flow across sphere at Re=10,000
 

Good Day,

I am trying to simulate flow over a sphere(using water 998.2 kg/cubic m) at Re=10,000 (V=0.01005 m/s) using Fluent. The following is my simulational setup:

• unsteady LES with dynamic smagorinsky model
• PISO (pressure velocity coupling)
• standard pressure discretization and second order upwind for momentum and energy
• time step size is 19.9005 sec (based upon assumed St=0.2, giving 25 steps in one cycle)

The problem is that my drag convergence history is a bit offset from those calculated by other people. Its a bit high, varying between 0.5~0.4 whereas it should be in the range of 0.42~0.37
I would be greatful if somebody can guide me in this regard.
Best Regards,
Raheel

Second order upwind is not a good choice when performing LES. The Bounded Central Scheme is the suggested one for LES simulations, for momentum and energy.

Also, the time step should be calculated from the Re number. Actually, when performing LES, it is assumed that time filtering is not present so the time step should be chosen following this reasoning (probably a courant number less than 1 is a safe condition). Obviously, the grid spacing should be selected in the same way.

Finally, i'd choose NITA- Fractional Step instead of PISO and the PRESTO! interpolation instead of the Standard one.

turbulent pipe flow
 

hi all,

I want to simulate turbulent in pipe flow. Could you guys give me some references so that i can validate my data. Thanks in advanve.

Thanx Paolo for your suggestions. I applied the suggested modifications and the results are much better now.

Thanx Again...