Impact of several coarse-graining models on a pilot-scale fluidized bed behavior using discrete element method–computational fluid dynamics.

This article focuses on assessing the effects of particle coarse-graining and mesh coarsening in coarse-grained discrete element method (CGDEM) simulations of a 3D pilot-scale gas-fluidized bed containing 9.6 million Geldart B-type particles. The study compares two additional dissipative mechanisms (ADMs) from the literature—Lu et al.'s restitution coefficient reduction model and Radl et al.'s relaxation force model—aimed at compensating for the reduced collision frequency and kinetic energy dissipation inherent to coarse-graining. Results indicate that a coarse-graining factor (fCG) of 2 combined with Lu et al.'s model yields the best fidelity to full discrete element method (DEM) simulations, while higher fCG values (3 and 5) lead to increasing discrepancies such as elevated bed height, reduced bubble population, softened fluid fraction gradients, and diminished vertical solid velocity. Radl et al.'s model, although dissipative, is computationally costly and tends to negatively impact particle vertical velocity, especially for fast-moving particles, by applying a non-conservative force that slows them down. The study finds that mesh coarsening has a negligible effect compared to particle coarse-graining, and that while ADMs improve some mesoscopic and macroscopic features, they are insufficient to fully recover DEM-level accuracy at high coarse-graining factors.