Computational Tools for Predicting Ground Vibration from Railways

Vibration in the ground causes problems. At large amplitudes, vibration causes damage, the most obvious example being earthquake damage. Seismic engineering is now a well-developed discipline but the ground motion in an earthquake is typically low-frequency (below 1 Hz) and large-amplitude (above 10 mm), and the computational tools developed to deal with this are not immediately transferrable to vibrations from railways. The latter generate higher frequencies (above 10 Hz) and lower amplitudes (0.01 mm). These vibrations rarely cause damage but they can nevertheless have significant social and economic consequences, and there remains a pressing need for predictive tools to guide design. This chapter reviews the various computational and predictive tools available to engineers. It summarises some of the available methods, such as the finite-element and boundary-element methods; it reviews particular models, which range from the purely theoretical to the purely empirical, and include combinations of the two; and it looks at practical considerations faced by practicing engineers, such as the impact of modelling assumptions and uncertainty