Learning Forum

[M_abati]

I'm working on an Embedded Large Eddy Simulation (ELES). Here's what I've done so far:

• Geometry Setup:

  • Created two box domains:
    • A smaller box for the LES (Large Eddy Simulation) region.
    • A larger box for the RANS (Reynolds-Averaged Navier-Stokes) region.
  • Both domains have been added as "Frozen" bodies.
  • Performed a Boolean subtract operation where the LES domain is subtracted from the RANS domain, keeping the tool body intact.

• Mesh Definition:

  • Defined the interface between the LES and RANS regions.
  • Specified the LES-RANS inlet, LES-RANS outlet, and the sides.

• Solver Settings:

  • Set the LES-RANS inlet as a RANS-LES interface and chose the Spectral Synthesizer as the synthetic turbulence generator.
  • Kept the sides as internal boundaries.
  • Set the LES-RANS outlet with No Perturbations.

My Concern:

I'm uncertain if this setup is correct for an ELES simulation, especially regarding the interface definitions and boundary conditions. Resources on this specific setup are quite limited.

Could anyone clarify whether this is the correct approach? Any guidance or suggestions would be greatly appreciated.

[Federico]

What is the flow that you are simulating? What does your domain look like? 

I don't believe the sides as internal is the correct approach. Are you able to constrain your model in the spanwise direction for periodicity?

[M_abati]

 

 

The red box is the body of influence geometry, I am simulating a bluff body (square cylinder). I have not applied periodicity on the sides/top. Is it not possible to accurately simulate without periodicity?