Probabilistic characterization of Young's modulus using Markov Chain Monte Carlo Simulation

This paper presents a Markov Chain Monte Carlo Simulation (MCMCS)-based approach for probabilistic characterization of the undrained Young's modulus, E_u, of clay, which utilizes both the prior knowledge and project-specific SPT data to generate a large number of equivalent samples of E_u for its probabilistic characterization. The proposed approach combines the prior knowledge and project-specific SPT data systematically under a Bayesian framework and allows general choices of realistic prior knowledge (e.g. an arbitrary histogram type of prior distribution). Equations are derived for the proposed approach, and a sensitivity study is performed to explore the effects of prior knowledge on probabilistic characterization of soil properties. It is shown that the proposed equivalent sample approach integrates the information provided by different types of prior knowledge with project- specific information in a rational manner and improves significantly probabilistic characterization of soil properties by incorporating consistent prior knowledge. © 2014 Taylor & Francis Group, London.