[A.D.E]

Hi,

I am currently having difficulties obtaining a reasonable agreement for Cd between experimental and numerical results (Re approximately 1 million).
The numerical analysis is 2d and in the experimental procedure end-plates are used to achieve similar effect. The results are corrected for interference blockade etc!!! Does anyone know any reasons for the big difference obtained in Cd (more than 40%)? Is there anything in the published literature describing similar disagreement??

The values of Cl agree within 5%.

I would be grateful for any comments.

Sincerely,

A.D.E

[naimishharpal]

Which turbulence models are you using for this analysis? Make sure all your residuals are well converged below three magnitude. BTW, do you have enough mesh resolution in boundary layer as well as in wake region?

- NHarpal (www.cfdmax.com)

[A.D.E]

I am using K-w SST with y+ <1. My grid is clustered close to the airfoil and wake with more than 20 points in the boundary layer! My residuals drop 5 orders of magnitude and there is no change thereafter.

[Martin Hegedus]

At what angle of attack is the difference of 40% and was the test tripped or untripped? Is your answer high or low? Is the problem incompressible or compressible? Can you describe your cross section a little? Is it an airfoil? Is it a cylinder? Can you give a reference for the experimental data?

You mentioned interference blockage. Why was the data corrected for that? I associate that with high subsonic/transonic issues. Or, when the wind tunnel has slots, even at low subsonic speeds. I don't recall coming across that for low subsonic flows with solid wind tunnel walls.