Turbulence boundary conditions
 

I am trying to simulate an air flow in a pipe, steady state, k-e turbulence model. Let's just say that the mesh is appropriate with the appropriate y+ for the wall functions.

At the inlet, I specified a turbulent intensity of 10% and hydraulic diameter of 0.3556m (14" pipe). Though I know that the T.I. might be a little on the high side, I would like to know if the calculation can sort of self-correct this T.I. to an accurate value down the pipe?

In fact, I think I have an answer for my own question, that is, my results shows that the T.I. drops to about 0.04-0.05 (4-5%) down the pipeline. However, here comes the puzzling portion: the T.I. at the inlet (and a little bit downstream) is 0.56 (56%)? This is far more than what I specified (10%). Why did the turbulence increase (to unphysical values) and then drop later on even further downstream?

I also want to know (but can't seem to find online) what does FLUENT do with the turbulence boundary conditions? Be it turbulent intensity and (hydraulic diameter/length scale/viscosity ratio). How does FLUENT use these values for its calculation?

Edit: If it matters, it's a pressure inlet of 1.8bars to pressure outlet of 0bars over a pipe length of 140m.

Not sure if this is the cause but:
Depending on the method that you use, you are specifying a constant value of TI at the inlet which is non-physical (since TI% increases towards the wall, reaches a maximum, and then decays to 0 at the wall). Garbage in, garbage out.

Also, TI% is a local variable and TI% of 25-50% near walls is not unusual. I guess your concern is that the TI% over a cross-section is too high?

Fluent uses these values to compute the k and epsilon at the inlet. In principle, the correct way to specify inlet turbulence boundary conditions is to specify k and epsilon directly. The details can be found in the Fluent User Guide .3.2.1 Determining Turbulence Parameters.