Induced Voltage of Electrical Machines

BY L. V. BEWLEY1
Associate, A. I. E. E.

Synopsis.-The object of this paper is to describe and discuss a general equation for the induced voltage of electrical machines having parallel coil sides, and which includes as special cases single and polyphase induction motors, synchronous generators, d-c. generators, synchronous converters, and static transformers. The application of the general equation to most of these cases is illustrated, and a number of interesting problems which may be solved by means of it is pointed out. A characteristic of this equation is that no restrictions are placed on the velocity of the moving con-

ductors, or on the rates of pulsation and rotation of the flux; but these may vary in any arbitrary manner which can be given a suitable analytic expression. The several methods for the reduction of harmonic voltages are classified and their limitations discussed in such a way as to leave in mind a vivid picture of the process. Tables and curves have been prepared for comparing the effects of the skew, pitch, distribution, and phase connection harmonic reduction factors. A new method for summing the finite series of the distribution summations is given in Appendix II.

I. INTRODUCTION T HE purpose of this paper is to derive and discuss a general equation for the induced voltage of electrical machines having uniformly distributed parallel coil sides moving through a distribution of flux which can be represented by a Fourier series. The equation is general in that the circuit considered may have any number of phases connected in series in any arbitrary manner, and any particular phase may consist of any number of uniformly distributed, fractional pitch, skewed coil sides moving at variable speed through a distribution of flux which may pulsate and rotate at different rates. This equation includes as special cases the procesess of induction found in the more familiar types of electric power apparatus, such as synchronous motors and generators, indu tion motors, d-c. motors and generators, synchronous converters, and static transformers. Bescause of its generality, this equation may prove awkward and unwieldy in dealing with a particular type of machine, since it was not derived with the idea in mind of adapting it to some special condition of symmetry affecting the choice of a coordinate system, or to the necessity of correlating with the related phenomena of currents, m. m. fs., and fluxes. There is an essential distinction between developing an equation for the complete analysis and determination of characteristics of a particular type of machine on the one hand and deriving an equation whose purpose is to study the variation in process of some specific phenomena which occurs in several types of apparatus, on the other. In the latter case, greater flexibility is required, and the expression should be purged of all possible detail irrelevant to an understanding and classification of the ideas under consideration. For instance, certain facts become m asked and obscured by temsi array of symbols associated with the complete harmonic analysis of a synchronous machine, involving the specification and manipulation of the Fourier series of
Pittsfield, Mass.'thntemsgeea Presented at the Winter Convention of the A. I. E. E., New York,
1. General Transformer Engg. Dept. General Electric Co.,

currents, m. m. fs., permeances, fluxes, and voltages. Incidentally, a study of the general equation derived in this paper permits making a critical review and classification of the available means for suppressing harmonics in the induced voltage by special arrangements of the windings. These special arrangements give rise to certain functions known as harmonic reduction factors, which enter the equation as ordinary coefficients. They are called, corresponding to the particular arrangement of the winding to which they apply, the skew, pitch, belt distribution, and phase connection harmonic reduction factors respectively. These same factors have an almost exactly analogous effect in suppressing the harmonics of armature reaction, but their consideration in that respect is outside the scope of the present paper. The full limitations of these factors are not generally appreciated. For instance, it is usually assumed that it is impossible to have a third harmonic in the line voltage of a threephase synchronous generator if it is connected Y, or has coils of 23 pitch, or 120 deg. phase belts, or conductors skewed 120 deg. with respect to the pole; because each of these conditions is supposed to completely eliminate the third harmonic and its multiples. Yet such a harmonic will appear in the line voltage of a machine having all of these conditions, if there happens to be an even time harmonic in the air-gap flux. On the other hand, a third harmonic due to such a cause can be eliminated by a suitable arrangement of the winding. Now this is not a new discovery, but only the specialist would notice the fact in a treatise on the harmonic analysis of synchronous machines. II. GENERAL DISCUSSION If it is possible to express the distribution of flux in an electrical machine having parallel coil sides by the c oresre ( x ^,S sin k~t T + z J B =~ (1)

N. Y., Jan. 27-31, 1930.

qaonfrhendc le voltage, regarding x0, A*k, and 'yk as functions of time, is shown in Appendix I to be

hntems

eea

qainfrteidcdln

456
30-11

April 1930

BEWLEY: INDUCED VOLTAGE OF ELECTRICAL MACHINES

457

Eli, =- 2 L T, 108
r1

n cN'
1
T

Cck Cdkl Cpk Csk

Lk

dkXnk k ( d t sin
d xow dt

7r +
r xo X

Ic

+ 7k +

lk', + k

If the conductor is skewed or spiralled so that half of it is cutting through a positive loop, and the other half through a negative loop of the harmonic of space distribution, then the voltage induced in the two halves of the conductor are always in direct opposition and therefore completely cancel within the conductor it-

k 7d + +td t

+ k+k+

where the C coefficients are the harmonic reduction factors depending on the arrangement of the windings, defined in Table I.
TABLE I

(17)

S&OW

.0

Reduction factor
Skew

Seat of the reduction the coil side

Lo _---a

factor...kCi

sin k X/2

/2In

Pitch factor Distribution factor of belt....

Cpk=
Cdk
C

sin sin k c d/2

.

__

Between coil ] sides of the coil

Between coils of
h,

CO/

CO

* / sinl ob]t:Cdk=C k 6/2

(

.|the phase belt

+

*s,

Connection
r

ac

sin k

Or)2
n

(z

cos

Or)2

Between

phases

Con

If the n phases of the connection factor are uniformly displaced by the angle then
sin kn/2 n sin k c/ which is then in the same form as Cdk, and the same set

FIG. 1-SEAT OF HARMONIC REDUCTION

of curves will serve for both. Moreover,

Cdk

sin k 6c/2
k ac/2

sink n

if ->OandC0, =

k n-/2

if

+0.

4aA

Thus under these limiting conditions, the three reduction factors C,k, Cdk, and C,k have the same form. Associated with Cdk, and Cck are the distortion angles k, ' and l,k defined in Appendix I, and these cannot, in general, be made to disappear by a convenient choice of reference axes. In other words, a lack of symmetry and uniformity in the windings, or in the way that they are connected in series, will cause a distortion of the voltage wave by an angular shift in the relative positions of the harmonics, as well, as by the reduction in amplitudes characteristic of symmetrical arrangements
While these reducing factors have different names, they are in fact due to the same essential cause-the arrangement of the circuit so that the harmonic voltages produced in different parts are in partial or complete opposition and thus tend to cancel out over the complete circuit. Fig. 1 illustrates how this is accomplished in the case of a third harmonic.
of the windings.

03u

FIG. 2-THE SKEW COEFFICIENT

self. The next simplest method of cancellation is to use a coil of fractional pitch so that both coil sides are cutting through positive ioops of flux, but the voltages generated thereby cancel in the coil. In this case the coil may be either of short or long pitch. The distribution of the winding in more than one slot per phase

458

BEWLEY: INDUCED VOLTAGE OF ELECTRICAL MACHINES

Transactions A. I. E. E.

As a rule it is desirable to suppress as many harmonics gives rise to a reduction of the harmonic voltages beas possible at the least cost to the fundamental. tween the coils which make up the phase group. Finally, it is possible to eliminate those harmonics For all harmonics are more or less objectionable, which are multiples of the number of machine phases, by connecting phases in series. Representative curves of the skew, pitch, and distri'11 bution factors are shown in Figs. 2, 3, and 4 respectively, 1I9 9 and the phase connection factor for a few simple cases OA t

FIG.

3-THE PITCH COEFFICIENT

monics has been plotted to indicate the general nature of the functions. It will be noticed that the maximum value which any of these coefficients may have is unity, or more appropriately, their values lie within the range (- 1 < C ' 1), because they have been defined as geometrical averages.

iS given in Table II. Only a sufficient number of har-

FIG. 4THE DISTRIBUTION COE:FFICIENT

but when they are suppressed in such a way as to reduce materially the amplitude of the fundamental it means more flux in the air gap, more excitation, and possibly higher losses, or even a larger sized

THE PHIASE CONNECTION COEFFICIENT AND ANGLE

Connection|
3 fDelta

TABLE II 7 1

k^=
Cek

1 1 0

3 1

5 1

9 1 0

11 1 0

13 1 0

115
1

117
1 0

19 1 0

M~

{k

10

10

,9 IkM

+300

2/
.-JT

Cck
Ck
Cck

t
I+45
1

+30

1-

30

+300
1
450
1

-30

1
+450
|

1
450

1
+45 |
|

1
+450
1

1
|
|

l
450
1

+30

30

1
+450
|

1
|
|

1
45
1

450
1

X ~2 3 fZigzag |Cck |3
6'E A
k

6 fStar

2
0

2
0

2
4
-60'

60

4
0

4
-60' 0

4
60

+600

-60

11
0

60'

April 1930

BEWLEY: INDUCED VOLTAGE OF ELECTRICAL MACHINES

459

machine. It must be understood that other considerations than the relative efficiency have a decided bearing on the choice of a method of harmonic reduction. Thus if either skewing or fractional pitch is an available method in a given case, the fractional pitch would undoubtedly be the choice; because it is less expensive to make straight slots and coils, there is considerable saving in copper and copper losses on account of shorter end windings, and for the same reason a possibly shorter machine with less windage loss. While3k, yk, and xo in Equation (17) may be any arbitrary functions of time, yet the electromagnetic and mechanical characteristics of electrical apparatus are such that each of these items may, for practical purposes, be taken as composed of a series of terms of the type: Ok =O130k + l3k fa + 132k sin w t +33k Ebt sin co t (22) d 'Yk dtd t = Vk + Vlk 6"' + V2k Sin w' t + V3k fb ' sin ' t (23)

constant velocity; then only those time harmonics can appear in the induced voltage which are present in the space distribution of the flux, and which are not wiped out by the harmonic reduction factors. In this case, if any reduction factor for the kth harmonic be made equal to zero, then the kth time harmonic will be canceled out of the induced voltage. But every harmonic, including the fundamental, will suffer a reduction in magnitude if the slots are skewed or if the coils are made either short or long pitch, or if the total turns are distributed, or if phases are connected in series-in brief, any departure whatsoever from a straight-sided, full-pitch, concentrated winding will cause a reduction of all harmonics. In particular if the fundamental is canceled by any reduction factor, so will every harmonic be canceled. (b) The presence of exponentials will cause either damped or cumulative oscillations, depending on the sign of the exponents, to appear in the line voltage. The abrupt change of the excitation on a machine, either by rheostat adjustment or by an automatic d xo voltage regulator, causes the flux to build-up (or decay) Ifat + V2 sin w" t + V3 eb't sin co" t dl= Vo + V, d t exponentially. This change in 0 and its derivative is (24) reflected in the equation for the induced voltage, as Thus immediately after an unsymmetrical short cir- exponential-trigonometric products, that is, as damped cuit on a synchronous motor the flux in the air-gap or cumulative oscillations. An exponential variation of 3k may also be caused by a symmetric short circuit, or abrupt change of load occurring on a generator. If the torque of the prime mover driving a generator, or the mechanical load on a motor changes, an adjustment ____\___,__ in speed takes place. This speed transient will contain / r _-\ 7' an exponential term due to the mechanical inertia of the rotating parts. In the case of power suddenly off from an open-circuited generator, the speed t 5411T rf decrement is practically a pure exponential (not rigort-Xt yzrously so, because the friction is not constant throughout lthe speed range). These exponentials in speed cause < ' /| a :-/-2.. the amplitudes of the harmonics in the induced voltage *-tJtO SIDr to change proportionally, and the wavelengths of these

t-

t7-

{TH2shut

harmonics either
exponential law.
nature.

contract

or

expand according

to

an

consists of terms such as given in (22). If the short circuit is severe enough to cause hunting, the constant velocity of the armature will have superimposed thereon an oscillation which may, if the conditions are right, continue indefinitely, die out exponentially, or even increase in amplitude until the machine falls out of step. The oscillatory passage of the teeth and sls across the flux. Thus pole faces will cause the distort . Equations (22), (23), and (24), represent >tetual possibilities. The direct substitution of (22), (23), and (24) in (17) yields a representative equation for the induced voltage in electrical machinery, but for the purpose of this discussion it is not necessary to carry out the actual substitution. It is sufficient to notice the following facts: (a) When the flux distribution contains only rigid space harmonics, and when the coils are moving at a

(c) Perhaps the most important class of time variations in flux and speed are those of a purely oscillatory

flux are caused by unbalanced short circuits and unbalanced loads on polyphase machines, the time harmonics of polyphase armature reaction; variation in the air-gap reluctance due to the passage of teeth and slots, and the elliptial rotating fields of Single-phase motors. Any possible product of sine and cosine terms may be expressed as the sum or difference of sines or of cosines, for example 2 sin w t sin k w0 t = cos (Ic wo -w) t-cos (kc wo + w) t It is of importance to observe from relationships of this type that it is quite possible to produce a harmonic of any order K (integer or fractional) in the induced voltage, even though the reduction factors have been

sustained

Periodic fluctuations in

460

BEWLEY: INDTJCED VOLTAGE OF ELECTRICAL MACHINES

Transactions A. I. E. E.

adjusted for its complete cancellation. These possibili- and taking the armature as the rotating element, there ties are given by is: = (k wo0 w) woK t wheref0 = frequency. For instance, suppose that the reduction factors will Xo = 2 fo r t = coo wipe out the 5th time harmonic due to a 5th space d /kld t = 0 since the flux distribution is steady. harmonic of flux distribution. But as a possibility 5 = (k + wlwo) d-Y kld t = 0 since the flux distribution is fixed. w/ from which it1S evident that the 5th time harmonic Substituting these values in Equation (17) there is is from whlch lt that the ,x, harmonic , n N' may nevertheless appear in the line voltage if the air CckCdk Cpk Csk O gap flux contains a space fundamental, k = 1, pul- Eline = Lfo- 108 sating at wlwo = 4 times normal frequency, or a 3rd 1 cos k (co t + 7Y ;k' + Ak) space harmonic pulsating at w/wo = 2 times normal + frequency; or in general a kth harmonic pulsating at Now in commercial machines the field pole flux is w/wo = (k - 5) times normal frequency. More gener- rigid and symmetrical. The space harmonics are thereally, since the periodicity of the flux pulsation may be fore of odd order and fixed in a symmetrical position. arbitrary, it is possible to produce frequencies in the in- Consequently for a machine on open circuit (or supplyduced voltage which are not multiples of each other. ing zero power factor current), k is odd, and either The only common sinusoidal fluctuations in speed 7k = 71 or 7k = (71 + w). When the machine is that occur in practise are those due to sustained supplying an energy component of load the air-gap hunting of synchronous machines, and prime movers flux is distorted and unsymmetrical, so that the zeros of or drives with cyclic torques (reciprocating engines, the flux harmonics are not necessarily coincident; and only consideration of the effects of armature compressors, etc.). It is seen from Equation (17) that such cyclic varia- reaction and air-gap permeance will permit /k and k tions in speed affect both the amplitudes and wave- to be completely specified. If all phase belts of an alternator are alike (that is an lengths of the induced voltage. there iS the class of time variations in integer number of slots per phase per pole) and if Q is (d) Finally flux and speed represented by damped (or cumulative) the number of phases per I pair of poles, then from Equation (9) of Appendix oscillations. Each of these appears in the voltage as a pair of damped (or cumulative) oscillations of different sin (2 k - 1) r/Q - C sin (2 k - 1) v/c Q frequencies. They are caused by hunting, pulsating cdk speed regulation by the governors of prime movers, but a uy ecsriu oi. = 0 if (2 k - 1) = multiple of QcQ not of (c Q) and unsymmetrical short circuits of polyphase mnach ines. 1i 2k-1 utpeo III. APPLICATIONS ~~= 1 if (2 k -1) = multiple of (c Q) III1. APPLICATIONS It is rather interesting that the distribution factor of Under this section of the paper, the application of a Q phase machine will eliminate the multiple of Q Equation (17) to some of the more familiar types of harmonics if there is more than one coil per phase belt, electrical machines will be illustrated. In order to but will not eliminate those harmonics which are also a conserve space, certain simplifying assumptions will be multiple of the product of the number of phases by adopted in making these applications. But it must be the coils per phase belt, (c Q). For instance, if a true borne in mind that the generality of the equation is not three-phase machine (120 deg. phase belts) has three limited by any such restrictions as may be invoked in coils per phase belt, then the third harmonic will be the interests of simplicity and brevity. An extension completely wiped out, but the ninth harmonic and its of the analysis to include more detailed considerations multiples will not be reduced by the distribution. will be obvious from the simple specifications given However, in a double layer winding the ninth would be here. wiped out by the pitch factor in this case. A single Synchronous Generator. If all of the phenomena layer wi-nding would have to be used to take advanwhich occur in a synchronous machine are to be included tage of tivin its analysis, then every term of Equation (17) will be ior Z1line given above is based on the Then"the required for the description of the induced voltage. assumptionl that the armature is the revolving member. Thus an unsymmetrical short circuit will cause a If the field revolves, then x0 = 0, and 7k = Yk + 00 t damped pulsating armature reaction with a correspond- (since the flux distribution moves as a rigid system), ing variation of flux; the passage of teeth and slots so that d 7k/d t-=c0. Making these substitutions, the cause the teeth harmonics of flux distribution to sweep same equation is obtained as that given above for the across the pole face; and the hunting of a synchronous case of a revolving armature, except that 7k replaces machine superimposes on the constant speed of rotation 7k as the constant angle fixing the relative position of an oscillation which may be damped, sustained, or even the harmonics in the rigid distribution of flux. cumulative. But neglecting such abnormal conditions, Polyphase Induction Motor. Consider for the sake

(kident

tim)

April 1930

BEWLEY: INDUCED VOLTAGE OF ELECTRICAL MACHINES

461

of simplicity only the space harmonic of fundamental frequency f in the revolving field of the air-gap flux, and let fo be the mechanical frequency of rotor rotation. If the direction of rotation of the revolving field be taken as positive, then the angle -yi in the Fourier series for the flux distribution, Equation (1), is negative, so that

Xo Substituting these values in Equation (17) there is n N' Erotor = 4 L - 108 CCl Cdl Cp1 Csl l

t yh =-27rft,

.. dyl/dt -27rf 2 ir fo t; .-. d xo/d t = 2 rfo

possibility of taking either sign is necessary in order to neutralize a change in sign caused by the harmonic reduction factors. Since the "phase belt" of a d-c. machine spans a pole pitch, the distribution factor is sin k 7r/2 2 k7r . sin if c + OC Cdk = 2 Cdncknk7r/2c - k r Thus even harmonics in the flux distribution will not contribute to the induced voltage. For a full pitch, infinitely distributed winding, without skew, . kw

(fu du From the standpoint of telephone interference, a rather interesting problem in the design of d-c. commutating machines is the calculation of the voltage pulsation due to the oscillation of the armature conductors about their mean position with respect to the brushes. These ripples may be computed by allowing 4 to vary over the oscillation range, in the above equations. d x/dt - velocityfof conductors While the segregation and ultimate specification of = (no.iy poles)t(rev, prmn/0=2' o these pulsations as Fourier series is usually desirable, Vlk = Vlk = 0 since there is only one belt between space will not be taken up here for such a study. brushes. Synchronous Converter. The a-c. and d-c. induced voltSubstituting these values in Equation (17), the induced are found from the voltage across an adjacent pair of brushes is found to be ages of a synchronous converter and D-C. Generators equations for the Synchronous Ebrukes 2~ LN' (RPM) (oe)respectively. 108 60 Substituting in the equation for the Synchronous Generator the values
Fo

(f fo) [2 (fo f) t + V/i' + V/] This equation shows that at standstill, fo 0, the voltage induced in the rotor is of stator frequency, f. If the rotor is running at synchronism the voltage is zero and direct current. For speeds above synchronism the voltage increases in magnitude and frequency with the speed. The equation for the stator counter e. m. f. is, of course, of the same form as that derived in the previous case of a synchronous generator. D-C. Generator. The exact calculation of the voltage across the brushes of a d-c. generator would take into consideration such things as the variation of the airgap flux due to the moving teeth and the pulsation in armature reaction caused by the oscillation about its mean position of the belt of conductors between brushes; the sector of brush overlap; the slight cyclic variation of field excitation when under automatic regulator control, etc. Assuming, however, that the number of armature conductors per pole is reasonably large, so that the effects of the belt oscillation may be neglected, and disregarding the teeth pulsations, brush overlap, etc., the following conditions obtain: d _Yk d t = 0 since there iS no movement of flux... d ,Okld tt = O since there iS -no pulsation of flux nopulsat o do 7r/T = 4 s angle breish x0 71/d = 0=inceh of brush shift-... = t
-

C,k -sin2
N'

and Csk=1.

cos

Eirusies --2 i L 108

2 r
co

(RPM) (Poles) 60
(2 k - 1)7r 2

siw-

2k-1

2 k-

cos (2 k-1) ('2k-1 4) N' (RPM) N' -2 108 4 (Poles) = -4 Jos (f 60 where > =

'rL r
2 2

> 1(2k-31) sin (2 k-1) ('Y2k-1 + ) d 4 ____ 2k-1 Cos (2 k-1) ( y2

I

Wr

2k-i

btweenb

toltheoddeharmncs)

00

n=1,

4k =O, V/k'= O, N'==c N, Cdk'= Cdk = C ink 3/2

Cdk k fcCos k ( yk + 4) (36) where 6 = 2w/c Q = slot angle; the phase voltage is CpCOk =T L i 8" Ephase = - 4 LTf0 N 10 8 evdn ~ It is eintthat the voltage is a maximum when ,sin k wr/Q zb= -(4+mw), for then cos k QYkb+ 4) = 1 2> sin k5/2 CPk CSk1k COSk (X0t+ yk) and each harmonic contributes a maximum. The

462

BEWLEY: INDUCED VOLTAGE OF ELECTRICAL MACHINES

Transactions A. I. E. E.

The effective value of the fundamental (the applied e. m. f.) is 4 sin 7r/Q E1 = - L fN 10-8 sin 6/2 C1PI Csl (31 V/2

of a standard converter supplied with a-c. field excitation, or for that matter, an excitation varying in any arbitrary way. The Transformer. There is no practical advantage in showing that Equation (17) includes the special case Substituting in the equation for the D-C. Generator of the static transformer. The demonstration is insin k 7r/2 cluded here merely for the sake of completeness and to the value Cdk= si k6/2 gives for the d-c. voltage show the general validity of the equation. Consider an ideal transformer having a rectangular core of section Ed = --4 L foNl-8 (N L); and neglect the leakage flux. Then at any ino . stant the flux density is uniform over the core, and the k 7r/2 Cpk C,k /k cos k (yk + t) the direction of is equation of its distribution inin a rectangular wavethat sin k /2 of a rectangular wave. But the I
sin
r

Therefore the ratio of the d-c. induced voltage to harmonies are all of odd order, and their amplitudes are the effective value of the fundamental component of inversely as their order, so that Equation (1) becomes the a-c. induced voltage-the ratio of conversion-is oT Edc B= 2k- 1si(2 1 CosQ(y 1-I E1 10 sin 7/Q
Since the turns are full pitch, and the conductors are not skewed or distributed with respect to the distribution of flux, it follows that C,k = 1, Cpk = sin k r/2, Cdk = 1, 41, = 0. Both the conductors and space distribution of flux are stationary, so that d xo/d t = 0 and d yk/d t = 0. The coordinate to the midpoint of the coil is xo = T/2. The induced voltage therefore is, by Equation (17) and the above conditions: T N' 4 dB sin2(2k-i) w/2 t (2 k-1)2 E =-2 L a N' d B N' d f - r - T 108 d t 108 d t 8 1 since (2k 1)2 = l and (r L B) = 1 Although rather irrelevant to the subject matter of The condition that this voltage must have the same this paper, a discussion of the theory of induction in a frequency f2 as that applied to the slip rings, requires transformer by "cutting action" may not be too much that out of place. 4 f2 = (fO -fl) or fo = (fi 4 f2) In the above equations B is the instantaneous value Then xo = 2 r fo t, d xold t = 2 r fo of the flux density pulsating at frequency f, so that B = Bosin wct and therefore -2 2wf,t, d oyi/dt = -2 2wf y The voltage at the brushes, as in the analysis for the 4 sin w t . sin (2 k-1) x 7rlr B = Bo Tw D-C. Generator is (2 k - 1)
EF,e Ar wr 10 CdlCPlC =- L l-3(fO-fl)eos (t-2f17rF t)
z

2 = /2 A/sin (7r/Q) For a detailed discussion of the voltage ratio of a converter, based on this equation, reference may be made to The Synchronous Converter by T. T. Hambleton and L. V. Bewley, A. I. E. E. TRANS., Vol. XLVI, 1927, p. 60. Synchronous Converter with Polyphase Field. Consider a synchronous converter armature with a frequency f2 applied to its slip rings, revolving in the polyphase stator field of an induction motor supplied with currents of frequency fl. If fo denotes the frequency of mechanical rotation of the armature, then from the equation for the rotor voltage of a Polyphase Induction Motor, there is (considering only the fundamental) N' Eroor=4L rn 108 Cdl Cpi Cs,ii(fi-fo)cos2 7(fo-fi)t

sin k ir/2 . sin 6/2 . Cpk C0,kk k 5msi rQ . sin k /2 . CplC5,u1 cos k ('y1 t) J1
k

4_

sin(2k-1)7rx/r
(2 k - 1)

N'c

-B0

'A

{COS[ Cot- (2k- 1)

- ]

Thus the voltage at the brushes is of the same fre1 TX 1 1 quency as that of the field, but the amplitude is pro-r portional to the frequency applied at the slip rings. - cos L X t + (2 k-i) T J = E + z (3 If fi = 0, the field is d-c. and the equations reduce to those for the ordinary Synchronous Converter. The term [co t (2 k -1) iv x/1-] = constant defines Equation (17) is also readily applicable to the case the position of any given point on the cosine wave at

April 1930

BEWLEY: INDUCED VOLTAGE OF ELECTRICAL MACHINES

463

any time t, and this point moves across the core section exaggerated cyclic torque. Segregate out the time +: -, harmonics for different amplitudes and periods of dx at a velocity d t = ( k d (2 1) 7r ' thus inversely oscillation. phase of a two-phase induction motor (d) Each proportional to the order of the space harmonic. The stator is excited at a different voltage and frequency. pulsating flux wave has thus been decomposed into two Find the slip ring voltage induced by the stator flux systems E ,B' and E 3" of traveling waves moving at in a three-phase Y-connected rotor driven externally equal speeds in opposite directions. The flux density at constant speed. (e) A d-c. generator field is excited with alternating at the conductors, located at x = 0 and x = r, is found by substituting these limits in the expression for B. current. Calculate the voltage at the brushes. (f) The field of a separately excited d-c. generator The induced voltage, calculated by the (B L V) is closed as the machine starts to accelerate from method, then is standstill. For a reasonable field time constant and - N rate of acceleration plot the voltage at the brushes. e 108 (g) Study the effects of slots and teeth in producing NL voltage ripples, by analyzing flux plots and specifying -_N _X=0 + Ox_=-_x=o_x=_ as Fourier series. In particular, decide whether these 108 7r 2k-I flux harmonics are standing or traveling waves. (h) Calculate the voltage ripples of a d-c. generator -Nw when the oscillation about its mean position of the belt 2 4 cos w t 7 L Bo2 = of conductors between brushes is taken into account. ( -)2 i Specify as Fourier series. Nw (r LBo) cos Xo t The author wishes to thank Mr. S. T. Maunder who verified the equations with independent derivations. -N co N do - 108 f Cos w t LIST OF SYMBOLS 108 d t B = instantaneous flux density If the core section is not rectangular it may be re- Bo = maximum flux density = coil sides per phase belt placed by any rectangular section having the same effec- c tive area, r L = A, and operating at the same density. C,, = skew factor for the kth harmonic Although the correct result has been obtained by Cr, = pitch factor for the kth harmonic considering the standing waves to be composed of two Cdhdistribution factor for the kth harmonic = belt series of traveling waves, and then computing the Cdk' = equivalent distribution factor for dissimilar induced voltage as due to "cutting action," yet at the phase belts in series conductors themselves the flux density is zero, for C,, = phase connection factor for the kth harmonic O =0 + Ox == 0 em = voltage induced in the mth coil of a phase belt + x= 0x eg = voltage induced in a single-phase group Thus in an application of the (B L V) method, the ep = voltage induced in a complete phase component flux densities and their individual velocities Eiine = induced line voltage relative to the conductor must be considered separately, f = frequency and only those flux densities superimposed which have k = order of the harmonic equal velocities and in the same direction at the L = effective length of stacking point of superposition. This interesting situation was m = position of coil in a phase group previously pointed out by Dr. H. B. Dwight in the April n = number of phases in series between lines 1928 issue of the Electric Journal. N = number of turns per coil Additional Applications. For those wishing to be- N' = number of turns per phase come familiar with the use of the general equation, p = per cent pitch the following problems may prove interesting and P = number of poles instructive. = number of phase groups per phase q (a) Calculate and plot representative ranges of the Q = number of phases per pair of poles voltage wave of a synchronous generator on open cir- Rev. per min. = revolutions per minute cuit brought up to speed by a prime mover delivering t = time constant torque, assuming friction and windage inde- v = velocity of moving flux pendent of speed. V = velocity of moving conductors (b) Plot the open-circuit voltage of a synchronous x = coordinate of any point in the direction of generator supplied with field excitation from a rectimotion fying device having a second time harmonic. x0 = coordinate of the midpoint of the first coil group (c) Plot the open-circuit voltage of a synchronous y = coordinate of any point perpendicular to the generator driven by a reciprocating engine having an motion

-(BLV)

464
a

BEWLEY: INDUCED VOLTAGE OF ELECTRICAL MACHINES

= =

Transactions A. I. E. E.

/k
'yk

5
E

6' 6 X
Of r b 'Pk

mechanical angle of skew instantaneous amplitude of the kth harmonic of flux density = instantaneous displacement angle of the kth harmonic of flux density = electrical angle between coil sides = base of Naperian logarithms = angle between adjacent phases = displacement angle of a phase group = displacement angle of a phase = electrical angle of skew = angle of brush shift = slot pitch distance -pole pitch distance

c = number of coils in the coil group u- = distance between adjacent coil sides = slot pitch a = angle of skew y = coordinate at right angles to the direction of motion, reckoned from midpoint of the coil side. The total flux included by the mth coil is, by Equations (1), (2), and (3), L/2 Xm' = r O B d y dx
-L/2 xm

= shift of the kth harmonic due to phases in series shift of the kth time harmonic due to dissimilar belts = 2 wr f = angular velocity where L = sign of summation, effective as implied.

==flux

2L

2 1

3k x 7r k Cpk C,k sin k T

k

C + -2

(4)

effective length of the stacking

sin k (L r tan a/2 r) sin k X/2

Appendix I

GENERAL EQUATION FOR THE INDUCED VOLTAGE OF sk k (L 7r tan a/2 r) k X/2 ELECTRICAL MACHINES duction factor for the kth harmonic (5) The following analysis applies to electrical machines having uniformly distributed parallel coil sides moving p 7r = pitch reduction factor for the kth through a distribution of flux which is uniform in a C2J = sin k direction in the plane of the coil perpendicular to its harmonic motion. (6) Now in general, X0, Ok, and 'Yk are all functions of Let the space distribution of flux density due to the combined effects of field excitations, armature reactions, time t; so that the induced voltage of the mth coil of and leakage reactances be specified by the Fourier N turns is series: -N -N d m O/m d xo e BU)108 X dt 108 dt xo
co

where k = order of the space harmonic Ok = amplitudeofthetkthrharmonic x = coordinate of any point wavelength of the fundamental = pole pitch = displacement angle of the kth harmonic. #yk If the coil sides situated in this field are skewed with respect to the poles, the equations of the trailing and leading coil sides of the mth coil, counting from the trailing end of the group of coils, are respectively (see Fig. 5)
=

Ok sin k t

ah J

+
=V

, a
b
1

dYjk
dt

0) ,

akm dIk l dt

108

1 k
d

dk. . 7r c + 1-2 m odtsint k Ix2. -r

N A 1

(___ +
t

d ) 07 COS k( dt

7 + 2 xm =xo-p 2T (2) a-+(m-1) -+ytan 2The total voltage for a phase group consisting of c r c-i such coils uniformly distributed by the pitch cf is, by Xm' =x0+P 2- 2 a-+(m-1) L+Y tan a=Xm+p r the method of summation given in Appendix II,
Ir

c- I

c+1-2m

r-w

'

O -coordinate of the midpoint of the coil group p = pitch of the coils expressed as a fraction

where where

result ~~~~~~~~~~~~~*This is obtainable Flux Linkages and Electromagnetic (19) of ~~~~~~~~~~~~~~tiontlie autbor's paper by direct substitution in Equa-

Indulction in Closed Circusits, A. I. E. E. Quarterly TRANS., Vol. 48, April 1929, p. 327.

April 1930

BEWLEY: INDUCED VOLTAGE OF ELECTRICAL MACHINES

465

e=
1

c~~~~~
em = -2 L
rr108
1

Cdk Cpk Csk

Eiine

2L

- n108 C,N
0
1

Cdk Cpk C8k Ck

p
s

[1ldf3k k

k dt

sint k (r

Xo1r .Yk + k)

1 df3k [ k dt sin k (

'r_ + tk + 4k + 41k
k Xo

(djk -d jOk cOk s ) To ; -+ t J

where

X d t Yk X Ir +o
(8)
where

+d t

j dXt

'Y)co

7(+ Yk+ ;+

(17)
n

Cdk

S1fl c sin k c* sn

sin k (c o-7r/2 r) (o ir/2 r) 7r/2 r)

c c

sink c8/2 sin k 8/2 = distribution sin 6/2

Cci

reduction factor for the kth harmonic. (9) (18) If the complete phase is made up of q phase groups ds displacesin k r)/( E cos k Or connected in series, but unsymmetrically arranged and 'Pk = tanwith a different number of coils per group (as in the mentangleforthekthharmonic case of fractional slots/pole/phase) then the phase (19) If the n phases are displaced by the same amount voltage is, by the method of summation given in from adjacent phases, Equations (18) and (19) reduce, Appendix II. as shown in Appendix II, to q k n Cpk N' N 2 (20) ep= e,, -2 L ~~r 108 NsinCdk' ~/2 C8k C,k = nsink //2
c

+ ( sin k Or) ] -connection reduction factor for the kth harmonic

[(1E c k r) Cos

[ k
+ (d d

d t sin k (
t

+ + k ) k + k+ d t

tk +

dx) , cos rk ( t

{k')]
(11)

where

Cdk'

V\Ak' + Bk'/N' = equivalent distribution factor of the phase, for the kth harmonic 46,b' = tan-' (Bk/Ak) = displacement angle of the kth harmonic in the phase voltage (13)
=

Ak =

qf

[c N Cdk], cos k 6,'

(14) (15)

Bk = I [cN Cdkl, sin k 0,

N' = I [c N],
q

1~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~nSince 1+ Zn

(21) for the Equation (17) is the most general equation induced voltage of electrical machines, if xo, f3k, and 'Yk rare perfectly arbitrary functions of time. Its application to the special cases of the more ordinary types of electrical machines is discussed in the main body of t pap Appendix II THE DISTRIBUTION SUMMATIONS In the harmonic analysis of electrical machines it is repeatedly necessary to find the sum of certain finite trigonometric series. Those required in this paper are given below. n-1
z

1lk= k (2>)

r=0 total number of turns in the phase 1 ~~~~~~~~~~~~~~~~ (16) it follows, upon replacing e ja by z, that 0,' = position of the midpoint of phase group g in -1 n-I r a) = electrical radians reckoned from the reference i(x+ra) - i-A(x+ra) sin (x point on the armature. If all groups are 9 ( alike, and displaced by multiples of 2wu, then r 2=n Cdk' reduces to Cdk. EJX I I E--Jna e- x i 1 -E-o \ 1 -ia J Two or more phases may be connected in series, 2] 21 t 1 - C-ia / j , i-a&(+n-)-C(x+)-C-x either additive or subtractive, to form the line voltage. If the displacement angles of the n phases as connected = 1 +n I in series are (6,, 62 ... n) the total line voltage is, by =21 + +e<+ +e +the method of summation in Appendix II. L1-Cm"i - Ca< + 1
=

r CaX+1Xa) -+ -j(x+naa)+C-ej(x ,-a)j

466

BEWLEY: INDUCED VOLTAGE OF ELECTRICAL MACHINES

Transactions A. I. E. E.

sin x-sin(x- a) +sin (xn a- a)-sin (x+n a) 2 (1 - cos a) 2 sin a/2 cos (2 x - a)/2 -2 sin a/2. cos (2 x + 2 n a- a)/2
cos (2 x- a)/2- cos (2 x- a + 2 n a)/2 2 sin a/2ErCSO

By the symbolic operators of complex notation: 2 Er E Er eJOr = E Er COS Or+j z Er sin Or= Eo eiOo

E2
00

where = V( EEcos Or)' + ( Er5sin

= tan-'

(6)
(7) (8)
7

Or)2

4S Er. EsOr Ers O,

sin n a/2

sin x +

2 1a

(2)
I an 2

Or transforming to trigonometric notation

2

Replacing (a) by (- a) gives

n-1

.sin (x - ra)
r=O

sin n a/2 ~~~~sin a/2 sin

x-

(3)

22 E, cos (x Or) = Eo cos (x + 0,) (10) = Er If E, = E2 = .E. = E and 0, = r a then (7) and (8) reduce to (2) and (9) as follows: 2 n-l n2

Er

sin (X 4 0r) = Eo sin (x +O0)

(9)

-12
sin r a)

Replacing (x) by (x + 90 deg.) gives

X-n

Eo =E

( >Zcosr) +r(
a
o

r =O

cos

( sin n a/2 n-i (x + r a) = sin a/2 cos x + 2 I a)

n ~~~~~sinae/2 r a) =Osinn/2 c

E sin na/2 sin a/2

n-1
r=0

(\(cos n-i

N (x cos

=sina/2

(x

n- 1

a)'+(sin2nsin a/2
( )a

a)2

The more general sums, in which the angles do not differ by a regular amount, and in which different Ho = tan' (tan amplitudes are involved, are found as follows:

(5)

E 2sinna/2

ni2