The effectiveness of an H 2 /O 2 torch igniter, which injects H 2 /O 2 combustion gas, and a plasma torch igniter in promoting ignition of an H 2 /air mixture within a supersonic combustor was investigated analytically and experimentally. First, bulk temperature and bulk radical concentration in an O 2 plasma torch igniter were estimated by asymptotic analysis. Those in the H 2 /O 2 torch igniter were also obtained by equilibrium calculation. Second, gas sampling was conducted to determine the concentration of igniter injectant and the equivalence ratio in the region of ignition, namely, the recirculation region at the base of a rearward-facing step. Then, ignition time of the mixture in the region was analytically estimated by using a two-step H 2 /O 2 reaction mechanism and the results just-mentioned. The estimation showed that the ignition promotion effect of the plasma torch igniter was higher than that of the H 2 /O 2 torch igniter at a given input energy. However, the H 2 /O 2 torch igniter was also expected to sufficiently promote ignition because it was easy to increase the input energy by increasing the mass flow rate. Thus, an ignition experiment with the H 2 /O 2 torch igniter was conducted in Mach 2.5 airflow. Comparison of the results with those of the plasma torch igniters showed that the H 2 /O 2 torch igniter also had a sufficient ignition promotion effect as expected from analytical results.

Supersonic Flow Ignition by Plasma Torch and H2/O2 Torch

Flameholding in supersonic flow depends on the local conditions in the recirculation region and the mass transfer into and out of this region. Large gradients in local gas composition and temperature exist in the recirculation region; hence, stability parameter correlations developed for premixed flames cannot be used to determine the blowout stability limits for nonpremixed flames encountered in practical devices. In the present investigation, mixture samples were extracted at different locations in the recirculation region and shear layer formed behind a rearward-facing step and were analyzed by a mass spectrometer to determine the distribution of species concentration in the region. Both nonreacting flow tests and combustion experiments were performed for a range of fuel-related parameters such as injection location, injection pressure, and fuel type. The difference between the local fuel mole fraction within the recirculation region determined from mass spectrometry and the global fuel mole fraction based on the total moles of air and fuel injected was identified. Planar-laser-induced fluorescence was used in nonreacting cases to provide a two-dimensional image of fuel distribution and complement the mass spectrometry measurements. Differences between local concentrations and estimates based on overall global injected mass flows were large, amounting to an order of magnitude. This implies significant differences in flame stability limits of a nonpremixed flame in supersonic flow compared to premixed flame.

Concentration Distribution in a Supersonic Flow Recirculation Region

Experiments were performed to study the effects of toe-in angle on the exterior ports of a three-hole, lowdownstream-angleinjector array in a supersonic e ow. The three jetswere arranged e ush with the test section e oor so that the interference between the three plumes and the main e ow would instill vortical motions. Three different cone gurations were tested involving three, 30 deg downstream angled, round injection holes in a row with sonic air and helium injection into a Mach 3.0 air cross stream. The two exterior holes of the row injector were set at toe-in angles of 0, 15, and 30 deg. The primary qualitative observations were nanoshadowgraphs and surface oil e ow visualization. A novel vertical plate e ow visualization concept was also tested. Aerothermodynamic probe measurementswereperformedata downstream location with heated airinjection to assessthetotalpressurelosses and mixing potential of the e ow. Results showed improved mixing and core penetration with an increase of jet momentum and toe-in angle. The observed oblique shock angle also increased with jet to freestream momentum ratio and toe-in angle, but the integrated total pressure losses associated with the injector shock structure were minimized with the 15 degree toe-in array injector.

Flowfield near a Mulitiport Injector Array in a Supersonic Flow

Although self-sustaining combustion has often been termed auto ignition in studies on scramjet combustors, it should be interpreted as a flame-holding phenomenon rather than ignition. Criteria for flame holding in H2-fueled scramjet engines have been discussed by approximating the recirculation zone in the combustor with a perfectly stirred reactor. The influence of mass flow rates and reactor volumes on flame holding can be summarized by a critical reaction time in the recirculation zone. It was found that the reaction time was proportional to pressure, with P−13 for H2 reactions, and the recirculation zone should be occupied with reactants having equivalence ratios in the range 0.4

Criteria for flame holding in H2-fueled scramjet engines

Flow field at pre-ignition phase in a supersonic combustor with perpendicular injections behind a backward facing step was simulated in cold flow tests with inert gas injection, to investigate the mechanism of the ignition enhancement observed in the ignition tests with the same combustor. The interaction, namely the merging, of separation at step base with that upstream of the injector caused enlargement of ignition region and the enhancement. The change in ignition ability was estimated based on ignition parameters. Based on the experimentally obtained ignition parameters, the ignition limit for the interacted separation region was predicted and showed reasonable agreement with the experimental results.

Auto ignit-ion in a supersonic combustor with perpendicular injection behind backward-facing step

An experimental study of pre-ignition phase flowfields in a supersonic combustor with backwardfacing steps was conducted. Inert gas was perpendicularly injected to coldairstream from four injector orifices downstream of thestep. Wall pressures andinjectant concentrations of the ignition region were measured and their dependence on dynamic pressure ratio and mass flux ratio of the injectant to mainstream, and on unit Reynolds number of the airflow were investigated. It was found that the wall pressure in the region between the step and the injection orifices was well correlated with dynamic pressure ratio but was little influenced by mass flux ratio and Reynolds number. Injectant concentration was influenced by both the dynamic pressure ratio and the mass flux ratio.

Hiroyuki Shiba

Papers

Auto ignit-ion in a supersonic combustor with perpendicular injection behind backward-facing step

Similarity parameters of pre-ignition flowfields in a supersonic combustor