I am using MFiX-24.1 to run a DEM simulation. I encountered a problem when simulating particles with diameter of 80 μm (8.0e-5 m), that is, DT<DT_min, and the recovery is impossible.
I have tried several modifications to fix this problem, such as reducing the min_time_step to 1e-10, increasing the spring constant, and changing the gas density to Ideal_Gas_Equation. However, the problem still exists. My work used a cuboid bed and there is no need to consider no cut cell problems.
Interestingly but confusingly, the simulations with particles in the diameter of 300 μm runs well, despite all simulation parameters kept the same.
The simulation files are attached. I would appreciate it if anyone could discuss about this problem. FLUID_BED_DEM_3D.zip (594.6 KB)
When troubleshooting, I find it useful to increase the verbosity of the output: Go to Output>Residuals and check “Display residuals”.
When I did that I got more details about the lack of convergence:
Velocity exceeds limit: 0.10033
in cell: I = 10 J = 9 K = 10
Epg = 0.92243 Ug = 0.0000 Vg = 0.10033 Wg = 0.0000
To change the limit, adjust the scale factor MAX_INLET_VEL_FAC.
The max velocity is set to 0.1m/s which is actually not large. The max velocity is based on the inlet velocity, multiplied by a factor (100 by default). You reach the max velocity because you have an inlet bc ramp that starts with a very small velocity (0.001 m/s). Here the report of large velocity is not alarming so you can turn it off. Go to Numerics> Advanced and set “Maximum inlet velocity factor” to 1E6.
Thanks for your kind reply!
I didn’t realize it could be caused by this problem, and your suggestion matters to me.
After adjusting the Maximum inlet velocity factor, the problem is solved. It runs well now.
Thanks again!
