Surface and subsurface attenuation of machine-induced vibrations through field experiment and ANN modeling

Abstract

This study examines attenuation characteristics of machine-induced vibrations at both the ground surface and different subsurface depths through large-scale field block vibration tests. The experimental results indicate that ground vibration amplitudes diminish with increasing distance and depth from the vibration source. Moreover, the vibration attenuation with depth becomes more pronounced at higher loading frequencies. Additionally, an artificial neural network (ANN) model is developed based on experimental data to predict the vibration attenuation behavior within the soil medium. The ANN model demonstrates excellent predictive performance and exhibits satisfactory alignment with reported studies, thereby validating its reliability. The proposed experimental-artificial intelligence (AI) based framework provides a useful approach to assess surface and subsurface vibration attenuation characteristics where conventional field testing is not feasible.