[dfonseka]

Hello,

I am keen to know how to simulate periodic boundary conditions in Fleunt for turbulent pipe flow. I understand turbulent flow is not steady and periodic implies steady but is it possible to get RANs or avergaed quantities for turbulent flow for this case. I know for normal boundary conditions the turbulent intensity can be specified at the inlet and outlet but this option is not available in k-epsilon model.

In addition, anyone have pointers on the method to ensure convergence. My mesh is really fine with nearl 200000 elements but it takes ages to run and the residuals don't converge below tolerance.

I am simulating flow in a wavy channel.

[LuckyTran]

Periodic does not imply steady.

The point of Reynolds-averaging is so you get a laminar-like equation after you take the time average. There are no "velocity fluctuations" in RANS. Remember what you solve for in RANS is the mean flow, not the instantaneous flow.

There are not explicit inlet BC's for periodic boundary condition option, because it's periodic. If you have an inlet/outlet then you can specify the turbulence intensity and so on, but you do not have an inlet or an outlet, you have a periodic BC.

200 000 is (frankly) a baby mesh. It shouldn't take ages to run.

[dfonseka]

Originally Posted by LuckyTran

Periodic does not imply steady.

There are not explicit inlet BC's for periodic boundary condition option, because it's periodic. If you have an inlet/outlet then you can specify the turbulence intensity and so on, but you do not have an inlet or an outlet, you have a periodic BC.

200 000 is (frankly) a baby mesh. It shouldn't take ages to run.

Thank you and noted. Just to confirm, so if I just specify k-epsilon and run the simulation, it should be good? Are there any other settings I should look out for?

I have been using SIMPLE thus far for methods.