Current proposed shapes for single-stage-to-orbit vehicles like the Lockheed-Martin X-33 and VentureStar reusable launch vehicle have extremely large base areas when compared with previous hypersonic-vehicle designs. As a result, base drag - especially in the transonic flight regime - is expected be very large. The unique configuration of the X-33, with its very large base area and relatively low forebody drag, offers the potential for a very high payoff in overall performance if the base drag can be reduced significantly. This article presents results of a base-drag-reduction experiment that was performed in the X-33 Linear Aerospike SR-71 (LASRE) flight program.

Figure 1. For the LASRE Experiment, a half-span model of an X-33 forebody with a single aerospike rocket engine at the rear was mounted on top of an SR-71 airplane.

The experiment was a flight test of a roughly 20-percent half-span model of an X-33 forebody with a single aerospike rocket engine at the rear. As shown in Figure 1, the test model was mounted on top of an SR-71 airplane. It was intended that the LASRE flight-test data would be used to define the aerospike-engine performance under realistic flight conditions and to determine interactions of the engine plume with the base and engine cowl areas.

In order to measure performance of the linear aerospike engine under a variety of flight conditions, the model was mounted on the SR-71 with a pylon that was instrumented with 8 load cells oriented to measure total forces and moments in six degrees of freedom. The model was also instrumented with surface pressure ports on the forebody, boat tail, base, engine ramps, and lower engine fence. By numerically integrating the surface pressure distributions obtained from measurements at these surface pressure ports, it was possible to calculate the model profile drag.

Baseline drag measurements on the LASRE configuration demonstrated a large transonic-drag rise that is significantly larger than the wind-tunnel value predicted for the X-33. It is likely that the observed transonic-drag difference is an effect of the sting mount used to support the X-33 wind-tunnel model. With increasing mach number in the subsonic flight regime, base drag (referenced to the LASRE base area) was found to be relatively constant at a base-drag coefficient of approximately 0.38 until the divergence-drag-rise mach number of approximately 0.90 is reached.

It was found that after the divergence mach number is reached, compressibility effects dominate and the base-drag coefficient rises rapidly. Above mach 1, the base-drag coefficient decreases steadily with increasing mach number. In the subsonic flight regime, base drag constitutes approximately 125 percent of the overall model drag. (In the subsonic flight regime, there was considerable suction present on the model forebody. The forebody suction induced a negative forebody pressure drag coefficient of approximately -0.075. The forebody suction results in a drag coefficient for the entire body of approximately 0.30. Thus, in the subsonic flight regime, the base drag was nearly 25 percent larger than the total drag of the vehicle.) Approximately 80 percent of the transonic-drag rise can be attributed to effects of compressibility on base drag. Baseline LASRE drag data clearly support the assertion that base drag dominates the overall drag. If one is to reduce the overall drag of the vehicle, then the base area is clearly the place to start.

Figure 2. Base Drag Was Reduced by roughening the forebody with grit. The measured reduction exceeded the reduction predicted for three different roughnesses of the order of magnitude of the actual roughness. All of the coefficient values plotted here are referenced to the LASRE base area.

In the case of blunt-based objects that feature heavily separated base areas, a clear relationship between base drag and "viscous" forebody drag has been demonstrated. Generally, as the forebody drag on such an object is increased, base drag tends to decrease. This reduction of base drag is a result of boundary-layer effects at the base. The shear layer generated by rubbing of the free-stream flow against the dead, separated air in the base region acts as a jet pump and serves to reduce the pressure coefficient in the base areas. The surface boundary layer acts as an "insulator" between the external flow and the dead air at the base. As the forebody drag is increased, the thickness of boundary layer at the aft end of the forebody increases, with a consequent reduction in the effectiveness of the pumping and a reduction in the base drag. For subsonic flight conditions, it has been demonstrated that for objects with base drag coefficients greater than 0.30, the forebody/base drag relationship is extremely sensitive. For these flow conditions, a small increment in the forebody friction drag will result in a relatively large decrease in the base drag of the object. Since the subsonic LASRE base drag coefficient is 0.38, it is expected that the LASRE base drag/forebody drag relationship should exhibit a similar high sensitivity. Conceptually, if the increment in forebody skin drag is optimized with respect to the reduction in base drag, then it may be possible to reduce the overall drag of the configuration.

In the LASRE drag-reduction experiment, researchers sought to increase the forebody skin friction and modify the boundary layer at the back end of the LASRE model. One of the most convenient methods of increasing the forebody skin drag is to add roughness to the surface. Such other methods as the use of vortex generators to energize the boundary layer would probably work more effectively, but the intrusion of vortex generators into the airflow precludes the use of them on hypersonic re-entry vehicles. The benefits of using surface roughness are nonintrusiveness (minimal heating), small weight penalty, mechanical simplicity, and low cost.

For the LASRE drag-reduction experiment, # 24 silicon carbide [0.035 in. (0.9 mm)] grit was glued to the skin by use of spray-on adhesive, and the surface was sealed by use of a high-tensile-strength, heat-resistant, white enamel paint. The resulting surface had an equivalent sand-grain roughness that varied between approximately 0.02 in. (0.5 mm) and 0.05 in. (1.3 mm). In an attempt to avoid inducing additional flow separation at the boat tail or along the forebody, only the flat sides of the LASRE model were gritted. The gritted area covered approximately one-third of the forebody wetted area.

Results of the experiment verified that surface roughness can be effective in reducing base drag. Figure 2 shows the measured reduction in base drag, in comparison with the reductions in base drag predicted for surface roughnesses of 0.02 in. (0.5 mm), 0.05 in. (1.3 mm), and 0.10 in. (2.5 mm). The predicted reductions in base drag ranged from 8 to 14 percent. The base-drag reduction calculated from flight data peaked at 15 percent. The base-drag reduction also persisted well out into the supersonic flight regime. Since base drags of supersonic projectiles had never before been correlated with viscous forebody drags, the sizable reduction in supersonic base drag in this experiment was a significant positive result.

Unfortunately, flight-test results for the rough-surface configuration did not demonstrate an overall net reduction of drag. The surface grit caused a rise in forebody pressures. Coupled with increased forebody skin drag, the forebody pressure rise offset benefits gained by reducing base drag. Clearly the techniques used to apply the surface grit must be refined. In addition, the existence of an optimal coefficient of viscous forebody drag must still be proven.

This work was done by Stephen A. Whitmore and Timothy R. Moes ofDryden Flight Research Center.

This invention is owned by NASA, and a patent application has been filed. Inquiries concerning nonexclusive or exclusive license for its commercial development should be addressed to the Patent Counsel, Dryden Flight Research Center; (805) 258-3720. Refer to DRC-99-01.