[MikePo]

Dear Foamers,

I am trying to evaluate the performance of the sonicFoam solver when applied to a transonic aerofoil case (RAE2822). Using a reasonably fine mesh with the k-epsilon turbulence model I have obtained the attached result (compared with experimental data). The free stream Mach number is 0.75, alpha=2.81, Re=6.5e6.

As you can see, the pressure distribution shows good correlation to the experimental data, however the shock is quite far downstream. Is this a problem with the turbulence model I have used? If so, has anyone any advice on a more suitable model to apply to this problem??


Michael

[CedricVH]

Wouldn`t SpalartAllmaras be a better turbulence model for your case?

Edit: As cboss said, your mesh can be to fine for the normal k-e model. The Spalart-Allmaras model is modelled to work everywhere (in the regions where your mesh is coarse and in the regions where your mesh is fine)

[cboss]

Just to be sure. "Fine mesh" lets me think of a low-Re approach. Are you using OF 1.6, or OF 1.5-dev? I am not that familiar with the k-epsilon model. But I think it is not that easy to set up a low-Re calculation with the k-epsilon model in OF 1.6.(?) (In theory damping has to be included in this case. Not sure how big the error without that will be)

Regards,
Christoph

[MikePo]

I am using 1.6 and have taken onboard your comments about a low-Re case using this release.

Following Cedric's suggestion I am running a case using the SA model. I will let you know my results.

[MikePo]

The results using the SA turbulence model are much smoother, however the shock is still captured too far downstream. In addition, the upper surface suction is over estimated, see below.

Any ideas? Perhaps trying a different solver?

[KrisT]

Have you tried rhoCentralFoam?

Check

http://openfoamwiki.net/index.php/TestLucaG

Don't know about simpleFoam, but sonicFoam seems to give the wrong position of the shock.

[MikePo]

The link mentions that rhoCentralFoam has been extended to viscous flows, but only laminar... is this still the case or can I use turbulence models in OF 1.6?

[truong_nm]

Just a question: why are you trying a incompressible solver (simpleFoam) when Mach number is .75 ? Are you using simpleFoam or rhoSimpleFoam ?

[MikePo]

Originally Posted by truong_nm

Just a question: why are you trying a incompressible solver (simpleFoam) when Mach number is .75 ? Are you using simpleFoam or rhoSimpleFoam ?

Apologies, meant sonicFoam!