[tikulju]

Hi all!
I've been working on multiphaseEulerFoam, and incorporated the dynamic mesh refinement on it. I have received some questions on it and would like to share them with you guys, hoping to get some answers to me too:

In the main code, I think putting

if (mesh.moving())
{
        Urel -= fvc::reconstruct(fvc::meshPhi(U));
}
mesh.update();

does the mesh movement and refinement correctly. Correct me if I'm wrong.

In the dynamicMeshDict, fluxes needed to be corrected. There's 'phiAlpha<phase>' stuff, which are needed. I'm not quite sure, what they are, so I've left them away. If someone do know what they are and what should be inserted on them, could shed some light on this matter.

And the third thing is the load balancing. Does anybody have an idea, how the load balancing used e.g. in snappyHexMesh, could be incorporated into this solver? I've tried it in several ways, but so far haven't succeeded on it. I've already started one thread on this matter (http://www.cfd-online.com/Forums/ope...amic-mesh.html), so maybe it's better if this matter would be discussed there...

- Timo

[kmooney]

Hi Timu,

Send me a PM and I'll share a solver which has the appropriate rel/absolute flux corrections.

As for the load balancing, I can tell you there is a fully functional load balancer that works with dynamicRefineFvMesh developed in a few academic circles, it just hasn't been released yet. I'm thinking we'll make an announcement on the forums when it is released with instructions on where to download it.

[tikulju]

Hi Kyle,

I sent you an PM, thanks already in advance! And it would be good, if you guys could share the load balancing stuff, since the load imbalance really slows down the simulations.

And when i browsed through my solver, I actually had some flux corrector stuff (similar as in interDyMFoam). I wrote these stuffs some time ago, so I forgot them. Sorry for that.

- Timo

[tikulju]

Hi Kyle (and others),

I did put the flux corrections as it was also in your solver, but it seemed that things went from bad to worse . Previously my 'bad' velocities were around 50 m/s, now they're around 200. I think the thing is, that in interFoam the velocity field is shared between the phases, but in multiphaseEulerFoam it is per phase. This 'velocity overshoot' happens between the phases, for which have interphaseCompression defined (i.e. VOF type of interaction) and high density ratio (~5000). Any ideas?

- Timo

[tikulju]

I think the problem is not in the code, but in the case setup. Since the mesh is not moving, I don't think the absolute/relative are doing anything (see $FOAM_SRC/finiteVolume/lnInclude/fvcMeshPhi.C). Of they should be there, if the solver would be used in the moving mesh case.

- Timo