Defect quantification evaluation of a rolling element bearing based on physical modelling and instantaneous vibration energy investigation

Abstract

The estimation of defect size of rolling element bearing (REB) is expected to remain a significant and challenging task in vibration-based condition monitoring of rotating machinery. The conventional approaches focus on extracting the time spacing of the double impulse signal to estimate the defect size of the REB which is subsequently found to be inaccurate and arbitrary. In this paper, a novel model is developed to estimate the defect size of the REB based on the theories of physics and kinematics, in which the multi-event excitations, i.e. the entry, destressing and collision generated by the rolling element-defect interaction, are considered and investigated. The kinematic and geometric contact mechanism, which is mapped onto the rolling element-defect interaction as the rolling element passes over the defect zone, are analysed and modelled as a function of the time information. In view of the fact that the variation of the vibration energy contained in the multi-impact vibration response is deeply mapped to the time information, a novel idea is proposed to extract the time information by analysing the instantaneous energy variation of the defect-induced impulse using the scheme of the autoregressive model and the smooth pseudo Wigner–Ville distribution. The proposed model and idea are validated by experiments under different defect size and rotational speed scenarios. The results calculated by the proposed model show good agreement with the real defect size. Experimental and modelling comparisons show that the proposed model has reliable effectiveness and good performance and confidence in quantifying the size of the defect area localized on the outer raceway of the REB, and can provide some theoretical guidance for damage detection and remaining useful life prediction of the REB.