Transition on elliptic cones at Mach 8.

摘要

Flow visualizations of boundary layer transition on two sharp-nosed elliptic cones at Mach 8 are presented. The elliptic cone is a relevant three-dimensional flow field since it represents a generic hypersonic lifting vehicle shape. Experiments utilize carbon dioxide enhanced Filtered Rayleigh scattering to produce planar single-shot and motion picture images. CO₂ is injected into the flow upstream of the tunnel stagnation chamber and subsequently condenses into nanoscale clusters during the nozzle expansion process. The clusters sublimate as they enter the hot boundary layer, and Rayleigh images capture the interface that exists between the regions of condensed (freestream) and sublimated (boundary layer) carbon dioxide.; Boundary layers ranging from laminar to late-transitional in character are imaged using streamwise, spanwise, and planform laser sheet orientations. Characteristics of observed instabilities are quantified using pdf profiles, power spectrum analysis, and autocorrelation results derived from single-shot images. A new MHz-rate imaging system is also used to produce motion pictures images and volumetric reconstructions of the boundary layer.; The pressure gradient and associated crossflow from the major axis to the minor axis of the cone causes increased growth and subsequent early transition of the centerline boundary layer. The convection velocity and temporal evolution of structures appearing on both the centerline and off-axis regions is studied using cross correlation procedures. Volumetric image sets of the centerline reveal hairpin structures characteristic of the early stages of subsonic turbulent spot formation.; In the off-axis regions, planform single-shot images reveal a pattern of finger-like crests in the boundary layer. At higher Reynolds numbers the breakdown of these crests involves the formation of a series of chain-like structures. The behavior appears qualitatively similar to visualizations of crossflow vortex breakdown in incompressible flows. Subsequent motion picture visualizations determined that the crest pattern was a traveling disturbance, and moved predominantly in the streamwise direction. This result indicates that the observed pattern does not represent active crossflow instability waves. Additional computational evidence suggests that these structures form due to crossflow instabilities occurring near the apex of the model where the pressure gradients are strongest.