[Hillie]

Dear Foamers,

I have been working on rocket nozzle simulations, and after some effort got them to work. (once have a comparison with my current commercial code CFX i will post a reference here)

I use coarse mesh to get a initial result, then map that solution to a fine mesh, on which i generate my final result.

What I have noticed is that when I use mapFields to map my coarse mesh to my fine mesh I get oscillations/steps in the mapped solution. And certainly in the subsonic region of the rocket nozzle it takes a while to get rid of those using the compressible solvers.

I was therefore wondering if anybody has an suggestion/experience with a smoother that i can use to smooth out my solution on the fine grid before i start iterating on my fine grid?

[vinz]

Hi,

Which solver do you use for your nozzle simulations?
Are they steady state or transient?

In MapFields you have different methods to make the mapping:
- cellVolumeWeight
- direct
- mapNearest

Did you try all of them? Maybe one is better than the others for your case.

[Hillie]

thanks for your reply.

I know mapfields has different mapping options. The 2 non default ones did work. (there was an error in one, and the other had a non-physical result.)

I am currently working with the transient solvers. Correct me if i am wrong here, but the steady state compressible solvers aren’t set up with hyperbolic equations and therefore cannot handle supersonic flow? (If you do know of a good steady state supersonic solver I am all ears).

What I currently do is run the startup simulation with sonicFoam, and run the dense mesh with rhoCentralFoam

[vinz]

Hi,

You may want to try the dbnsFoam (or dbnsTurbFoam) solvers available in foam-extend.
They are supposed to do a better job for supersonic flows.