A Comparative Study of Acoustic and Vibration Signals for Bearing Fault Detection and Diagnosis Based on MSB Analysis

Bearing is of a crucial part/element of all rotating machines. The main purpose of the rolling element in any rotating machine is to maintain rotating parts in their correct arrangement relative to stationary parts when under the action of radial and axial loads. The centrifugal pump is generally used to transfer fluid from one place to another and early detection of bearing faults allows proper action to be taken to maintain and/or improve performance and reduce lifetime cost. The focus of this study is to enhance the condition monitoring of the bearings in centrifugal pumps. The airborne sound signature and its remote measurement and analysis does away with the need to mount additional sensors as is the case of surface vibration measurement. However, a serious difficulty in detecting and diagnosing machine faults via traditional approaches is the high level of background noise, particularly with airborne sound signals. Thus, here the modulation signal bispectrum (MSB) which has ability to substantially reduce noise and detect nonlinear components has been used. The results obtained show that the MSB applied to airborne sound and the vibration signatures enhances detection and diagnosis of outer race bearing faults. Also, for detecting inner race defects, the results show that the diagnostic features obtained applying the MSB to the acoustic signal are more accurate and show a better separation between healthy and faulty cases than those obtained using the vibration signal.