Investigating sound absorption in rail tunnels using wave decomposition

Abstract

Acoustic noise in trains is a more prevalent problem in tunnels as compared to open track scenarios. This is mainly due to less acoustic radiation relative to increasing contributions of reflections and reverberation. Sound absorbing panels on a tunnel wall and in the track four-foot perform better above 1 kHz with absorption coefficients larger than 0.8. To investigate sound absorption below 1 kHz, we employ wave decomposition into incidence, reflection and absorption components for a section of a given underground tunnel design. A Finite Element (FE) 2D model of a carriage and tunnel is developed, representing a tangent portion of a rail track and including the noise power spectra from the rail-wheel interactions for three different roughness scenarios. The FE model, compared to a Ray Tracing one, provides precisely imposed boundary conditions and the pressure field of the entire tunnel interior. Our results can identify the performance of current panels, absorbing significantly less noise power in the lower frequency range, especially within the 0.3–0.5 kHz interval. The insights from wave decomposition analysis can lead to solutions to increase absorption by changing the reflection pattern below 1 kHz band, improving the passenger comfort during a longer train ride.