Learning Forum

[cnmatu]

Hi,

I have set up a cylinder to contain particles which when coupled with fluent in a 2-way CFD-DEM coupled simulation, are eroded into an exit tube. The result shows movement of the particles but they do not go through the exit tube. Is there anything wrong that I am doing in the building up of the exit geometry? Any advice?

[Rob]

Check in Rocky that you just have the in & out tubes and main body. Depending on how you built the model there may be another interior (for Fluent) at either end of the centre section. 

[cnmatu]

Hi Rob,

I have tried your suggestion but particles still do not exit through the exit tube. In Rocky I need the "interior" to create a flow path for the fluid which I disable later. However in Fluent, it is not required. So I have been importing a case file from Fluent into Rocky with the interior in it (Geometry - Import wall..). And when I am going to run a 2-way coupled simulation, I import another similar case file; however without the interior in it. Still this has not worked. Any further advice? What am I doing wrong?

 

[Rob]

Rocky will need a wall to avoid particles falling out when you initially fill the annulus. After that Rocky should just have the outer walls with no inlet or outlet surface. Check surfaces aren't being re-added to Rocky once you import Fluent case in for the coupled run - you should only have the outer casing when you hit "run". 

[cnmatu]

If I only have the outer casing in the Fluent case file that I import for coupling when I hit "run", won't the side exit wall of the outer casing stop particles from flowing out? How do I create an outer casing for the coupling Fluent case file for that not to occur?

[Rob]

I think when you link the two models the Rocky geometry handles that side and passes data back to Fluent. So Fluent's case shouldn't overwrite the fact that Rocky has fewer boundaries. I assume the particles can actually flow and the system hasn't blocked it self up?