CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   Low Reynolds k-e turbluence model (http://www.cfd-online.com/Forums/cfx/18721-low-reynolds-k-e-turbluence-model.html)

Bolster November 13, 2001 10:41

Low Reynolds k-e turbluence model
 
Does anybody know how to use a low Reynolds number k-e turbulence model in CFX 5.5. I have been looking for it and am unable to find it. My project is related to airflow in an atrium and hence it is important to me that I have a model that can accurately predict low as well as high Reynolds number flows in order to accurately predict heat transfer convection coefficients. Thanks if anyone can help

Herve November 14, 2001 05:53

Re: Low Reynolds k-e turbluence model
 
Why don't you consider the Shear Stress Transport model of Menter with scaleable wall functions? If the B.L. flow is important to your application this model is extremely good, and it is also able to deal with the flow in the far field. This makes it a good model in both regions (well it's been designed for that purpose), and has been tested quite thoroughly in various applications (I saw Menter's presentation at one CFX conference).

Herve November 16, 2001 08:06

Re: Low Reynolds k-e turbluence model
 
Ooops sorry. made a mistake. The SST does not use scaleable wall functions. This is used for standard k-e to ensure no node encroach on the region of application of the law of tha wall.

Dan Williams November 16, 2001 18:28

Re: Low Reynolds k-e turbluence model
 
Sounds like you need to look at using k-omega or SST. k-omega in CFX-5.5 is the classical wilcox k-omega model, and SST is Florian Menter's blended k-epsilon/k-omega model. You can access all the turbulence models by hitting the "Fluid Models..." button on the "Domains" form.

The basic idea with SST is k-omega in the near wall region and k-epsilon in the free stream. This is not the whole story however. There is a special limiter applied to the eddy viscosity which accounts for the transport of turbulent shear stress in the boundary layer. Standard k-epsilon and k-omega tend to overpredict the eddy viscosity, hence the limiter. In addition, with the k-omega model, including SST, the 5.5 solver will automatically integrate to the wall (use a low-re formulation) where you have enough grid to resolve the boundary layer appropriately, and use wall functions otherwise. This really reduces the need to worry about y+ for example.

For more details on SST see "F. Menter, Two equation eddy viscosity models for engineering applications, AIAA Jounral 32(8), 1994"

Dan.

Steve November 19, 2001 09:29

Re: Low Reynolds k-e turbluence model
 
By "Low Reynolds number model" do you mean a model designed for flows at low global Reynolds number (i.e. a transition model)?

Or do you mean a model that integrates down to the wall?

I agree that the SST model is the model of choice for high reynolds number wall bounded flows.


All times are GMT -4. The time now is 09:47.