Current signal analysis of an induction machine with a defective rolling bearing

Abstract

With the ultimate goal of rotating machinery diagnosis using Motor Current Signal Analysis (MCSA), this paper provides a coupled electro-magnetic-mechanical model of a rotating shaft supported by rolling bearings and driven by a three-phase squirreled cage motor. The modeling is based on the hypothesis that a bearing defect causes torque and then Instantaneous Angular Speed (IAS) variations associated to air-gap eccentricity of the induction machine rotor. Dynamic analysis of the multiphysic system highlights the sub-systems interactions, especially, angular periodicities and frequency modulations. The global model can be characterized by a unique set of non-linear state equations which are solved iteratively by an angle-step scheme while considering the angle-time relation. The major interest of presenting this model is that it allows to decrypt the transfer path from a small localized bearing defect until its manifestation on electrical signals. Analysis of bearing defects were performed by applying Fourier Transform on current per-phase signals.