Uncertainty quantification of heterogeneous mass concrete in macro-scale

Abstract

Various levels of uncertainties exist in the structural mechanics problems which need to be properly quantified. In the case of the brittle materials, the uncertainties might be addressed in the multi-scale level using random variables (RV), random fields (RF), and their correlation. This paper presents the results of a study that considers the spatial distribution of several mass concrete parameters in the context of the macro-scale RF theory.Three-dimensional slice finite element model of a dam is prepared as case study. The modulus of elasticity, tensile strength, and mass density are assumed to be either homogeneous or heterogeneous within the body. Appropriate RF realizations are generated based on the mid-point discretization technique, and covariance matrix decomposition.First, a deterministic reference coupled system model is analyzed, and the damage is verified according to the literature. Next, several stochastic finite element simulations are performed to understand the progressive failure of the heterogeneous and homogeneous mass concrete in the macro-scale. The impact of correlation length, and multiple RVs are also investigated. The uncertainty and dispersion of the intensifying dynamic responses, as well as the failure modes are quantified. It is observed that concrete heterogeneity affects the progressive failure analysis, and should be included in any detailed risk assessment framework.