Publication

Abstract : Probabilistic shakedown analysis is performed on spatially varied cohesive soil medium under two-dimensional moving surface load in the present study by combining lower bound finite element limit analysis, random field modelling, and Monte-Carlo simulation technique. The efficiency of the analysis is substantially increased by substituting the linearised yield criterion with a second-order conic function. Log-normally distributed random field of undrained shear strength is generated by Karhunen-Loève expansion method. Probabilistic design charts are presented in terms of mean shakedown limit and probability of failure of pavement for different wheel-soil interface frictions. The results show that an increase in wheel-soil interface friction decreases the effect of spatial variability, thereby reducing the probability of failure of the pavement. Distribution of elastic, residual, and total stresses within the soil mass is obtained to further understand the shakedown behaviour of soil mass under moving loads.