Learning Forum

[Sky0101]

Hello. I want to simulate a turbulent pipe flow by LES Kinetic-Energy Transport, where D= 2mm, L/D=10, Re=5400.

The geom and mesh as follow.

[Amine Ben Hadj Ali]

I would recommend Using the Synthetic Turbulence Generator and you need ti provide RANS Inlet Profiles. Increase the number of vortices for the VM will require some hidden commands which I cannot share here on this open platform.
I recommend however that you rather keep your fingers away from pure LES here and rely on Hybrid approach or at least on WMLES.

[Amine Ben Hadj Ali]

The subgrid scale can be estimated from the relationship with the eddy viscosity as mu_t=C_K*density*subgrid_K^0.5*Delta and you know know from Samgorinsky model that mu_t is density *(cs*Deta)^2*S

[Sky0101]

Hi, DrAmine´╝üThanks for your reply.
What if I decompose the original inlet into several independent inlets, and define the number of vortices for the VM respectively? I don't know if the idea works well.
I tied to use spectral synthesizerearly, but this method returned higher local velocity beyond my expectations. Does Synthetic Turbulence Generator perform better than spectral synthesizer or VM?
For the first question, how to calculate the initial subgrid kinetic energy?
Thanks!

[Amine Ben Hadj Ali]

It will work but I recommend the STG Method. Do not use SS. STG is the recommended approach.
For the first question: I made already a comment.