[MachZero]

I have found on the Fluent Users Guide that when I am preparing to run an LES simulation, it is recommended to run the simulation out in steady state, then to initialize the flow field using the command

solve init init-instantaneous-velocity

I notice that If I display contours of velocity before and after this step, the flow field is very different. I understand that a steady state simulation of a turbulent and unsteady flow is representing more of an average behavior, rather than one that every actually exists, but I feel like I must have missed something in how steady state simulations function and are calculated. My question is, where did this new velocity field come from? Are steady state calculations using this field (and they are just hidden from display), but simply averaging several together? Kind of like a running average?

[blackmask]

I have not done LES before but they seems to be some artificially generated inertial range to me. Those scales are not resolved by RANS so they do not exist for RANS. The artificially added fluctuations simply help to develop the small scales so they do not have to be realistic. Usually their spectra might subject to some constraints but ignore that for your purpose.