Comparison of Vibration and Electrical Signal Analysis for Induction Motor Single-Point Bearing Defect Diagnosis

Abstract

Among many possible failure causes, bearing faults are reported to occur most often in induction motors (IMs). Signature-based and diagnostic analysis can be applied to vibration and electrical signals for early detection of the bearing fault. Vibration monitoring is widely used and has proven effective but relies on additional sensors that may undergo faults in harsh environments. On the other hand, the signature analysis of IM currents and voltages can be performed more easily as it requires the collection of electrical signals that are usually already available. However, their efficacy in detecting the fault signature may be lower compared to vibration analysis. In this paper, different vibration and electrical diagnosis techniques are applied to the detection of a single-point fault in the bearing of an induction machine. To this end, a small laboratory testbed comprising an IM with an artificially-damaged bearing is built. The results are compared to highlight the efficacy and limits of each method.