Novice question
 

Hi All,

I am trying to get an LES case to converge, and I am having some difficulty. The case runs fine in FLUENT, but the pressure residual blows up in OpenFOAM. My mesh is great, timestep is really small (Co <0.05), so I believe my BCs are the problem. In FLUENT the outlet is set to outlet-vent. Can someone tell me how to set that in OpenFOAM. I jus want to double check what I've done.

Currently for U I have:

Outlet
type inletOutlet;
inletValue uniform (0 0 0);
value uniform (0 0 0);

For P I have:


OUTLET
{
type fixedValue;
value uniform 0;
}


Cheers

Hello Anonymous,

It seems indeed your boundary conditions may lead to these divergence problems, similar problems where found here:

http://www.cfd-online.com/Forums/ope...buildings.html

Maybe they have found a solution,

regards,
Tom

Thanks
 

Thanks Tom,

I had a look at the forum, and it helped. I had a question...what outlet Boundary condition in OpenFOAM would match an outlet-vent condition in FLUENT. In FLUENT, the outlet-vent condition allows backflow, but both the outflow and backflow have a pressure drop across the vent, proportional to the velocity (either in or out). Like a loss coefficient for a valve.

Cheers

I'd take a look at this:

http://www.foamcfd.org/Nabla/guides/...ml#x32-1640293

Maybe pressureInletVelocity or something similar for U?

I don't think there is a standard boundary condition available that is similar. I did experiment a bit with LES, but I am not an expert. Also I have no experience with Fluent, so I guess I am not the best person to give you more advice.

I only have one more suggestion:

The way you prescribe the Fluent boundary condition sounds like a sponge layer with backflow calculated based on pressure difference, maybe it helps to use

pressureInletOutletVelocity
instead of
inletOutlet with inletVelocity uniform(0 0 0)?

Regards,
Tom

Reply
 

Thanks Tom,

I will give that a try, they both sound similar. Boundary conditions for LES are very difficult, especially if they arent cyclical.

Thanks again,

Cheers