Importance Sampling‐Enhanced Reliability Analysis of Double‐Row Piled Earth Slopes Accounting for Soil Spatial Variability

Abstract

The application of stabilizing piles to increase slope stability is a common practice in slope engineering. Most related studies primarily focus on single‐row stabilizing piles; however, for large‐scale slopes, single‐row stabilizing piles may fail to meet the stability requirement, necessitating the use of double‐row or even multiple‐row stabilizing piles. Specifically, for large‐scale earth slopes where double‐row stabilizing piles are installed, the uncertainty and spatial variability of soil properties in the slope region may exert some probabilistic influence to the stabilizing efficiency. To address this probabilistic issue, this paper presents a reliability‐based analysis framework enhanced by importance sampling (IS) to quantify the stability of double‐row piled earth slopes, accounting for the soil spatial variability. The reliability analysis is efficiently realized by the first‐order reliability method (FORM) with iHLRFx iterative algorithm and the accuracy is enhanced using the IS method. Additionally, the Karhunen–Loève (KL) expansion method is used for efficient random field simulations. In connection with the standard Optum G2 slope stability modeling program, the integrated KL‐iHLRFx‐IS reliability method is then used to calculate reliability indices of double‐row piled earth slopes with various combinations of piling design schemes, followed by probabilistic sensitivity analyses of various soil uncertainty and spatial variability parameters.