A numerical study published in Results in Engineering demonstrates that combining passive geometric features with an active secondary air jet significantly modifies hydrogen fuel dispersion and aerodynamic efficiency in a Mach 2 supersonic crossflow. Utilizing three-dimensional compressible Reynolds-Averaged Navier-Stokes simulations, researchers evaluated how ramp-mounted chevron injectors equipped with internal air jets influence vortex formation, shear-layer deformation, and species distribution downstream of the injection plane. The aerodynamic investigation revealed that injector performance does not improve monotonically with geometric complexity or added air injection. Instead, coupling a four-sided chevron nozzle with an internal air jet produced a 35.3 percent increase in peak mixing efficiency alongside the highest total pressure recovery coefficient among the tested configurations, establishing a practical compromise between turbulent fuel dispersion and aerodynamic retention within high-speed propulsion environments.

Supersonic combustion ramjet engines encounter severe operational constraints because incoming air travels through the combustor at hypersonic velocities, leaving an airflow residence time typically on the order of milliseconds. Within this brief interval, gaseous fuel must be introduced, thoroughly blended with supersonic air, and prepared for stable combustion. Traditional wall-flush injectors provide structural simplicity but suffer from shallow fuel jet penetration and restricted cross-stream mixing. Strut injectors introduce bluff-body wakes that penetrate deeper into core flows, yet their presence imposes significant structural complexity and substantial total pressure losses. Cavity-based configurations provide flameholding capability but can generate excessive aerodynamic drag and wall thermal loads. Ramp injectors offer a balanced alternative by generating oblique shock waves that direct fuel into core streams while shedding streamwise vortices that enlarge the contact area between fuel and air, producing effective momentum exchange with moderate total pressure penalties.

To accelerate mixing rates beyond standard ramp capabilities, researchers have integrated serrated chevron geometries and internal active flow control mechanisms. While ramp structures generate primary compression waves and large-scale streamwise vortices that elevate fuel away from the combustor wall, chevron nozzle exits divide the surrounding shear layer into organized counter-rotating vortex pairs that promote peripheral stretching and fine-scale turbulent entrainment. Overlaying an active internal air jet modifies the momentum distribution within the ramp wake, energizing the wall boundary layer and suppressing intense recirculation zones. Prior research has frequently examined ramp geometries, lobed nozzles, or secondary air injection in isolation, but relatively little data has characterized their simultaneous physical coupling. The current study systematically investigated three dissimilar chevron exit configurations featuring three, four, and six lobes, both with and without internal air-jet assistance, to determine whether increasing nozzle complexity necessarily enhances supersonic mixing.

The computational methodology employed a density-based finite-volume solver within ANSYS Fluent to solve the three-dimensional compressible conservation equations for mass, momentum, total energy, and individual chemical species. The multi-species transport model treated the gaseous hydrogen and air as a non-reacting, compressible ideal-gas mixture, allowing local gas density to adjust dynamically with variations in pressure, temperature, and mixture composition. Turbulence closure was achieved using the Shear Stress Transport k-omega formulation, which provides robust predictive accuracy across wall-bounded flows experiencing adverse pressure gradients, shock interactions, and separated shear layers. Second-order spatial discretization schemes were applied across all governing equations to mitigate numerical diffusion. Computational solutions were considered converged when normalized residuals for all governing equations declined below 10 to the power of negative five and monitored integral flow quantities displayed negligible numerical variation with ongoing iterations.

The numerical domain modeled a Mach 2 supersonic inflow with a prescribed static pressure of 100 kilopascals, a total temperature of 340 Kelvin, and flow direction aligned with the freestream axis. Hydrogen fuel and secondary air were discharged at identical total pressure ratios of 0.1, calculated relative to freestream total pressure, with equal surface areas designated for the outer fuel nozzle and inner air orifice. Boundary walls, including the ramp surface, combustor walls, and sidewalls, were modeled as no-slip adiabatic surfaces, while the combustor exit was treated as a supersonic outflow where flow variables were extrapolated from interior cells. Grid-independence assessments evaluated four successive refinement levels for each geometry. For the three-sided chevron, grid resolution ranged from 704,000 to 1,844,000 cells; the four-sided configuration spanned 884,000 to 1,924,000 cells; and the six-sided configuration ranged from 984,000 to 2,220,000 cells. The third mesh level was selected across all cases because wall y-plus values remained below 3.5 while average hydrogen mass fractions 15 millimeters downstream exhibited negligible sensitivity to further refinement.

To validate the baseline numerical framework, simulation results obtained with a circular nozzle of 2 millimeters diameter were compared directly against experimental benchmark measurements reported by Waidmann and colleagues. Velocity comparisons along a symmetry line situated 78 millimeters behind the strut demonstrated close agreement across various heights above the combustor floor. At a vertical height of 10 millimeters, the experimental velocity of 671 meters per second matched the numerical prediction of 659 meters per second with an error of 1.7 percent. At 15 millimeters, the values were 710 meters per second experimentally and 701 meters per second numerically, representing a 1.3 percent error. Additional heights of 20, 25, 30, 35, and 40 millimeters exhibited experimental-versus-numerical values of 790 against 752 meters per second, 590 against 551 meters per second, 780 against 753 meters per second, 620 against 608 meters per second, and 610 against 599 meters per second, respectively. Centerline temperature comparisons 56 millimeters downstream confirmed consistent physical trends, though simulated temperatures remained below experimental values because chemical combustion reactions were intentionally omitted from the numerical model.

Examination of three-dimensional flowfields revealed distinct topological differences across the three chevron configurations. For the three-sided chevron, a strong oblique shock formed behind the ramp followed by an isolated recirculation zone with Mach numbers below unity, creating a narrow hydrogen core that exhibited strong vertical penetration but restricted spanwise dispersion. The four-sided chevron generated a symmetric and expanded flowfield with multiple interacting vortex pairs that energized the shear layer, promoting broader lateral fuel spreading while maintaining cohesive penetration into the Mach 2 freestream. In contrast, the six-sided chevron produced an intricate shock structure accompanied by highly diffused low-Mach regions. The elevated number of chevron lobes divided available jet momentum among numerous fine-scale vortices, which intensified local shear deformation near the nozzle exit but caused rapid core dissipation and diminished vertical penetration farther downstream.

Cross-sectional evaluations conducted at planes 5 millimeters and 20 millimeters downstream from the ramp underscored the structural evolution of the vortex systems. At 5 millimeters, the three-sided injector exhibited two dominant counter-rotating vortices confining a dense hydrogen core. The four-sided injector established a more symmetrical vortex pair that sustained larger high-Mach core dimensions and elevated cross-stream interaction. By 20 millimeters downstream, the four-sided geometry preserved cohesive, well-organized vortical structures, whereas the six-sided injector displayed fragmented, decaying vortices that hindered far-field fuel transport. Quantitative analysis of helicity and turbulent kinetic energy at 10 millimeters downstream corroborated these dynamics: the four-sided nozzle generated the highest localized turbulent kinetic energy along its shear layers, reflecting vigorous vortex stretching and turbulence production. Circulation strength calculations revealed that without air assistance, increasing chevron sides beyond three increased near-field circulation at 5 millimeters by more than twofold. However, when the internal air jet was introduced, peak circulation decreased and decayed more gradually downstream, resulting in only an 18 percent circulation difference between the three-sided and six-sided designs.

Secondary air injection exerted contrasting effects on fuel-air mixing efficiency across the different chevron geometries. Without air assistance, maximum mixing efficiencies evaluated across a 40-millimeter development length reached approximately 0.467 for the three-sided injector, 0.289 for the four-sided design, and 0.492 for the six-sided configuration. Introducing the internal air jet altered these mixing characteristics significantly. The three-sided injector achieved a maximum mixing efficiency of approximately 0.503, representing a 7.7 percent enhancement. The four-sided injector demonstrated the most dramatic relative improvement, climbing from 0.289 to 0.391, which corresponds to an increase of approximately 35.3 percent. Conversely, the six-sided injector experienced a performance decline from 0.492 to 0.326, representing a 33.7 percent reduction. This non-monotonic response indicates that excessive vortex fragmentation in the six-sided geometry, when combined with secondary air momentum, disrupted large-scale fuel-air interpenetration and suppressed downstream mixing growth.

Aerodynamic performance assessments based on combustor total pressure loss and exit pressure recovery highlighted critical engineering trade-offs. Downstream total pressure loss steadily increased across all non-assisted configurations due to shock formations, turbulent shear mixing, and boundary-layer friction, with the six-sided design incurring the greatest aerodynamic penalties. Activating the internal air jet reduced total pressure losses across all geometries by stabilizing the wall boundary layer and smoothing adverse pressure gradients behind the ramp wake. Evaluated at the combustor exit plane 40 millimeters downstream, the coefficient of total pressure recovery under air-assisted conditions was 0.248 for the three-sided chevron, 0.273 for the four-sided chevron, and 0.261 for the six-sided chevron. Although the three-sided configuration achieved the highest absolute mixing efficiency, the four-sided chevron delivered the greatest relative mixing gain while preserving the highest total pressure recovery, proving superior in minimizing total drag penalties.

The study notes several physical and computational boundaries that frame these aerodynamic conclusions. The numerical simulations were restricted to non-reacting cold-flow conditions to isolate fluid dynamic mixing mechanisms from chemical heat release, meaning that ignition delays, combustion pressure rise, and flame-vortex feedback were not simulated. Furthermore, the Reynolds-Averaged Navier-Stokes SST k-omega model cannot capture instantaneous multiscale turbulent eddies with the resolution provided by large-eddy simulation or hybrid methods. Operating conditions were confined to a single freestream Mach number of 2, a fixed total pressure ratio of 0.1, and an abbreviated downstream measurement domain of 40 millimeters that reflects near-field development rather than combustor-scale homogenization. The authors emphasize that subsequent research must evaluate reacting flows, comprehensive flameholder cavity geometries, and broader momentum ratios to verify the stability of these hybrid injection mechanisms under realistic scramjet operating conditions.

Subject of Research: Numerical investigation of hybrid ramp-chevron injectors with internal air-jet assistance for supersonic hydrogen mixing

Article Title: Vortex-enhanced hydrogen mixing in supersonic crossflow using a hybrid ramp–chevron injector with internal air-jet assistance

Article References: Houria, Z. B., Ghodratallah, P., Abdullah, M., Naik, M.-U.-D., Alshammari, S., Sadeq, A. M., Singh, N. S. S., Shaban, M., Rajab, H., & Aich, W. (2026). Vortex-enhanced hydrogen mixing in supersonic crossflow using a hybrid ramp–chevron injector with internal air-jet assistance. Results in Engineering, 32, Article 113845. https://doi.org/10.1016/j.rineng.2026.113845

Image Credits: AI Generated

DOI: 10.1016/j.rineng.2026.113845

Keywords: Scramjet Propulsion, Supersonic Mixing, Computational Fluid Dynamics, Hydrogen Fuel, Chevron Injectors, Vortex Dynamics, Aerospace Engineering, Active Flow Control, Vortex-enhanced, hydrogen, mixing, supersonic