Evaluation of the seismic site response in randomized velocity profiles using a statistical model with Monte Carlo simulations

Abstract

Abstract Uncertainty associated with the soil properties inevitably affects the seismic ground response, which is unable to be properly assessed using a deterministic seismic site response analysis. In this paper, an attempt has been made to investigate the possible impact of the shear-wave velocity ( V s ) variation on the predicted ground motions using a real site in Italy. A statistically-based randomization model is adopted to describe the uncertainty of the V s profiles for various scenarios. The equivalent-linear method combined with the direct Monte Carlo simulation is employed to perform the stochastic seismic site response analysis. The results are presented in terms of amplification function ( AF ), spectral acceleration ( S a ), fundamental frequency ( f 0 ) and amplification factor ( F a ). The mean and standard deviation of these four factors are evaluated for a wide range of shear-wave velocity variations, interlayer correlations, and input motion intensities. This study also explores the accuracy of adopting truncated bounds as an approximate process of preventing the generation of unrealistic soil profiles. The results show that the variability of the predicted ground motion is significant in some situations and can be decreased by reducing the shear-wave velocity uncertainty.