[jck]

Hello,

i already talked about my problem in another thread called "airfoil pressure coefficient in pisoFoam kOmegaSSt". I have to analyse the separation position of a NACA 4412 airfoil. For that i want to calculate the pressure coeffient. But the cp value at the trailing is to small.
I nearly checked every possible sources of error (mesh, /system-files, simulation with and w/o wallfunctions ...).

Yesterday i found a cp-plot similar to mine in the following book:

"https://books.google.de/books?id=5JaiBgAAQBAJ&pg=PA464&lpg=PA464&dq=druckv erteilung+naca4412&source=bl&ots=1KpjMcSVLh&sig=cW iqjot0_tyevA_tcJt7b5ugH58&hl=de&sa=X&ved=0ahUKEwiH q7fqo4TNAhXFaRQKHSLxDEgQ6AEIIzAB#v=onepage&q=druck verteilung%20naca4412&f=false"

I know the dependency of the Reynolds number on the pressure coefficient but i wanted to check it. Yesterday i made 3 calculation. All of them with a velocity of 27.13 m/s, a viscosity of nu=1.5e-5 m^2/s but with a converttometers of 0.1,1 and 10. According to that 3 different Reynolds numbers of 1.8e+5, 1.8e+6 and 1.8e+7 and so different pressure coefficients.
But in the results the cp-plots were nearly identical.

Is it possible that i'm calculating with a wrong Reynolds number?
Where do i have to change something?

My airfoil geometry is from 0 to 1.

Regards
Claus

[ottodrissen]

I know the dependency of the Reynolds number on the pressure coefficient but i wanted to check it

The Reynolds number does not depend on any pressure coefficient.

The pressure coefficient is not a good parameter to detect flow separation. To detect the separation point you should use Cf.

http://www.cfd-online.com/Wiki/Skin_friction_coefficient

As Cf depends on the Reynolds number you will find different separation points for different Reynolds numbers.

P.S.: If your angle of attack is moderate, your TE not too thick and your flow low subsonic, I'd rather use X-foil for this task.

[jck]

No i mean the other way around.
If you change the Reynolds number there is a big influence on the pressure coefficient.
But this dependency does not exist in my simulation and this must be wrong.

I will try to do both ways of separation detection, cp and Cf.