Detecting mechanical property anomalies along railway earthworks by Bayesian appraisal of MASW data

Abstract

The techniques traditionally used to estimate the mechanical properties of railway earthworks (RE) are costly and of low performance. There is a great need for the development of non-destructive methods, which would allow a fast and efficient diagnosis of RE. Seismic surface-wave (SW) methods, also known as multi-channel analysis of surface waves (MASW), were rescaled here for there systematic implementation on RE. A test site was chosen along the North European High Speed Line (HSL) for its well-known correspondence between discontinuity in mechanical properties and anomalous maintenance forces. With the proposed acquisition and processing strategy, a contrast in SW propagation velocity was correlated with the change in shear modulus of the soil layers beneath the track, previously evaluated in the laboratory. Bayesian inversion of the SW data also allowed to integrate the strong a priori knowledge available on these HSL lines and to provide quantitative results with a confidence index to help end-users in their decisions.