An experimental study of three-dimensional (3-D) shock wave turbulent boundary layer interaction has been carried out. Interactions generated by fin models having sharp and hemi-cylindrically blunted leading edges have been studied. Tests have been made using incoming turbulent boundary layer varying in thickness in the ratio of about 4:1. Extensive surface property measurements have been made on the test surface on which the incoming boundary layer developed and on the fin itself. All of these tests were carried out at a nominal freestream Mach number of 3, a freestream unit Reynolds number of about 63 million per meter, and under approximately adiabatic wall conditions. The emphasis in the study reported on in this paper was on two main areas. First, to determine the key geometric and/or flow parameters controlling the overall scaling and characteristics of both blunt and sharp fin-induced interactions. Second, to identify the conditions under which both blunt and sharp fins induced interactions have the same local scale and characteristics. (Author).

Shockwave boundary layer interactions (SBLI) occur on both internal and external surfaces and adversely affect both the structural and propulsive performance of high-speed flight vehicles operating in the trans/super/hypersonic flow regimes. In the absence of a comprehensive understanding of the flow physics associated with SBLI, the most common approach to mitigating the negative ramifications is structural over-design, often resulting in reduced aero-propulsion efficiencies and excessive cost. SBLI have been the subject of numerous experimental and numerical investigations focusing on simplified two-dimensional (2-D) canonical configurations derived from relatively complicated aircraft/turbomachinery components. A few recent studies have focused on addressing the knowledge gaps by examining component geometries that produce three-dimensional (3-D) SBLI and therefore a closer representation of real-world configurations. The current experimental investigation explores the viscous/inviscid interaction of an incoming supersonic turbulent boundary layer and a single, sharp unswept fin generated shockwave. This kind of SBLI is of keen interest to the high-speed aerodynamics community as the separated flow induces a strong crossflow component, giving rise to a highly 3-D flowfield. Although previous studies on 3-D SBLI have provided a substantial knowledge base, there are still a number of consequential questions pertaining to the flowfield topology and dynamical behavior that remain unanswered. First, what is the effect of Reynolds number on SBLI flow features, in particular, the length scales associated with the shock-induced separation region and its interaction with the shock generator (sharp-fin)? Second, what is the extent of facility dependence on the 3-D SBLI? Which, if any, component(s) of the unsteadiness is inherent to the interaction and which are facility dependent and therefore limit or bias the flowfield? Are the geometric and boundary layer constraints imposed by the size of the facility necessary for numerical simulations to ensure the proper development of scaling parameters as experiments shift from the laboratory scale to flight testing. Finally, how do the spatio-temporal scales associated with SBLI vary with the interaction strength? The main objective of the present experimental study is to answer the posed questions by conducting a detailed flowfield analysis of the sharp fin induced SBLI over a range of Reynolds numbers and interaction strengths. The research methodology involves high-fidelity experiments at the state-of-the-art wind tunnel facilities housed at the Florida Center for Advanced Aero-Propulsion at Florida State University and the data available from previously published literature. Cutting-edge global flowfield diagnostics allow for the full-field reconstruction of both skin friction (mean) and pressure (time-averaged/unsteady) underneath the single fin SBLI as the incoming Mach number (M[infinity] = 2 - 4), fin angle of attack ([alpha]F = 10° - 20°), and unit Reynolds number (Re/m 17 x 106 - 108 x 106) are parametrically varied. Reynolds number sweeps, spanning nearly an order of magnitude, illustrate that the interaction footprint is distinctly affected by the Reynolds number, with the effects being most prominent near the fin/surface junction and the outer edges of the interaction near the freestream boundary. The results indicate that the interaction flowfield becomes less receptive to Reynolds number variations as the Reynolds number continues to increase. This shrinking dependence indicates that there may be a point beyond which any further increases to the Reynolds number produce negligible differences in the flowfield id est Reynolds number independence. Identical surface oil flow and pressure measurements carried out in facilities of different scale/size compare favorably throughout the interaction region with Reynolds number based scaling. However, different incoming boundary layer thicknesses impose limitations on the extent of the inception region and the onset of finite fin effects. When investigating the mean skin friction between different scale facilities, the Reynolds number scaling could not be assessed due to limitations of the available data sets. An angular scaling was applied to enable proper inter-facility comparison between the conical regions of both identically matching and nominally equivalent interaction strength test cases. The results showed trends similar to those seen in the pressure measurements, with skin friction matching well between the facilities across the interaction with minor divergences in the near fin region, where viscous effects become more prominent. Simultaneously sampled high-speed pressure transducers and fast response PSP measurements allowed for a full-field investigation of the flow dynamics. The RMS pressure field highlights regions of increased unsteadiness along the interaction boundary, inviscid shock line and at/upstream of the fin tip vertex. Increased coherence levels indicate a communication mechanism is present between the inviscid shock and the interaction boundary. When compared with studies conducted in a smaller facility, findings of the current work are consistent in both the locations of increased unsteadiness and their respective magnitudes. In addition to illustrating the robustness of these dynamical features between differing size facilities, the current work identifies the presence of elevated levels of low-frequency content. The presence of this low-frequency content has been observed in investigations associated with 2-D SBLI, but has been absent in the 3-D SBLI studies conducted in smaller facilities. The present study has contributed significantly to a better understanding of swept 3-D SBLI, in particular, the role of Reynolds number and the size of facility on the interaction characteristics. The flowfield analysis has discovered the underlying physics associated with the fin induced SBLI. The high-fidelity experimental database generated will be very useful for the validation of numerical tools and the development of flight vehicle design guidelines.

For this thesis an experimental investigation of sharp fin-induced shock/boundary layer interactions was carried out at a Mach number of 2.95, unit Reynolds numbers ranging from 1 to 4 million per inch, boundary layer thicknesses of 0.14 and 0.50 inches, and fin angles of attack between 12 and 22 degrees. Surface pressure and surface flow visualization data were collected. Results showed that high shock strength interactions were qualitatively similar to those at low shock strengths. When compared to two-dimensional test results, the present three-dimensional data were seen to have a similar dependence on Reynolds number but a different sort of dependence on shock strength. The data were also seen to possess conical similarity. As was the case at lower shock strengths, the interactions could be scaled using unit Reynolds number, boundary layer thickness, and normal Mach number. The appearance of the feature termed secondary separation was noted to depend on boundary layer thickness. Competing feature and turbulence scales were hypothesized to produce this dependence.

An experimental study of three-dimensional (3-D) shock wave turbulent boundary layer interaction has been carried out. Interactions generated by fin models having sharp and hemi-cylindrically blunted leading edges have been studied. The emphasis in this particular study was twofold. First, the influence of incoming turbulent boundary layer thickness delta on the streamwise, spanwise and vertical scaling of the interaction was examined. Turbulent boundary layers varying in thickness from .127 cm (.05 in.) to 2.27 cm (0.89 in.) were used. In addition, a study has been conducted to examine the effects of the ratio D/delta (where D is the blunt fin leading edge diameter) on the interaction properties and scaling. Second, an investigation has been started to examine the unsteady shock wave-boundary layer structure and the resulting high frequency, large amplitude pressure fluctuations which occur ahead of and around the blunt fin leading edge. This is an area which in the past has been largely ignored, yet has important implications, since it is not clear that any mean surface property or flowfield measurements have any real physical significant. To date, measurement techniques and computer software have been developed and exploratory measurements made in the undisturbed turbulent boundary layer and also on the plane of symmetry ahead of the blunt fin.