Identification of Defect Size in Rolling Element Bearing using Audible Acoustic Sound Signature

Abstract

Prognostics predict the remaining useful life of machine with the use of current machine condition and future operating condition. Bearing is essential component in rotating machine. Defect in bearing leads to break down of machine. Condition based monitoring is used for finding RUL of bearing. In this experimental work, artificial defect is created on 6205 deep groove ball bearing with EDM on inner race and outer race of different defect size. Acoustic based method is used for finding defect size of bearing. Sound signature coming from faulty bearing is captured using microphone during experiment. Faulty sound signal is decomposed using discrete wavelet transform and this decomposed signal splits the sound signal into peak corresponding to ball entry into the fault and exit out to fault. Experiment is conducted for outer race defect and inner race defect. Actual size of defect is measured with optical microscope. Defect size obtain by this acoustic method is quite similar to that of actual size for 2mm and 1mm defect size while some difference in 0.3mm defect size.