Motor bearing damage detection via wavelet analysis of the starting current transient

Abstract

Preventive maintenance of induction motors plays an important role in avoiding expensive shut-downs due to motor failures. Motor Current Signature Analysis, MCSA, provides a non-intrusive way to assess the health of a machine. In this paper, the starting current transient of an induction motor is analyzed via discrete wavelet transform to detect bearing faults. The frequency subbands for bearing pre-fault and post-fault conditions are compared to identify the effects of bearing/machine resonant frequencies as the motor starts.