ALBERT

Description: The use of the External propulsion Accelerator (EPA) for launching models of hypersonic aerodynamic configurations into an instrumented ballistic range is discussed. The aerodynamic model is encased inside an axisymmetric projectile designed to be accelerated to high speed in the EPA. Accelerator lengths required to achieve hypersonic speeds are estimated to vary from 10 meters for Mach 7, 40 meters for Mach 10, 150 meters for Mach 15, and 700 meters for Mach 30, assuming a limit of 50,000 g's acceleration. For a model span of 10 cm to 25 cm, the launch tube diameters are 40 cm and 100 cm, respectively. Using this EPA launcher will enable exact simulation of hypersonic flight in ground facilities where both the gas composition and pressure can be controlled in the ballistic range.

Description: This note reports tests in a shock tunnel in which a fully integrated scamjet configuration produced net thrust. The experiments not only showed that impluse facilities can be used for assessing thrust performance, but also were a demonstration of the application of a new technique to the measurement of thrust on scramjet configurations in shock tunnels. These two developments are of significance because scramjets are expected to operate at speeds well in excess of 2 km/s, and shock tunnels offer a means of generating high Mach number flows at such speeds.

Description: In recent years, active research has been conducted to study the technological feasibility of supersonic laminar flow control on the wing of the High Speed Civil Transport (HSCT). For this study, the F-16XL has been chosen due to its highly swept crank wing planform that closely resembles the HSCT configurations. During flights, it is discovered that the shock wave generated from the aircraft inlet introduces disturbances on the wing where the data acquisition is conducted. The flow field about a supersonic inlet is characterized by a complex three dimensional pattern of shock waves generated by the geometrical configuration of a deflector and a cowl lip. Hence, in this study, experimental method is employed to investigate the effects of the variation of deflector configuration on the flow field, and consequently, the possibility of diverting the incoming shock-disturbances away from the test section. In the present experiments, a model composed of a simple circular tube with a triangular deflector is designed to study the deflector length and the deflector base width variation in the flow field. Experimental results indicate that the lowest external pressure ratio is observed at the junction where the deflector lip and the inlet cowl lip merge. Also, it is noted that the external pressure ratio, the internal pressure ratio, the coefficient of spillage drag, and the shock standoff distance decrease as the deflector length increases. In addition, the Redefined Total Pressure Recovery Ratio (RTPRR) increases with an increase in the deflector length. Results from the study of the effect of the deflector's base width variation on the flow field indicate that the lowest external pressure ratio is observed at the junction between the inlet cowl lip and the deflector lip. As the base width of the deflector increases, the external pressure ratio at 0 rotation increases, whereas the external pressure ratio at 180 rotation decreases. In addition, the internal pressure ratio and the coefficient of spillage drag decrease as the base width of the deflector increases. However, RTPRR and shock standoff distance increase as the base width increases. In conclusion, as deflector dimensions vary, distinctive patterns in the pressure variation around the inlet deflector are observed. With an increase in the deflector length and base width, the magnitude of shock-disturbances are weakened due to a decrease in the external pressure ratio. Also, as the deflector length and base width increase, a smaller bow shock angle is formed. Therefore, the inlet shock wave formation would be significantly altered, and consequently, shock disturbances on the wing test section can be avoided through appropriately designing the deflector.

Description: An investigation was conducted from October 1, 1990 to May 31, 1994 on the development of methodologies to improve the designs (more specifically, the shape) of aerodynamic surfaces of coupling optimization algorithms (OA) with Computational Fluid Dynamics (CFD) algorithms via sensitivity analyses (SA). The study produced several promising methodologies and their proof-of-concept cases, which have been reported in the open literature.

Description: The overarching theme was the domain decomposition, which intended to improve the numerical solution technique for the partial differential equations at hand; in the present study, those that governed either the fluid flow, or the aeroacoustic wave propagation, or the sensitivity analysis for a gradient-based optimization. The role of the domain decomposition extended beyond the original impetus of discretizing geometrical complex regions or writing modular software for distributed-hardware computers. It induced function-space decompositions and operator decompositions that offered the valuable property of near independence of operator evaluation tasks. The objectives have gravitated about the extensions and implementations of either the previously developed or concurrently being developed methodologies: (1) aerodynamic sensitivity analysis with domain decomposition (SADD); (2) computational aeroacoustics of cavities; and (3) dynamic, multibody computational fluid dynamics using unstructured meshes.

Description: The NPARC code, a Reynolds-averaged full Navier-Stokes code, was validated for nozzle afterbody (boatail) flow fields at transonic speeds. The flow fields about three geometries were studied: an axisymmetric nozzle with attached flow; an axisymmetric nozzle with separated flow: and a two-dimensional (rectangular) nozzle with separated flow. Three turbulence models, Baldwin-Lomax, Baldwin-Barth, and Chien k-epsilon, were used to determine the effect of turbulence model selection on the flow field solution. Static pressure distributions on the nozzle surfaces and pitot pressure measurements in the exhaust plume were examined. Results from the NPARC code compared very well with experimental data for all cases. For attached flow fields, the effect of the turbulence models showed no discernable differences. The Baldwin-Barth model yielded better results than either the Chien k-epsilon or the Baldwin-Lomax model for separated flow fields.

Description: The effects of grid resolution and specification of turbulent inflow boundary conditions were examined using the NPARC code with the Baldwin-Lomax and Chien k-e turbulence models. Three benchmark turbulent test cases were calculated: two were wall bounded flows and the third was a compressible mixing layer. The wall bounded flows were essentially insensitive to axial grid density; however, the location of the first point off the wall had a substantial effect on flow solutions. It was determined that the first point off the wall must be in the laminar sublayer (y+ less than or equal to 5) for the entire boundary layer. For the compressible mixing layer cases, the axial grid density affected the capturing of oblique shock waves in the mixing region, but the overall mixing rate was not strongly dependent on grid resolution. In specifying the inflow turbulent boundary conditions, it was very important to match the boundary layer and momentum thicknesses of the two flows entering the mixing region; calculations obtained with smaller or no boundary layers resulted in substantially reduced mixing. The solutions were relatively insensitive to freestream turbulence level.

Description: The transonic performance of a dual-throat, single-expansion-ramp nozzle (SERN) was investigated with a PARC computational fluid dynamics (CFD) code, an external flow Navier-Stokes solver. The nozzle configuration was from a conceptual Mach 5 cruise aircraft powered by four air-breathing turboramjets. Initial test cases used the two-dimensional version of PARC in Euler mode to investigate the effect of geometric variation on transonic performance. Additional cases used the two-dimensional version in viscous mode and the three-dimensional version in both Euler and viscous modes. Results of the analysis indicate low nozzle performance and a highly three-dimensional nozzle flow at transonic conditions. In another comparative study using the PARC code, a single-throat SERN configuration for which experimental data were available at transonic conditions was used to validate the results of the over/under turboramjet nozzle.

Description: An experimental study is conducted to examine the crossplane structure and streamwise decay of vortices shed from airfoil-type vortex generators. The vortex generators are set in a counter-rotating array spanning the full circumference of a straight pipe. The span of the vortex generators above the duct surface, h, is approximately equal to the local turbulent boundary layer thickness, delta. Measurement of three-component mean flow velocity in downstream crossplanes are used to characterize the structure of the shed vortices. Measurements in adjacent crossplanes (closely spaced along the streamwise coordinate) characterize the interaction and decay of the embedded vortices. A model constructed by the superposition of Oseen vortices is compared to the data for one test case.

Description: The NASA Steering Committee for Surface Modeling and Grid Generation (SMAGG) sponsored a workshop on surface modeling, grid generation, and related issues in Computational Fluid Dynamics (CFD) solutions at Lewis Research Center, Cleveland, Ohio, May 9-11, 1995. The workshop provided a forum to identify industry needs, strengths, and weaknesses of the five grid technologies (patched structured, overset structured, Cartesian, unstructured, and hybrid), and to exchange thoughts about where each technology will be in 2 to 5 years. The workshop also provided opportunities for engineers and scientists to present new methods, approaches, and applications in SMAGG for CFD. This Conference Publication (CP) consists of papers on industry overview, NASA overview, five grid technologies, new methods/ approaches/applications, and software systems.