Abstract

Transient events in railway rolling noise, such as the characteristic impulsive noise at switches and crossings, can significantly contribute to the perceived annoyance, despite being difficult to detect in the standard frequency-domain methods to analyse rolling noise. Studying these transient effects and their perception requires predicting the noise in the time domain. While several time-domain approaches exist for predicting the dynamic interaction of wheel and rail, predicting the associated rolling noise with adequate accuracy is computationally costly. The lack of a model for transient noise and the need for studying its perception was recently identified. Aiming for a comprehensive time-domain radiation model that includes the wheel and track contributions to rolling noise, this work focuses on the track radiation. The modelling approach taken here is based on a 2.5D formulation for the acoustic radiation and moving Green’s functions in the air. The computational cost, which lies mainly in the 2.5D BE calculations, is addressed by pre-calculating acoustic transfer functions. These transfer functions can be combined with different dynamic track models. Different dynamic track models in turn affect radiated sound field in different ways. Here, the sound fields produced by six different track models are compared, including different support types and analytical and numerical rail models. Several descriptors of the sound field are analysed. In terms of the radiated sound power and radiation efficiency, modelling the rail as a simple beam leads to similar results as elaborate numerical models up to about 5 kHz. In terms of the track-side sound pressure, simple beam models can provide similar results only up to 2.5 kHz. Euler-Bernoulli (E-B) beams seem unfit for time-domain predictions of the radiated noise as they over-estimate the bending wave speed at high frequencies. The results also show that the standard track decay rate (TDR) and the decay of acoustic sound pressure along the track are comparable.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call