Model Uncertainty for Predicting the Bearing Capacity of Sand Overlying Clay

Abstract

In this paper, 62 centrifuge tests are collected from the literature and used to perform a statistical evaluation of the model factor for conventional methods to calculate the bearing capacity of dense sand overlying clay. The model factor is defined as a ratio of measured capacity to calculated capacity. The variations of the model factor with input parameters are established as regression equations based on the results of finite-element limit analysis (FELA). Conventional methods that multiply by regression equations are shown to be more accurate on average for all data sets. Further verification exercise from 27 additional centrifuge tests indicates the modified method could also be applicable for medium dense sand overlying clay. The mean and coefficient of variation of the model factor for the modified methods are finally characterized as a lognormal random variable with a mean of 1.06 and coefficient of variation of 0.17.