[Ivangzp]

Hello,

I’m trying to validate the case NACA0012, incompressible at hight Reynolds number (6 Millions) at different angles of attacks.
I will focus on the case of 3 degrees, although the kind of errors that I’m getting are much the same for the rest (i.e. 6º,9º,12º).

I got the validation with OpenFoamversion v4.
Turbulence model: Spalart Allmaras
nu = 1.6666666666666668e-07;
U = 1m/s
Re = 6*10^6

Real Value for lift coefficient
Cl(3º) = 0,32341
Value obtained
Cl(3º) version OF4 = 0,32863

I want to insist on it, it happen the same for the rest of de AoA, the error is very short when I use OF4. So I can conclude I validate it with that OF version.

I tried to do the same with other version of OpenFOAM (OF v2006), and the values are deviated around 10%.

Real Value for lift coefficient
Cl(3º) = 0,32341
Value obtained
Cl(3º) version v2006 = 0,29336

As you can see, between both calculated, Cl(3º) = 0,32863 and Cl(3º) = 0,29336, there is an "appreciable" difference.

Same mesh, FvSchemes, FvControlDict, FvSolution, Initial Conditions, I mean every file is the same. (Find attached a picture of that mesh. Mesh1).

I tried other mesh (Find attached a picture of that mesh. Mesh2), probably with better quality (I tried dozens of them, but this is a good one), and the issue persists.
In that case I had to change the boundary conditions (Change from OGrid-Mesh to CGrid-Mesh), but phisically are exactly the same problem.
Real Value for lift coefficient
Cl(3º) = 0,32341
Value obtained
Cl(3º) version 4 = 0,32889
Cl(3º) version v2006 = 0,29693


I have been a long time with that problem and I couldn’t find a solution. I’m not sure if it is merely a difference between both version or whether I can change something to get closer results.

Each mesh has different boundary conditions and the results are pretty close, so the problem is not in the mesh or in the BC. The problem should be in FvSolution and FvSchemes which are identicall in both cases. For that reason, I copy it there here:

--------------------------------------------------------------------------------------
FvScheme:

FoamFile
{
version 2.0;
format ascii;
class dictionary;
object fvSchemes;
}


snGradSchemes{
default corrected;
}


laplacianSchemes{
default Gauss linear corrected;
}


fluxRequired{
p ;
}


gradSchemes{
default Gauss linear;
}


interpolationSchemes{
default linear;
}


divSchemes
{
default none;
div(phi,U) bounded Gauss linearUpwind grad(U);
div(phi,nuTilda) bounded Gauss linearUpwind grad(nuTilda);
div((nuEff*dev2(T(grad(U))))) Gauss linear;
}


ddtSchemes{
default steadyState;
}


wallDist {
method meshWave;
}


----------------------------------------------------------------------------------------

----------------------------------------------------------------------------------------
Fv Solution:

FoamFile
{
version 2.0;
format ascii;
class dictionary;
object fvSolution;
}


SIMPLE{
pRefPoint (-1.000000 4.990000 0.500000);
pRefValue 0;
nNonOrthogonalCorrectors 1;

residualControl{
p 1e-07;
U 1e-07;
}

}


relaxationFactors{

fields{
p 0.7;
}


equations{
U 0.7;
k 0.3;
omega 0.5;
nut 0.8;
nuTilda 0.8;
}

}


potentialFlow{
PhiRefPoint (0.000000 14.990000 0.500000);
nNonOrthogonalCorrectors 20;
PhiRefValue 0;
pRefPoint (0.000000 14.990000 0.500000);
pRefValue 0;
}


solvers{

U{
relTol 0.1;
preconditioner DILU;
tolerance 1e-10;
solver PBiCG;
}


nut{
relTol 0.1;
preconditioner DILU;
tolerance 1e-10;
solver PBiCG;
}


nuTilda{
relTol 0.1;
preconditioner DILU;
tolerance 1e-10;
solver PBiCG;
}


p{
relTol 0.1;
preconditioner DIC;
maxIter 10000;
tolerance 1e-10;
solver PCG;
}

Phi
{
solver GAMG;
smoother DIC;
cacheAgglomeration on;
agglomerator faceAreaPair;
nCellsInCoarsestLevel 10;
mergeLevels 1;

tolerance 1e-10;
relTol 0.01;
}
}



----------------------------------------------------------------------------------------

Annotations:
- The y+ number of both meshes is less than 1 in the wall.
- Force calculation is donde considering Lift in axis (0 1 0) and Drag (1 0 0). In order to have real values of Lift and Drag I rotate the axis to consider the angle of attack of 3 degrees after getting the results.
- The results are obtained after 2.000 iterations.
- The value of "Real Value for lift coefficient " it's from the NASA : https://turbmodels.larc.nasa.gov/naca0012_val.html
- I checked the official source, for that case https://www.openfoam.com/documentation/guides/latest/doc/verification-validation-naca0012-airfoil-2d.html , It didn't help me.
- I tried also with the version v2012 and the results are pretty similiar to the version v2006.
- Files (with the mesh and log files) can be find here in that link:
https://drive.google.com/drive/folders/1myUIeofFi-6YVKu9eMQrwIzhRoj5n4PA?usp=sharing

- Find attached the 0 and system folders for the case of Mesh2.

Any kind of help would be very appreciated.