Dear forum,
I am having some issues with the convergence of a simple CGP case involving heat transfer. The model I am trying to simulate is a quasi-2D system. I have oversimplified it to the maximum, so there is only a thin layer of CGP-DEM particles (Gunn Nu number) and a gas injection at 600 K, while the system is at 293.15 K. My real particles have a diameter of 292 um, which I am increasing to 1 mm with CGP. Even though the case is simplified to the maximum, I am unable to reach convergence (T0 residual does not go down) and the case crashes at t<0.1 s (T below 250 K if I try to resume). Noticeably, if I increase the real particle diameter up to 750 um (for example), the case converges with no apparent problem.
I have tried changing the inital conditions (particles falling from a given height insteat that at the bottom of the bed), the lateral walls BCs, the number of iterations, discretization, mesh, relaxing the convergence and some others, but the case always fails. I have tried in another computer with a different gnu version and I am still having the same issues.
Could you please give me a hint on how to proceed?
Thank you!
test_ht.mfx (17.6 KB)