Knock torque loss) for the following reasons. One is that different commercial gasolines occurs when the end-gas autoignites prior to its being burned up by the normal with the same research octane number can respond differently to variations in flame-propagation process. The tendency to knock depends on engine design and engine operating conditions. Calibrating the engine (i.e ., specifying the schedules operating variables which influence end-gas temperature, pressure, and time for spark advance, A/F, and EGR) must be done with sufficient margin of conser- spent at high values of these two properties before flame arrival. vatism to avoid objectionable knock with the normal range of commercial gas- The presence or absence of knock in an engine depends primarily on the olines over the full operating conditions of the engine. A second reason is antiknock quality of the fuel, which is defined by the fuel's octane number (see Sec. engine-to-engine production variability despite the close dimensional tolerances 9.6.3). It determines whether or not a fuel will knock in a given engine under of modern production engineering. For example, the effective compression pres- given operating conditions: the higher the octane number, the higher the resist- sure in each cylinder of a multicylinder engine is not identical, due to geometric ance to knock. The octane number requirement of an engine is defined as the and ring-pack behavior differences. The cylinder with the highest compression minimum fuel octane number that will resist knock throughout its speed and pressure is most knock-prone. Allowing for corresponding effects of cylinder-to- load range. The following factors affect an engine's octane requirement: (1) com- cylinder variations in A/F, EGR rates, and spark timing, it is obvious that for a position of the fuel; (2) chamber geometry and size; (3) charge motion; (4) spark- given operating condition in a multicylinder engine, one cylinder is more likely to advance curve; (5) inlet air, intake manifold, and water jacket temperatures; (6) knock than the others. It is that cylinder which limits the spark advance.+ A third carburetor or fuel-injector air-fuel ratio calibration; (7) the ambient conditions- reason for the discrepancy between actual spark-advance calibration and the pressure, temperature, and relative humidity-during the requirement determi- knock limit for a given engine and fuel is the octane requirement increase associ- ated with the buildup of deposits on the combustion chamber walls over nation. The following illustrates the interaction between fuel factors and engine extended mileage (see Sec. 9.6.3). operating variables. Figure 15-17 shows the relation between spark advance, torque, and speed in an engine operating at wide-open throttle. The dashed lines, determined with a fuel of sufficiently high octane rating to avoid knock, show + There is no assurance that the same cylinder will be the principal offender in all engines of the same MBT timing as a function of speed, along with the spark-advance limits for con- model, nor in a given engine at all operating conditions. 854 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 855 In the example shown above, it was the problem of knock at low engine Very open chambers - Disc: Low turbulence High turbulence speed which required the spark advance calibration to be retarded. Whether low -, Moderately open Bathtub o medium -, or high-speed knock is the limiting factor in a particular engine chambers 4-valve A depends on the sensitivity of the fuel, on engine design features, and especially Compact chambers Piston O Head upon the engine's spark-advance requirements for MBT. The knock-limited 110 spark advance determined from road octane rating tests will vary with engine speed and fuel sensitivity, as shown in Fig. 15-18. Low sensitivity fuels will toler- 100 ate more severe engine operating conditions and vice versa. Figure 15-18b, c, and d shows a typical engine spark-advance characteristic superposed on the knock- Octane requirement, RON limited spark-advance plot. Depending on the fuel sensitivity and shape of the 90 spark-advance curve, the knock region may occur at low, medium, or high speed (or not at all). 30 FIGURE 15-19 It will be apparent from the above discussion that defining the effect of Octane requirement (gasoline research octane combustion chamber geometry on knock can only be done in an approximate number), at wide-open throttle and MBT fashion. The importance of fuel composition details, differences in engine design, 70 timing, to avoid knock as a function of com- 9 11 13 pression ratio for various combustion chamber the variability between engines of the same type, and the effect of deposits all Compression ratio designs. 10 make the quantification of trends as chamber design is varied extremely difficult. One of the most important chamber variables is the compression ratio. Figure 15-19 shows the relationship between the octane requirement and com- (or light) knock coinciding with MBT timing at the given speed. As is well pression ratio for a number of combustion chambers. The octane requirement known, the octane requirement increases with increasing compression ratio; there was defined as the research octane number of the fuel required to operate the are, however, differences in the octane requirement between different types of engine at WOT with the weakest mixture for maximum power with borderline chamber at the same compression ratio. The chambers studied were disc-shaped chambers, bathtub and four-valve (open chambers with squish) and compact high compression ratio chambers (bowl or cup-type chambers in the piston crown or in the cylinder head around one of the valves). In the 9 to 11 compression ratio Low range there are only modest differences between the chambers studied. At higher sensitivity Low sensitivity compression ratios, 11 to 13, the compact chambers show a lower octane require- Knock-limited spark advance Knock-limited spark advance Engine spark ment which gives them a 1 to 2 compression ratio advantage over the more open advance chambers. This advantage for the compact (and high-turbulence) chambers comes High sensitivity characteristic largely from the increased heat-transfer rates in these chambers. Whether the * Region of knock higher turbulence is generated during intake or at the end of the compression Engine speed Engine speed stroke, it increases the heat transfer from the end-gas, reducing its temperature and therefore its propensity to knock. However, this higher heat transfer also reduces engine power and efficiency, so the benefits of the compression ratio advantage are reduced. There is some increase in the knock-limited compression ratio with a given fuel as burn time is decreased by using one, two, three, and Medium sensitivity High sensitivity then four spark plugs simultaneously, with a given chamber geometry, but the Knock-limited spark advance Knock-limited spark advance Engine Region effect is much smaller than the differences suggested by Fig. 15-19.23 advance Engine characteristic spark of knock Spark plug location within the chamber is an important factor affecting Region of knock advance octane requirement. More centrally located plug positions shorten the flame characteristic travel path to the cylinder walls and decrease the time between spark discharge Engine speed Engine speed and flame arrival at the end-gas location. This decreases the octane requirement. FIGURE 15-18 The position of the spark plug in relation to the exhaust valve is also important: Diagrams showing knock-limited spark-advance curves for fuels of different sensitivity and how these it is advantageous to burn the unburned mixture which has been heated by can give low -, medium -, and high-speed knock in the same engine.29 contact with the hot exhaust valve early in the combustion process. 856 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 857 TABLE 15.2 advance has a major impact on knock; since it is also easy to adjust, it is the Engine conditions affecting octane number requirement engine variable most commonly used to control knock. Studies show that typi- Octane number requirement cally 0.5 to 1.0 RON reduction is achieved per degree of retard.30 Atmospheric Octane number requirement tends to go up when: tends to go down when: conditions-pressure, temperature, and humidity-all affect the octane number requirement. 31 1. Ignition timing is advanced. 1. Ignition timing is retarded. The fuel/air equivalence ratio affects the octane requirement of an engine. 2. Air density rises due to 2. Engine is operated at higher The highest requirement is for slightly rich mixtures; increasing richness and supercharging or a larger altitudes or smaller throttle throttle opening or higher opening or lower barometric leanness about this point decrease the octane requirement substantially. barometric pressure. pressure. Figure 15-20 shows the knock-limited compression ratio as a function of the rela- 3. Humidity or moisture content 3. Humidity of the air increases. tive air/fuel ratio (1 = 1/0; 1 > 1 for lean mixtures) for conventional and high- of the air decreases. turbulence chambers, for two fuels with different octane ratings. Substantially 4. Inlet air temperature is 4. Inlet air temperature is higher compression ratios can be used with lean mixtures, especially with the increased. decreased. 5. Coolant temperature is 5. Fuel/air ratio is richer or high-turbulence chamber which extends the lean limit. The coolant temperature raised. leaner than that producing affects the octane requirement. Higher coolant temperature increases the inlet maximum knock. mixture temperature, and reduces heat losses from the end-gas to a modest 6. Antifreeze (glycol) engine 6. Exhaust gas recycle system degree. coolant is used. operates at part throttle. 7. Engine load is increased. 7. Engine load is reduced. 15.4.5 Chamber Optimization Strategy Operating variables that affect the temperature or pressure time histories of The discussion in the previous sections suggests that the following sequence of the end-gas during combustion or the basic autoignition characteristics of the steps in a combustion chamber development process is most logical. First should unburned fuel, air, residual mixture will also affect the engine's octane require- come the selection of the best chamber geometry. Geometric optimization can ment. The most important additional variables which increase or decrease octane result in substantial benefits and carries no significant penalties. Chamber requirement in a consistent manner are listed in Table 15.2. Relative spark geometry involves cylinder head and piston crown shape, and plug location. Open chambers such as the hemispherical or pent-roof cylinder head, and clam- shell, with near central plug location, give close to the maximum flame front Moderate turbulence High turbulence 15 surface area (and hence a faster burn), have the lowest chamber surface area in 15 Fuel Fuel contact with the burned gases (and therefore the lowest heat transfer), and have RON = 100 RON = 100 - RON = 92 inclined valves which give high volumetric efficiency. Spark plug location close to RON = 92 13 13 the center of the chamber is especially important in obtaining a fast burn rate. More compact chamber shapes than the open chambers listed above, such as the Knock limit bowl-in-piston or chamber-in-head designs, do produce a somewhat faster burn, 11 /11 but with lower volumetric efficiency and higher heat losses. Following this first step, two problem areas may remain: the chamber may Compression ratio have a higher octane requirement than is desired and the burn rate may not be 19 fast enough to absorb the high dilution required at part load to meet the emis- sions and fuel consumption goals. Lean mixture limi Inflammability limit Positioning the spark plug as close to the center as possible will have 7 -7 reduced the octane requirement for that particular chamber shape. Depending on chamber design details, some squish area could be introduced. However, the per- ceived octane advantage of chamber designs which contain substantial squish is 0.4 0.8 1.2 1.6 2.0 0.4 0.8 1.2 1.6 2.0 offset, at least in part, by their higher heat losses. A unit compression ratio Relative air/fuel ratio > increase results in a 3 percent or less increase in efficiency at part load. However, if the measures required to increase the compression ratio from, say, 9 to 10 FIGURE 15-20 Knock limits and lean engine operating limits as function of compression ratio and relative air/fuel resulted in a 10 percent increase in heat transfer, engine efficiency would not ratio 1 (1 = 1/o) for moderate and high-turbulence engine combustion chambers.32 improve. 858 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 859 The next step should be to reduce the cyclic variability in the combustion Mean piston speed, m/s process to the maximum extent feasible, by improving the uniformity of the 4 6 8 10 intake fuel, air, and EGR mixture. Delivery of equal amounts of each of these 800 -420 constituents to each cylinder, provision for good mixing between constituents in the intake manifold and port, and accurate control of mixture composition 380 700L during engine transients are all especially important. Also important is achieving -340 closely similar flow patterns within each engine cylinder during intake so as to 600 220 300 obtain equal burn rates in all cylinders. Attention to these intake process and bmep, kPa 225 g/kW . h Torque, N.m 500 260 mixture preparation details will always improve engine operation and carries no 230 performance penalties. 400 220 240 FIGURE 15-21 However, the burn rate may still not be fast enough, especially during the 250 180 Performance map for 6.54-dm3 eight-cylinder air- critical early stages of flame development, and cyclic variability may still be too 300 260 cooled naturally aspirated medium-swirl DI diesel high to meet the engine's performance goals. Then higher turbulence levels 140 engine. Contours of constant bsfc in grams per during combustion must be achieved. This is usually best done by creating swirl 1000 2000 000 3500 kilowatt-hour shown. Bore = 102 mm, stroke = during the induction process. The appropriate method for introducing swirl will Engine speed, rev/min 100 mm, r = 18. Multihole fuel nozzle.33 depend on any geometric manufacturing constraints and cost issues. With no geometric constraints, use of helical swirl-generating ports or a divided intake- discussed in Sec. 15.2. Here we examine the part-load behavior of various types of port system with valves to control the flow at light load are likely to have the naturally aspirated diesel engines. lowest power penalties. It is especially important that only the minimum addi- As with SI engines (see Sec. 15.3.3), performance maps where bsfc contours tional turbulence required to achieve the performance objectives be added at this are plotted on a graph of bmep versus engine speed are commonly used to stage. Higher than necessary gas velocities within the cylinder result in excessive describe the effects of load and speed variations. Figure 15-21 shows the per- heat losses and low volumetric efficiency. formance map for an air-cooled four-stroke cycle medium-swirl naturally aspirat- In summary, to meet the objectives of a fast, repeatable, and robust com- ed DI diesel (similar to the engine in Fig. 1-23). Maximum rated power for this bustion process with high volumetric efficiency, low heat transfer, and acceptable 6.54-dm3 displacement engine at 3200 rev/min is 119 kW, maximum bmep at octane requirement, combustion chamber development should proceed through 2000 rev/min is 784 kPa, and minimum bsfc (at 1600 rev/min and 580 kPa bmep) the following steps. is 220 g/kW . h, which corresponds to a brake fuel conversion efficiency of 38.5 percent. The gross indicated fuel conversion efficiency would be about 48 percent. 1. Optimize the chamber geometry within the design constraints for the Figure 15-22 shows the performance map for a small high-swirl DI diesel maximum flame front area, minimum burned gas/chamber wall contact area, which uses the M.A.N. combustion system with a single fuel jet sprayed tangen- and largest valve size. tially into the swirling air flow. Due to the higher speed and higher swirl than the 2. Obtain additional reductions in the cyclic combustion variability by improv- ing mixture distribution and uniformity and by creating flow patterns into each cylinder that are essentially identical. Mean piston speed, m/s 8 12 3. Achieve any additional improvement in burn rate and cyclic variability 800 6 10 14 required to meet objectives by increasing turbulence to the minimum extent. This is usually best done by creating swirl during the induction process. 600 bsfc = 246 g/kW .h 250 15.5 VARIABLES THAT AFFECT CI bmep, kPa 400 260 ENGINE PERFORMANCE, EFFICIENCY, 270- 280 FIGURE 15-22 AND EMISSIONS 200H 320 Performance map for 1.47-dm3 four-cylinder natu- 350 15.5.1 Load and Speed 290 300 rally aspirated DI diesel engine with high-swirl single-hole-nozzle M.A.N. combustion system. The performance of a naturally aspirated DI heavy-duty truck diesel engine and OL 1000 2000 3000 4000 5000 Contours of constant bsfc in grams per kilowatt- hour shown. Bore = 76.5 mm, stroke = 80 mm, a small IDI engine at full load over the engine speed range have already been Engine speed, rev/min r = 18.5.34 860 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 861 15 larger DI engine in Fig. 15-21, the maximum bmep is slightly lower. The best bsfc is about 10 percent higher largely due to higher friction mep, but in part due to Effects of: higher heat losses resulting from the less favorable surface/volume ratio of the Retar smaller bore engine and higher swirl, and lower heat-release rate of the M.A.N. ₹ 10- Late burn system. Note that the maximum mean piston speed for this engine, 13.3 m/s Heat loss Improvement, % at 5000 rev/min, is comparable to that of the larger medium-swirl engine in Pumping Fig. 15-21 (10.7 m/s). 5 Figure 15-23 gives the performance characteristics of an automotive natu- FIGURE 15-24 rally aspirated swirl-chamber IDI diesel engine. Maximum bmep values are Factors which improve the indicated efficiency of usually higher than those of equivalent size DI engines because without the need naturally aspirated small DI diesel combustion 10 20 30 40 50 60 70 to generate swirl during the intake process, the intake port and valve are less 80 systems relative to IDI swirl-chamber combustion Air/fuel ratio system, as a function of A/F or load.36 restrictive and volumetric efficiency is higher, and because the IDI engine can be run at lower A/F without smoking. The best bsfc values are usually some 15 percent higher than values typical of equivalent DI engines. The best brake fuel shape and when plotted against S ,, are quantitatively comparable. The increase in conversion efficiency of the engine of Fig. 15-23 is 32.5 percent. bsfc from the minimum value with increasing speed at constant load is due to the Comparisons between naturally aspirated DI and IDI diesel engines of increase in friction mep, partly offset by the effect of decreasing importance of closely comparable design and size indicate that the DI engine is always more heat losses per cycle on efficiency. The increase in bsfc with decreasing load at efficient, though the benefit varies with load. At full load, differences of up to 20 constant speed is dominated by the decreasing mechanical efficiency as bmep is percent in bsfc have been noted, especially in engines with larger displacement reduced. The indicated fuel conversion efficiency increase as the fuel/air equiva- per cylinder. At part load, the gain is less-of order 10 percent. Comparisons lence ratio is decreased partly offsets this. The trends in bsfc when increasing load should be made at equal emission levels, a task that is difficult to accomplish in at constant speed and increasing speed at constant load from the minimum are practice. Emission control with the DI engine is more difficult, so this constraint more modest. They are the net results of (1) the increase in mechanical efficiency reduces the benefit somewhat. Figure 15-24 shows a breakdown of the indicated and decrease in indicated fuel conversion efficiency as the load increases and (2) efficiency differences between the two systems. At full load (A/F = 18 to 20) the decreasing indicated efficiency due to increasing importance of heat losses and IDI suffers a penalty of about 15 to 17 percent due in large part to the retarded increasing mechanical efficiency as the speed decreases. The enrichment of the timing of the IDI combustion process and its long, late-burning, heat-release mixture at high load and low speed of spark-ignition engines is, of course, absent. profile. At light load, about 300 kPa bmep (A/F = 50), these combustion effects Figure 15-25 shows the effect of load on NO, and HC emissions for natu- are small and the indicated efficiency penalty of the IDI (around 5 to 7 percent) is rally aspirated DI and IDI diesel engines. For the DI engine NO, concentra- due to the higher heat losses associated with the larger surface area and high- tions rise steadily as the fuel/air ratio increases with increasing bmep at constant velocity flow through the connecting nozzle of the divided-chamber geometry injection timing. The increasing quantity of fuel injected per cycle results in an and due to the pumping pressure loss between the main and auxiliary cham- increasing amount of close-to-stoichiometric combustion products near the peak bers, 36 pressure and temperature (see Sec. 11.2.4). The IDI engine shows a similiar trend Note that all these diesel engine performance maps are similar in general except that, at high load, NO, concentrations level off. These characteristics do not change substantially with engine speed. The IDI engine shows significantly lower HC emissions than the DI engine. The high HC at idle and light load are Mean piston speed, m/s thought to result from fuel mixing to too lean an equivalence ratio. If diesel 6 8 10 12 14 engines are overfueled at high load, HC emissions then rise rapidly. These HC 800 mechanisms are described in Sec. 11.4.4. Injection timing affects NO, and HC 260 g/kW . h emissions significantly, as discussed in Sec. 15.5.2 below. 500 280 Figure 15-26 shows smoke and particulate mass emissions from a naturally bmep, kPa 400 .300 FIGURE 15-23 aspirated IDI engine. Rapidly increasing black smoke at very high load limits the 320 360 Performance map for 1.987-dm3 five-cylinder natu- 200 420 maximum bmep that a diesel engine can produce. On a specific emission basis 520 rally aspirated IDI swirl-chamber diesel engine. Road load .1100- Contours of constant bsfc in grams per kilowatt- [Eq. (2.36)], the particulates typically show a U-shaped behavior due to the pre- 1000 2000 3000 4000 5000 hour shown. Bore = 76.5 mm, stroke = 86.4 mm, dominance of hydrocarbons in their composition at light load and of carbon at Engine speed, rev/min re = 23.35 high load. 38 862 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 863 Indirect injection Static injection timing 280- Direct injection -1º 275 - +1º. Base bsfc, g/kW .h bsHC, g/kW.h Base 270- Pintle nozzle 265- 10 20 30 0 20- 30 EGR rate, % Orifice EGR rate, % +1º Base TT TT M bsNO ,, g/kW.h 2 - 1000 2000 +1º - smoke number Bosch Base. Injection timing -1º Injection timing 18º BTC of 18º BTC - 12º BTC 10 20 30 0 10 20 30 800 · 12º BTC 1500 EGR rate, % EGR rate, % FIGURE 15-27 600 IDI Brake specific HC, NO, and fuel consumption, and smoke emissions, as a function of percent recycled HC, ppm C1 1000 NOx, ppm (dry) exhaust for 2.4-dm3 four-cylinder high-swirl DI diesel engine at 1250 rev/min and 255 kPa bmep.39 400 - -- 500 Recycled exhaust gas, at part load, can be used to reduce diesel engine NO, 200 DI emissions. Note that since diesel engines operate with the air flow unthrottled, at part load the CO2 and H2O concentrations in exhaust gas are low; they are essentially proportional to the fuel/air ratio. Because of this, high EGR levels are 200 400 600 200 400 600 0 required for significant reductions in NO, emissions. Figure 11-18 shows how bmep, kPa bmep, kPa NO, concentrations decrease as a DI diesel engine inlet air flow is diluted at a FIGURE 15-25 constant fueling rate. The dilution is expressed in terms of oxygen concentration Effect of load on naturally aspirated diesel engine NO, and HC emissions at rated speed, with two in the mixture after dilution. Figure 15-27 shows how the EGR affects specific injection timings. Direct-injection and indirect-injection (prechamber) combustion systems. Six- cylinder, 5.9-dm3 displaced volume, engine. DI: re = 17, rated speed = 2800 rev/min; IDI: r = 16.7, NO, and HC, fuel consumption, and smoke for a small high-swirl DI diesel engine at typical automobile engine part-load conditions. Effective reduction of rated speed = 3000 rev/min.37 bsNO, is achieved and modest reductions in bsHC, with only a slight increase in bsfc. However, smoke increased as the EGR rate increased.39 3 60 15.5.2 Fuel-Injection Parameters 2º Bosch smoke number Fuel-injection timing essentially controls the crank angle at which combustion 10º starts. While the state of the air into which the fuel is injected changes as injec- 3 O FIGURE 15-26 tion timing is varied, and thus ignition delay will vary, these effects are predict- 2 Smoke (Bosch smoke number) and able (see Sec. 10.6.4). The fuel-injection rate, fuel nozzle design (including number - 60 particulate mass emissions (in grams of holes), and fuel-injection pressure all affect the characteristics of the diesel fuel Particulates, g/kW . h 10º per kilowatt-hour) as a function of spray and its mixing with air in the combustion chamber. load and injection timing for six- cylinder 3.7-dm3 IDI swirl-chamber Figure 15-28 shows the effect on performance and emissions of varying 0 800 diesel engine at 1600 rev/min (no injection timing, in (a) a medium-swirl DI diesel engine and (b) an IDI engine. At 0 200 400 600 bmep, kPa EGR). 38 fixed speed and constant fuel delivery per cycle, the DI engine shows an optimum 864 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 865 15 Swirl-chamber IDI diesel Medium-swirl DI diesel Retarding timing generally increases smoke, though trends vary signifi- Load: - 100% --- 0% cantly between different types and designs of diesel engine. Mass particulate emis- 120 sions increase as injection is retarded. 10 NO, 80- Particulates The injection rate depends on the fuel-injector nozzle area and injection Particulates, g/h 40 pressure. Higher injection rates result in higher fuel-air mixing rates, and hence 5t higher heat-release rates (see Sec. 10.7.3). For a given amount of fuel injected per of cylinder per cycle, as the injection rate is increased the optimum injection timing Bosch smoke number CO Smoke HC 600 moves closer to TC. The effects of injection rate and timing on bsfc in a naturally 0 bsfc, g/kW .h smoke bmep, kPa Brake specific emissions, g/kW.h aspirated DI diesel engine are shown in Fig. 15-29. The higher heat-release rates 800! HC, ppm C1 400 bmep and shorter overall combustion process that result from the increased injection 700 TTTTT 200 600 HO rate decrease the minimum bsfc at optimum injection timing: however, a limit to 3 of 400 these benefits is eventually reached. number Bosch 2 Smoke 200 NOT, ppm Increasing the injection rate increases NO, emissions and decreases smoke NO, or particulate emissions. The controlling physical process is the rate of fuel-air 400 0 260 T bsfc 9.0 mixing in the combustion chamber so, at constant fuel injected per cylinder per 240 bsfc bsfc, g/kW .h 350 cycle, both increased injection pressure at fixed nozzle orifice area (which reduces 8.0 Fuel, mm3/stroke Fuel rate 220 injection duration) and reduced nozzle area at fixed injection duration produce 300L 7.0 30 25 20 15 10 20 15 10 : 0 these trends.42 Injection timing, deg BTC Injection timing, deg BTC The engine designer's goal is obviously to achieve the best bsfc possible (a) (b) FIGURE 15-28 Effect of start-of-injection timing on diesel engine performance and emissions. (a) Medium-swirl DI Injection rate, Injection period diesel engine with deep combustion bowl and four-hole injection nozzle, 2600 rev/min, fuel delivery mm3/deg for dn = 0.28 mm 75 mm3/cycle, fuel/air equivalence ratio 0.69.37 (b) Swirl-chamber IDI engine, 2500 rev/min, 0 and 100 400 2.5 24ºCA percent load.40 3.0 20 3.5 17.1 4.0 15 bsfc and bmep at a specific start of injection for a given injection duration.+ The 4.5 13.3 5.0 IDI engine experiments are at fixed bmep; here, bsfc at full load and fueling rate 12 350 at idle show a minimum at specific injection timings. Injection timing which is more advanced than this optimum results in combustion starting too early before o TC; injection retarded from this optimum results in combustion starting too late. Injection timing variations have a strong effect on NO, emissions for DI bsfc, g/kW . h engines: the effect is significant but less for IDI engines. Retarded injection is 300 commonly used to help control NO, emissions. It gives substantial reductions, initially with only modest bsfc penalty. For the DI engine, at high load, specific 10 % HC emissions are low and vary only modestly with injection timing. At lighter 3.3 % loads, HC emissions are higher and increase as injection becomes significantly 250 Minimum bsfc locus retarded from optimum. This trend is especially pronounced at idle. For IDI diesel engines HC emissions show the same trends but are much lower in magni- tude than DI engine HC emissions.41 Figure 15-25 supports this discussion.37 30 20 10 0 Injection timing, deg BTC FIGURE 15-29 + This optimum injection timing gives maximum brake torque, though the designation MBT timing Effect of injection timing and injection rate on bsfc for 0.97-dm3 single-cylinder naturally aspirated DI is less commonly used with diesels than with SI engines. diesel engine with swirl. 2000 rev/min, 60 mm3 per stroke fueling rate.42 866 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 867 swirl level. Figure 15-31 shows the effects of swirl and injection-timing variations on bsfc and emissions of a DI engine of 1.36 dm3 per cylinder displacement with a toroidal bowl-in-piston chamber (see Fig. 10-3b). The swirl ratio [Eq. (8.28)] 5 was varied using shrouded inlet valves with shrouds of different subtended angle (60 to 120º). The injection timing which gives minimum bsfc shifts toward TC as 4 the swirl ratio increases due to the decreasing total combustion duration. The Bosch smoke number minimum bsfc was achieved with a swirl ratio of 6 to 7: while higher swirl levels continue to increase fuel-air mixing rates, heat transfer increases also and even- tually offsets the mixing rate gain. Particulate and CO emissions decrease as swirl FIGURE 15-30 increases due to more rapid fuel-air mixing. NO, emissions increase with increas- Tradeoff between NO ,, and smoke emissions for ing swirl. At constant injection timing, however, about half the increase is due to 1 quiescient single-cylinder DI diesel engine with bore = 140 mm, stroke = 152 mm, the effect of injection advance relative to the optimum timing and half to the Te = 14.3, eight-hole injector nozzle. Various speeds, fueling shorter combustion process.44 Similar trends have been observed as swirl is 200 400 600 800 1000 1200 rates, injection timings, injection pressures, % varied with the M.A.N. single-hole-nozzle diesel combustion system of Fig. 10-1c. NO ,, ppm EGR; constant A/F = 25.43 In production engines, the various types of port design shown in Fig. 8-13 can be used to generate swirl during the induction process. Of these, the helical ports are most effective at producing relatively uniform high swirl with the with emission levels low enough to satisfy the constraints imposed by emission minimum loss in volumetric efficiency. standards. The variations of bsfc, NO ,, and particulate emissions described The geometry of the bowl-in-piston combustion chamber governs the above involve tradeoffs that make achieving this goal especially difficult. One extent to which induction-generated swirl is amplified during compression. The well-established tradeoff is between bsfc and bsNO ,. Injection retard from flow field in the bowl during fuel injection is also dependent on the interaction optimum injection timing decreases bsNO, at the expense of an increase in bsfc. between this swirling flow and the squish motion which occurs as the top of the A second important tradeoff is that between NO, and particulate emissions, illus- piston crown approaches the cylinder head (see Sec. 8.4). Various types of bowl- trated for a DI diesel engine in Fig. 15-30. Smoke is plotted versus NO, for a in-piston design for multihole fuel nozzle DI engines are shown in Fig. 15-32 (for range of speeds, loads (fuel per cycle), injection timings, injection pressures, and the M.A.N. single-hole-nozzle system a spherical bowl is used; see Fig. 10-1c). EGR rates. The air/fuel ratio was maintained constant at 25 (( = 0.58). The More conventional designs (e.g ., Fig. 15-32a) have the bowl sides essentially figure indicates that for a well-optimized DI diesel engine, the smoke nitric oxide parallel to the cylinder liner. Note that it is often necessary to offset the bowl axis tradeoff is relatively independent of engine speed, injection rate, injection timing, from the cylinder axis and the injector nozzle hole locations from the bowl axis, and amount of EGR. A given reduction in one of these pollutants through chang- ing any one of these variables results in a given increase in the other pollutant. This tradeoff exists for essentially all types of diesel engine, though the magnitude 10.6 De depends on engine details. 2000- 0.2 8 1600- NO ,, ppm 15.5.3 Air Swirl and Bowl-in-Piston Design 1200- 800 Increasing amounts of air swirl within the cylinder (see Sec. 8.3) are used in "Swirl ratio = 2 41.0 direct-injection diesel engines, as engine size decreases and maximum engine 4 ASwirl ratio = 4| speed increases, to achieve adequately fast fuel-air mixing rates (see Sec. 10.2.1). -0.6 Smoke, g/m3 In these medium-to-small size engines, use of a bowl-in-piston combustion -0.2 Swirl ratio = chamber (Fig. 10-1b and c) results in substantial swirl amplification at the end of 500- FIGURE 15-31 the compression process (Sec. 8.3.3). Here, the impacts of varying air swirl on the Effect of air swirl on bsfc and emissions of single- performance and emissions characteristics of this type of DI engine are reviewed. bsfc, g/kW.h . 400 cylinder DI diesel engine with toroidal bowl-in- Since air swirl is used to increase the fuel-air mixing rate, one would expect piston chamber. 1.36-dm3 displacement, re = 16, bowl diameter/bore = 0.5, 2000 rev/min, full load. the overall duration of the combustion process to shorten as swirl increases and 300+ 30 20 10 TC Swirl ratio measured in bowl-in-piston at injec- emissions that depend on the local fuel/air equivalence ratio to be dependent on Injection timing, deg BTC tion.44 868 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 869 15.º into the expansion stroke.47 Reentrant chambers usually achieve lower HC and smoke emissions and slightly lower bsfc, especially at retarded injection timings. 4-hole Square cavity chambers (see Fig. 15-32c) are also used with swirl to achieve nozzle. low emissions in smaller-size DI diesel engines. The interaction between the swirl Piston and the chamber corners produces additional turbulence which, with fuel injected 750 into the corners as shown, achieves a more uniform mixture within the bowl. The air flow field within bowl-in-piston combustion chambers when fuel Swirl direction injection occurs is highly complex. Certain generalizations hold: e.g ., reducing the Exhaust Helical port Combustion bowl diameter at a constant compression ratio increases the swirl levels in the chamber bowl at TC [see Eq. (8.35) and the accompanying text] which decreases smoke and increases NO, and HC emissions.37 However, the squish-swirl interaction is Crankshaft centerline Swirl difficult to unravel, especially with the off-center bowls often required due to the constraints on injector location caused by the valves. Figure 14-33 gives an example of such a flow. It shows velocity vectors and turbulence intensities in `Inlet two orthogonal bowl-diametral planes within an off-center reentrant bowl as TC c) is approached in a small high-swirl DI engine. The off-center bowl location (a) (b) (c) coupled with the swirl-squish interaction cause substantial asymmetry in the flow FIGURE 15-32 within the bowl. Various bowl-in-piston chamber designs for DI diesel engines with swirl: (a) conventional straight- sided bowl,37 (b) reentrant bowl,45 (c) square reentrant bowl.46 . ... 15.6 SUPERCHARGED AND due to the geometric constraints imposed by the valves. An alternative design TURBOCHARGED ENGINE PERFORMANCE with a reentrant bowl (Fig. 15-32b) is sometimes used to promote more rapid fuel-air mixing within the bowl. The squish-swirl interaction with highly reen- The equations for power, torque, and mep in Sec. 2.14 show that these engine trant bowl designs differs markedly from the interaction in nonreentrant bowls. performance parameters are proportional to the mass of air inducted per cycle. Figure 15-33 shows the two different flow patterns set up in a diametral plane. This depends primarily on inlet air density. Thus the performance of an engine of With a conventional bowl, the swirling air entering the bowl flows down to the given displacement can be increased by compressing the inlet air prior to entry to base of the bowl, then inward and upward in a toroidal motion. In reentrant the cylinder. Methods for achieving higher inlet air density in the gas exchange bowls the swirling air entering the bowl spreads downward and outward into the processes-mechanical supercharging, turbocharging, and pressure-wave undercut region, and then divides into a stream rising up the bowl sides and a supercharging-are discussed in Sec. 6.8. The arrangements of the various practi- stream flowing along the bowl base. Reentrant chambers generally produce cal supercharging and turbocharging configurations are shown in Fig. 6-37. higher swirl at the end of compression, and maintain a high swirl level further Figures 1-11, 6-40, 6-43, 6-49, 6-53, and 6-58 show examples of the different devices used to achieve higher inlet air densities. In this section the effects of boosting air density on engine performance are examined. Spark-ignition and compression-ignition engines are dealt with separately. Power boosting via super- charging and/or turbocharging is common in diesel engines: few spark-ignition engines are turbocharged. Knock prevents the full potential of boosting from being realized in the latter type of engine. A more extensive discussion of turbo- charged engine operation is provided by Watson and Janota.48 FIGURE 15-33 15.6.1 Four-Stroke Cycle SI Engines Flow pattern set up in diametral plane by squish-swirl interaction in (a) conventional The bmep of most production spark-ignition engines at wide-open throttle is and (b) reentrant bowl-in-piston combus- knock-limited over part of the engine speed range (see Sec. 15.4.4). The compres- (a) (b) tion chambers. È cylinder axis.47 sion ratio is usually set at a sufficiently high value so that some spark retard from 870 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 871 MBT timing is needed to avoid knock for the expected range of available fuel 1.9 1.9 RON = 100 octane rating and sensitivity (see Fig. 15-17). The propensity of the end-gas to &RON = 100 1.8 knock is increased by increases in end-gas temperature and pressure (see Sec. ¢ = 1.1 With charge air cooling 9.6.2). Hence attempts to boost the output of a given size spark-ignition engine 1.7 1.7 by an inlet air compression device that increases air pressure and temperature ¢ = 0.9 will aggravate the knock problem, since end-gas pressure and temperature will 1.6 .6 increase. However, the potential advantages of power boosting are significant. @ = 1.1 Charge pressure, atm Charge pressure, atm RON = 91 The higher output for a given displaced volume will decrease engine specific 1.5- weight and volume (Sec. 2.11). Also, if the power requirements in a specific appli- ¢ = 0.9 cation (such as an automobile) can be met with either a naturally aspirated SI 1.4- 1.4- engine of a certain size or with a smaller size engine which is turbocharged to the Without charge air cooling same maximum power, the smaller turbocharged engine should offer better fuel 1.3- Tc = 7 1.3 ¢ = 1.1 economy at part load. At a given part-load torque requirement, the mechanical efficiency of the smaller turbocharged engine is higher, and if the gross indicated 1.2 1.2. 20 60 80 100 120 140 6 7 efficiencies of the engines are the same, the smaller engine will show a brake Charge temperature, ºC Compression ratio efficiency benefit. In practice, it proves difficult to realize much of this potential (a) (b) efficiency gain for the reasons described below. FIGURE 15-34 While a naturally aspirated spark-ignition engine may have sufficient Dependence of SI engine knock limits on: (a) charge pressure, temperature, and equivalence ratio ¢, margin of safety relative to knock to allow modest inlet-air boost, any substantial with re = 7, 2500 rev/min, MBT timing, 91 and 100 research octane number fuel; (b) charge pressure air compression prior to cylinder entry will require changes in engine design and compression ratio, without and with (to 60ºC) charge air cooling, 2500 rev/min, MBT timing, and/or operating variables to offset the negative impact on knock. The variables ¢ = 1.1, 100 RON fuel.49 which are adjusted to control knock in turbocharged SI engines are: compression ratio, spark retard from optimum, charge air temperature, and fuel/air equiva- lence ratio.+ Figure 15-34 shows how the knock limits depend on charge pres- effective pressure achievable at a fixed compression ratio as a function of charge sure, temperature, fuel/air equivalence ratio and compression ratio for given pressure and ignition timing with and without charge-air cooling. Additional octane rating fuels. The difference in boost achievable with the premium and the retard allows higher boost pressures to be utilized; however, at a constant safety regular quality gasoline is significant, as expected (Sec. 9.6.3). Charge-air tem- margin from the knock limit, the resulting gains in bmep decrease as retard is perature has a strong influence on allowable boost levels: lowering the com- increased. To avoid an unnecessary fuel consumption penalty, retarded timing pressed air temperature prior to entry to the cylinder with a charge-air cooler should only be used when the turbocharger does develop a high boost pressure. allows a substantially higher compression ratio to be used at a given boost level, The above discussion illustrates why turbocharged spark-ignition engines with a corresponding impact on engine efficiency.+ The boost pressure benefits of normally have lower compression ratios than naturally aspirated engines, use the richer mixtures in Fig. 15-34a (¢ = 1.1 compared with 0.9) are largely due to substantial mixture enrichment (up to o ~ 1.3) at high boost to cool the charge, the cooling effect of the additional fuel on the air charge. For example, Fig. often use an intercooler to reduce the charge-air temperature, and operate with 15-34b shows that, with a rich mixture and charge cooling to 60ºC, a charge pressure of 1.5 atm can be utilized at optimum spark timing with a compression 1.6 ratio of 8. Without charge cooling, the same charge pressure can only be used with a compression ratio of 6.49 1.5 Pi = 1.6 atm With CAC In turbocharged SI engines, the knock limit is usually reached at spark 1.5 timings retarded from the MBT optimum. Figure 15-35 shows the brake mean 1.4 1.4 bmep, MPa 1.3- Knock limit 1.6 FIGURE 15-35 1.2 1.5 1.4 Brake mean effective pressure and knock limits for without CAC turbocharged SI engines as a function of spark t Valve timing changes are often made too. These are done primarily to improve low-speed torque 1.1 advance and inlet pressure p; (in atmospheres). 2500 where turbocharging has a limited impact. 12 20 30 38 rev/min, re = 7, ¢ = 1.1, 99 RON fuel, without and + The turbocharged engine in Fig. 1-10 has an intercooler to reduce the inlet charge temperature. Spark advance, deg BTC with (AT = 45ºC) charge-air cooling.49 872 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 873 retarded timing at high boost pressures. Since compression ratio reductions and 120| retarded ignition timings result in losses in efficiency, and unintended knock with 2.3-dm3 TC high boost pressures would be especially damaging, precise control of ignition 110- -.- 2.1-dm3 TC timing is critical. Most turbocharged SI engines now use a knock sensor and --- 2.3-dm3 NA 100 ignition-timing control system so that timing can be adjusted continuously to Power, kW avoid knock without unnecessary retard. The sensor is usually an accelerometer 90+- which senses above-normal vibration levels on the cylinder head at the character- 80- istic knock frequency. With a knock sensor, ignition timing can be automatically adjusted in response to changes in fuel octane rating and sensitivity, and ambient 70 conditions. Turbocharged SI engines where fuel is mixed with the air upstream or 250 downstream of the compressor, using carburetors or fuel-injection systems, have been developed and used. Most modern turbocharged engines use port fuel injec- 200 Torque, N . m tion. This provides easier electronic control of fuel flow, avoids filling most of the 150 pressurized manifold volume with fuel-air mixture, and improves the dynamic FIGURE 15-37 response of the system by reducing fuel transport delays. 100 Power and torque as a function of engine speed We now consider the performance of actual turbocharged spark-ignition 0 1 40 60 80 100 for two turbocharged and one naturally aspi- engines. Examples of compressor outlet or boost pressure schedules as a function rated four-cylinder spark-ignition engine. See Engine speed, rev/s Table 15.3.52 of speed at wide-open throttle for three turbocharged engines are shown in Fig. 15-36. The essential features of the curves are the same. Below about 1000 high engine speed; the details of this problem have already been discussed above. engine rev/min the turbocharger achieves negligible boost. Boost pressure then Even with the use of very rich mixtures and spark retard at WOT, lower com- rises with increasing speed to 1.4 to 1.8 atm (absolute pressure) at about 2000 pression ratios for turbocharged engines, and intercooling, knock avoidance rev/min. Boost pressure then remains essentially constant with increasing engine requires that boost pressures (which would continue to rise with increasing speed. The rising portion of the curve is largely governed by the relative size of engine speed in the absence of any control) be maintained approximately con- the turbine selected for a given engine. This is usually expressed in terms of the stant. This is normally achieved by reducing the exhaust flow through the turbine A/R ratio of the turbine-the ratio of the turbine's inlet casing or volute area A as speed increases by bypassing a substantial fraction of the exhaust around the to the radius of the centroid of that area. Lower A/R values (smaller-capacity turbine through the wastegate or flow control valve (see Sec. 6.8.4). A wastegate is turbines) give a more rapid boost pressure rise with increasing speed; however, a spring-loaded valve acting in response to the inlet manifold pressure on a con- they give higher boost pressures at high engine speed, which is undesirable.48, 50 trolling diaphragm. Although other methods of controlling boost can be used, 48 Avoidance of knock is the reason why boost must be limited at medium to the wastegate is the most common. About 30 to 40 percent of the exhaust bypasses the turbine at maximum engine speed and load. Figure 15-37 compares the performance of two turbocharged spark-ignition 60 (a) (C ) engines (four-cylinder, 2.1- and 2.3-dm3 displacement) with that of the base 2.3-dm3 engine in its naturally aspirated form. Table 15.3 gives details of these 50- ( b ) 40- TABLE 15.3 Turbocharged spark-ignition engine performance52 Boost pressure above atm, kPa 30 Type 2.1-dm3 TC 2.3-dm3 NA 2.3-dm3 TC/AC 20- FIGURE 15-36 Boost pressure schedules for three turbocharged Displacement, dm3 2.127 2.316 2.316 10- spark-ignition engines: (a) 3.8-dm3 V-6 engine, Bore x stroke, mm 92 × 80 96 × 80 96 × 80 86.4 mm stroke, re == 8;50 (b) 2.2-dm3 four-cylinder Compression ratio 7.5 9.5 8.7 engine, 92 mm stroke, re = 8.1;51 (c) 2.32-dm3 four- Maximum power, kW at rev/min 98 at 5400 83 at 5400 117 at 5300 O 1000 2000 3000 4000 5000 6000 cylinder engine, 80 mm stroke, re == 8.7.52 All sched- Maximum torque, N . m at rev/min 210 at 3800 184 at 2800 250 at 2900 Engine speed, rev/min ules are wastegate controlled. Maximum bmep, kPa 1241 998 1356 874 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 875 Turbocharged --- Naturally aspirated thermal loading of critical components can become limiting too. As boost pres- 100 sure is raised, unless engine design and operating conditions are changed, 80 260 maximum pressures and thermal loadings will increase almost in proportion. In 270 practice, the compression ratio is often reduced and the maximum fuel/air equiv- 60 -280 270 280- alence ratio must be reduced in turbocharged engines (relative to naturally aspi- -300 Percent maximum torque rated engines) to maintain peak pressures and thermal loadings at acceptable 40 .300 -400_. FIGURE 15-38 levels. The fuel flow rate increases at a much lower rate than the air flow rate as Comparison of bsfc contours (in grams per boost pressure is increased. Limitations on turbocharged engine performance are 20F-400- 400 bsfc, g/kW.h kilowatt-hour) on performance maps of turbo- discussed more fully by Watson and Janota.48 charged and naturally aspirated versions of the Small automotive indirect-injection (IDI) turbocharged engines are limited 20 30 40 50 60 70 80 90 100 same spark-ignition engine, scaled to the same by structural and thermal considerations to about 130 atm maximum swirl- or Percent maximum mean piston speed maximum torque and mean piston speed.53 pre-chamber pressure, 14 m/s maximum mean piston speed, and 860ºC maximum exhaust temperature.54 Smoke and NO, emission standards are addi- three engines. The 2.1-dm3 turbocharged but not intercooled engine (which also tional constraints. Figure 15-39 shows the full-load engine and turbocharger per- does not have a knock sensor to control spark advance) requires a lower com- formance characteristics of a six-cylinder 2.38-dm3 displacement Comet V pression ratio and achieves less of a bmep gain than the 2.3-dm3 turbocharged swirl-chamber automobile diesel engine. The maximum boost pressure is con- intercooled engine with its knock-sensor spark-advance control, which together trolled by a poppet-valve-type wastegate to 0.75 bar above atmospheric. The fuel permit use of a higher compression ratio. Turbocharging the naturally aspirated consumption map for this engine is shown in Fig. 15-40. Superimposed on the 2.3-dm3 engine, with the modifications indicated, results in a 36 percent increase turbocharged engine map is the map for the base naturally aspirated swirl- in maximum engine torque and a flatter torque-versus-speed profile. chamber IDI engine of the same geometry and compression ratio (rc = 23). The The brake specific fuel consumption contours of an engine produced in both turbocharged engine has a maximum torque 46 percent higher and a maximum naturally aspirated and turbocharged versions are shown in Fig. 15-38. The data power 33 percent higher than the naturally aspirated engine. The best bsfc values have been scaled to represent engines of different displaced volume but the same are closely comparable. maximum engine torque. The smaller-displacement low-compression-ratio turbo- The different methods of supercharging internal combustion engines were charged engine (re = 6.9) shows a reduction in bsfc at low speed and part load reviewed in Sec. 6.8. Turbocharging, mechanical supercharging with a Roots due to improved mechanical efficiency. At high speed and load the larger- blower, and pressure wave supercharging with the Comprex are alternative displacement naturally aspirated engine has an advantage in bsfc due to its methods of boosting the performance of a small automotive swirl-chamber IDI higher compression ratio (8.2), less enrichment, and more optimum timing. 53 In a vehicle context, the low-speed part-load advantage of the smaller size but equal 80 power turbocharged engine should result in an average fuel economy benefit rela- 60 tive to the larger naturally aspirated engine. This benefit has been estimated as a kPa 40 PENWAI function of load. At full load the average efficiencies should be comparable; at 20 Boost pressure o Bosch smoke number half load, the turbocharged engine should show a benefit of about 10 percent, the 100 benefit increasing as load is decreased.49 oc 80- 60 Compressor discharge temperature 350 Full load 40 15.6.2 Four-Stroke Cycle CI Engines 800 300- The factors that limit turbocharged diesel engine performance are completely ºC 600|- different to those that limit turbocharged spark-ignition engines. The output of Turbine inlet temperature 250 bsfc, g/kW.h naturally aspirated diesel engines is limited by the maximum tolerable smoke 400 20 30 40 50 60 70 80 20 30 40 50 70 80 emission levels, which occur at overall equivalence ratio values of about 0.7 to 60 Engine speed, rev/s Engine speed, rev/s 0.8. Turbocharged diesel engine output is usually constrained by stress levels in critical mechanical components. These maximum stress levels limit the maximum FIGURE 15-39 Engine and turbocharger characteristics of six-cylinder 2.38-dm3 swirl-chamber IDI automotive diesel cylinder pressure which can be tolerated under continuous operation, though the engine at full load.54 876 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 877 Mean piston speed, m/s diesel engine. Figure 15-41 compares the torque and bsfc values obtained with 1.2,2 6 10 12 14 each of these supercharging methods on a performance map for a 1.2-dm3 engine. Values for a 1.6-dm3 naturally aspirated IDI diesel engine are also shown. All three approaches achieve close to the desired maximum power of the 1.6-dm3 NA 1.01 260 engine (40 kW at 4800 rev/min): e.g ., 1.2-dm3 turbo, 41.2 kW at 4500 rev/min; 250, 255 270 290 1.2-dm3 Comprex with intercooler, 42.3 kW at 3500 rev/min; 1.2-dm3 Roots, 245 310 37.6 KW at 4000 rev/min. The Comprex system produces the highest torque at 0.8 325 - 240 low engine speeds, even under unsteady engine operating conditions. The density 250- of the charge air determines the amount of charge, and hence the torque. Charge- 25 245 air pressure and temperature for the three supercharging systems are shown in bmep, MPa 0.6- Fig. 15-42. The Comprex (here without an intercooler) must have the highest 270. charge pressure because it has the highest charge temperature. Intercooling 0.4 would be particularly effective in this case.55 310 Small high-speed high-swirl turbocharged direct-injection diesel engines 325, (e.g ., suitable for automobile or light-truck applications) have similar per- 0.2H formance maps to those of equivalent IDI engines (Figs. 15-39 and 15-40). bsfc contours in g/kW . h Maximum bmep values are closely comparable: usually slightly higher boost is 20 30 40 50 50 70 required to offset the lower volumetric efficiency of the high-swirl-generating port 80 and valve of the DI engine. Best bsfc values for the DI engine are usually about Engine speed, rev/s 15 percent lower than of comparable IDI engines (see Ref. 56). The operating characteristics of larger medium-swirl turbocharged DI FIGURE 15-40 Fuel consumption map (bsfc in grams per kilowatt-hour) for turbocharged (-) and naturally aspi- diesel engines are illustrated by the data shown in Fig. 15-43. The engine is a rated ( --- ) versions of 2.38-dm3 six-cylinder swirl-chamber IDI diesel engine.54 12-dm3 displacement six-cylinder heavy-duty truck engine. The combustion chamber is similar to that shown in Fig. 15-32c, with a square combustion cavity and relatively low levels of swirl. The swirl is generated by a helical port in one of the two intake ports and a tangential port in the other in the four-valve cylinder - 1.2-dm3 Baseline (1) --- 1.2-dm3 Turbo (4) head. Both the engine's operating map and the turbocharger compressor map -- 1.2-dm3 Roots (2) - 1.6-dm3 NA engine (5) with the boost pressure curve superposed are shown for two different compressor 1.2-dm3 Comprex, (3) impellors. The adoption of the backward-vaned rake-type impellor compared to -.-. with intercooler --- without intercooler a more conventional design significantly increases low- and medium-speed per- 140; 120 2.0 450 100 5 1.8 80 1.6 100 Torque, N . m 280 Pressure ratio 60 FIGURE 15-41 Charge air temperature, K g/kW.h Torque and brake specific fuel consump- 1.4 tion of naturally aspirated and super- 350 charged 1.2-dm3 swirl-chamber IDI diesel 1.2 Turbo - 20 400 engine. Baseline (1): naturally aspirated. ....... Comprex g/kw.h Supercharged with (2) Roots blower; (3) 1.0 Roots 300 FIGURE 15-42 0 Comprex (with and without intercooler); Charge pressure and temperature with the IDI 1000 2000 3000 4000 5000 (4) turbocharger. Larger displacement 0 0.01 0.02 0.03 0.04 0.05 diesel engine and different supercharging methods Engine speed, rev/min 1.6-dm3 naturally aspirated engine (5).55 Air flow rate, m3/s of Fig. 15-41.55 878 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 879 2800 Torque 3.8 N . m With backward-vaned impellor 2400 - With conventional impellor 3.4 2000 Conventional Backward vaned 500 Surge line. 3.0 TT 1100 kw 2.5} 90,000 300 2.6 1000 Power Pressure ratio 200 900- 2.2 --- - Engine full load 80,000 Smoke -- - - 0.76- 0.74- 800 210 number g/kw . h OHNWA - - 0.70- 2.0 Surge line 1.8 - - - 0.65 -- 700 F g/kW . h Bosch smoke 220 0.77 Pressure ratio 65 bmep, kPa 220 600 210 bsfc 1.4 =-- 230: 60,000₮ Efficiency 200 500} 190 1.01 240 == 1.5 11 12 13 14 15 16 17 18 19 20 21 x 102 0.1 0.2 0.3 0.4 0.5 0.6 0.7 400 Engine speed, rev/min Air flow, m3/s (a) 300- 270 (b) 40,000 At 20ºC FIGURE 15-44 200 80 1.0 40 60 100 5 10 15 20 25 Performance characteristics of medium-speed turbocharged aftercooled DI diesel engine. (a) Torque, Percent maximum engine speed Air flow rate, m3/min power, smoke number, and bsfc for V twelve-cylinder version. (b) Compressor characteristics and (a) (b) engine full-load line for V-8 cylinder version. Bore = 128 mm, stroke = 140 mm, r = 15.58 FIGURE 15-43 Performance characteristics of turbocharged 12-dm3 six-cylinder medium-swirl heavy-duty truck DI Examples of values of combustion-related parameters for this type of engine diesel engine, with two different compressor impellors: (a) fuel consumption maps; (b) compressor over the load range at its maximum rated speed are shown in Fig. 15-45 for a maps with full-load boost operating line for engine with backward-vaned impellor superposed. 14.6-dm3 six-cylinder turbocharged aftercooled DI diesel engine with a boost Bore = 135 mm, stroke = 140 mm, r = 16.57 pressure ratio of 2 at rated power. The ignition delay decreases to about 10º (0.9 ms at 1800 rev/min) as load is increased. The bmep at 100 percent rated load at this speed is 1.2 MPa. Exhaust temperature increases substantially with formance by improving the compressor efficiency over the engine's boost pressure increasing load: maximum cylinder pressure increases to about 10 MPa at the curve (Fig. 15-43b). A wastegate is then used to control the boost level at high rated load. In this particular study it was found that these operating parameters engine speeds. The improvement in low-speed engine torque is apparent in were relatively insensitive to fuel variations. The cross-hatched bands show data Fig. 15-43a. The dependence of the maximum torque curve on both engine and for an additional nine fuels of varying sulfur content, aromatic content, 10 and 90 turbocharger design details is clear. With boost pressure ratios limited to below percent distillation temperatures.59 2, in the absence of air-charge cooling, maximum bmep values of 1.1 MPa are Higher outputs can be obtained with two-stage turbocharged aftercooled typical of this size and type of diesel engine. diesel engines, the arrangement shown in Fig. 6-37d. The performance character- With structurally more rugged component designs, aftercooled turbo- istics of such a high bmep (1.74 MPa) six-cylinder engine of 14-dm3 displacement charged medium-speed diesel engines with swirl in this cylinder size range can are shown in Fig. 15-46. The high air flow requires an overall pressure ratio of 3 utilize higher boost and generate much higher bmep. Wastegate control of boost at sea level ambient conditions (rising to 4 at 3658 m altitude). This was obtained is no longer required. Figure 15-44 shows the performance characteristics of a at lower cost with two turbochargers in series than with a multistage single tur- V-8 cylinder engine with its compressor map and full-load boost characteristic. bocharger. At rated conditions, the maximum cylinder pressure is 12.7 MPa and This turbocharged intercooled engine achieves a maximum bmep of about the maximum mean piston speed is 10.6 m/s. 1.5 MPa and bsfc below 200 g/kW . h between the maximum torque speed and Additional gains in efficiency with these heavy-duty automotive diesel rated power. Boost pressure at full load increases continuously over the engine engines can be achieved with turbocompounding: some of the available energy in speed range. 58 the exhaust gases is captured in a turbine which is geared directly to the engine 880 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 881 800- drive shaft. The above discussion indicates that typical turbocharged DI diesel engines achieve bsfc levels of 210 to 220 g/kW ·h (brake fuel conversion effi- 600 assimsin == = > Exhaust temperature ciencies of 0.4 to 0.38). With the increased cylinder pressure capability, higher fuel-injection pressures, and lower-temperature aftercooling of the above higher oC 400- bmep engines, bsfc values of 200 g/kW . h (0.42 brake efficiency) or lower can be Cylinder pressure achieved. With turbocompounding, bsfc values can be reduced another 5 to 6 12000 - 200 - percent to about 180 g/kW . h, or a brake efficiency of 0.47, at rated power.61 The largest four-stroke cycle DI diesel engines are used for marine propul- 8000 20 sion. An example is the Sulzer 400 mm bore 480 mm stroke engine which pro- MPa duces 640 kW per cylinder at 580 rev/min (S) = 9.3 m/s). Very high bmep levels 4000 Ignition delay-10 Ignition delay, deg (2.19 MPa) are achieved at maximum continuous rated power through progress 10 in turbocharger design and engine improvements which allow higher maximum Exhaust opacity, % cylinder pressures. These, combined with optimization of gas exchange and com- 5 hoffmann Smoke bustion processes, achieve bsfc values of 185 to 190 g/kW . h (45 to 46 percent brake efficiency).62 350 - Many diesel system concepts are being examined which promise even higher output and/or efficiency. Variable-geometry turbocharger-turbine nozzles 250- improve utilization of exhaust gas available energy at low engine speeds. The g/kw . h hyperbar turbocharging system-essentially a combination of a diesel engine 150 - 0 400 1600 2000 2400 with a free-running gas turbine (a combustion chamber is placed between the 800 1200 engine and the turbocharger turbine)-has the potential of much higher bmep. bmep, kPa Diesel systems with thermally insulated combustion chambers which reduce heat FIGURE 15-45 losses and increase the available exhaust energy have the potential for improving Operating parameters of 14.6-dm3 six-cylinder turbocharged aftercooled DI diesel engine as a func- efficiency and for increasing power through additional exhaust energy recovery in tion of load at maximum rated speed of 1800 rev/min. Maximum rated power = 261 kW at devices such as compounded turbines and exhaust-heated Rankine cycle bmep = 1192 kPa. Points: standard diesel fuel. Shaded band: nine fuels of varying sulfur content, systems. 48 aromatic content, 10 and 90 per cent distillation temperatures.59 1800 15.6.3 Two-Stroke Cycle SI Engines bmep 1600 The two-stroke cycle spark-ignition engine in its standard form employs sealed bmep, kPa crankcase induction and compression of the fresh charge prior to charge transfer, 400 1400 Power with compression and spark ignition in the engine cylinder after charge transfer. 1200 The fresh mixture must be compressed to above exhaust system pressures, prior 300|- 1600 to entry to the cylinder, to achieve effective scavenging of the burned gases. Two- Air flow Power, KW stroke cycle scavenging processes were discussed in Sec. 6.6. The two-stroke 400 Air flow, dm3/s 200 spark-ignition engine is an especially simple and light engine concept and finds 200 its greatest use as a portable power source or on motorcycles where these advan- 100 Air/fuel tages are important. Its inherent weakness is that the fresh fuel-air mixture which 30 20 short-circuits the cylinder directly to the exhaust system during the scavenging 260 10 FIGURE 15-46 process constitutes a significant fuel consumption penalty, and results in excessive 0 240 Operating characteristics of 14-dm3 six- unburned hydrocarbon emissions. bsfc, g/kW . h cylinder two-stage turbocharged after- 220 bsfc This section briefly discusses the performance characteristics of small crank- cooled quiescent-chamber DI diesel 200 engine. Maximum bmep = 1.74 MPa. case compression two-stroke cycle SI engines. The performance characteristics 800 1000 1200 1400 1600 1800 2000 2200 Boost pressure ratio at rated power = 3. (power and torque) of these engines depend on the extent to which the displaced Engine speed, rev/min Bore = 140 mm, stroke = 152 mm.60 volume is filled with fresh mixture, i.e ., the charging efficiency [Eq. (6.24)]. The 882 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 883 fuel consumption will depend on both the trapping efficiency [Eq. (6.21)] and the 20 Powe charging efficiency. Figure 15-47a shows how the trapping efficiency nt, varies KW with increasing delivery ratio A at several engine speeds for a two-cylinder 16- 347-cm3 displacement motorcycle crankcase compression engine. The delivery ratio increases from about 0.1 at idle conditions to 0.7 to 0.8 at wide-open throt- Torque 50 12 40 130 N . m 1.0 20 Speed, rev/min 8 Perfect mixing 10 2000 0.9 3000 4000 700 5000 600 0.2 bsfc 500 g/KW . h 6000 FIGURE 15-48 0.8 0.3 A 400 Performance characteristics of a 300 three-cylinder 450-cm3 two-stroke 0.7 7ch = 0.1 E4- 1000 3000 5000 7000 cycle spark-ignition engine. Maxi- Trapping efficiency ntr mum bmep = 640 kPa. Bore = 58 mm, Engine speed, rev/min stroke = 56 mm.64 0.6 tle. Lines of constant charging efficiency nch [which equals Ant; see Eq. (6.25)] 0.5 are shown. Figure 15-47b shows bmep plotted against these charging efficiency values and the linear dependence on fresh charge mass retained is clear. 0.4 Performance curves for a three-cylinder 450-cm3 two-stroke cycle minicar 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 engine are shown in Fig. 15-48. Maximum bmep is 640 kPa at about 4000 rev/ Delivery ratio A min. Smaller motorcycle engines can achieve slightly higher maximum bmep at (a) higher speeds (7000 rev/min). Fuel consumption at the maximum bmep point is about 400 g/kW . h. Average fuel consumption is usually one-and-a-half to two 500 times that of an equivalent four-stroke cycle engine. CO emissions from two-stroke cycle engines vary primarily with the fuel/air 400 equivalence ratio in a manner similar to that of four-stroke cycle engines (see Fig. 11-20). NO, emissions are significantly lower than from four-stroke engines 300 due to the high residual gas fraction resulting from the low charging efficiency. Unburned hydrocarbon emissions from carbureted two-stroke engines are about 200} five times as high as those of equivalent four-stroke engines due to fresh mixture bmep, kP short-circuiting the cylinder during scavenging. Exhaust mass hydrocarbon emis- 100 sions vary approximately as A(1 - nu)o, where o is the fuel/air equivalence ratio.63 15.6.4 Two-Stroke Cycle CI Engines -100. 0.1 0.2 0.3 0.4 0.5 0.6 Large marine diesel engines (0.4 to 1 m bore) utilize the two-stroke cycle. These Charging efficiency 1Ich low-speed engines with relatively few cylinders are well suited to marine propul- (b) sion since they are able to match the power/speed requirements of ships with simple direct-drive arrangements. These engines are turbocharged to achieve high FIGURE 15-47 (@) Trapping and charging efficiencies as a function of the delivery ratio. (b) Dependence of brake brake mean effective pressures and specific output. The largest of these engines mean effective pressure on fresh-charge mass defined by charging efficiency. Two-cylinder 347-cm3 can achieve brake fuel conversion efficiencies of up to 54 percent. An example of displacement two-stroke cycle spark-ignition engine.63 a large marine two-stroke engine is shown in Fig. 1-24. Over the past 25 years 884 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 885 the output per cylinder of such engines has increased by a factor of more than 145 two, and fuel consumption has decreased by 25 percent. These changes have been Rate of pressure rise 12/ 306 kPa/deg Peak pressure achieved by increasing the maximum firing pressure to 13 MPa, and by refining 13.3 MPa critical engine processes such as fuel injection, combustion, supercharging, and scavenging. The uniflow-scavenging process is now preferred to loop scavenging 10 Combustion begins 13º BTC since it achieves higher scavenging efficiency at high stroke/bore ratios and allows increases in the expansion stroke.62 8 Cylinder pressure, MPa The performance characteristics of a 580 mm bore Sulzer two-stroke marine diesel engine with a stroke/bore ratio of 2.9 are shown in Fig. 15-49. The solid lines show the standard turbocharged engine characteristics. The rated Injection begins 15.8º BTC. Injection ends speed for the engine is 125 rev/min, corresponding to a maximum mean piston 4 0.8º ATC speed of 7.2 m/s. The rated bmep is 1.66 MPa. The minimum bsfc is 175 g/kW . h which equals a brake fuel conversion efficiency of 48 percent. For larger N lower-speed engines, the efficiency is higher. The dashed lines show how the per- -50 -40 -30 -20 -10 0 10 20 30 40 50 formance of this engine can be improved by turbocompounding. A proportion of Crank angle, deg the engine's exhaust flow, at loads higher than 50 percent, is diverted from the turbocharger inlet to a separate turbine coupled to the engine power takeoff gear via an epicyclic speed-reduction gear and hydraulic coupling. The additional power recovered in this manner from the engine exhaust flow improves bsfc by -2600 5 g/kW . h. At part load, when the full exhaust flow passes through the turbo- charger, an efficiency gain is also obtained, due to the higher scavenging pressure -2200 (and therefore increased cylinder pressure) obtained with the full exhaust flow. 0.25 0.24 -1800 FIGURE 15-50 1400 Brake power, kw 0.23 Injection, combustion, and per- bsfc, g/kW . h formance characteristics of inter- 0.22 1000 mediate-size turbocharged two- 1301 0.21 stroke cycle uniflow-scavenged DI 120 -600 diesel engine. Bore = 230.2 mm, 110 0.20 stroke = 279.4 mm and r = 16. Maximum cylinder pressure, atm 100 - 3 Shallow dish-in-piston combustion 90 200 chamber with swirl. At maximum 80 - 2 Air pressure, atm 70- 340 420 rated power 500 at 900 rev/min, 580 660 740 820 900 --- bmep = 0.92-1.12 MPa depending Engine speed, rev/min on application.65 - 0 10 190 Both two-stroke and four-stroke cycle diesel engines of intermediate size FIGURE 15-49 bsac, kg/kW . h Performance characteristics of large (200 to 400 mm bore) are used in rail, industrial, marine, and oil drilling applica- - 8 185 marine t two-stroke cycle uniflow- tions. The performance characteristics of a turbocharged two-stroke cycle scavenged DI diesel engine. Bore == uniflow-scavenged DI diesel engine (similar to the engine in Fig. 1-5), with bsfc, g/kW . h 180- 580 mm, stroke/bore = 2.9, maximum 230.2 mm bore, 279.4 mm stroke, and a compression ratio of 16, are shown in rated speed = 125 rev/min (mean piston Fig. 15-50. Combustion in the shallow dish-in-piston chamber with swirl occurs speed = 7.2 m/s), bmep (at rated power) 175- = 1.66 MPa. Solid line: standard turbo- smoothly yielding a relatively low rate of pressure rise. The pressure curve shown charged configuration. Dashed lines: with peak pressure of 13.3 MPa is for full-load operation. The bmep at rated 170+ parallel turbocompounded configuration power at 900 rev/min is 0.92 to 1.12 MPa depending on application. The 30 40 50 60 70 80 90 100 at greater than 50 percent load. bsac: maximum mean piston speed is 8.4 m/s. The bsfc of 200 g/kW . h corresponds to Percent maximum power brake specific air consumption.62 Mf, b = 0.42. Brake power, kw 886 1401 120 100 40 60 80| 20 1200 FIGURE 15-51 stroke = 114.3 mm, r = 18.66 1400 1600 Maximum rated power the maximum boost pressure ratio is 2.6. INTERNAL COMBUSTION ENGINE FUNDAMENTALS 1800 228 225 231 243 237 256 268 15.7 ENGINE PERFORMANCE SUMMARY Engine speed, rev/min 2000 292 g/kW . h 2200 uniflow-scavenged two-stroke cycle DI diesel engine. Engine turbocharged at mid and high loads; Roots blown at low loads. Maximum boost pressure ratio = 2.6. Bore = 98.4 mm, Brake power and specific fuel consumption (grams per kilowatt-hour) map of four-cylinder 3.48-dm3 four-stroke cycle engines in the marine, industrial, and construction markets. The fuel consumption map of such a four-cylinder 3.48-dm3 displacement uniflow- a Roots blower to provide the required scavenging air pressure for starting and cient boost and the blower is not needed; the blower is unloaded (air flow is scavenged two-stroke cycle diesel engine is shown in Fig. 15-51. The engine uses light-load operation. At moderate and high loads the turbocharger supplies suffi- maximum bmep of 951 kPa at 1500 rev/min. The best bsfc is 225 g/kW . h and bypassed around the blower) under these conditions. The engine generates 138 kW at its rated speed of 2500 rev/min (mean piston speed of 9.5 m/s) and a ignition engines described in previous sections of this chapter are summarized The major performance characteristics of the spark-ignition and compression- mean piston speed S, at maximum rated power, and the minimum value of bsfc here to highlight the overall trends. Table 15.4 lists the major design features of these engines, the bmep at maximum engine torque, bmep and the value of the 2400 Smaller turbocharged two-stroke cycle DI diesel engines also compete with 2600 2800 TABLE 15.4 Performance of representative engines in different categories Volume Maximum torque Rated maximum power Maximum efficiency per Number Engine Bore , Stroke, Stroke/ cylinder, of bmep, Speed, bmep Speed, Boost pres S. befc type mm mm bore re dm3 cylinders kPa rev/min kPa rev/min sure ratio m/s g/kW .h Reference SI/4S/NA 96.8 86 0.88 8.6 0.632 910 2500 750 4300 12.3 SI/4S/NA 84.5 88 1.04 8.5 0.494 966 2800 767 5200 15.3 67 SI/4S/NA 86* 86* 1* 8.5 0.5 910 3500 758 5000 14.3 274 0.30 13 SI/4S/NA 96 80 0.83 · 9.5 0.579 998 2800 796 5400 14.4 52 SI/4S/TC 92 80 0.87 7.5 0.532 1241 3800 1024 5400 1.6* 14.4 52 SI/4S/TCAC 96 80 0.83 8.7 0.579 4 1356 290 1144 5300 1.6 14.1 52 SI/2S/C 58 0.97 0.144 m 654 3500 575 4500 8.4 ~400* ~0.2 64 SI/25/C 64 0.84 0.174 686 7000 590 8200 14.8 ~ 340* ~0.24 63 IDI/4S/NA 76.5 86.4 1.13 23 0.397 850 3100 670 4800 13.8 280 0.30 35 IDI/4S/NA 84 82 0.98 22 0.454 675 2000 502 5000 13.7 n+ + 0 IDI/4S/NA 102 100 0.98 19 0.817 848 2200 743 3500 11.7 251 0.34 46 IDI/4S/TC 76.5 86.4 1.13 0.397 1080 240 840 4800 1.7 13.8 240 0.35 DI/4S/NA 76.5 80 1.05 18.5 0.368 735 2800 600 5000 13.3 246 0.34 34 DI/4S/NAA 102 100 0.98 18 0.817 784 2000 682 3200 10.7 220 0.39 33 100 0.98 0.817 886 2200 782 3500 11.7 221 0.38 46 DI/4S/NA 102 m DI/4S/NA 115 135 1.17 16 1.40 851 1400 777 2700 12.2 204 0.42 DI/4S/NA 135 140 1.04 2.00 862 1400 763 2500 11.7 DI/4S/TC 115 135 1.17 1.40 1098 1500 941 2500 11.2 203 0.42 DI/4S/TCAC 115 135 1.17 1.40 1344 1600 1240 2300 10.4 DI/4S/TCAC 128 140 1.09 15 1.8 6-16 1560 1500 1280 2100 2.5 9.8 195 0.43 58 57 DI/4S/TC 135 140 1.04 16 2.00 1087 1300 911 2300 10.7 210 0.40 DI/4S/2TCAC 140 152 1.09 2.33 1740 1400 1445 2100 3 10.6 207 0.41 60 DI/4S/TCAC 400 480 1.20 60.3 6-18 2190 580 9.3 185 0.46 62 DI/2S/TC 98.4 114.3 1.16 18 0.870 3, 4, 6 1065 1500 952 2500 2.6 226 0.37 66 DI/2S/TC 230 279.4 1.21 16 11.6 8-20 920-1122 900 2.8 8.4 200 0.42 65 380-840 1100-2900 2.9-3.4 125-1607 4-12 1660 196-90 35 DI/2S/TCAC 7.2 180-160 0.47-0.53 62 t Engine type: SI = spark-ignition; IDI = indirect-injection compression-ignition; DI = direct-injection compression-ignition; 4S = four-stroke; 2S = two-stroke; NA = naturally aspirated; NAA = NA and air-cooled; 887 C = crankcase compression of scavenging mixture; TC = turbocharged; TCAC == turbocharged and aftercooled; 2TC = two-stage turbocharged. * Denotes estimated value. 888 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 889 and the corresponding brake fuel conversion efficiency. It should be stressed that (a) Rank the chambers 1, 2, 3 in the order of their volumetric efficiency (1 = highest there are many different engine configurations and uses, and that for each of no ) . these there are variations in design and operating characteristics. However, these (b) Rank the chambers in order (1, 2, 3) of their flame frontal area (1 = highest) when the mass fraction burned is about 0.2 and the piston is at TC. representative values of performance parameters illustrate the following trends: (c) Given this relative flame front area ranking, discuss whether the ranking by mass burning rate dm,/dt will be different from the flame area ranking. 1. Within a given category of engines (e.g ., naturally aspirated four-stroke SI (d) Briefly discuss the knock implications of these three chamber designs. Which is engines) the values of maximum bmep, and bmep and S ,, at maximum rated likely to have the worst knock problem? power, are closely comparable. Within an engine category where the range in size is substantial, there is an increase in maximum bmep and a decrease in Spark plug minimum bsfc as size increases due to the decreasing relative importance of E friction and heat loss per cycle. There is also a decrease in S, at maximum 30 22 power as engine size increases. Note the higher bmep of naturally aspirated SI E E mm mm -43-+ 36 -43- engines compared to equivalent NA diesels due to the fuel-rich operation of mm mm mm mm -30. 22. the former at wide-open throttle. 16 mm mm mm 2. Two-stroke cycle spark-ignition engines have significantly lower bmep and 13 mm higher bsfc than four-stroke cycle SI engines. 100 mm- 100 mm 100 mm- 3. The effect of increasing inlet air density by increasing inlet air pressure A. 2-valve B. 2-valve C. 4-valve Side plug Plug 16 mm from axis increases maximum bmep values substantially. Turbocharging with after- Center plug Normal port Helical port Normal ports FIGURE P15-1 cooling gives increased bmep gains relative to turbocharging without after- cooling at the same pressure level. The maximum bmep of turbocharged SI 15.2. Figures 15-23 and 15-10 show the variation in brake specific fuel consumption engines is knock-limited. The maximum bmep of turbocharged compression- (bsfc) for a swirl-chamber IDI automobile diesel (D) and a conventional automobile ignition engines is stress-limited. The larger CI engines are designed to accept spark-ignition (SI) engine as a function of load and speed, respectively. From these higher maximum cylinder pressures, and hence higher boost. graphs determine, and then plot, brake fuel conversion efficiency: (1) as a function 4. The best efficiency values of modern automobile SI engines and IDI diesel of speed at full load and (2) as a function of load at a mid-speed of 2500 rev/min. Both engines are naturally aspirated. Assume the engine details are: engines are comparable. However, the diesel has a significant advantage at lower loads due to its low pumping work and leaner air/fuel ratio. Small DI diesels have comparable (or slightly lower) maximum bmep to equivalent IDI Compression Equivalence Displacement, diesels. The best bsfc values for DI diesels are 10 to 15 percent better, however. ratio ratio range dm3 5. In the DI diesel category (which is used over the largest size range-less than Diesel 22 0.3-0.8 2.3 100 mm bore to almost 1 m), maximum bmep and best brake fuel conversion SI engine 1.0-1.2 1.6 efficiency steadily improve with increasing engine size due to reduced impact of friction and heat loss per cycle, higher allowable maximum cylinder pres- (a) List the major engine design and operating variables that determine brake fuel sure so higher boost can be used, and (additionally in the larger engines) conversion efficiency. through turbocompounding. (b) Explain briefly the reasons for the shapes of the curves you have plotted and the relative relationship of the D and SI curves. (c) At 2500 rev/min, estimate which engine will give the higher maximum brake PROBLEMS power. 15.1. The schematics show three different four-stroke cycle spark-ignition engine com- 15.3. The diesel system shown in the figure consists of a multicylinder reciprocating bustion chambers. A and B are two-valve engines, C is a four-valve engine (two diesel engine, a turbocharger (with a compressor C and turbine Tr mechanically inlet valves which open simultaneously, two exhaust valves). Dimensions in milli- connected to each other), an intercooler (I), and a power turbine (Tp) which is meters are indicated. A and C have normal inlet ports and do not generate any geared to the engine drive shaft. The gas and fuel flow paths and the gas states at swirl, B has a helical inlet port and generates substantial swirl. Spark plug locations the numbered points are shown. You can assume that the specific heat at constant are indicated. All three engines operate at the same speed (3000 rev/min), with the pressure c, of the gas throughout the entire system is 1.2 KJ/kg . K and y = cp/Cv = same inlet mixture composition, temperature, and pressure, and have the same dis- 1.333. The engine operates at 1900 rev/min. The fuel has a lower heating value of placed volume. 42 MJ/kg of fuel. 890 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 891 (a) What is the power (in kilowatts) which the turbocharger turbine (Tr) must The combustion efficiency: ne [Eq. (3.27)] produce? What is the gas temperature at exit to the turbocharger turbine? The indicated fuel conversion efficiency: n. : [Eq. (2.23)] (b) What is the power turbine power output? The indicated thermal conversion efficiency : n ., ; [Eq. (3.31)] (c) The heat losses in the engine are 15 percent of the fuel's chemical energy (a) Derive a relation between the variables ny, i, nc , and nt, i. (m/ @LHv). Find the engine power output, the total system power output, and (b) Derive an equation which relates the brake power P, to no, "Im, ne, "t, , and any the total system brake fuel conversion efficiency (friction effects in the engine other engine and fuel parameters required. and power turbine are internal to these devices and do not need to be explicitly (c) Explain briefly why the variations of no, nm, ne, n, i, n ., : with equivalence ratio evaluated). in the figure have the form shown (e.g ., why the parameter is approximately constant, or has a maximum/minimum, or decreases/increases with increasing Air richness or leanness, etc.). 1 atm 0.53 kg/s Turbocharger 300 K 15.5. The diagram shows the layout of a low heat loss turbocharged turbocompounded diesel engine. The engine and exhaust system is insulated with ceramics to reduce C heat losses to a minimum. Air flows steadily at 0.4 kg/s and atmospheric conditions into the compressor C, and exits at 445 K and 3 atm. The air is cooled to 350 K in 2 . 2.5 atm the intercooler I. The specific heat of air, c ,, is 1 kJ/kg . K. In the reciprocating 430 K 75, 1.5 atm Power diesel engine, the fuel flow rate is 0.016 kg/s, the fuel heating value is 42.5 MJ/kg, 5 turbine and the heat lost through the ceramic walls is 60 kW. 4 850 K 4 atm The exhaust gases leave the reciprocating engine at 1000 K and 3 atm, and 2.5 atm enter the first turbine T , which is mechanically linked to the compressor. The 3 1 320 K ip pressure between the two turbines is 1.5 atm. The second turbine T is mechani- cally coupled to the engine drive shaft and exhausts to the atmosphere at 800 K. Fuel Engine 6 675 K The specific heat of exhaust gases, c ,, is 1.1 KJ/kg . K. 0.018 kg/s 1 atm (a) Analyze the reciprocating diesel engine E and determine the indicated power Exhaust Geared to drive shaft FIGURE P15-3 obtained from this component of the total system. If the engine mechanical efficiency is 0.9 what is the brake power obtained from component E? 15.4. The attached graph shows how the brake power and specific fuel consumption of a (b) Determine the power obtained from the power turbine TB. four-stroke cycle single-cylinder spark-ignition engine vary with the fuel/air equiva- (c) Determine the total brake power obtained from the complete engine system lence ratio at wide-open throttle. It also shows how the following efficiencies vary and the fuel conversion efficiency of the system. You can neglect mechanical with equivalence ratio: losses in the coupling between the power turbine and the engine drive shaft. The volumetric efficiency : n. The mechanical efficiency: n. [Eq. (2.17)] Air 0.4 kg/s 1 - 1 atm 7 bsfc 700 300 K 6 600 PD, KW bsfc, g/kW . h 5 500 4 Lot C 400 TA 100 3 atm 2 : 445 K 1.5 atm 80 5 I moto of 1000 K 3 atm 3 atm TB 40 nt, i 350 K 3 Efficiency, % Reciprocating 6 + 1 atm engine ¥ 800 K 20 E 0.8 1.0 1.2 1.4 1.6 Equivalence ratio FIGURE P15-4 Fuel, 0.016 kg/s FIGURE P15-5 892 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 893 15.6. New automobile spark-ignition engines employ "fast-burn technology" to achieve engines to that of the gasoline-fueled engine, allowing for energy density effects an improvement in fuel consumption and reductions in hydrocarbon (HC) and at intake (at 1 atm and 350 K), at the knock-limited compression ratio for each oxides of nitrogen (NO,) emissions. This question asks you to explain the experi- fuel, for stoichiometric mixtures. You can assume that the fuel-air cycle results mental data which shows that faster-burning combustion chambers do provide for isooctane apply also for methanol and hydrogen cycles to a good approx- these benefits relative to more moderate burn-rate chambers. imation, when the energy density is the same. (a) Figure 9-36b shows the effect of increasing the percent of the exhaust gas recy- (c) The lean operating limit for the three fuels is different as indicated. Estimate the cled to the intake (for NO, control) in a moderate burn-rate engine at constant ratio of indicated fuel conversion efficiency for methanol and hydrogen at their speed and load, stoichiometric air/fuel ratio, with timing adjusted for maximum lean limit and knock-limited compression ratio, relative to gasoline at its lean brake torque at each condition. COVimep is the standard deviation in imep limit and knock-limited compression ratio, at the same inlet pressure (0.5 atm). divided by the average imep, in percent. The different types of combustion are: Under these conditions, rank the fuel-engine combinations in order of decreas- misfire, partial burn, slow burn, normal burn, defined in Sec. 9.4.3. Frequency is ing power output. percent of cycles in each of these categories. Use your knowledge of the spark- ignition engine flame-propagation process and HC emission mechanism to explain these trends in COVimep , HC, and frequency as EGR is increased. Gasoline (b) The fast-burn combustion chamber uses two spark plugs and generates swirl (isooctane) Methanol Hydrogen inside the chamber by placing a vane in the inlet port to direct the air to enter C,H1. CH3OH H, the chamber tangentially. The swirl angular velocity in the cylinder at the end of intake is six times the crankshaft angular velocity. There is no swirl in the Stoichiometric F/A 0.066 0.155 0.0292 moderate burn-rate chamber which has a single spark plug and a relatively Lower heating 44.4 20.0 120.1 quiescent in-cylinder flow. The table shows spark timing, average time of peak value, MJ/kg 114 pressure, average flame-development angle (0 to 10 percent mass burned) and Molecular weight of fuel 32 Molecular weight of 30.3 29.4 21 rapid burning period (10 to 90 percent mass burned) for these two engines. stoichiometric mixture Figures 11-29 and 15-9 show how the operating and emission characteristics of Research octane number 95 106 ~ 90 the fast burn and moderate burn-rate engines change as percent EGR is Knock-limited 9 12 increased. Explain the reasons for the differences in these trends in COVimeo, compression ratio bsfc (brake specific fuel consumption), and HC, and similarity in NO ,. The Equivalence ratio 0.9 0.8 0.6 operating conditions are held constant at the same values as before. at lean misfire limit Fast Moderate 15.8. Small-size direct-injection (DI) diesel engines are being developed as potential burn burn replacements for indirect-injection (IDI) or prechamber engines in automobile Spark timing 18 applications. Figures 10-16 and 10-2 show the essential features of these two types 40º BTC Crank angle for 15º 16º ATC of diesel. The DI engine employs high air swirl, which is set up with a helical swirl-generating inlet port (Fig. 8-13). The injector is centrally located over the average Pmax 0-10% burned 24º 350 bowl-in-piston combustion chamber and the injector nozzle has four holes, one in 10-90% burned 200 500 each quadrant. The IDI engine (a Ricardo Comet swirl chamber), in contrast, has no swirl in the main chamber, but generates high velocities and a rotating flow in the prechamber during compression. 15.7. Two alternative fuels, methanol and hydrogen, are being studied as potential future Figures 15-21 and 15-23 show performance maps for typical versions of spark-ignition engine fuels which might replace gasoline (modeled by isooctane these two types of engines. Bmep, brake mean effective pressure, is plotted against C8 H18). The table gives some of the relevant properties of these fuels. engine speed. Brake specific fuel consumption contours are shown with the (a) For each fuel calculate the energy content per unit volume (in joules per cubic numbers in grams per kilowatt-hour. meter) of a stoichiometric mixture of fuel vapor and air at 1 atm and 350 K. The heat-release-rate profiles for these two types of engine at a typical mid- The universal gas constant is 8314 J/kmol . K. What implications can you draw load mid-speed point are shown versus crank angle in the sketch. O has units of from these numbers regarding the maximum power output of an engine of fixed joules per second. geometry operating with these fuels with stoichiometric mixtures? (a) Explain the reasons for the differences in shape and relative timing in the cycle (b) The octane rating of each fuel, and hence the knock-limited compression ratio of the heat-release-rate profiles. of an engine optimized for each fuel, is different. Estimate the ratio of the (b) Suggest reasons for the differences (magnitude and shape) in the maximum maximum indicated mean effective pressure for methanol- and hydrogen-fueled bmep versus mean-piston-speed line for the DI and IDI engines. 894 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 895 (c) Evaluate the brake fuel conversion efficiency of each engine at its maximum (b) Estimate the ratio of the maximum indicated power of the engine operating efficiency point, and at 2000 rev/min and road load (road load is the power with natural gas to the maximum power of the gasoline engine. requirement to maintain a vehicle at constant speed; it is 2 bar bmep at 2000 (c) Estimate the ratio of the gross indicated fuel conversion efficiency of the natural rev/min). Explain the origin of the observed differences in efficiency at these two gas engine to that of the gasoline engine, at the part-load conditions given. operating conditions. (d) Explain whether the NO, CO, and hydrocarbon specific emissions (grams of pollutant per hour, per unit indicated power) at part-load conditions of the natural gas engine will be higher, about the same, or lower than the NO, CO, and HC emissions from the gasoline engine. Explain briefly why. DI engine You can assume that the fuel-air cycle results derived for isooctane-air mix- tures are also appropriate for methane-air mixtures. Heat release rate Q IDI engine 15.11. Spark-ignition and prechamber diesel engines are both used as engines for pas- senger cars. They must meet the same exhaust emission requirements. Of great importance are their emission characteristics when optimized for maximum power at wide-open throttle (WOT) and when optimized at cruise conditions for TC Crank angle FIGURE P15-8 maximum efficiency. 15.9. A four-stroke cycle naturally aspirated direct-injection diesel is being developed to (a) Give typical values for the equivalence ratio for a passenger car spark-ignition provide 200 KW of power at the engine's maximum rated speed. Using information engine and a prechamber diesel optimized for maximum power at WOT and available in Chaps. 2, 5, and 15, on typical values of critical engine operating 2000 rev/min, and optimized for maximum efficiency at part load parameters at maximum power and speed for good engine designs, estimate the (bmep = 300 kPa) and 1500 rev/min. Briefly explain the values you have chosen. following: (a) The compression ratio, the number of cylinders, the cylinder bore and stroke, (b) Construct a table indicating whether at these two operating conditions the spe- and the maximum rated speed of an appropriate engine design that would cific emissions of CO, HC, NO ,, and particulates are low (L), medium (M), or provide this maximum power. high (H) relative to the other load point and to the other engine. Explain your (b) The brake specific fuel consumption of this engine design at the maximum reasoning for each table entry. power operating point. 15.12. For a naturally aspirated four-stroke cycle diesel engine: (c) The approximate increase in brake power that would result if the engine was (a) Show from the definition of mean effective pressure that turbocharged. bmep oc nm ns. in.( F/A) 15.10. Natural gas (which is close to 100 percent methane, CHA) is being considered as a spark-ignition engine fuel. The properties of methane and gasoline (assume the where bmep = brake mean effective pressure same properties as isooctane) and the engine details for each fuel are summarized "= = mechanical efficiency below (¢ is the fuel/air equivalence ratio). ns, ¿ = indicated fuel conversion efficiency no = volumetric efficiency F /A = fuel/air ratio Natural gas Gasoline (b) Sketch carefully proportioned qualitative graphs of nm, ny, i, n ,, and Composition CH4 C8H18 (F/A)/(F/A)stoich versus speed N at full load, and explain the reasons for the Heating value, MJ/kg 50.0 44.3 shapes of the curves. Then explain why the maximum bmep versus speed curve Research octane number 120 94 has the shape shown in Fig. P15-12. Compression ratio 14 Displaced volume, dm3 2 Lean misfire limit ¢ = 0.5 ¢ = 0.8 Full load line Part-load equivalence ratio @ = 0.6 ¢ = 0.9 21 Full-load equivalence ratio ¢ = 1.1 ¢ = 1.2 As indicated in the table, the displaced volume of the engine is unchanged when bmep the conversion for natural gas is made; however, the clearance height is reduced to increase the compression ratio. (a) Estimate the ratio of the volumetric efficiency of the engine operating on natural gas to the volumetric efficiency with gasoline, at wide-open throttle and 2000 rev/min. Both fuels are in the gaseous state in the intake manifold. Speed - - FIGURE P15-12 896 INTERNAL COMBUSTION ENGINE FUNDAMENTALS ENGINE OPERATING CHARACTERISTICS 897 (c) The minimum brake specific fuel consumption point is indicated by the asterisk 22. Kuroda, H ., Nakajima, Y ., Sugihara, K ., Takagi, Y ., and Maranaka, S.: "Fast Burn with Heavy (*) in Fig. P15-12 (see Figs. 15-21 and 15-22). Explain why brake specific fuel EGR Improves Fuel Economy and Reduces NO, Emission," JSAE Rev ., no. 5, pp. 63- 69, 1980. consumption increases with (1) increasing speed, (2) increasing bmep, (3) 23. Thring, R. H.: "The Effects of Varying Combustion Rate in Spark Ignited Engines," SAE paper 790387, 1979. decreasing bmep. 24. Harada, M ., Kadota, T ., and Sugiyama, Y.: "Nissan NAPS-Z Engine Realizes Better Fuel Economy and Low NO, Emission," SAE paper 810010, 1981. 25. Poulos, S. G ., and Heywood, J. B.: "The Effect of Chamber Geometry on Spark-Ignition Engine REFERENCES Combustion," SAE paper 830334, SAE Trans ., vol. 92, 1983. 1. Armstrong, D. L ., and Stirrat, G. F.: "Ford's 1982 3.8L V6 Engine," SAE paper 820112, 1982. 26. Heywood, J. B.: "Combustion Chamber Design for Optimum Spark-Ignition Engine Per- 2. "Engine Rating Code-Spark-Ignition," SAE Standard J245, in SAE Handbook. formance," Int. J. Vehicle Des ., vol. 5, no. 3, pp. 336-357, 1984. 3. Okino, M ., Okada, K ., and Abe, M.: "Isuzu New 8.4L Diesel Engine," SAE paper 850258, 1985. 27. Novak, J. M ., and Blumberg, P. N.: " Parametric Simulation of Significant Design and Operating 4. Higashisono, M ., Takeuchi, K ., and Hara, H.: “The New Isuzu 1.8 Liter 4-Cylinder Diesel Engine Alternatives Affecting the Fuel Economy and Emissions of Spark-Ignited Engines," SAE paper for the United States Market," SAE paper 820116, SAE Trans ., vol. 91, 1982. 780943, SAE Trans ., vol. 87, 1978. 5. General Motors Automotive Engine Test Code For Four Cycle Spark Ignition Engines, 6th ed ., 28. Amann, C. A.: "Control of the Homogeneous-Charge Passenger-Car Engine: Defining the Problem," SAE paper 801440, 1980. 1975. 6. Heywood, J. B ., Higgins, J. M ., Watts, P. A ., and Tabaczynski, R. J.: " Development and Use of a 29. Bell, A. G.: "The Relationship between Octane Quality and Octane Requirement, " SAE paper Cycle Simulation to Predict SI Engine Efficiency and NO, Emissions," SAE paper 790291, 1979. 750935, 1975. 7. Heywood, J. B ., and Watts, P. A.: " Parametric Studies of Fuel Consumption and NO Emissions 30. Leppard, W. R.: "Individual-Cylinder Knock Occurrence and Intensity in Multicylinder of Dilute Spark-Ignition Engine Operation Using a Cycle Simulation," paper C98/79, in Pro- Engines," SAE paper 820074, 1982. ceedings of Conference on Fuel Economy and Emissions of Lean Burn Engines, Institution of 31. Ingamells, J. C ., Stone, R. K ., Gerber, N. H ., and Unzelman, G. H.: "Effects of Atmospheric Mechanical Engineers, London, 1979. Variables on Passenger Car Octane Number Requirements," SAE paper 660544, SAE Trans ., vol. 8. Quader, A. A.: "The Axially-Stratified-Charge Engine," SAE paper 820131, SAE Trans ., vol. 91, 75, 1966. 32. Gruden, D.: "Performance, Exhaust Emissions and Fuel Consumption of an IC Engine Oper- 1982. 9. Robison, J. A ., and Brehob, W. M.: "The Influence of Improved Mixture Quality on Engine ating with Lean Mixtures," paper C111/79, in Proceedings of Conference on Fuel Economy and Exhaust Emissions and Performance," J. Air Pollution Control Ass ., vol. 17, no. 7, pp. 446-453, Emissions of Lean Burn Engines, Institution of Mechanical Engineers, London, 1979. 33. Siezak, P. J ., and Vossmeyer, W.: "New Deutz High Performance Diesel Engine," SAE paper July 1967. 10. Thring, R. H ., and Overington, M. T.: “Gasoline Engine Combustion-The High Ratio Compact 810905, 1981. Chamber," SAE paper 820166, SAE Trans ., vol. 91, 1982. 34. Neitz, A ., and D'Alfonso, N.: "The M.A.N. Combustion System with Controlled Direct Injection 11. Hamburg, D. R ., and Hyland, J. E.: " A Vaporized Gasoline Metering System for Internal Com- for Passenger Car Diesel Engines," SAE paper 810479, 1981. 35. Sator, K ., Buttgereit, W ., and Sturzebecher, U.: "New 5- and 6-Cylinder VW Diesel Engines for bustion Engines," SAE paper 760288, 1976. 12. Nakajima, Y ., Sugihara, K ., and Takagi, Y.: "Lean Mixture or EGR-Which is Better for Fuel Passenger Cars and Light Duty Trucks," SAE paper 790206, 1979. Economy and NO, Reduction?," paper C94/79, in Proceedings of Conference on Fuel Economy 36. Monaghan, M. L.: “The High Speed Direct Injection Diesel for Passenger Cars," SAE paper and Emissions of Lean Burn Engines, Institution of Mechanical Engineers, London, 1979. 810477, 1981. 13. Wade, W ., and Jones, C.: "Current and Future Light Duty Diesel Engines and Their Fuels," SAE 37. Pischinger, R ., and Cartellieri, W.: “Combustion System Parameters and Their Effect upon Diesel Engine Exhaust Emissions," SAE paper 720756, SAE Trans ., vol. 81, 1972. paper 840105, SAE Trans ., vol. 93, 1984. 14. Lavoie, G. A ., and Blumberg, P. N.: “ A Fundamental Model for Predicting Fuel Consumption, 38. Ball, W. F ., and Hil, R. W.: "Control of a Light Duty Indirect Injection Diesel Engine for Best NO, and HC Emissions of a Conventional Spark-Ignited Engine," Combust. Sci. Technol ., vol. 21, Trade-Off between Economy and Emissions," paper C122/82, in Proceedings of Conference on Diesel Engines for Passenger Cars and Light Duty Vehicles, Publication 1982-8, Institution of pp. 225-258, 1980. 15. Caton, J. A ., Heywood, J. B ., and Mendillo, J. V.: "Hydrocarbon Oxidation in a Spark-Ignition Mechanical Engineers, London, 1982. Engine Exhaust Port," Combust. Sci. Technol ., vol. 37, nos. 3 and 4, pp. 153-169, 1984. 39. Wade, W. R ., Idzikowski, T ., Kukkonen, C. A ., and Reams, L. A.: "Direct Injection Diesel Capa- 16. Caton, J. A ., and Heywood, J. B.: "Models for Heat Transfer, Mixing and Hydrocarbon Oxida- bilities for Passenger Cars," SAE paper 850552, 1985. tion in an Exhaust Port of a Spark-Ignited Engine," SAE paper 800290, 1980. 40. Greeves, G ., and Wang, C. H. T.: "Origins of Diesel Particulate Mass Emission," SAE paper 17. Caris, D. F ., and Nelson, E. E ., "A New Look at High Compression Engines," SAE Trans ., vol. 810260, SAE Trans ., vol. 90, 1981. 41. Greeves, G ., Khan, I. M ., and Wang, C. H. T.: "Origins of Hydrocarbon Emissions from Diesel 67, pp. 112-124, 1959. 18. Kerley, R. V ., and Thurston, K. W.: "The Indicated Performance of Otto-Cycle Engines," SAE Engines," SAE paper 770259, SAE Trans ., vol. 86, 1977. 42. Greeves, G.: "Response of Diesel Combustion Systems to Increase of Fuel Injection Rate," SAE Trans ., vol. 70, pp. 5-30, 1962. 19. Muranaka, S ., Takagi, Y ., and Ishida, T.: "Factors Limiting the Improvement in Thermal Effi- paper 790037, SAE Trans ., vol. 88, 1979. ciency of S.I. Engine at Higher Compression Ratio," SAE paper 870548, 1987. 43. Yu, R. C ., and Shahed, S. M.: "Effects of Injection Timing and Exhaust Gas Recirculation on 20. Gruden, D. O.: "Combustion Chamber Layout for Modern Otto Engines," SAE paper 811231, Emissions from a D.I. Diesel Engine," SAE paper 811234, SAE Trans ., vol. 90, 1981. 44. Khan, I. M ., Greeves, G ., and Wang, C. H. T.: " Factors Affecting Smoke and Gaseous Emissions 1981. 21. Barnes-Moss, H. W.: " A Designers Viewpoint," paper C343/73, in Proceedings of Conference on from Direct Injection Engines and a Method of Calculation," SAE paper 730169, 1973. Passenger Car Engines, pp. 133-147, Institution of Mechanical Engineers, Conference publication 45. Bassoli, C ., Cornetti, G. M ., and Cuniberti, F.: "IVECO Diesel Engine Family for Medium Duty Vehicles," SAE paper 820031, 1982. 19, London, 1973. 898 INTERNAL COMBUSTION ENGINE FUNDAMENTALS 46. Kawamura, H ., Kihara, R ., and Kinbara, M.: "Isuzu's New 3.27L Small Direct Injection Diesel," SAE paper 820032, 1982. 47. Arcoumanis, C ., Bicen, A. F ., and Whitelaw, J. H.: "Squish and Swirl-Squish Interaction in APPENDIX Motored Model Engines," ASME Trans ., J. Fluids Engng, vol. 105, pp. 105-112, 1983. 48. Watson, N ., and Janota, M. S.: Turbocharging the Internal Combustion Engine, Wiley-Interscience Publications, John Wiley, New York, 1982. 49. Hiereth, H ., and Withalm, G.: "Some Special Features of the Turbocharged Gasoline Engine," A SAE paper 790207, 1979. 50. Wallace, T. F.: "Buick's Turbocharged V-6 Powertrain for 1978," SAE paper 780413, SAE Trans ., vol. 87, 1978. 51. Allen, F. E ., and Rinschler, G. L.: "Turbocharging the Chrysler 2.2 Liter Engine," SAE paper 840252, SAE Trans ., vol. 93, 1984. UNIT 52. Andersson, J ., and Bengtsson, A.: "The Turbocharged and Intercooled 2.3 Liter Engine for the CONVERSION Volvo 760," SAE paper 840253, SAE Trans ., vol. 93, 1984. 53. Watson, N.: “Turbochargers for the 1980s-Current Trends and Future Prospects," SAE paper FACTORS 790063, SAE Trans, vol. 88, 1979. 54. Grandinson, A ., and Hedin, I.: "A Turbocharged Engine for a Growing Market," paper C119/82, in Diesel Engines for Passenger Cars and Light Duty Vehicles, Institution of Mechanical Engi- neers, Conference publication 1982-8, London, 1982. 55. Walzer, P ., and Rottenkolber, P.: "Supercharging of Passenger Car Diesels," paper C117/82, in Diesel Engines for Passenger Cars and Light Duty Vehicles, Institution of Mechanical Engineers, Conference publication 1982-8, London, 1982. 56. Carstens, U. G ., Isik, T ., Biaggini, G ., and Cornetti, G.: "Sofim Small High-Speed Diesel Engines-D.I. Versus I.D.I ., " SAE paper 810481, 1981. 57. Okada, K ., and Takatsuki, T.: "Isuzu's New 12.0L Turbocharged Diesel with Wastegate Boost Control for Fuel Economy," SAE paper 820029, 1982. 58. Schittler, M.: "MWM TBD 234 Compact High-Output Engines for Installation in Heavy Equipment and Military Vehicles," SAE paper 850257, 1985. 59. Barry, E. G ., McCabe, L. J ., Gerke, D. H ., and Perez, J. M.: "Heavy-Duty Diesel Engine/Fuels Combustion Performance and Emissions-A Cooperative Research Program," SAE paper 852078, 1985. This table provides conversion factors for common units of measure for physical 60. Robinson, R. H ., and Schnapp, J. P.: "Cummins NTC-475 Series Turbocharged Engine," SAE quantities to the International System (SI) units. The conversion factors are pre- paper 820982, 1982. sented in two ways: columns 2 and 3 give the conversion to the base or derived 61. Wilson, D. E.: "The Design of a Low Specific Fuel Consumption Turbocompound Engine," SAE SI unit with the conversion factor as a number between one and ten with six or paper 860072, 1986. fewer decimal places, followed by the power of ten that the number must be 62. Lustgarten, G. A.: "The Latest Sulzer Marine Diesel Engine Technology," SAE paper 851219, 1985. multiplied by to obtain the correct value; columns 4 and 5 provide conversion to 63. Tsuchiya, K ., and Hirano, S.: "Characteristics of 2-Stroke Motorcycle Exhaust HC Emission and a recommended multiple or submultiple of the SI unit with the conversion factor Effects of Air-Fuel Ratio and Ignition Timing," SAE paper 750908, 1975. given as a four-digit number between 0.1 and 1000. 64. Uchiyama, H ., Chiku, T ., and Sayo, S.: " Emission Control of Two-Stroke Automobile Engine," SAE paper 770766, SAE Trans ., vol. 86, 1977. 1 3 65. Kotlin, J. J ., Dunteman, N. R ., Chen, J ., and Heilenbach, J. W.: "The GM/EMD Model 710 G To convert from TO 4 5 Multiply by To Series Turbocharged Two-Stroke Cycle Engine," ASME paper 85-DGP-24, 1985. Multiply by 66. Fellberg, M ., Huber, J. W ., and Duerr, J. W.: "The Development of Detroit Diesel Allison's New Area Generation Series 53 Engines," SAE paper 850259, 1985. foot m2 9.290 304 x 10-2 67. Hisatomi, T ., and lida, H.: "Nissan Motor Company's New 2.0 Liter Four-Cylinder Gasoline inch2 cm 929.0 m 6.451 600 x 10-4 cm2 Engine," SAE paper 820113, SAE Trans ., vol. 91, 1982. 6.452 Energy, heat, and work Btu (International Table) 1.055 056 x 103 1.055 calorie (thermochemical) 4.184000 x 10º KJ 4.184 erg 1.000 000 × 10-7 0.1000 foot pound-force (ft . Ibf) --- - --- 1.355 818 x 10º J 1.356 horsepower-hour (hp . h) 2.684 520 x 106 MJ 2.685 kilowatt-hour (kW . h) 3.600 000 × 106 MJ 3.600 metre kilogram-force (m . kgf) 9.806 650 x 10º J 9.807 900 INTERNAL COMBUSTION ENGINE FUNDAMENTALS APPENDIX A UNIT CONVERSION FACTORS 901 1 2 3 5 1 2 3 4 5 To convert from To Multiply by To Multiply by To convert from To Multiply by To Multiply by Energy (specific, specific heat) Pressure, stress (force per unit area) Btu (IT)/lb J/kg 2.326 000 × 103 KJ/kg 2.326 atmosphere (normal, 760 torr) Pa 1.013 250 x 105 kPa 101.3 Btu (IT)/1b . ºF J/kg . K 4.186 800 x 103 KJ/kg . K 4.187 inch of mercury (60ºF) Pa 3.37685 x 103 kPa 3.377 calorie (thermo.)/g J/kg 4.184 000 x 103 KJ/kg 4.184 kilogram-force/centimeter2 Pa 9.806 650 × 104 kPa 98.07 calorie (thermo.)/g . ºC J/kg . K 4.184 000 x 103 KJ/kg . K 4.184 mm of mercury, 0ºC (torr) Pa 1.333 224 x 102 Pa 133.3 Force pound-force/foot2 Pa 4.788 026 x 101 Pa 47.88 dyne 1.000 000 × 10-5 10.00 pound-force/inch2 (psi) Pa N 6.894 757 x 103 UN kPa 6.895 kilogram-force N 9.806 650 x 10º N 9.807 Temperature interval pound-force N 4.448 222 x 10º N 4.448 degree Celsius Force per unit length (includes surface tension) degree Fahrenheit K 1.000000 × 10º 5.555 556 x 10-1 K 0.5556 dyne/centimeter N/m 1.000 000 × 10-3 mN/m 1.000 Temperature pound-force/inch N/m 1.751 268 x 102 N/m 175.1 temperature (ºC) ºC + 273.15 pound-force/foot N/m 1.459 390 x 101 N/m 14.59 temperature (ºF) (ºF + 459.67)/1.80 (CF - 32)/1.80 Fuel consumption (economy) Torque pound/horsepower-hour kg/J 1.689 660 x 10-7 g/kW . h 608.3 kilogram-force meter N . m 9.806 650 x 10º N .m 9.807 gram/kilowatt-hour kg/J 2.777 778 x 10-10 ug/J 0.2778 pound-force foot N . m 1.355 818 x 10º N . m 1.356 mile/gallon (U.S.) m/m3 4.251 437 × 105 km/dm3 0.4251 Velocity mile/gallon (Imp.) m/m 3.540060 × 105 km/dm3 0.3540 foot/second m/s 3.048 000 x 10-1 m/s 0.3048 Heat flux (includes thermal conductivity) kilometer/hour m/s 2.777 778 × 10-1 m/s 0.2778 Btu (IT) . in/h . ft2 . ºF W/m . K 1.442 279 x 10-1 W/m . K 0.1442 mile/hour m/s 4.470 400 x 10-1 km/h 1.609 Btu (IT)/ft2 J/m2 1.135 653 x 104 KJ/m2 11.36 Btu (IT)/h . ft2 . ºF Viscosity W/m2 . K 5.678 263 x 10º W/m2 . K 5.678 centipoise calorie (thermo.)/cm2 J/m2 4.184 000 x 104 KJ/m2 Pa . s 1.000000 × 10-3 41.84 mPa . s 1.000 centistoke m2/s 1.000 000 x 10-6 mm2/s 1.000 Length poise Pa . s 1.000 000 × 10-1 Pa . s 0.1000 foot 3.048 000 × 10-1 m 0.3048 stoke m2/s 1.000 000 × 10-4 mm2/s 100.0 inch m 2.540 000 x 10-2 mm 25.40 Volume micron m 1.000 000 x 10-6 um 1.000 barrel (42 U.S. gallon) mile 1.609 344 x 103 ms 1.589 873 x 10-1 m km 1.609 0.1590 foot3 m3 2.831 685 x 10-2 dm3 28.32 Mass gallon (Imp.) m3 4.546 092 x 10-3 im3 4.546 ounce kg 2.834 952 x 10-2 28.35 gallon (U.S.) By 3.785 412 x 10-3 dm3 3.785 pound kg 4.535 924 x 10-1 kg 0.4536 inch3 m3 1.638 706 × 10-5 16.39 ton (long or Imp ., 2240 lb) kg 1.016047 x 103 Mg 1.016 liter m3 1.000 000 × 10-3 dm3 1.000 ton (short, 2000 lb) kg 9.071 847 x 102 Mg 0.9072 Volume per unit time tonne (metric) kg 1.000 000 x 103 Mg 1.000 foot3/minute (cfm) m /s 4.719 474 × 10 -4 dm3/s 0.4719 Mass per unit time (flow) foot3/second m3/s 2.831 685 x 10-2 dm3/s 28.32 pound/second kg/s 4.535924 x 10-1 kg/s 0.4536 gallon (U.S.)/minute (gpm) m3/s 6.309 020 × 10-5 cm 3/s 63.09 pound/minute kg/s 7.559 873 × 10-3 &/s 7.560 pound/hour kg/s 1.259 979 x 10 ·4 g/s 0.1260 Notes: Mass per unit volume 1. Derived units such as that for torque (newton-metre, N . m) are written with a period between each component unit for clarity. In practice, the period is often omitted. gram/gallon (U.S.) kg/m3 2.641 724 × 10-1 g/dm3 0.2642 2. Derived from Mobil Technical Bulletin SI Units, The Modern Metric System. Copyright Mobil Oil Corporation, pound/foot3 kg/m3 1.601 846 × 101 kg/m3 16.02 1974. Sections reproduced courtesy Mobil Oil Corporation. pound/inch3 kg/m3 2.767 990 x 104 kg/dm3 27.68 pound/gallon (Imp.) kg/m3 9.977 644 × 101 kg/dm3 0.0998 pound/gallon (U.S.) kg/m3 1.198 264 x 102 kg/dm3 0.1198 Power, heat flow Btu (IT)/hour W 2.930711 × 10-1 W 0.2931 horsepower (550 ft . lbf/s) W 7.456 999 x 102 kw 0.7457 horsepower (metric, CV, PS) W 7.35499 × 102 KW 0.7355 APPENDIX B IDEAL GAS RELATIONSHIPS 903 B.2 THE MOLE APPENDIX It is convenient to introduce a mass unit based on the molecular structure of B matter, the mole: The mole is the amount of substance which contains as many molecules as there are carbon atoms in 12 grams of carbon-12.+ Thus, the number of moles n of gas is given by IDEAL GAS n = = M (B.4) RELATIONSHIPS and Eq. (B.3) becomes pV = nRT (B.5) Values for the universal gas constant in different units are given in Table B.1. In the SI system, the value is 8314.3 J/kmol . K. TABLE B.1 Values of universal gas constant R 8314.3 J/kmol . K 8.3143 J/mol . K 1.9859 Btu/lb-mole . ºR 1543.3 ft . 1bf/lb-mole . ºR B.1 IDEAL GAS LAW The gas species which make up the working fluids in internal combustion engines (e.g ., oxygen, nitrogen, carbon dioxide, etc.) can usually be treated as ideal gases. This Appendix reviews the relationships between the thermodynamic properties B.3 THERMODYNAMIC PROPERTIES of ideal gases. The pressure p, specific volume v, and absolute temperature T of an ideal It follows from Eq. (B.1) that the internal energy ut of an ideal gas is a function of temperature only: gas are related by the ideal gas law pu = RT (B.1) u = u(T) (B.6) For each gas species, R is a constant (the gas constant). It is different for each gas Since the enthalpy h is given by u + pu, it follows also that and is given by h = h(T) R (B.7) R = (B.2) M where R is the universal gas constant (for all ideal gases) and M is the molecular weight of the gas. Since u is given by V/m, where V is the volume of a mass of gas t This is the SI system definition of the mole; it was formerly called the gram-mole. The kilogram- m, Eq. (B.1) can be rewritten as mole (kmol) is also used; it is 1000 times as large as the mole. + The symbol u will be used for internal energy per unit mass, u for internal energy per mole, and U PV = mRT - MRT (B.3) for internal energy of a previously defined system of mass m. Similar notation will be used for en- M thalpy, entropy, and specific heats, per unit mass and per mole. 902 904. INTERNAL COMBUSTION ENGINE FUNDAMENTALS APPENDIX B IDEAL GAS RELATIONSHIPS 905 The specific heats at constant volume and constant pressure of an ideal gas, B.4 MIXTURES OF IDEAL GASES c, and c ,, respectively, are defined by The working fluids in engines are mixtures of gases. The composition of a Ou mixture of ideal gases can be expressed in terms of the following properties of AT (B.8) each component: oh dh Cp (B.9) Partial pressure p;. The pressure each component would exert if it alone AT occupied the volume of the mixture at the temperature of the mixture. Parts by volume V/V. The fraction of the total mixture volume each com- From Eq. (B.1) it follows that ponent would occupy if separated from the mixture, at the mixture temperature Cp - C, = R (B.10) and pressure. Mass fraction x;. The mass of each component m;, divided by the total The ratio of specific heats, y, is a useful quantity: mass of mixture m. Mole fraction 5;. The number of moles of each component n;, divided by y = P (B.11) the total number of moles of mixture n. An additional restrictive assumption is often made that the specific heats are con- From Eq. (B.5) it follows that stants. This is not a necessary part of the ideal gas relationships. Pi M In general, the internal energy and enthalpy of an ideal gas at a temperature == Xi (B.16) p T relative to its internal energy and enthalpy at some reference temperature To are given by The thermodynamic properties of mixtures of ideal gases can be computed from the following relationships: u = uo + | c. (T ) aT (B.12) Molecular weight and h = hot | cp (T ) dT (B.13) JTO M = - [ n. M, = [ x, Mi (B.17) The entropy at T, v, and p, relative to the entropy at some reference state To, Do, Internal energy, enthalpy, and entropy Po, can be obtained from the relationships On a mass basis: du dp u = [ x,us h = [xchi s = Exist (B.18a, b, c) ds = 2 IT + R -= Ce IT - R (B.14) T C P On a mole basis: which integrate to give u = [xqui (B.19a, b, c) T S = SO @ dT + R in (B.15a) JTO T and s = SO - dT - R in P (B.15b) ITO T Po The properties u, h, and s can be evaluated on a per unit mass or per mole basis. On a mass basis, co, Co, and R would have the units J/kg . K (Btu/lbm . ºR); on a mole basis u, h, and s are replaced by u, h, and 3. R is then the universal gas constant R, c, and c, are replaced by c, and cp, and co, c ,, and R would have the units J/kmol . K (Btu/lb-mol . ºR). APPENDIX C EQUATIONS FOR FLUID FLOW THROUGH A RESTRICTION 907 Orifice APPENDIX 2 C FIGURE C-1 Schematic of liquid flow through orifice. EQUATIONS FOR C.1 LIQUID FLOW FLUID FLOW Consider the flow of a liquid through an orifice as shown in Fig. C-1. For the THROUGH A ideal flow, Bernoulli's equation can be written RESTRICTION P1 + P 2 1 = P2 + P 2 For an incompressible flow, continuity gives V141 = 12 42 and the ideal mass flow rate through an orifice is given by imideal = A2 2p(P1 - P2) 71/2 [1 - (A2/ A1)2 (C.1) The real mass flow rate is obtained by introducing the discharge coefficient: A. 2P(P1 - P2) 71/2 meal = CDA2| 1 - (A2/A1)2] (C.2) The discharge coefficient is a function of orifice dimensions, shape and surface roughness, mass flow rate, and fluid properties (density, surface tension, and viscosity). The use of the orifice Reynolds number In many parts of the engine cycle, fluid flows through a restriction or reduction in flow area. Real flows of this nature are usually related to an equivalent ideal flow. Re = DV2 D2 V2 D2 The equivalent ideal flow is the steady adiabatic reversible (frictionless) flow of an ideal fluid through a duct of identical geometry and dimensions. For a real fluid flow, the departures from the ideal assumptions listed above are taken into as a correlating parameter for the discharge coefficient accounts for the effects of account by introducing a flow coefficient or discharge coefficient CD, where m, p, v, and D2 to a good approximation.1 actual mass flow CD = C.2 GAS FLOW ideal mass flow Consider the flow of an ideal gas with constant specific heats through the duct Alternatively, the flow or discharge coefficient can be defined in terms of an shown in Fig. C-2. For the ideal flow, the stagnation temperature and pressure, effective cross-sectional area of the duct and a reference area. The reference area To and po, are related to the conditions at other locations in the duct by the AR is usually taken as the minimum cross-sectional area. The effective area of the steady flow energy equation flow restriction Ar is then the cross-sectional area of the throat of a frictionless nozzle which would pass the measured mass flow between a large upstream V 2 reservoir at the upstream stagnation pressure and a large downstream reservoir To = T + 2 cp **** at the downstream measured static pressure. Thus and the isentropic relation AE (7-1/7 CD - AR 906 908 INTERNAL COMBUSTION ENGINE FUNDAMENTALS APPENDIX C EQUATIONS FOR FLUID FLOW THROUGH A RESTRICTION 909 This ratio is called the critical pressure ratio. For (PT/Po) less than or equal to the critical pressure ratio, imideal VVRTo _ 2(+1)/2(7-1) ATPo (x + 1 ) (C.7) The critical pressure ratio is 0.528 for y = 1.4 and 0.546 for y = 1.3. Throat For a real gas flow, the discharge coefficient is introduced. Then, for sub- Po ideal critical flow, the real mass flow rate is given in terms of conditions at the P minimum area or throat by Po real p imreal = - CDAT PO VRTO (Pr ) (C.8) For a choked flow, inreal =- CD AT PO ,1/2( 2 ) (+ 1)/2(Y- 1) FIGURE C-2 (x + 1) (C.9) Pressure distribution for gas flow through a nozzle. VRTO Equation (C.8) can be rearranged in the form of Eq. (C.2) (with A2 < A1) as imreal = CDAR[2Po(Po - PT)]1/20 (C.10) By introducing the Mach number M = V/a, where a is the sound speed where @ is given by (= yRT), the following equations are obtained: Q = { [x (G - ! ][(PT/Po )2/7 - ( PT/Po) (" + 1 /2] ] 1/2 7 = 1 + /-1 M 2 2 (C.3) 1 - PT/Po (C.11) Figure C-3 shows the variation of " and (m/m*)ideal with (po - pr)/po . m* is the Do = ( 1 + " 1 M 2 ) ( Y - 1 ) (C.4) mass flow rate through the restriction under choked flow conditions (when the P Mach number at the throat is unity). For flow rates less than about 60 percent of The mass flow rate m is the choked flow, the effects of compressibility on the mass flow rate are less than 5 percent. m = pAV With the ideal gas law and the above relations for p and T, this can be rear- 1.00 ranged as mideal VYRTe - VM( 1 +2=1 M2) ("+1)/27-1) 0.80 (C.5) Apo 0.60 m or imideal VORTe - ( R ) { 2 [1 -( 2 ) 23/5 |8 212 (C.6) 0.40 Apo Choked flow For given values of po and To, the maximum mass flow occurs when the 0.20H- Y = 1.4 velocity at the minimum area or throat equals the velocity of sound. This condi- tion is called choked or critical flow. When the flow is choked the pressure at the OS 0.10 0.20 0.30 0.40 0.472 FIGURE C-3 throat, pr, is related to the stagnation pressure po as follows: Po - PT Po Relative mass flow rate m/m* and compressible PT =( 2) /(y -1 ) 1.0 0.9 0.8 0.7 0.6 0.528 flow function @ [Eq. (C.11)] as function of nozzle PT or restriction pressure ratio for ideal gas with Po Po y = 1.4. (From Taylor.2) 910 INTERNAL COMBUSTION ENGINE FUNDAMENTALS Flow coefficients are determined experimentally and are a function of the shape of the passage, the Reynolds number and Mach number of the flow, and APPENDIX the gas properties. For a Mach number at the throat less than about 0.7 and for passages of similar shape, the flow coefficient is essentially a function of Reynolds number only. Orifice plates are frequently used to measure gas flow rates. Standard D methods for determining flows through orifice plates can be found in Ref. 3. REFERENCES DATA 1. Lichtarowicz, A ., Duggins, R. K ., and Markland, E.: "Discharge Coefficients for Incompressible ON Non-Cavitating Flow through Long Orifices," J. Mech. Eng. Sci ., vol. 7, no. 2, pp. 210-219, 1965. 2. Taylor, C. F.: The Internal Combustion Engine in Theory and Practice, vol. I, p. 506, MIT Press, WORKING 1966. FLUIDS 3. Marks' Standard Handbook for Mechanical Engineers, 8th ed ., McGraw-Hill, 1978. 011 912 INTERNAL COMBUSTION ENGINE FUNDAMENTALS APPENDIX D DATA ON WORKING FLUIDS 913 TABLE D.1 TABLE D.2 Thermodynamic properties of air at low densityt Standard enthalpy of formation and molecular weight of species Cp Molecular K weight KJ/kg KJ /kg KJ/(kg . K) P. KJ/(kg . K) Y Species Formula g/mole Statet MJ/kmol kcal/mol 250 109.9 338.1 4.4505 7.660 38.81 1849.0 1.003 0.715 .. 401 Oxygen 32.00 275 O2 0 135.0 356.0 4.5187 7.7559 54.14 gas 1458.0 1.003 0.71 .401 Nitrogen N2 28.01 gas 0 300 460. 374. 4.5811 7.8432 73.39 1173.0 .004 0.717 1.400 Carbon C 12.011 solid 0 325 485.2 391.9 4.6385 7.9236 97.13 960.6 .. 006 0.718 1.400 Carbon co 28.01 gas -110.5 -26.42 350 510.4 409.9 4.6919 7.9982 125.9 797.8 .007 0.720 1.399 monoxide 375 535.6 427.9 4.7416 8.0678 160.5 670.8 1.010 0.723 1.397 Carbon CO2 44.01 gas -393.5 dioxide -94.05 400 560.8 146.0 4.7884 8.1330 201.4 570.0 1.013 0.725 1.396 Hydrogen 425 4.8324 H, 2.016 586.2 464. 8.1945 249. 488.9 1.016 0.729 .394 gas 0 0 Water 450 611.6 305.6 H2O 4.8742 18.02 482.5 8.2527 422.7 1.020 0.733 1.392 gas -241.8 Water -57.80 475 637.2 4.9139 H2O 500.8 8.3079 18.02 370. 368. 1.024 0.737 1.390 liquid -285.8 -68.32 Methane CH. 16.04 gas -- 74.9 -17.89 500 662.8 519.3 4.9518 8.3606 445.0 322.6 1.028 0.741 1.387 Propane C3Hg 44.10 gas - 103.8 525 688.6 -24.82 537.9 4.9881 8.4109 530.2 284.3 1.033 0.746 .. 385 Isooctane C3H18 114.23 gas -224.1 550 -53.57 714.5 556.6 5.0229 8.4590 627. 251. 1.03 0.752 .382 Isooctane C3H18 114.23 liquid -259.28 -61.97 575 740.5 575.5 5.0565 8.5053 736.8 224.0 1.044 0.757 1.379 Cetane C16H34 226.44 liquid -454.5 108.6 600 766.7 594. 5.0888 8.5499 860.6 1.050 0.763 Methyl 200.1 .376 CH3OH 32.04 gas -201.2 -48.08 625 793.0 613.6 5.1201 8.5929 alcohol 999.5 179.5 1.056 0.768 1.374 650 319.5 632. 5.1503 8.6344 1155.0 161.5 1.061 Methyl 0.77 1.37 CH3OH 32.04 liquid -238.6 -57.02 675 846.1 alcohol 552.3 5.1796 8.6745 1329.0 145. 1.067 0.780 1.368 Ethyl C2H,OH 46.07 700 872.9 671.9 5.2081 8.7135 1521.0 132.1 gas -234.6 1.073 0.786 1.365 alcohol -56.08 725 899.8 691.7 5.2358 8.7512 1735.0 119.9 1.079 0.792 1.362 Ethyl C2H,OH 46.07 750 926.8 711.5 liquid 5.2628 8.7879 -277.0 1972.0 109.2 1.085 0.798 1.360 alcohol -66.20 775 954. 731.6 5.2891 8.8236 2233.0 99.63 1.09 0.804 1.357 800 981.4 751.7 5.3147 8.8584 2520.0 91.12 1.09 0.81 1.354 + At 298.15 K (25ºC) and 1 atm. 825 1008.9 772.1 5.3397 8.8922 2836.0 83.52 1.103 0.81€ 1.352 850 1036.5 792.5 5.3641 8.9252 3181.0 76.7 1.10 0.82 1.350 875 1064.3 813. 5.3880 8.9574 3559.0 70.58 1.114 0.827 1.347 900 1092.2 833.8 5.4114 8.9889 3971.0 65.07 1.119 0.83 1.345 925 1120.2 854.7 5.4342 9.0196 4419.0 60.08 1.124 0.837 1.343 950 1148.4 875. 5.4566 9.0496 4907.0 55.58 1.129 ).842 1.341 975 1176.7 896.8 5.4786 9.0790 5436.0 51.49 1.134 0.847 1.339 1000 1205.1 918.1 5.5001 9.1078 6009.0 47.77 1.139 0.852 1.337 1025 1233.7 939.4 5.5212 9.1360 6629.0 44.39 1.144 0.856 1.335 1050 1262.3 960.9 5.5419 9.1636 7299.0 41.30 1.148 0.861 1.333 1075 291.1 982.5 5.5622 9.1907 8020.0 38.48 1.152 0.865 1.332 1100 1319.9 1004.1 5.5821 9.2172 8797.0 35.9 1.157 0.870 .330 1125 1348.9 1025.9 5.6017 9.2432 9632.0 33.53 1.161 0.874 .329 1150 378. 1047.8 5.6209 9.2688 10529.0 31.35 1.165 0.87 1.327 1175 1407.1 1069.8 5.6399 9.2939 11490.0 29.36 1.168 0.881 1.326 1200 1436.4 1091.9 5.6585 9.3185 12520.0 27.51 1.172 0.885 1.324 t Abstracted with permission from Thermodynamic Properties in SI (Graphs, Tables, and Computational Equa- tions for Forty Substances), by W. C. Reynolds, Published by the Department of Mechanical Engineering, Stanford University, Stanford, CA 94305, 1979. 914 INTERNAL COMBUSTION ENGINE FUNDAMENTALS 82-89 MON 97 - 115 120 100 109 89 92 TABLE D.3 Fuel octane rating Enthalpy of C, CO, CO2, H2, H2O, N2, O2 RON 91-99 120 112 fo(T)-Ro(298.15), kcal/mol 100 106 120 107 7(K) C CO CO2 H, H,O N2 02 (FIA). 0.0685 0.0690 0.0697 0.0580 0.0638 0.0661 0.0675 0.0753 0.0741 0.069 0.0869 0.0292 0.155 0.111 0.405 298 0.000 0.000 0.000 0.000 0.000 0.000 0.000 300 0.004 0.013 0.016 0.013 0.015 0.013 0.013 (4|F), 15.67 9.00 2.467 14.5 17.23 15.13 14.82 6.47 14.4 13.27 14.6 14.5 13.50 11.51 400 0.250 0.711 0.958 0.707 0.825 0.710 34.3 0.72 500 0.569 1.417 1.987 1.406 1.654 1.413 1.455 600 0.947 2.137 3.087 2.106 2.509 2.125 2.210 LHV of mixture, MJ/kg stoich. 700 1.372 2.83 2.873 2.853 2.79 4.245 3.390 2.988 2.85 2.72 2.75 2.808 2.75 2.82 2.78 2.79 2.68 2.70 2.9 2.69 3.40 2.91 800 1.831 3.627 5.453 3.514 4.30 3.596 3.786 900 2.318 4.397 6.702 4.226 5.240 4.355 4.600 heating MJ/kg Lower value, 44.0 1000 2.82 5.183 50.0 10.6 7.984 42.8 4.944 6.20 5.120 5.427 43.2 46.4 44 .3 44.0 40.2 33.8 45 20.0 26.9 10.1 120. 1100 3.347 5.983 9.296 5.670 7.210 5.917 6.266 1200 3.883 6.794 10.632 6.404 8.240 6.718 7.114 heating Higher MJ/kg 47.3 value, 46.1 50.4 1300 45.5 47.8 47.3 41.9 55.5 42.5 1.432 7.616 22.7 11.988 7.148 9.298 7.529 7.971 50 29.7 33.8 10.1 142.0 1400 4.988 8.44 13.362 7.902 10.384 8.350 .835 1500 5.552 9.285 14.750 8.668 11.495 9.179 9.706 Vapor C. 1600 KJ/kg . K 6.122 10.130 16.152 9.446 12.630 10.015 10.583 1.63 1.6 1.72 ~1.7 2.2 1.6 1.93 1.05 1.44 1.1 ~ 1.7 - ~ 1.7 1700 6.696 10.980 17.565 10.233 13.787 10.858 11.465 Specific heat 1800 7.275 11.836 18.987 11.030 14.964 11.707 12.354 1900 7.85 12.697 20.418 11.836 16.160 12.560 13.249 2000 8.442 13.561 21.857 12.651 17.373 13.418 14.149 KJ/kg . K Liqui 19 1.72 1.68 0.63 2100 2.4 2.2 2.5 9.029 14.430 23.303 13.475 18.602 2.1 14.280 2.6 2.5 15.054 2200 9.620 15.301 24.755 14.307 19.84€ 5.14 15.966 2300 10.212 16.175 26.212 15.146 21.103 16.015 16.882 2400 10.80 17.052 27.674 15.993 22.372 17.804 vaporization, 16.886 2500 Heat of KJ/kgt 11.403 17.931 29.141 16.848 23.653 17.761 18.732 350 270 230 426 308 509 358 433 412 840 1103 2600 12.002 18.813 30.613 17.708 24.945 18.638 19.664 J. W. Rose and J. R. Cooper (eds.), Technical Data on Fuel, 7th ed ., British National Committee, World Energy Conference, London, 1977. 2700 12.602 19.696 32.088 18.575 26.246 19.517 20.602 2800 13.203 20.582 33.567 19.448 27.556 20.398 21.545 E. M. Goodger, Hydrocarbon Fuels; Production, Properties and Performance of Liquids and Gases, Macmillan, London, 1975. (density, t 0.51 (2.0+) 0.72-0.78 0.82-0.88 0.78-0.84 (~0.79+) kg/dm3) Specific gravity: 2900 13.807 21.469 35.049 20.326 28.875 (0.72+) 21.280 22.493 (0.0901) (1.25t) 0.692 0.773 0.879 0.867 0.792 0.785 ~ 26 3000 14.412 22.357 36.535 21.210 30.201 22.165 23.446 E. F. Obert, Internal Combustion Engines and Air Pollution, Intext Educational Publishers, 1973 edition. Source: JANAF Thermochemical Tables, National Bureau of Standards Publication NSRDS- 16.04 44.10 2.01 78.11 92.14 114.23 32.04 226.44 46.07 NBS37, 1971. Molecular 12.01 28.01 weight ~18 ~110 ~200 ~ 170 C. F. Taylor, The Internal Combustion Engine in Theory and Practice, vol. I, MIT Press, 1966. + At 1 atm and 25ºC for liquid fuels: at 1 atm and boiling temperature for gaseous fuels. C,H3.8.No.1.(g) RON, research octane number; MON, motor octane number. C.H1.87.(1) C.H1.7,(1) C.H1.8.(1) C16H34(1) Formula C3H3(8) C3H18(1) (phase) C2H6O(1) CH (g) C6H6(1) C,H3(1) CH,O(1) C(s) CO(g) H2(8) (1) liquid phase; (g) gaseous phase; (s) solid phase. + Density in kg/m3 at 0ºC and 1 atm. Data on fuel properties Pure hydrocarbons Carbon monoxide Practical fuelss TABLE D.4 Heavy diesel Light diesel $ Typical values. Natural gas Gasoline Other fuels Isooctane Methane Methanol Hydrogen Propane Benzene Alcohols Toluene Ethanol Cetane Carbon Fuel Sources: 915 INDEX Adiabatic flame temperature, 81, 94 cycle, 184-186 Air: James, 3 constituents of, 65 Atomization of sprays: table of thermodynamic properties, 912 regimes of, 525-529, 532 viscosity, 143 secondary, 532-533 Air/fuel ratio: Autoignition, 462-470, 542-545 definition, 53 chemistry of, 463-467 feedback control, 301-304 cool flames, 465 lambda sensor, 301-303 of hydrogen, 463-464 relative, 71 induction-time correlations, 468, 543-545 stoichiometric, 69-70, 915 Shell model, 469-470 of gasoline, 280 single -, two-stage, 465-466 Air pollution, automotive: Availability: emissions mechanisms, summary, 568-572 analysis, 186-193, 792-797 nature of problem, 5-6 balances, 191-192, 196, 793, 795 sources of emissions, 567-568 combustion loss, 192-193, 196, 795 (see also Carbon monoxide; NO,; conversion efficiency, 84-85 Particulates; Unburned HC emissions) definitions, 84, 186-188, 792-793 Alcohols: distribution, by category, 796-797 antiknock rating, 476-477 losses, actual cycle, 196, 794-797 composition, 68 steady-flow function, 187, 793 methanol combustion, 382 oxygenates: Balance, 20 emissions of, 598 Bearings: as extenders, 476-477 journal, 716 stoichiometric equation, 72 eccentricity, 735 Aldehyde emissions, 568, 598 friction, 717, 736-737 Alkyl compounds (acetylenes, napthenes, olefins, load diagram, 734-735 paraffins), 67-68 schematic, 735 Aromatics, 68 slider, 716 Atkinson: Beau de Rochas, Alphonse, 2-3 917 918 INDEX INDEX 919 Blowby, 361-365 Cetane: Combustion (SI engines): Coolant heat flow, 673-675 unburned HC emissions, 6n, 567, 606-607 index, 542 abnormal phenomena, 451 Courant number, 760 Brake parameters, definition, 46, 48-49 n-hexadecane, 541, 915 burned gas: Crank angle, definition, 44 Burn angles (SI engines): number, 541, 550-552 mixed model, 378-381 Crevices : flame development, 389-390, 421-423, 777 fuel structure dependence, 550-551 temperatures, 379-380, 383 effect on performance, 195, 387-388 overall, 389-390 Charts (see Thermodynamic charts) unmixed model, 378-381 flows in/out, 360-365 rapid-burning, 389-390, 422-423, 777-778 Chemical equilibrium: composition effects, 395 geometry, 361-362 variations in, 415, 422-423 computer codes, 90-92 cycle-by-cycle variations, 413-424 model for, 387-388 Burn rate (SI engines): constants, 87-90 burn rate effects, 415-417, 845-846, 849-850 piston/ring assembly, 361-363 and combustion chamber geometry, 847-848 general principles, 86-94 causes of, 282, 419-422 unburned HC emissions, 604-608 effect on cycle-by-cycle variations, 415-417, Chemical reaction: description of, 372-373, 413-415, 829, 832, Critical pressure ratio, 909 845-846, 849-850 rate constants, 96 849-850 Cycles: effect on performance, 195 rates, 92-96 measures of, 415-418 Alkinson, 184-186 EGR, effect of, 837-839 steady state assumption, 96 cylinder-to-cylinder variations, 282, 413, 420, constant-pressure, 163-164, 178 turbulence effects, 846-848 Combustion: 829-830, 831 constant-volume, 163-164, 169-172, 178 (see also Burn angles; Heat-release rate) constant pressure, 74-75, 126, 172 description of, 371-373, 376 availability analysis, 189-195 Burned gas: constant volume, 73-74, 125, 169 factors that control, 846-850 fuel conversion efficiency, 170, 197 composition: efficiency, 81-83, 509, 601 lean/dilute operating limits, 424-426 four-stroke, 10-11 equilibrium, 93 inefficiency, 154, 195, 509 misfire, 414-415, 424-427, 611 fuel-air, 162, 177-183 low temperature, 104 products composition: motion produced by, 380, 411 assumptions, 177 fraction, 102-103 equilibrium, 93 partial burning, 414 415, 424-427, 611 CI engine, 181 Burning rate analysis, Krieger and Borman, low temperature, 104 speed, effects on, 394, 400-402, 411-412 results, 181-183, 197 511-514 stoichiometry, 68-72 stages of, 372, 389-390, 397-402, 412 SI engine, 178-180 (see also Heat-release analysis) (see also Flame development; Flame thermodynamics of, 376-383 ideal gas standard, 162, 169-177 propagation; Flame structure; Flames) turbulent flame regimes, 396-397 availability analysis, 189-192 Combustion chambers: (see also Flame propagation relations; Flame comparison, 173-177 Carbon monoxide: bowl-in-piston, 342, 353-357, 811, 866-869 structure; Flames; Heat release; Knock; entropy changes, 188-189, 192 background, 6, 567-571 design of, SI engines: Spark ignition) limited-pressure, 163-164, 178 diesels, 592 air breathing, 220-222, 846, 850-851 Compression, crankcase, 11, 238, 244 Otto, 11 oxidation kinetics, 593-596 burn rate, 844-846 Compression-ignition engines: overexpanded, 183-186 SI engines, 592-596, 836 common types, 845 operating cycle, 25-31 two-stroke, 11-12 (see also Catalytic converters) knock, 854-857 (see also Diesel engines) Cylinder pressure: Carburetors, 16-17, 282-294 objectives, 844-846 Compression ratio: analysis of: accelerator pump, 286, 291 optimization strategy, 857-858 definition, 43 CI engines, 508-517 air-bleed compensation, 287-290 swirl, 846, 850-852 effect on efficiency, 170, 172, 175, 182, 197, altitude compensation, 286, 292-293 surface area, 44 841-844 Rassweiler and Withrow method, 385-386 SI engines, 384-389 boost venturis, 16, 286-287 Combustion (CI engines): effect on mep, 176, 183, 842 (see also Heat release) choke, 286, 291-292 consequences, of 492-493 knock limited (critical), 470-472, 854-857 cycle sample size, 418 elementary, 282-285 fuel-air mixing and burning rates, 558-562 typical values, 43, 58, 492, 887 data (CI engines), 504, 513, 885 idle system, 286, 290-291 models for, 504-508, 779-780, 782-784, Compressors: data (SI engines), 162, 372-374, 384, 414 main metering system, 286-290 786-788, 816-818 centrifugal, 238, 258-262, 877-878 with knock, 453-454, 459-462 modern design, 16-17, 285-294 phases of, 505-506 corrected mass flow, 255, 262 measurement of, 384 multiple-barrel, 287 mixing-controlled phase, 506, 558-562 corrected speed, 255, 262 Pmax, 0, „, 415-418, 829 power enrichment, 286, 291 premixed, rapid burning phase, 505, isentropic efficiency, 251 Cylinder volume: transient effects, 293-294 558-560 performance maps, 255, 257-258, 261-262, equation for, 43-44 Catalytic converters: (see also Ignition delay) 270, 273, 878-879 catalyst conversion efficiency, 651-652, 656 photographs of, color plate (between 498 and roots blower, 15, 256-258, 886 Damkohler number, 396, 399 catalytic materials, 649, 651-652, 654, 655-657 499), 497-502 screw, 256-258 Delivery ratio, 237-240, 244, 882-883 degradation, poisoning, 651-653 role of, 493, 555-558 sliding vane, 255-257 Diesel combustion systems: design of, 649-650 summary of, 491-493 velocity diagrams, 260 direct-injection, 32-37, 493-494, 496 light-off temperature, 651, 653 (see also Diesel combustion systems; Ignition Comprex (see Supercharging) bowl-in-piston, swirl, multihole nozzle, NO reduction, 654-655 delay; Fuel sprays) Conservation equations, open system : 493-494, 496-501 oxidation, 649-654 Combustion modeling, 766-778, 779-780, energy, 751-753 M.A.N. " M", 494, 496-501 three-way, 655-657 782-784, 786-788, 816-818 mass, 750-751 quiescent, 493-494, 496-500 920 INDEX INDEX 921 Diesel combustion systems (continued): prechamber (see Prechamber engines) Flame propagation relations, 406-412 mixed lubrication, 718 indirect-injection, 33-34, 494-496 spark-ignition, 15-25 characteristic length, 410, 412 Stribeck diagram, 716-717 swirl prechamber, 494-502 stratified-charge (see Stratified-charge engines) "entrainment" burning laws, 771-778 crankshaft, 734-737 turbulent prechamber, 494-497 Wankel (see Wankel engines) flame areas, 406-409, 766-767 difference, motoring/firing, 720-721 (see also Combustion (CI engines); Fuel Enthalpy, 108-111, 116, 123-127, 903-905 turbulent burning speed, 408-409, 411-412 losses, categories of, 713 sprays) sensible, 113-114, 122 velocity parameters, 408-412 measurement methods, 719-721 Diesel emissions: tables of, 914 Flame quenching, 599-601 piston assembly, 730, 732-734 NO,/particulates trade off, 865-866 stagnation, 251 Flame structure (CI engines): pumping, 47, 168-169, 713-715, 725, 726-728 (see also Carbon monoxide; NO,; Enthalpy of formation: ignition, location of, 502, 556-557 throttling work, 727-728 Particulates; Unburned HC emissions) datum reference state, 77 species concentration data, 557-559 valve flow work, 727-728 Diesel engines: definition, 76 spray/flame photos, color plate (between 498 turbulent dissipation, 719 four-stroke cycle standard values, individual species, 77, 124, and 499), 502, 523, 525, 527, 537 valve train, 737-739 air-cooled, 35-36, 859 913 (see also Combustion (CI engines); Diesel Friction correlations: examples (DI), 32-36, 877-881 of unburned mixture, 123-125 combustion systems) Bishop, 727, 733, 736, 738 examples (IDI), 33-34, 875-877 Equilibrium (see Chemical equilibrium) Flame structure (SI engines), 390-402 crankshaft, con rods, 736 large (marine), 36-37, 883-886 Equivalence ratio (see Fuel/air equivalence ratio) flame area, 394, 406-410, 846-847 piston and rings, 733-734 two-stroke cycle, 14, 37, 883-886 Evaporative HC emissions, 6n, 567 flame thickness, 398-402, 410 total friction mep, 719, 722 Diesel, Rudolf, 4 Exhaust: swirl, effect of, 393 valve train, 738 Droplets: blowdown process, 166, 206, 231-233, 613 Flame volume (SI engines): Friction data: equations for individual, 814-815 displacement process, 167, 206, 231-233 data, 372-373, 409 diesels, 722, 724, 725 Sauter mean diameter, 434-436 (see also Intake and exhaust flow models) relationships, 406-410 engine breakdown tests, 722, 725-726 size distribution, 352-354 Exhaust gas: Flames: SI engines, 721, 723 vaporization, 536-539, 814-815 composition classification, 62-64, 395-397 Friction definitions: Dynamometer, 45-46 data, diesels, 148-149 diffusion, 63 accessory mep, power, work, 714-715 data, SI engines, 146-148 laminar, 63 pumping mep, power, work, 47, 714-715 Efficiency, definitions of: equivalence ratio determination, 148-152 premixed, 63 rubbing friction mep, power, work, 714-715 availability conversion, 84 F/A nonuniformities, 152-154 turbulence, effect on, 390-392, 398-402 total friction mep, power, work, 714-715 catalyst conversion, 651-652, 656 measurement, 145-146 turbulent, 63, 395-397 Friction/lubrication regimes: charging, 239, 244 mass flow rate, 231-232 Flow modeling (see Models, fluid-dynamic boundary, 716-718 combustion, 81-83 recycle, recirculation (EGR), 103 based) mixed, 716 compressor isentropic, 251 temperatures, 232-234, 648 Flows (in-cylinder): hydrodynamic, 716 fuel conversion, 52, 85, 164, 169 enthalpy-averaged, 234 exhaust stroke vortex, 365-367 Fuel-air cycle (see Cycles) mechanical, 49, 723 equivalence ratio effects, 834-835 laser doppler anemometry, 336, 808-809 Fuel-air equivalence ratio, 71 scavenging, 238, 244 EGR, effects of, 837-838 piston/cylinder corner, 365-367, 613-614 availability analysis, effect on, 192-193 thermal conversion, 85 thermodynamic state, 167 through intake valve, 224-225, 227, 229, from exhaust composition, 148-152 trapping, 238, 244 EGR tolerance, SI engines, 837-839 326-330 for optimum SI engine efficiency, 831-835 turbine isentropic, 253 Exhaust gas treatment, 648-660 valve-jet driven, 327-330, 807-809 Fuel-air mixing, diesels, 493, 504-508, 555-558 volumetric, 53-54 (see also Catalytic converters; Particulate velocities at intake valve, 326-327, 808-810 and burning rates, 558-562 Emissions (see Carbon monoxide; NO,; traps; Thermal reactors) (see also Blowby; Crevices; Squish; Swirl; Fuel/air ratio: Particulates; Unburned HC emissions) Exhaust manifold pressures, 214 Turbulence) definition, 53 Emissions index, 56 Flows through nozzles, orifices, restrictions, stoichiometric, 69 Energy, available (see Availability) Flame development process : 906-910 Fuel conversion efficiency: Energy balance, engine, 673-676 effects of combustion chamber geometry, Four-stroke cycle: constant-volume cycle, 170, 182 Engine processes: 846-847 definition, 10-11 compression ratio effects, 170, 175, 182, 197 availability analysis of, 186-193 effects of mixture composition and state, 846, exhaust process, 206-208 equivalence ratio effects, 182, 197 thermodynamic relations for, 164-169 848-849 inlet process, 206-208 constant-pressure cycle, 172, 175 Engines: effects of turbulence, 846-849 p-V diagram, 47, 162, 284, 727 definition, 52, 85 classification, 7-9 factors that control, 846-850 Friction: DI vs IDI diesel, 860-861 components, 12-15 Flame ionization detector, 145-146, 597n, 620 accessory requirements (fan, generator, limited-pressure cycle, 170, 175 compression-ignition (diesel), 25-37 Flame photographs: pumps), 739-740 overexpanded cycle, 184-185 energy balance, 673-676 CI engines, color plate (between 498 and 499) background, 712-713 Fuel conversion efficiency (SI engines): historical, 1-7 SI engine, 390-394, 397-399, 401, 458-460, coefficient of, 716 effect of: maximum work, 83-85 color plate (between 498 and 499) boundary lubrication, 716-718 burn rate, 832-833 multifuel, 39 Flame propagation data, 409, 412, 773-774 hydrodynamic lubrication, 718 compression ratio, 841-844 922 INDEX INDEX 923 Fuel conversion efficiency (SI engines) enthalpy of formation, 913 Heat-release rate (continued): effect on air properties, 67 (continued): gasoline: SI engines: psychrometric chart, 66 equivalence ratio, 830-834 composition, 280, 915 cycle-by-cycle variations, 414-415 relative, 65 Fuel injection (diesels): equilibrium vaporization, 314-315 results, 390, 413-414 Hydrocarbon emissions (see Unburned HC distributor pump, 30-32, 518 - heating values, 78-90, 915 Heat transfer: emissions) in-line pump, 28, 30, 518 hydrocarbons: characteristics of, in engines, 668-670, 672-673 Hydrocarbons: nozzle flow rate, 521 classes of, 66-68 coefficient, 671 burnup, 600, 614-618 nozzle geometry, 526-529 knocking tendency, 470-472 conduction, 670 classes of, 66-68 nozzles, 29, 31, 519-520 hydrogen, 915 convective, 670-671 knocking tendency, 470-474 objectives, 518 autoignition of, 463-464 dimensional analysis, 676-677 oxidation mechanism, 467 single-barrel pump, 518-519 combustion, 398, 773-774 cycle-simulation predictions, 702-704, 707, Hydrogen: systems, 27-31, 517-522 stoichiometric equation, 72 769-771 autoignition of, 463-464 unit injectors, 520-521 ignition quality of, 492, 541-542, 550-552 effect of engine variables, 701-707 combustion, 773-774 Fuel injection (SI engines): isooctane, 67, 915 compression ratio, 703-704 stoichiometric equation, 72 injection timing, 298-299 octane rating, 471 coolant temperature, 704-705 injector design, 295-296 stoichiometry, 69-71 equivalence ratio, 703 Ideal gas: multipoint port systems, 16-17, 294-299 laminar flame speeds, 395, 402-406 load, 702, 796-797 analytic model for, 109-112 air-flow meter, 297-299 primary reference, 475 spark timing, 704 law, 64, 902 fuel transport, 320-321 properties: speed, 673, 675, 679, 702-703, 797 mixtures, 905 mechanical, 298 table of, 915 squish, swirl, 704 relationships, 107-109, 902-905 speed-density, 294-296, 299 thermodynamic, 77, 130-133 wall material, 705-707 thermodynamic properties, 903-905 single-point systems, 299-301 stoichiometric A/F, 70, 915 effect on performance, efficiency, 194-195, (see also Burned gas; Gas properties; Fuel sprays: (see also Alcohols; Cetane; Octane) 770-771, 851-852 Unburned mixture; Working fluids) adiabatic saturation, 538-539 exhaust system, 682-683 Ignition delay: breakup, 522-523, 530-531, 532 Gas constant, 903 evaporation, 535-539, 814-815 Gas properties: intake system, 682 correlations for, 543-545 computer routines for, 130-140 (see also Intake manifold) in engines, 553-554 and flame structure, 555-558 isentropic compression functions, 113-115 prechamber diesels, 787 definition, 505, 539-540 ignition sites, 525, 556-557 molar and mass basis, 107, 904-905 radiation: factors affecting: modeling, 538-539, 780-784, 813 816 apparent emissivity, 684-688 air temperature, pressure, 547-548 equations for droplets, 814-815 molecular weight, 106, 136, 905, 913 chamber wall, 548-549 1-D turbulent jet, 780-781 polynomial functions apparent flame temperature, 685-686 fuels, 130-133 from gases, 683-684 injection timing, 546 multidimensional, 813-816 gas species, 130-131 monochromatic absorption coefficient, load, 546-548 multipackage, 782-784 ratio specific heats, 134, 137, 139, 904 687-688 oxygen concentration, 549-550 multizone, 781-782 prediction formula, 688-689 specific heats, 132, 134, 136, 138, 904 speed, 548 penetration, 529-532 relative importance, 693-694 photographs of, color plate (between 498 and stagnation values, 251, 907 spray parameters, 546-547 499), 523, 525, 527, 537 tables, 127-129, 912, 914 from soot, 683, 684-689 swirl, 549 Heat transfer correlations: unburned mixture, 130-135 fuel property effects, 550-553 spray angle, 526-528 evaluation of, 694-696 (see also Ideal gas; Thermodynamic charts; processes occurring, 540-541 structure, 522-527, 529, 535-537, 555-558 exhaust port, 682-683 swirl, effect of, 524-525, 531-532, 558 Transport properties) (see also Autoignition; Cetane) instantaneous local, 681-682, 695-696 Gasoline (sce Fuels, gasoline) Indicated parameters, definition: temperature distribution, 538-539 instantaneous, spatial average, 678-680, gross, 47-49, 714-715 wall interaction, 523-524 Heat-release analysis: 694-695 net, 47-49, 714-715 (see also Atomization; Droplets) gross, net, 387-388, 510-511 Annand, 678-679, 695 Intake and exhaust flow models: Fuels: IDI diesel engines, 514-517 Woschni, 679-680, 694-695 boundary conditions, 761 additives: one zone, 386-388, 508-511 time-average, 677-679 example results, 311-312, 761-762 antiknock, 473, 475-476 problems with (diesels), 508-509 zonal models, 682, 696, 768-769 finite-difference methods, 759-762 ignition-accelerating, 551-552 two zone (SI engines), 376-382 Heat transfer measurements: gas dynamic models, 313, 756-762 octane improvers, 476-477 Heat-release rate: methods, 689-690 homentropic flow, 758 API gravity, 542 diesels: results, diesels, 692-694 1-D unsteady flow equations, 756-758 distillate: apparent, 509 results, SI engines, 690-692 manifold models: cetane rating, 541-542 data, 504, 511, 516-517, 560-561 Heating values, 78-90, 915 filling and emptying, 311-312, 753-755 diesel index, 542 definition of, 497 higher heating value, 78 Helmholtz resonator, 312-313 nitrogen content, 577 mixing-controlled, 560-562 lower heating value, 78 method of characteristics, 759 sulfur content, 568 variables, effects of, 560-562 Humidity: quasi-steady models, 232, 753-754 924 INDEX INDEX 925 Intake manifold: SAE viscosity classification, 744-745 Models of engine processes, rationale, 748-750 SI engines, 579-581 air flow, 309-313 Lubrication: Models, fluid-dynamic based (multidimensional): NO formation: description of, 309-311 of bearings, 715-718, 734-737 for boundary layers, 803 description of, CI engines, 586-587 transient behavior, 310-312 of piston assembly, 729-734 of combustion, 816-818 equivalence ratio, effect of, 575-576 (see also Intake and exhaust flow models) regimes of, 716-718 flow field predictions, 360, 807-813 kinetics of, 572-576 design, 308-309 system layout, 740-741 concentration distributions, 811-813 rate constants, 573 fuel transport, 314-321 (see also Friction; Bearings) particle traces, 810-812 Zeldovich mechanism, 572 droplet behavior, 316-317 velocities, 807-811 model for, SI engines, 578-581 liquid films, 315-316, 318-320 Manifolds: governing equations, 798-799 rate equation for, 573-574 transient behavior, 310-312, 318-321 tuning of, 215, 217-218 KIVA code, 803, 804, 815-816 temperature, effect of, 574-576 vaporization, 314-321 (see also Exhaust manifold; Intake manifold) numerical methodology, 798, 803-807 NO from fuel N, 577 pressure variation, 212-214, 216, 310-311 Mass fraction burned, SI engines: computing mesh, 804-805, 810 Nitrogen, atmospheric, 65 IC engines (see Engines) data, 372-373, 382, 777-778 discretization practices, 804-806 NO2 formation, 577-578 Internal energy, 108-111, 116-127, 903-905 equations for, 377-378, 381-382, 390 solution algorithms, 806-807 NO ,, definition of, 567, 572 sensible, 113-114, 122 Wiebe function, 390, 768 overview, 797-798 NO, emissions (diesels): Internal energy of formation: Maximum work, 83-85 sprays, 813-816 effect of: standard values, 124 (see also Availability) turbulence models: diluents, 590-591, 861-863 of unburned mixture, 123-125 Mean effective pressure: full-field modeling, 799-802 equivalence ratio, 588-589 coefficient of variation (COV), 417, 424 425 k-e models, 775-777, 801-802, 808, 810 Knock: EGR, 590-591, 861-863 antiknock additives, 4, 473, 475-476 cycle-by-cycle variations, 417, 425 large-eddy simulation, 799 injection parameters, 863-867 Reynolds stress, 802 lead alkyls, 4, 473, 475-476 definitions, 50, 714-715 load, 861-862 subgrid scale, 802-803 MMT, 475-476 friction, 714-715, 825-827 swirl, 866-867 characterization of, 454-456 fuel-air cycle results, 183 Models, thermodynamic-based flame temperature correlation, 591-592 ideal cycles, 171, 173, 176 (phenomenological, zero-dimensional): combustion photographs, 458-461 (see also NO formation) importance of, 59, 823-824 complex engine systems, 789-792 damage, 456-457 NO, emissions (SI engines): transient behavior, 791-792 deposit effects on, 477-478 overexpanded cycle, 185-186 effect of: description of, 375, 450-457 pumping, 169, 714-715 turbocharged/turbocompounded, 789-792 compression ratio, 844 detection, 454 relationships for, 50-51, 57, 823-824 DI diesels : diluents, 582-584 effect on heat transfer, 707 Mean effective pressure (DI diesels): model structure, 784-785 equivalence ratio, 581-585, 835-836 effect of: simulation results, 784-785 end gas, 457-462, 467-470 EGR, 582-585, 836-838 injection parameters, 863-864 single zone combustion models, 778-780 load, speed, 840-841 temperature, 468-469 spray models, 780-784 impact of, 456-457, 852-854 load and speed, 858-860, 877-878 spark timing, 585-586, 829 full load, 826-827 (see also Fuel sprays) intensity, 455-456 (see also Catalytic converters; NO formation) Mean effective pressure (IDI diesels): emissions, 765, 787-788 Noise, 5, 571-572 pressure waves, 461-462 effect of: IDI diesels, 784-788 sensors, 872 injection parameters, 863-864 prechamber phenomena, 784-787 theories: load and speed, 860, 875-876 simulation results, 787-788 Octane: autoignition, 457-458, 462 full load, 826-827 open system conservation equations, 763-765 detonation, 457-458 antiknock index, 474 Mean effective pressure (SI engines): overall structure, 762-763 (see also Autoignition; Compression ratio; fuel sensitivity, 473-474 effect of: SI engines: Fuels; Octane) and knock, 854-855 compression ratio, 842 combustion models, 766-778 number, 471-474 Laminar flame: equivalence ratio, 830-832 cycle simulation results, 769-771 motor method, 471-473 speed, 395, 402-406 heat transfer, 851 "entrainment " burning laws, 771-778 requirement, 453, 474, 478, 852-857 correlations, 403-406 wide open throttle, 824-827, 839-840 flame geometry models, 766-768 research method, 471-473 data, 403-405 Mechanical efficiency: stochastic models, 787-788 road, 474 and SI engine combustion, 771-775, definition, 49 turbulence intensity models, 775-777 oxygenates as extenders, 476-477 777-778, 848-849 values, SI engine, 722-723 (see also Mass fraction burned) ratings of fuels, 915 straining effects, 406 Methanol combustion (SI engine), 382 Mole, 903 Odor, diesel, 568 thickness, 395, 402 Mixture nonuniformity, quality, 152-154, 282, Molecular weights, values, 913 Oil consumption rate, 610 Langen, Eugen, 2 314, 829-832 Moment of inertia of charge, 353 Organic compounds : Lead alkyls, 4, 473, 475-476 Mixture requirements (SI engines), 279-282, classes of, 66-68 Lenoir, J. J. E ., 2 833-834 NO concentrations, in-cylinder: (see also Fuels) Lubricant: steady, transient, 834 DI diesel, 587-589 Otto: requirements, 741-745 typical schedules, 281-282 IDI diesel, 589-590 cycle, 11 ... .. .... 926 INDEX INDEX 927 Otto (continued): description of, 13-14, 729 Pressure (continued): Nicolaus, A ., 2-3 forces on, 731-733 energy, by phase, 429-431 stagnation, 251, 907-908 Oxygen (lambda) sensor, 301-303 friction, 729-734 expansion velocities, 430-431 (see also Cylinder pressure) Oxygenate emissions, 568, 598 Piston rings: phases (arc, breakdown, glow), 427-429 Pressure-volume diagram: functions, 729-730 plasma volumes, 430, 434 four-stroke cycle, 47, 162, 384 Paraffins: lubrication, 730-732 temperature distributions, 431-434 ideal cycles, 163, 176, 194 ignition limits, 465-466 oil film thickness, 731-732 Spark ignition: log p vs log V, 384-385 knocking tendency, 470-472 Reynolds equation, 731 current and duration effects, 445-446 polytropic relation, 385, 554 molecular structure, 67 nomenclature, 729 description of, 17, 397, 427 pumping loop, 727 Part-throttle (SI engines): sealed ring-orifice design 605-606, 610 flow effects, 435-436 two-stroke cycle, 47 efficiency, 833-834, 843-844 Piston speed: fundamentals, 427-437 Pumping mean effective pressure, 169 mixture requirements, 280-282, 834 instantaneous, 45 models of, 433-435 Pumping work: performance, 833 mean: requirements for, 437-438 definition, 47, 714 Particulate emissions (diesels): definition, 44 (see also Spark ignition systems, Spark plugs) diesel engines, 492 effect of: importance of, 59, 839 Spark ignition engines: ideal cycle, 168 injection parameters, 864-866 maximum values of, 45, 887 examples, 13, 21-24 SI engines, 827 load, 861-862 Pollutant formation mechanisms: mixture requirements, 279-282, 833-834 (see also Friction, pumping) Particulate traps, 659-660 equivalence ratio, effect of, 570-571 operating cycle, 15-19 Particulates, diesels: summary (CI engine), 571-572 Spark ignition systems: Rapid compression machines: Ames test, 631 summary (SI engine), 568-571 available voltage, 439 results from, 466, 502, 523-524, 556, 560-561 composition of, 627-630, 647-648 (see also Carbon monoxide; NO formation; breakdown, 446 Residual gas flow, 206, 224, 327 distribution in cylinder, 631-635 Particulates; Unburned HC emissions) capacitive-discharge, 441 Residual gas fraction: HC absorption, condensation, 646-648 Polytropic compression relation, 385, 554 coil, 438-440 data, 230-231 measurement techniques, 626-627 Ports (four-stroke cycle): flame-jet, 447-450 definition, 102 dilution ratio effects, 646-647 effect on: higher energy, 445-446 Residual gas mixing, 420, 811-813 oil, contribution from, 629-630, 647-648 flow discharge coefficient, 229-230 magneto, 442 Roots blowers, 15, 256-258, 886 size, 628-631 valve flow area, 222-224 plasma-jet, 446-447 soluble fraction, 629, 646-648 geometry of, 220-224 required voltage, 439 Scavenging, 235-245 soot formation fundamentals, 635-642 Ports (two-stroke cycle): transistorized coil, 440-441 charge short-circuiting, 240, 242 soot oxidation, 642-646 discharge coefficients, 247-248 Spark kernal, photographs, 397-398, 436 crankcase, 11-12, 881 specific surface area, 631, 646 flow through, 246-248 Spark plugs: cross-scavenged, 235-236 spherules, 627-628 geometry of, 245-248 design, 442-443 data, 244-245 structure, 627-631 timing, 237 electrode geometry effects, 443-445 flow visualization of, 240-242 (see also Particulate emissions; Soot Power: fouling, 437 loop-scavenging, 235-236, 242-243 formation; Soot oxidation) brake, definition of, 46 heat rating, 443 models for, 239-240, 811 Particulates, SI engines, 626 correction factors for, 54 Spark timing: uniflow, 235-237, 243, 245, 884-886 Peclet number, 599-600 friction, 48, 825-827 EGR, effect of, 838 Second Law analysis (see Availability) Performance of engines, summary, 58, 866-888 full load: and knock, 852-854 Smoke, effect of: . Performance maps: DI diesels, 826-827, 879-890, 885 EGR 863 maximum brake torque (MBT), 18, 373-375, description of, 839 IDI diesels, 826-827 827-829 injection parameters, 864-866, 867 DI diesels, 858-860, 878-879 SI engines, 824-827, 873, 883 rules for optimum, 375, 828-829 load, 861-862 IDI diesels, 860, 875-876 rated, definition of, 43 Specific emissions: swirl, 867 SI engines, 839-840, 874 relationships for, 45-46, 49, 823 definition, 56 (see also Particulate emissions) Performance parameters: road-load, 49 importance of, 59 Soot formation: importance of, 42-43, 59, 823-824 specific, definition of, 57 Specific fuel consumption, definition, 51, 59 particle formation, 636, 638-639 relationships for, 56-57, 823-824 Prandtl number, expressions for, 142, 144 Specific fuel consumption (diesels): particle growth, 636, 639-642 typical values, 58, 824-827, 887 Prechamber engines: effect of: polycyclic aromatic HC, 636, 639 EGR, 863 Piston: designs, 33-34 pyrolyses, 633, 635, 638 acceleration, 732 flows, 357-360 injection parameters, 863-865, 867 regions of, diesel, 498-502, 536-537 heat outflow, 701 gas displacement, 359-360 load and speed, 858-860, 875-881 Soot limit on diesel performance, 492 temperature distribution, 698-699, 700-701, nozzle throat, 358-359 swirl, 866-867 Soot oxidation, 642-646 full load, 826-827 705 swirl velocities, 360 Soot, radiation from, 684-689 velocity, 44-45 Pressure: Spark discharge: Specific fuel consumption (SI engines): mean effective (see Mean effective pressure) effect of: Piston assembly: chemistry, 431-433 A/F, equivalence ratio, 831-835 928 INDEX INDEX 929 Specific fuel consumption (SI) (continued): squish interaction, 810-811, 868-869 Turbocharged diesels: intensity at TC, 341 burn rate, 832-833 velocity distribution, 351-353, 809, 810-812 combustion characteristics, 879-880 laser doppler anemometry, 336, 808 EGR, 837-839 Swirl generation: DI engine performance, 877-881 mean, 331-332, 336 heat transfer, 851-852 during compression, 349-353, 496-497 different supercharging methods, 249-250, with swirl, 342, 353 spark timing, 828 during intake, 345-349, 496-497 875-877, 879-881 Two-stroke cycle, 11-12 part throttle, 839-840 with ports (4-stroke): hyperbar system, 881 charge compression, 237-238 wide-open throttle, 824-827 helical, 346, 348-349, 810-812 IDI engine performance, 875-877 diesels: Specific power, 57, 59 tangential, 345-346, 348-349, 812 two-stage, 249-250, 879-880 combustion characteristics, 885 Specific volume, 54, 59 valve masking, 346-347 two-stroke, 883-886 efficiency of, 883-886 Specific weight, 54, 59 turbocompounding, 249-250, 789-791, performance of, 883-886, 887 Speed, rated engine, 43, 493 879-881, 884 scavenging data, 244-245 Squish: Temperature-entropy diagram, 188-189, 793-794 Turbocharged SI engines: charge purity, 238, 244-245 area, 353 Temperatures: advantages of, 870 p-V diagram, 47 impact of, 851-852 combustion chamber, 672 boost pressure, 872-873 scavenging, 235-245, 881-884 motion, 353-354 components, 698-707 charge, p, T, 870-871 SI engines: swirl interaction, 868-869 cylinder head, 699, 705 compared with NA engines, 873-874 bsfc, 883 velocity, 353-357: cylinder liner, 699-700, 705 compression ratio, 871-872, 874 charging efficiency, 243-244, 881-883 bowl-in-piston, 354-357, 810-811, 868 exhaust valve, 700, 705 knock impact, 869-872 emissions, 883 decrements, 355-356 piston, 698-699, 700-701, 705 power, torque, 873-874 examples, 24, 881-883 wedge chamber, 354 Thermal boundary layers, 697-698, 768-769 spark advance, 871-872 performance of, 881-883, 887 Stagnation pressure, 251, 907-908 Thermal insulation of engines, 705-707, 881 Turbocharger: trapping efficiency, 243-244, 882-883 Stratified-charge engines, 37-40 Thermal properties: dynamics, 789-792 direct-injection, 38-39, 815-816 ceramics, 706 layout, 20-23, 208, 790 Unburned HC emissions (diesels): M.A.N ., 38-39 metals, 706 matching, 791 contribution to particulates, 620 Texaco type, 38, 815-816 Thermal reactors, 648, 657-659 operating characteristics, 269-270, 877-879 effect of: prechamber designs, 39-40, 448-450 Thermodynamic charts: thermodynamic relationships, 249-254, 259, EGR, 863 Sulfate emissions, 568, 626, 653-654 burned mixture, 116-123 264-265 injection parameters, 863-864 Supercharging: datum, 116 Turbocharging, 249-250 load, 861-862 aftercooling, intercooling, 249-250, 870-871, low temperature, 122 constant pressure, 263 mechanisms, 620-622, 625 873-874, 876-877, 878-881 isentropic compression, 115 pulse, 263 nozzle sac volume, 623-624 charge cooling, 22, 249-250 mixture composition for, 113 two-stage, 249-250, 791, 879-880 overleaning, 622-623 Comprex, 249, 270-273 for unburned mixture, 112-115 wastegate, 270, 873, 875, 878 quenching and misfire, 625 diesel performance, 875-877 Thermodynamic relations: (see also Supercharging) undermixing, 623-625 performance map, 273 engine processes, 164-169 Turbocompounding, 249-250, 789-791, 881, 884 Unburned HC emissions (SI engines) : wave processes in, 271-273 ideal gas, 107-109, 902-905 Turbulence: absorption/desorption in oil, 608-610 mechanical, 249-250, 875-877 Throttle plate: character of, 330-331, 339-340 burn rate effects, 611-612, 845-846 pressure wave, 249, 270-273, 875-877 flow through, 305-308 flames, effect on, 390-392, 410-412, 771-778, combustion quality effects, 610-612 methods of, 248-250 fuel atomization at, 317 847-850 composition of, 597, 614-615 Roots blower, 256-258, 876-877, 886 geometry, 304-306 models, 775-777, 799-803 crevice mechanism, 604-608 (see also Turbocharging) Torque: scales: deposit mechanism, 612 Surface ignition, 375 brake, data, 828, 859, 873, 876, 883 data, 342, 401, 410, 412 effect of: different types of, 450-453 brake, definition, 46 integral, 333-334 compression ratio, 844 preignition damage, 456-457 relationships for, 45-46, 823-824 Kolmogorov, 335 Swirl, 342-353 microscale, 335 equivalence ratio, 570-571, 835-836 spark timing, effect of, 828, 853 load, speed, 840-841 amplification, 349-353 Transport properties, 141-145 velocities : spark timing, 829 and bowl-in-piston chambers, 349-353, 496, thermal conductivity, 141 autocorrelations, 333-334 exhaust concentrations, 602-603 866-869 viscosity, 141-144 data, 338-342, 410-412, 809 flame quenching, 603-604 coefficient, 344-345 Turbines: definitions, 331-333 mechanisms, 568-570, 601-603, 618-619 definition of, 342 A/R ratio, 872 ensemble-averaged, 331-333, 336-337, in diesels, 493-497, 866-869 axial, 263, 266-269 410-411 oxidation, 600, 614-618 oxygenates, 598 flame structure impact, 393-394 isentropic efficiency, 253 individual-cycle mean, 332-333, 337, reactivity, 597-598 friction effects on, 349-351 performance maps, 266-269 410-411 secondary air, 616-617 measurements, 343-349 radial, 263-267, 269 intensity, 331, 336 337, 353, 410-412, ratio, 344-345, 352 775-777, 809, 811 transport mechanisms, 612-614 velocity diagrams, 265-268 (see also Catalytic converters) 930 INDEX Unburned mixture: fuel factors, 210-211 composition, 102-107 heat transfer, 211, 217-218 properties, 112-116, 130-135 manifold pressures, 211-216 Unit conversion factors, 899-901 runner length, 218 speed, 216-220, 761-762 Valve: valve timing, 215, 217-220 choking, 217, 228 ideal cycle, 179, 209-212 curtain area, 226 and performance, 850-851 diameters, 222 discharge coefficient, 226-230 Wankel: flow area, 222-224 Felix, 4 flow pattern, 227, 229, 327-329, 333, 807-808 rotary engine, 4 flow rate, 225-226, 231-232 components, 23-25 geometry, 220-224 example, 26 Mach index, 228 operation, 24 -25 mean inlet Mach no ., 228 Wastegate, 270, 873, 875, 878 overlap, 206, 224 Weber number, 532 pseudo flow velocity, 224-225 Work per cycle, indicated, 46-47, 164 timing, 224-225 Working fluids: velocities, through, 327, 348-349, 808-810 constituents, 100-102 Valve train configurations, 737-739 properties and composition: Valve train friction, 737-739 computer routines, 130-139 Volumetric efficiency : data on, 911-915 correction factors for, 54-55 thermodynamic models, categories of, 101-102 definition of, 53-54 (see also Air: Burned gas; Exhaust gas; Fuels; effects of: Gas properties; Unburned mixture)