[dvanhauwermeiren]

Hi everyone,

I am simulating the flow over a wire-wrapped tube bundle in a hexagonal domain using overset grids (see pictures). Because I use a 'pressure outlet' boundary condition at the outlet, I'm encountering reversed flow. As the quality of the mesh is high enough, I think this is due to the fact that the flow hasn't stabilized enough at the outlet as the wire induces a swirl in the flow. A possible way to overcome this would be to change the boundary condition at the outlet to 'outflow' but when I do this, I get an error saying that this is not compatible with overset.

So if I have to use a 'pressure outlet' boundary condition at the outlet, I don't see any other way than to extend the hexagonal background grid to give the flow the chance to stabilize. The question I have is: is there a way to only extend the hexagonal background grid without extending the component grids? This would mean that the two wire-wrapped tubes would stop somewhere in the bulk of the hexagonal domain and not at the outlet of the hexagonal domain. This raises the question of what boundary conditions have to be set at the outlet of the two wire-wrapped tube grids. I tried to set those at 'overset' but this creates orphan cells at the outlet boundary.

It would complicate things if the component grids (wire-wrapped tubes) also have to be extended so this is the reason why I ask if there is a less complex solution to this.

I included several pictures to describe the setup and also the situation in which the two wire-wrapped tubes stop somewhere in the bulk of the hexagonal fluid domain.

Any help with this is welcome. Thanks

Single wire-wrapped tube.PNG

Overset mesh Inlet.jpg

Fluid domain 3D.PNG

Original boundary condition setup.jpg

Alternative boundary condition setup.jpg.