[kdneroorkar]

I have been trying to run a nozzle test case using simpleFoam with FixedValue pressure at inlet and outlet boundaries and zeroGradient velocity at both boundaries. I am getting a strange acceleration zone at the exit of the nozzle due to a very high pressure cell.

I know that a better boundary condition is an outlet pressure and an inlet velocity condition but I donot have this information for this case. Could someone please tell me if I am applying some inconsistent boundary condition or does simpleFoam have a problem with sharp corners.
I am attaching my case folder below for anyne who wishes to run it

[kdneroorkar]

sorry here is the image
the test case is too big to post on the board so if someone wants to see the case I can email it to them.

[schmidt_d]

Kshitij,

Interesting problem. It might help to list your email address:

Kshitij Neroorkar <kneroork@engin.umass.edu>

-David

[louisgag]

Kshitij Neroorkar: Is this a laminar simulation? Otherwise, what turbulence model are you using?


regards

[kdneroorkar]

it is laminar

[louisgag]

maybe if you try the same problem with a coarser or finer mesh your results will improve

[msrinath80]

I would first run checkMesh and ensure that your mesh passes all checks.

[kdneroorkar]

i ran checkMesh and the mesh did pass all tests.
Also, the mesh is pretty fine with 140,000 cells

[msrinath80]

Is this some kind of axisymmetric case? Can you go to www.rapidshare.com and upload your file there and paste the download link (that it gives you) here. The case might help give a better idea on what could be wrong.

[kdneroorkar]

here is the link

http://rapidshare.com/files/14553772...le.tar.gz.html

[msrinath80]

I just checked your mesh and case setup. I could not identify any obvious errors. However, your log snippet worries me a lot.

Listed below are the snippets for time step continuity errors at the 1st, 10th, 20th and 100th, 500th, 2000th and 4000th iteration. Your cumulative error does not look good, neither does the 'sum local'.

time step continuity errors : sum local = 1.02549, global = -0.0188398, cumulative = -0.0188398

time step continuity errors : sum local = 0.0699843, global = -0.000584316, cumulative = 0.0335601

time step continuity errors : sum local = 0.00333286, global = 0.000201054, cumulative = 0.0354029

time step continuity errors : sum local = 0.000304726, global = -3.63236e-06, cumulative = 0.0340224

time step continuity errors : sum local = 0.000303121, global = 3.93966e-06, cumulative = 0.0318931

time step continuity errors : sum local = 0.000308258, global = -1.41759e-05, cumulative = 0.0297984

time step continuity errors : sum local = 0.00030529, global = 1.66529e-05, cumulative = 0.0286185

I recommend going over the first few posts of this thread[1] to clarify what each error term means.

Things you may want to try:

1. Increase the number of non-orthogonal correctors (at least 1 corrector to start with).

2. Like Hrv recommends in that other[1] post, converge the pressure more tightly (at least in the beginning). Set the tolerance in fvSolution to (say) 1e-08 instead of 1e-06.

I am no expert in CFD, so you may want to recheck your Boundary conditions as well.

References:

[1] http://www.cfd-online.com/OpenFOAM_D...es/1/1671.html

[msrinath80]

Oh, and one more thing. You can always use icoFoam to run your laminar simpleFoam problem in a transient manner. The transient simulation should converge to steady state if the problem is indeed steady. I recommend maintaining a Max. Courant number of at most 0.75 even if you are not interested in resolving any transient behavior.

[kdneroorkar]

Hi Srinath
I have tried to run icoFoam using the same boundary conditions and i got a very similar result. In this case the continuity errors look alright. I am just pasting the last few time step continuity errors below

time step continuity errors : sum local = 3.42216e-13, global = -1.07768e-14, cumulative = 8.9745e-10

time step continuity errors : sum local = 4.29076e-12, global = -8.72646e-15, cumulative = 8.97432e-10

time step continuity errors : sum local = 3.58808e-13, global = -8.47658e-15, cumulative = 8.97417e-10

this was using 2 non-orthogonal correctors.

Thanks a lot for your help
Kshitij

[msrinath80]

How many time steps did you run and what was your Max. Co? Try plotting the time variation of velocity or pressure at a particular point in the domain to see if the variable has indeed converged. See the oodles/pitzDaily tutorial to see how to add probes to your simulation.