Transonic aircraft problem
 

Hi,




I have a transonic aircraft problem for which i want to run a RANS steady state simulation.
The boundary conditions I got are :
Mach 0.95
Static pressure 5520Pa
Static Temperature 216.65K
I have meshed the aircraft in star-ccm+ and have converted the mesh to openfoam format. ( the geometry is too complex for snappy)


I have several questions :
1. what is the best solver to use for this problem?
2. How should I set up the boundary conditions? Can i make the whole control volume a free-stream?













Thanks

Tom

Hello Tom,


I suggest you use rhosimpleFoam solver, since you wish to do a steady state analysis. But I highly doubt your final force coefficients would be steady at the transonic regime.



If you find that the coeffs are oscillating you might want to switch to sonicFoam solver which provides transient solution.


And by the way how did you convert your star ccm mesh to openfoam mesh? Did you use a CGNS method?


Thank you

Vicky

Hi Vicky,


i used ccmToFoam to convert the mesh https://www.openfoam.com/documentati...C.html#details .


I have tried using rhoSimpleFoam but the solver seems to crash after a few hundred iterations...giving me some error related to sutherland model.




The aircraft is failry large 11m in length , and from what i read in the literature you need to have 5-10 aircraft lenghts in all directions for your control volume.
The problem now is that my coarsest mesh ends up being 20m cells and my finest 100m+. I have looked into running sonicFoam and rhoCentralFoam but due to the mesh size i fear this is not an option as it would take months to run.

Is your aircraft symmetrical? If your aircraft is symmetric, you can half the number of cells.



My suggestion is to take a simple aircraft model and try to mesh it in starccm and import to openfoam first. Try the DLR/F6 model or any DPW model available online in AIAA website. Make sure the openfoam case setting works using that simple model. Then you can go for a complicated model or a full sized model.

Yes, symmetric. i'm only simulating half body.
That's a good idea, i might try that.

Great! Let me know if you make it work.

Hi,

- High-lift, and drag prediction workshops provide committee and participant grids. I wondered if you have considered to use them?
- I also suggest you to use Common Research Model (CRM) instead of DLR-F6 geometry, which is older than the previous one.
- There should also be a participant to Workshop-6, AFAIK, from Oxford Uni, using OpenFOAM - I think his surname was Ashton or something. He also published his results in a journal paper.

Hi,
Have you solve this problem yet? I'm having a similar problem when running a DLR-F6 at M0.78.
looking forward to your reply