Quantitative analysis of a short-circuited laminated high speed (faulty) switching reluctance motor yoke using a multi-harmonic multi-zone predictive loss modeling

This paper relates the principle and the development of a multi-harmonic multi-zone method in order to predict iron losses in short-circuited laminated yokes in a switching reluctance machine. Default like short circuit increases iron losses in machine yokes due to eddy current circulation. The method proposed to estimate losses thanks to a unidimensional field zone division and a Fourrier analysis of magnetic flux density extract from Finite Element computation and to apply those data on short-circuited laminated ring core. The ring core measurements are used to estimate iron losses in the machine with a relative error of 12.7% at 60000 rpm. In an industrial context of fault analysis, such a relative error is acceptable to predict losses in case of electrical insulation failure.