Computational divergence

Dear developer,
Excuse me for interrupting you. Recently, I have been simulating continuous pyrolysis reaction moving bed, as shown in the figure. However, it always calculates divergence in a short time. Could you help me find the reason? My mfix versions are 21.4 and 22.3.1. Thank you!
Uploading: image.png…

Best Regards.
continuous_moving_bed_2023-02-20T180537.130739.zip (112.8 KB)

There could be some bad cells in there. It appears to run better with a mesh of 22x50x10.
Please not that here you could set up the simulation without any stl file, and carving out rectangular blocks of cells (NSW BC) to achieve a cleaner mesh.

OK, do you mean that I can set up the simulation without using the functions in “Geometry-Geometry”? If so, the mesh quality is better.