Upscale rarefied volatile diffusion in porous media: A pore network modeling approach

On airless bodies such as the Moon and cometary nuclei, volatiles redistribute and escape through the surface porous regolith layer. Under a highly vacuum environment, volatiles migrate through porous media via Knudsen diffusion, where volatile molecules rarely collide with each other and instead jump directly from one solid surface to another. Although Knudsen diffusion in microscopic confinement can be numerically resolved by Monte Carlo-based methods, studying its upscaled behavior at the representative elementary volume scale remains challenging due to the high computational cost and inherent geometric and kinetic randomness. In this paper, we develop a probability-based pore network modeling (p-PNM) approach to model upscaled Knudsen diffusion behaviors in porous media. This approach is inspired by the duality of the molecular motion trajectories: volatile transport in pore-bodies satisfies the Markov property (the exit probability of molecules is independent of their residence time), while that in pore-throats does not satisfy the Markov property. This p-PNM approach is validated in ball-stick geometries and typical sedimentary geometries, with geometric randomness, surface adsorption energy heterogeneity, and temperature gradients. It enables upscaled modeling of volatile transport through regolith on airless body surfaces, for geoscience research and for in situ recovery of resources on the Moon and other airless bodies.