. 
 
 . . 
 
 . 
 
SHORT CIRCUIT TRANSIENTS ON 
 ELECTRICAL MACHINES 
 
 BY 
 
 WILLIAM MORRIS YOUNG 
 
 B.S. University of Illinois, 1921 
 
 THESIS 
 
 SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS 
 FOR THE DEGREE OF MASTER OF SCIENCE IN ELECTRICAL 
 ENGINEERING IN THE GRADUATE SCHOOL OF THE 
 UNIVERSITY OF ILLINOIS, 1922 
 
 URBANA, ILLINOIS 
 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

 UNIVERSITY OF ILLINOIS 
 
 THE GRADUATE SCHOOL 
 
 Tune - 2 - 192 - 2 . 
 
 1 HEREBY RECOMMEND THAT THE THESIS PREPARED UNDER MY 
 
 SUPERVISION BY Willia m Mor ris Young 
 
 ENTITLED Short CirouiU-T ran s i en t s - - on — Sle e trUeal 
 
 M achines . 
 
 BE ACCEPTED AS FULFILLING THIS PART OF THE REQUIREMENTS FOR 
 THE DEGREE OF Master of Scienc e in Electrical Engineering. 
 
 In Charge of Thesis 
 
 Head of Department 
 
 Recommendation concurred in* 
 
 Committee 
 
 on 
 
 Final Examination* 
 
 Required for doctor’s degree but not for master’s 
 

 
 Digitized by the Internet Archive 
 in 2016 
 
 
 
 
 
 
 https://archive.org/details/shortcircuittranOOyoun 
 
TABLE OF • CONTENTS. 
 
 I INTRODUCTION 1 
 
 II GENERAL DISCUSSION OF SHORT CIRCUITS 2 
 
 1. Short Circuits on Alternators 2 
 
 3. Short Circuits on D.C. Generators 4 
 
 III SHORT CIRCUIT TESTS. 
 
 1. D.C. Short Circuits on A.C. Driven Sets 11 
 
 A. festinghouse 85 K.W. Set 11 
 
 a. Old Breakers 
 
 b. Meter Disconnected from Power Plant 
 
 c. Automatic Breakers 
 
 d. Field Flux Shift 
 
 e. Fuse 
 
 B. Westinghou.se 9.5 K.W. Set. . . 21 
 
 C. General Electric 15 K.W. Set 25 
 
 3. D.C. Short Circuits on D.C. Driven Sets 28 
 
 A. Northern 8.5 X. V, . Set 38 
 
 B, Westinghcuse 3 K.W. Set 30 
 
 3. A.C. Short Circuit on D.C. Driven Set 33 
 
 A. Three Phase Short Circuits 33 
 
 a. General Electric 15 K.W. Set 
 
 b. West i ■ "house 9.5 K.W. Set 
 
 B. Sin 'le Phase Short Circuits 35 
 
 a. General Electric 15 K.W, Set 
 
 b. Westinghouse 9.5 K.W. Set 
 
 4. D.C. Short Circuit on Synchronous Converter 
 
 Westing-house 10 K.W. Converter 37 
 
 IV CONCLUSION 44 
 

 
 
 
 . 
 
 
 
 
 ! 
 
 
ILLUSTRATIONS 
 
 Fl£‘ 8 * Page 
 
 I Circuit Diagram - Wes .ting-house 85 K . I . Set XI 
 
 II Search Coils cn Field. Poles 18 
 
 III Circuit Diagram - Westinghouse 9.5 K.W. Set 33 
 
 IV Circuit Diagram - wort hern Electric 5.5 K.F. Set 38 
 
 V Circuit Diagram - General Electric 15 K.W. Set 33 
 
 VI Circuit Diagram - restinghcu.se 10 K.W. Converter 37 
 
 Oscillograms 
 
 1-155 Short Circuit Tests 4?~75 
 
 16 Apparatus and Arcing Fuse 7^ 
 
SHORT CIRCUIT TRANSIENTS ON ELECTRICAL MACHINES 
 I INTRODUCTION 
 
 Tne Electrical Engineering Laboratory receives direct current 
 from an 85 kilowatt. West inghouse, .350 volt, 540 ampere, compound 
 generator, that is coupled by a flexible leather coupling to a ten 
 pole, two phase, 440 volt. West inghouse synchronous motor. The gene-; 
 rator has been protected against sustained short circuit by two 
 type CC, 300 ampere, 250 volt, hand reclcsed circuit breakers. Re- 
 cently the Electrical Department installed two 400 ampere, 110 volt, 
 type HRL1, automatic reclcsing circuit breakers in series with the 
 West inghouse breakers. The generator is frequently short circuited 
 accidentally during laboratory periods and at such times beta sets 
 of circuit breakers have opened. It was therefore of interest to 
 determine tne length of time required for tne different circuit 
 breakers to open under severe snort circuit. An oscillographic stud;' 
 of this natter lead to the mere interesting study of the phenomena 
 that took place in motor and generator when a severe short circuit 
 was applied. Studies were therefore made of the short circuit phe- ! 
 nomenu in direct current generators driven by both synerrenous motcrii 
 and direct current motors, in alternating current generators driven 
 by direct current meters and in synchronous converters, the short 
 circuit being applied on the direct current side. 
 
IT GENERAL DISCUSSION OF SHORT CIRCUITS 
 1. Shcrt Cir cuit s c n A lternators.- In the case of the alter- 
 nating current generator, the maximum short circuit current is lim- 
 ited, by armature reaction and armature self-induction. The anr ature 
 current represents a magnetomotive force which is demagnetizing with 
 respect to the impressed or field flux. The armature magnetomotive 
 force results in a decreased resultant flux which causes the ^enerat' 
 ed voltage to <lecrea.se from E to e. There is another flux produced 
 by the armature current which links only with the pole face and 
 armature iron and is a self inductive flux. Therefore, the armature 
 impedance may be represented by 
 
 Z 1 = r l"i x l or z i = yFf + xj 
 
 where r^ is the effective resistance of the armature and short cir- 
 cuit path, and x^ is the armature reactance. 
 
 At shcrt circuit the impedance, Zi , consumes all the generated i 
 e.m.f. , and if the effective resistance of the armature and short 
 circuit path is small compared to x^ it may be neglected and the 
 snort circuit current is 
 
 e 
 
 Armature reaction, under steady conditions, may be represented by 
 an effective reactance Xg. The total reactance of the armature is 
 
 x c = X 1 + x 3- 
 
 At the instant of short circuit the only impedance in the cir- 
 cuit is that due to the resistance rp and the reactance x-j. because 
 Xg requires time to develop, as it represents the change in the 
 field flux produced by the armature magnetomotive force. Any change 
 in the field flux generates an e.m.f. in the field coils which 
 changes the field current so as to retard the change of field flux. 
 
3 
 
 Hence, at the moment of short circuit, the magnetomotive force of 
 armature reaction opposes that impressed by the field excitation and 
 the magnetic flux begins to decrease. This decrease of flux estab- 
 lishes an e.m.f. in the field coils in such a direction that the 
 field current, and therefore the impressed magnet omot ive force, is 
 increased. This field m.m.f, combines with that of armature reac- 
 tion to give tne instantaneous values of field flux. Therefore, 
 before the field flux nas appreciably decreased the generated volt- 
 age is E and tne short circuit current is limited only by the arma- 
 ture reactance. As the field decreases, due to the demagnetizing 
 action of armature reaction, the generated e.m.f., and hence the curj 
 rent, decreases until it reaches a value 
 
 o 
 
 11 " 
 
 which is its steady short circuit value. At tne same time the field 
 
 value 
 
 current reaches its normal full load and is constant. The amount of 
 decrease of tne field m.m.f. is tne same as tne m.m.f. of armature 
 reaction. 
 
 In the t nree pnase alternator the m.m.f. of the armature is i 
 constant in intensity and revolves at synchronous speed with respect 
 tc tne armature. It is, therefore, stationary with respect to tne 
 field in the stable condition cf short circuit. At tne instant of 
 short circuit the resultant armature m.m.f. changes in intensity and 
 velocity and approaches a steady value by a series of oscillations. 
 Hence, with respect to tne field, the armature reaction is pulsating 
 and oscillating so that it generates e.m.f. 's in the field coils 
 w hich cause variations in the field current and field terminal volt- 
 age. This pulsating gradually dies out as conditions stabilize. 
 
 Wnat nas been said in regard to three pnase short circuits on 
 
4 
 
 alternators also applies, in part, to single phase short circuits. 
 
 Of course, in the three phase short circuit the currents in the three 
 phases are different at any instant, whereas in the single phase 
 short circuit the currents in the two lines snort circuited must 
 nave tne same value. The armature current sets up an armature re- 
 action which alternates with respect tc the armature and hence pul- 
 sates with double frequency with respect to the field. This double 
 frequency pulsation of the flux of armature reaction induces a 
 double frequency e.m.f. in the field coils, ana hence there is a 
 double frequency pulsation in the field current and field terminal 
 
 v? i j 
 
 voltage. This pulsation is proportional to the current flowing in 
 the armature but, unlike the case of a three phase short circuit, it 
 dees not die away as time goes on and the transient term disappears. 
 
 Superimposed on tne double frequency pulsations is a single 
 frequency pulsation due to the transient in the armature current. 
 
 The initial value of tms pulsation depends on the instant on tne 
 current wave at which the short circuit is applied. If the short 
 circuit occurs at a maximum on the current wave, the single fre- 
 quency pulsations will be maximum, and, in like manner the pulsation; 
 will be zero if the short circuit occurs at a zero point on the cur- 
 rent wave. This occurrence of the short circuit at a maximum causes 
 the peaks of tne double frequency pulsations to be widely different 
 in size, but in the ether case the peaks will all be the same size. 
 
 2 • Short Circuits on Direct Current Generators . - In the direct 
 current generator, conditions are considerably different from those 
 in the alternators, Beth have the same essential parts and function; 
 in the same fundamental ;anner but conditions at short circuit are 
 quite different. The field is ordinarily supplied 'oy a shunt, or 
 

 
5 
 
 shunt and series winding, with or without interpoles. At no load 
 the field is supplied by current from the brushes and the terminal 
 e.m.f. is maintained by the e.m.t’.s induced in the armature conduc- 
 tors as they pass under the field poles and cut lines of force. If 
 the armature is slotted for the windings, as is most generally the 
 case, the teeth draw the lines of magnetic force from the field 
 poles in the fori; of tufts, and these tufts of flux sweep across tne 
 pole raceB with the same velocity as the peripherial speed of the 
 armature. 
 
 Suppose, now, a very low noninduct ive resistance is connected 
 across the terminals of the machine. The terminal voltage will at 
 once fall because it is only equal to the IR drop in the resistance. 
 When tne short circuit is applied the current mounts toward its 
 final value. This final value cannot be assumed at once, as would 
 be expected cf a current flowing; through a high resistance with the 
 same e.m.f. impressed, but must mount up cn a leg. curve ned 
 
 by the constants of the armature circuit. 
 
 At tne instant the short circuit is applied, then, the terminal 
 voltage of the machine is 
 
 E = Ri i + L § . 
 
 where i^_ is the instantaneous current in the armature, L the coef- 
 ficient or self induction cf the winding, and R the total resistance 
 
 of the armature ana external resistance. As time goes on the 
 
 di 
 
 transient term, will nave less and less effect, but the rate at 
 
 wnich the current builds up will depend on this term or on tne in- 
 ductance in the circuit. 
 
 After the transient term has disappeared tne current 
 
 E 
 R 
 
 i = E 
 
3 
 
 As the terminal voltage falls the shunt field excitation also 
 tends to decrease. There will be no immediate change in the form 
 of a decrease in the field current due to a fall in the terminal 
 e.m.f. because the shunt field of any generator is a highly induc- 
 tive circuit , and when the impressed voltage f alls there is induced 
 in tne field coils an e.m.f. in the same direction as the failing 
 impressed voltage. This causes the field current, at tne instant cf 
 short circuit, to tend to remain constant and the length cf time it 
 is constant depends on the inductance of the field circuit. The 
 field current may be obtained from the expression 
 
 E = Ri-, 4 L^il 
 dt 
 
 ri 3 4 
 
 L'di 2 
 dt ' 
 
 where r is the resistance of the shunt field circuit. U is induc- 
 tance, and i^ the current flowing. The other symbols were given 
 above. This indicates that the field current and current tnrough 
 the external circuit change at ditrerent rates and the rate of chang 
 of one is a multiple of the rate cf change of the other. 
 
 When the current rises in the armature it sets up a m.m.f. of 
 armature reaction. Thi3 opposes the field m.m.f. and causes an 
 e.m.f. to be induced in the field coils since the field flux cannot 
 change instantly. This e.m.f. causes a decided change in the field 
 current and the magnitude of this change depends on the strength cf 
 tne series field. It is maximum for a shunt generator and less for 
 an overcompcund generator. This rise in field current is only mcmer{- 
 tary, it tnen decreases, due to decreased voltage impressed on the 
 
 field. 
 
 If the field current falls the field flux also .falls and if the 
 
 / 
 
 terminal e.m.f. is tc remain constant a series field must operate tc 
 supply the needed flux. The series field has a low inductance sc 
 
7 
 
 it readily responds to changes in the armature current. At the De- 
 ment after the machine is short circuited, then, there is a field 
 acting which is tne resultant of tire sum of the shunt and series 
 m.m.f. minus the demagnetizing rn.m.f of armature reaction. Just 
 after short circuit more energy is being given up per unit time than 
 later when the shunt field excitation has fallen. 
 
 The direction of the resultant field, of course, changes great- 
 ly due to armature reaction. The shunt field flux is decreased by 
 the demagnetizing component and shifted from its normal position by 
 the cr^os magnetizing component. If the current in the armature is, 
 very great, as in the case of short circuit, it is almost certain 
 that there will be sparking at the brushes, or arcing at the commu- 
 tator, The cross magnetising action of armature reaction causes 
 e.m.f!s to be generated in the armature conductors at a different 
 position than at no load. Tne shift of this commutating zone will 
 oe in the direction of the brushes sc that the brushes will pass ove 
 commutator bars whose coils are still generating voltage, and that 
 coil short circuited will have a considerable current flowing in it. 
 As tms snort circuited coil leaves the brush a spark will result. 
 This spark ionizes the air and may even form carbon and copper ions. 
 When mors bars pass tne brush there is an ionized path ever which 
 the spark can easily travel sc that tne spark is drawn cut a little 
 farther and more ions are formed. This process continues until ther 
 is a path of air, carbon, and copper ions, from one brush to the 
 next of opposite polarity. At the potential difference existing, it 
 is very easy for the current tc travel over this path so that an arc 
 is established and maintained as long as the terminal voltage has 
 any appreciable value. Interpoles tend tc prevent arcing by 
 

8 
 
 maintaining the commutating zone in its normal position, but if the 
 current in tne interpole coils cannot build up with the armature 
 current, or if the interpoles have considerable inductance, the com- 
 mutating zone will shift at the instant of short circuit and may net 
 be restarted to its normal position before an arc is established. 
 
 It might be well to state that before the short circuit is ap- 
 plied there is a fairly uniform distribution of flux over the pole 
 face except for the tufting produced by the teeth of the armature. 
 When the short circuit is applied the flux is crowded over to the 
 trailing pole tip so that the iron in that re ion becomes saturated, 
 but in tne leading pole tip there are comparatively few lines of 
 force. At tne instant the short circuit current is applied, current 
 flows through tne armature conductors and for a very short period of 
 time tnere is a shifting of the lines of flux from the field over 
 the field pole face. This is due to the interaction between the 
 magnetic field surrounding the armature conductors and the field 
 lines of force. This interaction is in tne nature of a repulsion, 
 so for a short time there is nc cutting cf the lines of force by the 
 armature conductors. When the field flux has traveled as far as it 
 can then there is cutting cf the field flux again and tne terminal 
 e.m.f. begins to rise toward a maximum value determined by the IR 
 drop through the external resistance. 
 
 If tne short circuit is broken, the magnetic field around each 
 armature conductor is broken and the only field remaining is that 
 of the shunt field. When tnere is nc longer any crowding of the 
 field flux to the trailing pole tip it tends to whip back ir.to the 
 no load position. As it shifts the armature conductors for a very 
 short time, cut through the field flux at a much higher rate than 
 
normally is caused by armature rotation alone, because the field 
 flux is moving in a direction opposite to that of the armature rota- 
 tion. This results in a very nigh e.m.f. induced in the armature 
 which may be even double the rated armature voltage. However, this 
 voltage is only momentary and the terminal e.m.f. rapidly settles 
 down tc its rated value or the maximum value it can nave with that 
 given field excitation. 
 
 In obtaining data two oscillographs were used so that the vari- 
 ation in current and voltage could be obtained on both motor and 
 generator. In general the short circuit current, generator field 
 current, and a 60 cycle timing wave were taken on one oscillograph 
 and the field current, armature current, and generator terminal volt 
 age were taken on tne second oscillograph. Both film drums were 
 driven by one motor ana only one device was used to open the shutter^ 
 of both instruments. The short circuit was applied through a 100 
 ampere, 250volt, 4 pole, single-throw knife switch which was closed 
 by the operator when the shutters of the oscillographs opened. A 
 circuit diagram for each short circuit and a complete explanation 
 of the events recorded on tne film is given. Nc attempt is made to 
 calculate any of tne currents analytically, the maximum reached be- 
 ing estimated from references taken before the short circuit was 
 applied. 
 
 The general procedure in takin illograms was, first tc 
 
 take zero lines cf the waves, then a reference line for all quanti- 
 ties that varied, and lastly tc take the transient. References 
 were of such a size that they were comparable with the values reach 
 by the variaoles during the transient. Wherever possible, full load 
 values of the machines were used as references even though they 
 
' 
 
 . 
 
 
 
 
 ■ 
 
 • ■ • 
 
 . 
 
 
 
 
 
 
 ’ 
 
 ■ 
 
 , ." 1 , 
 
 
 
 
 
 
 
 
 
 
 . 
 
 1 1 II I] * I I 
 
 
 
10 
 
 were small compared to the maximum reached in some cases. The tim- 
 ing reference was always taken during the transient and was supplied 
 from a 110 volt, SO cycle line designated as Phase A in the electri- 
 cal laboratory. A 7500 chm variable resistance was used to control 
 the size of the wave in tne oscillograph in tne case of all voltage 
 waves. For the current curves the IR drop ever an instru it shunt 
 was used wherever possible. In the case of alternating currents, 
 current transformers of a 1200-5 ratio were used and a potentiometer 
 was connected on the secondary side. The IR drop from this potenti- 
 ometer was then impressed on the oscillograph. 
 
A cross (x) indicates an oscillograph vibratar in the circuit 
 as shown in the diagram. The generator armature current , terminal 
 voltage, and time were taken on one oscillograph and the motor arma- 
 ture current in phase A and phase B, and the voltage impressed on 
 the field were taken on the other oscillograph. Both shutters were 
 operated by one mechanism for the same length of time, ana, as the 
 field impressed e.m.f. of the synchronous motor was supplied by the 
 E. C. generator, there was a tie between the two oscillograms. The 
 fi r .id voltage of the motor dropped with the terminal e.m.f. or the 
 generator ana had the same general characteristics during the remain 
 der of the short circuit. 
 
 Oscillograms la and lb were taken on the old manual circuit 
 breakers before the new automatic breakers were installed* Tne bus 
 breakers were held closed by tying down the armature of the relay, 
 contained in tne circuit breaker, so that it could net operate. 
 Oscillogram la shows the transient m tne generator armature wave. 
 
 A 150 ampere reference was used and from this reference the indi- 
 cating lines of current were drawn. Oscillogram lb shows tne tran- 
 sient in tne currents in the two phases of the synchronous motor 
 and also that in tne field voltage. The field voltage has tne same 
 characteristics as the terminal voltage of the generator as it was 
 supplied by tne generator. 
 
 In oscillogram la, the short circuiting switen was closed at M. 
 The terminal voltage at once dropped to a value of about 40 volts an[i 
 then began to rise. The sudden drop was due to tne lines of force 
 of the field being carried across tne pole face and no lines of fore 
 being cut for a short time. When tne armature conductors began to 
 cut lines cf force tne current rose rapidly through the armature 
 
13 
 
 and also through the series field. The series field tended to de- 
 stroy the demagnetizing effect of armature reaction so that the 
 voltage slowly rose to a value of about 90 volts. After the current 
 ceased to rise rapidly, the rapid increase in voltage also ceased 
 due to less effective action of tne series field and also due to the 
 fact that the shunt field, which received a lower excitation at re- 
 duced voltage, had decreased the shunt field flux. These two factor 
 tne decrease in the shunt field excitation and the demagnetizing ef- 
 fect of . ture reaction, contributed t armature current 
 
 at the value shown. In the current curve tne increase was on a lag. 
 curve determined by the armature reactance. It required about .025 
 second for the current to cease its rapid rise. The maximum value 
 attained was about 3550 amperes or almost 7.5 times rated full lo^d 
 current . At 0 cn the current curve the circuit breakers opened and 
 the copper contact points separated. At Q tne carbon points sepa- 
 rated and the stored electrical energy in the armature caused an 
 arc to for rhich rapidly carried the current to zero at N. The tim 
 required for the circuit breakers to open was .041 second. In the 
 voltage curve, at a point directly above 0 on the current curve, the 
 voltage mounted rapidly due to the decreasing current and when the 
 arc broke the field flux whipped back across the pole face into 
 normal position. This induced a high momentary voltage in the arma- 
 ture conductors which produced a peak cn the voltage curve of 400 
 volts when tne current finally reached zero. The terminal voltage 
 again decreased but to a value below normal because the field cur- 
 rent had not had time to build up to its rated value in the snort 
 time taken for the voltage to mount when the armature current fell 
 to zero. The voltage never rose to normal because at N, or a little 
 
14 
 
 later, the synchronous motor was pulled out of sync Lsi . 
 speed cf the set decreased and decreasing speed meant less lines of 
 flux cut by tne armature conductors per unit time and aence a lower 
 induced voltage in the armature. The impedance of the field coils 
 acted to decrease the build up of the field current and hence cf the 
 field flux. The timing wave was taken from phase A cf the labora- 
 tory supply. A slight drop in the amplitude of the wave is notice- 
 able a little after N and this is due to a decreased voltage when 
 the synchronous motor drew a heavy current in trying to remain in 
 synchronism. Again, just below P, the reference or timing wave 
 show's evidence of a third harmonic of varying frequency. This thirc j 
 harmonic is in the magnetizing current of tne transformers that sup- 
 ply the laboratory. At P the circuit breaker on phase B opened 
 while that on phase A held in. The timing wave therefore shows a 
 third harmonic cf voltage whose frequency varies with respect to 
 the 60 cycle wave. The power plant supplying the electricity ran 
 at very nearly 30 cycles whereas the synchronous motor frequency 
 was less than 60 cycles after it was pulled cut of synchronism. It j 
 was later found that a 6S5 KVA three phase turbo alternator was 
 supplying power, and in the space cf the oscillogram the frequency 
 dropped from 60 to 58 cycles per second and the voltage decreased 
 very materially. 
 
 Oscillogram lb shows the current in the two phases cf the 
 synchronous motor and the voltage impressed on the motor field. The 
 voltage wave needs no explanation as the field was excited from the 
 D.C. generator and the voltage impressed on the field cf the motor 
 nas the same characteristics as the terminal voltage of the gene- 
 rator. When the voltage decreased on the field of the motor tne 
 
15 
 
 power factor decreased to a low value lagging as it was running as 
 an induction meter. It is quite probaole that the power factor did 
 not fall with the voltage but at some later point, for example, 
 about P, and the current in the two phases began tc increase to sup- 
 ply the power necessary tc deliver the D. C. power. 
 
 It will be noticed that the armature current dees not begin to 
 increase as scon as the short circuit is applied but actually begins 
 tc increase .05 second later. This shows that the snort circuit 
 power was supplied by the rotation of the armatures of the generator 
 and motor and only when the speed of rotation decreased was it that 
 the motor armature current mounted to hold the set in synchronism. 
 
 As the motor torque was not sufficient tc supply the power of short 
 circuit on the generator the set was pulled cut of synchronism but 
 attempted tc go each into synchronism between D and E and between E 
 and F. A little before F, phase B circuit breaker opened and the 
 phase circuit fell to zero. The meter was then operating single 
 phase and cut of synchronism sc that, if the phase A had not been 
 opened the points such as E and F would have become more and mere 
 frequent as the set slowed down. They represent points in time wher 
 tne motor is in pnase opposition with the supplying source. The 
 maximum value reached by the alternating current when the motor at- 
 tempted to remain in synchronism was, as indicated, about 575 amp- 
 eres. 
 
 b) Motor Disconnected from Power Plant. 
 
 Tc prove that the energy dissipated in the short circuit was 
 supplied by the kinetic energy of the revolving armatures cf the 
 two machines, another oscillogram was taken on the old breakers. 
 
 The same references and curves of generator armature current and 
 
16 
 
 voltage were taken. When everything was ready for the short circuit 
 to be applied, the power supply to the synchronous motor was broken 
 and the field circuit was opened. The short circuit was then applie 
 with the result as indicated in oscillogram 3. Tne current maximum 
 is about 3400 amperes and the voltage minimum about 35 volts. The 
 time required for tne breakers to open was .041 second as in oscil- 
 logram la and the total time required for the current to reach zero 
 after tne short circuit was applied was .058 second, the same as in 
 la. The voltage curve nas tne same values and general characteris- 
 tics a3 in la so it is thus proved that tne energy for short cir- 
 cuits of short duration is supplied by the revolving armatures cf 
 the machines. Two other oscillograms were taken h the field cf 
 the synchronous motor excited by a separated source but the curves 
 are identical with these of la and lb except that the e.ra.f. im- 
 pressed on the field is constant. These are not shown because cf 
 their similarity to la and lb. 
 
 c) Automatic Breakers 
 
 The automatic reclosing circuit breakers were then installed 
 in series with tne old manual breakers. It was desired to know tne 
 length of time required for these breakers to open under short cir- 
 cuit so the same connections were used as in Figure 1. Both the 
 bus ana the aid manual breakers were tied so that the armatures 
 pf one relays in tne breakers could net operate. In this way the 
 
 only protecting device on the generator was the automatic breakers. 
 These breakers open both cn overload and low voltage or zero volt- 
 age. In other words, a voltage must be impressed on the breakers 
 before they can close or stay closed. The same procedure as before 
 was followed in taking the oscillograms and the same references were 
 
1 ? 
 
 used. The transient is shown by oscillograms 2a and 2b. 
 
 In oscillogram 2a the voltage curve has the same charact eristicf 
 as that of la so the same explanation holds that was given before. 
 
 In the current curve the actual maximum obtained was less than in tl 
 snort circuit c old breakers due to more resistance in the line 
 
 there being several more contacts and a considerable length of wire 
 added. However, the short circuit represented the same conditions 
 that occur when a short circuit is plugged on the switchboard. The 
 time required for these breakers to open was .05 second against 
 .041 second for the old breakers. In the case of the old breakers 
 it required .053 second for the current to reach zero after the 
 snort circuit was applied but in the case of the automatic breakers 
 it required .035 second. From a comparison of the oscillograms it 
 is apparent that a longer arc is drawn by the automatic breakers. 
 
 The old circuit breakers have a magnetic blowout feature while this 
 is apparently lacking in efficiency in the automatic breakers as 
 shown by the much longer time required for the arc to break. 
 
 The oscillogram 3b requires no special comment as it is almost 
 identically the same as la. Phase B breaker opened after the same 
 length of time but tne maximum instantaneous current reached in 
 Phase A was 800 amperes as against 675 amperes in lb. From a 
 consideration of the oscillograms, it will be noted that more elec- 
 trical energy is given up in the short circuit in 2a than in la be- 
 cause 
 
 Q = idt 
 
 or the area under the current tine curve, and that is wnat is repre- 
 sented in the oscillogram. 
 
- 
 
 
 
 
 
 
18 
 
 d) Field Flux Shift 
 
 It was thought for some time that the sudden rise in the gene- 
 rator terminal voltage when the short circuit "/as removed, might he 
 due to armature reactance rather than to a shifting of the field 
 flux from its position during load back to its no load or normal 
 position. To prove that the second was the real cause, three search 
 coils were wound of 30 turns of double cotton covered wire and these 
 were placed in the air gap of the generator, one at each pole tip, 
 and one in the center of the pole face as shewn. 
 
 When the behavior of the induced e.m.f. in the search coil was notec 
 at no lead it was seen that there was an e.m.f. generated in the 
 coils due to the tufts of flux carried across the pole face cy the 
 armature teeth. Oscillogram 8c shows the induced e.m.f. in one of 
 the search coils with reference to the 60 cycle power wave. The 
 oscillogram shows a very nearly sine wave, hence the flux distri- 
 bution on the armature must be almost semisoidal with respect to the 
 teeth. This gave a basis for believing that the high voltage at 
 break of the short circuit was due tea shift in the field flux. 
 Accordingly the generator was again short circuited, having the 
 
13 
 
 circuit breakers set nigh but leaving both the automatic and manual 
 breakers free to operate. A 60 cycle timing wave, generator arma- 
 ture current, and one search coil were cc .ted to one oscillo- 
 graph and the other two search coils and the generator terminal 
 voltage were connected to the other. Oscillograms 6a and 6b snow 
 this short circuit transient. The e.m.f. in all the search coils 
 before short circuit is of constant magnitude. When the short cir- 
 cuit was applied in M of 6a, there was a very high e.m.f. generated 
 in all the search coils showing that there had been a great change 
 in the flux distribut er the pole face. The cur cent had net 
 
 had time to change in the field winding, nor had the armature cur- 
 rent mounted to a sufficient value to cause the series field to have 
 effect, so that the e.m.f. must have been generated by the shifting 
 of the field flux distribution. This also occurs at the moment the 
 voltage falls so no lines of force are cut by the armature conduc- 
 tors for a short time. It will be noticed that after the flux has 
 ceased to swing over the pole face there is a higher e.m.f. gene- 
 rated in all the search ceils than before short circuit. This is 
 because the series. field is adding its flux to that of the shunt 
 field and giving a greater flux density at the pole face and at the 
 armature teeth. The increase in generated e.m.f. in the search 
 coils is greatest at the trailing pole tip and least at the leading 
 pole tip, wnicn Dears cut the theoretical consideration of flux dis- 
 tribution under load. When the circuit breakers open ana the cur- 
 rent dies down in the armature, the field flux slowly comes back to 
 its no load distribution. This is indicated by the continually in- 
 creasing e.m.f. generated in the search coils as the armature cur- 
 rent dies down. When that current reaches zero and the flux is 
 

 
 
 
 
 
 . 
 
 . 
 
 
30 
 
 restored to normal no load distribution the e.m.f. generated in the i 
 search coils also falls back to the original values. Careful rr.eas- 
 uring of the e.m.f. generated in the coils both before and after 
 snort circuit showed that in all coils the e.m.f. was less after 
 short circuit. This indicates a smaller flux density at the pole- 
 face and hence there must be a smaller e.m.f. generated in the arn.a- 
 t ure . 
 
 A glance at the armature current curve shows the same charac- 
 teristics and values as in la. The time required for the breakers 
 to open is the same as in la and, as the automatic breakers have an 
 inherently slower speed of operation, it is concluded that the manu- 
 al breakers opened before the automatics. 
 
 s) Fuse 
 
 It was also desirable to note the difference between the char- 
 acteristics of a current broken by a circuit breaker and that brok- 
 en by a fuse. A 13 inch fuse was constructed of two strips of 80 
 ampere fuse wire in parallel, giving a capacity of 160 amperes. 
 
 All circuit breakers were securely tied to a nonoperate position anc. 
 the short circuit was applied with the results as shown in oscil- 
 lograms 4a and 4b. Comparison of tnese with 3a and 3b show exactly 
 the same phenomena except the armature D.C. current has .0086 seconc 
 longer duration in 4a. As all the circuit breakers were tied in tuf 
 nonoperate position and as the fuse was still intact after the shor’; 
 circuit it was assumed that the fuse carried the short circuit cur- 
 rent. Further investigation revealed the fact that the short cir- 
 cuit current was actually interrupted by the automatic circuit 
 breakers which opened when the voltage dropped far below normal, 
 indicates the opening of the circuit breakers and Q the point at 
 
0-1 
 
 3J. 
 
 wmcn tile arc across the carbon point broke. 
 
 A fuse composed of 4 strands of 15 ampere fuse wire was then 
 inserted for the 160 ampere fuse and a short circuit applied under 
 tne same conditions as before. A kodak was set before the fuse to 
 take a picture cf the arc drawn and this is shown in 16. The short 
 circuit is recorded in oscillograms 5a and 5b. 5b is similar to 
 other oscillograms of the reaction on the synchronous motor and 
 needs no comment here. 5a is also the same as 2a with the exception 
 of the D.C. current through the armature. The resistance of the 
 fuse increased as the temperature rose hence the current fell slight- 
 ly and in this case there is no hump in tne curve due tc the open- 
 ing of the circuit breaker. There is only tne formation or the arc 
 across the sections of fuse ana. this is indicated at Q. It seemed 
 that the fuse exploded and melted in many points rather than in only 
 one. Tne points of melt probably were determined by slight kinks 
 in the fuse wnen it was stretched between the points on the wooden 
 fuse block. 
 
 There is, therefore, a difference in tne current curve of a 
 short circuited D.C. generator when the circuit is broken by a fuse 
 and when it is opened by a circuit breaker. A circuit breaker inter- 
 rupts the current twice, once when the copper contacts separate, and 
 
 again when the carbon tips separate and draw an arc, D'ith a fuse 
 
 « 
 
 the current is interrupted only once, i. e. , when the arc formed by 
 a melted fuse, is broken. 
 
 B. Westinghcuse 5.5 K.W : . Set 
 
 Data of Set 
 
 Generator 
 
 Type SK 
 
 320 volts 
 
la+If 
 
 Frame SQ 
 12 Horse power 
 Style 78E37 
 Serial 4180299 
 4 poles 
 
 Style S078B36 
 Serial 41803S5 
 3 phase 
 30 cycle 
 3.5 K.W. 
 
 Field 
 
 6 pole-revolving 
 
 47 amperes 
 1200 r.p.m. 
 
 50° C temp, rise in 34 1 
 under full load 
 
 Field 
 
 Shunt ana series windings 
 Interpoles 
 
 220 volt 
 
 25 amps, per phase 
 1200 r.p.m. 
 
 Armature winding 
 Double circuit wye 
 
 110 volt excitation 
 
 Circuit Diagram 
 
 Gen era tor. 
 
 f\ Of on 
 
 22 
 
 Fig m 
 
 2.2.0 Volt 3 Phase 
 Scott ~BanA of Trons. 
 
33 
 
 Generator field current, generator armature current, and a 60 
 cycle timing wave were recorded on one oscillograph and motor arma- 
 ture current, motor field current, and generator terminal, or crush, 
 voltage were recorded on the other. Reference lines were taken usiq? 
 full load on the generator ana unity power factor on the motor. Zero 
 lines were also taken and then the short circuit was applied. A 
 circuit breaker set for 100 amperes ms used in tne generator cir- 
 cuit and the motor ms protected by the circuit breakers on this 
 primary side of the transformers. When the short circuit was appliei. 
 the armature current began to rise but did not reach any definite 
 value. It rather rose and fell in a series or irregular saw-toothed i 
 oscillations. This is because the generator arced over at tne com- 
 mutator and the arc path, when once established, had so much less 
 resistance than the short circuiting circuit that, nearly all the 
 current flowed through the arc. At times the arc woul i wreak and I 
 the current would flow through the four pole switch and thus give a 
 momentary deflection in the oscillograph. As Ion- as the arc main- 
 tained the oscillograph showed no current flowing in the external 
 circuit. 
 
 The voltage curve also shews these saw-toothed irregularities. 
 Whenever the arc broke and current flowed in the external circuit 
 that current was less than the arc current hence the field flux, 
 which had been badly distorted at short circuit, would attempt to 
 regain its no load distribution and would cause a momentary in- 
 creased e.m.f. in the armature. This increased e.m.f. would again 
 cause the arc path to be established and the voltage would fall 
 once more. It finally reached a value of about 30 volts because of 
 
 tne added effect of tne series field. 
 
24 
 
 The field current of the generator rises on a lag curve to 
 three times its normal value. Tnis increase in the field current is 
 a result of the armature reaction of the generator* The armature 
 reaction sets up a demagnetizing m.m.f. (demagnetizing with respect 
 to the shunt field) and the flux produced threads through the field 
 coil and induces an e.m.f. therein. This e.m.f. causes the field 
 current to mount up rapidly, its rate of rise being determined by 
 the reactance of the field coils. When there is no more relative 
 motion of the field coil and the flux cf armature reaction, the 
 field current begins to die down toward its normal value. The field 
 is excited by the terminal or brush e.m.f. and this terminal e.m.f. 
 is very small. Therefore the field excitation decreases and the 
 field current within the time indicated by the oscillogram, nearly 
 reaches zero. The very small ripples in the field current curve 
 are caused by the segments of the commutator passing under the brush- 
 es. The larger ripples are a result cf the very large current in 
 the motor armature. This current mounted up tc a very large value 
 and one phase was taken through the switchboard to a current trans- 
 former located near the oscillograph. It was necessary tc take the 
 leads for the two field currents through the switchboard to bring 
 them to the oscillograph. When the armature current increases it 
 builds up a large magnetic field about the switchboard wires which 
 threads through the conductors carrying a part cf the field current 
 tc the oscillograph; This induces an e.m.f. in the f . 
 leads, the magnitude depending on the distance between the wires, 
 
 ... the result is the ripples in the field current lines. It will 
 be noticed that these ripples are of the same frequency as the arma- 
 ture current and begin and end with the increase and decrease of the 
 
25 
 
 armature current. 
 
 In oscillogram 7b the armature current of the motor began to 
 rise at D and, unfortunately, the oscillograph vibrator shifted its 
 zero position as is indicated by the dotted line. The circuit 
 breakers opened at R. The field current be to mount as scon .. 
 the armature current increases because of the voltage induced in 
 field coils by armature reaction. This induced voltage folic. vs the 
 fluctuations of tne peak values of the armature current and not the 
 instantaneous values. Between D and S are the large ripples in the 
 field current of the generator. Tnese did not actually exist in the 
 machine but only in the- oscillograph circuit. The large fluctu- 
 ations betv -en S and T are caused by the armature current. The 
 fluctuations are of a double frequency nature and cease as scon as 
 tne circus area sned. It may 03 explained on the assumption 
 
 that at £ one of the breakers of the Scott bank of transformers 
 opened and for a snort time tne three phase motor ran single phase. 
 This would induce such a double frequency pulsation in the field cur 
 rent and i8 much the same condition as in a single phase short cir- 
 cuit cn a three phase alternator. The small ripples that occurred 
 in the 4.2 ampere reference of field current are probably due to an 
 unbalancing of the currents in the three phases of tne motor. This 
 is probably because the transformers forming the Scott bank are of 
 tne same size and will give an unbalanced current supply tc a bal- 
 anced three phase circuit. Other phenomena have been dealt with 
 before and tne discussion will net be repeated. 
 
 C. General Electric 15 K. V. Set 
 
 Data of Set 
 
Generator 
 
 Type EFI 
 No. F567357 
 30 Horse power 
 
 Field 
 
 36 ' 
 
 85 amps, 
 
 220 volts 
 1800 r.p.m. 
 
 4 poles and 4 interpoles 
 
 Shunt 1300 turns 
 
 I tterpoles 50 tui 
 Armatur e 
 
 Slots 5/ IS" x 15/16" 
 Length 4.5" 
 
 Nc.ll D. C.C. ..'ire Resist. 300 u>. 
 
 No. 5 D. C.C. ire Resist. .019 to. 
 
 Diameter 12". Field bore 12 7/52" 
 finding - 2 turns per slot 2 Nc.ll D. C.C. wires 
 2 coils per slot Resist. .085 Ifo. 
 
 ! E rushes 
 
 7/16" x 1 1/.-" x 3" 
 
 Motor 
 
 3 per stud. 
 
 4 studs 
 
 Type EFA -4-30-1800 Form 1. 06 amp. per phase 
 
 No. F300398 
 15 K.W. 
 
 F ield 
 
 4 poles 
 
 Shunt 
 
 2500 turns No. 13 C.C. 
 
 240 volts 
 1800 r.p.m. 
 
 C. wire 
 
 Series. 3 in series with motor armature (disconnected) 
 6 turns 1 layer 5 No . 6 E.C.C. wire 
 
 Armature 
 
27 
 
 48 slots 2/6” x 15/16” 
 
 Length 4.5” Diameter 12” Field bore 12.35" 
 
 Winding 
 
 5 turns per coil 
 
 1 coi]. per slot 
 
 Double circuit wye 
 Collector 
 
 240 volt. 7” Diameter 
 
 The circuit diagram was the same as that used for the Westinghouse 
 set in the last discussion. Oscillograms 8a and Sb show the tran- 
 sient clearly. In this case there was no arcing at the commutator 
 except for a very short time between M ana but even then the 
 current does not fall below 150 amperes at any time. The maximum 
 reached was about 435 amperes. All small irregularities in the cur- 
 rent wave are due to sparking at the brushes of s all sections of 
 the commutator being arced over. The irregularities in the gene- 
 rator field current are due to rapid changes in the e.m.f, induced 
 in the field by the ampere turns of armature reaction. At X tne 
 armature current has become fairly stable and the transient or 
 shifting of the armature reaction has decreased sc that the field 
 cur . falls to a value caused by the brush voltage. All the oth- 
 er phenomena on 8a have been discussed under 7a and much in 8b has 
 been discussed in 7b. The current in the motor armature mounts to 
 a very high value as the motor is pulled cut of synchronism, reach- 
 ing a value between 1000 and 1200 amperes (allowing for inaccuracies 
 in measuring the reference). Hence, tnere large ripples in- 
 duced in the motor field current between 8 and T, The circuit 
 breakers both opened at R so there is no double frequency pulsation 
 
26 
 
 in the field current at any time except that due tc tne transformers 
 as explained before. In this oscillogram, as in the case cf the 35 
 K.W. set, the energy of short circuit is supplied by the kinetic 
 energy cf the rotating armatures so the motor armature current does 
 not mount up until the armature tends to slow down below synchronous 
 speed. 
 
 2. D.C. Short Circuits on D.C. Driven Sets .- 
 A. Northern Electric 6.5 K.W. Set 
 
 Data cf Set 
 
 Generator 
 
 135 volts 52 amperes 
 
 6.5 K.W. No. 10368 
 
 Speed 1035 r.p.m. 4 poles shunt wound 
 
 Mot or 
 
 320 volts 
 .5 Horse power 
 Speed 1200 
 
 6 0 amp s , 
 
 No. 10S0S 
 
 4 poles shunt wound 
 Circuit Diagram 
 
 Generator. 
 
 y^otor. 
 
 / 4 Pole Knife 
 Switch. 
 
 AA a - 
 
 
 lotif 
 
 
 Fig. Ht. 
 
 Zf 
 
 220 Volts 
 
 D.C. 
 
 lot If 
 
29 
 
 Oscillogram Sa shews the generator armature and field currents, 
 and a 00 cycle timir g wave., 9b shows the transient in the generator 
 terminal voltage, motor armature current, and motor field current. 
 When the short circuit was applied the armature current rose rapidly 
 attaining a maximum momentary value of almost 800 amperes. The 
 maximum could net be estimated as the curve went off the film, for a 
 short time. After going through the maximum the current, shown in 
 oscillogram 9a below the field current line, decreased in a series 
 of jumps and irregularities which were caused by sparking at the ! 
 commutator. The decrease in the armature current is caused by a ver ' 
 great decrease in the field flux. The field circuit responded rap- i 
 idly tc the m.m.f, of armature reaction and the field current rose 
 with tns armature current to a maximum value of 5.5 amperes, whereas 
 its full load value was only 1.65 amperes. Tne demagnetizing action 
 of tne armature reaction greatly decreased the effective flux per 
 pole so tne armature current decreased. The series winding was not 
 used, hence there was no flux established to counteract that caused 
 by armature reaction. For this reason the terminal e.m.f. remains 
 low and the armature current decreases. The short circuit was ap- 
 plied at 156 volts or with normal full load field excitation sc the 
 armature current may be rather higher than if only 135 volts term- 
 inal were used. 
 
 The short circuit was applied at G, oscillogram 9b, and the 
 generator voltage follows its usual short circuit cnaracterist ic. 
 
 The very low value was that caused by no lines of force being cut 
 by the armature conductors for a short time. The voltage then 
 builds up, after there is no more shifting of tne field flux, to a 
 value of about 40 volts. In the motor ^mature current there was a 
 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ■ 
 
 
 
 
 
 
 
 I ill 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ■ 
 
 
 
30 
 
 short time, after the short circuit was applied, during which it 
 did not rise. This was, as has been stated before, because the kin- 
 etic energy of the rotating armatures supplied the energy of short 
 circuit. n the rotation decreased the armature current rose to 
 
 supply the required energy of short circuit. The final decrease in 
 tne armature current was caused by a decrease in the generator arma- 
 ture current, less energy being required to supply the short circuit 
 energy. The slight rise in the motor field current is due to arma- 
 ture reaction which induces an e.m.f. that aids the e. m.f. impressed 
 on the field. As the impressed e.m.f. is constant the resultant de- 
 crease of tne field current to its lead value was determined by the 
 impedances of the field circuit. 
 
 The motor armature current was taken through the switch board 
 to an instrument shunt near the oscillograph. The field current 
 leads to the oscillograph also came through the switchboard from a 
 low resistance shunt in the field circuits. It will be noticed 
 that there are no periodic ripples in the field current after short 
 circuit as there were in 8a and 7a because tne set was driven by a 
 direct current motor and any tendency of the motor armature current 
 to induce an e.m.f. in the leads carrying the field current to the 
 oscillograph would have resulted in a small increase in the field 
 current while the motor armature current rose, ana a decrease in the 
 field current when the motor Current reached a steady value. Be- 
 cause of the low maximum to which the armature current rose no ef- 
 fect is noticeable in the field current. 
 
 B. vest inghouse 6 K.W. Set 
 
 Data of Set 
 
 Generator 
 

31 
 
 Mo. 40 Type 8K 
 
 6 K.W. 
 
 135 volts 
 Serial No. 1836192 
 No. 48-3T-99B 
 4 poles 
 
 1750 r.p.m. 
 48 amps. 
 
 4 interpoles 
 
 Field 
 
 Shunt 
 
 3 enes 
 
 Motor 
 
 
 
 
 
 No. 40 Type SK 
 10 Horse power 
 230 volts 
 Serial No. 1336197 
 No. 40-3J-SSB 
 
 Constant speed 
 1750 r.p.m. 
 37.7 amps. 
 
 4 poles 
 
 Field 
 
 Shunt 
 
 4 interpoles 
 
 S eries 
 
 
 The circuit diagram was the same as that used for the short cir 
 suit on the Northern Electric set. -Oscillogram 10a shows the gene- 
 rator armature and field current and a 60 cycle timing wave. Oscil- 
 logram. 10b snows the motor armature and field currents, and the gene 
 rater terminal voltage. 
 
 The generator armature current rose at M when the short circuit 
 was applied, reaching a maximum of about 325 amperes about 6.5 times 
 full load current. At no place on the oscillogram did it decrease 
 below 235 amperes because the series field was very strong and over- 
 came the field setup by armature reaction. This is further illus- 
 trated in the field current. if there had been a large flux from 
 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
armature reaction threading through the shunt field. coil3 there 
 would have been a large increase in the shunt field current. If the 
 series field was of sufficient size to neutralize the effect of arma 
 ture reaction, there would have been induced still another e.m.f. in 
 the field coils which opposed and neutralized the e.m.f. caused by 
 the armature ampere turns.. In that case there would be no rise in 
 the field current. The oscillograms shew this condition to exist 
 as there is but a very slight increase in the field current. It de- 
 creases to its zero because of the very low voltage inpress ed upon 
 it and the main field comes from the series coils and interpoles. 
 
 The interpoles prevented sparking flash-over at the commutator 
 even at short circuit. It is an inherent characteristic of the 
 interpoles to give the current in the conductor approaching a brush 
 a steep wave front but apparently in this case the brushes were set 
 in such a position and the interpoles were cr a sufficient strength 
 that commutation was excellent at all conditions of lead. 
 
 The generator was short circuited at 135 volts or rated voltage 
 which dropped to about 50 volts terminal, then rose slightly due to 
 a small effect of the flux shift, and continued to fall as the speed 
 decreased. 
 
 In the motor, the armature current did not begin to rise for 
 .05 second after the short circuit was applied. Its increase was 
 even then very gradual. The kinetic energy of the rotating arma- 
 ture wires supplied the energy of short circuit for an appreciable 
 time. When the rotational speed began to decrease the motor current 
 rose and a stable condition was soon reached at which the energy of 
 short circuit and machine losses was supplied by the power from the 
 line. There was a gradual transition of power delivered to the 
 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
34 
 
 Tiie three short circuit currents were recorded on cne oscillo- 
 graph and the field current, voltage between lines, and a tilling 
 reference were recorded on the other. Oscillogram 13a shows clearly 
 tne change in the armature current during short circuit. At the in- 
 stant of short circuit the armature current was opposed only by trie 
 self inductive reactions of armature reaction and sc the current 
 mounted to a very large value. The short circuit occurred when the 
 current in phase B was almost zero sc it shows the highest initial 
 value of almost 400 amperes. The effective reactance of armature 
 reaction then comes into effect and slowly decreases the effective 
 field flux, and hence the generated e.m. f. , and the short circuit 
 current. Tne armature current dies down to a steady value of about 
 70 amperes or twice full load current. The transient term has en- 
 tirely died out at X so that, from that point on, there is no change 
 in the wave except, pernaps, a decrease in the frequency. Within 
 the limits of the oscillogram the frequency has decreased from 60 tc 
 53 cycles per second. 
 
 In oscillogram lib tne line voltage fell first tc a value of 
 about 50 volts and then slowly decreased as tne transient in the 
 field current died down to give a steady average value of field cur- 
 rent. The maximum instantaneous value reached by the field current 
 was 7 amperes or seven times normal full lead current. The sing-le 
 frequency pulsations in the field current were due to the transient 
 term of armature reaction. The m.m.f. of armature reaction was at 
 first pulsating in intensity and oscillating with respect to the 
 field. This gradually settled down to a value which was constant ir 
 intensity and position with respect to the field but this is not 
 shown on the oscillogram. The variation in the maximum reached by 
 
35 
 
 the pulsations of the field current was caused by the e.m.f. s in- 
 duced in the field coils, by the . m.m.f..of .... '.ire reaction which 
 decreased with decreasing armature current, hence, generated a low- 
 er and lower e.m.f. in the field circuit. 
 
 cO Westinghcuse 9.5 K.W. Set 
 
 Oscillogra s 13a and 13b were taken on the 9.5 K.W. testing- 
 house set with the same circuit diagram as in the case of the Gener- 
 al Electric 15 K.W. set. They need no explanation in the light of 
 what has just been given. Normal field current on the alternator is; 
 4.7 amperes and normal impressed voltage 110 volts. It will be not- 
 ed that the maximum value of field current attained is about 38 
 amperes or about 5.5 times full lead value. 
 
 B. Single Phase Short Circuits. 
 
 a) General Electric 15 K.W. Set 
 
 Oscillograms 13a and 13b were taken on the General Electric 15 ; 
 K.W. set using the same circuit diagram as for the polyphase short 
 circuit but only short circuiting one phase. In this case the cur- 
 rents in the two lines were the same and need no comment. The cur- 
 rent flowing through the armature set up an armature reaction which 
 was of double frequency with respect to the field and induced a ; 
 double frequency voltage therein. This is very evident in the field; 
 current curve. When the short circuit was applied, the armature 
 as in such a position that the m.ffl.f. of armature reaction acted 
 to produce a flux in the same direction as the field flux. By 
 transformer action tms flux produced an e.m.f. in opposition to 
 that impressed on the field and actually caused the field current 
 to become negative. Tne further rotation ot the armature caused 
 the m.m.f. of armature reaction to pulsate with respect to the f itflc 
 
33 
 
 and so generate the double frequency pulsations in the field current. 
 Superimposed on this double frequency pulsation was a single fre- 
 quency the same as in the polyphase short circuit. In this case, 
 however, the single frequency transient depends on the point on the 
 current wave at which the short circuit was applied. It is zsrc if 
 the current in the armature is zero at the instant of short circuit, 
 and a maximum if the armature current is a maximum. In oscillo- 
 gram 13b it i3 evident that the short circuit was applied at a point 
 near tne maximum of the current wave and hence the field current 
 mounted to about 7.5 times its normal full load value. 
 
 The voltage wave was taken between one of the short circuited 
 lines and the undisturbed line. There is a noticeable third har- 
 monic in tne terminal voltage which is displaced in time phase with 
 reference to the fundamental. This third harmonic voltage is gene- 
 rated by the double frequency pulsation of the field and is there- 
 fore initially caused by armature reaction. This third harmonic is 
 
 also visible in the armature current but in a much smaller degree. 
 The zero line of terminal voltage was displaced upwards about one 
 eighth inch when it was taken due to a small set of the oscillograph 
 vibrator mirror. 
 
 b) West inghpuse 9.5 K.W. Set. 
 
 Oscillograms 14a and 14b were taken on the Westinghcuse 3.5 
 K.W. set, and are explained in the previous discussion on tne Gen- 
 eral Electric 15 IC. W . set. It might be remarked, in passing, that 
 the short circuit in this case was applied about 130 electrical de- 
 grees later than that in 13a and 13b and hence the field current 
 variation is almost of the same magnitude. Comparing oscillograms 
 Idb and 14b it will be noticed that in 13b the field current has a 
 
 much larger negative value than in 14b. This was probably due oc 
 

 
 
 
 
 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ill 1 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3? 
 
 the much higher armature current in the case .of 13fc. it would seem 
 tc indicate that the General Electric generator has a relatively 
 lower effective armature reaction than the Westinghouse generator. 
 4. D, C. Shor t Circuit on Synchronous Converter . - 
 Westinghouse Synchronous Converter 
 10 K.W. 330 volts D.C. 45.5 amps. D. C. 
 
 1-5-8 phase 60 cycle 1800 r.p.m. 
 
 Serial No. 1396553 
 
 Field 
 
 .Shunt and series 'windings. 
 
 Int erpoles 
 
 Circuit Diagram 
 
 complicated, yet most interesting, results. The machine pulled out 
 of synchronism and nearly came to a stop, then began to accelerate 
 and moon. It seemed to be coming up to speed as an induction motor 
 but as the energy was spent in the short circuit it would again die 
 down* This dying down and accelerating continued until the A. C. 
 circuit breakers were opened. The transient is recorded on 
 
38 
 
 oscillograms 15a and 15b and represents the conditions in the ma- 
 chine for only one-third second after the short circuit ?as applied. 
 
 The elements of the two oscillographs were placed as indicated 
 by the crosses on the circuit diagram. D. C. armature current , A.C. 
 voltage between slip rings, and a 60 cycle timing -wave were recorded 
 illograph and D.C. voltage, shunt field current, and A.C. 
 armature current were recorded on the other. The procedure was to 
 take zero lines, reference lines and lastly the short circuit. 
 
 In oscillogram 15b the short circuiting switch was closed at G. 
 For a very short time the energy of short circuit was supplied by 
 . rotating armature. As the armature decreased in speed there 
 was a ] displacement bet e A.C. arma ) flux and 
 
 . field flux so that 1 ,C. armature current began to rise, 
 
 tending tc held the machine in synchronism. The torque was not suf- 
 ficiently large to hold the machine in synchronism so the A.C. cur- 
 rent reached a maximum when the phase displacement was 180° and de- 
 creased as the ph .. . ement went cn to 330° . At 330° the arma 
 
 ture has a very small A.C. current flowing. through it because it is 
 then in phase opposition with the supply for an instant. As the 
 phase displacement increases there is the alternate rising and fall- 
 ing of the armature current. The theoretically maximum torque that 
 can be developed depends cn the constants of the converter. If 
 there is reactance alone in the circuit, the maximum torque will be 
 developed at S0 C phase displacement and if there is only resistance 
 the machine cannot operate as maximum torque is developed at 0° 
 phase displacement. At zero speed the machine would be a static 
 transformer because there would be no shunt field. There would be 
 a tendency toward rotation because of induction motor action on tne 
 
3S 
 
 amort isseur winding. At stand-still all the peaks of the current 
 wave would be the same height. 
 
 At the instant the short circuit was applied the D. C. current 
 mounted to a tremendous value, over 30 times rated full load current 
 of the machine. At the. same time the rotational speed of the arma- 
 ture decreased and the A, C. current began to increase. As the 
 phase displacement of the machine increased to 90 electrical degrees 
 less and less voltage was generated in the armature conductors. 
 Neglecting reactions at 90° this generated D.C. e.m.f. was zero be- 
 cause the armature conductors were net cutting lines of .force. On 
 account of the inductance cf the field circuit, there was aliays a 
 flux from the shunt field, as the current in the winding could net 
 creas v it h terminal e.m.f. Therefore, the armature conductors 
 always cut lines of force and generated an e.m.f. until the rotor 
 became stationary and there was no shunt field excitation. As the 
 phase displacement increased to 180° the E. C. voltage generated be- 
 came negative and caused the D.C. current tc still further decrease 
 until it reached a minimum at 180° phase displacement. This was 
 also a point at whict A.C. cur-re . as ... maximum. As the phase 
 displacement still further increased the generated voltage again 
 approached zero at 270° and a positive maximum at 380°, or a slip of 
 two poles. From there on the cycle cf events repeated itself. As 
 the machine continued to slow down the negative maximum decreased 
 from an actual positive value tc a negative value and in the limit, 
 cr when the rotor was at standstill this current would be alter- 
 nating 80 times a second with equal positive and negative values. 
 
 The very jagged parts cf the E.C. current curve are caused by spark- 
 ing cf the brashes. The commutating; zone was shifted backwards with 
 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
40 
 
 respect to the rotation cf the armature so that the coils snort cir- 
 cuited by the brush were generating voltage of the same polarity as 
 the next brush to be encountered. This, then, was not conducive to 
 vicious sparking. When the commutating zone shifted forward with 
 respect to the rotation cf the armature (really only a continued 
 backward shift) the short circuited coils were generating an e.m.f. 
 of the same polarity as the brush just left. In that case there 
 would be a spark drawn but this sparking would decrease on account 
 cf the continual backward shift cf the commutating zone. Hence, 
 there would be no tendency, or at least very little tendency, for 
 flashover on short circuit. If this short circuit were broken the 
 commutating zone would swing forward as the machine accelerated and 
 an arc over could easily be established. Hence with respect to the 
 two armature currents the E.C. current is a maximum when the A.C. 
 current is ~ mil u and the A.C. current is a maximum when the D.C. 
 current is a minimum. 
 
 In the converter used there were series coils wound ever the 
 shunt field, interpoles for aiding commutation, and an amort is seur 
 winding in the pole faces to eliminate hunting. When the short 
 circuit was applied a very strong current flowed through the series 
 winding. This current also gave rise to a high armature reaction 
 m.m.f. and. induced a current in the amort isseur winding. There 
 were, then, the four m.m.f.s acting. Under normal conditions cf 
 load armature reaction is very small but at the i istant of short 
 circuit very little alternating current was flowing and there was 
 a strong m.m.f. of armature reaction at once set up. The m.m.f, of 
 armature reaction tended to set up a flux in opposition to the flux 
 from tne shunt field, and the series coils and amort isseur winding 
 

41 
 
 tended to set up fluxes to aid the shunt field flux. The result was 
 that the . m.m, f. of armature reaction was completely op by that 
 
 of the other coils and a flux from the series and amort is seur wind- 
 ings threaded through the shunt coils and induced an e.rn.f. therein. 
 This e.m.f. was in opposition to the impressed e.m.f. and so the cur 
 rent in the shunt field dropped very rapidly to zero. When the D.C. 
 current passed through its maximum and began to decrease the m.m. f. 
 of the series coils decreased until, when the D. C. current was a. 
 minimum, the shunt field current was a maximum. In this position 
 the current through the series coil is almost zero hence there is 
 little effect from the series coil in aiding the shunt coil. As thi 
 . f. decreased an e.m.f. was generated in the shunt field coils by 
 
 transformer action tending to aid the impressed e.m.f. This' was fur- 
 ther increased by the similar effect produced by the armature re- 
 action. Although the D. C. current had decreased to almost zero the 
 A. C. current . Increased very materially, . i t, to a peree, 
 so that there was an armature reaction produced. This was in a di- 
 rection tending to demagnetize the shunt field so, by transformer ac 
 tion, it toe acted to increase the shunt field current. At X, the 
 D. C. current began to increase and the A.C. current began to de- 
 crease so that the series coils were excited, the armature reaction 
 caused by the D. C* current grew and the cycle just explained repeatei 
 When the short circuit was impressed the D.C. voltage fell to 
 a value of about 60 volts. This meant a much lower impressed e.m.f. 
 on the shunt field so chat its m.m.f. began to decrease at a rate 
 
 depending on the impedance of the shunt field circuit. As the m.m.f 
 decreased, the m.m.f. of the series field, when the D.C. current was 
 I naxi um, was large enou h to induce an e.m.f. in the shunt field 
 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
coils that actually caused the shunt field current to become nerativ 
 In the limit, or at stand still, there would be no m.m.f. produced 
 in the shunt field as there was no D.C. voltage generated and the 
 impedance of the shunt field is so high that it would allow only a 
 very small A. C. current to flow. There would be A.C. current flow- 
 ing from the brushes through the series field and this would induce 
 a normal frequency voltage and current in the shunt field. In oth- 
 er words, the shunt and series field coils would form a step-up 
 transformer and unless the shunt field circuit was opened the insu- 
 lation might be broken down due to high induced voltages. 
 
 As noted before, the D.C. voltage decreased to a value deter- 
 
 .R. _ ■ the external short circuiting connec- 
 tors. The terminal D.C. voltage never became negative e, a 
 
 long as conductors cut through a magnetic field that was stationary, 
 there was a positive D.C. voltage generated. This voltage increased 
 and decreased with the rate of cutting lines of force ana was a 
 maximum when the phase displace:’ ent was zero and minimum when it was 
 
 . i 
 
 180 electrical degrees. The pulsations in the D.C. voltage followed 
 those in the D.C. current and in the limit or at rotor stand still 
 would ce z-rrc . 
 
 ’• e last curve to be considered is that of the impressed . 
 
 '' tge. After the short circuit was applied the i 3 s.m.f. 
 
 continued to create a magnetic field which rotated at synchronous 
 speed. If, then, the armature decreased in speed the armature con- 
 ductors cut through this revolving field and generated an e.m.f. in 
 the a 2 conductors. The frequency of this e.m.f. was the same 
 
 as the frequency indicated in tne direct current curve and it was 
 superimposed on the impressed A.C. voltage of the slip rings. The 
 
43 
 
 distance between X and Y on oscillogram 15a is equal to the length 
 
 f the 50 cycle wave. hence the frequency of XY 
 SO per second, and is superimposed on a wave cf 60 cycles so that 
 the characteristics of a third harmonic wave should appear in the 
 voltage wave. This is lally the case in the voltage wave direct- 
 ly above. Then, if the speed of the rotor still further decreased, 
 so that the armature conductors were cut by the revolving field at 
 30 cycles, it would be expected that the voltage wave would nave 
 second harmonic characteristics. The two points of minimum D. C. cur 
 rent following Z are about two 60 cycle periods apart and the volt- 
 age wave immediately above has very decided second harmonic charac- 
 teristics. In the limit, or at rotor stand still, the revolving 
 A. C. field should cut the conductors on the armature SO times per 
 second ana the A.C, impressed voltage should retain its 60 cycle 
 sine characteristics. This is the case in a transformer with open 
 circuited secondary. It seems, therefore, that a synchronous con- 
 verter on short circuit has some of the characteristics of a short 
 circuited r .C. generator, synchronous motor, and induction motor, as 
 well as induction generator, and transformer. 
 
 I 
 
44 
 
 IV CONCLUSION 
 
 From a consideration of the oscillograms it is evident that any 
 electric machine is a very closely interconnected system of electric 
 and magnetic circuits. The change, or disturbance of any one f hes« 
 circuits results in the change of all the others in such a direction 
 that they tend to come into stable equilibrium. The smallest change 
 in the electric or magnetic circuit of the machine, either internal 
 or external, causes such a change to a degree depending on the magni- 
 tude of the disturbance. The application of a load to an electric 
 generator causes a change in the armature current, armature m.m.f., 
 field current, field m.m.f., generated voltage, and flux distributer 
 in poles, pole faces, and armature. As the machine increases in tem- 
 perature, its electrical resistance changes and there results a ! 
 change in the electric and magnetic quantities. 
 
 In the short circuit te ie on D. C. generators driven by 
 
 synchronous motors the armature current cf the generator rose to six 
 cr seven times its full load value. This would have been larger had 
 the short circuit been applied directly to the brushes but conditions 
 were made so that they represented, as nearly as possible, those ob- 
 tained in the laboratory. The terminal voltage dropped to approxi- 
 mately ZQfjo of its normal value. It, too, would vary with the point 
 on the line at which the short circuit was applied and would be zero 
 for a short circuit across the brushes. The field current attained 
 a momentary value of approximately four times its normal value. This 
 could be varied by over compounding; the generator, changing the series 
 coils so as to establish a greater flux from the series field and so 
 neutralize the m.m.f. of armature reaction. The values obtained 
 agreed very well considering the great difference in the generator 
 capacity and generator design. 
 
! 
 
 I 
 
 
 
 
45 
 
 In all instances where a short circuit was applied tea generate: , 
 there was a change in the flux distribution at the field pole face 
 caused by an actual shifting of the lines of magnetic force. When a 
 current flows through a conductor that is moving in a magnetic field 
 there is a tendency for the lines of force to accumulate on one side 
 of the conductor ana be neutralized on the other side. As the lines 
 are supposed to be in longitudinal tension and lateral compression 
 they are carried along by the conductor until the longitudinal and 
 lateral strength limits are passed. The lines then tend to redis- 
 tribute themselves over the pole face with their original no load 
 distribution. Other conductors carrying current then produce a re- 
 action on the lines of force and keep them crowded to one pole tip. 
 When tne current ceases to flow in the conductors the lines of force 
 will at once assume no load distribution. This is accomplished be- 
 cause of tne lateral compression of the lines. As this movement of 
 the lines of force is in a direction opposite to the rotation of the 
 armature, a high momentary voltage results in the armature. 
 
 In the short circuits on the A.C. generators, the aximum cur- 
 rent reached in the armature was about 12 to 15 times full load cur- 
 rent. In the stable condition tnis settled down to about three times 
 f'dH load currenu . 1 ne voltage between lines fell to a low value 
 
 and continued to decrease as the armature current decreased. Both 
 current and voltage attained during short circuit depend on the 
 constants of the short circuit path. Approximately the same conditions 
 were used in the short circuit on A.C. and D.C. machines, so that a 
 Short circuit directly at the brushes of the machine would have 
 caused a higher value of armature current. The field current of the 
 is tested had peak values of six and seven times full lead 
 
46 
 
 value, just after the short circuit was applied Out the pulsations in 
 the three phase short circuit gradually died out as toe conditions 
 became steady. 
 
 In the single phase short circuit the armature current mounted 
 again to about ten or twelve times full load value and the maximum 
 positive peak in the field current was about six or seven times full 
 load value. The armature currents became steady at about four or f iv< 
 tines full load current. The terminal voltage between a live line 
 and a short circuited line showed distinct evidence of third harmon- 
 ics caused by armature reaction. 
 
 In a D.C. short circuit on a synchronous converter that pulled 
 cut of step, the characteristics of the transient are, first those 
 of a short circuited D.C. generator and then those of a synchron- 
 ous motor. These sets of characteristics alternate with increasing 
 frequency until the D.C. generator characteristic has disappeared 
 and the rotor is stationary. In this condition the machine has ten- 
 dencies to motor as an induction motor yet acts as a step up trans- 
 former, the armature and shunt field forming the two coils. 
 
 Whenever a short circuit was applied to an electrical machine, 
 the energy of the short circuit was supplied, for a short time, by 
 the kinetic energy of the rotating armatures of motor and generator. 
 This was true for all motor generator sets whether the coupling be- 
 tween the machines was flexible, semi-rigid, or rigid. It was net 
 until the a r nature speed decreased that the motor supplied any short 
 circuit power. 
 
 The writer desires to express his appreciation to Professor 
 E.B. Paine ana Mr. C.A. Keener for their interest in the work and 
 for tneir valuable surges t ions. 
 

 
 
 
 
 
 
 
 
 
 
^7 
 
 Westinghouse E>5 K.W. Set. Short Circuit, Old Breakers. Generator. 
 
I. 
 

 
 
 

 
 Weatinghouse $5 K.W. Set . Short Circuit, Automatic Breakers. Generator 
 
Westinghouse S5 K.W. Set. Short Circuit, Automatic Breakers. Motor. 
 
SJ 
 
 Weatinghouse S5 K.W. Set. Short Circuit, Old Breakers. Generator. 
 

 ■ 
 
 / 
 
 
 
 
 
Westinghouse 3 5 K.W. Set. Short Circuit, 160 ampere fuse. Generator. 
 
Wes tinghouse 2>5 K.W. Set. Short Circuit, 160 ampere fuse. Motor. 
 

 
c! 
 
 SB 
 
 Westinghouse $5 K.W. Ret. Short Circuit. 60 ampere fuse. Generator, 
 

 
 
Weatinghouse S5 K. f. Set. Short Cirauit. Field Flux Shift. Generator. 
 

sst inghouse 9*5 K.W. Set. Short Circuit. Generator. 
 
■\ 
 
 
Westinghouse 9.5 K.W. Set. Short Circuit. Motor. 
 

 
General Electric 15 K.W. Set. Short Circuit. Generator. 
 
Northern Electric 6l 5 K.W. Ret. Short Circuit. Generator. 
 

 
 
stinghouse 6 K.W. Set. Short. Circuit. Generator. 
 
Westinghouae 6 K.W. Set. Short Circuit. Motor. 
 
General Electric 15 K.W. Set. } Phase Short Circuit. Phase Current 
 
General Electric 15 K.f. Set. 3 Phase Short Circuit. Generator 
 

Westinghouse 9.5 K.W. Set. 3 Phase Short Circuit. Phase Currents. 
 

 
Westinghouse 9*5 K.W. Set, 3 Phase short Circuit. Generator. 
 

 
General Electric 15 K.W. Set. Single Phase Short Circuit. Phase Currents. 
 

General Electric 15 K.W. Set. Single Phase Short Circuit. Generator. 
 
Westinghouse 9.5 K.W. Set. Single Phase Short Circuit. Phase Currents. 
 
Westinghouse 9.5 K.W. Set. Single Phase Short Circuit. Generat 
 
Westinghouse 10 K.W. Synchronous Converter. D.C. Short Circuit. 
 
Westinghouse 10 K.W. Synchronous Converter. D.C. Short Circuit.