Detection of bearing faults in mechanical systems using stator current monitoring

Abstract

Induction motors have been responsible for running mechanical systems in the industry for many decades. Their diagnosis still remains a hot quest for the researchers using various techniques. In this study, motor current signature analysis (MCSA) technique has been used to detect the faulty bearing installed in load machine (coupled to an induction motor). It has been seen that faulty bearings installed in load machines do not directly alter airgap eccentricity of an induction motor. In fact, these bearing faults affect the resultant torque of an induction motor. As modulating fault components show very low amplitude, these are usually masked by noise. This paper is devoted towards extracting features of faulty components efficiently from stator current using continuous wavelet transform. This methodology is assessed for detecting outer race faults in bearings installed in load machines using MCSA.