Synchronous Machines IV

BY R. E. DOHERTY* and C. A. NICKLE*

Member, A. I. E. E. Associate, A. I. E. E.

Synopsis.-The special case of cylindrical rotor machines has in the open phase, and for voltages across external reactances in the
been treated previously by Boucherot and others. The present armature and field circuits. Comparison of calculated and test
paper solves the general case, including salient pole machines. curves for these quantities are shown in Figs. 5 to 8, for a 7000-kv-a.
The cylindrical rotor type thus becomes merely a limiting case. 375 rev. per min. salient pole alternator, and in Figs. 11 to 25 for a
The principal assumption which distinguishes the present theory 20-hp. 1800-rev. per min. induction machine (cylindrical rotor).
from the extensively studied cylindrical rotor theory is that the These show satisfactory agreement.
total armature self-inductance is here taken as variable with respect The form of the results is interesting and instructive. The
to rotor position, whereas the previous theory of short circuits, as short-circuit currents are expressed, as in equation (9), as the sum
represented by Boucherot, for instance, assumes this inductance to of two series, one of odd, the other of even, harmonics, in which
be constant-in other words, that the air-gap is uniform. the magnitudes of the harmonics, in ascending order, are in geomet-
The four basic concepts underlying the improvement in theory, ric ratio. The value of that ratio is
both as applied here and in the authors' previous work on Syn-
chronous Machines, are: Vx - XD
1. Characterization of the machine by four reactance coefficients, b , zj + g
two corresponding to the main pole axis, i. e., direct axis, and two to Q + XD
the interpolar axis, i. e., quadrature axis. These are: x0, x0', where XD and xQ are transient reactances in the direct and quadra-
xQ, XQ'. (See notation.) Thus the theory has been referred to, ture axes.
more or less aptly, as the "Four-Constant Theory." The effect of salient poles in reducing the relative magnitude of
2. Resolution of flux and m. m. f. waves traveling with respect the harmonics with respect to the fundamental in the short-circuit
to the rotor into stationary, pulsating components in line with the current is clearly shown by the equations. This is important both
direct and quadrature axes. The theory involves also, of course, the in connection with the calculation of mechanical forces, and of the
usual Blondel resolution of the stationary fundamental waves. voltage induced in the open phase during single-phase short circuit.
3. That the variable component of armature inductance varies It is shown that although this voltage may be 15 or 20 times normal
between the direct axis value and the quadrature axis value as a in a laminated, cylindrical rotor machine without amortisseur
second harmonic function of the electrical space angle. This circuits, it may be only 2 to 4 times normal in a salient pole machine.
applies to all variable components of inductance which, with arma- An amortisseur winding greatly reduces it in either case.
ture current of fundamental frequency, produce fundamental voltage. The fundamental component of this voltage across the open phase
4. The complete use of a fractional system of notation throughout is shown to be approximately equal to (1 + b) times the voltage
the equations, thus expressing each quantity as a decimal fraction of before short circuit.
some definite, convenient value, such as "normal" value. This is, of The mathematical work and definitions are in Appendixes as
course, an outgrowth of the familiar practise of expressing such follows:
quantities as a per cent of normal value, but the latter involves A Summary of equations.
carrying along the "100" in calculations, which is both cumbersome B Armature current.
and conducive to numerical error. Experience has established a very C Field current.
definite prefe&ence for the former-the only difference between them, D Voltage across external reactance in armature circuit.
of course, being the position of the decimal point. Thus, with E Voltage across external reactance in field circuit.
unity as the base, instead of "100," it is proposed to denote the F Voltage across the open phase.
former by "per unit" system, instead of "percentage" system. G Variation of armature inductance with position.
Turning to the results of the paper, expressions are derived for the H Decrementfactors.
short-circuit currents in the armature and field, for voltage induced I Reactances, definitions, and relations.

HISTORICAL Certain special cases of short circuits have been

A T the White Sulphur Springs Convention in 1926
the authors presented the first of a series of papers
covering their investigations of the Synchronous
Machines. Part I was an extension of fundamental
theory; Part II, a treatment of torque-angle character-
istics under steady state conditions; and Part III,
presented at the 1927 Winter Convention, a study
of torque-angle characteristics under transient con-
ditions. In the present paper, an analysis is made of
single-phase short circuits. In the future, the treat-
ments of other characteristics, including polyphase
short circuits and the torque existing under short-
circuit conditions, may be given,
*Both of the General Electric Co.
1.~~~~~Fo.eeecsse ilorpy
Presented at the Winter Convention of the A. It E. E., N~ew Vork, circuits. Working from the same fundamental prem-
N. V., February 13-17, 1928.
treated previously. Following the early work by
Steinmetz1 and Berg, Boucherot2 presented his classical
paper in 1912, dealing with alternators of the laminated,
cylindrical rotor type, with and without amortisseur
windings. In 1915 Diamant4 reviewed the work of
Berg and Boucherot, and, among other mathematical
extensions, gave expressions for the envelope of the
short-circuit current. In 1918 he presented an analysis
of sustained short circuits, dealing principally with the
nature of flux distribution under that condition. In
the same yearsa one of the authors proposed the use of
the Constant ixnkage Theorem in dealing with short
circuits, and in 19219b and 19239c illustrated its use.
Also in 1923 Franklin5 applied this method to a large
number of cases of single-phase and three-phase short
ises, Laffoon7 solved additional cases in 1924. The
457F

28-15
458 DOHERTY AND NICKLE: SYNCHRONOUS MAC'HINES-IV
following year Karapetoff6 analyzed, in the main, the
same cases as those of Franklin and Laffoon; but he
started from a somewhat different theoretical basis,
i. e., from Kirchoff's Laws. But he, like Franklin and
Laffoon, neglected resistance, which made the funda-
mental premises the same-since the Constant Linkage
Theorem is merely a corollary of Kirchoff's Second Law,
for the special case of negligible resistance. That is,
his equations involved voltages, instead of magnetic
linkages, and were therefore the first derivatives of the
corresponding equations in Franklin's work. Involv-
ing the same assumptions regarding the circuits, the
results naturally also agreed. In 1923, Lyon"3 outlined
a method according to which "certain low-frequency
transient conditions in electric machinery may be
analyzed by a vector method." From the vector
point of view, first proposed by Dreyfus'4'15,16 in 1912,
the transient voltage or current may be represented by a
series of vectors which shrink exponentially as they
rotate. The thorough, mathematical treatments of
cylindrical rotor machines both by Shimidzu and Ito'
and by Biermanns,3 in which the effect of resistance is
approximately taken into account, are especially note-
worthy. All of the foregoing investigations have dealt
with cylindrical rotor machines; and, granting the
simplifying assumptions which have appeared to be
practically necessary, such as those made by Shimidzu
and Ito, and Biermanns, relating to resistance and
saturation, it may be said that a practical solution of
short circuits for the cylindrical rotor type has been
attained.*
SCOPE
These previous analyses, however, have not solved
the important case of salient pole machines. The
present paper treats the general case, including both
types. Proceeding from the same basic point of view
from which the authors' previous work on Synchronous
Machines9d,9c has been developed, the present treatment
takes up single-phase short circuits and partial short
circuits, i. e., with external reactance in the armature
and field circuits, developing expressions for armature
current, field current, armature voltage across the open
phases, voltage across reactance in the armature circuit,
and the voltage across reactance in the field circuit.
These expressions apply not only to salient pole
machines, but also to the cylindrical rotor type,
since under the present theory, the latter type merely for instance, where saturation exists, or where there are short-
becomes a special case of the former. Thus, the.
scope includes the development of expressions, appli- desirable to use an additional constant, namely, the zero phase-
incl te
cable alike to salient poles and eylindrical rotors, for sequence reactance. The positive phase and negative phase values
the voltage and current phenomena under single-phase
short circuit.
*After submitting the present manuscript, a copy of S. Bekku's
treatment of "Sudden Short Circuit of Alternator," Bibliography
12, has been received. This treats the case of the cylindrical
rotor machine with a three-phase winding on the rotor, each
rotor circuit being closed either directly or through the exciter.
He employs "Heaviside" in the mathematical work.
Transactions A. 1. E. E.
Although the single-phase phenomena of three-phase
machines only are considered here, the theory is, of
course, applicable also to two-phase machines.
DISCUSSION OF THEORY
The four basic concepts underlying the improvement
in theory, both as applied here and in the authors'
previous work, are: (1) characterization of the machine
by four reactance coefficients, two corresponding to the
main pole axis, i. e., direct axis, and two to the inter-
polar axis, i. e., quadrature axis. These are XD, XD',
XQ, XQ'. (See notation.) Thus the theory has been
referred to, more or less aptly, as the "Four-Constant
Theory.t
2. Resolution of flux and m. m. f. waves traveling
with respect to the rotor into stationary, pulsating
components in line with the direct and quadrature axes.
The theory involves also, of course, the usual Blondel
resolution of the stationary fundamental waves.
3. That the variable component of armature induc-
tance varies between the direct axis value and the
quadrature axis value as a second harmonic function of
the electrical space angle. This applies to all variable
components of inductance which, with armature cur-
rent of fundamental frequency, produce fundamental
voltage.
The general aspects of this premise should be care-
fully considered. In a salient pole machine, if the axis
of the armature fundamental m. m. f. is not in line with
either the pole axis or the interpolar axis, the resulting
fundamental flux wave, of course, will not be in space
phase with the m. m. f. wave; and therefore calculations
involving the total waves would be very complicated.
By symmetry, the flux wave will be in space phase with
the m. m. f. wave producing it, if the latter is in line
with either of these axes. The purpose of resolving
the fundamental wave of m. m. f. along these two
particular axes is thus to obtain component flux waves
tThere are, of course, more than four. The four constants
which characterize the machine under single-phase operation are
not the same as the four which characterize it under polyphase
operation. They are, however, of the same general nature, thus
corresponding each to each, the chief difference being due to the
mutual inductance between phases. Nor are the four constants
the same for a line-to-neutral, as for a line-to-line, short circuit.
Thus, there are twelve constants in this particular connection,
and even some of these may be split up under special conditions;
eircuited windings on the rotor in addition to the main field
winding, etc. Moreover, in certain network calculations, it is
are also used in the latter connection, but these are expressed in
terms of the "four constants." And, if one should include also
the calculation of all the possible phenomena due to harmonics,
the number of necessary constants could be extended almost
indefinitely. Thus while it is correct to say that the macehine,
under single-phase or polyphase operation, is characterized
under the premises of this paper by four constants, there are, all
told, many more than four values of reactance to be considered
in general.
Feb. 1928 DOHERTY AND NICKLE: SYNCHRONOUS MACHINES-IV 459

in space phase with the respective m. m. f. components
producing them.
Now it is a mere incident of machine design that there
is symmetry about the two axes mentioned, and that
these axes happen to be 90 electrical degrees apart.
But these facts underly the possibility of expressing the
variation of armature inductance as a simple second
harmonic, as in equation (3). For derivation of this,
see Appendix G.
Such a variation is shown to apply not only in the case
where the field winding is open circuited, i. e., to the case
of pure reluctance symmetry of the rotor magnetic
circuit, but also to the case of artificial reluctance sym-
metry as produced by short-circuited rotor windings,
such as the field winding. A short-circuited secondary
sAxis of Single Arrnature/\
Phase Winding Winding
Thus, considering the variable components of induc-
tance, it is permissible, under the premises, to regard the
variation as a second harmonic of the electrical space
angle, regardless of whether the variation is produced
by the shape of the iron magnetic circuit, by short-
circuited secondary windings on the rotor, or by both;
and regardless of whether the variation affects the
fundamental or the harmonic flux waves, so long as the
reactive voltages produced by such waves are of funda-
mental frequency. Fig. 2 shows the variation of the
4.0
9,'
20-
/\
,

838i3;B4g FJ X~~~~~~~~~~~~~F V
CURVE A

Ax IS Artr¢ Ais Winding

A 4.0

2.0

9?J 217
Axis of ~~~~~~~C
URVE B

FIG. 2-VARIATION IN TRANSIENT INDUCTANCE, 20-Hp.

MACHINE
Curve A-No saturation
Curve B-With saturation

Motion fPA.
C .6
FIG. 1

on the rotor merely increases the reluctance of the path v22

of the armature flux; and in machines as ordinarily
constructed, the artificial reluctance symmetry is about
the same two axes as when the secondary windings are
open-circuited, since the axes of the latter windings
coincide with the direct and quadrature axes. See
Fig. 1.
There is also a relatively small component of armature
leakage inductance, which varies as a second harmonic.
Although this component is due to space harmonics of
flux, it nevertheless produces a reactive voltage of the
same frequency as the armature current. This has
been considered by KarapetoffGa and Arnold,'1 its
significance is fully discussed in Synchronous Machines,
_I,9d and its magnitude is given by Alger.t
t' C8alculation of Reactance of Synchronous Machines,"
by P. A. Alger, p. 493.
0
FIG. 3-VARIATION IN TRANSIENT INDUCTANCE-4-POLE,
15-KV-A., 1800-REV. PER MIN., SALIENT-POLE MACHINE
transient inductance of a 20-hp. machine, on which
many of the tests described herein were taken. Curve
A was taken under the condition of no saturation, and
is practically a sine wave of double frequency, whereas
curve B was taken under the condition of partial
saturation, and departs considerably from a sine wave-
thus containing significant even harmonics. Fig. 3
shows the same curve for a four-pole, 15-k-va., 1800-
rev, per mmn. salient-pole machine without saturation.
Fig. 4 shows the curve for a 7000-k-va. machine. This
460 DOHERTY AND NICKLE: SYNCHRONOUS MACHINES-IV
was obtained from an oscillographic record taken in
connection with the "slip method." '
4. The complete use of a fractional system of notation
throughout the equations, thus expressing each quantity
as a decimal fraction of some definite, convenient value, harmonic reactive voltage due to an nth harmonic current
such as "normal" value. This is, of course, an out-
growth of the familiar practise of expressing such
quantities as a percentage of normal value, but the
latter involves carrying along the "100" in calculations, sponding per-unit reactive voltage. For the funda-
which is both cumbersome and conducive to numerical
error. Experience has established a very definite
preference for the former, the only difference between
them, of course, being the position of the decimal
point. Thus with unity as the base, instead of "100,"
it is proposed to denote the former by "per-unit"
system, instead of "percentage" system.
The use of the "per-unit" system has distinct
advantages. In the first place, the results are in forms
which have a definite significance in giving at once an
idea of relative values; for instance, a current of 0.5
means 50 per cent of normal current. Another advan-
LO
| \ /tion, include these linkages; it would include only those
o 0z,
FIG. 4-VARIATION IN TRANSIENT INDUCTANCE-7000-KV-A.
MACHINE
tage is that cumbersome conversion factors are thus
eliminated. For instance, as per-unit quantities, cur-
rent is taken as numerically equal to the m. m. f. of
armature reaction, and voltage, at normal frequency,
as numerically equal to the magnetic linkages which
produce it, and in many cases, as noted below, reactance
as numerically equal to inductance. It should be
mentioned that in equating per-unit current to per-unmt
armature reaction, or per-unit linkages to per-unit
voltage or field current, dimensional relations are not
disturbed, because all values are ratios; i. e., all val-
ues are numerics. Conversion factors thus become
unnecessary.
At some risk of confusion, attention is also called to
the fact that, according to the definitions, reactance and
inductance, as per-unit quantities, become identical
numerically in many special cases where the reactance
is based on fundamental frequency. Reactance is still
reactance, and inductance is still inductance; but with
each expressed as a ratio, i. e., as a numeric, they
become, by definition, numerically equal under the
special condition referred to. The definitions are:
tSee R. H. Park & B. L. Robertson, p. 514.
Transactions A. I. E. E.
per-unit inductance for the nth harmonic is the ratio of the
nth harmonic linkages, produced by an nth harmonic
current of normal value, to normal linkages. And, per-
unit reactances for the nth harmonic is the ratio of the nth
of normal value, to normal voltage. Thus, if 1 is the
circuit inductance, i,, 1 gives the per-unit linkages due
to the nth harmonic current in, and i,, x,. gives the corre-
mental component of current i,
ix = il
provided all of the linkages due to i produce voltage of
fundamental frequency; and if so, x = 1. But this is
not always so. If the inductance 1 is constant, the
linkages must be proportional to the current (under the
premises), and the fundamental current could not
produce any but fundamental linkages; but if 1 is a
function of time, as it is during rotation, a current of
one frequency may produce linkages of another fre-
quency. For instance, the fundamental component
of current in the armature may produce 5th harmonic
(in time) linkages in the armature. But the reactive
coefficient x for the fundamental would not, by defini-
of fundamental frequency. Thus, as defined, the
per-unit reactance for the fundamental, and the per-unit
inductance become numerically equal when 1 is con-
stant, and the voltage across the inductance, regardless
of the character of the current, is given by
di di
e = Id = xd
However, for rotation, the voltage is given only by
d
e = (l i)
In other words, the only difference between per-unit x
for the fundamental, and per-unit 1, is the difference
between the total linkages produced by the funda-
mental current i, and that component of the total
which is of the same frequency as i. Hence when there
is no difference, that is, when the linkages are at all
instants proportional to the component of current
under consideration, x and 1 may be taken as numeri-
cally identical.
So, considering the system of notation as a whole,
although it may possibly be a bit confusing at first, it is
nevertheless sound, has very great advantages in simpli-
fying equations, and in giving an added significance to
the results, and is time-tried; i. e., has proved its value
over years of day-to-day practical use in the solution of
complicated problems.
The four basic concepts outlined in the foregoing
paragraphs, afford the basis for symmetrical, relatively
simple, and very interesting expressions for the voltage
and currents under single-phase operation. It has been
pQssible in this way to take into account, without undue
mathematical complication, the predominating factors
Feb. 1928 DOHERTY AND NICKLE: SYNCHRONOUS MACHINES-IV 461

of the problem, and therefore to obtain a very satis- equal, the equations then apply to cylindrical rotor
factory agreement between calculated and test results, machines.
as may be seen by reference to Figs. 11 and others What difference does it make in the results if that
which follow. inductance is assumed to be variable? It makes a large
These extensions in theory have required not only difference in the relative magnitudes of the fundamental
additional characteristic machine constants, as already and harmonics, and thus also in the ratio of the ampli-
mentioned in a footnote, but also have made it neces- tude of the fundamental to the peak value, both in the
sary, in order to avoid confusion and misunderstanding, short circuit current, and in the open phase voltage.
to classify and more carefully define them. In the One important result of the theory is that the peak
early days, a synchronous machine had a reactance; value is the same for salient poles as for cylindrical
now it has many. Moreover, our ideas about some of rotors, when the transient reactance, direct axis, is the
the well established terms, for instance, armature leakage same. Thus equations (18b) and (19b) show, as indi-
reactance, and field leakage reactance, have needed cated graphically in Fig. 9, that the peak is independent
some revision. As a rough illustration in connection of XQ'. The magnitudes of the harmonics and of the
with the general situation; in running a polygonal fundamental are very important, both in determining
boundary, an error in the length or bearing of one mechanical stresses and angular vibrations, and also in
of the sides is not so apparent, unless it becomes determining the voltage in the open phase and across
necessary to "close" the polygon. The error is merely external reactance. These are greatly affected by
passed on the next side, letting the end whip about as salient poles, and can be predetermined with practical
occasion may require. In synchronous machines, it accuracy from the present theory. But they obviously
was formerly necessary to "run" only one or two sides; cannot be correctly calculated from cylindrical rotor
now, with the necessity of calculating practically all of theory based on the assumption of constant total
the electrical characteristics, we must "close the inductance with respect to rotor position.
polygon." So, an orderly classification, description The second distinguishing premise is that there are
and careful definition of terms is quite in order, as is two rotor windings, one in each of the conventional
also the presentation of methods of calculating some of axes, and having different constants. It is thus possi-
these factors. The former is given in a companion ble to take into account, at least in an approximate
paper by R. H. Park and B. L. Robertson; and the way, the effect of any short-circuited winding which
latter, by P. L. Alger. may exist in the quadrature axis, such, for instance, as a
Returning to the general aspects of the theory, non-uniform squirrel-cage winding, as used in laminated
consider now the distinguishing differences between the cylindrical rotors, or in salient-pole machines, and also
present treatment and those already referred to, which such as massive steel rotors in which the transient
have preceded it. Boucherot's work is the outstanding reactances XD' and XQ' are different. And if there are
early contribution on the subject. Others have actually only two distinct rotor circuits in the two axes,
attacked essentially the same problem in different and even though they have different constants, the theory is
perhaps more convenient ways, and have made note- completely applicable. Moreover, for a special case of
worthy extensions. But they, like Boucherot, have all cylindrical rotor in which there are two rotor windings
treated cylindrical rotor machines only, whereas the at 90 deg., i. e., one in each axis, and also in which
present investigation has dealt with the general case XD' > XQ', the characteristics can not be calculated, so
including salient pole machines, the cylindrical rotor far as the authors are aware, by the previous cylindrical
type thus becoming merely a limiting case. It may be theory. But they are easily calculated and analyzed
of interest, therefore, to draw some specific comparison by the present theory. This is illustrated by reference
relating to the differences in fundamental premises, and to Figs. 13, 14, and 15, which apply to such a case.
hencealso in the resulting equations. These are explained under "Results." Thus, the
There areoin twofesultindequame onta assumptsecond distinguishing assumption broadens the appli-
cylipn dhi- cation of the theory over the previous theory, even in
cal rotors. The first relates to the character
cylndr-the
distinguish the general treatment from that of of the case of cylindrical rotors.
total armature self-inductance; Boucherot's L2, the ASSUMPTIONS
authors' £D and 0Q. In the cylindrical rotor theory, 1. Negligible saturation. This is, of course, not true
this inductance is assumed to be constant with respect practically. However, as mentioned in Synchronous
to position-in other words, that the air-gap is uniform; Machines-Ill, this does not prevent the use of the
in the present theory, it is assumed to be variable,* results in cases where saturation is present, by judicious
as it actually is in salient pole machines. In the limit, shading of the constants, or by shading the calculated
when there is no variation, that is, when £D and £Q are results with respect to the degree and distribution of
*The total variable component comprises not oniy the variable thsautinre mbig,owv,tatheehd
term of the inductance of armature reaction, but also the small thus applied is an approximation.
variable term of armature leakage inductance. 2. The open-circuit voltage is a sine wave. This is
462 DOHERTY AND NICKLE: SYNCHRONOUS MACHINES-1V Transactionis A. I. E. E.

practically true of most modern alternators. The peared, and are thus obtained by setting the transient
combined effect of coil pitch, distribution of the wind- terms equal to zero.
ing, Y-connection, and perhaps "fractional slot" wind- The decrement factors are obtained by an approxi-
ing, practically eliminates all harmonics in the voltage mation. It seems to be Well nigh hopeless, as a practi-
wave. So this assumption is reasonably correct for cal matter, to take into account rigorously the armature
most synchronous machines. The only significant and field resistances in the general case, but it happens
practical departure might be in the case of the line-to- fortunately that it is possible to obtain results which are
neutral voltage of a machine with a relatively small sufficiently accurate for any practical purpose, by
number of full pitch coils per pole, in which case there making simplifying assumptions.
might be a significant third, and possibly other odd Before stating these specifically, the physics of the
harmonics. transient currents will be briefly considered. When the
3. The machine has a short-circuited rotor winding in armature winding is short-circuited, the magnetic flux
the quadrature axis, as well as the main field winding in linked with the closed field circuit can not change signif-
the direct axis. (See Fig. 1.) icantly in the first moment, yet, since it can not now
4. Resistance of the armature and field circuits is enter the short-circuited armature winding, it is forced
by the demagnetizing action of the armature currents,
neglected in calculating the magnitude of the initial short-
to pass through paths (the leakage paths) of greater
circuit current; and the armature resistance, in calculating
the magnitude of the sustained current. But both resis- reluctance; that is, the short-circuited armature is
tances are taken into account in the decrements. The equivalent to an increased magnetic reluctance. Hence,
effect of resistance in the impedance is usually insignifi-an additional current must appear in the field circuit in
cant, but its effect on the decrement is, of course, very order to sustain the flux in those new paths, and this
pronounced. Indeed, there would be no decrement spontaneous additional direct current, being unsup-
without resistance. ported by the exciter voltage, is, of course, transient in
5. The machine is short-circuited at no-load. A sug- character; that is, the voltage which supports it through
gested extension of the theory is to take into account the resistance of the field circuit, is generated by the
the relatively small effect of load current. decay of the flux through that circuit. Considered
thus, the problem of determining the transient becomes
FUNDAMENTAL EQUATIONTS practically identical with that of determining it for the
The general plan of attack and the general considera- simple inductive circuit, namely, to obtain the simple
tions leading up to the fundamental equations will be r
discussed here, and the final equations will be derived decrement factor L , in which r is the resistance of the
in the Appendix.
Superposition is used in obtaining the initial values circuit through which the transient flux is decaying, and
of armature and field current. The plan is first to find L is the total circuit inductance defined by the ratio of
a general expression for the magnetic linkages of the the total linkages of the circuit, to the current required
armature winding, due to any transient armature cur- to sustain them, this decrement factor being applied to
rent, and with no field current except that induced by the transient component of current. This general view
the armature current; and then to determine the arma- of decrements was presented by one of the authors9a in
ture linkages due to any other current in the field 1918, and is essentially the same as that used by
winding, and superpose these to find the resultant Biermanns.
value of armature linkages. Under short circuit, and How does armature resistance affect the field decre-
neglecting resistance, the latter value of linkages must ment factor? From the limit of zero armature resis-
be equal to the known value existing at the instant of tance, on the one hand, for which all flux linkages are
short circuit.* These relations are sufficient to solve forced into the high reluctance leakage paths, to
for the initial armature current i. infinite resistance on the other, that is open circuit,
The initial value of field current is obtained by under which condition none of the flux is forced out of
adding the component induced by the armature wind- the normal magnetic path of relatively low reluctance,
ing, to the steady value existing the instant before short there are, as far as the field transient is concerned, all
circuit. possible intervening values of effective reluctance and
There arecertaincomponents of .th t u corresponding inductance. Thus, the effect of arma-
.ture resistance on the field transient may be taken into
which are known to be transient. These, as a function account in the field decrement factor by appropriate
of time, are obtained by multiplying the individual moiiaino.heidcac.Ti is doewenta
transient components by the appropriate decrement reitnei.ageeog oafc sinfcatyth au
factor. of the direct axis component of armature current. It is
The sustained values, of course, comprise those rarely necessary to do this, but when it is, it is a much
components which exist after the transients have disap- easier task than to attempt a rigorous solution, which,
*Constant Linkage Theorem. See Bibliography 9e. to the author's knowledge, has been obtained only in the
Feb. 1928 DOHERTY AND NICKLE: SYNCHRONOUS MACHINES-IV 463
special and limited cases treated by Biermanns,3 both the armature winding and the short-circuited field
Shimidzu and Ito,8 and Bekku.12 winding. Corresponding to these fluxes there will be a
The same general argument may be applied to the direct static transient inductance 2D' and corresponding
effect of field resistance upon the armature decrement reactance XD'; similarly, when the quadrature axis is
factor, that is, with respect to the transient flux linkages lined up with the axis of the armature winding, as in
which are "trapped" in the short circuit armature 1B, there will be a quadrature static transient inductance
winding. lQ and corresponding reactance XQ'. For any other
Thus, the approximation is made that, unless the position, as in ic, the inductance will vary, as pre-
armature resistance is large enough to become signifi- viously shown, between XD' and XQ' as a second
cant in the impedance, i. e., large enough to affect the harmonic function of the angle -y between the armature
demagnetizing current in the direct axis, it is neglected winding axis and the direct axis. Thus
in the field decrement factor; and when it is taken into 1
account, this is done by an appropriate modification of 1' _ [ (lD' + lQ') + (lD'- lQ) cos 2 y]
the inductance. And under similar limitations, the 2
field resistance is neglected in the armature decrement Or, since XD' and xQ' are "static" values of per-unit
factor. reactance, they may be taken, as already explained, as
The field decrement factor is applied to the induced equal to the corresponding inductance. Hence,
direct component of current in the field, and to all other 1
phenomena which depend on that current in direct 1' = [(XD' + XQ') + (XD'- XQ') cos 2 -y] (3)
proportion, such as the transient alternating component
of current in the armature, and the armature circuit Let the transient armature current be due to a single-
voltages which are due to the latter. And likewise the phase short circuit occurring during rotation at uni-
armature decrement factor is applied to the d-c. form speed, and at the instant when the direct axis is
component in the armature and to all other phenomena displaced from the armature axis by the angle a.
which thus depend upon it, such as the transient a-c. Counting time from this instant, the angle -y becomes a
component in the field circuit, and the voltages whichfunction of time.
are due to the latter. For the derivation of decrement -y = t + a (4)
factors see Appendix H. Therefore, any transient current, i, which exists at
The armature decrement factor is, any time t, will produce the following flux linkages with
r the armature winding,
XVXDXQ (1I
cos 2 (t + a)]
2 i [(XD' + XQ') + (XD' -xQ')
-

and the field decrement,

XD + V XDI XQ' (5)
afX= VD' + + XDD XQ
XQ' (2) The current existing in the field winding before
short circuit, would produce in the open circuited
I

where o- is the decrement factor of the field corre-
sponding to total field self-inductance.
Another item for consideration is the voltage across
an external inductance in either the armature or field
circuit. This is found by differentiating the particular
current with respect to time, and taking the product
of this by the external inductance coefficient. These
voltages are derived in Appendixes D and E.
Still another is the voltage across the open phase of a
three-phase machine during short circuit of one of the
phases. This is determined by first obtaining an
armature winding at any time t the alternating mutual
flux linkages,
= k e0 cos (t + a) (6)
where eo represents the peak voltage, line-to-neutral,
generated before short circuit, and therefore also the flux
linkages; and since on open circuit, terminal voltage is
also nominal voltage, eo also represents the correspond-
ing field current. k is a factor whose value depends
upon the nature of the short circuit, i. e., whether line-
to-neutral or line-to-line. Thus, k 1.0 for line-to-
=

expression for the linkages with the open phase in neutral, and 1.73 for line-to-line.
question, and differentiating this with respect to time. The sum of (5) and (6), must equal the linkages
This is given in Appendix F. existing at the instant before short circuit, namely,
Turning now to the details of the fundamental ¢t'm = k e0 cos ae (7)
equations, refer to Fig. lA. If a transient current i is The above relations determine the total armature cur-
caused to flow in the armature winding in the position rent, as derived in Appendix B.
shown, that is, with the direct axis and the armature The total field currents are given by
winding axis coinciding, a current will, of course, appear Id-e O+^e
in the field winding which is short-circuited either at Id=e e Se
the terminals or through an external inductance, and I,= eql = A\ eq (8)
Thus, fluxes will be produced in the leakage paths of where ed represents direct nominal voltage, and therefore
 464 DOHERTY AND NICKLE: SYNCHRONOUS MACHINES-IV Transactions A. I. E. E.
also the total field current in the direct axis winding, and XD' = 0.4
A ed the increment induced by the transient armature XQI = 1.0
current. Similarly, A eq represents the increment b 0.23
induced in the quadrature axis winding. It is assumed
inducedhinothe quadrature axismwinding.Itis assdumed current In this machine, the 3rd harmonic in the short-circuit
tharet.Thefonlyquadrurercomponent thrive
is peindud C would be 23 per cent of the fundamental, the
5th 23 per cent of the 3rd, the 7th 23 per cent of the
RESULTS 5th, and so on. Likewise for the even harmonics. The
The foremost result is the method of analysis which total wave, both calculated and tested, is shown in
makes possible, in a practical way, the calculation of Fig. 5 for armature current, and Fig. 6 for field current.
transient phenomena in any synchronous machine, Asan extreme case
regardless of whether it is of the salient pole or the XD' = 0.21
cylindrical rotor type. The equations thus derived XQ' = 3.77
have made it possible to calculate complicated transient b = 0.618
characteristics which check, in minute details, with the These constants apply to an actual induction machine
oscillographic records of tests; and these equations have with a coil wound secondary. t The armature is three-
revealed a number of important and very interesting
facts relating to the form of the current waves, the phase, the rotor two-phase; the latter, in this particular
resultin posblt of reaivl sipeaayssoh case, being connected two phases in series with direct-
waves, and the induced voltage in the open phase during current excitation. The machine is rated 20-hp., 1800-
A summary of equatontrv.60cyle mmn., 6-yl,10vl.Frcnezne
single-phase short circuit. Asummaryofequation re.per
11-ot.o
sisngiven-nAphasendhortcircuixtA. the per-unit quantities are
cneine
based on 10-kv-a., three-
phase, 57.7-volt line-to-neutral.
Referring to the short-circuit current waves, the With the rotor winding thus connected, the harmonics
current is expressed as the sum of two harmonic series, are relatively 2.67 times as large as in the previous case
one of odd, the other of even, harmonics; and the of the 7000-ky-a, salient-pole machine. The waves of
magnitudes of the harmonics in each series, in ascending both calculated and test, are shown in Figs.
order, are in geometrtc ratio. Thus, the single-phase current, 11-13, for the armature current, Figs. 16-18 for the
short-circuit current of the armature is field current.
k e, cosa [ + 2
XD=
V
, 2 + XDIt is interesting to note that if b = 0, that is if
XDX changes from the peaked
= XQ', all harmonics
disappear, and the current
+ 2 b2 cos 4 (t + ae) + 2 b3 cos 6 (t + az) +
wave wave in Fig. 13 to
.. . ] approximately a sine wave, as in Fig. 14. The values of
2 k eo XD' and XQ' were made practically equal by short cir-
D+ V XQ [cos (t + a) + b cos 3 (t + a) cuiting one of the two rotor phases, and exciting the
other, keeping the external impedance of the latter as
+ b2 os 5 (t + a) + b3 cos 7 (t + a) + ] (9) low as possible.
The geometric ratio b is given by Going a step further, XD' can be made greater than xQ'
-,- ,- by placing external inductance in the field winding
b = v XQ' v XD which is excited, i. e., the direct axis winding. In this
V,/ XQ' V XD'
+ case the sign of the harmonics is reversed, i. e., b becomes
. . negative,
where XQ' IS the
where x
transient reactance In the quadrature in Fig. 15. thus giving a dip in the wave. This is shown
the transientreactanceinthequadratur
axis, and XD' the corresponding quantity in the direct Referring again to equation (9), it is interesting to
axis. consider how limiting values of b modify the shape of the
For instance, a 7000-k-va., 375-rev. per min., 25- current waves. The first series, i. e., even harmonics,
cycle, 2200-volt salient-pole machine of usual construc- has the general form shown in Fig. 9A. The second
tion has the following constants.* series has the general form shown in Fig. 9B. The total
*It should be remembered that these constants represent the current, the sum of the two series, has the general
static transient linkages (or, at normal frequency, the correspond- form of Fig. 9c.
ing reactive voltage) as a fraction of normal linkages, line-to Now if XD' and XQ' are equal, b becomes zero, and all
neutral, (or, at normal frequency, the line-ta-neutral normal
voltage). The values given look high compared with the usual tThis machine was chosen with the two points in mind; one
three-phase values, because they are line-to-line constants. If was that our premises should be closely approximated; the other,
the line-to-line voltages were taken as the per-unit base, instead that it was desirable, in the interest of an exacting test of the
of line-to-neutral, the values would be divided by W J. Also, theory, to employ extreme limits in reactance coefficients thus
the corresponding line-to-neutral, three-phase values would be involving harmonics of large magnitude, which harmonics,
just one-half the values given. See companion paper by Park nevertheless, would be capable of control by different connections
and Robertson regarding the relations between three-phase and of the windings, and by variation of .constants in the external
single-phase constants. circuits.
Feb. 1928 DOHERTY AND NICKLE: SYNCHRONOUS MACHINES-IV 465

_~
U~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-3l

-2

FIG. 5-SIXG.LE-PIIASE LINE TO LINE StHI)I (CEcuITr AT a = 55) I)F(;., 7000-KV-A. IACHINE
Equation:--
A-Test cuive of armature current. 0.59 0.3:5s Cos (I + .5')
B-Test curve of open-phase voltage 1.4 - (i. Cos (2 t + 1100) 1.4 - 0.6 cos (2 t + 1100)
C-Calculated cur ve of armature current with test points inilicatilug
the peaks 0._41 cos (I +550) .0M12-7
1.4 - O.ti cos (2 t + 110°)

-S

140
/5 ~~~~~~~~~~~~~Curve
D

FIG. 6-SINGLE-PHASEF LINE-To-LINE, SHORT CIRCUIT AT a =55 DEG., 7000-Ky-A. MACIIINE

A-Test curve of field Current
B-Test curve of open-phase voltage 0.624 [3~.24 + 1.24 cos (2 t + 1101)]
C-Test curve of voltage across external reactor in field circuit + -1.4 - 0.6 cos (2 t + 1100)
Dl-Calculated curve of fleld current with test points indicating thle
peaks
Equation:- 1.109 13.24 + 1.24 cos (2 1 + I1010)l
1.4
0.364 cos (1 + 550)
- 0.6 cos (2 t + 1100)
+
.01691.4 - 0.6 cos (2 t + 1 100)
466 DOHERTY AND NICKLE: SYNCHRONOUS MACHINES-IV Transactions A. I. E. E.

harmonics vanish, as already indicated. Taking a = 0,
the first series then reduces to only a d- c. component of
an amplitude
k eo
o= V XD' (10)
XQ
And the second series then contains only a fundamental
term of amplitude
2k eo
tl = - (11)
XD' + / XD XQ
However, if XQ' is increased until it is very large
compared with XD', b will approach unity-that is, all
harmonics will approach equality-and the three waves
shown in Fig. 9, will approach respectively those shown
in Fig. 10. The value of b for the 20-hp. induction
machine is very high compared with usual machines,
but its value of 0.62 is considerably below unity. Yet,
the current wave in Fig. 11 indicates a definite approach
to Fig. lOc. In the first four cycles, that is, up to
t = 8 7r, the peaks of the component corresponding to
Fig. 10A slightly predominate over those of the com-
ponent in Fig. 1OB. But the former, being subject
to a more rapid decrement, become less after the fourth
cycle, and the little resultant "tip" reverses direction.
Note that the calculated wave in Fig. 11 checks the test
perfectly in this detail.
Another interesting and useful feature of the expres-
sions for short-circuit currents is that they open up new
possibilities of analyzing the current waves. In addi-
tion to the relations shown in Fig. 9c and Fig. lOc, and
the expression for the fundamental component of
current in equation (10), it is shown in Appendix B
that the r. m. s. value of the first series, i. e., even har-
monics, of equation (9), is
k eo cos a
/XD' XQ' N 1-b
1 + b2
(12)
Vnd harmonics
and of the odd harmonics
oftheodd
mental. Hence this might be a useful approximation.
Another possibility of convenient analysis of the
oscillograph current wave is afforded by the fact that,
with the harmonic ratio b, if one harmonic in each series
is determined, all of them are known. The convenient
relations exist that the peak value of the odd harmonic
series is 1/(1 - b) times the peak value of the funda-
mental component, and the peak value of the even
harmonic series is (1 + b)/(1 - b) times the direct
current component. It is shown in Appendix B that
since the three quantities-fundamental current i1,
direct-current component io,and the harmonic ratio b-
completely determine the wave, it can be analyzed by
scaling off three current values from the oscillogram at
(t + a) = 45 deg., 90 deg., and 180 deg.
So much for the results regarding short-circuit
current. Turning now to the next major point in the
results, the equations show that under single-phase
short circuit the peak voltage across the open phase may
be extremely high; many times normal in cylindrical
laminated rotor machines without amortisseur winding.
This voltage depends upon a very simple relation,
namely the ratio of XQ' to XD'. Thus, for a short circuit
at a = 0-that is, at maximum flux enclosure or
maximum flux linkages --the ratio of the voltage across
the open phase after short circuit (initial value) to the
peak voltage before short circuit is
-
PO = ( 2 - 1 ) sin: (14)
x XD
and for short circuit at zero flux linkages, i. e., a = 7r/2.
'
p 7r/2 = Q, sin (15)
XD
where d is the angle between the axis of the short-
circuited winding, and the axis of the open-circuited
winding considered.
Thus, as an extreme case, consider the

chine. Here the short circuit was line-to-line, and the

20-hp. ma-

_ 2 k eo 1 voltage recorded was across the open phase line-to-

ib
-XD'IZ +-\V XD'I X;
A/ ZD' XQ' V 1-b2(13)
A/ 1 2
-flux
neutral. Thus A = 90 deg. The ratio for maximum
linkages (a = 0), therefore, is,
Since equation (11) gives the amplitude of the funda-
mental term, it follows that the first fraction in (13) p = 2 -1 = 34.9
is the r. m. s. of the fundamental. Hence the r. m. s. 0.21
of the whole odd harmonic series is different from the Reference to Fig. 19 will show that this initial value, of
r. m. s. of the fundamental by the factor over 34 times the voltage before short circuit, was
1 obtained. The transient, of course, reduced the value
V 1-b2 of the first peak to obtain 28 times the value existing
This is usually*not significantly different from unity, before short circuit, which value is also shown on the
and hence an ammeter reading will be approximately calculated wave.
the fundamental. For instance, b = 0.5 would be an When XD' = XQ', the maximum voltage for the case
extremely high value for any commercial synchronous considered iS unity.
machine. The corresponding value of the factor isl1.15. Turning to the test on the 7000-ky-a, salient-pole
However, a representative value of b might be 0.2, for machine, the short circuit was line-to-line, and the open-
which the corresponding ratio is 1.02; that is, the phase voltage recorded was also line-to-line. Thus
ammeter would read 2 per cent more than the funda- ,B = 60 deg., and from (14),
Feb. 1928 DOHERTY AND NICKLE: SYNCHRONOUS MACHINES-IV 467

p = 3.46 for a = 0
and p = 2.16 for a = 7w/2
These are limiting values. In the test, the short circuit
occurred at a = 2.2 (i. e., about 125 deg.) and therefore
the peak should fall between those indicated by the
above values of p. Using a = 125 deg. (or, for con-
venience, the corresponding value, 55 deg.) in the
equation of Appendix F, the calculated wave Fig. 7
phase short circuit may rise to a peak value many times
the value existing before short circuit. It varies almost
directly with the ratio XQ' to XD'. Hence, it would be
extremely high-possibly 15 or 20 times normal voltage
-on a cylindrical, laminated rotor machine without a
squirrel-cage or metal wedges, serving the same pur-
pose; but is practically no greater than normal when
XQ' = X'D, as in a machine which has an amortisseur
winding or massive steel rotor. Most of the cylindri-

~0

2
3
FIG. 7-CALCULATED CURVE OF TRANSIENT OPEN-PHASE VOLTAGE, SINGLE-PHASE. LINE-TO-LINE, SHORT CIRCUIT AT
a = 55 DEG., 7000-KV-A. MACHINE. (FOR TEST RESULTS REFER TO FIG. 5B.)

0.746 sin (2 t + 170) cos (t + 55)

0.624 sin (t + 1150) +
Equation:- 1.4 -0.6 cos (2 t + 1100)
[ 10.35 -0.3 cos (2 t + 1700)] sin (2 t + 1100)
e = 2.38 _ 1.4 -0.6 cos (2 t+ 110o)2 1 0.35 -0.3 cos (2 t + 170°) ] sin (t + 550)
L 1406s +10-121.4 -0.6 cos (2 t + 1100)
1.194 sin (2 t + 1700) 1
1.4 - 0.6 cos (2 1 + 110) J e-0 01l68t + (1 + 1.77 e-0-00272t) 1.49 [0.35 - 0.3 cos (2 t + 1700) 1 sin (2 t + 110°) cos (t + 550)
[ 1.4 - 0.6 cos (2 t + 110°) ]2

/6
FIG. 8-TRANSIENT VOLTAGE ACROSS AN EXTERNAL REACTOR IN THE FIELD CIRCUIT, SINGLE-PHASE LINE-TO-LINE
SHORT CIRCUIT AT a = 55 DEG., 7000-KV-A. MACHINE. (FOR TEST RESULTS REFER TO FIG. 6C.)
Equation:- 0.75 sin (2 t + 1100) [3.24 + 1.24 cos (2 t + 1100)1
{ 0.0425 cos (t + 550) + 3.64 sin (t + 550) + [1.4 - 0.6 cos (2 t + 110o)]2
1.4 -0.6 cos (2 t + 1100)
2.75'sin (2 t + 110°)
4.37 cos (t + 550) sin (2 t + 1100) -6f 2.7 csi (2 ±t 1100)
± [1.4 - 0.6 cos (21 + 1100)]2 J 1

r
1.55 sin (2 t + 1100)
1-4 -0.6cos (2 t + 1100)
was obtained. The caleulated first peak is 2.73 times
the peak voltage before short circuit, and the corre-
sponding test value, 2.97. The calculated value is thus
less than the test by 8 per cent, which is a reasonable
check, considering the probable error in scaling the
angle from the oscillogram, and that due to the simpli-
fying assumption made in Appendix F. And it will be
noted that the shape of the curve checks satisfactorily,
Thus, the voltage across the open phase during single-
1.33 sin (2 t + 1100) [3.24 + 1.24 cos (2 t + 1100)]
+ [1.4 - 0.6 cos (2 1 + 1100)12 eJ0.00272t
cal rotors in usual practise have one or the other of
these features. In salient-pole machines, the ten-
dency toward this high voltage is less, because XD 'is
relatively higher, and XQ', on account of salient poles, is
relatively lower than in cylindrical rotor machines; but
as shown in the foregoing actual illustration, the peak
voltage, even in this case, may be from 2 to 3.5 times
normal. A very important point to observe in this
connection is that this high voltage across the open
468 DOHERTY AND NICKLE: SYNCHRONOUS MACHINES-IV Transactions A. 1. E. E.

phase is due, not to high flux density, but rather to a
high rate of change, and hence saturation would not
limit it, unless by tending to equalize XD' and XQ'.
Boucherot2 called attention to the possibility of high
voltage across the open phase, and to the favorable
effect of an amortisseur winding in reducing it. The
value of the present treatment in this connection is that
the equations, in addition to being generally applicable,
show the very significant extent to which salient poles
reduce this voltage.
It may be of practical importance to know the value
of the fundamental component of the open phase
voltage. The ratio of this to the voltage existing across
the open phase before short circuit is given approxi-
mately by the simple form (see Appendix F),
pi = 1 +b-----
A careful inspection of each of the figures will clearly
show a satisfactory agreement between the calculated
and test curves, even with respect to the minute Cetails.
It will be noted that the decrements agree well.
The ordinates of the curves, it will be remembered,
are scaled as the ratio of the actual value of the quantity
to the value taken as the per-unit base. Ordinarily,
normal value is chosen as that base. In the present
case, however, the base for the 20-hp. machine was
taken as 10 kv-a., 57.7 volts line-to-neutral, thus mak-
ing the per-unit value of reactance numerically the same
as the value in ohms-merely as a matter of conveni-
=7

where b =
_V XQ- V/XD
7VX-XD',
e

N\/ XQ + AV/ XD 7Zt , 7JT4ff

For instance, in the case of the 20-hp. machine, Fig. A
19, pi would be 1.628; and in the salient-pole machine,
Fig. 7, for corresponding conditions, 1.23.
COMPARISON OF CALCULATED AND TEST RESULTS
Calculations and tests were made on the two machines
already referred to, namely, a 20-hp., 1800-rev. per r t 27T 37T
min., 60-cycle, 110-volt induction machine, and a
7000-kv-a., 375-rev. per min., 25-cycle, 2200-volt
salient-pole alternator. The four major per-unit coeffi-
cients (measured) for the 20-hp. machine, on the basis of b
10 kv-a., three-phase, 57.7 volts line-to-neutral, are:
XD = 3.77, XD' = 0.21, XQ = 3.77, XQ' = 3.77 2
The auxiliary coefficients are:
XAD = 3.64, XAD' = 0.076, XAQ = 3.64, XAQ' = 3.64 l
of = 0.132 0a = 0.0474
For the 7000-kv-a. alternator the corresponding
per-unit coefficients based on 7000 kv-a., three-phase,
1270 volts line-to-neutral, are: IT . 27 3 471
XD = 2.24, XD' = 0.40, XQ 1.00, XQ'
= = 1.00 C
XAD = 2.09, XAD' = 0.25, XAQ = 0.85, XAQ' = 0.85 FIG. 9
of= 0.00272 O-a 0.0117
=
ence. The normal values were taken as the basis for
A comparison of calculated and test results is shown the 7000-kv-a. machine. These points should be kept
in Figs. 11 to 25, inclusive, for the 20-hp. machine, and in in mind in reading the ordinates.
Figs. 5 to 8 for the 7000-kv-a. machine. Each illustra- In the case of the open-phase voltage during short
tion shows, circuit, Figs. 7 and 19, the ordinate gives the ratio of the
a. Calculated curve, from equations in Appendix A. voltage before short circuit to the voltage after short
b. Oscillograph record of test. circuit.
c. Equation, with numerical values substituted,* With reference to the 7000-kv-a. alternator, this
from which the calculated curve was derived, machine has no closed circuits on the rotor except the
The reason for making extensive tests on the induc- main field winding and the massive steel of the rotor
tion machine, excited by direct current as a synchronous spider. The effect of the latter is probably negligible.
machine, has been stated in the second footnote under The poles themselves are, of course, laminated. The
"Results." tests were taken at a relatively low voltage-at 300
*Tes eqaios ben.eue oumrclcluain volts; hence the premises were pretty closely approxi-
are in a somewhat different form from those in the summary of mated, anld consequently the calculated results agree
equations in Appendix A. closely with the tests.
Feb. 1928 DOHERTY AND NICKLE: SYNCHRONOUS MACHINES-IV 469

In the case of the voltage across the field terminals, These expressions are derived in the appendixes
i. e., across the external impedance, the value of this mentioned before, but for ease of reference, they are
impedance was not definitely known. Hence with the tabulated here. The order is as stated: First, the
estimated value, the general shape of the calculated general equation for the quantity is given. Equations
curve is all that may be expected to agree with the test. for special conditions follow readily from this. Thus, for
The calculated curve for this is shown in Fig. 8, the test the condition of maximum flux linkages at the instant of
curve in Fig. 6c. short circuit, a = 0; while for minimum flux linkages
ACKNOWLEDGMENTS a = 7r/2 and the armature decrement is absent.
Steady
The authors gratefully acknowledge the valuable ac = ir 2,state conditions are obtained by placing
and omitting the armature and field decre-
assistance of Mr. B. L. Robertson in the preparation of ments.
the paper, of Mr. C. C. Herskind in tests and calcu- ments.
lations, and of Messrs. W. G. Scarborough and R. Since the equations are somewhat long it is found
Hammar in the preparation of data and illustrations. advisable to put them in a simpler form. By inspec-
They also wish to acknowledge the helpful suggestions tion, it is observed that they contain small groups which
of Professor W. V. Lyon of M. I. T., and of Messrs. are more or less common to all of the expressions.
R. H. Park and P. L. Alger. These few groups, therefore, may be represented by a
Appendix
A single symbol, as follows:
APPendiX A
SUMMARY OF EQUATIONS A = XD' + XQ'
In the first part of the paper, under "Scope," it was
stated that the present treatment develops expression C = XD ± XQ
D XQ,
for, F x,- Q'
1. Armature Current-Appendix B. G = XD - XD'
2. Field Current-Appendix C. H = (XD' + XQ') + (XD'- XQ') cos 2 (t + a)
= A + B cos 2 (t + a)

r I l | J = (XD' + XQ') - (XD'- XQ') cos 2 t

. = A- Bcos2t

XDI + XD XQ
7lt27 37T XD + V/XD XQ
____ I K

47T
_____

N= |XD
1L Fl I AXQ/
I
| X I ~ ~ ~~~~~~~~P
XD- XDI
7t 3V 7T4XD+ XD XQ

r
V/XD'IXQ'

, r= K (general)
X
iff = C Q (for cylindrical, laminated rotor only)
V 1?
= X (If R and X are in ohms, 0 will be a numeric,
I t ef21T 377 47T in proper terms to use in the above
expressions).
FIG. 10
1. ARMATURE CURRENT
3. Voltage Across Reactance in Armature Circuit - Transient.
Appendix D. 2 k e0
4. Voltage Across Reactance in Field Circuit- i = H [e'a cos a-K cos (t +a)-P eof' cos (t +a)]
Appendix E.
5. Voltage Across Open Phase-Appendix F. For maximum flux linkages ae = 0 in the above
470 DOHERTY AND NICKLE: SYNCHRONOUS MACHINES-IV Transactions A. I. E. E.

equation, but for minimum flux linkages a = 7r/2 and 5. VOLTAGE ACROSS THE OPEN PHASE
the armature decrement does not appear. Thus, for the 2 k ki B eo
latter case, Transient. e1 - H2 cos a [H cos 2 (t -+ a)
2 k eo
i= skin t [K +
snt[ -t
P fft]+ B sin2 2 (t + a) ] E--Ot
k, eO
Sustained.2ke0Ksint
2ke i
-kH + kH2 [H2k- 2kBHcos2 (t + a)
Sustained. z=
-2 k B2 sin2 2 (t + a)] cos (t + a)
2. FIELD CURRENT + kB H sin(t+ a)sin2 (t+ a) } [K- (1-K)cE-Jft]
Transient. For maximum flux linkages a = 0, while for mini-
eo mum flux linkages a = 7r/2 and
I= H {[C+Fcos2(t+a)] [K+Pe'f'] kieo
e1 = 2 {[H2k + 2kBHcos2t + 2kB2sin22t)] sint
-2 G cos (t + a o eea
For maximum flux linkages a = 0; and for minimum + k B H cos t sin 2 t [(1 - K) e- ftK
linkages, k, eo K
eO Steady State. e- =- k H2 { kBHcostsin2t
I = {[C -F cos2 t] [K + P E-Of']}
+ [H2 k + 2k B H cos 2 t + 2kB2 sin22 t)] sin t
Sustained. I = K. (C - F cos 2 t) Appendix B
3. VOLTAGE ACROSS REACTOR IN ARMATURE CIRCUIT ARMATURE CURRENT
Refer to "Premises" and "Fundamental Equations"
eze H2 [2 B cos a sin 2 (t + a) e-aalin the main body of the paper. Equating (5) and (6)
and solving for i,
+ sin (t + a) [H-4B cos2 (t + a)] (K + P ef)] 2k eo[cosa- cos (t + a)]
For maximum linkages a = 0; and for minimum i = (XD'+XQ') ± (xD'-xQ')cos2(t4-a) (lb)
linkages,
2 e0,X
2k Resolved into the harmonics, this current becomes,
e= [cos t (J -4 B sin2 t) (K + P E"')] kefcos a
2kexi V= D',[1+2bcos2(t+a)
Sustained. exe =
2keoxeK XQ A/~~~~~~~XD'
COSt (J-4 B sin2 t) +2 b2cos 4(t + a) +2 b3c os 6 (t + a) + ..
4. VOLTAGE ACROSS REACTANCE IN FIELD CIRCUIT 2k , [cos (t + a) + b cos 3 (t + a)
2GEoX,cosasin(t+a) [H-4Beos2 XD' + V/XD XQ'
Exe= H2 (t+a)] E-0af -a) +
+ b2 cos 5 (t + .] (2b)

2PEoXesin 2 (t + a) where b= (3b)

+ H2 (B C- A F) e'ft VXQ + v7
XD
k = a factor depending upon the phase which is short
2 K Eo Xe sin 2 (t + a) circuited. It is equal to 1.0 for line-to-neutral,
H2 and 3 for line-to-line. V/
For maximum flux linkages co = 0; while for mini- eo = peak voltage, line-to-neutral, before short circuit.
mum linkages a = 7r/2 and When the short circuit occurs at maximum flux
linkages, that is at a = 0, (Ib) becomes,
Exe 2PEoXe J2 sin 2 t(A F-B C) e~' .- 2 keO (1-cos t) (4b)

2 KEo Xe = ° (XD'+ XQ') + (XD-XQ') COS2 t

+ J2 sin 2 t (A F -B C) and (2b) becomes,
2KEoXe i= -,ke
, [1 +2bcos2t +2b2cos4t
Sustained. Exe = J2 sin 2 t(A F-BC) ax=0 V/XD' XQ'
+ 2b3cos6t + . ...]
Feb. 1928 DOHERTY AND NICKLE: SY1NCHRONOUS MACHINES-IV 471

2keo + b2sin5t-b31sin7t + . .1 (7b)

XD, + XD XQb2COS5t It is interesting to note that from equation (2b),
.
ZDb XXthe reactance which, with the voltage k eo, determines
(5b) the fundamental component of short-circuit current is
When the short circuit occurs at zero flux linkages, XD' + / ZD' XQ'
that is at a = 2- (lb) becomes, 2
and that which, with the same voltage, determines

A - AA A . A,AA .A- .
2T 47T 67T 77 /0? Curve A

FIG. 11-TRANSIENT ARMATURE CURRENT, MAXIIMUM INITIAL ARMATURE LINKAGES ON

SINGLE-PHASE LINE-TO-LINF SHORT-CIRCUIT, 20-Hp. MACHINE
A-Calculated curve
B-Test curve
Equation:- 2.64 cos It 0.132t
3.4fi
i
3.98 -3.56 cos 2 t
e-0.0474- 0.818 Cos t 3.98 - 3.56 cos 2 1
3.98 -3.56 cos 2/

8.V ~~~~~~~~~~~~Curve
A

FIG. 12-TRANSIENT ARMATURE CURRENT, ZERO ARMIATURE LiNKAGES ON SINGLE-PHASE

LINE-TO-LINE SHORT-CIRCUIT 20-Hp. MACHINE
A-Calculated curve
B-Test curve
Equation:-
0.818 sin t 2.64 sin t
3.98 + 3.56 cos 2 t 3.98 + 3.56 cos 2 t e

2 k eO sin t the direct-current component is \/ XD'X Q'

ir/2 + XQ') - (XD' - ,Q') cos 2 t (6b)
a =r/2 (xD + xQ -(xD'-XQ') cos 2 t The peak value of the fundamental is
and (2b) becomes, 2keo (8b)
2 k eO XD- ' + / XD ZQ.
i = ± Q and the direct-currt c[sin
to-ibsin 3 t
a =7r/2 XD I+ \/XD' XQ' and the direct-current component is,
472 D)OHEIRTY AND N'ICKLE: SYNCHRONOUS MACHINES-IV Transactions A. 1. E. E.

ke, cos a x/2Ikeo 1

io= V-XD'XI
A/XD XQ
The effective value of the series in equation (2b) will
now be derived. This value is the square root of the
sum of the squares of the individual effective values.
The coefficient of each harmonic term represents the
peak value. Thus, dividing each harmonic term by
(9b) or i. = (llb)
XD' + \VXD' XQ' v 1 - b2
Equations (lOb) and (llb) give the effective value of
the respective series as a fraction of the peak value of
normal fundamental current. To place the results in
terms of the effective value of normal current, multiply
each equation by N' 2.
Turn now to the relations which make it possible to
analyze a single-phase short-circuit current wave from
only three values taken from the wave. Referring to
equation (2b), the sum of each series is a maximum
(but the total is zero) when
(t + a) = 2 7r
k eo cos a

V\XDIVTX_'XQ/[
Lt A\iCarve \g V i = +/, , ~~[1 + 2 b + 2 b2 + 2 b3 + 2 b4 + ...
2 ek [1 + b + b2 + b3 + .] (1b

XD + XQ' . (12b)
Taking the summation of these series,
ke,,cosa 1 + b 2keo 1
vXD'XQ' 1 -b XD + \/XD XQ 1- b
(13b)
Substituting (8b) and (9b) in (13b), the peak value of

FIG. 13-SUSTAINED ARMATURE CURRENT ON SINGLE-PHASE I

LINE-To-LINE SHORT-CIRCUIT 20-Hp. MACHINE I
A-Calculated curve Equation:- 0.818 sin I
B-Test curve i 3.98 + 3.56 cos 2 t

FIG. 15-SUSTAINED ARMATURE CURRENT SINGLE-PHASE LINE-

TO-LINE SHORT CIRCUTIT WITH ZD' > xZ' 20-Hp. MACHINE

the odd harmonic series is thus 1/'(1 - b) times the peak

value of the fundamental component i,, and the peak
value of the even harmonic series is (1 + b)j(1 - b)
times the direct-current component i0. Thus, for the
sum of the odd harmonics, the peak value is
FIG. 14-SUSTAINED ARMATURE CURRENT, SINGLE-PHASE . 1 b 14b
LINE-TO-LINE SHORT CIRCUJIT WITH XD' = xQ' 20-Hp. MACHINE odd = i1/( - ) (1)
Similarly, the peak value of the even harmonic series is
-Vo2 and squaring, and taking the square root of the ieD = io, [(1 + b)/(1 - b) ] (15b)
sum, the effective value of the even harmonic series is Substituting for b its value from (3b), some additional
k eo cos a ____________________________ interesting relations are incidentally obtained. Thus
ff O v 1 + 2 b2 + 2 b4+ 2 b6+ Q,+'/XXD
VXD IXQ = (16b)
1 -b 2 V/XD'
keocosa 1 +b2
or i,.,, = V XD' XQ 1 1-b2 4/(1Ob) and 1±+b
1-b _ Q_
XD'b)

And for the odd harmonic series, Substituting (8b) and (16b) in (14b),

tellt = XD' +V~/

V2e A/1 + b2 + b4 + b6 + i = lceo (18b)
\/XD'
±D V XQI'X
Feeb. 1928 DOHERTY AND NICKLE: SYNCHRONOUS MACHINES-IV 473

and (9b) and (17b) in (15b), b = 0.333 io = 1.0 i, = 2.667

k eo cos a (19b) Hence, by (2b) the wave is
1b
XDI
2
The relations are shown graphically in Fig. 9. i = 1 + cos 2 (t + a) ± + cos 4 (t + a)
Consider now the value of the sum of the two series
in (2b) at three different values of t + a, namely,
2
- '
ir Ir7+-o
2- and r. First, when t + a = -,or45deg.
6( + cos6(t+) +
a

i45= 1b - 1t1 + bb (20b) 1 cos 3 (t + a)

-2.667 [cos (t + a) + L3-o3ta
1+ b2

'50

to JlK 21 L4 Curve A

FIG. 16-TRANSIENT FIELD CURRENT, MAXIMUM INITIAL ARAMATURE LINKAGES ON

SINGLE-PHASE LINE-TO-LINE SHORT-CIRCUIT 20-Hp. MACHINE
A-Calculated curve Equation: - 12.3 cost 3.08 9.97
B-Test Curve I 3-Tes.nCurve8I=
- 366 cos
_C74t 3-98 - 3.56 cos 2 t + 3.98 - 3.56 cos 2 t e-o32

when t + a 2 or 90 deg., + cos 5 (t + a)+.

1 -b When the wave has no d-c. component io, the analysis

i90 = io 1 + b (21bj is much simpler, since only two values from the wave
are now needed; namely, the value at 45 deg. and at
and when t + a = 7r, or 180 deg., 90 Transient
deg. Armature Currents. The armature current
1+b 1 immediately after short circuit is given by equation
_ + 1-b
1i80 = o11 - (22b) (lb). The term which disappears when a = r/2
-b + 1-b
contains the direct current component and the even
From these three equations, the entire wave can be harmonics. As time elapses, these currents decay to
analyzed. For illustration, let i45 = - 0.332, iDo = 0.5, zero according to the armature decrement factor, a.,
and i180 = 6.0. The three equations thus determine derived in Appendix H. The other term contains the
b, io, and il. They are odd harmonics which are proportional to the field
474 DOHERTY AND NICKLE: SYNCHRONOUS-MACHINES-IV Transactions A. I. E. E.

current at every instant. Hence this component of cos a

equation (lb) decays from the value shown to a (XD' +
sustained value
XQ') (XD'- xQ') cos 2 (t + a)
+

XD' + V XD IXQcoa
<XZD'XQ'
Zr, ++ V
XD XD ZQ' Xa
C°S X
/ XDI ZQ
[even series] (25b)
times as great, according to the field decrement factor,
of, derived in Appendix H. and,
The sustained armature current is thus cos (t + a)
i = - 2k eO [ X + ]XDXQ 1 (XD' + XQ') + (ZXD - XQ') COS (t + a)
sustained ZD ± V ZD ZQ
cos (t + a) , , [odd series] (26b)
(XD' + XQ') ± (XD' - xQ') cos 2 (t + a) (2bZr'+Vr,Q
The component which decays to zero according to the expression for armature current at any instant
the field decrement factor is then the difference between after short circuit is
30 .
I
20
AAAAAAAAAA~~urveAAAAAAA
I0kkA
0 2T
7 UI77....M 0

FIG. 17--TRANSIENT FIELD CURRENT, ZERO INITIAL AfRtMATURE LINKAGES ON SINGLE-PHASE

LINE-TO-LINE SHORT-CI[ICUIT 20-HP. MACHINE
A-Calculated curve
1B-Test curve (not quite zero flux linkages)
Equation:-
3.08 9.97 e-0 132t
3 98 + 3.56 cos 2 1 + .98 + 3.56 cos 2 1

the second term of equation (lb) and equation (23b). 2 k eO cos a

The total armature current at any instant after short i = -x, [even series] e-at
circuit is then V XD ZQ
2 k eo cos a 2 k eo (XD-XD [odd series] eft)
(ZD' + ZQ') + (ZD' - ZQ') cos 2 (t + a) (ZD + VXD' ZQ') (XD' + VZD ZQ')
2 k eO (ZD - ZD') COS (t + a) -2 k eO [odd series] (27b)
(ZD + V ZD'ZQ') [(ZD' + X) Z V ZD ZQ
+ (ZD' - xQ') Cos 2 (t + a) ]Appendi C
ZD + V ZD ZQ FIELD CURRENT
-k XD + \/ XD'XQ' The equation for the initial field current will be
obtained in accordance with the plan outlined under
cos (t + a) "Fundamental Equations."
(XDr' + XQ') + (XD' XQ') cos 2 (t + a)
- (24b) It is convenient to take the three-phase normal
values of current and armature reaction as the bases
If we substitute from equation (lb) and (2b) the of the per-unit quantities, both three-phase and single-
relations phase. Thus a three-phase per-unit current i will
 Feb. 1928 DOHERTY AND NICKLE: SYNCHRONOUS MACHINES-IV 475
produce a rotating space fundamental m. m. f. whose the direct axis is displaced from the axis of the single-
amplitude, expressed as a fraction of the amplitude phase winding by an angle a, the component of A1
corresponding to normal current, is numerically equal which remains in the direct axis is
to the current i. The single-phase m. m. f. and current AD = 2/3 k i cos (t + a) (2c)
will be expressed in terms of these. and in the quadrature axis
It will be recalled that the amplitude of the syn- AQ = 2/3 k i sin (t + a) (3c)
chronously rotating component of space fundamental
m. m. f. due to normal frequency current in one phase It has been shown in Synchronous Machines-III
of a three-phase machine, is one-half of the amplitude that unit m. m. f-, in suddenly appearing over the direct
of the total pulsating space fundamental produced by axis, will induce a per-unit field current
that current, the other half rotating oppositely; and A Id = Xd - Xd' (4c)
that the corresponding synchronously rotating com- or, since per-unit field current is numerically equal to
ponents of the other two phases happen to be in space per-unit nominal voltage,
phase with that of the first. That is, the amplitude of A Id = A ed
the total three-phase space fundamental m. m. f., i. e., Thus, a sudden m. m. f. Ad will induce a field current
the m. m. f. of armature reaction, is 1.5 times the total given by
amplitude Al of the pulsating space fundamental due Aged Ad (Xd-Xd') (5C)
to one phase only. Or, conversely, the pulsating space A e =e-pAdi(el the (tn )
fundamental due to a line-to-neutral single-phase cur- where Xd and Xd' are respectively the line-to-neutral,
rent is 2/3 of the m. m. f. of armature reaction A three-phase values of synchronous and transient per-
funit reactances corresponding to the direct axis. (The
81 negative sign is taken in accordance with the convention
eA that the space fundamental m. m. f. due to the armature
.4 is considered to be in the same direction as the field
/2 \__________
m. m. f. when the two axes coincide).
Similarly, for the quadrature axis,
PT 4T 67r Ae = -A,, (X -Xq,') (6c)
CUrVe A
Since the quantities Ad and Aq in equations (5c) and
(6c) each represent the amplitudes of a space funda-
mental m. m. f., expressed as a fraction of the amplitude
of the space fundamental m. m. f., due to normal three-
phase current, it follows that the amplitude of any
space fundamental m. m. f., so expressed, such as AD
or AQ, may be directly substituted. Hence, from
(2c) and (3c),
FIG. 18--SUSTAINED FIELD CU-RIRENT ON SINGLE-PHASE LiNE- 2
TO-LINE SHOItT CIRCUIT 20-Hp. MACHINE A ed = - - k i (Xd- Xd') COS (t + a) (7c)
A-Calculated curve
B-Test curve 2
Equation:- and A e, =- - k i (xe- xe') cos (t + a) (8c)
3.08
3.98 + 3.56 cos 2 1 If the main field winding is the only rotor winding
in the direct axis, the field current A Id, or A ed, will
due to a three-phase current of the same value. Since exist entirely in the winding. However, if there are
the space fundamental A1 at any instant is independent other short-circuited rotor windings in the direct axis,
of the nature of the variation of the current i which only a part of A ed will appear in the main field windings,
exists at that instant, it follows that, in general, any the remainder appearing in the other closed rotor
single-phase current i will produce a space fundamental circuits. For the case of a single rotor winding, the
A1 = 2/3 k A = 2/3 k i (Ic) field current A Id which is produced by the armature
where k = unity for line-to-neutral current, and 1.73 m. m. f. due to the armature current i, is obtained by
for a line-to-line current. And obviously Al will vary substituting in (7c) the value of i given by equation
in time according as i varies. (lb) in Appendix B. Thus,
When the rotor is moving, the space fundamental A Id =
m. m. f. A1 will have relative motion with respect to the 4
rotor poles, but at any instant it may be resolved into - - k2 eo (Xd - Xd') cos (t+ ±a) [cos a - cos (t + a) ]
the two conventional components; AD in the direct
axis, and AQ in the quadrature axis. If the single-phase (XD' + XQ') + (XD' -xQ') cos 2 (t + a)
armature current is due to a short circuit at the instant (9c)
476 DOHERTY AND NICKLE: SYNCHRONOUS MACHINES-IV Transactions A. 1. E. E.

It is desirable to express (Xd - Xd') in terms of voltage immediately after short circuit, that is the
XD- XD'). initial value, is
Referring to the definitions and equations in Id = (Io + A Id) = ed= (eo + A ed) (12c)
Appendix I, eO (XD + XQ') + (XD -XQ') cos 2 (t + a)
3 - 2 (XD - XD') cos a cos (t + a)
2 k2 (XD' + XQ') + (XD' XQ') cos 2 (t + a) -

Substituting this in (9c), (13c)

A Id = A ed When the harmonic series form for i from equation
-2 eO (XD - XD') cos (t + a) [cos a - cos (t + a) (2b), Appendix B, is substituted in (7c) and remember-
(XD' + XQ') + (XD' XQ')
- cos 2 (t + a) ing the relation in (lOc),
(llc) Id =ed = eO (XD + V/ XD' XQ')
The total per-unit field current, or per-unit nominal XD' + \ XD' XQ'
M bp
20

/G
/0
7 3 jI
_ 7A Curve A A
157
AA
e0 ¢
A
v-A
A

FIG. 19--TRANSIENT OPEN-PHASE VOLTAGE, MAXIMUM INITIAL ARMATURE LINKAGE ON

SItHGLE-PHASE LINE-TO-LINE SHORT-CIRCLIT 20-Hp. MACHINE
A-Calculated curve,
B-Test curve
Equation:-
(3:56 -3.9S cos 21) 1.78 Cos t -3.36 cos t sin2 t 5.9S cos t sin2 21
e - -7.1
7.12 "+0044
-t
(3.98 - 3.5icos 2) 3.98 -3.56 Cos 2 t (3.98 - 3.56 Cos 2 1)2
(5.75 cos t - In.9 cos I sin2 1) 0 12
(19.3 cost sin2 2!)
(3.98 -3.56 cos 2 1) ' e (3.98 - 3.56 cos 2 t)2

20[

AA , A
ss'A .. A A. A
r - 0iA f ioR77 _LVv
.
Tl8 - y v~5~V- V v V
2o Curve A

FIG. 20-TRANSIENT OPEN-PHASE VOLTAGE, ZERO INITIAL ARMATURE LINKAGES ON

SINGLE-PHASE LINE-TO-LINE SHOnT-CIItCUIT 20-Hp. MACHINE
A-Calculated curve 5.75 sin I - 10.9 sin I cos2 t
B-Test curve 3.98 + 3.56 cos 2 ! -0.1321
Equation: -
-1.78 .sin I - 3.36 sin t cos,2 t 5.98 si sin2 21 19.3 sin t sin2 2 1
3.98 + 3.56 cos 2 t (3.98 + 3.56 cos 2 tl3 3.98 + 3.56 cos 2t2 .1
Feb. 1928 DOHERTY AND NICKLE: SYNCHRONOUS MIACHINES-IV 477

eO (XD - XD') (1 + b) cos a portional to the average field current at every instant
- aX [cos (t + a) after short circuit, they will also decay to their sustained
A/ XD ZXQf values in the same manner. The odd harmonics in
+ b cos 3 (t + a) + b2 cos 5 (t + a) + . . ] the field current are proportional to the direct-current
eO (XD - XD) component in the armature at every instant and thus
+ ± IX,'(1 + b) [cos 2 (t + a) decay to zero according to the armature decrement
XD' + N/XD' XQ factor Ja- derived in Appendix H. In equation (13c),
+ b cos 4 (t + a) + b2 cos 6 (t + a) + . . .] (14c) the odd harmonics comprise those terms which disap-
The quadrature component A I, may be obtained in
a similar manner. pear when a = 2 i.e., when there is no direct
The field current eomprises a direct current com- 2
ponent, a series of odd harmonics, whose amplitudes are current component in the armature. The remaining
in geometric ratio, and a series of even harmonies whose terms comprise the average value and all even har-
5 monics. Thus
4 I - 2 eO (XD - XD') COS a COS (t + a)
23 A II W1 oId d (XD' + XQ') + (XD' - XQ') cos 2 (t + a)
(15c)
and
I II (XD - xQ') cos 2 (t + a)] (16c)
IIeO [(XD + XQ') +
eve (Xd' + XQ') + (XD' - XQ') cos 2 (t + aL)
.4 Cuirve A The sustained value of (15c) is zero and the sustained
value of (16c) bears the same relation to (16c) as the
final average value bears to the initial average value.
Hence the sustained field current is

Id = XD + v XD XQ
sustained XD + V XD' XQ'
e (XD + XQ) + (XD - XQ) cos 2 (t + a) (17c)
L (XD' + XQ') + (XD' XQ') cos 2 (t + a)
-

The differences between (16c) and (17c) thus decays

to zero according to the field decrement factor, and the
FIG. 21-St-STAINED OPEN-PIIASE VOLTAGE ON SINGLE-PIIASE expression for the total field current at any instant
L,INE-TO-IN,TE SHORT-CIRCuIT 20-Hp. AIACHINE after short circuit is
A-Calctlated curve
Bf-Test curve - 2 eO (XD - XD') cos a cos (t + a)
Equation:- Tecr=
Sin 1 Cos21
7t.,sin I - 3.36 2.9\slutsin22 (XD' + XQ) + (XD' - XQ') cos 2 (t + a)
e = 3.9s + 3.56 cos 2 t + [3.98 + 3.56 cos 2 1j2
XD- XD
+
amplitudes are in geometric ratio. The geometric XD + VA XD XQ'
ratio is the same in all cases, for the armature current ( + XQ'), + (XD - xQ') cos 2 (t + a)'
as well as for the field current, and its value is given by eo r XD + Q) ± (XD Q) ( + ) e')
equation (3b), Appendix B. L (XD/ + XQ) + (XD/- XQ)COS 2 (t + a)]
Transient Field Current. It will be noticed that the XD' + V XD' XQ'
average field current or nominal voltage increases at +
short circuit in the ratio XD + V XD/XQ/

XD + v XD' XQ' e (XD + XQ') + (XD - XQ') cos 2 (t + ) 1 (18c)

XD' + v XDI XQ eL (XD' + XQ') + (XD' - xQ') cos 2 (t + a) ]

As time elapses, this average field current will decay If we substitute the odd harmonic series from
to the value existing before short circuit according to equation (14c) for (15c), and the average value and
the field decrement factor af derived in Appendix H. even harmonics from equation (14c) for equation
Since all even harmonics in the field current are pro- (16c), equation (18c) may be written
478 DOHERTY AN\D NICKLE: SYNCHRONOUS MACHINES-IV Transactions A. I. E. E.

-eO (XD - XD') (1 + b) cos a Appendix D

I=ZD ZQI [odd series] %t VOLTAGE ACROSS EXTERNAL REACTANCE IN ARMATURE
CIRCU1T
(XD - XD )2 (1 + b)CIUT
eO
+ - [even series]rzf The reactances, XD' and xQ', may in general include
(XD+ V\XD' XQ') (XD + V\XD XQ') an external reactance xe. It is then of interest to find
e0 (XD - XD') the voltage which exists across this reactance under
+ XD' (19c) conditions of sudden short circuit. The per-unit
XZD + VA'XD'XQ' value of voltage will be given by the product of the
eO (XD - XD') (1 + b) external inductance and the rate of change of current
+ XD + XD' XQ'
; [even series] with respect to time. Since per-unit reactance at
normal frequency is numerically equal to per-unit
+ eo inductance, the voltage is,

40-

C0 \ #) - Ar F t J4 ,$4 .
10 2T 41T 6T 8oT1 2011
. ~~~~~~~~~~~CtrveA

FIG. 22-TRANSIENT VOLTAGE ACROSS AN EXTERNAL REACTOR IN THE ARMATURE CIRCUIT, MAXIMII M INITIAL
ARAIATITRE LINKAGES ON SiNGLE-PHASE LINE-TO-LINE SHORT-CIRCUIT 20-Ip. MACIIIXE
A-Calculated curve 7.12 cos I sin 2 t + (3.98 - 3.56 cos 2 t) sin t
B-Test curve + 0.815 (3.98 - 3.56 cos 2 t)2
Equation:-
0 e=4
- 14 e 0474t
e = - 3.98 -
24.6 sin 21t 0.0474t
~~~~+
7.12 cos t sin 2 t + (3.98 cos 2 t) (sin I + 0.132 cos t) 132t
3.98 - 3.56 cos 2 t 3.9828 - 2.64. 3.56 cos 2 )i) (.8-356CS2)

1. exe,\-A 4 -A

2* ~~~~~~~~CurveA

FIG. 23-TRANSIENT VOLTAGE ACROSS AN EXTERNAL REACTOR IN THE ARMATURE CIRCUIT, ZERO INITIAL ARMATURE
LINKAGES ON SINGLE-PHASE LINE-TO-LINE SHORT-CIRCUIT 20-Hp. MACHINE
A-Calculated curve
B-Test curve
Equation:- 7.12 sin 2 t sin I + (3.98 + 3.56 cos 2 t) (cos t - .132 sin t) 0132
7.12 sin 2 t sin t + (3.98 + 3.56 cos 2 1) cos t (3.98 + 3.56 cos 21)2
e - -0.815 (3.98 + 3.56 cos 2 1)2
Feb. 1928 DOHERTY AND NICKLE: SYNCHRONOUS MACHINES-IV 479

di The remainder of (2d) is that portion which contains

x6 d
eS6 = (1d)
the field decrement o-f, and which decays from the value
t shown to a sustained value
The initial value of short-circuit armature current is
given by equation (lb) of Appendix B. Hence, from
XD' + XD'IXQ -K
this expression, the voltage across the external reac- XD + A/ XD' Q'
tance, assuming no decrements in the current wave, times as great, according to this decrement. It is
will be divided into the alternating component which finally
2 k eO xe dies away, and the sustained value, in the same manner
e, = H2 [H sin (t + a) as its corresponding currents of equation (lb) were
formed into the second and third terms of equation
+ 2 B sin 2 (t + a) [cos a- cos (t + a)]] (2d) (24b).
The symbolic form is that outlined in Appendix A. As an alternative to decomposing (2d) into the above
mentioned components, the transient voltage across an
exe external reactance in the armature circuit may be more
5 easily obtained by differentiating (24b). Neglecting
5 V A\/ T A/ 41r
AL Li
V V 6T V we have,
L A ~~~~~~those
terms concerned with derivatives of the exponents

0 Curve A 10
= 2k e0Xe
H2 [ 2 B cos a sin 2 (t + a) e-at
+[H sin (t+a)-2 B cos (t+a) sin 2 (t+a)] (K+P e-fi]
(4d)
It is interesting to treat the initial voltage in series

20 Ex.
FIG. 24-SUSTAINED VOLTAGE ACROSS AN EXTEIjNAI, L E . A .
REACTOR IN TIIE AiMIATUItE CIICUIT ON SINGLE-PIIASE 10
f r461r
LINE-TO LINE SIIORT-CIRCUIT20-HP. MACHINE .20 Ct.rve A
A-Calculated curve
B-Test curve
Equation:-
[ -7.12 sin 2t.dIn t -(3.98+ 3. co2) cos 1
e = 0.815 (3.98 + 3.5G cos 2 t)2 J

Under the development of Transient Armature

Current of Appendix B, it was shown that part of the
current wave decays to zero according to the armature
decrement or., and that second part decays to zero
FIG. 25SuSTAINED VOLTAGE ACROSS AN EXTERNAL
a REACTOtUINTHE FIELD CIRCUIT ONSINGLE-PHASE LINE-To LINE
according to the field decrement o-f. The third part is SHORT-CIRCUIT 20 Hp. MACIIINE
the steady-state value. Thus, the above voltage A-Calculated curve
equation will have components which also decay to zero B-Test curve (inverted)
according to these two decrements. 21.9 sin 2 1
That part of equation (2d) which contains the arma- (3.98 + 3.56 cos 2 )2
ture decrement corresponds to the first term of equation
(24b) and is form and observe the shapes of the total and component
4 k eO xe B sin 2 (t + a) cos a waves. Thus, using the series expression for current as
H2 ea (3d) given by equation
reactance becomes,(2b), the voltage across the external
With reference to (24b), it will be noticed that a second k eo X. cos a
term, containing the derivatives of the exponential, e, =- [4bsin2(t+a)
should be introduced in order to be exact. However, A XDZ XQ
since this quantity is very small in comparison with + 8 b2 sin 4 (t + a) + 12 b3 sin 6 (t + a) + . . .1
(3d) and will never be noticeable in practical cases, it 2 k eo xe
is not considered here. That is, the voltage due to the + , [sin (t + a) + 3 b sin 3 (t + a)
rate of change of current which in turn is due to the XD' + V/XD' XQ'
armature decrement is neglected. + 5 b2 sin 5 (t + a) + . . . (5d)
480 DOHERTY AND NICKLE: SYNCHRONOUS MACHINES-1V Transactions A. I. E. E.
When a = 0, the first series will have the general C + F cos 2 (t + a)
form shown by Fig. 25 in the main body of the paper, Id = Eo K A +B cos 2 (t + a) ] (3e)
while the second series will have the form shown by sustained
Fig. 24. The total wave is given by Fig. 22. or Id = Eo + Eo P (1 + b) [cos 2 (t + a)
Appendix E sustained + b cos 4 (t + a) + . . .] (4e)
VOLTAGE ACROSS EXTERNAL REACTANCE IN FIELD The sustained voltage across the field reactance is
CIRCUIT thus
Under conditions of sudden short circuit, alternating
currents are produced in the field circuit. When this
EXe 2 Eo Xe K sin 2 (t + a) (B C- A F)
H2
(5e)
circuit contains some reactance external to the field or Ee -2 Eo P (1 + b) [ sin 2 (t + a)
winding proper, such as the exciter armature, alter-
nating voltages exist across the reactance. It may + 2 b sin 4 (t + a) + 3 b2 sin 6 (t + a) ± ] (6e)
sometimes be desirable to determine such voltages. Equation (le) can be decomposed into three com-
As before, for a pure reactance, the per-unit voltage ponents; that which dies away according to the arma-
is given by the product of the per-unit external induc- ture decrement (a, that which decays to zero according
tance (which is numerically equal to the per-unit to the field decrement of, and the sustained voltage.
reactance Xe), and the rate of change of current; in Since, however, the transient field current is known,
this case, field current. equation (18c), the voltage across the reactance can
In evaluating this voltage, it is convenient to express more easily be obtained by the product of the derivative
the external reactance in field terms. Unit field voltage of this expression and Xe. Thus,
is defined as the d-c. voltage corresponding, by Ohm's 2 Eo Xe G
Law, to unit field current. The unit value of either EXe = H2 cos a sin (t + a)
resistance or reactance is thus the ratio of unit field
voltage to unit field current. Any other value of ohms [ H-4 B COs2 (t + a) ] E-al
could be chosen as the unit base, but this particular 2 Eo Xe P
value affords special convenience in that the calculated + H2 sin 2 (t + a) [B C - A F] E-ft
results are in terms of the known field excitation
voltage. If, on this basis, the per-unit external 2 Eo Xe K
reactance in the field circuit is Xe, the per-unit voltage + H2 sin 2 (t + a) [ B C - A F ] (7e)
across its terminals under sudden short circuit con-
ditions is the product of Xe and the rate of change of As in the case of the voltage across an external
field current. reactance in the armature circuit, the terms containing
The initial field current is given by equation (13c), the derivatives of the decrements have been omitted
hence the voltage across the reactance at the instant of because they are quite inappreciable. It will be ob-
short circuit is served that the last term of (7e), representing the sus-
tained voltage across a field reactance, is exactly
Exe - 2E 'X { H G cos a sin (t + a) equation (5e).
H2
+ sin2 (t + a) [B C-HF + Fcos2 (t + a) Appendix F
-2 G cos a cos (t + a)] } (le) VOLTAGE ACROSS THE OPEN PHASE
Expressed in series form (le) becomes
in It has been shown that the current in the short-cir-
EpEssed seriesGform le)
( becomes cuited phase produces a space fundamental m. m. f.
A 2/3 k i (if)
E E XeG (1 + b)cosa
) =

E,/ = XQ'
VXD' [sin(t+a)of which the component in the direct axis is
+ 3bsin3 (t + a) + 5b2sin5 (t + a) + . ] AD = 2/3kicos (t + a) (2f)
2 Eo Xe G (1 + b) and the component in the quadrature axis is
XD/ + N/ XD/ XQ' [sin 2(t + a) AQ = 2/3 ki sin (t + a) (3f)
+ 2 b sin 4 (t + a) + 3 b2 sin 6 (t ± a) + . . . ] (2e) Under sudden short-circuit conditions, the space
fundamental fluxes produced are
Referring to Appendix B, it will be remembered that, SD=23kiXd o t+o)(f
under sustained short circuit, only the odd harmonics D=2/kiXa'cs(+a)4f
are present in the armature current. Hence, only the SQ= 2/3 k 'i Xaq' sin (t ± ax) (5f)
d-c. component and the even harmonics will exist in the The components of these fluxes which an open
field current for sustained short circuit, as explained in winding whose axis is at an angle:g from the axis of the
Appendix C. These field currents, given by their short-circuited winding are respectively,
respective components of equations (13c) and (14c) are 'pi = 2/3 k i Xad' cos (t + az) cos (t + ae + I3) (6f)
Feb. 1928 DOHERTY AND NICKLE: SYNCHRONOUS MACHINES-IV 481

and 02 = 2/3kixa' sin (t + a) sin (t + a +±) (7f) 2 k eo I ki

The flux linkages due to these fluxes are - H sin (t + a) Xm + 2 k
l = 2/3 k kii Xad' cos (t+ a) cos (t + a I + ) (8f)
and i,V2 2/3 k k i xaq' sin (t + a) sin (t + a + ,) (9f)
=
where k1 is a constant by which per-unit fundamental J
flux must be multiplied to give the flux linkages in the
open winding under consideration. For a line to _ 4 k e Bin (t + )
neutral winding, k1 is unity and for a line to line winding H2
k, is N/3. ki
The current in the short-circuited winding also
produces mutual flux linkages with the open-circuited
[cos a- cos (t + a) ]
X 2
k1 [(XAD + XAQ3
winding such as those due to space harmonies, mutual
slot fluxes, mutual fluxes in the end turn leakage paths, cos 0 + (XAD'- XAQ') cos (2 t + 2 a + d)] (17f)
etc. The flux linkages from this source are }
3
=
i Xm (lOf) When normal current flows through one winding of
The flux linkages in the open-circuited winding due
to the rotor current existing before short circuit are .o O armature, fluxOf linkages
the will be produced in the
= k1 e0 cos (t ± ae413± A) k, eo os (t(llf)
= + a +otherwindings. these linkages, the larger part is
due to space fundamental fluxes. The remainder are
The total flux linkages in the open winding at any due to non-space fundamental fluxes such as mutual
instant after short circuit are then slot fluxes, mutual fluxes in the end turn leakage paths,
-t
1 1 + 412 + VI3 ± 4 (12f) etc. These linkages are represented by xm. In many
Using the relations, similar to equation (13i), that cases, this is small compared with the space fundamental
3 flux linkages and by assuming it to be zero many of the
Xad' + Xa,' =~k2 (XAD' + XAQ') (13f) equations are greatly simplified.
For ordinary values of f which exist in synchronous
3 machines (60 deg. and 90 deg.) equation (17f) essen-
Xad' - Xaq' 2k2 (XAD' - XAQ') (14f) tially has its maximum when t = 180 - a. For this
value of t, all terms but the first and second disappear
the total flux linkages are and
ki eo cos (t +
a ±i3)
k1eocos(t±
a
± ~ ki
+ A) + i x +2k
ki
em =-.k eo sin [ XQ'
, (1 + cos a)-cosa] (18f)
cs)

[(XAD' + XAQ') COS 3 + (XAD'- XAQ') cos (2 t + 2 a + )] When 3 is very small this value is in error since for
(15f) f = 0 deg. there will be a voltage in the open phase
The voltage induced in the open winding when corresponding to leakage reactance voltage in the
another winding in the machine is short-circuited is short-circuited phase. For values of : approaching
given by the negative rate of change of equation zero it is necessary to plot the voltage curve to find its
(15f), or maximum value.
e= k1eo sin (t + a ±f) The open phase voltage can also be expressed in the
ki harmonic form by substituting the current expression
+ -yk (XAD' - XAQ') sin (2 t + 2 a + 13) j from equation (2b), and its derivative. In this har-
monic form, if we make the approximation that
k1 XAD - XAQ' = (XD - XQ') (19f)
Xm + 2k [ (XAD' + XAQ')COS
and let = 2 then xm = 0, and the fundamental
d is
+ (XAD' - XAQ') cos (2 t + 2 a + A)] (16f) voltage in the open phase is found to be
d= kte(1 + b)cos(t + a) (20f)
Substituting the value of current from equation (lb), The fundamental voltage in an open-circuited phase
and its derivative, and using the notation of Appendix 90 deg. away from the short-circuited phase will thus
A, this becomes
voltage
2 k e0
~~bemainincreased by the ratio (1 + b), as mentioned in the
body of the paper.
el= k1 e0 sin (t ± at + A3) + H When the short circuit occurs when maximum flux
links the winding, i. e., when ae = 0, the vroltage on the
(XAD - XAQ') sin (2 t + 2 ae + j3) [cos a - cos (t ± at)] open phases will have the general form shown in Fig. 19
482 DOHERTY AND NICKLE: SYNCHRONOUS MACHINES-IV Transactions A. I. E. E.

7r Equation (22f) has a maximum when t = 7r/2.

When a = 2 , no flux links the winding at the XQ' 1
Then e1=-kekL X'Q
instant of short circuit and the voltage on the open
phase may be of the form shown in Fig. 20. 7r
If the short circuit occurs line to neutral, the winding a = 2
comprised of the two unshort-circuited phases in series
is in space quadrature with the short-circuited phase. 7r
Then A = </2 = 2
k1 = V3
Xm 0
°

k =1 t = (24f)
For a = 0, equation (17f) then becomes, (using the 2
simplified notation), Equation (17f) may be decomposed into a form which
2 k eo clearly shows the components affected by the armature
el = k eo cos t + H (XAD' - XAQ') cos 2 t (1 - cos t) and field decrements, and the steady-state term. Thus,
substituting for (XAD' - XAQ') and omitting xm, equation
k eo (17f) becomes
± H (XAD'- XAQ') sin t sin 2 t 2 k k, B cos eo [ o
el e1 = H2
H2 H 2 (t +ao)
[cos
2 k B eo
+ H2 (XAD' - XAQ') sin2 2 t (1 - cos t) (21f) + B sin2 2 (t + a) ] E-¢a'
k, eO
w + kH2 [H2k- 2kBHcos2 (t + a)
When a = 2 equation (17f) becomes
- 2 k B2 sin 22 (t + a) ] cos (t + a)
2keo + kBHsin(t+±a)sin2(t+ a)} [K- (1-K) e-yf']
(25f)
el = k eo sin t- J (XAD- XAQ') sin t cos 2 t
This is the open phase voltage for any condition of
k eo transiency, i. e., for any flux originally trapped in the
- J (XAD' - XAQ') cos t sin 2 t armature at the time of short circuit. For maximum
flux linkages a = 0 and all terms appear. For minimum
2 k B eo
+ J2 (XAD -XAQ') sin t sin2 2 t (22f) flux linkages a = 2 and that term containing the
As pointed out before, practically all the difference armature decrement does not appear. The steady-state
between XD' and XQ' is due to the difference in space voltage is obtained by neglecting all terms containing
fundamental flux, a negligible amount being due to the either the armature or field decrements.
variation in leakage fluxes in the two axes. Hence, for
all practical purposes, (XD' - XQ') may be taken equal Appendix G
to (XAD' - XAQ'), which simplifies equations (21f) and VARIATION OF ARMATURE INDUCTANCE WITH POSITION
(22f), some of the terms combining into more compact The object of resolving a sine wave of m. m. f. into
terms. Equation (21f) has a maximum when t = 7r. two components is to obtain fluxes which are in space
For this value of t phase with the m. m. fs. In dealing with synchronous
[2 XQ' 1 machines, resolution is made about the direct and
el =-k eo XD I (23f) quadrature axes, and the reactance of a single-phase
armature winding in any position with respect to these
a =0 axes, due to the space fundamental flux produced by
7r the space fundamental m. m. f., may be obtained as
,3 =
2 follows.
Assume a single-phase armature winding with the
t =w7 axis of the winding at anangle zyfrom the direct axis
The voltage on this winding before short circuit is of the machine. When a current i of normal frequency
k e0. Hence, the voltage at short-circuit line-to-neutral flows through this winding a space fundamental m. m. f.
2xX' of amplitude
may rise to Q-1 times the voltage before short 2
XD ~~~~~~~~~~~~A
= kicost (lg)
circbuit on the open phases.
Feb. 1928 DOHERTY AND NICKLE: SYNCHRONOUS MACHINES-IV 483

is produced. The component of A in the direct axis is as the m. m. fs. The flux linkages, produced by the
2 2
AD = 3 k i cos -y cos t 3 (2g) single-phase current i will then be - k2 times the linkages
~~~~~~~~~~~~~~~~3
and the component in the quadrature axis is produced by the same three-phase current i, and
2 2
AQ = 3 k i sin y cos t (3g) XAD = 3 k Xad (9g)
The m. m. f. in the direct axis produces a space 2
fundamental flux in this axis proportional to the and XAQ = 3 k2 Xaq (lOg)
m. m. f. If there are no short-circuited secondary
windings in this axis, the proportionality factor between Substituting these relations in (8g)
m. m. f. and line-to-neutral linkages is Xad,* or, in general
k Xad. =[XAD + XAQ XAD - XAQ 1
a L 2~~ ~2 ~~~~~
+ 2 -ylcos t
cos
2 L 2
Then, {D = k2 i Xad Cos -Y cos t (4g) (lg)
the quadrature
Likewise, inLikewise, inquadrature
the axisd
axis
The induced voltage is d t Thus, remembering
2 dt
4,Q = k2 i Xaq sin -y cos t (5g)
that i is taken here as the amplitude of normal fre-
The component of ,tD which is in space phase with the quency current:
armature winding, thus linking the winding, is e4 XAD + XAQ XAD -XAQ c
2 e,= lL 2 + 2 cos 2 y jsin t(12g)
A1/='AD cOS y =3 k2 i xad Cos' 'Y cos t
The coefficient of i sin t is the per-unit reactance of the
2 r1 1 1 single-phase winding due to the space fundamental flux
3 k2 i Xad + - COS 2 y cos t produced by the space fundamental m. m. f., as a func-
tion of position when both the rotor and stator are
(6g) stationary. Since per-unit reactance of a stationary
Also, the component of 'PQ in space phase with the circuit at normal frequency is numerically equal to the
armature winding is per-unit inductance of the circuit, this coefficient
2 gives also the per-unit inductance; and when used as an
'P2 = k2 i xaq sin2 cyos t inductance, is independent of the relative motion of the
rotor and stator.
2 ri 1 - As pointed out in a previous paper,9" the leakage
= - k2 i Xa,L - 2 COS 2 y cos t (7g) reactance also varies with position as a second harmonic
of the angular displacement of the winding from the
The linkages due to the total flux in space phase with direct axis, i. e.,
the winding is XLD + XLQ XLD - XLQ
2 Xad + Xaq Xad Xaq XL 2 + 2 COS 2 y (13g)
=k-k2 2 cos2y
C y t (8g)
The total reactance may then be written
The three-phase reactances, Xad and Xaq may be XD + XQ XD - XQ
expressed in terms of the corresponding reactances for x 2 + cos22 y (14g)
the single-phase winding in question. When a current 2 2
i flows in a single-phase winding, the alternating where, XD = XLD + XAD (15g)
2 and, XQ = XLQ + XAQ (16g)
stationary m. m. f., fundamental in space is - k times
Appendix H
the rotating m. m. f. produced when a three-phaseDERMNFATS
current i flows through a three-phase winding. IfDERMNFATS
these m. m. fs. act on the same reluctance paths, the Field Decrement. The theoretical basis for deter-
space fundamental fluxes will have the same relation mining this is given in the main body of the paper
- ~~~~~~~~~~~under
"Fundamental Equations." It is shown in
*xd is the three-phase reactance of armature reaction andApedxCtathavrgfilcuen i
applies here since the single-phase space fundamental m. m. fs. Apni htteaeaefedcreti . h
are expressed in terms of normal three-phase spacye fundamental d-c. component, spontaneously increases at short
m. m. f. circuit to a value
484 DOHERTY AND NICKLE: SYNCHRONOUS MACHINES-IV Transactions A. I. E. E.

XD V ID Q The average current required to sustain these linkages

XD' + V
XD'
XQ' ~ is,
by equation (9b) in Appendix B,
times the value existing immediately before short =o k eocos a
circuit. Thus the increase in equivalent magnetic V\/ XD'I XQ'I
reluctance, as discussed in the above reference, is given Thus, the inductance to be used in the decrement
by the same ratio, and the corresponding change in factor is t'+io, or
inductance is given by the reciprocal. Hence, assuming L=v XD' XQ'
negligible armature resistance, as may be done without adteamtr atri
appreciable error in commercial machines, the field adteamtr ermn atri
ermn

decrement factor is r
XD ± V\/XD'/XQ' a
VXD'/XQ' (4h)
=XD'+VN/XD' XQ'(h where r is the per-unit resistance of the short-circuitedl
where o-o is the open circuit, field decrement factor, armature circuit, and XD' and xQ' represent the per-unit
static transient inductance, (and also the corresponding
0,0
20
(R
~~~~~~per-unit
~~~~~The
reactance).
decrement factor 0- applies to the d-c. com-
where (R is the ohmic resistance of the field circuit, and ponent of the total armature current, the envelope of
2o the total inductance, in henrys, of the field circuit, the even harmonic currents in the armature, and the
corresponding to the condition of 'open circuited arma- envelope the odd harmonics in the field current.
of
ture, and w is 2 wr f, where f is normal frequency. Apni
These three quantities cii, £~, and w are not per-unit Apni
quantities, but o-0 is. REACTANCEs-DEFINITIONS AND RELATIONS
If appreciable resistance exists in the armature circuit Xd = synchronous reactance, three phase, line-to-
the value of o-f given by (1h) will be decreased to the neutral, direct axis, corresponding, to open-
extent (discussed under "Fundamental Equations") circuited rotor windings.
which the equivalent inductance is thereby increased. Xad = reactance of armature reaction, thus due to the
It is of interest to note that in the ideal case of a direct axis space fundamental air-gap flux,
laminated cylindrical rotor where produced by three-phase currents. It is one
XQ =XQ = XD componentf Xd.
XDa- (h Xld = armature leakage reactance, three-phase, line-to-
equation (1h) reduces to o-f - (h)neutral, and is the other component Of Xd.
XD
The decrement factors (1h) and (2h) give, of course,
~~~~~Thus,
X
Xecadce +hreX hae,lin(tnut)l
l

the average decrement, thus determining the direct Xd

rnin ectacertree-pdnt hae, lieto-nleutaral
trnsen
component of transient current in the field, the envelope Xd
of the even harmonics in the field current, and the direct axis, corsodnIothnoa ekg
envelope of the odd harmonics of current in the arma- flux between the -armature and field windings.
ture. It can be shown that these decrement factors are It thus corresponds to the total flux linkages
rigorous for the envelopes and average component of the armature winding occasioned by the
referred to, for the assumed case of zero armature sudden application of three-phase currents,
resistance. Thus the only error involved is the slight when the field winding is closed, such flux
difference in wave shape, the peaks being correct. The likgsbigolCtoewihgnrt
same thing is true of the armature decrement, derived voltage of fundamental frequency. It com-
below, under the assumption of zero field resistance. prises X,d-that is, the leakage flux of the
Armature Decrement. The armature decrement ap- armature plus additional space fundamental
plies to the treatment flux linkages which happen to be fuwhccrseteaIr-ga,adcre
caught and "trapped" in the armature circuit when the spnsi mutt hefedlaaefu
latter is s,hort-circuited. It is,- of course, altogether a produced by the induced field current. The
Feb. 1928 DOHERTY AND NICKLE: SYNCHRONOUS MACHINES-IV
current under those conditions. It will, of
course, have a different value if the single-
phase winding is considered to be line-to-line,
from that if the winding is taken as line-to-
neutral, the per-unit value being about three
times as much in the former as in the latter
case. It is analogous to the three-phase
total reactance u to
XAD = static reactance, single-phase, corresponding to
the space fundamental flux, direct axis under
the above conditions. It is a component of
XD, and is analogous to the three-phase reac-
tanee of armature reaction Xad.
XLD = static armature leakage reactance, single phase,
direct axis, corresponding to the conditions
specified under XD. It is analogous to Xld and
is the other component of XD, that is, corre-
sponds to all flux linkages not contained in the
space fundamental. Thus,
XD = XAD + XLD (3i)
XD' static transient reactance, single-phase, direct
axis, corresponding to the conditions specified
under XD, except that the rotor windings are
short-circuited. It corresponds to the total
leakage flux between the armature and rotor
windings, and is thus analogoustOXd. It thus
comprises XLD, and some additional space
fundamental flux crossing the air-gap and
corresponding in amount to the field leakage
produced by the induced field current. The
latter component is denoted by XAD'. Thus,
XD' XLD + XAD' (4i)
The corresponding symbols for all of the foregoing
ten quantities for the quadrature axis are denoted by
the subscript q instead of d for three-phase quantities,
and by Q instead of D for the static single-phase quanti-
ties. Thus small letter subscripts indicate three-phase
quantities, and capitals indicate static single-phase
quantities.
There are certain additional important relations
which exist between the above quantities. These
relations will now be derived.
Subtracting equation (4i) from equation (3i),
XD - XD' = XAD -XAD' (5i)
and, subtracting equation (2i) from (Ii),
Xd - Xd' = Xad - Xad t (6i)
Now the same space fundamental m. m. f., whether
produced by a three-phase or a single-phase current, will
produce in the same axis the same space fundamental
linkages in a given winding. Thus, since by definition,
the product of Xad by the amplitude of a space funda-
mental m. m. f. produced by a three-phase current
direct axis will give line-to-neutral linkages of armature
reaction, it follows that the product of Xaci by any
space fundamental m. m. f. in the same axis will give
line-to-neutral linkages. Hence, the line-to-neutral
linkages, maximum value, due to a suddenly appearing
485
space fundamental m. m. f. AD, the latter being pro-
duced by a single-phase current, are,
AD = AD Xad
when the rotor circuits are open-circuited, and are,
tkD AD Xad'
when the rotor circuits are closed. The line-to-line
linkages would be k times these values, where k = 1.0
for line-to-neutral, 1.73 line-to-line. Thus in general
{D = k AD Xad (7i)
and 6D' = k AD Xad' (8i)
Subtracting (8i) from (7i),
'PD - ,D' = k AD (Xad -Xad') (9i)
But, by definition of x, 1D = iD XD (IOi)
and iD! = tD XD' (lli)
Subtracting these
{D -'D = iD (XD - XD') (12i)
Equating (9i) and (12i), and substituting the general
relation from (1c) between the amplitude of the space
fundamental
andathetaluem. m.ft f. produced by current,
single-phase current,
2/3 k2 iD (Xad - Xad') = iD (XD - XD')
Therefore, substituting (6i),
3
Xd - Xd' = 2k2 (XD - XD') (13i)
Nomenclature
omencla ure
Per-unit quantities, as outlined under "Discussion
of Theory, are li2ted and defined below. The general
scheme of notation involves the use of capital letters for
field circuit quantities, small letters for the armature.
For the Armature. All voltages are expressed as a
fraction of the peak value of normal armature voltage;
currents, as a fraction of the peak value of normal
armature current; flux, as a fraction of normal space
fundamental flux, the latter corresponding to normal
voltage at normal frequency on open circuit; linkages,
as a fraction of the linkages corresponding to normal
flux; inductance, as a fraction of unit inductance, the
latter being defined by the ratio of normal linkages to
normal current; and reactances (at fundamental
frequency), as a fraction of unit reactance, the latter
being defined by the ratio of normal voltage line-to-
neutral to normal current.
For the Field. All voltages are expressed as a fraction
of unit field voltage, i. e., the direct voltage across the
field corresponding to unit field current; currents, as a
fraction of unit field current, i. e., the direct current
corresponding to normal a-c. terminal voltage on open
circuit (no saturation); flux, as a fraction of normal flux,
as defined above for the armature circuit; linkages, as a
fraction of unit field linkages, the latter being defined
as the linkages produced in the field winding by normal
flux; inductance, as a fraction of that indluctance which
is defined as the ratio of unit field linkages to unit field
current; reactance, as a fraction of unit field reactance,
486 DOHERTY AND NICKLE: SYNCHRONOUS MACHINES-IV Transactions A. I. E. E.

which is defined as the ratio of unit field voltage to unit
field current.
With regard to the per-unit bases for field voltage and r = per-unit resistance of the armature circuit
field reactance, it should be mentioned that there are R - resistance, in ohms, of field circuit. Not a per-
other bases which, in certain cases, it might be more
convenient to adopt. It is altogether a matter of con- t
venience. For instance, if it should be of greater con-
venience to have symmetrical equations for the arma-
ture and field, than to have, as in the present case, the
calculated results of field circuit voltage come out in x
terms of the known excitation voltage, then unit field Xad = three phase, line to neutral reactance of arma-
voltage might be taken as the voltage which unit field
linkages would generate in the field winding at normal Xld = three phase line to neutral armature leakage
frequency. Unit field reactance and resistance would
then be changed accordingly.
A = amplitude of space fundamental of armature
m. m. f. Specific values of A are:
Ad = m. m. f. in direct axis, produced by
balanced polyphase current
Aq = corresponding value in quadrature axis
A1 amplitude
= of m. m. f. produced by
current any
at instant in a single-phase
winding.
AD = component of A1 in direct axis
AQ component of A1 in quadrature
= axis
b = geometric ratio defined by the relation following XD'
equation (9)
e = instantaneous value of any armature voltage
eo= line-to-neutral voltage before short circuit
ed = nominal voltage due to direct axis excitation
eq = nominal voltage due to quadrature axis
excitation
A ed increment in ed induced by transient field A3
current
A eq = increment in eq induced by transient field
current
E = field voltage in general
Eo = field voltage preceding short circuit
i = armature current in general
I field current in general
Id - field current in direct axis winding
Iqz = field current in quadrature axis winding
A Id increment in Id induced by transient armature
current
A Iq = corresponding increment in Iq
Io = field current at the instant before short circuit p = ratio of open phase armature voltage to peak
k = voltage conversion factor for the short-circuited
phase. k =1 for line-to-neutral short cir-
0i
cuit, and = ,/3 for line-to-line short circuit
kL = similar factor for the voltage generated in the
open circuited phase
I = inductance, in general, of the armature circuit
I' - transient inductance of the armature, variable
with respect to rotor position
L c inductance, in general, of the field circuit
.1 - inductance, in henrys, of the field circuit, with
open circuited armature. That is, the to-
tal field self-inductance. Not a per-unit
quantity
unit quantity
= time, expressed as a fraction of the time corre-
sponding to one radian at normal frequency.
It is thus numerically equal to the elapsed
radians
armature reactance in general
ture reaction, direct axis (See Appendix I)
reactance, direct axis
Xd Xad + Xld = three phase, line to neutral syn-
chronous reactance, direct axis
Xd' = three phase line to neutral transient reactance,
direct axis
XAD = static reactance single phase, corresponding to
the space fundamental flux, direct axis.
(See Appendix I)
XLD = single phase armature leakage reactance, direct
axis
XD XAD + XLD = total static reactance, single phase
direct axis. (See Appendix I)
single phase static transient reactance, direct
axis
The corresponding reactances in the quadrature axis
are denoted by q and Q, instead of d and D.
a displacement between armature winding axis
and axis of field pole at the instant of short
circuit. See Fig. 1
angle between the axis of the short circuited
winding and any open circuited winding
considered
Py - angle between armature winding axis and direct
axis at any time t. See Fig. 1
OD = flux in the direct axis due to current in a single
phase winding
Q= corresponding flux in the quadrature axis
= flux linkages the armature, in general
in
= flux linkages, produced in the armature by any
transient armature current
Vt'm armature linkages due to current in the field
winding
value of unit armature voltage
decrement factor of the field circuit, correspond-
ing to the -total self-inductance (i. e., with
armature open circuited). It is thus,
____
= Wo
where 6R is the total resistance (in ohms), 4s is
the total inductance (in henrys), of the field
circuit; and X is the angular velocity corre-
sponding to normal frequency-thus 2 w f.
Feb. 1928 DOHERTY AND NICKLE: SYNCHRONOUS MACHINES-IV 487

Although, M, ,e, and w are not per-unit 12. Bekku, S., "Sudden Short Circuit of Alternator,"
quantities, oro nevertheless is. The usual Researehes of the Electrotechnical Laboratory No. 203, June
simple exponential is of the form 1927.
13. Lyon W. V., Transienit Conditions in Elec. Machinery,
e'
Etowhrt'adto'
where t' and to' are
aeiin seconds.
eod.II the h TRANS. A. I. E. E., L.,
14. Dreyfus, Vol. 42,Magnetische
1923,"Freie p. 157. Energie Zwischen
present system t represents the elapsed Verketteten Mehrphasensystemen," E. U7. M. 29 S. 891, 1911.
radians-i. e., time is expressed as a fraction 15. Dreyfus, L., "Auisgleichvorgange in der Symmetrischen
of the time of one radian at normal frequency. Mehrphasenmaschine," E. U. M1. 30 S 25, 121, 139, 1912.
Thus, referring to the exponent 16. Dreyfus, L., "Ausgleiehvorgange Beim Plotzlichen
Kurzsehluss von Synchron Ceneratoren," Arch
it fur
Electro-
t' t' t' technik, 5 S 103, 1916.
= =
tot -
co - u
Discussion
af = decrement factor of the field corresponding to C. H. Linder. In Synchronous Machines-Part IV, mathe-
the condition of single-phase short circuit. matical relations are presented which permit the calculation of
See equation (2) the short-circuit currents and voltages at the instant of single-
a. = corresponding decrement factor of the armature
circuit. This factor is on a per-unit basis
similar to ao. It is merely the ratio of per-
unit r to per-unit x. If these were each
multiplied by unit ohms, and in the resulting
ratio of resistance to reactance the factor wo
were taken out of the reactance, the decre-
ment factor would be back on the usual basis.
o = 2 7r f, wheref = normal frequency
Bibliography
1. Steinmiietz, C. P., "Transient Electric Phenomena and
Oscillations," 1909, Chapter XIV.
2. Boucherot, P., "Electromagnetic Phenomena Resulting
from the Sudden Short-Cireuiting of an Alternator," Trans.
International Elec. Congress, Vol. II, p. 49-93, Turin, 1912.
3. Biermnanns, J.,(a) E. T'. Z., 1915, p. 579. (b) "Uber-
strome in Hochspannungsanlagen," 1926, p.49 et seq.
4. Diamant, N. S., (a) Calculation of Sudden Short-Circuit FI. I-ARMATUIE CURRENT; THE SHORT CIRCUIT OCCURRING
Phenonoena . . . ,TRANS. A. I. E. E., Vol. 34, p. 2237, 1915., AT THE INSTANT OF MAXIMU m LINKAGES WITH TIIE SHORT-
(b) Sustained Short Circuit Phenomena. . ., TRANS. A. I. E. E., CIRCUITET I PHASE
Vol. 36, Plart 2, p. 1141, 1918. Upper curve-Test
5. Franklin, R. P., (a) Thesis (Master's Degree), Union Lower curve -Calculation
College, 1923, Calculation of Alternator Shortt-Circuit Cur-
rent.s. (b) Short-Circuit Currents of Synchronous 3Machines,
TRANS. A. I. E. E., Vol. 44, p. 420,1925.
6. Karapetoff, V., (a) Initial and Sustained Short Circuits
in Synchronous MlIachines," TRANS. A. I. E. E., Vol. 44, p. 403,
1925. (b) Variable Arnmature Leakage Reactance in Salicnt-Pole
SynchronouTs Ma(ichines, TRAN S. A. I. E. E., Vol. 45, p. 729, 1926.
7. Laffoon, C. AM., Short Circuits of Alternating-Current
Generators, TR1ANS. A. I. E. E., Vol. 43, p. 356,1924.
S. Shimidzu and Ito, "Theory of Single-Phase Generators,"
Memoirs of College of Engineering, Kyoto Imperial University,
Vol. 2, No. 7,192,5.
9. (a) Doherty and Shirley, Reactance of Synchronous
Machines and its Application, TRANS. A. I. E. E., Vol. 37,
Part 2, p. 1323, 1918. (b) Doherty, R. E. Shoort-Circuit Current
of Induction Motors and Generators, TRANS. A. I. E. E., Vol. 40,
p.547, 1921. (C ) Doherty, R. E., A Sinplified .Method of A nalyzing
Short-Circait Problems, TRtANS. A. I. E. E., Vol. 42, p. 849, 1923.
(d) Dolierty and Nickle, Synchronous Machines-I and II, FIG. 2-FIELD CURRENT; TIIE SHORT CIRCUIT OCCURRING
TRANS. A. I. E.E.,Vol.45,p.912,1926. (e)DohertyandNickle, AT THE INSTANT OF MNAXIMUm LINKAGES WITII TlHE SHORT-
Synchronous Machines-IlI, Winter Convention 1927. CIRCUITED PHIIASE Upper Ctlrve-Test
10. \Newbury, F. D., Discussion. TRANS. A. I. E. E., Vol. 37, Lower curve-Calculation
Part 2, p. 1304 1918 and Laffoon, C. M., p. 1316.
11. Arnold, E., Die l1echselstromntechnik, Vol. 4, p. 37-40, phase short circuit, and at any time thereafter. Experimental
1913. data are included in the paper which check the calculated curves