ABSTRACT
Recurrent train-induced vibrations can affect the long-term stability of rock slopes located in proximity of railways, where rock falls can cause high-risk conditions for transit and passengers. Seismic monitoring of the unstable rock walls was experienced as a risk-mitigation strategy, in order to detect variations in the vibrational behaviour over time that can be related to microcracking able to modify the pre-existing crack-net. This effect, known as “rock-mass damaging”, can justify changes of rock mass mechanical parameters and consequently of rock mass rheology and they can lead toward slope failures (i.e. rock falls or slides). By deriving mean ambient noise levels from signals recorded over a representative time window and by comparing time histories and related Fourier spectra, potential changes in the rock mass vibrational trend can be observed. Moreover, by spectral analyses of seismic ambient noise records it is possible to distinguish the contributions to vibration due to natural sources (generally generating low-frequency signals), including wind, thermal effects or earthquakes, respect to the ones induced by artificial sources (generally generating high-frequency signals), including the train transit.
