CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > Software User Forums > ANSYS > FLUENT

turn off wall functions in Transition SST model?

Register Blogs Members List Search Today's Posts Mark Forums Read

Like Tree3Likes
  • 1 Post By Far
  • 1 Post By vinerm
  • 1 Post By vinerm

Reply
 
LinkBack Thread Tools Search this Thread Display Modes
Old   May 22, 2012, 18:23
Default turn off wall functions in Transition SST model?
  #1
Member
 
john
Join Date: Aug 2010
Location: pasadena, va, usa
Posts: 36
Rep Power: 13
johnp is on a distinguished road
Hello,

I would like to turn off the wall functions in the new Transition SST turbulence model in FLUENT. Most of my experience has been with the SST k-omega model. When I turn off the wall functions, I use the command:

(rpsetvar 'kw-set-wall-w? #t)

Is there a similar command to turn off the wall functions for the transition model? I can't seem to track down the source for the command for the k-omega (but it works).

The flow problem I am looking at is the impingement of a submerged circular jet on a normal, flat plate (jet height/nozzle diameter = 15). The fluid is water and the jet Reynolds number is roughly 40,000. Since this problem should contain both laminar and turbulent flow, I am curious how the transition model will compare with the SST k-omega. The main quantity of interest is the wall shear stress in the impingement region. any suggestions are welcome. Thanks,

john
johnp is offline   Reply With Quote

Old   May 23, 2012, 02:20
Default
  #2
Far
Super Moderator
 
Sijal
Join Date: Mar 2009
Location: Islamabad
Posts: 4,549
Blog Entries: 6
Rep Power: 51
Far has a spectacular aura aboutFar has a spectacular aura about
Send a message via Skype™ to Far
Yes. Use y+ below 1 and wall function will be turned off!!!


PS. Transition model works only for Y+ < 1
metmet likes this.
Far is offline   Reply With Quote

Old   May 23, 2012, 13:05
Default
  #3
Member
 
john
Join Date: Aug 2010
Location: pasadena, va, usa
Posts: 36
Rep Power: 13
johnp is on a distinguished road
Thanks for the reply. I'm glad to hear that the wall functions turn off automatically. I was surprised when I found that FLUENT will use wall functions with the SST k-omega regardless of y+.
johnp is offline   Reply With Quote

Old   May 23, 2012, 13:31
Default
  #4
Far
Super Moderator
 
Sijal
Join Date: Mar 2009
Location: Islamabad
Posts: 4,549
Blog Entries: 6
Rep Power: 51
Far has a spectacular aura aboutFar has a spectacular aura about
Send a message via Skype™ to Far
http://www.ansys.com/Products/Simula...lence+Modeling


Quote:
4.13.6.*Enhanced Wall Treatment ω-Equation (EWT-ω)


Unlike the standard epsilon-equation, the omega-equation can be integrated through the viscous sublayer without the need for a two-layer approach. This feature can be utilized for a y*-insensitive wall treatment by blending the viscous sublayer formulation and the logarithmic layer formulation based on y*. This formulation is the default for all omega-equation based models.
Far is offline   Reply With Quote

Old   May 24, 2012, 10:28
Default
  #5
Senior Member
 
Bionico's Avatar
 
Flavio
Join Date: Sep 2011
Location: Brescia, Italy
Posts: 181
Rep Power: 12
Bionico is on a distinguished road
Hello!
I'd want to try SST Transitional model, but my system has y+ 2.5 in some zones: does this situation hardly affect the results? or can I use this model anyway? (I think that's impossible to refine more the mesh because of memory limits...)

Thanks in advance!
Bionico is offline   Reply With Quote

Old   May 24, 2012, 10:47
Default
  #6
Far
Super Moderator
 
Sijal
Join Date: Mar 2009
Location: Islamabad
Posts: 4,549
Blog Entries: 6
Rep Power: 51
Far has a spectacular aura aboutFar has a spectacular aura about
Send a message via Skype™ to Far
No need to refine the mesh. Just put the first node below Y+< 1. This is mandatory. There are other restrictions are also there e.g. stream-wise refinement
Far is offline   Reply With Quote

Old   November 13, 2014, 05:40
Default
  #7
Member
 
Naimah
Join Date: Nov 2011
Posts: 58
Rep Power: 12
wanna88 is on a distinguished road
Hi

May I know,

a) For transition SST 4 equations, in what conditions we should use the "Roughness correlation" as well as " Curvature Correction"?
b) Besides, how to calculate/specify for Geometry roughness height?

Thank you

Naima
wanna88 is offline   Reply With Quote

Old   May 23, 2020, 07:50
Default
  #8
New Member
 
AmiN
Join Date: Nov 2014
Posts: 13
Rep Power: 9
AmiN.D is on a distinguished road
Quote:
Originally Posted by Far View Post
Yes. Use y+ below 1 and wall function will be turned off!!!


PS. Transition model works only for Y+ < 1

You mean if our mesh is that fine than we get yPlus<1 and use K-e model the wall functions are automatically off? then what is the boundary condition for k,e at walls?

I have to simulate a plate with yPlus>30 and yPlus<1 with k-e Model in OpenFOAM
for yPlus>30 i set the boundary condition at walls " kRewallFunction and epsilonWallFunction
but when I fine the mesh to get yPlus<1 then it doesn't seem right to set them like i mentioned. I tried to put the zerogradient which i thought it's more reasonable[is it?]
+and my case didn't converge after 10000 steps. the Re Number is 1e+6.


any thoughts on this?

Last edited by AmiN.D; May 23, 2020 at 10:27.
AmiN.D is offline   Reply With Quote

Old   May 25, 2020, 13:47
Default Yplus
  #9
Senior Member
 
vinerm's Avatar
 
Vinerm
Join Date: Jun 2009
Location: Nederland
Posts: 2,946
Blog Entries: 1
Rep Power: 33
vinerm will become famous soon enough
A value of y^+ lesser than 1 has no meaning; nor is it required by any model. It would be akin to saying that turbulence is important even when Re is lesser than transition Re. Transition models also require y^+ = \mathcal{O}(1). Wall treatment does not get disabled automatically based on y^+, at least, not in Fluent, though EWT kind of does something similar. But for any CFD tool, use SWF only if y^+ > 20, else, the results would be completely unrealistic. And for any simulation, maintain y^+ >= 1, else, convergence would be difficult.
AmiN.D likes this.
__________________
Regards,
Vinerm

PM to be used if and only if you do not want something to be shared publicly. PM is considered to be of the least priority.
vinerm is offline   Reply With Quote

Old   May 25, 2020, 18:18
Default
  #10
New Member
 
AmiN
Join Date: Nov 2014
Posts: 13
Rep Power: 9
AmiN.D is on a distinguished road
Quote:
Originally Posted by vinerm View Post
A value of y^+ lesser than 1 has no meaning; nor is it required by any model. It would be akin to saying that turbulence is important even when Re is lesser than transition Re. Transition models also require y^+ = \mathcal{O}(1). Wall treatment does not get disabled automatically based on y^+, at least, not in Fluent, though EWT kind of does something similar. But for any CFD tool, use SWF only if y^+ > 20, else, the results would be completely unrealistic. And for any simulation, maintain y^+ >= 1, else, convergence would be difficult.

That's the problem i've been given and can not decide about yPlus. Although we know that K-e model is just introduced to get faster result [because of the bigger mesh] near the wall, so I think y=<1 can be meaningful because we want to resolve the B.L completely and dont use the approximation.


anyways I got the case converged but i'm not still sure about the e,k B.C in case of y<5 .
Any help would be appreciated
AmiN.D is offline   Reply With Quote

Old   May 26, 2020, 11:10
Default Rans
  #11
Senior Member
 
vinerm's Avatar
 
Vinerm
Join Date: Jun 2009
Location: Nederland
Posts: 2,946
Blog Entries: 1
Rep Power: 33
vinerm will become famous soon enough
That's a wrong notion that RANS or EVM models are introduced to get faster results or are expected to be used with coarse mesh. There is no such assumption behind development of these models. The only assumption in EVM is that the turbulence is isotropic and non-EVM RANS, such as, RSM don't even have that assumption.

And when it comes to wall treatment, it is not directly linked with turbulence model; even LES requires wall treatment. y^+ is a non-dimensional (Reynolds) number and for almost all industrial fluids, theoretically as well as experimentally, it is found that u^+ = y^+ up to y^+ of 5. And if it is linear within this limit, it does not matter if you have 10 points or just 1 point, the line would be same. So, y^+ being smaller than 1 is an overkill and does not help within anything.

Boundary conditions for both k and \varepsilon at the wall is 0.
AmiN.D likes this.
__________________
Regards,
Vinerm

PM to be used if and only if you do not want something to be shared publicly. PM is considered to be of the least priority.
vinerm is offline   Reply With Quote

Old   May 26, 2020, 14:57
Default
  #12
New Member
 
AmiN
Join Date: Nov 2014
Posts: 13
Rep Power: 9
AmiN.D is on a distinguished road
Quote:
Originally Posted by vinerm View Post
That's a wrong notion that RANS or EVM models are introduced to get faster results or are expected to be used with coarse mesh. There is no such assumption behind development of these models. The only assumption in EVM is that the turbulence is isotropic and non-EVM RANS, such as, RSM don't even have that assumption.

And when it comes to wall treatment, it is not directly linked with turbulence model; even LES requires wall treatment. y^+ is a non-dimensional (Reynolds) number and for almost all industrial fluids, theoretically as well as experimentally, it is found that u^+ = y^+ up to y^+ of 5. And if it is linear within this limit, it does not matter if you have 10 points or just 1 point, the line would be same. So, y^+ being smaller than 1 is an overkill and does not help within anything.

Boundary conditions for both k and \varepsilon at the wall is 0.



thank you
I would try also to discuss this with my prof. maybe there is a reason is just to test us!
AmiN.D is offline   Reply With Quote

Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
k-omega SST and wall functions in OpenFOAM-1.5-dev vaina74 OpenFOAM Running, Solving & CFD 21 September 29, 2016 11:26
incorrect to use nuTilda wall functions for k-epsilon? newToOpenFoam OpenFOAM Pre-Processing 0 December 14, 2011 06:15
Implementation details of SST k-omega model Niels Main CFD Forum 3 July 3, 2008 18:45
Multicomponent fluid Andrea CFX 2 October 11, 2004 06:12


All times are GMT -4. The time now is 12:19.