[vinz]

Hi everybody,

I'm new in OpenFOAM and I'm tryin to solve the flat plate case. The laminar case works fine but I've some troubles with the turbulnt case. I use the k-epsilon or the launder sharma k-eps model and try to compare my Cf with experimental results obtained by Weighardt.
However, either my results fits the laminar case(instead of being turbulent) either the flow is turbulent but the Cf obtained is to high (by a factor 1.5 to 2).
Does someone here allready try to solve a turbulent flat plate? What did you choose as tubulence model? DO you still have you results for Cf?

Thanks
Best regards

Vincent

[pierre]

Are you using the lowRe k-eps model?

Pierre

[vinz]

I tried a lot of models: k-eps, Launder-Sharma k-eps, RNG k-eps. Non of them is giving good results.
I also tried to use first order or second order schemes. Second order gives obviously better results for laminar case but is still not able to cope with turbulence.
Did you allready run this case?

[dmoroian]

How good is your mesh? What is your y+?

Dragos

[vinz]

I have two meshes.
One mesh is realy fine with y+ of the order of 1e-2 or 1e-3. I use this one with turbulence models without wallfunctions.
The other one is a bit less refined at the wall, and is used with k-eps model using wallfunction. This one is giving the best results.
However, I tried both meshes with the different models and as I said, results are not good enough.

Vincent

[cosimobianchini]

I was testing Low Re k-epsilon (Chien, LienLeschziner, LamBremhorst, AbeNagano) compressible (inlet Ma = 0.2) turbulence models , over a flat plate with no pressure gradient.
I was checking the behaviour of turbulence quantities near the wall at the section with Rex = 1.03*10^7.
Unfortunatly I cannot post process my data now to obtain the Cf as a function of x but the skin friction coefficient on my test section seems to be consistent with the experimental data from Wieghardt: I got a mean value of 0.0028.
Hope this can help.
Cosimo

[marhamat]

Hi everyone

which boundry conditions is proper for turbulent flat plate modeling?
can we use empty for front and back wall?(can we model this case in 2D ) .and symetry for upper wall?
Do anybody have experience in this field?
What is your conditions for correct result?

Best Regards
Marhamat

[cosimobianchini]

Hi Marhamat, that's exactly what I did:

symmetryplane on the upper boundary,
empty for left and right boundaries,
obviously wall, inlet and outlet for the other three.

I was testing some Low-Reynolds k-epsilon compressible models in order to choose only few of them for more advanced tests.
The test with experimental data has been done properly only for kplus and epsilonplus behavior normally to the wall, with just a rough estimate for the other quantities of interest.
Aiming at catching turbulent peaks, the grid was very fine with an yplus of almost 0.1.

Hope this will help,
Best regards
Cosimo

[vinz]

Hi Marhamat,

I would also advise a slip plate just before the wall plate to avoid singularities. But as you see, I didn't get right results for the turbulent plate. At that time, I didn't have the time to go deeply in the problem so I gave up.

Then, I waiting forward to see the results that you will obtain. It would be great if you could post your results and the parameters values that you used (k, eps, Re...).

Good luck.
Best regards,

Vincent

[marhamat]

Hi

I thank you a lot for your answer
I'm interested in LES of this case
So if i get any right result i will inform you.

Best Regards
Marhamat

Anyone knows how to extract skin friction Cf from simpleFoam?

[marhamat]

Hello everybody
I used channelOodles for this case(turbulent flat plate) with below settings.
Dimentions (1000 10 1)mm for 2D analysis with two blocks
(
hex (0 1 2 3 4 5 6 7) (20 100 1) simpleGrading (1 1000 1)
hex (1 8 9 2 5 11 10 6) (80 100 1) simpleGrading (1 1000 1)
);
In block0 bed(with L=200mm) is symmetryplane

Inlet Velocitt(U)=10m/s & nu=1e-6
So: Re=8*10e6 in the end of plate

You see distribution of U after runnig and initial perturbed condition that imposed to this case.




I don't see any turbulence structures in my case.
What is your idea about this?

Best Regards
Marhamat

[ivan_cozza]

Hi foamers!
I'm having some troubles with a turbulent flat plate calculation. I'm using simpleFoam with SST k-Omega model, for a Mach 0.11 and Re 5x10^5, but I can't get the correct values for k and it seems that the boundary layer is not represented, but my y+ is up to 8!
Any idea? Maybe SST is not the best model for the flat plate problem?