request for help with convergence problem

[mattkoch]

OK, so I really like the way Salome and code_saturne interact. When a problem converges, I feel very comfortable with this software combination now. The sad thing is, the fewest of the problems I have tried to tackle do actually converge.

The latest problem I am trying to look at is the underfloor plenum of a server room. I have it set up quite simply for now. One opening blows air into the plenum (inlet, downward), and fifteen openings ("perforated floor tiles") let the air escape from the plenum (outlet, upward). The rest is wall. Notably, I wanted to model the effect of all the struts and rails of 2 inch square cross section and spaced every 24 inches in the plenum.

No matter what I do in the code_saturne setup (4.0 or 5.0), it inevitably blows up. In the end, I would like to be able to put K values on the outlets, too, in order to more accurately model the effect of the flow restriction they represent.

Is there any way I could share my CGNS Salome mesh file and my XML code_saturne input file with anyone, so that they may take a look at these with a better-trained eye? In the past, I have been told that the mesh can cause a lot of trouble, but from what I can see, the mesh looks rather clean? If there is a taker, please advise on how to communicate the files.

Many thanks in advance.

[Yvan Fournier]

Hello,

Did you check the best practices in the documentation part of the forum ?

Before you post the mesh, please post the other recommended info of a failed run (using the forum usage guidelines).

If the mesh is small you may post it too. Otherwise wait for a first iteration or post it through a large file exchange service.

Regards,

Yvan