Validation of a railway rolling noise model on a system with slab track and resilient wheels

Abstract

This study investigates rolling noise emissions using an analytical model of the Seattle Sound Transit Light Rail system developed using Train Noise Expert (TNE) software. Measurements of track frequency response, decay rate, rail and wheel roughness were used to determine model input parameters. Wheel modal test results were used to characterize the wheels. Wayside noise and track vibration measurements during revenue service train passbys were used to experimentally validate the model. The average difference between measured and predicted overall passby LAeq noise at a position immediately adjacent to the track was 0.1 dBA (average from four scenarios, i.e. two surface track sites with two train types). Inspection of the predicted versus measured noise spectra indicates the model does not exactly match the measured spectrum in every frequency band, however the correlation with overall spectrum shape is good in the frequency bands that contribute most to the overall A-weighted level. It is concluded that the model provides an accurate representation of the most relevant physical rolling noise factors and is reliable with resilient wheels and light rail vehicles on ballast and slab track.