The autonomous near-wall turbulent cycle

Abstract

The regeneration cycle of near-wall turbulence is investigated by numerical experiments in which different possibly-important effects are selectively removed either from the equations of motions or from the boundary conditions. A candidate is considered important to the cycle if its removal results in substantial damping or decay of the turbulence intensity. Other effects, although probably present, are considered secondary. It is shown in this way that neither the presence of a turbulent core nor the generation of secondary vorticity at the wall are important in the generation of turbulence, while the presence of low-velocity streaks is crucial. Damping the streaks, even if the quasi-streamwise vortices are not directly modified, leads to full relaminarisation. Interrupting the process by which streaks are created by advection of the mean profile by the quasi-streamwise vortices has the same effect. This suggests that drag control strategies could be equally directed to damp the streaks or, as is more common, to damp directly the quasi-streamwise vortices.