[cheng1988sjtu]

I've just run the channelFoam using dynSmagorinsky model.

I just run the original case in tutorials, and found that the result seems to be more like laminar, which gives a Poiseuille flow.

Maybe that's because I didn't define a perturbation in channelFoam, in k file in 0 folder, I can see that it's uniform 0. Can I just change it to be nonzero?

Also the boundary type is fixed value of 0, can I change it to what's in RAS, i.e. kqRwallfunction?

In the attachment, the dot stand for the classic profile of turbulence. 2.5log(y+) +5.5, the solid line stand for the result without defining perturbation.

Does anyone have a similar experience? How can I set up the cse correctly to reproduce the log law? Thank you!

[cheng1988sjtu]

Forget it, I figured out that perturbU can do that, hopefully, I can post some results after a while.

[aka]

I have the same problem. My channel flow is laminar and I am getting a problem when I run LES. Do you get any idea how to get turbulent flow in your channel simulation?

Thanks

[cheng1988sjtu]

you could use perturbU to generate perturbations based on the laminar solution, this will make your simulation to fully turbulent faster.

Basically, the perturbation is added following a Gauss normal distribution, you can define the magnitude of the perturbation, and you may need a few tries.

you are also trying to implement LES in sediment transport? which model do you want to use?

Originally Posted by aka

I have the same problem. My channel flow is laminar and I am getting a problem when I want to apply LES for sediment transport mainly due to the diffusion part of the scalar sediment equation. Do you get any idea how to get turbulent flow in your channel simulation?

Thanks

[aka]

Thanks. Yes I am trying to implement a sediment transport model.

[aka]

one quick question: perturbU needs 'retau' in transport properties. Is there any thing else I have to put in the 'U' or 'k' files inside '0' folder?

Thanks