Pore-scale flow simulation in anisotropic porous material via fluid-structure coupling

Abstract This paper describes a novel hybrid method for fluid simulation of saturating anisotropic porous material via fluid-structure coupling. Our framework employs particle finite element method (PFEM) that not only adopts Lagrangian scheme to model the motion of freely-moving particles, but also produces the extended Delaunay Tessellation to furnish the governing equations with FEM discretization. We first employ adaptive smoothed particle hydrodynamics (SPH) to simulate porous flow respecting the anisotropic permeability with little cost. Second, the extended Delaunay Tessellation is obtained to solve differential equations for skeletal deformation. Third, a hybrid particle system is adopted to track the surface and topological changes. At the physical level, we introduce dynamic permeability considering skeletal deformation via fluid-structure coupling. At the geometric level, PFEM reduces the computational cost and effectively tracks topological changes. Moreover, our implementation on CUDA improves the performance in high-quality physics-based graphics applications. Consequently, the proposed method realistically reproduces interactions between pore-scale flow and anisotropic porous material.