~ AND Drive SYSTEMS SECOND EDITION Paul C. Krause Oleg Wasynczuk Scott D. SudhoffANALYSIS OF ELECTRIC MACHINERY AND DRIVE SYSTEMS oadIEEE Press 4445 Hoes Lane, PO. Box 1331 Piscataway, NJ O8855-1331 ‘Stamatios V. Kartalopoulos, Edluor in Chief M. Akay M. Eden M. Padgett, J.B. Anderson ME. Elllawary M.S, Newman RJ. Baker RJ. Herrick W.D. Reeve LE, Brewer RF Hoyt G. Zobrist D. Kirk Kenneth Moore, Director of IEEE Press Catherine Faduska, Senior Acquisitions Editor John Grftin, Acquisitions Eaitor Anthony VenGraitis, Project Editor IEEE Power Engineering Society, Sponsor PE-S Liaison to IEEE Press, Chanan Singh Cover Design: William 7; Donnelly, WT Design BOOKS IN THE IEEE PRESS SERIES ON POWER ENGINEERING Analysis of Faulted Power Systems P.M. Anderson 1995 Hardcover $36pp——_(0-7803-1145-0 Power System Protection P.M. Anderson 1999 Hardcover 1344pp—0-7803-3427-2 Power and Communication Cables: Theory and Applications Edited by R. Bartnikas and K. D. Srivastava 2000 Handover -896pp—_0-7803-1196-5, Understanding Power Quality Problems: Vollage Sags and Interruptions Math HI. Bollen 2000 Hardcover S76pp——-7803-4713-7 Electric Power Applications of Fuzzy Systems Edited by M.E. El-Hawary 1998 Hardcover 384pp_——_(0-7803-1197-3ANALYSIS OF ELECTRIC MACHINERY AND DRIVE SYSTEMS Second Edition PAUL C. KRAUSE OLEG WASYNCZUK SCOTT D. SUDHOFF —“| Purdue University, wer Engineering Society, Sponsor MG IEEE Power Engineering Society, Sponsor peg EXSRERING IEEE Press Power Engineering Series Mohamed E. ElHawary, Series Editor Ieee PRESS wom LEY- INTERSCIENCE ‘A JOHN WILEY & SONS, INC. PUBLICATION‘This oak i primed a ace paper Copyright 2002 by The nite of Fler and Electonics Engin, In. Al sighs eee, [No prt of hi publication may be repradoed, sre ina eteva sono anid in any form orby any meas, else, mechanical phtoeopying, recording, samsing or there eicept pei under Seton 17 o 108 of the 197 Unit Stats Copycat At, wien ‘iter th prior writen prison of he Pike ot atharicaton ough payment ofthe npoprit per-copy fet the Copyright Claaoce Cente, 22 Rosod Dive, Daven MA (01928 078)750.8400, fax (978)750474. Rees othe Pblsher for permission sold be ales othe Permissions Deparment, fn Wey & Sons, I, lS Tid Avenue, New Yok, ANY 101580012 (212)90-0t fax (212880 608 E Mai: PERMREQS WILEY.COM, For ering and cestom servic, cll SO0-CALL WILEY. {Library of Congress Cataloging Publication avaiable, ISBN 0471-18326 ‘ited in he Unie Snes of Ameri, woaTesTo our familiesCONTENTS PREFACE x Chapter 1 BASIC PRINCIPLES FOR ELECTRIC MACHINE ANALYSIS 1 LAL Introduction / 1 1.2 Magnetically Coupled Circuits / 1 1.3, Electromechanical Energy Conversion J U1 14 Machine Windings and Air-Gap MMP / 35 LS Winding Inductances and Voltage Equations / 47 References / $8 Problems / 58 Chapter 2 DIRECT-CURRENT MACHINES 67 2. Introduetion £67 2.2 Elementary Direct-Current Machine / 68 2.3. Voltage and Torque Equations / 76 24 Basie Types of Direct-Current Machines / 78 2.5. Dynamic Characteristics of Permanent-Magnet and Shunt de Motors / 88 2.6 Time-Domain Block Diagrams and State Equations / 92 2.7 Solution of Dynamic Characteristics by Laplace Transformation / 98 References J 104 Problems / 105vili contents Chapter 3 REFERENCE-FRAME THEORY 109 3.1 Introduction / 109 32. Background / 109 3.3 Equations of Transformation: Changes of Variables J 111 34 Stationary Circuit Variables Transformed to the Arbitrary Reference Frame / 115 35 Commonly Used Reference Frames / 123 3.6 Transformation Between Reference Frames J. 124 3.7 Transformation of a Balanced Set / 126 38 Balanced Steady-State Phasor Relationships 127 39 Balanced Steady-State Voltage Equations / 130 3.10. Vatiables Observed from Several Frames of Reference J 133 References J 137 Problems / 138 Chapter 4 SYMMETRICAL INDUCTION MACHINES 1 4.1 Introduction / 141 42 Voltage Equations in Machine Variables / 142 43° Torque Equation in Machine Variables / 146 44 Equations of Transformation for Rotor Circuits / 147 4.5 Voltage Equations in Arbitrary Reference-Frame Variables / 149 46 Torque Equation in Arbitrary Reference-Frame Variables / 153 47 Commonly Used Reference Frames / 154 48 Per Unit System / 155 49 Analysis of Steady-State Operation J 157 4.10 Free Acceleration Characteristics / 165 4.11 Free Acceleration Characteristics Viewed from Various Reference Frames / 172 4.12 Dynamic Performance During Sudden Changes in Load Torque 174 4.13 Dynamie Performance During a 3-Phase Fault atthe ‘Machine Terminals J. 181 4.14 Computer Simulation in the Arbitrary Reference Frame / 184 References / 187 Problems / 188 Chapter 5 SYNCHRONOUS MACHINES 191 5.1 Introduction / 191 5.2 Voltage Equations in Machine Variables / 192 53. Torque Equation in Machine Variables / 197cconrents ix 5A Stator Voltage Equations in Arbitrary Reference-Frame Variables / 198 55 Voltage Equations in Rotor Reference-Frame Variables: Park's Equations 7 200 5.6 Torque Equations in Substitute Variables / 206 5.7 Rotor Angle and Angle Between Rotors / 207 5.8 Per Unit System / 209 59 Analysis of Steady-State Operation / 210 5.10 Dynamie Performance During a Sudden Change in Input Torque / 219 5.11 Dynamie Performance Dusing a 3-Phase Fault at the Machine Terminals / 225 5.12 Approximate Transient Torque Versus Rotor Angle Characteristics / 229 5.13 Comparison of Actual and Approximate Transient Torgue-Angle Characteristies During a Sudden Change in Input Torque: First Swing Transient Stability Limit / 232 5.14 Comparison of Actual and Approximate Transient Torque-Angle (Charactristies During a 3-Phase Fault atthe Terminals: Critical Clearing Time / 239 5.15 Equal-Area Criterion / 242 5.16 Computer Simulation / 246 References / 255 Problems / 256 Chapter 6 THEORY OF BRUSHLESS de MACHINES 261 6.1 Introduction / 261 62 Voltage and Torque Equations in Machine Variables / 261 63 Voltage and Torque Equations in Rotor Reference-Frame Variables / 264 64 Analysis of Steady-State Operation / 266 65 Dynamic Performance / 274 References / 281 Problems / 281 Chapter 7 MACHINE EQUATIONS IN OPERATIONAL aM 72 13 14 15 16 IMPEDANCES AND TIME CONSTANTS 283 Iniroduetion / 283 Park's Equations in Operational Form / 284 Operational Impedances and G(p) for a Synchronous Machine with Four Rotor Windings / 284 Standard Synchronous Machine Reactances / 288 Standard Synchronous Machine Time Constants 290 Derived Synchronous Machine Time Constants / 291x contents 7.7 Parameters from Shor-Circuit Characteristics / 294 7.8 Parameters from Frequency-Response Characteristics / 301 References J 307 Problems / 308 Chapter 8 LINEARIZED MACHINE EQUATIONS ai 8.1 Inoduetion / 311 8.2 Machine Equations to Be Linearized / 312 83. Linearization of Machine Equations / 313 84 Small-Displacement Stability: Eigenvalues / 323, 8.5 Eigenvalues of Typical Induction Machines / 324 8.6 Eigenvalues of Typical Synchronous Machines 327 8.7 Transfer Function Formulation / 330 References J 335 Problems / 335, Chapter 9 REDUCED-ORDER MACHINE EQUATIONS. 337 9.1 Introduction / 337 9.2 Reduced-Order Equations / 338 9.3 Induction Machine Large-Excursion Behavior Predicted by Reduced-Onder Equations / 343, 94 Synchronous Machine Large-Excursion Behavior Predicted by Retuced-Order Equations / 350 95 Linearized Reduced-Order Equations / 354 9.6 Eigenvalues Predicted by Linearized Reduced-Onder Equations / 354 9.7 Simulation of Reduced-Order Models / 355 9.8 Closing Comments and Guidelines / 358 References / 358 Problems / 359 Chapter 10. SYMMETRICAL AND UNSYMMETRICAL 2-PHASE INDUCTION MACHINES. 361 10.1 Introduction / 361 10.2 Analysis of Symmetrical 2-Phase Induction Machines / 362 10.3 Voltage and Torque Equations in Machine Variables for Unsymmetrical 2-Phase Induction Machines / 371 10.4 Voltage and Torque Equations in Stationary Reference-Frame Variables for Unsymmetrical 2-Phase Induction Machines / 373contents xi 10.5 Analysis of Steady-State Operation of Unsymmetrical 2-Phase Induction Machines 377 10.6 Single-Phase Induction Machines / 383 References / 393 Problems / 393 Chapter 11 SEMICONTROLLED BRIDGE CONVERTERS, 305 11.1 Introduction / 395 11.2. Single-Phase Load Commutated Converter / 395 11.3 3-Phase Load Commutated Converter / 406 References 1 42: Problems / 425 Chapter 12 dc MACHINE DRIVES ar 12.1 Introduction / 427 12.2 Solid-State Converters for de Drive Systems 427 123 Steady-State and Dynamic Characteristics of aclde Converter Drives / 431 124 One-Quadrant defde Converter Drive / 443 125 Two-Quadrant de/de Converter Drive 460 126 Four-Quadrant defde Converter Drive / 463 12.7 Machine Control with Voltage-Controlled de/de Converter / 466 128 Machine Control with Current-Contolled de/de Converter / 468 References / 476 Problems / 476 Chapter 13. FULLY CONTROLLED 3-PHASE BRIDGE ‘CONVERTERS 481 13.1 Inoduction / 481 132 ‘The 3-Phase Bridge Converter / 481 133 180° Voltage Source Operation / 487 1344 Pulse-Width Modulation / 494 135 Sine-Triangle Modulation / 499 136 Thied-Harmonie Injection / 503 13.7 Space-Vector Modulation J 506 138 Hysteresis Modulation / $10 139 Delta Modulation / 512 13.10. Open-Loop Voltage and Current Control J $13 13.11 Closed-Loop Voltage and Current Controls / 516xii contents, References / 520 Problems / $21 Chapter 14 INDUCTION MOTOR DRIVES 525 14.1 Introduction / 525 14.2 Volts-Per-Hertz Control / 143 Constant Slip Current Control / 332 144 Field-Oriented Control 540 14.5 Direct Rotor-Oriented Field-Oriented Control / $4 14.6 Robust Direct Field-Oriented Control / 546 14.7 Indirect Rotor Field-Oriented Control / 550 148 Conclusions / 554 References / 554 Problems / $55 Chapter 15 BRUSHLESS de MOTOR DRIVES 587 15.1 Introduction / 537 15.2 Voltage-Source Inverter Drives / $58 153, Equivalence of VSI Schemes to Kalized Source 1 560 154 Average-Value Analysis of VSI Drives / 568 15.5 Steady-State Performance of VSI Drives / 571 186 Transient and Dynamic Performance of VSI Drives / S74 15.7 Consideration of Steady-State Harmonies / 578 158 Case Study: Voltage-Source Inverter-Based Speed Control / 582 159 Current-Regulated Inverter Drives / 586 15.10 Voltage Limitations of Current-Souree Inverter Drives / 590 15.11 Current Command Synthesis / 591 15.12. Average-Value Modeling of Current-Regulated Inverter Drives / 595 15.13 Case Study: Current-Regulated Inverter-Based Speed Controller / $97 References / 600 Problems / 600 ‘Appendix A Trigonometric Relations, Constants and Conversion Factors, and Abbreviations 603, INDEX 605PREFACE ‘The first edition of this book was written by Paul C. Krause and published in 1986 by ‘MeGraw-Hil. Eight years later the same book was republished by IEEE Press with (leg Wasynczuk and Scott D. Sudhoff added as co-authors, The focus of the first dition was the analysis of electric machines using reference frame theory, wherein the concept of the arbitrary reference frame was emphasized. Not only has this approach been embraced by the vast majority of electric machine analyst, it has also become the approach used in the analysis of electric drive systems. The use of reference-frame theory o analyze the complete drive system (machine, converter, and contro) was not emphasized inthe frst edition. The goal ofthis edition i to fill this void and thereby meet the need of engineers whose job itis to analyze and design the complete drive system, For this reason the words “and Drive Systems” have been added to the tite [Although some of the material has been rearranged or revised, and in some eases «eliminated, such as 3-phase symmetrical components, most ofthe material presented inthe fist ten chapters were taken from the original edition. For the most part, the ‘material in Chapters 11-15 on electric drive systems is new. In particular, the an lysis of converters used in electric drive systems is presented in Chapters 11 and 13, hile d, induction, and brushless de motor drives are analyzed in Chapters 12, 14, and 15, respectively Central to the analysis used inthis text isthe transformation to the arbitrary refer cence frame. All eal and complex transformations used in machine and drive ana lyses can be shown to be special cases ofthis general transformation. The modern electric machine and drive analyst must understand reference frame theory. For this reason, the complete performance ofall electric machines and drives considered are illustrated by computer traces wherein variables are often portrayed in differentframes of reference so that the student is able to appreciate the advantages and sig- nificance of the transformation used, ‘The material presented inthis text can be used most beneficially if the student has had an introductory course in eleciic machines. However, a senior Would be com= fortable using this textbook as a first course. For this purpose, considerable time should be devoted to the basic principles discussed in Chapter I, perhaps some of ‘Chapter 2 covering basic de machines, most of Chapter 3 covering reference frame theory, andthe beginning sections of Chapters 4, 5, and 6 covering induction, syn- cthronous, and brushless de machines. ‘Some of the material that would be of interest only tothe electric power engineer has been reduced or eliminated from that given in the fist edition. However, the material found in the final sections in Chapters 4 and S on induction and synchronous ‘machines as weil as operational impedances (Chapter 7), and reduced-order model- ing (Chapter 9) provide an excellent background for the power utility engineer, ‘We would like t acknowledge the efforts and assistance of the reviewers, in par- ticular Mohamed E. El-Hawary, and the staff of IEEE Press and Jobn Wiley & Sons, Paut.C. Keaust cee Wasyncaux Scorr D. Suonore Wea Lafeene, tine November 200)ANALYSIS OF ELECTRIC MACHINERY AND DRIVE SYSTEMSChapter 1 BASIC PRINCIPLES FOR ELECTRIC MACHINE ANALYSIS 4.1 INTRODUCTION ‘There are several basic concepts that must be established before the analysis of elec- tic machines can begin. The principle of electromechanical energy conversion is perhaps the comerstone of machine analysis. This theory allows us to establish an ‘expression of electromagnetic torgue in terms of machine variables, generally the ‘currents and the displacement of the mechanical system. Other principles that must be established are (1) the derivation of equivalent circuit representations of ‘magnetically coupled circuits, (2) the concept ofa sinusoidally distibuted winding, (G) the concept ofa rotating airgap magnetomotive force (MME), and (4) the deri- vation of winding induetances. The above-mentioned basic principles are presented in this chapter, concluding with the voltage equations of a 3-phase synchronous ‘machine and a 3-phase induction machine. Ic is shown thatthe equations, which describe the behavior of altemating-current (ac) machines, contain time-varying coeflcients due to the fact that some of the machine inductances are functions of the rotor displacement. Tis establishes an awareness of the complexity ofthese vol- tage equations and sets the stage for the change of variables (Chapter 3), which reduces the complexity ofthe voltage equations by eliminating the time-dependent inductances, 1.2. MAGNETICALLY COUPLED CIRCUITS Magnetically coupled electric circuits are central to the operation of transformers ‘and electric machines. In the case of transformers, stationary cirewits are 12 BASIC PRINCIPLES FOR ELECTRIC MACHINE ANALYSIS Figure 1.21 Magnetically coupled circuits ‘magnetically coupled fr the purpose of changing the voltage and current level. In the case of electric machines, citcuits in relative motion are magnetically coupled for the purpose of wansfersing energy between mechanical and electrical systems Because magnetically coupled circuits play such an important role in power vans- mission and conversion, itis important to establish the equation that describe thet ‘behavior and to express these equations in a form convenient for analysis. These goals may be achieved hy starting with two stationary electric circuits that are mag ‘eially coupled as shown in Fig. 1.21. The two eols consis of turns Ny and N respectively, and they are wound ona common core that is general a ferromagnetic ‘material with permeability large celatve to that of ait. The permeability of fee space, , is $x % 10-7 Him. The permeability of other materials is expressed as A= pip wher pis the relative permeability. Inte ease of transformer stel the relative permeability may be as high 2000 to 4000, In general, the hx produced by each coil can be separated into two components 8 Teakage component denoted with an I subscript and a magnetizing component ‘denoted by an m subscript. Each ofthese components is depicted by asingle steamn- line withthe postive direction determined by applying the right-hand rule to the dlieetion of current flow inthe col. Often, in transformer analysis, fis elected positive out ofthe tp ofeol 2, and a dot is placed at that termina ‘The fs linking each coil may be expressed a8 Dy + Oy + Gao (124) 1D; = Op + yo + Bq (1.22) ‘The leakage Mux 4 is produced by current flowing in coil 1, and it Tinks only the ‘tums of col . Likewise, the leakage flux is produced by current lowing in coil 2, and it inks only the turns of coil 2. The magnetizing ux yy is produced by current owing in coil 1, and it links all rns of coils 1 and 2. Similarly, the magnetizing ffux da is produced by current flowing in coil 2, and it also Tinks ll tums of coils 1 nd 2, With the selected postive direction of current flow and the manner in whichMAGHETICALLY COUPLED CIRCUITS 3 the coils are wound (Fig. 1.2-1), magnetizing flux produced by positive current in ‘one coil adds to the magnetizing Aux produced by positive current in the other cil. In other words, if both currents are actually flowing in the same direction, ‘the magnetizing fluxes produced by each coil are in the same direction, making the total magnetizing flux or the total core ux the sum of the instantaneous magni tudes of the individual magnetizing Muxes. Ifthe actual currents are in opposite directions, the magnetizing uxes are in opposite directions. In this case, one coil is said to be magnetizing the core, the other demagnetizing Before proceeding, i is appropriate to point out that this is an idealization ofthe actual magnetic system. Cleary, all of the leakage ux may not Kink all the tums of ‘the coll producing it. Likewise, all ofthe magnetizing fux of one coil may not Tink all ‘of the turns ofthe other coil, To acknowledge this practical aspect of the magnetic system, the numberof turns is considered to be an equivalent number rather than the ‘actual number Ths fet should cause us litle concern because the inductances ofthe electric circuit resulting from the magnetic coupling are generally determined from tests, ‘The voltage equations may be expressed in matrix form as varie (si ae 123) where # = diag [re a dagonal mates, and ()" =li Al (12-4) where frepresents voltage, curent or ux linkage. The resistances rj and r2 and the flux linkages 2, and 2 are related to coils 1 and 2, respectively. Because it is assumed that links the equivalent turn of eo! and‘ links the equivalent tums of col 2, the ux linkages may be writen as A=Ma, (125) Aa = Ny: (126) where and are given by (1.2-1) and (1.2-2), respectively Linear Magnetic System If saturation is neglected, the system is linear and the fluxes may be expressed as Mi 0-9 (127) Bq = Mit (128) b. (129) oq = Bt (1240)4 SASIC PRINCIPLES FOR ELECTRIC MACHINE ANALYSIS Where @y and dz are the reluctances ofthe leakage paths and ys the reluctance of the path ofthe magnetizing fluxes. The product of N times i (ampere-turs) isthe (MME, which is determined by application of Ampere’s law. The reluctance of the leakage paths is dificult to express and impossible to measure. In fact, a unique determination of the inductances associated with the leakage flux cannot be made by tests instead, itis either calculated or approximated from desi ‘The reluctance ofthe magnetizing path ofthe core shown in Fig. 12-1 puted with sufficient accuracy from the well-known relationship ‘where isthe mean or equivalent length ofthe magnetic path, is the cross-sectional area, and jis the permeability Sabutting(12-7)-0.210) ino (12-1) and (12-2) yields Nut Md, aya, ta ee Mi Bs, i al a (213) Sting (12-12 and (12-13) int (12-5) and (12-6) yes a= My ML, 4 MMe (21) Ry” Rw Ba MONE. NaN Bn a oy (1248) ‘When the magnetic system is linear, the lux linkages are generally expressed! in terms of inductances and currents. We see thatthe coelicients of the first two terms ‘0 the right-hand side of (1.2-14) depend upon the turns of coil | and the reluctance of the magnetic system, independent of the existence of coil 2. An analogous state ‘ment may be made reganding (1.2-15). Hence the selfsinductances are defined as (1.216) (12417) Where Ly and Lp are the leakage inductances and Ly and Lo are the magnetizing inductances of coils 1 and 2, respectively. From (1.2-16) and (1.2-17) it follows that the magnetizing inductances may be related as (248)MAGNETICALLY couPLED CIRCUTS 5: ‘The mutual inductances are defined as the coefficient of the third term of (12-14) and (1.2-15). NN bas (1249) (1.220) Obviously, L12 = La). The mutual inductances may be related to the magnetizing inductances. In particular, la (1228) “The Mux inkages may now be writen a8 neu (12-20) where (tele Be] ee far tn)” | Mla bet bao Although the voltage equations with the inductance matrix L incorporated may be used for purposes of analysis, itis customary to perform a change of variables that yield the well-known equivalent T circuit of two magnetically couple coils. To set, the stage for this derivation, let us express the lux linkages from (1.2-22) as nis ba(u 88s) a2 a sr stal tins) was [Now we have two choices, We can use a substitute variable for (N2/N;)i2 oF for /N2)i. Let us consider the fist of these choices (1.226) whereupon we are using the substitute variable i that, when Rowing through coil 1, produces the same MMF as the actual / flowing through coil . This is said to be referring the current in coil 2to coil 1 whereupon coil I becomes the reference col ‘On the other hand, if we use the second choice, then (1227)6 BASIC PRINGIPLES FOR ELECTRIC MACHINE ANALYSIS Here, is the substitute variable that produces the same MMF when lowing through coil 2 asi, does when flowing in coil 1. This change of variables is said to refer the current of coil 1 to coil 2 We will demonstrate the derivation of the equivalent T circuit by refering the ceurreat of coil 2 to coil 1; thus from (1.2-26) we obtain a=My, (1.228) (1.2.29) ‘whereupon voiz = v4i). Flux linkages, which have the units of volt-second, ane related fo the substitute flux linkages in the same way as voltages. In particular, My, (1.230) If we substitute (1.2-28) into (12-24) and (12-25) and then multiply (1.2-25) by Ni/No to obtain 2 and we further substitute (N3/N7 ma for Laz into (12-24), then Ay = Lats + La ln +8) (1230) A= ligiy tll +8) (12:32) where (1.233) ‘The voltage equations become (1.234) (12:35) where a= (Mn (236)MAGNENCALLY COUPLED CIRCUTS. 7 Figure 12-2. Fguvalenteteuit with col I selected as the reference col, “The above voltage equations suggest the T equivalent ciruit shown in Fi is apparent that this method may be extended to include any number of eoils wound fon the same core. Example 1A Its instructive to ilustrate the method of deriving an equiva- lent T circuit from open- and short-circuit measurements. For this purpose let "us assume that when coil 2of the two-winding transformer shown in Fig. 1.2-1 is open-circuited, the power input to coil 2 is 12 W with an applied voltage is 100 V (ems) at 60 Hz and the current is 1A (rms). When coil 2 is short revited, the curent flowing in cil Lis 1 A when the applied voltage is 30 V at 60 Hz. The power during this testis 22 W. IF we assume Ly = Lis. ‘an approximate equivalent T circuit can be determined from these measure ‘ments with col I selected asthe reference coil ‘The power may be expressed as = |Vilfhi}eos @ (1A) ‘where and 7 are phasors and ds the phase angle between V and 7 (power factor angle). Solving fr 6 during the open-circuit text, we have 12 Tot =r (a2) With 7 asthe reference phasor and assuming an inductive circuit where T, lags Vy, we obtain =12+j10939 (a3) If we neglect hysteresis (core) losses, which in effect assumes a linear mag- netic system, then ry = 120, We also know from the above calculation that Xi +Xqi = 109.32, For the short-circuit test we will assume that j; = — because transformers are designed so that Xt “> [r+ JN. Hence using (TA-1) again we obtain a 0x1 = co = 408° a4)8 GASIC PRINCIPLES FOR ELECTRIC MACHINE ANALYSIS Inthis ease the input impedance is (7, +r) +j(Xn + Xf). This may be deter: mined as follows: 7 =24j2049 (As) = 1092 and, because it is assumed that Xj = Xf, both ae 10.20, ‘Therefore Xqi = 109.3 — 10.2 = 99.10. In summary, = 120 Ly =262.9mH wo In = 27.1 ma 27.1 mH Nonlinear Magnetic System Although the analysis of transformers and electric machines is generally performed assuming a linear magnetic system, economics dictate that in the practical design of. ‘these devices some saturation occurs and that heating of the magnetic material exists ‘due to hysteresis losses. The magnetization characteristics of transformer or ‘machine materials are given in te form of the magnitude of flux density versus mag- nitude of feld strength (B-H curve) as shown in Fig, 1.23, If itis assumed thatthe ‘magnetic flux is uniform through most ofthe core, then B is proportional to and #7 is proportional to MMF. Hence a plot of flux versus current is of the same shape as 1a) 8, Woin — 7: i 23 co CO Atm Figure 1.23. B-H curve for typical silicon tel used in uansformersMAGNENCALLY CouPLED cIRCUTS the B-HT curve. A transformer is generally designed so that some saturation occurs during normal operation. Electric machines are also designed similarly in that a ‘machine generally operates slightly in the saturated region during normal, rated ‘operating conditions. Because saturation causes coefficients ofthe differetial equ tions describing the behavior of an electromagnetic device to be functions ofthe coil ‘currents, a transient analysis is difficult without the aid of a computer. Our purpose here is not to set forth methods of analyzing nonlinear magnetic systems. This Pro- ‘cedure is quite straightforward for steady-state operation, butt cannot be used when analyzing the dynamics of electromechanical devices [1]. A method of incorporating the effects of saturation into a computer representation is of interest Computer Simulation of Coupled Circuits Formulating the voltage equations of stationary coupled windings appropriate for ‘computer simulation is straightforward and yet this technique is fundamental to the computer simulation of ac machines. Therefore i isto our advantage to consider this method here. For this purpose fetus first write (1.2-31) and (1.2-32) as Ay Lain the (1237) Any (1.238) where An = kali) (1.239) Solving (1.2-37) and (12-38) forthe currents vields (12-40) (24 Wf (1.240) and (1.2-41) are substituted into the voltage equations (1.2-34) and (1.2.35) and we solve the resulting equations for ux linkages, the following equ tions are obtained: a= |[+Ze-aa (12-42) Zz Boy a-ffuege-ae (2s ‘Substituting (1.2-40) and (1.2-41) into (1.2-39) yields (12-44) where (12-45)10 BASIC PRINCIPLES FOR ELECTRIC MACHINE ANALYSIS. ‘We now have the equations expressed with 2; and 2 a tate variables. Inthe com puter simulation, (12-42) and (12-43) are used to solve for 2: and 2 and (12-44) is {sed 10 solve for The currents can then be obtained from (12-40) and (1.24). I is clear that (12-44) could be substituted into (12-40)-1.2-43) and that could be ‘climinatd from the equations whereupon it would not appear inthe computer sm- lation. However, we Will ind iq (he mutual flux linkages) an important variable when we include the effects of saturation Ir the magnetization characteristics (magnetization curve) of the coupled winding is known, the effects of saturation of the mutual flax path may be readily ineoxpo- ‘ated into the computer simulation. Generally, the magnetization curve can be ade- quately determined from a test wherein one of the windings is open-circuited {Ginding 2, for example), and the input impedance of the other winding (winding 1) is determined from measurements asthe applied voltage is increased in magni- tue from ze0 to say 150% of the rated value. With information obtained fom this type of test, we an plot zy versus (+) a8 shown in Fig, 12-4 wherein the slope ofthe linear portion ofthe curve is Zn From Fi, 1.24, iti clear that inthe region of saturation we have diy = Lgl +4) ~ fll) (12-46) $G+i) tm + 8) Figure 12-4 Magnetization curveELECTROMECHANICAL ENERGY CONVERSION 11 10 Figure 2-5 f(g) vers from Fig 124 ‘where f/q) may be determined from the magnetization curve for each value of dn. In particular, (2) is function of Zy as shown in Fig. 1.2-5. Therefore, the effects ‘of saturation ofthe mutual flux path may be taken ito account by replacing (1.2-39) With (1.2-46) for Zy. Substituting (1.2-40) and (1.2-41) for jy and f, respectively, into (1.2-46) yields the following computer equation for Ay rante(2ea) Bes oa Hence the computer simulation for including saturation involves replacing Ay given bby (1.2-44) with (1.2-47) where f(g) i generated function of A determined from the plot shown in Fig. 12-5. 1.3. ELECTROMECHANICAL ENERGY CONVERSION Although electromechanical devices are used in some manner in a wide variety of systems, electric machines are by far the most common. I is desirable, however, to establish methods of analysis that may be applied to alleleetromechanical devices, Energy Relationships Electromechanical systems are comprised ofan electrical system, « mechanical sys ‘em, and a means whereby the electrical and mechanical systems can interact. Inter- action can take place through any and all electromagnetic and electrostatic fields that are common to both systems, and energy is transferred from one system to the other asa result of this interaction. Both electrostatic and electromagnetic coupling fields ‘may exist simultaneously, and the electromechanical system may have any number of electrical and mechanical systems, However, before considering an involved sys tem itis helpful to analyze the electromechanical system in a simplified form. An electromechinical system with one electrical system, one mechanical system, and12 BASIC PRINCIPLES FOR ELECTRIC MACHINE ANALYSIS. “Bleach, Coupling Mechanical scm wad" Sen Figure 13-1 Block diagram of an elementary elactomechanica system, ‘one coupling field is depicted in Fig. 3-1. Electromagnetic radiation is neglected, and iti assumed thatthe electrical system operates at a frequency sufficiently low so thatthe electrical system may be considered as s lumped parameter system, Losses accu in all components of the electromechanical system. Heat loss will, ‘cur in the mechanical system due to frietion, and the electrical system will dissi- pate heat due to the resistance ofthe current-carrying conductors. Fdy current and hysteresis losses occur in the ferromagnetic material ofall magnetic fields, whereas dielectric losses oceur in al electric fields. If We is the total energy supplied by the electrical source and Wa, is the total energy supplied by the mechanical source, then the energy distribution could be expressed as We = W, + Wa + Was (34) Woy + Won + Was (132) In (13-1), Wes isthe energy stored in the elctic or magnetic feds that are not coupled with the mechanical system. The energy Wa. isthe heat losses associated with the electrical sytem, These losses occur de 0 the resistance of the eurent- carrying conductors aswell sth energy dissipated from these fel in te form of, heat de to hysteresis edly cuments, and delet Tose. The energy W, istheenersy transferred othe coupling eld by the electrical system, The eneses common to the ~~ _< mechanical sytem may be defined in similar manner. In (1.32), Was isthe energy gored inthe moving member and compiances ofthe mechanical syst, Wy is the “Regnery losses ofthe mechanical system inthe form of heat, and Wn isthe enerzy 2. franstered to the coupling eld. Is important to note that with the convention adopt SA fat the energy supplied by either source is considered positive. Therefore, Ws( Wx) fs negative when energy is supplied to the electrical source (mechanical source. PI Wis defined asthe total energy transfered to the coupling fed, then V+ Wr, (133) Wr ‘where Wi energy stored in the coupling fied and Wy. isthe energy dissipated inthe form of heat due to losses within the coupling field (eddy current, hysteresis, or dielectric losses), The electromechanical system must obey the law of conservation of energy; thus We ~ Wor ~ Wes) + (Wie ~ Want ~ Was) (13-4) Wy + Wye which may be written Wy Wy = Wes Wy (135) ‘This energy relationship is shown schematically in Fig, 13-2ELECTROMECHANICAL ENERGY CONVERSION 13 Figure 13-2 Energy balance ‘The aciual process of converting electrical energy to mechanical energy (or vice versa) is independent of (1) the loss of energy in either the electrical othe mechan ical systems (We, and Wa). 2) the energies stored inthe electic or magnetic fields that are not common to both systems (W,s), or (3) the energies stored inthe mechan- ical system (Ws). I the losses of the coupling field are neglected, then the field is conservative and (1.3-5) becomes Wy = We + Wy (13-6) Examples of elementary electromechanical systems are shown in Figs. 13-3 and 1.34, The system shown in Fig. 1.33 has a magnetic coupling field while the elec- tromechanical system shown in Fig. 13-4 employs an electric field as a means of ‘ansfersing energy between the electical and mechanical systems, In these systems, is the voltage ofthe electric source anf isthe external mechanical force applied t0 the mechanical system. The electromagnetic or electrostatie force is denoted by fi The resistance ofthe current-carrying conductors is denoted by r, and ls the induc tance of a linear (conservative) electromagnetic system that does not couple the ‘mechanical system. In the mechanical system, M isthe mass ofthe movable member hile the linear compliance and damper are represented by a spring constant Kand a Z Cy ! Ci i ae Figure 133 Elecwomechanial system witha magnetic Hel1 aksic PRINCIPLES FOR ELECTRIC MACHINE ANALYSIS. Figure 1.34 Electomechanical sytem with an elect fed ‘damping coefficient D, respectively. The displacement 1p isthe zero foree or equli- brium postion of the mechanical system whieh is the steady-state position of the mass with f. and f equal to zero. A series or shunt capacitance may be included in the electrical system wherein energy would also be stored in an electric feld ‘extemal to the eleetomechanical process. ‘The voltage equation that describes both electrical systems may be writen as di vedsiite 39) ‘where ¢, is the voltage drop across the coupling field. The dynamic behavior ofthe translational mechanical system may be expressed by employing Newton's law of ‘motion. Thus, dx nde SoM + DE 4K 30) he (138) Thal ty pl ye wi we fa «a9 Teale spy ie hi son Wu = fra (1310) ‘which may also be expressed as y= [yaEELECTROMECHANICAL ENERGY CONVERSION 15, Werf af sa sy ‘The first term on the right-hand side of (1.3-12) represents the energy loss due tothe resistance ofthe conductors (WWz,). The second term represents the energy stored in the linear electromagnetic field external to the coupling field (W,s). Therefore, the total energy transferred tothe coupling field from the electrical system is fev (1313) Similarly, forthe mechanical system we have "\e Here, the first and third terms on the right-hand side of (1.3-14) represent the energy stored in the mass and spring, espectively (Wa). The second tern is the heat loss doe to friction (Wy). Thus, the total energy transferred to the coupling field from ‘the mechanical system is Ww +0 {(S) ac {amare a) Wa = Joa (13.13) kis important co note that a positive force fis assumed tobe in the same direction 48. positive displacement, dr. Substituting (13-13) and (1.315) into the energy bal- ance relation, (13-6), yields feriar— freae (13:16) ‘The equations et forth may be readily extended 10 include an eleetromechanical system with any number of electrical and mechanical inputs to any number of cou ling fields. Considering the system shown in Fig. 1.3-5, the energy supplied tothe ‘coupling fields may be expressed as Wy = Oo + Woe asin) me ‘wherein J electrical and K mechanical inputs exist. The total energy supplied tothe ‘coupling fcld atthe electrical inputs is Em [Sone ass)16 ASIC PRINCIPLES FOR ELECTRIC MACHINE ANALYSIS. Wo Was Wa Was 1] counting sens [| ' i ' ' wt i Figure 135 Mule lca and Inechancal inputs ‘The total energy supplied to the coupling field from the mechanical inpus is (1349) (13.20) In differential form we obsin aM, = Deptt ~ Shade (1321) Energy in Coupling Fields ‘Before using (1.3-21) to obtain an expression forthe electromagnetic force fe tis necessary to derive an expression for the energy stored in the coupling fields. Once ‘we have an expression for W;, we can take the total derivative to obtain dW;, which can then be substituted into (1.3-21). When expressing the energy in the coupling fields itis convenient to neglect all losses associated withthe electric and magnetic fields whereupon the fields are assumed to be conservative and the energy stored therein i a function of the state ofthe electrical and mechanical variables. Although the effects of the field losses may be functionally accounted for by appropriately introducing a resistance inthe electric circuit, this refinement is generally not nece~ ssary because the ferromagnetic material is selected and arranged in laminations so as to minimize the hysteresis and eddy curret losses. Moreover, nearly all of the tenergy stored in the coupling fields i stored in the air gaps ofthe electromechanical device. Because air is a conservative medium, al ofthe energy stored therein can beELECTROMECHANICAL ENERGY CONVERSION 17 returned to the electrical or mechanical systems. Therefore, the assumption of loss- Jess coupling fields is not a restrictive as it might fist appear “The energy stored in a conservative field is a function ofthe state of the system variables and not the mannee in which the variables reached that sate. Ie is conve nient to take advantage ofthis feature when developing a mathematical expression forthe field energy. In particular, itis convenient to fix mathematically the position of the mechanical systems associated with the coupling fields and then excite the electrical systems with the displacements of the mechanical systems held fixed. During the excitation ofthe electrical systems, Wi is zero even though electroma: _gnetic or eleetrostatie forces occur. Therefore, with the displacements held fixed the ‘energy stored inthe coupling fields during the excitation ofthe electrical systems is {equal to the energy supplied to the coupling fields by the electrical systems, Thus, With Wig =0, the energy supplied from the electrical system may be expressed from (113-20) as We J Seine (1322) Icis instructive to considera singly excited electromagnetic system similar to that shown in Fig, 13-3, In this case, ey = d/dt and (1,3-22) becomes w= fia (13:23) Here J = 1; however, the subscript is omitted for the sake of brevity. The area vo the left ofthe 2 relationship (shown in Fig. 1.36) fora singly excited electromagnetic device i the area described by (1.3-23) In Fig. 13-6, this area represens the energy stored inthe field at the instant when 2 = 2, and. = iy. The 4~relationship need not be linear; it need only be single-valued, a property that is characteristic to a conser ‘ative or lossless field. Moreover, because the coupling field is conservative, the energy stored in the field with 2 2, and = jg is independent of the excursion ‘of the electrical and mechanieal variables before reaching this state ‘The area tothe right of the Z-i curve is called the coenergy and. is expressed as we= adi (13:24) Which may also be written as i= Wy (1325) Although the coenergy has litle or no physical significance, we will find ita conve- nent quantity for expressing the electromagnetic force. It should be cleat that for & linear magnetic system where the 2~iplos are straight-line relationships, W; = W.18 BASIC PRINCIPLES FOR ELECTRIC MACHINE ANALYSIS at a Figure 13.6 Stored energy and coenergy in magnetic device netic field ofa singly excited elect ‘The displacement x defines completely the influence of the mechanical system ‘upon the coupling field; however, because 4 and i are related, only one is needed in addition toxin order to describe the state of the electromechanical system. There- Tore either J and x or i and x may be selected as independent variables, If and x are selected as independent variables, itis convenient to express the field energy and the ux linkages as W; = Wii,x) (13-26) A= Aisa) (1327) With ‘and » as independent variables we must express din terms of di before sub- stituting into (13-23). Thus from (1.3-27) Alia), O2i a 4 Da (1328) aili,x) In the derivation of an expression forthe energy stored in the fel, dis set equal to zero, Hence, in the evaluation of feld energy, dis equal to the frst tem on theELECTROMECHANICAL ENERGY CONVERSION 19 righthand side of (1.3-28). Substituting into (1.3-23) yields Wylie) = J 216.) f HEA) 0 a | ea (1329) ‘where £ is the dummy variable of integration, Evaluation of (1.3-29) gives the ‘energy stored in the field of the singly excited system. The coenergy in terms of f and x may be evaluated from (1.3-24) as Welix) = Juena = [asnae (1.3.30) With 4 and as independent variables we have We = Wi) (331) i= i@,3) (13.2) ‘The fled energy may be evaluated fom (13-23) a8 Wns) = [ini da= [ tgarae (1333) Tn order to evaluate the coenergy with J and x as independent variables, we need to express di in terms of di; thus from (1.3-32) we obtain ait ila di il2,x) = (1334) Because d= 0 hiro (.8.24 somes mi) = fGen fBEPa aay Fora linear electromagnetic system the 2-i plots are straight-line relationships; thus forthe singly excited system we have (1336) (13.37)20 GASIC PRINCIPLES FOR ELECTRIC MACHINE ANALYSIS Let us evaluate W,(i,2). From (1.3-28), with dx = Aii,x) = Ls) di (1.3.38) Hence, from (1.3-29) wea = [eoyeetaye 3% Us ee en ct (225 Yl ad We) Te le a et rein rig lc Shenton Tas amas aibceny mpetee ares aber ia aubckea i aaancrtcnonn Cheer ree (eae cea cane ne ter cut ys beast a W; Jove, (13-40) Because the coupling fields are considered conservative, (1.3-40) may be evaluated independently of the order in which the ux linkages or currents are brought to their final values. To illustrate the evaluation of (1.3-40) for a multiexcited system, we will allow the currents to establish their final states one at atime while all other cur ‘ents are mathematically fixed in either ther final or unexcited state, This procedure may be illustrated by considering a doubly excited electric system with one mechan ical input. An electromechanical system ofthis type could be constructed by placing a second coil, supplied from a second electrical system, on either the stationary or movable member of the system shown in Fig, 13-3. In this evaluation itis conve- hient to use currents and displacement as the independent variables. Hence, for a doubly excited electric system we have Wylintor)= [ldiliina) Fiediaivins)) (13-40) In this determination of an expression for W;, the mechanical displacement is held ‘constant (dx = 0); thus (1.3-41) becomes wie = faa a « Md] taints), Bisinins) , PEG gy aC gy] (3.97ELECTRONECHANICAL ENERGY CONVERSION 21 ‘We will evaluate the energy stored in the field by employing (1.3-42) twice. First we ‘will mathematically bring the current i to the desired value while holding iy at zero, ‘Thus, i is the variable of integration and diz = 0. Energy is supplied tothe coupling field from the source connected to coil 1. As the second evaluation of (1.3-42), fps brought to its desired! current while holding iy at its desired value. Hence, fs isthe variable of integration and di, = O, During this time, energy is supplied from both sources tothe eoupling field because id is nonzero, The total energy stored inthe ‘coupling field is the sum of the two evaluations. Following this two-step procedure the evaluation of (1.3-42) for the total field energy becomes Minin) gy , Willits) fe a ao] 3.) which shouldbe writen as = fh -Baalbina) Weis) = [eM ag HT Odulinsdsx) 99 , , BlalisEs) : wf fag ae Malet ae] a. ‘The frst integral on the right-hand side of (1.343) or (J.3-44) results from the first step of the evaluation with jas the variable of integration and with i = 0 and diz =0. The second integral comes from the second step of the evaluation with fy = fydly = Oand is asthe variable of integration. tis clear that the order of allow ing the currents to reach ther final state is irelevant; that is, as our first step, we could have made és the variable of integration while holding i at zero (dy and then let; become the variable of integration while holding gaits final variable, ‘The results would be the same. Ii also clear that for three electrical inputs the eva uation procedure would require three steps, one foreach current to be brought math ematically to is final state Let us now evaluate the energy stored in a magnetically linear electromechanical system with two electrical inputs and one mechanical input. For ths let An(iisas4) = La lain + Laalelin (13-48) Jallnsla) = Layla)in + Laalala (13-46) With that mechanical displacement held constant (lx = 0), we obtain dyin a.) = Lay (a) di + Lint) dir (347) dali iy) = Lralx) dy + Lanls) di (1348)22 BASIC PRINCIPLES FOR ELECTAIC MACHINE ANALYSIS eis clear thatthe coefficients on the right-hand sie of (13-47) and (1.348) are the panial derivatives, For example, (1) i the parial derivative of Zy (ff) with respect oj. Appropriate substitution ito (13-44) gives Waive) = [tala +f tata) rentals (1349) which yields Wolivis) =ZLvG)A + enti +hlaw® —— (13-50) ‘The extension to linear electromagnetic system with J electrical inputs is straight- ‘orward whereupon the following expression forthe total field energy is obtained: leh (asi) Wis It's left to the reader to show that the equivalent of (1.3-22) for a multexcited electrostatic system is w= [Sat (as) Graphical Interpretation of Energy Conversion ‘Before proceeding othe derivation of expressions forthe electromagnetic force, itis instructive to consider briefly a graphical interpretation of the energy conversion process. For this purpose let us again refer to the elementary system shown in Fig, 1.3-3 and let us assume that as the movable member moves from x= x, 10 x=, where. < Xy the 2 characteristics are given by Fig. 1.3-7. Let us further assume that as the member moves from x, to 5 the Zi trajectory moves from point Ao point B. Its clear that the exact trajectory from A to Bis determined by the ‘combined dynamics of the electrical and mechanical systems. Now, the area (OACO represents the original energy stored infleld; area OBDO represents the final ‘energy stored in th field, Therefore, the change infield energy is AW; = area OBDO ~ area OACO (13-53) ‘The change in W., denoted as AW,, is aw,= |" idi =a case (1354)ELECTROMECHANICAL ENERGY CONVERSION 23, ° Figure 13-7 Graphical representation of eletromechnical energy conversion for J path MOB. ‘We know that AW, = AW; — AW, (1355) Hence, ‘AW = area OBDO ~ atea OACO ~ area CABDC = ~area OABO _(1.3-56) ‘The change in Wy, denoted as AW, is negative; energy has been supplied to the mechanical system from the coupling field part of which came from the energy stored inthe field and part from the electrical system. Ifthe member is now moved back to 4, the / trajectory may be as shown in Fig. 1.3-8. Hence the AWy i still area OABO but itis positive, which means that energy was supplied from the ‘mechanical system to the coupling field, part of which is stored in the field and part of which is transferred to the electrical system. The net AW, for the eyele from A 10 B back to A is the shaded area shown in Fig. 1.39, Because AW; is 2210 for this eycle AW, = AW, (ssn) For the cycle shown the net AIV, is negative, thus AWy, is positive; we have gen erator action. Ifthe trajectory had been in the counterclockwise ditection, the net AW, would have been positive and the net AW, would have been negative, which ‘would represent motor action,24 BASIC PRINCIPLES FOR ELECTRIC MACHINE ANALYSIS. o 7 Figure 13-8 Graphical representation ofelctomechanical energy conversion for 2 path Bios ° —7 Figure 1.89 Graphical representation ofcloctromechanical energy conversion for 2~ path Ato B tod,ELECTROMECHANICAL ENERGY CONVERSION 25 Electromagnetic and Electrostatic Forces ‘The energy balance relationships given by (1.3-21) may be arranged as Shady = Sepia aw, (1358) In order to obtain an expression for fi, it is necessary t first express Wy and then take its total derivative. One is tempted to substitute the integrand of (1,3-22) into (13-58) for the infinitesimal change of field energy. This procedure is, of course, incorrect because the integrand of (1.3-22) was obtained with all mechanical displacements held fixed (ds, = 0), where the total differential of the field energy’ is required in (13-58). ‘The force or torque in any electromechanical system may be evaluated by ‘employing (1.3-58). In many respects, one gains a much better understanding of the energy conversion process of a particular system by starting the derivation of the force or torgue expressions with (1.3-58) rather than selecting a relationship from a table. However, for the sake of completeness, derivation of the force equa tions will beset forth and tabulated for electromechanical systems with K mechan- {cal inputs and J electrical inputs (2. For an electromagnetic system, (1.3-58) may be writen as Shade Siw, aw, (13-59) With i and x selected as independent variables we have Wy = Wisi stot) (13.60) in obs oo 8) (13-61) From (1.3-60) and (1.3-61) we obtain a =F 2H SIM) g,—3 Dify.n) Aifiom) aay = ad 5 Y ai = y aly + y Mt) ay (1363) Jn (13-62) and (1.3463) and hereafter inthis development the functional notation of| (is. sia,-—-.2e) is abbreviated a8 (jy). The index misused 0 a8 to avoid confusion withthe index j because each di, must be evaluated for changes in all curents in order to account for mutual coupling between electrical systems.26 BASIC PRINCIPLES FOR ELECTRIC MACHINE ANALYSIS. [Recall that we did this in (1.3-42) for J = 2] Substituting (1.3-62) and (13-63) ito (1.3-59) yields Y fala (13-64) By gathering terms, we obtain y fain) = > ee in isi fis) ty a gia) se Saliva) = Yo es) Gen) (13-66) i (i307) Although 1.67 js fit rata! impoance (3-65) can be use ev site fc ath hha oil of anemia! tm with nly magni euping lian with and sete independent wares A secon foe eaaton wth ands independet variables maybe obtained rom (1369) incorporating he expression for cents. Fora mulexcted sem he comer maybe expend x w= Daa Wy (13.68) Because iy and x; are independent variables, the partial derivative with respect o.xis Wey) Aina) _ ON ylinm) ts Pili.) _ OW.) (1360) Os Os Hence, substituting (13-69) into (13-66) yields amen Saline) = Mes) (13-70)ELECTROMECHANICAL ENERGY CONVERSION 27 Table 131 Electromagnetic Force at kth Mechanical Input” sans = fi awl Sf to 2h on ou, Fer oaional yeas replace fa with Tan wih 1 should be recalled that positive fg and postive dy are in the same if the magnetic system i linear, then W, = W By a procedure similar to that used above, force equations may be developed for ‘magnetic coupling with 2, and.x, as independent variables. These relations are given n Table 1.3-1 without proof. In Table 1,3-1 the independent variables to be used are designated in each equation by the abbreviated functional notation, Although only {ranslational mechanical systems have been considered, all force relationships {developed herein may be modified for the purpose of evaluating the torque in rota- tional systems. In particular, when considering a rotational system, fis replaced ‘with the electromagnetic torque Toa, and is replaced with the angular displace- rent Oj. These substitutions are justified because the change of mechanical energy ina rotaional system is expressed as AWya = Tar dl, ‘The force equation for an clectromechanical system with electric coupling fields may be derived by following a procedure similar to that used in the ease of magnetic coupling fields. These relationships are given in Table 1.3-2 without explanation. tis instructive to drive the expression forthe electomagnetic force ofa singly excited electrical system as shown in Fg. 1.33. Its clea thatthe expressions given in Table 13-1 are valid for magnetically linear or nonlinear systems. If we assume thatthe magnetic system is linea, then Ai,x) is expressed by (1.336) and W(x) is expressed by (13-39), whichis also equa othe coenergy: Hence, either the fis or second entry of Tale .3-1 can be used to expres In paticular, aWali ‘ox L(x)2B BASIC PRINCIPLES FOR ELECTRIC MACHINE ANALYSIS ‘able 13-2. Hlectrostatie Force at hth Mechanical Input” feslepa El» Sul) een) faleosX) ic) falayn) = — Me) fala rf Megan) Aa.) Foe atonal pers epic Wh Tad WB With the convention established, a positive electromagnetic force is assumed 10 ‘act in the direction of increasing 2. Thus with (1.3-15) expressed in differential orm as Wy = fede (13-73) \we see that energy is supplied to the coupling field from the mechanical system when {and dt are opposite in sign, and energy is supplied to the mechanical system from, the coupling field when f, and dx are the same in sign. ‘From (1.3-72) it is apparent that when the change of L(x) with respect to x is negative, f- is negative. In the electromechanical system shown in Fig. 1.3-3 the change Lx) with respect tox is always negative; therefore, the electromagnetic force is in the direction so as to pull the movable member to the stationary member. In ‘other words, an electromagnetic force is set up so as to maximize the inductance of the coupling system, or, since reluctance is inversely proportional to the induc- tance, the force tends to minimize the reluctance. Because fis always negative in the system shown in Fig. 13-3, energy is supplied to the coupling field from the ‘mechanical system (generator action) when dis postive and is supplied from the ‘coupling field to the mechanical system (motor action) when dr is negative. ‘Steady-State and Dynamic Performance of an Electromechanical System Itis instructive to consider the steady-state and dynamic performance ofthe elemen- tary electromagnetic system shown in Fig. 1.3-3. The differential equations that describe this system are given by (J.3-7) for the electrical system and by (13-8) for the mechanical system. The electromagnetic force f. is expressed by (I.3-72). Ifthe applied voltage, v, and the applied mechanical force, f, are constant, all,ELECTROMECHANICAL ENERGY CONVERSION 29 derivatives with respect to time are zero during steady-state operation, and the beha- vior can be predicted by ven (13-14) f= Kie-m) fe (13-75) Equation (13-75) may be writen as “fe=f- Kix) (13-76) The magnetic core ofthe system in Fig. 1-3 is generally constructed of ero magnetic material witha relative permeability onthe order of 2000 to 4000, In this ‘ase the induetance (x) can be adequately approximated by way =t am In the actual system the inductance wil bea large finite value rather than infinity, as predicted by (13-77), when x = 0. Nevertheless, (13-77) is quite sufficient to illus- trate the action of the system for x > 0. Substituting (13-77) into (1.3-72) yields os) wom plot of (1.3-76), with f, replaced by (1.3-78), is shown in Fig. 1.310 for the following system parameters: r= 109 K = 2667Nim In Fig .3-10, the plo ofthe negative ofthe electromagnetic force is for an applied voltage of 5 V and a steady-state current of 0.5 A. The straight lines represent the right-hand side of (1.3-76) with f = 0 (lower straight line) and f =4N (upper straight Ine). Both lines intersect the ~f. curve at two points. In particular, the upper Tine imersects the —f, cutve at 1 and I’; dhe lower line intersects at 2 and 2. Stable ‘opetation occurs at only points I and 2. The system will not operate stably at points and 2. This can be explained by assuming that the system is operating at one of these points ("and 2) and then showing that any system disturbance whatsoever ‘ill cause the system to move away from these points. If, for example, x increases Slightly from its value corresponding to point, the restraining force f — K(x ~ xy) is larger in magnitude than —f., and x will continue to increase until the system reaches operating point I. Ifxincreases beyond its value corresponding to operating point 1, the restraining force is less than the electromagnetic force. Therefore, the system will establish steady-state operation at I. If, on the other hand, x decreases30 BASIC PRINCIPLES FOR ELECTRIC MACHINE ANALYSIS. Force, N Figure 13-10 Steady-state operation of electromechanical system shown in Fig. 13-3 from point 1 the electromagnetic force is larger than the restraining force. There- fore, the movable member will move until it comes in contact with the stationary ‘member (x = 0). Iis clear that the restraining Force that yields a straight line below the ~f. curve will not permit stable operation with x > 0. ‘The dynamic behavior of the system during step changes in the source voltage vis, shown in Fig, 13-11, and Figs. 13-12 and 1.3-13 show dynamic behavior during step changes in the applied force f. The following system parameters were used in addition to those given previously: 0 M=0055kg D N-sim "The computer traces shown in Fig. 13-1] depict the dynamic performance of the example system when the applied voltage is stepped from zez0 to 5 V and then ‘back to zer0 with the applied mechanical force held equal to zero. The system vari- ables ae ey i fo.%, We, Wy, and Wy. The energies are plotted in milljoules (ml). Initially the mechanical system is at rest with x= xy (3 mm). When the sourcew, Was ah ELECTROMECHANICAL ENERGY CONVERSION 91 5 Source votage decreed h from sw 0 —— : Sous vlogs insead i Homoio st A eons {—~ ———— ol 4 Figure 13-11 Dynamic performance ofthe elacuomechanical sytem shown in Fig, 133 uring sep changes inthe source voltage. \oltage is applied, x decreases; and when steady-state operation is reestablished, 8 approximately 2.5 mm, Energy enters the coupling field via I. The bulk ofthis energy stored inthe field (W)) with smaller amount wansfered othe mechanical system, some of which is dissipated in the damper during the tansient period wlile the remain is stored inthe spring When the applied voltage is removed, the ele: trical and ‘mechanical systems retum to their original states. The change in Wy82 BASIC PRINCIPLES FOR ELECTRIC MACHINE ANALYSIS. as; ev -asl on Wor oo ba N al serait Fore dec Wem) = W,mt se Worms — -4 igure 13-12 Dynamic peformance ofthe electromechanical system shown in Fig. 13.3, uring step changes inthe applied force i small, increasing only slightly. Hence, during the transient period there is an interchange of energy between the spring. and mass which is finally dissipated in the damper. The net change in W; during the application and removal ofthe applied voltage is zero; hence the net change in W. is positive and equal tothe negative ofthe net change in Wy. The energy transfered to the mechanical systom ducing this eycle is dissipated in the damper, because fis fixed at zero, and the mechanical system. returns tits initial est position with zero energy stored inthe spring. In Fig. 13-12, the initial state is that shown in Fig. 1.3-11 with 5 V applied to the electrical system. The mechanical force fis increased from zero to 4 N, whereupon ‘energy enters the coupling field from the mechanical system, Energy is wansfetredELECTROMECHANICAL ENERGY CONVERSION 38 0.004 oi a as ass ia [igure 1.343 System response shown in Fig. 1.33 from the coupling field to the electrical system, some coming from the mechanical system and some from the energy originally stored in the magnetic field. Next the force is stepped back to zero from 4 N. The electrical and mechanical systems return to their original states. During the eycle a net energy has been transferred from the ‘mechanical system tothe electrical system which is dissipated in the resistance. This ceyele is depicted on the 2-i plot shown in Fig, 1.3-13 Example 1B [cis helpful to formulate an expression forthe electromagnetic ‘orque of the elementary rotational device shown in Fig. 1B-I. This device consists of two conductors. Conductor 1 is placed on the stationary member (stator; conductor 2 is fixed on the rotating member (rotor). The crossed Tnes Inside a circle indicate thatthe assumed direction of positive current ow is into te paper (we are seeing the tai of the arrow), whereas a dat inside acirele indicates positive current flow is out ofthe paper (the point ofthe aerow). The length ofthe ar gap between te stator and rotor is shown exaggerated relative {o the inside diameter of the stator34 BASIC PRINCIPLES FOR ELECTRIC MACHINE ANALYSIS. Stator o cy Figure 18-1. Elementary rotational eletromechanical device. (a) End views (b) eros sectional view. ‘The voltage equations may be written as iy wean + (1B) ” (18.2) Where rj and ry are the resistances of conductor 1 and 2, respectively. The magnetic system is assumed linear; therefore the flux linkages may be expressed as Ay Enh bLake (18:3) d= Layhh + Laake (18-4) ‘The selfinductances Ly, and Las are constant. Let us assume thatthe mutual inductance may be approximated by Liz = La) = Mos 6, (Bs) ‘where 0, is defined in Fig. 1B-1. The reader should be able to justify the form of (IB-5) by considering the mutual coupling between the two conductors as (4, vaties from 0 to 2x radians. Weis) Tal sag le) = ay (18-6) Because the magnetic system is assumed tobe linear, we have Wel ies0.) = lai + Linh + Sat (1B) Substituting into (1B-6) yields Te= ~iieMsind, (1B8)Figure 18-2. Stator and rotor poles for constant currents Consider for & moment the orm of the torque if and iy ae both constant. FFor the positive direction of current shown, the torgue is of the form T. = ~Ksind, (1B.9) where K is positive constant. We can visualize the production of torque by considering the interaction of the magnetic poles produced by the current flowing in the conductors If both i and f; ate positive, the poles produced are as shown in Fig. 1B-2. One should recall that flux issues from a north pole. Also, the stator and rotor each must he considered as separate electro- ‘magnetic systems. Thus, lux produced by the =I" winding issues from the north pole of the stator into the airgap. Similarly, the flux produced by the 2-2' winding enters the air gap from the north pole of the rotor tis left 0 the reader to justify the fact that the range of 6, over which stable operation ‘can occur for the expression of electromagnetic torque given by (1-9) is 2/258, < 1/2, 1.4 MACHINE WINDINGS AND AIR-GAP MMF For the purpose of discussing winding configurations in rotating machines and the resulting air-gap MMF as well as the calculations of machine inductances, it is con- venient to begin with the elementary 2:pole, 3-phase, wye-connected salient-pole synchronous machine shown in Fig. 1.4-1. Once these concepis are established for this type of a machine, they may be readily modified to account for all types of induction machines and easily extended to include the synchronous machi \with short-cicuited windings on the rotor (damper windings). The stator windings of the synchronous machine are embedded in slots around the inside circumference of the stationary member. In the 2-pole machine, each phase winding of the 3-phase stator winding is displaced 120° with respect to each other a illustrated in Fig. 14-1. The field orf! winding is wound on the rota ing member. The as, bs, es, and faxes denote the positive direction of the flux peo duced by each of the windings. The as, bs, and cx windings are identical in that each ‘winding has the same resistance and the same numberof tums. When a machine has thre identical stator windings arranged as shown in Fig, 1.41, iis often referred to36 GASIC PRINCIPLES FOR ELECTRIC MACHINE ANALYSIS Figure 14-1. Elementary, 2-0 ‘machine Sephase, wye-comected saliene-pole synchronousACHINE WINDINGS AND AIR-GAP MME 37 as a machine with symmetrical stator windings, We will find thatthe symmetr- cal induction machine has identical multiphase stator windings and identical mul phase rotor windings. An unsymmetrical induction machine has nonidentical ‘multiphase stator windings (generally 2-phase) and symmetrical multiphase rotor windings In Fig. 14-1, tis assumed that each eo spans x radians ofthe stator fora 2-pole ‘machine. One side of the col (oil side) is represented by a © indicating thatthe assumed positive direction of current is down the length ofthe stator (nto the paper) ‘The © indicates that the assumed positive direction of current is out of the paper. ach coil contains n- conductors. Therefor, in the case of the as winding, postive ‘current flows in a conductor of coil ay, then axially down the length ofthe stator and ‘back the length of the stator and out at coil side a. This is repeated for n, conduc tors. The last conductor of the coil ai is then placed in the appropriate slot so as {start coil a>-<, wherein the eurrent flows down the stator via coll side a> and then ‘back through a and so on until a. The bs and cs windings are arranged similarly, andthe last conductors of coil sides, Hand care connected together to form the ‘wye-connected stator. The end tums (looping ofthe cols) at both ends of the stator soas to achieve the span of radians are not shown in Fig. 14-1. As mentioned, each coil consists of n. conductors, each of which makes up an individual single conduc- {or coil within the main coil. Thus the numberof turns of each winding is determined by the product of n, and the number of coils or the product of n- and the number of coil sides carrying current in the same direction. In the case of the fd winding, each coil (fi=f), for example) consists of ny conductors. It should be mentioned that in Example IB, the stator and rotor coils each consisted of only one coil side with ‘one conductor (7, = 1) in each coil side. ‘One must realize that the winding configuration shown in Fig. 1.4-1 is an over: simplification ofa practical machine. The col sides of each phase winding are con sidered to be distributed uniformly over 60° ofthe stator circumference, Generally, the coil sides of each phase are distributed over a larger area, perhaps as much as 120°, in which ease it is necessary for some of the coil sides of 1Wo ofthe phase ‘windings to occupy the same slot. In some cases the coil sides may not be distrib ‘uted uniformly over the part of the circumference that it occupies. For example, it ‘would not be uncommon, in the case of the machine shown in Fig. 14-1, to have ‘more tum in col sides ay and as than in ay and a. (Similarly forthe bs and es winnd- ings.) We will find that this winding arrangement produces an air-gap MME which ‘more closely approximates a sinusoidal air-gap MMF with respect to the angular displacement about the air gap. Another practical consideration is the so-called fractional-pitch winding. The windings shown in Fig, 14-1 span x radians for the 2-pole machine. This is refered to as full-ptch winding. In onder o reduce voltage and current harmonies, the windings are often wound so that they span slighty less ‘than x radians fora 2-pole machine. Ths is refered to a a fractional-pitch winding, All of the above-mentioned practical variations from the winding arrangement shown in Fig. 14-1 are very important to the machine designer; however, these fea tures are of less importance in machine analysis, where in most cases a simplified approximation ofthe winding arrangement is sufficient,38 BASIC PRINCIPLES FOR ELECTRIC MACHINE ANALYSIS A salient-pole synchronous machine is selected for consideration because the analysis ofthis type of machine may be easily modified to xecount for other machine types. However, a salient-pole synchronous machine would seldom be a 2-pole ‘machine except in the case of small reluctance machines which are ofthe synchro- nous class but which do not have a field winding. Generally 2-and 4-pole machines are round-rotor machines with the field winding embedded ina solid steel (nonlam- nated) rotor. Salient-pole machines generally have a large number of poles com- posed of laminated steel whereupon the field winding is wound around the poles ‘Similar to that shown i Fig. [4-1 For the purposes of deriving an expression for the air-gap MMF, itis convenient to employ the so-called developed diagram of the cross-sectional view of the ‘machine shown in Fig. 14-1. The developed diagram is shown in Fig. 1-2, The length of the air gap between the stator and rotor is exaggerated in Figs. 14-1 and 1.4-2 for clarity. The fact thatthe air-gap length is small relative to the inside diameter of the stator permits us to employ the developed diagram for analysis pur- poses. In order to relate the developed diagram to the cross-sectional view of the ‘machine, it i helpful to define a displacement to the left of the origin as postive. The angular displacement along the stator circumference is denoted ¢, and along the rotor circumference. The angular velocity of the rotor is, and 8, isthe angular displacement of the rotor. Fora given angular displacement relative tothe as axis we can relate $,, and 8, a5 b= 6,48, say (Ouranalysis may be simplified by considering only one ofthe stator windings at a time. Figure 1.4-3 is a repeat of Figs. 14-1 and 14-2 with only the as winding shown. Due to the high permeability ofthe stator and rotor steel the magnetic fields cons begs aranis fais KKM a gee a Figure 1442 Development ofthe mache shown in Fig. 1.