[sadjad.s]

Hi there,
I'm modelling 2d axisymmetric turbulence flow in a channel.
Solution is converged when I use kOmegaSSTF model but no matter what i do, I can't get converge with kOmegaSST or kEpsilon RNG models.
I tried changing the solver to pisoFoam or pimpleFoam or decreasing under relaxations, but sill no answer.
I would appreciate any help.
Thanks in advance.

[kebsiali]

case, mesh, boundary conditions, and turbulence parameters?
when, how does divergence start?

[roadRunner]

Originally Posted by sadjad.s

Hi there,
I'm modelling 2d axisymmetric turbulence flow in a channel.
Solution is converged when I use kOmegaSSTF model but no matter what i do, I can't get converge with kOmegaSST or kEpsilon RNG models.
I tried changing the solver to pisoFoam or pimpleFoam or decreasing under relaxations, but sill no answer.
I would appreciate any help.
Thanks in advance.

Wait, you said axisymmetric and channel? Am I wrong or this is a conflict? It may be relevant.

Did you setup the BC correctly? Are you using wall functions? What is your y+?
What did you use for initial conditions?

Is your simulation steady?

I find it very strange that kOmegaSST doesn't converge in a simply channel/pipe flow.

Hope it helps

[sadjad.s]

Thanks for reply and sorry for delay.
Case seems well because i could get convergence for 3d and 2d kOmegaSSTF.
BTW this is my boundary conditions: (i can share 0-folder files if it's necessary)
U:
profile1DfixedValue for inlet, (this boundary is used for cylindrical cases)
profile1DfixedValue for outlet,
fixedValue for walls
zeroGradinet for outlet
wedge for frontAndBack (symmetry faces)
also for k, I had experimental data.
with kOmegaSSTF this is residuals:

and for kOmegaSST:

Flow is in axial and radial directions in inlet so i think axsymmetric assumption is right...isn't it?
Also for boundary condition of U and k, i had put a probe in inlet and checked them with experimental data which exactly fit.
I haven't checked y+ yet, but due to good results of U profile of kOmegaSSTF i think that is not a matter...BTW i will check it.
Simulation is steady.