[ksmithgall]

I have a surface mesh that I want to build a case using the interfoam solver to simulate multiphase flow.

I’m having a hard time wrapping my head around the best way to generate the mesh for this, I'm assuming snappyhexmesh. I have a watertight .stl representing the ground/wall, but I'm not sure what I need to do next.

Do I need to create the background mesh that's larger than the surface file I have? or does there need to be a flat plane to act as a clipping boundary to create the patches for the inlets / outlets? The surface I have curves to follow the channel so I assume there would need to be multiple blocks.

Any help is appreciated

[Tobi]

Good morning Kurt,

you next is to generate the background mesh. Normally it is a good advice to make the mesh larger than the enclosed STL file. I prefer that way but it depend on the geometry. Sometimes you get screwed cells at the patches (depend on the position of the cells) and then I move my mesh in the corresponding direction, make it finer or coarsen it or make it just a bit smaller than the patches (which are aligned with the background patches). However, my first suggestion is to make the backgroundmesh larger but keep in mind that you should have pure cube cells. Well you also can make some aspect ratio but the best results are always with pure cube cells.

Hope this give you some idea. Further information can be found in the UserGuide or here

Good luck

[ksmithgall]

I have looked at the snappyhexmesh wiki it still does not make sense to me.

For the example with the car in the background cube, the bottom of the mesh corresponds to the ground. If it does not hit the .stl the whole z range is included in the computational domain, which is accurate.

Whereas for the case of the open channel, if snappyhexmesh does not hit the .stl, and the whole z range is included, wouldn't it simulate water flowing under the surface rather than on top of it? Or is there a way to clip back the mesh to generate the boundary patches afterward?