Fluent Vortex Method Turbulence/General NRBC Compatibility
 

I'm working on an LES model (pressure-based, coupled solver, WALE) of a rather small premixed methane-air burner with a ~1" diameter intake system. It uses a velocity inlet with a defined velocity profile and vortex method turbulence generation with defined Reynolds stress distributions. The setup works very well and is in excellent agreement with PIV data. The problem arises when I try to apply the general non-reflecting boundary condition (NRBC) to the inlet. When I do, the synthetic turbulence "freezes" and all of the turbulent fluctuations at the inlet maintain their last value.

There are three relaxation terms in the documentation in reference to the general NRBC (sigma, sigma2, relax). The "sigma2" term has a default value of 5 and applies only to density-based solvers and the "relax" term has a default value of 0.5 and affects overall convergence. The "sigma" term has a default of 0.15 and is only explicitly described under the pressure outlet section where it is used to calculate the relaxation factor. This "sigma" is the relaxation factor I expected tune; however, I can't seem to find a value that works. There is also one additional option for "tangential-source?" (yes or no) in the TUI that doesn't show up anywhere in the documentation aside from an unrelated "fan" model. It isn't even included in the list of TUI commands in the documentation under "general-nrbc." I haven't been able to determine exactly what it does. These appear to be all the "levers" I have access to regarding the NRBC.

Aside from impacting the synthetic turbulence, the NRBC does function and incoming waves pass through the boundary as expected. I am simply unable to get the NRBC to function without affecting the synthetic turbulence generation. If anyone here has experience with this I'd appreciate your suggestions. Let me know if you need additional details regarding the model. Thanks in advance!