Propulsion Systems for Hybrid Vehicles [2nd ed.] 1849191476, 9781849191470

Worldwide the automotive industry is challenged to make dramatic improvements in vehicle fuel economy, some already legi

516 119 7MB

English Pages xiv+594 [610] Year 2010

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Propulsion Systems for Hybrid Vehicles, 2nd Edition......Page 4
Contents......Page 8
Preface......Page 14
1 Hybrid vehicles......Page 16
1.1 Electric engine hybrids 2010......Page 26
1.2 Limits of engine-only actions......Page 27
1.3 Vehicle electrification and more electric vehicle......Page 30
1.4.1 Partnership for new generation of vehicle goals......Page 33
1.4.2 Engine downsizing......Page 34
1.4.3 Drive cycle characteristics......Page 37
1.4.4 Hybrid vehicle performance targets......Page 41
1.4.5 Basic vehicle dynamics......Page 42
1.5.1 Components of road load......Page 47
1.5.2 Friction and wheel slip......Page 53
1.6 Predicting fuel economy......Page 56
1.6.2 Brake specific fuel consumption......Page 57
1.6.3 Fuel economy and consumption conversions......Page 58
1.7 Internal combustion engines: A primer......Page 60
1.7.1 What is brake mean effective pressure (BMEP)?......Page 62
1.7.2 BSFC sensitivity to BMEP......Page 64
1.7.3 ICE basics: Fuel consumption mapping......Page 66
1.7.4 Emissions regulations......Page 67
1.8.1 The connected car, V2G......Page 71
1.8.2 Grid connected HEV20 and HEV60......Page 74
1.8.3 Charge sustaining and charge depleting......Page 77
1.9 Exercises......Page 79
References......Page 80
2 Hybrid architectures......Page 82
2.1.1 Locomotive drives......Page 86
2.1.2 Series–parallel switching......Page 89
2.1.3 Load tracking architecture......Page 91
2.2 Pre-transmission parallel configurations......Page 92
2.2.1 Energy recuperator systems......Page 94
2.2.2 Micro hybrid......Page 95
2.2.3 Mild hybrid......Page 96
2.2.4 Power assist......Page 99
2.2.5 Dual mode......Page 100
2.3 Pre-transmission combined configurations......Page 101
2.3.1 Power split......Page 103
2.3.2 Power split with shift......Page 108
2.3.3 Continuously variable transmission derived......Page 111
2.3.4 Integrated hybrid assist transmission......Page 112
2.4 Post-transmission parallel configurations......Page 115
2.4.1 Post-transmission hybrid......Page 116
2.4.2 Wheel motor hybrid......Page 117
2.5.1 Launch assist......Page 119
2.5.2 Hydraulic–electric post-transmission......Page 120
2.5.3 Very high voltage electric drives......Page 121
2.6.1 Texas A&M University transmotor......Page 122
2.6.2 Petrol electric drivetrain......Page 123
2.6.3 Swiss Federal Institute flywheel concept......Page 124
2.7.1 Catenary powered vehicles with ultra-capacitors......Page 125
2.7.2 Catenary powered vehicles with wayside ultra-capacitors......Page 126
2.8 Electric four wheel drive......Page 128
2.8.1 The E4 system......Page 129
2.9 Exercises......Page 130
References......Page 132
3 Hybrid power plant specifications......Page 134
3.1 Grade and cruise targets......Page 139
3.1.2 Wide open throttle......Page 142
3.2.2 Lane change......Page 143
3.3.1 Series RBS......Page 144
3.3.3 RBS interaction with ABS......Page 148
3.3.4 RBS interaction with IVD/VSC/ESP......Page 149
3.4.1 Types of drive cycles......Page 150
3.4.2 Electric vehicle and regenerative electric vehicle cycles for PHEVs......Page 151
3.4.3 Average speed and impact on fuel economy......Page 153
3.4.5 Wide open throttle launch......Page 154
3.5.2 Range and performance......Page 155
3.6.2 Electrical burden......Page 156
3.6.4 Neutral idle......Page 157
References......Page 158
4 Sizing the drive system......Page 160
4.1 Matching the electric drive and ice......Page 162
4.1.2 Gear step selection......Page 164
4.1.3 Automatic transmission architectures......Page 166
4.1.3.1 Simpson type......Page 167
4.1.3.3 Lepelletier type......Page 169
4.2 Sizing the propulsion motor......Page 170
4.2.1 Step 1......Page 173
4.2.2 Step 2......Page 174
4.2.3 Step 3......Page 177
4.2.4 Torque and power......Page 179
4.2.5 Constant power speed ratio (CPSR)......Page 183
4.2.6 Machine sizing......Page 185
4.3 Sizing the power electronics......Page 188
4.3.1 Switch technology selection......Page 190
4.3.2 kVA/kW and power factor......Page 191
4.3.3 Ripple capacitor design......Page 195
4.3.4 Switching frequency and PWM......Page 198
4.4 Selecting the energy storage technology......Page 199
4.4.1 Lead–acid technology......Page 209
4.4.3 Lithium ion......Page 211
4.4.5 Fuel cell......Page 214
4.4.6 Ultra-capacitor......Page 219
4.5 Electrical overlay harness......Page 223
4.5.1 Cable requirements......Page 224
4.5.2 Inverter bus bars......Page 227
4.5.4 Power distribution centres......Page 230
4.6 Communications......Page 231
4.6.1 Communication protocol: CAN......Page 234
4.6.2 Power and data networks......Page 235
4.6.3 Future communications: TTCAN......Page 237
4.6.4 Future communications: FlexRay......Page 238
4.6.5 Competing future communications protocols......Page 241
4.6.6 Diagnostic test codes (DTC)......Page 242
4.7.1 Steering systems......Page 243
4.7.2 Braking systems......Page 244
4.7.4 Thermal management......Page 245
4.7.5 Human–machine interface......Page 248
4.8.1 Cost analysis......Page 249
4.8.2 Weight tally......Page 251
4.9 Exercises......Page 252
References......Page 253
5.1 Permanent magnets......Page 258
5.1.1 Permanent magnets: A primer......Page 259
5.1.2 What happened to Alnico?......Page 262
5.1.3 Rare earth permanent magnets......Page 263
5.2 Brushless machines......Page 267
5.2.1 Brushless dc......Page 271
5.2.2 Brushless ac......Page 276
5.2.3 Design essentials of the SPM......Page 280
5.2.4 Dual mode inverter......Page 289
5.3 Interior permanent magnet......Page 292
5.3.1 Buried magnet......Page 293
5.3.2 Flux squeeze......Page 297
5.3.3 Mechanical field weakening......Page 302
5.3.4 Multilayer designs......Page 304
5.4.1 Classical induction......Page 305
5.4.2 Winding reconfiguration......Page 308
5.4.3 Pole changing......Page 309
5.4.3.1 Hunt winding......Page 310
5.4.3.2 Electronic pole change......Page 311
5.4.3.3 Pole–phase modulation......Page 313
5.4.3.4 Pole changing PM......Page 319
5.5 Variable reluctance machine......Page 321
5.5.1 Switched reluctance......Page 322
5.5.2 Synchronous reluctance......Page 326
5.6.1 Dynamic performance comparisons......Page 328
5.6.2 Comparisons for electric vehicles......Page 330
5.6.3 Comparisons for hybrid vehicles......Page 331
5.7 Exercises......Page 334
References......Page 335
6 Power electronics for ac drives......Page 340
6.1 Semiconductor device technologies......Page 341
6.1.1 Trends in power semiconductors......Page 342
6.1.2 Wide bandgap devices......Page 343
6.2 Essentials of pulse width modulation......Page 345
6.3 Resonant pulse modulation......Page 350
6.4 Space vector PWM......Page 352
6.5 Multilevel inverters......Page 361
6.6 Comparison of PWM techniques......Page 363
6.7 dc/dc converters......Page 364
6.8 Thermal design......Page 366
6.9 Reliability considerations......Page 371
6.10 Sensors for current regulators......Page 374
6.11 Interleaved PWM for minimum ripple......Page 376
6.12 Exercises......Page 378
References......Page 380
7 Drive system control......Page 382
7.1 Essentials of field oriented control......Page 383
7.2 Dynamics of field oriented control......Page 388
7.3 Sensorless control......Page 394
7.4 Efficiency optimization......Page 399
7.5 Direct torque control......Page 403
References......Page 406
8.1 Traction motor......Page 410
8.1.1 Core losses......Page 413
8.1.2 Copper losses and skin effects......Page 418
8.2.1 Conduction......Page 421
8.2.2 Switching......Page 423
8.2.3 Reverse recovery......Page 424
8.3 Distribution system......Page 425
8.5 Efficiency mapping......Page 427
8.6 Exercises......Page 431
References......Page 432
9 Hybrid vehicle characterization......Page 434
9.1 City cycle......Page 442
9.2 Highway cycle......Page 443
9.3 Combined cycle......Page 444
9.4 European NEDC......Page 445
9.5 Japan 10–15 mode......Page 447
9.6 Regulated cycle for hybrids......Page 448
Exercises......Page 450
References......Page 453
10 Energy storage technologies......Page 454
10.1 Battery systems......Page 456
10.1.1 Lead–acid......Page 462
10.1.2 Nickel-metal hydride......Page 464
10.1.3 Lithium ion......Page 469
10.2 Capacitor systems......Page 476
10.2.1 Symmetrical ultra-capacitors......Page 481
10.2.2 Asymmetrical ultra-capacitors......Page 485
10.2.3 Ultra-capacitors combined with batteries......Page 488
10.2.4 Hybridized battery example......Page 496
10.2.5 Ultra-capacitor cell balancing......Page 497
10.2.5.1 Dissipative cell equalization......Page 499
10.2.5.2 Non-dissipative cell equalization......Page 500
10.2.5.3 Electrochemical double layer capacitor specification and test......Page 503
10.3 Hydrogen storage......Page 508
10.3.1 Metal hydride......Page 510
10.4 Flywheel systems......Page 511
10.6.1 Battery model......Page 514
10.6.2 Fuel cell model......Page 519
10.6.3 Ultra-capacitor model......Page 522
10.7 Exercises......Page 531
References......Page 534
11 Hybrid vehicle test and validation......Page 538
11.1 Vehicle coast down procedure......Page 540
11.2 Sports utility vehicle test......Page 542
11.3 Sports utility vehicle plus trailer test......Page 544
11.4 Class-8 tractor test......Page 547
11.5 Class-8 tractor plus trailer test......Page 550
11.6 Exercises......Page 554
References......Page 556
12 Automated electrified transportation......Page 558
12.1 Personal rapid transit......Page 561
12.2 Automated highway system......Page 562
12.3 Non-contacting power transfer......Page 564
12.3.1 Inductive coupling technology......Page 565
12.3.2 Radiated near-field power transfer......Page 567
12.4 Transporting cargo......Page 568
References......Page 576
Appendix A......Page 578
References......Page 580
Index......Page 582
Recommend Papers

Propulsion Systems for Hybrid Vehicles [2nd ed.]
 1849191476,  9781849191470

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Renewable Energy Series 7

Dr John M. Miller PE

Propulsion Systems for Hybrid Vehicles 2nd Edition John M. Miller

IET RENEWABLE ENERGY SERIES 7

Propulsion Systems for Hybrid Vehicles

Other volumes in this series: Volume 1 Volume 6 Volume 7 Volume 8 Volume 9 Volume 11

Distributed generation N. Jenkins, J.B. Ekanayake and G. Strbac Microgrids and active distribution networks S. Chowdhury, S.P. Chowdhury and P. Crossley Propulsion systems for hybrid vehicles, 2nd edition J.M. Miller Energy: resources, technologies and the environment C. Ngo Solar photovoltaic energy A. Labouret and M. Villoz Cogeneration: a user’s guide D. Flin

Propulsion Systems for Hybrid Vehicles 2nd Edition

John M. Miller

The Institution of Engineering and Technology

Published by The Institution of Engineering and Technology, London, United Kingdom The Institution of Engineering and Technology is registered as a Charity in England & Wales (no. 211014) and Scotland (no. SC038698). † 2004 The Institution of Electrical Engineers † 2008, 2010 The Institution of Engineering and Technology First published 2004 (978-0-86341-336-0) Paperback edition 2008 (978-0-86341-915-7) Second edition 2010 This publication is copyright under the Berne Convention and the Universal Copyright Convention. All rights reserved. Apart from any fair dealing for the purposes of research or private study, or criticism or review, as permitted under the Copyright, Designs and Patents Act 1988, this publication may be reproduced, stored or transmitted, in any form or by any means, only with the prior permission in writing of the publishers, or in the case of reprographic reproduction in accordance with the terms of licences issued by the Copyright Licensing Agency. Enquiries concerning reproduction outside those terms should be sent to the publisher at the undermentioned address: The Institution of Engineering and Technology Michael Faraday House Six Hills Way, Stevenage Herts, SG1 2AY, United Kingdom www.theiet.org While the author and publisher believe that the information and guidance given in this work are correct, all parties must rely upon their own skill and judgement when making use of them. Neither the author nor publisher assumes any liability to anyone for any loss or damage caused by any error or omission in the work, whether such an error or omission is the result of negligence or any other cause. Any and all such liability is disclaimed. The moral rights of the author to be identified as author of this work have been asserted by him in accordance with the Copyright, Designs and Patents Act 1988.

British Library Cataloguing in Publication Data A catalogue record for this product is available from the British Library ISBN 978-1-84919-147-0 (paperback) ISBN 978-1-84919-148-7 (PDF)

Typeset in India by MPS Ltd, a Macmillan Company Printed in the UK by CPI Antony Rowe, Chippenham

For Quentin Patrick Miller† 1983–2005 You will always be missed

Contents

Preface

xiii

1 Hybrid vehicles 1.1 Electric engine hybrids 2010 1.2 Limits of engine-only actions 1.3 Vehicle electrification and more electric vehicle 1.4 Performance characteristics of road vehicles 1.4.1 1.4.2 1.4.3 1.4.4 1.4.5

1.5

Calculation of road load 1.5.1 1.5.2

1.6

Emissions Brake specific fuel consumption Fuel economy and consumption conversions

Internal combustion engines: A primer 1.7.1 1.7.2 1.7.3 1.7.4

1.8

Components of road load Friction and wheel slip

Predicting fuel economy 1.6.1 1.6.2 1.6.3

1.7

Partnership for new generation of vehicle goals Engine downsizing Drive cycle characteristics Hybrid vehicle performance targets Basic vehicle dynamics

What is brake mean effective pressure (BMEP)? BSFC sensitivity to BMEP ICE basics: Fuel consumption mapping Emissions regulations

Grid connected hybrids 1.8.1 1.8.2 1.8.3

The connected car, V2G Grid connected HEV20 and HEV60 Charge sustaining and charge depleting

1.9 Exercises References 2 Hybrid architectures 2.1 Series configurations 2.1.1 2.1.2 2.1.3

2.2

Locomotive drives Series–parallel switching Load tracking architecture

Pre-transmission parallel configurations 2.2.1 2.2.2 2.2.3

Energy recuperator systems Micro hybrid Mild hybrid

1 11 12 15 18 18 19 22 26 27 32 32 38 41 42 42 43 45 47 49 51 52 56 56 59 62 64 65 67 71 71 74 76 77 79 80 81

viii

Propulsion systems for hybrid vehicles 2.2.4 2.2.5

2.3

Pre-transmission combined configurations 2.3.1 2.3.2 2.3.3 2.3.4

2.4

Texas A&M University transmotor Petrol electric drivetrain Swiss Federal Institute flywheel concept

Ultra-capacitor-only vehicles 2.7.1 2.7.2 2.7.3

2.8

Launch assist Hydraulic–electric post-transmission Very high voltage electric drives

Flywheel systems 2.6.1 2.6.2 2.6.3

2.7

Post-transmission hybrid Wheel motor hybrid

Hydraulic post-transmission hybrid 2.5.1 2.5.2 2.5.3

2.6

Power split Power split with shift Continuously variable transmission derived Integrated hybrid assist transmission

Post-transmission parallel configurations 2.4.1 2.4.2

2.5

Power assist Dual mode

Catenary powered vehicles with ultra-capacitors Catenary powered vehicles with wayside ultra-capacitors Ultra-capacitor trolley bus vehicles

Electric four wheel drive 2.8.1 2.8.2

The E4 system Production ‘Estima Van’ example

2.9 Exercises References 3 Hybrid power plant specifications 3.1 Grade and cruise targets 3.1.1 3.1.2

3.2

Launch and boosting 3.2.1 3.2.2

3.3

Series RBS Parallel RBS RBS interaction with ABS RBS interaction with IVD/VSC/ESP

Drive cycle implications 3.4.1 3.4.2 3.4.3 3.4.4 3.4.5

3.5

First two seconds Lane change

Braking and energy recuperation 3.3.1 3.3.2 3.3.3 3.3.4

3.4

Gradeability Wide open throttle

Types of drive cycles Electric vehicle and regenerative electric vehicle cycles for PHEVs Average speed and impact on fuel economy Dynamics of acceleration/deceleration Wide open throttle launch

Electric fraction 3.5.1 3.5.2

Engine downsizing Range and performance

84 85 86 88 93 96 97 100 101 102 104 104 105 106 107 107 108 109 110 110 111 113 113 114 115 115 117 119 124 127 127 128 128 128 129 129 133 133 134 135 135 136 138 139 139 140 140 140

Contents 3.6

Usage requirements 3.6.1 3.6.2 3.6.3 3.6.4

Customer usage Electrical burden Grade holding and creep Neutral idle

3.7 Exercises References 4 Sizing the drive system 4.1 Matching the electric drive and ice 4.1.1 4.1.2 4.1.3

4.2

Sizing the propulsion motor 4.2.1 4.2.2 4.2.3 4.2.4 4.2.5 4.2.6

4.3

Cable requirements Inverter bus bars High voltage disconnect Power distribution centres

Communications 4.6.1 4.6.2 4.6.3 4.6.4 4.6.5 4.6.6

4.7

Lead–acid technology Nickel-metal hydride Lithium ion Metal–air batteries Fuel cell Ultra-capacitor Flywheels

Electrical overlay harness 4.5.1 4.5.2 4.5.3 4.5.4

4.6

Switch technology selection kVA/kW and power factor Ripple capacitor design Switching frequency and PWM

Selecting the energy storage technology 4.4.1 4.4.2 4.4.3 4.4.4 4.4.5 4.4.6 4.4.7

4.5

Step 1 Step 2 Step 3 Torque and power Constant power speed ratio (CPSR) Machine sizing

Sizing the power electronics 4.3.1 4.3.2 4.3.3 4.3.4

4.4

Transmission selection Gear step selection Automatic transmission architectures 4.1.3.1 Simpson type 4.1.3.2 Wilson type 4.1.3.3 Lepelletier type 4.1.3.4 Summary of transmission types

Communication protocol: CAN Power and data networks Future communications: TTCAN Future communications: FlexRay Competing future communications protocols Diagnostic test codes (DTC)

Supporting subsystems 4.7.1 4.7.2

Steering systems Braking systems

ix 141 141 141 142 142 143 143 145 147 149 149 151 152 154 154 155 155 158 159 162 164 168 170 173 175 176 180 183 184 194 196 196 198 199 204 208 208 209 212 215 215 216 219 220 222 223 226 227 228 228 229

x

Propulsion systems for hybrid vehicles 4.7.3 4.7.4 4.7.5

4.8

Cabin climate control Thermal management Human–machine interface

Cost and weight budgeting 4.8.1 4.8.2

Cost analysis Weight tally

4.9 Exercises References 5 Electric drive system technologies 5.1 Permanent magnets 5.1.1 5.1.2 5.1.3

5.2

Brushless machines 5.2.1 5.2.2 5.2.3 5.2.4

5.3

Classical induction Winding reconfiguration Pole changing 5.4.3.1 Hunt winding 5.4.3.2 Electronic pole change 5.4.3.3 Pole–phase modulation 5.4.3.4 Pole changing PM

Variable reluctance machine 5.5.1 5.5.2 5.5.3

5.6

Buried magnet Flux squeeze Mechanical field weakening Multilayer designs

Asynchronous machines 5.4.1 5.4.2 5.4.3

5.5

Brushless dc Brushless ac Design essentials of the SPM Dual mode inverter

Interior permanent magnet 5.3.1 5.3.2 5.3.3 5.3.4

5.4

Permanent magnets: A primer What happened to Alnico? Rare earth permanent magnets

Switched reluctance Synchronous reluctance Radial laminated structures

Relative merits of electric machine technologies 5.6.1 5.6.2 5.6.3

Dynamic performance comparisons Comparisons for electric vehicles Comparisons for hybrid vehicles

5.7 Exercises References 6 Power electronics for ac drives 6.1 Semiconductor device technologies 6.1.1 6.1.2

6.2 6.3 6.4

Trends in power semiconductors Wide bandgap devices

Essentials of pulse width modulation Resonant pulse modulation Space vector PWM

230 230 233 234 234 236 237 238 243 243 244 247 248 252 256 261 265 274 277 278 282 287 289 290 290 293 294 295 296 298 304 306 307 311 313 313 313 315 316 319 320 325 326 327 328 330 335 337

Contents 6.5 Multilevel inverters 6.6 Comparison of PWM techniques 6.7 dc/dc converters 6.8 Thermal design 6.9 Reliability considerations 6.10 Sensors for current regulators 6.11 Interleaved PWM for minimum ripple 6.12 Exercises References

xi 346 348 349 351 356 359 361 363 365

7 Drive system control 7.1 Essentials of field oriented control 7.2 Dynamics of field oriented control 7.3 Sensorless control 7.4 Efficiency optimization 7.5 Direct torque control 7.6 Exercises References

367 368 373 379 384 388 391 391

8 Drive system efficiency 8.1 Traction motor

395 395 398 403 406 406 408 409 410 412 412 416 417

8.1.1 8.1.2

8.2

Core losses Copper losses and skin effects

Inverter 8.2.1 8.2.2 8.2.3

Conduction Switching Reverse recovery

8.3 Distribution system 8.4 Energy storage system 8.5 Efficiency mapping 8.6 Exercises References 9 Hybrid vehicle characterization 9.1 City cycle 9.2 Highway cycle 9.3 Combined cycle 9.4 European NEDC 9.5 Japan 10–15 mode 9.6 Regulated cycle for hybrids 9.7 Exercises References 10 Energy storage technologies 10.1 Battery systems 10.1.1

Lead–acid

419 427 428 429 430 432 433 435 438 439 441 447

xii

Propulsion systems for hybrid vehicles 10.1.2 10.1.3

10.2

Nickel-metal hydride Lithium ion

Capacitor systems 10.2.1 10.2.2 10.2.3 10.2.4 10.2.5

10.3

Symmetrical ultra-capacitors Asymmetrical ultra-capacitors Ultra-capacitors combined with batteries Hybridized battery example Ultra-capacitor cell balancing 10.2.5.1 Dissipative cell equalization 10.2.5.2 Non-dissipative cell equalization 10.2.5.3 Electrochemical double layer capacitor specification and test

Hydrogen storage 10.3.1 10.3.2

10.4 10.5 10.6

Metal hydride High pressure gas

Flywheel systems Pneumatic systems Storage system modelling 10.6.1 10.6.2 10.6.3

Battery model Fuel cell model Ultra-capacitor model

10.7 Exercises References

449 454 461 466 470 473 481 482 484 485 488 493 495 496 496 499 499 499 504 507 516 519

11 Hybrid vehicle test and validation 11.1 Vehicle coast down procedure 11.2 Sports utility vehicle test 11.3 Sports utility vehicle plus trailer test 11.4 Class-8 tractor test 11.5 Class-8 tractor plus trailer test 11.6 Exercises References

523 525 527 529 532 535 539 541

12 Automated electrified transportation 12.1 Personal rapid transit 12.2 Automated highway system 12.3 Non-contacting power transfer

543 546 547 549 550 552 553 561 561

12.3.1 12.3.2

Inductive coupling technology Radiated near-field power transfer

12.4 Transporting cargo 12.5 Exercises References Appendix A

563

Index

567

Preface

This second edition of Propulsion Systems for Hybrid Vehicles represents a major revision to the earlier book first published in 2004. A lot has changed over the past seven years and a good deal of that progress has been captured in this work. There is also significant structure change to the chapters with the inclusion of worked examples in all the chapters to further clarify and expand on the material presented. Exercise problems are provided at the end of each chapter that are intended to solidify the materials covered. Answers are provided to these exercises so that the approach taken can be validated. In this book attention is focused on hybrid technologies that are combined with gasoline internal combustion engines, or spark ignited, SI, engines as they are known. When an SI engine is direct injected it is more appropriate to refer to it as spark ignited direct injected, SIDI. The SIDI is being replaced by a new acronym, GDI-gasoline direct injection. Hybrid compression ignition direct injection, CIDI, engines operating on diesel fuel have been demonstrated, but the efficiency gained by adding electric fraction will be modest since the diesel is already a very efficient energy converter. When SIDI and CIDI engines are downsized and hybridized the effect is that of electric supercharging as Toyota Motor Co. once referred to this electrification process. Electrification brings with it energy conversion in the form of electric machines and power conversion in the form of power electronic converters. Both of these topics were covered in the 1st edition and have been updated in this 2nd edition. Electric drive system control, energy management and electric energy storage comprise the remainder of the electrification chapters in this book. The topic is then completed by treatment of standard drive cycles, emissions considerations and efficiency. Finally, the book concludes with an all new chapter on AET – automated electrified transportation, what many view as the future of hybrid vehicle evolution. As a text to supplement an existing automotive course at college senior level or graduate level it is recommended that chapters 1 through 4, along with portions of chapters 8, 9 and 11 be used. For more in depth treatment of hybrid and plug-in vehicles, battery electric vehicles and fuel cell vehicles the materials in chapters 5, 6, 7, 8 and 9 are most relevant along with chapter 12 for a vision of the role of these systems in future transportation systems. Chapter 12 on automated electrified transportation has been included in the 2nd edition to introduce this topic that is now receiving more interest, especially for the transporation of freight. The material in this book is recommended primarily for practicing engineers in industrial, commercial, academic, and government settings. As already said, it can

xiv

Propulsion systems for hybrid vehicles

be used to complement existing texts for a graduate or senior-level undergraduate course on automotive electronics and transportation systems. Depending on the background of the practicing engineers or university students, the material contained in this book may be selected to suit specific applications or interests. In more formal settings, and in particular where different disciplines such as electrical and mechanical engineering students are combined, course instructors are encouraged to focus on material presented in Chapters 2, 3 and 4. Instructors also have the flexibility to choose the material in any order for their lectures. Most of the material in this book has been developed by the author during active projects and presentations at conferences, symposia, workshops, and invited lectures to various companies and universities. I wish to acknowledge my wife JoAnn for her encouragement and assistance in the preparation of this second edition. Also, to my many good friends in the automobile and supplier industry to whom I owe so much for the opportunity to educate myself on hybrid propulsion technology. My special thanks to Ms. Lisa Reading, and the staff at the Institution of Engineering and Technology, Stevenage, UK, and S. Ramya at MPS Ltd, a Macmillan Company for all their help on making the publication of this book possible. John M. Miller March 2010

Chapter 1

Hybrid vehicles

When the first edition of this book was written in 2003, the following statement was made: ‘More visionary companies see hybrid vehicles as viable long term environmental solutions during the period when ICEs1 evolve to cleaner and more efficient power plants.’ At the time of writing, some six years later, this statement has proven correct as the following citations will illustrate. In Reference 1, Rishi et al. state, ‘A new era is rapidly approaching in which the very definition of personal mobility will change. Multi-modal transportation will become increasingly common, and intelligent vehicles will cater to diverse consumer needs for information, environmental responsibility and safety.’ This was followed by an even more provocative statement that the next 10 years will see more change than the previous 50 years. The pace of change is accelerating. This in fact is the motivation for releasing this second edition to Propulsion Systems for Hybrid Vehicles. On the topic of change and hybridized vehicles, Raskin and Shah [2] state, ‘The world is on the cusp of a major transition to hybrid power vehicles, which use highly efficient electric motors to boost the fuel efficiency of vehicles powered by internal combustion engines.’ They also qualified this statement somewhat by stating, ‘Over the last 30 years, many industries have either dramatically improved their energy efficiency or shifted to alternative fuel sources; transportation has been an exception.’ This last comment is sobering because it illustrates just how reluctant the automotive industry had been to adopt hybridization. Until recently that is. Climate change has become a very forceful motivator for energy efficiency and emission reductions. Informed persons across the globe are now well aware of the findings of the Intergovernmental Panel on Climate Change (IPCC), especially its finding that severe climate change is primarily the result of anthropogenic emission of greenhouse gases (GHGs), principally carbon dioxide (CO2). Severe climate change exemplars include melting glaciers across the globe, loss of polar ice much faster than previously thought and a host of other factors. The IPCC estimates that it is necessary to reduce global CO2 emissions by up to 85%, relative to year 2000, by 2050 in order to limit global temperature rise to less than 2.4  C.

1

Internal combustion engine.

2

Propulsion systems for hybrid vehicles

Holding to this temperature rise will require reducing the rate of CO2 accumulation from 380 ppm in 2005 to less than 450 ppm by 2050 [3]. As we proceed through this introductory chapter much of the material from the first edition will be retained, but augmented with new material to highlight the progress made over the past six years in hybrid electric vehicles (HEVs). In the first edition, it was noted that all the major automotive manufacturers had announced plans to introduce HEVs. Now they have and are moving into more electric vehicles (MEVs) such as plug-in and battery electrics. Technology leadership in hybrid technology continues to be dominated by the Japanese. According to the US National Research Council [4], North America ranks nearly last in all areas of hybrid propulsion and its supporting technologies. Table 1.1, extracted from Reference 4, is a condensed summary of their rankings. Table 1.1 Advanced automotive technologies supporting hybrid propulsion ranked by geographical region Technology

North America

Europe

Asia-Pacific

Internal combustion engine: compression ignited direct injection (CIDI) Internal combustion engine: spark ignited Gas turbine Fuel cells* Flywheel Advanced battery Ultra-capacitor Lightweight materials

3

1

2

2 1 2 1 1 3 2

2 1 2 1 2 3 1

1 1 1 3 1 1 1

*Author’s assessment

North America ranks high in energy storage technologies primarily because of developments by the National Laboratories for application to spacecraft use and by the US Advanced Battery Consortium (US ABC). Past introductions of gasoline–electric hybrid concept vehicles [5] and announcements of production plans show that most of the major global automotive manufacturers have plans to introduce hybrids between 2003 and 2007. Many of these introductions will be first generation hybrid propulsion technologies, and in the case of Toyota Motor Co. (TMC) their third generation of products and in 2010 a fourth generation. The prevailing system voltage for hybrid electric personal transportation vehicles is 300 V nominal. The 300 V level is a de jure standard adhered to by most manufacturers because it offers efficient power delivery in the automobile for power levels up to 100 kW or more while meeting the constraints of power electronic device technology (currently 600 V) and electrolytic bus capacitor ratings (450 V). TMC has deviated from this system voltage level in their announcement for the Lexus RX330, Hybrid Synergy Drive (HSD) concept vehicle. Figure 1.1 shows the display model that is said to deliver V8 performance with

Hybrid vehicles

3

Figure 1.1 TMC’s Hybrid Synergy Drive concept vehicle [5] (left) and 2010 Prius IV (right) a V6 power plant using nickel-metal hydride (NiMH) battery technology and a system voltage of 550 V. The HSD represents TMC’s third generation of hybrid technology after their Toyota Hybrid System THS 1 and 2, plus the THS-C for continuously variable transmission (CVT) architecture. The previous generation THS-C on display at the 2002 Electric Vehicle Symposium is shown in Figure 1.2 and consists of a small 4-cylinder ICE and twin electric motors integrated into the transmission. The NiMH battery pack is to the right in the picture.

Figure 1.2 CVT hybrid powertrain (THS-C) [5] The orange coloured (bundle of 3 cables and bundle of 2 cables) high voltage distribution and motor harness wire shown in Figure 1.2 are consistent with industry cable identification and markings. A power electronic centre is mounted above the transmission. The power electronic centre receives dc power from the battery pack via the two cables and processes this into ac power for both

4

Propulsion systems for hybrid vehicles

motor–generators (M/Gs) required by the CVT transmission (shown front centre and centre right). Table 1.2 is a fact sheet on the Toyota Prius, the hybrid vehicle introduced into mass production in Japan in 1997 and into the North American market in 2000. Prius implements the THS first and second generation hybrid propulsion systems. Table 1.2 Toyota 2002 Prius fact sheet Features and benefits Warranty Mechanical specifications

Battery pack Brakes

Incentives

THS hybrid system: improved fuel economy and range. Reduced emissions. Seamless operation and no change in driving habits necessary. Basic: 3 year/36,000 mi Drivetrain: 5 year/60,000 mi THS M/G and battery pack: 8 year/100,000 mi Engine: 1.5 L, I4 Atkinson Cycle, DOHC 16 valve with VVT-I, rated 75 hp at 4,500 rpm and 82 ft-lbs torque at 4,200 rpm M/G: Permanent magnet synchronous (interior magnet), 44 hp at 1,040–5,600 rpm, 258 ft-lb torque at 5,600 rpm Drivetrain: front wheel drive with THS power split transmission Curb weight: 2,765 lb Fuel tank: 11.9 US gallons Nickel-metal hydride, NiMH, 35.500 W  1200 H  6.500 D Weight: 110 lb Voltage: 274 V Regenerative braking system (RBS). Captures up to 30% of energy normally lost to heat. M/G operates as a generator above speeds of 5 mph to replenish battery ABS supercedes THS regeneration Federal tax deduction of $2,000 Some states in the US permit single occupant HOV lane access with hybrid vehicles

Honda Motor Co. is aggressively introducing HEVs following the success of their insight with integrated motor assist (IMA). Honda has taken a different tact on hybrid propulsion than Toyota. Honda integrates a permanent magnet synchronous motor into the transmission. The IMA operates under torque control from stall to wide open throttle speed of the engine. This enables electric torque assist over the complete engine operating speed range. Figure 1.3 shows the Honda IMA system integrated into the powertrain. In Figure 1.3 the Honda IMA M/G, rated 10 kW, 144 V, is sandwiched between the inline 4-cylinder engine and the CVT transmission. The CVT belt is clearly visible in Figure 1.3. In particular, notice the presence of a ring gear to the immediate right of the IMA M/G. Honda continues to use the 12 V starter motor for key starts and only uses the IMA for warm restart in an idle–stop strategy. With this choice of architecture, the IMA is not required to meet cold cranking torque needs of the engine so that it can be designed to operate over the 6:1 torque augmentation speed range.

Hybrid vehicles

5

Figure 1.3 Honda Motor Co. integrated motor assist (IMA) [5] The IMA motor is unique in that it is a novel heteropolar permanent magnet synchronous machine having bobbin wound stator coils and surface inset magnet rotor. Today we know this electric machine as a concentrated fractional pitch winding machine. Figure 1.4 is a computer aided design graphic of the IMA system and cutaway of an actual 18 coil and 12 pole version. The IMA M/G is designed to provide 13.5 hp at 4,000 engine rpm to assist the 85 hp, 1.4 L, VTEC I4 engine, supported by a 144 V (120 cells), NiMH battery with an 8 year, 80,000 mile (mi) warranty.

Figure 1.4 Honda IMA synchronous motor–generator and in cut-away (right) Honda’s S2000 Roadster shown in Figure 1.5 (top) achieves more than 650 mi cruising range on a single fill up of 13 US gallons. The S2000 Roadster has a fuel economy of 51 mpg from its 1.3 L gasoline engine. It has room for five adults.

6

Propulsion systems for hybrid vehicles

Figure 1.5 Honda Motor Co. 2002 Civic Hybrid (top); 2010 Civic Hybrid (bottom) The Civic Hybrid with IMA (Figure 1.5, bottom) claims a 66% torque boost by the 144 V permanent magnet M/G. The Civic internal location of engine with IMA and hybrid traction battery is shown in Figure 1.6. A 144 V, 120 cell NiMH battery pack is located behind the rear passenger seat. Power distribution is via high voltage shielded cables from the trunk area to the power electronic centre under-hood. A simple charge or assist indicator is included into the instrument cluster to inform the driver of IMA performance. An indicator lamp is used to signal idle–stop function. There is no SOC indication

Figure 1.6 Honda Civic Hybrid battery location and instrument panel layout

Hybrid vehicles

7

on the battery. To date, SOC algorithms are unreliable and prone to misjudge battery available energy due to charge/discharge history and ageing effects. Table 1.3 is a side-by-side comparison of the 2003 model year Toyota Prius and Honda Civic hybrid vehicles. Table 1.3 Comparison of Prius and Civic hybrids (MY2003) Comparison

Honda Civic Sedan

Toyota Prius Sedan

Base price (MSRP) Fuel economy – city Fuel economy – highway Warranty: Powertrain months Powertrain miles Engine # cylinders Driveline Engine displacement (cc) Valve configuration Engine horsepower at rpm Engine torque at rpm Fuel system Brakes – front – rear Steering Climate control Curb weight (lb) Passenger compartment volume (ft3) Cargo volume (ft3) Headroom (in) Wheels and tyres

$19,550 46 51 36 36,000 4 Front wheel drive 1,339 SOHC 85 at 5,700 87 at 3,300 Multipoint injected Disc Drum Rack and pinion Standard 2,643 91.4 10.1 39.8 1400 alloy 70R14

$19,995 52 45 96 100,000 4 Front wheel drive 1,497 DOHC 70 at 4,500 82 at 4,200 Multipoint injected Disc Drum Rack and pinion Standard 2,765 88.6 11.8 38.8 1400 alloy 65R14

The comparisons in Tables 1.3 and 1.4 are interesting because they show how similar the two vehicles are in style, occupant room and powertrain. It is also insightful to compare earlier versions of the Prius such as the Prius II in Table 1.2 with the new Prius IV in Table 1.4. As a general trend, the vehicle mass has gone up, interior volumes and headroom have somewhat reduced and power plant ratings have increased. This mass increase, however, has been compensated by higher power electronic integration, lower battery mass with improved cell and pack designs, higher speed electric machines, and hence higher specific power and more use of lightweight materials. New IMA has 50% more power and 14% higher torque at the same mass. The 2010 Civic Hybrid battery has 25% more specific power for same mass. It has recently been made public that the 2010/2011 Civic Hybrid will introduce a lithium ion pack (LNMCO cathode, 3.7 V, 6.0 Ah) battery. This pack will be manufactured for Honda by Blue Energy Corp., a 51:49 joint venture between GS Yuasa and Honda. Production volumes target 100,000 units per year (APV). North American automobile manufacturers are beginning to build their hybrid portfolios with product offerings targeting sport utility vehicles (SUVs) and

8

Propulsion systems for hybrid vehicles

Table 1.4 Comparison of Prius and Civic hybrids (MY2010) Comparison

Honda Civic Sedan

Toyota Prius Sedan

Base price (MSRP) Fuel economy – city Fuel economy – highway Warranty: Powertrain months Powertrain miles Engine # cylinders Driveline Engine displacement (cc) Valve configuration Engine horsepower at rpm Engine torque (Nm) atrpm Electric motor

$26,850 40 45 36 36,000 4 Front wheel drive 1,339 SOHC 110 at 6,000 123 at 1,000 ? 2,500 15 kW (20 hp) 103 Nm

Battery Hybrid warranty CVT ratio drive/(Rev)

NiMH 158 V, 5.5 Ah, 8 year/100 000 mi 2.526 ? 0.421 (4.511 ? 1.875) 4.94 Multipoint injected AT PZEV Disc Drum Rack and pinion Standard 2,877 90.9

$26,200 51 48 60 60,000 4 Front wheel drive 1,798 DOHC 98 at 5,200 144 at 4,000 650 Vmax, 60 kW (80 hp) 207 Nm NiMH 201.6 V, 6.5 Ah, 44 kg 8 year/100 000 mi eCVT k = 2.60 4.11 Multipoint injected AT PZEV Disc Drum Rack and pinion Standard 3,042 93.7

10.4 39.4 1500 alloy P195/65R15 89S 12.3

21.6 38.6 1500 alloy P195/65R15 11.9

Final drive Fuel system Emissions Brakes – front – rear Steering Climate control Curb weight (lbs) Passenger compartment volume (ft3) Cargo volume (ft3) Headroom (in) Wheels and tyres Fuel tank capacity (gallons)

pick-up trucks. Ford Motor Co. announced its hybrid Escape SUV at the 2000 Los Angeles auto show. The hybrid Escape powertrain, derived from earlier work by Volvo Car Company and Aisen Warner transmissions [6], requires an electric M/G and a starter–alternator (S/A). The powertrain on the hybrid Escape is a version of power split similar to that employed in the Toyota Prius. Figure 1.7 shows the hybrid Escape SUV that went into mass production in mid-2004 and is now in its second generation, and the Mariner hybrid was introduced in 2005. The Escape claims a fuel economy of 40 mpg city driving from its 2.3 L Atkinson I4 engine augmented with a 65 kW M/G and 28 kW S/A powertrain. Some of the fuel economy will be evident on highway driving because the Atkinson cycle (late intake valve opening) delivers approximately 10% higher economy than normally aspirated ICEs. Operating from a 275 V NiMH advanced battery pack capable

Hybrid vehicles

9

Figure 1.7 Ford Motor Co. MY2004 hybrid Escape SUV and MY2010 Mariner hybrid SUV of 27 kW power transfer, the hybrid Escape and Mariner deliver the same performance as their conventional vehicle (CV) sister with a 200 hp V6 power plant. General Motors Corp. made the most sweeping announcement of vehicle hybrid powertrain line-ups than any other major automotive company [5]. In addition to its already unveiled Silverado pick-up truck with a 42 V crankshaft mounted ISG, the company migrated this technology cross-segment to its Sierra pick-ups during 2003. Initial offerings of the Silverado 42 V ISG will be as customer options. Figure 1.8 is an illustration and cut-away of the crankshaft ISG manufactured by Continental Group for GM for use on their Silverado pick-up truck.

Figure 1.8 General Motors Corp. crankshaft ISG used in the Silverado pick-up [5]

10

Propulsion systems for hybrid vehicles

Following this product introduction, the company introduced a hybrid SUV, the Chevrolet Equinox, in 2006. The Equinox is equipped with a CVT transmission so the system may be similar to Toyota’s THS-C. GM also introduced hybrid passenger vehicles beginning in 2007 with its hybrid Chevy Malibu. Also in 2007, GM introduced the first hybrid full size SUVs – a hybrid Tahoe and Yukon. Both of these vehicles are in the 5,200 pound class so the ac drives will most likely be at 100 kW plus power levels. Figure 1.9 is the ParadiGM hybrid propulsion system concept that uses twin electric machines in an architecture that permits power split like performance yet accommodates electric drive air conditioning with one of the M/Gs when the vehicle is at rest in idle–stop mode. Typically, cabin climate control in summer months in hybrid passenger vehicles is not available unless the engine is running. The ParadiGM system changes that constraint by making dual use of one of the M/Gs in much the same way that Toyota does on its THS-M class of hybrids (M for mild as in 42 V hybrid).

Figure 1.9 General Motors ParadiGM hybrid propulsion system and 2-mode eCVT [4] Nissan Motor Co. introduced their capacitor hybrid truck, a 4 ton commercial delivery vehicle based on a parallel diesel–electric hybrid propulsion system [7]. A prototype of the capacitor hybrid was designed in 2000 that used an I4, 4.6 L CIDI with purely hydraulic valve actuation running the Miller cycle. The engine produced 55 kW and was used to drive a 51 kW permanent magnet synchronous generator. Propulsion was provided by twin 75 kW synchronous motors. The ultracapacitor pack was rated 1,310 Wh and weighted 194 kg. In a more recent incarnation, the Condor capacitor hybrid truck is derived from that prototype but uses a 346 V, 583 Wh, 60 kW ultra-capacitor built in-house at Nissan’s Ageo factory. The Condor capacitor hybrid is designed to meet the demands of in-city deliver routes of up to 2.4M cycles of braking and stop–go traffic during its expected 600,000 km lifetime. Testing validates a 50% improvement in fuel consumption and reduction in CO by 33%. Fuel cell hybrids are now available in limited production quantities. Honda Motor Co. has begun selling fuel cell electric vehicles (FCEVs) for city use in Los Angeles where 5 vehicles were delivered in 2002 and 30 more in the next 5 years.

Hybrid vehicles

11

The fuel cell experimental vehicle (FCX), shown in Figure 1.10, is similar in appearance to a conventional minivan but that is where any further similarity ends.

Figure 1.10 Honda Motor Co. fuel cell hybrid, FCX A detailed discussion of the Honda fuel cell hybrid is presented in Chapter 10. When the body skin is removed from an FCEV, there are virtually no moving parts. Under-hood layout consists of air induction and compression for the fuel cell stack, thermal management for the fuel cell stack (and water management) as well as cabin climate control functions. There is a conventional radiator, electric drive pumps and fans. Beneath the floor pan resides a 78 kW Ballard Power Systems’ fuel cell stack. Compressed gas hydrogen storage cylinders with a capacity of 156 L are located behind the rear passenger seat. Steering is electric assist, brakes are regenerative with ABS override and suspension is standard with an integrated shock in strut. With this brief introduction of hybrid vehicles, which are either now available in the marketplace or soon will be, we start our discussion of understanding the basics of vehicle propulsion and target setting. Chapter 2 will then take a more detailed look at hybrid propulsion architectures. Later chapters will develop the details of ac drives necessary for an understanding of hybrid propulsion and its attendant energy storage systems.

1.1 Electric engine hybrids 2010 To expound on this topic, it is insightful to list some statistics from EDTA2 on what our current situation is with imported oil. 2

Electric Drive Transportation Association, www.electricdrive.org

12 ●









Propulsion systems for hybrid vehicles Transportation accounts for 50% of US urban air pollution and 33% of US GHG emissions. Today 50% of all Americans live in areas that don’t meet air quality standards. Global oil consumption is expected to reach 119M bbl a day in 2025, 60% of which will be supplied by the Organization of the Petroleum Exporting Countries (OPEC). The Energy Information Administration (EIA) estimates that imports will constitute 61% of US liquid fuel demand in 2030. Electric drive vehicles can reduce petroleum consumption by 40% and emissions by 50%. Hybrid electric transit buses see reductions in particulate matter by nearly 90% and nitrous oxides by up to 50%.

Vehicle electrification is now a key aspect of virtually all automotive and heavy transportation manufacturers’ product development and marketing plans. Table 1.5 highlights this electrification path for the automotive case [8]. Many studies put the sort of engine actions and vehicle electrification activities listed in Table 1.5 into the perspective of installation cost, on-cost, versus their CO2 reduction performance. Figure 1.11 illustrates this for the case of engine actions designed to improve efficiency through higher thermodynamic efficiency and emission reductions through more complete combustion process and after treatment as needed. Figure 1.12 extends this same format to the case of vehicle electrification and notes, in particular, the existing case of hybridization as one path to electrification and a second, much higher opportunity, for the case of MEVs, including plug-in electric vehicle (PHEV), range extended electric vehicle (REV) and battery electric vehicle (BEV). In Figure 1.11 the engine optimization technologies include electric spin-up turbo charger, cylinder deactivation or variable compression ratio, GDI and electromechanical valve actuation (EVA), now being approximated by variable valve timing and lift (VVTL).

1.2 Limits of engine-only actions As Figure 1.11 shows, the low-hanging technology fruits are already being harvested in the $50/%CO2 to $100/%CO2 range. Beyond this, more advanced technologies such as HCCI that blurs the distinction between Otto cycle and diesel cycle operation are being pursued. As Figure 1.12 demonstrates, the capability to push deeper into CO2 reduction requires higher levels of vehicle electrification and the introduction of MEVs: PHEV, REV and BEV. Coinciding with engine actions to lower CO2 emissions are the quest for cleaner fuels, including ●



Ethanol, bio-diesel, cellulosic diesel and ethanol, jatropha-derived diesel and other crop-to-wheels fuels. Electricity and hydrogen top the list of energy carriers.











More complete combustion Thermodynamic efficiency

CNG Bio-diesel Hydrogen

Advanced ICE













Downsize engine GDI HCCI

Opportunity charging Idle–stop Enhanced starter motor

Recuperator and micro hybrid

















Downsize engine GDI HCCI eTurbo

Stop–start Boosting Regen Crankshaft ISG

Mild hybrid















Downsize engine GDI HCCI eTurbo

Torque assist Limited AER New transmission

Strong hybrid (HEV)













Clean fuels GDI HCCI eTurbo

Larger battery 10–20 mi AER

Plug-in hybrid (PHEV)









Clean fuels Minimize brake loss

AER of 40 mi or more Fast charge

Range extended vehicle (REV)







Efficient thermal management

Energy-optimized battery AER over 100 mi

Battery electric vehicle (BEV)

CNG, compressed natural gas; AER, all electric range; GDI, gas direct injection; HCCI, homogeneous charge compression ignition (also, CAI – controlled autoignition); eTurbo, electric spin-up turbo charger

CO2 reduction opportunity

Technology

Automotive electrification path

Table 1.5 Vehicle electrification roadmap

14

Propulsion systems for hybrid vehicles $200

$250

% CO2

% CO2

Engine actions to reduce emissions

$150 % CO2

2,500 ON cost ($)

$100 % CO2

2,000

HCCI

1,500 $50

ICE optimization e-Turbo GDI

1,000 500

Thermal 0

2

% CO2

Cyl-deact EVA 4

6

8

10

12

14

16

18

20

CO2 reduction (%)

Figure 1.11 Engine optimization actions to reduce CO2 emissions Vehicle electrification to reduce emissions

ic ve hic le

$300 PHEV - 2010

ctr

15,000

$250

REV

%

ON cost ($)

M

ore

ele

12,000

% BEV

$200 %

9,000 $150

PHEV 2020

6,000 on if icatiStrong HEV

Electr

3,000 1,500 Engine opt

0

5

CNG

%

Adv diesel

$50

Mild HEV

HCCI Micro HEV

10

% $100

15

20

% 25

30

35

40

45

50

CO2 reduction (%)

Figure 1.12 Vehicle electrification actions to reduce CO2 emissions However, there are pressing issues of storage and infrastructure for all these alternative fuels and especially the energy carriers. The overarching need then is how to supply energy for 1010 people in 2050. Hydrogen generation, distribution and storage remain questionable. The better option remains electricity generation, distribution and storage, along with a better battery. Example 1: Compute the power plant CO2 emissions that are the result of a 40 mi daily commute in a PHEV vehicle with a specific consumption of 250 Wh/mi and 100% driveline efficiency. Assume transmission and distribution (T&D) losses at

Hybrid vehicles

15

10% for electricity charging and a conventional coal fired plant that emits 950 g CO2/kWh produced.

PHEV at 4 mi/kWh

Solution: The total commute of the PHEV at standard consumption and assuming that it operates in only charge depleting (CD) mode implies that it will take 10 kWh. This amount of consumed energy translates back to the generating station as 10 kWh/0.9 = 11.11 kWh. At the stated power plant’s emission level, this results in 950 g CO2/kWh  11.11 kWh = 10.5 kg CO2.

1.3 Vehicle electrification and more electric vehicle The electrification actions discussed in Section 1.1 transform the automobile from its historical heat engine power plant, transmission, driveline and axles into dual power plant in the case of hybrids to electric-only power plant in case

6,000 Total Petroleum Coal Natural gas Cement production

5,000 4,000 3,000 2,000 1,000

1800

1850

1900

1950

Figure 1.13 Total CO2 emissions

2000

Million metric tons of carbon/year

7,000

16

Propulsion systems for hybrid vehicles

of BEV. In this section the progression from CV to BEV will be investigated in more detail. Reducing CO2 is paramount to any other innovation in the automobile that an environmentally conscious manufacturer could pursue. Figure 1.13 from EIA, the energy information agency of the US government, shows the components of annual CO2 emissions and some of the key contributors. What is revealing in this chart is that petroleum (mainly transportation) and coal (mainly electricity generation) are so close in recent times for annualized CO2 emissions. Sadly, petroleum now outpaces coal in emissions. Getting back to vehicle electrification actions, Figure 1.14 traces the energy flow from 100 units of energy in the vehicle fuel tank through the engine, transmission and driveline to the wheels. At the wheels the remaining energy must overcome aerodynamic drag, tyre and driveline rolling resistance, inertial effects and braking energy loss (which a hybrid can partially recuperate). In the hybrid vehicle, aerodynamic and rolling resistance losses remain unrecoverable losses. Engine thermal losses 62

Engine mechanical and pumping losses 12

Driveline losses 3

Engine Fuel gasoline, diesel, methanol, CNG

Fuel - 100 Start with 100 ‘units’ of gasoline fuel energy

Transmission

Transmission losses 9 Accessories 1

Kinetic energy braking losses 8 Rolling resistance 3 Aerodynamic losses 2

Figure 1.14 Energy flows in the conventional vehicle Thermal management in the automobile is crucial because out of the 62 units lost in the ICE approximately 60% are rejected to the radiator and 40% to the exhaust. This is important for energy harvesting where thermoelectric or bottom cycle conversion can be used to recuperate some of this energy. Transmission losses are mainly torque converter in the automatic transmission (AT) plus gear mesh losses (~1.5%/mesh/stage), plus bearing and clutch friction losses. Out of the 100 units energy input to the CV, only 13 units are available at the wheels. For BEV the energy flow diagram is far different, as shown in Figure 1.15. Here, the energy originates at the charging plug (see Example 1) and is stored in the vehicle on-board energy storage system, an electrochemical battery of an appropriate technology. The energy available at the wheels of the BEV is now 73 units out of a stored 100 units. But caution is in order regarding the energy storage component as the following example will illustrate. Currently, there is considerable

Hybrid vehicles Ess one way 6 Energy storage system

Converter loss 6 Utility

On-board charger

100 Units electricity

Conversion loss 3

Power converter

M/G

Transmission

Veh. sys.cnt’l

Driveline loss 5

Traction motor loss 7

Units loss 27

17

Kinetic and braking including regeneration 4 FD

Rolling resistance 22 Aerodynamic at 45mph average 47

Units to wheels 73

Figure 1.15 Energy flow in the BEV activity on clarification of electric energy usage by plug-in and BEV and how fuel economy is to be computed when a portion of the driving is CD.3 Example 2: Compute the regenerative energy storage system (RESS) total losses over the SAE J1634 drive cycle for a Camry Hybrid vehicle without accessories (no air conditioning, lights or fans) for the case of an NiMH battery versus a lithium ion battery (lithium–nickel–manganese–cobalt oxide type) (Figure 1.16). The NiMH pack is rated 244.8 V, 6.5 Ah, 1.59 kWh and will be characterized at a nominal 60% SOC. The lithium ion battery has similar ratings 244.2 V, 6.5 Ah, 1.59 kWh and will be characterized at 60% SOC and is assumed to have a one-way efficiency of 93%. Both vehicles operate in CS mode and are exercised over a UDDS + UDDS + Hwy + Hwy drive cycle sequence according to SAE J1634.

(a)

(b)

Figure 1.16 (a) Camry Hybrid under-hood; (b) Panasonic battery pack (representative) Test results for Camry Hybrid with NiMH pack: Input charge (Ahin = 13.18 Ah); output charge (Ahout = 9.18 Ah) for a difference of 4 Ah. 3

M. Duoba, T. Bohn, E. Rask, SAE Surface Vehicle Recommended Practice: J1711, J2841 and J1634. Team led by M. Duoba at Argonne National Laboratory, Advanced Powertrain Research Facility, APRF.

18

Propulsion systems for hybrid vehicles

Solution:

Efficiency, one-way Efficiency, turnaround Discharged Ah Difference

NiMH battery

Li ion battery

(9.18/13.18) = 0.8346 hta = (how)2 = 0.83462 = 0.6965 9.18 (and = 0.6965  13.18) N/A

0.93 (given) hta = (0.93)2 = 0.865 11.4 (= 0.865  13.18) 21

where how = one-way efficiency and hta = turnaround efficiency. The lithium ion pack in the same application will be 21% more efficient in recuperating energy and making it available for reuse.

1.4 Performance characteristics of road vehicles The vehicle attributes foremost in customers’ minds when contemplating a purchase, other than cost and durability,4 are its performance and economy or P&E. Performance generally relates to acceleration times, passing manoeuvres and braking. Economy has metrics of fuel economy (North America) or fuel consumption (Europe and Asia-Pacific), as well as emission of GHGs.

1.4.1

Partnership for new generation of vehicle goals

It is prudent to start a discussion of hybrid vehicle P&E by stating the goals of the US Partnership for a New Generation Vehicle (PNGV) [9]. PNGV Goal 3 sets a vehicle mass target for a 5-passenger vehicle at less than 1,000 kg from its CV production mass of 1,472 kg. The vehicles targeted in North America were the GM Chevrolet Impala, Ford Taurus and Chrysler Concorde, which are all high volume, mid-sized passenger cars. Table 1.6 presents a summary of their performance targets. Table 1.6 Vehicle performance goals (US PNGV targets) – 5 passenger, 1,472 kg base, 26.7 mpg Vehicle attribute

Parameter

Acceleration Number of passengers Operating life Range Emissions Luggage capacity Recyclability Safety Utility, comfort, ride and handling Purchase price and operating costs

0–60 mph in 12 s Up to 6 total occupants >100,000 mi 380 mi on combined cycle Meet or exceed EPA Tier II 16.8 ft3, 91 kg Up to 80% Meets Federal Motor Vehicle Safety Standards Equivalent to conventional vehicle Equivalent to conventional vehicle, adjusted to present economics

4

Safety and security systems are foremost in consumers’ minds when technology is used as a product differentiator.

Hybrid vehicles

19

The interesting items in Table 1.6 are that customer expectations for price, operating costs and ride and handling must be comparable to a CV. In comparison the economy goals are very straightforward as noted in Table 1.7. Table 1.7 Vehicle economy goals (US PNGV Goal 3) Attribute

Baseline vehicle

Weight-reduced baseline vehicle

Hybrid vehicle

PNGV goal

Fuel economy (mpg) Fuel usage during lifetime of 150 000 mi (gallons) Lifetime fuel savings over baseline (gallons)

26.7 5,600

10. Power assist architectures do not have electric-only range capability, or if designed with more energy storage capacity the electric-only range may be just a few kilometres, perhaps as many as 7 km. These architectures typically have electric fractions of only 10–30% and a modestly downsized engine. In the US Department of Energy study by its Office of Automotive Transportation Technology, the power assist mode hybrids are said to cost incrementally more than a conventional vehicle. Figure 2.14 illustrates the cost components in conventional vehicles, power assist hybrids (HEV0) and dual mode hybrids 3

Power to energy ratio of a regenerative energy storage system (RESS) defines the capacity to sustain a high power pulse for time T ¼ (P/E) in (W*kg)/(Wh*kg) ¼ h 1, or convert to seconds, 3,600 s/h.

Hybrid architectures

Component retail price equivalent

$30,000 $25,000

85

Average of base and On vehicle charging system ANL methods Energy storage system Electric traction Accessory power Transmission Engine + exhaust Glider

$20,000 $15,000 $10,000 $5,000 $0

CV

HEV0

HEV20

HEV60

Figure 2.14 Major components in conventional vehicle, power assist and dual mode hybrids (mid-size vehicle, retail price equivalents; from Reference 16)

(HEV20 and HEV60), accounting for glider cost (the base vehicle shell including chassis), engine and exhaust system, transmission and accessories. The major differences between a power assist hybrid and a CV is the lower cost of powertrain components due to downsized engine and different transmission. However, total vehicle cost is higher in power assist because of the added electric traction and energy storage components. These same added components in dual mode vehicles are higher still due to their increased rating. The final distinction between dual mode and power assist is the additional on-board charger necessary in dual mode for utility charging.

2.2.5 Dual mode Dual mode is still a pre-transmission architecture but with a very capable ac drive system having electric fractions of greater than 30% and sufficient on-board energy storage for sustained electric-only range. The dual mode hybrid electric-only range can be 20, 40 or as high as 60 mi in NA. Because of the EV like energy storage system levels, the battery technology will generally have P/Es less than 10. Dual mode is a connotation for engine power only, electric propulsion power only or both. According to the OATT report, a dual mode vehicle will cost from $3,000 to $5,000 more than its CV counterpart as noted in Figure 2.14. The battery alone will represent a sizeable fraction of this cost as well as of the added mass. Battery warranty, due to high replacement cost, must be 10–15 years. The warranty should also be transferable to the second owner (typically past year 6 for the vehicle) in order for the vehicle to hold its residual value. High residual value is an investment benefit for the first owner and transferable battery warranty a benefit to the second owner. Figure 2.15 illustrates the cost increments if battery replacement is required.

86

Propulsion systems for hybrid vehicles

Vehicle retail price equivalent

$40,000 $35,000 $30,000

Base Base battery replacement ANL ANL battery replacement

$25,000 $20,000 $15,000 $10,000 $5,000 $0 CV

HEV0

HEV20

HEV60

Figure 2.15 Conventional and hybrid vehicle cost increment with and without battery replacement (from Reference 16 where ANL is Argonne National Laboratory model)

The initial on-cost would be higher based on retail price equivalents. The second column is based on data in Reference 16 taken by US Argonne National Laboratory cost methodology. Differences in battery costs are due to different models used to estimate costs of advanced battery chemistries such as NiMH and lithium ion. The batteries used in Figure 2.15 are not yet in mass production nor are their full manufacturing costs clear. The second difference in cost is due to the battery capacity involved. Data shown in Figure 2.15 assume a mass production cost for NiMH of $250/kWh. Furthermore, battery cost (in $/kWh) is estimated as being inversely proportional to specific energy density (in kWh/kg). The higher the specific energy content, the less the material consumed and the lower the total battery cost. Today, lithium ion battery for consumer electronics will cost approximately $0.50/Wh, whereas a large format lithium ion pack still costs $1.20/Wh when fully packaged. Cell level cost is approximately 55% of pack cost. A packaged lithium ion module today includes the cell compliment, interconnects, battery management electronics, thermal management and CAN communications. Only with more refined cell production and lower cost of materials, or new materials, will large format lithium ion packs be capable of realizing manufacturing costs of $340–200/kWh and lower.

2.3 Pre-transmission combined configurations Power split electronic continuously variable transmission (eCVT) is the architecture of choice for passenger sedans, light trucks and SUV hybrids because it offers CVT like performance. CVT like performance in a non-shifting, clutchless transmission is enabled through the control of two M/Gs. The concept of power

Hybrid architectures

87

split has been known since the early 1970s, particularly in the work by the TRW group [17]. In Reference 17, Gelb et al. describe a dual M/G architecture having electric machine functions of ‘speeder’ and ‘torquer’ in what was then called an electromechanical transmission. In this precursor to power split, the speeder M/G acted as a generator and the torquer M/G as a motor in the driveline. The engine crankshaft was connected to the sun gear of an epicyclic gear set. Input power to the epicyclic gear set is divided in direct proportion to the respective speeds of the sun, planets (carrier) and ring gears. The speeder M/G is connected to the carrier and the torquer is connected to the ring gear via an additional gear ratio. Figure 2.16 illustrates the mode of operation of the TRW electromechanical transmission.

FD

R ws Speeder M/G

C S

ICE

Torquer M/G

wT

3f

we

3f

Battery pack

Figure 2.16 TRW electromechanical transmission (precursor to power split, from Reference 17) The electromechanical transmission has five modes of operation associated with the engine and both speeder and torquer M/Gs (Figure 2.17). Mode 1. Low acceleration events for which engine power exceeds the road load and the remainder is used to charge the vehicle battery. Torquer and speeder act as generators sending excess engine power to the battery. Mode 2. Engine power equals road load demand but the engine has insufficient torque. Torquer acts as a motor. The speeder M/G accepts excess engine power and transfers this power to the torquer and to the battery. Mode 3. Road load torque and power exceed the available engine torque and power. In this mode the battery delivers peaking power to the speeder and torquer combination. Mode 4. Higher speed cruising, the scenario in Mode 3 changes and shifts to Mode 2 and the speeder is taken out of the loop (locked) and the engine throttled up. The torquer absorbs or delivers power to the battery.

Propulsion systems for hybrid vehicles

Vehicle accel (m/s2)

88

Mode 3 Mode 2 Mode 1

Mode 4

Vehicle speed (mph)

Mode 5

Figure 2.17 Operational map of electromechanical transmission Mode 5. All deceleration events are used to replenish the vehicle battery by regenerating in either Mode 1 or Mode 2 depending on vehicle speed. Both speeder and torquer M/Gs act as generators. The modern incarnation of the original TRW electromechanical transmission system is available to the public in the Toyota Prius hybrid.

2.3.1

Power split

Power split is a dual M/G architecture depicted functionally in Figure 2.18. The basic functionality is that of engine output shaft connected to the carrier of an epicyclic gear set. Mechanical power is transferred to the driveline from the engine through the planetary gear set via the ring gear to the final drive and to the wheels. To effect this mechanical path, an electric path is split off via the MG1, or S/A, operating as a generator so that its reaction torque is developed against the carrier.

Wheels

3f

3f

M/G

R

ICE

C

S/A

S

Gearbox

FD

Battery pack

Figure 2.18 Power split single mode eCVT functional architecture

Hybrid architectures

89

Electric power from the S/A is then routed to the dc bus and consumed by the main M/G for propulsion or sent to the battery. The M/G is the main traction motor used for propulsion and for regenerative braking. There are no driveline clutches in a power split propulsion system, only indirect mechanical paths from the engine to the driven wheels. Figure 2.18 illustrates the basic functionality of power split architecture. In Figure 2.18 it can be seen that the M/G has its rotor connected to the planetary set ring gear and from there directly to the wheels via gearing and the final drive. In the actual implementation there is a chain drive link between the ring gear and the input to the fixed ratio gearbox. The electric fraction of power split is determined by the peak power rating of ICE and M/G. The S/A is a torque reaction source necessary to hold the ICE speed within a confined range, via its sun to carrier gear ratio. The benefit of constraining engine speed to more restricted ranges within fuel islands was described in Chapter 1. In power split operation, engine torque is delivered to the wheels for propulsion by first splitting off a portion and converting it to electricity. This diverted power is then recombined with engine mechanical power at the planetary set ring gear. In the process of power splitting, the engine speed becomes decoupled from vehicle speed through the action of MG1. This balancing act is best described using stick diagrams as shown in Figure 2.19.

Engine speed

Accelerat

ion

Vehicle speed

S/A speed

Engin

e cran

king

Cruise S

C

R

unch

le la

ic Veh

Figure 2.19 Power split ‘stick’ diagram of speed constraints Figure 2.19 is a static illustration of power split operation at a static operating point. Before describing the dynamic operation we examine the kinematics of the epicyclic gear set to illustrate the power splitting function. The epicyclic gear set is illustrated in Figure 2.20 showing its respective components and speeds. Torques at each gear are in proportion to the respective speeds and the power being transmitted.

90

Propulsion systems for hybrid vehicles R C S ws

wr

wc

(a)

(b)

Figure 2.20 Epicyclic gear set and definitions: (a) definition of planetary gear components and (b) illustration of planetary gear set (from UQM) A planetary gear set is composed of sun gear at the centre of the diagram, a set of pinion gears (planets) arranged around the sun and held by the carrier and the ring gear. Bearings support the three sets of gears. The basic ratio, k, of an epicyclic gear set is defined as the ratio of ring gear teeth to sun gear teeth, or the ratio of their corresponding radii, Rr and Rs, as k = Rr/Rs. Given the basic ratio, the governing equation for epicyclic gear as a speed summing device is ws þ kwr

ðk þ 1Þwc ¼ 0

ð2:2Þ

In (2.2), ws, wc and wr correspond to angular speed of the sun, carrier and ring gears, respectively. S/A sun gear speed, ws, runs backward with respect to the engine speed, wc, in order to match road dependant speed at the ring gear. Controlling the power transferred via the S/A at a given speed sets its reaction torque level against the ICE via the ratio Gcs. The torque levels at sun and carrier gear can be expressed in terms of the ring gear torque, M, the gear mesh efficiencies, h (generally a loss of 2%/mesh), the polar inertias, J, and accelerations as shown in (2.3): 1 1 hr Mr Js w_ s þ Jr w_ r ¼ 0 k k kþ1 kþ1 hr Mr Jc w_ c Jr w_ r ¼ 0 hc Mc þ k k hs Ms

ð2:3Þ

Figure 2.21 is a plot of vehicle speed, V, versus ws, wc and wr to illustrate how each of these propulsion components responds during acceleration at WOT. For example, suppose the ICE is rated 80 kW peak power for a Focus sized 4 or 5 door passenger vehicle and further suppose that MG1 is rated 32 kW and MG2 is rated 16 kW. The vehicle accelerates from standstill to 60 mph (26.82 m/s) in 6.8 s with the ICE operating at approximately 2,500 rpm.

Hybrid architectures Component speeds (m/s, rad/s)

292.004136

91

300 wr(i)

10 V(i)

200

wc(i)

100 ws(i) 0

0

0

2

4

6

8

t(i) Time (s)

0

10 10

Figure 2.21 Acceleration performance of power split ‘electric CVT’

For the stated conditions, and for k ¼ 2.8, the S/A speed remains positive and governed by (2.2) for the given ring gear (vehicle) and carrier (engine) speeds. For this particular choice of engine speed, both the S/A and M/G operate with positive speeds and in generally efficient torque–speed regions. If the engine speed were reduced somewhat during this acceleration event, the sun gear speed would actually decrease to zero and reverse direction. Figure 2.22 illustrates this behaviour.

Component speeds (m/s, rad/s)

292.004136 400

–167.170167

w r(i) 10 V(i)

200 w c(i) 0 w s(i) –200 0

0

2

4

6 t(i) Time (s)

8

10 10

Figure 2.22 Acceleration of power split transmission when engine speed is lowered Depending on choice of planetary and final drive ratios it is possible for the power split S/A to assume inefficient operating points, particularly if its speed were to dwell near the zero crossing point. This could be caused, for example, by poor

92

Propulsion systems for hybrid vehicles

choice of ratios and operating the vehicle in slow traffic. The power split operating modes are explained: Mode 1. Vehicle launch: The engine is off, carrier speed is zero and only the M/G propels the vehicle. Battery power is discharged through M/G to the wheels. This mode persists to approximately 20 kph. Mode 2. Normal cruise: Engine power is delivered to the wheels via the planetary gear. S/A operates as a generator and the M/G operates as a motor. The battery does not participate in propulsion. S/A electrical power is summed at the driveline by the M/G. Mode 3. Full throttle acceleration: Conditions are the same as for Mode 2 with the exception of M/G power being augmented by input power from the battery. The battery discharges in this mode. Mode 4. Deceleration/regenerative braking: Engine is stopped, S/A is stopped and kinetic energy from the vehicle is recuperated via the M/G back to the battery.

Example 4: Using the 2010 Camry Hybrid eCVT powertrain architecture shown in Figure 2.23 and given that the maximum angular speed of MG1 = 1,465.8 rad/s, what is VWOT=? 2010 Camry Hybrid eCVT

ICE MG1 Batt

Pb

78 23 C 1

R C2

30 S 1

S2

E1

E2

Final drive 57 18 23

axle Diff

80 23 55 54 gfd = 3,542 MG2

P215/60R17

Pcir

Figure 2.23 Camry Hybrid eCVT architecture Solution: In Figure 2.23 the applicable gear ratios are defined as follows: 78 57 ¼ 2:60; E2:kE2 ¼ ¼ 2:478; 30 23    80 55 ¼ 3:542 Final drive: gfd ¼ 23 54 E1:kE1 ¼

Using the defined gearing (Table 2.5), the tyre characteristics and the stated maximum speed of the main traction M/G, MG1 yields   rW wMG1 ¼ 59:29 m=s ¼ 132:6 mpg VWOT ¼ gfd kE2

Hybrid architectures

93

Table 2.5 Camry Hybrid 2010 model specifications Engine Tyres Dynamic rolling radius Vehicle curb mass Tyre roll resistance 0–60 mph time

2.4 L P215/60R17 rw Mv Crr tz60

110 kW 0.355 m 1,673 0.008 8.9 s

Aerodynamic Frontal area Battery voltage MG1 power Battery power MG2 power

Cd Af Ub PMG1 Pb PMG2

0.27 ~2.38 m2 244.8 V 105 kW 30 kW 75 kW

The vehicle is therefore capable of operating to VWOT provided its engine and electric traction motor can supply the needed propulsion power. See ‘Exercises’.

2.3.2 Power split with shift The basic architecture of power split can be augmented with a gear shift after the torque summation point.4 With a shift point in the ring gear to vehicle speed plots there must be a fast speed transition in the sun and ring gear speeds if the engine speed is held steady and there can be no discontinuities in vehicle speed. The behaviour depicted in Figures 2.21 and 2.22 for the same component speeds, ws, wc and wr, and vehicle speed, V, versus time shows tendency for the S/A speed to cross zero or to hover near zero. Now suppose a single gear shift event is assumed to occur sometime during the vehicle’s acceleration (Figure 2.24). The consequent speed transitions are shown as the sun and carrier speeds slew to new speeds to maintain the power flow constant prior to and subsequent to the shift.

Component speeds (m/s, rad/s)

408.805791

–410.018516

500 w r(i)

10 V(i) w c(i)

0

w s(i) –500 0

0

2

4

6

8

t(i) Time (s)

10 10

Figure 2.24 Power split dynamics during acceleration when a single gear shift is assumed 4

Alternative architectural concept explored by J-N-J Miller Design Services, PLC.

94

Propulsion systems for hybrid vehicles

What the gear shift event does in a power split transmission is to cause the sun gear speed to toggle from clockwise rotation to counter-clockwise rotation while remaining well away from zero speed. This ensures higher operating efficiency and no stalled operation of the S/A (i.e. as it would be, had its speed been commanded to zero while holding torque level). A second rationale for providing a gear shift to a power split is to implement a high/low range feature. With the added gear ratio active, the driveline is essentially given a different, much shorter final drive. For example, if the inserted ratio is 1.4:1 and the final drive, FD, was 3.5 then the new, equivalent final drive will be 4.9. This much higher final drive is typical of towing applications and provides the vehicle with a low range function. When engaged, the vehicle has much higher launch traction for grade climb, deploying or launching a boat, or for driving in deep snow for example (see also Chapter 11 for towing example). Disengaged, the final drive reverts back to its normal setting, or the equivalent of high range transmission. The terminology of shorter and longer final drive has been used in the above discussion without explanation. It may be clearer if this terminology is defined in the context of driveline revolution counts per mile. Vehicle speedometers and odometers rely on a signal taken from the transmission output shaft that delivers a pulse per wheel revolution. This means, for example, that a vehicle having a production tyre will turn a prescribed number of revolutions per mile of travel, typically 850 for the production final drive ratio. Now, if the customer changed tyres to a different aspect ratio, but the same rim size, the count would be off and so would the speedometer and odometer readings. For illustration, suppose the final drive ratio is changed from 3.5 to 3.7. With this increased ratio, the propeller shaft to rear wheel drive or half-shaft speed in the case of front wheel drive will spin 5.7% faster for the same distance travelled. The final drive is thus said to be shorter because each revolution of the propeller shaft results in a proportionally shorter distance travelled. Had the final drive ratio been decreased from 3.5 to perhaps 3.2, then the propeller shaft would only turn 91.4% of a revolution to traverse the same distance, in effect a longer final drive. Longer final drive means the engine speed is lower for a given vehicle speed. As a consequence, the engine exhibits more lugging behaviour. The disadvantage of shifting a power split transmission is the increased control complexity of blending torque from three sources plus the need for a driveline clutch. A clutch in the driveline always introduces a torque hole in propulsion while the clutch disengages, the component speeds re-establish themselves to new equilibrium points and the clutch ceases to slip. The resulting loss of transmitted torque, or a torque hole, may persist for 150–300 ms. In addition to interruption of tractive effort, a clutch event introduces power loss and contributes to driveline shudder and potentially to driveline oscillations if left unchecked. Toyota Motor Company incorporated a gear shifting function to their single mode eCVT with the introduction of the GS450h in 2006. The reason for the gear shift was to equip this luxury vehicle with autobahn cruising speed capability. The single mode eCVT discussed to this point is restricted in WOT speed by the upper limit of its main traction M/G, MG1, angular speed. In the GS450h the 14,000 rpm

95

Hybrid architectures

MG1 enables the vehicle to reach approximately 75 mph prior to a compound planetary gear shift by a ratio of 1.6:1. After the shift down in MG1 angular speed, the vehicle can reach cruising speeds of 250 kph or 150 mph. Development of the GS450h [18] has been described as electric turbo charging and the means to extend the high speed capability of a single mode eCVT to autobahn speeds. This is the first production roll-out of a shifting eCVT, the powertrain architecture of which is shown in Figure 2.25.

rw B1

B2

E1 R ICE

N2

ETC

E2

FD

R

N1

P C

M/G2 Motor

C

S

P

P

Sm

R

P

P

Wheels

Sr P

B1

B2

Ratio

High

1

0

1.9

Low

0 0

1

3.9

C

ECU Exec. Cnd

N1

cA/C Brakes Electrical Loads EPS hotel brakes, etc.

R

= ESS

V

=

Uess Ravigneanx gearset ‘shifts’ MG2 speed range

Figure 2.25 Shifting type eCVT: Lexus GS450h

Table 2.6 Lexus GS450h hybrid model specifications Engine Tyres Dynamic rolling radius Vehicle curb mass Tyre roll resistance Wide open throttle Generator power

3.5 L P245/40ZR18 rw Mv Crr VWOT PMG2

218 kW 0.327 m 1,879 0.008 131 mph

Aerodynamic Frontal area Battery voltage MG1 power Battery power Final drive 0–60 mph

Cd Af Ub PMG1 Pb gfd tz60

0.27 ~2.34 m2 288 V 147 kW 36 kW 3.769 5.2 s

Example 5: Using the GS450h architecture and specifications in Table 2.6 and given that the main traction motor, MG2, has a maximum speed of nmax = 14,400 rpm, find the vehicle speed shift point range.

96

Propulsion systems for hybrid vehicles

Solution: Using the same expression developed in Example 4 results in   rW wMG1max ð0:327Þð1507:7Þ ¼ ¼ 33:54 m=s ¼ 75 mph VLow ¼ gfd gLow ð3:769Þð3:9Þ   rW wMG1max ð0:327Þð1507:7Þ ¼ ¼ 68:8 m=s ¼ 154 mph VWOT ¼ gfd kE2 ð3:769Þð1:9Þ Clearly, the introduction of gear shifting into the single mode eCVT has enhanced its capability for high speed operation. This, coupled with a higher output V6 engine and more powerful battery, yields a vehicle net propulsion power of 254 kW sufficient to sustain the high speeds. However, in the production of GS450h Toyota has introduced a rev-limiter that restricts top end speed to 131 mph.

2.3.3

Continuously variable transmission derived

The driveability concern of torque holes in a step ratio transmission or the corresponding losses in an automatic transmission are partially offset in a CVT. The CVT adjusts its ratio continuously over its gear shift ratio coverage range, Gsrc. In the CVT, Gsrc is maximally 6:1. In Figure 2.26, the belt type CVT has the engine input applied to its primary side through a mechanical clutch and an M/G connected permanently at the primary, but outboard of the engine. The secondary side of the CVT is connected via gears to the transmission final drive as shown. The CVT itself can be either a belt (Reeves) or a toroidal (Torotrak) system. Engine

CVT Motor Accessory drive

Figure 2.26 CVT hybrid (from Reference 16) The Reeves type CVT with a rubber belt is commonly found in snowmobile transmissions. The steel belt (Van Doorne) CVT having offset axes is ideal for mounting in small front wheel drive vehicles. The Van Doorne is a steel compression

Hybrid architectures

97

belt and is most popular as the transmission in subcompact and compact passenger cars. This type of CVT will exhibit a fuel savings of 8% when compared to a conventional 4-speed automatic transmission. This fuel savings is the same for a 6-speed automatic transmission, but the CVT is claimed to offer better acceleration performance. The toroidal CVT is better suited to larger passenger vehicles with high displacement engines (400 Nm torque range). Fuel economy in larger cars is improved because the CVT offers wider gear shift ratio coverage that can push the ICE farther into its lugging range than a conventional transmission. The limitation of toroidal CVTs in the past has been the design of the variator, particularly its limited cross-section space allocation due to vehicle design. Dual cavity toroidal CVTs are most suitable for rear wheel drive vehicles, hence larger passenger cars, light trucks and sports utility vehicles. Low variator efficiency occurs when there is excessive contact pressure on the torus rollers in the low ratio position and when large ratio spreads are demanded. Efficiency at ratio spreads greater than 5.6:1 can fall from 94% to 89% at full load. A novel dual cavity, toroidal CVT has been announced by Torotrak and is called the IVT (infinitely variable transmission) [19] shown in Figure 2.27.

ICE

Wheel Low clutch

High clutch

Figure 2.27 Torotrak IVT, toroidal CVT architecture and production model In the Torotrak IVT with epicyclic gearing, the system has high and low operating regimes. In the low regime, the IVT covers low speed, reverse and neutral. In the high operating regime, the IVT covers all forward speeds including overdrive.

2.3.4 Integrated hybrid assist transmission The transmission manufacturer JATCO developed a hybrid automatic transmission termed integrated hybrid assist transmission (IHAT) that works on the epicyclic gear principle of speed summing. Rather than employing dual M/Gs for power split operation, the IHAT uses a single M/G in a unique architecture. Figure 2.28

98

Propulsion systems for hybrid vehicles Wheels OWC R

ICE

L/U C

M/G

Gearbox

FD

S

3 Battery pack

Figure 2.28 IHAT architecture of power split with single M/G

illustrates the IHAT driveline architecture having an M/G connected to the sun gear, engine at the ring gear and output from the carrier. The one-way clutch (OWC) grounds the transmission input shaft to chassis for park and engine cranking by the M/G as well as preventing reverse rotation of the carrier. The IHAT architecture has six operating modes: Mode 1. Idle–stop: In this mode, the OWC is activated and M/G torque is amplified by the basic ratio of the epicyclic gear for cranking the engine. Mode 2. Vehicle launch and creep: With the ICE running, the M/G torque reverses to generating quadrant so that reaction torque is applied to the sun gear, enabling vehicle creep while ICE speed is held constant. Mode 3. Vehicle launch: When the accelerator pedal is pressed, the engine produces higher torque, IHAT torque increases in generating mode and engine torque is applied to the wheels via the ring gear. At some point during launch, the M/G torque reverses sign and enters motoring quadrant. M/G speed approaches engine speed. When M/G speed and ICE speed are approximately equal, the lockup clutch (L/U) is applied connecting the M/G and ICE to the transmission input shaft. Mode 4. Power assist and regenerative braking: The L/U is applied and the system operates in ISG mode. Power from the ICE is summed with M/G power to meet the road load requirement. During braking, regenerative power is supplied via the M/G to the vehicle’s traction battery. Mode 5. Generating mode: With automatic transmission selector lever in N or P, the L/U engages with the engine running, causing the M/G to generate electricity. Mode 6. Hill holding: Again in ISG mode, an issue with non-level terrain is roll back when the driver attempts to launch the vehicle on a grade following engine idle–stop mode. In the IHAT system during the time between release of the brake pedal and application of tractive power, the OWC supplies sufficient hydraulic pressure to engage the AT gears.

Hybrid architectures

99

Table 2.7 Single M/G power split architecture component ratings Internal combustion engine

Engine type Max torque and power

Electric motor–generator

M/G type Torque and power

Battery

Type Voltage Power Automatic transmission ratios

AT

Gasoline, V6, 2 L 172 Nm at 4,400 rpm 96 kW at 5,600 rpm Permanent magnet 122 Nm at 1,000 rpm 41 kW at 4,000 rpm Nickel–metal hydride 288 V (40 modules of 7.2 V each) 22 kW maximum First: 3.027 Second: 1.619 Third: 1.000 Fourth: 0.694 Rev: 2.272 Final drive: 4.083

Ratings of the IHAT system are described in Table 2.7 for a 1,650 kg curb weight vehicle. With this propulsion system architecture, the engine speed can be held constant during the entire launch interval. However, the M/G must slew from generating to motoring mode and from reverse rotation to forward rotation in the process. Figure 2.29 illustrates the velocities of the planetary gears and vehicle speed versus time during vehicle acceleration.

Launch mode

ICE Carrier = G ⫻ Uveh

M/G

M/G = Generating

M/G = Motoring

Figure 2.29 Single M/G power split acceleration performance

100

Propulsion systems for hybrid vehicles

In Figure 2.29 the M/G rotates in reverse direction during generating mode so that constant speed ICE operation is possible. The generating mode continues until the M/G speed crosses through zero, at which point it enters motoring mode. This is an inefficient operating point requiring M/G stall torque. Once through zero speed the M/G is motoring with positive rotation until the L/U clutch engages, pinning the ring (ICE) and carrier (AT) gears together.

2.4 Post-transmission parallel configurations The second option for locating the ac drive in a hybrid vehicle is to insert the M/G at the transmission output shaft, but ahead of the final drive. In this posttransmission configuration the M/G does not have the benefit of gear ratio changes; therefore, it must operate over the very broad vehicle speed range. This demands a high torque ac drive that can function over wide CPSR. The disadvantages of post-transmission hybrids are the high torque levels, impact of continuous engagement spin losses on fuel consumption and package difficulty. Higher torque M/Gs are always physically larger since more rotor surface area is needed to develop surface traction. Larger moment arms to this surface traction are more restricted because the package diameters are usually constrained to fit within transmission bell housing diameters (200–350 mm OD). An example of a post-transmission hybrid would be an in-wheel motor or hub motor hybrid. The GM Autonomy, for example, could be classified as a posttransmission hybrid because the hub motor is separated from the wheel by a nonshifting epicyclic gear. The Autonomy (Figure 2.30) is a concept automotive chassis designed for wide-ranging body style flexibility and cross-segment application. All propulsion, energy storage, chassis functions and wiring for power distribution and communications are packaged within the skateboard-like chassis. Communications is via controller area networks (CAN). Power for propulsion is at high voltage, 300 V typical or higher when fuel cells are used. Chassis and passenger amenities are powered by 42 or 12 V for lighting. A concern with hub motors is their higher unsprung mass, a tendency for torque steering and durability. Because of lower speeds and high torques, a hub motor will be inherently heavier than its higher speed axle or pre-transmission equivalent. Torque steer is a phenomenon due to steering and suspension geometry design. Durability is a persistent issue with hub motors because of simultaneous vibration, temperature, water/salt spray ingress, and sand, dust and gravel impingement. Torque steer can be understood by recognizing that vehicle steering geometry will generally have non-zero scrub radius. When the suspension king pin axis intercepts the tyre-road patch inside the plane of the wheel, the distance from the wheel plane to the king pin axis is referred to as the tyre scrub radius. If the intercept point is inside the wheel plane the scrub is positive and if outside it will be negative. A negative scrub radius puts the wheel turning axis outside the wheel

Hybrid architectures

101

Figure 2.30 Autonomy with in-hub M/G plane on which the corner mass of the vehicle sits. The wheel torque develops a longitudinal component of tractive effort at the wheel plane that is in board of the steering axis. This off-axis steering moment due to tractive effort tends to re-align the wheel so that the axis of applied wheel torque and the king pin axis align.

2.4.1 Post-transmission hybrid There has been work on electric M/Gs connected to the vehicle propeller shaft ahead of the final drive, but these programmes were discontinued in the case of electric propulsion due to the high demands on machine power density, speed range and physical size. Figure 2.31 illustrates the concept of a post-transmission hybrid in which an electric machine is interfaced to the driveline via a gear reduction.

E-steer

M/G ICE

FD

XM 3 Battery pack

Figure 2.31 Post-transmission hybrid architecture

102

Propulsion systems for hybrid vehicles

The speed range concern with a post-transmission hybrid has to do with operating deep into field weakening of the electric machine and not incurring electrical and mechanical spin losses when the M/G is unenergized. If spin losses become a major fuel economy issue, the post-transmission M/G would require an additional clutch to remove it from the driveline during coasting periods. Wide CPSR is more problematic. With a post-transmission M/G there is no option – it must possess CPSRs greater than 6:1 and preferably 10:1 in order to deliver both high torque at low speeds for tractive effort plus constant power at higher speeds for optimum propulsion. Figure 2.32 illustrates the motor capability curves required from a post-transmission hybrid. A high torque, in the vicinity of 300 Nm, is necessary to deliver low speed tractive effort and wide CPSR is necessary to hold shaft power at high vehicle speeds. Efficiency contours are estimated for such a post-transmission electric M/G to illustrate the placement of peak plateaus. An even more advantageous efficiency contour map would have high efficiency islands extending towards zero on the chart so that best operation would be available at low demands regardless of speed as well as at higher demands. The capability curve mapped in Figure 2.32 is also needed for in-wheel motors. Such hub motors have no option for gear shifting and generally are direct drive units.

2.4.2

Wheel motor hybrid

There have been many projects over the years to adapt hub motors as posttransmission wheel motors. DOE has funded some of these activities and others T (Nm) 300

0

1

2

3

4

5

6 Speed (krpm)

Figure 2.32 Post-transmission hybrid capability curves

Hybrid architectures

103

have been privately funded. Ontario Hydro developed an in-wheel motor for hybrid propulsion. Volvo Car Company examined hub motors to determine the package benefits of fully packaged wheel assemblies that contained propulsion, steering, suspension and braking all integrated. The recent GM Autonomy is a similar concept. The University of Sheffield in the UK developed a demonstration wheel hub motor for application in their Bluebird EV formula 3000 vehicle. The hub motor is a direct drive, f310 mm by L220 mm, capable of delivering 382 Nm of continuous torque. Toyota Motor Co. has unveiled a wheel motor fuel cell hybrid called the FINE-S (Fuel Cell Innovative Emotion – Sport). Toyota has already leased four FINE-S vehicles to city officials in Japan for operational use. The design goal of FINE-S is to focus on modularity of components and subsystems. The fuel cell components and wheel motors permit versatile packaging freedom not available in conventional cars. Individual wheel motors in the FINE-S fuel cell hybrid vehicle enable low centre of gravity, high performance handling and smooth ride qualities. Fine-tuning the wheel motor torque levels provides high dynamic response traction control and longitudinal stability. The four seat FINE-S concept vehicle is shown in Figure 2.33. A very recent illustration of in-hub motors can be found in Reference 20 in a concept demonstration motorcycle having the complete power plant housed inside the rear wheel hub. Developed by Franco Sbarro, and unveiled at the 2003 Geneva

Figure 2.33 Toyota fuel cell vehicle with individual wheel motors – the FINE-S concept

104

Propulsion systems for hybrid vehicles

Motor Show, the semi-encapsulated motorcycle has a 160 hp (119 kW) Yamaha engine plus 5-speed gearbox, including radiator, exhaust, brake, battery, fuel tank and suspension all packaged within a single 2200 wheel. The system is cited as being an autonomous motor unit, or independent wheel drive.

2.5 Hydraulic post-transmission hybrid Architecting a post-transmission hybrid with hydraulic M/G is probably the most sound engineering approach. Not only will a hydraulic motor have the necessary torque and power density, but will also offer dramatic launch and acceleration performance.

2.5.1

Launch assist

Figure 2.34 illustrates the concept of hydraulic propulsion in which a hydraulic motor–pump (M/P) is connected at the transmission output shaft. Hydraulic fluid flow is managed at a valve head within the M/P and includes a reservoir located beneath the chassis at the vehicle’s rear axle. A comparison of hydraulic to electric drives is given in Reference 21. In this reference the author points out that power densities of the hydraulic systems will be higher than that of the electric systems because hydraulic pressures can be increased to achieve more performance. Hydraulic system pressures of 5,000 psi (350 bar) are containable and provide M/P performance at levels of 0.5 kW/kg. Working pressures and speeds of, for example, an axial piston pump have not increased much beyond 350 bar due to issues with noise and vibration. A relative comparison of hydraulic versus electric systems is shown in Figure 2.34.

Specific power density (kW/kg)

10 Hydraulic (swach plate, bent axis pumps)

1 Electric machines Automotive starter: 0.3 kW/kg Hybrid S/A(8 kW): 0.5 kW/kg Hybrid M/G (50 kW): 1.2 kW/kg 0.1 0

50

100

150 200 250 Nominal power (kW)

300

350

Figure 2.34 Specific power density (kW/kg) of hydraulics versus electric systems

Hybrid architectures

105

However, with fluid power, the mass of system components such as reservoir, lines, fittings and fluid is generally more than doubled. In the Airbus 320, for example, the hydraulic components in one system weigh 200 kg while the plumbing, fittings and pressure containment mass adds an additional 240 kg. In total, the redundant hydraulic system weighs some 560 kg. In an electric system operating at high voltage, the dominant mass will be contributed by the M/Gs themselves and very minimal contribution will come from wiring harness, connectors and cable shielding. The hydraulic launch assist hybrid is an excellent example of hydraulic M/Ps applied to the propeller shaft of a truck or SUV. During decelerations, the hydraulic launch assist accumulator is charged by a hydraulic pump driven by, and directly connected to, the vehicle’s propeller shaft. Then, on subsequent acceleration, the accumulator hydraulic pressure is discharged through the same M/P operating as a motor, thereby adding propulsion power during acceleration. Such systems operate at 350–420 bar and require substantial containment structure around the accumulator and M/P.

2.5.2 Hydraulic–electric post-transmission There are proponents of hydraulic energy storage in an electric drive system. The concept is sound, and reminiscent of flywheel systems, but most likely not economical. In order to deplete/replenish, the hydraulic storage and electric M/G connected to a hydraulic M/P is necessary, as illustrated in Figure 2.35 according to Reference 22. The hydraulic system described in this section has been called an offgrid power boost, or mechanical capacitor, by its inventor, Steven Bloxham. Supervisory controller

E-steer

M/G ICE

FD

XM

M/G

High pressure accumulator

M/P

Low pressure accumulator

Figure 2.35 Hydraulic–electric post-transmission hybrid In Figure 2.35 the presence of two energy conversions sets an upper bound on system efficiency at approximately 76% each way, or 58% turnaround, assuming the following reasonable component efficiencies. Efficiency of the hydraulic M/P is taken at 90% and electric M/G at 92%. A 58% turnaround efficiency is less than

106

Propulsion systems for hybrid vehicles

the storage efficiency of lead–acid battery systems. The second issue with two energy conversions is the necessity to size the M/G and M/P to the maximum power levels needed. Hydraulic components can achieve very high power densities, in the range of 1.3 kW/kg or higher, depending on system pressure levels. The issue with operating at pressures of 5,000 psi or higher is the level of safety afforded by containment structures and the attendant weight added.

2.5.3

Very high voltage electric drives

This section is included to accommodate the views of some that high voltage vehicular ac drives operating from 600 V to 3 kW or higher provide the performance demanded by next generation hybrid vehicles [23]. The premise that higher voltage electric machines are more efficient than lower voltage machines, all else equal, is generally not true. The most efficient electric machines are large turbogenerators rated up to 600 MW for utility generation operating under load at 99% efficiency – at a single operating point, 3,600 rpm, 60 Hz and fixed voltage! In the past 5 years there has been considerable interest in composition modified barium titanate (BaTi) sintered in a void-free matrix of very high purity. With this ceramic material, the goal is to realize a dielectric withstand voltage of greater than 3.5 kW for 1 mm thick sandwiches having an aluminium flash current collector. The US start-up company EESTOR [24,25] had claimed to have a production version of this ceramic ultra-capacitor available by the end of 2009, but till the time of writing, that is the end of 2009, no samples of the EESU appeared. The ceramic ultra-capacitor EESU [25] is claimed to store 52 kWh in a package of 281 pounds and 2.63 ft3 of volume. The energy stored in this ultracapacitor will be: each of the 31,351 slabs contain 100 CMBT capacitor chips making a grand total of 3,135,100 chips in the EESU. The EESU consists of some 31,351 thin slabs of CMBT (composition modified BaTi) dielectric that is capable of 610 V/mm at +85 C. This means the 1 mm thick slab can easily withstand the 3.5 kV that will appear across the 31,351 slabs in parallel having a combined capacitance of 30.693 F. The energy stored in this ultra-capacitor will be 1 30:693ð3,500Þ2 ¼ 188 MJ We ¼ CUc2 ¼ 2 2

ð2:4Þ

And, a specific energy, SE, of SE ¼

We 52:2 kWh Wh ¼ 407:8 ¼ 128 kg kg M

ð2:5Þ

This SE amounts to a truly phenomenal energy storage component if it materializes. Technically, this ceramic ultra-capacitor will require ultra-pure dielectric material that is absolutely void free and protected from mechanical fracture. Electrically the dielectric constant published to be in the range of 15,000 or higher must not saturate much with voltage or the SE mark will not hold. Moreover, if this unit

Hybrid architectures

107

were to be used in a plug-in hybrid or BEV, the claimed charge time of 6 min will demand an electrical service of 624 kW. In Chapter 1 it was noted that a level 3 charger is designed to operate from 240 Vac and 7 kW or higher. Other classes within level 3 will elevate this to 250 kW for dc chargers. Connectors will be the real concern and are being addressed by standards setting organizations.5 At the time of this writing EESTOR has not met their final milestone for scaled sample testing.6

2.6 Flywheel systems Flywheel energy storage has been promoted by some for several years as a viable, high cycling, storage medium. ORNL did considerable work on 500 Wh flywheel units for automotive use during the last decade. With the availability of high tensile strength fibres, it is possible to develop high energy density storage systems suitable for vehicular use. This topic is discussed more under energy storage system in chapter 10. A second application of flywheel technology has been to use the M/G itself as the flywheel.

2.6.1 Texas A&M University transmotor Figure 2.36 is a functional diagram of a flywheel hybrid system developed at Texas A&M University referred to as a ‘transmotor’ system. Transmotor is an electric motor suspended by its shafts and having both rotor and stator in motion. As a speed reducer or for speed increase, the transmotor permits constant speed operation of the Battery

Motor controller Clutch 3 Clutch 1 Clutch 2

Stator Rotor

Gearbox Engine

Housing

Figure 2.36 Transmotor basic configuration (with permission from Texas A&M University) 5

6

SAE J1772 specifies 80 A dc connector and IEC18651 an 80 A ac rating connector. Connectors for charging up to 250 kW are in the development process. Public announcements that scaled EESU samples would be available by the end of 2009 did not materialize, nor has there been any public announcement by October 2010.

108

Propulsion systems for hybrid vehicles

engine when used in conjunction with a torque splitting device. The governing equation for the transmotor, assigning wr to the rotor and ws to the stator, is Pe ¼ Tr ðwr þ ws Þ

ð2:6Þ

where Pe is the electric power at port 3 of the transmotor. Ports 1 and 2 are the rotor and stator mechanical connections. The physical arrangement of this configuration is shown in Figure 2.36 where clutches are used to connect and disconnect the engine and driveline so that all operating modes can be met. Clutch 1 connects or disconnects the transmotor rotor from or to the engine. Similarly, clutch 2 connects or disconnects the transmotor stator from or to the engine. Clutch 3 is used to lock the transmotor stator to the chassis. The transmission input shaft is permanently connected to the gearbox input shaft.

2.6.2

Petrol electric drivetrain

Clutch One-way clutch

A concept that has been explored for some time is the work of P. Jeffries in the UK [26] that he refers to as petrol electric drivetrain (PEDT). The PEDT frees the M/G stator assembly to rotate so that both rotor and stator are free to move. Doing so permits the stator to function as a flywheel and accumulate mechanical energy from the drivetrain and store it in the same form. This concept of storing and delivering the energy in the same form in which it is being used has considerable merit. For example, it has already been pointed out that ultra-capacitors make good storage systems because electric energy is stored in the same form in which it is being used. Figure 2.37 illustrates the PEDT concept. The ICE in Figure 2.37 couples to the PEDT electric machine stator/flywheel through an automatic clutch, an OWC and speed increasing gears. When the engine

I

C

E

Inverter

Batteries

Electrical machine Gears

Wheel

PEDT system configuration

Figure 2.37 Mechanical energy storage and consumption in flywheel ( from Reference 26)

Hybrid architectures

109

is switched off, the OWC prevents it from being back driven by the flywheel and ensures that the stator/flywheel can only rotate in the same direction as the rotor. For its part, the rotor is connected to the wheels through reduction gears, and a differential, but no clutch. Operation of this PEDT is in either all electric or dual mode. In electric-only mode with wr > ws, the battery is discharged into the stator via the power inverter. Reaction torque on the stator extracts energy from the flywheel, adding to the battery supplied energy. Both battery and flywheel energy accelerate the vehicle, delivering rather brisk performance. When the flywheel energy is bled off, the stator is clamped to the chassis by the OWC. Power transfer in this mode is limited by the peak coupling torque existing between stator and rotor. Battery current enables the power transfer according to the action of the power electronic control strategy. The second condition presents itself when wr < ws and the stator/flywheel rotates faster than the rotor. In this mode, the machine operates as a generator sending power to the battery. During vehicle braking the same set of conditions apply, depending on the relative speed between the stator and rotor. Vehicle kinetic energy is delivered to either the flywheel, the battery or both during braking. In dual mode the ICE remains coupled to the PEDT via the drive clutch. The ICE operates in thermostat mode when engaged, sending its power to the battery/ flywheel when not needed to meet road load. When the flywheel is charged, the ICE is shut down and the clutch deactivated.

2.6.3 Swiss Federal Institute flywheel concept A flywheel storage system based on CVT transmission technology has been described by the Swiss Federal Institute of Technology (ETHZ) based on a small gasoline engine rated 50 kW, a 6 kW synchronous M/G, a 0.075 kWh flywheel and 5 kWh of batteries in a CVT architecture [27,28]. The architecture shown in Figure 2.38 has four modes of operation: Mode 1. Normal drive: The ICE and the CVT are used for propulsion. Clutches C1 and C2 are closed. Clutch C3 to the flywheel is open and the electric M/G is coasting. Mode 2. Flywheel assist: Clutches C2 and C3 remain closed so that the flywheel and CVT maintain propulsion. When the flywheel speed drops below 1,800 rpm, the engine is started and C1 engaged to spin up the flywheel to 3,800 rpm and to deliver propulsion power to the wheels. When the flywheel speed exceeds 3,800 rpm, the engine is cut off and clutch C1 opened. Mode 3. Electric machine and flywheel assist: This is the zero emissions operating mode. Power is delivered to the wheels by the M/G, augmented by decelerating the flywheel if necessary. Mode 4. M/G and CVT only: This is the zero emission operating mode of a conventional EV. Because of limited rating of the M/G and battery, vehicle acceleration performance is limited. Without a higher rated M/G, this mode will be used only for low speed operation.

110

Propulsion systems for hybrid vehicles

Supervisory control

C1 C3 Flywheel C2

e-mtr

Energy storage

Power inverter

CVT trans. & FD

Figure 2.38 Swiss Federal Institute of Technology flywheel CVT hybrid The supervisory controller operates the ICE, CVT, M/G and clutches for proper operation in the four hybrid modes. The controller also monitors the energy storage system, power inverter and ICE for state of charge (available energy), inverter modulation settings and engine power.

2.7 Ultra-capacitor-only vehicles There is growing interest in the application of carbon–carbon symmetric and asymmetric ultra-capacitors as the main energy storage component in hybrid vehicles – primarily for heavy transportation such as transit bus, metro or subway and light rail. The reason is that most of these vehicles are route following and can have infrastructure for charging stations along the route, at minimum one at the beginning and one at the terminus. Across the globe municipal transit authorities (MTA) have invested in various efficiency optimization options including biofuel and compressed natural gas (CNG), diesel power plants to hydrogen fuelled and recently to batteries, hybrid electric diesel and most recently ultra-capacitor energy storage buses. The following sections look at each of these in somewhat more detail.

2.7.1

Catenary powered vehicles with ultra-capacitors

A US and Chinese venture company Sinautec Automotive Technologies7 has developed an ultra-capacitor-only powered transit bus that relies on overhead catenary lines (dual lines for safety) that charge the bus at each stop. Along the route the bus can navigate for several kilometres on energy from the ultracapacitor-only RESS. The transit bus is of course a series electric. Figure 2.39 illustrates the Sinautec bus at a stop. For charging from the catenary, the bus must be positioned directly below the pair of overhead lines and its pantograph raised to make connection. At the bus stop a utility 480 Vac, 3-phase service with appropriate switchgear energizes an ac/dc 7

Linked with Shanghai Aoweii Technology Development Company.

Hybrid architectures

111

Number of passengers = 41 Bus length = 12 m Bus mass = 12.5 ton Maximum vehicle speed = 30 mph Air conditioning load = 15 kW Energy consumption: W/mi with A/C = 1,265 W/mi W/mi without A/C = 805 W/mi Range: With A/C = 3.5 mi Without A/C = 5.5 mi Ultra-capacitor pack voltage = 700 Vdc nominal Charging requirement = 200 A for 3 min

Figure 2.39 Ultra-capacitor-only transit bus converter with current regulation. The ultra-capacitor packs, typically a series string of 15 standard 48 V modules, are charged under constant current until their voltage reaches their rated value, at which point the charger enters constant voltage mode. Indicators on the bus instrument panel alert the driver when charging is complete, the pantograph is lowered and the bus resumes its route. It is typical of 12 ton class transit buses to consume from 700 to 1,500 Wh/mi energy. The benefit is in reduced CO2 emissions and no pollution along the bus routes. Operators are pleased with the quiet electric drive, passengers are impressed with transportation without diesel fume smell and the transit authorities are pleased with the significantly improved fuel economy and reduced maintenance.

2.7.2 Catenary powered vehicles with wayside ultra-capacitors Potentially the greatest benefit to transportation systems when it comes to the application of ultra-capacitors is that ultra-capacitors provide cyclable energy storage for heavy hybrids. Subway and metro-rail operators are finding that trackside, or wayside, energy storage facilitates stabilization of the catenary, reduces the transient burden on the power supply and improves schedules because increased ridership has resulted in higher catenary loading, shorter headways and higher voltage transients on the catenary. One application of ultra-capacitors in metro-rail is the Bombardier system of putting the RESS on the train car roof. With this system the metro is capable of running up to 0.5 km without catenary power, crossing intersections or passing through a tunnel. Figure 2.40 is one example of such a system. Trackside, or wayside, energy storage technologies currently in use include ultra-capacitor and flywheel systems. Pentadyne,8 for example, has in production high speed flywheel systems capable of 10 M cycles in 570–900 V applications at 1–6 MW pulse power and 5.2 MJ useable energy. Figure 2.41 illustrates a typical trackside energy storage using flywheel modules. 8

www.pentadyne.com /uploads for reference materials on trackside flywheel energy storage.

112

Propulsion systems for hybrid vehicles

Figure 2.40 RESS pack on-board, Bombardier system Top bearing Steel tube container Carbon and glass fibre composite rotor Stator Bottom bearing

Electric utility GRID Meter

Direct link in dc

Third rail or catenary

Multiflyheel unit configuration

Urban rail ac/dc substation

Figure 2.41 Trackside RESS using flywheels – Pentadyne system Trackside energy storage has benefits of minimum capital expenditure, lower operating costs, higher operating efficiency and less environmental impact due to reduced transient loading on the catenary power supply. For light rail the regenerative braking feature provided by ultra-capacitor RESS gives significant cost savings, improved network reliability via third rail or catenary voltage stabilization. The local transport system in several cities such as

Hybrid architectures

113

Cologne and Madrid are using the Sitras system that applies ultra-capacitor modules to a trackside unit, as shown in Figure 2.42. Each trackside unit consists of 1344 Boostcap ultra-capacitor cells rated 2,600 F, operating at 2.3 V and delivering a power level of 1 MW at 95% efficiency. The system reliability requirement is 22 h/day operation and maintenance free for 10 years.

Figure 2.42 Trackside RESS using ultra-capacitors – Siemens Sitras system

2.7.3 Ultra-capacitor trolley bus vehicles Trolley bus is another example of where wayside or third rail regenerative energy storage can make a substantial impact on system reliability, up-time as well as substantial energy savings. Many trolley buses were converted to dual mode for flexibility in navigating tunnels and intersections. Later the dual mode was abandoned in favour of hybrid technology and more recently of battery-electric mode. The 2008 Beijing Olympic buses are one notable example of battery-electric buses used to transport teams, managers and officials to events. The battery only electric bus developed for the 2008 Beijing Olympics was the result of a 10 year effort by the Beijing Institute of Technology (BIT) and 20 other organizations, including the Beijing Bus Corp. The team focused on overall energy efficiency, the need for zero emissions and fast turnaround of depleted batteries. To do this the BIT team devised the robotized battery swap system shown in Figure 2.43 that pulled a depleted lithium ion battery tray from the bus and replaced with a fresh charged pack from the charging racks as shown. This swapping operation of battery packs took less than 8 min per bus and the bus was back in service. The battery recharge and swap facility was located along the bus route.

2.8 Electric four wheel drive Another advancement that has been facilitated by electric propulsion systems is the flexibility to drive one or more axles using independent electric traction motors, either as stand-alone units with half-shaft coupling or as in-wheel motors as discussed earlier in this chapter.

114

Propulsion systems for hybrid vehicles

Figure 2.43 Trolley bus example and Beijing Olympics battery-electric bus

2.8.1

The E4 system

A hybrid propulsion architecture that physically decouples the ICE propulsion and electric propulsion systems is an electric four wheel drive, or all wheel drive system. In the E4 architecture the existing ICE driveline remains unchanged, except perhaps for up-rated electrical generation. The electric drive system is then implemented on the non-driven axle as illustrated in Figure 2.44, where the M/G is connected to the axle through a gear ratio and differential. Depending on design, a clutch at the gearbox input may be necessary to mitigate the effects of M/G spin losses.

Alt. Power inverter Energy storage

e-mtr Gear

Figure 2.44 Electric four wheel drive architecture It is also interesting to investigate the range of options in an E4 system. Not only are the axle power levels variable in the range from 10 to 25 kW, but the amount of storage dedicated to E4 is variable from 0 to 1 kWh or higher. To explain power level demands, it is only necessary to realize that four wheel drive on demand systems are not engaged frequently and when it is engaged the power level rarely exceeds 20 kW on passenger sedans to medium and full sized SUVs. This is a case of a little traction on the non-driven axle being far better than no traction. A more interesting concept is the fact that E4 can be implemented as a completely autonomous system with stand-alone transient energy storage to a fully integrated system sharing energy storage with the main hybrid propulsion system.

Hybrid architectures

115

2.8.2 Production ‘Estima Van’ example Adding on demand electric propulsion to the non-driven axle benefits the overall vehicle performance, depending on whether this axle is the front or rear. In a front wheel drive vehicle the ICE driveline remains at production level but an electric M/G is then added to the rear axle. Rear axle propulsion power levels are significantly lower than front axle levels due to a need to maintain longitudinal stability. A power of 15–20 kW peak provides adequate axle torque for grade, split mu and stability enhancement in a passenger sedan to medium SUV class of vehicle. Recall that regeneration levels at the given power levels recuperate most of the available energy (Figure 2.45).

Figure 2.45 Electric four wheel drive (Toyota Estima Van)

2.9 Exercises A narrow lane vehicle of the type shown in Figure 2.2 is driven at steady speed of 55 mph on level grade with no headwind. (a) What is the total propulsion power required? (b) What is the energy consumption rate (Wh/mi)? Given that Crr = 0.007 kg/kg, Af = 1.12 m2. Hint: Set P(V) = Pinertial + Proll + Paero + Pgrade Q1:

(1) (2) (3) (4)

Pinertial ¼ ðMv þ Mpass ÞV V_ Proll ¼ gMtot Crr V Paero ¼ 12 rair Cd Af V 3 Pgrade ¼ gMtot sin q

A1:

(a) 2,195.15 W; (b) 40 Wh/mi

Q2:

Repeat Q1 for the case of 15 mph headwind. What will be the total range in this case before fuelling up is necessary? For the case of Vw = 15 mph, P(V ) = 2195.15(31.3/24.6)3 = 4521.6 W a factor of 2.06 higher, therefore, the range is reduced to 1/(2.06) = 0.485. So, less than half the range without headwind.

A2:

116

Propulsion systems for hybrid vehicles

Q3:

Referring to Example 4 and Table 2.5 for the Camry Hybrid, the vehicle propulsion power, P(V), stated here must match the road load. Determine whether or not the vehicle’s net propulsion power is sufficient to match the road load at VWOT and if not, then to what speed can it match road load. Assume level terrain, no headwind and no driveline losses. Also, (2.7) uses M = Mv + Mpass and VWOT = 132.6 mph:   1 _ r Cd Af V 3 þ ðgM sin qÞV ð2:7Þ PðV Þ ¼ MV V þ ðgCrr MÞV þ 2 air

A3:

For a single occupant at standard mass P(59.29) = 0 + 8,129.3 + 81,698.7 + 0 = 89.828 kW. The net propulsion power of the 2010 Camry Hybrid: Pnet = Pe + Pb = 110 kW + 30 kW = 140 kW. So, yes, the propulsion power is sufficient to sustain this speed under ideal conditions.

At what steady speed can the Camry Hybrid in Q3 sustain for the case of four occupants, 20 mph headwind and 6% grade? Hint: Convert the %grade to angle using (2.8) as we’ll see in Chapter 3:   1 % grade ð2:8Þ q ¼ tan 100

Q4:

A4:

Since the net available propulsion power is 140 kW, the maximum speed can be stated as (2.9):   1 3 PðV Þ ¼ Pnet ¼ 140 kW ¼ ½gCrr M þ gM sin qŠV þ ð2:9Þ rair Cd Af V 2 V 3 þ 26:82V 2 þ 26,055:6V

356,428:36 ¼ 0

Or, V = 13.57 m/s = 30.36 mph. For these conditions the vehicle can sustain only 30 mph. Q5:

Much has been said of energy storage for hybrid vehicles and more will be covered in later chapters. For this question, consider the power and energy metrics of the regenerative energy storage system (RESS) pack for various classes of hybrid vehicles listed in Table 2.8. For each class of vehicle,

Table 2.8 Approximate peak power and energy metrics by hybrid vehicle class Class

Peak power (kW)

Total stored energy (kWh)

Power/energy ratio (P/E)

Open circuit voltage

Mild hybrid Strong PHEV BEV

12 27 130 136

1 1.3 16 42

12 20 8.12 3.3

158 330 330 335

Hybrid architectures

117

estimate the time its RESS can sustain peak power when at 60% SOC start and not drop below 25% SOC. A5: Mild = 105 s; Strong = 63 s; PHEV = 155 s; BEV = 390 s Note: Since P/E [=] h 1 the direct approach is to just solve: T = 0.35(3,600)/(P/E). Q6:

A6:

U2

J1798 defines battery peak power, Ppk ¼ 29 Roci where Ri is battery internal resistance and Uoc is the battery open circuit voltage. For each of the vehicle classes listed in Table 2.8, (a) estimate the battery pack internal resistance, Ri, and (b) the short circuit current for each case. Ri_mild = 0.462 W; Ri_strong = 0.895 W; Ri_PHEV = 0.186 W; Ri_BEV = 0.183 W And Isc_mild = 342 A; Isc_strong = 368.7 A; Isc_PHEV = 1,774.2 A; Isc_BEV = 1,830.6 A From SAE J1798, Ppk = (2/9)(Uoc2/Ri).

References 1. US DOE in cooperation with Office of Transportation Technologies and National Renewable Energy Laboratory. Future US Highway Energy Use: A Fifty Year Perspective, 2001. 2. Daimler Chrysler, Hightech report, Research and Technology, no. 2, 2002. 3. Volkswagen Motor Co. Press release. Available from www.vwvortex.com/ news/04_02/04_17, 6 June 2002. 4. Kornbluth K., Burke A., Wardle G., Nichell N. Design of Freeway Capable Narrow Lane Vehicle, SAE technical paper, SAE 2004-01-0760, 2009. 5. Stephan C.H., Miller J.M., Davis L.C. ‘A Program for Individual Sustainable Mobility,’ Journal of Advanced Transportation, Calgary, Alberta, Canada, 2003. 6. Available from http://www.autoshuttle.de. 7. Boys J.T., Shuzo N. Primary Inductive Pathway, US Patent 5,619,078, 8 April 1997. 8. Ansorge U., Wunderlich H., Aldinger M., Seelig A., Huder B. Track-guided Transport System with Power and Data Transmission, US Patent 6,089,512, 18 July 2000. 9. Rahimo M., Kopta A., Schlapbach U., Vobecky J., Schnell R., Klaka S. Power Semiconductor Devices & IC’s, 2009, 21st International Symposium, ISPSD 2009, 2009. pp. 283–86. 10. Backlund B., Rahimo M., Klaka S., Siefken J. Topologies, Voltage Ratings and State of the Art High Power Semiconductor Devices for Medium Voltage Wind Energy Conversion, Symposium on IEEE Power Electronics and Machines in Wind Applications, PEMWA2009, Lincoln, NE, 24–26 June 2009. 11. Miller J.M., McCleer P.J., Cohen M. ‘Ultra-capacitors as energy buffers in a multiple zone electrical distribution system’. Presented at the Global PowerTrain Conference, Advanced Propulsion Systems, Ann Arbor, MI, 23–26 September 2003. 12. Gay S.E., Gao H., Ehsani M. Fuel Cell Hybrid Drive Train Configuration’s and Motor Selection, Advanced Vehicle Systems Research Group, Texas A&M University, 2002 Annual Report, paper 2001–02.

118 13.

14.

15.

16. 17.

18.

19. 20. 21.

22. 23.

24. 25.

26. 27.

28.

Propulsion systems for hybrid vehicles Miller J.M., Stefanovic V.R., Levi E. ‘Prognosis for 42 V integrated starter alternator systems in automotive applications’ IEEE EPE 10th International Power Electronics and Motion Control Conference, Cavtat & Dubrovnik, Croatia, 9–11 September 2002. Richard D., Dubel Y. ‘Valeo StARS technology: A competitive solution for hybridization’. Presented at the IEEE Power Conversion Conference, PCC2007, Nagoya, Japan, 3–5 April 2007. Leonardi F., Degner M. ‘Integrated starter generator based HEV’s: A comparison between low and high voltage systems’. Presented at the IEEE International Electric Machines and Drives Conference, IEMDC2001, MIT, 3–5 June 2001. Electric Power Research Institute. Comparing the Benefits and Impacts of Hybrid Vehicle Options, final report no. 1000349, July 2001. Gelb G.H., Richardson N.A., Wang T.C., Berman B. An Electromechanical Transmission for Hybrid Vehicle Power Trains – Design and Dynamometer Testing, Society of Automotive Engineers, paper no. 710235, Automotive Engineering Congress, Detroit, MI, 11–15 January 1971. Kamichi K., Okasaka K. Hybrid System Development for GS450h, The 22nd International Electric Vehicle Symposium, EVS22, Yokohama, Japan, 23–26 October 2006. Tech briefs, SAE Automotive Engineering International, Tototrak, February 2003, pp. 65–66. Available from http://www.sae.org. Tech briefs, SAE Automotive Engineering International, May 2003, pp. 10–11. Available from http://www.sae.org. Bracke W. ‘The present and future of fluid power’. Proceedings of the Institution of Mechanical Engineers, Part I: Journal of Systems and Control Engineering, 1993, vol. 207, pp. 193–212. Bloxham S.R. ‘Off-grid electro-link power boost system’. Personal conversations with Mr Steve R. Bloxham in July 2002. Louches T. ‘PaiceSM HyperdriveTM, its role in the future of powertrains’, Presented at the Global PowerTrain Conference, Advanced Propulsion Systems, Ann Arbor, MI, 24–26 September 2002, pp. 86–94. MIT Technology Review, 5 August 2008. Weir R., Nelson C. US Patent 7,466,536 B1. Utilization of PET plastic with composition modified barium titanate powders in a matrix that allows polarization and use of integrated circuit technologies for production of ultralight weight and ultrahigh energy storage units, EESU’s, 16 December 2008. Miller J.M. Personal conversations with Peter Jeffries on petrol electric drive train concept, 1999–2000. Shafai E., Geering H.P. Control Issues in a Fuel-optimal Hybrid Car, International Federation of Automatic Control, IFAC 13th Triennial World Congress, San Francisco, CA, 1996. Guzzella L., Wittmer Ch., Ender M. Optimal Operation of Drive Trains with SI-Engines and Flywheels, International Federation of Automatic Control, IFAC 13th Triennial World Congress, San Francisco, CA, 1996.

Chapter 3

Hybrid power plant specifications

The vehicle power plant is designed to deliver sufficient propulsion power to the driven wheels to meet performance targets that are consistent with vehicle brand image. The previous chapters described how conventional engines and electric drive systems are matched to meet performance and economy targets. In this chapter, we continue to evaluate the matching criteria between combustion engines and ac drives for targeted road load conditions. The reader is no doubt aware of the various powertrain configurations available in the market place, including small in-line 3- and 4-cylinder engines with ISA type ac drives matched to the driveline with 5-, 6- and now 7-speed manual or automatic transmissions or even with continuously variable transmission engines such as V6 and V8 with inherently higher torque are typically matched to the driveline with 3- and 4-speed transmissions. At the high end, V10, V12 and even V16 engines with their available torque ranging from 350 to nearly 1,400 Nm explain why such drivelines can pull ‘tall’ gear ratios. Some examples of this are as follows: the Daimler Chrysler V10 Viper engine is an aluminium block 8.3 L, overhead valve (OHV), 10-cylinder power plant rated 373 kW (500 PS) and 712 Nm torque. The Jaguar XJ-S V12 is a 5.3 L, 12-cylinder power plant rated 208 kW (284 PS) with 415 Nm torque at 2,800 rpm. General Motors Corp. (GM) in January 2003 introduced its Cadillac Sixteen with a V16 aluminium block engine. The 13.6 L, OHV, V16 delivers 746 kW (1,000 PS) and 1,356 Nm of torque at just 2,000 rpm. The XV16 Cadillac engine has a mass of 315 kg and is designed to operate with cylinder deactivation. Cylinder deactivation means the V16 engine can run on eight or as few as four cylinders delivering an impressive 20 mpg fuel economy in the 2,270 kg GM flagship vehicle. In 2009 Chevy introduced the 2010 Camaro SS designed for auto enthusiasts and performance aficionados (Figure 3.1). It uses the Zeta platform having improved torsional stiffness to house the LS3 6.2 L V8 engine rated 426 hp (318 kW) and Tremec TR6060 6-speed automatic transmission (Table 3.1). In the following sections the various trade-offs between power plant torque and power rating are illustrated in regard to transmission selection and vehicle performance. To further illustrate the process, suppose the power plants described above are placed into passenger vehicles of size and weight recommended by the manufacturers. Table 3.2 shows the specifics of the vehicle propulsion system. Furthermore, because of lack of transmission data, the driveline gearing in low gear

120

Propulsion systems for hybrid vehicles 2010 Chevy Camaro SS Zeta platform Engine: 426 hp (318 kW), 570 Nm Transmission: 6L80E six speed Trans. Input shaft torque: 553 Nm tz60 = 5.0 s Standing ¼ mi: 13.38 s Speed in ¼ mi: 105.5 mph

Figure 3.1 The 2010 Camaro SS (SAE Automotive Engineering International, October 2009, pp. 29–31) Table 3.1 Specifications of the Chevy Camaro SS Transmission 6L80E 1st 2nd 3rd 4th 5th 6th Reverse Final drive

Vehicle specifications

4.03 2.36 1.53 1.15 0.85 0.67 3.06 3.27

Engine Tyres, rw = Curb mass Aerodynamic drag Frontal area Roll resistance Max engine speed Max veh. speed

kW P275/40R20 Mv Cd Af Crr neng VWOT

318 0.364 m 1,754 kg ~0.25 2.37 m2 ~0.009 rpm mph

is taken as the tyre to road adhesion limit. The axle torque breakpoint shown in Figure 3.2 is taken as the wheel speed representing this traction limit when the tyre to road friction coefficient is 0.85. In comparison, the Jaguar XJ-S (3-speed automatic: first 2.5:1, second 1.5:1 and third 1.0:1 with 2.88:1 final drive) has a driveline gear ratio of 7.2:1 in low gear. In the second row of Table 3.2, the Jaguar XJ-S is taken as the reference point since the gearbox ratios are known and the larger engines are then used to replace the production V12 to determine their effect on driveline matching. The immediate difference is the fact that driveline gear ratio in low gear (maximum ratio) quickly trends towards an overall ratio of 2:1. The second Table 3.2 Characterizing the vehicle power plant Engine Vehicle Engine Engine Gear Gear Traction Axle Speed Max. 0–60 type mass power torque ratio ratio limited torque low vehicle time (s) (kg) (kW) (Nm) high low force (N) limit gear speed (Nm) V10 V12 V16

1,800 1,900 2,270

373 208 746

712 415 1,356

2.88 3.8 7,500 2.88 7.2 7,914 2.1 2.23 9,454

2,370 2,500 2,787

107 58 180

200 162 260

8.6 7.2 8.4

Hybrid power plant specifications Power plant capability reflected to axle

2,500 Axle torque (Nm)

2.37096 ×

121

103 2,000

T(x) 1,500

1.138061 × 103 1,000

0 0

(a)

Axle torque (Nm)

3,200

50 x Angular speed of axle (rad/s)

100 100

Power plant capability reflected to axle 3,000

T(x) 2,000 1,000 717.12 0 0 0

(b)

50 x Angular speed of axle (rad/s)

100 100

Power plant capability reflected to axle Axle torque (Nm)

3,400 3,400 2,550 T(x) 1,700 850 0

0 0 0

(c)

50

100 x Angular speed of axle (rad/s)

140

Figure 3.2 Vehicle power plant torque–speed capability for large engines: (a) V10 engine rated 373 kW and 3.33:1 driveline gear ratio, (b) V12 engine rated 208 kW and 7.2:1 driveline gear ratio and (c) V16 engine rated 746 kW and 2.23:1 driveline gear ratio observation is that as power plant torque rating increases, the low to high gear ratio also quickly converges to the value of 2:1. The third observation from this exercise is that as power plant torque rating (maximum power rating) increases, the gear

122

Propulsion systems for hybrid vehicles

shift ratio coverage (ratio of low to high gear) for constant power speed range decreases dramatically. In summary, Figure 3.2 highlights a very important fact in propulsion system sizing and driveline matching: As power plant rating increases for the same application, the need for large gear shift ratio coverage decreases, fewer gear steps are necessary, and power plant torque is sufficient to meet vehicle acceleration targets to very substantial speeds. Figure 3.3 illustrates the acceleration performance for vehicles equipped with these large engines. Data for Figure 3.3 are taken from a vehicle simulation using fourth-order Runge–Kutta integration of the net tractive force available at the driven wheels while accounting for all road load conditions. Figures 3.2 and 3.3 convey a strong message about setting vehicle propulsion targets. For the conventional vehicle listed and for three very different high performance engines, it can be seen that vehicle acceleration is only loosely connected to gross power plant rating but very intimately tied to powertrain matching. A huge engine does not even require a transmission – simply connect it to the wheels and it has sufficient torque to launch the vehicle with more than

Speed (mph)

202.807781

Vn,1

Vehicle acceleration

300

225 1 0.447 150 75 0

0

0

10

20

30

0 (a)

Speed (mph)

165.86241

Vn,1

40 Vn,0 Time (s)

50

60

70

80 80

Vehicle acceleration

200

150 1 0.447 100 50 0 0 0 0

(b)

10

20

30

40 Vn,0 Time (s)

50

60

70

80 80

Figure 3.3 Vehicle acceleration performance for three engine types: (a) vehicle with V10 engine, 0–60 time 8.6 s and maximum speed 200 mph; (b) vehicle with V12 engine, 0–60 time 7.2 s and maximum speed 162 mph; (c) vehicle with V16 engine, 0–60 time 8.4 s and maximum speed 260 mph

Hybrid power plant specifications

Speed (mph)

263.35658

Vn,1

123

Vehicle acceleration 400

300 1 0.447 200 100 0 0 0 0

10

20

30

(c)

40 Vn,0 Time (s)

50

60

70

80 80

Figure 3.3 Continued adequate acceleration and has sufficient power to sustain high speed cruise. However, even a large engine can be stalled out from standstill without proper gearing and clutch. Consider a Top Fueler dragster with a 6,000 hp engine: it requires a 5-speed transmission. With large displacement engines, the transmission ratios were forced to fall within specific bounds due to tyre adhesion limits for the normal production tyres and maximum engine rpm at cruise. It was shown that as engine capability increased, the demand for wide gear shift ratio coverage diminished dramatically because either the tractive force would be too great for the tyre to road friction or the ratio would be too great for the engine red line limit. The Cadillac V16 turned out to require virtually no gear shifting whatsoever due to its extreme torque. Of course, changing to higher road adhesion tyres will change this scenario, and those familiar with the dragster class of vehicles know that 5-speed shifting transmissions are needed in a vehicle having a 6,000 hp engine. More economical engines such as in-line V4s, V6s and even V8s have crankshaft torque ratings in the range of less than 100 Nm to perhaps 350 Nm. Because the vehicle performance targets are not set differently for smaller passenger cars, the demand for wider gear shift ratio coverage increases as engine torque rating decreases. This is necessary in order to meet vehicle performance targets, particularly the 0–60 mph acceleration time and the 50–70 mph passing times. The following sections will elaborate more on these points. Example 1: A Top Fueler dragster can accelerate through the quarter-mile in 3.1 s reaching a top speed of 310 mph. Suppose the dragster is controlled near peak tyre slip. How much acceleration does the driver experience in the 3 s? Solution: The driver is accelerated along with the vehicle and so experiences the same acceleration 1 d ¼ V0 t þ at2 ¼ ð1;609Þ m; 4



402:25 m ¼ 41:86 m=s2 3:12

The driver experiences an acceleration of 41.86 m/s2 or 4.27 g.

124

Propulsion systems for hybrid vehicles

Example 2: Assuming this acceleration is constant, what is the dragster’s speed when d = 100 ft? Solution: Using the relations given in Example 1 the time for the dragster to cover 100 ft can be found to be 1.205 s. Speed is the integral of acceleration, so the velocity at T = 1.205 s is given by rffiffiffiffiffiffi rffiffiffiffiffiffiffiffiffiffiffi 2d 60:8 ¼ 1:205 s; ¼ T¼ a 41:86

0 V100 ¼

ðT 0

adt ¼

ð 1:205 0

41:86dt ¼ 50:45 m=s

The dragster’s speed at 100 ft is therefore 50.45/0.447 = 112.8 mph!

3.1 Grade and cruise targets In order to assess the vehicle performance on grades and during cruise, a representative vehicle is selected to carry out the analysis. Here, the Ford Focus is used since it represents a mid-sized vehicle capable of comfortably carrying four passengers and meeting customer expectations on performance and economy. For this class of vehicle, 0–60 mph acceleration performance in the 9–12 s range is adequate. Table 3.3 summarizes pertinent attributes for the Focus 5-door, 4-passenger, mid-size passenger car. Table 3.3 Passenger vehicle attributes Vehicle attribute

Value

Curb weight, kg Gross vehicle weight, kg (fully loaded vehicle limit) Frontal area, m2 (length: 4.178 m  height: 1.481 m – ground clearance) Aerodynamic drag coefficient, Cd Acceleration time, 0–60 mph, seconds (0–100 kph) with 1.8 L Zetec (4 V, DOHC, 85 kW at 5,500 rpm, 160 Nm at 4,400 rpm) Elasticity, 30–60 mph, seconds (in fourth gear for passing, lane changing) Maximum vehicle speed, mph/kph Fuel consumption, L/100 km (on NEDC cycle) Specific fuel consumption, min: 230 g/kWh Emissions, g CO2/km Gsrc, gear shift ratio coverage (1st/4th) transmission, 4-speed automatic 1st 2.816:1, 2nd 1.498:1, 3rd 1.000:1, 4th 0726:1, FD 3.906:1 rw, m, rolling resistance of P185/65 R 14 tyres. Tyre code: P = passenger, Wsection = 185 mm, c = aspect ratio = 65%, R = speed rating, rim OD = 14 in.

1,077 1,590 2.11 0.335 10.3 13.5 124/198 7.5 181 3.88 0.285

The dynamic rolling radius of the vehicle’s tyres is computed according to (3.1), given the adjustment factor taking static unloaded radius to dynamic loaded

Hybrid power plant specifications

125

radius is e = 0.955. Equation 3.2 defines the derivation of section height from the given section width and tyre aspect ratio: rw ¼ 0:5eðODrim þ 2Hsection Þ Hsection ¼

ð3:1Þ

c Wsection 100

ð3:2Þ

From the data given in Table 3.3 and from (3.1) and (3.2) the tyre rolling radius is found to be rw = 284.6 mm. For the given data the 1.8 L Zetec is modelled as a torque source with breakpoint at 85 kW corresponding to a vehicle speed of 26.2 mph according to the definitions in (3.3), where driveline efficiency is the composite of automatic transmission efficiency, propeller shaft plus CV joints, and final drive, all approximately equal to 0.85: hdl Pe Gr Te Gr ¼ z1 zFD hdl ¼ hAT hprop hFD wa rw V¼ 0:447

wa ¼

ð3:3Þ

When these data are plugged into the vehicle dynamic simulation model to account for driveline tractive effort and road load, the chart shown in Figure 3.4 results. In this chart aerodynamic drag is taken at the standard 200 m elevation, still air and level grade.

Speed (mph)

111.085664

Vehicle acceleration

150 112.5

Vn,1

1 0.447

75 37.5

0

0

0 0

10

20

30

40

50

60

70

Vn,0

80 80

Time (s)

Figure 3.4 Vehicle performance on level grade, no headwind and 500 W accessory load

In Figure 3.4 the vehicle accelerates to 60 mph in just under 10 s, very close to the listed elapsed time. The vehicle’s maximum speed is realistic at 112 mph with the driver as the only occupant.

126

Propulsion systems for hybrid vehicles

To see how this vehicle performs under cruise conditions of 55 mph in fourth gear (24.6 m/s) and for the consequent load imposed on the engine, we start with the road load equation: Fr ¼ R0 mv g cosðaÞ þ 0:5rCd AðV   %grade a ¼ arctan 100

V0 Þ2 þ mv g sinðasÞ

ð3:4Þ

Following this, the resultant loading at the driven wheel axle is reflected back to the engine’s crankshaft, or engine plus hybrid M/G, output shaft. In (3.4), grade is converted to an angle in radians, wind speed (headwind or tailwind) in m/s (mph in (3.1.5)), and the remaining parameters are listed in Table 3.3. Equation (3.5) explains the procedure: Fr rw hdl z4 zFD 0:447z4 zFD we ¼ V rw

Te ¼

ð3:5Þ

Taking the cruise conditions as 55 mph, driving the Focus vehicle reflects a load of Te = 38.4 Nm and we = 244 rad/s (2,330 rpm) at the crankshaft. Figure 3.5 shows the road load for 0% and 7.2% grades. At 33% grade, the vehicle can sustain 23 mph by completely loading its 85 kW ICE. These data are shown scaled in Figure 3.5 to illustrate that it would take nearly 600 kW of engine power to sustain top speed over a 33% grade. Road load, grade, no headwind

Tractive effort (N)

5.862618 6

4

Fe33(k) × 10–5

Fr7(k) × 10–3

2

0

Fr0(k) × 10–3 0 0 0

20

40

60 80 k Vehicle speed (mph)

100

120 120

Figure 3.5 Vehicle cruise performance on level terrain and on 7.2% grade Grade climbing presents a particular challenge to hybrid vehicles because the main source of sustainable propulsion power is the ICE and not the hybrid battery.

Hybrid power plant specifications

127

3.1.1 Gradeability Hybrid power plants must have their heat engines sized to meet sustained performance on a grade. As we saw above for the Focus vehicle, a 33% grade consumes its entire engine output in order to sustain 23 mph. If there was some headwind or more vehicle occupants, this would not be possible in high gear. The conclusions above have taken into account that the driveline will downshift as appropriate to move the full engine power to these lower speeds.

3.1.2 Wide open throttle Analysis of vehicle propulsion under wide open throttle (WOT) conditions is generally the approach taken to illustrate the best vehicle performance in acceleration times and passing. When the vehicle’s accelerator pedal is pressed completely to the floor, the engine controller senses the demand for full performance and autonomously declutches the vehicle’s air conditioner compressor by deenergizing its electromagnetic clutch. In vehicles with controllable fans and water pumps, there may be some further gains by restraining the power consumed by these ancillaries. Example 3: Using the 2010 Camaro SS specifications given in Table 3.1: (a) compute the maximum tractive force at the rear wheels when accelerating from standstill and (b) compute the ratio of traction limited force to the maximum engine delivered traction force from part (a). Given: The wheel base L of the Camaro SS is specified as 2,852.4 mm and the weight balance is 52%/48% (front/rear). Solution: ● Assuming the Camaro SS launches in first gear and for the specifications listed in Table 3.1, the maximum engine delivered tractive force to the rear wheels is Fteng ¼ ●

ðg1st gfd Þ ð4:03Þð3:27Þ me ¼ 553 ¼ 2;0020:6 N rw 0:364

From Chapter 1 the front weight balance (B/L) = 0.52 and the rear (A/L) = 0.48, where A is the moment arm from vehicle centre of gravity (cg) to front axle centreline and B is the moment arm from vehicle cg to rear axle centreline. Therefore, A = 1,369.2 mm and FNr = gMr = g(A/L)(Mv + Mpass) = 9.8066(0.48) (1,754 + 75.5) = 8,611 N. Given that the maximum slip coefficient is 0.85, the traction limited force, Ftlim = 0.85(FNr) = 7,320 N: Ftlim 7;320 ¼ ¼ 0:365 Fteng 20;020:6 Therefore, launch in first gear demands only 36.5% of available tractive force.

128

Propulsion systems for hybrid vehicles

3.2 Launch and boosting Virtually the complete impression of vehicle performance is gleaned during the first couple of seconds of a brisk take-off and acceleration. How smoothly the vehicle accelerates, whether or not shift events are noticeable and if any driveline shudder or vibration is present contribute to the overall impression. Beyond the initial launch, and particularly when the automatic transmission torque converter is transitioning out of torque multiplication, the benefit of hybrid boosting becomes noticeable. In ISG type of direct drive systems having power levels less than 10 kW, the boost impact is not noticeable above 3,000 engine rpm. But up to that speed, boosting by the electric M/G is noticeable and does benefit vehicle acceleration because engine output torque is augmented. We saw in the introduction to this chapter how adding engine torque to the driveline dramatically improved the total vehicle capability, provided the correct matching is employed. If the transmission and final drive gear ratios are too ‘tall’, the acceleration will not be as brisk, even with torque augmentation.

3.2.1

First two seconds

The most noticeable launch feel occurs during the first two seconds when the automatic transmission torque converter is delivering double the engine torque to the driveline. Manual transmissions generally require 18% higher transmission gear ratios to ‘make up’ for the torque boost of the hydraulic torque converter. The initial performance feel creates an image of the vehicle in the driver’s mind, an image that defines the particular vehicle, what some would call brand DNA.

3.2.2

Lane change

Another measure of vehicle performance is lane change during passing and the attendant need for acceleration varies from either 30 to 60 mph or from 50 to 70 mph depending on geographical location and driving habits. To illustrate the vehicle’s performance during passing manoeuvres, we take the same Focus five door and compare its WOT performance in terms of time to accelerate for both of the speed intervals noted. Table 3.4 summarizes the findings of running the simulation for three cases: Case (1) 0% grade and single occupant, Case (2) 7% grade and single occupant and Case (3) again 0% grade but with four occupants. Table 3.4 Vehicle lane change and passing manoeuvres Manoeuvre

30–60 mph time (s)

50–70 mph time (s)

Case 1: 0% grade, 1 occupant Case 2: 3% grade, 1 occupant Case 3: 0% grade, 4 occupants

6.1 11.6 7.0

5.9 21.0 7.2

Hybrid power plant specifications

129

Table 3.4 presents some interesting data. Again, performance on grade is much more demanding than adding occupants as can be seen from the passing manoeuvre times. The change from one to four occupants for the same manoeuvre makes only a 17% increase in acceleration times, but climbing just a 3% grade results in a very noticeable 90% and 256% increase in times.

3.3 Braking and energy recuperation The performance of vehicle hybrid propulsion systems is strongly dependent on the type of brake system used. The energy recuperation component of fuel economy gain depends on (1) the hybrid M/G rating and (2) the types of regenerative brake systems employed. It is far more important to implement active brake controls on rear wheel drive versus front wheel drive hybrid vehicles when regenerative braking is employed. Vehicle rear brakes tend to lock and skid far easier than front wheels due to normal weight balance allocation and dynamic weight shifting during braking. Once the rear wheels lock up, the vehicle anti-lock braking system (ABS) will engage and start to modulate the brake line pressure at approximately 15 Hz. Engagement of ABS pre-empts regenerative brakes in a hybrid vehicle regardless of architecture. There are several versions of regenerative brakes available depending on the level of energy recuperation anticipated. Series regenerative braking systems (RBS), as the name implies, engage the electric M/G first in generating mode. Then, if the brake pedal is depressed further, and/or faster, the brake controller adds in the vehicle’s service brakes to gain a more rapid deceleration. Lastly, if the brake pedal is depressed hard, the ABS controller engages and controls braking using the service brakes only. Parallel RBS is the system most commonly employed in mild hybrids. Parallel RBS does not require electrohydraulic braking (EHB) or electromechanical braking (EMB) systems, but uses both M/G braking and service brakes in tandem. An algorithm in the M/G controller proportions braking effort between regenerative and service brakes. Split parallel RBS is another transitional system wherein the service brakes are not engaged for low effort braking but the hybrid M/G is. Again, ABS pre-empts any RBS actions. Interactions with vehicle longitudinal stability programmes, such as interactive vehicle dynamics (IVD) (as used in NA), electronic stability programmes (ESP) (as used in Europe) and vehicle stability controls (VSC) (as used in Asia-Pacific), are all coordinated by the vehicle system controller. The following subsections elaborate on each of the RBS discussed above.

3.3.1 Series RBS Series RBS introduces electrical regeneration sequentially with the vehicle’s service brakes in proportion to the brake pedal position. Figure 3.6 illustrates how the

130

Propulsion systems for hybrid vehicles

Brake effect (g)

0.45

0.30

Service brakes engaged

0.15 M/G regeneration 0 0

25

Compression braking 50 75 100 Brake pedal position (%)

Figure 3.6 Series RBS at vehicle speed, V total brake force is proportioned in the series RBS configuration. When the brake pedal is depressed, the vehicle experiences velocity retardation in proportion to engine compression braking. This is a very mild deceleration effect and one that drivers expect. As the brake pedal is further depressed, an algorithm computes the M/G braking torque (regen) so that the vehicle kinetic energy recuperated and sent to the traction battery is controlled to mimic normal foundation brake feel. If the brake pedal is depressed further, the deceleration effect is more noticeable, and at some pre-defined point, the service brakes are blended in with M/G torque. Had a third axis been included in Figure 3.6, it would show how overall brake effort is apportioned between M/G torque and service brakes as the vehicle decelerates. As vehicle speed decreases, and accounting for transmission down shifts, there will come a point when M/G efficiency is too low for regeneration so that it will be commanded off and only service brakes left in effect. The strictly linear, and snapshot, nature of Figure 3.6 does not convey this information. Series RBS requires active brake management on all four wheels so that the total braking effort is coordinated. For example, the M/G may impose braking force on the front axle wheels only and none on the rear axle. Therefore, a hydraulic system would be necessary to actuate the rear brakes in the proper proportion to front axle brakes so that vehicle longitudinal stability is maintained. Brake coordination is a complex function of brake pedal position, rate of pedal application and vehicle speed. Properly coordinated front–rear braking force optimizes stopping distance without loss of tyre adhesion. Series RBS is typically implemented with EHB hardware. EHB consists of the hydraulic control unit that interfaces to the driver foot. The second component is the electronic control unit that manages brake cylinder pressures and front–rear axle brake balance. It is insightful to examine how vehicle speed of a hybrid changes when commanded to generate at some fixed power level into the RESS (i.e. the battery pack). The capability for constant power deceleration is unique to hybrid electric vehicles

131

Hybrid power plant specifications

and will be examined here for completeness. Consider again the hybrid vehicle on level terrain and no wind. The propulsion power for this case is PðV Þ ¼ MV V_

gMCrr V

1 rCd Af V 3 2

ð3:6Þ

where M = (Mv + Mpass) and let Cr = gMCrr and Ca = ½CdAf. Furthermore, note that the propulsion system driveline power encounters at least three dissipative elements to its flow, the losses due to electrical to mechanical energy conversion in the main traction motor, MG1, the power inverter losses and the loss due to gear meshing in the final drive. Accounting for these means that to sustain a constant regenerative power, Pb, into the battery requires a power Pv at the wheels of Pv ¼

Pb

ð3:7Þ

hdl hmg1 hINV

Let the regen power be very modest, and within the M/G regeneration region of Figure 3.6 at a level of Pv = 500 W. Later in Example 4 the power will be increased dramatically to illustrate the difference. For this case one must solve (3.6), and this is challenging because there is no closed form solution. One approach would be to refer to tables of integrals,1 but this gives time as an implicit function of velocity, V, which is the sought-after dependent variable. In this case the usual approach is to employ numerical integration. A numerical solver for (3.6) is shown in Figure 3.7 using the values for a Camry Hybrid listed earlier and for the case of Pv = 500 W at the wheels. The power set point for constant power regen is set in the lower right of Figure 3.7 in the box showing an example case Pv = 29,200 to be discussed in the V_Dot

Simulation of Camry Hybrid vehicle speed to regen 25 kW into NiMH battery

GAIN

1 Velocity and acceleration outputs Cr_by_M

SUM2

SUM1

INTG1

CONST

I

0.0785

29.06

Vel GAIN

1 Mass MUL1

MUL3

CONST

MUL2

1748.5 CONST1

CA_by_M

SUM3

CONST

0.000224

MUL4

Reciprocal Rec

Given: Driveline efficiencies: eta_dl = 0.93 eta_mtr = 0.95 eta_INV = 0.97

0.2

Regen power setpoint Power_Val

Figure 3.7 Numerical integration of (3.6) in Simplorer 1

See, for example, equation 2.172 in Reference 1.

CONST

CONST

29200

132

Propulsion systems for hybrid vehicles

example. For a very light load of 500 W regen, the vehicle speed versus time and deceleration are shown here. Notice that after 169 s of deceleration at this light load, the vehicle speed, once it reaches just under 5 m/s, collapses very quickly. The briskness of vehicle speed deceleration near 165 s is evident in the plot of V_Dot. Since mechanical ‘jerk’ is defined as the rate of acceleration or deceleration change, the fast upsweep in V_Dot for the Camry Hybrid case noted in Figure 3.8 is apparent. When the vehicle kinetic energy can no longer sustain the demanded constant power, its rate of deceleration, or jerk, becomes very large. All hybrid powertrains must release the demand for brake energy recovery before this threshold, and most will relax brake recuperation power as vehicle speed drops below 10 mph and disengage completely at 5 mph. V_dot (m/s 2)

Vehicle speed (m/s) 35.00

5.00

30.00

4.00 V_Dot.VAL

Vel.VAL

25.00 20.00 15.00

3.00 2.00

10.00 1.00

5.00 0

0

0

25.00

50.00

75.00

0

100.00 125.00 150.00 175.00

25.00

50.00

t

75.00

100.00 125.00 150.00 175.00 t

Figure 3.8 Camry Hybrid deceleration characteristics for Pv = 500 W Example 4: A Camry Hybrid described in Example 3 with NiMH energy storage pack rated 25 kW will use series regen braking to slow from an initial vehicle speed of 65 mph (29.06 m/s). Determine to what vehicle speed it can sustain a battery charging power of 25 kW before the onset of vehicle jerk. Solution: The numerical integration shown in Figure 3.7 is used. In ANSYS/Ansoft the numerical integration routine is an adaptive trapezoidal method and is used to solve this example. The following results for vehicle speed, V, and deceleration, V_Dot, are found: V_dot (m/s2)

Vehicle speed (m/s) 10.00

35.00 30.00

8.00 V_Dot.VAL

Vel.VAL

25.00 20.00 15.00 10.00

6.00 4.00 2.00

5.00

0

0 0

5.00

10.00

15.00 t

20.00

25.00

0

5.00

10.00

15.00

20.00

25.00

t

At Tf = 21 s the vehicle jerk becomes very excessive and V ~ 2.5 m/s (5.5 mph).

Hybrid power plant specifications

133

3.3.2 Parallel RBS A parallel RBS is easier to implement than series RBS because full EHB is not required. As Figure 3.9 illustrates, the parallel system immediately activates the vehicle service brakes anytime the brake pedal is depressed. An algorithm is needed to blend M/G torque with service brake force so that the total deceleration effect is smooth and seamless to the driver. Front–rear brake coordination is similar to that for series RBS so that vehicle stability is maintained.

Brake effect (g)

0.45

0.30

Service brakes

0.15 M/G torque 0 25

Compression braking 50 75 100 Brake pedal position (%)

Figure 3.9 Parallel regenerative brake system With parallel RBS, the fuel economy benefit is not as pronounced as for series RBS because some fraction of vehicle kinetic energy is always dissipated as heat rather than used to replenish the battery.

3.3.3 RBS interaction with ABS For normal rate of application of pedal position, the total braking effort in a hybrid vehicle will be as shown in either of Figures 3.6 or 3.9 depending on whether the brake system implemented is series or parallel RBS, respectively. However, if the pressure applied to the brake is brisk, there is a tendency for wheel lock-up and skid. The ABS was introduced as a mechanism to intervene in the braking process should the operator engage the brakes too harshly and cause loss of longitudinal stability as a result of a skid. With ABS the brake line pressure is modulated at a frequency of about 15 Hz so that wheel skid is avoided. Also, braking distances are shorter, vehicle stability is better managed and the tendency to excite vehicle yaw motion is minimized, especially on low mu surfaces (i.e. snow, ice, wet pavement etc.). In all hybrid propulsion systems, the engagement of ABS pre-empts M/G regeneration torque. The M/G is commanded to free-wheel as brake line pressure is modulated by the ABS system. Another important aspect of any vehicle, including hybrid electric, is the stopping distance from some initial velocity, V0. In the case of braking, all the friction loss components contribute to stopping and that includes

134

Propulsion systems for hybrid vehicles

the brakes and aerodynamic losses. The driveline losses are negligible since there is no power flow upstream in the vehicle driveline when the foundation brakes are engaged, regen is disabled, ABS is active and all braking energy is dissipated in the wheel hub brake assembly. For this case (3.6) is written as follows at the level of force components: MV

dV ¼ Fb þ Ca V 2 dx

ð3:8Þ

Since all four wheels are engaged in braking, the total braking force at the wheel tyre patches can be very high. In general, the braking force Fb = mgM and (3.8) can be integrated in closed form as ð0 VdV 2 ð3:9Þ MVdV ¼ ðFb þ Ca V Þdx; SD ¼ M 2 V0 ðFb þ Ca V Þ Example 5: Solving (3.9) compute the stopping distance, SD, for the Camry Hybrid with four passengers on level terrain from an initial speed, V0 = 29.06 m/s, when tyre coefficient of friction m = 0.85. Solution:

  M ðFb þ Ca V02 Þ ln SD ¼ Fb 2Ca   ð1;673 þ 4  75:5Þ 16;462 þ 0:392ð29:062 Þ ln ¼ 50:15 m ¼ 2ð0:392Þ 16;462

and Fb = (0.85)(9.8066)(1,975) = 16,462 N. The Camry Hybrid will therefore stop at 50.15 m (165 ft) from 65 mph on level ground.

3.3.4

RBS interaction with IVD/VSC/ESP

The previous sections have described how RBS, a necessity for hybrid functionality, reacts with the vehicle’s longitudinal stability functions (i.e. ABS) during extreme manoeuvres. Also, the introduction of EHB hardware into the vehicle platform brings with it additional stability features. In both CVs and HVs having EHB (in the future EMB), it is possible to further enhance overall stability during vehicle handling manoeuvres. This author has test driven EHB and EMB equipped test cars on a handling course to more fully appreciate the benefits of dynamic stability programmes. In this series of tests [2], vehicles equipped with outriggers are driven at speeds of 50–60 mph on a marked course that forces a brisk lane change manoeuvre on wet pavement. The active stability programmes (IVD, ESP or VSC algorithm) initiate appropriate control of vehicle throttle and brakes (independent control is possible) so that manoeuvre induced yaw motion is damped. An EHB or EMB system

Hybrid power plant specifications

135

modulates the out-board wheels, slowing them down, so that oscillatory motion is avoided. Regardless of whether the vehicle is a low cg passenger vehicle, or a higher cg SUV, the stability programme reacts far faster than a human operator and corrects the yaw motion. With the system disabled, each lane change manoeuvre at the same speed resulted in a complete loss of vehicle control and some very aggressive side skids.

3.4 Drive cycle implications The US Federal Urban Drive (FUD) cycle was created during the early years of battery electric vehicle (BEV) development to model urban driving conditions (Figure 3.10).

Speed (rad/s)

700 600 500 400 300 200 100 1, 0 08 1, 8 15 1, 6 22 1, 4 29 1, 2 36 0

2

02 1,

4

95

6

88

8

Time (s)

81

0

74

2

68

4

61

6

54

8

47

0

40

2

34

4

27

8

20

13

0 68

0

Figure 3.10 FUD standard drive cycle used for BEV development (wheel speed) The FUD cycle is the first 1,369 s of the Federal Test Procedure FTP75. FUD cycle represents an urban drive of 7.45 mi at an average speed of 19.59 mph.

3.4.1 Types of drive cycles A great number of drive cycles have been developed to mock-up the driving habits of large populations of drivers in particular geographical areas. The main drive cycles of interest to hybrid propulsion designers are the Environmental Protection Agency (EPA), city and highway cycles used in North America. In Europe the New European Drive Cycle (NEDC) is used extensively. In Asia-Pacific, and particularly Japan, the 10-15 mode is used almost always. There are other related, but revised for some particular attribute, drive cycles, for example US-06 or HWFET for highway fuel economy test. There are cycles that exaggerate the vehicle’s acceleration demands and are known as real world drive cycles. The drive cycles listed in Table 3.5 show distinct geographical character. The first three rows, for example, define the average and maximum speed of how large groups

136

Propulsion systems for hybrid vehicles

Table 3.5 Types of standard drive cycles by geographical region Region

Cycle

Time idling %

Max. speed (kph)

Avg. speed (kph)

Max. accel. (m/s2)

Asia-Pacific Europe NA-City NA-Hwy NA-US06 Industry

10–25 NEDC EPA-city EPA-hwy EPA Real world

32.4 27.3 19.2 0.7 7.5 20.6

70 120 91.3 96.2 129 128.6

22.7 32.2 34 77.6 77.2 51

0.79 1.04 1.60 1.43 3.24 2.80

of the population are assumed to drive their vehicles. Notice that maximum speeds increase going from Japan, through North America to Europe being the highest.

3.4.2

Electric vehicle and regenerative electric vehicle cycles for PHEVs

The resolution of fuel economy testing standards for plug-in and BEVs is undergoing signification updating as noted in Chapter 1. It appears that fuel economy will be replaced with limits on CO2 emissions per mile, for example, > ; : 3  104 IM=VRM 4:5  104

ðArms =mÞ

ð4:9Þ

Notice that the electric loading definitions in (4.9) are in Arms/m, not in peakamperes. This is well defined for sinusoidal machines, but somewhat limited for non-sinusoidal machines such as the VRM. The machine sizing procedure using (4.9), and supported by the definitions of electric and magnetic loading, permits the first approximation to machine sizing to be accomplished without resorting to finite element or detailed computer design since the lamination design has not been fixed at this point, only the major packaging dimensions. The process of working with electromagnetic surface traction as just described is akin to having a detailed lamination design, imposing the electric loading, and then using a magnetics finite element solver to find the flux

172

Propulsion systems for hybrid vehicles

and from this using a post-processing calculation of the Maxwell shear force at the rotor surface (after averaging over the pole pair area). The machine design is further constrained by a mechanical limit – the rotor burst condition. For this constraint it is common design practice to limit the machine rotor tangential velocity to 20 by reading right to left in the plot in Figure 4.18. For example, the minimum length, package OD constrained M/G will have 10 or 12 poles. If fewer

Sizing the drive system

173

1,000

Diameters of stator and rotor

800

600 Dro_lim

Dsoi 400

200

Droi

Dso_lim Drii

0 0

20

40

60

80 100 h FCAS stack length (mm)

120

140

160

Figure 4.18 Variation of hybrid M/G diameters versus stack length poles are used the aspect ratio trends to less than one and if more poles are used the aspect ratio trends to values larger than 3/2 for drum type designs. Drum type electric machine designs have radial flux from the rotor surface on its OD. Axial type machines have axial flux emanating from side faces of the rotor. Even though the hybrid propulsion M/G just elaborated on was pancake shaped, it was nonetheless a radial design. Axial design machines were not considered here, although much the same rationale applies, because in a powertrain the need to restrict axial movement of the disc rotor becomes a major challenge. It would require some elaborate design to restrain the crankshaft end play in an engine after it had been in service for some years to guarantee that the crankshaft did not move axially more than 0.25 mm. If it did, then a fair portion of the axial airgap would be intruded into, with possible rotor impingement on the stator and subsequent damage to the machine. The various types of electric machines used in hybrid propulsion are covered in more detail in Chapter 5.

4.3 Sizing the power electronics All of the electrical power directed to the hybrid propulsion M/G must pass through the power electronics. It has been said that control electronics uses power to

174

Propulsion systems for hybrid vehicles

process information and that power electronics uses information to process power. In this section, we describe how power electronics is sized to match the electric machine to the vehicle ESS. Figure 4.19 is a schematic of a hybrid ac drive system consisting of on-board energy storage, power processing according to control algorithms and traction actuation via the M/G and vehicle driveline.

Rd

Power electronics

Ri Vf , If Vb

w Transmission T Driveline

C’mds

Control electronics Controller, comm. gate drives, pwr supply

Figure 4.19 Schematic of hybrid ac drive system The essentials of ac drive system operation are that power from a dc source such as a fuel cell, battery or ultra-capacitor is converted to variable voltage, variable frequency ac power at the M/G terminals, Vf and If. The M/G then converts this electrical power to mechanical power in the form of a torque and speed at the transmission input shaft, T and w. The power electronics is an electrical matching element in much the same manner that a gearbox processes mechanical power to match the engine to the road load requirements. The power inverter matches the dc source to the mechanical system regardless of torque or speed level, provided these quantities are within its capability. The power processing capability of power inverters is directly related to the dc input voltage available. Higher voltage means more throughput power for the same gauge wiring and semiconductor die area. Figure 4.20 captures the power throughput versus voltage given the system constraint on current of 250 A due in part to cable size, connector sizes and contactor requirements. The reader will appreciate that practical contactors rated in excess of 250 Adc interruption capability are far too bulky and expensive for hybrid vehicle applications. In the case of BEVs, contactors using high energy permanent magnet arc suppression are used effectively to 500 Adc. In Figure 4.20(a) notice that as automotive voltages move to 42 V PowerNet, the sustainable power levels approach 10 kW. For hybrid propulsion the chart illustrates the recommendation that voltages in excess of 150 V are advisable. With recent advances in power semiconductors there is ample reason to move to voltages beyond 300 V, provided ESS does not suffer and complexity is manageable.

Sizing the drive system 80 kW

80 70

175

Vsystem for Isys = 250 A

Capacity (kW)

60 50 40 kW

40 30 20 10 0

10 kW 3.5 kW 14 V (12 V) Conventional

42 V (36 V) Low power hybrid

150 V (144 V) Mild hybrid

330 V (288 V) Full hybrid

System voltage (V)

Figure 4.20 (a) Power throughput capability versus voltage and (b) cable size In the US the National Electrical Code (NEC) article 625 defines all requirements for electric vehicle (PHEV, BEV) charging systems including cabling, connectors, disconnects, equipment ratings, ventilation requirements of charging equipment and much more – for example, the proper locations of charging equipment, cables and ventilation system for residential garage charging station. Specifically, the vehicle supply equipment cable must be type EV, EVJ, EVE, EVJE, EVT or EVJT flexible cable. Article 400, Tables 400.4 and 400.5, of the NEC provides detail on cable gauge (AWG), voltage class (97% efficiency nominal, we see that our distribution system must have >97% efficiency. In Figure 4.31 it is easy to see that this means system voltage levels of several hundred volts (off the chart in Figure 4.31 having a logarithmic abscissa). For this example, a nominal system voltage, Ubnom = 500 V, is assumed. From this we calculate the required number of cells in a series string in the battery module. Equation (4.21) quantifies the procedure: Ub nom Nbc ¼ ð4:21Þ Uc nom

190

Propulsion systems for hybrid vehicles

According to (4.21) we calculate that Nbc = 400 cells in series per string. From this and using (4.18) and (4.19) we can state that the battery module will have maximum and minimum voltage levels of Ub min ¼ Nbc Uc min ¼ 440 V

ð4:22Þ

Ub mx ¼ Nbc Uc mx ¼ 580 V

ð4:23Þ

For a direct parallel combination of battery and ultra-capacitor, there is no isolation between the system bus and the ultra-capacitor so that it must function within the stated voltage swing limits. In this case, as given by (4.20), only 71% droop is permitted. In a lead–acid system, for example, this droop from maximum (2.56 V/cell) to minimum (1.75 V/cell) or a ratio of 0.68. This low percentage of voltage droop will not extract the maximum energy from the capacitor bank. Typically an ultra-capacitor bank can deliver 75% of its energy for a voltage droop of 50%. This fact can be verified by substituting the values given in (4.20) and (4.23) into (4.24): 1 Euc ¼ Cuc ð1  s2u ÞUb2mx ðJÞ 2

ð4:24Þ

In practical systems the working voltage swing factor, su, is dictated by the storage system technology, and the maximum bus voltage, Ub mx, is set by the power electronics device technology. From (4.24), it can be seen that stored energy in the ultra-capacitor is maximized when the swing voltage factor is minimal (i.e. 0) and the bus voltage is as high as possible. When the capacitor energy is determined from the drive scenario, we will use (4.24) to calculate the required capacitance. Our ultracapacitor cells in the resultant string must adhere to the following constraints, just as the battery cells have limitations on charge and discharge potential extremes. For the ultra-capacitor, (4.18) and (4.19) become Uucc min  0 Uucc mx  2:7

ðVÞ ðVÞ

ð4:25Þ ð4:26Þ

There is not a nominal open circuit voltage for the ultra-capacitors, so we define the necessary number of cells per string using the maximum working voltage. Note that there is some tolerance in the maximum working voltage for an ultra-capacitor cell stated in (4.26). Ultra-capacitors can be operated with >3.0 V/ cell for short periods of time. When the voltage across the cell exceeds 4 V, the cell is strongly in overvoltage and likely to rupture from overpressure. The number of ultra-capacitor cells in series for the hybrid bus example becomes Nuc ¼

Ub mx Uucc mx

ð4:27Þ

Sizing the drive system

191

For the values given in (4.23) and (4.26) the required number of series connect ultra-capacitor cells is Nuc ¼

580 ¼ 214:8 2:7

ð4:28Þ

Relative number of events

which means the ultra-capacitor bank will consist of a series string of 215 cells. To complete the hybrid strategy used in the energy management controller, we constrain the engine driven generator to only those periods for which the vehicle engine is required to run. The engine is not running during regenerative braking and during stops. The generator power is zero when the engine is off. During such keyON stops, the engine remains off (idle–stop) and the battery plus ultra-capacitor support the continuous loads. When the vehicle accelerates from a stop, the engine is started and participates in acceleration and recharging of the battery and ultracapacitor packs. The engine control strategy participates with the EMS to maintain the system bus voltage within the prescribed working voltage swing minimum because the ESS requires some unfilled capacity to absorb regeneration energy. It should also be understood at this point that the statistics of the customer drive cycle will have a dramatic impact on storage system component rating. The reason for this is that stopped periods vary considerably for different drive cycles and the storage system must maintain all connected loads during engine off. For passenger cars and light duty trucks this is not so much an issue. For instance, the time spent stopped for existing standard drive cycles has been quantified and illustrated in Figure 4.32.

Truck/police/ fleet

Suburban/light commuter

0

1 Stop time (min)

10

Figure 4.32 Statistics of stop time for various drive cycles (courtesy Ford Motor Co.) The trend line in Figure 4.32 represents a compilation of standard drive cycles for North American passenger vehicles averaged over several geographical locations. For light commuters as suburban dwellers, the trend curve will move left to

192

Propulsion systems for hybrid vehicles

indicate fewer and shorter stops on average per commute. Commercial truck, bus, fleet and police vehicles, on the other hand, experience more stops of longer duration. Our city bus, for example, falls on the rightmost trend line. Regeneration of the braking energy to replenish the storage system is likewise strongly dependent on the drive cycle statistics. For the bus example we are interested in what portion of the fixed loads, or all of it, that can be supported by the recuperated energy from braking only. The un-realized energy of recuperation (i.e. the shortfall in storage system SOC) must be replenished by the engine driven generator. The term ‘un-realized’ is used because energy recovery through regenerative braking is typically limited to 30% of the available kinetic energy due to storage system charge acceptance, driveline mismatch such as lack of transmission torque converter lock-up in lower gears and generator inefficiency at low speeds. Figure 4.33 illustrates regeneration potential for various standard drive cycles for a 42 V PowerNet vehicle that is applicable to our present study. Regeneration duration for 50 V < Vbatt < 55 V

Number of events

20 15

EPA-city EPA-hwy

10

ATDS US06

5 0 0

10

20 Time (s)

30

40

Figure 4.33 Statistics of regeneration duration for standard drive cycles (42 V system) In Figure 4.33 the ATDS is known as ‘real world’ customer usage and gives fuel economy predictions that come closer to matching driver experience. The ATDS cycle approximates our city bus highway portion since there are few stops and longer cruise portions. The EPA city cycle comes closest to our city bus drive cycle because of the frequent stops and relatively long duration of braking time. Other cycles such as the New European Drive Cycle (NEDC) are more representative of European city driving and US06 is more representative of North American commuting. These drive cycles are covered in more detail in Chapter 9. A simulation of the hybrid city bus was performed using the forward modelling technique to track energy expenditures. In this simulation, the engine driven generator is controlled for maximum power only when the propulsion system demands power. Energy recuperation is done according to the charge acceptance limits noted in (4.16) and (4.17). When the battery is unable to discharge (or charge) at the demanded rate, the ultra-capacitor will source or sink the excess power. Given this strategy, and for a 31 Ah, 510 V traction battery, and a 37.6 F ultra-capacitor

Sizing the drive system

193

module, each set to an initial SOC of 80%, it is shown that a charge sustaining mode can be realized for the battery, in the presence of random passenger loading overdrive cycle noted in Figure 4.28. However, the ultra-capacitor bank will be in charge depletion over this cycle for the selected strategy. Figure 4.34 illustrates the city bus drive schedule already defined, the propulsion power, energy storage system SOC, plus the battery and ultra-capacitor pulse power. The drive schedule specifications are listed in Table 4.14 for the 110 min urban and highway fixed route (total stopped time = 48.5 min or 44% of total time). Table 4.14 City bus drive schedule Event #

t1

t2

t3

t4

t5

t6

t7

t8

t9

t10

Event mark (min) Event time (min) Number pass Accel (m/s2) Stop time (s)

12 12 43 0.174 330

26 14 39 0.149 370

40 14 31 0.10 370

47 7 20 0.199 180

62 15 9 0.093 400

72 10 28 0.139 260

84 12 37 0.116 320

95 11 41 0.126 290

103 8 36 0.174 210

110 7 30 0.199 180

Note: The number of passengers, Np, is a random number with mean value according to (4.11).

City bus drive schedule

50 0 0

2,000

–50

6,000

8,000

Time (s) Battery and UC SOC

1.0 SOC

4,000

0.5 0.0 0

2,000

4,000

6,000

300.0 200.0 100.0 0.0 −100.0 0 −200.0

8,000

Time (s)

2,000

4,000

6,000

8,000

Time (s)

100.0 50.0 0.0 −50.0

–0.5

Propulsion power

Pulse power, battery and UC

150.0 Power (kW)

Speed (kph)

100

M/G power (kW)

In Figure 4.34 the battery SOC sags noticeably during the highway and early urban stop–go events because of the extended stop times and the burden of a constant accessory load for cabin climate control and entertainment features.

0

2,000

4,000 Time (s)

6,000

8,000

Figure 4.34 City bus simulation of propulsion power and energy storage system performance For the engine driven generator strategy selected the city bus returns to its starting point with the battery replenished, but with the ultra-capacitor depleted. It is clear from the drive schedule that the second, third and fourth stop–go events

194

Propulsion systems for hybrid vehicles

with high non-idling times result in battery energy drain from SOC = 0.8 to 0.35. Referring to Table 4.14 it can be seen that the first five events have the highest passenger loading, the highest stop times and modest acceleration (and braking). Since the ultra-capacitor in this architecture relies on the availability of braking energy to replenish its SOC, the fact that it becomes fully depleted means that energy balance between the generator, battery and ultra-capacitor is suboptimal. Simply increasing the ultra-capacitor size will not remedy the situation. We saw earlier in (4.16) and (4.17) that the battery charge acceptance is low, so most of the regeneration energy is already being directed into the ultra-capacitor bank. There are simply not enough regeneration events to maintain its SOC. The strategy would therefore require further manipulation to include opportunity charging during decelerations so that instead of shutting off the engine it would continue to run at some lower, and more efficient, power level, with that output being directed into the ultra-capacitor bank. The variations on control strategy of such multiple source hybrid systems are too great to explore here. The salient point is that beyond component sizing based solely on physics there must be a control algorithm designed around anticipated usage and projected passenger occupancy in the case of a city bus, to further refine the system.

4.4.1

Lead–acid technology

The most cost effective secondary storage battery is the flooded lead–acid battery. This technology costs approximately $0.50/Ah for a 6-cell module. Maintenance free, valve regulated, lead–acid, valve regulated lead acid (VRLA), and absorbant glass mat (AGM) lead–acid batteries are capable of higher cycle life than the flooded lead–acid type. The main disadvantage of lead–acid for hybrid vehicle traction application is its low cycle life. Even deep discharge lead–acid batteries such as those used in BEV traction applications are not capable of much beyond 400 cycles (to 80% depth of discharge (DOD)). Table 4.15 lists the differences between BEV and hybrid vehicle batteries. In this table a thin-metal-foil (TMF) lead–acid battery is listed that was developed during the mid-1990s by Johnson Controls and Bolder Technologies Corp. as a very high power (thin electrode) secondary cell. A typical 1.2 Ah, 2.1 V cell in cylindrical package, f22.86  L72.26 mm, has a foil thickness of 0.05 mm, a plate thickness that is less than 0.25 mm and a plate to plate spacing when spiral wound of less than 0.25 mm. The cell ESR is 500,000 units/year – APV). In costing studies the various representations are (1) APV (annual production volume) or actual vehicles produced and sold, and (2) FPV (financial planning volumes) for the more upstream accounting and budgeting to meet specific corporation goals such as CAFE´ and brand image. Table 4.26 Fuel cell stack cost (DOE goal for 50 kW, 500,000 APV) Component

Anode and cathode layers Electrolyte Gas diffusion layers Bipolar plates Gaskets Other Total

Cost

2004

%

$

$/kW

$/kW

50 20 5 15 5 5 100

3,625 1,310 420 1,035 380 280 7,050

75 25 5 20 10 5 140

5 5 5 N/A 5 N/A N/A

Sizing the drive system

235

Costs associated with hybridization are significantly increased by the addition of electric drive subsystems and their supporting components. To illustrate this case we assume a 42 V PowerNet enabled, ISG system installed in a mild hybrid. Table 4.27 is presented here as a cost walk of the ISG system and its supporting subsystems for three specific cases of vehicle power supply: (1) 42 V PowerNet, (2) 150 V hybrid and (3) 300 V or higher voltage hybrid. In all three cases the costs associated with installation of the vehicle power supply are included in the form of battery, wiring harness and thermal management. Table 4.27 presents some interesting insights into the economics of hybridizing a conventional vehicle. At the relatively low power level of 10 kW, the installation costs are nearly equally split between the costs associated with adding the ISG hardware (battery, inverter and machine), and the necessary supporting subsystems for steering, braking and cabin climate control, along with thermal and power management. Increasing the system voltage shifts the relative proportions, but does not change the cost breakdown; it is still virtually an even split between the hybrid technology and subsystems needed to support it. It should also be apparent that had hybridization taken place after x-by-wire functionality was already in production, the installation costs of hybridizing such a vehicle would be cut in half. This is because all the major supporting subsystems would already be in place for steering, braking (including vehicle stability), cabin climate control and thermal management. Table 4.27 Mild hybrid vehicle cost walk (P = 10 kW) Cost walk ISG system Battery Inverter Electric machine Wiring and conn. Subtotal Supporting systems Electric water pump + electric fans dc/dc converter Electric assist steering Electrohydraulic brakes Electric assist air conditioner Subtotal Total

42 V (%)

150 V (%)

300 V (%)

9 21

11 19

14 18

11 2

12 4

11 5

43

46

48

2

2

2

9

10

11

13 16

12 15

11 14

17

15

14

57 100

54 100

52 100

Comments

VRLA MOS – 42 V and 150 V, IGBT at 300 V Asynchronous Special insulation required above 60 V Thermal management components Required in dual-voltage systems Motor on steering rack EHB hardware with ABS cost offset 2 kW electric drive components added

236

Propulsion systems for hybrid vehicles

4.8.2

Weight tally

An assessment of vehicle mass comparing a conventional mid-sized vehicle and its hybridized sister version was carried out by the Electric Power Research Institute (EPRI) and documented in their final report [46]. The hybrid vehicle considered is under the EPRI designation HEV0, which has a downsized engine that is augmented with an M/G rated for an electric fraction, EF = 30%. Table 4.28 is extracted from the EPRI study and modified for this mass budget example. The salient features in Table 4.28 are the following: ● ●

The downsized engine introduces significant mass savings to the total. Hybridization, including the traction battery, adds significant mass that virtually displaces the entire mass gained through engine downsizing.

Table 4.28 Mass budget of CV and HV compared Component

Conventional vehicle (mass, kg)

Hybrid vehicle (mass, kg)

Engine Engine thermal Lubrication Engine mounts and cross members Engine subtotal Exhaust and evaporative system Transmission Alternator Air conditioner compressor Air conditioner condensor Air conditioner plumbing + coolant, mounts Accessory power module Climate control and accessory module subtotal Cranking motor Hybrid M/G Power electronics/inverter M/G and inverter thermal management Electric system subtotal Fuel system, tank + lines 12 V battery Traction battery Battery tray(s) Installation hardware Battery climate control Energy storage system subtotal Total power train Glider (with chassis subsystems) Fuel mass Total curb mass Occupant (1) plus cargo Total vehicle mass

155.6 8.1 7.8 37.7 209.2 41.0 97.9 4.7 6.2 2.2 12.6 – 25.7

87.1 4.7 7.0 15 113.8 31.6 50 – 11.2 2.3 12.6 10.0 36.1

6.1 – – – 6.1 13.4 14.8 – – – – 28.2 408.1 1,053 38.4 1,499.5 136 1,636

– 23.5 5.0 16.6 45.1 9.0 5.0 75.2 7.0 13.5 14.6 124.3 400.9 1,053 27.7 1,481.6 136 1,618

Sizing the drive system ●



237

Fuel tank and conventional vehicle accessory battery (SLI) both represent mass savings in the hybridized vehicle. Gross vehicle weight is only slightly less for the hybrid vehicle.

The bottom line is that hybridizing a conventional vehicle can easily consume all weight reduction actions taken before the hybrid components are installed. Rather than retrofit and modify a conventional vehicle, it would be more appropriate to design a hybrid vehicle (fuel cell vehicle for that matter) from the ground up.

4.9 Exercises Q1:

A1: Q2:

Using the mid-sized vehicle data given in Table 4.1, the result of Example 1 and a simple calculation based on the equation in Example 1, show that the ratio of vehicle peak accelerating power to cruise power is >10:1. 11:1 for this particular set of parameters. For the duty cycle values calculated from the equation in Example 3, determine what the converter output current, Id, and ultra-capacitor current, Ic, are when the battery voltage Ud = 150 V and for a. Motoring mode power of 12 kW. b. Regenerating mode power of 6 kW.

A2:

(a) 162.3 A and (b) 78.9 A.

A plug-in hybrid electric vehicle is charged using Class II equipment supplied from a 240 Vac line and delivers a constant 7 kW over a 12 AWG cable. According to NEC 400.5(B) a 2500 2 wire conducting cable is rated for 31 Arms and the load here is 30 Arms. If the charging equipment is located in a residential garage and the cable is rated 60  C and well insulated (electrically and thermally), then will the cable conductors remain below 60  C in a 30  C ambient after charging for 8 h? Hint: For copper conductors a20 = 0.00427 K 1, specific heat, cp = 385 kJ/kg-K, 12 AWG wire is 29.42 kg/km and 5.21 W/km. Use Q = McpdT for heat input. A3: Yes, the near adiabatic heat input into the copper conductors of the cable is well below the thermal capacity of the total cable mass. Q3:

Q4:

A4: Q5:

A PHEV lithium ion pack of LFP chemistry is rated 15 kWh and is assumed to have 42% of its total elemental lithium participate in energy storage. Using the results of Example 4, calculate how much total lithium metal is needed in this PHEV battery and what is relative equivalent lithium per kWh? 2,642 g (2.64 kg) and 176 g/kWh. Find the energy stored in (J) and (Wh) and calculate the following metrics, SP, SE, PD and ED (specific P & E, and E & P density), given an ultra-capacitor cell having C = 3,000 F, Umx = 2.7 V, M = 0.54 kg and Vol = 0.475 dm3.

238

Propulsion systems for hybrid vehicles Using the following relations: 2 E ¼ 1=2CUmx ;

A5:



2 Umx 4ESRdc

E = 10,935 J (3.04 Wh); P = 6.5 kW; SP = 12 kW/kg, SE = 5.63 Wh/kg, PD = 13.68 kW/L and ED = 6.4 Wh/L.

References 1.

2.

3.

4.

5.

6.

7.

8.

9.

Scherer H. ZF 6-Speed Automatic Transmission for Passenger Cars, SAE technical paper, 2003-01-0596, Society of Automotive Engineers 2003 World Congress, Detroit, MI, 3–6 March 2003. Nozaki K., Kashihara Y., Takahashi N., Hoshino A., Mori A., Tsukamoto H. Toyota’s New Five-speed Automatic Transmission A750E/A750F for RWD Vehicles, SAE technical paper, 2003-01-0596, Society of Automotive Engineers 2003 World Congress, Detroit, MI, 3–6 March 2003. Yamamoto Y., Nishida M., Suzuki K., Kozaki S. New Five-speed Automatic Transmission for FWD Vehicles, SAE technical paper 2001-01-0871, Society of Automotive Engineers 2001 World Congress, Detroit, MI, 5–8 March 2001. Burke A. ‘Cost-effective combinations of ultra-capacitors and batteries for vehicle applications’. Proceedings of the Second International Advanced Automotive Battery Conference, AABC, Las Vegas, NV, February 2002. Gonder J., Pesaran A., Lustbader J., Tataria H. Fuel Economy and Performance of Mild Hybrids with Ultracapacitors: Simulations and Vehicle Test Results, The 5th International Symposium on Large EC Capacitor Technology and Application, ECCAP, Long Beach, CA, 9–10 June 2009. Miller J.M. ‘Combination ultracapacitor-battery performance dependence on drive cycle dynamics’. The 18th International Seminar on Double Layer Capacitors and Hybrid Energy Storage Devices, Deerfield Beach, FL, 8–10 December 2008. Miller J.M. ‘Energy storage system technology challenges facing strong hybrid, plug-in and battery electric vehicles’. Presented at the 5th IEEE Vehicle Power and Propulsion Conference, VPPC2009, Fairlane Technical Center, Dearborn, MI, 7–10 September 2009. Miller J.M., Bohn T., Dougherty T.J., Deshpande U. ‘Why hybridization of energy storage is essential for future hybrid, plug-in and battery electric vehicles’. The 1st IEEE Energy Conversion Congress and Exposition, ECCE2009, San Jose, CA, 21–24 September 2009. Miller J.R., Burke A.F. ‘Electrochemical capacitors: Challenges and opportunities for real-world applications’. The Electrochemical Society Interface, 2008, vol. 17, no. 1, pp. 53–7.

Sizing the drive system 10.

11.

12.

13.

14. 15. 16.

17.

18.

19.

20.

21.

22.

23.

239

Miller J.R., Klementov A.D. ‘Electrochemical capacitor performance compared with the performance of advanced lithium ion batteries’. Proceedings of the 17th International Seminar on Double Layer Capacitors and Hybrid Energy Storage Devices, Deerfield Beach, FL: Redox Engineering LLC, 10–12 December 2007. Miller J.M., Deshpande U., Dougherty T., Bohn T. ‘Power electronic enabled active hybrid energy storage system and its economic viability’. Presented at the IEEE 24th APEC Conference, Washington, DC: IEEE, 15– 19 February 2009. Miller J.M., Deshpande U., Dougherty T., Bohn T. ‘Combination ultracapacitor-battery performance dependence on drive cycle dynamics’. The 18th International Seminar on Supercapacitors and Hybrid Energy Storage Devices, Deerfield Beach, FL, 8–11 December 2008. Miller J.M., Bohn T., Deshpande U. ‘Dc-dc converter buffered ultracapacitor in active parallel combination with lithium battery for plug-in hybrid electric vehicle energy storage’. SAE World Congress, SAE paper 2008-01-1003, Detroit, MI, 17 April 2008. Bohn T., Miller J.M. Ultracapacitor Energy Storage Methods for PHEVs, SAE Hybrid Symposium, San Diego, CA, 14 February 2008. Burke A. ‘Ultracapacitor technologies and application in hybrid and electric vehicles’. International Journal of Energy Research, July 2009. Burke A., Miller M. ‘Testing of electrochemical capacitors: Capacitance, resistance, energy density, and power capability’. Electrochemical ACTA, Nantes, France, June 2009. Burke A., Zhao H., Gelder E.V. Simulated Performance of Alternative Hybrid-electric Powertrains in Vehicles on Various Driving Cycles, International Electric Vehicle Symposium, EVS-24, Stavanger, Norway, 13–16 May 2009. Burke A. Performance Testing of Lithium-ion Batteries of Various Chemistries for EV and PHEV Applications. Presented at the 2009 ZEV Symposium, Sacramento, CA, 22 September 2009. Miller J.M., Prummer M., Schneuwly A. Power Electronic Interface for an Ultracapacitor as the Power Buffer in a Hybrid Electric Energy Storage System, Power System Design Europe, July 2007. Auer J., Sartorelli G., Miller J.M. ‘Ultracapacitors – Improving energy storage for hybrid vehicles’. Presented at the EET-2007 European Ele-Drive Conference, Brussels, Belgium, 2007. Bohn T. Plug-in Hybrid Vehicles: Decoupling Battery Load Transients with Ultracapacitor Storage. Presented at the Advanced Capacitor World Summit, San Diego, CA, 25 July 2007. Verbrugge M., Liu P., Soukiazian S., Ying R. ‘Electrochemical energy storage systems and range extended electric vehicles’. The 25th International Battery Seminar & Exhibit, Fort Lauderdale, FL, March 2008. Ostovic V. Personal discussion by author, 19 February 2002.

240 24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34. 35.

36.

37.

Propulsion systems for hybrid vehicles Lorenz L., Mitlehner H. Key Power Semiconductor Device Concepts for the Next Decade. IEEE Industry Applications Society 37th Annual Meeting, IEEE IAS, Pittsburgh, PA, 13–18 October 2002. Francis R., Soldano M. ‘A new SMPS non punch thru IGBT replaces MOSFET in SMPS high frequency applications’. Presented at the IEEE Applied Power Electronics Conference and Exposition, APEC 2003, Miami, FL, 9–13 February 2003. Miller J.M., Bohn T., Dougherty T., Deshpande U. ‘Why hybridization of energy storage is essential for future hybrid, plug-in and battery electric vehicles’. IEEE 1st Energy Conversion Congress and Exposition, San Jose, CA, 20–24 September 2009. Wang J., Zhou X., Li J., Zhao T., Huang, A.Q., Callanan R., Husna F., Agarwal A. ‘10 kV SiC MOSFET based boost converter’. Proceedings of the 1st Annual FREEDM Systems Center Conference, Raleigh, NC, 18–19 May 2009. Briere M.A. GaN-based Power Devices Offer Game-Changing Potential in Power-Conversion Electronics. A technology report from ACOO Enterprises LLC on a new semiconductor process from International Rectifier. Backlund B., Rahimo M., Klaka S., Siefken J. ‘Topologies, voltage ratings and state of the art high power semiconductor devices for medium voltage wind energy conversion’. IEEE Power Electronics and Machines in Wind Applications, PEMWA 2009, Lincoln, NE, 24–26 June 2009. Kieferndorf R. Active Rectifier Controlled Variable DC Link PWM Drive. Wisconsin Electric Machines and Drives Consortium, WEMPEC, Annual Review Meeting, Madison, WI, 22–23 May 2002. Taraba G.M., Cebreiro J.P., Tacca H.E. ‘Batteries and hypercapacitors selection criteria for a series hybrid bus’. IEEE Workshop on Power Electronics in Transportation, No. 02TH8625, Auburn Hills, MI, 24–25 October 2002, pp. 17–23. Brost R.D. ‘42V battery requirements from an automaker’s perspective’. Presented at the Ninth Asian Battery Conference, Indonesia, 10–13 September 2001. Sutula R., Heitner K., Rogers S.A., Duong T.Q. Electric and Hybrid Vehicle Energy Storage R&D Programs of the US Department of Energy, Electric Vehicle Symposium, EVS16, September 1999. Ballard Power Systems. Available from www.ballard.com. Smith B.C. Next Energy Initiative. Positioning the State of Michigan as a Leading Candidate for Fuel Cell and Alternative Powertrain Manufacturing. Report by the Michigan Economic Development Corporation and the Michigan Automotive Partnership, August 2001. Burke A. ‘Cost-effective combinations of ultra-capacitors and batteries for vehicle applications’. Proceedings of the Second Advanced Automotive Battery Conference, Las Vegas, NV, February 2002. Lugert G., Knorr R., Graf H.-M. ‘14V/42V PowerNet and ISG – A solution for high dynamic energy supply suitable for mass market’, Proceedings of the

Sizing the drive system

38. 39.

40.

41.

42.

43. 44.

45.

46.

241

Second Advanced Automotive Battery Conference, Las Vegas, NV, February 2002. ELDRE Corporation. Available from www.busbar.com. Dimino C.A., Dodballapur R., Pomes J.A. ‘A low inductance, simplified snubber, power inverter implementation’. Proceedings of the High Frequency Power Converter HFPC Conference, IEEE PESC - power electronics specialists conference, Piscataway, NJ, 1994, pp. 502–09. Huang H., Miller J.M., Degner M.W. Method and Circuit for Reducing Battery Ripple Current is a Multiple Inverter System of an Electric Machine, US Patent 6,392,905, 21 May 2002. Leteinturier P., Kelling N.A., Kelling U. TTCAN from Applications to Products in Automotive Systems. SAE technical paper, SAE World Congress, 2003-01-0114, Detroit, MI, 3–6 March 2003. Fuehrer T., Hugel R., Hartwich F., Weiler H. FlexRay – The Communications System for Future Control Systems in Vehicles. SAE technical paper 2003-01-0110, SAE World Congress, Detroit, MI, 3–6 March 2003. Lupini C.A. Multiplex Bus Progression 2003. SAE technical paper 2003-010111, SAE World Congress, Detroit, MI, 3–6 March 2003. Miller J.M. ‘Barriers and opportunity for power train integrated power electronics’. Center for Power Electronic Systems, CPES, Invited Paper, Virginia Technological and State University Seminar, Blacksburg, VA, 16 April 2002. Jaura A.K., Miller J.M. HEVs – Vehicles that Go the Extra Mile and are Fun to Drive, SAE Convergence on Transportation Electronics, paper no. 202-210040, Cobo Exposition and Conference Center, Detroit, MI, 21–23 October 2002. Graham R. Comparing the Benefits and Impacts of Hybrid Electric Vehicle Options. EPRI final report 1000349. Available from www.epri.com.

Chapter 5

Electric drive system technologies

This chapter explores the four classes of electric machines having the most bearing on hybrid propulsion systems: the brushless permanent magnet machine in its surface permanent magnet (SPM) configuration; the interior permanent magnet (IPM) synchronous machine in either inset or buried magnet configuration; the asynchronous or cage rotor induction machine (IM); and lastly, the variable reluctance or doubly salient machine (VRM). There exists a mountain of books, papers and training workshops dedicated to the design and study of these four classes of machines. The purpose of this chapter is to present a design perspective of these electric machines in the context of hybrid vehicle propulsion. This is a topic that is still not as well defined, for example, as the design of industrial electric machines. In this chapter, we focus attention on the design characteristics of the four classes of electric machines of most interest for hybrid propulsion. To amplify the reasons for these choices, consider the following products now available in the market: ●

● ●



SPM (Honda Insight and Civic mild hybrids, FCX-V3 FCEV; Mannesmann Sachs ISA) IPM (Toyota Prius, Estima and Ford Motor Co. hybrid Escape) IM (GM Silverado ISG, Continental ISAD, Delphi Automotive ISG Valeo ISA) VRM (Dana ISA)

These machines in current use are all of the drum design – that is, rotor flux is radial across a cylindrical airgap, versus axial designs that have a distinct pancake appearance with axial flux across an airgap that separates the disc-shaped stator(s) and rotor(s). A plural connotation is used on axial machines because most will have a single stator and twin rotor discs or vice versa.

5.1 Permanent magnets Permanent magnet materials comprise various alloys of ferrous and non-ferrous metals to achieve a strong ferromagnetic effect. Key attributes include high remanence flux, Br, and high coercivity, Hc, to yield the highest energy product, BHmx.

244

Propulsion systems for hybrid vehicles

For example, permanent magnets exhibit strong flux retention in their Martensite state of tetragonal crystals and strong ferromagnetic phase. In this phase, iron will gradually lose its flux carrying properties as the temperature rises to the Curie temperature of approximately 1,100 K, at which point its magnetization is zero. A material that exhibits a strong phase transition, such as nickel–manganese–gallium, will gradually lose magnetization as the Curie temperature (~320 K) is reached, then drop precipitously at the Curie temperature and then reduce fairly quickly to zero within a relatively narrow temperature band of only 100 K. Phase transitions involve crystallographic or crystal symmetry transformation, which leads to sharp transitions in magnetism from ferromagnetic to paramagnetic. For example, a Martensite material heated above its Curie point will exhibit crystallographic change from tetragonal to cubic crystal structure in its Austenite phase. Ferrites and samarium– cobalt (SmCo) magnets have hexagonal crystal structure and the high energy rare earth neodymium iron boron (NdFeB) magnets have tetragonal structure as noted. Ferrites such as barium (~1955) and strontium (~1970) mixes are the cheapest, but also the lowest energy, permanent magnets available. Alnico was commercialized (~1935) and is about twice the price of ferrite and is the most stable of the permanent magnet materials. Of the cobalt types, SmCo2 is stable to 1,150  C. NdFeB grades (~1980) have been developed to a stored energy level of 58 MGOe. Sumitomo Special Metals developed NEOMAX, an NdFeB permanent magnet with Br ¼ 1.53 T (15,200 G), Hci ¼ 784 kA/m (9.8 kOe) and BHmx ¼ 460 kJ/m3 (57.7 MGOe). The ceramic ferrites have low magnetic properties (Br, Hc), but are lowest in cost too, have good temperature stability to >250  C and are corrosion resistant. Sintered and hot formed fully dense anisotropic neodymium magnets are superior to SmCo below 180  C. There needs to be solid technical justification such as corrosion resistance and very high temperature (approaching 300  C) to justify the expensive SmCo magnets. Alnico magnets were the first type used in electric machines. This is because of their high flux that approximated the fields possible in shunt wound dc motors of the day. However, because of very low coercive force of alnico magnets, these early electric machines used novel pole magnetic structures such as soft iron pole pieces adjacent to the armature or soft iron pole pieces bonded to the alnico magnet and machined to the arc of the rotor (armature of dc motor) [1]. With soft iron pole shoes of this design, the alnico magnet motor could operate at up to six times the normal armature current of alnico only before demagnetization. Even if the alnico did become demagnetized, it would be an easy matter to simply re-magnetize it using coils designed just for this purpose as was done in these early days.

5.1.1

Permanent magnets: A primer

With the availability of higher performance and low cost ceramic magnets, the alnico magnet was displaced from use in electric machines but remained useful in electronic metres as noted earlier. The industry compares various grades of permanent magnets based on their maximum energy product of remnant flux density,

Electric drive system technologies

245

Table 5.1 Performance of popular permanent magnet types Type

Remnant Br (T)

Coercive Hc (kA/m)

Energy (MGOe)

Recoil permeability (#)

Alnico 5 Alnico 6 Alnico 9 Ceramic 5 Ceramic 6 Ceramic 8 Magnequench I Magnequench II Magnequench III NeoMax 27H NeoMax 35

1.35 1.05 1.06 0.38 0.32 0.40 0.68 0.8 1.31 1.1 1.25

58.9 62 119.3 190.8 190.8 222.6 390 517 979 811 882

7.5 31 9 3.4 2.5 4.1 9.8 13 42 28 36

17 13 7 1.1 1.1 1.1 1.22 1.15 1.06 1.05 1.05

Conversion: (Hc in A/m)/79.6 = Oersted (Oe); (Br in T)  104 = Gauss (G).

Br, and coercive force, Hc, at a permeance coefficient that falls on the magnet recoil line. Table 5.1 lists magnetic properties of magnets. Permeance coefficient (Pc) Also referred to as the operating slope or the load line. The slope of the load line is the permeance coefficient (Pc) ad

Lo e

lin

Demag curve second quadrant

Operating point

Intrinsic curve

Br

Bd

Normal curve

H-axis (Oersteds or A/m)

Hci Hc

B-axis (Gauss or Tesla)

Slope of the normal curve as it passes through the B-axis is called the recoil permeability, mr Typical values of mr for sintered NdFeB are 1.05. Bonded NdFeB range from 1.12 to 1.70.

Hd

Figure 5.1 Illustration of permanent magnet intrinsic and normal curves The better grades of permanent magnets used for electric machines tend to have recoil permeability approach the permeability of air (1.0), which is why the best motor magnets are the RE neodymium types such as NeoMax 27H. Figure 5.1 illustrates this situation for RE magnets. In summary, we can say the following about permanent magnets: ● ●

Alnico for highest temperature stability Ceramic for best energy at lowest cost

246 ● ●

Propulsion systems for hybrid vehicles SmCo for compactness and good thermal stability NdFeB for compactness, robustness and low cost

Example 1: For the 57.7 MGOe NEOMAX permanent magnet described earlier, calculate the B versus H characteristic in the third quadrant and estimate Bd and Hd at BHmx. Assume: mr ¼ 1.05 and given m0 ¼ 4p  107 H/m. Solution: This can be solved in Excel by inputing a range for H (kA/m) and solving the following equation for B(H) (Figure 5.2): BðHÞ ¼ Br þ mr m0 H The result is shown in the following chart for B(H) and the locus of BHmx ¼ 57.7 MGOe. From the tangent point of BHmx with the B(H) characteristic, the maximum energy product points are found to be ● ●

Bd ¼ 0.76 Hd ¼ 600 RE permanent magnet B vs. H 3.50 B(H)

3.00

BHmx = 57 MGOe Magnetization (T)

2.50 2.00 1.50 1.00 0.50 0.00 –1,400

–1,200

–1,000 –800 –600 –400 Coercive force, Hc (kA/m)

–200

0

–0.50

Figure 5.2 Plot of rare earth NdFeB normal characteristic and BH

mx

trace

In reality, the normal characteristic shown in the equation in Example 1 will exhibit a knee in the second quadrant. To make a more accurate calculation of B(H) requires that more detail is known about the magnet so that a bilinear relation can be applied to find B(H).

247

Electric drive system technologies

5.1.2 What happened to Alnico? The real concern with Alnico, a very high flux permanent magnet having typical remanence flux of the order of the highest grade rare earth NdFeB, is that its coercive force is very low, less than that of ferrite magnets. This means that in electric machine applications where demagnetizing force via armature reaction appears through the magnet, it will easily become demagnetized. Alnico 5, for example, has an energy product of 5 MGOe that is stable and flat from 100 to þ300  C, and its intrinsic coercivity, Hci, is flat at a low value of ~1 kOe over this temperature range. In comparison with other magnet materials, Alnico was introduced relatively early on and found widespread application in D’Arsonval meter movements as a flux source. Figure 5.3 illustrates the types of permanent magnets by introduction date and energy capability.

Improvements in magnet strength 60

480 OTHER IMPORTANT CHARACTERISTICS

BHmx (MGOe)

40

440

REQUIRED MAGNETIZING FIELD THERMAL STABILITY MECHANICAL PROPERTIES CORROSION RESISTANCE MANUFACTURABILITY COST

400 360 320

NdFeB

280 30

240

SmCO 1-5 and 2-17

200 Bonded isotropic NdFeB Sintered ferrite

20

160 120

Columnar alnico 10

0 1900

Ks steel

80

Mk steel Alnico 5

1920

BHmx (kJ/m3)

50

40 1960

1940

1980

0 2000

Year

Figure 5.3 Illustration of permanent magnet types (Arnold Magnetics)

Figure 5.3 also shows by scale the phenomenal energy product of NdFeB relative to other types of magnets. The magnitudes of energy are truly phenomenal, with NdFeB reaching close to the theoretical limits of magnetic energy storage in a PM. Not all electric machines require, or use, rare earth magnets. Some examples of various novel electric machines are shown here for perspective. The investigation of alternative stator steels has also been ongoing over the years with amorphous iron of particular interest. Figure 5.4 provides some examples of such novel electric machines.

248

Propulsion systems for hybrid vehicles

1 cm Ferrite magnet rotors (a)

Coiled iron core made of amorphous metal (b)

(c)

Figure 5.4 Novel electric machines: (a) amorphous iron (Hitachi motor), (b) very high speed machine (1M rpm PM machine) and (c) concentrated winding servo motor The Hitachi machine shown in Figure 5.4 was built by Hitachi Industrial Equipment Systems Co. Ltd with a core of amorphous metal and ferrite magnet rotor. It does not require neodymium or dysprosium rare earth magnets. Because amorphous metal is a supercooled liquid, its disordered atomic structure, versus crystalline structure of steels, provides high tensile strength and low magnetic losses. Researchers at ETH Zurich, and later the company Celerotron, fabricated such titanium-encapsulated rotor, and special ball bearings facilitated this extreme speed machine. The prototype shown in Figure 5.4 is of the size of a small matchbox. The concentrated winding stator shown is typical of such designs. Here, an m ¼ 3 phases, Q ¼ 9 slots and P ¼ 6 poles having a slots/pole/phase (SPP) ¼ Q/mP ¼ 1/2.

5.1.3

Rare earth permanent magnets

The NdFeB permanent magnet has been discussed in previous sections in terms of energy and its relation to other types of magnets. In this section, some detail of NdFeB in regard to electric machine use is cited. In particular, ●







NdFeB rare earth magnets have high electrical conductivity and will therefore experience higher rotor heating from PWM drives than lower conductivity magnets such as ceramics and ferrites. The rare earth magnet is also very dense, close to the density of iron and as such is subject to corrosion in humid environments. This has led to a need to coat the magnet with nickel and aluminium. Earlier designs relied on hermetically sealed cans to cover the NdFeB magnets. Otherwise, they would slowly disintegrate and rust away. It is typical of RE magnet use to have smaller slabs cut and insulated from each other to minimize eddy current heating on the rotor. There is still interest in using bonded RE magnets because of their low cost, but the very low energy product (see the lower right box in Figure 5.3) has been a performance deterrent. The real advantage of NdFeB in addition to very high energy product and high remanence is its very high coercivity. This is a strong advantage in electric machines so that demagnetization due to armature (stator) reaction will not occur, even at higher temperatures.

Electric drive system technologies ●

249

The temperature capability of REs is now of the order of 180  C when trace amounts of dysprosium are added to the alloy mix.

Table 5.2 summarizes some of the shock and bending moment characteristics of the more popular permanent magnets. Table 5.2 Mechanical properties of common permanent magnets Type of permanent magnet

Shock resistance

Bending moment resistance

Alnico Ferrite SmCo NdFeB Bonded types

Strong, but brittle Chips or fractures easily Chips or fractures Chips or fractures Resists chipping

55–206 MPa Brittle 35–42 MPa 69–97 MPa 20–82 MPa

The rare earth PMs are the most widely used magnets and should be understood before elaborating more on electric machines. A generally useful method to understand the differences between normal and intrinsic flux density is to consider a gapped toroid with a PM inserted into the physical gap and a current carrying winding on this toroid that produces a flux in the same direction as the PM. A simple calculation of Ampere’s circuital law around the closed contour of the flux path resolves to (5.1), where Lm is the magnet length, Lg the airgap length, Iarm the total armature winding current (i.e. the coil of Narm turns) and I0 the coil current: H m Lm þ 2

X Bg Lg ¼ I arm ¼ N arm I 0 m0

ð5:1Þ

A straightforward calculation shows that (5.1) reduces to (5.2), assuming no fringe flux at the physical airgaps, which, in this case, are assumed to be very small relative to the magnet length (i.e. the dimension aligned with its magnetization vector). The normal flux density is therefore given by B m ¼ m0 ¼

P

I arm  Hm Lm



Lm 2Lg



ð5:2Þ

The intrinsic flux density, Bi, accounts for flux density that would exist in vacuum for the same applied magnetic field strength, H, and is given as Bi ¼ B  m0H. The intrinsic flux density is normally displayed in permanent magnet characteristic curves and is given by (5.3), which is Bm  m0Hm: B m  m0 H m ¼ m0

 P

I arm  Hm Lm þ 2Lg



 Lm þ1 2Lg

ð5:3Þ

250

Propulsion systems for hybrid vehicles

A useful analysis and graphical tool is the permeance coefficient, Pc, when any PM is placed into some magnetic circuit. The permeance coefficient represents the magnetic conductance of the circuit and is typically set to obtain as high a value of airgap working flux in an electric machine as possible (i.e. minimal physical gaps). Figure 5.5 illustrates the second quadrant characteristics of a typical RE magnet in both intrinsic and normal magnetization. In Figure 5.5 the normal curve (5.2) and the intrinsic curve (5.3) traverse the B–H second quadrant from the remanence flux density, Br, to coercive force, Hc, and intrinsic coercive force, Hci, respectively. A particular magnetic circuit design determines the permeance coefficient at the static condition. When excitation current is applied to an electric machine stator or armature, a demagnetization field is impressed across the magnet, and if this field is sufficient to push the permeance coefficient (i.e. load line) across the knee of the curve, there will be permanent demagnetization. Figure 5.5 illustrates the case of normal operation at some elevated temperature where the knee of the normal curve has moved up into the second quadrant. Since operation is along the recoil line, mr, of the normal curve, the magnetic circuit remains intact over its useful operating life. However, in the case of abnormal currents or faults, the PM may become demagnetized sufficiently to render the machine unuseable.

rc

(rc + 1) rc

B Br Bm = Bg

Hci

Bd

Norm al

Intrinsic

mr

Hc Hd

Hm

O

∑Iarm Lm

Figure 5.5 RE permanent magnet characteristics An example will suffice to illustrate these points. In the following example, a simple numerical calculation is performed in Excel, where the spacial and temporal characteristics of an electric machine are approximated by columns and rows, respectively. The column headings represent phase windings in a 3-phase brushless dc machine of the type to be considered in the following section and the rows represent a particular time point of the 3-phase currents. These currents are temporarily displaced by 120 electrical degrees. The machine itself is assumed to be a Q ¼ 12 slots, m ¼ 3 phases, P ¼ 4 poles and q ¼ 1 SPP. Example 2: In this example, the brushless dc motor (BDCM) is assumed to be excited by 80 A of armature current into phase coils of 80t (80 turns) each. The SPM rotor consists of four poles, of which only two are highlighted in this example. The convention here is that a positive clocked winding, such as phase A, is noted by

Electric drive system technologies

251

AP and AN for the respective coil sides, P for positive magnetomotive force (mmf) when a positive current flows into that slot. Solution: The reader can verify the mmf waveform in the spacial direction (x-direction) by plotting out the current amplitudes beneath a particular row. Be careful to take into account the current sign and slot conductor clocking orientation. Time 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01 0.011 0.012 0.013 0.014 0.015 0.016 0.017 0.018 0.019 0.02 0.021 0.022 0.023

BN −49.691 −23.121 6.696 35.573 59.454 74.984 79.982 73.747 57.153 32.532 3.342 −26.317 −52.281 −70.901 −79.563 −77.050 −63.715 −41.431 −13.328 16.647 44.284 65.701 77.890

AP 29.451 54.765 72.387 79.842 76.083 61.638 38.536 10.021 −19.901 −47.029 −67.550 −78.585 −78.581 −67.541 −47.014 −19.884 10.039 38.552 61.650 76.089 79.841 72.380 54.752

CN 79.142 77.886 65.691 44.269 16.629 −13.346 −41.446 −63.726 −77.055 −79.561 −70.893 −52.267 −26.301 3.360 32.549 57.166 73.754 79.983 74.978 59.442 35.557 6.679 −23.138

BP 49.691 23.121 −6.696 −35.573 −59.454 −74.984 −79.982 −73.747 −57.153 −32.532 −3.342 26.317 52.281 70.901 79.563 77.050 63.715 41.431 13.328 −16.647 −44.284 −65.701 −77.890

AN −29.451 −54.765 −72.387 −79.842 −76.083 −61.638 −38.536 −10.021 19.901 47.029 67.550 78.585 78.581 67.541 47.014 19.884 −10.039 −38.552 −61.650 −76.089 −79.841 −72.380 −54.752

The two colour shaded blocks are approximately the PM span. For example, in the t ¼ 1 ms row (top row) and for the case where radian frequency is 377 rad/s, the north magnet pole is just left of the CN conductors and the south magnet pole just approaches the CN conductors. The reader can verify that the mmf over each magnet is given by mmf N ¼

X ðBN  I B þ AP  I A Þ;

mmf S ¼

X ðBP  I B þ AN  I A Þ

mmf N ¼ 40ð49:69Þ þ 40ð29:45Þ ¼ 3,165:6 At mmf S ¼ 40ð49:29Þ  40ð29:45Þ ¼ 3,165:6 At The stator windings produce an mmf of proper polarity to cause the rotor to spin in the counterclockwise direction, as seen by the left shifting of the magnet pole shaded areas above as the time evolves towards 20 ms.

252

Propulsion systems for hybrid vehicles

The above example serves as a good transition to discussion on the various types of electric machines. The following sections of this chapter carry this analysis to greater depths. The types of machines considered in this chapter are contained in a broader scope or taxonomy of electric machines shown in Figure 5.6.

Electric machines for hybrid vehicle ac drives

Synchronous

Asynchronous

Induction cage rotor

Induction wound rotor

Induction doubly fed

Brushed DC

Unipolar

Interior PM

IM

Permanent magnet

Inset PM

IPM

Variable reluctance

Surface PM

Doubly fed reluctance

SPM

Switched reluctance VRM

Figure 5.6 Survey of electrical machine types

5.2 Brushless machines The electromagnetic interaction responsible for torque production is the Lorenz force defined with the help of Figure 5.7. In this illustration, a pair of magnets force magnetic flux across a gap in which reside a pair of conductors that are free to rotate. When current is injected into the conductor turn, there will be a magnetic flux encircling the conductor that interacts with the field flux (depicted as lines), resulting in a force on the currents in that conductor and oriented orthogonal to both the flux and the current. The resultant Lorenz force is a vector cross-product of the flux and the current, with the seat of the force resting on the current in the conductors. The term seat of the force is used to dispel the notion that the force acts on the copper conductor or on some other member. For example, the electron beam in a cathode ray tube (CRT) is formed by thermionic emission of electrons from a caesium-coated tungsten wire cathode. The electron cloud is subsequently focused into a beam and accelerated by a high potential at the CRT anode (a conductive coating on the inside walls of the tube). A raster is scanned on the CRT face by horizontal and vertical deflection coils placed around the neck of the CRT that form a cross-field into which the electron stream passes. When encountering the magnetic field (shaped very similar to that in Figure 5.7), the electrons experience the Lorenz force and are deflected orthogonal to both their velocity vector (initially down the z-axis of the tube) and to the field itself (the x- and y-axes). As further

Electric drive system technologies

253

illustration of what is meant by ‘seat of the electromagnetic force’, consider the superconducting motors and generators now being developed by the American Superconductor Corporation in ratings of up to 5 MW. In a superconducting M/G, the rotor contains high temperature superconducting wire (BSCCO-2233, an acronym for the alloy Bi(2x)PbxSr2Ca2Cu3O10 high temperature superconductor (HTS) multifilament wire in a silver matrix that superconducts up to 110 K). The stator is conventional design of laminated steel with copper conductor coils. Because the rotor flux is so high, the airgap flux density will typically be in the range 1.7–2.0 T, negating the need for stator teeth. The resulting coreless design requires nonmagnetic wedges or teeth to provide a restraint for the stator reaction force acting on the conductors during operation.

L

Magnet Axis of rotation N

F

B

X r

+I S T

–I

Figure 5.7 Torque production mechanism in the basic electric machine The illustration in Figure 5.7 is the basic dc brushed motor in which the conductors carrying currents labelled þI and –I would be connected to commutator segments upon which current carrying brushes would ride. In the process, the brushes would ensure that current was always injected into the conductor at the top and extracted from the conductor at the bottom in Figure 5.7. The brushes and commutator therefore maintain the current in the armature conductors so that the armature field produced by the current carrying loop is maintained orthogonal to the magnet produced field. In Figure 5.7 the magnet field is shown filling the gap between the arc shaped pole pieces in the north to south direction. The plane of the conductor loop lies within this flux field so that when current is flowing in the conductors as shown, this armature winding produces a magnetic field with its north pole oriented to the left and orthogonal to the magnet field lines. The Lorenz force equation applied to this geometry results in a force, F, on both conductors that is tangential to the loop

254

Propulsion systems for hybrid vehicles

axis of rotation at a distance, r, from its axis of rotation. Torque on each conductor is F  r, so the total motoring action becomes T ¼ 2Fr ¼ 2BILr

ð5:4Þ

U ¼ 2BLv ¼ 2BLrw

Equation (5.4) also includes the definition of the back electromotive force (back-emf) induced in the same conductors as they rotate through the same field, and according to Lenz’s law, produce an opposition to the current being injected. In practical electric machines, the armature conductors reside within slots in an iron core so that the reluctance to the permanent magnet flux across the gap is minimized. Similarly, in order to provide a low reluctance magnet flux return path, an iron sleeve is present around the magnet arcs. In ac electric machines, the mechanical commutator is replaced by power electronics switches that regulate the injection of currents into ‘armature’ windings, or more appropriately, into the fixed stator windings. The windings are composed of a number of conductor turns per coil, and multiple coils are present around the periphery of the stator according to the number of magnetic pole pairs. Figure 5.8 illustrates a ‘rolled-out’ stator in which the conductors are evenly distributed per unit length.

I

L X

X X X X

F

X X

b

F m=∞ I

Ax

ds = rdq Bz

Bz Right hand rule

Figure 5.8 Rolled-out stator showing electric loading, Ax The effect of conductors placed into slots in stator iron is an approximation to a current sheet. The rotor magnets develop magnetic poles that interact with the stator current sheet. The electromagnetic traction developed by the stator current sheet, Ax, and the magnet flux, Bz, located at an angle b from the current sheet vector is a surface traction, gs oriented in the y direction (tangential in a rotary motor). The surface traction is the product of stator current sheet, Ax, also referred to as the electric loading, and Bz, the magnetic loading. The Lorenz force is produced by the interaction of the electric and magnetic loading per unit surface area of the rotor. A term ‘Z’ is used to count conductor turns. It is evident from (5.4) that the orientation angle b should be held as close to 90 as possible for maximum

Electric drive system technologies

255

torque per ampere. This means that the spacial displacement from flux field and armature current sheet must be maintained orthogonal: ds ¼ rdq Z ¼ 2I Ax ¼

Z 2pr

S r ¼ 2prL gs ¼ Ax Bz cos ðbÞ ¼ F ¼ gs S r ¼

Bz Z cos ðbÞ 2pr

ð5:5Þ

Bz Z 2prL ¼ Bz ZL cos ðbÞ 2pr

T ¼ Fr ¼ BZLr cos ðbÞ Deviation from orthogonality between field flux distribution and armature current sheet results in a loss in torque and torque ripple components. The brushed dc machine, which the reader is no doubt familiar with, has Nb torque pulsations per mechanical revolution of the armature, where Nb is the number of commutator segments. The number of torque pulsations is the same regardless of the number of magnetic poles in the brushed dc machine. The mechanical angle of the armature, b, is inversely proportional to the number of commutator segments so that the following holds: p/Nb < b < p/Nb, where this range defines the peak-to-peak torque pulsation magnitude according to (5.5). The common 12-bar commutator has a torque ripple magnitude of   DT p ¼ 1  cos T pk Nb

(Nm)

ð5:6Þ

For the example of 12 segments, (5.6) predicts a variation in torque of (1  0.966) ¼ 3.4%. However, if only eight commutator bars are used, the pulsation torque increases to 7.6%. Of course, if only two segments are used, the torque ripple is 100%. Regardless of technology, the purpose of any electric machine used for hybrid propulsion is to develop the highest level of torque for a given current in the smallest package possible. Torque and power density are absolutely essential in order to maximize performance without incurring excessive weight and its attendant impact on fuel economy. As (5.5) shows, torque production depends on as high a level of flux in the machine as possible for a given magnitude of current (ampere-turns) and a value of rotor radius and length that meets vehicle packaging constraints. Each electric machine technology, when coupled with its electronic commutator, the power electronics inverter, has relative merits and disadvantages. The

256

Propulsion systems for hybrid vehicles

following sections explore the merits and disadvantages of the four machine technologies in the context of hybrid propulsion.

5.2.1

Brushless dc

The electronically commutated motor most closely related to the brushed dc machine described earlier is the brushless dc machine configured with SPMs. A BDCM may be of either 120 or 180 current conduction in the stator windings. When the machine’s back-emf due to the permanent magnet rotor has trapezoidal shape, the machine will be brushless dc and having current conduction in block mode of 120 duration. If the rotor magnets are designed for sinusoidal back-emf, the machine will be of the brushless ac variety and stator currents should be in 180 conduction. Figure 5.9 illustrates both types of brushless dc machines.

Voltage U, Current I

U(qe) I(qe)

qe

(a)

Voltage U, Current I

U(qe) I(qe) qe

(b)

Figure 5.9 Brushless dc motors: (a) brushless dc or trapezoidal flux and (b) brushless ac or sinusoidal flux

Electric drive system technologies

257

Both types of electronically commutated dc motors require electronic controls. A question that should immediately come to mind is how does the flux generate trapezoidal versus sinusoidal voltage? The answer is the rotor magnet design and magnetization orientation will determine the character of the voltage, and to some extent the slot design. Permanent magnets used in electric machines are invariably parallel magnetized with the magnetic field intensity lines oriented across the magnet length, which is generally of the order of 8–15 mm for ceramic magnets and 4–7 mm for rare earth magnets. Figure 5.10 illustrates the trend in back-emf as magnetization orientation proceeds from parallel to radial, that is sinusoidal to trapezoidal waveform.

(a) More sinusoidal

(b)

(c)

(d)

More trapezoidal

Figure 5.10 Illustration of permanent magnet magnetization orientation: (a) Halbach, (b) parallel, (c) tapered (breadloaf) and (d) radial Some comments on Figure 5.10 are necessary to explain the magnet configurations. In Figure 5.10(a) the SPM is made up of individually magnetized segments, each having a slightly different magnetization orientation so that the resulting flux is more dense at the centre and tapered towards the magnet ends. The length of the magnet is in the direction from the rotor back iron core (shown hashed) along a radius line to the rotor OD defined as the surface of the permanent magnet facing the airgap. The Halbach array, as this orientation is known, comes from nuclear physics focusing magnet arrays wherein the flux on the inside of the array extends to the centre, but on the outside of the annular magnet array, the flux is zero. Halbach arrays are self-shielding and require no back iron or minimal back iron on the self-shielding side. Electric machines have been fabricated with Halbach array techniques in an attempt to minimize rotor mass and inertia. The magnet orientation in Figure 5.10(b) is the conventional parallel magnetized magnet arc segment. This is the most common magnetization orientation found in dc brushed motors and brushless ac motors having sinusoidal back-emf. Ceramic magnet brushless ac motors are magnetized in situ by placing either the rotor or the entire motor in the proper orientation into a magnetizing fixture and applying a very high magnetizing intensity pulse. Rare earth permanent magnets such as SmCo of NdFeB have such high intrinsic coercivity that in situ magnetization is not possible and individual magnet segments must be pre-magnetized. One reason for this is that magnetizing fixtures are unable to supply the intense

258

Propulsion systems for hybrid vehicles

fields of the order of 2.8 MA/m (35 kOe, i.e. cgs units are more common in the magnet industry). The second reason is that NdFeB magnets have relatively high bulk conductivity so that the fast magnetizing transient induces high levels of eddy currents into the magnet slab, thus inhibiting penetration of the magnetizing flux. Magnetizers for NdFeB magnets tend to require higher pulse durations (higher stored energy) to ensure sufficient levels of magnetization. Regardless of the magnet material, if the magnetizer does not have sufficient magnetizing intensity to push the magnet well into first quadrant saturation (in its induction, B, versus magnetizing force, H, plane), the value of remanence induction will be low and/or there will be too much variation part to part from the process. Second, if the magnetizing pulse has insufficient dwell, the magnet may not be uniformly magnetized. The highest power density electric machines are the brushless dc type. This is because for a given value of flux in the machine, the flat top of the trapezoid results in much higher rms value than a sinusoidal flux for the same iron saturation limited peak value. The same applies for the current – block mode conduction with flat top waveform has a higher rms value than its corresponding sinusoidal cousin for the same current limit in the power electronics inverter. For this reason, brushless dc machines have found use in industrial machine tools and some traction applications. In Figure 5.10(c) the tapered magnet geometry is shown that tends to a more trapezoidal back-emf. This breadloaf style of magnet is typical of tapered designs for which the gradual magnetization, through gradual increase in the magnet thickness, yields a smooth shape for the reluctance torque. Reluctance torque in brushless machines of either variety is a serious noise issue, particularly for highenergy rare earth magnets. A motor design with NdFeB can produce three times the commutating torque than a ferrite ceramic design. The NdFeB design therefore has far more reluctance, or cogging, torque. The motor cogging torque gives the feeling of detents as the rotor is turned. The spectrum of reluctance torque effects is linearly decreasing for parallel magnetization (sinusoidal back-emf) designs with harmonic number. For a gradual magnetization, the effect is a similar linear decrease with harmonic number, but the initial value of reluctance is some 30% higher. For radial magnetization (trapezoidal back-emf), the reluctance torque increases with harmonic number, peaks for the second and third harmonics and then decreases linearly with higher harmonics. This harmonic flux is a serious issue with brushless dc machines: the trapezoidal back-emf causes very significant detent torque and consequent vibration. For traction applications, the inertia of the driveline may or may not swamp out the reluctance torque induced vibrations. There have been many techniques proposed for minimizing reluctance torque production in brushless dc machines, such as skewing the magnets along the rotor axis length, and careful design of the magnet pole arc and interpolar gap. The magnet pole arc can be visualized as the circumferential span of the magnet in Figure 5.10 versus the pole arc (in the 1-pole case shown this would be p-radians). It is most common to have magnet pole arcs of 0.7–0.8 times the pole span in order to minimize harmonic production. One of the more effective means to reduce detent torque in a brushless dc machine has been the implementation of stator pole notching. The effect is to have the magnet edges pass evenly spaced discontinuities

Electric drive system technologies

259

in airgap rather than just the stator slot gaps at the edges of full-pitched coils. The details of these techniques are outside the intent of this book. However, because of the issue with cogging torque, brushless dc machines have not found widespread acceptance as a hybrid propulsion technology, but rather are relegated to electrified ancillary drives where very high power density, low cost and compact packaging are the overriding considerations. Power electronics control of BDCMs is generally accomplished through classical 120 current conduction, or what has been referred to as block mode. Figure 5.11 illustrates the architecture of the BDCM with trapezoidal back-emf and rectangular current control. Power electronics stage and heat sink D

D

G

G S

D G

S

S Surface permanent magnet (SPM)

+ dc link cap.

Vehicle energy storage system

D D

G

D

G G

S S

S

Logic power supply

Gate driver Current sense

Gate driver

Back-emf voltage sensing

Gate driver

Microcontroller and communications

Communications network

Figure 5.11 Brushless dc motor control Figure 5.11 shows in schematic form the major components of a BDCM drive: (1) power electronics inverter stage and thermal management cold plate, (2) gate driver assemblies for controlling the power switches, (3) communications, current sensing and controller, (4) logic power supply for powering the controller, gate drivers and sensors, (5) the dc link capacitor necessary to circulate ripple currents from the motor and (6) the SPM motor. BDCMs for position control and applications requiring operation at zero speed will require an absolute rotor position sensor. The most economical choice for rotor position sensing is the use of Hall element sensors placed at 120 electrical intervals near the rotor magnets. The position information from these three Hall transducers provides the microcontroller commutation logic its timing and rotor direction information. For applications not requiring operation at or near zero speed, it is very common for BDCMs to rely on sensorless techniques such as back-emf sensing of the inert phase, the use of phase voltage and bridge current signals to infer position, and various techniques based on development of an artificial neutral.

260

Propulsion systems for hybrid vehicles

Current is injected into the BDCM in one phase and extracted from a second phase. This 2-phase excitation means that at any given time, one of the motor phases is available for use as a position sensor. Sensing back-emf is the most common form of sensorless BDCM control. Also, because of 2-phase excitation, a single current sensor in the bridge return path is all that is needed to regulate the currents in block mode.

A+ B+

C+ A–

B– C–

Vab Vbc Vca Van Vbn Vcn Ia Ib Ic

Figure 5.12 Brushless dc motor commutation signals (120 conduction) Figure 5.12 shows the 2-phase motor current conduction and the corresponding inverter switch commands. The switches are labelled Aþ, Bþ, Cþ across the top in Figure 5.11 and correspondingly across the bottom. When the switch gate command is logic 1, the switch is ON, connecting the mid-point of the phase leg to either þVd or 0. The line-to-line voltages can be used to flag the inert phase for back-emf sensing.

Electric drive system technologies

261

The features of brushless dc machines are high start-up torque and high efficiency. Voltages are nominally up to 60 Vdc, or 100–240 Vac, at power levels from 5 W (for computer disk drives) to 2.5 kW for CVT transmission oil pump drive. Speed ranges up to 30 krpm have been attained (e.g. as a sub-atmospheric refrigerant pump). High rotor speeds and high power are problematic because of the concern with rotor surface magnet retention.

5.2.2 Brushless ac When the stator back-emf is sinusoidal, the inverter is controlled in 180 conduction mode. The switch commands listed in Figure 5.12 have a dwell of 180 electrical degrees. For this conduction interval, the commands to switches in the A-phase leg, for instance, are auto-complementary. When switch Aþ is ON, then switch A must be OFF because bus shorting is not permitted. During the switch commutation interval, a built-in dead time of 3–5 ms is used as a guard band to prevent switch shoot through conduction. The commutation logic for a brushless ac motor differs in several key respects from its brushless dc cousin. Because of 180 conduction, the actual phase current must be monitored rather than bus current to regulate the sinusoidal waveshape. Also, due to the need for precise rotor position information, some form of mechanical rotor sensing such as an absolute encoder or resolver is required. Figure 5.11 is a schematic for the brushless ac motor having the features noted above. When the SPM motor back-emf voltage is lower than the dc link, the power inverter operates in pulse width modulation (PWM) mode to synthesize sinusoidal current waveforms in each of the motor phases with an amplitude set by the current regulator. Current regulator command magnitudes can be in response to torque if used in torque control mode, which hybrid propulsion motors are generally designed for, or to motor speed when used in speed control mode if used to drive electrified ancillaries such as pumps, fans or compressors. For example, the hybrid air conditioning used at times in hybrid electric vehicles uses a brushless motor rated 1.5–2.0 kW to maintain cabin climate control during engine-off periods. Brushless ac motors are also used in electric assist steering, a necessary feature when idle stop mode is used while the vehicle is still in motion. Other applications of brushless ac motors are active suspension actuators in a fully active, wide bandwidth, wheel position controller. Brushless ac motors are used in this application because of its smoother operation and softer detent torques than the BDCM. Application voltage ranges of the brushless ac motor controller range from 60 to 600 V (IGBT power inverter). An example of a pre-packaged motor controller is the International Rectifier Plug-N-Drive module [2] IRAMS10UP60A. This module is rated 10 A at 600 V and is capable of switching up to 20 kHz PWM. The power electronics relies on non-punch-through (NPT) (motor drive) IGBTs in a single in-line package (SIP). The module is rated for direct control of 750 W brushless motors. Compared to Figure 5.13, it contains all components except the motor, current sensors, power supply and the part of the microcontroller handling communications and outer loop control. Internal to the module is an IR21365C

262

Propulsion systems for hybrid vehicles Power electronics stage and heat sink D G

D G

S

D G

S

S

+ dc link cap.

Vehicle energy storage system

SPM

D D

G

D

G G

S

Rotor position sensor (encoder resolver)

S S

Current sense

Logic power supply

Gate driver

Communications network

Gate driver

Gate driver

Microcontroller

Figure 5.13 Brushless ac motor control integrated commutation logic controller. External phase leg current sense resistors are recommended for low cost applications. Overcurrent, overtemperature (via internal NTC thermistor) and undervoltage lock-out are built-in features. Six-step mode in 180 conduction is illustrated in Figure 5.14, where the top three traces show the gate driver signals to the power switches. Since conduction is 180 , the bottom switch command for each phase leg is the complement of the signal shown. For example, Aþ is the command to switch S1 in the power inverter and the complement of Aþ is impressed on switch S2 in phase leg A. During current regulation, the phase currents are as depicted in Figure 5.8(b) and phase shifted 120 for phases B and C, respectively. The total rms voltage and fundamental rms voltage of the brushless ac motor controller can be calculated easily from the fourth, fifth and sixth traces in Figure 5.14 by noting that the pulse takes on a magnitude of the dc link voltage, Ud, and has duration 2p/3 for every half-cycle interval of duration p electrical radians. The total rms value of the lineto-line voltages, Uab, Ubc and Uca, are calculated as

U ab

U ab

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð 2p=3 1 U 2d dq ¼ p 0 ðVolts11 , total rms) rffiffiffi 2 ¼ Ud 3

ð5:7Þ

The fundamental component of Uab is calculated from its first harmonic value as pffiffiffi 6 U an ¼ U d ðVolts11 , fundamental rms) ð5:8Þ p

Electric drive system technologies T1

T2

T3

T4

T5

T6

263

T1

A+

B+

C+

Vab

Vdc

Vbc

Vdc

Vdc Vca

Van

Vbn

Vcn

Vno

2/3 1/3

2/3 1/3

2/3 1/3

2/3 1/3

Figure 5.14 Brushless ac motor waveforms For quasi-square waveforms, the total rms and fundamental rms are very similar. Equation (5.4) predicts a total rms value of 0.816 Ud and (5.8) predicts a fundamental rms content of 0.78 Ud. Using the same procedure for the line to neutral voltages shown as traces 7, 8 and 9, the corresponding values are rffiffiffi 2 Ud U an ¼ pffiffiffi 3 3 ð5:9Þ pffiffiffi 2 U an1 ¼ Ud p where the line-neutral rms voltage given by (5.9) equates to 0.471 Ud and its fundamental component, Uan1 ¼ 0.45 Ud. The relations given by (5.8) and (5.9) will be important in later sections.

264

Propulsion systems for hybrid vehicles

Example 3: A 3-phase sine wave BDCM has the properties listed in Table 5.3 and is operated in a power inverter configured as shown in Figure 5.13 with waveforms according to Figure 5.14. Referring to the phasor diagram for this BDCM shown here (Figure 5.15) and the parameters listed, calculate: ● ● ●

The d- and q-axis components of the input stator voltage, Us The d- and q-axis components of the input stator current, Is, when g ¼ 15 The input electrical power at nb (calculate the magnet voltage first)

Table 5.3 Sine wave PM machine, 3-phase, inset magnet design Ur Ir Ipk Rph P Flux linkage (lm)

110 Vrms phase 255 Arms 600 A 3 mW 8 poles 0.03 Wb-t

Ld Lq Torque (mr) Peak torque (mpk) Base speed (nb)

0.1 mH 0.1 mH 65 Nm 155 Nm 4,400 rpm

g

jXqIq

jXdId

Us

Rs Is Ug

Eg Is = Id + jIg

Us = Ud + jUg Ud = Us sind Ug = Us cosd

Iq

Id = Is sing

d

Ig = Is cosg

g f

Ud

d Id

Figure 5.15 Phasor diagram for the sine wave BDCM

In the following solutions, the machine voltage expression for stator phase voltage Us and electromagnetic torque, mem, will be used: U s ¼ Eq þ Rph I s þ jX d I d þ jX q I q     3 P U s Eq us 2 1 1 sind   sin2d Pe ¼ 2 Xq Xd 2 2 Xd

Electric drive system technologies

265

Solution: (a) The d–q components of the stator voltage. Referring to the phasor diagram: U q ¼ Eq þ Rph I q þ X d I d ;

U d ¼ Rph I d þ X q I q

Substituting into the above equation after solving for Eq ¼ 55 Vpk, Uq ¼ 73.3 Vpk and Ud ¼ 64.13 Vpk. (b) For the stated value of g ¼ 15 , the rated stator current given results in Iq ¼

pffiffiffi 2I r cosg;

I q ¼ 348 Apk ;

Id ¼

pffiffiffi 2I r sing;

I d ¼ 92:5 Apk

(c) The input power requires first a calculation of the power angle, d, which can be found easily from the components of Us in the equation in (a) as d ¼ sin 1

Ud 64:1 ¼ 1:04 ðradÞ ¼ sin 1 Uq 73:3

Next, compute the input electrical power using the classical synchronous machine relation shown as Pe ¼

3 U s Eq sin d 2 Xd

Substituting into the above equation results in Pe ¼ 37.5 kW of input electrical power. The power inverter shown must be sufficiently rated to process this continuously.

5.2.3 Design essentials of the SPM In this section, the SPM machine will be treated from a hybrid design vantage point. The objective is to design an SPM as an M/G for a mild hybrid vehicle. The design process will illustrate the important features of machine target setting, electromagnetic design and modelling. First, we will briefly review the types of electric machines available for the hybrid propulsion M/G set. Figure 5.16 illustrates six types of electric machines that should be considered for this application. The machine types are as follows: ●

Surface permanent magnet. This is the most basic of permanent magnet electric machine designs. PMs are bonded to the surface of the solid rotor back iron that is, in turn, fitted to the high carbon steel (4,150 or equivalent steel) shaft. Rotor back iron is necessary as a flux return path, and this core may be either solid or laminated depending on application.

266 ●









Propulsion systems for hybrid vehicles Interior permanent magnet, or buried magnet design. The original IPM, known variously as buried magnet, or interior PM machine, consisted of a single magnet layer, tangentially oriented along chord lines in the rotor, when it was first conceived1 about 30 years ago. The buried magnets provide magnetization of the stator and minimize the reactive power needed. This design has been used for line-start appliance adjustable speed applications because of its high power factor and good torque performance. Asynchronous or induction machine design – the industrial workhorse electric. The induction design dates to 1888 when Nikola Tesla first conceived of what he termed his ‘current lag’ motor [3]. Trapped in his fourth floor room at the Gerlach Hotel in New York City during the blizzard of 1888 – the location of his newly formed Tesla Electric Company funded by venture capitalists when he resigned from the Edison Electric Company just 3 days earlier – he sketched out the design of the world’s first asynchronous electric machine. His design was that of an alternating current, 3-phase asynchronous machine – the IM. Interior permanent magnet – flux squeeze design with radial PMs. Rather than burying the rotor permanents beneath a soft iron pole shoe, the magnets can either be inset into the surface with the interpolar gap filled with soft iron, or they can be sandwiched between larger soft iron wedges as shown in Figure 5.16(d). This design is particularly attractive for ceramic magnet machines because of the flux squeezing effect at the airgap of flux collected from the larger surface of the magnets and effectively focused into the rotor to stator air gap. Synchronous reluctance (SyncRel design). By simply swapping out the cage or wound rotor of an IM and inserting a rotor of laminated saliencies, one obtains the synchronous reluctance design. These ‘rain-gutter’ style laminations have low reluctance to stator flux in the direct axis, but high reluctance to flux in its quadrature axis. The SyncRel machine has received renewed interest in recent years for use in machine tools and factory automation because of its inert rotor and synchronous torque–speed control. Variable reluctance machine. When both stator and rotor laminations are salient, a class of doubly salient machines is realized. The reluctance machine is renowned for its completely inert rotor and easy to install bobbin wound stator coils. The VRM has the power density of IMs but continues to have audible noise problems unless the stator back iron is reinforced, for example, by being excessively thick, or includes bosses or other structural enhancements.

Electric machines with smooth stators, that is, slotted designs with distributed windings, can be either ‘copper dominated’ or ‘iron dominated’. When the iron fraction, or ratio of stator tooth width to stator tooth pitch, b ¼ Wt/ts > 0.55, the machine can be called iron dominated. For the most part, hybrid propulsion machines are iron dominated in order to operate at magnetic loading values >0.7 T. This trend results from the fact that hybrid propulsion M/Gs have high peak to 1

US Patent 4,139,790 by C.R. Steen, ‘Direct Axis Aiding Permanet Magnets for a Laminated Synchronous Motor Rotor’, Issued, February 13, 1979.

Electric drive system technologies

(a)

(b)

(c)

(d )

(e)

(f)

267

Figure 5.16 Types of electric machines for hybrid propulsion M/G: (a) surface permanent magnet (SPM), (b) interior permanent magnet (IPM), (c) induction machine (IM), (d) interior PM – flux squeeze, (e) synchronous reluctance, SyncRel and (f) variable reluctance machine (VRM) continuous usage, so that putting more iron into the design permits operation at high flux levels for efficient inverter and battery operation. When high torque is demanded, the machine currents are driven to high values, but generally for short durations such as 10–30 s, and to somewhat lower values for up to 3 min. The fact

268

Propulsion systems for hybrid vehicles

that hybrid M/Gs are generally liquid cooled (use of engine coolant and/or transmission oil) completely supports this trend. Generally speaking, liquid cooling boosts the torque production, so that the output of liquid-cooled machines is four times that of air-cooled machines, all else being equal. The fundamental purpose of any electric machine is to deliver torque. If the machine package volume is constrained, then a metric of torque per litre is valid, but this is not as universally applicable as the more specific torque per unit mass, Nm/kg. High torque to mass implies high power density and also high acceleration capability. Some comments about the application choice of the electric machine types listed in Figure 5.16 are in order before proceeding with a more detailed design of the SPM. The SPM design may be sinusoidal or trapezoidal back-emf. Sinusoidal, brushless ac designs tend to require higher inverter rating than the trapezoidal, brushless dc designs. Both brushless dc SPM and the IPM designs tend to have high torque ripple. The IM has the lowest torque ripple of any other type, but it requires a supply of magnetization current from its inverter, thus increasing the inverter kVA rating. IMs for hybrid propulsion require that a rather large fraction of input VAs be dedicated to magnetizing the machine. IMs of several hundred to thousands of horsepower have a much small fraction of input VAs dedicated to magnetizing the machine. The doubly salient machines have many desirable features for hybrid propulsion and are beginning to be looked at more seriously. The VRM has had hybrid propulsion advocates for many years, but structural design to maintain the tight airgaps necessary and controller algorithms capable of real-time current waveshape control based on rotor position have been problematic. The VRM is capable of the wide CPSR as is the IPM, and so should find application to power split and other hybrid propulsion architectures. To summarize the comparisons of the various electric machine technologies for application to hybrid propulsion, it is necessary to comment on their torqueproducing mechanisms. We close with comparisons of electromagnetic torque production in the machines evaluated so far (Table 5.4). Table 5.4 Electric machine torque production Definitions P = number of poles M = number of electrical phases Ip = phase current Lp = phase inductance l = flux linkage Lm = magnetizing inductance Lr = rotor inductance = Lmr + Llr q = rotor angle (rad) Synchronous machine (SPM, IPM, SyncRel) Asynchronous machine (IM) under rotor field oriented control (i.e. lqr = 0) Variable reluctance machine (VRM)

Expression for torque

T em ¼ m2  P2 ðldr I qs  lqr I ds Þ T em ¼ m2  P2  LLmr ldr I qs T em ¼ 12 I 2p

dLp ðqÞ dq

Electric drive system technologies

269

The specific torque density of the electric machines shown in Figure 5.16 is the most important metric of general applicability to hybrid propulsion. With the aid of Reference 5 and prior developments in this book, a short summary is made of the specific torque density of these electric machines. Table 5.5 Specific torque density figure of metric Electric machine type

Specific torque density (Nm/kg)

SPM – brushless ac, 180 current conduction SPM – brushless dc, 120 current conduction IM, asynchronous machine IPM, interior permanent magnet machine VRM, doubly salient reluctance machine

1.0 0.9–1.15 0.7–1.0 0.6–0.8 0.7–1.0

Table 5.5 lists the IM and VRM machines as having very comparable specific torque. The range is included to offset the differences in power electronics requirements. In Reference 6 a detailed comparison of both machines was made when the package volume was held constant for a hybrid propulsion application. In this work, the power inverter was remote from the M/G and not included in the metric. f 295.0 mm

D 295.0 mm

+

+

Induction machine

Variable reluctance machine

Hub Rotor Stator Iron Copper Adaptor Total

Torque density: 9.24

Electromagnetic mass: 3.6 3.6 8.35 4.6

kg kg

8.6 3.8 6.5 30.85

kg kg kg kg

8.4 3.4 6.5 26.5

S/A performance attributes: IM VRM Airgap 0.6 0.6 mm Torque 285 187 Nm At speed 500 500 rpm I bus 102 75 Adc 7.06*

Nm/kg

Torque constant: 2.8 2.5

Nm/A

Polar inertia: 0.086

kg m2

0.047

* Limited by inverter power switch rating.

Figure 5.17 IM versus VRM when machine volume is held fixed In Figure 5.17 it is instructive to note that for the same package dimensions, stator OD and length, the VRM is somewhat lower in mass (26.5 kg versus 30.85 kg) but that the IM developed higher specific torque because of limitations in the VRM power electronics at that time. The IM and VRM have comparable torque per ampere, but the VRM has much lower rotor inertia.

270

Propulsion systems for hybrid vehicles

The permanent magnets used in this design study will be sintered rare earth type. Table 5.6 lists the RE-magnet properties of most interest in a hybrid propulsion application, and these are its remanence as a function of temperature, temperature coefficient of remanence flux and bulk resistivity. Table 5.6 Properties of permanent magnets Magnet type

NdFeB SmCo5 Sm2Co17 Alnico Ceramic

BHmx (kJ/m3)

200–290 130–190 180–240 70–85 27–35

Br (T)

1.2 1.0 1.05 1.2 0.4

Hc (kA/m)

870 750 660 130 240

T op, max ( C)

180 250 250 500 250–300

Reversible Temperature Coefficient Br (%/ C)

Hc (%/ C)

0.13 0.045

0.60 0.25

0.02 0.20

+0.01 +0.40

The best magnet for an electric motor would be SmCo, owing to its high induction and simultaneous high coercive force and high operating temperature. Moreover, its reversible temperature coefficient on induction is sufficiently low to hold airgap flux density nearly constant over the normal operating temperature range of most M/Gs in use. The issue is cost; SmCo permanent magnets cost from two to three times as much per unit energy than rare earth, NdFeB. This has resulted in SmCo magnets being applied in only the most performance sensitive applications such as aerospace and spacecraft. The discussion to follow is meant as a brief introduction to the overall process of designing an M/G for a hybrid propulsion system, in this case, an integrated starter generator (ISG). For ease of explanation, an SPM machine is selected. The permanent magnet material will be NdFeB having a remanence of 1.16 T, a coercive force of 854 kA/m and a recoil permeability, mr ¼ 1.08. Equation (5.10) summarizes the calculation of induction, Bd, versus applied field intensity, Hd, due to current in the stator windings: Bd ¼ mr m0 ðH c  H d Þ Ni Hd ¼ le

ð5:10Þ

where Hc is the magnet coercive force. With a suitable magnet mounted to the SPM rotor, the resultant flux induces a voltage into the stator windings, E0, when the rotor speed is at its corner point, n0. For a given rotor speed in per unit, pu, the d- and q-axis voltages are is ¼ iq  jid pP E0 ¼ n0 ldr 60 uqs ¼ npu ðE0  X d id Þ

uds ¼ npu X q iq

ð5:11Þ

Electric drive system technologies

271

Equations (5.11) can best convey their meaning through a vector diagram in the d–q plane according to the convention for d- and q-axes given for the stator current, is. Us

jXdids uds jXsis

q

ids

is

d g

q-axis iqs

uqs

Es

ldr

d-axis

Figure 5.18 Vector diagram for the SPM machine The electrical power associated with the SPM machine is calculated in the d–q frame as shown in (5.12) and (5.13). In the latter relationship, (5.11) for uqs is substituted into (5.12): Pe

pu

¼ Refuqds iqds g

Pe

pu

¼ uq iq þ ud id

Pe

pu

ðWÞ

¼ npu ½E0 iq þ ðX d  X q Þid iq  ðWÞ

ð5:12Þ ð5:13Þ

where * is the conjugate. The currents given in (5.13) can be converted back to the stator current by using the definition of current angles shown in Figure 5.18 relative to the q-axis, and in doing so, we obtain the more common expressions for electrical power in a synchronous machine. Equation (5.14) can form the basis of the M/G sizing operation necessary to design for a specific power level, for example, peak regenerating power. To proceed from this point, it is necessary to have an understanding of what constitutes the back-emf, E0 and the expressions for d- and q-axis reactances (inductances). In a practical machine, the rotor magnets are separated from the stator bore by a physical airgap, g, in which the electromagnetic interaction takes place. For a permanent magnet of remanence, Br, the airgap flux density, Bg is given as Bg ¼

Br 1 þ ðmr g=Lm Þ

ðWb=m2 Þ

ð5:14Þ

272

Propulsion systems for hybrid vehicles

where Lm is the magnet length in the direction of magnetization (along rotor radius). A number of refinements are generally made to (5.14) to account for stator slotting (Carter coefficient), rotor curvature, magnet fringing and leakage, and other non-ideal factors. For the purpose of this development, (5.14) is sufficient. Because rotor magnets have finite interpolar gap (if made from a ring magnet), or an intentional circumferential gap to minimize the magnet material and to develop a desired flux pattern, it is necessary to calculate the fundamental component of the magnet produced flux density for a given arc segment of material. The arc segment length is taken as the ratio of magnet pitch, tm, to stator pole pitch, ts. For this discussion, the magnet to pole pitch ratio is am. From this consideration, the fundamental component of magnet flux density in the airgap becomes, from (5.14): p  4 ð5:15Þ Bg1 ¼ Bg sin am p 2 The back-emf according to Faraday’s law is due to the rate of change of total flux linking the stator coils. In the M/G development under consideration, the stator coils in a phase are assumed to be all connected in series. The total flux per pole is now Dsi am s1 hBg P s1 ¼ 0:97 fp ¼ p

ð5:16Þ

0:7 < am < 0:9 where Dsi is the stator bore diameter, s1 the stacking factor of stator laminations, h the stator stack length and am ¼ 0.8 typically. The speed voltage induced into the stator coils is composed of a stack up of individual coil turn emfs having various angular relations to the composite voltage due to their placement in slots, whether the coils are full pitched over a pole or short pitched, and whether the stator slots are skewed, or more practically, whether the rotor magnets are skewed in the axial direction. Derivations for distribution, pitch and skew factors can be found in many texts on machine design. For the purpose here it is important to realize that the winding factor, kw, is less than unity. The SPM internal emf is now rffiffiffi 3 2pf k w N s fp E0 ¼ 2 ðVrms , line-to-line) ð5:17Þ kw ¼ kd kpks N s ¼ PN c where Nc is the number of turns per coil, per phase, per pole, and Ns is the total turns in series per phase. It is still not possible to evaluate the M/G power capability since the variables for machine reactance (speed times inductance) listed in (5.13) are not known. Therefore, the next step in the design of the SPM machine is a determination of the

273

Electric drive system technologies

stator inductance. The total self-inductance of a stator winding is taken as total turns in series squared times the magnetic circuit permeance, which, in terms of its constituent parts, can be stated as Lp ¼ N 2s G Lp ¼ Lms þ Lsl þ Let

ðHyÞ

ð5:18Þ

where phase inductance, Lp, is composed of magnetizing inductance defined as that fraction of the total stator flux that links the rotor, a slot leakage term for flux that crosses the stator slots transversely and does not cross the airgap, and an end turn leakage flux due to flux on the ends of the machine that neither crosses the airgap nor links the rotor. At this point, it is essential to clarify what is meant by airgap. Let kc be the carter coefficient, the modifier to physical airgap that accounts for the presence of open slots. Then the magnetic equivalent airgap, g0 , for the various machine types is as given in Table 5.7. Table 5.7 Air gap of various electric machines BDCM

SPM 0

0

g = kcg + Lm

g = kcg + Lm

IPM 0

g = k cg

IM 0

g = k cg

The constituents of phase inductance listed in (5.19) are ð m0 Dsi h 2p 2 Lms ¼ N ðqÞdq 2g 0 0 ðHyÞ  2 pm Dsi h N s Lms ¼ 0 0 P 2g

VRM 0

g =g

SRM g0 = g

ð5:19Þ

There can be some discrepancies in the interpretation of (5.19), particularly in the definition of the winding function, N(q), in the case of a P-pole machine. The second expression in (5.19) gives the magnetizing inductance of a P-pole machine in terms of its stator bore, stack length, h, and total series connected turns: Lsl ¼ 12N 2s h

rs Qs

ð5:20Þ

where the variable rs is the slot geometry describing the slot permeance and Qs is the number of stator slots, which is equal to the number of coils in a 2-layer winding. Slot leakage inductance is very design dependent, but the relationship given in (5.20) is what is typically used to compute its value:   m0 PN 2c Det Det ln 4 2 Let ¼ 2 rffiffiffiffiffi GMD ð5:21Þ Sa GMD ¼ 0:447 2

274

Propulsion systems for hybrid vehicles

where Det is the end turn diameter assuming circular end turn geometry and GMD is the geometric mean distance of conductors within a bundle of square cross section taken as half the slot area. The expedient of taking the end turn bundle as having a cross section equal to one-half the slot area from which it connects is true for double layer windings in which two coil sides are present per slot. Stator resistances in the machine are dependent on the machine geometry and number of turns. In addition, factors such as stranding (i.e. number of conductors in hand) and end turn design all impact the calculation of winding resistance. The machine parameters developed above give a complete picture of the design process necessary to develop an M/G for hybrid propulsion. The next steps in this process are to assess the torque, power and speed capabilities of the machine to determine its performance against the hybrid M/G targets. If the performance is adequate, then the machine is simulated for performance and economy over regulated drive cycles. In this case, an accurate efficiency map of the machine is necessary to account for losses during the dynamic drive cycle. Machine losses and efficiency mapping procedures are given in Chapter 8.

5.2.4

Dual mode inverter

A very recent innovation to SPM brushless dc or ac motor control has been the development of the dual mode inverter control (DMIC) concept by engineers at Oak Ridge National Laboratory and the University of Tennessee [3]. In this concept, a cascade converter is used wherein the base converter, a power MOSFET design rated 1 pu voltage and 1 pu current, is controlling the SPM motor well beyond base speed. This is made possible by the cascaded thyristor stage having rating 6 pu voltage and 1 pu current. When the motor voltage rises to 6 pu (CPSR ¼ 6:1) at six times base speed, the thyristor stage begins to block braking current developed by the motor as it tries to back-drive the base inverter through that inverter’s inverse diodes. By inhibiting flow of braking current, the SPM motor torque can be held more or less uniform under phase advance control, up to 6:1 base speed. Theoretically, infinite speed is achievable, but in practice, speeds of 6:1 have been demonstrated in the laboratory. T1

Power electronics Rd

T2 Ri

T3

SPM

Vb T4 T5

Cmds

Control electronics Controller, Comm. gate drives, Pwr supply

T1 T6 T6

Figure 5.19 Dual mode inverter concept

w Transmission T Driveline

Electric drive system technologies

275

Notice in Figure 5.19 that in addition to the normal inverter bridge and its gate drive and controller, there is a requirement for a second set of controller commands and gate driver signals for the six thyristors. The distinction at this point is that the cascade thyristor gates must be driven by fully isolated gate drives. Since thyristors such as the SCRs shown require only a gate pulse (typically 10 ms) of several amperes magnitude, a relatively compact isolation transformer and driver transistor suffice. In operation, the DMIC controller commands the MOSFET bridge gates in the same fashion as depicted in Figure 5.20, except that when commands Aþ and A are sent to the MOSFET gates, this same gating signal (leading edge pulse) is sent to the corresponding thyristor gates, T1 and T2. Once fired, thyristor T1 conducts motor drive current for the positive half cycle (and T2 for negative half cycle). However, at some point into the conduction of phase A current, the motor back-emf will equal the supply voltage, and thyristor T1, for example, will then naturally commutate off and regain its forward and reverse voltage blocking capability. Once recovered, thyristor T1 will stand off the motor back-emf potential. The same procedure applies to the remaining phases so that motor braking current is inhibited, there is no diode conduction, hence no loss of torque and full function is maintained. In the generating mode, the thyristors must be again gated on to permit current flow out of the motor phases that is 180 shifted from its motoring polarity. Under regeneration mode, a conventional thyristor is unable to naturally commutate off, so operation into field weakening range should be blanked. This is a disadvantage of the DMIC concept, but not a strict liability, since replacement of the SCRs with GTOs or other device capable of being force commutated will provide full 4-quadrant capability to 6:1 CPSR. Also, various ac switches are under development that would make an excellent match to the DMIC inverter cascade stage. The limitations of the DMIC with SCR thyristors are shown in Figure 5.20 with solid and dotted traces for motoring and generating capability curves. Without the feature of forced commutation, the cascade converter cannot block braking currents from the SPM motor when its speed is in an overhauling condition. The SPM back-emf in that case will exceed the bus voltage, and once an SCR is gated on, there will not be an occasion for natural commutation off during the half cycle before the 1 pu bridge is into an overvoltage condition. With force commutated thyristors, the cascade stage is commanded on with gate pulses and commanded off with negative gate pulses (GTO switch), or ac switches with forward and reverse voltage blocking capability are used. The ac switches are capable of bidirectional current conduction and bidirectional voltage blocking. Figure 5.21 illustrates five classes of ac switches that are available for use in the DMIC inverter. Transistor-based ac switches maintain conduction while base current or gate voltage persists. Thyristor-based ac switches conduct after being pulsed on, and conduction is only extinguished when the circuit current naturally reverses or when the gate electrode is pulsed negative. There are two disadvantages associated with thyristor ac switches: low switching speed and turn-off capability. Thyristors are known to have latch-up problems or commutation failures (GTO). Lack of switching frequency is a major

276

Propulsion systems for hybrid vehicles T (Nm) P (kW)

Motor

Power

Torque

0

1

2

With SCR switches

3

4

5

6

Speed (pu)

With GTO or ac switch

Generator

Figure 5.20 DMIC capability curves with SCR and with ac switches A

C

B

D

E

Figure 5.21 Classes of ac switches for use in DMIC concern in the DMIC; because electric machine speeds are high, the base frequency can be in the kilohertz range. Transistor ac switches are preferred. Topology ‘C’ in Figure 5.21 is a conventional ‘totem’ pole phase leg arrangement that comes with simple and cost-effective

Electric drive system technologies

277

gate drivers. Furthermore, topology ‘C’ has low conduction losses and is amenable to solid-state integration. A disadvantage of ‘C’ may be the need for switching snubbers.

5.3 Interior permanent magnet There has been a great deal of writing on IPM machines during the past three decades since their inception as an energy conservation improvement of line-start IMs. During the 1970s, the buried magnet machine was subject to intense research, leading to it being proposed as an alternative to high efficiency IMs in low power applications of 2–25 horsepower (hp) [7,8]. Conventional line-start IMs suffer from continuous I2R losses in the rotor and the consequent I2R losses in the stator necessary to supply the rotor magnetizing currents. This, coupled with the availability of improved ferrite magnets, and then of NdFeB rare earth permanent magnets, contributed to the increased interest in a line-start buried magnet machine for commercial and industrial low power applications. In this early work, attention was focused on inrush current demagnetizing effects on the buried magnets, particularly at elevated temperatures. Ferrite magnets were susceptible to demagnetization effects at cold temperatures, and rare earth magnets were susceptible to demagnetization effects at hot temperatures in a line-start application. In recent years, attention has shifted to use of the IPM as the machine of choice for electric traction applications, particularly in the power split hybrid propulsion architecture. This choice is motivated by the IPM’s wide CPSR under field weakening control and the inherent need for wide CPSR in the power split architecture [9–20]. The most pervasive application of IPMs has been in white goods applications, such as refrigerators, washing machines and other household appliances, where the losses noted above from IMs were counter-productive in an energy conscious environment and where the durability issues of brush-type universal motors are questionable. The IPM gives the white goods designer the flexibility to eliminate rotor losses, reduce stator losses and realize approximately 10% additional torque from the reluctance component inherent in the IPM. In applications where adjustable speed is required, such as the appliances noted as well as in air conditioning equipment, these benefits are cost-effective. IPMs today fall into two broad categories depending on the permanent magnet employed: weak magnet IPMs and strong magnet IPMs. The strong magnet IPM may be more suitable to line-start applications such as large fans and industrial equipment for which asynchronous start-up and synchronous running is beneficial. Since continuous operation is at a synchronous speed, the magnets can be sized to provide ac synchronous machine performance at near unity power factor at rated conditions. Traction drives, on the other hand, have gravitated to the weak magnet IPM. This has been a somewhat surprising trend because a weak magnet IPM is in reality a VRM or, more precisely, a reluctance machine that happens to have some magnet content. The reasons for this are threefold: (1) hybrid propulsion

278

Propulsion systems for hybrid vehicles

architectures require the traction motor to operate well into field weakening, particularly in fuel cell architectures having a single gear reduction between the traction motor and wheels, (2) safety reasons so that over-speeding the traction motor by the engine in a gasoline–electric hybrid will not backfeed the dc link in the event of loss of d-axis current regulation by the inverter. Concurrent with this requirement is the corollary that the IPM does not develop braking torque should the inverter switches all turn off, regardless of speed and (3) a weak magnet IPM is more cost-effective because ferrite magnets, bonded rare earth and other ceramic magnets are sufficient to provide the d-axis magnetization needed. The conditions stated in condition (2) above are also known as uncontrolled generation (UCG) mode and represent a significant design constraint on the application of IPMs in traction drives. The following subsections will treat the various IPM designs in more detail based on the above constraints for vehicle hybrid propulsion systems.

5.3.1

Buried magnet

The most common IPM machine has the rotor geometry of the original buried magnet design. This single buried magnet layer design is illustrated in Figure 5.22 along with a dimensioned magnet slab for reference. q ym

d

Lm

Dm Wm Permanent magnet

Figure 5.22 Original buried magnet rotor design In Figure 5.22 the magnet length, Lm, is its thickness in the direction of magnetization. Each face of the buried magnet is aligned with the rotor d-axis so that alternating north and south poles are equally spaced about the rotor circumference. The interpolar gap of a buried magnet IPM is filled with soft iron as shown in Figure 5.22 along the machine’s q-axis. The magnet slabs themselves are most suitably inserted into the rotor magnet cavities in an as-pressed state so that no additional machining is required during assembly. Ceramic magnets (barium and strontium ferrite) are easily magnetized in situ. Rare earth magnets, including bonded rare earth, are best pre-magnetized and then inserted into the rotor cavities.

Electric drive system technologies

279

The caption of Figure 5.22 describes the magnet length dimension along its direction of polarization, the width as lying in the rotor tangential direction and depth as its axial length along the rotor axis. Recall that an IPM can have almost an infinite variety of magnet to reluctance torque ratios as the magnet strength ranges from weak to strong. It is common in fact to describe the buried magnet machine from the perspective of its characteristic current, Ic, defined according to (5.22) as the ratio of magnet flux to stator direct axis inductance: Ic ¼

ym Lds

ðApk )

ð5:22Þ

where the magnet flux linkage times speed corresponds to the machine back-emf voltage as follows: U oc ¼ wym

ðVpk )

ð5:23Þ

In (5.23), the inverter must supply d-axis, or demagnetizing current, of this magnitude to suppress the magnet voltage given by (5.23) in field weakening. For a given rated inverter current, there are now three variations in the IPM design, according to (5.22), that must be considered. In these cases, the value of the IPM characteristic current is compared to the inverter rated current, Ir. ●





When ym/Lds < Ir, the inverter has sufficient current overhead to source q-axis current and hence produce torque at high speeds while the d-axis component of inverter current sustains the field weakening. In this regime, IPM power at high speeds drops below its peak value but does not decrease to zero. When ym/Lds ¼ Ir, the output power of the IPM is sustained at high speeds and monotonically approaches its maximum value. This is the important class of theoretical infinite CPSR that the IPM is noteworthy of. When ym/Lds > Ir, there is a finite speed above which the IPM output power has peaked and decreases monotonically to zero. This is understandable because according to (5.22) the inverter has insufficient current rating to completely suppress the magnet emf. The inverter simply cannot deliver q-axis current to the machine, so its output decreases monotonically.

This analysis simply illustrates the fact that for buried magnet designs the inverter must be overrated, or the machine must be overrated, in order to develop the targeted CPSR desired. A valuable metric to assess buried magnet machines is its saliency ratio, x, defined as the ratio of q-axis stator inductance to its d-axis inductance as follows: x¼

Lqs Lds

ð5:24Þ

The influence of saliency ratio, x, on IPM machine performance in the case of an inverter fault was studied by Tom Jahns [9] and was found to have an unsettling effect on UCG mode operation when x > 2. In this regime, all IPM machines for

280

Propulsion systems for hybrid vehicles

which x is a number greater than 1 are prone to UCG when the rotor speed is above some threshold that is less than the speed for which Uoc ¼ Udc, suggesting that some chaotic behaviour has set in. By chaotic behaviour, it is meant that the IPM can either generate in the UCG mode a current given by (5.22) or not generate any current at all should the inverter switches be all gated OFF. Figure 5.23 is a reconstruction of a figure used in Reference 9 to explain this effect. During the UCG mode of operation, the inverter active switches are gated OFF so that only the uncontrolled rectifiers, the switch inverse diodes, are able to function in a normal Graetz bridge fashion. This means that the IPM stator current and its terminal voltage are operating at unity power factor the same as, for example, in the Lundel automotive alternator fitted with a diode bridge. The exception in the case of IPM, however, is that rather than balanced d- and q-axis inductances the IPM has a saliency ratio, and therein lies the difference. Depending on the IPM speed and loading, the vectors in Figure 5.23 assume different proportions so that above a threshold speed it is possible to enter into UCG mode. jXqIrq

q-axis wym jXsIr

jXdIrd Vs

–Ird d-axis Ir

Irq

Figure 5.23 IPM machine phasor diagram during UCG mode (derived from Figure 4 in Reference 8) As can be seen in Figure 5.23, the IPM internal voltage due to magnets can be lower than the terminal voltage, Vs, during UCG mode of operation. This is possible in this situation because of the large q-axis inductance and low d-axis inductance. The stator current through the inverter diodes, Ir, is, of course, 180 out of phase with the terminal voltage during UCG mode, just as it is in a synchronous alternator. An important consideration for hybrid propulsion when using the buried magnet variety of IPM and for which a CPSR of not greater than 4–5 is desired

Electric drive system technologies

281

according to Reference 9 is that the saliency ratio must be at least 9:1. This is, of course, very difficult to achieve in practice due to the need for rotor iron bridges and posts to secure the soft iron pole shoes over the magnets and not have issues with rotor retention. The second design constraint for hybrid propulsion is that for the buried magnet IPM to be immune to UCG will require that it be designed for CPSR < 4. This means the lowest possible magnet flux that will suffice to meet application design torque requirements. Otherwise, the IPM will have some speed regime where an inverter shutdown may result in UCG output and consequent overvoltage being imposed on the traction battery, other energy storage system, or concerns with active device voltage ratings (aluminium electrolytics fall into this category as well).

Example 4: The Camry Hybrid traction motor, MG1, is directly coupled to the Camry wheels via a compound planetary, with stage 2 epicyclic gear E2 ratio of kE2 ¼ 2.478. The total final drive ratio, gfd ¼ 3.542 and dynamic rolling radius, rw ¼ 0.355 for the tyres. Assess the following: (a) The vehicle speed, V, for which MG2 angular speed is 5,000 rpm. (b) Given that MG2 open circuit voltage, Uoc ¼ 150 Vrms line-to-line, at 5,000 rpm, determine the PM voltage constant for this IPM electric machine (i.e. use (5.23)). MG2 is a P ¼ 8 poles, 3-phase IPM with Q ¼ 48 slot stator. (c) What is the voltage of MG2 at vehicle WOT? (d) Using data on the Camry MG2 magnets (IPM with single layer chevron geometry, 145 ) Lm ¼ 6.6 mm, g ¼ 0.73 mm and NdFeB with recoil permeability mr ¼ 1.05, calculate the airgap flux during static conditions. Solution: (a) For the single-mode eCVT with the stated gearing, and MG2 geared from the driveline, the vehicle speed corresponding to MG2 angular speed n ¼ 5,000 rpm is V¼

rw wMG2 0:355ð523:5Þ ¼ 21:2 m=s ð47:4 m=hÞ ¼ 3:542ð2:478Þ g fd k E2

(b) This is a straight application of (5.1) and with care about units: ym ¼

U oc phase , we

p we ¼ wMG2 ¼ 4ð523:5Þ ¼ 2,094 rad=s 2

282

Propulsion systems for hybrid vehicles pffiffiffi pffiffiffi pffiffiffi 2U oc phase 2ð150= 3Þ ym ¼ ¼ 0:058 Wb-t ¼ we 2,094

(c) At VWOT the line-to-line voltage of MG1 is wMG2

WOT

¼

P g fd k E2 4ð3:542Þð2:478Þ  ð59:3Þ V WOT ¼ 2 0:355 rw

¼ 5,863 rad=s U oc

WOT

¼ wMG2

rffiffiffi 3 ¼ 416:5 Vrms WOT ym ¼ 5,863ð0:058Þ 2

(d) Recalling the relationship for B(H) from Exercise 1, and referring to figure 5.1, then the B–H characteristics given the fact that this NdFeB magnet has Br ¼ 1.19 T, Hci ¼ 14.46 kOe at moderate operating temperature and for the specifics of the magnet stated the static permeance coefficient, Pc, can be found, and from this an approximation of airgap flux density, Bg ¼ Bm. Pc ffi

  Lm Ag 6:6 ð< 1Þ  5 ¼ 0:733 g Am

Bg ¼

Pc 5 1:19 ¼ 0.98 ðTÞ Br ¼ Pc þ mrec 6:05

In part (d), there are approximations based on geometry and leakage factors influencing the IPM magnet flux that are outside the scope of this text. Suffice to say that leakage due to magnet end effects, flux shunting via magnet pocket bridges and other non-ideal factors all act to decrease the permeance coefficient – the goal of course being to realize as high an airgap flux density as possible. More is discussed on this topic of IPM electric machines in the exercises at the end of this chapter. The reader is advised to consider these reinforcements on the topics discussed here.

5.3.2

Flux squeeze

It is somewhat misleading that the flux squeeze rotor geometry is believed by many to be superior to the buried magnet design. This is true, but for some very restricted applications to be discussed shortly. The flux squeeze design appears tailored to ceramic magnets because the large magnet faces are available to force significant levels of flux density in the machine airgap. Figure 5.24 illustrates the flux paths in the flux squeeze geometry.

Electric drive system technologies

fp

283

fp

fp fp

Figure 5.24 Flux squeeze interior permanent magnet machine The rotor centre of the flux squeeze design must be made of non-magnetic material or some design of iron bridges and cavities so that the ends of the rotor magnets are not shorted out. In the flux squeeze IPM, the magnets are mounted in the tangential direction and the flux path is completed as shown by the soft iron wedges set between the magnets. By designing the rotor so that the magnet face area is large compared with the airgap surface, the flux in the gap may be very high. For example, with ceramic magnets having remanence Br ¼ 0.25 T, the flux density in the gap may be 0.82 T. This provides rare earth magnet performance for approximately a tenth of the magnet cost, but at the expense of a much larger diameter rotor. Because a portion of the magnet lies at the machine airgap, the magnets themselves must have sufficiently high coercivity so as not to demagnetize beneath the strong demagnetizing fields of the stator. High coercivity magnets such as barium ferrite and NdFeB rare earth magnets do well in this geometry. It is also noteworthy that for the flux squeeze design, the saliency ratio x < 1 since Lds > Lqs. The d-axis in fact lies completely in rotor iron, and the q-axis interestingly lies completely in magnet material. The saliency ratio can be very low in this regard, or viewed from another perspective, x1 . . . 1. Another difference of the flux squeeze compared to the buried magnet IPM is that now the mmf across each magnet is twice the mmf across the airgap. This is true because the airgap flux over the soft iron pole face is the composite of flux from two magnets. This type of machine has many proponents for various hybrid propulsion systems. Honda, for example, uses a variation of the flux squeeze IPM in its hybrid designs. The reason for this is that the volumetric and gravimetric power output of electric machines for hybrid propulsion, as well as aerospace, must be as high as possible and the flux squeeze IPM does deliver high specific output, but for relatively small ratings. This latter fact does not appear to have been made sufficiently clear in the hybrid propulsion design camp. In Reference 11 the investigators proposed an optimization procedure in which contours of constant volume are presented for IPM machines in the 0.25–10 kW range and for speeds in the 10–100 krpm range. It is

284

Propulsion systems for hybrid vehicles

further interesting that rare earth magnets in the flux squeeze design do not fare significantly better than ferrite magnets. The rare earth magnet design gained only 10% higher power density. The ferrite magnet design, however, must be cooled so that the rotor temperature does not exceed 30  C of ambient since its temperature coefficient of remanence is 0.17%/ C. Compared to a conventional synchronous machine, SPM design in this case, the flux squeeze IPM has higher specific output over the range of 1.3–4.2 kW when the speed regime ranges from 10 to 100 krpm in aerospace applications.

Speed (krpm)

100

Vol. ratio 0.8

0.9

1.0

1.1

80 60 40 20 10 0.25

1

2 1.3 kW

3

4

5

4.2 kW

6 7 8 Power (kW)

Figure 5.25 Constant volume contours of flux squeeze IPM/SPM designs

The flux squeeze design has an advantage over the SPM in terms of specific power density only for small size machines, 1.3–4.2 kW, as shown in Figure 5.25, and for high speeds. The design size of larger machines is restricted more by winding temperature than by airgap flux density, so that conventional large IPM designs tend to have higher power density than SPM designs. Comparisons of motor performance based on machine volume are common in the automotive and aerospace industry. This is because package volume is generally very restricted and costly in both industries. Other investigators have used various techniques to compare various electric machines for specific power output [21]. Comparisons of performance based on flux–mmf diagrams have also come to the same conclusions as those stated above regarding the IPM in contrast to SPM and other machines. In the analysis and design experiments performed by the authors of Reference 21, all the machines studied were designed to occupy the same volume so that valid comparisons of specific power, torque and torque ripple could be realized. Figure 5.26 illustrates the relative ranking of the machine types studied thus far in this chapter. The basis for comparison is that the machines fit the standard D132 IM frame size, airgaps are identical at 0.5 mm, slot fills are held fixed at 40% and total copper losses are fixed at 634 W (115  C rise and 7.5 kW continuous power).

Electric drive system technologies

285

Comparison of machine types Brushed-dc IPM SPM-dc SPM-ac VRM SREL IM 0

20

40 60 Torque (Nm)

80

100

Figure 5.26 Relative ranking of machine types based on peak torque ( from Reference 21) In Figure 5.26 the IPM machine compares very favourably with the surface PM designs operating with sinusoidal flux (ac) and trapezoidal flux (dc) for a fixed frame size and total electromagnetic volume. The torque for these designs is given in absolute terms. Torque ripple is a major consideration in hybrid propulsion and must be included in any comparison of machine types. In Figure 5.27 the corresponding torque ripple is plotted, again from data presented in Reference 21. Comparison of machine types Brushed-dc IPM SPM-dc SPM-ac VRM SREL IM 0

20 40 60 Torque ripple (%)

80

Figure 5.27 Machine comparison based on output torque ripple ( from Reference 20) It is much clearer from the discussion above and by reviewing Figure 5.27 that the IPM machine has the torque ripple character of a VRM, since it in essence is a reluctance machine. This characteristic has significant bearing on its application in hybrid propulsion not only because of its high ratings but also because driveline inertia will suppress torque ripple to be a minor issue. But the fact remains that IPMs still have more torque ripple than brushless ac machines and certainly more ripple torque than an IM.

286

Propulsion systems for hybrid vehicles

Another recent variant on the IPM has been a novel rearrangement of the permanent magnets to reside in cavities that closely resemble the pattern of a wound field synchronous machine. The rotor continues to have magnet cavities and iron bridges to support the structure. Figure 5.28 illustrates this unique design referred to as a permanent magnet reluctance machine (PRM). Magnet flux in the PRM follows a high reluctance path as it does in the normal IPM, but q-axis flux from the stator follows an iron path through the rotor so that there is a cross-field in the rotor laminations.

S S N

N

S

N N S

S

N N

N S

N S

S

Figure 5.28 Permanent magnet reluctance machine In the PRM, the combination of permanent magnet torque and reluctance torque are suitable for hybrid propulsion systems. This can be seen from the fact that conventional IPM machines require substantial d-axis current to realize field weakening, but the PRM realizes flux control with an inverse relationship of stator current to achieve the same results. This rather obscure behaviour can be seen by comparison of d-axis currents in the plot of voltage versus speed for both no load and full load conditions shown in Figure 5.29.

PRM

Uload Id

Uload

Voltage

Uoc

Voltage

IPM

Uoc Id

Speed

Speed

Figure 5.29 Illustration of d-axis current behaviour in PRM versus IPM

Electric drive system technologies

287

The PRM is claimed to offer a CPSR of 5:1 with high efficiency of 92–97% over this range, and power levels of 8 kW and up to 250 kW are said to be possible. The reluctance torque of the PRM is 1.5 times the permanent magnet torque [22]. At maximum speed, the back-emf of the PRM is 1.3 times the rated voltage, so that minimal d-axis current is needed to perform field weakening. The reason for high efficiency in the PRM is the fact that field weakening current is only 14% of maximum inverter current at no load versus 86% in the case of a buried magnet IPM. This is significant and the reason for the high efficiency noted in Figure 5.29 for the PRM.

5.3.3 Mechanical field weakening In addition to purely electronic means of field weakening of permanent magnet machines, there have been, and continue to be, notable mechanical field weakening designs during the past few decades. This section will discuss two of the more interesting field weakening schemes. In the first scheme proposed by M. Lei and others at the Osaka Prefecture University in Japan, a moveable magnetic shunt is arranged so that as the speed increases, the IPM rotor flux is reduced [23]. The basic concept is illustrated in Figure 5.30 for the buried magnet IPM design on which it has been carried out.

(a)

(b)

Figure 5.30 Mechanical field weakening by moving iron shunt: (a) low speed position of iron shunt and (b) high speed position of iron shunt The moving iron shunt is in effect a magnetic governor that has a defined position–speed dependency set by the mechanical design and spring constant (which can be non-linear). At low speed, the spring is relaxed and the movable iron shunt is out of the flux bypass cavity. At high speed, the spring compresses due to centrifugal force, causing the iron shunts to move into the bypass cavities, thereby shunting magnet flux through the rotor iron bridge instead of allowing it to link the stator. The benefit of this mechanical field weakening scheme is that the machine efficiency is improved in the field weakening region, unlike the conventional IPM for which field weakening efficiency is low due to high d-axis currents. Figure 5.31

288

Propulsion systems for hybrid vehicles

compares the efficiency of the mechanical field weakening method with that of electronic field weakening IPM. There are obvious mechanical disadvantages with a scheme such as that shown in Figure 5.30 (such as lubrication, striction, unbalance and oscillation if not properly damped).

Efficiency

100 90 Mechanical

80 70

Electronic Speed

Figure 5.31 Efficiency comparison of mechanical versus electronic field weakening The second mechanical field weakening method has been more recently described [24] for which a mechanical spring and cam assembly is employed to shift the relative position of the magnet discs in an axial flux permanent magnet (AFPM) machine. Figure 5.32 illustrates the cam and spring mechanism as well as the implementation on an AFPM rotor. Rotor Stator Rotor Low speed

Phasor

Rotor Stator Rotor High speed

Phasor

Figure 5.32 Cam spring method of mechanical field weakening ( from Reference 23) In this mechanical field weakening scheme, the airgap flux density is not altered, so no mechanical work is done by the rotor phasing. The cam and spring mechanism need only phase the two rotor sections as a function of mechanical speed to realize field weakening. Output regulation due to loading will, of course, need to be accomplished using electronic controls. In Figure 5.32 the high speed

Electric drive system technologies

289

configuration shows that the two rotor discs have been phased such that the flux linkages in the stator are diminished. The net voltage induced into the stator coils is the vector sum of voltages due to flux from each rotor disc magnet, but the flux linkages are now out of phase, resulting in lower net induced voltage, hence the equivalent of field weakening. The conceptual lever arm shown at the right in Figure 5.32 is meant to illustrate a mechanical cam and spring assembly that actuates the rotor disc phasing. As a point of reference, this technique can be traced back to work performed by Dr Izrail Tsals, c. 1989, while he was with the PA Consulting Group, Hightstown, NJ. Dr Tsals later joined the Arthur D. Little Company in Cambridge, MA. In a mechanical field weakening scheme devised by Dr Tsals, the two layers of permanent magnets in a drum machine were phased in a manner very similar to the scheme illustrated in Figure 5.32. Mechanical springs and counterweights attached to the rotor hub were used to effect field weakening by masking off magnet flux as speed increased.

5.3.4 Multilayer designs Single buried layer IPM designs were the first to be implemented and put into production for white goods, industrial use, and electric and hybrid vehicle propulsion. Many investigators have since implemented various multilayer designs in which the magnets are inserted into radial cavities separated by soft iron and supported by iron bridges and posts. In some of this work, it has been reported that torque was improved by 10% and the high efficiency contour was also expanded by 10% when the total rotor magnet volume was held constant [25]. Figure 5.33 illustrates the improvement over a single buried magnet layer, total magnet volume constant, ferrite magnets are used. The test machines were 3-phase, 4-pole, 24-slot IPM designs having 60 mm rotor OD, a 0.5 mm mechanical airgap and remanence of 0.42 T and 280 kA/m (3.5 kOe) coercivity.

210

Double layer

180 T (Nm)

Single layer 93% 90%

0

2

4 6 Speed (krpm)

8

10

Figure 5.33 Multiple layer IPM machines for battery electric and hybrid propulsion The power curve for the single and double layer designs considered here are for x > 2 so that high speed power is somewhat lower than peak power at the corner point

290

Propulsion systems for hybrid vehicles

speed. Corner point speed in Figure 5.33 is 4,000 rpm and the torque is 205 Nm at stall for the double layer rotor. Maximum speed is 10,000 rpm.

5.4 Asynchronous machines The most convenient definition of an asynchronous machine is that of a singly fed ac machine in which the rotor currents are also alternating [4]. All synchronous machines have dc rotor currents from either field windings or permanent magnets. Recall that a permanent magnet may be modelled as an equivalent current sheet that produces the intrinsic coercive force exhibited by the magnet. In this section, the various types of asynchronous machines that are considered for hybrid propulsion systems are evaluated. It should be emphasized that in hybrid propulsion the need persists for machines having wide CPSR. Traditionally, asynchronous machines are capable of operating over a range of 2.5–3:1 in CPSR. This is due in some respects to the specification based on thermal constraints for peak to continuous rating of line-start applications and in some respects to the fact that inverter driven asynchronous machines are limited by the resolution of currents injected into the d-axis of the machine at high speeds. Magnetizing current requirements are low at high speed, and regulating a 10 A d-axis current in the presence of 350 A q-axis current is constrained by the sensor resolution, A/D word length and microcontroller limitations. This section starts with a brief overview of the classical IM having a cast rotor and then elaborates more on the research activities directed at improving the operating speed envelope of IMs in general. The wound rotor and other doubly fed asynchronous machines are noted but are generally not of high interest in hybrid propulsion systems.

5.4.1

Classical induction

The cage rotor IM is durable, low cost and relatively easy to control for fast dynamic response under vector control. In hybrid propulsion systems, the availability of such a rugged electric machine is very beneficial to designs in which the M/G is located inside the transmission or on the vehicle axle in the case of electric four wheel drive. There exists voluminous literature on IM design, modelling and control [26]. Our interest here will be on those attributes of IMs that make them attractive for hybrid propulsion and how this machine compares with other types. It has already been noted in section 5.3.2 that the IM does not possess the torque density of a permanent magnet design, and that is quite true because the IM must receive its excitation from the stator leading to higher VA requirements on both the stator windings and inverter drive to deliver this excitation. The IM itself is low cost for this reason, and all excitation costs are passed on to the user in the form of reactive kVA requirements. In the permanent magnet design, the machine excitation is provided during manufacturing and therefore represents a first cost to the manufacturer.

Electric drive system technologies

291

Stator yoke Stator slots Airgap Rotor slots Rotor yoke Shaft

Figure 5.34 Classical induction machine cross section Figure 5.34 shows the construction of an IM in cross section. This is a smooth rotor design in which rotor slots are typically semi-closed or fully closed for inverter drive, and the slots are relatively deep. Line-start IMs, on the other hand, will have open slots that are shallow or double cage designs in order to improve starting torque from a fixed frequency supply. With inverter supply, the frequency of the rotor currents is controlled in response to the rotor mechanical speed so that flux penetration into the rotor is not restricted by eddy currents as it would be for fixed frequency starting. To explore the IM further for application as a hybrid propulsion system starter– alternator for a pre-transmission, parallel hybrid as discussed in Reference 27, it is important to understand the slot design for both stator and rotor. Figure 5.35 is used to illustrate a practice of stator design having parallel sided teeth (iron intensive) and parallel sided rotor slots (iron intensive). The machine is designed for high overdrive conditions to meet the vehicle driveline package constraint in both axial and radial dimensions. For the lamination design illustrated in Figure 5.35 and with a 60 mm stack, the machine develops 300 Nm of torque to 1,000 rpm at the engine crankshaft. It should be noted in this IM design that stator slots are consistent with a 3-phase, 12-pole design for which q¼

Qs mP

ð#Þ

ð5:25Þ

where q ¼ 2 SPP for the parameters given. The stator winding is double layer, 5/6 pitch and has 2 turns/coil with all coils in a phase belt connected in series. The operation is heavily in saturation under this level of overdrive. The developed torque peaks at 300 Nm for 360 Apk of inverter drive at 1,000 rpm (100 Hz). In this application, it was found that the rotor teeth saturate first and to the greatest extent followed by the stator teeth second. The rotor slots should be ‘coffin’ shaped to enhance the rotor flux and limit rotor teeth saturation, but in this test machine, the original design was made using copper bars for the rotor cage and, hence, parallel sided rotor slots.

292

Propulsion systems for hybrid vehicles f 295.0

6.67

0.6

3.18

f 175.0 All dimensions in mm 72 Stator slots 87 Rotor bars Stator I.D.: 235 mm Stator slot depth: 17.23 mm Rotor slot depth: 15.40 mm Stator slot opening: 2.0 mm Rotor slot opening: 2.0 mm

Figure 5.35 Induction machine for hybrid propulsion starter–alternator ( from Reference 27) Without accounting for magnetic saturation in the stator yoke and teeth, and rotor teeth and yoke, the agreement between experiment and model would not match as well as it does in Figure 5.36. In Figure 5.37 the non-linear model of the IM is illustrated on a per-pole basis by reordering the detailed homeloidal model 350

Shaft torque (Nm)

300 250

380 348 358 360 Exp.

306

200

Model 248

150 248

100

269

50

233 163

175

0 0

1,000

2,000 3,000 Shaft speed (rpm)

4,000

5,000

Figure 5.36 IM starter–alternator torque versus speed performance

293

Electric drive system technologies

over a pole pitch with boundary conditions of 0 mmf at the q-axis of mmf. When this procedure is followed, the resulting model is obtained. Rsy/4

Rst

Rg Fm

Fsm ⫹ Fsm ⫺

Stator end winding leakage

Rrt

Stator slot and diff. leakage

Rry/4 Frm

Rotor slot and diff. leakage

Rotor Frm ⫹ end ring leakage ⫺

Figure 5.37 IM non-linear model, per pole This analysis illustrates that the IM is capable of very high torque density if overdriven well into magnetic saturation. The downside of doing this is that efficiency in the low speed, high torque regime is very low, of the order of 35%. However, because the M/G is used only transiently at these conditions (engine cranking), it is a very practical approach to meeting strict package limitations with a rugged electric machine.

5.4.2 Winding reconfiguration Expanding the constant power speed range of an IM has traditionally been accomplished using mechanical contactors. Everyday examples of such approaches include multi-speed ceiling fans that have separate stator windings for each pole number. The industrial machine tool industry uses this technique for high speed spindle applications for which CPSRs > 10:1 are required. In some spindle applications, speed ratios of up to 30:1 are necessary [28] with low speed for ferrous metal cutting and high speed for aluminium alloys. Conventional means of winding changeover have been delta-wye switching to realize a sqrt(3):1 speed change or to use series–parallel winding reconnection to realize a 2:1 speed ratio. The series–parallel winding change is most often used in industrial drives, especially for spindle applications, and it does have merit in hybrid propulsion systems. Figure 5.38 illustrates the technique employed. Stator coils in all phase belts can be tapped windings or series–parallel changeover. Figure 5.38 shows series–parallel changeover for which stator currents can actually be increased in the high speed, parallel coil configuration. With tapped stator winding, the low speed and high speed powers are different:  NL 2 NH   NL ¼ NH

PH ¼ PL I sH I sL



ð5:26Þ

where Nx refers to the stator coil number of turns in low or high speed modes. The issue with arrangements as shown in Figure 5.38 is that mechanical contactors for doing winding changeover are typically bulky and not robust enough for

294

Propulsion systems for hybrid vehicles B

Series

Parallel A

C

Figure 5.38 IM stator winding changeover method

mobile applications. Furthermore, the inverter controller must coordinate the changeover very accurately to allow time for stator drive removal, time for the mechanical contactor opening or closing and time to re-excite the IM. Mechanical changeover is accomplished with the inverter in its high impedance state so that it is not damaged by overvoltage transients due to persistence of the rotor flux. There have also been numerous designs of winding changeover that employ ac switches such as the thyristor combinations described in this chapter. Many of those schemes follow the same procedure in terms of inverter protection and application as the mechanical changeover techniques. The most common again are delta-wye, delta-2delta etc.

5.4.3

Pole changing

With IMs, it is possible to establish rotor flux having arbitrary pole number and in doing so obtain discrete steps in rotor mechanical speed. The cage rotor of an IM can be viewed as either a continuous conducting surface into which eddy current patterns can be established via excitation from the stator pole number or an m-phase winding where m equals the number of rotor bar circuits around the periphery of the machine. In either case, the pole pattern established in rotor flux is that due to the stator impressed pattern. There have been numerous attempts in the past to produce discrete speed control using an IM such as the 2:1 pole change technique developed by Dahlander in which the entire winding is utilized. Unlike conventional tapped windings and winding reconfiguration techniques, pole changing provides discrete steps in mechanical speed of 2:1, 3:1 or at arbitrary pole number ratios. Pole–amplitude modulation is another technique employed in synchronous machines for large fan drives in which the speed could be reduced by some fraction of the designed synchronous speed via a winding change. In this section, it will be shown that pole–phase modulation (PPM) is the more general class of discrete speed change for which both pole number and phase

Electric drive system technologies

295

Pole–phase modulation: p1 and p2 arbitrary; m1 and m2 arbitrary

Pole–amplitude modulation: p1 : p2 = n : (n ⫺1) m1 = m2

Dahlander connection: p1 : p2 = 1 : 2; m1 = m2

Separate windings: p1 and p2 arbitrary; m1 and m2 arbitrary Poor slot area utilization

Figure 5.39 Hierarchy of discrete speed control methods number are arbitrary. In Figure 5.39 the hierarchy of discrete speed control methods is listed for an ac machine. In general, discrete speed change by winding reconfiguration has been applied to conventional drum type machines with single, double or higher number of layer windings. The PPM technique can be applied equally well to such machines, but it has been found to be more flexible when applied to toroidally wound IMs. In Figure 5.39 px is the pole number and mx is the phase number.

5.4.3.1 Hunt winding A unique winding for IMs was discovered by L.J. Hunt and published in 1914 that described a self-cascaded IM in which windings of different pole numbers were wound on the same stator. The schematic in Figure 5.40 has become known as the Hunt winding. The rotor of the Hunt motor is wound with a pole number different from either stator windings. A typical Hunt wound, self-cascaded, IM may have two sets of stator windings, one with p1 ¼ 4 poles and the second with p2 ¼ 8 poles. If the rotor is wound having p3 ¼ 6 poles, the machine will function as a 12-pole IM. In this novel machine, the p1 winding acts as the source of excitation and the second stator winding behaves as if it were a rotor winding. The rotor p3 winding itself interprets the stator p1 and p2 fields and develops a torque corresponding to p1 þ p2. The original Hunt winding was a very early attempt at IM speed control for low speed applications from a fixed frequency supply. With resistance loading on the wound rotor, it was possible to realize high starting torque and low speed operation.

296

Propulsion systems for hybrid vehicles A

B

a

b

C

c

Frequency converter

to A

to B

to C

Figure 5.40 Schematic of Hunt winding, self-cascade induction machine Subsequent applications of the Hunt winding have led to numerous developments of a class of doubly fed induction and most recently, doubly excited reluctance machines [28–30]. The interested reader is referred to those references.

5.4.3.2

Electronic pole change

The concept of electronic pole changing has been described by various authors since the early 1970s after the availability of bipolar electronic switches. The fact that many industrial applications require operation over vastly different speed regimes had been the early motivation for such research. More recently, and especially after the early years of research and development on battery electric vehicles, it became important to extend the limited CPSR of IMs to enable high torque for vehicle launch and grades, yet maintain sustainable power at relatively high speeds. The work by Osama and Lipo in the mid-1990s was one such example of electronic pole changing in which contactorless changeover was realized by purely electronic control of machine currents [31]. In their work, Osama and Lipo described a contactorless pole changing technique that was capable of extending the field weakening range of a 4-pole IM. The machine itself was wound with six coil groups (e.g. phases) with two sets of three phases each connected to their respective power electronics inverters as shown schematically in Figure 5.41. The machine described in Figure 5.41 must be designed to sustain the stator and rotor flux of its lowest pole number operating mode. For example, if the motor is a conventional 4-pole IM and 2-pole operation is required, the injected currents must conform with the values given in Table 5.8. In this table, a minus sign signifies polarity reversal at the inverter group leg associated with that coil group. If a single dc supply is used as is shown in the figure, then the neutrals of the two 3-phase groups must remain isolated. That means that neutrals in group (1, 3, 5) must be isolated from the group (2, 4, 6). Furthermore, the rotor of the IM used for

297

Electric drive system technologies Power electronics – 2nd group

Vb

(Vf , If)2

6

4

2

5

Cmds

2

Control electronics Controller, Comm. gate drives, Pwr supply

3

1

4 6 1

3

5 (Vf , If)1

Power electronics – 1st group

Figure 5.41 Electronic pole change technique for 2:1 speed range increase 2:1 pole change must be somewhat larger than a 4-pole rotor in order to remain unsaturated under 2-pole operation. Electronic pole changing of an existing IM design often leads to oversaturation of the machine’s stator yoke (back iron) or rotor yoke. This is the case because a high pole number machine is often reconfigured electronically to a lower pole number machine – in the case of the electronic pole changing scheme covered here, by a factor of 2:1. In order to realize pole change of 3:1 or in fact arbitrary pole number changing, the techniques of PPM must be employed as will be seen in the next section [32]. Before closing this topic on electronic pole changing by a factor of 2:1, consider the impacts on the IM as illustrated in Figure 5.42. A 4-pole IM is electronically changed to a 2-pole IM by redirecting the stator currents by appropriate commands (Table 5.8). Table 5.8 Electronic pole changing technique

Inverter group 1 Inverter group 2

Ref. current

4-pole

2-pole

i1 i3 i5 i2 i4 i6

ia ib ic ia ib ic

ia ic ib ia ic ib

298

Propulsion systems for hybrid vehicles 2.5

2.0 Variables (pu)

Swe 1.5 I

1.0

Byoke 0.5 T, Bg 0

0

1,800

3,600

5,400 Speed (rpm)

7,200

9,000

Figure 5.42 IM electrical and magnetic variables during electronic pole change In Figure 5.42 the machine current is held constant at 1.0 pu over the entire speed range of >4:1 for an IM that has a breakdown torque to rated torque ratio of 2:1. The machine torque and airgap flux density over the complete range are shown as the same trace, but yoke flux density encounters a step change in magnitude when the machine is reconfigured from 4-pole to 2-pole at 3,600 rpm. The stator yoke flux in a 2-pole machine is two times the flux density in a 4-pole machine at 3,600 rpm. Lastly, the machine slip is shown rising from rated slip at the 4-pole machine configuration to twice rated slip at 3,600 rpm, the extent of its field weakening range. The slip in 2-pole configuration is shown rising to nearly 2.5 pu to illustrate the fact that IMs are capable of this range in slip control. Not shown in Figure 5.42 is the inverter control required to manage the flux linkages in the machine during the pole changeover. The controller must regulate d-axis current into the machine to restore the flux to rated value of a 2-pole machine at 3,600 rpm. With such continuous excitation, there will be a transient in currents and flux linkages lasting for approximately one rotor time constant as the flux re-establishes itself to a new steady state. The ability to reconfigure pole number without loss of excitation is one of the major benefits of electronic pole changing, a benefit that is as important for industrial machine tool drives as it is for battery electric and hybrid electric propulsion systems.

5.4.3.3

Pole–phase modulation

PPM is the most general method for discrete speed control of an ac machine fed from a constant frequency source. Referring again to Figure 5.39, let p1 denote the number of pole pairs and m1 the number of phases at one synchronous speed and p2 the number of pole pairs and m2 the number of phases at the second synchronous speed, then the various combinations of PPM can be explained as follows. Not only does it enable a variation of pole numbers, but the number of phases can change along with the number of poles. PPM can be implemented in machines with

Electric drive system technologies

299

conventional windings having both sides of coils in airgap slots, as well as in toroidally wound machines, with only one coil side in airgap slots. The implementation consists of selecting the number of pole pairs by controlling the phase shift between currents in the elementary phases, where each elementary phase consists of a coil, or a group of coils connected in series. As opposed to Dahlander’s connection, which allows only one, 2:1, ratio between the number of pole pairs created by a single winding, the number of pole pairs in PPM is arbitrary. The Dahlander winding is usually built with full pitch at lower speeds of rotation, and, therefore, with half the pole pitch, that is y ¼ tp/2 at higher speeds of rotation ( y denotes here the winding pitch and tp the pole pitch, both expressed in the number of slots). The PPM winding with conventional coils, on the other hand, is always built to have full pitch at higher speeds, when the number of pole pairs at lower speeds is odd, and a shortened pitch at higher speeds of rotation, when the number of pole pairs at lower speeds is even. The number of pole pairs PP is a function of the total number of stator slots N, the phase belt q and the number of phases m according to (5.27): PP ¼

N 2qm

ð5:27Þ

where PP and m must be integers, and q is usually an integer. This means that an m-phase machine with N slots can be built having several pole pairs, the numbers of which depend on the value of q. In this example, a 72-slot toroidal stator is assumed, and a 12-pole/4-pole toroidal winding is used, because a toroidal winding allows much more freedom in PPM design than a conventional one. In this example, the IM is connected to a 9-leg, 18-switch inverter. The toroidal machine phase belts for 12-pole and 4-pole configurations are defined as q12 ¼

72 6 ¼ , 12m12 m12

q4 ¼

72 18 ¼ 4m4 m4

ð5:28Þ

where m is the number of phases, 72 is the number of stator slots and q is the corresponding phase belt, expressed in a number of slots. An additional constraint is that q12 ¼ nq4

ð5:29Þ

where n is an integer. Finally, the last condition is that the sum of all line currents is zero. Equation (5.28) shows that with the 72-slot machine, the maximum number of phases (neglecting all other considerations) can be 6 for a 12-pole connection and 18 for a 4-pole connection. Having different number of phases for these two configurations would lead to an inefficient use of current sensors. In order to minimize

300

Propulsion systems for hybrid vehicles

the number of current sensors, a 3-phase winding was selected for both configurations, that is m12 ¼ m4 ¼ 3

ð5:30Þ

The phase belts for 12-pole and 4-pole configurations are consequently: q12 ¼ 2,

q4 ¼ 6

ð5:31Þ

Before presenting the complete winding diagram, it is useful to define the elementary machine coil and the coil polarity as illustrated by referring to Figure 5.43. Coil #

71 +

72 +

1 +

2 +

Figure 5.43 Schematic representation of elementary coils in a toroidally wound machine. All coils are wound in the same directions; the front end of the coil is designated as (þ); the back end is (). Each winding is obtained by connecting coils as in Figure 5.45 It is important to recall that in a toroidal machine, each coil side in the airgap conducts current in one direction only as denoted in Figure 5.43. To obtain the same effect as in a standard IM (where each coil conducts the current axially in both directions; see Figure 5.44), here, two elementary toroidal coils are connected in series. These two coils then form a coil group, so that the number of coil groups (i.e. effective coils) in a 72-slot machine is 36.

Figure 5.44 Schematic representation of a conventional coil. The current flows in both directions – compare with toroidal coil connection, as shown in Figure 5.43 One possible winding connection, for a 3-phase toroidal machine, which satisfies all requirements, is shown in Figure 5.45. Coils #1 and #2, #19 and #20, #55 and #56 form one branch etc. The top point of each branch (þ1, þ21, þ5, þ25, þ9, þ29,

301

Electric drive system technologies

þ13, þ33 and þ17) is connected to a corresponding inverter totem pole mid-point as shown in more detail in Figures 5.46 and 5.47. A () phase sign for the 4-pole connection means that the current entering the connection point of the corresponding branch (þ21 for –A, þ9 for –B and þ33 for –C) has the opposite direction from the currents entering the other two branches belonging to the same phase. Pole changing is performed by inverter control, by re-assigning coil strings to different phases without the need of any mechanical contactors as shown in Figures 5.45–5.47. Indirect vector control, with a shaft encoder, is used for both motor and generator operation. Because the changeover from 12-pole to 4-pole is performed by reconnecting the stator coils, the machine flux is reduced to zero during the transfer. When the ratio of two speeds is 1:3, 1:5, 1:7 etc. (1:odd number), the coils usually have full pitch at lower polarity. When the ratio of two speeds is 1:2, 1:4, 1:6 etc. (1:even number), the coils must have shortened pitch at lower polarity. A special case of PPM winding with speed ratio 1:2 is Dahlander connection, in which the coil pitch at lower polarity is 50% shortened in order to give a full pitch at higher polarity.

p = 12 A Phases: p =4A Inverter totem-pole + connection 1 points – + 2 – – 19 + – 20 + + 37 – + 38 – – 55 + – 56 +

B

C

A

B

C

A

B

C

–A +

A +

B +

–B +

B +

C +

–C +

C +

5

21 – + – –

– –

+ –

+ –

+ +

+ +

– +

– +

– – + –

+ –

+

+ –

+ –

+

+

– –

+ –

– – 71

+ –

+ – 72

52 +

– + 54

51

68

48 +

– +

– –

+ –

+ + 53

70

67

+ – 36

+ +

– +

– –

35

69

50

47

64

44 +

– – 63

+ +

– +

– –

+ – 16

49

66

18

15

32

– +

– –

+ –

+ +

– +

– –

31

65

46

43

60

40

– +

– – 59

39

+ + 45

62

42

58

+ +

– –

+ – 12

– + 34

14

11

17

33 – +

– –

+ – 28

61

41

57

– – 27

8

24

4

– + 30

10

+ –

13

29 – +

– – 7

23

3

– + 26

6

22

9

25 – +

+

+

Figure 5.45 Winding connection for 4-pole and 12-pole configurations. The coil polarity is shown in Figure 5.40. The coil connections and the inverter connection points (indicated above) are fixed. Pole change is performed by assigning coil strings to the appropriate phases, through inverter control (see Figures 5.46 and 5.47.)

302

Propulsion systems for hybrid vehicles B

C

+ 1 –

+ 21 –

5

– 56 +

– 40 +

– 60 +

Phases: A

+ –

A

B

+ 25 –

9

– 44 +

C

A

B

C



+ 29 –

+ 13 –

+ 33 –

+ 17 –

– 64 +

– 48 +

– 68 +

52

+



– 72

+

+

Figure 5.46 12-pole inverter connections and control. The elementary coils are connected in series, as shown in Figure 5.45. All nine inverter totem poles are used

Phases: A

–A

A

B

–B

B

C

–C

C

+ 1 –

+ 21 –

+

+ 25 –

+

5

9 –

+ 29 –

+ 13 –

+ 33 –

+ 17 –

– 56 +

– 40 +

– 60 +

– 64 +

– 48 +

– 68 +

52



– 44 +



– 72

+

+

Figure 5.47 4-pole inverter connections and control. The elementary coils are connected in series, as shown in Figure 5.42. All nine inverter totem poles are used The conventional winding connections for (1:odd number) and (1:even number) combinations will be illustrated by two examples. ●

Speed ratio (1:odd number) – full pitch winding at lower polarity. Consider an AC winding that has to operate in 4-pole and in 20-pole connection. For PPM implementation, the winding will be double layered and will have 5 phases at 4 poles (m4 ¼ 5) and 2 phases at 20 poles (m20 ¼ 2). Phase belt at 4 poles is q4 ¼ 2, and at 20 poles q20 ¼ 1. Coil pitch is expressed as the number of teeth, y ¼ tp,4 ¼ 10, and the winding is placed in 40 slots.

Winding configuration at two polarities is shown in Figure 5.45. For purpose of clarity, only coils belonging to 1-phase, that is, carrying the same current, are shown in this figure. Current direction in the coils is given by the arrows. For the 4-pole connection illustrated schematically in Figure 5.48 (top), the adjacent two coils belong to the same phase (q4 ¼ 2). The pole areas are denoted by N4 and S4.

Electric drive system technologies

303

S4

N4

tp 1

3

5

7

9

11

13

15

17

19

21 slot #

N20 S20

tp,2

Figure 5.48 PPM windings: 4-pole (top) and 20-pole (bottom) For the 20-pole connection shown in Figure 5.48 (bottom), the phase belt is equal to one and the pole pitch is equal to two. Pole areas in this connection are denoted by N20 and S20. ●

Speed ratio (1:even number) – shortened winding pitch at lower polarity. Assume that an ac machine has to operate in 4-pole and in 16-pole connection (Figure 5.49).

The double layer winding for PPM in this case will have 6 phases at 4 poles (m4 ¼ 6) and 3 phases at 16 poles (m16 ¼ 3). Phase belt at 4 poles is q4 ¼ 2, and at 16 poles q16 ¼ 1. Coil pitch expressed in number of teeth is y ¼ 9, and the winding is placed in 48 slots. Again, as shown in Figure 5.48, only the coils belonging to the same phase are shown in Figure 5.49 for the case when the speed ratio is an even number. PPM is the most generic method of arbitrary pole number change in an IM. When the IM is constructed having toroidal windings, the ability to redefine phases electronically makes this scheme even more flexible. However, there are issues with a PPM machine just as discussed for the 2:1 electronic pole change method in the previous section. The machine must be designed for flux patterns of the lowest pole number, or some compromise between the low and high pole numbers, otherwise the machine will be undersized magnetically for the low pole number and oversized magnetically for the high pole number. Figure 5.50 illustrates the flux patterns in a PPM machine that executes a 3:1 pole change electronically by inverter phase group control in the most general sense [31]. It must be pointed out that current research into PPM is focused on minimization of the current sensor requirements of high phase order systems. Recall that if the

304

Propulsion systems for hybrid vehicles

S4

N4

tp 1

3

N16

5

7

9

11

13

15

17

19

21

23

25 slot #

S16

tp,1

Figure 5.49 PPM winding connection for 4-pole (top) and 16-pole (bottom) operation

Figure 5.50 Flux patterns of PPM machine for toroidal IM: (a) 2-pole flux pattern of PPM and (b) 12-pole flux pattern of PPM 9-phase machine is considered to be three sets of 3-phase machines, a single sensor per set will suffice provided the proper voltage control is applied to the remaining 2 phases per set. One technique being investigated is to augment the three physical current sensors with machine flux sensors in the remaining 6 phases.

5.4.3.4

Pole changing PM

Several years ago, there was work performed on what was termed a ‘written’ pole machine in which the permanent magnet rotor was re-magnetized periodically via the stator into a new set of permanent magnet poles. This was done in order to realize frequency control from a prime mover that was less regulated than would otherwise be necessary.

Electric drive system technologies

305

60° 45°

F

(a)

F

(b)

Figure 5.51 Pole change PM machine in 8-pole (left) and 6-pole (right) configurations ( from Reference 33): (a) PCPM machine having four magnets/pole in an 8-pole configuration and (b) PCPM machine in a 6-pole configuration (one magnet remains un-magnetized, shown in white with black dots) In recent years, some dramatic improvements over that concept have come to light [33,34]. In the work by Prof Ostovic, it is proven that discrete speed control is achievable in permanent magnet machines via control of stator current so that such machines can be operated at discrete speeds in much the same manner as squirrel cage IMs. Initially, all 32 of the permanent magnet slabs in Figure 5.51 are magnetized tangentially and oriented such that pole flux exits the soft iron rotor wedges as shown. Next, suppose the prior 8-pole stator winding is reconfigured electronically using the techniques described for PPM into a 6-pole stator. Next, a short pulse of magnetizing current is fed into the stator windings that re-magnetizes the rotor magnets into the 6-pole pattern illustrated in Figure 5.51(b). Since 32/6 is not an integer, some of the magnets in the 6-pole configuration remain non-magnetized; one such magnet is shown in Figure 5.51(b). In another variation on the PCPM, the magnet polarization is modulated so that true field weakening can be achieved. This is possible in the ‘memory’ motor or, more appropriately, variable flux memory motor (VFMM). In Figure 5.52 the variable flux PM machine is shown in a full flux state (a) and at partial flux (b). Note that in Figure 5.52, when under partial magnetization, portions of the permanent magnets become reverse magnetized so that the net field entering the airgap is diminished. This feature of the VFMM yields efficient field weakening by a short pulse of stator current. This machine in effect combines the high efficiency of a PM machine with the airgap flux control of a wound field synchronous machine. The rotor magnets can be bonded rare earth, ferrite, or in this special class of machine, Alnico. Alnico is relatively easy to magnetize and re-magnetize and at the same time a high flux magnet. The rotor geometry is amenable to such magnets and the flux squeezing design means that very high airgap flux can be realized. It is anticipated that CPSRs of >5:1 can be achieved with this class of machine. This is particularly important for hybrid propulsion systems, ancillary systems control and other applications.

306

Propulsion systems for hybrid vehicles

Fmax

Fmax

(a)

F

F

ro (b)

Figure 5.52 Variable flux PM machine (from Reference 34). (a) VFMM when fully magnetized (dark grey-black shows PM, light grey shows non-ferrous); (b) VFMM when partially magnetized (light grey shows soft iron)

5.5 Variable reluctance machine The VRM is one of the oldest machine technologies known but also one of the slowest to be integrated into modern products. One reason for this is the lack of a unified theory of electromagnetic torque production as exists for dc and ac machines. Since the VRM does not conform to d–q theory owing to its double saliency and the fact that its torque production occurs in pulses has resulted in its analysis being limited to energy exchange over a stroke. VRMs have characteristics that make them very amenable to use in mobile, hybrid propulsion systems. The most notable characteristics are the facts that the VRM rotor is inert and thus easy to manufacture, has no permanent magnets nor field windings and so is robust in high speed applications, and due to its double saliency has stator coils that can be simple bobbin wound assemblies. For all its benefits the VRM has been slow to be adopted due in part to its need for precise rotor position detection or estimation, the requirement for close tolerance mechanical airgap and a proclivity to

Electric drive system technologies

307

generate audible noise from the normal magnetic forces in the excited stator. The issue with audible noise is particularly irksome in hybrid vehicle applications because any periodic noise source has the potential to excite structural resonances or to be the source of structure borne noise and vibration. The tendency of VRMs to generate noise comes from the passage of magnetic flux around the perimeter of the stator from one magnetic pole to its opposite polarity pole, sometimes diametrically opposed. The high normal forces under current excitation then cause the stator to deform into a number of modal vibrations. A second contributor to audible noise comes from the tendency of the rotor and stator saliencies, for example, the teeth, to deform under high tangential forces and when allowed to decay to start vibrating. Many attempts have been made to quiet the VRM including structurally reinforced stators, shaped stator and rotor teeth to minimize noise and inverter current shaping to mitigate any tendency to produce noise. Many of these techniques are proving the point that VRMs should be treated as viable hybrid propulsion system alternatives to asynchronous and permanent magnet synchronous machines. In the following subsections, two versions of reluctance machine will be described: the switched reluctance and the synchronous reluctance types. Each has its particular advantages and merits in a hybrid propulsion system.

5.5.1 Switched reluctance The switched reluctance machine is now the common terminology for the conventional double salient VRM. Figure 5.53 illustrates this structure in a 6/4 geometry. Because of the industrial acceptability of 3- and 4-phase systems, it is natural for VRMs to exist in 6/4 and 8/6 geometries. The numerator, Ns, in the expression, Ns/Nr, is the number of stator teeth, where Ns ¼ 2 m and the denominator is simply the number of rotor teeth. A machine of this construction will have mNr pulses per revolution, hence the descriptive term, a 12-pulse VRM or a 24-pulse VRM. In Figure 5.53 the energy enclosed by the shaded area represents the co-energy as a rotor tooth moves from unaligned position as shown by the position of the rotor in the figure as phase A is energized. Current is injected into the phase A winding in

W

C⫹

C⫺

B⫺ A⫺

Flux linkage l

A⫹

B⫹

W

I Current limit

Figure 5.53 Classical 6/4 variable reluctance machine

Current, i

308

Propulsion systems for hybrid vehicles

this unaligned position. The inductance is low, so the rise time of the current is very rapid until the current amplitude reaches the inverter current limit, I. At this point, the inverter current regulator begins to PWM the current, so that it remains at its commanded value while the rotor moves from unaligned to aligned position. After sufficient dwell, the current is switched OFF and allowed to decay along the flux linkage–current trajectory for fully aligned position back to the origin. The counterclockwise motion about the l–I diagram encloses an area, W, representing the net energy exchanged to mechanical energy at the shaft. The area beneath the unaligned position line represents phase A leakage inductance. The area from the fully aligned curve to the l-axis represents energy stored in the winding and returned to the supply, the reactive kVA in other words. The spacing between the lI curves in Figure 5.53 is a function of rotor angle, q. Bear in mind that as the inverter sequences between the phases in the order A–B–C, the rotor indexes clockwise as noted by W. Torque production in the VRM is given by (5.32) for average and instantaneous values: mN r W 2p ðNmÞ ð5:32Þ @W ði, qÞ T¼ @q where W equals the energy converted per working stroke of the machine. That is the energy converted per phase excitation. It requires excitation of all m-phases to move the rotor by one rotor tooth pitch – hence, the quantity in the numerator for average torque of the number of ‘working’ strokes per revolution times the work performed per stroke. The flux linkage in the VRM is given by (5.33) and the expression for induced voltage. The variable L(q) is the inductance variation with rotor position: T avg ¼

l ¼ LðqÞi E¼

dl di dL dq ¼L þi dt dt dq dt

ð5:33Þ

dq ¼w dt From these expressions, it is easy to compute the electrical power as the product of back-emf, E, and current and obtain   d 1 2 1 dL Li þ i2 w Pe ¼ dt 2 2 dq ð5:34Þ 1 2 dL T¼ i 2 dq where the first term in the expression for electric power represents the reactive voltamps, that is, derivative of the stored field energy, and the second term is the mechanical output power. Instantaneous torque is rewritten in (5.34) to highlight the fact that it is a function of current squared. This latter point can be interpreted to mean that part of the input current is used to excite, for example, magnetize, the machine and part is used to develop mechanical work.

Electric drive system technologies

309

20.0 mm

16.0 mm 60.0 mm j 295.0 mm

R = 87.5 mm

+

Figure 5.54 VRM machine used as hybrid vehicle starter–alternator A VRM hybrid vehicle starter–alternator was designed and fabricated to assess its merits as a viable hybrid propulsion system component [35]. This section will close with a description of the VRM constructed for a hybrid electric vehicle starter–alternator rated 8 kW and 300 Nm of torque. Figure 5.54 is an illustration of the VRM lamination design and stack. Because of the very large diameters involved, it was necessary to develop a 5-phase machine in which the 10/6 geometry was used as a repeating pattern that was then replicated three times about the circumference of the machine. The 5-phase converter for this machine is illustrated in schematic form in Figure 5.55. Each phase was constructed with pairs of individual phase leg power modules as shown. Phase configuration repeated 5 times

Phase : 1–5

SR power converter topology

Figure 5.55 VRM power electronics

310

Propulsion systems for hybrid vehicles New bidirectional current drive A B C D E

Stator A

B

C

D

E

A

Rotor (a)

Standard unidirectional current Phase A

A B

Phase B Phase C

C D

Phase D

E Phase E Stator A

B

C

D

E

A

Phase –A

Rotor (b)

Figure 5.56 Unidirectional versus bidirectional current drive cases for VRM: (a) unidirectional current drive and machine flux pattern and (b) bidirectional current drive and flux pattern It was also determined that bidirectional current would be advantageous since machine torque would be higher and inverter switch current would be reduced. The drawback was the need for twice as many active switches as shown in the schematic in Figure 5.55. Figure 5.56 illustrates the differences between conventional VRM unidirectional current drive and bidirectional current drive cases. Figure 5.57 is a picture taken on a test dynamometer of the VRM starter– alternator with an attached rotor position sensor. The machine is designed for stator liquid cooling via the mounting housing shown.

Electric drive system technologies

311

Figure 5.57 VRM prototype starter–alternator The reality of a 5-phase stator in a bidirectional current drive inverter is that a large bundle of stator leads are necessary (10 in total). This has the potential to be a packaging concern unless the power electronics box is mounted in close proximity to the high phase order VRM.

5.5.2 Synchronous reluctance The synchronous reluctance, SyncRel, machine is more directly analysed using d–q theory, and in fact the torque of this machine is given by T av ¼ mpðLd  Lq ÞI d I q

ðNmÞ

ð5:35Þ

where Id and Iq are rms amps. The number of pole pairs is given by p. It is very interesting to compare the torque production of the SyncRel relative to the IM. In the IM the average torque is T av i ¼

3 P Lm ðLm ids Þiqs 2 2 Lr

ðNmÞ

ð5:36Þ

where ids and iqs are in peak amps, and Lm and Lr are IM inductances. Taking the ratio of (5.35) for the SyncRel machine to (5.36) for the IM results in T r ð1  ðLmq =Lmd ÞÞðLmd =Lm Þ ¼ Ti Lm =Lr

ð5:37Þ

Taking representative ratios for inductance in the SyncRel of ~8:1, the expression in parentheses in the numerator is (1  1/8). The second numerator inductance

312

Propulsion systems for hybrid vehicles

Figure 5.58 Schematic of the SyncRel machine ratio is ~0.94 and the denominator ratio is typically 0.91 for an IM. This results in the expression in (5.37) equalling 0.904. In other words, it would appear that the SyncRel machine is only capable of 90% the torque of the IM for identical stator currents; but there is more. Notice by inspection of the rotor construction illustrated in Figure 5.58 that a SyncRel machine has no rotor losses. In the SyncRel machine, the rotor is constructed of axial laminations (rain-gutter geometry) that lie in the direction of d-axis flux. The q-axis inductance is very low because of the large saliency and also because q-axis flux must cross the many lamination to lamination insulation layers. The IM rotor losses are 3 Pr ¼ i2r Rr 2

(W)

ð5:38Þ

This means that unlike the IM, the SyncRel has no rotor losses to contend with, so only stator copper losses must be accounted for. The IM stator current consists of both magnetizing and load current components, for which magnetizing current is approximately 15% of the total. Taking this into account results in the SyncRel machine having only 63% the losses of the IM. So now, if the ratio in (5.37) is again computed but for equal losses, the result is Tr ¼ 1:26 Ti

ð5:39Þ

This is a completely different picture of the SyncRel in comparison to the IM, but it does not account for the fact yet that inverter losses in the SyncRel are higher that for an IM because the power factor is lower in the SyncRel. Typical power factors for IMs are 0.85, whereas for the SyncRel it is 0.80. The advantage of the SyncRel machine over the IM diminishes as rating increases. For larger machines, generally >20 hp, the magnetizing reactance of the

Electric drive system technologies

313

SyncRel increases, thereby diminishing the efficiency advantage it held over the IM. This means that for machine ratings suitable for hybrid propulsion systems the SyncRel really has no efficiency advantage even given its inert rotor. The other drawback of the machine technology is the fact that its axial laminated rotor structure has real issues at higher speeds where rotor retention becomes a major concern.

5.5.3 Radial laminated structures Radial laminated SyncRel machines are generally not as widely used as axial laminated designs because with multiple flux barriers the mechanical constraints on ribs and posts to support the soft iron pole shoes above and in between the flux barriers become major design challenges. It has been shown that for the same total losses a SyncRel machine can have 20% more torque than a corresponding IM. However, losses in the SyncRel machine increase substantially with increasing pole number, so these machines are restricted to pole numbers of less than 4. There is strong interest in SyncRel for machine tool applications because it can deliver overload torque ratings of >3:1. It is also recommended to drive the radial laminated designs with current controlled inverters and to avoid delta connected stator windings. In general, this design has not found much favour with hybrid propulsion designs because it cannot compete with its cousin, the IPM.

5.6 Relative merits of electric machine technologies This chapter has reviewed a great many electric machine technologies and several methods of controlling such machines. It is now important that this vast amount of material on electric machines be summarized into a more cohesive framework from which hybrid propulsion system designers can choose. For this summary, two seminal papers will be cited that give clear insights into machine comparisons for the two important categories of ac drives: battery electric vehicles and hybrid propulsion systems. The next two subsections will summarize the material in this chapter in the context of machine technology comparisons of the leading three major categories: IM, IPM machine and the VRM. Furthermore, each of these machine types was compared based on a representative vehicle and performance specifications. The interested reader is referred to the appropriate reference at the end of this chapter.

5.6.1 Dynamic performance comparisons Volumes have been written on the topic of electric machine comparisons, so this section will be brief. In Reference 36 Melfi et al. look at industrial induction and permanent magnet machines from the view of power density and energy savings. Their findings cover salient pole PM machines of the inset magnet design (similar to what Honda Motor used in early IMA systems), plus single buried magnet to

314

Propulsion systems for hybrid vehicles

single and multilayer chevron shaped buried magnet (IPM) designs. The industrial motors are all rated 75 hp (56 kW), 400 V with class H insulation (180  C) or 125  C temperature rise in a 55  C environment (Table 5.9). What these authors find is very relevant to the theme of this text: ●





IM NEMA 250 frame TEFC (totally enclosed fan cooled), 4-pole and 60 Hz base frequency SPM machine, same size as the IM, but wound for 8-pole, 120 Hz base frequency, fan cooled Salient pole design using same stator as the IM, but different winding as 4-pole and fan cooled

Table 5.9 Industrial motor comparison, 75 hp (56 kW) tested at full load, 1,800 rpm Attribute

Unit

IM

SPM

IPM

Number of poles Voltage* Base frequency Torque Current* Power factor Efficiency Temperature rise

# V Hz Nm A # %  C

4 459 60 304 92 82 93.6 111

8 405 120 298 85 98 96.2 71

4 395 60 298 90 93 96.8 90

*Voltage and current are rms values, line-to-line and line respectively.

For the same shaft power (~56 kW), the SPM runs the coolest, has the best power factor, lowest line current and high efficiency. The IPM comes in second and lowest in the performance category is the IM. Dynamic performance is a particular requirement, especially for the generator, MG2, in an eCVT hybrid vehicle. Typically, in hybrid vehicle systems, the controller must incorporate inertia compensation to correct for generator rotor inertia. In a detailed analysis, these rotating component inertias can be reflected to the drive wheels and contribute to higher inertial mass to be accelerated. Polar moment of inertia is given as (5.40) J 0 ¼ M rot r2ro

ð5:40Þ

Initial acceleration of an electric machine commanded to peak torque is strictly dependent on its polar inertia: m ¼ J 0w ¼ J 0a

ð5:41Þ

Electric drive system technologies

315

Example 5: The Camry hybrid MG1 and MG2 rotor mass and outside diameter are approximately: MMG1 ¼ 9 kg and MMG2 ¼ 5.2 kg and diameters DrMG1 ¼ 160.46 mm and DrMG2 ¼ 160.45 mm. Compute their polar moment of inertias. Solution: Applying (5.40) results in J1 ¼ 0.029 (kg-m2/rad) and J2 ¼ 0.017 (kg-m2/ rad)

5.6.2 Comparisons for electric vehicles It is well known that the choice of ac drive system for European and North American battery electric vehicle traction systems is the ac IM drive. For AsiaPacific the choice is and continues to be the ac permanent magnet drive system. This section looks at the reasons for this particular choice of ac drive system. The variable reluctance drive system is and continues to receive favourable reviews but as yet remains on the sidelines in vehicle propulsion areas. In the comparisons to follow, the relative rankings are based on the availability of a 400 Apk and 400 Vdc inverter as the power driver. The battery is sized for 250 Vdc. Inverter control is based on field orientation principles for fast dynamic response and efficient use of the power silicon. The vehicle itself is assumed to have a mass of 1,500 kg, a drag coefficient of 0.32 with a frontal area of 2.3 m2 and an overall transmission gear ratio of 8:1 to 12:1. The vehicle performance targets are >30% grade, and 135 kph maximum speed. Table 5.10 lists the key machine parameters needed to determine the machine’s continuous torque rating. Based on thermal limitations, the continuous torque is set by the dissipation limits on conductor current density, which for battery electric vehicles is restricted to 5 A/mm2 except for the IPM machine where it is 6.5 A/mm2. Peak torque is typically 2.5–3 times this value, and in general 1 1.12 58

93/200 1.33 3,000 5:1 for IPMs is often times not observed in practical systems, and in this figure it was as limited as the IM.

Electric drive system technologies

319

5.7 Exercises Q1:

Repeat example 2 by reversing any two of the phase currents and show that the rotor will reverse its rotation direction. In this exercise the currents are defined as ia ¼ I 0 sin wt; ib ¼ I 0 sin ðwt þ 120 Þ; ic ¼ I 0 sin ðwt  120 Þ

A1:

Q2:

Normal Connections in Block Mode A B B C C C A A reverse phase B and C connections A A C C B C B B A A A B

Q3:

A3: Q4:

A4: Q5:

A5: Q6: A6: Q7:

B C

The following exercises are meant to expand on the topic of IPM machines and their torque production via magnet and reluctance components. Example 2 should also be used as reference. In the following exercises, the IPM torque expression will be needed: mIPM ¼

A2:

C B

3 P  ½ym I d þ ðLd  Lq ÞI d I q  2 2

Given that a stator max current, Is ¼ 450 A, and a current angle d ¼ 35 compute the Id and Iq components of stator current at maximum loading. Id ¼ 258 A and Iq ¼ 369 A For the IPM of Q2 and given that its phase winding flux linkage, km ¼ 0.058 Wb-t, and the fact that this is a 3-phase, 8-pole machine, compute the magnet torque at maximum load. mmag¼129.5 Nm (note, the first term in the equation given in Q2) Given that the total torque, mIPM, of (Q2) is 270 Nm at 450 Adc and for the stated current angle of maximum torque production compute the quantity (Ld  Lq). Ld  Lq ¼ 246 mH At maximum loading the inductance ratio of an IPM decreases significantly from its unexcited value. Taking the ratio, Ld/Lq ¼ 2.7 at maximum loading and using the result of Q4, compute the values of Ld and Lq. Ld ¼ 391 mH and Lq ¼ 145 mH. Apply (5.22) from this chapter and calculate the characteristic current of this IPM. 148 A Compute the quantity (Ld  Lq) for the Synchronous Reluctance machine of section 5.5.2 if the same torque target (mSyncRel ¼ 270 Nm) is imposed

320

Propulsion systems for hybrid vehicles

for the same Id and Iq currents found in (Q2) and the SyncRel is also 8-pole, 3-phase machine. A7: 0.473 mH Hint: The Id and Iq for the IPM are peak quantities and in the SyncRel are rms. The SyncRel is a much higher reactance machine than the IPM, making it similar to the asynchronous machine as noted. Q8: A8: Q9:

A9:

For the electric machine performance comparison in Table 5.9 calculate the input electrical power and compare to the shaft power of 56 kW. Pe ¼ 64 kW (IM); 60.7 kW (SPM); 59.2 kW (IPM) Using data listed in Example 1 for the Camry Hybrid compute the MG1 and MG2 rotor peak tangential velocity and peak acceleration given that MG1 torque is 270 Nm and MG2 torque is approximately 200 Nm. Vt1 ¼ 117 m/s and Vt2 ¼ 92.4 m/s a1 ¼ 9,310 rad/s2 and a2 ¼ 11,764 rad/s2

References 1. 2.

3.

4. 5.

6.

7.

8.

Toliyat H.A., Kliman G.B. Handbook of Electric Motors, 2nd edn. Marcel Dekker Inc., 2004. Wood P., Battello M., Keskar N., Guerra A. Plug-N-DriveTM Family Applications Overview. IR Application Note, AN-1044, February 2003. Also see www.irf.com. Lawler J.S., Bailey J.M., McKeever J.W. ‘Theoretical verification of the infinite constant power speed range of the brushless DC motor driven by dual mode inverter control’. IEEE 7th Workshop on Power Electronics in Transportation, WPET 2002, DaimlerChrysler Technical Center, Auburn Hills, MI, 24–25 October 2002. Wise T. Tesla – A Biographical Novel of the World’s Greatest Inventor, Atlanta: Turner Publishing, Inc., 1992. EPRI TR-101264, Project 3087-01, ‘Assessment of Electric Motor Technology: Present Status, Future Trends, and R&D Needs’ prepared by McCleer Power Inc., Jackson, MI, December 1992. Miller J.M., Gale A.R., McCleer P.J., Leonardi F., Lang J.H. StarterAlternator for Hybrid Electric Vehicle: Comparison of Induction and Variable Reluctance Machines and Drives, IEEE Industry Applications Society Annual Meeting, Saint Louis, MO, 11–15 October 1998. Schiferl R., Lipo T.A. ‘Core loss in buried magnet permanent magnet synchronous motors’. IEEE Proceedings of the IEEE Power Engineering Society Summer Meeting, July 1988. Honsinger V. ‘The fields and parameters of interior type AC permanent magnet machines’. IEEE Transactions on Power Apparatus and Systems, April 1982, vol. PAS-101, no. 4, pp. 867–76.

Electric drive system technologies

321

9. Jahns T.M. Component Rating Requirements for Wide Constant Power Operation of Interior PM Synchronous Machine Drives, 2000 IEEE Industry Applications Society Annual Meeting, Rome, Italy, 8–12 October 2000. 10. Jahns T.M. Uncontrolled Generator Operation of Interior PM Synchronous Machines Following High-Speed Inverter Shutdown, IEEE Industry Applications Society Annual Meeting, Saint Louis, MO, 12–15 October 1998. 11. Amaratunga G.A.J., Acarnley P.P., McLaren P.G. ‘Optimum magnetic circuit configurations for permanent magnet aerospace generators’. IEEE Transactions on Aerospace and Electronic Systems, March 1985, vol. AES-21, no. 2, pp. 230–55. 12. Kume T., Iwakane T., Sawa T., Yoshida T. ‘A wide constant power range vector-controlled ac motor drive using winding changeover technique’. IEEE Transactions on Industry Applications, September/October 1991, vol. 27, no. 5, pp. 934–39. 13. Saki K., Hattori T., Takahashi N., Arata M., Tajima T. High Efficiency and High Performance Motor for Energy Saving in Systems, IEEE Power Engineering Society Winter Meeting, 2001, pp. 1413–18. 14. Osama M., Lipo T.A. ‘Modeling and analysis of a wide-speed-range induction motor drive based on electronic pole changing’. IEEE Transactions on Industry Applications, September/October 1997, vol. 33, no. 5, pp. 1177–84. 15. Miller J.M., Stefanovic V.R., Ostovic V., Kelly J. Design Considerations for an Automotive Integrated Starter-Generator with Pole-Phase-Modulation, IEEE Industry Applications Society Annual Meeting, Chicago, IL, 30 September– 4 October 2001. 16. Wai J., Jahns T.M. A New Control Technique for Achieving Wide Constant Power Speed Operation with an Interior PM Alternator Machine, IEEE Industry Applications Society Annual Meeting, Chicago, IL, 30 September– 4 October 2001. 17. Ostovic V. Memory Motors – A New Class of Controllable Flux PM Machines for a True Wide Speed Operation, IEEE Industry Applications Society Annual Meeting, Chicago, IL, 30 September–4 October 2001. 18. Ostovic V. Pole-Changing Permanent Magnet Machines, IEEE Industry Applications Society Annual Meeting, Chicago, IL, 30 September–4 October 2001. 19. Caricchi F., Crescimbini F., Giulii Capponi F., Solero L. Permanent-Magnet, Direct-Drive, Starter-Alternator Machine with Weakened Flux Linkage for Constant Power Operation Over Extremely Wide Speed Range, IEEE Industry Applications Society Annual Meeting, Chicago, IL, 30 September– 4 October 2001. 20. Tapia J.A., Lipo T.A., Leonardi F. CPPM: A Synchronous Permanent Magnet Machine with Field Weakening, IEEE Industry Applications Society Annual Meeting, Chicago, IL, 30 September–4 October 2001. 21. Staton D.A., Deodhar R.P., Soong W.L., Miller T.J.E. ‘Torque prediction using the flux-MMF diagram in AC, DC, and reluctance motors’. IEEE

322

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

Propulsion systems for hybrid vehicles Transactions on Industry Applications, January/February 1996, vol. 32, no. 1, pp. 180–88. Sakai K., Hattori T., Takahashi N., Arata M., Tajima T. High Efficiency and High Performance Motor for Energy Saving in Systems, IEEE Power Engineering Society Winter Meeting, 2001. Lei M., Sanada M., Morimoto S., Takeda Y. ‘Basic study of flux weakening for interior permanent magnet synchronous motor with moving iron piece’. Transactions of IEEE Japan, 1998, vol. 118-D, no. 12. Translated from Japanese by Dr Samuel Shinozaki for the author. Caricchi F., Crescimbini F., Giulii C.F., Solero L. Permanent Magnet, Direct-Drive, Starter/Alternator Machine with Weakened Flux Linkage for Constant-Power Operation Over Extremely Wide Speed Range, IEEE Industry Applications Society Annual Meeting, Chicago, IL, 30 September–4 October 2001. Honda Y., Nakamura T., Higaki T., Takeda Y. Motor Design Considerations and Test Results of an Interior Permanent Magnet Synchronous Motor for Electric Vehicles, IEEE Industry Applications Society Annual Meeting, New Orleans, LA, 5–9 October 1997. McCleer P.J., Miller J.M., Gale A.R., Degner M.W., Leonardi F. Nonlinear Model and Momentary Performance Capability of a Cage Rotor Induction Machine Used as an Automotive Combined Starter-Alternator, IEEE Industry Applications Society Annual Meeting, Phoenix, AZ, 3–7 October 1999. Kume T., Iwakane T., Sawa T., Yoshida T. ‘A wide constant power range vector-controlled ac motor drive using winding changeover technique’. IEEE Transactions on Industry Applications, September/October 1991, vol. 27, no. 5, pp. 934–39. Xu L., Liang F., Lipo T.A. Transient Model of a Doubly Excited Reluctance Motor. Wisconsin Electric Machines and Power Electronics Consortium, WEMPEC, research report 89-29. ECE Department, University of WisconsinMadison, Madison, WI, 1990. Liang F., Xu L., Lipo T.A. d-q Analysis of a Variable Speed Doubly ac Excited Reluctance Motor. Wisconsin Electric Machines and Power Electronics Consortium, WEMPEC, research report 90-16. Department of Electrical and Computer Engineering, University of Wisconsin, Madison, WI, 1990. Lao X., Lipo T.A. A Synchronous/Permanent Magnet Hybrid ac Machine. Wisconsin Electric Machines and Power Electronics Consortium, WEMPEC, research report 97-14. Department of Electrical and Computer Engineering, University of Wisconsin, Madison, WI, 1997. Osama M., Lipo T.A. ‘Modeling and analysis of a wide-speed range induction motor drive based on electronic pole changing’. IEEE Transactions on Industry Applications, September/October 1997, vol. 33, no. 5, pp. 2366–73. Miller J.M., Stefanovic V.R., Ostovic V., Kelly J. Design Considerations for an Automotive Integrated Starter-Generator with Pole-Phase-Modulation, IEEE Industry Applications Society Annual Meeting, Chicago, IL, 30 September–4 October 2001.

Electric drive system technologies 33. 34. 35.

36.

37.

38.

323

Ostovic V. Pole-Changing Permanent Magnet Machines. Ibid. Ostovic V. Memory Motors – A New Class of Controllable Flux PM Machines for a True Wide Speed Operation. Ibid. Miller J.M., Gale A.R., McCleer P.J., Leonardi F., Lang J.H. StarterAlternator for Hybrid Electric Vehicle: Comparison of Induction and Variable Reluctance Machines and Drives. IEEE Industry Applications Society Annual Meeting, Saint Louis, MO, 12–15 October 1998. Melfi M., Evon S., McElveen R. ‘Induction versus permanent magnet motors’. IEEE Industry Applications Magazine, December 2009, vol. 15, no. 6, pp. 28–35. Winter U. ‘Comparison of different drive system technologies for electric vehicles’. Proceedings of Electric Vehicle Symposium, EVS15, Brussels, Belgium, 30 September–2 October 1998. Conlon B. ‘A comparison of induction, permanent magnet, and switched reluctance electric drive performance in automotive traction applications’. PowerTrain International, 2001, vol. 4, no. 4. Available from www.powertrainintl.com.

Chapter 6

Power electronics for ac drives

Power electronics and its control fall within what is known as ‘inner-loop’ control of the hybrid ac drive system. Starting with solid state, or brushless, commutators, for permanent magnet electric machines in the 1970s, the techniques of solid state motor control have been applied to industrial induction machines in the form of adjustable speed drives (nearly 90% of all electric machines produced and sold are induction machines) and recently to interior permanent magnet and reluctance machines for hybrid propulsion. Without the advancements and miniaturization efforts of the semiconductor community, the hybrid vehicle would not be market ready. As the highest cost component of the hybrid propulsion system, with the possible exception of the vehicle battery, the power electronics represents one of the most complex power processing elements in the vehicle. In this chapter the various types of semiconductor devices are summarized along with their applicability for use as invehicle power control. The assessment of power electronics for ac drive systems then continues with discussion of various modulation techniques, thermal design and reliability considerations. Modulation techniques are important for many reasons. Most of the present modulation methods are capable of synthesizing a clean sinusoidal ac waveform from the vehicles on-board energy storage system (ESS), but not all do so with equal efficiency, noise emissions or dc voltage utilization. Integration of power electronic systems today is at the stage where a single integrated power module consisting of a full active bridge of power semiconductors, integrated gate drivers, and fault detection and reporting logic is available off the shelf. At higher powers (>50 kW), the power electronics may consist of individual phase leg modules. At low powers ( 1, sinusoidal modulation is no longer possible because of pulse dropping. In some systems, predefined modulating waveforms are used for overmodulation. Consider a square wave of amplitude Udc. The square wave has an rms value of Udc and a peak fundamental value of 4 U dc p U s1 ¼ 1:27 m0 ¼ U dc

U s1 ¼

ð6:13Þ

338

Propulsion systems for hybrid vehicles Us* * Ush



Ua

Figure 6.11 Synchronized sampling of modulating wave Now, the sine wave that can be synthesized from a fixed dc bus of magnitude Udc is constrained to a peak value of Udc, so it will have a modulation index equal to the reciprocal of the modulation index listed in (6.13) or mmax ¼

1 ¼ 0:787 m0

ð6:14Þ

Referring again to Figure 6.8 one can say that with sinusoidal, or regular, modulation the limit on depth of modulation within the zone marked full PWM is the value given by (6.14). In the transition region of pulse dropping, the modulation depth exceeds 0.787 and reaches a value of 1 at the boundary of six step square wave. The value mmax is a real and significant limitation of sinusoidal modulation and it results in less than optimum bus voltage utilization. Later in this section it is shown that with SVPWM, this limitation is increased significantly so that maximum bus utilization is realized. To further illustrate sinusoidal, regular sampling, refer to Figure 6.12, where it is shown what the switch patterns will appear as for three different voltage vectors in sector I and one vector in sector II of the inverter state space. The points illustrated are angle of 20 with the a-axis and modulation index of 0.35 (A), angle of 40 with the a-axis and modulation index of 0.8 (B), angle of 55 from the a-axis and

Power electronics for ac drives

339

1.5 Tk(i)

1

Amps

Ia(i)

Ib(i)

Ic(i)

0

⫺1 ⫺1.5

0 0

50

100 i

150

200 199

(a) Point

g (°)

A

20

Switching patterns 5 4 Ua(i) + 3 Ub(i) + 1.5

2

Uc(i) + 0.2 0.2 0 0 0

B

40

10

20

30 i

40

50

60 60

30 i

40

50

60 60

30 i

40

50

60 60

5 4

2

Ua(i) + 3 Ub(i) + 1.5 Uc(i) + 0.2

0.2 0 0 0

C

55

10

20

5 4

2

Ua(i) + 3 Ub(i) + 1.5 Uc(i) + 0.2

0.2 0 0 0

D

75

10

20

5 4

Ua(i) + 3 Ub(i) + 1.5

2

Uc(i) + 0.2 0.2 0 0 0

10

20

30 i

40

50

60 60

Figure 6.12 Sinusoidal synchronous (regular) sampling: (a) regular sampling for point (A) (b) corresponding space vector waveforms for regular sampling at angle 20 , 40 , 55 and 75

340

Propulsion systems for hybrid vehicles

modulation index of 0.8 (C), and an angle of 75 with the a-axis and modulation depth 0.8 (D). Figure 6.12(a) shows the sample points for a 3-phase balanced set at an angle of 20 from the phase a axis and for a modulation depth 0.35. Regardless of sector, modulation depth, or vector angle, the sinusoidal synchronous (regular) sampling exercises all the inverter switches during each switching cycle. This incurs higher switching losses than are necessary to synthesize the voltage waveforms required. A more optimum switching strategy is provided by SVPWM in which a switch is not exercised unless needed and in which only vectors adjacent to the reference voltage vector are selected. In sinusoidal synchronous PWM, all voltage vectors are selected regardless of the operating sector. In SVPWM, it is convenient to represent inverter voltage vectors in the a–b plane as shown in Figure 6.13. In this figure, the six sectors of the inverter output voltage states are listed along with a representative inverter source vector, U s , in the first sector. Im-axis, b U3

U2 II Us* Uk + 1 I p /3

III U4

U1 Re-axis, a

U0, U7 VI IV V U5

U6

Figure 6.13 Derivation of space vector PWM As an illustration, in sector I the only available voltage vectors from which to construct an arbitrary reference vector are U1, U2 and either of U0 or U7. As shown earlier in Figure 6.12, each of these vectors is the result of a discrete switching pattern of the inverter poles as follows: ða, b, cÞ ¼ ð1, 0, 0Þ ¼ U 1 ða,b,cÞ ¼ ð1, 1, 0) ¼ U 2 ða, b, c) ¼ ð0, 0, 0) ¼ U 0

ða, b, cÞ ¼ ð1, 1, 1) ¼ U 7

Each of these inverter states applies full voltage magnitude at the inverter phase leg mid-point as discussed earlier. In order to synthesize a voltage vector of arbitrary magnitude and of arbitrary angle within sector I (and by extension, in any

Power electronics for ac drives

341

other sector), all that is necessary is to use appropriate durations of vectors U1, U2, and U0 or U7. It can also be shown that SVPWM is optimum from the standpoint of minimization of ripple current in an L–R load. Because the load ripple current of an electric machine ‘sees’ the machine leakage inductance, the ripple current is a scaled version of the flux linkage ripple since the leakage inductance is basically linear. Flux linkage ripple is in essence a deviation in volt-seconds appearing across an inductor, so minimum ripple current exists only if minimum voltage ripple is impressed. SVPWM can be said to be optimum if this deviation in the load current vector, due to ripple current, for several switching states becomes small and if the cycle time is as short as possible. These conditions are met if ●



Only the four inverter states adjacent to the reference vector are used. In reality, only three inverter states are used since it is desirable to permit only a single switch transition per switching cycle. A cycle is defined by three successive switching states only. During a given cycle, the average voltage vector is equal to the reference vector as shown in Figure 6.14. U1

t U2

U3

T0/2

Tk

Tk + 1 T0/2 T

2T

Figure 6.14 Switching pattern of SVPWM To visualize this in more detail, refer again to Figure 6.13 and note that a succession of the vectors U1, or (a,b,c) = (1,0,0), U2, or (a,b,c) = (1,1,0), and U7, or (a,b,c) = (1,1,1,), will be used to synthesize the reference vector U k . In this instance the nomenclature Uk is used to connote which vectors in the given sector are adjacent to the reference. Also, only a single switch transition takes place in moving from one state to the next at the inverter output. In general, SVPWM seeks

342

Propulsion systems for hybrid vehicles

to solve for the time increments given in (6.15) with the proviso that the reference vector, U k , remains constant during a switching cycle: U k T ¼ U k T k þ U kþ1 T kþ1

ð6:15Þ

where (6.15) must be solved for Tk and Tk+1 as fractions of the switching cycle period, T. The null vector, U0 or U7, must persist for a time increment given by (6.16): T0 ¼ T

Tk

ð6:16Þ

T kþ1

When the null vector on time, T0 = 0, SVPWM has reached its limit of applicability and its modulation depth is maximized for full PWM. Beyond this limit, pulse dropping must occur just as it does for sine-triangle PWM (see Figure 6.11). Now, by referring again to Figure 6.13 and rewriting (6.15) and (6.16) as integrals it can be seen that the reference vector U k must satisfy ðT

U k dt

¼

Tðk

U k dt þ

0

0

Tð kþ1

U kþ1 dt þ

Tk

ðT

U 0; 7 dt

ð6:17Þ

T k þT kþ1

where k is index of vectors adjacent to the sector in which the reference vector is located. Because the switching frequency is at least an order of magnitude greater than the fundamental frequency, the transitions between states in (6.17) will occur for essentially quasi-static behaviour of the commanded reference. Refer again to Figure 6.13 and note the angle that the vectors Uk, Uk+1 and U s make with the a-axis and then rewrite (6.15) as (6.18) in terms of actual magnitudes: h p

p i 3 3 mi U dc ½ cos ð0Þ þ jsinð0ފT k þ mi U dc cos þ jsin 2 2 3 3 3 ¼ mi U dc ½ cos ðgÞ þ jsinðgފ 2

ð6:18Þ

where the modulation index has been defined previously as the ratio of the maximum sinusoidal synthesized wave amplitude to that of an equivalent square wave magnitude, and g = wt. Therefore, solving (6.18) results in expressions for Tk and Tk+1 as T k ¼ mi T T kþ1 ¼ mi T T0 ¼ T

sin ðp=3 gÞ sin ðp=3Þ sinðgÞ sinðp=3Þ Tk

T kþ1

ð6:19Þ

Power electronics for ac drives

343

Equation (6.19) summarizes the calculations leading to SVPWM. It is also evident that with a fast digital processor these calculations can be processed online and quickly since half the coefficients are constants or easily obtained from look-up tables as the reference vector moves from sector to sector in the inverter state diagram. It is also clear from (6.19) how the modulation depth impacts the time intervals. Higher levels of modulation index result in a larger fraction of the total switching period being devoted to adjacent state vectors and a diminishing amount of time for the null vector. Next, it is instructive to illustrate the particular switching pattern characteristic of SVPWM during one switching cycle. It is also important to recognize that in SVPWM the inverter gating frequency is one-half the state clock frequency. For example, if the clock frequency is fclk then the inverter switching frequency fs = fclk/2. The average value over a switching cycle of inverter phase to negative bus voltages U1, U2 and U3 are calculated with the aide of Figure 6.14. To clarify, the voltages given by (6.20) are from the inverter phase leg to negative bus where subscripts 1, 2 and 3 are used in lieu of a, b and c phases for convenience:

U1 ¼

  U dc T0 T0  þ T1 þ T2 þ T 2 2

U2 ¼

  U dc T0 T0   T1 þ T2 þ T 2 2

ð6:20Þ

  U dc T0 T0   T1  T2 þ ¼ U 1 U3 ¼ T 2 2 When the relations given in (6.19) are substituted into (6.20) with appropriate subscript change, the inverter to negative rail voltages are 2 p

U 1 ¼ pffiffiffi mi U dc sin g þ 3 3 p

U 2 ¼ mi U dc sin g  6

2 p

U 3 ¼  pffiffiffi mi U dc sin g þ 3 3

ð6:21Þ

344

Propulsion systems for hybrid vehicles

Because we know the inverter phase leg voltages, it is a straightforward calculation to arrive at the line-to-line voltages at the inverter output. These are the voltages that will be applied to the electric machine as a load: 4 p

U ab ¼ pffiffiffi mi U dc sin wt þ 6 3 4 p

U bc ¼ pffiffiffi mi U dc sin wt  ð6:22Þ 2 3   4 7p U ca ¼ pffiffiffi mi U dc sin wt  6 3

The line-to-line voltages are sinusoidal but the modulating function that will produce these output voltages is not sinusoidal. First, to clarify the amplitudes of the line-to-line, U1–1, phase to negative bus, U1, and peak of the modulating function, Um, one finds 4 jU 11 j ¼ pffiffiffi mi 3 4 jU 1 j ¼ mi 3 mi mi < jU m j < pffiffiffi 2 3

ð6:23Þ

The double-valued nature of the SVPWM modulating function is characteristic and due to the fact that harmonics of triple order may be added to phase voltages without affecting the Park transformation from stationary to synchronous reference frames. Therefore, there is an infinite variety of modulating functions that generate a sinusoidal line-to-line voltage but in themselves are not sinusoidal. To illustrate this fact it is necessary to generate such an SVPWM modulating function. This function, if applied to the sinusoidal synchronous modulator discussed earlier, will in fact generate SVPWM gating waveforms. If the synchronous sampling process described in Figure 6.11 is applied here but with the SVPWM modulating function shown in Figure 6.15 for phases U, V, W 1.732051 2 eU(i)

1 eW(i)

eV(i)

0 0 0 0

50

100

150

200 i

250

300

Figure 6.15 SVPWM modulating functions

350

400 359

345

Power electronics for ac drives 1.984824

2

0

⫺1.994946 ⫺2

0 0

100

200 i eU(i)

300

400 359

eC(i)

(a) 1.5 1 U1(i)

0 ⫺1.5 (b)

0 0

100

200 i

300

400 359

Figure 6.16 Synchronous sampling of SVPWM modulating function: (a) synchronous sampling and (b) inverter phase a mid-point voltage to negative bus

and moreover, if the carrier triangle waveform is set to an integer of nine samples per fundamental, the PWM patterns for inverter output phases U1, U2 and U3 can be obtained. The PWM waveform in Figure 6.16 is very similar to that obtained for sinusoidal synchronous PWM. However, the line-to-line voltages for SVPWM are now very different as can be seen in the waveforms of Figure 6.17. In this instance, the inverter pole (U1 is the phase a pole composed of switches S1 and S2, and so on for the others) to negative bus voltages are subtracted, which eliminates the interim negative bus reference leaving only the corresponding voltages. It is apparent from Figure 6.17 that the PWM pattern is symmetrical in the pulse placement as expected and, furthermore, there is no redundant switching in a phase leg. Also, some of the line-to-line pulses have two levels. The SVPWM is the most efficient in terms of low switching losses in the inverter and maximum utilization of the dc link voltage. In fact, the maximum modulation index for SVPWM is

mi

SVPWM

2 ¼ pffiffiffi ¼ 1:15 3

ð6:24Þ

346

Propulsion systems for hybrid vehicles 1.5 U12(i) 0

⫺1.5

0 0

100

200 i

300

400 359

(a) 1.5

U23(i) 0

⫺1.5

0 0

50

100

150

200 i

250

300

350

400 359

(b) 1.5

U31(i) 0

⫺1.5

0 0

100

200 i

300

400 359

(c)

Figure 6.17 SVPWM line-to-line voltages: (a) phase a–b voltage ( U12), (b) phase b–c voltage ( U23) and (c) phase c–a voltage ( U31). A comparison of various PWM schemes is given in Section 6.6 for completeness.

6.5 Multilevel inverters In HV systems such as light rail and other utility interface applications such as the SST, discussed in Section 6.1, the trend is towards both higher voltage devices and

Power electronics for ac drives

347

multilevel inverters. Although not entirely relevant to vehicle propulsion systems because of relatively low voltage (LV), there are instances where multilevel (matrix) converters have found application in the 0. (a) Synchronous frame commands when ramping from motoring to generating, (b) response of M/G currents to the commanded ramps and (c) response of M/G phase-a current only, Ide > 0 case

Drive system control

379

entered into. The second fact to notice is that because of the non-zero flux command the phase relations of the M/G currents are now displaced from what they were when the flux command was zero. This phase shift of stator currents relative to a rotor position is the means of building flux. The final point to note is that at the ramp edges the M/G currents execute some continuous phase changes needed to ensure that flux does change. Yet another dynamic scenario to illustrate is the situation in which the M/G is given a speed reversal command as it will experience in the power split hybrid architecture when connected to a planetary gear set sun gear (e.g. the S/A in the THS system). In this situation the torque command is held steady while the speed is executing a reversal. During a speed reversal the synchronous frame commands execute a swap in phase. The M/G 3-phase currents then execute a sequence change from a–b–c to a–c–b as shown in Figure 7.11(c). This is consistent with the manner in which a 3-phase ac machine executes a direction change. If any two phases are swapped the machine rotor spins in the reverse direction, which is exactly what the FOC controller has electronically commanded the M/G to do. Notice that during the phase sequence change one of the phase currents is completely undisturbed while the remaining two phases slew very rapidly to their new sequence.

7.3 Sensorless control The topic of position sensorless control of ac drive systems is particularly relevant to hybrid propulsion systems. Not only are mechanical position sensors difficult to integrate into and package within the vehicle driveline, but they are fragile and susceptible to EMI and signal distortion. It would be a great advantage to minimize position sensor requirements or to eliminate the need entirely if adequate software algorithms were available to perform the function of tracking the M/G rotor position accurately. Not only is rotor position sensor degradation an issue in maintaining smooth control of the hybrid M/G system, but corruption of the position signal introduces disturbances into the voltage and current controllers that have a tendency to unbalance the machine excitation and cause noise and vibration. An intermittent sensor is even more insidious because the effect may come and go from just driving over a pothole or other road disturbance. Many investigators have tackled the problem of sensor elimination for the various types of electric machines. Before noting what has been done to eliminate position sensors, it should be noted that different machines require fundamentally different types of rotor position sensing. Synchronous machines, such as permanent magnet types, require a very accurate indication of where the rotor magnet is so that armature current can be maintained in quadrature to the rotor flux. This requires an absolute position sensor that resolves shaft position to typically 13

135 >15

81 >10

570

656

720

570

14.4

22.8

72.1

275.8

1.0

3.4

10.2

>120

exceeding 200  C. Tz QS was approved in 1997 and shown to have better than acceptable chemical resistance to automotive fluids (gasoline, oil and freon). Magnetek Motors and Controls offers a line of ‘corona-free’ inverter driven motors in their E-Plus line that are designed to withstand 1,600 V spikes in applications rated 600 V or lower [6]. Additional losses occur in the copper conductors of electric machines due to eddy currents and proximity effects [7,8]. As early as 1912, investigators made empirical measurements of what was, and continues to be, referred to as stray load losses in electric machines. These early investigators noted that even with very accurate measurements of resistances, currents and voltages, the efficiency measurements did not agree with measured electrical power input and mechanical power output, so these additional losses were treated as stray losses in the machine. The definition and origin of stray load losses continued to be puzzling well into the 1960s and even to this day, although now we have a more fundamental understanding of these losses. The sources of stray loss are considered to consist of flux pulsation in the core, tooth, slot leakage and surface or tooth tips. A result of flux crossing the conductor slots, hence the conductors transversely, is that the current distribution in the conductor is not uniform, nor even in phase in different sections. Hence, the conductor resistance is higher by an amount due to skin effect caused by transverse slot leakage flux and consequent eddy currents in the central portions of the conductors. Giacoletto [9,10] has analysed the issue of skin effect losses and, in particular, has shown that the voltage rise across the conductor reaches 2.5 pu during the leading and falling edge transients for an inverter drive operating at 10 kHz, with 5 ms rise time of voltage as depicted in Figure 8.5. In his derivation, Giacoletto uses a 1 A current source inverter driving the electric motor when the 2.5 pu overvoltages are generated. His general observation was that a hollow tubular conductor is more effective, on a mass basis, than a thin rectangular conductor at low frequencies, but that at higher frequencies it becomes less effective. In large turbo-generators for utility power (250, 600–1,000 MW rating), it is customary to take precautions against stray load losses. In small machines the stray load losses are noticeable and to some degree negligible. In very large machines the second and third order effects contributing to stray load losses become significant thermal design considerations. Armature winding stray load losses in large

406

Propulsion systems for hybrid vehicles

generators consist of circulating current and eddy current loss originating from cross-slot leakage flux. A remedy has been to transpose the armature bars along the length of the stator so that their position changes from top of the slot at one end, to middle of the slot in the central section of the stator, and on to bottom of the slot at the opposite end. The typical armature bar transposition is 540 so that end turn coupling is also minimized. Even with such transposition of a conductor in very large machines, it is common to still have a 20  C temperature difference between conductors at the bottom of a slot and those at the top of a slot. The reason is that cross-slot leakage is higher at the top coils so that higher eddy current losses are experienced.

8.2 Inverter Losses in the electronic power processor can be grouped into active component (semiconductor) losses and passive component losses. Passive components experiencing losses related to power throughput are the link capacitors, device snubbers if used and current shunts if used. Active device losses are decomposed into conduction, switching and reverse recovery losses. This section gives a brief introduction to inverter losses and some of the traditional methods used to quantify inverter losses.

8.2.1

Conduction

Conduction loss in a power inverter is due to the power dissipated in the semiconductor chip by the simultaneous current and voltage stress. During ON-state conduction, a majority carrier device such as a MOSFET will experience a voltage drop that is linearly proportional to the current through the device and the resistance of the device. In the OFF-state the resistance increases by six orders of magnitude or more. Minority carrier devices, on the other hand, experience conductivity modulation during the ON-state and have a voltage drop across the device terminals, which is a logarithmic function of the current through the device. IGBTs are representative of minority carrier devices as are diodes, bipolar transistors and thyristors. The simplest device, the bipolar diode, has a voltage–current characteristic given by the ideal diode (8.7), where k = 1.38  10 23 J/K (i.e. the Boltzmann constant), q = 1.602  10 19 coulomb is the electronic charge, K = 298 is the nominal temperature in Kelvin, and I0 is the diode saturation current ~10 14A. At room temperature the diode voltage coefficient is 0.026 and at a forward current of 10 A the diode voltage is 0.9 V. Power diodes at higher currents will have a different value of saturation current:   KT I ln VD ¼ q I0

ðVÞ

ð8:7Þ

Drive system efficiency

407

In addition to the voltage polarization, the diode also has bulk resistance and transport phenomena that can be modelled according to (8.8) for a more accurate assessment of diode conduction losses [11]: 

 T þ 273 PD ¼ 0:026 f ln jij þ c1 ig þ c2 i2 300

ð8:8Þ

where representative values for the constants c1 and c2 are 37 and 0.003. For the power MOSFET transistor the conduction loss can be modelled as PMOS ¼ i2 Rds ðTÞ PMOS ¼ i2 fRds ðT ¼ 20Þ½1 þ gðT 1 Rds ðTÞ ¼ 0 ðZ=LÞme ðT ÞC g ðU gs

20ފg

ð8:9Þ

U gsðTHÞ ðT ÞÞ

where the temperature coefficient of resistance (i.e. second term in the Fourier expansion) for a majority carrier device, g = 0.0073. Rds(T) is shown to be a function of the device active source perimeter, Z, and channel length, L, with multipliers of carrier mobility (m2/V-s), gate oxide capacitance per unit area (F/m2) and effective voltage at the gate [12]. The ON-state losses for an IGBT device can be developed from (8.8) since its behaviour is similar to that of a diode consisting of minority carrier injection, bulk resistivity and contact resistance. Power loss of the IGBT device is given in (8.10) where the dynamic resistance accounts for the MOS channel and contacts: PIGBT ¼ U ceðSATÞ i þ i2 Rd

ð8:10Þ

In reality, the IGBT is more closely approximated using (8.11) in which the exponent on device current is approximately 1.7 or less. The first term is the collector–emitter saturation voltage and the last term the dynamic resistance. At a junction temperature of 100  C the loss equation for an IGBT can be written as PIGBT ¼ U ceðSATÞ i þ Rd ih U ceðSATÞ ¼ 0:6 Rd ¼ 0:135

ð8:11Þ

h ¼ 1:645 The parameters in (8.11) are for a 130 A IGBT or the parallel combination of IGBTs necessary to sustain that current when hot.

408

8.2.2

Propulsion systems for hybrid vehicles

Switching

In power electronic systems the switching loss accounts for a significant fraction of inverter dissipation. For the power MOSFET and IGBT the turn-ON and turn-OFF switching energy is calculated based on the dc link voltage and load current. The switching power loss is then the switching energy times the switching frequency as follows: Psw ¼ f ðEON þ EOFF Þ

1 EON ¼ U dc tr i 6 1 EOFF ¼ U dc tf i 6

ð8:12Þ

where a hard switched inverter is assumed and the inverter current and voltage during the transitions are triangular. Switching waveform rise time, tr, and fall time, tf, determine the switching energy when the dc link potential is Udc. Figure 8.6 illustrates the derivation of (8.12) when the current and voltage transitions are linear. A bipolar junction transistor (BJT) will have an additional turn-OFF switching loss due to the phenomena of charge-storage ‘walk-out’, an effect of full current being sustained even as the device voltage begins to rise. The impact of temperature on the BJT stored charge is to support load current until the junction charge is depleted, and then the current begins to tail off. Since the link voltage is fixed, some investigators approximate the switching loss using a pair of exponents on current as shown in (8.13), which may be useful in U, I

Udc i

Switching event tr, tf

t

Figure 8.6 Switching waveforms for power semiconductors in hard switched inverter

Drive system efficiency

409

a computer simulation to get approximate results without a substantial amount of device characterization: Psw ¼ f ðc1 iu1 þ c2 iu2 Þ c1 ¼ 0:012

c2 ¼ 0:0042

u1 ¼ 1:25

(Hz, mJ/cycle)

ð8:13Þ

u2 ¼ 1:3

For the case of 100  C junction temperature and a fixed link voltage, c1 and u1 are for turn-ON and c2 and u2 are for turn-OFF. For different link voltages or different temperature the coefficients will need to be recomputed based on some device characterization. A similar procedure will work for MOSFET devices [13].

8.2.3 Reverse recovery In a hard switched converter with reactive current flow, there will be diode conduction when the transistor opposite the conducting diode is gated ON. Circuit current will switch to the transistor gated ON, but an additional component of current will flow through the switch in a shoot-through fashion until the diode is commutated OFF. The diode current is quickly reversed, but persists, for a duration of time necessary to sweep all the stored charge from the p–n junction. The time to accomplish this is the reverse recovery time during which an amount of charge Qrr is cleared. As with the active devices, the power dissipation in the diode during reverse recovery is calculated as Prr ¼ f Err

1 3U dc trr i Err ¼  2 6

ð8:14Þ

In many power electronic circuits, particularly those built with thyristors, it is necessary to add snubbers across the active device to limit dU/dt, dI/dt or some combinations. In the MOS controlled thyristor (MCT), it is necessary to add dI/dt snubbing in series with the anode to limit the rate of rise of current to prevent device damage. Gate turn off thyrsitors (GTOs) also require dU/dt snubbers to limit the rate of rise of voltage during forward recovery to allow the device sufficient time to internally stabilize. If a snubber capacitor is used, its switching power dissipation is calculated as shown in (8.15): Psnub ¼ f Esnub   1 3U dc 2 Esnub ¼ C 2 2

ð8:15Þ

410

Propulsion systems for hybrid vehicles

Other elements in the switching circuit may be analysed in a similar manner as done in the above three subsections. Link capacitors, for example, have losses equal to the circulating current and capacitor equivalent series resistance (ESR), or i2ESR.

8.3 Distribution system Losses in the power distribution system comprise harness and cable resistive losses, connector losses and fuse or contactor or other protective device losses. Fuses are sized to protect the downstream wiring from damage due to wire shorts to ground, or to other circuits, or faults at the load. In a fuse itself a ‘weak link’ or series of ‘weak links’ are regions of the ribbon element that are narrower than the fuse stock. At sufficient current, Joule heating proportional to I2t causes the weak link to begin to melt, at which point two unstable phenomena contribute to very rapid fuse link metal vaporization. First, the weak link itself begins to melt. Melting is followed by a surface tension effect in which the molten portions tend to form droplets that are still joined. At the napes of this series of droplets, or unduloids, the cross-sectional area becomes constricted below that of the original fuse thus further accelerating fuse vaporization. The second phenomenon that accelerates fuse clearing is the pinch effect whereby current flow through a liquid conductor reacts with its own magnetic field further constricting the conductor material into even smaller cross section [14]. The constricting pressure at the unduloid is proportional to the square of the current flowing in the cross section and inversely proportional to the diameter squared of the effective cross section. The effect is that of an Amperean force (as opposed to Lorentz force) that tends to separate the liquid conductor into individual balls. Fuse elements are made of low melting point materials such as silver or its alloys. Semiconductor protection fuses such as those used in each pole of a thyristor inverter, GTO, for example, are made of pure silver to achieve the fastest clearing time. McCleer [14] develops a concise theory of the fuse from both an electrical and a thermal model perspective. Electrical contacts are also analysed by McCleer [14] where he shows that contact resistance of a relay or contactor can be modelled as a constriction resistance due to a large number of asperities, N, having an average radius, ra, at the points of contact. The contact resistance is calculated as shown in (8.16) where rr is the contact material resistivity: Rc ¼

2

rr N P

ak

ð8:16Þ

k

Contact voltage drop is nearly self-regulating at 0.1–0.2 V per interface, with most instances of contact voltage drop in automotive circuits being in the vicinity of 0.025 V. Vehicle harness cables are sized to conform with allowable temperature rise in confined spaces and because of this they tend to follow the industrial practice of 3–5% line drop at rated load from source to point of load. For instance, the circuit

Drive system efficiency

411

feeding the tail lamps in an automobile would experience up to a 5% voltage droop under steady state loading. Depending on the chassis return path integrity, it is possible for this value to increase with ageing. As an example, suppose the tail lamps require 40 W of power to be delivered at a load voltage of 13.5 Vdc. In order to meet a less than 5% line drop, the source voltage (i.e. alternator regulated output voltage) must be 14.2 Vdc. The harness resistance is thereby constrained to be less than the value given in (8.17): Rc ¼ Rc ¼

1

h h

1

RL ð8:17Þ

h V L2 h PL

According to (8.17) for the example noted, the cable resistance would have to be less than 0.24 W. Example 1: Electrical distribution system in a vehicle is sized so that voltage drops at load points do not introduce more than a 5% droop. Calculate the maximum branch wire resistance for each of the following electrical loads given a power law relationship for load power sensitivity to system voltage of 13.5, 14.2 and 15.1 V: P2 ¼ P1



U2 U1

a

(a) Headlamps: P1 = 120 W; U1 = 12.8 V and a = 1.6 (b) HVAC blower fan on high setting: P1 = 300 W; U1 = 13.5 V; a = 2.4 (cage fan) (c) Cooling fan motor: P1 = 600 W; U1 = 13.5 V; a = 2.6 (axial flow fan) Solution: All the calculations of branch wire resistance will assume a line drop of 5% voltage. For the three cases of system voltage stated, Table 8.6 summarizes the interim calculations and wire resistance maximums for headlamps (a). The reader is requested to apply the same methodology to (b) and (c) using the stated power law. Table 8.6 Solution for headlamps in Example 1 System voltage (V) Load power (W) Load current (A) Wire resistance (mW)

13.5 120.37 9.386 71.9

14.2 130.52 9.675 73.4

15.1 144 10.04 75.2

The branch circuit wire must therefore be sized for the lowest system voltage in order to meet the criteria of 400 Wh/kg and energy density of >700 Wh/L. Solution: Given that each capacitor cell has L ¼ 5.08 mm, W ¼ 11.43 mm, t ¼ 9.732 mm and dielectric k-factor of 18,543: (a) How many individual parallel plate capacitance cells are necessary to form a C0 ¼ 30.693 F capacitor when all the cells are connected in parallel? (b) What is the electric field across each capacitive element when charged to maximum potential, Umx ¼ 3,500 V?

440

Propulsion systems for hybrid vehicles (c) Calculate the total package volume of this EESU if the packing factor PF ¼ 69%. (d) Using the volume calculated in (c) and given that the effective package density, including electronics is rEESU ¼ 6.2, calculate the total package mass.

Solution: (a) Let the total number of individual cells be Nc. Then C cell ¼

12

e0 KA ð8:854  10 ¼ t

Þð18:543  103 Þð5:806  10 5 Þ 9:732  10 6

¼ 0:9795 mF Nc ¼

C tot 30:693 ¼ C cell 0:9795  10

6

¼ 31:335  106

(b) The electric field in each cell for the stated dimensions is E¼

U mx 3,500 ¼ t 9:732  10

6

¼ 3:596  108 ;

or

3:596 MV cm

(c) Given that the aluminium flash coating on each dielectric slab of thickness, t, is 1 mm: Volcell ¼ Aðt þ 1 mmÞ ¼ ð5:806  10 5 Þð10:732  10 6 Þ ¼ 62:31  10 VolEESU ¼ ¼

11

N c Volcell 31:335  106 ð62:31  10 ¼ Pf 0:69

11

Þ

0:01953 m3 ¼ 0:0283 m3 0:69

This means a 28.3 L EESU package volume to store 52.22 kWh. (d) The package mass of the EESU is therefore M EESU ¼ rEESU VolEESU ¼ 4:06ð28:3Þ ¼ 114:9 kg This example serves to illustrate two important aspects of all ESS. First, it is possible to realize very large energy storage by the combination of many smaller cells. In the case of the EESU it required over 31M such cells all parallel connected. Second, combinations of elementary cells are normally arranged into modules and then modules are connected into packs. Reasonable packing factor for cells in modules and packs is generally on the order of 70%.

Energy storage technologies

441

Table 10.1 Energy storage mediums and their relative ranking Energy storage technology

Energy density – gravimetric (J/kg)

Energy density – volumetric (J/m3)

Nuclear fusion Nuclear fission Reformulated gasoline Ideal battery (Li-F) Fuel cell (Li-hydride) Lead–acid battery Flywheel Compressed gas at 35 kpsi Rubber spring Electric field in mylar* capacitor at E = Ebd = 16.5 kV/mi Electric field in barium titanate ceramic at 3 kV Magnetic field dipole–dipole interaction in iron at 2 T

3.4  1014 2.89  1012 4.4  107 2.19  107 9.2  106 1.6  105 5.3  104 10  104 6.2  103 4.3  103

2.37  1016 1.0  1017 3.3  1010 1.89  1010 8.6  109 4.6  108 8.1  108 3.0  108 6.2  106 6.0  106

2.5  103

1.5  104

2.0  103

2.4  104

*High pulse power electrostatic energy storage mediums consist of polycarbonate (dielectric constant = 3.2 and dielectric strength = 5 kV/mi), fluorene polyester (FPE, dielectric constant = 3.4 and dielectric strength = 10 kV/mi) and diamond-like carbon (DLC, dielectric constant = 3.5 and dielectric strength = 25 kV/mi). The energy storage density of these electrostatic mediums is >1  103, >2  103 and >4  103 J/kg, respectively.

10.1 Battery systems The world battery market is approximately a $30B business, 30% of which is devoted to motive power applications in the form of starting–lighting–ignition (SLI) batteries ($3B market) and the remainder to primary cells and sealed rechargeable cells. Global lithium ion battery market is some $8B in 2009 with over 2.8B small format cells sold annually. In 2009 the total market for large format hybrid electric vehicle nickel-metal hydride (NiMH) batteries was $1B worldwide for a gross market of some 815,000 hybrid vehicles. The hybrid vehicle breakdown by company in 2009 was Toyota Motor Co. 580,000, Honda Motor Co. 180,000, Ford Motor Co. ~30,000, General Motors ~15,000 and Nissan ~10,000. Taking the ratio of NiMH battery market to total hybrid vehicles sold in 2009 yields a pack price of $1,200/vehicle on average. Projections are for the lithium ion battery market for hybrid electrics of all configurations to grow from a very minor market in 2010 to $1B in 2015. Where it took 12 years for the NiMH market to reach $1B globally, it is expected that Li ion will reach this level in 5 years. A battery is a collection of electrochemical cells that convert chemical energy directly to electrical energy via an isothermal process having a fixed supply of reactants. The battery is self-contained and generally has constant energy density for the particular choice of active materials. We can view the battery as illustrated in Figure 10.1 to consist of an anode, cathode and electrolyte in a suitable

Propulsion systems for hybrid vehicles

Reduction

Cathode +

Anode –

Oxidation

442

Electrolyte ionic conductor

Figure 10.1 Cell construction container. Electrons are transported through the electrolyte from cathode to anode inside the cell, generating a potential across the cell as shown (cathode plate becomes positive and anode plate becomes negative). The origin of potential in an electrochemical cell can be viewed as the oxidation of fuel, resulting in displacement of charge. Figure 10.2 illustrates this process and some representative materials used for the anode and cathode. During discharge, active material, the fuel, is oxidized at the anode where it takes up electrons from the external circuit. Current flow in the external circuit releases energy. Electrons exit via the cathode where a reduction process ensues and after passing through an external circuit return at the oxidizing electrode, the anode. During recharge, the process is reversed as is the nomenclature of anode and cathode – electrons return via the anode, reconstituting the active materials at each of the electrodes. It must be pointed out at this point that the reactions involved in discharge and charge of battery systems may not be completely reversible, nor do

Fuel

Oxidant

Electron state ‘high’ Lead Zinc Hydrogen Lithium Anode

Electron state ‘low’ Lead dioxide Manganese dioxide Oxygen Cathode

Electron transport electrical energy output

Figure 10.2 Development of voltage in a cell

Energy storage technologies

443

the reactions necessarily proceed at the same rate in both directions or at both electrodes. This unsymmetrical reaction rate process will give rise to a different charge acceptance rate (generally much lower) than the charge release rate (generally high), and noticeably temperature dependent. Thermodynamics of battery systems are developed around the Gibbs free energy of the constituent materials used in the electrodes and electrolyte. In the ideal case this energy content, DG  (Btu, Cal, Joules) is defined as DG ¼

nFE

ðJ=molÞ

ð10:1Þ

where n is the number of electrons involved in the reaction, F is the Faraday’s constant1 (96,474 C/mol, 26.8 Ah/equivalent, 23.06 kcal where 1 cal ¼ 4.186 J) and E is the voltage. In an electrochemical cell the reaction kinetics determine the potential, with available energy dependent on the amount of materials present to take part in the reaction. As an example of cell voltage we consider a nickel– cadmium (NiCd) system. The reaction is a two-electron exchange process that can be written as follows: Cd þ 2NiOOH ¼ CdðOHÞ2 þ 2NiðOHÞ2

ð10:2Þ

DG þ 0 þ 2ð 129:5Þ ¼ ð 112:5Þ þ 2ð 108:3Þ

ð10:3Þ

DG ¼

ð10:4Þ

70:1

ðkcalÞ

Using the energy value obtained in (10.4) in (10.1) results in the cell potential of E¼

DG 70:1  103 ¼ 1:52 ¼ nF 2  23:06  103

ðVÞ

ð10:5Þ

According to (10.5) the NiCd system has a theoretical cell potential of 1.52 V. Of course we don’t get something for nothing. There are kinetics involved that will diminish this internal potential, resulting in a lower potential of 1.35 V at the terminals. The predominant loss kinetics involved in electrochemical cell thermodynamics can be grouped into potential losses resulting from electrode reaction kinetics (activation polarization), the availability of reactants (concentration polarization) and Joule losses (ohmic loss in electrodes and electrolyte). These polarization effects can be defined in terms of physical constants, number of electrons involved in the reaction, exchange current and electrolyte concentration. Activation polarization arises from hindrances to kinetic transport in the electrolyte of charge exchange during the reaction. The reaction rate at equilibrium

1

The Faraday constant is a convenient way to represent the ratio of Avagadro’s number to the electron charge, NA/q = 6.0221  1023/1.602  10 19 = 96,474 C/mol.

444

Propulsion systems for hybrid vehicles

determines charge flow, which in turn defines the exchange current, I0, as shown in (10.6): Ea ¼

RT ln ðI=I 0 Þ ðVÞ nF

ð10:6Þ

Equation (10.6) can be rewritten as a linear equation from which a Tafel plot can be constructed: E¼a

b logðIÞ ðVÞ

ð10:7Þ

Current

where a is a constant, b ¼ 2.303 RT/anF2 and a ¼ ~0.5 the charge transfer coefficient. By extrapolating (10.7) to zero in the Tafel plot, the value of the equilibrium exchange current at the given system temperature is obtained. An example of a Tafel plot is given in Figure 10.3 for representative charge transfer coefficients. Activation polarization has relatively fast time dynamics for build-up and decay.

Log (I) a2 1 – a1 a1 1 – a2

I0

Eeq.

Potential

Figure 10.3 Charge–potential behaviour of a cell electrode Concentration polarization is strongly dependent on the supply of reactants in the cell and how the by-products are removed or displaced. This effect is defined in (10.8), where C is the concentration in solution and Ce is the concentration of the electrolyte at the electrode surface, both in units of mol/L or mol/cm3: Ec ¼

2

  RT Ce ln C nF

ðVÞ

ð10:8Þ

The reader will recognize that a logarithmic base change was used: b  ln(M) = b  log(M)  ln(10) = 2.3026  b  log(M).

Energy storage technologies

445

The electrolyte concentration, C, can be expressed in Fick’s law form as a diffusion process having diffusion coefficient, De (cm2/s), and thickness of the diffusion zone, d (cm), as C¼

d  Jd nFDe

(mol=cm3 )

ð10:9Þ

In a one molal aqueous solution, the diffusion current density, Jd, or charge transport, is limited to J ~ 25 mA/cm2 with relatively slow time dynamics for build-up and decay. Ohmic polarization results from resistance of electrode materials, electrode current collectors, the terminals and contact resistance between the electrode active mass and electrolyte diluents. Accurate representation of ohmic polarization is modelled according to the cell geometry, material used and design of the current collectors (surface finish, coatings, bosses on electrode plates etc.): Er ¼ IðRelectrode þ Rcollector þ Rsurface Þ

ð10:10Þ

Ohmic polarization has instantaneous time dynamics for build-up and decay. Thermal effects also result from changes in the internal energy of the system due to temperature variations. This effect can be explained by expanding (10.1) into its thermodynamic equivalent expression relating to enthalpy and entropy change: DG ¼

nFE ¼ DH

DG ¼ DH

TDS

ð10:11Þ

nFTðdU =dT Þ

ð10:12Þ

When the change in internal energy, dU/dT, is >0, the ideal cell will heat up during charge and cool during discharge (Pb–acid is a representative case, and so are electrochemical capacitors, ultra-capacitors). When the internal energy changes, dU/dT is 0, meaning it is to be dissipated. This behaviour is explained by the strong irreversible nature of the polarization phenomena of heat flux Q as a function of entropy and Joule heating processes: Q ¼ T DS

IðEoc

Epol-total Þ

ð10:13Þ

As a result of the combined effects of polarization, the voltage–current behaviour of any electrochemical cell can be described as having three phenomenological regions as illustrated in Figure 10.4. In Figure 10.4 the end of useful life of the cell is defined as the terminal potential dropping to approximately 80% of its open circuit potential. This boundary is marked E-discharged. Temperature moves the voltage–current curve as shown by the diagonal trace. Lower ambient temperatures result in less useful

Propulsion systems for hybrid vehicles Voltage

446

Activation polarization

Eoc Ohmic polarization

Useful power

Concentration polarization

E-discharged

Current

Figure 10.4 Voltage–current behaviour of electrochemical cells power delivered by the cell. The normal operating voltage, shown as the region dominated by ohmic polarization, will generally have a very shallow slope until its end of life. The terminology ‘end of useful life’ is used to quantify primary cells. In secondary cells this metric defines the lower limit of state of charge (SOC), typically 10%, at which point the cell must be recharged. In lithium ion cells, for example, the discharge must stop before the cell terminal voltage reaches approximately 68% of open circuit potential. Going below this potential invites the onset of side reactions if the anode current collector metal (copper) and the electrolyte may irreversibly damage the cell. Typical lower bounds on discharge of lithium ion cells are 2.5–2.2 V or 2.0 V depending on specific chemistry. A useful relation for predicting the end of life of an electrochemical cell is the Peukert equation. This is an empirical relationship in which the current discharged over a time interval is shown to equal a constant: I nt ¼ k

ð10:14Þ

which is generally plotted according to (10.15) for values of n ¼ 1.0 for Pb–acid and NiCd, and n ¼ 0.95 for Li systems: LogðtÞ þ nLogðIÞ ¼ k

ð10:15Þ

More will be said later of useful life, discharge characteristics, and gravimetric and energy metrics for several common battery systems used in hybrid propulsion. A great deal has been written on battery electrochemistry, reliability and modelling. Principal considerations should be given to long term mechanical and chemical stability, temperature range of operation (in the case of automotive systems being 30 to þ70  C), self-discharge (shelf life), cell reversal, cost, cycle life (ability to reform electrode materials during recharge) and how well the battery can tolerate overcharge and overdischarge, both of which are abuse conditions. Battery ambient temperature is probably the most problematic on this list of attributes because

Energy storage technologies

447

environmental conditions can subject it to lows of 40  C and under-hood conditions can raise it well above the 70  C limit. Generally, vehicle batteries are located in more benign locations such as in the trunk, behind or beneath the rear seat or beneath the vehicle floor pan. The trend in battery electric vehicles (BEV) is to package beneath the vehicle floor pan or centre tunnel, over most of the vehicle length and protected from water splash and gravel impacts by a shield plate. BEV and plug-in hybrid electric vehicle (PHEV) battery packs are now being installed with provision for cell or module heating for colder climate operation. In Figure 10.5 the General Motors Chevy Volt, a range extended vehicle, has its 16 kWh lithium ion pack in T-shape along the centre tunnel. Some battery chemistries are very sensitive to overcharge and overdischarge. Lithium ion systems are a good example. Lithium ion has far more chemical energy than electrical energy storage, so its stability and charge/discharge must be carefully monitored and controlled, especially in high cell count series strings where any imbalance may take certain cells to overvoltage (OV) or undervoltage (UV) conditions.

Figure 10.5 Chevy Volt PHEV car and phantom view and its battery pack

10.1.1 Lead–acid Lead–acid secondary cells are used pervasively in automotive systems as standard SLI batteries in conventional vehicles, BEV, low speed neighbourhood EVs, NEV and low end hybrid vehicles. Recent improvements to SLI batteries since the development of maintenance free batteries in the 1970s have been the use of calcium as a hydrogen getter, other additives such as antimony for sulphation control, better current collectors and expanded grid assemblies. The typical Pb–acid system has a cell potential of 2.1 V, gravimetric energy content of 35–50 Wh/kg and volumetric energy of 100 Wh/L: Pb þ PbO2 þ H2 SO4 ¼ 2PbSO4 þ 2H2 O

ð10:16Þ

Lead–acid batteries are typically characterized at a C/20 discharge rate, where C is the capacity of the battery (Ah) at 20 h rate. Higher discharge rates incur higher internal losses and lower resultant useful power. Figure 10.6 illustrates the voltage– current discharge behaviour of a Pb–acid battery with discharge rate as a variable and temperature as a fixed parameter.

Propulsion systems for hybrid vehicles

Battery voltage (normalized)

448

1 0.9

0.8

C/20

C/8 0.7

Discharge limit 0.6 2

5

3C

2C

10

20 minutes

1C

30

60

2

3

5

10

20

hours

Discharge time at 20 °C (68 °F)

Figure 10.6 Discharge behaviour of lead–acid battery for various discharge rates

The discharge behaviour described in Figure 10.6 will shift left (shorter time intervals) as the battery is cooled. For example, at 0  C the 3C rate will result in a discharged battery in approximately 5 min on this same scale. Lead–acid batteries are amongst the oldest known rechargeable electrochemical couples. During discharge both electrodes are converted to lead sulphate and during charge the electrodes are restored. However, on charge, oxygen is liberated at the positive electrode and hydrogen at the negative electrode. This side reaction causes disassociation of water by electrolysis, resulting in water loss that must be periodically replenished, and hence the battery requires maintenance. During the early 1970s, maintenance free batteries were introduced that reduced water loss through oxygen recombination with freshly formed elemental lead on the negative electrode. In the presence of the sulphuric acid electrolyte, this oxygen combines with lead to form lead sulphate, causing depolarization of the negative electrode and thereby effectively suppressing hydrogen formation. Oxygen released through electrolysis is able to accomplish this because it has access via voids between the electrodes to react with the lead where the electrolyte is immobilized. Immobilization of electrolyte in the inter-electrode spaces was accomplished in two ways: (1) by use of an absorbent glass mat of a highly porous microfibre construction that is only partially saturated with electrolyte and (2) with gelled electrolyte. Adding fumed silica to the electrolyte causes it to congeal into a gel. When the battery is recharged, some water is lost, the gel dries and on subsequent recharging cracks and fissures propagate in the gel, thereby acting as channels for oxygen to find its way to the negative electrode and recombine. Use of a pressure relief valve helps to further regulate the flow of oxygen from positive to negative electrode. The large, prismatic, maintenance free or valve regulated lead–acid (VRLA) batteries normally have 1–2 psi gauge (psig) pressure thresholds on the relief valve. Smaller, spiral wound VRLA batteries can have pressures as high

Energy storage technologies

449

as 40 psig. Due to their novel construction, VRLA batteries are orientation flexible and can operate lying on their side, or under rotation when used in wind turbine blade pitch adjustors. But these applications either have or are being replaced with ultra-capacitors.

10.1.2 Nickel-metal hydride It is derived from what are commonly referred to as mischmetal compositions of either lathium–nickel (AB5–LaNi5) or titanium–nickel (AB2–LaNi2) alloy. If these alloys are referred to as ‘M’, the NiMH cell with potassium hydroxide (KOH) electrolyte becomes ð10:17Þ

M(H) þ 2NiO(OH) ¼ M þ 2NiðOHÞ2

Battery voltage (per cell)

where the mischmetal, or hydrogen electrode, coupled with the nickel oxyhydroxide electrode reacts with base mischmetal and nickel hydroxide. The capacity of NiMH cells is relatively high, but its cell potential is low, only 1.35 V as it was with NiCd systems, in fact as it is in all the nickel chemistries such as NiZn. Gravimetric energy density is ~95 Wh/kg and volumetric energy is ~350 Wh/L. NiMH does not have the high discharge rate capability of NiCd, but it shares a cell structure similar to NiCd. NiMH also suffers from relatively high selfdischarge, it is more sensitive to overcharge/overdischarge than NiCd, it requires constant current charging and it is more problematic for hybrid propulsion systems because of very reduced performance at cold temperatures. The problems with overcharging and discharging mean that some form of battery management system is necessary, as with all high performance advanced batteries. Figure 10.7 illustrates the discharge behaviour of NiMH cells, where nominal potential at 20  C is 1.35 V. The NiMH cell capacity diminishes rapidly as discharge rate is increased.

1.35 1.2 1.1 1.0 0.9 4C 0

20

2C

1C C/8 C/10

40 60 80 100 120 %DOD Discharged capacity (%) at 20 °C (68 °F)

Figure 10.7 Discharge curve for nickel-metal hydride (NiMH) cell

450

Battery voltage (per cell)

Propulsion systems for hybrid vehicles

1.35 1.2 1.1 1.0 0.9 0 °C

–20 °C 0

20

40 60 80 100 Discharged capacity (%) at 1C

20 °C 120 %DOD

Figure 10.8 Discharge curve for NiMH with temperature as parameter Figure 10.8 describes the rather poor temperature behaviour of NiMH systems. Because of this poor temperature characteristic, the use of NiMH in hybrid propulsion generally requires some form of climate control system such as heaters for cold operation and chillers for hot environments. Charge acceptance of NiMH is another concern for hybrid propulsion systems. At nominal temperature the cell potential is a strong function of the charge rate, particularly when the cell SOC exceeds 80%. This is illustrated in Figure 10.9 for various rates of charge. For higher voltage systems, NiMH cells are typically connected in series strings of modules, each module consisting of 6–10 cells. Nominal voltage for NiMH systems is 1.25–1.28 V/cell (some applications higher) and the nominal variation of 22% to þ16.7%. Self-discharge is high, typically 30% per month at 20  C. 1.6 Cell voltage (V)

1C 1.5 C/2 C/3

1.4 1.3 1.2 1.1

0

20

40

60

80 100 Capacity

120 140

% SOC

Figure 10.9 NiMH voltage versus SOC with rate as parameter at 20  C

Energy storage technologies

1.3

1.4

NiMH

1.2

dV

1.1

Cell voltage (V)

451

0

20

40

60 SOC (%)

80

100

Figure 10.10 NiMH charge characteristic The charge characteristic for an NiMH cell is illustrated in Figure 10.10. Because of the shallow slope in voltage for SOC values from 40% to 80%, it becomes difficult to implement a simple charge controller. Because the cell voltage increment is very shallow with increasing SOC, and perhaps not even monotonic, charge control of NiMH is more difficult than for lithium ion systems. Even more serious of an issue with NiMH is their precipitous drop in pulse power with decreasing temperature. It is common for NiMH to be limited to less than 40% of its 20  C capacity at 20  C. This is illustrated in Figure 10.11, where both discharge and charge power characteristics are shown. In Figure 10.11 a 30 cell, 16 Ah, 42 V nominal pack having pulse power capability of 15C (3 s at 20  C) is shown to decrease to 30% of this when cold. At 20  C the pack is only capable of ~5 kW. A 36 cell module, on the other hand, is capable of 6 kW at 20  C for 3 s. The module’s internal resistance is on the order

15

NiMH pack at 50% SOC 30 cell 50 V, 36 cell 58 V 30 cell

10

Power (kW)

36 cell

5 0 –5 –10 –15 –30

–20

–10

0 10 20 Temperature (°C)

30

40

Figure 10.11 NiMH power capability versus temperature (30 cell versus 36 cell string)

452

Propulsion systems for hybrid vehicles

of 36 mW, or 1 mW per cell including interconnects. Because of the serious limitations of NiMH at cold temperatures, some form of climate control system is necessary (heating element that discharges the battery) or some additional energy buffering is needed, such as an ultra-capacitor. Two examples are presented here of hybrid electric vehicle performance in different conditions [3,4]. In the first example Nissan performed test drives of identical Tino hybrids, one with NiMH and one with lithiuim ion battery packs in a cross-country route to examine ESS efficiency. The second example illustrates cold weather influence on a Toyota Camry hybrid and Ford Escape hybrid during winter testing at the Argonne National Laboratory. Performance was improved by replacing the NiMH pack in the Tino hybrid with Nissan’s in-house spinel based lithium ion cells processed with thin electrodes and laminate structure that provided twice the power (2.5 kW/kg) of NiMH, twice the energy (140 Wh/kg) in half the pack size [5]. Thermal management was facilitated by the laminate cell design. Two of the Tino hybrids were driven the same route from Los Angeles to Las Vegas to Phoenix and back to Los Angeles and their battery performance compared. The results are listed in Table 10.2 for reference.

Table 10.2 Status of electric energy storage systems in Tino hybrid vehicles

NiMH Li ion

d SOC (%)

Peff (kW)

Tbatt ( C)

Aheff (%)

Wheff (%)

C/D power restraint frequency (%)

40–80 30–85

4.25 4.28

52 49

91 98.8

83 95.1

44 2.2

Table 10.2 shows that for this field trial both the NiMH and lithium ion packs operated over a similar state-of-charge window (dSOC), delivered roughly the same average power (Peff), and all the while the lithium ion pack ran slightly cooler (Tbatt) and had substantially higher Coulombic (Aheff) and energy efficiency (Wheff). What is most revealing is what Nissan engineers refer to as the battery pack charge/discharge (C/D) power restraint frequency (last column in Table 10.2). That is, the number of times (as a percentage) that the battery management system (BMS) was required to limit the battery discharge power or to limit the rate of braking energy recuperation due to various factors, including the pack temperature. It can be seen in Table 10.2 that lithium ion delivers far superior performance than NiMH in real world driving conditions. Argonne National Laboratories also performed road evaluations on commercially available hybrid vehicles to determine the impact of battery temperature on vehicle fuel economy and braking energy recuperation limits [6]. Figure 10.12 illustrates the rate at which the vehicle NiMH packs curtail fuel economy as the pack temperature decreases in cold weather to 10  C and below. This is essentially the

Energy storage technologies

453

point made earlier in reference to Figure 10.11. The key finding now is that once lithium ion chemistries trend below approximately 0  C, their power capability becomes dramatically reduced just as it does in NiMH systems. Two significant aspects of advanced chemistry battery performance are revealed in these charts. First, vehicle fuel economy at 10  C is reduced to 50% of its 25  C normalized value. Second, the ability of the battery pack to accept charge decreases strongly for both cold and hot temperatures in the Camry and very dramatically for cold temperatures in the Escape hybrid. The differences in battery pack performance between these two vehicles are not known, but speculated to be associated with their respective thermal management systems and energy management strategies (EMS). Regardless, the fact remains that at cold temperature the vehicle performance and economy will be dramatically restricted due to limitations on the energy storage battery.

Normalized FE

1.6 1.4

Poly. (Camry)

1.2

Poly. (Esc)

1 0.8 0.6 0.4 0.2 0 –15

–5

5 15 25 Battery (NiMH) temperature (çC)

35

45

Regen braking energy (Wh)

350 300

Camry Esc

250 200 150 100 50 0 –15

–5

5 15 25 Battery (NiMH) temperature (çC)

35

45

Figure 10.12 Fuel economy and brake energy recuperation versus battery temperature

454

Propulsion systems for hybrid vehicles

Thermal management of ESS packs remains a matter of OEM preference, but air cooling prevails due to cost considerations. Some examples can be cited in support of this assertion: ●





The Ford Escape hybrid discussed here uses Sanyo NiMH cells and a separate dedicated air cooling loop. The 2010 Fusion hybrid with similar electric drive and ESS does not require a separate cooling loop. Recent introductions of battery electrics such as the Ford Focus EV, BMW Mini PHEV conversion and Nissan EV may also use air cooling of lithium ion packs.

Mild hybrids such as the Mercedes S Class and GM’s next generation BAS (belt alternator starter) are, or will be, designed around 120 V, 18–20 kW lithium ion packs.

10.1.3 Lithium ion Lithiated transition metal oxides are used as the cathode (positive terminal) in a lithium ion cell. The metal is typically bound within a host lattice during discharge and released during charge with no real change or damage to the electrode host. These lithium ions form the basis of the lithium ion cell chemistry as follows: LiMn2 O4 , Li1 x Mn2 O4 þ xLiþ þ xe

ð10:18Þ

C þ xLiþ þ xe , Lix C

ð10:19Þ

LiMn2 O4 þ C , Lix C þ Li1 x Mn2 O4

ð10:20Þ

Cathode (negative terminal) chemistry is defined by (10.18) and anode chemistry (positive terminal) by (10.19). Notice in these two expressions that some fraction of lithium metal is released into solution with an equivalent electron release to the external circuit at the cathode. Only the lithium ions are able to cross the separator and fill into pores in the anode host lattice. The anode (10.19) illustrates how lithium ions entering the host lattice reunite with electrons from the external circuit to form a carbon compound. Equation (10.20) illustrates the overall reaction, known as ‘rocking chair’ chemistry. The reversible parameter, x, in these equations is on the order of 0.85. On recharging, the carbon compound releases lithium ions back into solution that traverse the separator and combine with electrons at the cathode, reconstituting the lithium manganese oxide. Figure 10.13 shows that during charge a lithium ion oxidizes at the positive lithium–metal oxide (LiMO2) electrode and enters solution (the electrolyte consisting of LiPF6 plus additives) simultaneous with its orbital electron being removed from the cathode and delivered under external electromotive force (emf) to the anode. At the negative electrode, the free electron enters the graphitic structure simultaneous with a lithium ion being removed from solution and reduced in the anode to metallic lithium between the graphitic carbon layers such that three

Energy storage technologies

455

Charge A

e

Graphene structure

e

LiMO2 structure

x

x x

x

e

Li+ ion

x x x

x

x

x x

e

Lithium-ion battery

LiMO2

e

Electron

x

Solvent

Cathode Li1–x MO2 + xLi+ + xe–

Discharge

x Porous separator

PF–6 ion

Cathode (+)

Anode (–)

x

Al current collector

Anode Cn+ xLi+ + xe–

Charge CnLix

Figure 10.13 Lithium ion electrodynamics at cell level (from Reference 7) carbon atoms above and three carbon atoms below trap this lithium atom, hence LiC6. During discharge, the lithium de-intercalates from the anode graphite and reenters the electrolyte as a free ion. At the cathode, another free lithium ion enters the electrode structure and reconstitutes it in a self-assembly manner. This is why lithium ion cells require excess LiMO2 and binders – the mass transfer must not be excessive to collapse its structure. Roughly 50% of the cathode mass is therefore involved in reactions, and one useful rule of thumb is that energy storage is approximately 1 kWh per 170 g of metallic lithium. Lithium systems have a nearly reciprocal charge/discharge characteristic, or ‘rocking chair’ behaviour. A lithium system exhibits very high energy density, very good pulse power, highest cell potential and excellent cycle life. However, like the NiMH cell, it requires more capable charge/discharge management, generally under microprocessor control. A Li ion cell has a potential of 4.1 V open circuit, a gravimetric energy density of 125 Wh/kg and in excess of 300 Wh/L. Discharge potential is generally from 4.1 to 3.0 V or 73%. Cycle life at 100% depth of discharge (DOD) can exceed 1,000 cycles with a charge retention of 94%. Operating temperature for Li ion systems is only 20 to þ40  C on charge and 20 to þ45  C on discharge. It is of interest that Li ion has a useable SOC that is four times that of an SLI Pb–acid battery. This is because where a Pb–acid battery may only be operated from 90% to 40% SOC or less, the Li ion battery can easily operate from 100% to 10% or less SOC before recharge is necessary. This makes the Li ion very suited to hybrid propulsion. Li ion has the discharge behaviour illustrated in Figure 10.14. In a lithium ion cell the anode (negative terminal during discharge) is generally made of carbon (graphite), whereas the cathode (positive terminal during discharge) consists of a lithium–metal oxide composition. The electrolyte is an

Propulsion systems for hybrid vehicles

Battery voltage (per cell)

456

4.0

3.6

3.2

2.8 2C 0

20

40 60 80 100 Discharged capacity (%) at 1C

1C C/5 120 %DOD

Figure 10.14 Li ion discharge characteristic with rate as a parameter at 20  C organic mixture of lithium, phosphorous and other materials in a solvent, typically LiPF6 in solvent plus additives. The anode must readily release/accept (i.e. di-intercalate/intercalate) lithium ions for this type of cell to have superior electrical performance, mechanical ruggedness and long life. Significant advantages of lithium ion over NiMH are a significant weight reduction (~30% for same energy storage), much higher pulse power capability, self-discharge at least 20% less and future potential for lower cost. Li ion is also volumetrically smaller than NiMH, so vehicle integration is not an issue. Li ion systems that do not have free lithium metal present, other than trapped in the electrode lattice, are generally safe. These batteries are sensitive to overdischarging or over charging, particularly if the cells in a string are unbalanced. There is potential for fire, and if this occurs it is not advisable to use water. Extinguishers for Li ion systems are CO2 based or dry chemical types. Battery voltage range of Li ion systems should remain from 2.5 to 4.2 V per cell with a 3.68 V/cell nominal ( 30 to þ17.6%). Ambient temperature must be maintained within the range 30 to þ50  C. Long term storage is best done with discharged cells (80–90% DOD) and kept at an ambient temperature within the range cited, and dry. Self-discharge at 20  C is 80%, and especially when SOC > 90%. Li ion/polymer, on the other hand, requires continuous reduction of charge rate from low SOC all the way to 80% SOC. Above 80% its charge rate must be closely monitored, particularly cell potential, so that overcharging is not encountered.

0

Pb–acid (6×)

0

10

20

30

40

50 60 SOC (%)

70

80

90

Figure 10.16 Comparison of specific power of NiMH, Li ion and Pb–acid batteries (42 V nominal)

100

458

Propulsion systems for hybrid vehicles

More recently there has been experimental work on comparing the pulse discharge characteristics of a lithium ion battery with a high power ultra-capacitor [8]. The graphic in Figure 10.17 compares a commercial 12 Ah lithium ion cell versus a production 3,000 F, 2.7 V ultra-capacitor cell in the ability to capture regeneration energy in an HEV and then discharge it back into the vehicle system. Notice that at 100 s the lithium cell will capture five times more energy than the ultra-capacitor and return this with relatively high efficiency (the dotted trace). However, at 10 s both capture the same energy, but the ultra-capacitor discharges this back at >95% efficiency whereas the lithium ion can only discharge 50%. Therefore, for 10 s power the ultra-capacitor is twice as effective as the lithium ion in cyclable energy transfer. It is also evident that the ultra-capacitor applicability extends up to 20 s versus lithium ion. 1,000

Specific energy (kJ/kg)

battery 100 captured capacitor 10 stored

1 1

10

100 Charging time (s)

1,000

10,000

Figure 10.17 Energy shuttling comparison of lithium ion versus ultra-capacitor (from Reference 8 with permission) Today there are considerable varieties of lithium cell chemistries, most deal with replacing expensive cobalt in the cathode active material with less expensive elements such as nickel, aluminium, manganese, iron vanadium and to a lesser extent with the types of conductive and binding agents used [9]. To improve safety, by enhancing thermal stability, iron, magnesium and copper are used. Non-graphite anodes continue to receive significant attention with Li4Ti5O12 for improved safety and Li1.1V0.9O2 for improved capacity. For any of these cell types, cost is driven to a first order by the separator material with polymers, ceramics, non-wovens and metal oxide composites. Alternate materials for the separator section include natural as well as synthetic materials. For natural separators, the products are mostly cellulose products or their chemically modified derivatives. For the synthetic counterparts, polymers of many types including nylons, polyolefins, PTFE and polyamides (to name a few) are used, as well as synthetic inorganic forms such as ceramics and

Energy storage technologies

459

glasses (mostly for high temperature applications). For automotive applications where ESS performance and durability are paramount, the Li(Ni, Co, M)O2 (M ¼ Al or Mn) based positive electrode is being actively investigated; however, durability of this electrode at high temperatures is not known [10]. It was found that electrode structural changes occurred in the LiNiCoAlO2 positive particles, leading to increase in resistance. Kim and Cho [11] argue that replacement of cobalt in the popular LiCoO2 cell with nickel, to take advantage of its lower cost and high capacity, introduces a number of significant problems. First is the moisture intake into the cathode slurry, leading to gelation of the positive electrode mix and non-uniformity of the final electrode film. Second is the issue of structural instability above 60  C. This is confirmed by storage at 90  C, where it is found that nickel and other active metal ions dissolve into the electrolyte, resulting in cathode particle coating with Li2CO3 and the attendant process of gas evolution. Third is the problem of strong O2 gas evolution above 200  C, leading to thermal run-away. The authors demonstrated that use of lithium-reactive Co3(PO4)2 nanoparticle coating completely blocked the metal ion dissolution during storage at 90  C. Metal doping and coatings are also reported for the LiNi1/3Co1/3Mn1/3O2 chemistry because of its cost and performance, but its low electrode density and low specific capacity compared to higher nickel content electrodes remains an issue. In Reference 12 EnerDel reported on the safety of lithium ion for automotive applications and noted that LiNi1/3Co1/3Mn1/3O2 is slated for production in 2010 as cell to module to systems for EVs and HEVs. To improve capacity, layered oxides (Co, NMA, NMC), spinel (LiMn2O4) and olivine (LFP) structures are being pursued, leading to 100% SOC capacity of 278 mAh/g. Henriksen [13] describes the development of high capacity positive electrodes for application to PHEV that have a capacity of >150 mAh/g, typical of conventional lithium ion, which is realized from (1) increased particle density (1.3 to >2 g/cc), (2) increased rate capability (C/24 ! C/3) and (3) optimized processing conditions for higher capacity density and stability. Henriksen’s work points out that the major contributor to impedance rise is the interfacial impedance of the positive (þ) electrode. The most probable mechanisms are reduced interfacial diffusion due to changes in surface films or oxide surface layers and the preferential isolation of small active material particles. The problems encountered with present lithium ion processing amount to challenges facing those pushing a now two decades old slurry process to meet the demands of vehicle application in plug-in and BEVs. The performance of the battery is simply inadequate to meet vehicle demands for power and energy. Until this dilemma is solved, the prospects of a commercially viable PHEV or EV remain unknown. To conclude these sections on battery systems, a compilation of advanced battery system technologies is listed in Table 10.3 that gives specific attributes of each technology, representative cycling capability and a metric listed as energy-life to quantify the throughput energy of each battery system until it enters wear-out mode. Wear-out mode of a battery system is taken as the point at which its capacity has diminished by 20% of rated. Table 10.3 contains some interesting comparative data. The table itself is composed of ESS technologies that have been optimized either for energy storage

VRLA TMF NiMH Li ion Li Po EDLC

Type

250

180 220 300

70 90 140

Power (W/kg)

35

Energy (Wh/kg)

1,200 600 800

400

Cycles (# at 80% DOD)

Battery electric vehicle

2.6 2.4 2.1

7

P/E (#)

67,200 43,200 89,600

11,200

Energy-life (#Wh/kg)

Table 10.3 Summary of battery storage system technologies

80 800 1,000 1,500 9,000

4

Power (W/kg)

25 30 40 65

Energy (Wh/kg)

500 k

300 – 5,500 2,500

Cycles (# at 80% DOD)

Hybrid vehicle

2,250

3.2 27 25 23

P/E (#)

1,600,000

6,000 ? 1,76,000 1,30,000

Energy-life (#Wh/kg)

30, þ70 0, þ60 0, þ40 0, þ35 0, þ40 35, þ65

Range ( C)

Temp

Energy storage technologies

461

(BEVs) or for high pulse power applications (i.e. hybrid electric vehicles). In the case of energy optimized systems, the cycle life is quantified at 80% DOD. Note that a lead–acid battery fabricated as a thin metal foil (TMF) structure is not suited to EV applications, and in general, neither is the electronic double layer capacitor (EDLC) nor, more generally, the ultra-capacitor. Energy capacity multiplied by deep cycling capability (0.80) multiplied by cycles to wear out gives a metric of energy-life. For example, if a vehicle consumes 250 Wh/mi on a standard drive cycle, and it uses a Li ion traction battery (listed in Table 10.3) that is capable of 43,200 Wh/kg of energy throughput, and if a battery life of 8 years (100,000 mi) is specified, then a range of 172.8 mi/kg can be anticipated. To meet the range goal requires some 578.7 kg of battery. If the average consumption increases to 400 Wh/mi, then a battery mass of 926 kg is necessary. Battery systems for hybrid vehicles are optimized for shallow cycling (1% to perhaps 4% of capacity per event) and have comparably higher cycling. In Table 10.3 the hybrid system cycling capability has been projected from its low DOD cycling capability to a comparative value had its DOD been 80% on a log–log plot. This said, the metric of energy-life of the hybrid vehicle, advanced battery technologies, will be typically more than double that of EV batteries. Now, it can be seen from Table 10.3 that ultra-capacitors (EDLCs) are capable of ten times the energy-life of even pulse power optimized advanced batteries. The temperature application range of ultra-capacitors is also much better than that of battery systems. Hence, there is resurgence of interest in ultra-capacitors as energy buffers in hybrid and fuel cell applications. The next section expands on the topic of ultra-capacitor ESS.

10.2 Capacitor systems Table 10.1 listed the ideal conventional capacitor as storing 4.3  103 J/kg (1.2 Wh/kg) in its electric field. Practical conventional capacitors, of course, are not capable of even this amount of energy storage. A conventional capacitor achieves high capacitance by winding great lengths of metal foil plates separated by a dielectric film. The voltage rating is determined by the dielectric strength (V/m) and its thickness. An ultracapacitor works differently. Instead of metal electrodes separated by a dielectric (sheet or film) that facilitates charge separation across its thickness, an ultra-capacitor ˚ ). The achieves charge separation distances on the order of ion dimensions (~10 A ultra-capacitor’s charge separation mechanism, or double layer capacitor model, was described by Helmholtz in the late 1800s. In Figure 10.18 the construction of an ultracapacitor is shown to consist of carbon (AC) film electrodes that are impregnated with conductive electrolyte and laminated on current collector metal foils. Positive and negative foils with this carbon mush have an electronic barrier or separator that is porous to ions between them, generally of 50–80% porosity. The electrolyte materials commonly used in ultra-capacitors are propylene carbonate (PC) or with acetonitrile (AN, i.e. 10–20% by mass, but generally 75% by volume) and the quaternary salt tetraethylammonium tetrafluoroborate (TEATFB, 5–15% by mass and 25% by volume), AC (10–20% by mass) and the remainder the

462

Propulsion systems for hybrid vehicles

Current collector Electrode Carbon matrix

d

2Ceq

Electrolyte (ACN + TEATFB)

Ceq

Separator Electrolyte (ACN + TEATFB) 2Ceq

Carbon matrix Electrode Current collector

Figure 10.18 Ultra-capacitor cell construction metallic cell package, plastic covering, end seal and terminations. AN is a toxic substance on its own, but in the ultra-capacitor it is in solution with other organic constituents and in low free volume since it is mostly absorbed into the AC electrodes and separator. There is generally no safety concern with AN even if the ultra-capacitor is in overvoltage and outgassing of the electrolyte occurs. However, should the gas effluent be burned in an oxygen starved atmosphere then there is the potential to generate cyanide gas, HCN. Application of ultra-capacitors into vehicles must take into account proper installation, crash worthiness and abuse, just as lithium ion and advanced battery systems. Today, there are numerous toxicology tests being performed by various testing institutions that validate claims of its safe use. The ultra-capacitor gets its enormous surface area from the porous carbon based electrodes that can provide nearly 2,000 m2/g. The charge separation distance is not dependent on any dielectric paper or film or ceramic, but by the size of the ions in the organic electrolyte that is on the order of angstroms. Figure 10.19 is an illustration of ultra-capacitor carbon electrode porosity and ion size, including an illustration of how charged ions accumulate into the various regions of activated carbon electrode pores. Pores on the nanoscale can have a diameter on the order of the ions, so that accumulation of ions into these pores is blocked. If this is the case, the EDLC effect is not seen for either aqueous or organic electrolytes. Ion diameter ~1 nm Nano pore 1.5 nm diameter

+ –

+ Activated carbon electrode deposit

+

– –

+

+V Activated carbon electrode deposit

– Meso and macro pores

+

+ + + + Meso and macro pores

Figure 10.19 Illustration of an electronic double layer capacitor system

Energy storage technologies

463

As seen in Figure 10.19 the ions in meso and macro pores will accumulate into layers resulting in an electric field within the electrolyte. This phenomenon results in the EDLC having a capacitance that is somewhat voltage dependent. The electric field across an isotropic dielectric is linear, but when in the presence of distributed charge within the electrolyte, it obeys the Poisson equation. As the capacitor is charged, the electrolyte becomes depleted of ions and further layering is slowed down. To illustrate an ultra-capacitor’s energy storage mechanism, consider a popular production cell rated 2,700 F, 2.5 V (2.7 V maximum), 625 A pulse discharge and having an internal resistance of 1 mW  25%. The capacitance dispersion on ultracapacitors is typically 10%, þ30% of rated, and its operating temperature is 40 to þ70  C. The production cell has a mass of 725 g in a 0.6 L prismatic package. From these data its energy rating, gravimetric energy and power density, and volumetric energy density are determined: 1 2,700ð2:5Þ2 ¼ 8,400 J Ec ¼ CV 2mx ¼ 2 2

ð10:21Þ

Charge separation distance, d, can be approximated using the facts listed and substituting into (10.21). Before proceeding we note that from the ultra-capacitor construction that the terminal capacitance used to calculate the stored energy is actually the equivalent of two electrolytic double layer capacitors, 2Ceq, connected in series, each having the charge separation distance d. Using this new insight into the ultra-capacitor, we approximate the ionic separation distance, d, as d¼

er e0 re mc 2C eq

ð10:22Þ

where the dielectric constant er ¼ ~3, permittivity e0 ¼ 8.854  10 12 F/m, specific area density re ¼ 2,000 m2/g and cell mass mc ¼ 725 g. The capacitance C for this unit is given as 2,700 F. From this, (10.22) predicts that d¼

3ð8:854  10 12 Þ2,000ð725Þ ¼ 3:57 nm 2ð2,700Þ

ð10:23Þ

For this relatively crude illustration, (10.23) yields an ionic separation of just ˚ . If the activated carbon pore size is 2 nm, both aqueous and organic electrolytes in the activated carbon electrode structure will exhibit the EDLC effect. There is considerable research at present into the ion kinetics and adsorption effect that leads to electronic double layer capacitance (EDLC) in aqueous and organic electrolytes. Figure 10.20 is a taxonomy of energy storage technologies that is a useful guide to where in this overall picture the EDLC resides. Our focus will be on electrochemical capacitors of symmetric design in organic electrolyte.

464

Propulsion systems for hybrid vehicles Energy storage components

Capacitor

Battery

Primary

Lead– acid

Secondary (rechargeable)

NiCd

NMH

Electrostatic Electrolytic Electrochemical Li ion Symmetric Aqueous electrolyte

Organic electrolyte

Asymmetric Organic Aqueous electrolyte electrolyte

Most popular today Potential for bulk storage

Active research

Figure 10.20 Family tree of electrochemical energy storage technologies To understand the EDLC effect in more detail, consider the graphics in Figure 10.21 in which an electron distribution is assumed in the activated carbon (AC) and cations are present in the electrolyte, TEATFB/AN (reads tetraethylammonium– tetrafluoroborate in acetonitrile as the salt in a solvent). We are interested in understanding in more depth the reason for the compact layer (i.e. the Helmholtz layer) and for this a digression into electrochemistry is in order, specifically the governing relationship for the electron–ion charge separation distance. This distance is now more appropriately referred to as the Debye length, dc [14]. The electric field strength shown in Figure 10.21 across this boundary is high as depicted in the right hand vector plot for charges in non-conducting medium. The electrostatic forces are very high for nanoscale charge separation distances, but are balanced by van der Waals nuclear forces as the spacing becomes very small (nanometres). For this situation the Debye length is defined as sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi er e0 RT ¼ 6:67  10 dc ¼ 2F 2 C 0

9

ð10:24Þ

where er ¼ 37.5, e0 ¼ 8.854  10 12, R ¼ 8.314 J/K-mol, T ¼ 300 K, F ¼ 96,474, C0 ¼ 1 mol/L. From (10.24) it is clear that a charge separation distance of 6.67 nm can be expected in an electrolyte concentration of monovalent salts when the salt concentration is 1 molar at room temperature. An example will help illustrate the applicability of Debye length in the calculation of electrode capacitance of a carbon–carbon symmetric ultra-capacitor.

Energy storage technologies

465

Vector field of the electric field Charge: electron –

cation +

+

5–

+

4– 3–

+ +

2– 1–

+ 0

1

2

3

4

5

6

(V1, V2)

x (10–10 m)

Figure 10.21 Illustration of an electron distribution in ac and ions in solution Example 2: Using the Debye length, dc, calculated above find the EDLC capacitance for a symmetric carbon ultra-capacitor in organic electrolyte, given the mass of AC, MAC, on the electrode and the specific area, SA, of carbon surface. Given MAC ¼ 72 g (effective considering total electrode film mass) SA ¼ 1,620 m2/g of specific area in the carbon measured by BET method. Solution: Using the definition of capacitance when plate area is decomposed in this case to a porous electrode having the given mass and specific area and using the Debye length for charge separation distance (see graphics in Figure 10.21 showing spacing of 2 nm as example), we obtain C elec ¼

er e0 M AC S A 37:5ð8:854  10 12 Þð72Þð1,620Þ ¼ ¼ 5,806 F dc 6:67  10 9

For this electrode capacitance the cell effective capacitance is Ceq ¼ Celec/2 ¼ 2,903 F, and since each electrode is only half the terminal voltage, or 1.35 V for organic electrolyte, the terminal voltage is 2.7 V. Hence, this would be approximately a 3,000 F, 2.7 V ultra-capacitor. To further illustrate the equivalent plate size of this production ultra-capacitor, we put the separation distance and capacitance value into the formula for a classical two-plate capacitor and solve for the area, getting: d c C elec 6:67  10 9 ð5,906Þ ¼ 1:186  105 m2 ¼ er e0 37:5  8:854  10 12 A ¼ 11:86 ha



A  30 acre

ð10:25Þ

466

Propulsion systems for hybrid vehicles

The area given by (10.25) is enormous. To put this into perspective we divide by 104 m2/ha and obtain 11.86 ha! Or, in English units, this amounts to some 30 acres. In other words, the acreage of a small farm rolled up into a small canister with a volume of 0.6 L.

10.2.1 Symmetrical ultra-capacitors The ultra-capacitor just described is a symmetrical type because both of its electrodes are composed of the same porous carbon matrix ingredients. Capacitance is purely a double layer effect, there is no ionic or electronic transfer as in electrochemical cells, only polarization. Unlike an electrochemical cell that functions by virtue of the Faradic process of ionic transfer, an ultra-capacitor is a non-Faradic process that is simply charge separation and no electronic transfer. In a conventional capacitor the energy storage effect is purely a surface phenomenon, so most of the materials used are there for structure, not for energy storage [15,16]. Ultra-capacitors, however, achieve phenomenal surface area for rather finite plate areas by having porous electrodes that are dense with crevices and pores. A battery makes the best use of available materials because the electrode mass contributes in a Faradic process to the energy storage task. However, the Faradic process involves ion transfer, so there are transport delays and time dynamics to contend with. Capacitors and ultra-capacitors have very fast pulse response times because only stored charge is removed or restored at the interfaces rather than reactions occurring in the bulk electrode material. By extension, this means that ultra-capacitors have cycle life orders of magnitude greater than electrochemical cells. It is not unreasonable to expect an ultracapacitor to provide several million cycles in use. The Nissan production Condor capacitor-hybrid truck, a commercial 4 ton load capacity, 7 L diesel series electric hybrid with twin 55 kW ac synchronous motors, relies on a 583 Wh, 346 V ultracapacitor module. A commercial truck such as this is designed for urban stop–go driving with a durability target of 600,000 km of driving over which it is expected to encounter 2.4 M braking cycles. The reason for using an ultra-capacitor is to store regenerated braking energy and deliver it to the electric drive system for vehicle launch and acceleration. Moreover, Nissan Motor Co. [17] claims a 50% improvement in fuel consumption and CO2 reduction of 33% with its capacitorhybrid system. A passenger car hybrid would target roughly one-third of the mileage and stop–go events as the delivery truck, or 200,000 km of driving and 800,000 stops [18,19]. Energy and power density of ultra-capacitors is of utmost importance. Energy density, for example, translates into the ability of the storage system to source vehicle power demands for protracted lengths of time – tens of seconds instead of just fractions of a second. Power density provides an indication of how well the ultra-capacitor can deliver pulse power when needed. The classical relation between energy and power density is the Ragone plot in which a collection of data points are plotted with specific energy density (Wh/kg) on the ordinate and power density (W/kg) on the abscissa. Each point on the Ragone plot is the result of a

Energy storage technologies Electrochemical storage

1,000

NaS Li ion Pb–acid

100 Energy density (Wh/kg)

467

10

Ultra-capacitor

1

2,700 F, 2.5 V

0.1

Film capacitor

0.01 100

10

1,000 Power density (W/kg)

10,000

Figure 10.22 Ragone plot of energy storage systems

constant power discharge experiment. Figure 10.22 is a Ragone plot for some representative ESS. Empirical specific energy density versus specific power density trend lines are depicted in Figure 10.22 for various energy storage technologies. These trends have been characterized in the laboratory and derived from constant power test data. The following set of equations gives energy density, gE, as a function of power density, gP, which will be useful in sizing operations to be described later. The trend line data fit consists of two parameters: k1 the high power rate discharge test and k2 the slope derived from high minus low power rate test data: gE ¼ k 1

k 2 gP

ð10:26Þ

For a sodium–sulphur advanced battery system (flow type battery [6]): gE ¼ 130

0:034gP

ð10:27Þ

For a lithium ion battery system: gE ¼ 75

0:025gP

ð10:28Þ

For an NiMH battery system: gE ¼ 50

0:035gP

ð10:29Þ

For a lead–acid battery system, assuming valve regulated technology: gE ¼ 50

0:2gP

ð10:30Þ

468

Propulsion systems for hybrid vehicles For a carbon based, symmetrical ultra-capacitor system: gE ¼ 3

0:00127gP

ð10:31Þ

The specific power density range used in (10.27) through (10.31) is 300– 500 W/kg except for the lead–acid battery, a valve regulated lead–acid (VRLA) design for which the data sets are restricted to 20–80 W/kg discharge rates. Ultracapacitors modelled by (10.31), on the other hand, have specific power density values ranging to over 10 kW/kg with newer units now approaching specific energy values of 6 Wh/kg and specific power densities of 12–15 kW/kg and higher. Dual mode hybrid (i.e. hybrids having electric-only range) and battery electric vehicles have characteristically different energy versus power relationships. Advanced batteries for dual mode operation have the following Ragone relationships: Lithium ion dual mode battery has the following specific energy versus specific power (refer (10.28)): gE ¼ 125

0:48gP

Similarly, the NiMH dual mode battery is characterized as follows (refer (10.29)): gE ¼ 63

0:145gP

In dual mode application, the ESS battery is designed for higher specific energy at the expense of power. Consequently, the sensitivity of specific energy to specific power is dramatically higher. Furthermore, cell design is often tailored to specific applications with prismatic designs having higher efficiency than cylindrical designs. For example, an NiMH cell at 80% SOC and discharged at a C/5 rate will exhibit the following typical efficiencies: cylindrical ¼ 82%, prismatic ¼ 94%. In these expressions relating data curve fits within regions of the Ragone plot, the ranking from top to bottom has been intentionally done in terms of highest to lowest specific energy density. It is apparent that electrochemical cells are from one to two orders of magnitude more capable than ultra-capacitor cells, which in turn are superior to parallel plate capacitors. Today, the utility of Ragone charts is still appropriate to contrast different energy storage technologies as demonstrated in Figure 10.23, where the Ragone line is shown intentionally as a band to reflect an E(P) ¼ Esto tPln{Ui/Uf} characteristic as the storage system capacity changes from top of charge (Ui) to some lower value represented here by Uf. By taking the ratio of E/P in these log–log Ragone charts, a new set of parameters can be inserted that reflect time duration of energy exchange. For example, an energy storage unit that stores 1 kWh of energy and then discharges this at a rate of 100 kW will be capable of sustaining this power level for E/P ¼ 1 kWh/100 kW ¼ 36 s. Consider the Ragone chart in Figure 10.24 that is complete with a set of characteristic time diagonal lines (lower left to upper right slant). Notice here that the case in point has a P/E ¼ 100, which represents a very high power lithium ion cell. In the Ragone chart below, compiled using manufacturer data (the star points are

469

Energy storage technologies

Specific power (W/kg at cell level)

1,00,000 Li ion Super very high power capacitors 10,000 Lead–acid Li ion spirally wound high power Ni-Cd

1,000

Ni-MH

Li ion high energy

Na/NiCl2

100

LiM-polymer Lead–acid

10

1 0

20

40

60 80 100 120 140 Specific energy (Wh/kg at cell level)

160

180

200

Figure 10.23 Ragone chart for various electrochemical energy storage components (from Reference 20)

1×103

10

h

36

0.

s

6

3.

s

36

s

0.

Li ion

100 Specific energy (E, Wh/kg)

1h

1h

s

36

m

NiMH Pb–acid s

10

6

3.

m

acap

Ultr

1

0

36

0.1 s an Ev

0.01

1

10

p

36

ca

µs

µs

ec -el s Al cap m f il

1×104 1×105 100 1×103 Specific power (P) at 90% efficiency (W/kg)

Figure 10.24 Ragone chart with characteristic time as parameter

1×106

470

Propulsion systems for hybrid vehicles

production cells), the 36 s time line crosses right through these VHP cells and also the very limits of NiMH. Ultra-capacitors lie midway between batteries and the aluminium electrolytic film and ceramic capacitors (~36 ms). The Evans Capacitor is a novel hybrid of tantalum with carbon ultra-cap and is used primarily in pulsed radar applications. Example 3: Clarify what is meant by the United States Advanced Battery Consortium (USABC) goal for high power, lower energy, energy storage system (LEESS) for power assist hybrid electric vehicles – specifically, the terminology ‘Energy over which pulse power goals simultaneously met’. The scenario is this: ● ● ●

Discharge at 20 kW for 10 s (therefore, 20 kW  10 s ! 56 Wh) Charge at 30 kW for 10 s (therefore, 30 kW  10 s ! 83 Wh) Specify the energy window for vehicle use: Euse ¼ 165 Wh Discharge pulse = 56 Wh 26 Wh

Charge pulse = 83 Wh Energy to meet both power goals

Vehicle energy window = 165Wh Chg: 56 Wh + 26 Wh = 82 Wh Dchg: 83 Wh + 26 Wh = 109 Wh

Therefore, the ESS must supply 26 Wh of energy in order to meet a vehicle energy window of 165 Wh. The following graphic will help clarify this in terms of hybrid vehicle regeneration and boosting events. Cumulative energy (Wh) to Vehicle Boosting Time (s)

Regen

In use energy window

10.2.2 Asymmetrical ultra-capacitors A variant of symmetrical, carbon–carbon electrode ultra-capacitors is the asymmetrical carbon–nickel super-capacitor. Presently considered for engine cranking

Energy storage technologies

471

Nickel oxyhydroxide

Carbon

on large trucks and hybrid vehicles having electric-only range capability, the asymmetrical super-capacitor, or pseudo-battery, has the capability of very high pulse power and somewhat more energy than an ultra-capacitor. A 60 kJ supercapacitor, charged from a small 12 V lawn tractor battery, was demonstrated to crank a 15 L over the road truck diesel engine for 15 s, several consecutive times before its voltage was too low [21]. This testing was done even with the cold capacitor, and it outperformed even a parallel connection of Group 31 lead–acid modules. When super-capacitors are combined with batteries, such as those described in References 22 and 23, it is common to use energy storage capacity in the range 30–100 kJ. Combined with a 12.6 V lead–acid battery, this means a super-capacitor of 378–1,260 F would be required. A typical super-capacitor module would have one-half the volume of a Group 31 Class A truck battery. The super-capacitor structure and chemistry is shown in Figure 10.25. Supercapacitors consist of one polarizable electrode, carbon, and one Faradic electrode made of nickel oxyhydroxide (NiO(OH)) in a potassium hydroxide aqueous solution.

KOH solution

Figure 10.25 Electrochemical capacitor or super-capacitor As an illustration of production super-capacitors, or pseudo-batteries, one manufacturer, ESMA, located in Russia produces these cells in modules for automotive applications. A 14 V super-capacitor module can deliver very substantial peak power. Characteristics of two ESMA production modules are listed in Table 10.4

Table 10.4 Electrochemical capacitor parameters ESMA capacitor type

Voltage range at given energy

Peak power (kW)

Energy (kJ)

Resistance, Ri, at 25  C (mW)

Resistance, Ri, at 30  C (mW)

10EC1024 20EC402

14.5 V ! 4 V 14.5 V ! 4 V

8.7 35

30 90

6 2

9 3

472

Propulsion systems for hybrid vehicles

Simulation is used to determine ultra-capacitor pack sizing for engine cold cranking. The worst case is cold cranking a compression ignition direct injection (CIDI) engine (e.g. diesel engine) at 30  C. Such a simulation was performed in ANSYS Simplorer using an ultra-capacitor model, solenoid model, permanent magnet starter motor model, gearing and a simple engine friction and inertia model. Figure 10.26 illustrates the simulated ultra-capacitor module voltage (a 6S  1P  3,000 F module) charged to 16 V and applied to the engine starter motor for approximately 1.5 s. The starter motor is assumed to have a torque constant, UC volts

Str_Mtr_Amps

16.00

2.52k 2.00k 1.50k

AM1.I [A]

14.00

AM1....

VM1....

VM1.V[V]

15.00

1.00k 500.00

13.00

0 –480.00

12.08 0

250.00 m 500.00 m 750.00 m 1.00

1.25

1.50

1.75

0

2.00

250.00 m

750.00 m 1.00

1.25

1.50

1.75

2.00

t Engine crank speed (r/s)

20.00 15.00

WM_...

WM_crank.OMEGA [rad/s]

24.80

10.00 5.00 –1.98n 0

(a)

500.00 m 1.00 250.00 m 750.00 m

1.25

1.50

1.75

2.00

Battery

BOOSTCAPS +

Starter

Engine

(b) Isolating relay +

5 A fuse +

Cranking motor

Required 10psi oil pressure switch

Batteries

BOOSTCAP* module –

(c)

Required relay

Figure 10.26 (a) Ultracapacitor voltage, ultracapacitor current and engine speed, (b) Block diagram of ultracapacitor assisted battery for engine cold cranking, (c) Architecture of ultracapacitor plus battery for engine cold cranking

Energy storage technologies

473

kt ¼ 0.0085 Nm/A. The isolation relay shown at the bottom of Figure 10.26 is left open during the following simulation.

10.2.3 Ultra-capacitors combined with batteries Capacitors must be combined with electrochemical storage systems in most vehicle applications to realize maximum benefit. For instance, an ultra-capacitor may be used independently as an energy recovery and delivery device for capturing vehicle kinetic energy in an idle–stop situation. Some of the energy may be used during vehicle launch and acceleration, but in general such applications require suboptimal capacitor mass since the traction inverter is rated over a relatively narrow voltage window of 2:1 or less. Capacitors, in combination with batteries, are the most common architectures and there are two possible connections: (1) parallel battery and capacitor (read this as ultra-capacitor or super-capacitor/electrochemical capacitor, a tandem connection) or (2) capacitor with independent power processor, active parallel. These two cases are illustrated in Figure 10.27. Vd +0.16%, –22%

Ub mb gEb gPb

Vd +0.16%, –22%

Uc

Ub mc gEc gPc

(a)

mb gEb gPb

Uc

mc gEc gPc

(b)

Figure 10.27 Battery–capacitor combinations: (a) direct parallel, (b) independent power converter architectures of active parallel type For the analysis of what is the optimal sizing of battery and capacitor combinations, the following assumptions apply: M stor ¼ mb þ mc 0:78 pu  V dnom  1:16 pu Ppk ¼ 40 kW

ð10:32Þ

gFC ¼ 0:30 kWh=mi Furthermore, it will be assumed that the mid-sized sedan under consideration will have its energy storage mass restricted to 75 kg for the specific fuel consumption noted in (10.32). The nominal system voltage will be taken as Udnom ¼ 550 V, the open circuit voltage of a 440 cell NiMH battery pack. The vehicle has attributes listed in Table 10.5, including performance targets.

474

Propulsion systems for hybrid vehicles

Table 10.5 Vehicle attributes for battery–capacitor combination study Attributes

Performance targets Vehicle mass Drag coefficient Frontal area Rolling resistance Wheel radius Glider mass Accessory power

mv Cd Af R0 rr mglider Pacc

1,610 kg 0.327 2.17 m2 0.008 kg/kg 0.313 m 1,053 kg 500 W

0–60 mph accel. 50–70 mph passing Max speed 50 mph for 15 min E-only range, H0 E-only range, H20 E-only range, H60

rx rp

tc

Cpr A fp ~ 10 mHz

rs

Lpr

Lsr Csr RL Us

Figure 12.7 Near-field power transfer with intervening resonator coils The wireless power transfer, augmented with slightly detuned primary and secondary side resonators, that can also have angular displacement relative to their adjacent coil are shown by Kurs et al. to be capable of delivering 60 W to the load when 400 W is pulled from the wall plug. Given a 40% efficient wireless mode over D ¼ 2.4 m (eight times the secondary coil radius), the total transmission efficiency is 15%. For closer spacing, the efficiency will be higher. McKeever et al. [16] expanded on the work of previous investigators, including Kurs, to analyse, model, simulate and prototype a large evanescent wave power transfer structure similar to Figure 12.7 but with large rectangular copper coils separated by 17 cm without the intervening detuned resonators. The rectangular coil geometry is meant to facilitate a variable overlap, such as when a vehicle mounted coil moves over a road bed, or garage floor, mounted transmitting coil. The prototype system exhibited 82% power transmission efficiency, exclusive of the primary side amplifier at 58 kHz and secondary side power conversion stage. More recently, Keeling et al. [17] described a guideway track primary conductor of hairpin design that is excited via a utility side converter at 10–50 kHz and 300–400 A of primary current. Ferrite core secondary side inductors are placed along the primary cable and designed for 300 V operation at 1 kW. The compensating system demonstrated that power could be corrected over a 10% to þ15% variation in compensating capacitor.

Automated electrified transportation

553

In 2009 Yu et al. [18] described a wireless power transfer apparatus similar to Kurs, but without compensator resonators placed 5–40 cm apart operating at 760 kHz. At 40 cm the unit transmitted over 18 W power to a lamp at a voltage of 102 Vrms. This coil spacing is sufficient for transmitting power to a secondary coil in the vehicle floor pan while parked in a garage or moving along a roadbed. During the near term, there will be further improvements to non-contacting power transfer as a convenient means to charge PHEVs and BEVs. This would indeed add to the consumer acceptance of such vehicles by taking out the requirement that users remember to plug-in. With increasing development of vehicle to charger via Zigbee (IEEE 802.15.4-2003 protocol) and the ability to communicate time of day rates and allowable power transfer levels, this would mean the user simply park over the primary coil so that sufficient overlap exists to transfer peak power. Such systems would essentially conform to smart grid sensor and control paradigms for active load control. In addition, wireless power transmission eliminates many of the safety concerns of cable and connector systems such as those covered by the US National Electrical Code – 2008, Article 625, Electric Vehicle Charging System Equipment.

12.4 Transporting cargo Potentially one of the most promising near-term applications of AHS is the transportation of cargo. A good way to start this discussion on AHS transportation of freight is to cite the words of a meeting of the Texas Department of Transportation [19]: Freight is an immensely important component of our thinking. In the next reauthorization, freight will be dominant in the Congress. People say that freight doesn’t vote, but it has immediate, immense, and financial, job, and cost implications. And people are beginning to recognize this. It makes the point very real that transportation policy and planning is not a parlor game; it’s a very serious thing that we have to address. One question I always ask is: how many ton-miles are in your breakfast? If you look a pineapple, it’s probably two ton-miles to get that here. If you think of what we bring to our homes, the number of ton-miles per capita is an extraordinary statistic. Those comments coming from a policy-making component of state government should be a real awakening for us. To paraphrase another of their comments is something to the effect that today we are making things worse a little slower. This is simply not sustainable and with AHS, which is now evolving to automated electrified transportation (AET), is taking on more importance to move cargo and people. Moving cargo as the first application of AHS/AET is a prudent approach to introduce this technology as it would prove out and validate the claims of economics and safety. Economides and Longbottom [20] have the following to say about AET. A bold solution to these challenges may exist in the form of a dual mode electrified transportation network. This concept merits a critical mass research effort to evaluate the cost/benefit balance, identify and address the technology challenges, analyze the transition pathways to the alternative

554

Propulsion systems for hybrid vehicles architecture, and ascertain the policies and coalition support mechanisms that would enable the vision to become reality. A proposed new national network uses single and dual mode vehicles to provide mobility for freight, private cars, and mass transit vehicles. In single mode, the vehicles will be captive to an electric guideway from which they draw propulsion energy in real time as the vehicle moves at high speed under automated computer control. Single mode applications could include fully automated (driverless) terminal-to-terminal freight transport and PRT. In dual mode operation, driver-controlled vehicles will be able to travel the first and last miles offguideway using onboard energy storage as one mode and then enter the guideway in a second mode for high speed automated travel.

Their report goes on to outline the need to engage the private sector, leverage political support and give a positive cost/benefit analysis to then develop design and operations standards to ensure interconnection between inter-city systems justified by freight and intra-city systems justified by transit that promotes adoption of dual mode private vehicle access. The authors in Reference 20 address the issue of congestion and highway network cost by making a comparison of average vehicle-per-hour (VPH) capacity per $1,000 of highway infrastructure investment. In Table 12.3 [20] the dual mode vehicle is compared to conventional vehicle and highway and also to conventional vehicle and highway but with vehicles equipped with cooperative adaptive cruise control (CACC). Table 12.3 System cost of dual mode with convention vehicle and highway System

Speed (mph)

Capacity (VPH)

Cost ($1M/lane-mile)

Capacity/$invest (VPH/$1,000)

Average dual mode Conventional highway Conventional highway with 100% CACC

158 70 70

13,349 2,174 4,550

10.3 5 6

2,360 435 758

Table 12.3 clearly shows that dual mode has a capacity factor (VPH/$1,000 infrastructure) of 2,360 versus only 435 for today’s vehicles and highways. By adopting CACC the conventional highway and vehicle metric increases a good 74%, but is still a long way from the dual mode metric. This analysis demonstrates that moving to an AET infrastructure makes economic sense, far more so than adding more lanes to our existing highway network. Starting AET off with freight movement will therefore cut the cost of goods delivery, the ton-mile metric noted earlier. In closing, a list of dual mode and related PRT, AHS and AET activities worldwide, as compiled by Mr Jerry Schneider,1 University of Washington, USA, at [email protected], is provided as a quick reference in Table 12.4 for those interested in learning more. 1

Visit URL: http://faculty.washington.edu/jbs/itrans/ for an exhaustive listing of PRT, AHS and dual mode systems.

Location

USA, TX

USA, TX Netherlands France

France UK Australia Germany USA, VA New Zealand USA USA, CA UK

Korea

USA, MI

USA, CA USA, TX South Africa Netherlands

System name

Aerobus

Aerorail Aerorider Aerotrain

Air Car Atmostrack Austrans Autoshuttle Autoway ATN Autran AVT-Train Blade Runner

BT

Cabintaxi

Car Bus CargoRail Capsi City Mobility

Guideway

C&C software

Test program

Cost target

Active marketing?*

Operating?**

(Continues)

Currently for sale, was operational several years ago in Europe and Canada. System currently being constructed in China Conceptual only, prototype being developed Three-wheel bicycle with 1-passenger enclosure, for commuting History of efforts to develop an air-cushion, high-speed, jet-propelled train in the 1970s; illustrated and includes English version Small auto that runs on compressed air, availability pending Conceptual only – would use compressed air for propulsion M/H H M H M H M M M L L M/MH M M PRT concept – development funding being sought Automated transportation network – a PRT/dual mode concept M M M N L L N Conceptual only – high-speed train that carries autos and people Uses vehicles with rubber tyres and retractable steel wheels for dual mode capabilities, cargo and passenger modes High capacity dual mode concept under development, website features excellent animations of system in operation, many application illustrations Extensive test facility and development program completed in Germany in 1979. Shuttle system in operation since 1976. US company pursuing private sector applications Conceptual only, other versions called autobus and Personal Mass Transit Cargo carrier version of MegaRail, under development Conceptual only, PRT approach uses small vehicles in a tube Conceptual only

Vehicle

Status of design eng. and testing

Table 12.4 Reference list of AHS systems (http://faculty.washington.edu/jbs/itrans/)

Test program

Operating?**

H

USA, WA

Higherway

N

N

H

N

L

N

M

L

VL

L

M

M – ops. prototype available N

Minibus that can be operated on conventional rail as well as roadways remains in experimental and pre-production phase A dynamic carpool service now in operation in Amsterdam L/M L/M L/M L VL-MH M/H L/M

USA

Finland USA, FL

Japan

M M M L VL M L Conceptual only, small supported vehicle L L L N VL M L One of FTA’s Urban Maglev contractors, prototype and test track operational, demonstration planned at California University in Pennsylvania

Germany Austria USA Netherlands Finland USA, CA, NY UK

C&C software

Active marketing?*

Conceptual only, designed to provide a shuttle service for major activity centres and other high density areas in the city Conceptual only, small electric cars on automated guideway, dual mode possible eventually Test track and vehicles developed and undergoing testing, both alpine and urban versions being developed L L L L L L L H H H H L H H Concept only, as described in 1996 book about the future H H M/H M/H L M/H M/H Suspended monorail, final stages of design, prototype to follow, patents pending

Guideway

Cost target

New Zealand USA, OK Sweden USA, CA

USA, GA

City Shuttle

Vehicle

Status of design eng. and testing

CompuCar Coaster CULOR CyberCab CyberCab CyberTran Dragonfly MonoMetro Dual mode Vehicle Easy-Rider Evac. Tube Transport FlexiTrain Flash Flyway General Atomics, Urban Maglev Gimbal Craft

Location

(Continued)

System name

Table 12.4

Maglev concept under development, includes palletized dual mode, cargo, solar driven hydrogen production and municipal utility capabilities Dual mode concept, extensive documentation and video available Small, suspended, computer controlled vehicles, development underway, very good slide show at their website Prototype 6-passenger vehicle being developed, uses permanent magnets for suspension and linear motors for propulsion, test track being constructed near Port Angeles, Washington

UK, London

USA, MI

Mezzanine Transit

LEVX Maglev by Magnaforce MAGLEV 2000 Magnemotion, Urban Maglev Magnetrans Magplane pipeline Magplane passenger Magtube, Inc. MegaRail

An area-wide, dial-a-ride concept, using advanced communications technologies – called Taxibus – utilizes relatively small vehicles

USA, IL USA, OR

Hytran Individual Mass Transit Intelligent Grouping Transport Interstate Traveler InTransSys Jpods

USA, TX

MH

H

H

Conceptual only, being developed in Houston, TX, unique horizontal switching method

L

(Continues)

L/M, video available

N

M H

Evacuated tube concept using Maglev, focused on moving freight M M/H L L VL

M

H M

USA, CA USA, TX

M

M VL

M

M H

USA, MA

H H

H H

USA, CA USA, MA

M H

M M/H L L L/M L L One of FTA’s Urban Maglev contractors, working towards demonstration project planned for near future, prototypes operating

USA, FL USA, MA

USA, WA

USA, CO USA, MN

L H H N MH H N National dual mode system with high capacity synchronous Maglev guideways (conceptual only) For sale, extensive test and demo program continuing, first application in Japan is almost completed. Has been studied in the US under FTA Urban Maglev contracts. First public service scheduled on Linimo line for March 2005 Suspended monorail – conceptual, seeking development funding A 3-tier dual mode concept, conceptual only

USA, GA USA, WA Japan

HighRoad HiLoMag HSST Maglev Linimo line

Parry People Mover Pathfinder POSTECH PRT Project PRISM PRT 2000 Personal Transportation System (PTS)

USA, MI USA, MA USA, CA

USA, MI Korea

UK

Two systems in full operation (Amsterdam Airport since late 1997 and Rotterdam since early 1999); more systems being deployed Features trams of several sizes, powered by a flywheel, no overhead wires. First application now (2003) underway in the UK M L L L M M L Research and development project at Pohang University, includes 40 m test track and operational vehicle, Phase II program completed 2003 Dual mode concept being developed at Ford Research Laboratory Development and test program completed in 2000, remains inactive since 2003 A dual mode concept that features small dual mode vehicles that can also be operated on conventional city streets

Netherlands

L

M

L

VL/L

M

M

N – good documents

Prototype early 2002 N – but ops. video available M

L

L

H

European

H

L

VL

M M L M VL M Developing novel Maglev system, some prototype components currently operational

M

L

Test program

USA, OH USA, CA

H

H

C&C software

Monomobile Modern Transport System Corporation Modular Automated Individual Transport Park Shuttle

M

M

Guideway

USA, OR

Operating?**

Mitchell

Active marketing?*

USA, TX

Cost target

MicroRail

Vehicle

Status of design eng. and testing

Location

(Continued)

System name

Table 12.4

USA USA, MN USA, CO USA, MD USA, MD Switzerland USA, SC UK USA, TX USA, NJ

SkyTran Skyweb Express SmartSkyways SMRTram Surrey System SwissMetro System 21 Taxibus IGT TriTrack TubeXpress

USA Switzerland USA

New Zealand Russia USA, FL

(Continues)

Conceptual only, two-way travel on one monobeam Conceptual, small rail vehicles on interesting elevated guideway, in-vehicle switches planned Redesigned to incorporate fully proven light rail MH M/H N components Conceptual only – high speed, small vehicle, low cost, Maglev H H H M L H M L L L N L L L Conceptual – electric, large-vehicle, two-way travel with one lane Conceptual – small, automated dual mode vehicles operating in a tube Maglev vehicle in a tube, considerable research has been conducted M H M L MH H M/L Minibuses operated so as to intelligently group passengers using modern telecommunications technology H L N L VL L L Prototypes built and tested, produce market-ready

A dual mode concept that utilizes a modification of existing freeways Modular automated railway system that combines a sophisticated undercarriage with the advantages of Maglev; using existing railways – for both people and cargo Conceptual only, moving beltway with passive vehicles Small automated vehicles on exclusive guideway, prototype vehicle constructed and being tested A very large bus concept M/H H/M L H M M H Conceptual only, tube transport concept System for helping pedestrians get across heavily travelled roadways and other barriers to pedestrian movement Conceptual only – a palletized dual mode concept H H H M VL M M Conceptual only – specially designed facilities for serious bike transport

USA, WA Germany

USA, CA USA, MD USA/UK Denmark Germany Germany

Maglev, dual mode concept, developed at VPI, scale model constructed, video available (website down as of 13 January 2005 and not planned to be put back in service)

USA, VA

Rideway Robocab Roadrunner RUF RUMBA Schmid Peoplemover Segway Serpentine Skybikes, Bike Trains SkyCabs SkyTaxi Sky Train

Personal Electric Rapid Transit System (PERTS) Puget Pullway RailCab

Germany USA, TX

UK Russia USA, WA USA, WA Germany UK UK China

TubeWay Tubular Rail

ULTra Unitran Urbanaut VMTS Velotaxi Whoosh York PRT Zhonghua-06

Guideway

C&C software

For vehicles, hardware and software, testing: H: Highly developed, fully built, being tested, or ready for testing M: Partially developed, some components/reduced scale prototype built and tested L: Still mostly on paper, some engineering studies completed N: All on paper or elsewhere

For cost: H: More than $30 million/mile ($18.75/km) MH: $20–30 million/mile ($12.5–$18.75/km) M: $10–20 million/mile ($6.25–12.5/km) L: $5–10 million/mile ($3.125–6.25/km) VL: Less than $5 million/mile ($3.125/km)

Test program

Cost target Active marketing?*

Operating?**

Concept only, uses air pressure and capsules for passenger and cargo A unique, monorail-type system that does not require a conventional guideway, for high-speed mass transit applications H H H H L H M Test facility constructed, testing under way. See website for details M M/H M/H MH H M M Conceptual only – uses large truck to haul small electric vehicles on freeways For 2 people, muscle-powered with electrical assist, available now Conceptual only – monorail that uses compressed air for propulsion Appears to be a reincarnation of Raytheon’s PRT 2000 technology Suspended, lightweight, Maglev concept, initial testing under way

Vehicle

Status of design eng. and testing

* Indicates whether or not the systems described have activity to promote in marketplace. **Indicates whether or not the system is actually ready for being put into service.

Location

(Continued)

System name

Table 12.4

Automated electrified transportation

561

12.5 Exercises Q1:

A1: Q2:

A2: Q3:

A3:

Q4:

Repeat the analysis given in (12.2) and (12.3) for the case of methanol fuel, CH3OH, in air. Given that the HHV for methanol is 22.7 kJ/g, calculate the energy release for this exothermic reaction. 1,452.8 kJ Calculate the vehicle propulsion power that must be supplied by a CWT to a PRT size vehicle at steady state cruise on a guideway for the case of intracity speed of 95 mph and inter-city speed of 150 mph. Use the same Excel file developed for Q7 in Chapter 7, but with the following modifications: Mv ¼ 600 þ 2(75.5) ¼ 751 kg, Cd ¼ 0.22, Af ¼ 2.3 and all remaining parameters the same. P(V ¼ 95) ¼ 26.14 kW and P(V ¼ 150) ¼ 97 kW For the CWT secondary voltages calculated in Example 1, make the following calculations based on a CWT power loss of 3% and the PRT guideway power levels calculated in Q2: a. Find the secondary current when f ¼ 15 kHz and f ¼ 50 kHz given P(V) ¼ 26 kW. b. Find the secondary current when f ¼ 15 kHz and f ¼ 50 kHz given P(V) ¼ 97 kW. a. 26 kW case: Is( f ¼ 15 kHz) ¼ 179.3 Arms and Is( f ¼ 50 kHz) ¼ 53.8 Arms b. 97 kW case: Is( f ¼ 15 kHz) ¼ 668.9 Arms and Is( f ¼ 50 kHz) ¼ 200.7 Arms Cite some societal benefits of widespread automated highway system (AHS) for intra-city and inter-city commuting and freight transport.

References 1. National Hydrogen Association. The Energy Evolution: An Analysis of Alternative Vehicles and Fuels to 2100, National Hydrogen Association, 2009. Available from www.hydrogenassociation.org. 2. Fahimi B. ‘On the efficiency of the fuel cell vehicle with on-board hydrogen generation’. IEEE Power Electronics Society, PELS Newsletter, 2009, vol. 21, no. 1, pp. 14–17. 3. Stephan C.H., Sullivan J. ‘Environmental and energy implications of plug-in hybrid-electric vehicles’. Environmental Science and Technology, 2006. 4. Anderson J.E. ‘A Review of the State of the Art of Personal Rapid Transit.’ Journal of Advanced Transportation, 2000, vol. 34, pp. 3–29. 5. Anderson J.E. ‘How innovation can make transit self-supporting’. Presented at the Conference of Georgist Organizations, Chicago, IL, 19–23 July 2006.

562 6.

Propulsion systems for hybrid vehicles

Thornton R., Clark T., Bottasso M. Maglev PRT, MagneMotion, A Force in Electromagnetic Systems, www.magnemotion.com, 2007. 7. Lowson M. Engineering the ULTra System, University of Bristol and Advanced Transportation Systems Ltd. 8. Kornbluth K., Burke A., Wardle G., Nickell N. ‘Design of a freeway-capable narrow lane vehicle’. Presented at the SAE World Congress, Society of Automotive Engineers, SAE 2004-01-0760, 8–11 March 2004. 9. Stephan C.H., Miller J.M., Pacheco J., Davis L.C. ‘A program for individual sustainable mobility’. Presented at the Global Powertrain Congress, GPC2003, Ann Arbor, MI, 23–25 September 2003. 10. Stephan C.H., Miller J.M., Davis L.C. ‘A program for individual sustainable mobility’. International Journal of Vehicle Autonomous Systems, 2004, vol. 2, pp. 255–77. 11. Stefan C., Miller J.M. ‘A program for individual sustainable mobility, frontiers in transportation: Social interactions’. Amsterdam, the Netherlands, 14–16 October 2007. 12. Stephan C.H., Miller J.M., Davis L.C., Anderson R.D. System and Method for Controlling Flow of Vehicles, US Patent 6,637,343B2, 28 October 2003. 13. Klontz K.W., Divan D.M., Novotny D.W., Lorenz R.D. ‘Contactless power delivery for mining applications’. Presented at the IEEE Industry Applications Society Annual Meeting, Dearborn, MI, 28 September– 4 October 1991. 14. Stephan C.H., Miller J.M., Davis L.C., Anderson R.D. Achieving and Maintaining Desired Speed on a Guideway System, US Patent 6,619,212B1, 16 September 2003. 15. Kurs A., Karalis A., Moffatt R.M., Joannopoulos J.D., Fisher P., Soljacic M. ‘Wireless power transfer via strongly coupled magnetic resonances’. Science, 2007, vol. 317, p. 83. 16. McKeever J.W., Scudiere M.B, Bigelow T.S., Caughman J.B., Ryan P.M. ‘Evanescent power transfer for electric vehicles’. Oak Ridge National Laboratory, Energy and Transportation Science and Fusion Energy Division, March 2009. 17. Keeling N., Covic G.A., Hao F., George L., Boys J.T. ‘Variable tuning in LCL compensated contactless power transfer pickups’. IEEE 1st Energy Conversion Congress and Exposition, ECCE2009, San Jose, CA, 20–24 September 2009. 18. Yu C., Lu R., Mao Y., Ren L., Zhu C. ‘Research on the model of magneticresonance based wireless energy transfer system’. IEEE 5th Vehicle Power and Propulsion Conference, VPPC2009, Dearborn, MI, 7–10 September 2009. 19. Dual mode systems for cargo can be found by consulting: www.txdot.gov for the Horizon: The Future of Transportation, a publication of the Texas Department of Transportation, 2007. Also, consult the company INNOV8 Transport, Inc. 20. Economides C.E., Longbottom J. ‘Dual mode vehicle and infrastructures analysis’. Report 0-5827-1, Texas A&M University, April 2008.

Appendix A

The following tables are provided in support of the materials presented in this book and can be obtained from the USABC website (www.uscar.org) and US Department of Energy website (www.eere.energy.gov/) for public use.

Table A1 USABC requirements of end of life storage systems for PHEVs Characteristics at EOL

Units

High power/energy ratio battery

High energy/power ratio battery

Equivalent electric-only range Peak pulse discharge power (10 s) Peak regen pulse power (10 s) Available energy for CD, charge depleting, mode, 10 kW rate Available energy for CS, charge sustaining, mode Minimum round trip energy efficiency (USABC HEV cycle) Cold cranking power at 30  C, 2 s, 3 pulses CD life/discharge throughput CS HEV cycle life, 50 Wh profile Calendar life, 40  C Maximum system weight Maximum system volume Maximum operating voltage Minimum operating voltage Maximum self-discharge System recharge rate at 30  C Unassisted operating and charging temperature range Survival temperature range Maximum current (10 s) pulse Maximum system production price at 100,000 APV

miles kW

10 45

40 38

kW kWh

30 3.4

25 11.6

kWh

0.5

0.3

%

90

90

kW

7

7

Cycles/MWh No. of cycles Years kg L Vdc Vdc Wh/day kW  C

5,000/17 300,000 15 60 40 400 >0.55  Umx 50 1.4 (120 V/15 A) 30 to þ52

5,000/58 300,000 15 120 80 400 >0.55  Umx 50 1.4 (120 V/15 A) 30 to þ52

 C A $

46 to þ66 300 1,700

46 to þ66 300 3,400

564

Propulsion systems for hybrid vehicles

Table A2 Theoretical and practical properties of batteries System

Negative Positive electrode electrode (mat’l) (mat’l)

Open Theoretical Theoretical Practical circuit capacity energy energy voltage (V) (Ah/kg) (Wh/kg) (Wh/kg)

Lead–acid NiCd NiMH NaS (300  C) NaMCl2 (300  C) Li ion

Pb Cd MH alloy Na

PbO2 NiO–OH NiO–OH S

2.1 1.35 1.35 2.1–>1.78

83 162 178 377

171 219 240 754

20–40 40–60 60–80 120–150

Na

NiCl2

2.58

305

787

80–100

LixC6

Li1 xMO2 4.2–>3.0 M = Co, Ni, Mn 3.3–>2.0 VOx

380

150–200

884

150

Li polymer Li

95 x = 0.6 340

Table A3 Estimates of storage system capital costs Battery technology

Current cost ($/kWh), 2010

Projected cost at 10 year ($/kWh)

Flooded lead–acid (PbA) Valve regulated lead–acid (VRLA) Nickel cadmium (NiCd) Nickel metal hydride (NiMH) Lithium ion (Li ion) large format Zinc–air (ZnAir) Sodium sulphur (NaS, 300  C) Sodium/nickel chloride (NiCl2) Vanadium Redox at 100 kWh Zinc bromine (ZnBr) Pseudo-capacitor (Pb-C) Flywheel – low speed Flywheel – high speed

150 200 600 800 1,300 150 450 800 600 500 500 380 1,000

150 200 600 350 780 100 350 150 500 250 250 300 800

Table A4 Power electronic inverter targets for HEV, PHEV, BEV (DOE) Specification

Units

2010

2015

2020

Specific power (peak) Power density (peak) Cost

(kW/kg) (kW/L) ($/pk kW)

>10.8 >8.7 7.9

>12 >12 5

>14 >13 3.3

Appendix A

565

Table A5 Electric machine targets for HEV, PHEV, BEV (DOE) Specification

Units

2010

2015

2020

Specific power (peak) Power density (peak) Cost

(kW/kg) (kW/L) ($/pk kW)

>1.2 >3.7 11

>1.3 >5 7

>1.6 >5.7 4.7

For industrial electric machines, prompted by the interest in renewable energy, more energy efficient buildings, products and vehicles in line with objectives of the US Energy Independence and Security Act 2007, a higher level of efficiency is required for 2011 and beyond. Electric machines manufactured after December 2010 will need to comply with the efficiency levels of the energy independence and security act, EISA. In Reference 1, the author states that the average industrial motor consumes electricity worth 50–60 times its initial purchase price over a 10 year lifetime. This means that capital cost is only 2–3% of its lifecycle cost, mainly electricity consumption. Prior to the EISA ruling, an industrial induction motor having an efficiency as targeted by the EPAct is now reclassified as a new subtype I of electric machine, which requires these general purpose motors to now meet the efficiency levels of previous NEMA premium efficiency electric machines. Electric machines for hybrid vehicle application by the very nature of their use demand very high efficiency. In Reference 2, Savagian describes the new electric machine to be used in hybrid electric vehicles and battery electric vehicles that employs copper bar wound stator to realize 30% lower resistance. This electric machine in a BEV drive cycle exhibits 20  C lower operating temperature than a conventional strand wound electric machine.

References 1. Butler K. What Motor Users Need to Know, published in Intech. Available from www.isa.org, November 2009. 2. Berry D., Hawkins S., Savagian P. Motors for Automotive Electrification, SAE Hybrid Vehicles Technology Symposium, San Diego, CA, 10–11 February 2010.

Index

Aachen model 509–10 Aachen University of Technology 409 ABB bimode IGT (BIGT) 72 ABS: see anti-lock braking system (ABS) ac 169 brushless 261–5 discrete speed control methods for, hierarchy of 295 ac drives 10, 100, 167, 174 battery electric 315–16 and diesel locomotive drives 74 dual mode with 85 ISA type 119 power electronics for 325–63 AC Propulsion 56–7 ac switches, thyristors 275–6 acceleration 124–5, 127 brisk 128 dynamics of 139 full/wide open throttle 127 high capability 268 maximum time 474 smooth 128 straight and level 419 test 525, 534 acceptance rate 443 accumulators 229, 499 acetonitrile 461, 464 Ackerman angle 30–1 activated carbon (AC) 462–4 activation polarization 443–4 actual fuel economy 41 see also fuel economy actuator control unit (ACU) 229

actuators electric 499 hydraulic 499 ACU: see actuator control unit (ACU) adjustable speed drive (ASD) 497 ADVISOR (mathlab program) 422 aerodynamic drag 419 aerodynamic drag coefficient 125, 523, 525, 532, 537 closely spaced vehicles and 531 drag semi-tractor–trailers optimized for 532–3 Escape SUV for 527 for full rig 536 road load components and 526–7 SUV class vehicle and 529 SUV with trailer and 530–1 aerodynamic drag force 36 aerodynamic loading 550 aerodynamics 549 aerospace 270, 356 AET: see automated electrified transportation (AET) AFMA: see Alternative Motor Fuels Act (AFMA) AFPM: see axial flux permanent magnet (AFPM) AFV: see alternative fuel vehicle (AFV) AHS: see automated highway systems (AHS) air conditioning 127, 277, 538 air cooling 454

568

Propulsion systems for hybrid vehicles

air density 524, 533 Airbus 320, 105 airgaps 270–1, 273, 283, 298–9, 305 sinusoidal 369 Aisen Warner Navimatic transmission 142 alcohol 496 alloys 293, 449, 498 Alnico 244–5, 247–8, 305 alternative fuel vehicle (AFV) 496 Alternative Motor Fuels Act (AFMA) 496 alternators current ripple 383 Lundel 412–13 aluminium 293, 368, 397 American Superconductor Corporation 253 amorphous metal 248 Amperean force 410 Ampere’s circuital law 249 amplitude 331 Anderson, J. Edward 546 angstroms 462 ANL: see Argonne National Laboratory (ANL) anodes 442, 455–6, 458, 505 chemistry of 454 anti-lock braking system (ABS) 129, 222, 229 RBS interaction with 133–4 antimony 447 Apollo mission 546 APTz insulation systems 404 APU: see auxiliary power units (APU) architectures 414, 472–3, 481 automatic transmission 151–2 brushless dc motor 259 fuel cell 278 power split 379 Argonne National Laboratory (ANL) 420, 422, 452 armature current 367–8, 379 armature windings 405–6

aromatic polyamide 498 ASD: see adjustable speed drive (ASD) Asia-Pacific 368, 420, 432 see also Japan NEDC in 135 see also Japan asymmetrical ultra-capacitors 470–3 asynchronous machines 367 design 266 ATDS cycle 25–6 Atkinson cycle 140 atmospheric conditions 524 Austenite phase 244 automated electrified transportation (AET) 553 automated highway systems (AHS) 545, 547–9, 555–60 example of 548 ingress and egress of PRISM vehicles and 549 automatic transmission 16, 119, 125, 128, 142, 527 architectures 151–2 Lepelletier type 143, 154–5 Simpson type 152–4 types of 153 Wilson type 143, 154 automotive reliability 357 autonomy 100–1 auxiliary power units (APU) 538 available energy model, storage system 501–2 average speed 135–6 on fuel economy 138–9 axial flux permanent magnet (AFPM) 288 axial laminated rotor 313 back-emf sensing 259–60, 272 balance of the plant (BOP) 506 balanced delay error 360 balanced offset 360 Ballard Power Systems 11 band pass filtering (BPF) 383 bandgap materials 356

Index bandwidth 348, 361 barium 244, 278 Barkhausen, Heinrich 396 BAS: see belt alternator starter (BAS) batteries conductance of 500–1 defined 441 ultra-capacitors with 473–81 battery electric vehicle (BEV) 12, 135, 157–64, 315–16, 372, 447 energy flow diagram for 16–17 EPRI survey 59 battery immittance method 501 battery management system (BMS) 452 battery model, storage system 499– 504 battery systems 441–61 capacitor systems 461–93 see also capacitor systems lead-acid 447–9 lithium ion 454–61 market for 441 nickel-metal hydride 449–54 thermodynamics of 443 BDC: see bottom dead centre (BDC) BDCM: see brushless dc motor (BDCM) ‘beaming’ 28 Beijing Bus Corp. 113 Beijing Institute of Technology (BIT) 113 Beijing Olympic buses 113 Bell, F.W. 361 belt alternator starter (BAS) 80 BEV: see battery electric vehicle (BEV) BIGT: see ABB bimode IGT (BIGT); bimode insulated gate transistor (BIGT) bimetallic effects 359 bimode insulated gate transistor (BIGT) 328 bipolar diode 406 bipolar junction transistor (BJT) 408 BIT: see Beijing Institute of Technology (BIT)

569

BJT: see bipolar junction transistor (BJT) Bloxham, Steven 105 BMEP: see brake mean effective pressure (BMEP) BMS: see battery management system (BMS) Boltzmann constant 406 boosting, launch and 128–9 BOP: see balance of the plant (BOP) bottom dead centre (BDC) 46–7 boundary conditions 293 BPF: see band pass filtering (BPF) ‘brake’, defined 48 brake mean effective pressure (BMEP) 43–4, 47–9 BSFC sensitivity to 49–51 brake specific air consumption (BSAC) 49 brake specific fuel consumption (BSFC) 42–3, 495 sensitivity to BMEP 49–51 braking 452, 466, 498 dynamic effects of 30 energy recuperation and 129–35 braking systems 229–30 brushed machine 255, 257, 368 brushless dc motor (BDCM) 496–7 brushless machines 252–77 ac 261–5 dc 256–61, 326, 367, 384 BSAC: see brake specific air consumption (BSAC) BSFC: see brake specific fuel consumption (BSFC) buck–boost converter 482, 485–6 advantage of 486 ultra-capacitors and 486 Buller, S. 509, 511 buried magnet design 266, 278–82, 287 bus bars 212–15 butane 496 BYD Auto 59 bypass cavity 287

570

Propulsion systems for hybrid vehicles

cabin climate control 230, 538 cable 210–12 CACC: see cooperative adaptive cruise control (CACC) calcium 447 California Air Resources Board (CARB) 57–8 Camaro SS 119–20 specifications of 120 Camry Hybrid vehicle 131 V_Dot for 132 CAN: see controller area networks (CAN) capacitor systems 461–93 flywheel systems 496–9 hydrogen storage 493–6 pneumatic systems 499 storage system modelling 499–516 ultra-capacitors 466–93 capacity point 502 CARB: see California Air Resources Board (CARB) carbon 454–5, 461–2, 508, 514 activated 462–4 porous 462, 466, 506 carbon dioxide 421, 493, 496 cargo, transporting 553–4 Carter coefficient 272 catenary powered vehicles with ultra-capacitors 110–11 with wayside ultra-capacitors 111– 13 cathode ray tube (CRT) 252 cathodes 252, 442, 455, 505 chemistry of 454 Cauer I network 512–3, 515 C/D: see charge/discharge (C/D) power Celerotron 248 cell balancing, ultra-capacitors 482–93 dissipative cell equalization 484–5 electrochemical double layer capacitor 488–93 non-dissipative cell equalization 485–8 centre of gravity, of vehicle 29

Ceramic magnet 245, 257, 278 CFD: see computational fluid dynamics (CFD) characterization, device 409 characterization, hybrid vehicles 419–37 charge constant current 488 separation distances 461–4 stored 408 test 490 charge acceptance 502 charge depleting 62–4 charge sustaining (CS) and grid connected hybrids 62–4 plug-in hybrids and 60 charged voltage 502 charge/discharge (C/D) power 452, 455, 475 chassis 411 chassis systems 425 Chevrolet Equinox 10 Chevy 119 Chevy Malibu (hybrid) 10 Chevy Volt, General Motors Corporation 447 CIDI: see compression ignition direct injection (CIDI) engines city cycle 137, 427–8 CIV: see corona inception voltages (CIV) Civic Hybrid (Honda) 6 vs. Toyota 2002 Prius 7–8 class-8 tractor test 532–5 Clayton twin roller dynamometers 430 climate change, energy efficiency and 1 climate control system 450 CNG: see compressed natural gas (CNG) CO2 emissions annual, components of 15–16 calculations 42 as function of fuel heat rate 42

Index CO2 emissions, reduction of engine optimization actions and 12, 14 vehicle electrification actions and 12, 14 coast down procedure 523, 525–7, 538 section of road for 528–9 semi-tractor in 533 in tractor-only testing 534 coast down test procedure 419 coaxial winding transformer (CWT) 548, 550–1 coefficient of thermal expansion (CTE) 359 coercive force 270 coils 300 combined cycle 429–30 communications 216–27 CAN 219, 222–3 class A networks 226 class B networks 226 class C networks 226 DTC 227–8 FlexRay 223–6 OSI 7-layer model 217–18 power and data networks 220–1 protocols 219, 226–7 TTCAN 222–3 commutators 253, 255, 373 comparisons, machine technology dynamic performance 313–15 electric vehicles 315–16 for hybrid vehicles 316–18 comparisons, PWM techniques 348–9, 350 compressed natural gas (CNG) 419, 496 compression ignition direct injection (CIDI) engines 21–2, 21n5, 421, 423, 472 diesel fuelled, with SCR and DPF 58 torque curve versus speed for 43 computational fluid dynamics (CFD) 232

571

concentrated winding stator 248 concentration polarization 444–5 condition factors 502 Condor capacitor hybrid truck 10 conductance 500–1 conduction 258, 261, 406–7 diode 409 conductivity 397, 399 high bulk 258 thermal 353 conductors 252–4 resistances 405 temperature difference between 406 constant current 488, 491 constant offset 384 constant power speed range (CPSR) 78, 168–70, 268, 277, 279, 287, 296, 305, 318 wide 290 consumption conversions, fuel economy and 43–5, 46 Continental Group 523 continuous rating 167 continuously variable transmission (CVT) 3, 86, 142 see also electronic continuously variable transmission (eCVT) belt type 96–7 Reeves type 96 toroidal 97 Van Doorne 96–7 controller area networks (CAN) 100, 216–17 conventional vehicle (CV) 9, 125, 134, 139, 545 EPRI survey 59 vs. dual mode vehicle 554 vs. plug-in hybrids 60 cooling 351 torque and 268 cooperative adaptive cruise control (CACC) 554 copper losses 284 and skin effects 403–6

572

Propulsion systems for hybrid vehicles

core losses 396, 398–402 components of 397 net 397 corona discharge 404 corona inception voltages (CIV) 404 corporate fuel economy standards (CAFE´ ) 53 costs 234–5, 352, 355, 454, 499, 507 coupling 292, 406 covalent bonds, energy storage in 439 CPSR: see constant power speed range (CPSR) crankshaft mounted starter-alternators 538 crankshaft torque 123 creep 142 Crown, Toyota Motor Company 432 CRPWM: see current regulated pulsewidth-modulation (CRPWM) CRT: see cathode ray tube (CRT) cruise conditions, grade and 124–7 cryogenic process 493–4 crystal symmetry transformation 244 CS: see charge sustaining (CS) CSI: see current source inverter (CSI) CTE: see coefficient of thermal expansion (CTE) Curie temperature 244 current regulated pulse-widthmodulation (CRPWM) 331, 385, 388 current sensors 361 current source inverter (CSI) 347 CV: see conventional vehicle (CV) CVT: see continuously variable transmission (CVT) CVT hybrid powertrain (THS-C) 3 CWT: see coaxial winding transformer (CWT) cyanide gas 462 cylinder deactivation 119 Dahlander connection 294, 299, 301 DaimlerChrysler 67, 69 F400 69 V10 Viper engine 119

DaimlerChrysler ESX3 19 and engine downsizing 20 Dana 243 D’Arsonval meter 247 dc brushed 255, 257, 368 brushless 256–61, 326, 367, 384 link capacitor 361–2 dc/dc converters 349, 350–1 De Doncker, R. W. 351 Debye length, defined 464–5 decay 444–5 deceleration 525–6, 534, 536 depressed brake pedal and 130 dynamics of 139 deceleration fuel shut off (DFSO) 22 delay error 360 Department of Energy and the Office of Transportation Technologies (US) 503 depth of discharge (DOD) 455 Dewar containers 494, 496 DFSO: see deceleration fuel shut off (DFSO) diagnostic test codes (DTC) 227–8 diamond 356 dielectric strength 461 diesel 421, 423, 429, 539 diesel cycle operation 12 diffusion 445, 459, 503 direct torque control (DTC) 388–91 DIS: see discrete immittance spectroscopy (DIS) discharge 442, 445–9, 452, 455–7, 500, 503 depth of 455 efficiency 475, 477, 485, 490–1, 493 matched impedance 490, 493 self 446, 450, 456, 486, 493 discrete immittance spectroscopy (DIS) 501 displacement 255, 538 engines 140 dissipative cell equalization 484–5

Index Divan, D. M. 335 DMIC: see dual mode inverter control (DMIC) DOD: see depth of discharge (DOD) double layer capacitor, electrochemical specifications for 491–3 testing 488–91 downsizing 539 engine 140 DPF: see dual particulate filter (DPF) drag 534 see also aerodynamic drag coefficient pressure wave 532 roof equipment 532 skin friction 532 wheel 532 dragster class 123 drive cycle 424–5 characteristics 22–6 characteristics of 136–8 and fuel economy 22–3 implications 135–40 power and energy distribution in 24–5 real world customer 25–6 types of 135–6 vehicle braking events duration by 23 drive systems control 367–91 efficiency 395–416 electric: see electric drive system sizing 145–237 Drive-by-Wire protocol 227 DTC: see diagnostic test codes (DTC); direct torque control (DTC) dual mode inverter control (DMIC) 274–7 capability curves 275 dual mode pre-transmission architecture 85–6 dual mode vehicle vs. conventional vehicle 554 dual particulate filter (DPF) 57 DuPont 505–6

573

durability 277, 352, 383 dynamic index 29 dynamic performance 348–9 comparisons 313–15 dynamometers 430 E85 ethanol 420–1 E4 system, the 114 economical engines 123 eCVT: see electronic continuously variable transmission (eCVT) eddy currents 258, 291, 294 anomalous 397 classical 397, 399 magnetic domain-model of 396–7 Edison Electric Company 266 EDLC: see electronic double layer capacitor (EDLC) EDS: see electrical distribution system (EDS) EDTA: see Electric Drive Transportation Association (EDTA) EESTOR (US) 106 EESU: see electric energy storage unit (EESU) efficiency 420, 433 mapping 412–16 optimization 384–8 WTW 421–3 EGR: see exhaust gas recirculation (EGR) EHB: see electrohydraulic braking (EHB) systems EIA: see Energy Information Administration (EIA) EIS: see electrochemical impedance spectroscopy (EIS) electric actuators 499 electric drive system 412, 414 technologies 243–318 Electric Drive Transportation Association (EDTA) 11n2 statistics from 11–12

574

Propulsion systems for hybrid vehicles

electric drives battery-only 433 matching 147–55 electric energy storage unit (EESU) 440 ceramic ultra-capacitor 106–7 electric engine hybrids (2010) 11–12 electric four wheel drive architecture 113–15 E4 system 114 Estima Van 115 electric fraction (EF) classifications 140 engine downsizing 140 range and performance 140–1 electric loading 170 electric locomotive propulsion system 72 electric power 538 electric power assist steering (EPAS) 35, 426–7 voltage vs. rack loading 427 Electric Power Research Institute (EPRI) 59, 236 electric spin-up turbo charger 12 Electric Vehicle Symposium, 2003 3 electrical burden 141–2 electrical distribution system (EDS) 78 electrical overlay harness 208–16 electrification effort 538 vehicle: see vehicle electrification electrochemical cell 16 charge–potential behaviour of 444 life, end of 445–56 Peukert equation and 446 loss kinetics of 443–5 voltage development in 442–3 voltage-current behaviour of 445–6 electrochemical double layer capacitor 488–93 specifications for 491–3 testing 488–91 electrochemical impedance spectroscopy (EIS) 504

electrohydraulic braking (EHB) systems 129–30, 134–5, 208–9 ACU 229 components 229–30 HECU 229 electrolytes 441–2, 455 immobilization of 448 electromagnetic interference (EMI) 325, 348, 379 electromechanical braking (EMB) systems 129, 134–5 electromechanical transmission operating modes 87–8 operational map of 88 electromechanical valve actuation (EVA) 10 electronic continuously variable transmission (eCVT) 21, 527 architecture 86 power split: see power split architecture electronic double layer capacitor (EDLC) 461–4 electronic pole change 296–8 electronic stability programme (ESP) 70, 129 RBS with 134–5 electrons 252 EMB: see electromechanical braking (EMB) systems EMI: see electromagnetic interference (EMI) emission pollutant mass 42 emissions of CO2: see CO2 emissions limits, Euro 3 and Euro 4 58–9 minimal 420 testing 430 V2G 57–8 emissions tax 53 emitter turn-off thyristor (ETO) 176 EMS: see energy management strategies (EMS) Endura 1.8 L CIDI engine 43–4

Index energy consumption see also fuel consumption comparative analysis of 423–4 energy distribution 24–5 energy efficiency, climate change and 1 energy flow in BEV 16–17 in conventional vehicle 16 Energy Information Administration (EIA) findings 12 energy loss 425 energy management strategies (EMS) 453 energy recovery 141 energy recuperator system 79–80 and braking 129–35 energy storage 385 in covalent bonds 439 flywheels 208 fuel cells 199–204 lead-acid batteries 194–6 lithium ion 196–8 mass and 461, 466, 475–7, 481 mediums 439, 441 NiMH 196 in nuclear bonds 439 propulsion system power and 78–9 technologies 184–208, 439–519 ultra-capacitors 204–7 energy storage system (ESS) 325, 347, 439–41, 545 aspects of 440 thermal management of 454 energy use, upstream 420–1 engine downsizing, vehicles size and 19–22 engine optimization actions CO2 emissions reduction due to 12, 14 limits of 12, 14 engine peak power rating, of vehicle mass 20–1

575

engine plus electric power calculation of 20–1 engine power 126 calculation 42 engine speed 384 and vehicle speed, relationship between 37–8 enthalpy 445 entropy 445 Environmental Protection Agency (EPA) 22, 53, 135 city cycle 427–8 on fuel economy 53–4 EPA: see Environmental Protection Agency (EPA) EPA75 cycle 137 EPAS: see electric power assist steering (EPAS) epicyclic gear set, power split architecture 90 EPRI: see Electric Power Research Institute (EPRI) equivalent mass 28 calculation 33, 35 equivalent series resistance (ESR) 410, 488 ES3 experimental hybrid vehicle (Toyota) 75–6 Escape SUV (hybrid) 8–9 ESMA capacitors 471 ESP: see electronic stability programme (ESP) ESR: see equivalent series resistance (ESR) ESS: see energy storage system (ESS) Estima Van (Toyota) 115 ETH Zurich 248 ethanol 419–20, 423 ETO: see emitter turn-off thyristor (ETO) Euro 3 (CY2000) emission limits 58 Euro 4 (CY2005) emission limits 58 EVA: see electromechanical valve actuation (EVA) Evans Capacitor 470

576

Propulsion systems for hybrid vehicles

excitation 295, 298, 307, 372, 381 constant 368 frequency 369, 398 permanent magnetic 373 exhaust gas recirculation (EGR) 47 F400 69–70 failure in time (FIT) 357–8 FAKRA (Germany) 82 Faraday constant 443, 443n1 Faraday’s law 272 fault management 367 FCEV: see fuel cell electric vehicles (FCEV) FCHV: see hybridized fuel cell vehicle (FCHV) FCHV, Toyota Motor Company 506–7 FCT: see field controlled thyristor (FCT) FCV: see fuel cell vehicles (FCV) FCX 11 FCX-V3 67, 67n1, 68 Federal Test Procedure (FTP75) 135, 137, 427 Federal Urban Drive Cycle (FUDC) 24, 135 Federal Urban Drive Schedule (FUDS) 63, 425 feedback control 383, 390 ferrite magnets 244, 277, 284, 289 field controlled thyristor (FCT) 429 field oriented control (FOC) 367, 376, 388 condition of 371 dynamics of 373–9 essentials of 368–73 rotor flux 370–1 figure of merit (FOM) 329 FINE-S: see Fuel Cell Innovative Emotion-Sport (FINE-S) first two seconds, launch and boosting 128 FIT: see failure in time (FIT) FlexRay protocol 223–6

flux density 249, 298, 401 intrinsic 249 flux linkage 308, 341, 384 flux penetration 401–2 effect of skin depth on 402 flux squeeze 282–7 flyback converter 485–6 flywheel systems 107–10, 208, 367, 439, 496–9 advantage of 498–9 petrol electric drivetrain 108–9 Swiss Federal Institute flywheel concept 109–10 Texas A&M University transmotor 107–8 FOC: see field oriented control (FOC) FOM: see figure of merit (FOM) Ford Focus 5 door specifications 31 vehicle mass breakdown 34 Ford Motor Co. 8 Escape SUV 8–9, 243, 524, 527–32 Focus 124, 126–8 Mariner SUV (hybrid) 8–9 Model U concept vehicle 74, 76–7 series-parallel switching architecture 74, 76 Ford Prodigy 19 and engine downsizing 20 42 V PowerNet standard mild hybrid architecture 81–2 voltage definition of 82 forward converter 487 Foster II network 512, 514 four stroke gas exchange process 46–7 Frankfurt auto show 2009 70 FREEDM: see Future Renewable Electric Energy Delivery and Management (FREEDM) friction 120, 123 defined 48 and wheel slip 38–41 FTP: see Federal test procedure (FTP) FUDC: see Federal Urban Drive Cycle (FUDC)

Index FUDS: see Federal Urban Drive Schedule (FUDS) fuel, oxidation of 442 fuel cell electric vehicles (FCEV) 10–11 fuel cell hybrids 10–11 Fuel Cell Innovative Emotion-Sport (FINE-S) 103 fuel cell model, storage system 504–7 fuel cell vehicles (FCV) 419, 423–5 vs. FCHV 421, 424 fuel cells 199–204 fuel consumption 420–1, 423–4 brake specific 495 mapping, ICEs and 51–2 percentage of all energy usage 423 fuel economy 129, 135–6, 255, 419, 427–9, 433, 538, 549 actual 41 average speed on 138–9 brake specific fuel consumption 42–3 calculations 41–5 CO2 emissions and 42 combined mode 429–30 and consumption conversions 43–6 drive cycle and 22–3 factors impacting 41, 44 and hybrid functionality 23 with parallel RBS 133 of plug-in electric vehicle 54 powertrain matching and 44 predicting 429, 433 real world 41 testing 430–1 transmission gear selection and 44 utility factor 429 fuel injector failures 494 fuel processing pathways 420–1 fuel savings 531, 538–9 fuses 410

577

Future Renewable Electric Energy Delivery and Management (FREEDM) 177–8 GaAs: see gallium arsenide (GaAs) Gage, Tom gain error 360 gallium arsenide (GaAs) 328 gasoline 419, 421, 423–4, 429 chracteristics 23–4 performance 495 price of 493 Gasoline–electric hybrid concept vehicles 2 gassing 502 gate turn-off thyristors (GTO) 72, 176, 275, 349, 409–10 gear ratios 33, 36 gear shift, power split with 93–6 gear step selection 149–51 General Electric Company 395 General Motors Corp., 9–10, 80 Chevrolet Equinox 10 Chevy Malibu (hybrid) 10 ParadiGM hybrid propulsion system 10 Silverado pick-up truck 9 SUVs by 10 General Motors Corporation, Chevy Volt 447 General Motors (GM) Corporation Cadillac V16 119, 123 Cadillac XV16 119 Silverado 243 Geneva Motor Show 2003 103–4 geometric mean distance (GMD) 274 GHG: see greenhouse gases (GHG) Gibbs free energy (DG ), defined 443 global oil demand, US and 68 GM Covair 28 GM Precept 19 and engine downsizing 20 GMD: see geometric mean distance (GMD)

578

Propulsion systems for hybrid vehicles

goals vehicle economy 19 vehicle performance 18 good ride performance, dynamic index for 29 gradeability 127 grades, and cruise targets 124–7 gravimetric energy 447, 449, 463, 496 Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation (GREET) 420 greenhouse gases (GHG) 1, 420 GREET: see Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation (GREET) grid connected hybrids 56–64 charge sustaining and charge depleting 62–4 HEV20 and HEV60 59–62 vehicle to grid (V2G) vehicles 56–9 GS450h 94–6 development of 95 hybrid model specifications 95 GTO: see gate turn-off thyristors (GTO) guideways 548–9 H0 vehicle 60 H20 vehicle 60 Halbach array 257 Hall effect 361, 382 Hall transducers 259 hard switched inverter 330 harmonics 332, 344, 348, 398–9 HECU: see hydraulic electronic control unit (HECU) helium 496 Helmholtz layer 464 heterodyning techniques 383–4 HEV0 236 HEV inverter: see hybrid electric vehicle (HEV) inverter

HEVs: see Hybrid electric vehicles (HEVs) HFEDS: see highway fuel economy drive schedule (HFEDS) HFT: see high frequency (planar) transformer (HFT) HHV: see high heating value (HHV) high frequency (planar) transformer (HFT) 327 high heating value (HHV), of fuel 543–4 high pressure gas, energy storage 496 high temperature superconductor (HTS) 253 high voltage disconnect 215 high voltage (HV) 330, 346 highway cycle 428 highway drive cycle (Hwy) 137 highway fuel economy drive schedule (HFEDS) 63 Highway fuel economy test (HWFET) 63, 135 Hino company 75 Hitachi corporation 412 Hitachi Industrial Equipment Systems Co. Ltd 248 Holtz, J. 348 Honda by Blue Energy Corp. 7 Honda Insight and engine downsizing 20 as lowest mass vehicle 20 Honda Motor Co. 67, 283, 313 Civic 139, 243 Civic Hybrid 6–8 FCX 11, 243 fuel cell electric vehicles by 10–11 Insight 243 integrated motor assist (IMA) 4–6 S2000 Roadster 5–6 Honeywell 361 HSD: see Hybrid Synergy Drive (HSD) HTS: see high temperature superconductor (HTS) hub motor 100 human-machine interface 233–4

Index Hunt, L.J. 295 Hunt winding 295–6 HV: see high voltage (HV) HWFET: see Highway fuel economy test (HWFET) Hwy: see highway drive cycle (Hwy) hybrid electric vehicle (HEV) inverter 347 dc/dc converters in 349–51 hybrid electric vehicles (HEVs) 2 hybrid functionality, and fuel economy 23 Hybrid Synergy Drive (HSD) 2–3 hybridized battery 481–2 hybridized fuel cell vehicle (FCHV) 321, 424–7 vs. FCV 421, 424 hydraulic actuators 499 hydraulic electronic control unit (HECU) 229 hydraulic launch assist hybrids 104–5 hydraulic post-transmission hybrid architecture 104–7 see also post-transmission parallel hybrid architecture hydraulic-electric posttransmission 105–6 launch assist 104–5 very high voltage electric drives 106–7 hydraulic storage systems 439 hydraulic-electric post-transmission hybrid 105–6 hydrogen 422–5 liquid 423, 493–4 in transportation systems 543 hydrogen storage 493–6 high pressure gas 496 metal hydride 495 hysteresis 388, 396–8 current regulators 349 loops 398–9

579

ICE: see internal combustion engine (ICE) idle times 432 idle–stop 432 functionality 22, 142 IGBT: see insulated gate bipolar transistor (IGBT) ignition timing (IP) 47 IHAT architecture: see integrated hybrid assist transmission (IHAT) architecture IM: see induction machine (IM) IMA: see Integrated motor assist (IMA) system; integrated motor assist (IMA) system IMEP: see indicated mean effective pressure (IMEP) impedance 459, 493, 504, 507, 509–10 IMST: see insulated metal substrate technology (IMST) ‘indicated’, defined 47–8 indicated mean effective pressure (IMEP) 47–8 indicated specific air consumption (ISAC) 47 defined 49 indicated specific fuel consumption (ISFC) 47 inductance 271–2, 308, 403 leakage 341, 383, 386 self 368 induction machine (IM) 243, 268, 296, 315, 318, 368, 373, 384 block diagram model of 370 cage rotor 368–70 classical 290–3 current feeding 372 design 266 FOC and 372 line-start 277, 291 non-linear model 292–3 remanence 258 self-cascaded 295 stator currents 374

580

Propulsion systems for hybrid vehicles

stator winding changeover 293–4 SyncRel vs. 312–13 vs. VRM 269 inductive coupling technology 550–2 inertia 430 calculation 34–5 component 34 defined 33 driveline 285 polar 314 rotor 257, 269 infotainment 141 initial IR 502 ‘inner-loop’ control 325 insulated gate bipolar transistor (IGBT) 261, 325, 328, 348, 356 device technology 406–7 reverse conducting 328 insulated metal substrate technology (IMST) 325 integrated electric traction system (INTETS) 412–13 integrated hybrid assist transmission (IHAT) architecture 97–100 operating modes 98 ratings of 99 single M/G power split architecture 99–100 integrated motor assist (IMA) system 4, 139 Honda 4–6 integrated power and attitude control system (IPACS) 497 integrated starter-alternator (ISA) 70, 119 Mannesmann Sachs 243 integrated starter-generator (ISG) 81–4, 128, 208, 270 idle stop technologies 432 intelligent power electronic module (IPEM) 325–6, 328–9, 354 interactive vehicle dynamics (IVD) 129 RBS with 134–5

Intergovernmental Panel on Climate Change (IPCC) 1 interior permanent magnet (IPM) 243, 266, 268, 277–90, 315, 318, 385 flux squeeze 282–7 multilayer designs 289–90 intermittent overload operation 167 internal combustion engine (ICE) 1, 1n1, 45–55, 126, 327–8, 422–4 BMEP 47–9 BSFC sensitivity to BMEP 49–51 emissions regulations 52–5 four modes (strokes) of 47–8 four stroke gas exchange process 46–7 fuel consumption mapping 51–2 functions 45–6 limitations of technology 422 pressure-volume (P-V) diagram 47–8 internal energy, change in 445 International Rectifier Plug-N-Drive module 261, 325 SuperTab devices 353 International Space Station (ISS) 497 INTETS: see integrated electric traction system (INTETS) inverters 278 corona-free 405 dual mode 274–7 hard switched 330 hybrid electric vehicle 347 losses 406–10 modulation index 330, 333, 337 multilevel 346–7 power 349 power electronics 296, 332 resonant dc link 335–6 resonant pole 330, 335 switching states 331–2 very high power 349 voltage source 335–7 ionic/elect 502 IP: see ignition timing (IP)

Index IPACS: see integrated power and attitude control system (IPACS) IPCC: see Intergovernmental Panel on Climate Change (IPCC) IPEM: see intelligent power electronic module (IPEM) IPM: see interior permanent magnet (IPM) iron 244, 265, 395, 398 ISA: see integrated starter-alternator (ISA) ISAC: see indicated specific air consumption (ISAC) ISFC: see indicated specific fuel consumption (ISFC) ISG: see integrated starter-generator (ISG) ISS: see International Space Station (ISS) IVD: see interactive vehicle dynamics (IVD) Jaguar XJ-S 120 V12 engine 119–20 Jahns, Tom 279 Japan 287 10–15 mode 432–3 NEDC in 135 Jeffries, P. 108 J-N-J Miller Design Services 93n4 Joule heating 395, 397 kinetics 502 L1 (Volkswagen) 70 LA4 cycle 137 Lambilly, H. 359 lamination 312, 397, 399–402 lane change, launch and boosting 128–9 lattice defects 397 launch and boosting, first two seconds 128 and boosting, lane change 128–9 vehicle 296 WOT 139–40

581

lead-acid batteries 194–6, 447–9 discharge behaviour of 447–8 lead-acid technology 385 Lee, F. C. 327 Lei, M. 287 LEM 361 Lenz’s law 254 Lexus RX330, as HSD concept vehicle 2 LHV: see lower heating value (LHV) LiMO2: see lithium–metal oxide (LiMO2) line-to-line voltage 260, 262, 332, 335 SVPWM 346 line-to-neutral voltage 263, 332 link capacitors 361–2, 410 Lipo, T. A. 296 liquefied natural gas (LNG) 496 liquefied petroleum gas (LPG) 419–20 liquid hydrogen 423, 493–4 production of 493–4 properties of 494 lithium ion 196–8 lithium ion battery 454–61 charge characteristic for 456–7 electrodynamics of 455 ultra-capacitors and 481–2 vs. nickel-metal hydride batteries 456 vs. ultra-capacitor 458 lithium ion pack (LNMCO cathode, 3.7 V, 6.0 Ah) battery 7 lithium–metal oxide (LiMO2) 454–5 LNG: see liquefied natural gas (LNG) load tracking series hybrids 76–7 locomotive drives important aspects about 73–4 series hybrid propulsion systems 71–4 Lorentz force 252, 254, 410 low pass filtering (LPF) 383 low voltage (LV) 347

582

Propulsion systems for hybrid vehicles

lower heating value (LHV) 543–4 LPF: see low pass filtering (LPF) LPG: see liquefied petroleum gas (LPG) lumped parameter battery model 503 Lundel alternator 280, 412–13 LV: see low voltage (LV) machine sizing 170–3 Magnetek Motors and Controls 405 magnetic loading 170 magnetic saturation 292–3 magnetoelastic effects 397 magnets magnetic properties of 245 permanent: see permanent magnet maintenance free batteries 447–8 mandrel flexibility 404 marginal efficiency 23 Mariner SUV (hybrid) 8–9 Martensite 244 mass 257 coast down testing 523, 527, 536 energy storage and 461, 466, 475–7, 481 ‘mass factor’ 28 calculations 33–5 Matic, P. 389 maximum speeds 135–6, 138 Maxwell Technologies Inc. 491 McCleer, P.J. 410 MCT: see MOS controlled thyristor (MCT) mean effective pressure (MEP) 47 as function of BMEP 49–50 mean time before failure (MTBF) 357–8 mechanical efficiency (hm) 47–8 mechanical field weakening 287–9 mechanical ‘jerk’, defined 132 mechanical storage systems 439 medium voltage (MV) 347 melting 410

MEP: see mean effective pressure (MEP) metal hydride 495 metal-air batteries 198–9 metal-oxide-semiconductor field effect transistor (MOSFET) 83, 274, 351, 356, 406–7, 409, 486 MEVs: see More electric vehicles (MEVs) M/G braking torque (regen) 130, 133 micro hybrid reversible alternator system 80–1 microcontroller 326 mild hybrids 20, 22, 81–4 mileage weighted probability (MWP) 429–30 Miller cycle 10 minority carrier devices 406 mischmetal compositions 449 mobile media 227 2-mode eCVT 10 modulation index 330, 333, 337, 345 more electric vehicles (MEVs) 2 MOS controlled thyristor (MCT) 409 MOSFET: see metal-oxidesemiconductor field effect transistor (MOSFET) Motloch, Chester 503–4 motor-generators (M/G) machine sizing 170–3 sizing 155–73 torque and power 156–7, 164–8 MultiCAN: see time triggered CAN (TTCAN) multilayer designs 289–90 MV: see medium voltage (MV) MWP: see mileage weighted probability (MWP) Nafion 505–6 NAIAS: see North American International Auto Show (NAIAS) Nana Electronics 361

Index narrow lane vehicles (NLV) 547–9 National Electrical Manufacturers Association (NEMA) standards 403 National Hydrogen Association (NHA) 543 National Personal Transportation Survey (US) 508 National Renewable Energy Laboratory (US) 422 National Transportation Highway Safety Agency (NTHSA) 53 natural gas, compressed 419 NdFeB magnet: see neodymium iron boron (NdFeB) magnet NEDC: see New European Drive Cycle (NEDC) NEDO ACE (Japan) 74 NEMA standards: see National Electrical Manufacturers Association (NEMA) standards neodymium iron boron (NdFeB) magnet 244–5, 247–8, 257–8, 270, 277, 283 NEOMAX 244 NeoMax 27H 245 neutral idle transmission 142–3 New European Drive Cycle (NEDC) 22, 135, 424, 430–1 New York City 266 NHA: see National Hydrogen Association (NHA) NiCd: see nickel-cadmium (NiCd) system nickel-cadmium (NiCd) system 443 nickel-metal hydride (NiMH) batteries 3, 70, 196, 441, 449–54 charge characteristic for 451 discharge behaviour of 449–50 limitations of 449 voltage versus SOC 450 vs. lithium ion battery 456 NiMH: see nickel-metal hydride (NiMH) batteries

583

Nissan Condor Super-Capacitor truck 75 Nissan Motor Co. 10 Nixon, Richard 546 NKK steel 399 NLV: see narrow lane vehicles (NLV) NMHC: see non-methane hydrocarbons (NMHC) noise 404 structure borne 318 non-contacting power transfer 549–53 non-dissipative cell equalization 485–8 types of 485–7 non-methane hydrocarbons (NMHC) 57 non-punch-through (NPT) 261, 325 normal force 527 North America, ranking 2 North American International Auto Show (NAIAS) 59 novel electric machines 247–8 NPT: see non-punch-through (NPT) nuclear bonds, energy storage in 439 Oak Ridge National Laboratory 274 OATT report 85 OCV/SOC slope 502 offset error 360 ohmic polarization 445 oil shock (1970s) 433 one-way clutch (OWC) grounds 98 Ontario Hydro 103 OPEC: see Organization of the Petroleum Exporting Countries (OPEC) Open Systems Interconnection (OSI) 7layered model 217–18 application 218 data link 218 network 218 physical 217 presentation 218 session 218 transport 218

584

Propulsion systems for hybrid vehicles

operating time, warranty vs. 358 Organization of the Petroleum Exporting Countries (OPEC) 12 Osaka Prefecture University 287 Osama, M. 296 oscillation 288 exciting driveline 318 oscilloscopes 404 OSI 7-layered model: see Open Systems Interconnection (OSI) 7-layered model Ostovic, V. 305 OTR: see over the road (OTR) trucks Otto cycle 12 OV: see overvoltage (OV) over the road (OTR) trucks 419, 532–9 overcharge 502 overcharge CA 502 overmodulation 337 overvoltage (OV) 447 OWC grounds: see one-way clutch (OWC) grounds oxidation, of fuel 442 ozone odour 404 P2000 low storage requirement (LSR) vehicle 25 PA Consulting Group 289 ParadiGM hybrid propulsion system 10 parallel diesel–electric hybrid propulsion system 10 parallel hybrid architectures fuel economy benefits of 78 post-transmission: see posttransmission parallel hybrid architecture pre-transmission: see pretransmission parallel hybrid architecture parallel RBS 129, 133 Partnership for a New Generation Vehicle (PNGV) 18–19 goals 18–19

passenger vehicle 124 dynamic attributes 28–9 PC: see propylene carbonate (PC) PCPM: see pole changing permanent magnet (PCPM) PCT 429 peak overload operation 167 peak-to-peak torque 255 PEDT: see petrol electric drivetrain (PEDT) PEI: see pulse endurance index (PEI) PEM: see proton exchange membrane (PEM) fuel cell Pentadyne 111 website 111n7 performance targets 26–7 permanent magnet 243–52, 254, 257, 290, 317, 368, 381 excitation 373 interior 277–90 mechanical properties of 249 performance of 245 pole changing 304–6 properties of 270 rare earth 248–52 synchronous motor 412 types of 247 permanent magnet reluctance machine (PRM) 286–7 permanent magnet synchronous machine (PMSM) 496–7 permeability 396 permeance coefficient 250 personal rapid transit (PRT) 545, 546–7 characteristics and parameters for 546 cost of 546 reasons for 546 petrol electric drivetrain (PEDT) 108–9 Peukert equation 446 Peukert’s slope 502 Phelps Dodge magnet wire company 404

Index PHEV: see plug-in hybrid electric vehicle (PHEV) pitch axis (Y-axis), of vehicle 29 planetary gear, power split architecture 90 planetary gear sets 379 plasma technique, reformation of hydrogen from ethanol and 543 platooning 531 plug-in hybrid electric vehicle (PHEV) 12, 136–8, 447, 543, 545 battery packs 447 fuel economy of 54 plug-in hybrids charge sustaining and 60 vs. conventional vehicles 60 Plug-N-Drive series, International Rectifier 325 PMSM: see permanent magnet synchronous machine (PMSM) pneumatic systems 499 PNGV: see Partnership for a New Generation Vehicle (PNGV) polarity consistency rule 348 polarization 279, 305 activation 443–4 concentration 444–5 ohmic 445 pole changing 294–306 pole changing permanent magnet (PCPM) 304–6 pole-phase modulation (PPM) 294–5, 298–304 polyamideimide 404 polyester 404 polyvinyl chloride (PVC) 211 porous carbon 462, 466, 506 positive temperature coefficient (PTC) 485 post-transmission parallel hybrid architecture 100–4 see also hydraulic post-transmission hybrid architecture Autonomy 100–1 capability curves 102

585

disadvantages of 100 overview 77–8 speed range concerns 102 wheel motor hybrid 102–4 power and data networks 220–1 distribution in drive cycle 24–5 torque and 164–8 power assist architectures, pretransmission parallel hybrids 84–5 power cycling 359 power distribution centres 215–16 power distribution system 410–11 losses 410 power electronics 258 for ac drives 261, 325–63 goal of 326 inverter 296, 332 sizing 173–84 variable reluctance machine 309 power factor 312 power plant specifications 119–43 power split 379 power split architecture background 87 basic functionality of 88–9 ‘electric CVT’, acceleration performance of 90–1 electromechanical transmission and 87–8 epicyclic gear set 90 with gear shift 93–6 IHAT architecture 97–100 operating modes 92 overview 86–7 planetary gear 90 single mode eCVT functional 88–9 ‘stick’ diagram of speed constraints 89 power split hybrid propulsion systems 142 PowerNet (Boardnet) 78 PPM: see pole-phase modulation (PPM)

586

Propulsion systems for hybrid vehicles

predictive controllers 349 pressure wave drag 532 pre-transmission parallel hybrid architecture 77–86 combined configurations 86–100 dual mode 85–6 energy recuperator systems 79–80 fuel economy benefits of 78 micro hybrid 80–1 mild hybrid 81–4 power assist 84–5 PRISM: see program for individual sustainable mobility (PRISM) Prius 21, 31, 140–1, 433 PRM: see permanent magnet reluctance machine (PRM) prognostics 367 program for individual sustainable mobility (PRISM) 548 propane 419, 496 propulsion system and energy storage 78–9 tractive effort, defined 32–3 propylene carbonate (PC) 461 proton exchange membrane (PEM) fuel cell 504–6 reaction of 505 PRT: see personal rapid transit (PRT) PSAT simulation tool 422–3 PTC: see positive temperature coefficient (PTC) pulse endurance index (PEI) 404 pulse-width modulation (PWM) 261, 325–6, 330, 332, 401 comparisons of techniques 348–9, 350 essentials of 330–5 interleaved, for minimum ripple 361–3 sine-triangle 348–9 voltage wave form 404 PVC: see polyvinyl chloride (PVC) PWM: see pulse-width modulation (PWM)

quantization error 361 quantum shield layer 404 radial flux 243 radial laminated structures 313 radial ply tyres 529 radiated near-field power transfer 552–3 Ragone plot 466–9, 478 Randle’s equivalent 503 range extended electric vehicle (REV) 12 rare earth (RE) permanent magnets 248–52, 277–8, 284, 305 characteristics 250 RBS: see regenerative braking systems (RBS) RC-IGBT: see reverse conducting IGBT (RC-IGBT) R/D converters: see resolver-to-digital (R/D) converters RE permanent magnets: see rare earth (RE) permanent magnets real world drive cycles 135 real world fuel economy 41 see also fuel economy rectifier diode operation 383 recuperator systems 79–80 architecture 79 ‘RED Pipe’ 74 redundant systems 356 Reeves type CVT 96 regen (M/G braking torque) 130, 133 power 131–2 regenerative braking systems (RBS) with ABS 133–4 with IVD/VSC/ESP 134–5 parallel 129, 133 series 129–32 split parallel 129 regulated cycle for hybrids 433–5 reliability considerations 356–9 reluctance machines doubly excited 296 permanent magnet 286–7

Index switched 307–11 synchronous 266 variable 243, 266, 277 remanence 258, 270 resistance ac 214 bulk 407 chemical 405 conductor 405 constriction 410 contact 410 dynamic 407 harness 411 rolling 430, 523–7, 529–30, 534, 536–8 stator 385–6 thermal 352 winding 368, 403 resistive cell balancing network, 484 resolver-to-digital (R/D) converters 383 resonant dc link inverter 335–6 resonant pulse modulation 335–7 REV: see range extended electric vehicle (REV) reverse conducting IGBT (RC-IGBT) 328 reverse recovery 409–10 road grade 29 calculation 37, 39–40 road load 125–6 in coast down testing 525–7 road load calculation 32–41 components 32–8 friction and wheel slip 38–41 road surface, coast down testing and 525, 528 road vehicles performance characteristics 18–32 basic dynamics 27–31 drive cycle 22–6 engine downsizing 19–22 hybrid vehicle performance targets 26–7

587

Partnership for a New Generation Vehicle goals 18–19 ‘rocking chair’ chemistry 454 roll axis (X-axis), of vehicle 29, 32–3 rolling resistance 28, 33, 430, 523–7, 529–30, 536 calculation 36–7 for semi-tractor 534 of tyres 537–8 roof equipment drag 532 rotor flux 370–2 rotors 243, 258–9, 270, 289–90, 305, 312 AFPM 288 coupling between stator phases and 368 of Hunt motor 295 inert 313 parallel sided slots 291 variables 369 Runge–Kutta integration 122 S2000 Roadster 5–6 SAE J1772 56–9, 57n7 safe operating area (SOA) 73 Safe-by-Wire Consortium 223 safety bus 227 saliency ratio 279 samarium-cobalt (SmCo) magnet 244–5, 257, 270 Sbarro, Franco 103 SCR: see selective catalytic reduction filter (SCR) selective catalytic reduction filter (SCR) 57 semiconductors 326–9, 348, 351–2 performance 329 power, trends in 327–8 protection fuses 410 voltage rating 73 wide bandgap devices 328–9 sensorless control 367 sensors for current regulators 359–61 separation distances 533, 535

588

Propulsion systems for hybrid vehicles

series propulsion system architecture 71–7 load tracking architecture 76–7 locomotive drives 71–4 series–parallel switching 74–6 series RBS 129–32 series-parallel switching 74–6 shallow cycle life 502 Shanghai Aoweii Technology Development Company 110n6 shift, power split with 93–6 shore power requirements 538 shorting 261 SiC: see silicon carbide (SiC) signal injection 383 silicon 327–8, 399 silicon carbide (SiC) 328 Silverado pick-up truck 9 Simplorer, numerical integration in 131 Sinautec Automotive Technologies 110, 110n6 single in-line package (SIP) 261 single M/G power split architecture 99–100 acceleration performance 99–100 component ratings 99 sinusoidal modulation 334, 337–8 limitation of 338 sinusoidal synchronous (regular) sampling 339–40 SIP: see single in-line package (SIP) Sitras system 113 six step square wave mode 335 skin friction drag 532 SLI: see starting–lighting–ignition (SLI) batteries SmCo magnet: see samarium-cobalt (SmCo) magnet SMES systems: see superconducting magnetic energy storage (SMES) systems snubbers 409 SOA: see safe operating area (SOA) SOC: see state of charge (SOC)

software 414 SOH: see state of health (SOH) solder contact surfaces 359 solid state transformer (SST) 327 space vector PWM (SVPWM) 331, 337–46, 348–9, 385 derivation of 340 line-to-line voltages 346 modulating function, synchronous sampling of 344–5 switching pattern of 341 spacecraft 270 spacing 530–1 class-8 semi-tractor–trailer rig and 535–6 tractor-to-trailer 533 vehicle–vehicle 531–2 spin losses 318 split parallel RBS 129 SPM: see surface permanent magnet (SPM) sport utility vehicle (SUV) 7 dynamic rolling radius of 524 by Ford Motor Co. 8–9 frontal area of 524 by General Motors Corp. 10 test 523, 527–32, 535 trailer test and 529–32 sprung mass 29 defined 30 SST: see solid state transformer (SST) stability programmes 129, 134–5 Stacking factor, of magnetic cores 400 standard temperature and pressure (STP) 524 StARS: see Starter–Alternator Reversible System (StARS) Starter–Alternator Reversible System (StARS) 80 starter-alternators 291, 311 crankshaft mounted 538 starting–lighting–ignition (SLI) batteries 441 state of charge (SOC) 446, 501 state of health (SOH) 499, 501

Index Stationary reference frame 373–4 stator current 370, 372, 379, 381–2, 386 stators 243, 253, 258–9, 261, 269, 277, 296 coupling between rotor phases and 368 parallel sided teeth 291 resistances 274 rolled-out 254 slotting 272 smooth 266 toroidal 299 windings 254, 270, 273, 291, 293 steam methane reformer 544 steel 396–7 good quality 398 lamination 400–1 non-oriented, grades M15–M47 400 steel belt (Van Doorne) CVT 96–7 steering angle 29–30 steering systems 228–9 Steinmetz, Charles Proteus 395, 397 storage system modeling, capacitor system 499–516 battery model 499–504 fuel cell model 504–7 ultra-capacitor model 507–16 storage systems hydraulic 439 mechanical 439 STP: see standard temperature and pressure (STP) stress 356 thermal 359 strontium 244 strontium ferrite 278 Sumitomo Special Metals 244 super-capacitor 471 superconducting magnetic energy storage (SMES) systems 412 surface permanent magnet (SPM) 243, 257, 259, 274, 284, 314–15, 359 design essentials of 265–74 vector diagram for 271

589

surface tension 410 SUV: see sport utility vehicle (SUV) SVPWM: see space vector PWM (SVPWM) Swiss Federal Institute of Technology flywheel CVT hybrid 109–10 operating modes 109 switchable series–parallel hybrid architecture 74–5 energy storage in 74–5 ultra-capacitor storage 75 switching 334, 341, 383–4, 408–9 resonant 335 states, inverter 331–2 switching frequency 348, 401, 408 symmetrical ultra-capacitors 466–70 synchronous machine, torque of 373 synchronous reluctance (SyncRel) 311–13 design 266 vs. induction machine 312–13 SyncRel: see synchronous reluctance (SyncRel) Tafel plot 444, 444n2 Tahoe (hybrid) 10 tank-to-wheels (TTW) 420–3 TDC: see top dead centre (TDC) TDI: see Turbo Direct Injection (TDI) TDMA: see time division multiple access (TDMA) TEATFB: see tetraethylammonium tetrafluoroborate (TEATFB) technologies, energy storage 439–519 technology leadership ranking 2 telematics instrumentation 141 Tempel M19 398 Tesla, Nikola 266 Tesla Electric Company 266 testing class-8 tractor 532–5 coast down 419, 523, 525–9, 533–4, 538

590

Propulsion systems for hybrid vehicles

full charge 429 partial charge 429 validation and 428, 433, 523 tetraethylammonium tetrafluoroborate (TEATFB) 461 Texas A&M University transmotor 107–8 thermal capacitance 353 thermal cycling 356, 359 thermal design 351–6 thermal management 16, 167, 230–3, 259 thermal time 502 thermaleze ‘quantum shield’ (Tz QS) 404–5 thermistor 262 thermodynamics of battery systems 443 thin metal foil (TMF) 461 three box structure 28–9 THS: see Toyota Hybrid System (THS) THS-C: see CVT hybrid powertrain (THS-C) THS-II: see Toyota Hybrid Synergy Drive (THS-II) thyristors 175–6 ac switches 275–6 gate turn off 275, 409–10 MOS controlled 409 time division multiple access (TDMA) 224 time triggered CAN (TTCAN) 222–3 Tino hybrid vehicles 452 TMC: see Toyota Motor Co. (TMC) TMF: see thin metal foil (TMF) top dead centre (TDC) 46 Top Fueler dragster 123 toroidal CVT 97 limitation of 97 Toronto model 508 Torotrak IVT 97 torque 125, 253–4, 274, 296, 316, 374–6, 385–6 braking 278 and cooling 268

crankshaft 538 density 269 driveline 139 electromagnetic 268, 369–70, 372, 389, 391 engine 123, 128 good performance 266 high 256, 267–8, 293, 296 improved 289 instantaneous 308 M/G 130, 133, 375, 403 peak 291, 314–16 peak-to-peak 255 permanent magnet 286 in permanent magnet dc machine 368 power and 164–8 production, in variable reluctance machine 308 reluctance 258, 279, 286 ripple 284–5 stall 290 synchronous machine 373 torque converters 139, 142, 395 torque steer 100 Toyota 2002 Prius and engine downsizing 20 fact sheet 4 vs. Civic hybrids 7–8 Toyota Group switchable series–parallel hybrid 75 Toyota Hybrid Synergy Drive (THS-II) 433 Toyota Hybrid System (THS) 3, 139–40, 351 Toyota Motor Co. (TMC) 2, 103, 351 Crown 432 ES3 experimental hybrid vehicle 75–6 Estima Van 115 FCHV 506–7 FINE-S 103 gear shifting function 94

Index GS450h 94–5 Prius 140, 243, 433 Toyota 2002 Prius 4, 7–8 Trackside RESS 111 Bombardier system 112 using flywheels 112 using ultra-capacitors 113 Trackside/wayside energy storage technology 111–12 traction motor 395–406 tractive effort 125 calculation 35–6 defined 32–3 limits 38–9 tractor (class-8) test 532–5 trailer test 529–32, 535–9 transformations 373–6 transistor stack 352–3, 359 transmission 119–20, 122–3, 149–51 transmotor 107–8 transporting cargo 553–4 trolley bus, ultra-capacitor 113–14 TRW electromechanical transmission 87 Trzynadlowski, A. M., 334 Tsals, Izrail, Dr 289 TTCAN: see time triggered CAN (TTCAN) TTW: see tank-to-wheels (TTW) Turbo Direct Injection (TDI) 70 turbo-generators 405 tyre rolling radius, of vehicle 523 tyre track, defined 31 tyres 419, 536 radial ply 529 rolling radius 124–5 rolling resistance for 537–8 UCG: see uncontrolled generation (UCG) UDDS: see urban dynamometer drive cycle (UDDS) ultra-capacitor model, storage system 507–16 ultra-capacitor storage 75

591

ultra-capacitor-only powered transit bus 110–11 ultra-capacitor-only vehicles 110–13 catenary powered 110–11 catenary powered with wayside/ trackside 111–13 trolley bus 113 ultra-capacitors 204–7, 412 asymmetrical 470–3 with batteries 473–81 buck–boost converter and 486 cell balancing 482–93 lithium ion battery and 481–2 symmetrical 466–70 vs. lithium ion battery 458 uncontrolled generation (UCG) 278, 280–1 undervoltage (UV) 262, 447 Unique Mobility Corporation 412 United States (US), global oil demand and 68 University of Sheffield (UK) 103 University of Tennessee 274 University of Wisconsin 383 unsprung mass 29–30 upstream energy use 420–1 urban dynamometer drive cycle (UDDS) 136–7 US ABC: see US Advanced Battery Consortium (US ABC) US Advanced Battery Consortium (US ABC) 2 US Argonne National Laboratory 54 US-06 drive cycle 135, 137 US National Research Council 2 usage requirements 141–3 customer usage 141 electrical burden 141–2 grade holding and creep 142 UV: see undervoltage (UV) 300V level 2 Valeo 80, 243, 412 valve regulated lead-acid (VRLA) batteries 448–9

592

Propulsion systems for hybrid vehicles

Van Doorne 96–7 variable flux memory motor (VFMM) 305 variable reluctance machine (VRM) 243, 266, 268, 277, 306–13, 318 characteristics of 306 IM vs. 269 power electronics 309 torque production in 308 variable valve timing and lift (VVTL) 12 V_Dot 132 vector control: see field oriented control (FOC) vector rotator matrix 373 vehicle acceleration 122 V10 engine 122 V12 engine 122 V16 engine 122 vehicle architectures 422 vehicle dynamic attributes battery package locations 30 ‘beaming’ 28 definitions of 28–9 equivalent mass 28 for hybrid propulsion simulation 27 ‘mass factor’ 28 packaging hybrid components 28–31 rolling resistance 28 three box structure 28–9 vehicle economy goals (US PNGV Goal 3) 19 vehicle electrification 12–13 CO2 emissions reduction due to 12, 14 and more electric vehicle use 15–17 vehicle longitudinal stability programmes 129 vehicle mass, breakdown 34 vehicle performance cruise, on level terrain 126 goals 18 on level grade 125

vehicle power plant characteristics of 120 torque–speed capability for large engines 121 vehicle speed calculation 37–8 and engine speed, relationship between 37–8 vehicle stability controls (VSC) 129 RBS with 134–5 vehicle to grid (V2G) vehicles 56–9 classification 56–9 emissions 57–8 vehicles size and engine downsizing 19–22 velocity 525, 527 deceleration 534 VFMM: see variable flux memory motor (VFMM) V2G: see vehicle to grid (V2G) vehicles Volkswagen L1 70 VW Tandem 69–70 voltage rating, for high power superconductors 73 voltage source inverter (VSI) 335–7 volumetric efficiency (hv) 47–9 defined 48 volumetric energy 447, 449, 463 Volvo Car Company 8, 103 VRLA: see valve regulated lead-acid (VRLA) batteries VRM: see variable reluctance machine (VRM) VSC: see vehicle stability controls (VSC) VSI: see voltage source inverter (VSI) VVTL: see variable valve timing and lift (VVTL) VW Tandem 1L/100 km concept vehicle 69–70 ‘walk-out’ 408 Walter, J. 351 Walters, Jim 19

Index Warburg impedance 503 warranty vs. operating time 358 water tower model 502–3 waveforms 261, 331, 337, 404 brushless ac motor 263 PWM 345 quasi-square 263 SVPWM gating 344 switching 408 ‘weak link’ 410 weight tally 236–7 wells-to-tank (WTT) 420–1 well-to-wheel (WTW) 421–3, 543, 545 wheel drag 532 wheel inertia 34 wheel mass 34 wheel motor post-transmission hybrid architecture 102–4 wheel slip curve 38–9 friction and 38–41 wide bandgap devices 328–9l wide open throttle (WOT) 127 launch 139–40 windings 362, 368, 373 armature 254, 405–6

593

connections 301–3 Hunt 295–6 PPM 299, 303–4 reconfiguration 293–4 stator 254, 270, 273, 291, 293 tapped 293 toroidal 299 wireless protocols 227 Word 1 parameters 501–2 Word 2 parameters 501–2 WOT: see wide open throttle (WOT) WTT: see wells-to-tank (WTT) WTW: see well-to-wheel (WTW) yaw axis (Z-axis), of vehicle 29 yaw motion 133–5 Yukon (hybrid) 10 Zener diode (sharp knee) 484–5 zero emission vehicle (ZEV) 53 Zetec 125 ZEV: see zero emission vehicle (ZEV) zinc 493

Propulsion Systems for Hybrid Vehicles 2nd Edition

Worldwide, the automotive industry is being challenged to make dramatic improvements in vehicle fuel economy. In Europe there are CO2 emissions penalties prorated by the degree to which vehicles exceed mandated CO2 levels. In the United States, vehicle fuel economy targets set by Congress in 2007 for 20 per cent fuel economy improvement by 2020 are now being accelerated by the Obama administration to 35.5 mpg by 2016 for a passenger car. Taking effect in 2012, the new rules set more aggressive fuel economy measures that will require significant gains in engine and driveline efficiency, better performance cabin climate control and the introduction of electric hybridization. This 2nd Edition of Propulsion Systems for Hybrid Vehicles addresses the electrification innovations that will be required, ranging from low end brake energy recuperators, idle-stop systems and mild hybrids on to strong hybrids of the power split architecture in both single mode and two mode and introducing new topics in plug-in hybrid and battery electrics. Important topics of the 1st Edition are retained and expanded and some outdated material has been replaced with new information.

Dr John M. Miller PE is vice president of systems and applications at Maxwell Technologies. He is also founder and principal engineer of J-N-J Miller Design Services, PLC. Dr Miller worked for 20 years in the automotive industry, leading several hybrid vehicle technology programmes including 42 V Integrated Starter Alternator (ISG) for application into a SUV. He has been active in collaborations at industry and government levels, including the NSF-funded systems center for Future Renewable Electric Energy Delivery and Management (FREEDM). He was actively engaged in MIT’s Consortium on Advanced Automotive Electrical and Electronic Components and Systems and has served as Adjunct Professor of Electrical Engineering at Michigan State University and at Texas A&M University. Dr Miller has authored over 160 technical publications, holds 53 US patents, and has authored or co-authored five books. He is a Fellow of the IEEE, Member of SAE, and 2009 recipient of the IEEE Kliman Innovator award.

The Institution of Engineering and Technology www.theiet.org 978-1-84919-147-0