Prediction of the effective diffusion coefficient in random porous media using the finite element method

A finite-element-based method is presented for evaluating the effective gas diffusion coefficient of porous solids. Using this method, the 3-D micro-scale geometries of the porous solids are constructed under the ANSYS platform by the parametric code; the relation between effective gas diffusivity and micro-scale features of random-distributed porous solids is established. The results show that in random-distributed pore media, there is a percolation threshold epsilon(p), and this percolation threshold decreases with increasing coordination number of the pore network. The relationship between the effective diffusivity and porosity is strongly nonlinear when the porosity, epsilon, is less than a certain value epsilon (L) ; for epsilon > epsilon (L) , the relationship becomes quasi-linear. This dividing point epsilon (L) decreases with increasing coordination number. The larger the coordination number of the pore network, the higher the effective gas diffusivity. Based on the simulation results and observations, a formula relating the effective diffusion coefficient with porosity is proposed.