Compressible calculation with Code_Saturne

Hello everyone,

I am a new user of Code_Saturne and I am currently working on a project regarding the computation of the flow around an airfoil (NACA23012 here). The mesh was created using Salome and I successfully launched some incompressible cases with Code_Saturne using different meshes and turbulence models (without the GUI interface).
The mesh width is only 1 cell. The boundary conditions are the following : "wall" on the airfoil, "inlet" on all the farfield (which is infow or outflow depending on the direction) and "symmetry" at z=0 and z=1.

I am now trying to activate the compressible module in Code_Saturne and I am currently having trouble to set up the environment. I modified the following user files for my purpose : uscfcl.f90, uscfxi.f90, usini1.f90, usppmo.f90 and usproj.f90.
In particular, I added the "coumax" keyword in usini1.f90 and set it to 0.4 as adviced in the user manual (with idtvar=1 for variable time step).
Most of the parameters remain the same between incompressible and compressible cases (initialisation ...) and I tried to follow the user manual for specific parameters.

I compute the lift coefficient in usproj.f90 at each time step so I can follow its convergence and here is the problem : The calculation doesn't converge toward a fixed value but oscillate. I tried to change as many options as I can but the problem remains. Obviously, pressure and velocity fields as well as pressure coefficients around the airfoil are wrong.

Could somebody give me a lead to investigate the problem ? Or maybe a tutorial or some fortran files suitable for compressible calculations ?

I hope I am specific enough. Do not hesitate to answer for any further information.

The fortran user files and the mesh I used :

ubuntuone.com/p/vx6/
ubuntuone.com/p/vx7/

Thanks in advance for any help.

Regards,

Clément

Post edited by: CROZON, at: 2011/05/27 19:13<br /><br />Post edited by: CROZON, at: 2011/05/27 19:18

Here is the mesh

Regards,

Clément

Hello,

What Mach number flow are you trying to simulate ?

The compressible option is mainly oriented towards slightly compressible flows (it is still an elliptic solver). Some experiments have been run with higher Mach flows, and I believe that some were successful, but those flows were not the original target of the code. Some colleagues could probably tell you more, so you may want to write saturne-support at edf.fr to ask to an account on the main Code_Saturne forum at code-saturne.info.

A few of us Code_Saturne developpers check the CAE linux forum occasionally, but we check the main forum more regularly. This doesn't gurantee you'll always have a useful answer, but it does increase the chances.

Best regards,

Yvan

Hello,

Thank you for your answer. I am trying to simulate subsonic flows (mach number between 0.3 and 0.5) but with sometimes a strong deflection that leads to high velocities so the compressible effects should be taken into account.

I will try to open an account at edf forum as you told me.

Thank you again,

Clément