Why does the simulation report an error when the inlet speed is set to 0?

MFiX Bug report
#1

Hello, developers! My simulation is divided into two parts, the first part is the case where the inlet speed is not 0, and there is no problem with this part. The second part is the case where the inlet speed is 0, which is to set the inlet speed to 0 on the basis of the first part (the rest of the settings have not changed) and resume the simulation. However, during the second part of the simulation, “DT<DT min” and “solver crash” errors appeared. What is the reason for this? The relevant files have been uploaded. Looking forward to your answer, thank you!
1.zip (4.6 MB)
6580859fb06462af3f4afce77d58aa4

#2

When you set the inlet velocity to zero, you are basically settling the bed under gravity, right? If so, maybe you have reached a steady state where nothing moves anymore and you are getting a false non-convergence.

#3

What you said should be right. The current state is that after a certain stage of calculation, an error is reported and the model has not completed all calculations. But I want the simulation to continue until it reaches the simulation time. How can I adjust to achieve it?

#4

When the inlet velocity is set to 0, particles settle under the action of gravity and reach a so-called stable state, which results in an error. At this time, the particle velocity is very small, but not 0. I want the simulation to continue until the particle velocity drops to 0. How can this be achieved?

#5

Can you please attach your particle_input.dat file?

#6

Here are the relevant documents. Looking forward to your answer, thanks！
particle_input.dat (1.7 MB)

#7

This is also a file that may be required. Thanks!
1.mfx (9.5 KB)

#8

I don’t think 1.mfx is the correct file. This one is set up for a granular flow (no fluid), so you can’t get a 'DT_MIN` error with it.

#9

Sorry, I didn’t explain it clearly. 1.mfx file is the process of simple particle falling (without fluid), and the export file for this process is the particle-input.dat file for the fluidized bed. The fluidized bed file is in the 1.zip , in which the simulation contains two processes: the first process (the first 30s of the simulation) is the process by which the fluid enters from the bottom, and there is no problem with this process; The second process (the last 10s of the simulation) is the process of no fluid entering and the particles settling freely, and the simulation duration I set for this process is 10s, but every time I calculate halfway, there will be errors that DT＜DT min and solver crash. What is the reason for this? Thank you!

#10

Thanks, it is a bit tedious to reproduce, but I think I can see similar failure with a simplified version. A few notes:

  1. If you are planning to reach steady state, you can set DT_MAX to a large value say DT_MAX=0.1 or DT_MAX=1.0
  2. In Solids>DEM pane, check the enable explicit coupling of interphase quantities. This should make the simulation run faster.
  3. In Numerics> Linear solver, set the max number of pressure equations to 200, and set the max number of u,v,w momentum equations to 50. This setting allows me to avoid the DT_MIN error.
  4. You have a rectangular geometry, you should use the regular BCs, and not need to use an STL file.
1 Like
#11

I tried as you said, but still got DT＜DT min. I am thinking the reason for the error is my working condition setting is wrong?

#12

Can you remove the STL file and set it up with regular BCs to see if this helps?

#13

How to set up with regular BCs?

#14

Instead of using an STL file for the walls, use the x=xmin, x=xmax etc. planes to set up the BC regions.

#16

Ok, I tryied as you said. And it worked sucessfully! Thank you very much!