nutRoughWallFunction v1.6 - Values for Ks and Cs
 

Hello everyone

There are already some posts about nutRoughWallFunction, e.g. "nutRoughWallFunction and Surface Roughness Documented". But still I have a problem.

The following piece of code

{
type nutRoughWallFunction::Ks;
value uniform 0.001;
type nutRoughWallFunction::Cs;
value uniform 0.5;
}

seems to be correct. If I leave the values for Ks or Cs I get an error message. Nevertheless, I always get the same result no matter which values are chosen. I have also used different meshes but the result is always the same.

Does anyone know a solution to this problem? I think I am not the only one who is interested in this topic.

Thanks in advance.


Kind regards,

Irina

nutRoughWallFunction bug in 1.6 fixed in 1.6.x
 

nutRoughWallFunction in OpenFOAM 1.6 doesn't work correctly due to a bug. Try the latest 1.6.x from the git repository for the corrected version.

For more details on using nutRoughWallFunction try:
http://www.cfd-online.com/Forums/ope...ocumented.html

Hope this helps.

Still a bug on nutRoughWallFunction?
 

Hello everybody,
I still have problems using nutRoughWallFunction with 1.6.x. More in detail, I have the same problems as posted above: whenever I change Ks (from extremely low to extremely high values), the results I get are always the same.

I am studying the flow around an airfoil at Re 1.5 millions and modeling turbulence with the KOmegaSST model. I defined the airfoil surface as:
Cs is not specified since I am using the default value of 0.5.

I tried different angles of attack and let the simulations converge till 1000 time steps. However, no changes on the aerodynamic coefficients while changing Ks. I tried different Ks, including 0 (smooth surface) and 1 meters, and also different turbulence models (kEpsilon). No changes. I am wandering if there is still some bug on the nutRoughWallFunction.

Any experience on that? Suggestions? Is there any incompatibility between schemes or parameter definition that let the simulation do not take into account the Ks?

Any help is appreciated.
Cheers,


mad

Hi,

I've already worked with rough wall function in 1.6.x in my results are fine. Changing Ks had a big influence in my case. But I'm not sure if your BC is defined correctly, because you give no value for Ks but just for nut, which is in my opinion completely ignored. I used something like
Regards,
Stefan

Hi Stefan and thanks for your answer.
I followed what posted by Irina above... However, using:
did not changed my results. Also keeping Ks fixed and changing Cs did not produce any variation.
Maybe I installed the new OF 1.6.x version not properly. Do you know how can I switch on some debug to check whether the nutRoughWallFunction references are correct? In order to check if is a problem with my openFoam, I could also run a case that is working for sure for someone else... Maybe you have a really simple case I can test... May I try it?
Thank you.

mad

Ok, switched on the debug flag (see Userguide!)
If I change the Ks, I can see that:

• KsPlus changes --> OK!
• Edash changes --> OK!
• nutw = 0 in any case!

As explained here: http://www.cfd-online.com/Forums/ope...on-v1-6-a.html, this is the cause of the bug fixing of OF 1.6.x: nutw always inizialised to zero.
Thus, probably there is something wrong with my installation...
What do you think?

mad

Hi Maddalena

I installed OpenFOAM 1.7.x and used

lowerWall
{
type nutRoughWallFunction;
Ks uniform 0.24;
Cs uniform 0.5;
value uniform 0.0;
}

as was suggested by Stefan. When changing Ks values my results changed as well. Finally, everything works fine. Maybe you should also install the latest version.


Cheers,

Irina

Hi Irina and thanks for your interest. My problem was something different, see http://www.cfd-online.com/Forums/ope...tml#post264132. Everything worked when I changed mesh resolution.
Regards,

maddalena

Does anyone of you tried Rough wall function in LES using OpenFoam?

Thanks

Hello!

I am also interested in Rough wall function in LES ...

Have you got any ideas?

Hi hfs,
my only idea is to use fixedValue for nuSgs and e.g. nuTilda or every other LES parameter. Maybe it's possible to calibrate this value to behave like a normal roughness parameter. What do you think?
Cheers

Thanks for the idea ... am not sure ... will get back to you!

Thanks!