Aaron H. Auslender

Hypersonic Boundary-Layer Trip Development for Hyper-X

Boundary-layer trip devices for the Hyper-X forebody have been experimentally examined in several wind tunnels.Fivedifferenttripconegurationswerecomparedinthreehypersonicfacilities:theNASALangleyResearch Center 20-Inch Mach 6 Air and 31-Inch Mach 10 Air tunnels and in the HYPULSE Reeected Shock Tunnel at the General Applied Sciences Laboratory. Heat-transfer distributions, utilizing the phosphor thermography and thin-elm techniques, shock system details, and surface streamline patterns were measured on a 0.333-scale model of the Hyper-X forebody. Parametric variations include angles of attack of 0, 2, and 4 deg; Reynolds numbers based on model length of 1.2 ££ 10 6‐15.4 £ 10 6 ; and inlet cowl door simulated in both open and closed positions. Comparisons of boundary-layer transition as a result of discrete roughness elements have led to the selection of a trip coneguration for the Hyper-X Mach 7 eight vehicle.

The Dynamics of Shock Dispersion and Interactions in Supersonic Freestreams with Counterflowing Jets

An active flow control concept using counterflowing jets to significantly modify the external flowfields and strongly weaken or disperse the shock-waves of supersonic and hypersonic vehicles to reduce the aerothermal loads and wave drag was investigated. Experiments were conducted in a trisonic blow-down wind-tunnel, complemented by pre-test computational fluid dynamics (CFD) analysis of a 2.6% scale model of Apollo capsule, with and without counterflowing jets, in Mach 3.48 and 4.0 freestreams, to assess the potential aerothermal and aerodynamic benefits of the concept. The model was instrumented with heat flux gauges, thermocouples and pressure taps, and employed five counterflowing jet nozzles (three sonic and two supersonic with design Mach numbers of 2.44 and 2.94) and nozzle exit diameters ranging from 0.25 to 0.5 inch. Schlieren data show that at low jet flow rates of 0.05 and 0.1lbm/sec, the interactions result in a long penetration mode (LPM) jet, while the short penetration mode (SPM) jet is observed at flow rates greater than 0.1lbm/sec, consistent with the pre-test CFD predictions. For the LPM, the jet appears to be nearly fully-expanded, resulting in an unsteady flow structure in which the bow shock becomes highly dispersed such that it is no longer discernable. Higher speed camera Schlieren data reveal the shock to be dispersed into striations of compression waves, which suddenly coalesce to a weaker bow shock with a larger standoff distance as the flow rate reached a critical value. The pronounced shock dispersion could significantly impact spacecraft aerodynamic performance (L/D) and heat transfer in atmospheric entry and re-entry, and could also attenuate the entropy layer in hypersonic blunt body flows. For heat transfer, the results show significant reduction in heat flux, even giving negative heat flux for some of the SPM interactions, indicating that the flow wetting the model is cooling, instead of heating the model, which could also significantly impact the requirements and design of thermal protection system. These findings strongly suggest that the application of counterflowing jets as active flow control could have strong impact on supersonic and hypersonic vehicle design and performance.

Numerical study of the performance of swept, curved compression surface scramjet inlets

A computational performance enhancement study was performed employing systematic modifications to a planarsidewall compression scramjet inlet operating at an entrance Mach number of 4 and at a dynamic pressure of 2040 psf. The variations included modifying the planar-side wall compression angle as a function of height, utilizing sidewall curvature, and employing, simultaneously, both forward-swept and reverse-swept compression surfaces. Turbulent flowfield solutions were generated by solving the Reynolds-averaged Navier-Stokes equations to obtain inlet performance parameters such as total-pressure recovery, mass capture, and flowfield pressure distortion (the ratio of maximum static pressure to minimum static pressure generated at the inlet exit plane). Additionally, an inviscid parametric study was performed by employing solutions to the Euler equations to optimize a cubic polynomial that defined the longitudinal sidewall geometry. A final viscous flowfield solution of the optimized inviscid inlet geometry yielded inlet performance improvements; however, inlet top-wall surface boundary-layer shock wave separation interactions persisted. Hence, this numerical study demonstrated that enhanced performance is obtainable via curved-wall geometric modifications to the standard planar-sidewall inlet design, although future work should employ constraints to mitigate detrimental flow separation effects.

Numerical Investigation of Shock-Train Response to Inflow Boundary-Layer Variations

A dataset of normal shock trains in a rectangular cross-section channel has been created from direct numerical simulations in an effort to quantify the impact of inflow confinement ratio on the sho...

Aeroheating Predictions for the X-43 Cowl-Closed Configuration at Mach 7 and 10

A pre-flight analysis was conducted for the Mach 7 and Mach 10 X-43 cowl-closed configurations to evaluate three-dimensional flow-field effects on localized heating in the vicinity of the closed cowl flap and sidewall. First, engineering-level analyses based on documented laminar and turbulent surface heating database were used to estimate localized heating amplification factors to account for turbulent flow effects, corner flow effects and the presence of forebody boundary layer trip devices. These estimated factors were applied to 2-D computational predictions to develop surface heat loads. Second, thin-film gauge surface heating data were obtained for a Mach 10 cowl-closed configuration in a reflected shock tunnel at simulated flight conditions. Third, computational fluid dynamics (CFD) flow-field predictions were obtained at Mach 10 ground test conditions to provide comparisons with the engineering database predictions and experimental measurements. Similarly, a CFD solution was also obtained at Mach 7 flight conditions to provide qualitative comparisons with engineering predictions. The Mach 10 CFD predictions compare favorably with the experimental measurements and the engineering estimates for surface heating amplification on the closed cowl. However, experimental measurements were not of sufficient quantity on the sidewall surface to capture the highest localized heating rates, and also the Mach 10 CFD predictions suggest higher peak heating on the sidewall compared to the engineering estimates. Also, the Mach 7 CFD predictions indicate qualitative agreement with the database estimates in terms of flow-field structure and peak heating locations on the closed-cowl flap and sidewall surfaces.

Feasibility and Performance of Atmospheric-Breathing Propulsion for Mars Descent

Analysis was performed to assess the impact of atmospheric-breathing supersonic retropropulsion as a technology solution for Mars descent. Vehicle models were developed for three architectures, employing an atmospheric-breathing engine for both descent and terminal maneuvers, an atmospheric-breathing engine for descent and rocket engine for the terminal maneuver, and a fully rocket propulsive vehicle. Investigations into design constraints showed the inlet area to dictate convergence for the all atmospheric-breathing architecture. These vehicles were limited by oxidizer ingestion for the terminal maneuver and the reduced propulsive descent timeline. Optimal configurations preferred lower-thrust, lower-propellant-usage designs, which were better able to mitigate the mass penalty of the low thrust-to-weight engine. The terminal rocket architectures were instead limited by rocket thrust, which compensated for marginal deceleration during descent. Optimal configurations tended toward large atmospheric-breathi...

Numerical simulation of shock trains in a 3D channel

A dataset of normal shock trains in a rectangular cross-section channel has been created from Direct Numerical Simulations (DNS) in an effort to quantify the impact of inflow confinement ratio on the shock train structure. To this end, only the inlet boundary layer momentum thickness was varied while the bulk inflow and outflow conditions remained constant. The fully-resolved 3D turbulent boundary layer inflows correspond to atmospheric air isentropically expanded to Mach 2 and were obtained from auxiliary DNS. The simulations show that a change of inflow confinement ratio has a nonlinear impact on the shock train location, with a reduction in boundary layer momentum thickness leading to a displacement of the shock train downstream inside the isolator. As expected, an increase in boundary layer momentum thickness results in a reduction of the normal-like portion of the lambda-shock structures in the tunnel core. This leads to more numerous but weaker bifurcating shocks as well as an increase of the shock train length. It is also found that the growth rate of the boundary layer past the first bifurcating shock is dependent on both the inflow momentum thickness and the relative speed of the shock train compared to the bulk flow. When the inflow boundary layer thickness is varied temporally, the complex shock train response depends strongly on the excitation frequency. Its location along the tunnel is as expected more sensitive to lower frequencies while the shock train length exhibits a band-pass filter behavior.

A wall function technique for accelerating PNS computations of compressible flows

AIAA, Fluid Dynamics Conference, 27th, New Orleans, LA, June 17-20, 1996 A description is given of the implementation of a wall function methodology in an existing PNS code. The wall function analysis used is a derivative of the defect stream function approach of Clauser (1956) and consists of an analytic description of the inner region based on a combined law of the wall and law of the wake. The outer region of the boundary layer is solved numerically using the algebraic turbulence model of Baldwin and Lomax. The wall shear stress predicted by the wall function method agrees well with that from a corresponding fully gridded calculation, while yielding considerable savings in computational cost. (Author)