Soil–structure interaction analysis in natural heterogeneous deposits using random field theory

Inherent variability of soil properties induces inevitable sources of uncertainties in geotechnical analyses. Hence, deterministic evaluation of geostructure performance may lead to undesired unconservative outcomes which in turn may result in the system failure. In this study, a numerical simulation approach is employed to examine the influence of variation of different random soil properties on the soil–structure interaction phenomenon as well as the subgrade reaction coefficient within the shallow foundation analysis framework. Implementing Monte Carlo simulations along with the random field theory into the finite difference numerical iterations, several probabilistic analyses are performed in an undrained condition. The results of stochastic simulations illustrate the significant influence of incorporating the spatial variability of index soil properties on the response analysis of shallow foundations above heterogeneous soil strata. In particular, heterogeneity of soil layer is observed to bear a remarkable role in the evaluation of subgrade reaction coefficient. Adopting the results of numerical analysis, it is observed that as the coefficient of variation and as a result, heterogeneity of soil layers increases, the mean subgrade reaction coefficient decreases. The descending rate of mean subgrade reaction coefficient decreases as the scale of fluctuation increases and the soil behaviour tends towards the homogeneous state.