Fusion of Microphone and Accelerometer Sensing for the Identification and Measurement of Inner Race Defect

Use of uni-axial accelerometer is simplest and cost effective way to monitor vibration in a single direction. Situation in which, fault is rotating/changing its direction extraction of defect frequency is difficult from it. However, some of the bursts can be properly analysed. The acoustic sensor is another cost effective device which is generally omnidirectional and can be used in identification of bearing defect. In this paper, data from both the acoustic sensor and accelerometer are utilised and fused to have complete information for identification of defect type and estimation of its width. The present work utilizes defect identification using envelope demodulation of acoustic signal. Subsequent measurement of defect width is carried out from accelerometer signal. Continuous wavelet transform (CWT) of vibration signal is carried out using adaptive wavelet to produce 2D scalogram. Finally, time marginal integration (TMI) of CWT coefficient is performed for the measurement of defect width.