Experimental and Computational Investigation of Light-Gas Injectors in a Mach 4.0 Crossflow

Abstract

An experimental and computational study of an aerodynamic ramp (aeroramp) injector was conducted at Virginia Polytechnic Institute and State University. The aeroramp consisted of an array of two rows with two columns of flush-wall holes that induce vorticity and enhance mixing. The holes were spaced four diameters apart in the streamwise direction with two diameters transverse spacing between them. For comparison, a single-hole circular injector with the same area angled downstream at 30 deg was also examined. Test conditions involved sonic injection of helium heated to 313 K to safely simulate hydrogen into a Mach 4 air cross stream with average Reynolds number 5.77 e+7 per meter at a jet to freestream momentum flux ratio of 2.1. Sampling probe measurements were utilized to determine the local helium concentration. Pitot and cone-static-pressure probes and a diffuser thermocouple probe were employed to document the flow. This allowed total pressure losses to be determined. The numerical flow solver used was GASP v. 4.2. The inviscid fluxes were computed in three dimensions using third-order AUSM+ with modified essentially nonoscillatory limiting. The AUSM+ algorithm was chosen because of its good resolution of shock discontinuities and its efficiency. The Wilcox k-ω (1998) turbulence model was used. The main results of this work can be summarized as follows: 1) the mixing efficiency value of this aeroramp design, which was optimized at Mach 2.4 for hydrocarbon fuel, was only slightly higher than that of the single-hole injector at these flow conditions; 2) the mass-averaged total pressure loss parameter showed that the aeroramp and single-hole injectors had the same overall losses; 3) the computational fluid dynamics (CFD) was unable to accurately predict the quantitative mixing data produced by the experiment, however, the qualitative comparisons of the injectors using the CFD predictions agreed with the experiment.