Diffuser Boundary Conditions
 

Dear all,
I am a mechanical engineering student attempting to obtain a very basic understanding of CFD. In my class we have a project that asks us to analyze an annular diffuser for separation along the wall. My issue is how to set up the boundary conditions specifically at the outlet. Hopefully the following will "paint the picture" well enough for some advice.

Using fluent for the solver
Axisymmetric
Pressure based due to low Mach number
Transient
k-epsilon turbulent model
working fluid of air

My current boundary conditions are as follows

Inlet --> velocity inlet, only thing I changed was the velocity, set to 0.1 m/s
Axis--> Axis, axis which geometry revolves around
Wall--> outer wall diffuser (sloped wall)

Outlet--> Currently set to pressure outlet with a gauge pressure of 0pa,
I have tried an outflow as well, with no change in my solution. It is my understanding that I need to apply a down stream reservoir of sort to allow the diffuser out to expand into a volume. I attempted to do this, with no change in my solution. Could someone please explain as to how to properly apply this outlet reservoir?

No matter which way I seem to set my configurations I cannot get a solution that will converge. Specifically the k and epsilon residuals. All residuals are set to 1E-5, using the second order upwind.

If it helps I am using ansys workbench for modeling and mesh generation. I have refined my

If someone could please just explain the use of an applicable outlet condition for a problem of this nature or in general I would be grateful. I am in no way asking someone to do my homework, just looking for a push in the right direction. I have spent so many hours on this with little to no progress.

Thank you,
CFDWTB

Hi Jack

I'm not familiar with Fluent, I have never used it.
Natural would be to use some sort of total states inlet boundary conditions and static pressure outlet boundary condition. I guess the inlet velocity BC you mention is
some kind of derivative of the total states BC.
As I say I'm not familiar with Fluent so I do not know what means the "pressure outlet with a gauge pressure of 0pa".
nor can comment on your residuals and convergence of your computational process. I would expect though that if you have the flow separation in your diffuser then the convergence is not going to be that low.

I found this pdf file with an annual diffuser problem, maybe it can help you
to set correct boundary conditions.

http://www.google.se/url?sa=t&rct=j&...t_89Mg&cad=rja

Good luck