[massive_turbulence]

Hello Foamers!

If anyone wants to see the project I'll try uploading it to github, many thanks for any suggestions on this problem.

Basically I'll just simply describe the problem and how I set it up. For starters I wanted to simulate a simple pipe with both endings similar in radii and in the middle of this cylinder sits a circular disk that is closing off both ends from "flowing".

The way I set this up was with Salome; I exported the valve mesh and the pipe mesh, merged them together and setup the AMI stuff and what not.
One of the things I noticed about the cellsets was that there were 3 zones( more on this later on as it's important to the geometry between the valve and how it touches the pipe almost really close).

From there I ran the setup using pimplefoam and even after 100 iterations it wouldn't converge for the velocity. It ( the residuals) would basically oscillate around 1/10th and 1/100th for the velocity and the tolerance was 10^-5.

So then I figured that something must be wrong with the geometry and I went back to Salome and redid the whole setup with the valve being smaller than the diameter of the pipe (around 10% smaller and now there were only two cellsets not three) and now after 17 iterations for the same exact dictionary files pimplefoam converges.

Does this mean that I cannot setup a case without the geometry "leaking" fluid from one end of the mesh to the other or does anyone think that there is a solution to this but I'm using the wrong solver (maybe interfoam is it?) I dunno.

Please help! Cheers!

Andrew