Quantification of cohesive fracture parameters based on the coupling of Bayesian updating and the boundary element method

Abstract

The fracture process in concrete involves nonlinear mechanical phenomena, which are accurately represented via the cohesive crack model. Due to the inherent randomness of this process, large scatter is observed in the experimental results. Therefore, significant uncertainties control the parameters that govern the theoretical approaches for concrete fracture modelling. In this study, a stochastic procedure for the parameter quantification of concrete nonlinear fracture models is presented. The Boundary Element Method is coupled to the cohesive model to model the nonlinear fracture phenomena. The Bayesian updating approach is subsequently applied to quantify the parameters that govern the cohesive laws based on the results of experimental analyses. The stochastic procedure enables the use of different cohesive laws to identify the law that provides the best agreement between numerical responses and experimental responses. A three-point bending notched test regarding different concrete mixtures is used to demonstrate the relevance of the proposed scheme.