Search

Vortex generators within a streamline-traced, external-compression supersonic inlet for Mach 1.66 were investigated to determine their ability to increase total pressure recovery and reduce total pressure distortion. The vortex generators studied were rectangular vanes arranged in counter-rotating and co-rotating arrays. The vane geometric factors of interest included height, length, spacing, angle-of-incidence, and positions upstream and downstream of the inlet terminal shock. The flow through the inlet was simulated numerically through the solution of the steady-state, Reynolds-averaged Navier-Stokes equations on multi-block, structured grids using the Wind-US flow solver. The vanes were simulated using a vortex generator model. The inlet performance was characterized by the inlet total pressure recovery and the radial and circumferential total pressure distortion indices at the engine face. Design of experiments and statistical analysis methods were applied to quantify the effect of the geometric factors of the vanes and search for optimal vane arrays. Co-rotating vane arrays with negative angles-of-incidence positioned on the supersonic diffuser were effective in sweeping low-momentum flow from the top toward the sides of the subsonic diffuser. This distributed the low-momentum flow more evenly about the circumference of the subsonic diffuser and reduced distortion. Co-rotating vane arrays with negative angles-of-incidence or counter-rotating vane arrays positioned downstream of the terminal shock were effective in mixing higher-momentum flow with lower-momentum flow to increase recovery and decrease distortion. A strategy of combining a co-rotating vane array on the supersonic diffuser with a counter-rotating vane array on the subsonic diffuser was effective in increasing recovery and reducing distortion.

Vortex generators within a two-dimensional, external-compression supersonic inlet for Mach 1.6 were investigated to determine their ability to increase total pressure recovery, reduce total pressure distortion, and improve the boundary layer. The vortex generators studied included vanes and ramps. The geometric factors of the vortex generators studied included height, length, spacing, and positions upstream and downstream of the inlet terminal shock. The flow through the inlet was simulated through the computational solution of the steady-state Reynolds-averaged Navier-Stokes equations on multi-block, structured grids. The vortex generators were simulated by either gridding the geometry of the vortex generators or modeling the vortices generated by the vortex generators. The inlet performance was characterized by the inlet total pressure recovery, total pressure distortion, and incompressible shape factor of the boundary-layer at the engine face. The results suggested that downstream vanes reduced the distortion and improved the boundary layer. The height of the vortex generators had the greatest effect of the geometric factors.

Surface pressure fluctuations of Mach 8 turbulent flow past a 2.5- and a 4.25-deg expansion corner maintained a Gaussian distribution but were severely attenuated by the expansion process. The pressure fluctuations did not recover to those of an equilibrium turbulent flow even though the mean pressures reached downstream inviscid values in four to six boundary-layer thicknesses. The fluctuations were convected with a velocity comparable to that on a flat plate, and they maintained their identities longer for the stronger expansion. The damping of pressure fluctuations at hypersonic Mach numbers, even by small corner angles, may be exploited in fatigue design.

Experiments were performed at Mach 8 in which a shock was reflected off a low Reynolds number, turbulent boundary layer past an expansion corner. The shock was generated by 2- and 4-deg sharp wedges, and the corner was either 2.5 or 4.25 deg. The inviscid shock reflection was one boundary layer thickness ahead or behind the corner or at the corner itself. All interactions were unseparated. The dynamic surface pressure distributions were examined together with the case of shock reflection on a flat plate. With shock reflection ahead of the corner, the mean surface pressure downstream was attenuated due to the proximity of the corner. With shock reflection downstream of the corner, the surface pressure distribution showed a reduced upstream influence. The highly swept expansion fan produced a surface pressure which rose gently downstream with no minima, unlike in supersonic flows with the same shock-corner separation distance. In many of the interactions, an anomalous pressure peak was found downstream. The surface pressure through the interaction exhibited unsteadiness.

Previous studies of the high-speed viscous inviscid interaction between a turbulent boundary layer and an expansion at a convex corner have noted that surface pressure decreases toward the downstream inviscid value yielded by a Prandtl-Meyer expansion. A downstream influence on the corner is presently identified which is based on the mean surface pressure distribution; a scaling law is proposed for this distance.

Mean and fluctuating surface pressure data were obtained in a Mach 8, turbulent, cold flow past an expansion corner subjected to shock impingement. The expansion corner of 2.5 or 4.25 deg was located at 0.77 m (30.25 in.) from the leading edge of a shape-edged flat plate while an external shock, generated by either a 2- or 4-deg sharp wedge, impinged at the corner, or at one boundary layer thickness ahead or behind the corner. The mean pressure distribution was strongly influenced by the mutual interaction between the shock and the expansion. For example, the upstream influence decreased when the shock impinged downstream of the corner. Also, the unsteadiness of the interactions was characterized by an intermittent region and a local rms pressure peak near the upstream influence line. The peak rms pressure fluctuations increased with a larger overall interaction strength. Shock impingement downstream of the corner resulted in lower peaks and also in a shorter region of reduced fluctuation levels. These features may be exploited in inlet design by impinging the cowl shock downstream of an expansion corner instead of at the corner. In addition, a limited Pitot pressure survey showed a thinning of the boundary layer downstream of the corner.

Two classes of shock-wave boundary-layer interactions were studied experimentally in a shock tunnel in which a low Reynolds number, turbulent flow at Mach 8 was developed on a cold, flat test surface. The two classes of interactions were: (1) a swept interaction generated by a wedge ('fin') mounted perpendicularly on the flat plate; and (2) a two-dimensional, unseparated interaction induced by a shock impinging near an expansion corner. The swept interaction, with wedge angles of 5-20 degrees, was separated and there was also indication that the strongest interactions prossessed secondary separation zones. The interaction spread out extensively from the inviscid shock location although no indication of quasi-conical symmetry was evident. The surface pressure from the upstream influence to the inviscid shock was relatively low compared to the inviscid downstream value but it rose rapidly past the inviscid shock location. However, the surface pressure did not reach the downstream inviscid value and reasons were proposed for this anomalous behavior compared to strongly separated, supersonic interactions. The second class of interactions involved weak shocks impinging near small expansion corners. As a prelude to studying this interaction, a hypersonic similarity parameter was identified for the pure, expansion corner flow. The expansion corner severely damped out surface pressure fluctuations. When a shock impinged upstream of the corner, no significant changes to the surface pressure were found as compared to the case when the shock impinged on a flat plate. But, when the shock impinged downstream of the corner, a close coupling existed between the two wave systems, unlike the supersonic case. This close coupling modified the upstream influence. Regardless of whether the shock impinged ahead or behind the corner, the downstream region was affected by the close coupling between the shock and the expansion. Not only was the mean pressure distribution modified but the unsteadiness in the surface pressure was reduced compared to the flat-plate case.

Initial experience in operating a small UTA (University of Texas at Arlington) shock tunnel at high Reynolds numbers for perfect gas simulation of hypersonic flow and boundary layer studies is described. Particular attention is given to tradeoffs between various constraints for achieving the test requirements, and methods for obtaining precise control of test conditions and protecting low-range pressure transducers from overpressure damage.

The boundary layer that developed on a one-meter long flat plate at a nominal Mach number of 8 and a Reynolds number of 10.2 million per meter was studied using a shock tunnel. The wall-to-stagnation temperature ratio was 0.35. The turbulent velocity profiles possessed very small wake components. The small wake component is characteristic of low Reynolds number flows and may indicate that a turbulent boundary layer was not fully developed. The magnitude of the surface pressure fluctuations in the turbulent part of the flow was found to be in agreement with semi-empirical predictions. These fluctuations were larger than those obtained at lower supersonic Mach numbers and indicated that pressure fluctuations may not be neglected at hypersonic Mach numbers. The fluctuations also possessed a convection velocity that increased with transducer spacing and followed the trend of previous investigations.