CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   Yplus (http://www.cfd-online.com/Forums/cfx/21483-yplus.html)

Kali Charan Nayak July 27, 2005 02:51

Yplus
 
Hi all I have a general query on wall function approach. Suppose I'm interested in getting a qualitative flow field and have yplus 300-400, how good is the predicted flow field using any turbulence model. I use RNG k-epsilon. I'm not interested in calculating wall shear stress also. In a nutshell, I want to know what is the guideline to use max yplus while using wall function approach.If one uses higher than that, what is the expected error & why?

Thanks, Kali

Glenn Horrocks July 27, 2005 18:25

Re: Yplus
 
Hi,

In general y+ can go up to 100-500, but it is case dependant. For instance near-wall resolution can affect the separation points predicted by the turbulence models, leading to large scale changes in the flow. Also, different turbulence models can predict different separation points which again lead to large scale changes in the flow. Therefore it is not possible to make generalisations as it is case dependant.

Can you describe your flow a bit more?

Glenn Horrocks

Kali Charan Nayak July 28, 2005 02:38

Re: Yplus
 
The domain consists of a rotor and stator. The rotor rotates at about 14000 rpm. The fluid will have high swirl. I'm using RNG k-e model. There are many recirculating zones. My question is if the yplus is high (300-500), how reliable is the predicted flow field. What is the recommended yplus and why is that? The energy equation is also solved. How good will be the rotor work done on the fluid with this high yplus. Hope this give you more info. to help me.

Thanks, Kali

Glenn Horrocks July 28, 2005 19:40

Re: Yplus
 
Hi,

Many turbulence models have a problem with swirling flows. You may have to use a Reynolds Stress model or DES/LES to get accurate answers as no two-equation turbulence model in CFX is strong in modelling swirling flows. This problem has nothing to do with y+, it is the inherent inability of the simple models to handle non-isotropic turbulence.

Regards, Glenn Horrocks


All times are GMT -4. The time now is 14:31.