Structural analysis with probability-boxes

Abstract

Probability-box (p-box) is a rigorous and practical way to represent epistemic sources of uncertainty where the available knowledge is insufficient to construct the required probability distributions. In this paper, interval finite element (FE) methods are combined with the concept of p-box to analyse structures subjected to uncertain loads modelled by p-boxes. Two methods, namely the discrete p-box convolution and interval Monte Carlo methods, are presented along with example problems. The computational efficiency of the p-box FE method is also presented.