Evaluation of Loess Collapsibility Based on Random Field Theory in Xi’an, China

The engineering properties of collapsible loess have significant uncertainty. Accurate prediction deformation is crucial for the safety construction in areas. Taking typical stratum as research object Xi’an, based on random field theory, combined with Monte Carlo strategy and modulus reduction method, stochastic finite element analysis self-weight collapsibility carried out to study influence spatial variability compression loess. results show that tends be stratified average along depth direction increase transverse correlation distance. result considering significantly greater than deterministic layered calculated value code. Considering low dominance effect positive skewed distribution field, equivalent characteristic by modeling evaluation strata Xi'an area proposed.