Introduction to Generalized Electrical Machine Theory. D. O'Kelly and S. Simmons Lecturer and Senior Lecturer, University of Bradford McGRAW-HILL Publishing Company Limited LONDON - NewYork - Toronto - Sydney shPublished by McGRAW-HILL Publishing Company Limited MAIDENHEAD - BERKSHIRE - ENGLAND 94050 Copyright © 1968 McGraw-Hill Publishing Company Limited. Ail rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior permission of McGraw-Hill Publishing Company Limited PRINTED AND BOUND IN GREAT BRITAINPreface A traditional course of study in electrical machines involves a con- siderable amount of physical and analytical detail which tends to obscure the fundamental concepts of machine operation. The procedure is piecemeal, a separate steady-state analysis being made for transformers and the more important types of rotating machines. With this approach the important aspect of transient machine performance is often neglected. The generalized machine theory which avoids many of the difficulties inherent in the traditional approach has been gaining favour in recent years for undergraduate courses in machines. The theory is based on original work by G. Kron and R.H. Park. This work enables most conventional machines to be represented, after suitable transformation, as part of a two-axis primitive machine known as the Kron Primitive. The performance equations deter- mined for the Kron Primitive may therefore be employed for the transient and steady-state analysis of most conventional machines. The contents of this book, based on undergraduate lectures given by the authors, is an introduction to the generalized theory of machines. The mathematics employed is of undergraduate level and wherever possible mathematical analysis is reinforced with physical concepts. The authors gratefully acknowledge the assistance of Miss D. M. Hart and Miss M. Broughton in typing the manuscript, and also that of the University of Bradford in granting permission to publish questions from past examination papers. D. O’Kelly S. SimmonsContents Preface Glossary of symbels and abbreviations Chapter 1 Conventions and basic principies 1.1 Complexors and phasors 1.2. Power 1.3 Units 1.4 Theorem of constant flux-linkages 1.5 Conventions 1.6 Primitive network Chapter 2 Coupled circuits and the two-winding transformer 2.1 Coupled circuits 2.2 The air-cored two-winding transformer 2.3. The iron-cored transformer PROBLEMS Chapter 3 Electro-mechanical energy conversion 3.1 Electro-mechanical transducers 3.2 Singly-excited system 3.3 Doubly-excited system PROBLEMS Chapter 4 Basic electrical machines 4.1 Fundamental machine features 4.2. Windings 4.3. Types of rotating machines PROBLEMSChapter 5 Elements of generalized theory 5.1 Simplifying assumptions 5.2 Concentrated and distributed winding inductances 5.3. Concentrated coi] representation for a rotational transducer 5.4 Commutator machines 5.5 The primitive multi-coil machine PROBLEMS Chapter 6 D.C. and crossfield machines 6.1 General analysis 6.2 D.C. machine steady-state analysis 6.3 Sudden short-circuit of a d.c. generator 6.4 D.C. motor transients 6.5 Crossfield machines PROBLEMS Chapter 7 Three-phase winding transformations 7.1 Transformation between three-phase (abc) and two- phase («By) windings 7.2. Transformation between three-phase (abc) and two- phase (dqy) windings 7.3. The polyphase magnetic field Chapter 8 Three-phase synchronous machine 8.1 Introduction to machine analysis 8.2 Idealized machine inductances 8.3 Steady-state analysis (balanced conditions) 8.4 Transient analysis PROBLEMS Chapter 9 Three-phase induction machine 9.1 Transformation to the primitive machine 9.2 Steady-state analysis 9.3 Steady-state torque analysis 9.4 Double-cage rotor 9.5 Machine inductances 9.6 Transient analysis PROBLEMS 91 91 91 105 112 118 123 125 125 126 137 142 150 157 162 163 174 179 187 189 190 203 220 247 248 252 266 272 272 277CONTENTS Chapter 10 Single-phase motors 10.1 Single-phase commutator motors 10.2 Single-phase induction motor PROBLEMS Bibliography Index 281 281 285 291 294 295Glossary of symbols and abbreviations Symbols PPB PARRA s PsP zBPEHrorm = & a AUB -s flux density capacitance damping coefficient co-ordinate axes magneto-motive force (m.m.f.) coefficient of rotational inductance magnetizing force steady-state value of current phasor current complexor current crest value of alternating current instantaneous value of current polar moment of inertia coefficient of coupling, coupling factor inductance mutual inductance co-ordinate axes number of turns turns ratio electrical power mechanical power reactive volt-ampere resistance volt-ampere phasor volt-ampere fractional slip electrical torque mechanical torque timeGreek x, B Aqsa rEEA Sons SES GLOSSARY OF SYMBOLS steady-state value of voltage phasor voltage complexor voltage crest value of alternating voltage instantaneous value of voltage electrical energy mechanical energy reactance co-ordinate axes impedance phasor impedance letters co-ordinate axes torque angle elemental quantity angle (electrical radian) angle (mechanical radian) permeance permeability magnetic space constant time constant magnetic flux phase angle flux-linkage leakage coefficient angular frequency of supply angular frequency of rotation (electrical rad/s) angular frequency of rotation (mechanical rad/s) Mathematical symbols the matrix A conjugate of the matrix A inverse of the matrix A transpose of matrix A real part ofimaginary part of Laplace transform of v initial value of V d/dt Laplace operator v(-)1. Conventions and basic principles This chapter contains essential material required for the analysis of electrical networks and machines. 1.1 Complexors and phasors The complexor is a convenient mathematical tool for representing the instantaneous value of a steady-state sinusoidal alternating quantity. A complexor is shown in Fig. 1.1 asa modulus /,, rotating anticlockwise with uniform angular velocity w equal to the angular frequency of the alternating quantity. Stationary orthogonal xy- y-axis magnitude I, X-axis Fig. 1.1 The complexor {= /,e""*”” axes define a stationary space-frame and with respect to the x-axis the modulus /,, is displaced an instantaneous angle (w/+ a). Instantaneous xy-axis components i, = I, cos (wt+a) (Ay i, = 1, Sin (t+)2 GENERALIZED ELECTRICAL MACHINE THEORY A complexor is a time-dependent complex quantity defined as 1 =i,+/;i, = 1,[cos (wt +a) +/ sin (wt +a)] T= Jeo (1.2) The instantaneous value of a sinusoidal alternating current of crest value /,, and supply angular frequency w may be expressed in one of the following ways, depending on the instantaneous value being a cos or sin function. Rel or Im] = i, = 1, sin (wt +a) = I,, cos (wt+a) (1.3) i where Rel and Im I denote, respectively, the real and imaginary parts y-axis neaxis 7 o . maxis Phasor I with respect to m,n axes x-axis 1.2. The phasor I= Je” of thecomplexor I. Similar expressions can be written far sinusoidal alternating voltages. In partially expanded form Te v4 oe a.4) The time-invariant complex quantity shown in brackets in (1.4) is called a phasor. Phasor I = Ie (1.5) where the r.m.s. current = Im “5 As shown in Fig. 1.2, orthogonal m, n-axes define a space-frame which rotates in synchronism with I and I. With respect to the m-axis, I maintains a fixed angular displacement. I (1.6)CONVENTIONS AND BASIC PRINCIPLES 3 Analysis of circuit response to steady-state sinusoidal alternating currents and voltages of common frequency is simplified by the use of complexors and phasors. Consider, for example, a series R, L, and C circuit with an instantaneous current i= 7, sin (wt+a) The corresponding complexor current T= /2Ie™ (1.7) Operational circuit impedance 1 Zp) = R+Lp+— 1.8 (p) P+ Ge (1.8) d he == id = \d where ea) oo { t Complexor voltage applied to the circuit V=Zp)i 1 ot = (Retp+a,\va Te! «a9 = : 1 v= (a+woe+cal (1.10) For steady-state sinusoidal conditions the impedance Z(jw) is a complex quantity obtained by substituting jw for p in the opera- tional impedance Z(p). P ; j = R+joL-—+ ZU) joL C = R+j(X_~Xo) ab Inductive reactance =X, = wL ohm i | $ 3 Capacitive reactance X¢ V = ZUo)i (1.12)4 GENERALIZED ELECTRICAL MACHINE THEORY ana Referring to Fig. 1.3, Z(jo) is représented as a modulus Z and an argument ¢ for the two conditions X,>X_ and X, Xc¢, isa positive angle, and X,< Xe, isa negative angle Ro GX, Xe Fig. 1.3. Impedance Z(jw) = Ze’* Substituting (1.7) and (1.13) in (1.12) V = Zei*/2 Tei = J2 (Zle)ei (1.14) The time-invariant complex quantity shown in brackets in (1.14) is a phasor voltage V = Zle'* (1.15) From a point of observation within the m, n-axis space-frame I and V, given by (1.6) and (1.15), are time-invariant, i.e., stationary,CONVENTIONS AND BASIC PRINCIPLES 5 and phase-separated by an angle y. Referring to Fig. 1.4, a phasor diagram is shown for the following conditions X,>Xc — Tlagging V Xp < Xe I leading V Phasors are employed for steady-state a.c. circuit analysis as the analysis is concerned with the relative phase separation of currents and voltages. In addition, mean or average circuit power will be neaxis _. X,X,), sing and Q are positive quantities and the load imports power and reactive volt- amperes from the supply. For a load of leading power factor (X, 0 which is equivalent to stating that the flux- linkages cannot change in zero time. The same conclusion can be deduced from (1.38) since a step- change of ¥, no matter how small in magnitude, would result in an infinite value for (v—iR). An interesting example is the sudden removal of a closed loop of wire from a magnetic field. To maintain the flux-linkage constant at the value existing at the instant of removal, a current is induced in the loop of wire. Due to resistance, this current and its associated flux decays with time.CONVENTIONS AND BASIC PRINCIPLES il As a further example, consider two identical mutually-coupled coils, one short-circuited and the other excited with current i. It is assumed that the flux-linkage of both coils is the same. Following v—iR _ area AY at : —> time Fig. 1.6 Incremental flux-linkage switching of i to zero, a current is instantaneously induced in the short-circuited coil. To maintain the flux-linkage momentarily unchanged immediately following switching, the initial value of induced current in the short-circuited coil is i. 1.5 Conventions In this section conventions are established for indicating the positive sense of action of various electrical and mechanical quantities sub- sequently employed for circuit and machine analysis. 1.5.1 Electrical and mechanical quantities (a) Voltage, current, and power Figure 1.7 represents a single-port energy device connected through an external resistance R to a voltage Power flow P, Energy device Fig. 1.7 Positive sense for voltage, current, and power source which raises the potential of terminal 1 with respect to éerminal 2 by a positive amount r. _A rise in potential between two points in a Circuit is indicated by an arrow.GENERALIZED ELECTRICAL MACHINE THEORY Electrical power P, flowing into ports is assumed positive. This establishes the positive direction for current flow iin Fig. 1.7 as being from terminal | to terminal 3. r ne Oy and Oy Fig. 1.8 Positive sense‘for torque and mechanical rotation (b) Torque and mechanical rotation Mechanical angular rotation ¢,,and angular velocity w,, are positive ina clockwise sense. Torque is assumed positive when acting in the sense of positive rotation, see Fig. 1.8. Mechanical power Pe eo With T,, and @,, positive quantities, P,, is positive and corresponds to power input to the shaft. Thus power flow into mechanical and electrical energy ports is positive. Considering motors and generators as two-port energy devices, electrical power P, flows into the energy port formed by the machine “terminals and mechanical power P,, flows into the energy port formed by the shaft, see Fig. 1.9. Terminals Shaft P. o— . Py cco Machine Q- 4 7 Electrical Mechanical energy port energy port Fig. 1.9 Two-port energy device 1.5.2. Dot convention For the study of magnetically coupled circuits information is required as to the relative sense of winding of individual coil pairs along a magnetic path linking both coils. TheCONVENTIONS AND BASIC PRINCIPLES 13 dot convention is a convenient method of providing a coil diagram with this information. An appropriate end of each coil is marked with a dot and the order in which the dots arise, when traversing the common magnetic path, is indicative of the relative sense of coil winding. Referring to Fig. 1.10 for symmetrical dot arising /7blanky, dot dot \ blank / coils are wound in the same sense, and with unsymmetrical dot arising 77 dot dot blank \olank coils are wound in an opposite sense. If both currents are shown simultaneously entering or leaving at dots the coil m.m.f.s and fluxes act in the same sense. Conversely, with one current entering and the other current leaving at a dot, the m.m.f.s and fluxes act in an opposite sense. ir blank oo — dot dot \ \\ blank “ : / coils wound ae ae \— \ in the same sense common magnetic path Lo ao poccccc -< dot blank i \ \ blank — _ _ coils wound in an opposite sense Fig. 1.10 Dot conventionis . GENERALIZED ELECTRICAL MACHINE THEORY: 1.6 Primitive network The analysis of an interconnected network may be rationalized by representing such a network in a modified form known as the primitive network. For the primitive network, interconnections are deleted. impressed voltages are shown with their positive potentials driving currents into the ends of coils marked with dots. The voltage equations for the primitive network, expressed in terms of i, OR, . vy M, 1 ; ; { Mo Ry Re ®, ®, o; Actual circuit Primitive network Fig. 1.11 The primitive network voltage, current, and flux of positive sense are completely general as they are independent of circuit connections. The primitive voltage equations may be used to provide actual circuit voltage equations for any arrangement of interconnections. This is achieved by relating primitive and actual circuit values followed by the substitution of these values in the primitive voltage equations. Figure 1.11 represents an interconnected coupled circuit and the corresponding primitive network. Leakage fluxes are ignored and related primitive and actual circuit quantities are distinguished by the use of different suffices. Primitive network voltage equations v, = i R, +N, WO, +2) by = 1,Ry+N,p(P, +2)CONVENTIONS AND BASIC PRINCIPLES 15 Relating primitive and actual values a - - =i Thus v, = iR,+N,p(P,— 4) v, = —iR,+N,p(?,—%,) = (R,+R,)+N,p(P, — 9,)—N,p(®, — Py) e2. Coupled circuits and the two-winding transformer In this chapter a study is made of magnetically coupled circuits which are stationary relative to one another and to the surrounding mag- netic medium. The magnetic medium is considered homogeneous and of constant permeability resulting in linear flux/current relation- ships for the coupled circuits and a linear theory describing their magnetic interactions. The linear theory, which is shown later to form the basis of a generalized theory for electrical machines, is employed for the steady- state and transient performance analysis of a two-winding trans- former. 2.1 Coupled circuits Consider a system of n separate circuits magnetically coupled to one another. Circuits are represented as magnetically equivalent con- centrated coils with fluxes linking total turns, i.e., no partial flux- linking. The effective circuit resistance is connected externally in series with the concentrated coil and a time-dependent voltage, applied to the series combination, results in a time-dependent current and flux. Currents are shown entering coils at dots and in accord- ance with the dot convention, described in chapter 1, self- and mutual fluxes link coils in the same sense. With linear conditions, the total flux linking any one coil of the n coil system is the superposition of the self-flux and n— 1 components of mutual flux. To show the significance of these fluxes attention is given to the fluxes linking the Ath coil. Figure 2.1(a) represents the flux distribution with only the kth coil excited. The total self-flux ®,, due to i,, comprises a magnetizing flux ®,,, which links the other coils and a leakage flux @, linking N, turns only. Dy = Opt Prk (2.1)COUPLED CIRCUITS AND TWO-WINDING TRANSFORMER 17 (a) Self-flux (b) Mutual flux Fig. 2.1 Self and mutual flux-linkage Figure 2.1(b) shows the mutual fluxes linking the Ath coil with all coils except the kth coil excited. The mutual flux ®,, comprises n—~1 components jen Om = YO; (2.2) int tk @,; isa mutual flux produced by the jth coil and linking the Ath coil.18 GENERALIZED ELECTRICAL MACHINE THEORY By superposition, the total flux ®, linking the kth coil Oy, = Py t Puy = Py t Om + Pup jen = Pyt Pmt Y Pej (2.3) jm1 ioe Flux-linkages jan Y, = NO = Fut Pmt Yo Pay (2.4) j= jek Expressing the flux-linkage components of (2.4) in terms of magnetic permeance and exciting current (2.5) Punt = NePmx = Ame Nike (2.6) Yay = My = AyNN,i; (2.72) Defining inductances: : 2 Nn Leakage inductance Ly, = Ay,Ng = —~ (2.8) he oe N®, Magnetizing inductance Lny = Apu? = ——™* (2.9) ®, Self-inductance Ly = (An+Am)NZ = Me (2.10) h Mutual inductance between the kth and jth coils NP My; = ANN; a (2.11) i and N,® My = AgNN, = (2.12) a but Any = Ajx thus My = My (2.13) It follows from (2.13) that for any pair of coils there is one value of mutual inductance.COUPLED CIRCUITS AND TWO-WINDING TRANSFORMER 19 Expressing (2.4) in terms of inductance-current products Y= Lyi t s M, ji; (2.14) jet The total flux-linkage for each of the n coils are determined from (2.14) with k=1, 2, 3,....n. In matrix form Y, Ly My. My3 77> My ces Milf i Y, My, Ly Moy 0° My o> Mall Y; M3, M3, Ly ste My ott May | ig ae | a 5 : E : (2.15) oe Myr Miz Maz 7 Ly “+ Men iy Pp Mar Maz Mas 077 Lg 77 Len in (Y= (44) (2.16) 2.1.1 Voltage equations The instantaneous voltage applied to the kth coil d % = Ri+pY, where p= 7 (2.17) For n coils (el = (RIL + C2 0C = ((R]+[Z]aE) (2.18) [R] is a diagonal resistance matrix and [L] is defined by (2.15). Relating [v] and [/] is an operational impedance matrix (Zio) = [R}+ Lele @.19) For sinusoidal voltages and currents p=jw [ZUe)] = [x] +L] (2.20) and for phasor quantities [V] = [2G)][1] (2.21)20 GENERALIZED ELECTRICAL MACHINE THEORY EXAMPLE 2.1 To determine expressions for the voltages v, and applied to the coupled circuit Fig. 2.2. Fig. 2.2. Coupled circuit, example 2.1 Commencing with the primitive circuit Fig. 2.3 ey R, 0 0 [Zi Miz Mis 4 mf=4{0 RR. 0 |+{Mi2 Le Maslpblis tv. 0 0 Rs} Lays Mos Ls is 3 vy = —Ryi,— (Ly —My2)pig— My spin 0, = Roi, +(L2—M42)pi,— Maspip v3 = —R3i,—L3pi,—(M3— M23)pi, [a s" a] |% +L,—2M,2) Morel (M,3—M25) Ls tyCOUPLED CIRCUITS AND TWO-WINDING TRANSFORMER 2 The corresponding phasor equations ‘| _ fey” 0 | vs 0 Ry 10 Oe Maree (Mis— Mas) Ly I, D L A eS, i 3 Fig. 2.3 Primitive circuit, example 2.1 2.1.2 Coefficient of coupling & and leakage coefficient ¢ Consider multiplying together (2.11) and (2.12) > _ NN Puy; Mi ie From (2.10) NM Ly i Oy Similarly, for the jth coil Thus Cr (2.22) The flux ratios represent the fraction of total self-flux which links a 322 GENERALIZED ELECTRICAL MACHINE THEORY secondcoil. Inthe absence of saturation and with a fixed separation between coils these ratios are constant values ® ®, ky = —* and kj = re (2.23) kk ‘i Constants k, and k; are called coupling factors which are said to give a measure of the ‘tightness’ of coupling between coil pairs. With decreasing separation the coupling factors increase in value and have a (theoretical) maximum value of unity. Substituting (2.23) in (2.22) Mi = Kk L ily Mx = Vkjky) JL iba) The geometric mean of the coupling factors is called the coefficient of coupling ky kn = VkKy) (2.24) 1>k,2>0 Thus My = Kjua/(LjLa) (2.25) Another constant used in coupled circuit theory is the leakage coefficient On = 1-ki (2.26) and from (2.25) = 1 Mik Op = Te (2.27) 2.2. The air-cored two-winding transformer An air-cored two-winding transformer, Fig. 2.4, comprises two con- centrated coils of N, and N, turns respectively, maintained at a fixed distance of separation in air. The winding connected to the supply is called the primary and the winding connected to the load, the secondary. Equations derived in the preceding sections, describing the magnetic interactions of n magnetically coupled coils, are em- ployed for analysing the performance of the transformer.COUPLED CIRCUITS AND TWO-WINDING TRANSFORMER 23 With n=2, the mutual and magnetizing fluxes are equal coil] 4. = Ono coil 2 G2, = Ont (2.28) The total mutual flux linking the primary and secondary windings Dy — Pz, 4+ Py, = Oy, +O, (2.29) From (2.3) #, = 442,492 (2.30) SS Px o, D2. + G1, +H, (2.31) ae Dy En =P, LF 92 =F 12 R: in [i rail Supply t, oy Wy M2 oy Load 2 Lis : ---4 Primary ~~ L Secondary by =O, +62, Fig. 2.4 Two-winding air-cored transformer From (2.8) through to (2.13), N,®, N2®, Ly = Ly = (2.32) a iy N\®. Wp a aa M,, = 2.33) 12 My, = My, =M Lg = NiPas Dug = Mat @.§M) Ly = Lith Ly = Latha (235)24 GENERALIZED ELECTRICAL MACHINE THEORY From (2.33) : Mi Mi. = ye Pn= We. (2.36) Substituting respectively in (2.34) Lay = —M ae m= Lm2 N, M (2.37) whence N,\2 Lm = ) Ema bm = ( aa . t a ; N, Ny 4 i, Ri (Ly-M) (L,-M) 2 i, Fig. 2.5. Transformer equivalent circuits From (2.18) v, = i,R, +L pi, + Mpi, (2.39) vb) = ipRy+Lapiz+Mpi, Rearranging (2.39) v, = Rk, +(L,— iy + Mpli, +i: ae Cy ees pli ee (2.40) 02 = IR + (L2—M)pin + Mpi, +i) Equations (2.39) and (2.40) are satisfied respectively by the circuits shown in Fig. 2.5, which are equivalent circuits for the two-winding air-cored transformer.bi COUPLED CIRCUITS AND TWO-WINDING TRANSFORMER 25 An alternative form of equivalent circuit, leading to the equivalent circuit for an iron-cored transformer, is now developed. Defining voltages vy,, and vy, which are respectively due to the primary and secondary voltages induced by the mutual flux ®y, ow = MipPa Oz = Nap Py (2.41) thus Pwr _ Ni ets 2.42 Um2 No ae From (2.30) and (2.31) Nip®, = Nip, +N pO N2p®z = N2p%.+N2pPy Substituting (2.41) in (2.18) vy = Ry +Lypiy toms . 5 (2.43) vg = 12Ry+Lppiz tome , From (2.33) Mi. Mi ®,,= We b= Ne Ou = Pi2t Or, = (Mi + Noir) (2.44) ww The mutual flux @,, is provided by an m.m-f. N,i,, where i,, is the primary magnetizing current ig, = Nyis +Noin A . N: oe in = ity ip = ith (2.45) where i3 =(N3/N;)iz is the secondary current referred to the primary ML. Oy = — i, = i nn N, im Thus Oger = Linrin (2.46)26 GENERALIZED ELECTRICAL MACHINE THEORY > “Equation, (2.46). shows that magnetization, for ©, can be repre- * sented in an equivalent circuit as a current j,, flowing in a magnetizing ‘: inductance L,,, connected across the voltage vy,. Between vy, and Uyg2 is an ideal two-winding transformer, so called as it has no losses, has no leakage fluxes, and does not require a magnetizing m.m_f. The equivalent circuit satisfying equations (2.43), (2.45), and (2.46) is shown in Fig. 2.6. Ideal transformer Fig. 2.6 Transformer equivalent circuit showing magnetization For the ideal transformer i,N,-ijN, =0 combining with (2.42) v, N, i aM 2 (2.47) tm2 No hy Pail = Umeati It is seen for the ideal transformer that power is transferred with- out loss. For the actual transformer, power to the load equals the power supplied at the primary circuit terminals minus the total winding resistance loss. 2.2.1 Referred values A simplified version of Fig. 2.6 is obtained by referring secondary values to the primary circuit. Commencing with vy = 1,Ry+Lypizt+vysCOUPLED CIRCUITS AND TWO-WINDING TRANSFORMER oe Expressed in terms of i; nM Ny : 02 = in Rata Lapin toma Ny, Nz N N vo) = N ry and ty = Om2 = ¥, m2 2 N,\? vy = (%) n+(¥ ) Lippi +0uyy Similarly, the referred value of load operational impedance : MV Zip) = (=) Z(P) thus vy = Ry tLypiy tomy 02 = ~Zi (pig = Ry + Lips toma (2.48) With secondary values referred to the primary, the turns ratio of the ideal transformer N,/N;=1 and its presence can be ignored. The equivalent circuit satisfying (2.45) and (2.47) is shown in Fig. 2.7. A i oe ty Ry aA Fig. 2.7 Transformer equivalent circuit, all values referred to the primary 2.2.2. Phasor equations and phasor diagram Expressing (2.45). (2.46), and (2.48) in terms of t.m.s. phasor quantities28 GENERALIZED ELECTRICAL MACHINE THEORY Vi = (Ri tol +Van ri a 2.49 Vi = (Ry +jeoLjos + Van ml Vari = JL pity Ty lags Vy, by 90° (2.50) 1, = 45 (2.51) As previously discussed, power transfer is from the supply to the load and to accord with this sense of power flow the secondary Fig. 2.8 Transformer phasor diagram, all values referred to the primary current is more usually represented as flowing into the load. De- fining primary and secondary referred currents as I, and I,COUPLED CIRCUITS AND TWO-WINDING TRANSFORMER 29 =I, b= -f V,= (Ry tfolaMy + Vite es) Vi = (RL +JOLiM.+V arp Vaep = JL mpln (2.53) 1, = 1,-f, (2.54) Primary leakage reactance X,,=@Lyp Referred secondary leakage reactance X,=wlLj, Primary magnetizing reactance Xjp=@Lmp Primary impedance Z, = RytiXy Referred secondary impedance Z, = Re+iXs Figure 2.8 isa phasor diagram for all values referred to the primary and / assumed to be lagging V; by an angle g. 2.2.3 Transient analysis The transient response of the air-cored two-winding transformer may be determined from the differential equations (2.39). Writing (2.39) in Laplace form and assuming zero excitation prior to the application of a forcing function of voltage at the primary or secondary terminals: dy = (Ry ues a (2.55) 3, = (Ry +Lys)i, + Msi, (a) Consider the secondary circuit terminals to be open-circuit i,=0 5, = (Ri +L,s)i, d, = Msi, b= Msi, a (44s) (2.56) where the primary open-circuit time constant ty=(L,/R,).20 GENERALIZED ELECTRICAL MACHINE THEORY (i) With r, a step of voltage V at ime 1=0 Vv uy and a, = TY (te) Y MY ony, 1) : yy = et = Vk £2\ ot acne all? : (2.57) With increasing L,, the amplitude of v, at =0 diminishes in value with a corresponding increase of t,, see Fig. 2.9. With ZL, small uh Fig. 2.9 Open-circuit response to a step function compared with L,, the secondary response approximates to an im- pulse. (ii) Forcing function v, = V sin wt, applied for 1>0 Vo Vo © Sa? ~ (6+ ju\(s—jo) >,COUPLED CIRCUITS AND TWO-WINDING TRANSFORMER 31 Substituting in (2.56) Mov s H (s +s +jo)(s— jo) b= In partial fractions _ Mov Tu jo i, = OS L, 1 1 I * (s+1)(4+0") (-ie +— \e +jo)(— 2ja) T/\TH u jo +7 (0+ )aiors ~ie) Mov ets : 1 = >_> + sin wt+— cos wt 2,1 as 1 Li o+5 1 Let 1 o . cos g = —~——_~_ and _ sing = ——_-——_ z+ : 1 /[o+5 Mera) ee . 1 Ng @ sin wt +— cos wt = (oa) sin (wt+ @) T Gt 1 and = el (sin (wf+g)—sin ge~"") (2.58) G +3) i Ly V A typical response, illustrating (2.58), is shown in Fig. 2.10. (b) The secondary circuit terminals are short-circuited 6, =0 5, = (R,+L,5)i, + Msi, 0 = (R,+L,8)i, + Msi, (2.59)#32 GENERALIZED ELECTRICAL MACHINE THEORY _ Ms i (R34 Ls) - Ms? \. ob, = (Rater) i,= (Ly Lz — M?)s? + (L, Ry +L,R,)8+ RR, - R,+L,8 _ L,\L)—M? = L,L,0 where a is the leakage coefficient. rt —MeV sind | dic. LL. component Fig. 2.10 Open-circuit response to a suddenly applied sinusoidal voltage With t,=L,/R, and t)=L,/R,COUPLED CIRCUITS AND TWO-WINDING TRANSFORMER 33 Roots of the polynominal, m and n 1/1 1 Ll iy = -5-(—+— —;(—+—) - 2.60 mon aes) Vleet) ae] ! + ,-L Ae) , 1 Lys (s—m)\(s—n) ' (2.61) : M s = a= eel L,L,0 (s—m)(s—n) Equations (2.91) can be solved for a variety of primary forcing functions v,. Consider v, a step change in voltage V at time :=0 (-) hoe - Vv ie L,o (s—m)(s—n)s a (2.62) : MV 1 he “L136 (s—m)(s—n) Expressing (2.62) in partial fractions 1 1 —+m —+ -_v (2 ) (2 n) nll i — ~ Lyo \(—m)(m—n)m * (n—m)(s—n)n* t2mns ia al 1 + 1 ) 2 * Ly Lzo \(s—m\(m—n) "(n= m\(s—n) The inverse Laplace transforms (2.63) ; eid ( mt 9) i, = —— A,le™-e 2? LL, °? As seen by (2.60) the roots m and n are both negative quantities with the result that the exponential terms in (2.63) are transient J = a I i, = iG (A\e™ + Byo™+C,) 444 GENERALIZED ELECTRICAL MACHINE THEORY currents which decay with time. When the transient currents have disappeared i, =0 but thus the steady-state primary current. 2.3 The iron-cored transformer Power transformers are designed for high efficiency, i.e., low loss, and for small on-load variations of secondary output voltage V,. To achieve these conditions low values are required for the primary and secondary winding resistance and leakage reactance. With the windings core Y) windings core (a) Core-type (b) Shell-type Fig. 2.11 Transformer cores primary and secondary windings linked by a closed iron core of high permeability, the leakage reactances are small as there is little ten- dency for flux to follow leakage paths in air. Two common forms of iron-cored transformer construction are shown in Fig. 2.11. With an iron-core the pulsating flux Py, gives rise to two types of iron-loss, known respectively as hysteresis and eddy-current loss. The hysteresis loss is associated with the hysteresis loop of the mag- netic material and the work expended in cyclic magnetization of the core. Eddy current loss is the result of currents induced within theCOUPLED CIRCUITS AND TWO-WINDING TRANSFORMER 35 iron, flowing in closed paths, giving rise to ’7R Josses. As the two Josses can be shown to vary with Prgmax) and, therefore Vyy,. the total iron loss can be represented in the transformer equivalent circuit as a resistance R,,, connected in parallel with X,,,, see Fig. 2.12. Total iron loss Vi = tt = ERup (2.64) wp With respect to Vy,, the iron loss component of current I,, is in phase and the magnetizing component I,, is 90° lagging. Fig. 2.12 Equivalent circuit for iron-cored transformer On no-load the primary current, known as the no-load current Ip, is given by Ig =1,ttn (2.65) On no-load V, = (Rpt), lot Vay (2.66) As previously stated, (R, +jX;,) is small for a power transformer and to a sufficient degree of accuracy V,~Vy, on no-load. Thus Vv, evpen or on no-load. (2.67) Vas Ns The voltage transformation ratio is defined as Vus _ V. vo7Vv (2.68) Pp with the transformer on no-load.36 GENERALIZED ELECTRICAL MACHINE THEORY For a step-up transformer For a step-down transformer Vs cy Vv, P The shunt admittance of R,,, and /X,,, can be placed directly across the primary supply, see Fig. 2.13, without appreciable error. This I, Ry = RytR,— iXir = KX + Xi) Fig. 2.13 Approximate equivalent circuit for iron-cored transformer leads to considerable simplification as the primary and referred secondary values of resistance and leakage reactance can be com- bined to give total effective values. Fig. 2.14 Phasor diagram corresponding to the approximate equivalent circuit, Fig. 2.13COUPLED CIRCUITS AND TWO-WINDING TRANSFORMER 37 I N\2 Ry = Ry +R, = &,+(*) R, Ny , N,\? (2.69) Nip = Xipt Xig = ot(™) Xs 2 Zp = RrtiXer and v Vv L=-/72 W=z% (2.70) Xnp Rup Lr 7) The phasor diagram satisfying (2.70) and (2.71) is shown in Fig. 2.14 for I lagging Vi by an angle g. PROBLEMS 2.1 Accoil A of resistance 15 Q and self-inductance 0-015 H is placed near toa short-circuited coil B of resistance 30 Q and self-inductance 0-03 H to give a mutual inductance of 0-009 H. Determine the effective resistance and reactance of coil A at an angular frequency of 1000 rad/s. (Ans. 109+ 9-1] 2.2. Plot to a base of log frequency the magnitude and phase-angle of the impedance of the circuit shown in Fig. Prob. 2.2. The coefficient of coupling is 0-1. A 1002 B self-inductance self-inductance 15 pH 15 pH Fig. Prob. 2.2 2.3 Show that the turns ratio of a single-phase transformer is the square- root of the ratio of the winding self-inductance when the winding coupling factors are equal. A 50-c/s, single-phase transformer has a loss-free core and the following parameters, primary winding: resistance 19 self-inductance 8 H coupling factor 0:9838 GENERALIZED ELECTRICAL MACHINE THEORY f secondary winding: resistance 102 self-inductance 128 H coupling factor 0:98 Derive the equivalent circuit in terms of (a) resistance, self- and mutual inductance; and (b) resistance, leakage reactance, and magnetizing reactance referred to the primary. [Ans. (b) R =0-625 Q; X,=503 Q; X,,,=2460 Q) Two coupled coils have the following parameters, L,=100mH = R, = 252 _—_N, = 1000 turns L,=25mH = -R, =5Q__N, = 500 turns M = 47-5 mH Core losses are negligible. Derive two forms of equivalent circuit. For an angular frequency of 1000 rad/s determine, (a) the equivalent impedance for coil 1 with a short-circuit on coil 2, (b) the open-circuit voltage of coil 2 with a voltage of 10 V r.ms. applied to coil 1. Use both equivalent circuits to calculate (2) and (b). (Ans. (a) 25:9 +133; (b) 4-47 +/1-12] The primary of a two-winding transformer is energized by a 50-c/s supply. The voltage, current, power and power factor are measured with (a) the secondary open-circuit, i.e., open-circuit te: (b) the secondary short-circuited, i.c., short-circuit test. Show that the values of the elements in the approximate equivalent circuit, referred to the primary winding, may be deduced from these tests. Explain why the transformer is energized at rated voltage for the open-circuit test and at reduced voltage for the short-circuit test. Tests ona 100-kVA, 3300/250-V, 50-c/s transformer gave the following results, ‘open-circuit test: 800 W, 0:5 power factor at normal voltage, short-circuit test: (data on h.v. side) 24 A, 300 V, 750 W. Determine the constants of the approximate equivalent circuit. The transformer supplies rated load current at 0°8 p.f. lagging. For an output voltage of 250 V calculate the primary voltage and the transformer efficiency. [Ans. 3590 V, 0-9756 p.u.] An iron ring, relative permeability 4, = 1000, has a mean diameter of 0-2 m and a square cross-section of side 0-05 m. Two coils are wound on the ring. Coil A has 1000 turns of resistance 30, coil B has 250 turns of resistance 2.Q and the coefficient of coupling is unity ;and i 'COUPLED CIRCUITS AND TWO-WINDING TRANSFORMER 39 2.8 2.9 Draw two equivalent circuits in terms of, (a) resistance, self- and mutual reactance, and (b) resistance and reactance referred to winding A. Assume an excitation frequency of 50 c/s. The ring is now sprayed with copper paint of resistivity | »Q— cm to a thickness of 0-1 mm and coil B is removed. (c) Derive the equivalent circuit referred to coil A for a frequency of 50 c/s. Assume the iron is lossless and all the flux is contained in the iron ring. [Ans. (a) ¥, =1570 Q, Xg=98 Q, Xy=392 Q; (b) Rg=32.0, Xya= 1570 Q; (©) (33 +/0-0057) Q] Show, from the approximate equivalent circuit for a single-phase transformer, that the losses at normal voltage comprise a fixed loss and a loss proportional to the square of the load current. Hence, show the fixed loss (iron loss) and the variable loss (copper loss) are equal for maximum transformer efficiency. Derive the voltage equations for three mutually coupled coils which include coil resistance. Assuming mutual fiux iscommon to all coils determine an equivalent circuit in terms of values referred to one coil. Assume some mutual flux links only two coils. What effect has this on the equivalent circuit?3. Electro-mechanical energy conversion Energy is the capacity to do work. In a mechanical system mass may have two types of energy: (a) potential energy due to its position, and/or (b) kinetic energy due to its motion. Similarly, there are two types of energy associated with electrical charges: (a) A static charge produces an electrostatic field which has electro- static energy due to its potential or voltage. (b) A moving charge produces an electromagnetic field which has electromagnetic energy due to the motion of the charge. 3.1 Electro-mechanical transducers An electro-mechanical transducer is a device for converting electrical energy into mechanical energy or the reverse, i-e., mechanical energy into electrical energy. The essential link in this chain is the energy of the electric and magnetic fields, Fig. 3.1. When a mechanical member is allowed to move under the action of a force produced by electrical means, then mechanical work is given out. The energy to do this work is ob- Motoring» Field storage — Electrical electrostatic and Mechanical energy electromagnetic energy fields ~<— Generating he Fig. 3.1. Electro-mechanical energy conversion __SststsésésS=ésesOONCSCitsCCNELECTRO-MECHANICAL ENERGY CONVERSION 41 tained from the electrical and magnetic fields. If no energy is given to the field from an electrical source, motion continues until either (a) the mechanical force becomes zero and motion ceases. or (b) the moving member reaches the limit of its permitted travel. With a continuously operating device such as a rotating machine, the field energy absorbed in doing work is replaced by energy from the electrical source. A similar process occurs for energy conversion from the mechani- cal state to the electrical state. In this instance an essential require- ment is that a magnetic or electric field must exist before an induced voltage can generate electrical power. For this reason a transducer may not be reversible if the field collapses when attempting to take electrical power from the device. 3.1.1 Conservation of energy An energy balance equation can be written for a transducer in which the input energy is equated to the sum of the following components: dissipation by losses, field storage, and useful output. With an electro-mechanical energy transducer Electrical Energy to Energy to field Electrical to energy | =| electrical }+] storage inthe |+] mechanical input losses electrical system energy 3.1) In symbolic form, with appropriate suffices, W. = WietWretWom (3.2) All the mechanical energy does not appear as useful output energy. A fraction is dissipated in mechanical losses W,,, and some energy Wm May be stored in the mechanical system as potential or kinetic energy. The net mechanical output W,, is given by Wi, = Wom—Wim— Wom (3.3) Substituting in (3.2) We = (Wiet Wim) + (W e+ Wom) + Won (3.4) Deriving an expression for a mechanical to electrical transducer in a similar manner: Wm = (Wiet Wim) + (W 40+ Wom) + W, (3.5)42 GENERALIZED ELECTRICAL MACHINE THEORY With energy flowing into ports taken as positive, a general equa- tion can be written as Wot Wy = (Wiet Wim) + (Wyret Wom) 3.6) The conditions described by (3.4) and (3.5) are satisfied. Electrical to mechanical energy conversion, W.istve, W,,is —ve Mechanical to electrical energy conversion, W,, iS +ve, W, is —ve 3.1.2 Types of transducers Possible methods for producing force by electrical means are summarized below: (a) Mechanical forces due to the interaction of magnetic fields. A typical example being a current-carrying conductor in a mag- netic field. (b) The forces exerted on ferro-magnetic material to move it in the direction of increasing field strength. Typical examples are a compass needle in the earth’s field and an armature-attracted type relay. (c) The electric field analogy of (a). (d) The electric field analogy of (b). (e} Magnetostriction, or the deformation ofa ferromagnetic material in a magnetic field. (f) The electric field analogy of (e), i.e., the piezoelectric effect. The power available from methods (c) to (f) is relatively small compared with methods (a) and (b). Consequently, the majority of applications use the first two methods. Typical examples are relays, solenoids, and, of course, rotating electrical machines. The main magnetic fields in practical devices and machines are invariably produced by electrically energized coil systems to make a versatile and economic arrangement. With a single energized coil a transducer is called a singly-excited system. Doubly-excited systems have two coils which are usually arranged with one on the stationary member and one on the moving member. Systems with more than two coils may be employed but practical transducers can often be reduced to singly- or doubly- excited systems.ELECTRO-MECHANICAL ENERGY CONVERSION 43 3.2 Singly-excited system The attracted-armature type relay, Fig. 3.2, is probably the common- est example of a singly-excited system. Another possible arrange- Stationary member Moving member Fig. 3.2 Typical singly-excited system ment, Fig. 3.3, has an action similar to a compass needle in a magnetic field. The ferromagnetic rotor tries to position itself to give minimum reluctance for the magnetic flux, i.e., torque is exerted to move it in the direction of increasing field strength. The torque depends on the shape of the rotor. With a cylindrical rotor there is no torque as the reluctance is unchanged by angular position. With Position A (torque) Position B (no torque) | SN \ Fig. 3.3. Rotational singly-excited system diametrically opposite projections on a round rotor, the reluctance is now dependent on the rotor angle. The rotor is said to have saliency. For the analysis of a singly-excited system a ‘dumb-bell’ shape rotor, with an exaggerated saliency effect, as shown in Fig. 3.3, is considered. The following assumptions are made:44 GENERALIZED ELECTRICAL MACHINE THEORY (a) Non-linear flux/current relationships for the magnetic circuit formed by the stator and rotor iron. (b) No leakage flux, i.e., the flux produced by the coil links with all the turns, and follows the main magnetic path. (c) A rigid coil for which no deformation occurs due to electro- magnetic forces. Without this assumption, the coil inductance changes with current and mechanical strain energy is stored in the coil. (d) Hysteresis and eddy current losses are neglected. (e) The magnetic field predominates and electric field phenomena are neglected. With a coil resistance R Q, the instantaneous voltage equation for the coil is dv = Ri+#— 3.7, v i 3.7) Multiplying (3.7) by i d¥ i= RP +i 3.8) vi PFE (3.8) Integrating with respect to time t from =0 to r=¢, and assuming the current and flux-linkage to be initially zero 7 7 ” [vee f Ri are [ idv 3.9) 0 ° o and Total Energy Useful electrical = to a electrical (3.10) input electrical energy energy losses input The Useful electrical vo energy ~{ id¥ (3.11) input Now the energy balance equation (3.2) is We = Wit Wyet Wem erin 4 at # i { eesELECTRO-MECHANICAL ENERGY CONVERSION 4s Referring to (3.9), (3.10), and (3.11) W, = Total electrical input energy . = Energy to electrical losses and ” { id P = Wye+ Wom (3.12) o 3.2.1, Static energization For any position of the rotor of the singly-excited system, Fig. 3.3, a characteristic of flux or flux-linkage against current can be drawn. Figure 3.4 shows the characteristic for two rotor positions. Position A: with the rotor inclined at an angle to the stator pole axis. Position B: with the rotor aligned with the stator pole axis. Elemental area i.d¥ —~ Elemental area ¥.di oO Fig. 3.4. Static flux-linkage against current characteristics for the system . of Fig. 3.3 Flux does not increase uniformly due to saturation of the iron. The air-gap reduces the flux density and also makes the relationships more linear. In position A the rotor asymmetry produces torque in the direction shown in Fig. 3.3. On reaching position B, there is no magnetic field distortion and the torque on the rotor is zero. With no movement there is no mechanical output and all the useful electrical energy input is converted into field energy;46 GENERALIZED ELECTRICAL MACHINE THEORY a ie, useful electrical energy = [ id¥ = W,, (3.13) 0 Referring to Fig. 3.4, the total magnetic field energy W,, corre- sponding to a current increase from zero to i is represented by the area of this integral. Position A: The field energy is area OALO. Position B: The field energy is area OCDO. The co-energy of the field is defined as the area under the flux-linkage/ current characteristic and is denoted by Wj,. Co-energy W’;, { Y di (3.14) 0 Position A: The field co-energy is area OEAO. Position B: The field co-energy is area OECO. 3.2.2. Dynamicenergization Before considering the general transi- ent operation of the device, it is useful to consider two types of con- trolled movement. The rotor moves either very slowly or instan- taneously from position A to position B. (a) Slow movement If the rotor is constrained to move very slowly the current i is sensibly constant as d'¥//dt is small. The flux- linkages may be assumed to increase from ¥, to ¥ at an assumed constant current i. The useful electrical input is _ v: [ i, d¥ = al a¥ = 1,(%,—¥)) G.15) ey ey Figure 3.5 shows how the field energy changes. The mechanical work is the difference between the energy input and the change in field energy, and equals the area OACHO. The average electromagnetic torque during movement is given by 7. (average) — chanical work _ area OACHO . Be) = angular movement —_ angular movement (3.16) Consider a small change of angle d@ during the movement. TheELECTRO-MECHANICAL ENERGY CONVERSION 47 incremental mechanical work is represented by area OGKO in Fig. 3.5. ol hanical a OGKO T, (average) = {mechanieet work) a (3.17) The instantaneous torque at any angle 9 is given by a << L=-f1 (ms gone) G.18) 50-0 60 Final stored v field energy D c y TOKO? S<8 Electrical 2) = K 0 I i i Energy to mechanical work Fig. 3.7 Flux-linkage against current characteristic for typical rotor movement mechanical work lies between the values obtained by moving the rotor instantaneously and very slowly. Fora small movement PQ of the rotor, see Fig. 3.7, the energy to mechanical work is represented by area OPQO. With controlled movement at P the energy to mechanical work is given by area OPRO for movement at constant ¥, and area OPSO for movement at constant i. The torque expressions for the three cases are as follows: (a) actual movement: _ dW y, [area OPQO a0 FAA 30 oS50 GENERALIZED ELECTRICAL MACHINE THEORY (b) controlled movement at constant current: _ [ewe To ma “ (= ree -#1 (= OPQO +area =) 0-0 50-0 60 (c) controlled movement at constant flux-linkage: OW ye Te= -| a | = P = #1 (= a) =-#1 (= OPQO area PQR 50-0 30-+0 4 (3.25) 50 50 (3.26) In the limit as 6@ > 0 the areas PQSP and PQRP vanish and the torque expressions converge to a common value 7 = Wie _ We « "d0 ~~ a0 om] [2% = =- 3.27 [ — oa ae current flux-linkage Equation (3.27) enables torque to be calculated at any angle 6 from a knowledge of the change of field co-energy or energy under con- trolled movement at constant i or ¥. These results may be obtained by a simple mathematical treatment. The field energy and co-energy are functions of flux-linkage ¥ and current i, eg., field co-energy = W;,(¥, i) (3.28) If there is a relationship between ¥, i, and @ which can be expressed mathematically, the field energy and co-energy may be expressed explicitly as functions of (i, 0) or (¥, 0), e.g., field co-energy = W;,(i, 0) (3.29)ELECTRO-MECHANICAL ENERGY CONVERSION 31 Differentiating (3.29) with respect to @ gives the expression for torque _ IW 5i, 0) Te = “a9 _ Wei, 0) dO OW yi, 0) dé a a re) (G30) With current i held constant di a7? and OW Ai, | T, =|—— 3.31 [PA Tena 8? (Note: Function W;,,(i, 0) has to be expressed in terms of iand and not flux-linkage ¥.) Ina similar manner, the expression for torque may be found with the flux-linkage held constant; Le, tee -| Pa 2 (3.32) 0 | EXAMPLE 3.} A rotating singly-excited transducer, Fig. 3.3, has a non-linear relationship between the flux-linkage Y, current i, and position 0, which may be expressed as i = (Ap—A, cos 20)¥1'° Derive the torque expression for any angle 0. Solution The torque may be found from either the field energy or the co-energy. 7 Wye -| idv 0 ” = { ¥'6(A—A, cos 20) d¥ 052 GENERALIZED ELECTRICAL MACHINE THEORY This expression is the stored energy for any angle @ which is con- sidered constant during integration val Wy. = (Ao— Ay cos 20) |" wre ay yee = (Ag— A, cos 20 [ yrs W,, = (Ao—A, cos 20) joules Torque T, = -[ eee) IW constant co a yrs = —=]| (Ap—A, cos 20)- | al = 2-6 by constant pre T.= Ee (2A, sin 20)N—m (3.34) An identical solution is obtained using the co-energy expression. Wie = ( Y di 0 i pes (A.A. cos 20)re di lo (Ag — A, cos 20) = 1 ‘ 1/16 ae ann) os a 1 16 j2-6/1-6 | Wie RK cos HT amie! joules (3.35) Whe (. 0}. 20 16 j2-6/1-6 slaw cos 20) 3-6 | _ 1 (=2A, sin | o- ow Torque T, = (° é a0 = 16 peng ea (Ag—A, cos 20)7 9S 16 = RAL sin sin 20)(?" cue) 2-6(Ay—~A, cos 20)? 1° N-m (3.36)ELECTRO-MECHANICAL ENERGY CONVERSION 33 Substituting for (Ay—Aj, cos 20)?!" in (3.36) gives the same expression as (3.34) for T,. —(2A, sin 20)i7°6/1°6 2:6(Ao — Ay cos 20)?"°/1'¢ = (2A, sin 2972-916 (prey 2611-6 ~ 26 ( ) —2A, sin 20976 1p g T,= oe . 26 Both the field energy and co-energy may be expressed solely as functions of ¥ and / by substituting for the term (Ay—A, cos 26). Equation (3.33): wre W,. = (Ap—-A as ‘re = (Ao 1 cos 20) 6 iowre we iv Ww, = 144 : fe = 3G joules (3.37) Equation (3.35): eres fee 1 16 j2-6/ 1-6 Wre = (KIA, cos 20)" Gs ‘ 2 (18 rons 7 \36 1-6Pi w= 3.38 be = Se (3.38) As expected, the stored energy and co-energy together equal Yi. 3.2.3. Linear system With a linear flux/current relationship for the singly-excited system Y= Li (3.39) where L is the inductance which is a function of 0 but is independent of current. 554 GENERALIZED ELECTRICAL MACHINE THEORY The field energy for any angle 0 (or inductance value L) is given by: . w= f id¥ F wa = [fee lo L 19? Wea Alternative forms for field energy using (3.39) are 1 Wy.=3¥i = ght (3.40) The field co-energy we nm f Y di 0 = [aia o / 1, Wye = 5 Li Hence the field energy and co-energy are equal in a linear system. Summarizing: eo le el te Wye = Wye = 5 LP = 5 ¥i= 5 (3.41) Electromagnetic torque OW, T, =| 0 [Reo esl =|—[=Li? [ole*)) 1, aL T.=5° 5 As L is not a function of / the partial derivative is unnecessary 1 dL _ (3.42) 62" do tEAM As ALM RE SON RRO ASO erecta eee at 8 lawyr rPrPrPreeFrrararrrFrtnnTns SF PPELECTRO-MECHANICAL ENERGY CONVERSION 55 EXAMPLE 3.2 A rotating transducer, Fig. 3.3, has a linear relation- ship between flux-linkage and current. The inductance varies as (Lyi + Ly 608 20). (a) Derive the general torque expression. (b) Calculate the average torque if the rotor has a constant angular velocity w rad/s; ie, 6 = qt, and the current varies as /,, cos (wt +6) Solution (a) Torque 7 ied ©" 3" a0 lid =F P 6 (Ly + L,2 cos 20) T, = —i(L,2 sin 20) (3.43) With a steady current the torque oscillates sinusoidally. The net torque is zero over one revolution. (b) With a sinusoidally-varying current, T. = —i?(L,z sin 20) = —[J, cos (wt+64)]?L,, sin 20 The average torque over a cycle 1 2" T, (average) = z| —[I,, cos (wt+6)]?L,2 sin 26 dé (3.44) T Jo as 9=wt the expression may be written Pin (ee e ~ a [ cos? (9 +6) sin 20 dO o 2 2e 7 Inka 1+ cos 2(6+6) sin 26 46 2n Jo 2 2 2x i — _ EL sin 20-4 sin 2(20+6)—sin 26 40 4x Jo 2 27. poe ‘ _ _ fae f [si 9 Si 2(20+5)—sin 2 a0 ano 2So GENERALIZED ELECTRICAL MACHINE THEORY RL [33 20 cos 2(204+6) Osin 26 P* 4n 2 8 2 Io PL, 22 a Iinbra 2h 4s dn Dn 20 T,, (average) = 0-25/2.L,2 sin 25. N~—ne (3.45) With an alternating current applied to the device a useful torque is developed when the angular velocity is equal to the angular frequency of the supply current. Single-phase reluctance motors use this principle, which is considered further in chapter 4. 3.3 Doubly-excited system In a doubly-excited system each winding has self flux-linkage and there is mutual flux-linkage between the coils. Unless the effects of these flux-linkages are independent of each other, an analytical treat- ment is extremely difficult. It is assumed, therefore, that the coil Ti _ Ne Fig. 3.8 Typicai doubly-excited rotating transducer systems are linear, flux is proportional to current and superposition applies. Other necessary assumptions are given in Section 3.2 as for the singly-excited system. Figure 3.8 defines the voltages. currents. etc,. for a typical trans- ducer. The flux-linkage equations for the two windings are Y, = Lyiy+Mi, (3.46) Wy = Lyis+ MiELECTRO-MECHANICAL ENERGY CONVERSION = a The instantancous voltage equations for the coils are ey i U0 eae dr 2 = Roig t " we dt Substituting for Y, and ¥, in (3.47) Lyi), dM a hee edre _ dia) , A(Min) Raita “dt (3.47) Now the inductances are independent of current but depend on position 6, which isa function of time. Similarly, currents are time- dependent and are not functions of inductance. Hence dL), dia, 4 de aa etc. Expanding (3.48) dLy , yp dis, , dM a art ar ;, 92 an dM din, = Ry + Ly 2 +i, 24M te Ret la gy te gy tM Gy th ar d ty = Rit Ly at (3.49) Multiplying these expressions by i, and i, respectively, gives the power equations for the coils di, dL dM vy, = Rit Li al 1432 ari wh tie Gr 2 db di, .. dM Pi, = Ry thai 2 +5 dr 2 hM gti a (3.50) Integrating (3.50) with respect to time and adding gives the energy balance equation (3.2), feiresin di = [ (Ry + Rpi3) de + I( Lyi; diy + Lip diy +i,M diy +2i,iy df +P db, +i3 dL, +i,M di, 3.51)58 GENERALIZED ELECTRICAL MACHINE THEORY Useful electrical : 2 2 energy input {~ (epi He giz) dt— | (Ry + R13) de and Energy to Electrical to field storage} + scan | in the energy electrical system. Lyi, diy + Lois diy +i,M di, + 2iip UM -{( Nes 3.3.1 Stored energy in the field The instantaneous value of energy stored in the magnetic field depends on the inductance and current values at the instant considered. This energy may be found by con- sidering the transducer to be stationary and the coil systems to be +i} dL; +73 dL, +i,.M diy 3p Few 38 vq arsed yo —— 22 R ron 0/5 energized from zero current to the required instantaneous values of ae current. There is no mechanical output and W,,, is zero. The Th =] inductance values are constant and termsindL,,dL,,and dM vanish. - é From (3.52) A t 7 , a iat a (aw. -{ Liiy a+ [ bated [ (4M dizti,Mdi,) weg gg & 5 5 5 5 eeraiel sod a 7 I l SRT EE [total W,,] = 3 Litts L313 + Mii, (3.53) % mre gb | é 3.3.2 Electromagnetic torque Equation (3.53) holds for any trans- ducer position. If the transducer rotates, the rate of change of field energy with respect to time is given by differentiating (3.53). a) 1 2 ahi 141 (G3) yipdh var 12a 132 ar 73? a dM di, di ti Gti S pti de aW,. _ diy 1 dL, dip 1, dLy aL Mb Liss di gy tg lt ge thee G98 G dM di di thi iM ae +i,M. aa (3.54) se ¥ = = =e eee 2 2 4 W ta Ppa wm) y Mand) IHDY Go Fs 72 ~Y. vee TIP yyELECTRO-MECHANICAL ENERGY CONVERSION sy. Integrating this expression with respect to time ', Us , . Lyi, diy +, idl, + Lyi, dat, dh, di, = Wye = 2 2 +iyiy AM +i,M di, +i,M di, (3.53) This expression is now general for a transducer with movement in which L,, £2). M. i;, and i; are all varying with time. Comparing (3.55) with (3.52) it is seen that Electrical to i ; Wom = | mechanical | = [(58 dL, +5 3 dLy+ijiz am) energy 7 7 (3.56) Differentiating this expression with respect to 0 dW.y 1 dL, 1 ,dl;,. dM do 2" go 73? aa"? a6 as only L,, L, and M are dependent on #. Now electromagnetic torque OW AW oy wo 00 dO rahe Slr yt pdb. aM 2 0) 2 Dee 08 aaa 0) vt (3.57) With steady values of currents maintained in the two coils of the doubly-excited system (3.54) reduces to d idl, 1 .dl: ., aM hts +i, dt 2! de 27? dr 8? de Integrating with respect to time ¢ i B W,, fi aty+hiat.—ii,am| (3.58) which is equal to the energy converted to mechanical work W,,, in (3.56) The useful energy input is divided equally between the field storage and the energy to mechanical work60 GENERALIZED ELECTRICAL MACHINE THEORY PROBLEMS 3.1 Show for an electromagnet that the electromagnetic force where v is the linear displacement. Hence. show an electromagnet of area Am? and flux density B Wb m’, exerts an electromagnetic force 3.2 A 5-A moving-iron ammeter has the following values on test. deflection (degrees) 0 2 40 60 80 100 120 140 inductance (WH) 100 105 10 11S 129 124 127 130 Estimate the required restraint torque for full-scale reading at 120° and draw a calibration curve assuming the restraining spring charac- teristic is linear. Assume the iron does not saturate [Ans, 107 x 107 N-ny] 3.3. Aniron ring, cross-sectional area 0-01 m? and mean circumference 5 has an idealized BH characteristic B=0005H for0 200 where Bis the flux density in Wb/m? and H is the magnetizing force in ampere-turn/m. A coil on the ring has 1000 turns and a resistance of 20 Q. (a) A step-voltage of 120 V is applied to the coil at r=0. Show the transient current is i= 6(1-e7*) for0<1 < 00912 = 6[l-— for > 00912 (b) Calculate the electromagnetic field energy for a steady coil current of 6A. (c) Repeat for (b) with an air-gap of 0-003 m in the iron [Ans. 40. 9-76 joules} = 10(¢- 0-07 3.4 The singly-excited system shown in Fig, Prob, 3.4 has a small air-gap compared with rotor diameter. Assuming no fringing. show the reluctance torque is either.ELECTRO MECHANICAL ENERGY CONVERSION al (sero, oF ONT ib) jr Determine the mean torque at synchronous speed and show it to be a maximum when 6=45 and (7, +7) =90 radius r Bap g rotor width / i= 1,, sin (wt—d) Fig. Prob. 3.4 Reluctance motor 3.5 A single-phase reluctance motor has the following relationships between the flux-linkage ¥. current i, and angular position 0 = «1 i = (Ag—A, cos 20)8!? (a) Derive the instantaneous torque expression. (b) Sketch the variation of current, flux-linkage, and torque with time over a cycle when the supply current is /,, cos (+30 ) 3.6 A doubly-excited transducer. Fig. Prob, 3.6, has a cylindrical rotor which rotates at an angular velocity @, rad's. The rotor winding. energized through slip rings. has a resistance of | ohm and a self inductance of (0-02 +0-01 cos 20) henry. where 0 is the angle between Fig. Prob. 3.6 Doubly-excited transducer39 GENERALIZED ELECTRICAL MACHINE THEORY the coilaxes. “The stator winding has a resistance of 100 obi and a sell-inductance of 0-12 henry, The mutual inductance between the rotor and stator windings varies as cos 0 and has a maximum value of 0-06 henry. For constant rotor and stator currents of 10 amperes aad Fampere respectively, derive expressions for the fastantancous voltages applied to the windings, the electromagnetic torque and the electrical input power. Determine their values for 0 =90 and 45 [Ans. r, 40, $7-62 0, 4. — 142: T, —6, — 1-42: P, 80. 848) A doubly-excited rotating transducer has stator and rotor coil self- inductances of 0-4 H and 0-2 H respectively, and a mutual inductance of (0-2 cos 0) H, where 0) is the angle between the coil axes, A steady current of 10 A flows in both coils. Determine the stored energy in the system and the instantancous torque for an angle 0. Consider a small rotor movement irom é@= 29 to0=31 Calculate the instantancous torque at 30 and verify the value from energy bal- ance considerations. {Ans. (30+ 20 cos 0) joules: — 100 sin 0 N=m; —50 N—m] A singly-excited rotating transducer has a permeance characteristic. Fig. Prob. 3.8. which is independent of current Permeance A 7 3 —s Fig. Prob. 3.8 Permeance-() characte! Deduce the instantaneous (orque for an angle 0=cr and calculate the mean torque as a reluctance motor energized by a current i= 1, 3in (ot+8) {Ans. zero] y-excited The expression, torque 2(4L dO) is not valid for a sing system with a non-linear current ‘ux relationship. Explain the reason for this and jilustrate your answer by selecting a suitable relationship for i, L. and 44. Basic electrical machines Allelectrical machines have many common properties and features. The basic common factors in rotating electrical machines are developed and described in the following sections. 4.1 Fundamental machine features Section 3.1,2 mentions the different methods of producing mechani- cal force, or torque, in electro-mechanical transducers. The two concepts used in rotating machines are: (a) the attempt to co-align two magnetic fields, and (b) the attempt to align a magnetic material in a magnetic field or move it into a stronger magnetic field. 4.1.1 Physical concepts of torque production Consider a magnet in a magnetic field, as shown in Fig. 4.1. It is readily demonstrated that a torque is exerted on the magnet in an attempt to align the lines of action of the magnetic fields. With the fields aligned, Fig. 4.1(a), there is no torque. Complete misalignment of the magnetic fields as in Fig. 4.1(b) also produces no torque. This position is one of unstable equilibrium, however, and a small displacement of the magnet produces torque increasing the displacement further. Maxi- mum torque occurs when the individual fields are orthogonal. An initially unmagnetized ferromagnetic material placed in a magnetic field may experience torque. Consider a member of *dumb-bell’ shape as shown in Fig. 4.2. Torque is exerted on the member unless the resultant magnetic field is symmetrical as in Figs. 4.2(a) and 4.2(c). This torque is commonly called a reluctance torque as the ferromagnetic material trices to reach a position giving minimum reluctance to the flux path. 4.1.2. Equivalent coil systems The magnetic fields in the previous section may be produced by coil systems energized with currents.64 GENERALIZED PLECTRICAL MACHINE THEORY TS = SED". — a TSS ee (a) No torque (b) No torque — unstable equilibrium ae (c) Torque (d) Maximum torque Fig. 4.1 Torque production by the attempted co-alignment of two magnetic fields The interacting magnetic fields of Fig. 4.1 are produced by the doubly-excited system shown in Fig. 4.3(a). A. singly-excited 4.3(b). produces an identical field to t : system, F Doubly-excited svstem, Fig. 4.3(a) The individual magnetic fields gths are depend on the coil currents. ff the magnetic field streng assumed lo increase in a linear manner with current, the resultantBASIC ELECTRICAL MACHINES, 6S torque 7, may be expressed as the product of the currents and a function of position T.. & (iyiz) x (function of 0) Gb Singly-excited system, Fig. 4.3(b) The field strength is proportional to /, and the poles induced on the ferromagnetic material depend on S/H mS \ lJ d induced poles (a) No torque (b) Torque (c) No torque Fig. 4.2. Torque production by the attempted alignment of a magnetic material in a field the field strength, if linearity is assumed. The resultant torque is the product of the field strength, the induced pole strength and a function of position. T, x (#2) x (function of 0) (4.2) Combining the dumb-bell shaped ferromagnetic material with a doubly-excited system, gives the general arrangement for a doubly- excited system including reluctance torque. Fig. 4.4. (a) Doubly-exeited system With suitable assumptions With suitable assumptions ie Ky Tr. = RAW) Fig. 4.3. Basic torque concepts66 GENERALIZED FLECTRICAL MACHINE THEORY Including suitable coeflicients A, and As. both functions of 0. in (4.1) and (4.2) the resultant torque is given by Ty = iyigK (+5 KM) (43) T, is the torque on the rotor. The winding and the ferromagnetic core are assumed to be mechanically rigid in themselves and as a unit. How the torque is shared between the winding and the core depends on the flux distribution and not on the relative values of the co-alignment and reluctance torques. In practice, nearly all the torque is exerted on the core. Comparing (4.3) with the torque expression (3.57) derived pre- viously. viz. dM ot, i Gg t5 ta The expressions are identical if K,(0) = am K,(0) wd" uw 200 and dL,/d0=0 as there is no saliency on the stator. Hence the physical concept agrees with the analytical treatment. Saliency is included only on one member as very few machines have saliency on both members. 4.1.3. Motoring and generating action To emphasize the torque concepts common to rotating machines. the preceding section con- siders only currents in the windings. When the rotor moves under the torque action, mechanical work is given out. Relative move- ment induces voltage in each winding and electrical power is required to drive currents into the windings against the induced voltage. Energy conversion occurs. With a machine supplied with mechanical power, the reverse process. electrical power generation, may occur. The principle of torque production applies for the reverse process as the generated current produces a magnetic field which interacts with the main magnetic field. The resultant torque opposes the mechanical driv- ing torque and energy conversion is present. All rotating machines do not have reversible power flow. Some types of machine will only motor as the magnetic field collapses if generation into a passive load is attempted.BASIC FEF CERICAL MACHINES: 07 The device, Fig. 4.4, includes the wo methods of producing torque, one or both of which are present in all electromagnetic transducers. When rotation occurs with this simple device, the field misalignment decreases to zero and hence the torque becomes zero. A practical With suitable assumptions Torque T, = iji.K 0) +8K (0) Fig. 4.4 Doubly-excited system including reluctance torque rotating machine requires a finite mean torque per revolution. Suitable methods are discussed later, following the sections dealing with windings and connections, as these provide the required facility, i.e., misalignment maintained with rotation. 4.1.4 Basic rotating machine The essential requirements for a practical rotating electrical machine, based on Fig. 4.4, are shown in Fig. 4.5. stator stator winding AY air-gap iy : Ce Potor fr Fs [| Qwinging 6 TisZKSrowor connections ~ rotor Fig. 4.5 Basie rotating machi8 GENERALIZED PELCERICAL MACHINE THEORY (a) Stationary member (stator core and windings). (b) Rotating member (rotor core and windings) (c) Air-gap to allow relative motion. (d) Rotor winding connections (via brushes if externally supplied). Solid or laminated ferromagnetic materials are used for the stator and rotor cores to reduce the reluctance of the flux paths. Pulsating fluxes require laminated materials to reduce eddy current losses. Most rotating machines have windings on the rotating and stationary members. Some small machines have only one exciting winding as one member is permanently magnetized. Reluctance machines have only one winding as they are singly-excited. 4.2. Windings In order to treat machine windings in a general manner, and emphasize common features, their properties are discussed in terms of basic physical and electrical arrangements. As in Section 4.1.2, an energized coil produces a magnetic field and magnetic poles are shown located at the points where tux leaves or enters the coil. A detailed treatment of winding m.m. and associated flux densities follows in Section 4.2.4. 4.2.1 Basic properties (a) Poles and pole-pitch The instantaneous picture of a rotating machine may be represented by magnetic poles as in Fig. 4.6(a). This diagram shows a four-pole machine. Equal numbers of pole- (a) Four-pole machine (b) Unequal rotor and stator pole-pairs ic poles Fig. 4.6 Instantaneous picture of machBASIC ELECTRICAL MACHINES 69 With pairs on the rotor and stator are necessary to produce torque: irs the resultant torque is zero, Fig. 4.6(b). iny number of pole-pairs. Poles are equally unequal pole-pa: Machines may hav spaced around the machine periphery ata pole-pitch apart. For convenience and clarity a two-pole machine is often shown but the results can be extended to any number of pole-pairs Two pole-pitches span 2x electrical radians or 360 degrees regardless of the number of pole-pairs. son the rotating or stationary trical val arrangement Windir may be arranged, (a) around pole-pieces (i.e., salient-pole type); or (b) distributed in slots along the core periphery. (b) Distributed in stots Fig. 4.7 Physical winding arrangements The instantaneous magnetic field patterns for both arrangements are shown in Fig. 4.7. Magnetic poles are produced at regular intervals, a pole-pitch upart. (c) Electrical connections Connections to the rotating winding ofa machine are from stationary brushes making surface contact with (a) slip rings, or (6) a commutator 670 GENERALIZED PEEOTRICAL MACHINE THEORY ectrical connections with slip rings are to fixed points on the rotor winding regardless of the winding position in space With acommutator the electrical connections are to points on the winding occupying a fixed position in space. Rigid connections are made to stationary windings in all cases. (d) Flux pattern The flux pattern of a winding may be moving or stationary with respect to the winding, depending on the type of winding and the method of excitation employed N allele | Hout these conductors commutator o Current ction in these changed by commutator Ss {e Fig. 4.8 Flux movement relative to winding by commutator action A winding energized by direct current or single-phase alternating current has a stationary magnetic flux pattern with respect to the winding. A polyphase winding energized by balanced polyphase currents produces a magnetic field which appears to travel with respect to the winding. A rotating commutator winding excited by current supplied via its brushes produces a space-stationary flux distribution. Therefore, a relative speed difference between the stationary flux and the moving winding. Figure 4.8 illustrates the commutator action 4.2.2. Commutator winding A two-pole ring ty pe winding illustrates the action of a commutator windin Current enters theBASIC LEECTRICAL MACHINES 7 winding through brushes making surface contact with the commuta- for, Currentand magnetic Hux patterns are established. Rotation of the winding does not change these patterns which remain station- ary relative to the brushes if practical effects are neglected, ¢.g., width of commutator segment. The ring winding is only suitable for very as itis difficult to wind and low voltage, heavy current machines uneconomic to manufacture. The commutator winding arrangement in general use is the drum type winding with coils spanning approximately a pole-pitch. Coils ‘Equivalent coil (a) Flux pattern (b) Winding schematic Fig. 4.9 Ring-type commutator winding may be either /ap-wound or wave-wound. The lap winding has the same number of parallel current paths as there are poles, whilst wave windings have only two parallel paths. Figure 4.10 shows a simple two-pole lap winding. The flux and current patterns are similar to the ring winding, see Figs. 4.10(b) and (c). The winding diagram, Fig, 4.10(a), is referred to as a dereloped diagram, as the conductors are shown in a horizontal tine instead of lying in a circle on the rotor seen thata coil fora lap winding spans a pole-pitch. As with the ring winding, the coils form a continuous closed winding. The brush-pair is now physically at right-angles to the Mux line of action but schematically the brushes are always shown along the axis of the equivalent coil, as in Fig. 4.10(¢).7 GENERALIZED FEECERICAL MACHINE THEORY pole pitch : A OI NS Oo eG ONE NNeG , VA APR RIN IS DPN, Vv K< YSN YOY MON y Coil connections. to commutator (a) Developed diagram J —— Typical coil Brushes on flux axis (b) Flux pattern (c) Winding schematic Fig. 4.10 Simple two-pole lap-wound commutator winding The developed diagram for a simple four-pole wave winding is shown in Fig. 4.11. With this winding an odd number of slots is necessary so that the winding progresses and forms a continuous closed winding. The coil-pitch is not exactly a pole-pitch and, as with the lap winding. the brushes are physically displaced from the flux axis. As there are only two parallel paths through this winding. In practice, however, machines only two brushes are necessary.BASIC ELECTRICAL MACHINES: 1 with four poles are often fitted with two brush-pairs to reduce brush current, The use of more than a pair of brushes reduces the effective on machines of four poics or commutator length and for this rea more require a design compromise between brush current and ade- quate effective commutator length. pole pitch typical coil Fig. 4.11 Simple four-pole wave-wound commutator winding (a) Torque and induced rotational voltage Consider a two-pole commutator winding with a brush-pair arranged at right-angles toa magnetic field. as in Fig. 4.12. Current entering the commutator commutator winding Induced voltae ‘fi across brush 1 pair 4.12 Two-pole commutator winding,74 GENFRALIZED PLECERICAL MACHINE THEORY winding through the brushes produces a magnetic field directed along ct and torque is exerted on the the brush-pair. The two fields int rotor, as described in Section 4.1.2 If the commutator winding is allowed to rotate under this torque, there is physical movement of the winding and core. The com- mutator action, however, maintains the rotor current pattern stationary in space regardless of rotor movement. The interacting flux pattern produced by currents in the stator and rotor windings is held in a misaligned state and continuous torque is exerted an the rotor. This torque is independent of the rotor movement and angular velocity and depends only on the current in the commutator winding. Voltages are induced in the coils of the rotating commutator winding due to the changing flux-linkage with the stator magnetic field. Atany instant the voltage across a brush-pair is the sum of the induced voltages in all the turns in series between the brushes, and is proportional to the rotor speed, the stator field. and the number of commutator turns. With an unchanging stator field and a steady angular velocity a d.c. voltage is generated across the brush-pair. Energy conversion occurs with current flowing in the commutator winding. With rotation the torque exerted on the rotor results in mechanical output power. Current flows into the roter winding against the generated voltage and is a condition of electrical power input. If the mechanical and electrical quantities remain steady the magnetic field energy and the mechanical kinetic energy are constant and the electrical input balances the mechanical output. neglecting losses. This corresponds to motor action. When current flows out of the rotor winding under the action of the generated voltage there is electrical power output and the machine is generating. If the brush axis is not at right-angles to the stator pele axis (or direct axis) the interaction between the rotor and stator fields is decreased and the torque is reduced. The induced voltage across the brush-pair alse decreases and the mechanical and electrical powers remain balanced With the brush axis along the direc induced voltage across the brush-pair and zero torque axis there is ne resultant (b) Pseudo-stationaey coil Wis seen in Section 4.3.2 that ste current are identics! to those and flux patterns for a commutator windin: obtained with a stationary coil along the brush avis. Fv the com-Pe ee BASIC ELECTRICAL MACHINES S mutator winding is rotating ina magnetic field with a Mux component at right-angles to the brush-pair a rotational voltage is induced in the winding. Therefore. to represent a rotating commutator winding completely the stationary coil has to include an induced voltage effect : due to the rotation ofits conductors. [tis called a pseudo-stationary coil and has all the properties of a static coil with the additional feature of a rotational voltage. This representation is important in machine analysis and will be considered further in chapter 5. 4.2.3 Polyphase distributed windings Practical polyphiase circuits use three phases for engineering and economic expediency. In this treatment only three-phase windings will be considered although the principles are general and apply to any number of phases. 4.13. Single-layer three-phase distributed winding A three-phase winding consists of three phase coils with their axes physically separated by 21/3 electrical radians or 2/3 of a pole-pitch. Figure 4.13 shows a typical winding with two conductors per phase. For convenience, winding diagrams are developed as in Fig. 4.14. Current and flux patterns are established when the winding is energized by balanced three-phase currents. If different intervals in time are considered the resultant flux pattern appears to move along the winding. see Fig. 4.14 and Section 4.2.4. This travelling wave of flux (usually called a rotating magnetic field) moves two pole- pitches in the time taken for the phase currents to alternate one com- plete cycle. An analytical treatment shows (a) the resultant flux distribution is sinusoidal if an individual phase winding has a sinusoidal flux distribution and a linear flux current relationship, and (b} he flux pattern travels along the winding at a constant speed called the svichronous speed.16 GENERALIZED ELECTRICAL MACHINE THEORY Additional harmonic travelling waves of flux are present with a ron- sinusoidal flux distribution. The harmonic flux components move se direction, at sub-synehronous along the winding. some in a revers speeds. time te time ty time t, time fete a Current vari. with tin sion —Time for one eyele-———. Fig. 4.14 Polyphase winding. Movement of flux pattern with time A polyphase winding on the machine stator has stationary electri- cal connections to the phases. With a rotating polyphase winding the phases are energized from stationary brushes through slip rings on the rotor. (a) Equivalent coil representation The fundamental field produced by a polyphase winding can be produced by an equivalent number of magnetic poles rotating at synchronous speed. Consider a two- pole polyphase winding, Fig. 4.15. A similar magnetic field is pro- duced by an equivalent coil energized by d.c. and retating at synchronous speed. Fig. 4.15¢b). Alternatively, a similar rotating field is established by a stationary commutator winding energized by brushes rotating at synchronous speed. This concept, 4.15(c). is extremely useful as a physical picture of the de. throu shown in FigBASIC EEPCERICAT MACHINES: anda equivalence of a commutator winding with rotating brushe: polyphase winding. The transformation of a polyphase winding to a pscudo-stationary coil system is an important part of generalized machine analysis. Rotating fie Rotating field (a) Stationary polyphase (b) Rotating coil (c) Stationary winding energized by energized by commutacor winding balanced polyphase dircet current energized by d.c currents through rotating brushes Fig. 4.15 Equivalent coil representation of a polyphase winding 4.2.4 M.M.F. and flux density diagrams Consider a machine with a uniform air-gap containing a single turn spanning 180 . as shown in Fig. 4.16(a). The m.m.f. Facting round a closed path contait current /, is given by i= $ H.d/ (4.4) where H is the magnetizing force at any point It will be assumed that: (a) the relative permeability of the ferromagnetic members is large and the magnetizing force is totally employed in driving flux across the air-gaps, (b) the air-gap flux is radial and field distortion in the vicinity of the conductor is ignored. (c) the conductor diameter is negligibly small, and (d) the gap length is small with respeet Co the rotor diameter and the flux density is constant along a radius,78 GENERALIZED ELECERICAL MACHINE THEORY The coil flux crosses the air-gap twice. The flux leaving the rotor surface sees an air-gap reluctance equal to that for the air-gap crossed Flux pattern | Flux density \73. t B, (a) Physical arrangement (b) Developed diagram Fig. 4.16 A single full-pitch turn in a uniform air-gap by the flux entering the coil. Hence, from (4.4), the mmf. F, required for each gap is half the total coil m.m.f. F. Foi, F=5 57 Meg and where g is the gap length gap flux density B, = woH, = Uo A (4.5) and is constant for a uniform air-gap. The m.m.f. and flux density diagrams are shown in Fig. 4.16(b). The convention adopted is positive m.m.f. for flux leaving the rotor surface. (a) Saliency Consider projecting pole-pieces on the outer member, see Fig. 4.17. The air-gap reluctances in the path of the flux leaving and entering the rotor are equal if the pole-pieces are identical to cach other. Hence the gap m.m.f.s are equal, see Fig. 4.17(b). The non-uniform idealized flux density ata point is obtained from (4.5), including the appropriate gap dimension for g.BASIC ELECTRICAL MACHINES 8 The m.m.f. and flux density diagrams may be extended to complete windings with any number of pole-pairs. Actual machines have symmetrical windings and pole-pieces, and the assumptions apply without appreciable errors, ignoring slotting effects. Short-pitched coils, spanning less than 180°, do not provide any difficulty, The current pattern for a symmetrical winding with short-pitched coils remains identical over every pole-pitch, and the winding may be considered as a group of full-pitched coils. Flux pattern (a) Physical arrangement (b) Developed diagram Fig. 4.17 A single full-pitch turn in an air-gap with saliency on one member Two examples are now given for typical machine arrangements. (b) D.C. machine on load Figure 4.18 shows developed diagrams for a machine with a salient-pole stator winding and a distributed rotor winding. This corresponds to a d.c. machine on load. In practice the flux pattern is distorted, non-radial, and results in the fairly smooth flux density curve shown in Fig. 4.18(c). (c) Distributed polyphase winding Figure 4.19 shows a double- layer, three-phase distributed winding with nine slots per pole-pitch and short-chorded by one slot. The winding is energized by balanced currents and developed diagrams are shown in Fig. 4.19(a) for an instant when the current in phase A is a maximum. With an increasing number of conductors per phase. the steps ofso GENERALIZED ELECTRIC) MACHINE THEORY smooth ‘alled sult ina relatively m.m.f. become less pronounced and m.m.f. waveshapeas shown in Fig. 4.!%b). Such a winding isc unijormdy distributed. Two are shown: (a) current in phase Aa maximum (b) zero current in phase B. Salient-pole + winding (b) Idealized flux density \ (c) Actual flux density | Fig. 4.18 Flux-density and m.m.f. diagrams for a d.c. machine on load 4.2.5 Summary of winding properties The two physical arrange- ments of windings are (i) concentrated around pole-picces. ; 2. satient poles: and, (b) distributed in slots.BASIC PLECTRICAL MACHINES XI The behaviour of the windings ina magnetic sense may be divided into three groups a) a flux pattern which is stationary relative to the winding, e.g. a salient-pole winding. (b) a flux pattern which is stationary relative to brushes but not to the winding itself, i.c., a commutator winding, AGB AD oc WD | ©99000808888288000 | ©OOOOGGOV®BB®EBOOOOO | xO BO KE N Cc 7 B [ Phase A “ Prose” / Yo Phase BT tant mame LL (a) Two-pole. double-layer. short with nine slots per po horded winding -pitch Current in A phase a maximum Current in B phase zero NL 4 vtechorded winding (b> Unitormiy-distributed, she 4.19 M.m.f. diagrams for a three-phase winding#2 GENERALIZED ELECTRICAL MACHINE THEORY (c) a flux pattern which moves with respect to the winding, Le., a polyphase winding. The cage winding mentioned in Table 4.1 consists of conductors with a shorting ring at each end. A magnetic ficld moving relative to this winding induces voltages in the conductors and alternating currents circulate. The currents in adjacent conductors have a phase shift between them and the complete winding establishes a flux pattern which rotates at an angular speed dependent on the frequency of the induced voltages. This winding is in common use for induction motor rotors and is considered further in Section 4.3. Table 4.1 Summary of winding properties Flux pattern Winding STATIONARY relative CONCENTRATED COIL to WINDING Salient pole. d.c. energized Distributed, dic. energized COMMUTATOR continuous closed STATIONARY relative to BRUSHES ROTATING relative to WINDING Squirrel-cage Damper winding PULSATING relative DISTRIBUTED. energized with to WINDING single-phase a.c. 4.3 Types of rotating machine Torque is exerted on a machine rotor when offset magnetic poles are present (or induced) on the stator and rotor surfaces. The simple arrangement of Fig. 4.4 is inadequate for most practical rotating machines as the torque oscillates with rotation and produces zero net torque. An exception is the single-phase reluctance motor. Commutator or distributed polyphase windings enable the simple device to be modified so that constant misalignment of the two fields is retained with rotation. Continuous uscful torque is exerted on the rotor. Most of the possible winding combinations are used in practice. The three common arrangements are: ee eeeBASIC PEECTRICAL MACHINES x2 (a) a commutator winding with a salient-pole stator. eg. a dc. machine, (b) a polyphase distributed stator winding with a concentrated coil a synchronous machine, rotor, i. (c) a polyphase distributed stator winding with a rotor having a distributed polyphase or a squirrel-cage winding, i-¢.. an indue- tion motor. 4.3.1. Commutator machines (a) D.C. machine The commutator machine, Fig. 4.12 in Section 4.2.2, is the basic form ofad.c. machine. The fundamental relation- ships, if the magnetic circuits are assumed linear, are as follows: torque % (rotor current)(stator current), i.e., % fi, induced rotor voltage ¥ (rotor angular velocity)(stator current) % @,/, The machine motors if electrical power is supplied to the rotor circuit and mechanical power is taken from the shaft. Sufficient rotor current flows to produce an electromagnetic torque equal to the mechanical load torque, neglecting losses. The supply voltage equals the rotor induced voltage, neglecting the resistance drop. which governs the running speed. If mechanical power is supplied to the shaft, the rotor speed in- creases until an equal electrical power is generated by the rotor current, neglecting losses. This power may flow into a passive load or regenerate into the rotor supply source. (b) Single-phase a.c. commutator motor The principle of torque production applies to the machine shown in Fig. 4.12 with both stator and rotor windings energized with single-phase a.c._ Induct- ance of the windings modifies the machine behaviour however and additional windings. brush-shift. etc. are often necessary for satis- factory performance. (c) Cross-field machine A d.c. machine with an additional brush- ield machine. Dillerent machine to act as a power pair on the direct-axis is called a cro: winding combinations now enable thi amplifier, constant current device. ete4 GENERALIZED PEECTRICAT MACHINE FHIEORY 4.3.2. Synchronous-speed machines (a) Svnchranous machine Consider a two-pole machine with a three-phase distributed winding on the stator and a concentrated coil winding on the rotor. The stator winding is energized by balanced three-phase currents ofangular trequeacy « which establish a magnetic field rotating at « rad/s in an anticlockwise direction. The rotor winding. in this machine arrangement called the field winding, is energized by d.c. Figure 4.20 shows a machine sche- matic. The rotor movement is anticlockwise and with an angular Instantaneous lines of action of rotor and stator fluxes. field winding Fig. 4.20 Synchronous machine schematic-stationary three-phase winding velocity of w rads the rotor and stator magnetic fields rotate to- gether at a common speed, viz. synchronous speed. The two fields interact and, in general, are misaligned. Torque is exerted on the rotor field to try and align it with the stator field. This torque is transmitted by the rotor field to the rotor itself. which consequently develops mechanical power as it is running at synchronous speed. On no-load with steady running conditions and neglecting losses, the rotor and stator individual tields are aligned and there is no out- put power. An applied mechanical load torque decelerates the rotor until the magnetic field misalignment produces sufficient electromagnetic torque to balance the applied mechanical torque. The device is called a synchronous machine as the rotor and stator magnetic fields have to be synchronized. i.e. rotate at the same speed, before useful torque is present. With different angular velocities of the magnetic fields, the torque pulsates with time and its mean value is zero. At synchronism, energy conversion is possible in. both directions (when the three-phsse winding is connected to an appro-BASIC PERCTRICAL MACHINES, ss priate supply). i.e. electrical power may be generated from. or con- hanical power to basic torque concepts discussed in Section 4.1.1. verted into, me With referen it is seen that the polyphase winding enables a rotating machine to have a steady flux pattern established which includes misalignment of the component fluxes and is capable, therefore, of exerting torque. The practical requirement for a rotating machine mentioned in Section 4.1.3, i.c.. misalignment maintained with rotation, is pro- vided by the polyphase winding. The induced voltage in the three-phase winding depends on the rotor speed and the magnetic field strength. The speed depends on the supply frequency. Consequently, for a fixed supply frequency the induced voltage in the stator winding is proportional to the magnetic field strength The machine on no-load requires zero inphase or wattful com- ponent of stator current as no power is required. neglecting losses. If the supply voltage and the stator induced voitage are not equal. a reactive current circulates. The twofold effects of this current are: (a) a change in the resultant magnetic field, called the armazure reaction effect : (b) an inductive voltage drop in the inherent leakage reactance of the winding. On load the power factor of the stator current is controlled by the relative values of the supply voltage and the open-circuit stator induced voltage. With a constant supply voltage the field winding excitation controls the machine power factor. The machine will operate as a generator when connected to a passive electrical load. The induced or generated stator voltage supplies load current; no synchronism is required. Oftena machine operating in this mode is called an alternator. The field winding may be either distributed in slots (round-rotor machine) or concentrated on pole-picces (salient-pole machine). The former has a nearly uniform air-gap but with salient-poles the reluctance varies considerably with angle, producing a relatively strong reluctance torque in addition to the cylindrical co-alignment torque. In machine analysis it is often convenient to interchange the rotor and stator windings as shown in Fig. 4.21. Representing the poly- phase winding by an equivalent coil representation, Fig. 4,22 gives86 GENERALIZED ELECTRICAL MACHINE THEORY a commutator machine with brushes inclined to the stator pole axis. An accurate representation is possible by brush-pairs along the a field LK winding fo Fig. 4.21 Synchronous machine schematic-stationary field winding Stationary field pattern direct and quadrature axes. Fig. 4.23. A detailed treatment is given in later chapters. b rotating commutator winding field field winding winding ose Zi / Stationary brushes Pseudo-stationary coil . 4.22 Synchronous machine—equivalent coil representation (b) Reluctance motor A salient-pole synchronous machine with the field winding omitted is called a reluctance motor as it relies on reluctance torque only. Polyphase reluctance motors are rarely used. Single-phase reluctance motors rely on the contra-rotating graxis fr deaxis Ly Fig. 4.23 Synchronous machine-equivalent coil representation along axes at right-anglesBASIC ELECTRICAL MACHINES 7 fields produced by a single-phase winding and are suitable for electric clocks, ete. 4.3.3. Asynchronous machines (a) Induction machine Consider a two-pole machine with a three- phase short-circuited rotor winding and a three-phase distributed stator winding energized from a supply of angular frequency « rad/s. The stator magnetic field rotates at rad/s. With the rotor running Rotor field Rotor field relative to relative to stator > =\ \ \ so o, +50 w rotor Stator! field Fig. 4.24 Induction motor magnetic fields below synchronous speed at «y, electrical rad/s, the stator rotating field cuts the rotor surface at (« —w,) rad/s or ws rad/s, where s de- fined as the ratio (w—@,)/@ is called the fractional slip. Relative motion of the rotor surface and the stator magnetic field induces rotational voltages of angular frequency se in the rotor phases. These voltages drive currents round the closed rotor circuits. With a symmetrical rotor winding the rotor phase voltages and currents are balanced and of angular frequency sw. The rotor cur- rents flowing in a three-phase winding produce a rotating field of angular velocity sc rad/s relative to the rotor surface and moving in the same direction as the rotor. This is shown in Fig. 4. quently, the rotor magnetic field rotates at w rad/s rela: stator, i.e., at the same speed as the stator magnetic field 5 rotor speed. The frequency of the rotor voltages and Jurrents. times supply frequency, is referred to as slip frequency Interaction of the rotor and stator magnetic fields produces electromagnetic torque on the rotor which is in the direction of rotor . a motoring action. The torque depends on the movement, ixk GENERALIZED ELECTRICAL MACHINE THEORY angle of misalignment between the component ficlds. Figures 4.25(a) to (c) show that the stator and roter individual magnetic (a) inclined at right-angles with a purely resistive rotor circuil, i.c., maximum torque for given field strength: (b) in direct opposition with a purely inductive rotor circuit, i.e. zero torque. instantaneous instantaneous: induced voltage ze current Torque maximum torque 1b) Purely resistive ta) Resistive-reactive rotor rotor no o torque Fig. 4.25 Induction motor. Instantaneous rotor and stator magnetic fields In general. the angle of inclination is equal to(= 2+ @) where cos 9 is the power factor of the rotor circuit. Machines with a polyphase winding on the rotor are called wound- rotor induction motors. Squirrel-cage motors have a rotor-cage winding. The mode of action is identical for both constructions. Motoring action occurs at speeds below synchronous speed. The torque action is zero at synchronous speed as there is no rotor induced voltage or current. Above synchronous speed the rotor induced voltage circulates slip-frequency currents. The fractional slip » is\ BASIC PLECERICAL MACHINES xo now negative and the current polarity is reversed producing an electromagnetic torque opposing the driving mechanical torque. The machine regenerates into the three-phase supply connected to the stator winding. With no energization of the stator there is no rotating field to induce voltage in the rotor winding and generating action is impossibie. 4.3.4 Other machines One or more of the electro-mechanical actions described for the machines in the preceding sections are embodied in aJJ rotating electrical machines. PROBLEMS 4.1 Describe the two basic techniques of torque production used in all rotating electrical machines. Indicate positions of maximum torque and show how a commutator winding enables continuous torque to be developed. A six-pole, 36-slot, three-phase stator is energized by balanced alternating currents. Draw the m.m.f. diagram over two pole-pitches at the instant of zero current in one phase. Describe how the m.m.f. pattern varies with time. 4.2, Show that both two- and three-phase systems of balanced m.m.fs applied to their appropriate a.c. windings produce a rotating magnetic field. 4.3 Describe the two physical concepts of torque production in a rotating electrical machine. Show, with the aid of diagrams. the winding arrangements of the following machines. State their essential features and briefly indicate how continuous torque is exerted in each case. (a) Commutator machine (b) Synchronous machine (c) Reluctance motor 4.4 Describe the principles of torque production in a d.c. machine in which the salient-pole stator winding and the commutator winding are both energized with steady current, How will the torque vary with angular displacement of the brush- pair with respect to the stator coil axis” 4.5 A uniform air-gap machine has a four-pole, 24-slor. three-phase two- layer winding on the stator energized by balanced alternating currents The winding is short-chorded by one slot (a) Draw the m.m.f. and Mux density dia at the instant of maximum current in one phase rams over two pole-pitches90 46 4.7 GENERALIZED ELECTRICAL MACHINE THEORY (b) Estimate the maximum flux linking one turn when the phase current is LOA rms. tive length of stator lom {ffective stator bore diameter Am Air-gap Tem 10 conductors/slot, and all phase conductors in series. The effect of slotting may be neglected [Ans. 0-45 Wb] A dc. machine has a lap-connected, four-pole, commutator winding with 400 conductors. Plot the m.m.f. diagram for a total armvature current of 100 A. Calculate the magnetizing (d-axis) and cross-magnetizing (q-axis) ampere-turns with the rotor brushes inclined at, (a) 10 clectrical degrees, and (b) 10 mechanical degrees to the neutral axis (q-axis). [Ans. (a) 1441, 139. (b) 972, 278] A four-pole, d.c. machine has 1200 turns on each pole winding, and the effective spread of a pole is 72 mechanical degrees. The commutator winding is wave-connected and has 768 conductors. Sketch the m.m.f. diagram over two pole-pitches for (a) a field current of 2-4 A, (b) a rotor current of 30 A, and (c) a combination of (a) and (b). Assume a uniform distributed winding for the rotor m.m.f. A current of 2-4 A in the field winding produces a maximum flux density of 0-5 Wb m? in the machine air-gap. The relative permeance of the air-gap is symmetrical about a pole-piece and has the following values, angle (mechanical degrees) 0 9 18 27 35 45 (polecentre-line) relative permeance 010108095 1 1 Sketch the flux density diagram for case (c). assuming radial flux and no saturation, and the rotor brushes (d) on the q-axis. or neutral axis, and (e) at an angle of 18 mechanical degrees to the q-axis.5. Elements of generalized theory The purpose of this chapter is to describe in mathematical terms the physical principles governing the action of most conventional rotating electrical machines and by so doing, develop a basic theory from which the performance of all such machines can be determined 5.1 Simplifying assumptions The theory, which is known in current machine literature as the generalized theory of electrical machines. considers all practical machines to be idealized by the simplifying assumptions listed and described below: (a) Magnetic saturation is neglected. Superposition of magnetic fields can be used, and the self- and mutual inductances of all machine windings are independent of the magnitude of winding currents. (b) Air-gap m.m.f.s and fluxes are represented by the fundamental components of their spatial distributions which are symmetrical about the magnetic axis of the windings of origin. (c) Slotting effects are ignored. Distributed windings comprise finely spread conductors of negligible diameter. (d) Commutation is ideal. The width of brushes and commutator segments is negligible and current reversal is instantaneous dur- ing commutation. (e) Magnetic materials are tree of eddy current and hysteresis losses 5.2 Concentrated and distributed winding inductances It is shown in subsequent sections of this chapter that the perform- ance of most conventional rotating electrical machines, following idealization, can be described by a set of linear equations comprising the voltage equations for windings. together with the equation for0 GENERALIZED ELECTRICAL MACHINE FHEORY neral electromagnetic torque. The cquations, known as the machine equations, hay ¢ cocflicients comprising the self- and mutual of machine windings. Thus, before proceedin: inductance the performance analysis of rotating machines in general. exp are required for the inductances of windings in common use with machines. Physical arrangements of practical machine windings are of two basic types, see Section 4.2.1; those which are distributed over a magnetic surface and housed in slots are known as distributed wind- ings, while those which assume the form of a multi-turn coil are called concentrated windings. The magnetic properties of con- centrated and distributed windings are considered in some detail in Sections §.2.1 and 5.2.2. Expressions for self- and mutual induct- ance, first determined for non-sinusoidal air-gap flux-density distributions, are subsequently modified to agree with the simplifying assumption of a fundamental air-gap flux density distribution. Finally, a distributed winding is shown to be transformable to an equivalent concentrated winding enabling all windings to be repre- sented in this one basic form. 5.2.1 Concentrated rotor coil self- and mutual inductance A simple two-pole machine, Fig. 5.1. comprises a cylindrical rotor of radius r and length /, and a salient-pole stator. Concentrated salient-pole windings of N, turns supplied with current i, provide air-gap excita- tion. The rotor coil of .V, concentrated turns is short-chorded and of angular pitch 2f. A d-axis and q-axis. in space quadrature, designate main axes of q-axis deaXis Nawrn dx coil axis i, turns Fig. 5.1 Concentrated short-chorded rotor ¢ELEMENTS OF GENERALIZED THEORY 93 symmetry for the machine and the rotor coil is located with its magnetic axis an arbitrary angle 6, electrical radian to the d-axis. (a) Mutual inductance The developed diagrams, Fig. 5.2. show the gap distributions of m.m.f. (a), and flux density Bla), to be sym- metrical about the d-axis, sce Section 4.2. Taking the d-axis as 4 | N Co Stator m.mf. distribution Fla) | yo fundamental gap flux density 3rd harmonic Fig. 5.2. Gap distributions of m.m.f. and flux density (stator excitation only) reference and uming B(x) to be a rectangular distribution of amplitude B,, and angular width (n—2y), B(a) can be described by the following Fourier-series expansion 4 I 1 Bla) = — B(cos 7 Cosa COS 3; cos sats cos 57 cos Sx 1 7 —jyoee ae cos (2 = Dy cos Qn— Nats) (5.1) 2n-1 4s feu! Bo =—-B, Yt, i cos (2n= 1); cos (20 = 1a 1X Qno4 GENERALIZED ELECTRICAL MACHINE THEORY Vhe harmonic crest values 4. (ne = — B, Ban = Be cos (2n—1); (5.3) diminish rapidly with ascending order n, sce Fig. 5.2. I is this fact which enables assumption (b). Section 5.f, to be made to a reason- able degree of accuracy in later work. Assuming the stator m.m.f. to be employed entirely in driving flux across an air-gap of length g and radial width (x — 2) Ho Bm = Holl = “* Nal (5.4) Substituting (5.4) in (5.2) Abo alee Bla) = g Nels 5 (& n= ~ cos (2n = Ly cos (21 = vy) oe Ba) =i, ¥ re! Wr Kay 1605 (21) (5.5) ee where the constant 40N, Koy = A=! cos (2n= 1p (5.6) "gy Consider an element of rotor surface r/ da, Fig. (5.1). The con- tribution made to the rotor coil flux-linkage a, == NirB(a) da (5.7) Total rotor coil flux-linkages a=MetB (ne (yet WY, = Nile { ( x Koy, Cos (20 — by) da a=o-p Nat 21 a= © /(_yyett ¥,=2Ni lr > ( “F Ky,—, cos [(2n—1)0,] sin [@»-9)) nat (5.8) Mutual inductance 1S (Ses Rae 1608 [=H] sin [2D] (Qn~1)7 (5.9)ELFMENTS OF GENERALIZED THEORY 95 It is seen from (5.5) and (5.9) that a particular harmonic of mutual inductance is associated with a harmonic of flux density of the same order. A particular harmonic of mutual inductance can be elimin- ated from (5.9) using a rotor coil of angular pitch 2 determined from Qn-Dp =x whence YB (5.10) ~ Qn=1) For third harmonic elimination 1 =2 is written in (5.10) and the required coil span 2n 2p = = 0 f 3 Special cases of (5.9): (a) Full-pitch rotor coil B=x/2 K. Ks M,, = 2nd cos 0,+ 3 cos 30, + = cos SO,4 0° ) (5.11) (b) Fundamental flux density distribution only. n= 1, M,, = 2N,IrK, sin B cos 0, = M,,sin Bcos0, (5.12) (c) Fundamental flux density distribution only and full-pitch coil p=n/2 M, (b) Self-inductance For the machine structure shown in Fig. 5.1, the rotor coil m.m-f. acts on an air-gap permeance varying in a com- plex manner with changing position of the rotor coil. Resulting from the permeance variations is a gap flux density distribution which changes in magnitude and symmetry. A typical gap flux density distribution is shown in Fig. 5.3. The very difficult task of writing an equation for the gap flux density distribution prevents using a direct process of integration for finding rotor coil flux-linkages. hence self-inductance. An alterna- tive procedure, employing ¢ ssible approximations, is described below. Consider the rotor coil to be full-pitch and its rectangular m.m.f. distribution to be described with sufficient accuracy by its space distributed fundamental F, of crest value F,. Resolving F, along = 2N,IrK, cos 0, = M,, cos 8, (5.13)96 GENERALIZED ELECTRICAL MACHINE THEORY the d-axis and q-axis gives instantancous crest values, Fy and £,,. of {wo space stationary, sinusoidally distributed m.m.f.s Fy and Fy respectively. Acting on paths of different permeance. Fy and F, produce space stationary, non-sinusoidal gap flux density distribu- tions which are symmetrical about their respective axes of reference. These can be represented with sufficient accuracy by their funda- mental! components of spatial distribution B, and B, of crest value q-axis gaaxis rotor gap flux density coil axis rotor m.m.f. geaxis | d-axis graxis i deaxis, ‘Two-axis air gap distributions of m.m.f, and fluy density (rotor excitation only) feaELEMENTS OF GENFRALIZED THEORY o7 B,,and B,,. see Fig. $.3. ‘Thus, the machine is considered to po: two fictitious uniform air-gaps, one of length g, associated with the d-axis permeance and another of length g, associated with the q-axis permeance. As the permeance of the d-axis is greater than that of the g-axis, g,>g4. Resolving F, along the d-axis and g-axis respec tively Fy = F, cos 0, (5.14) Meet ee etd YL LL LILLE ELL LLL hypothetical 44 9a uniform stator | surfaces \ de rotor surface Fig. 5.4. Hypothetical uniform two-axis air-gaps With the d-axis as reference, the m.m.f. components acting at a displacement « F,, = F,4cos 4 = F, cos 0, cos « 5.15 Fig = Fog COS («+3) = F,sin 0, sina a Referring to Fig. 5.4 The permeance of an element of area r/ dz aa, ~ Hols Ga 5.16) dA, = Hor! do ‘ Gq Element of flux-linkage with the rotor coil d¥, = N, (dA, Fu tdAy Fy) Ky F = N,fyrl dof 74 4° (5.17) Ya 4gON GENERALIZED ELECTRICAL MACHINE THEORY Permeance for a full coil span 4, = tetlt o (5.18) A, = Min whence da a, = NS (AgFuat Aga) (5.19) X Substituting (5.15) in (5.19) and integrating for total rotor coil flux- linkages EN, [2=0e+H2 . Ce £m f (Aq cos 0, cos 2+ A, sin 0, sin x) da (5.20) m de =0,-K/2 _ 2F,N, = Ne Ay cos? 0,4 A, sin? 0,) F,N, Y, = EN [Ay A) +(Ag— Ay) 605 20,] 6.21) Referring to Fig. 5.3: The rectangular m.m.f. distribution is of maximum value N,i, pa en ” 2 A Fourier series expansion for the rotor m.m.f. distribution can be written directly from (5.2) with y=0 and « replaced with («—0,) as the coil is full-pitched and the m.m-f. distribution is symmetrical about the axis of the rotor coil Fa— oy = *F, 5 ( nt net -1 cos (2n— I)(a— 0) Assuming the rectangular m.m.f. distribution to be described with sufficient accuracy by the space fundamental F(x—0,) = *k, cos (x—0,) (5.22) The crest value for the fundamental, obtained when «= 0, (5.23)ELEMENTS OF GENERALIZED THEORY 99 Combining (5.21) and (5.23) 2N2 ¥, =F i [(Aat Ay) + (Aa Ay) 608 24,] (5.24) Rotor coil self-inductance Y L, = 1 = Ly+L,2 cos 20, (5.25) i where N?(Ag+ Aq) and (5.26) Ni(Aa~ Aq) As seen by (5.25) and Fig. 5.5, the self-inductance of a concentrated coil rotating within a salient-pole structure comprises a second harmonic term superimposed on a constant term. d-axis Fig. 5.5 Variation of self-inductance of a rotor coil maving within a salient- pole structure 5.2.2 Distributed rotor winding self- and mutual inductance Figure 5.6 shows the surface of the cylindrical rotor covered with a uniformly distributed winding of Z series connected conductors per electrical radian. Initially, the winding is connected single-phase at dia- metrically opposite tapping points Pand P’. The magnetic axis of the single-phase winding is a line joining points P and P’ through the rotor centre displaced an angle 0, from the d-axis. (a) Mutual inductance With the stator salient-pole winding excited, the gap flux density distribution is given by (5.2) with the d-axis as reference. Referring to Fig. 5.6, the two elements of angle dj. each100 GENERALIZED ELECTRICAL MACHINE THEORY containing Z df conductors, are considered to form a concentrated coil of Z dp turns and angular-pitch 2p. The elementary coil fux- linkage can be written directly fri (5.8) with N,=Z df as /¢_pytd dW, = 2Z dp idr ¥ (5° iP Ky, —, cos (20 = 1)0, sin (2n— oy) nel eee (5.27) qraxis deaxis dp i rotor winding axis Fig. 5.6 Uniformly distributed rotor winding The total flux-linkage with the single-phase distributed winding _ net Y, = 2Zilr y (ae 1 = Koy 1 Cos (2n—1)0, sin (2 298) a8 n— =O n=1 (5.28) K,,-,c082n—1)0, } (5.29) n+l y v= SZir 5. ( nat Mutual inductance Wee (1! con M, = — =4Zlr Y (er Ko 1 cos (2n— 11H, } (5.30) s n= Expanding (5.30) Ky Ks M,, = 4Zir| K, cos 0,— 3 cos 30,4 gi 608 50, (5.31) Comparing (5.11) and (5.31) shows, for identical flux density dis- tributions, the amplitude of ascending order of mutual inductanceELEMENTS OF GENERALIZED THEORY lon harmonies decrease more rapidly for a distributed winding than for a concentrated winding. As for the concentrated winding, a harmonic of stator gap flux density is associated witha corresponding harmonic of mutual inductance. With the gap density distribution represented by a space fundamental M,, = 4ZIrK, cos 0, = M,, cos 0, (5.32) Referring to Fig, 5.6, the rotor winding is now considered 10 be connected three-phase at tapping points a, band c. To determine Ss phase spread Fig. 5.7. Expanded view of phase aa’ conductors an expression for the mutual inductance of any one of the three phases, (5.28) is employed with a change in the limits of integration and the magnetic axis of the phase taken as reference. Considering phase aa’ in Fig. 5.7, the changed limits of integration for (5.28) are B=(2n)/3 and B=2/3. The total flux-linkage with phase aa’ Vy. = 2Zi,lr ee a 3 (Ga 2 x os Qn—1) = cos Qn—1) 3h (5.33) an1 608 (20190, sin Qn—1)8 | dp Ky,-, cos (2n—1)0, Expanding (5.33) K. K May = 27 &, cos 0, +2 ay cos + . cos 50,+--~)102 GENERALIZED ELECTRICAL MACHINE THEORY with the axis of phase bb! at (@, — 27/3) 3 K + FP cos (0. ) . | (5.34) with the magnetic axis of phase cc’ at (0,— 22/3) I\ Ky af On = 2211 cos (0.277) 2 7 cos (03) ¢ 2 +E cos (0.47% )4 . al Assuming the stator gap flux density distribution to be described with sufficient accuracy by its space distributed fundamental and with M,, defined by (5.32) ” (On\ . K. 2 My: = 22k, cos (4 3)? 71 60s x 0-3) M,,y = 2ZIrK, cos 0, = My cos 4, 2 Myy = 2ZIrK, cos (.-3) =~ M,, cos (.-) (5.35) 2: 1 2; M.,. = 2ZIrK, cos (0.43) = 5 Mp cos (0+) (b) Self-inductance The developed diagram, Fig. 5.8, shows the uniformly distributed rotor winding, supplied at tapping points P and P’ with current i,, producing a triangular m.m.f. distribution. Enclosed by the dotted path is a total current Zz, and the triangular m.m_f. distribution is of maximum value Zi po zie G 2 (5.36) With the winding axis as reference (}=0). the m.m_f. distribution is described by the Fourier-series 8 F(B) = ) (5.37) ee lrrrCs—ses—Sei—e—C—NCsCOSCZSC;izazéOLO“OC =C«za 1 \ ofcos B+ 33 08 3f5— 5, 008 S++ELEMENTS OF GENERALIZED THEORY 103 Considering the triangular distribution to be represented with sufficient accuracy by the fundamental 8 FM) = | in COS fb The fundamental crest value (5.38) Since F, can now be resolved into d-axis and g-axis components, the method employed in Section 5.2.1 for finding the self-inductance of a winding qraris d-axis axis A 4, 7 re 5 ay actual rotor m.m.f. uniformly 2 fundamental m.m-f distributed va winding rotor surface Fig. 5.8 M.m.f. distribution for a uniformly distributed winding concentrated winding, can be used to find the self-inductance of a distributed winding. Referring to Fig. 5.8: The flux-linkage for the elementary coil of Z df turns and angular- pitch 28 can be found using (5.20) with a change in the limits of integration = Ont av, = zB) 7 [ (A, cos 0, cos x+ A, sin 0, sin a) da TH Jamnnp Zz =8 a dBi,( A, cos? 0,+ A, sin? 0,) sin f (5.39)lod GENERALIZED ELECTRICAL MACHINE THEORY Total flux-linkage with the rotor distributed winding 2 Wen Vos i {Ay cos? 0,4 A, sin? mf sin f dp Lae =o 82? Yoox = iL (Aut A,) + (Aa Aq) 608 20,] (5.40) Self-inductance L, = Lj +L, 608 20, (5.41) where 8 2, Ln = 3 2 Agt Ay) and (5.42) 8 La = 2A, A) Comparison of (5.24) and (5.40) shows for an assumption of sinusoidally distributed gap flux density, the self-inductance ex- pressions for a distributed and concentrated winding are of similar form. 5.2.3 Concentrated and distributed winding equivalence Sum- marizing the derived expressions for the self- and mutual inductances of concentrated and distributed windings with an m.m.f. distribution assumed sinusoidal: concentrated full-pitch rotor winding; from (5.13) and (5.24) M,, = 2N,IrK, cos 0, 2 L= a N?[(Aat Aq) + (Ay— 4) 608 20,] uniformly distributed rotor winding; from (5.32) and (5.40) M,, = 4ZirK, cos 0, 8Z b= 4 [Aa Ag) + (Ag~ Aq) £08 20,] PeELEMENTS OF GENERALIZED THEORY, tos Equating separately the expressions for self- and mutual inductance, blishes the result that a concentrated and distributed winding arc magnetically equivalent when =2Z (5.43) With (5.43) established, distributed machine windings are repre- equivalent concentrated coils. The diagrammatic form of shown in Fig. 5.9. the convention adopted for is that both pass through the coil in the same sented a this representation i: current and m.m.f. direction. Q222Qg Z conductors A per elec. rad. Fig. 5.9 Diagrammatic representation of equivalent concentrated’ and distributed windings 5.3 Concentrated coil representation for a rotational transducer Figure 5.10 represents a doubly-excited. two-pole, rotary transducer with concentrated stator and rotor windings. Externally applied qraxis mS a = Fig. 5.10 Concentrated cuit representation of a doubly-excited rotary transducer106 GENERALIZED ELECTRICAL MACHINE THEORY time-varying voltages », and r, produce time-varying currents, m.m.f.s, and fluxes. The rotor revolves clockwise at an angular velocity «, electrical rad/s and an instant is considered when the rotor and stator magnetic axes are separated 0, electrical radian While a two-pole transducer has been shown for ease of repre- sentation, the more genera} case of a transducer with severa) pole- pairs can be considered by writing 0, = (pole-pairs) Oy . (5.44) @, = (pole-pairs) «,, where ? @, and 0, are measured in electrical radian, and @,, and 0,, in mechanical radian. As the action of a multi-winding machine is the superimposed effect of winding pairs, equations of torque and voltage, derived for the transducer model Fig. 5.10, will form the basis of a generalized theory for rotating electrical machines. 5.3.1 Induced voitages The total flux-linkage of a coit comprises the sum of self- and mutual flux-linkages. Stator flux-linkages Y= Lit Mei, (5.45) Rotor flux-linkages Y, = Lyi, + Mui, (5.46) Instantaneous applied coil voltages v, = i,R,+p¥, (5.47) v, = iR,+p¥, (5.48) d where p= - 2, = i,R, + pLyi, + M,,i,) (5.49) v, = i,R,+p(L,i, + Myi,) (5.50) For generality it is assumed in (5.49) and (5.50) that L, and L, are functions of the angle #),, hence time ¢. Expanding. ty = GR + hpi + pl, + Mp, + ipM,, (5.58) v, = i,R,+L,pi,+ ipl, + M.,pi,tioM (5.52)ELEMENTS OF GENERALIZED THEORY 17 Considering induced voltage terms which are time-derivatives of inductance and taking as an example é,pL, where Rewriting (5.51) and (5.52) by dL dM. RL oj oes ay | AM 5 iRo+ Lypi,+ M,,pi, + Orig gg tee wi, (5.54) dM. dM, 53 Us (5.35) It v, dL, ER, + LePip + MaPigt Only Go" + Ol In matrix form []- ie elle tbr aela OB] b, o RI LM, L, "d0,\M, L, |fli (5.56) Induced voltage terms in (5.56) are seen to be of two distinct types: Those whichare functions of the time-derivative ofcurrent are mutual and self-induced voltages while those which are functions of the angular derivative of inductance are rotational voltages. Writing in (5.56) [R] = [‘s 2 (5.57) [4] = e | (5.58) (= ale, lao! with i-[e] om [1] [= TAL Hep rofe RU (5.00)108, GENERALIZED PLECERICAL MACHINE THEORY The physical significance of the rotational voltages can be more readily understood if the coefficients of the [G ] matrix are expressed in terms of permeance. Commencing with My = NN Ay Ly = N2A, (5.61) L, = N?A, where A,,. A, and A, are the gap permeances for mutual flux and stator and rotor self-flux. Writing the [G] matrix in terms of (5.61) dA dA 2 dA, 1s, . do, NN, do, Glee {c] wn da, yo lds (5.62) dO, It is seen from (5.62), rotational voltage terms are the result of permeance variations during rotation. The stator coil of a cylindri- cal rotor machine sees constant gap permeance, i.e.. dA,/d0, =0, the rotor coil of a cylindrical stator machine sees a constant gap permeance and d4,/d0,=0. For all stator and rotor configurations the effective area of the air-gap between the stator and rotor coils changes with rotation and A,, is a function of 0,. Substituting the mutual and self-inductance values established in Section 5.2 in (5.58), (5.59), and (5.60) gives the stator and rotor voltage equations. Case (a) Zero saiiency —~ cylindrical rotor and stator : L, = constant L, = constant M,, = M,, cos 0, ey) ffR, OF f ky M,, cos 0, rf ILO R, on a. 0 —M,, sin 0, If i +0 7 - M,, sin 0, 0 fLEEMENTS OF GENERALIZED THEORY log Case (b) One degree of saliency cylindrical retor and a salient- pole stator: = Ly +L,: cos 20, L, = constant M,, = M,, cos 0, ie R, 0 i M,,cos 0, re] Lo Rf * LM, cos, (2), +L,3 cos 20,) |” 0 —M,, sin 0, ]{fi —M, sin 0, —2L,, mR] oo The expressions for a machine with a salient-pole rotor and a cylindrical stator are obtained by interchanging s and r in (5.64). +0,] of machines with uch machines are Case (c) Two degrees of saliency. The analysis stator and rotor saliency will not be considered as s rarely encountered in practice. .2 Equivalent circuit With the rotor of the doubly-excited ansducer maintained stationary the coefficients of the [G] matrix are zero, there are no rotational voltages and (5.60) reduces to the form normally written for the primary and secondary terminal s ofan ideal two-winding transformer. It follows, to produce a0 voltages M,, cos 0, R R, . ; mo fy > ~ 5 4 rn Mi, sin 8, \ OM pig Si0 0, : ‘ ‘ L Ly $L,008 20, Y2o,L,, sin 20, Fig. 5.11 Equivalent circuit for a doubly-excited rotary transducer with stator saliency an equivalent circuit for the transducer, from which the current and voltage conditions can be determined for different modes of opera- tion, it is only necessary to include rotational voltages in the equivalent circuit of the two-winding transformer. Rotational voltages are represented as voltage generators of zero internal enerators are assigned instantaneous Impedance. The voltage polarities according to the sign of the rotational voltage terms.Ho GENERALIZED ELECTRICAL MACHINE THEORY Figure 5. ! fis the equivalent circuit for a doubly-cxcited transducer I with stator salicney, constructed in accordance with (5.64). 5.3.3 Electromagnetic torque Referring to (3.53), the energy stored within the air gap of a doubly-excited rotary transducer Vn, da oo. Wye = 5 BLAS Ly bid Moy Electromagnetic torque dW Te = G0, . dW, = (pole-pairs) “a0, Let iadle Mu = (pote pais So 5 at eae, ) (5.65) Tn matrix form dL, dM,, (pole-pairs). _.] dO, dd, T= [et] aM, dt, (5.66) “do, 40, or = eS el (6 Having constructed a transducer with a particular air-gap geo- metry and windings of specific numbers of turns, the [G] matrix is fixed and a change in the cyclic variation of T, can only result from changes made to the stator or rotor currents or both. Thus, the characteristic torque equation for a two-winding transducer with stator saliency is formed by writing in (5.66) the actual expressions for the G coefficients whilst leaving the currents quite general. In this way the characteristic torque equation can be used repeatedly to determine the torque characteristics for a variety of current con- ditions. Substituting the [G] matrix of (5.64) in (5.66) . _ (pole-pairs) 4 0 —M,, sin 0, || i, 7, = POPAMS Fi] 2 —M,,sin0, —2L,, sin 20, |] i, (5.68)ELEMENTS OF GENERALIZED THEORY mn ase of the stator and rotor coils supplied The angular variations of the Consider the particular with constant positive d.c. currents. torque components and the total electromagnetic torque is shown in Fig. 5.12 The following points of interest can be deduced from Fig. 5.12: (a) 0<0,<7 T.—ve torque opposes motion, generator action; —n<0,<0 T,+ve torque assists motion, motor action. (b) Motor or generator action can be achieved throughout the cycle if the direction of flow of the stator or rotor current is reversed at instants in the cycle 0,=0 and x. Stator coil axis Generator ner : —(pole-pairs) 2L,, sin 26, — 06, —(pole-pairs) ii, Moy Sin 0, Fig. 5.12 Variation of torque with rotation (c) A torque is produced with only one winding excited provided the non-excited magnetic member has saliency. (d) For given values of stator and rotor currents the peak value of T, increases with saliency. An example of interest considered in chapter 3 is one in which the transducer operates as a singly-excited reluctance motor. Assuming stator saliency and a rotor current = In, Sin (@,t +2) = —(pole-pairs) L,,/2, sin? ((o,1+ 2) sin 20,1 L T, = —(pole-pairs) 2p 5 nel |—cos 2(c,1 +4) ] sin 20,Wa GENERALIZED ELEC ERICAL MACHINE THEORY . 7 fe, a or 2 sin 20 ] —sin 20,4— “| d(,1) 2 wey Est! eosin 2a = (pole-pairs) —«,f sin 2x” &r Jow=0 ole-pai = — POE PANS) 78, sin 2a Mean torque is positive and a maximum when «= —45" a condi- tion achieved with a rotor current lagging the angular displacement 0, by atime phase of 45°. The reluctance motor is not self-starting and can only operate at synchronous speed. 5.4 Commutator machines An important group of a.c. and d.c. machines are those employing commutator windings. The simple two-pole commutator machine with stator saliency, Fig. 5.13. will serve as a model for determining the performance characteristics of commutator machines. For the q-axis \ pseudo- F. stationary 7 ‘Two-pole commutator machine quite general, the two purpose of making the subsequent an windings are assumed to be supplied with time-varying voltages and currents: the uniformly distributed commutator winding being con- nected to its supply by a brush-pair held stationary on an axis at 0, to the d-axis.ELEMENTS OF GENERALIZED THEORY 13 As previously described. Section 4.2, commutator action effec- tively divides the commutator winding along the brush axis to form two uniform current sheets of opposite polarity. From magnetic considerations the commutator winding behaves as a stationary con- centrated coil of N, turns carryingacurrenti,. For total equivalence the stationary concentrated coil generates the same rotational volt- age as the commutator winding. _ It follows, such a coil is fictitious and as it is to be used to represent commutator windings it is given the identifying name of pseudo-stationary coil. The two-pole commutator machine, Fig. 5.13, can now be repre- sented in diagrammatic form as two stationary concentrated coils. The two coils are considered to produce sinusoidally distributed air-gap m.m.f.s, space phase separated 0,, and in time phase with their respective currents. 5.4.1. Motor and generator action Before proceeding with the study of motor and generator action, it is necessary to establish a convention for coil currents and electromagnetic torque. Oo 5.14 Coil current convention Commencing with coil currents, an applied coil voltage is shown in every case with the positive potential assigned to the end of the coil which is nearest the machine centre. The reference currents, in this case i, and i,, are always shown to be positive, that is, entering at the positive potential end of coils. For the general case of current entering or leaving the point of positive potential, the current is represented by a separate symbol and equated to the reference cur- ig. 5.14 rent as shown in FM4 GENERALIZED ELECTRICAL MACHINE THEORY A similar convention is used for clectromagnetic torque where the reference torque T, is always considered to act in the positive clock- wise sense. For the general case of an electromagnetic torque T acting clockwise or anticlockwise, we equate it to T,,. see Fig. 5.15. tf i Fig. 5.15 Torque convention The merit of the above conventions, as will be shown later, is that a general set of machine equations can be formed in which the vari- ables are the positive reference values. A particular mode of operation can then be described by substituting for the reference variables the actual variables. Figure 5.16 shows two instantaneous commutator machine con- ditions which differ only with respect to the assumed directions of rs Fig. 5.16 Motor and generator actionELEMENTS OF GENERALIZED THEORY 115 current flow for the pseudo-stationary coil. Motor or generator action is determined by assigning to the coils magnetic polarities which accord with the directions of current flow. Where the mag- is motor action and when they oppose netic forces assist motion ther motion there is generator action. Referring to Fig. 5.16: i. i =i T,=T motion assisted, motor action i= -t, & =. T, = —T motionopposed, generatoraction. 2 Torque and voltage equations As the previously discussed rotary transducer and the commutator machine each have two windings the [L] matrix for each device will be of similar form. No such statement can be made for the [G] matrices and it is necessary that a [G] matrix be determined for the commutator machine before writing equations for torque and voltage. Commencing with the generai siator and rotor voltage equations (5.54) and (5.55), all the rotational voltages in (5.54) may be deleted a (a) the stator conductors and the pseudo-stationary rotor coil flux are space-stationary, and (b) the stator self-inductance L, is a constant value for a cylindrical rotor. v, = i,R,+ L,pi,+ M,,pi, (5.69) ; : 5 . dh, . dM,, i,R, + L,pi,+ M, pi,+o,i, at ae (5.70) The incremental energy supplied to the two windings in time dr dW, = (v,i,+ 0,1) dt (5.71) AW, = (2R, +7 R,) dt+[Lp(aig) + L,plaip) + M,plisi,)| dt aA Seek incremental incremental increase energy loss in in field storage energy winding resistance dW,, rola aM | 2 iz) di (5.72) incremental energy for torque production dW,16 GENERALIZED EEPCTRICAL MACIHNE THEORY Electromagnetic torque (pole-pairs) dW, 7, = (ole oO, dr ; IMy , 2 dL, = (pole-pairs (pole pairs ie Gg + gg) : or. af OO Wi . T, = (pole-pairs)[i,i,] dM, dt, [ (5.73) do, dO, T, = (pole-pairs)[i,][G][‘] (5.74) Comparing (5.74) with (5.67) shows the general expressions for electromagnetic torque differ by a factor 3. Thus, for a commutator machine with stator saliency m[ ] 0 0 (G]=|4dM, dL, d0, do, Voltage equations ry R, 0 L, = a (T-lo elsL, For a general brush displacement 0, with respect to the d-axis (5.75) L, = constant dL, L, = L,,+L,, cos 20, =~ = —2L,, sin 20, 0, dM, M,, = M,, cos 0, o“ = —M, sin 0, dO, Substituting the L and G coefficients in (5.73) and (5.75) ( 1, = (pole-pairs)fi, il] ) ° [i] (5.76) ty —M,, sin 0, —2L,) sin 20, feELEMENTS OF GENERALIZED THEORY u7 nfo of, f ry} (LO R,| 7 0 0 i 5.11) FO) vp sind, —2L,; sin 20, ]fLi] | With brushes on the q-axis, 0, = — 1/2 L M,, cos U, Ip M,, cos (1, L, T.. = (pole-pairs) i,i,M@,, | v, = i,R,+ Lypig = 1,R,+ Lpi, +o,M nig For steady-state d.c. conditions p=0,0,=V, and r,=1’, Motor action i,=/,. i,= 1, Generator action i,=J,,i,= —I, T, = +(pole-pairs) /1,M,, Vi = TpRy VYq= £1,R,+O,Mmls On no-load the terminal voltage, called the excitation voltage, is given by E = 0Myl; where M,, is the value of mutual inductance between the stator coil and the pseudo-stationary rotor coil with the brush axis at 0,=0. The equivalent circuits for steady-state d.c. motor and generator action are shown in Fig. 5.17. R, : R VF Wwe Vv. rotor rotor ; coil n coil oH OH 4 = 4, <= OM gly OMpl, Motor Generator Fig. 5.17 Equivalent steady-state circuits for a d.c. commutator machine 9118 GENERALIZED ELECTRICAL MACHINE THEORY 5.5 The primitive multi-coil machine It will be shown in subsequent chapters that the windings of a rotat- ing electrical machine and their associated electrical quantities can be transformed mathematically into a different arrangement of coils with new electrical quantities. The machine which results after transformation has performance characteristics identical to those for the original machine. A transformation of special interest is d-axis ive d-q axis machine one in which machine windings are transformed into a d-axis and q-axis array of stator and pseudo-stationary rotor coils. As this arrangement is the two-winding commutator machine repeated a number of times, the analytical results previously derived for this machine can be used to determine the performance characteristics of the transformed machine, hence the original machine The name primitive machine is given to the d-q axis array of coils for which connections between coils are absent and each coil is supplied with a separate externally applied voltage. This machine is commonly referred to as the Kron primitir A primitive machine with two pseudo-stationary rotor coils and two stator coils is shown in Fig. 5.18. The number of coils for each axis could be increased without limit but as an excessive number of coils would make subse- quent analysis laborious without adding to the understanding of the analytical technique employed, the number has been limited to that shown in Fig. 5.18. Each coil is identified by a sutlix notation givingELEMENES OF GENERALIZED THEORY Hg supporting magnetic structure, tor d-axis coil and 5, a rotor med to have the axial location of the coil and the stator or rotor, For example, } isa y-axis coil. The two pseudo-stationary coils are ass the same number of concentrated turns, Nj= Nj. Voltage coefficients: matrix form It is now proposed to construct {£] and [G] matrices for the d-g axis primitive machine Fig. 5.18, with stator saliency. The d-q axis coil array comprises a number of individual coil pairs. Relative to a d-axis coil, pairs are formed with coils displaced by angles of 0, =0, 0,= — 1/2and relative are formed with coils displaced by angles of Thus. for all coil pairs the displacement angles of concern are 0,=0 and 0,= +7/2. The rules for constructing [L] and [G] matrices for the primitive machine can be determined from th tor and rotor voltage equa- tions (5.77) for a two-sinding commutator machine. From (5.77) ie M,, cos 0, [4] = [M,,cos0, —L, The mutual inductance coefficient (5.78) Therefore, mutual inductance coefficients are present when a pair of coils are situated on a common magnetic axis. From (5.77) : 0 0 [¢] = [_ M,, sin 0, —2L,, sin »| For 0,=O and +7'2. sin 20,=0 and the only G coefficient of con- cern is G,, = —M,, sin 0, (5.79) 0,=0 G,. = 0,5 25 Gy = FMyLo GENERALIZED ELECTRICAL MACHINE FHEORY A G coeliicient is presem when a pair of coils are located in space quadrature and one of the coils is pseudo-stationary. Before proceeding further a suffix notation is required to identity the two coils associated with a particular 2 and G coefficient. Consider as a particular example Gig Ist column 4, the seat of induced voltage 2nd column § the source of flux inducing the voltage. For any coil pair the axis of the flux-producing coil is taken as reference. For the example considered, the axis of j is located + 7/2 from the axis of }. (a) The inductance coefficient matrix [L] Seat —— source of ofvoltage L | | Gf GO flux (5.80) (4] The matrix contains mutual inductance coefficients forming equal pairs Mi = Mis My, = Mi For a salient-pole machine LAL LEAL, and My = Mx as the d-axis and g-axis permeances are different. The correspond- ing inductances are equal for a machine with zero saliency.ELEMENTS OF GENERALIZED THEORY 42 (b) The rotational coefficient matriy |G | The [G] matrix in general form Seat source of of voltage | 5 oe flux a 8 ye I on | a 1 » | rr | ia Gh | \ ! [qe] = 4 Ga i | A 0 \ L { | . 0 | | ‘ Itisseen that G coefficients are only obtained for pseudo-stationary rotor coils. Coils acting as flux sources may be stator or pseudo- stationary rotor coils. The significance of the G coefficients is determined from Gy, = — My sin 0, The value of M,, is the coupling obtained when the axis of the seat of voltage coil is imagined to be aligned with the axis of the source of flux coil. Coefficients with stator coils as the source of flux (4k Gig = ~ Mg, sin (+3) =-ME 2) ars Soe % 1s Gy = ~Mig sin ( -5) = Mai Coefficients involving two pseudo-stationary coils oer mg z a Gi, = —M. sin( +5 = -My, 44 ) = Mi x On we ~ Mii sin ( = w Mi = VUE = LG a4 wer a = JUGl = 4fe2) GENERALIZED ELECTRICAL MACHINE THEORY Substituting the above inductance coeflicients in (5.81) a a q 4 13/0 oO| 0 0 [Fe aaleeee eee, i=) 510 0 io os (5.82) 7 0 oe fy | Mi 1 | 0 0 i a Combining the results obtained in (5.80) and (5.82). substituting in the general voltage equation [e] = (R]}+[2]et+o[6]}[] gives the complete impedance matrix for a four-coil. primitive machine, Fig. 5.18. % (Rit Lip) Miip 0 0 4 {=| Map (Ri+Lip) -oMg -0,Ly |i 2 0 0 (Ri+Lip) MEp |Lis | | Mas oy Mip — (Ri+Lip) |e (5.83) 5.5.2 Torque and power Substituting (5.82) in (5.74) 0 0 0 0 lfe T, = (pole-pairs)[iiisi7]] 0 0 —ME -Liiie} S84) oe ole ee o fr Expanding and collecting terms, (3.85) 1, = (pole-pairs)[ Miyiyi, — MIE + (Li = LivELEMENTS OF GENERALIZED THEORY 133 The torque component in (5 ~ 11) is only present in salient-pole machines and is called saliency-torgue. The remaining tarque com- ponents, which define the clectromagnetic torque for a uniform air- gap, are described as cylindrical torques. It will be noted that the cylindrical torque components comprise a positive motoring term and a negative generating term: the sum of all such terms determines the machine action, motor, or generator. The total internal electrical power developed by the machine ” Po= . 5 *™ (pole-pairs) oy) oo, Mais — Misi8i5 + (Lh — Li] (5.87) An alternative method of deriving (5.85) is to consider the torque exerted on each pseudo-stationary coil by all coils at right-angles. PROBLEMS 5.1 A rotating transducer has a salient-pole stator and a rotating coil connected to slip-rings as shown in Fig. Prob. 5.1. The mutual coupling varies as 41, cos @ where 0 is the angle shown. dd xe dt d-axis o, Fig, Prob. 5. Derive from first principles expressions for the instantancous torque and the rotor and stator voltges. How is the torque expression modified when the rotating winding is connected to a commutitor and current enters through a fixed brush- pair inclined at an angle 7 10 the salient-pote axis? Discuss the physical concepts of torque production for the two types. of rotor connection Explain how a rotitting commutator winding may be represented as an equivalent pscudo-stationary coil tlong the brush axis, and describe the special properties of this type of coilIM GUNTRATIZED PLECERICAL MACHINE THEORY Write the general voltage and torque equations for the two-pole commutator machine shown in Fig. Prob. 5.2. Explain how the rotational inductance coellicients (G) may be obtained in terms of the Land Mf inductance coeflicients 1g-AXis i Fig. Prob. 5.2 5.3. A four-pole commutator winding has 576 uniformly distributed con- ductors. How many turns has the equivalent concentrated coil if the winding is (a) lap-wound, and 1 (b) wave-wound? [Ans. 45, 90] 5.4 A two-pole commutator machine has a sinusoidal space-distributed stator flux with a maximum density of 1-5 Wb;m?. The effective length and radius of the rotor are 06 and 0:25 metres respectively (a) Calculate from first principles the flux-linkage of a full-pitch rotor coil with its axis inclined at an angle to the stator pole axis. (b) Hence calculate the maximum flux-linkage (per parallel path) of the complete rotor winding if there are 200 conductors. [Ans. (a) 06 cos x Wb-turns; (b) 19-2 Wb-turns] i6. D.C. and crossfield machines The physical arrangement of a d.c. or crossfield machine is identical to the primitive machine. The rotating commutator winding has one or more brush-pairs on orthogonal axes and stator coils are on the same orthogonal axes. The equations developed in chapter 5 for a primitive machine are directly applicable for these types of machines. 6.1 General analysis Machine voltage and torque equations may be written down by inspection when the ideas and rules formulated in chapter 5 are understood. For convenience, a two-pole machine is always schematically shown although the equations apply for any number of pole-pairs. graxis | = . { 7 (ny) os Fig. 6.1 Primi machine Consider a machine with one brush-pair on the quadrature-axis and two direct-axis stator coils as shown in Fig. 6.1. The voltage and electromagnetic torque equations are: ry (Roepe \iaep 0 fiz! eP l=] MP S'p | (RE +LEp) 0 Ge (6.1) vy oGy y OG OR +L pie rl 7, = (pole-pairs)| (G7 +Gi'y iP)] (6.2)2 GENERALIZED ELECTRICAL MACHINE THEORY The rotational inductance (G) coetlicients may be replaced by transformer inductance (M) coefficients as described in Section 5.5, ry! (RY +L'p) ME Sp 0 | is! Ppa] MPa (RP+LPp) 0 ie (6.3) | oMy st oMi ? eee Lie (pole-pairs)[i7'(.4j" 3 Mi ¢ iP)) (6.4) Assuming a mechanical damping coefficient D proportional to speed, the torque balance equation is. Tt To = Iptig + DOp (6.5) where, as previously, suffix 7 refers to a mechanical quantity and J is the. polar moment of inertia. The general equations (6.3) to (6.5) apply for steady-state or transient performance. For steady-state operation the derivative terms are zero and the equations are considerably simplified. For simultaneous changes of current and speed the solution of the equations may be extremely formidable, see Section 6.4.3. Simplified notation 1n the previous chapter each primitive coil is identified by two letters and a number. With a particular machine arrangement it is more convenient to use a single letter to identify a winding, especially with d.c. and crossfield machines in which primitive coils correspond to actual windings. For example, a coil designated /has resistance r,, self-inductance , and its mutual inductance with a coil s is M,,. 6.2 D.C. machine steady-state analysis D.C. machines have one or more stator windings for the required load characteristic. Additional stator windings are included to give satisfactory commutation. 6.2.1 Field windings The main stator or field windings are wound on salient poles. Coils may be connected: (a) in series with the rotor or armature winding. ie. a series field or (b) energized from a separate d.c. supply, z..a separately-excited field or (¢) across the armature brush-pair, ie. a shunt field.D.C. AND CROSSEIFED MACHINES. 17 A dc. motor with a shunt winding is identical to a separately- excited machine as in both cases the winding is energized from the supply voltage. With a shunt generator, the simplifying assumption of a linear flux/current relationship is not justified as the open-circuit vollage is limited only by magnetic saturation. The analysis of this machine therefore requires a more realistic treatment, see Section 6.2.4. A compound machine has a series field together with a shunt or separately-excited field winding. The following analysis is for a compound machine with series and separately-excited windings. 6.2.2 Compound machine analysis Figure 6.2(a) shows the primi- tive coil arrangement for acompound machine. For simplicity, the armature and the series and separately-excited fields are designated a. s, and f, respectively. The primitive voitage equations are, ey? (Ry+L yp) My sp 0 fee ry} = Mp (R,+ Lyp) 0 i (6.6) cA oMas OM (Rat Lab SLi The field coils may be connected to give m.m.f.s in the same sense or in opposition to each other, and are termed either cumulative connected or differential connected windings, respectively. The two alternative connections for a machine in the motoring mode are shown in Figs. 6.2(b) and 6.2(c). Reversal of the armature current also reverses the series field, and a differential connected motor operates as a cumulative connected generator. Similarly, a cumulative connected motor operates as a differential connected generator. Establishing relationships between the primitive and actual machine currents and voltages, rh 4a, = lq a = ti, a = iy In all cases the current i, is shown entering coil a at the dot. For motor action, i, is a positive quantity and for generator action is a negative quantily. The armature circuit voltage 1’ Pov 418 GENERALIZED ELECERICAL MACHINE THEORY Substituting in (6.6) for the primitive voltages and currents, Ly t by combining rf! and rj! and writing R=R,+R,. L [r, (RR, +L ;p) + Mop qtr, (6.7) = 5. LVL Loa Mop) (R4 Lp =O,Myg) NL i The electromagnetic torque equation is, T,. = (pole-pairs)[ 7)? Myyp+ 7, ij! Mac] Substituting-for actual currents, T, = (pole-pairs)[i, Myris +E Mua] (6.8) g-txis qraxis | NO, i we Jf (D} (Oye {a) Primitive machine (b} Cumulative connected i, motor {0) d-axis ‘y (c) Differential connected motor Fig. 6.2. Compound machine The alternative sign + is positive for a cumulative connected motor and a differential connected generator. and negative for a differential connected motor and a cumulative connected generator 6.2.3 Steady-state analysis The steady-state operation of a com- pound machine is determined from the general performance equa-D.C AND CROSSHIELD MACHINES, 120 tions. By omitting the appropriate winding from the results for the compound machine, steady-state solutions for series and separately- excited machines are readily obtained (a) Compound machine With constant speed and current values the pe, and pi terms are zero, and (6.7). (6.8). and (6.5) may be written, V = 0Mypl yp +(REOMa Me (6.9) EM] (6.10) (6.11) T, = (pole-pairs)[1,Myyl Tt T, = Dov, Typical characteristics are shown for a compound machine in Fig. 6.3. A level-compounded generator has no voltage regulation. + 4 e, Constant voltage Constant speed Generating 4 Fig. 6.3 Compound machine characteristics, cumulative motor connection In practice, saturation of the magnetic circuit increases the voltage regulation at high load currents and the term level-compounded describes a generator with zero voltage regulation at full load (b) Series machine Omitting the separately-excited winding gives the steady-state equations for a series motor Vs (R+oO,M J, (6.12) = (pole-pairs)(/2M,,) (6.13) The characteristics for a series motor, Fig. 6.4. showa high starting torque and the danger of running the machine on no-load. The series generator is seldom employed130 GENERALIZED ELECTRICAL MACHINE THEORY Torque, speed ee Fig. 6.4 Series motor characteristics (c) Separately-excited machine or shunt motor The steady-state equations are: V = RU, +0,Myyly (6.14) T, = (pole-pairs)(I,M,,1,) (6.15) Typical characteristics are shown in Fig. 6.5. + oO, constant speed constant voltage Generating Motoring Fig. 6.5 Shunt motor and separately-excited machine characteristics 6.2.4 Shunt generator The assumptions made in the previous work may give rise to appreciable errors in analysing the perform- ance of d.c. machines, However, the ability to derive and solve general equations for both steady-state and transient: behaviour often justifies these assumptions. With these simplitying assump-D.C) AND CROSSEIELD MACHINES 131 tions the performance of the shunt generator camnot be determined The analysis of this machine requires a more realistic approach employing the flux/excitation m.m.f. characteristic. (a) Actual characteristics The field current for a shunt generator upplied by the armature voltage, sce Fig. 6.6. With the armature rotating the residual flux in the stator poles generates a voltage across the brush-pair which drives current through the field winding in- creasing the pole flux, and hence, the armature voltage. The voltage i li, tiy) Fig. 6.6 Shunt generator continues to rise in a fairly linear manner with field current and the machine is said to se/f-excite. When the magnetic circuit of the pole flux begins to saturate, the output voltage does not increase as rapidly as previously, see Fig. 6.7. A point is reached at which the voltage characteristic cuts the field resistance line and this corre- sponds to the open-circuit voltage of the machine. Two necessary conditions for self-excitation are, (a) the residual flux is in the correct sense, i.¢.. field current due to rotational voltage reinforces the residual flux of the stator poles; and (b) the slope of the field resistance line is sufficiently low to intersect the voltage characteristic at a reasonable value above the knee of the curve. Fora given speed the critical field resistance is the maximum value of field circuit resistance at which self-excitation occurs, With a given value of field circuit resistance there is a critical speed belowI GENERALIZED ELECTRICAL MACHINE THEORY which the machine fails to excite. “The sudden short-circuit of an energized shunt generator reduces the excitation to zero and the steady-state value of short-circuit current, duc only to residual pole flux, ts small. oad outpul voltage Voltage due to residual flux | i Fig. 6.7 Voltage characteristic for a shunt generator, including saturation The solution of shunt generator problems requires graphical methods unless the magnetization characteristic can be expressed as an analytic function of the field current. (b) Method of analy. The non-linear voltage characteristic of a shunt generator, Fig. 6.7. may be represented with reasonable accuracy by two linear parts of different slope as Fig. 6.8. Rotational voltage Field current /, Fig. 6.8. Approximate representation of saturation for a shunt generatorDiC AND CROSSEIELD MACHINES 133 With incremental inductance coefficients of Gm, Mand Gt. Mr, for the two slopes, the rotational voltage term in the armature coil is given by in Dip > 0, rotational voltage term = 07,G";i, = OM, tp > tps rotational voltage term = 0,[G" ip, + G2) —ip1)] = Of Minn + Maisie] The voltage and torque equations in the operating region above the knee of the curve, part n, are v = (Ri + Lap)ig + OM a pip (6.16) ty = = (Rp +Lyp)i; (6.17) (pole-pairs)(.Mariaip) (6.18) As a function of i;, it, Mey) iy Eliminating i, from (6.16) and (6.17). Mas = Mast (Ret Lap )iat Oni py My a 6. OAL (6.19) For steady-state operation, py = Relat rip Mery Mis) (620) Equations (6.19) and (6.20) are valid for operation in region n only, corresponding to a range of field current i, >i,,. 6.2.5 Compensating winding A dic. machine with a severely fluctuating load, e.g., a rolling mill motor. may flashover across the commutator segments due to the high volt; induced in the armature coils by the rapidly changing armature current, A flash- over often results in severe damage as short-circuit current may flow 104 GENERALIZED FLECTRICAL MACHINE THEORY along the breakdown path. The large voltages induced in the armature coils may be climinated by including a compensating wind- ing in the machine, The voltage induced between two adjacent commutator segments. depends on the rate of change of flux-linkage with the armature coil connected to the segments. There is an induced voltage (nearly independent of load current) due to the main excitation flux (an ,M term). An additional induced voltage may appear due to a rapid change of armature current and, hence, self-linkage (an Lpi term). qvaxis compensating 7 compensating L winding winding 77 1 a\t oe \1 ie 1 | 1 1 1 (a) Physical arrangement (b) Primitive coil diagram Fig. 6.9 Compensating winding The compensating winding. connected in series with the armature, produces an m.m.f. equal and opposite to that of the armature. see Fig. 6.9(a). A rapid change of load current produces negligible change of flux along the q-axis. inducing little voltage in the armature coils. Representing a separately-excited machine with a compensating winding as a primitive machine, Fig. 6.9(b), the coil voltages are ve! (Ry +Lyp) 0 0 B! ays 0 (RetLp) Map fet} (6.20 | vy oMyy Mp (R,~ Lapse! Connecting the armature and compensating windings. armature current i, =f = -2'. armature circuit voltageD.C AND CROSSEIFLD MACHINES. 13s equation (6.21) may be written ry | _ [Rpt hyp) 0 Tir fo tf edgy (Rpt R) + (Ly + Le 2M,p ffi, | ©?) Equations (6.22) are identical to those for a separately-excited machine with an armature resistance and inductance of (R, + R,and (L,+L,—2M,,), tespectively. With equal turns on the armature and compensating windings, L and the effective armature inductance (L, + L,—2M,,) is negligible. With the compensating winding not extending to the interpolar regions, as shown in Fig. 6.9(a), i.e., restricted to the pole faces, the compensating winding inductance is less than the armature self- inductance and the effective armature inductance (L,+L,—2M,,) may not be negligible, M, 6.2.6 Interpole windings Satisfactory commutator action requires the current in a coil undergoing commutation to reverse in the com- mutation period, see Fig. 6.10. An induced voltage, produced by Brush Instant ¢, Commutator “V7 ore Tl bp tole | in coil Cstats JC T Commutation |e to decrease Le period i Ideal commutation i Incomplete Instant 4 RS Oa rent reversal shown for +o current in coil C Fig. 6.10 Commutation period changing self-flux in the coil, opposes the current change and may prevent complete current reversal in the time allowed for commuta- tion. The difficulty is overcome in al} modern machines by fitting interpoles to give practically spark-free commutation at all loads and speeds. Interpoles, located as their name suggests in the interpolar16 GENERALIZED ELECTRICAL MACHINE THEORY region, see Fig. 6.1 (a), are connected in series with the armature winding and produce fluxes which induce voltages in the coils under- going commutation. These voltages are approximately equal and opposite to the self-induced voltages and enable the current to reverse completely in the commutation period. Both self-induced and inter- pole induced voltages are proportional to speed and armature current, hence correct compensation is achieved for all values of speed and load, including rotation in either direction. The interpole acts as a compensating winding over a small angle and is sometimes termed a commutating pole or compole. interpole \ 1 D, ea | le @}s—— 7 ! ' , ' | (a) Physical arrangement (b) Primitive coil diagram Fig. 6.11 Interpole windings The analysis is identical for machines with interpoles or com- pensating windings. However, the inductance of an interpole winding, L;, is small compared with the armature inductance, L,, and the modified armature circuit inductance (L, + L;—2M,,) cannot be ignored. Interpoles are necessary for a machine with a compensating wind- ing restricted to the salient-pole faces but are not required for a com- pensating winding which extends over the interpolar regions. 6.2.7 Starting With full voltage applied to a stationary d.c. machine, the armature current, limited only by the low resistance of the armature circuit. would be dangerously la Consequently. d.c. motors are always started at reduced armature voltage. A common method is additional resistance in series with the armature which can be gradually removed as the speed increases. Full field current is applied to shunt machines to provide maximum rotational