Wayside detection of faults in railway axle bearings using time spectral kurtosis analysis on high-frequency acoustic emission signals

Abstract

Typical railway wheelsets consist of wheels, axle and axle bearings. Faults can develop on any of the aforementioned components, but the most common are related to wheel and axle bearing defects. The continuous increase in train operating speeds means that failure of an axle bearing can lead to serious derailments, causing loss of life and severe disruption in the operation of the network, damage to the track and loss of confidence in rail transport by the general public. The rail industry has focused on the improvement of maintenance and remote condition monitoring of rolling stock to reduce the probability of failure as much as realistically possible. Current wayside systems such as hot axle box detectors and acoustic arrays may fail to detect defective bearings. This article discusses the results of wayside high-frequency acoustic emission measurements performed on freight rolling stock with artificially induced damage in axle bearings in Long Marston, UK. Time spectral kurtosis is applied for the analysis of the acoustic emission data. From the results obtained, it is evident that time spectral kurtosis is capable of distinguishing the axle bearing defects from the random noises produced by different sources such as the wheel-rail interaction, braking and changes in train speed.

Perspectives on railway axle bearing condition monitoring

“…The use of AE for axle bearing monitoring was also carried out by Papaelias et al and Amini et al [10,62,34,49,63]. They have demonstrated the use of acoustic emission on test trains and in-service trains.…”

“…Examples of the use of spectral kurtosis in bearing fault detection in railways can be found in [34,35,36].…”

“…Envelope analysis can be applied to signals within frequency bands showing high kurtosis in the SK, forming a hybrid technique between HFRT and SK which has been shown to be effective for detecting and diagnosing bearing faults [34,36,21]. Crest factor can also be used to make a decision using the HFRT spectrum; the largest…”

“…The application of integrated high-frequency acoustic emission with vibration analysis has been extensively studied for both on-board and wayside measurements [10,34,49,62,95,96,97,98]. This research in the UK was issued a certificate of acceptance, which allows trackside instrumentation [97].…”

Adaptive fault feature extraction from wayside acoustic signals from train bearings

“…These provide high quality information but require sensors and equipment to be fitted to every bearing on every vehicle. Acoustic emission [4] and hot axle box [5] detection systems can be fitted to one location in the network, but the former requires phyical access to the track, while the latter is only suitable for detecting late stage faults [6]. Wayside acoustic detection is another technology that is becoming increasingly popular because one monitoring station will observe multiple vehicles, no physical track access is required in order to install the equipment, and detection is at an earlier stage than its thermal counterpart.…”