Global sensitivity analysis of probabilistic tunnel seismic deformations using sparse polynomial chaos expansions

摘要: Abstract There is an increasing interest in utilizing surrogate models to perform the global sensitivity analysis (GSA) for uncertainty quantification since they allow quantify efficiently relative importance of each input parameter with relatively small computational costs. This paper aims at investigating feasibility applying this method a stochastic soil-tunnel system seismic conditions. An advanced technique named sparse polynomial chaos expansions (SPCE) introduced build model that allows performing GSA combined Sobol’ indices. The accuracy and efficiency SPCE-GSA are validated by comparison true numerical predictions Monte Carlo simulations. Parametric analyses performed wide range soil shear wave velocities, ground motion intensities, probability distribution types, coefficients variation, assuming nonlinear behavior. influences sampling size, degree, characteristics on variability indices evaluated. results show velocity modulus reduction factor two variables significantly affect deformations tunnels.