Pressure gradient boundary condition for LES
 

http://www.cfd-online.com/Forums/mai...tau-600-a.html

See above post for details.

I am trying to set-up the LES simulation for the square pipe and having issues in defining boundary conditions. Can any body help me in implementing the pressure gradient across the stream-wise periodic boundaries?

Why do you need to define a pressure gradient on the periodic boundaries? Usually preiodic boundaries require no manipulation as the values just transfer across. The pressure gradient takes care of itself.

I am simulating the infinite long pipe. Therefore to save the computational cost, I have defined the inlet and outlet as translational periodic. For this I need to define the pressure gradient which will derive the flow from inlet to outlet.

PS. This is LES of simple square pipe.

I see.

The simple application of this is just to use a additional interface model and set a pressure change. This is trivial - I assume you can do this. This will simply increase the pressure by the amount you specify.

For LES you want the vortex features going out the outlet to come back through the inlet still coherent. I think the simple approach will allow this, vorticies will remain just with the pressure increased. This is probably not exactly correct but the degree of difficulty in gettign the exact approach working is likely to be pretty high :)