[nathanricks]

Hello,

I'm trying to simulate the classical case of turbulent channel flow at Re = 180 with Ansys Fluent 15.0, using LES (with dynamic smag SGS model, and an in-house SGS model).

The simulation is running, and seems to give trends as expected, however we are seeing an overprediction in the mean streamwise (x) velocity - ~21m/s compared to ~18m/s in the literature.

Does anyone have any experience with this particular case in Fluent, and can give me some hints/point me towards a tutorial, etc?

Some of the major setup points include:

- Periodicity in the x (streamwise) and z (cross-stream), with walls in the +/- y direction

The periodicity was defined in fluent, using modify zone. The pressure gradient is defined as -1 (from dP/dx = density*(u_fric)^2), with flow direction = [1,0,0] - this is one of our concerns, is this correct for Fluent?

- CFL and wall y+ < 0.5
- Grid and domain defined consistent with literature
- Flow initialised using k-epsilon RANS, and then init-instantaneous-vel

If any more details would be helpful I'd be happy to give. Any help would be much appreciated!

Thank you in advance,
Nathan.

[songsshan]

Hi, sorry to bother you, but has your problem been solved? I've just recently started dealing with this issue and am not sure how to work about it?Could you give some help?