Origin of Turbulence in Near-Wall Flows

Abstract

In experiments by Klingmann, Boiko, Westin, Kozlov and Alfredsson (1993) it became possible to remove the suction peak near the leading edge and to reduce the pressure gradient region. The experimental theoretical neutral disturbance data are given in Fig. 1. As seen, for the low frequencies, the difference between the theoretical curves appeared to be practically negligible. The experiments showed also that the linear parallel stability theory predicts well the distributions of amplitudes, wave numbers, growth rates and neutral stability of the two-dimensional Tollmien–Schlichting waves. The assumption of the boundary layer parallelity allows to acquire quite precise results for the disturbance growth rates, except for the most upper part of the neutral stability curve, where it is necessary to take into account the effects of flow nonparallelity which are quite small. The classical analysis of the linear stability treats the disturbances as separate modes of the Orr–Sommerfeld and Squire equations with exponential growth rates. However, the approach does not take into account the fact that these equations are not self-adjoint, i.e. the modes are not orthogonal. The phenomenon of the ‘lift-up’ effect (a redistribution of the streamwise momentum by small velocity perturbations in the direction normal to