CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   Mesh requirements for k-w models (http://www.cfd-online.com/Forums/fluent/108332-mesh-requirements-k-w-models.html)

Jeremie84 October 19, 2012 15:27

Mesh requirements for k-w models
 
Hello,

I am wondering if it is relevant to use the k-w models (standard or SST) with coarse mesh near the wall (y+>50). I read that in the Fluent User's guide and I am a little confused:

"Both k-w models available in ANSYS FLUENT are available as low-Reynolds-number models as well as high-Reynolds-number models. Therefore, the mesh guidance should be the same as for the enhanced wall treatment."

Also, what the advantage of using the "Low-Re corrections" option for the k-w models? Is it relevant only for fine meshes (y+<5)?

Thank you

Jérémie

syavash October 19, 2012 15:40

Quote:

Originally Posted by Jeremie84 (Post 387571)
Hello,

I am wondering if it is relevant to use the k-w models (standard or SST) with coarse mesh near the wall (y+>50). I read that in the Fluent User's guide and I am a little confused:

"Both k-w models available in ANSYS FLUENT are available as low-Reynolds-number models as well as high-Reynolds-number models. Therefore, the mesh guidance should be the same as for the enhanced wall treatment."

Also, what the advantage of using the "Low-Re corrections" option for the k-w models? Is it relevant only for fine meshes (y+<5)?

Thank you

Jérémie



If you have a High Reynolds problem, you may implement k-w with y+ between 30 and 300. If your problem regards low Reynolds, then you can check "Transitional flow" option to take into account low Reynolds correction. in this case y+ should be in order of 1.0.

Good luck

Jeremie84 October 19, 2012 16:14

Thank you Syavash for your answer.

I know that there is not an universal rule, but is there a typical value of Re from which we can consider a high-Re problem?
I am studying a VAWT, and I have a range of Re from 100 000 to 500 000.

Thanks again

Jérémie

syavash October 19, 2012 20:17

Quote:

Originally Posted by Jeremie84 (Post 387582)
Thank you Syavash for your answer.

I know that there is not an universal rule, but is there a typical value of Re from which we can consider a high-Re problem?
I am studying a VAWT, and I have a range of Re from 100 000 to 500 000.

Thanks again

Jérémie


Actually I'm not sure, I think we should consider transitional length in which flow turns into turbulence. If that would be substantial length in comparison with the characteristic length, we may assume it to be low Reynolds regime.
But I would rather prefer to compare the numerical results with experimental ones to make sure which model I should choose.

cfd seeker October 20, 2012 03:31

Quote:

Originally Posted by Jeremie84 (Post 387571)
Hello,

I am wondering if it is relevant to use the k-w models (standard or SST) with coarse mesh near the wall (y+>50). I read that in the Fluent User's guide and I am a little confused:

"Both k-w models available in ANSYS FLUENT are available as low-Reynolds-number models as well as high-Reynolds-number models. Therefore, the mesh guidance should be the same as for the enhanced wall treatment."

Also, what the advantage of using the "Low-Re corrections" option for the k-w models? Is it relevant only for fine meshes (y+<5)?

Thank you

Jérémie

Standard and SST versions of kw model by default use Enhanced Wall Treatment(B.L is resolved all the way upto viscous sublayer) and therefore wall y+<5 are generally recommended but theses models have the flexibility to switch over to Wall functions approach once a course mesh(wall y+ > 30) is used. So you can use it either way.
But if your sure that in your case transition is taking place then don't use "Low Re. No" option because in Fluent 13 they themselves have mentioned that "Low Re. No" does not perform good in such a situation. In Fluent 12 onward they have incorporated two new Turbulence models to properly capture Transition phenomenon, these models are Transition k-kl-w(3 eq. model) and Transition SST kw(4 eq. model). But again these models give you an extra headache because of their stringent mesh requirements i.e
1. Wall y+ <1
2. Sufficient Mesh resolution in streamwise direction
3. Proper estimate of Turbulence Intensity at the Inlet boundaries

Regards

Jonathan June 13, 2014 09:49

Hi Seeker,

I am looking at using the Transition SST model to predict transition on a turbine endwall and aerofoil.

I was wondering whether you knew whether it is a good idea to use the Low Re Correction's (i.e. Wilcox's damping functions) as well as the additional equations for transistion, or whether one should leave out the Low Re Corrections when using the additional transition equations.

I havent read anywhere that you shouldn't, but when i did, i didn't quite get intermittancy levels of '1' where i thought i would - i got a max of ~0.5 where i thought the BL should have been fully turbulent!

Any comments or thought based on your experience???

thanks very much
cheers
jonathan

Quote:

Originally Posted by cfd seeker (Post 387616)
Standard and SST versions of kw model by default use Enhanced Wall Treatment(B.L is resolved all the way upto viscous sublayer) and therefore wall y+<5 are generally recommended but theses models have the flexibility to switch over to Wall functions approach once a course mesh(wall y+ > 30) is used. So you can use it either way.
But if your sure that in your case transition is taking place then don't use "Low Re. No" option because in Fluent 13 they themselves have mentioned that "Low Re. No" does not perform good in such a situation. In Fluent 12 onward they have incorporated two new Turbulence models to properly capture Transition phenomenon, these models are Transition k-kl-w(3 eq. model) and Transition SST kw(4 eq. model). But again these models give you an extra headache because of their stringent mesh requirements i.e
1. Wall y+ <1
2. Sufficient Mesh resolution in streamwise direction
3. Proper estimate of Turbulence Intensity at the Inlet boundaries

Regards



All times are GMT -4. The time now is 18:51.