[laplacian]

Dear Foamers,
in the new release OF 2.2 new wall-functions are available for epsilon and K.
They seem damping-functions for both Low and High Re mesh.
My question are:
1) Which is the theoretical base for the equation of K in viscous sub-layer (i.e. where come from the numerical coefficients)?

PHP Code:

(kLowReWallFunctionFvPatchScalarField.C)
00208             scalar C = 11.0; 
00209             scalar Cf = (1.0/sqr(yPlus + C) + 2.0*yPlus/pow3(C) - 1.0/sqr(C)); 
00210             kw[faceI] = 2400.0/sqr(Ceps2_)*Cf; 


2) Is it valid for SST k-w model (y+<5)?

Thanks for any advise.

[fumiya]

Hi,

1) You might want to read the paper(eq. (28)):
http://www.os-cfd.ru/cfd_docs/wall_f..._functions.pdf

2) I think this wall function is designed for use with v2f turbulence model.
I don't know whether or not it is valid to use this wall function with other turbulence models.
Is there anyone familiar with this area?

Hope this helps,
Fumiya

[laplacian]

Originally Posted by fumiya

Hi,

1) You might want to read the paper(eq. (28)):
http://www.os-cfd.ru/cfd_docs/wall_f..._functions.pdf

2) I think this wall function is designed for use with v2f turbulence model.
I don't know whether or not it is valid to use this wall function with other turbulence models.
Is there anyone familiar with this area?

Hope this helps,
Fumiya

Thak you Fumiya for your fast reply.
Reading the paper really helped me.

[xgarnaud]

I am struggling to reproduce reference results for the skin friction on a Naca airfoil (http://turbmodels.larc.nasa.gov/naca0012_val.html), so I would like to understand better the wall functions in OpenFOAM, in particular for well resolved cases.

in kLowReWallFunction, the friction velocity is computed as
scalar uTau = Cmu25*sqrt(k[faceCellI]);
for both the viscous and log layers, while in nutLowReWallFunction it is computed as
sqrt(nuw*mag(Uw.snGrad()))

In the reference paper I did not find how uTau should be computed.

Are the two approaches consistent? If not, do you know how it should be done? Should the wall shear stress be computed from uTau or, as it is done in wallShearStress from (nu + nut)*grad(U)?

Thank you