The Testing of Transformer Steel

The object of this paper is to discuss the conditions which should be realized in the measurement of energy losses in sheet-iron and steel subjected to alternating magnetization, with a description of a modification of the Epstein method and apparatus. This method is believed better to satisfy the desired conditions and to give an accuracy of one per cent with the use of less than two kilograms (4.4 lb.) of material. Results are given showing a wide range in the quality of material in general use, quality having slight connection with price. Several foreign specimens of ordinary steel and silicon-steel are included for comparison with the American product. Silicon-steels contain from 3 to 4 per cent silicon, the quality not depending upon the exact percentage of silicon. By making measurements at two frequencies, the eddy-current and hysteresis losses have been separated and a study made of the variation of each with flux density. The values of the hysteretic constant and the total loss in watts per pound at 60 cycles and 10,000 gausses are tabulated. The effects of artificial aging are shown to depend upon the flux density selected for test, the hysteresis increasing more for 5000 than for 10000 gausses. Tests should therefore be made at the density to be used. The aging is usually negligible in silicon-steels.