CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (http://www.cfd-online.com/Forums/openfoam-solving/)
-   -   LowRe RANS turbulence models (http://www.cfd-online.com/Forums/openfoam-solving/58533-lowre-rans-turbulence-models.html)

luca January 10, 2006 14:22

Hi all, I have some doubts,
 
Hi all,

I have some doubts,

OpenFOAM supports Low-Reynolds RANS Turbulence Models?
How can I activate Low-Reynolds Turbulence Models in OpenFOAM for RANS calculations?!

Thanks

eugene January 10, 2006 14:34

Choose LienCubicKELowRe, Lie
 
Choose
LienCubicKELowRe,
LienLeschzinerLowRe
or SpalartAllmaras

as your turbulence model. Make sure all turbulent variables are set to zero (or 1e-20) on the boundary and off you go.

hjasak January 10, 2006 14:39

You're probably better off wit
 
You're probably better off with one of these instead:

- LaunderSharma. My choice - I wouldn't say it's special, but it's a very decent choice. Requires the first cell to be below y+ = 1
- LamBremhorstKE. That's the standard reference, but needs distance to the wall everywhere. More sensitive to the near-wall resolution
- QZeta. Nice - a reformulation of LaunderSharma in better variables (lower resolution requirements).

Enjoy,

Hrv

luca January 11, 2006 10:31

First, thank you very much for
 
First, thank you very much for the answers.

but when i write, for example, "LienCubicKELowRe" in turbulenceProperties file i receive this message:

FOAM FATAL ERROR : Unknown turbulenceModel type LienLeschzinerLowRe

Valid turbulenceModel types are :

7
(
kEpsilon
realizableKE
LaunderSharmaKE
laminar
RNGkEpsilon
LRR
LaunderGibsonRSTM
)

So it's not clear how can i set low-Re turbulence models in RANS compressible calculations.
Moreover, when i set LaunderSharma for Low-Re turbulence model simulation i have to digit "LaunderSharmaLowRe" in turbulenceProperties file ?!

sorry but i have some difficult

thanks in advance again

Luca

hjasak January 11, 2006 10:46

For compressible flows, we've
 
For compressible flows, we've only got LaunderSharmaKE for low-Re simulations.

constant/turbulenceProperties:

turbulenceModel LaunderSharmaKE;



I don't understand the second part of the question, sorry.

Hrv

luca January 11, 2006 11:57

sorry for the confusion in the
 
sorry for the confusion in the last phrase..

your answer is enough for my doubt.

Best regards

Luca

diego_n July 19, 2006 09:51

I am trying to use Launder-Sha
 
I am trying to use Launder-Sharma lo-Re compressible model which, standing on literature, requires a zero value for epsiln at wall.
By setting epsilon equal to 0 at wall I obtain a singularity error, while it runs using a zeroGradient boundary at wall.
Which boundary do you use for this model?

yohannperrot July 20, 2006 04:06

In any CFD software, avoid as
 
In any CFD software, avoid as much as possible setting 0 to values which are used as a denominator somewhere in the code (so as to avoid infinity values, undetermined forms and so on). Put a low number close to zero like 1.e-10 instead.

Good Luck

rolando October 4, 2007 11:06

Hi, can anyone give me the re
 
Hi,
can anyone give me the references of the following turbulence models?
-LienCubicKELowRe
-LienLeschzinerLowRe

The nomenclature of the models in the common literature might do as well.

Thanks

Rolando

nzy102 October 4, 2007 11:39

Hey Rolando: Check the link
 
Hey Rolando:

Check the link below:

http://www.cfd-online.com/cgi-bin/Op...cus/discus.cgi

Hope this will help.

rolando October 4, 2007 11:51

Hi Ning, something doesnīt wo
 
Hi Ning,
something doesnīt work with the link you posted.
Donīt copy paste the URL. Try to right click on the yellow icon in the header line of the message (beside the blue arrows) and copy the link. This should word.

Could you please try it again?

Rolando

nzy102 October 4, 2007 12:20

Do a search for liencubic and
 
Do a search for liencubic and you should be able to find my post.

Ning

rolando October 4, 2007 13:34

Thanks Ning Rolando
 
Thanks Ning

Rolando

bastil September 21, 2008 15:13

Sorry for refreshing this old
 
Sorry for refreshing this old post. I am wondering if OF has some kind of "blending" between Low- and Hi-Re formulations for turbulence as comercial codes have (all yplus wall threament in ccm+, hybrid wall function in STAR-CD or enhanced wall threadment in Fluent). I have not come over something like that so far.

braennstroem September 22, 2008 07:25

Hi Bastil, I am trying to i
 
Hi Bastil,

I am trying to implement a 'compound wall treatment' together with a low-Re turbulence model. I still got some trouble implementing it (http://www.cfd-online.com/OpenFOAM_D...tml?1219326882). Maybe it is a good case to take a look on at the 'Stammtisch' :-)

Fabian

sandy August 9, 2010 05:42

Quote:

Originally Posted by braennstroem (Post 197654)
Hi Bastil,

I am trying to implement a 'compound wall treatment' together with a low-Re turbulence model. I still got some trouble implementing it (http://www.cfd-online.com/OpenFOAM_D...tml?1219326882). Maybe it is a good case to take a look on at the 'Stammtisch' :-)

Fabian

Hi Fabian, I am interesting in this kind of blending wall fuction too. What about your work for it? Did you finish it? I am also wondering why there are not V2f-like turbulence models with the "compound wall treatment" in OF.

braennstroem August 11, 2010 13:57

I actually did not work von the v2f-model anymore as it did not converge for all my hvac test cases except for some 2D/3D diffuser cases. Did you have more success with the v2f-model and for which cases? I did some different approaches for the SST model, but never got as good results as Luca with his kOmega98-model...

DanielPT February 4, 2011 04:51

Hi,

my problem is an internal flow in a diffuser with Re_inlet=13000. My y+ is below 1, so I want to use LaunderSharmaKE as lowRe model.
Is it right to set k and epsilon to 0 at walls (at y=0) for this k-epsilon model?

Daniel

olivierG February 4, 2011 07:19

Short answer: no
(see previous post: do not use 0 but 1e-10 to avoid infinite value)
zerogradient may also work

regards,
olivier

DanielPT February 4, 2011 07:52

Ah ok, thanks a lot Oliver! I have overlooked the previous post.

Regards
Daniel


All times are GMT -4. The time now is 10:43.