Further investigation of the effects of 'aerodynamic ramp' design upon mixing characteristics

Abstract

This paper reports the findings of a mixing study involving a variety of “aerodynamic ramp” fuel injectors. Computational fluid dynamics analysis was employed to study various spanwise injector spacings, injector yaw angless, injector transverse angles, and dynamic pressure ratios. The resulting data were studied to determine the impact of each change on performance. One configuration was also tested on a coarser grid; grid convergence was considered acceptable. Analyses prove that neither spanwise injector displacement nor yaw is necessary for the formation of strong streamwise vortices. In general, the elimination of spanwise injector displacement and yaw reduces losses. Centerward displacement and yaw enhance penetration but reduce mixing; the converse is also true. Increasing the transverse angle of centerline injectors improves both penetration and mixing but strengthens shocks and enhances losses; increasing the dynamic pressure ratio produces a similar effect. INTRODUCTION The design of fuel injectors for supersonic combustion engines is among the most formidable tasks facing high-speed aerodynamicists. Achieving mixing adequate for combustion in an environment with millisecond residence times is only the first of many challenges the designer must meet. Drag, thermal degradation, and even machinability are all unavoidable factors in the design of a successful supersonic combustion engine. Because intrusive injectors tend to increase drag, reduce pressure recovery, promote thermal degradation, and complicate fabrication, many researchers have turned their attention to flush-mounted injectors. Direct comparisons between flush-mounted injectors and leading intrusive designs have demonstrated the viability of flush-mounted injectors for efficient fuel mixing. ’ Furthermore, flush-mounted injectors are * Research Scientist, Member AIAA. ’ Aerospace Engineer, Senior Member AIAA. particularly advantageous in combined-cycle systems and engine-off portions of supersonic flight, because the injectors are essentially invisible to the flow when not fueled. By contrast, intrusive injectors carry a substantial drag penalty, especially in unfueled flight modes (for example, during ejector or rocket modes or unpowered cruise). For these reasons flush-mounted injectors are strong contenders for the fueling systems of any supersonic combustion engine design. The advantages of flush-mounted fuel injectors have been known for many years, and various researchers have investigated a great number of flush-mounted injector designs.’ Of particular interest are those research programs which consider the performance of a group of flush-mounted injectors, arranged in such a manner as to allow the injectors to interact favorably with each other. One such design was originally conceived as an “aerodynamic ramp,” an attempt to mimic the geometry, and therefore the performance, of a swept-ramp injector with fluid from nine low-angled, flush-mounted fuel injectors arranged in three columns of three injectors each.3 (See Fig. 1.) Subsequent analysis indicated that, while the design showed considerable promise as a scheme for fuel injection and mixing, its performance was in fact quite different from that of an actual ramp injector.” Computational fluid dynamics (CFD) analysis of the same design provided a great deal of insight into the details of the jet interactions and into the dependence of injector performance upon these details.5 One outcome of the above-mentioned CFD analysis was a list of suggestions for design modifications and possible improvements. With these suggestions as a starting point, work is underway to improve the aerodynamic ramp. Early efforts in this research program were reported in Refs. 6 and 7, and support two conclusions. First, a design process based on the CFD analysis has indeed led to a number of Copyright