Bearing fault detection in induction machine based on stator current spectrum monitoring

Abstract

Bearing faults account for a large majority of the faults in a three-phase induction motor. Recently, many research activities were focused on the diagnosis of bearing faults by motor current signature analysis. However, the effective frequency band is not fully understood in terms of the diagnosis of the bearing fault. Moreover, the temporal change in motor stator current spectrum with faulty bearing has not been sufficiently investigated. The purpose of this paper is to clarify the characteristic frequency band and to evaluate the temporal change in the power spectral density of stator current by using powder contaminated bearing before complete halt of the motor. Experiments were performed with normal and powder contaminated bearings in the induction motor. The diagnosis technique for the powder contaminated bearing is discussed based on the experimental results.