[Kostas_K]

Hello everyone,

I want to simulate using Ansys Fluent a rectangular free jet and wall jet coming out of a sharp-edged rectangular orifice configuration with aspect ratio 10 (L=70mm and H=7 mm) at Re_h=23000 (Re_Dh=42000) so that it is fully turbulent.
My goal is to see the 4 vortices that appear at the 4 corners of the orifice configuration that are responsible for the development of a vena contracta in the plane of the major axis of the jet and the formation of saddle-backed velocity profiles in the characteristic decay region (typically, 5<x/h<30).

I have already created the geometry and mesh so my question is what turbulence model would you recommend ? I think k-e will not work good enough because the mesh is very fine at the orifice. So i was thinking using the K-omega SST model.

Thank you in advance.

[LuckyTran]

Originally Posted by Kostas_K

I think k-e will not work good enough because the mesh is very fine at the orifice. So i was thinking using the K-omega SST model.

Realizable k-epsilon with enhanced wall treatment is suitable for fine meshes. Personally I prefer the near-wall behavior of the EWT (i.e. the Wolfshtein model) more than SST, but I think k-omega SST is a good place to start. It's a flexible model. You also a free jet, so my preference for near-wall behavior is not at all justified for this configuration.

Just make sure you have a really fine grid with high quality. The SST model produces beautiful results if you're able to get the near-wall resolution to integrate all the way to the wall AND you have a nice smooth transition, i.e. don't have skewed cells, etc.

Originally Posted by Kostas_K

My goal is to see the 4 vortices that appear at the 4 corners of the orifice configuration that are responsible for the development of a vena contracta in the plane of the major axis of the jet and the formation of saddle-backed velocity profiles in the characteristic decay region (typically, 5<x/h<30).

I'm not familiar with these vortices. Are these streamwise vortices or are these like corner vortices in rectangular duct flows? Keep in mind that Boussinesq models lack the mechanism to produce secondary flows of the second kind (secondary flows driven by turbulence anisotropy).

[Kostas_K]

Thank you for your comment LuckyTran

I didn't reply sooner because i wanted to test what you said above. I tried solving a steady state simulation with k-e model and EWT as you said. The velocities are good but i have a problem. I can't visualize the vorticities values (Ωx, Ωy, Ωz) in the interior but only at the orifice.
So i tried a transient simulation with the delayed DES and Spalart Allmaras turbulence model and the vorticities are shown very nice. However the problem with this one is, the residuals are oscillating between the same values for about 3 days now.
If you are familiar with the DES and can be of any help it would be great.

The vortices you asked are being formed at the four corners and then they follow the flow. So you could call them streamwise.