Discrete element model for moisture diffusion of rocks during water absorption

Rock masses in a water-rich environment often destabilize due to the water-softening effect; therefore, identifying the migration and diffusion of water in rock is critical. Most current simulations of water migration in rocks rely on grids and driving conditions, which limit the practical application of numerical models. This paper presents a discrete element model to quantify the moisture diffusion processes in rock based on diffusion theory. In the model, the particles are divided into particle elements by the radial-Voronoi method, and moisture is assumed to transfer between the pores of these particle elements. Soaking tests are conducted to determine the natural water absorption properties of rock and to verify the rationality of the DEM. By using the proposed DEM, the effects of the diffusion coefficient and particle size are studied. The results show that the diffusion coefficient affects the saturation velocity but does not affect the water distribution at the same saturation states, while particle size and pore structure heterogeneity can influence the water distribution. The proposed model greatly exploits the properties of discrete contact networks in the DEM, which will lead to better computational efficiency and applicability for simulations of water migration in rocks.