Random finite element analysis of a clay-core-wall rockfill dam considering three-dimensional conditional random fields of soil parameters

Abstract

Structural reliability and safety have been investigated using the random finite element method (RFEM) based on conditional random field (RF) by many researchers. However, the constraint RFEM has not been studied sufficiently to assess its influences on dam structural behaviours, especially in clay-core-wall rockfill dams. In this study, based on Kriging theory and in-situ measurements, a method is applied for constructing conditional RFs of dam material parameters. The results demonstrate that conditional RF has little effect on the mean values of maximum settlement, maximum deformation gradient and hydraulic fracture area ratio; relative to the unconditional simulations, the relative differences are 1%, 3%, and 7%, respectively. By establishing an effective Kriging estimation, the standard deviations of conditional maximum settlement, maximum deformation gradient and hydraulic fracture area ratio are relatively reduced by 16%, 4%, and 24%, respectively. Thus, the conditional simulation can provide more accurate security estimations for the cumulative probabilities of the structural responses. Measurement layout has a great effect on variation accuracy of random structural responses. The measurements should be taken near the locations of the structural behaviours of interest so that the conditional RFs can more accurately represent the spatial variability of soil properties and reduce simulation variation.