CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   Turbulence Model for Rectangular Duct Flow (http://www.cfd-online.com/Forums/fluent/36752-turbulence-model-rectangular-duct-flow.html)

ken May 24, 2005 04:53

Turbulence Model for Rectangular Duct Flow
 
What's the most suitable turbulence model for simulating flow (Re ~ 2000 to 6000) in a rectangular duct with angled inlet and outlet?


Mehdi Ghoddosy May 24, 2005 11:18

Re: Turbulence Model for Rectangular Duct Flow
 
The k-e RNG seems to be the best. it contains enhancements for low reynolds number flows and aslo for swirling flow befor and next to your inlet and outlet(curved streamlines). However you know that RSM works better for these flows but solution time will be increased about 170-200%. Also be aware about you wall functions. If you have outlet that causes swiling flows you have to use non-equilibrium wall functions or enhanced wall treatment.

Best Regards

ken May 24, 2005 16:56

Re: Turbulence Model for Rectangular Duct Flow
 
Since wall functions are required, will they still be accurate, if I want to compute say the shear stress at the wall?

I was thinking perhaps v2f ro kw might be more appropriate. no?

Mehdi Ghoddosy May 25, 2005 00:51

Re: Turbulence Model for Rectangular Duct Flow
 
If your field is a near wall case and you just want to model near wall, using k-w could be the best. But, if you want to model core flow sorrounded by viscous sublayer and buffer layer k-w is not good selection. If you choose appropiate wall functions the wall shear stresses are accurate too. However, i think in duct flow there is a core flow soorounded by viscous near wall layers, but it may be just visous near wall layers depending on size and dynamic spc. of problem.

Regards


ken May 26, 2005 20:39

Re: Turbulence Model for Rectangular Duct Flow
 
Suppose the calculated Re is less than 1,000 but because of vibrations (say from the pump), a turbulence model is applied. In this case, are near wall models more appropriate (i'm thinking sa or kw)?


All times are GMT -4. The time now is 04:26.