[bgraves]

Hi guys,

I am currently trying to run a validation case of flow over a 2D bump in a channel, using simplefoam with the k-omega SST turbulence model. I am having trouble getting the results to match the existing results by NASA at the following website, http://turbmodels.larc.nasa.gov/bump.html. The analysis looks like it is going to converge and then after 2 to 3 thousand iterations taking most of a day the time step continuity errors blow up .

Attached is a picture of the skin friction coefficient before the blow up, the same result occurs during the entire length of the simulation, so something is obviously really wrong, I just cannot figure out what The grid was generated by NASA so I think it is safe to say that the grid is not the problem. Attached is the zipped file, any help or suggestions would be greatly appreciated. which has the boundary conditions file and all of the system files I have used

Thanks,
-Ben

[santiagomarquezd]

I think you should contact

http://www.cfd-online.com/Forums/members/nisi.html

he was solving this kind of problems with simpleFoam.

Regards

[andreamc]

Hello
I am trying to simulate the same case. My setup is compressible solver (rhoSimpleFoam) in OF3.1 and I have succeed at running the case for the three coarsest grids provided by NASA. However for the last two (refined mesh) my simulation crashed and I can not find the problem.
I have found that the relaxation factors play an important role in the convergence so I have changed to p=0.05 and rho=0.1. This has made my simulation a little slow however I achieve more iterations with this change. I am positive the problem is not in the boundary conditions because I have succeed with the same BC in coarse grids. And I have also changed the solvers and the discretization schemes. But nothing seems to work and I do not know what else can be the problem.
The odd thing is whatever I change on the setup I just get to make 29 iterations. I really have not clue what else can be the problem, and I would really appreciate if you can help me.
Thanks in advance

[piu58]

Dear Andrea,

without knowing the problem in detail:
A finer mesh requires a smaller time step too. In solvers where the time step can be set, make it smaller. Some solvers require to lower the maximal Co number. A larger relaxation factor is a third aid to (in fact) reduce the time step.

[andreamc]

Dear Uwe
Thank you very much for your answer but I'm using compressible steady-state solver (rhoSimpleFoam) so the time step shouldn't be a problem. As many people has suggested me that solution I've tried anyway but the problem persists and I still do not get any solution. My simulations crashes after 29 iterations no matter what I change.
Any other tip would be helpful
regards

[piu58]

Are you sure that you don't have problems with the mesh? Did you try checkMesh?

[andreamc]

Dear Uwe

Yes I am sure about the mesh, checkMesh just failed at high aspect ratio cells around 8% of cells are with high aspect ratio. But I do not think that is the problem.
I am absolutely clueless about this problem I've tried BC, schemes, mesh and the simulation keeps crashing after 29 iterations