Monitoring of rail/wheel interaction using acoustic emission

Abstract

The work presented in this thesis is related to the condition monitoring of rail-wheel interaction using Acoustic Emission (AE), the principle being that both normal and abnormal rolling give rise to AE, features of which are related to the mechanical intensity of the interaction and hence the stress range (or stress intensity factor range) to which sections of track is being exposed. Most of the work was carried out on a model wheel running on a model circular track, which was first characterised using a simulated source before studying the wave propagation from a continuously moving (wheel) source. Using a number of sensor arrays placed on the track, primary wave propagation characteristics such as wave speeds and attenuation coefficients and also secondary wave propagation characteristics such as reflection and transmission from and through the joint. A high speed camera was used to confirm, that wheel slip does not occur at the wheel speeds and loads of interest. A simple analytical model was derived using the measured wave propagation characteristics which describes the expected AE recorded at a track-mounted sensor as the wheel approaches and recedes. Using the analytical model, the effect of increasing wheel speed and axle load on the normal rolling signal was measured. Wheel rattling was observed, particularly at lower wheel speeds and loads, and this was eliminated in some trails by introducing a spacer. The effect of minor track defects and wheel flange rubbing on the track was also studied, where the comparison of the expected normal running signal with excursions above background allowed the locations of track defects to be identified. Finally, a set of experiments were carried out with simulated wheel defects. The signals were analysed using the principle of demodulated frequency resonance and matching to the expected pulse train spectra. . The findings of the work, along with a limited set of field tests on actual train-track interactions allows recommendations to be made for the deployment of sensors for cumulative damage monitoring on critical areas of track.