[ali]

Has anybody implemented turbulent inflow conditions for DNS or LES of jets in openFOAM yet? I would be grateful if there is somebody who has done this and can share it. Is the mapping between the inlet and outlet of a pipe to get a fully-developed turbulent pipe flow a sufficiently accurate approach or it is not as good as the following inflow conditions?

I guess some of the good inflow conditions are as follow:

1) N. Jarrin, S. Benhamadouche, Y. Addad and D. Laurence
Synthetic turbulent inflow conditions for Large-Eddy Simulation ,
in Turbulence, Heat and Mass Transfer 4, Hanjalic, Nagano, Tummers Edts. Begell House inc., 467-474, 2003

2) M. Klein, A. Sadiki und J. Janicka A Digital Filter Based Generation of Inflow Data for Spatially Developing Direct Numerical or Large Eddy Simulations. J. Comp. Physics,186:652--665, 2002.

[grtabor]

I have a PhD student who is working on these issues (hopefully as we speak). We have looked at 3 techniques: turbulence synthesis, library lookup and the mapping technique implemented by Eugene de Villiers in FOAM. Of these the mapping technique wins hands down; turbulence synthesis is very difficult to get the higher order moments correct (we have primarily tested in various channels) because of having to get longer range correlations correct; and the library lookup techniques are cumbersome. We are now investigating ways of generating non-fully developed inlets by driving the inlet mapping towards a pre-specified mean velocity profile (in our case with swirl, but it doesn't have to be). If you send me your email address I can send you a conference publication we had last year in ECCOMAS; we are also working on journal papers in this area. Thanks for the refs you gave though - I'll pass them on to my PhD student.

Gavin

[bernarde]

Hi Gavin

I have had a look at the documentation and the forum, but I can find a reference to the turbulence mapping you mentioned. How is this implimented in OpenFOAM? I would like to have a look at it and test it a bit.

Regards