CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   FLUENT (https://www.cfd-online.com/Forums/fluent/)
-   -   Aerodynamic 3D Wing simulation (https://www.cfd-online.com/Forums/fluent/148261-aerodynamic-3d-wing-simulation.html)

aleix.de.toro February 7, 2015 14:23

Aerodynamic 3D Wing simulation
 
3 Attachment(s)
Hi friends!

I am studying a 3D wing (NACA 0012 0.5 meters span 1 meter chord) on ansys fluent, but I am having some problems. The mesh has over 1.6 millions elements, the skewness is 0.89 and ten maximum aspect ratio is 32. You can see the attached mesh. Maybe the skewness is a little bit high, but I think aspect ratio is pretty good. I have include an inflation layer too.

On fluent, I am using a density based solver with Spallart Allarmas. The inlet velocity is 30 m/s at 10º angle of attack (yes I have introduced the correct x and z velocity projections for boundary inlet conditions). The references values are introduced too, the area of the wing (on the lift Z projection 0.5*1=0.5 m^2) as well as the chord 1 meter. On Solution monitors, I have added lift and drag coefficients with its proper projections for 10º angle of attack. Then I have initialized and calculated the solution.

The proper value of lift coefficient for 10 degrees of a NACA0012 is 1, and drag coefficient is 0.012 (http://turbmodels.larc.nasa.gov/naca0012_val_sa.html)

On my lift coefficient attached picture, the result of lift coefficient is 0.17 which has no sense. The drag coefficient is more accurate (0.036), but not perfect.

I could have let the computer rune more time, but the lift coefficient would not have differ more.

Has anyone of you tried to simulate a 3d naca 0012 and obtained proper results? Or what am I doing wrong :SS


Many thanks!!!!:)

ghost82 February 7, 2015 15:18

Hi!
I'm not expert with wings, but as I can see from your pictures your scaled residuals are too high.
I suspect your solution is not converged and this can derive from different reasons..you can try to improve the mesh, can't you generate a full hexa mesh?

aleix.de.toro February 8, 2015 05:43

Do you mean mesh methods and hex-dominant method?

I have tried but it appears some error. :/

ghost82 February 8, 2015 06:23

I don't use ansys meshing, but I mean a mesh with hexa elements in all domain.
Usually it provides better results and better convergence.

aleix.de.toro February 8, 2015 09:41

Yes thank you I have tried, but im not having good results. Thankyou anyway! :)


All times are GMT -4. The time now is 17:48.