A 3D random porous media model for cement mortar based on X-ray computed tomography

Abstract

Cement-based materials are typical porous media in which the spatial pores are of vital importance to flow problems, such as chloride ion penetration. This study proposes a three-dimensional (3D) random porous media model to characterize the spatial distribution associated with the porosity in cement mortar. The proposed model addresses the random heterogeneity in material properties using a Weibull random field and an exponential correlation function. The spatial distribution data associated with the porosity of the cement mortar, including the marginal distribution of the porosity and correlation structure of the Weibull random field, are obtained via X-ray computed tomography (XCT). The results indicate that the isotropic exponential correlation function is adequate for characterizing the correlation structure of the Weibull field with a maximum correlation length of 5.7515 mm. Moreover, a comparison of the simulated and measured porosity fields suggests that the proposed model can capture the characteristics of the porosity distribution in cement mortar. Considering the correlation in the porosity leads to a larger variance in the mean porosity of the mortar in the Monte Carlo simulation (MCS). In the example, the chloride diffusion in the cement mortar is illustrated using the proposed 3D random porous media model coupled with finite element (FE) analysis. The effect of the spatial correlation of the porosity on chloride diffusion in cement mortar was discussed from a probabilistic perspective. The proposed model provides insights into considering the properties of heterogeneous materials.