Differences in RAS models at transition
 

Hello all,

I did some simulations and saw that the outcome is different with different turbulence models.

Background:
In the attached picture you can see a inletcone with 50° which leads into a pipe.
The pipe has a diameter of 5mm and is 300mm long.
I can measure the flowprofile at the outlet and I need an undisturbed axissymmetric profile.
But this is hard to get. Reynoldsnumber needs to be in Range of ~500 - 25000.
I am not that familar with Openfoam so I use this case to get experience.
The simulation here does not give me the answer I need so I am still working on that with a 3D model.
Seems that the real hardware is extremly sensible for any disturbance in the boundary layer.
In that way - I am also happy for any experience...

Simulation set up:
The simulation in the picture is made with an 2D wedge in simpleFoam with Re of 13000.
The upper section shows a the kOmega simulation result.
The lower section shows a the kOmegaSST simulation result.

Outcome:
I think it might be due to the suitability of the different models.
The 3 plots are at the position of 0.05 / 0.1 / 0.2 mm after the inletcone.
I saw that on the kOmegaSST model there is a higher velocity in the middle of the pipe while the flowprofile is in transition to its turbulent profile while the kOmega models shows a smoother transition along the pipe.


Question on the RAS models:
1. Is it possible to say that one model is better than the other for that case?
2. Are the models good in predicting the transition to turbulent profile?

If some of the setup data is needed I can upload...

Thanks in advance for any help & input
Michael