Necessity of condition monitoring for estimating the residual lifetime of rolling element bearings

Rolling element bearings are the most widely used components in rotating machinery. Estimation of their remaining useful life in order to increase the reliability and availability of them is a critical issue in the field of condition monitoring of these machinery. The defect propagation in these components is inherently stochastic. This factor causes that their remaining useful life prediction be a challenging problem. In this paper a method to predict the defect area of rolling element bearings is proposed using RLS adaptive algorithm and fatigue damage mechanics approach. First, two models are developed to correlate the statistical features of vibration signals and the defect size on inner and outer races. After construction of diagnostic models, healthy bearings are tested under operating conditions and the generated vibration signals are collected from the time of initial defects until the final failure of the bearing. Using RLS adaptive algorithm, diagnostic models, and the extracted vibration signals, the parameters in the mechanistic model are updated and the area of defect can be predicted. The results of experimental tests show that this approach can determine the defect area on both inner and outer rings of roller bearings and as a result, their remaining lifetime can be predicted reliably.

Necessity of condition monitoring for estimating the residual lifetime of rolling element bearings Keywords:

rolling element bearing , adaptive algorithm , diagnostic models , remaining lifetime , defect size