Fault feature extraction of rolling bearing considering slippage influence based on a dynamic model

Abstract

Bearing spalling, pitting and other local faults are one of the common bearing faults, which are quite difficult to detect in the early stage. Fault characteristic frequency is the most widely used in fault diagnosis. However, bearings are likely to slip during operation, which will result in the deviation between theoretical and actual fault characteristic frequencies. This paper proposes a dynamic model of a defective rolling bearing considering slippage to evaluate the fault characteristic frequency. The interactions among inner ring, rolling body, cage, and outer ring, as well as the time-varying displacement excitation of the outer raceway spalling are considered in the constructed dynamic model. The effects of slippage on the fault characteristic frequency at different speeds and loads are investigated using the proposed dynamic model, and an experiment was conducted to validate the proposed model. The results show that the actual fault characteristic frequency will be smaller than the theoretical fault characteristic frequency at high speed and light load. The proposed model provides a new method for modelling bearing dynamics and a theoretical basis for monitoring and diagnosing bearing faults.