[giammy92]

Hi guys, i would like to understand if my assumptions are corrected. I have a mesh of an airfoil with y+ around wall <1. I want use k omega SST. This model considers flux turbulent fully developed, so i want to understand if i have to consider the flux like was in y+> 30 and then use wall functions?

[giammy92]

hi guys, i have a mesh with y+<1 a i use k omega SST. How you valutate this BC on wall? K_wall=1e-12 omega_wall=1e+7 according this post:

Originally Posted by vkrastev

We have to clarify this point, by distinguishing the local and global low-Reynolds number concept. The wall-issue, i. e. turbulence suppression by viscous (and also non-viscous) damping is a local feature of the flow; the turbulent nature of a fully developed boundary layer, with a viscous sub-layer, a buffer layer and a logarithmic layer is a global feature of the flow. In terms of recovering the correct wall asymptotic behaviour of a fully turbulent boundary layer, which means correctly describing the viscous sub-layer (where the local turbulent Reynolds number is very low), all the k-omega type models are perfectly able to do it on a sufficiently fine mesh (at least 4-5 nodes below y+=5) and provided the correct BC (sufficiently high omega value at the wall), without any damping functions (which is not the case, for instance, for the standard k-epsilon model): it is clearly verified and demonstrated in Wilcox's standard k-omega original paper (1988), it applies to the SST model as well, I have verified it with the standard OF implementation of the k-omega and k-omega SST models, you can also verify it by your own (just run a flat plate BL case, with k_wall=very low value and omega_wall=very high value, for instance 1e-12 and 1e+7 respectively, nut_wall should be "calculated"). Wilcox itself says this is most likely due to mathematical behaviour of the omega differential equation rather than the physical "soundness" of the model as a whole, but the practical consequence is there: it just works.

But i have a problem of floating point and my values become very high. This problem is due to fact that BC on wall are wrong or I have to change something in fvsolution?

[Kina]

You can use a very small, non-zero fixedvalue for k, omegawallfunction for omega and nutlowrewallfunction for nut. These are my preferred lowRe mesh wall BCs for the variables.

[giammy92]

I've set nut to calculated. And I have a doubt if my fvschemes and fvsolution are correct:

Fv schemes

ddtSchemes
{
default steadyState;
}

gradSchemes
{
default Gauss linear;
}

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

laplacianSchemes
{
default Gauss linear corrected;
}

interpolationSchemes
{
default linear;
}

snGradSchemes
{
default corrected;
}

wallDist
{
method meshWave;
}


Fv solution

solvers
{
p
{
solver GAMG;
tolerance 1e-06;
relTol 0.1;
smoother GaussSeidel;
nPreSweeps 0;
nPostSweeps 2;
cacheAgglomeration true;
nCellsInCoarsestLevel 10;
agglomerator faceAreaPair;
mergeLevels 1;
}

U
{
solver smoothSolver;
smoother GaussSeidel;
nSweeps 2;
tolerance 1e-08;
relTol 0.1;
}

k
{
solver smoothSolver;
smoother GaussSeidel;
tolerance 1e-8;
relTol 0.1;
nSweeps 1;
}

omega
{
solver smoothSolver;
smoother GaussSeidel;
tolerance 1e-8;
relTol 0.1;
nSweeps 1;
}
}

SIMPLE
{
nNonOrthogonalCorrectors 0;
pRefCell 0;
pRefValue 0;

residualControl
{
p 1e-5;
U 1e-5;
k 1e-5;
omega 1e-5;
}
}

relaxationFactors
{
fields
{
p 0.3;
}
equations
{
U 0.3;
k 0.3;
omega 0.3;
}
}