[jaydeep]

Hi,

I am referencing to : FVDOM instability in wedge-type mesh


I am attaching a test case. It has 2 concentric cylinders with diameters 22mm and 7.6 mm at temperatures 873.15K and 1673.15K respectively. Both have emissivities 0.7.

Answer for this problem is, you will see a heat flux of 260.9kW/m2.
Setup is assumed infinitely long.

Openfoam predicts the heat flux by ~ 15% wrong.

I looked up online to figure out the issues, and my guess is because of wedge patches there is some sort of flux leak (rays leaking) since the problem is solved as 3D instead of 2D.

I tried to compile fvDOM with basically copy pasting 2D part into 3D -- to force openfoam to solve in 2D mode even in wedge type cases. (stupid idea) But that threw me a duplicate library/model error. Since I don't know how to create copy of fvDOM as myfvDOM to avoid hash table duplicacy, I gave up the effort.

I also tried setting cyclic instead of wedge, but I get the same error.

Would anyone recommend any ideas?

Thanks!

[jaydeep]

The fvDOM solver can be modified with some stuff from polyMesh::nSolutionD()

     
forAll(boundaryMesh(), patchi)
   69     {
   70         if (boundaryMesh()[patchi].size())
   71         {
   72             if (isA<emptyPolyPatch>(boundaryMesh()[patchi]))
   73             {
   74                 nSolutionD = 2;
   76             }
   77             else if (isA<wedgePolyPatch>(boundaryMesh()[patchi]))
   78             {
   79                 nSolutionD = 2;
   86             }
   87         }
   88     }

followed by the fvDOM.C which determines if the solution is for 2D or 3D problems

if (nSolutionD == 3)
{
nRay_ = 4*nPhi_*nTheta_;
IRay_.setSize(nRay_);
scalar deltaPhi = pi/(2.0*nPhi_);
scalar deltaTheta = pi/nTheta_;
label i = 0;
for (label n = 1; n <= nTheta_; n++)
{
for (label m = 1; m <= 4*nPhi_; m++)
{
scalar thetai = (2.0*n - 1.0)*deltaTheta/2.0;
scalar phii = (2.0*m - 1.0)*deltaPhi/2.0;
IRay_.set
(
i,
new radiativeIntensityRay
(
*this,
mesh_,
phii,
thetai,
deltaPhi,
deltaTheta,
nLambda_,
absorptionEmission_,
blackBody_,
i
)
);
i++;
}
}
} ............



Any comments?

[ashrufsyed]

I know it has been a long time since you posted but, did you try to add a wedge specific boundary condition for fvDOM in OpenFOAM?

[Junjun Guo]

Hi, did anyone know how to solver this issue? wedge/Asix BC for fvDOM model?