CFD Online Discussion Forums

CFD Online Discussion Forums (
-   OpenFOAM (
-   -   Naca 0012 validation, Cd value is too high. (

Peter88 August 11, 2011 06:06

Naca 0012 validation, Cd value is too high.
He everybody,

I am busy doing a Naca 0012 validation. First I started with the zero incidence angle. A Cmesh domain, with a lot of cells near the trailing edge. The Cd result that I get is: Cd of 0.02, which should be 0.008. I use a Re of 5e6. Does somebody got any tips, or have done a naca validation. What kind of turbulence model and wall function?

Best Regards,


akidess August 11, 2011 07:25

Using the search function you can find hundreds of threads on NACA validation, see e.g.

gashishr August 15, 2011 10:54

Which solver are you using?

shubham jain January 14, 2014 09:55

i am validating my Cl and Cd values for NACA 0012. Can anybody provide me with a link or or exact Cd values?? I can find Cl values but i am unable to find Cd values of naca 0012.

Please help

GregorS January 21, 2014 17:10

Cd values calculated with RANS models are usually higher than measured values from wind tunnel. RANS models trat boundary layer as turbulent over whole airfoil surface, whichisnt the case in reality. In reality, boundary layer is laminar in the front portion of airfoils and then becomes turbulent. Because scin frctioninlaminar boundary layer is lower than inturbulent one, drag values measured with experimant are lower than ones from CFD simulation.

To overcome this difference, sometimes a trip wire is attached to the leading edge of airfoil, which makes boundary layer turbulent already at the trailing edge. These results match much better with simulation.

You can get tripped data for NACA 0012 in:
N. Gregory in C. L. O’Reilly, Low-speed Aerodynamic Characteristics
of NACA 0012 Aerofoil Section, Including the Effects of Upper-surface
Roughness Simulating Hoar Frost

But your value of 0.2 is still to high. For Re=3x10^6 should be about 0.11 for tripped airfoil. I got good results with Spalart-Allmaras and k-omega SST model, both in high-Re and low-Re mode.

On the other side, Cd values obtained with LienCubicLowRe and NonLinearKEShih matches pretty good with non-tripped experimental data.

All times are GMT -4. The time now is 18:44.