Non-convergence when calculating gas-solid heat transfer using the PIC method

Dear developers,
In MFIX, when I use the MP-PIC method to calculate a staged fluidized bed containing gas-solid heat transfer, the results always diverge suddenly after a short period of time (about 0.2 seconds). I have attempted to improve the mesh and replace other heat transfer models ( by setting the des_conv_corr keyword,e.g. GNUU or WAKAO), but the problem still persists. However, if I remove the particles and calculate only the flow field, or if I set both the particle and fluid temperatures to 293.15k, which means that there is no temperature difference between the gas and the solid, both test cases appear to be able to calculate continuously. I have been struggling with this problem for quite some time. Could you please provide me with some advice on this issue? Is this non-convergence caused by grids or other heat transfer models?
Thank you so much!



image
staged-3-test_2022-09-26T065113.651059.zip (329.0 KB)

I don’t see your stl file clc-staged-3-test.stl in the attachment. Can you please attach it?

Here is the stl file. Thank you very much!
clc-staged-3-test.STL (15.9 KB)

It failed for me due to y-velocity getting larger than the defined tolerance. Here velocity will get large due to the constriction. It runs better when I increase the max velocity. Go to Numerics pane, advanced pane and set “Maximum inlet velocity factor” to say 100.

Thanks for your help! I will try this way.
By the way, does using explicit coupling (solid-PIC-advanced pane) makes some difference in the calculation results as opposed to using implicit ? I find that when I use the energy equation in this example, if I do not turn on the explicit coupling, it is prone to non-convergence. Is this also related to the mesh or boundary conditions ? Thanks again for your advice