Convergence problem with TFM in 2D

MFiX How to
#1

Hi, I just want to solve the simple 2D model in TFM, but the residual is still too high with nearly 5*10^-3, and it can hardly reduce with the change of max-iteration and relaxation factor, the figure of residual is shown below. For supplemant, this model can be solved under lower gas velocity (0.3 m/s)

image
image1329×652 12.7 KB

with gepmetry and mesh below:

image
image2520×1422 240 KB

image
image1221×1476 36.1 KB

the property of fluid and solid:
image
image1221×1476 44.5 KB

image
image1221×1476 54.3 KB

Initial condition and BC:
image
image1221×1476 41.8 KB

image
image1221×1476 42 KB

image
image1221×1476 39.4 KB

image
image1221×1476 43.8 KB

Numeric:
image
image1221×1476 38.2 KB

image
image1221×1476 47.7 KB

image
image1221×1476 45.4 KB

could anyone give me some advices for this question?
tfm_2d_ug_0.90_solid_0.15_2023-11-29T100555.194981.zip (1.8 MB)

#2

Instead of the screenshots, please upload your case files - Main
menu/Submit bug report. That makes it easier for us to try running your case. Thanks.

#3

@cgw thank you for your reminder, the report has been upload now :grinning:

#4

The settings in your attached files are different from the screenshots.

My suggestion is to run with ideal gas law for the fluid. If you really need to run as incompressible, increase the pressure and velocity iterations in the linear solver, and lower the relaxation factors. You cam also change the initial velocity (say lower y-velocity, and leave the pressure blank so it uses hydrostatic values).

#5

thank you for the suggestions, i would try them later