Incipient Faults Detection In Rolling Element Bearing Using Vibration And Acoustic Based Method

Abstract

Health evaluation of rolling element bearings is of critical importance in reliable operation of rotating machinery. If a defect in such a bearing, not detected in time, may even lead to catastrophic failure of the machinery. The defect detected in the initial stage of the bearings may save considerable capital of any industry by felicitating proper maintenance scheduling. The present study aims at detecting the fault occurring at two specific locations- inner race and outer race of the bearing, at its initial stage of occurrence. The hairline cracks of various dimensions have been artificially induced to simulate the occurrence of a crack in its initial stage. The faulty bearings were tested for various operating conditions like variation of radial load and operation speed in the test rig available with the department. The response from the bearings under test was recorded in the form of vibration signals and acoustic signals using proper sensors. The signals were processed using FFT, Hilbert Transfer function and Spectral Kurtosis based Envelope analysis. The conclusion from the study has been drawn in terms of detectability of the faults under various operating condition and suitability of signal processing techniques for the same.