digital filter
 

Hi,

I am new in OpenFoam and I am starting my first LES simulation of a jet flame in turbulent co-flow with OpenFoam.

I am wondering if someone has successfully implemented digital filter based generation of inlet data by Klein (M.Klein, A. Sadiki, and J. Janicka, J. Comput. Phys. 186:652-665,2003) and would be kind to share it? It would be very helpful.

Also, is there some alternative for such kind of boundary conditions implemented already in OpenFoam 2.1.x?

Thanks.

Ivana

Hi,

there is the code from LTT-Rostock, built on the idea from Kornev... there is some paper available. The code is attached here.

I have a general feeling that the method of Klein is better, but I have not seen it in OpenFOAM, nor implemented it.

Regards,
Kalle

Thank you for this! I will try it.

Best regards,
Ivana

Hi Kalle,

Could you please tell me what are:

R uniform (1 0 0 1 0 1); and fluctuationScale (1 1 1)?

It is give as input to turbulence generator.

Second is turbulence intensity?

Thanks a lot.

Ivana

Hi!

Not too sure what the entries mean, it's been a while since I used this code. I'd guess both are related to turbulence intensity, and R meaning the reynolds stress tensor. I have no clue about dimension of these quantities. You can check the code to find out, or set up at small test case and run it to see what statistics it generates.

If you find out, please post it here.
K

Hi Kalle,

I have one more question. I think something is going wrong, maybe you could help me?! When I run simulation with this inflowGenerator number of spots increases in time!? Also fluctuations increase, what should not happen. Any idea what can be problem?

Thanks! I.

Hi!

Number of spots can probably fluctuate, but if they constantly increase in time it seem strange. I did not monitor this, so I do not know too much about expected behaviour.

What do you mean by increase of fluctuations? Can you sample in some points and plot the evolution of U during, say 20, integral time scales?

K

Hi,

Indeed number of spots can fluctuate I know. But at one point it starts only increasing, and it stays like that. So that velocity becomes almost 10 times higher where it should actually only increase for max 10%...

Thanks.

Dear All,

I am also trying implementing the digitial filter method into the openfoam. Is there anybody also working on that?

In the paper Klein et al 2003 Journal of Computational Physics, I have a question about why the following form of filter coefficients i.e. Eq. (14), are given (although the authors mentioned that it is only approximately valid).


best
flames

Hi,

I am also now looking into it, however I didn't get far yet. Would you maybe share your implementation with me? Thank you.

I am not sure why they choose these filter coefficients, but from my experience the method works good with them.

Regards,

Ivana