| Popis: |
Rotating Detonation Engines (RDEs) represent a promising pressure-gain combustion technology for improving the performance of existing rocket engines. While ample theoretical evidence exists for these benefits in ideal scenarios, additional research is needed to characterize the operational behavior of these devices at high pressure and validate the expected performance gains in practice. To this end, Purdue University developed a high-pressure experimental staged-combustion RDE with a supersonic plug expansion nozzle and conducted four test campaigns using this engine. The first two campaigns employed gaseous hydrogen fuel in conjunction with a liquid oxygen pre-burner. The final two campaigns employed methane and natural gas fuels. Propellant mass flows ranged from 0.47 lbm/s (0.21 Kg/s) to 8.41 lbm/s (3.8 kg/s) while mean chamber pressures ranged from 61 psia (4.1 atm) to 381 psia (25.9 atm). Results from tests conducted with hydrogen were mixed. Detonation briefly appeared at shutdown in some configurations, but the combustor behavior was generally dominated by flame holding instead of detonation. Injector erosion and instrumentation damage were also persistent challenges. Results from tests conducted with natural gas and methane were much more successful. Overall, several different types of detonation wave behavior were observed depending on test configuration and operating conditions. In all configurations, the engine thrust, chamber pressure, wave speed, and wave behavior were characterized for differences in injector orifice area, injection location, chamber width, pre-burner operating temperature, equivalence ratio, mass flow, and throat configuration. General aspects of the plume structure, startup behavior, and dynamic oxidizer manifold response were also characterized. Two configurations were also tested with a transparent combustor to characterize wave height and profile. These observations and measurements provided insight into the effects that high-pressures and rocket propellants have on RDE operating behavior. One of the more intriguing results from the experimental campaigns described above was the simple fact that natural gas and methane behaved so differently from hydrogen despite similar operating pressures, flow rates, and injector geometry. Simplified analysis and modeling of the injector dynamic response, mixing processes, and chemical kinetics provided insight into these differences and the scalability of these processes with pressure. In particular, the chemical kinetic analysis suggests that heat release during the injection and mixing phase can dominate the chamber behavior and prevent stable limit cycle detonation from occurring with certain propellant combinations above certain pressures. These results support the observed differences in engine operating behavior, and they provide insight into potential operability limits of gas-phase RDEs. In addition to the contrast between natural gas and hydrogen, several other important observations were made during the experimental RDE evaluation process. In particular, the installation of a convergent throat appeared to suppress detonation behavior. The number of waves was also invariant with respect to the mass flow and chamber pressure, and a natural transition into limit-cycle detonation modes (i.e. self-excited instabilities) appeared despite using a torch igniter with no initial detonation. Significant manifold interaction and an overall destabilizing effect in the limit-cycle detonation cycle tended to occur at low injector pressure ratios. The relationship between pressure, wave speed, and thrust did not follow the expected correlation and instead displayed a more complex configuration-dependent relationship. While the delivered thrust did not exceed theoretical values for a constant pressure cycle, thrust performance greater than 90% was achieved in configurations with simple injector geometries, simple expansion nozzle geometries and a chamber L* of only 2.75 inches. This suggests that further improvements are possible when heat loss into the wall is considered and improved injector designs are implemented. While heat flux was not measured during any experimental test cases, post-test analysis of the chamber environment using available data suggests that heat flux may be moderately higher in RDEs than in constant pressure combustors operating at the same mean flow conditions. Nevertheless, the computed heat flux was based on limited data and may have been affected by localized conditions near the injector face, so uncertainty remains in this area. Since appreciable uncertainty exists in the theoretical performance benefits relative to the measured experimental values, a detonation engine performance model was developed using modifications to existing zero-dimensional rocket performance relations. This approach made it possible to rapidly characterize the effects of different engine operating parameters on expected performance gains including propellant choice, equivalence ratio, initial propellant temperature, chamber pressure, nozzle configuration, nozzle expansion area, and ambient pressure. While the model was relatively simple, it captured the expected “DC shift” in mean chamber pressure between constant pressure combustors and combustors with steep-fronted non-linear instabilities. |
