difference between homogeneousDynSmag (OF-2.0.x) and dynSmag (OF-1.6-ext)

florian_krause · January 26, 2012, 08:53

Hi! I am refering to the homogeneousDynSmagorinsky model (named dynSmagorinsky in OF-1.6.-ext), more specifically to the following lines:

dynSmagorinsky in OF-1.6-ext, line 64, calculation of Smagorinsky constant:

http://openfoam-extend.git.sourcefor...955616;hb=HEAD

dynSmagorinsky in OF-2.0.x, line 72, calculation of Smagorinsky constant:

http://foam.sourceforge.net/docs/cpp/a08633_source.html

Both models are practically the same. The only differ by a factor of 1/2 in the above mentioned lines.

Does anyone have a clue why they differ?

Thanks,
Florian

January 30, 2012, 09:29

Hello Florian,

the factor 1/2 was missing in the older dynSmagorinsky models, since they were constructed based on the paper by Fureby where the factor is also missing. The new version (homogeneousDynSmagorinsky) is the correct version. I've compared the two versions in a minimal Channel flow and to be honest there is not much of a difference in the results. Well at least not for u+, I haven't compared anything else.

Hope this helps
Stefanie

florian_krause · February 1, 2012, 08:02

Hi Stefanie,
like couple of times before, I am wondering why they don't spend just a minute to include the publication reference in the header...

Anyway, thanks for the clarification!

Florian

January 26, 2012, 08:53	difference between homogeneousDynSmag (OF-2.0.x) and dynSmag (OF-1.6-ext)	#1
florian_krause Senior Member Florian Krause Join Date: Mar 2009 Location: Munich Posts: 103 Rep Power: 17	Hi! I am refering to the homogeneousDynSmagorinsky model (named dynSmagorinsky in OF-1.6.-ext), more specifically to the following lines: dynSmagorinsky in OF-1.6-ext, line 64, calculation of Smagorinsky constant: http://openfoam-extend.git.sourcefor...955616;hb=HEAD dynSmagorinsky in OF-2.0.x, line 72, calculation of Smagorinsky constant: http://foam.sourceforge.net/docs/cpp/a08633_source.html Both models are practically the same. The only differ by a factor of 1/2 in the above mentioned lines. Does anyone have a clue why they differ? Thanks, Florian

January 30, 2012, 09:29		#2
StSchiff Guest Posts: n/a	Hello Florian, the factor 1/2 was missing in the older dynSmagorinsky models, since they were constructed based on the paper by Fureby where the factor is also missing. The new version (homogeneousDynSmagorinsky) is the correct version. I've compared the two versions in a minimal Channel flow and to be honest there is not much of a difference in the results. Well at least not for u+, I haven't compared anything else. Hope this helps Stefanie

February 1, 2012, 08:02		#3
florian_krause Senior Member Florian Krause Join Date: Mar 2009 Location: Munich Posts: 103 Rep Power: 17	Hi Stefanie, like couple of times before, I am wondering why they don't spend just a minute to include the publication reference in the header... Anyway, thanks for the clarification! Florian