[cedric_duprat]

Hi Jianying and Ning,

I just want to add that the reference DNS calculation we used to compare our calculation was done with a second order code (in both time and space).
So if you use (real) second order scheme with an accurate mesh, there are no reason to get DNS data.

You can check easily your mesh (close to the wall) but for the numerics .... I don't know.

Cedric

[nzy102]

Hi Cedric:

Obviously if you refine the grid, you should be able to get better results. I compared my data with Moser's data, published in physics of fluids. I don't think they use a second-order code.

Ning

[nzy102]

Hi Cedric:

Obviously if you refine the grid, you should be able to get better results. I compared my data with Moser's data, published in physics of fluids (1999). I don't think they use a second-order code.

Ning

[jiao]

Hi Ning
I tried the one equation model, smagorinsky, and dynsmagorinsky, but above my results used the dynsmagorinsky model, not smagorinsky model. my results of dynsmagorinsky model is better than one equation model.
my mesh is poorer than yours, but others are the same as yours.

Hi Cedric
My results isn't better than DNS data, I know my mesh is poor. If increasing mesh, the time of computation increases. If a third-order of the temporal discretization term is used, the time of computation is less than the time of computation of a second-order of the temporal discretization term and increasing mesh.

gave me some advices.
Jianying

[jiao]

Hi Ning
I tried the one equation model, smagorinsky, and dynsmagorinsky, but above my results used the dynsmagorinsky model, not smagorinsky model. my results of dynsmagorinsky model is better than one equation model.
my mesh is poorer than yours, but others are the same as yours.

Hi Cedric
My results isn't better than DNS data, I know my mesh is poor. If increasing mesh, the time of computation increases. If a third-order of the temporal discretization term is used, the time of computation is less than the time of computation of a second-order of the temporal discretization term and increasing mesh.

gave me some advices.
Jianying

[sek]

Has anybody tried Re_tau = 180?

[santosh]

Just Curious... I'm a complete newbie out here.. Is there any tutorial to get started with DNS/LES of Channel Flow?? What machines do you guys run your code on???

[harishg]

I tried the Re_tau=180 case and it produced acceptable results. I did some analysis at Re_tau=395 using one eq eddy and dyn one eq model. The results obtained using the localized dynamic one equation model was much better than the one eq case with Van Driest damping on a 64 cube grid with 2pi*2*pi domain and the same numerical discretization as the tutorial. Has anyone tried using their own filter width expression instead of the smooth/cuberoolvol filter ? Also majority of the papers which i came across used higher order schemes with LES and that can be another reason for the problem.

[sega]

Originally Posted by cedric_duprat

the time discretisation is Crank-Nicholson. (I think it's backward in the tutorial).
For the numerical scheme I using only central diffential scheme (to keep second order accurate).
so there is no limitedLinear 1 for these terms (div(phi,k), div(phi,B))
Then, the time step is different to keep Co number less than 0.4.
For the solver, I'm using ICCG to solve the pressure (and the same PISO as the tutorial) and BICCG for the other quantities, which is quite different from the tutorial also.

Dear Cedric.

You say you are NOT using limitedLinear 1 for the two terms mentioned.
Can you tell me which one you are using?
Ore actually post your fvSchemes dictionary?

I'm having some problem using oneEqEddy in a square duct LES.
My results are poor and Smagorinsky is far better ...
I don't think thats correct ...

Have a nice day. Sebastian

[sega]

Hello World.

As mentioned above I am doing the channel flow simulation in a square duct with cyclic bc's for in- and outflow.
My Resolution is 56x56x70 with refined mesh towards the wall so there are 7 cells within y+ < 10.

I'm using two different LES models, namely

• Smagorinsky (SMG)
• One Equation Eddy

with two different filter-widths (Delta) each time, namely

• Cube Root Volume (CRV)
• van Driest Damping

Well, the results with Smagorinsky are looking more or less good, the results with the One Equation Eddy are scrap.

I'm even experiencing that the velocity profile is not symmetric.
Unfortunately the asymmetry looks to be getting worse when simulation longer and thus doing longer averaging.

Important information on how these plots are obtained: I'm using my own post-processing tool for averaging in the flow-direction (with MATLAB). I'm not primary doubting my own tool, but is there an OpenFOAM tool for post-processing a square duct channel?

Any ideas why the One Equation model is so bad compared to both DNS and Smagorinsky? Well, I expected vice versa.

[sega]

Originally Posted by philippebv

you use the final pressure gradient in the channel to get Tau_w and then U_tau?

I would like to know what is meant by the final pressure gradient?
Is this the averaged pressure gradient at the end of the simulation?

[lakeat]

I guess he meant gradP.raw

[sega]

Originally Posted by lakeat

gradP.raw

Which is written in the last uniform sub-directory of the last time-step?
Just to make sure ...

[lakeat]

Your objective is?? Why do you care about it?

[sega]

Originally Posted by lakeat

Your objective is?? Why do you care about it?

I think you have read my previous posts and know I'm having some difficulties with my LES simulations. Just to make sure I'm not getting something wrong on the way (for example how and where to read gradP to get utau) I'm just asking.

[lakeat]

Am I missing something??? As I remember, I have never used gradP to get U_tau.

[sega]

Originally Posted by lakeat

Am I missing something??? As I remember, I have never used gradP to get U_tau.

Well, but I do.

utau = sqrt( -D/(4*rho) * gradP )

with D beeing the diameter of the Duct.

[lakeat]

I see. ......
I calculate u_taw from a wallShearStress utility, of cause in an averaging sense.

^^^^^^^^^
I guess they are the same, since the essence are the same, right?

[sega]

Originally Posted by lakeat

I guess they are the same, since the essence are the same, right?

I hope so ...

[lakeat]

Hi, can anyone tell me how did you get the U+ versus y+ data?

My results is wrong, so I lost my idea about the procedure...

1. Get wall shear stress

wallShearStress.boundaryField()[patchi] =
sqrt
(
	nuEff.boundaryField()[patchi]
   *mag(U.boundaryField()[patchi].snGrad())
);

2. Calculate U_tau, (wSS denotes wall shear stress)

uTau= Foam::sqrt(wSS);

3. Plot U+ versus y+, (nuTmp denotes the value of nuEff)

scalarField UMeanXPValues = UMeanXvalues/uTau;

makeGraph(y*uTau/nuTmp, UMeanXPValues, "UP", UMean.path(), gFormat);

Is the procedure right?