CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM Pre-Processing (http://www.cfd-online.com/Forums/openfoam-pre-processing/)
-   -   Driver region for LES (http://www.cfd-online.com/Forums/openfoam-pre-processing/65078-driver-region-les.html)

prapanj June 3, 2009 05:51

Driver region for LES
 
Hi,

I would like to run a LES case(external flow) with a turbulence upstream. In order to create turbulence upstream, I thought I will build a driver region. that has the same cross section as my domain and same number of cell faces as my domains inlet. In this driver region I think of using standard K epsilon. The idea is to have a uniform flow at the inlet and then map the outlet to the inlet so as to make it a periodic(cyclic) flow. How can I do this?
Will the initial boundary condition at the inlet remain, not letting my cyclic condition influence it?

thank you
Prapanj

cedric_duprat June 30, 2009 13:28

Quote:

Originally Posted by prapanj (Post 218066)
Hi,

I would like to run a LES case(external flow) with a turbulence upstream. In order to create turbulence upstream, I thought I will build a driver region. that has the same cross section as my domain and same number of cell faces as my domains inlet. In this driver region I think of using standard K epsilon. The idea is to have a uniform flow at the inlet and then map the outlet to the inlet so as to make it a periodic(cyclic) flow. How can I do this?
Will the initial boundary condition at the inlet remain, not letting my cyclic condition influence it?

thank you
Prapanj

Hi Prapanj

Maybe you can first have a look on the boundary condition called directMappedPatch
check in the forum you'll find tutorial and details on this BC.

Then, if you want to simulate a LES case, running a cyclic RANS geometry is not enough (of course). You will have to add specific noise, fluctuations at your average parameter.
Otherwise, you can run a previous cyclic LES calculation there, there is no need to add fluctuations but you can get problem to match your mean flow.
Well in fact, the methods depend of your case.

I hope it will help

Cedric


All times are GMT -4. The time now is 23:58.