Multiphysical modeling for fault detection in Permanent Magnet Synchronous Motors

The aim of this paper is to present an analytical multiphysical model of Permanent Magnet Synchronous Motors (PMSMs) dedicated to fault detection purpose. The electromagnetic aspect is based on the analytical calculation of the Maxwell pressure in the air-gap. An existing mechanical approach is improved to compute natural frequencies of the machine taking into account its different parts. A better agreement has been obtained between analytical results and those issued from a Finite Element Analysis (FEA) applied to two different stator structures. The dynamic displacement signal of the stator is finally estimated in the space-time domain. Different operating modes of the machine can be simulated by changing some model parameters. In this paper, the rotor eccentricity fault is investigated and detected by the analysis of the vibration signal. Therefore, the proposed multiphysical model could provide healthy and faulty characteristics and could be suitably used for fault detection and diagnosis tasks.