[ngj]

Hi Daniel

As far as I can see, then your wallShearStress is actually U_tau, as you apply the "sqrt". Then you take the square root of the wSS (I assume it is the above mentioned wallShearStress), hence you have defined

tau = rho U_tau^4

which obviously leads to erroneous results.

Best regards,

Niels

[lakeat]

O, yes, my mistake.

Thank you Niels!

Another question, how did you get coefficient Cf? Using sum(wallShearStress) or using file gradP.raw, which is the easist way that you went?

[gaby]

Hi Daniel!

Do you use the "wallShearStress" utility for calculate U_tau in your LES cases? I've tried to use it in my LES, but that utility it is defined for RASproperties.....
Am I missing something?

Thanks

Gaby

[lakeat]

from my post " June 16, 2009, 03:26", you will see how to get wallshearstress;
then I do an average of wallshearstress;
then sqrt(wallshearstress);

[leonardo.morita]

Another option, as already suggested by Sebastian, is to get gradP and then:

Tau = (dP/dx)*h
utau = (Tau/rho)^(1/2)
Retau = utau*h/nu

where h is the half channel width. Note that OpenFOAM gives us gradP already divided by rho, so you can just do:

utau = (gradP*h)^(1/2)
Retau = utau*h/nu

Obs: these equations are valid for a flat plate; those presented by Sebastian are valid for a cylindrical tube.

[gaby]

Hi
Daniel and Leonardo, thanks for your answers.

The problem that I'm solving is about a flow with pressure gradient (from high P to low P). So, I have 2 questions:

1) In my case, some times gradP is > 0 , and the eq for calculate utau is:

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

and numerically this is no possible: sqrt(a), if a<0.

Am I doing something wrong? Can I solve with LES this kind of pressure gradient cases?

2) If I want to define the Ubar in transport properties, this is related to the U at the inlet or to the U at the Outlet?

Thanks in advance

Gaby

[lakeat]

Hi Gaby,

Sorry, I am not familiar with your case, so I cannot give much help.

But are you sure channelOodles is suitable for your FPG case?
I guess equations like "utau = sqrt( -D/(4*rho) * gradP )" should be best understood in a averaging sense, so I think my approach, using wall shear stress is more universal.
I believe LES (like oodles) is able to achieve your goal, so what's your b.c.?

Ubar means U_bulk, but eventually, we need to get a desired Re_tau based on u_tau, this can be only done in LES with fine mesh. Ubar gives the mean value of the initial velocity profile, note inlet and outlet are but periodic conditions.

I'm not sure this helped very much.

Regards,

[gaby]

Hi! Thnx for the answer again!

Now, I get the idea about Ubar… And after running the solver during the night I’m getting gradP<0 …

In my case I have a channel with high pressure at the inlet and low pressure at the outlet, also the channel has a decreasing diameter (like a nozzle).
Now, I think that I should use the oodles solver in spite of the channelOodles, like in the example of the diffuser (in Eungene’s thesis).

Do you have any suggestion??

Cheers!

Gaby

[sandy]

I know nothing about LES method, however, I want to use it as the tubulence model. What is the key I need to pay attention to? How to set the BC about it, for example k file. Could you give some advice?

[lakeat]

Hi Gaby!

I am not familiar with that case, but I think it's not that hard, you can figure it out, at least you can find reference from his paper.

And to Sandy, I think your question is too big to answer, if you are sure LES is really what you want, then pick a small standard test-case as starting point to taste it. For example, the turbulent channel flow.

Best,

[sandy]

Hi Daniel,

I am reading your past post (http://www.cfd-online.com/Forums/ope...043-les-7.html ), you gave the boudary conditions about k and nuSgs as follows:

<pre>*******************initial k*********************************start
internalField uniform 0;
boundaryField
{
cylinder
{
type fixedValue; //zeroGradient????
value uniform 0;
}
lateral1
{
type symmetryPlane;
}
lateral2
{
type symmetryPlane;
}
inlet
{
type fixedValue;
value uniform 2e-05; //zero or zeroGradient or a small value??
}
outlet
{
type inletOutlet; //zeroGradient or what..?? Why inletOutlet is preferred?
inletValue uniform 0; //If inletOutlet, then how to set these 2 value properly?
value uniform 0;
}
slip
{
type slip;
}
}

*******************initial k***********************************end

*******************initial nuSgs*****************************start
internalField uniform 0;
boundaryField
{
cylinder
{
type zeroGradient;
}
lateral1
{
type symmetryPlane;
}
lateral2
{
type symmetryPlane;
}
inlet
{
type zeroGradient;
}
outlet
{
type zeroGradient;
}
slip
{
type slip;
}
}

*******************initial nuSgs*******************************end</pre>

Then, finally how did you change them? I also have some difficulties to specify those BCs, and I don't know the relationship between the U and P's boudary condition with the k and nuSgs' BC. Could you talk about them? Thanks.

[sandy]

You think, what is the Eugene's viewpoint: ' Unless your walls are well resolved, you should put nuSgsWallFunction on the nuSgs BCs for walls.' ?

In addition, what is the defination of the boundary condition 'InletOutlet' ? When should we choose it as BC?

[lakeat]

Originally Posted by sandy

You think, what is the Eugene's viewpoint: ' Unless your walls are well resolved, you should put nuSgsWallFunction on the nuSgs BCs for walls.' ?

Sorry, I don't quite understand your question.

Originally Posted by sandy

In addition, what is the defination of the boundary condition 'InletOutlet' ? When should we choose it as BC?

Are you refering to Channel Flow case?

Anyway, you can read these two posts:

1. http://www.cfd-online.com/Forums/ope...onditions.html
2. http://www.cfd-online.com/Forums/ope...letoutlet.html

As for Channel flow, the inlet and outlet b.c. are cyclic.

If you are simulating the other cases, I suggest your question should be submitted in another post. But as for now, here are what I think,

1. Initial condition is different from boundary condition, I suggest the following link for a first view: Turbulence free-stream boundary conditions . Initial condition is transient. Eg., in my case, a small value is ok.
2. Wall b.c., is a hard issue, it depends on your first grid height, (zero or zeroGradient).
3. Inlet b.c. gives the information about inlet turbulence condition, some cases are very sensitive to the inlet b.c., others are not, so the inlet condition is different from one another. Sensitive in channel flow, diffuser, etc., but insensitive in most wind tunnel cases, ie. a small value is ok.
4. If you are doing a case sensitive to the inlet b.c., please refer to some papers on that, it's a hot topic.
5. Outflow, many wind tunnel cases set the out boundary in a distance far away, for example in a cylinder case, 6D or 10D or 15D, it depends on your interest, but the basic idea is by comparison to choose a sound distance to remove the undesired numerical influence. and then, zeroGradient (or zero for p) can be applied in most cases.

I'm not sure this helps...

Regards,

[sandy]

Hi Daniel,

Your advice is very very useful to me. Thank you a lot.

By the way, if I specify the inlet velocity (fixValue), can I use the expression

k = (2*ck/ce)*delta^2*||D||^2

to estimate the inlet k as BC. My LES model is oneEqEddy. Could you understand my question?

[And]

Hi OpenFOAM channel flow community,

I'm working on LES of turbulent channel flow and so I'm deeply interested in your discussions and in particular on the validation on the OpenFOAM results.

As I'm concerned in the study of both spatial than temporal discretizations OpenFOAM schemes for LES, my doubt is if selecting the flag CrankNicholson in ddt schemes would mean the correct use of such scheme as in the FOAM ProgrammersGuide P-43 it is said that:

"The Crank Nicholson scheme can be implemented by the mean of implicit and explicit terms:

solve
(
fvm::ddt(phi)
==
kappa*0.5*(fvm::laplacian(phi) + fvc::laplacian(phi))
)
".

If is this so, I'have to rewrite the left hand side of the corresponding UEqn.H exploiting both fvm and fvc operator in order to implement the Crank Nicholson scheme. Otherwise I would not have an effective Crank Nicholson time integration schemes and consequently losing the second order time accuracy.

Any hints about that

Regards

Andrea

[AirS]

Hi Foamers,

My simulation uses the LES Smagorinsky turbulent model. I would like to use cubRootVol for delta and the VanDriest damping function with it. Is that possible? How can I do that?

Here is my LESProperties file:
LESModel Smagorinsky;
delta cubeRootVol;
printCoeffs on;
...

Thanks for your help,

[lakeat]

Easy, you just need to choose VanDriest

Regards,

[AirS]

So the choice of vanDriest for delta permits to use cubeRootVol for delta along with the Van Driest damping function ?
Another question: Do you know where I can change the Cs coefficient ? I read somewhere it is equal to 0.13, but I'd like to change it to 0.1.
Thanks!

[lakeat]

1. Yes
2. By ajusting Ck you can change Cs, I suggest you read some papers to get an idea about their detailed relationship.

Regards,

[operita]

Dear Cedric, I want to run a LES of a turbulent channel flow and would need to figure out how the boundary conditions should be set to correctly represent the periodicity in x and the homogeneous span wise direction. Could you suggest where I can find info specifically about boundary conditions to be set in periodic channel flow? The version I use of OpenFOAM is 2.1.0. Regards