CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM Programming & Development (http://www.cfd-online.com/Forums/openfoam-programming-development/)
-   -   about programming a new B.C. (http://www.cfd-online.com/Forums/openfoam-programming-development/97899-about-programming-new-b-c.html)

garrison February 28, 2012 05:08

about programming a new B.C.
 
Hi, I'm trying to programming a new LES wall condition, It's formula is exactly the same as RAS wall function omegaWallFunction.

you may find it src/turbulenceModels/incompressible/RAS/derivedFvPatchFields/wallFunctions/omegaWallFunctions

This wall functions needs y+(yplus), but LESModel has not defined y+, I have to comply a new LESmodel using the y+ from RASModel (let's call it myLESModel, and I got a library called myLESModel.so), and after that, when complying the omegaWallfunction for LES, how can I including the head file?

for example, in RAS-omegaWallFunction, it begins with:
#include "omegaWallFunctionFvPatchScalarField.H"
#include "RASModel.H"
#include "fvPatchFieldMapper.H"
#include "volFields.H"
...
now I want to change RASModel.H to myLESModel.H, but it doesn't work..

and another question is, when i compile the new wall function, what should I write in make/file?

PeS February 28, 2012 06:22

There is an easy (but uncool) way how to do it. Download OpenFOAM 1.5, go to kOmegaSST source and you will find several *.H files. These files are included in kOmegaSST.C and work as omegaWallFunction so you can include them in the very same (there may be some changes in the code, depending on your OF version) way in your code. It is not nice... but easy.

Are you working on the SST-DDES model? Which version? With F1, F2 or the new rd blending? Do you intend to use some nuSgs wall function as well? And what is your application?

Good luck.
Petr

garrison February 28, 2012 08:34

Hi Petr, thanks for your suggestion. it sounds like kOmegaSST in openFoam 1.5 will act in the wall boundary like what omegawallfunction does, am I right? I have already change the omegawallfunction(RANS) to LES, but the problem is the function uses ypls which is not included in LESModel, so anyway I have to compile a new LESModel(say myLESModel).

Indeed I am working on SST-DDES with wing tip vortex, with omegafunction i think maybe i can use some relatively coarse grid to get the correct results.

I have no idea about F1,F2 or rd version, I am using the standard MSST model and I found the code in RANS-kOmegaSST and changed it into a DES one. the Delay blending function is with F2, i find the function in F.R. Menter's paper, 2004.

would you pls tell me what's the new rd bending? is it the same rd from SA-DDES?

Regards
Garriosn

PeS February 28, 2012 08:57

Hi, the rd function is very similar to SA - check this article:

Development of DDES and IDDES Formulations for the k-ω Shear Stress Transport Model
Mikhail S. Gritskevich,Andrey V. Garbaruk,Jochen Schütze, Florian R. Menter


If you include the .H files in your DES model you don't need to compile new LESModel, only your DES model. The included files will work as omegaWallFunction and the ypls is computed in the included file kOmegaWallFunctionsI.H .

garrison February 28, 2012 09:39

Thanks Petr, it sounds very attractive! just one more question, with your method, how should i set the B.C. for omega? should it be "caldulated"?

and do you have those files you mentioned above? would you please be so kind to send them to me ? it can really do me a huge favour! but anyway, thanks a million.

p.s. my email adress: garrison.liang@gmail.com

garrison February 29, 2012 08:45

Hi Petr, I already found it. And i think i know how to deal with it like you told me.
once again, thanks a million!


All times are GMT -4. The time now is 04:07.