Guide for Multigrid/Bottom Solver Selection: MLMG convergence failures in CFD-DEM

Hi MFIX-Exa Team,

I am looking for guidance on optimizing the selection of multigrid solvers and parameters based on specific problem physics (e.g., dilute vs. dense regimes or complex geometries).

I am currently running a couple of CFD-DEM cases where the MLMG is frequently reporting a failure to converge. I have tried different settings like mg_max_coarsening_level = 2 as suggested in this topic.

Any advice on how the different multigrid/bottom solver settings work and/or affect convergence would be very helpful.