Influence of slope property variabilities on seismic sliding displacement analysis

In this study we quantitatively examined how the variabilities of slope property parameters influence the seismic slope displacement predictions based on two commonly used methods, namely the Newmark's rigid-block and the fully coupled deformable methods. A suite of 20 acceleration time-series were selected as input motions, and Monte Carlo simulations were performed to account for the influence of slope parameter variabilities. The results show that, for both Newmark's and fully coupled analyses, modeling the variability of the effective friction angle phi significantly increases the geometric mean <(D)overbar> and standard deviation ai sigma(lnD) of the resultant displacements, while modeling the variability of the other slope parameters (i.e., soil cohesion c, thickness, and water table level) results in a similar estimate of <(D)overbar> and a slight increase of sigma(lnD). The other sources of uncertainty exist in fully coupled analysis are the characterizations of the average shear wave velocity V-s and the nonlinear soil properties. Modeling the variability in nonlinear soil properties yields a reduced <(D)overbar> estimate, and modeling the V-s variability causes a slight reduction of <(D)overbar>. Also, incorporating the variability of slope property parameters in fully coupled analysis consistently increases sigma(lnD), in which the variation of phi plays the predominant effect. This study thoroughly quantified the influence of slope property variabilities on the computed displacements, which could help engineers in addressing the uncertainty issue in seismic slope displacement analysis.