Diagnosis Technology Investigation on High-Speed EMU Axle-box Bearings with Local Damages

The EMU axle bearings’ damage can cause increasing of both the vibration and temperature of the shaft. Temperature monitoring is used for fault monitoring, but it often leads to false message. So the EMU axle bearing fault diagnosis based on vibration signals have important applications. In this paper, diagnosis investigation based on a bearing experiment rig is introduced. The tests indicate that because the fault locations are not fixed but varying with the rotating inner race and rolling element, all orders of fault characteristic frequencies are very complex due to the frequency modulation when the local damage happens on the inner race or the rolling element. It is hard to detect the local faults by the conventional FFT processing. So the SR spectra are adopted, based on the stochastic resonance method, to effectively extract the fault characteristic frequencies. Based on the above researches, a monitoring and diagnosis system was developed, which can be used for early warning and judgment for the axle bearing fault of the bullet trains.