[Jojell]

Dear Fellow CFDudes,

Currently I am working on a multiphysics problem using GeN-Foam, a neat extension of the OpenFOAM library. The problem concerns convergence for the P_rgh while solving the fluid flow field (single phase).

To boil down my problem; The entire flow regime of the primary circuit of a molten salt reactor is modelled, including pump and HX, with a momentumSource emulating the pumping pressure. So it is a 3D closed loop system, where the salt is going around. Now, the thing that stuns me a bit, is that with a salt density of 1000 or less, the convergence is just fine (residual ~10e-4 and going), but when I take density above ~1200 (density of salt is 4000+), the residual starts to plateau around 0.3 or higher, 0.95 when density is 2000+. I have changed a great many things, but the only crucial parameter seems to be the density, and I have somewhat run out of ideas about how to remidy the issue. The solved flow field for the non-converged problem cases seems oke, even with a very low convergence, which confuses me even more.

checkMesh gives Max skewness ~1.45, Non-ortho ~67.5 with the most skewed elements mostly near the wall of the pipes.

Using Under/Over relaxation does not work, using different meshes does not work, using different initial turbulence parameters does not work, changing to laminar does not work, changing pump speed does not work, changing porous media properties does not work.. It seems for real that density is THE crucial factor, for some CFDelirious reason.

I am running OpenFoam v2112 and have meshed with Salome.

I can post outputs and inputs on request. Any suggestions, leads, clues, will be greatly appreciated. Kind regards!