Modelling the sealing of a channel by a particle

Dear community,

the problem I want to model is based on the sealing of a fluidic channel by a large particle. I am quite new (and very fascinated) to the DEM-part of the CFD-DEM world and I need some input from you.

I prepared a small test-case that is modelling the closing of a ball check valve-type structure. I tried two solvers, Piso and IB. I figure IB is probably the right choice, but Piso is faster and I would prefer to use it. As you can see in the result plots, the channel isn't sealed fully by the particle with either solver:

- With the Piso solver there is still bypass flow around the particle, yet the particle is perfectly stopped at the correct position and everything else looks good.
- With the IB solver I have problems with the "interface" field. The particle is stopped too early before the channel and the surface is strangely deformed. The simulation is unstable, the flow field "pulsates" around the deformed interface.

Anyone ideas what parameters I need to change?

If someone wants to have a closer look, I uploaded the test cases with their latest time step to dropbox (I get some error when uploading it here). I used cfdem 3.0.1 with OpenFOAM 2.3.x, both about a week old.

https://www.dropbox.com/sh/w0nsg8w4vbd9nyc/AADt1E-wXYXKcD7k8lltf-aaa

(Sorry, I didn't make use of the rotational symmetry in my models. The prepare_mesh.sh script might not run on your computer, at least I had to change some cmake settings in the "gmsh" installation when testing it on another computer. I do not think however that it's a mesh related problem.)

Many thanks!

Daniéll