Basic experiments on the stability of three-dimensional boundarylayer flows

Abstract

For the investigation of the transition process in 3-D boundary layer basic experiments have been performed. Main objective has been to establish the physical features of the disturbance and experimental data for modelling and validation of primary and secondary stability analyses as well as of numerical simulations. For this reason the models have been designed to simulate idealised flows most accessbile to theorey. The experiments showed that the most amplified instabilities agree very well in wavelength, frequency, and direction of propagation with those predicted by local linear stability theory, but they do not agree in the downstream development. The latter depends strongly on the initial conditions. It is most affected by surface roughness and less affected by free-stream turbulence. As regards sound, no effect could be traced. In contrast to local linear stability theory, under environmental conditions usually existing streamwise vortices are the dominating primary instability. They give rise to nonlinear disturbance development as soon as they become of measurable size. In this case non-stationary modes rather appear as secondary instabilities. Secondary stability theory as well as direct numerical simulation of transition have proved to enable the description of those developments. Non-stationary modes as dominating promary instabilities could only be observed on smooth surface at turbulence levels Tu > 0.25%. The location of the final transition into turbulence depends mainly upon the saturation amplitudes of the streamwise vortices and therewith upon the surface roughness. Free-stream trubulence is reducing the saturation amplitudes and is observed to have a retarding effect on transition. Only at values Tu > 0.3% the transition location moves upstream. It is suggested that in this case by-pass effects come into play.