ALT KK\. \Tnii REGULATION AM) <H>EKATION 131 



produced by the armature current, the power represented by 

 this loss must be supplied by the armature current. The eddy- 

 current loss varies as the square of the flux density and the 

 resis loss varies as the !.(> power of the flux density. As 

 the leakage flux is nearly proportional to the current, the eddy- 

 current loss varies as the square of the current and the hyst- 

 loss as the 1.6 power of the current. The combined loss varies 

 nearly as the square of the current. 



The effect of these local iron losses is to increase the total 

 loss due to the flow of current through the armature. As these 

 local losses vary nearly as the current squared, their effect is 

 practically the same as if the resistance of the armature were 

 increased. 



Unless the armature conductors are small in cross-section, the 

 effect of the slot leakage flux is to force the current towards the 

 top of the slot, so that the current density in the portions of a con- 

 ductor near the top of the slot is greater than in those portions 

 near the bottom of the slot. This also increases the effective 

 re-istance of the armature. 



The effective resistance of an armature is therefore greater 

 for alternating than for direct current, due to the alternating flux 

 which accompanies the flow of the alternat ing current. The per- 

 N depends to a large extent on the shape of the slots 

 and the teeth and on the gfce <f the conductors, and ranges from 

 nt. As the armature resistance drop is very small 

 inpared with the voltage drops due to armature reactance 

 and armature reaction, considerable error in determining the 

 tnoe introduces little error in mo<t computations. The ef- 

 armature resistance may be measured by running the 

 inc as a generator, with weak field excitation. The input 

 is measured with tin- armature open-circuited and then short- 

 circuited through ammeters. ing the change in CO1 

 CftUfled by armature reaction, the difference of input divided by 

 the number of phases is equal to the / /,' foflfl per phase. The 

 r pha>e is found by dividing \\\\- difference 

 of input by th<- current per p liared. A M uuon. 



thoiiuh le<s arelirate method, i- to mea>lire the olmiie resMMiire 



with din '>ii- value by an estimated 



factor, such as -I' 1 the in<l i lo<>es. 



