[Quest.Qiu]

Dear All,

I am trying to simulate a two-phase pipe flow question with interFoam solver. I found the solution is quite good with circular section pipe, but is not good with triangular pipe. I guess it is the mesh quality problem and tried different way of meshing. But all of them are bad. Does anyone know how to fix it? Any advice would be useful.
http://m.qpic.cn/psc?/V11Cdl5T0ExPCp...A!&rf=viewer_4
http://m.qpic.cn/psc?/V11Cdl5T0ExPCp...A!&rf=viewer_4

[Tobi]

Pictures are not available.

[Quest.Qiu]

Originally Posted by Tobi

Pictures are not available.

Thanks for your answer. I have upgraded my picture.

[Tobi]

In that case, take care of your numerical schemes. Check your mesh using

checkMesh

and apply the appropriate numerical schemes onto your case.

[al.csc]

Originally Posted by Quest.Qiu

Dear All,

I am trying to simulate a two-phase pipe flow question with interFoam solver. I found the solution is quite good with circular section pipe, but is not good with triangular pipe. I guess it is the mesh quality problem and tried different way of meshing. But all of them are bad. Does anyone know how to fix it? Any advice would be useful.
http://m.qpic.cn/psc?/V11Cdl5T0ExPCp...A!&rf=viewer_4
http://m.qpic.cn/psc?/V11Cdl5T0ExPCp...A!&rf=viewer_4

Mesh looks good but the numerical recipe probably needs to be adjusted, as Tobi mentioned.
Have you tried pointCellsLeastSquares as gradSchemes? Increase nCorrectors to 3 and/or nNonOrthogonalCorrectors to 1.
With the latest ESI version, I personally would use isoAdvector, plicRDF reconstructionScheme and the hexahedra based mesh you posted. But be aware of max(Alpha)Co <1 limitation.

Cheers