Issues with Initial Convergence (DT < DT_MIN. Recovery not possible!) even with followed suggestions

MFiX How to
#1

Hello!

I created a pseudo 2D rectangular spouted bed under the DEM model. In the log file, I found the phrase “t= 0.0000 Dt= 0.1014E-05 NIT> 50 Sm= 0.0000 MbErr%=
DT < DT_MIN. Recovery not possible!”

I would like to make clear that even with the time step constant (DT_FAC=1), I still had the issue of nonconvergence. I have also tried increasing the number of gas iterations, using a line preconditioner, turning off stall detection, reducing under-relaxation factors, and removing small cut cells. I am also using the ideal gas law.

I have attached the MFIX file. I am genuinely stuck, and I am not too sure where to go beyond here.

SBP.mfx (22.7 KB)

#2

Also! the version of MFIX is 23.2

#3

My suggestion is to start with the simplest simulation and add up physics one at a time. This will help troubleshooting the issue. Here you should start without turbulence and cohesion model to make sure the basic simulation runs.

There is a typo in the initial gas pressure, there should not be a decimal point between 101 and 325:

image
image557×578 52.9 KB

Then, please double check the k-epsilon initial and boundary condition values.

The Hamaker constant is also many orders of magnitude off. It is typically around 1E-19 to 1E-21 J.
image

1 Like
#4

Thank you so much! I appreciate your guidance!