# twoPhaseEulerFoam: dispersed phase not conservative?

 Register Blogs Members List Search Today's Posts Mark Forums Read

March 9, 2018, 09:11
twoPhaseEulerFoam: dispersed phase not conservative?
#1
New Member

Lennart Steffen
Join Date: Mar 2017
Location: Braunschweig, Germany
Posts: 17
Rep Power: 6
Dear Foamers,

when simulating a liquid-solid system with twoPhaseEulerFoam, it appears that the mass of the solid phase within the system increases - without actual influx of sediment.

To visualise this phenomenon, a simple rectangular, two-dimensional mesh was created (height: 0.5m, width: 0.6m): The lateral BCs are defined as cyclic, and the top and bottom as walls (no slip at the bottom, slip at the top, for velocities of both phases). Thus, no influx of sediment into the system should be possible.
The lower 0.2m of the domain are filled with a solid phase fraction of 0.6. This defines the initial volume of the solid phase as 0.6*0.2m*0.6m = 0.072 m³/m.

All other BCs and initial values can be seen in the case, but are nothing fancy. The turbulence model for the solid phase is kinetic theory. Runs with OpenFOAM 3.0.

The expectation now would be that only a sedimentation process occurs, resulting in a stable phase fraction profile after some time. However, instead, the solid bed surface continually rises.

This can be seen in this video, and the attached phase fraction profile plot (alphaProf-time.png). The latter shows that the total solid phase volume in the domain increases, as the area below the curve clearly grows.

Does anyone know what might cause this?

Kind regards,
Lennart

P.S: I also tried setting all boundaries to walls, i.e. velocities to fixedValue (0 0 0). No significant difference in the described phenomenon.
Attached Images
 alphaProf-time.png (86.8 KB, 28 views)

 March 16, 2018, 11:56 #2 New Member   Lennart Steffen Join Date: Mar 2017 Location: Braunschweig, Germany Posts: 17 Rep Power: 6 Hmm. Feedback is not exactly overwhelming. Should I instead write this as a bug report or is that useless if I'm not using the current OF version?

 May 3, 2018, 02:33 #3 New Member   AliG Join Date: Mar 2018 Location: San Francisco Posts: 6 Rep Power: 5 Post your case files so we can get a better idea of what is going on.

 May 3, 2018, 04:06 #4 New Member   Lennart Steffen Join Date: Mar 2017 Location: Braunschweig, Germany Posts: 17 Rep Power: 6 I wrote a link in the OP where you can download the case. Here's that link again: https://www.dropbox.com/s/4ehmjvo3td...ation.zip?dl=0

 May 8, 2018, 03:49 #5 Senior Member   Join Date: Aug 2014 Location: Germany Posts: 292 Rep Power: 11 Have you tested it with 3D? Is gravity direction correctly applied?

 May 8, 2018, 08:32 #6 New Member   Lennart Steffen Join Date: Mar 2017 Location: Braunschweig, Germany Posts: 17 Rep Power: 6 I have not tested it in 3D. The case is 2D. I am quite sure that gravity is correctly applied, but even if it weren't, that would not explain the observable phenomenon: Apparently, judging by the concentration profiles, a phase change between liquid and solid phase occurs.

May 8, 2018, 08:46
#7
Senior Member

Join Date: Aug 2014
Location: Germany
Posts: 292
Rep Power: 11
Quote:
 Originally Posted by RL-S I have not tested it in 3D. The case is 2D. I am quite sure that gravity is correctly applied, but even if it weren't, that would not explain the observable phenomenon: Apparently, judging by the concentration profiles, a phase change between liquid and solid phase occurs.
If it's 2D then you are correct. If the case was 3D a concentration profile would not be enough information.

What is your mesh resolution? Have you checked refinement? What about convergence, time step size? What about boundedness of your alpha field during calculation?

I suspect something in the setup, but i can not download the case and check everything.

 Tags liquid-solid, sediment, sedimentation, twophaseeulerfoam