The semi-resolved CFD-DEM method has emerged as a prominent tool for
modelling particle-fluid interactions in granular materials with high
particle size ratios. However, challenges arise from conflicting
requirements regarding the CFD grid size, which must adequately resolve
fluid flow in the pore space while maintaining a physically meaningful
porosity field. This study addresses these challenges by introducing a
two-grid mapping approach. Initially, the porosity field associated with
fine particles is estimated using a coarse CFD grid, which is then
mapped to a dynamically refined grid. To ensure conservation of total
solid volume, a volume compensation procedure is implemented. The
proposed method has been rigorously validated using benchmark cases,
showing its high computational efficiency and accurate handling of
complex porosity calculations near the surface of coarse particles.
Moreover, the previously unreported impact of the empirical drag
correlation on fluid-particle force calculations for both coarse and
fine particles has been revealed.