I am using a rectangular channel with classic boundary conditions ( velocity_inlet 20m/s, wall, pressure_outflow and symmetry) and RSM model, to get a turbulent boundary layer. I compute the flow and then I extract the profiles for Ux, Uy, Uz and the 6 Reynolds stress u'², v'², w'², u'v', u'w' and v'w'. I want to compute again the flow with the extracted profiles as boundary conditions. But it seems there is a new transition to turbulence, although the precedent flow was already turbulent. That spoil the result for the 2nd computation.

Thanks for your help.

Same thing I observed on the same model. Had no clue virtually why does it happen.