low reynolds k omega / wall functions

stawrogin · November 25, 2009, 08:20

Hi,

I make my first steps with low-Reynolds models for heat transfer analysis.

When I understand the new wall function approach in OF 1.6. correctly, I “simply” can switch off the wall functions e.g. by using “calculated” in the nut-file and some reasonable bcs for omega and k at the walls. I build some test cases with grids of y+ below “1” and used fixed values for k (I used very low non-zero values) and zero gradient for omega. Does this make sense? But the convergence is quite poor and the results (e.g. when I estimate wallShear stresses from different models) differ strongly. Any hints or recommendations?
I also read that Fluent uses a low Reynolds correction for the k-omega model. Did anybody implement something like this in OF or compared the effect?

Thanks in advance

Stawrogin

Anne Lincke · March 30, 2011, 07:04

Hey Stawrogin,

" fixed values for k (I used very low non-zero values) and zero gradient for omega. Does this make sense?"

Fixed values for k which are non-zero but low are good but zero gradient for omega does not make sense.
Have a look at the post

http://www.cfd-online.com/Forums/ope...tml#post296705

November 25, 2009, 08:20	low reynolds k omega / wall functions	#1
stawrogin Member Join Date: Nov 2009 Posts: 36 Rep Power: 17	Hi, I make my first steps with low-Reynolds models for heat transfer analysis. When I understand the new wall function approach in OF 1.6. correctly, I “simply” can switch off the wall functions e.g. by using “calculated” in the nut-file and some reasonable bcs for omega and k at the walls. I build some test cases with grids of y+ below “1” and used fixed values for k (I used very low non-zero values) and zero gradient for omega. Does this make sense? But the convergence is quite poor and the results (e.g. when I estimate wallShear stresses from different models) differ strongly. Any hints or recommendations? I also read that Fluent uses a low Reynolds correction for the k-omega model. Did anybody implement something like this in OF or compared the effect? Thanks in advance Stawrogin

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Low Reynolds Number Instability in icoFoam - Cavity Case	kmooney	OpenFOAM Running, Solving & CFD	2	October 8, 2009 15:09
"OF uses logarithmic wall functions always" - does it changed in 1.6?	yuhai	OpenFOAM Running, Solving & CFD	2	September 17, 2009 13:16
LES with wall functions, wall boundary, Re 10000	subba	CFX	4	July 10, 2007 06:04
Wall functions	Confused	Main CFD Forum	1	August 14, 1998 10:31
Wall functions or low-Re models?	J. P.	Main CFD Forum	7	July 9, 1998 12:11

March 30, 2011, 07:04		#2
Anne Lincke Senior Member Anne Gerdes Join Date: Aug 2010 Location: Hamburg Posts: 168 Rep Power: 16	Hey Stawrogin, " fixed values for k (I used very low non-zero values) and zero gradient for omega. Does this make sense?" Fixed values for k which are non-zero but low are good but zero gradient for omega does not make sense. Have a look at the post http://www.cfd-online.com/Forums/ope...tml#post296705