No abstract

Experimental investigations in the three-dimensional boundary layer of a swept flat plate with the pressure gradient induced from outside are aimed at enhancing knowledge of the transition process in the presence of pure crossflow instability. The development of disturbances is characterized by the occurrence of both stationary and travelling instability modes, by early nonlinear development and by complex dependence upon the environmental conditions. Experiments under natural conditions of transition showed a good correspondence of the identified modes with those predicted by local linear stability theory. The disturbance growth, however, is generally overpredicted. Controlled excitation of crossflow vortices allowing measurements closer to the linear range of amplification confirmed this result. Nonlinear effects such as interaction between stationary disturbances and base flow and between travelling and stationary modes have already been observed when the naturally excited instabilities become of measurable size.The most striking feature of the disturbance development is the complex dependence on initial conditions. Experiments under systematically varied environments showed that surface roughness represents the key parameter responsible for the initiation of stationary crossflow vortices. In contrast to two-dimensional boundary layers, free-stream turbulence influences the transition process indirectly. Only for turbulence levels Tu > 0.2% and smooth surfaces do the travelling instability waves dominate. The location of the final breakdown of laminar flow is clearly determined by the saturation amplitude of crossflow vortices. The receptivity to sound, two-dimensional surface roughness and non-uniformities of the test-section mean flow was found to be very weak.

No abstract

The amplification of disturbances developing under conditions of natural transition in the unstable three-dimensional boundary-layer flow on a swept-back flat plate is measured with the aid of hot-wire anemometry. A detailed analysis of the experimental data allows identification of the most amplified instability modes and determination of their growth rates. The results are compared with linear stability theory. Although the amplification process is affected by nonlinearities starting a short distance downstream of the positions where the disturbances become of measurable size, in some essential respects the applicability of linear theory can be examined. It turns out that the initial amplification rates of the stationary instability modes are fairly well predicted whereas the amplification rates of the nonstationary modes are overestimated. A remarkable feature is that the disturbances with the largest amplitudes are not, in every case, the most amplified ones in theory as well as in experiment.

Stability features are studied experimentally for the unstable three dimensional boundary layer flow on a swept-back flat plate. A pressure gradient on the flat plate is induced by a displacement body. Infinite sweep conditions are approximated by means of contoured endplates. For the measurements, hot-wire and surface hot-film anemometry as well as flow visualization techniques are used. In addition to stationary waves, traveling waves are also traced. The cross-flow Reynolds numbers for the first appearance of either instability mode are of approximately the same magnitude. Wavelength and the direction of stationary vortices, as well as the frequencies of the most amplified traveling waves, are measured for different Reynolds numbers. The data obtained by the measurements are compared with the results of linear stability theory. The location of the final transition on the swept flat plate has proved to be fairly well predicted by the empirical transition criterion of Coustols (Thèse de Docteur Ingénieur, Ecole Nationale Supérieur de l’Aéronautique et de l’Espace, Toulouse, 1983).