ARMATURE LAMINATIONS 



15 



with the "residual deflections" produced by the samples having 

 known and stated amounts of hysteresis. In practice, therefore, 

 the method is absolutely empirical, and 

 is free from all theoretical assumptions. 

 In Fig. 10 is shown a scale calibrated 

 in the manner described. 



The deflection of the pointer to one 

 side should be read, and, again, the 

 deflection to the other side when the 

 direction of rotation is reversed. The 

 average of these two readings gives a 

 value proportional to the hysteresis 

 loss. The deflection of the instrument 

 is fairly independent of the speed at 

 which the carrier revolves. If the rate 

 is very slow, the magnet will show each 

 individual impulse which it receives as 

 the ends of the sample pass its poles ; 

 but when the speed is sufficiently in- 

 creased, these impulses blend together 

 into a steady deflection. A needlessly 

 high speed should, of course, be avoided, 

 as the windage set up would interfere 

 with obtaining a correct reading. 



For the test sample, as many strips 

 are to be - taken as will make up a 

 weight approximately equal to the 

 weight of one of the standard samples. 

 No error is introduced by reasonable 

 variation in this respect, and hence no 

 exact adjustment of the weight is 

 necessary. The length of the strips 

 must, however, be exactly 3 inches. 

 The number of strips required for the 

 test sample will, in general, be six or 

 seven pieces, i.e. for the usual gauges of 

 transformer iron. The most customary 

 thickness of each sheet is, as already 

 stated, 0*5 mm. The magnetic induc- 

 tion is practically the same in all the 

 specimens, notwithstanding the differences in the permeability of the 

 iron; this is on account of the comparatively large air-gap between 



