CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   NACA 0012 K-omega model (http://www.cfd-online.com/Forums/fluent/101304-naca-0012-k-omega-model.html)

 eishinsnsayshin April 30, 2012 14:05

NACA 0012 K-omega model

Hello everyone,

I'm struggling to get a k-omega model to run a FLUENT simulation of a naca 0012.

The 0012 is at zero degrees angle of attack, with a chord of 1 meter. The rectangular domain is a tet mesh whose inlet is 20 meters upstream and outlet is 40 meters downstream with the top and bottom of the domain at +/- 20 meters from the airfoil. It should be a big enough domain and a fairly simple and classic simulation.

This sounds like a very broad question but how would I get the k-w model to produce good results? My understanding is that k-w has the potential to predict transition so could k-w run at a Re# of say 1.3 million by specifying the inlet velocity of 20 m/s? It doesn't seem to work when I tried it. The velocity contours show a low velocity bubble around the airfoil and then acceleration way downstream.

k-w can also do compressibility apparently. So I tried running a density solver with the energy equation on, and the k-w compressibility option turned on. I used pressure-far-field conditions for the inlet, top and bottom of the domain (M=0.8) and a pressure outlet with a turbulent intensity of 0.1%. That solution diverged.

Sooooo it should be a fairly simple simulation. I'm not sure what I'm doing wrong here. I was able to get a k-epsilon model to run assuming a turbulent Re# (about 4.8 million; inlet velocity 70 m/s). If I step outside of this regime,I would have to change models because my flow would not be fully turbulent, which is my whole motivation for trying to get a k-w model to run at both low Re# and compressible flows.

One issue might be the turbulent intensity and viscosity ratio. I set 0.1% for the intensity because I read on cfd-online that's an estimate for intensity for external flows around aircraft/cars/etc. could be much less than 1%. I left the viscosity ratio as 10 and it worked out very well for the k-e turbulent case that I ran.

Any ideas of how to get some meaningful results?

Thanks!

 LuckyTran May 1, 2012 14:41

Quote:
 Originally Posted by eishinsnsayshin (Post 358517) This sounds like a very broad question but how would I get the k-w model to produce good results? My understanding is that k-w has the potential to predict transition so could k-w run at a Re# of say 1.3 million by specifying the inlet velocity of 20 m/s? It doesn't seem to work when I tried it. The velocity contours show a low velocity bubble around the airfoil and then acceleration way downstream.
the standard k-w model is not for transition. are you talking about the transitional k-w? it is another whole model in and of itself.

Quote:
 Originally Posted by eishinsnsayshin (Post 358517) k-w can also do compressibility apparently. So I tried running a density solver with the energy equation on, and the k-w compressibility option turned on.
It can, but that doesn't mean it's recommended. If you want best results with any CFD modelling, stick with incompressible pressure based solver unless it's required. whenever the density based solver w/ energy equation is used there is always the chance for the flow to diverge rapidly from having the wrong initialization. Your initialization has to be good, probably with a converged pressure-based solver solution.

 All times are GMT -4. The time now is 15:07.