Lance Jacobsen

Sonic injection from diamond-shaped orifices into a supersonic crossflow

The plume from a diamond-shaped, sonic injector orifice was studied experimentally in a Mach 3 crossflow. The structure of the plume, as well as the near injector flowfield, were examined by flow visualization techniques,and penetration height growth and maximum concentration decay were evaluated from aerothermodynamic probing measurements at two downstream stations. For the transverse injection, the jet-to-freestream dynamic pressure ratio was varied from 0.3 to 2.0. At lower dynamic pressure ratios, the plume from the diamond-shaped injector penetrated farther across the main flow compared to that from an equivalent circular injector, whereas an increase in the dynamic pressure ratio resulted in a deterioration of the plumes sharpness, and penetration of the plume became comparable to that from the circular injector. Angled injections were applied to the diamond-shaped orifice to enhance the penetration at a high dynamic pressure ratio of 2.0. Giving sweepback angle to the orifice was as effective as the case with circular injectors in enhancing the penetration. Adding a moderate yaw angle to the sweptback, diamond-shaped orifice resulted in greatly enhanced penetration, unlike the case with the circular injector. In all cases, the decay rate of the maximum concentration was almost insensitive to the orifice shape.

Improved Aerodynamic-Ramp Injector in Supersonic Flow

An experimental study was performed in the Virginia Polytechnic Institute and State University supersonic wind tunnel on a simplified and revised multiport aerodynamic-ramp injector array in a supersonic flow. The new aerodynamic-ramp injector consisted of four flush-walled holes, in contrast to the original nine-hole versions. For comparison, a single, low downstream-angled circular injector hole was examined. Test conditions included sonic air injection into a Mach 2.4 air cross stream with an average Reynolds number of 4.2 x 10 7 /m at jet-to-freestream momentum flux ratios from 1.1 to 3.3. Shadowgraphs and surface oil-flow visualization pictures were taken in the vicinity of the injectors to gain a qualitative assessment of the injector flowfields. Quantitative measurements of the pressure field on the surface near injectors and in a cross-stream plane downstream were conducted using pressure-sensitive paint and pitot/cone-static probes, respectively. The mixing characteristics of the injectors at three downstream stations were quantified using total temperature probes and a combination of heated and unheated injected air profiles to generate a mixing analog to concentration. Results showed that the aerodynamic-ramp mixed faster and had a larger plume area than the single-hole injector, while sustaining somewhat higher pressure losses due to increased blockage and a higher downstream-angled injector arrangement.

Plasma-Assisted Ignition in Scramjets

This study assesses the prospect of main-fuel ignition with plasma-generating devices in a supersonic flow. Progress from this study has established baseline conditions for operation, such as the required operational time of a device to initiate a combustion shock train as predicted by computational fluid dynamics computations. Two plasma torches were investigated: a direct current constricted-arc design and an alternating current unconstricted-arc design based on a modified spark plug. Both plasma torches are realistic in size and operate within the same current and voltage constraints, although differing substantially in orifice geometry. To compare the potential of each concept, the flow physics of each part of the igniter/fuel-injector/combustor system was studied. To understand the constraints involved with the ignition process of a hydrocarbon fuel jet, an experimental effort to study gaseous and liquid hydrocarbons was conducted, involving the testing of ethylene and JP-7 fuels with nitrogen and air plasmas. Results from individual igniter studies have shown plasma igniters to produce hot pockets of highly excited gas with peak temperatures up to 5000 K at only 2 kW total input power. In addition, ethylene and JP-7 flames with a significant level of the hydroxyl radical, as determined by planar laser-induced fluorescence, were also produced in a Mach 2 supersonic flow with a total temperature and pressure of 590 K and 5.4 atm. Information from these experiments is being applied to the generation of constraints and the development of a configuration with perceived high ignition potential in full scramjet combustor testing.

Integration of an Aeroramp Injector/Plasma Igniter for Hydrocarbon Scramjets

A hydrocarbon-fuel injection and ignition/e ame-holding system consisting of an aerodynamic-ramp injector and a DC plasma torch was designed for a scramjet vehicle operating from Mach 4 to Mach 8. It was tested in an unheated Mach 2.4 e ow for initial evaluation. The injector consisted of two rows of two holes, angled downstream, and toed in to create additional vorticity and enhance mixing. The plasma torch was placed downstream of the injector at three different locations. The experiments involved ethylene injection through the aeroramp at jet-to-freestream momentum-e ux ratios from 1.4 to 3.2. Methane and nitrogen were used as the main feedstocks for the plasma torch. The power output of the plasma torch varied from 1500 to 3000 W. Results showed that nitrogen outperformed methane, and increasing the oxygen content at the plasma/fuel-plume interface signie cantly improved the potential for ignition and e ame propagation. The methane and nitrogen feedstocks performed best at the closest and middle downstream torch stations relative to the ethylene-fuel plume, respectively. Because of the low static freestream temperature (131 K), very little heat release was produced under these cold-e ow conditions. Tests in a model scramjet combustor with hot e ow are needed to complete the evaluation of this system. In addition, at all three torch stations, the counter-rotating vortex motion of the fuel-injector plume lifted up the plasma-torch plume. As a result, the downstream temperature-plume cores were 2.5 and 3.5 times higher at injector jet-to-freestream momentum-e ux ratios of 1.5 and 3.0, respectively, compared to the torch alone.

Operational sensitivities of an integrated scramjet ignition/fuel-injection system

Results are presented of experiments conducted in a supersonic wind tunnel on an integrated fuelinjection/ignition system, consisting of an aeroramp injector and a plasma-torch igniter. The main goals of the work were to determine how the lifting effect of the aeroramp affected the plasma jet, to ascertain how the injection of fuel through the aeroramp and the power supplied to the torch ine uenced the distributions of excited species downstream of the device, and to investigate any synergistic effects from the combination. The aeroramp was observed to have a strong lifting effect on the plasma jet, especially for injector momentum-e ux ratios above 1.5. In addition, increases in the torch input power produced an exponential effect on the emission intensity of the excited-state species downstream of the plasma jet, but was not observed to ine uence the jet penetration height. The results demonstrate that the increased penetration of combustion enhancing radicals is largely a function of the e uidic mechanisms generated by the injector and, thus, aids the plasma torch in ine uencing the combustion kinetics farther into the freestream than would normally be possible.

CAVITY-BASED INJECTOR MIXING EXPERIMENTS FOR SUPERSONIC COMBUSTORS WITH IMPLICATIONS ON IGNITER PLACEMENT (POSTPRINT)

recirculating flow in a downstream cavity. The influence of wall effects were investigated with a similar single-hole injector placed 3.5 injector diameters from the sidewall of the rectangular duct. Simulated plasma torch holes were tested up and downstream of the injectors, all upstream of the cavity. Tests took take place in a uniform Mach-2 crossflow, and in a shock train generated by duct backpressure. The tunnel total pressure and total temperature were 1.7 atmospheres and 530 K. Results from the experiments demonstrated the difference in coupling of the simulated plasma torch plumes with the two types of injectors. Without backpressure only small differences were shown between the injector near the sidewall and the one on centerline. In the region investigated, the aero-ramp mixed faster than the 15-degree injector as inferred by NO PLIF maximum intensity ratio and relative size of the jet plumes. Further, the close proximity of the 15-degree injector to the corner was found to decrease the near-field mixing of the injector as compared to the centerline results.

Operational characteristics of a periodic plasma torch

Development of a plasma torch, which is intended as an ignition aide within a supersonic combustor, is studied. The high-voltage discharge and plasma plume generated by the torch module are described in a quiescent environment and in a supersonic crossflow. Voltage-current characteristics of the discharge and optical images of the plasma plume are used to characterize the operation of the torch module. The principal advantages of this torch module are its compact design, durability, and operational flexibility. The torch module can be operated in periodic or pulsed modes, depending on the power supply used. In the periodic mode presented in this paper, the capacitors are charged at the line frequency of 60 Hz resulting in a cyclical discharge at a frequency of 120 Hz. In this mode, peak and average powers reaching 8 and 2.8 kW, respectively, are demonstrated. The energy can be as high as 46 J per cycle, which is mainly limited by the power handling capability of the power supply. The penetration height and the volume of torch plume into a Mach 2.5 supersonic flow, typical for a supersonic combustor startup condition (vis-a-vis the crossflow velocity), are investigated. In addition, ignition of ethylene fuel in a Mach 2 supersonic flow with a total temperature of 590 K and pressure of 5.4 atm is demonstrated.

Experimental and Computational Investigation of a Dynamic Starting Method for Supersonic/Hypersonic Inlets

This paper presents the current results of experime ntal and numerical studies of a dynamic door starting method applied to a nominally two-dimensional scramjet inlet with variable contraction ratio (CR). The inlet was test ed in a Mach 4 wind tunnel with average Reynolds number of 5.77e+7 per meter, and a broad range of door rotation rates were investigated. The use of a non-ejectable, rotating inlet cover was found to improve the starting limit of overcontracted inlets by approxim ately 10%, which yielded an improvement of inlet pressure ratio of 58%. The numerical flow solver used to model each cont raction ratio was GASP v. 4.2 which uses a RANS formulation. A Roe flux-difference splitting scheme with min-mod limiting was chosen to compute the inviscid fluxes due to its strength in resolving discontinuities. The Wilcox k-ω (1998) turbulence model was used for viscous solutions and a dual-time stepping Euler implicit a lgorithm was used for time accurate predictions. Numerical results coincided well with experimental pressure measurements and shock patterns obtained from schlieren photography.

Flight Dynamics of a Hypersonic Vehicle During Inlet Un-start

The aerodynamics and the stability and control characteristics of a hypersonic vehicle that is has an un-started inlet are addressed. The vehicle considered is a generic scramjetpowered ∞ight test vehicle with a 3D, inward-turning inlet and cruciform flns for control. An analysis of aerodynamic prediction methods for the vehicle with inlet started are given and shows that engineering methods are satisfactory for predicting the external aerodynamics when compared to CFD. However, during un-start, the engineering methods do not capture the efiect of the normal shock that stands outside the inlet; therefore, CFD is necessary to assess the vehicle aerodynamics. An analysis of the open-loop ∞ight dynamics of the vehicle was performed for both the started and un-started inlet. A nominal control law to track a commanded ∞ight-path angle is designed using the started inlet aerodynamics. It is shown that the un-started inlet destabilizes the controller, with a rather large time-to-double, but the aircraft angle-of-attack transient settles rather quickly, which should allow for the inlet to be re-started.