Compressible model and rough wall

[Nicolas Merigoux]

Dear all,
 
I would like to model an atmospheric dispersion case using the compressible model in Code_Saturne (as I have to model a gas release of about 230 m/s).
I think the wall roughness will be very important in that case but it seems impossible to use a rough wall as a boundary condition with the compressible model. Perhaps I'm wrong, but if not, has anyone a suggestion on how to deal with this problem (I'm not really keen on explicitly modelling the wall roughness using uniform arrays on the whole domain) ?
 
Thanks a lot for your advices.
Regards,
 
Nicolas

[Mickael Hassanaly]

Hello,
You are right, you can't use rough wall with compressible module. However you can try to implement it by your own. You will have to modify mainly cfxtcl.f90 to reach your goal.
 
Regards
 
Mickaël

[Nicolas Merigoux]

Hello,
 
Thanks Mickael for the tip. I had a first look in cfxtcl.f90 and indeed your suggestion seems to be the best way to proceed.
 
But first (before spending some time on the rough wall part), I would like to be able to run the compressible module using a variable gamma (with perfect gas law) because what I have to model is a light gas release (ammonia) in the atmosphere. So I can't use a constant gamma on the whole domain because that's not correct all (I tried but my results are not exactly what is expected and that's normal).
In order to do so, I made a few changes, mainly in uscfth.f90 and cfxtcl.f90 in order to be able to run the model but it doesn't work at all for the moment. I tried to track the issues and made some "dangerous" changes to make it run but the first result I got are very very bad.
So if anyone as already experienced the problem and managed to make that module (with variable gamma) works, it would be very nice to give me some advice or example.
 
Thank in advance.
Regard,
 
Nicolas