Survey of basic stray losses in squirrel-cage induction motors

A study of the literature reveals that there is still considerable confusion on various aspects of stray losses, particularly with regard to their definition and origin. Accordingly the theoretical information is presented in a rationalised and systematic way to obtain a fully comprehensive approach, based on classifying stray losses into two groups, namely those due to main flux variations (stray no-load losses) and those due to leakage fluxes (stray load losses). In this way three and six main basic types of stray losses can be recognised under the two headings, respectively. Having identified the physical origin of these stray losses, their effect on motor design parameters can be ascertained qualitatively, and available methods of quantitative calculation, which are mostly semiempirical, can then be judged. A critical appraisal of two synthetic test methods for stray load losses, the reverse-rotation and d.c./a.c.-short-circuit tests, is made in the light of basic theory, and the fundamental deficiencies of these tests, particularly the reverse-rotation test, are demonstrated theoretically and by tests carried out on a number of medium-to-large induction motors ranging from 400hp, 4pole, to 1700hp, 8pole. The importance of insulating squirrel cages to avoid circulating-current losses deduced theoretically is confirmed by experiment. Additionally, a direct investigation into squirrel-cage circulating-current losses on a 1500hp 4pole motor shows how such losses can be controlled by a suitable current-displacement rotor design; these additional losses being directly measurable at no load on this machine with open stator slots. Consideration is also given to the fact that modern, high-efficiency (particularly totally enclosed) machine designs cannot cater thermally for substantially increased losses, e.g. excessive stray losses of any kind. There is therefore already an overriding and definite safeguard that the declaration of efficiency to BS 269 is substantially c- orrect, or, more specifically, that there cannot be any major error in the value of the declared losses. However, a modification of the fixed allowance for stray load losses is suggested. From the theoretical and practical investigations carried out, it is concluded that making a nominal allowance for stray load losses is at present the only reasonable and practicable method of declaring efficiency.