Wall-particle heat transfer in pure granular flow

MFiX How to
#1

Hi, everyone. recently i want to build a pure granular flow in DEM and discuss the wall-particle conduction heat transfer, and when i try to run, something wrong with check_gas_phase.f:312, Does anyone know how to solve it? I would appreciate for any answer from your!
here is the .fix file
0710.mfx (26.0 KB)

ć±ćč•æˆȘć›Ÿ 2021-07-22 165018
ć±ćč•æˆȘć›Ÿ 2021-07-22 1650182145×1399 206 KB

1 Like
#2

Hi ChenYuan.

Something very odd is going on here - somehow your input file is not even being found - note the “mfix.dat Not found” message. I was able to load and run your case here. So maybe something is wrong with your installation? What version of MFiX are you using and how did you install it? What solver module are you using? Can you submit a full bug report using the option in the main menu?

– Charles

#3

Hi again, I just noticed the screenshot says “pureflow.mfx” while the file you uploaded is 0710.mfx. Did you attach the wrong file?

#4

HiCharles, I noticed that I upload wrong file, and thanks for your reminder, here are the .mfx file with errors.
pureflow.mfx (11.4 KB)
Looking forward to your reply

#5

@ChenYuan

Thank you for the error report.

Even though you have not defined a gas phase, this is happening (Warning 1)

 Message: 2000 The energy equations are being solved (ENERGY_EQ) and
the constant gas specific heat is undefined (C_PG0). Thus, the thermo-
chemical database will be used to gather specific heat data on the
individual gas phase species.

Since you have not defined any gas phase or any species at all, this is an error. In addition to this there’s an

Error: Solver log message not formatted properly.

which means that one or more error messages are not being displayed at all.

So this reveals a couple of flaws in MFiX, which we will fix as quickly as possible. In the meantime, I find that if I disable the energy equations, your pureflow case runs without trouble.

– Charles

#6

Even with granular flow the DEM wall conduction involves a particle-fluid-wall component and this requires setting fluid properties. We probably should set some default values for granular flow so it doesn’t trigger a datacheck error.

I can run your case if I use the following trick:

  1. Go to Model Setup pane and uncheck “Disable fluid solver”. This will unlock the Fluid panel.
  2. Go to the Fluid panel. Set a constant specific heat, say 1005 J/(Kg.K). Either set a zero thermal conductivity if you don’t want the particle-fluid-wall component, or set a non-zero thermal conductivity if you still want the particle-fluid-wall component).
  3. Go back to the Model Setup pane and check back the “Disable flow solver” box.

Notes:

  1. Always include the particle diameter in the vtp files so the GUI can properly scale particles when visualizing the data.
  2. You can get the simulation to run a bit faster if you decrease the spring stiffness to say 100 N/m Solids > DEM pane, Normal spring constant values for particle-particle and particle-wall collisions)
1 Like
#7

Thanks, Jeff. I modified the file the way you said, and it’s really worked, thanks for your solution again!

#8

Thanks, Charles. I modified the file according to Jeff’ method, and it worked. Thanks for your reply, sincerely!