[tomf]

Hi all,

At the moment I am working on a packed bed case where a liquid is flowing through a packed bed. The situation is such that there is an inlet at the top where the liquid enters, a region with only liquid, than the packed bed which is located above the outlet. The real case is a bit more complicated geometrically, but basically it is the same. I want to have that the liquid can freely exit the domain, but that the particles of the packed bed remain inside the domain.

So a picture to show the simplified 2D situation (start.png), laminar flow of water should be from top to bottom, but after 2 minutes the situation is as shown in end.png. The case is using a uniform blockMesh mesh, with aspect ratio of 1 throughout.

I had looked around on the forum for degassing boundary conditions or similar set-ups and people have used a slip boundary condition for the particles with a zeroGradient or inletOutlet boundary condition for the water. Following the fluidised bed tutorial I used a prghpressure (also tried fixedValue) boundary condition for p_rgh on the outlet.

I have also attached the simplified case (for OpenFOAM version 4.x, probably works for 3.0.x with minor modifications). Right now it is configured to run on 10 processors, if you want to have a look, please modify the decomposeParDict and the runscript.sh according to your system.

I have tried to run the case with OpenFOAM v3.0.x, v4.x, openFoam-plus and foamExtend-3.2, which all gave similar results.

If anyone can help, I would much appreciate this.
Regards,
Tom

[tomf]

Small update:

Tried to run the same case using MFIX, which also gave similar results, however I have no experience with MFIX, so I am a bit unsure about the setup for it. I tried to match it as much as possible with the OpenFOAM cases.

Some other things I tried: use a porous model for the particle phase only, use explicitSetValue constraints for the particle phase, change the order of the phase definitions, but nothing seems to have a significant effect on the outcome.

Then I treid to give the particle phase a small upwards velocity from the outlet, this quite quickly (couple of tens of timesteps) causes a divergence in the solver after the alpha equation goes out of bounds.

Maybe I am trying something that is not possible? If you think so as well, please let me know.

Best regards,
Tom

[tomf]

Please find attached a sketch of the problem with the applied boundary conditions.