HOBART AND PUNGA'S METHOD 177 



102. Hobart and Punga's Method 



What promises to be an extremely convenient method of carrying 

 .nit the heat test of a large generator has recently been described by 

 Hobart and 1'unga.* The method consists essentially in running the 

 machine alternately on open circuit and on short circuit, the periods 

 of the two sets of runs and the values of the exciting current 

 corresponding to them being so chosen that the mean values of the 

 various losses during the entire time of the test are equal to the 

 actual values of these losses at full load. 



The losses may be regarded as made up of (1) mechanical friction 

 losses ; (2) armature copper loss, w a ; (3) iron loss, wi ; (4) field 

 copper loss, or excitation loss w e . 



The friction losses are independent of the excitation, and their 

 value may be determined by driving the unexcited alternator at its 

 normal speed by means of an auxiliary standardized motor. The 

 armature copper loss is easily calculated, as is also the field copper 

 loss, for a given value of the current. The iron loss is found by 

 driving the alternator on open circuit by means of a standardized 

 motor, the exciting current having its normal full-load value ; if from 

 the total power employed in driving the alternator we subtract the 

 friction losses as previously determined, we obtain the iron loss. For 

 the purposes of the heat test under consideration, it is necessary to 

 know the relation connecting the iron loss with the excitation loss, so 

 that a set of readings must be obtained connecting these two, and 

 the results plotted in the form of a curve as in Fig. 125.f 



Let the alternator be run for t\ minutes on open circuit with 

 a certain excitation, then for t% minutes on short circuit with a different 

 excitation, then again for t\ minutes on open circuit, for t% minutes on 

 short circuit, and so on, the open-circuit and short-circuit runs 

 alternating with each other. During the open-circuit runs, there 

 are only frictional, excitation, and iron losses taking place in the 

 machine, while during the short-circuit runs we have only frictional, 

 excitation, and armature copper losses, the iron loss being negligible. 

 The machine running at its normal speed, it is evident that the 

 frictional loss is identical, all the time, with the full-load frictional 

 loss. The other losses will depend on the excitation, and the problem 

 is to so choose the values of the exciting current corresponding to the 

 open-circuit and short-circuit runs, and the relative values of t\ and t i} 



* Elfttrical World and Engineer, vol. xlv. p. 759 (1905). 



f On account of armature reaction, which produces u distortion of the field, the full- 

 load iron loss wi will generally be in excess of the open-circuit iron loss with the same 

 exciting current. We have already indicated ( 98) how this increase in the iron losa 

 may bo determined. 



N 



