Rotor fault detection of electrical machines by low frequency magnetic stray field analysis

Olivier Chadebec (1) , Viet Phuong Bui (1) , Pierre Granjon (2) , Laure-Line Rouve (1) , Nicolas Le Bihan (2) , Jean-Louis Coulomb (1)

1 LEG - Laboratoire d'électrotechnique de Grenoble
2 LIS - Laboratoire des images et des signaux

This paper shows the reliability of fault detection on electrical machines by analysis of the low frequency magnetic stray field. It is based on our own experience about magnetic discretion of naval electrical propulsion machine. We try to apply the techniques developed in previous works on the subject to faults detection. In this paper we focus on rotor defaults in a synchronous generator (eccentricity and short-circuit in rotor). Two kinds of study are performed. The first one is numerical. Firstly, an adapted Finite Elements Method is used to compute the stray field around the device. However, this approach is difficult to apply to fault detection and not well-adapted. A new model, simpler and faster, is developed. Results are compared for both modelling. The second one is experimental and is driven thanks to a laboratory machine representative of a real high power generator and to fluxgate magnetometers located around the device. Both studies show good agreement and demonstate the reliability of the approach.

Domaines: Autre

Olivier Chadebec, Viet Phuong Bui, Pierre Granjon, Laure-Line Rouve, Nicolas Le Bihan, et al.. Rotor fault detection of electrical machines by low frequency magnetic stray field analysis. 6th IEEE International Symposium on Diagnostics for Electric Machines, Power Electronics and Drives (SDEMPED 2005), 2005, Vienna, Austria. ⟨hal-00021533⟩