Probabilistic bearing capacities of strip foundation on two-layered clay

The bearing capacity of a shallow foundation on two-layered soil has been widely studied using deterministic methods, without considering the spatial variation within layered soils. In this study, the bearing capacity of strip foundations on two-layered clay considering spatial variability is investigated by using the random finite element method (RFEM). The results show that the embedment ratio and the spatial variability parameters of the upper layer could significantly affect the normalized probabilistic bearing capacities. Based on the random forest algorithm, a model is established to predict probabilistic bearing capacities, which can be applied as an easy-to-use method without carrying out RFEM analysis.