Lateral Air Injection Angles Reshape Fuel Mixing in Trapped Vortex Combustor Simulations

Heval Serhat Uluk, Sam M. Dakka and Kuldeep Singh simulated the combustor under realistic idle conditions of an aeroengine, 3.66 bar and 375 K, and found that adding a lateral angle of 15 degrees to air injectors already angled 30 degrees transversely gave the best overall fuel-air mixing for all three fuels. The work is the first to introduce ammonia as an alternative fuel candidate for this combustor design, and it maps in detail how a small geometric change redirects fuel trajectories, turbulence and rotation inside the cavity.

Trapped vortex combustors stabilise flames differently from the swirl-stabilisers used in commercial gas turbine engines. Instead of relying on swirling main flow, the design creates a recirculation zone inside a cavity formed between two bluff bodies. Air and fuel are injected directly into this cavity, which lets engineers adjust the equivalence ratio locally, enhances air-fuel mixing through the trapped vortex, and shields the fuel injectors from overheating. The concept traces back to the bluff-body drag studies of Mair and of Little and Whipkey, who found an optimum cavity aspect ratio of 0.6. Later work by Hsu and colleagues confirmed that this ratio gives the lowest total pressure drop for non-reacting flows, and demonstrated low lean-blow-out limits, high combustion efficiency and low nitrogen oxide emissions across a range of operating conditions.

The motivation for revisiting the geometry now comes from the push toward alternative fuels. Emission restrictions and the growing carbon footprint of aviation have driven interest in fuels beyond methane, propane and JP-8. Hydrogen is attractive because of its low activation energy, high diffusivity and wide flammability range, although previous studies noted that high hydrogen concentrations raise temperatures and can cause instabilities. Ammonia offers easier storage in conventional fuel tanks, delivers more energy than hydrogen per unit injected, and can be produced and transported much like propane, yet it had never been tested in a trapped vortex combustor before this study. The authors set out to evaluate how both gases mix inside the cavity compared with propane, and whether angling the air injectors could improve that mixing.

The team adopted an axisymmetric combustor geometry from Hsu’s earlier design, with a 70 mm forebody diameter, a 50.8 mm afterbody, and a cavity whose length can be adjusted by sliding the afterbody along a 9 mm spindle. Eight fuel injectors feed fuel into the cavity, with sixteen air injectors above them and eight below to reinforce vortex creation. In the new configurations, the fuel injectors stayed perpendicular to the afterbody while the air injectors were angled 30 degrees transversely and then additionally tilted laterally by 0, 15, 30 or 45 degrees. The simulations used ANSYS Fluent 2022 R2 with the Reynolds-Averaged Navier-Stokes Realizable k-epsilon turbulence model, the PISO pressure-velocity coupling algorithm, QUICK discretisation and the PRESTO pressure scheme, with a species transport model handling the fuel-air mixing.

Before running the angled cases, the researchers validated their model against the experimental pressure-drop data of Hsu and co-workers, who measured total pressure drop with water manometers in non-reacting tests with the cavity air and fuel injectors turned off. The two-dimensional simulations reproduced the measured total pressure drop percentages for cavity lengths of 21, 42 and 70 mm at main air velocities of 0.33, 0.66 and 1.0 times the reference velocity of 42 m/s. The lowest pressure drop occurred at the 42 mm cavity length for all velocities, all values stayed below one percent, and pressure drop rose with increasing main air velocity, matching the experimental trends well. A three-dimensional grid independence study then settled on a poly-hexahedral mesh of 4.69 million cells, since axial velocity and turbulent kinetic energy results converged at that resolution.

The flow-field results revealed sharp differences between the fuels. Propane penetrated the cavity most rapidly, losing only 42.35 percent of its pace by the time it reached the mid-cavity measurement plane at half a forebody diameter downstream, whereas ammonia lost 83.55 percent of its velocity over the same distance. Ammonia compensated with high radial velocity that pushed fuel toward the bottom of the cavity and the highest tangential velocity overall, suggesting it suits configurations with extra rotation. Hydrogen behaved differently again: while transverse angling twisted its flow slightly in the negative tangential direction, further lateral angling drove a substantial twist in the positive direction, confirming that lateral angles add rotation on the tangential plane for hydrogen too.

Mixing efficiency, calculated from the fuel mass fraction relative to the stoichiometric value across cross-sections of the cavity, told a nuanced story. For propane, mixing efficiency was highest at the end of the cavity with no lateral angle and declined as the lateral angle increased, because less fuel travelled down toward the lower vortex and more exited through the large main vortex without mixing well. Yet the 45-degree laterally angled injectors produced the highest mixing efficiency near the fuel jets, around 0.24 to 0.3, while showing the lowest values downstream of the cavity at roughly 0.08. The 15-degree configuration showed low mixing efficiency around the fuel jets, between 0.21 and 0.28, but comparatively high mixing efficiency at the cavity end with well-distributed fuel.

Ammonia and hydrogen followed related patterns with fuel-specific twists. For ammonia, the 45-degree laterally angled injectors achieved the highest mixing efficiency around the fuel jets at 0.29, while the lowest value of 0.23 was recorded for the zero-degree injectors downstream. Beyond just over half a forebody diameter into the cavity, all angles converged to nearly identical mixing efficiency, with 30 degrees highest and zero degrees lowest. For hydrogen, all lateral angles peaked at the mid-cavity plane where the small vortex traps incoming fuel, the 45-degree case was highest overall, and the lowest value of 0.2 was measured for the zero and 15-degree cases further downstream. Hydrogen’s fuel trajectory carried it out of the cavity soon after entry, which explains why its mixing efficiency was lower than the other fuels, since some fuel left through the large vortex without mixing thoroughly.

Turbulence and rotation data added an important caveat. Increasing the lateral angle raised turbulent kinetic energy overall, with the most significant changes where fuel and air flows meet, and intensified velocity fluctuations tied to the growing tangential velocity. The authors warn that these high-turbulence regions and strong tangential rotation could produce localised high-temperature zones and combustion instabilities once burning occurs, so the 45-degree configuration must be treated with caution despite its strong mixing near the fuel jets. Fuel trajectory analysis also showed that lateral angling pushed propane and ammonia trajectories outward and upward, reducing fuel distribution toward the cavity end, while hydrogen’s trajectory moved inward as the angle increased, which the authors suggest makes the 45-degree arrangement a candidate specifically for pure hydrogen operation.

Weighing all the evidence, the team concluded that 30-degree transverse plus 15-degree lateral air injectors are the best option for propane, ammonia and hydrogen alike under these idle conditions, and they recommend that configuration for good air-fuel mixing. The study has clear limits: it is entirely non-reacting and numerical, the ammonia and hydrogen cases have not yet been tested experimentally under these configurations even though propane was validated against earlier experiments, and angles beyond 45 degrees would push jet impingement close to the fuel injectors and disrupt the fuel trajectory. The authors propose experimental construction of the combustor, species and temperature measurements with commercial sensors, and future large-eddy and reacting simulations using the eddy dissipation model to evaluate combustion efficiency, pattern factor and temperature distribution. Until then, the message for combustor designers is that a modest sideways tilt of the air injectors offers a simple geometric lever for tuning how alternative fuels behave inside a trapped vortex.

Subject of Research: Computational study of how lateral air injector angles affect fuel-air mixing of propane, ammonia and hydrogen in an axisymmetric trapped vortex combustor under realistic idle aeroengine conditions.

Article Title: Flow behaviour of propane, ammonia and hydrogen in axisymmetric trapped vortex combustor

Article References: Uluk, H. S., Dakka, S. M., & Singh, K. (2025). Flow behaviour of propane, ammonia and hydrogen in axisymmetric trapped vortex combustor. Discover Fluid Mechanics, 1(1), Article 2. https://doi.org/10.1007/s44369-025-00003-1

Image Credits: AI Generated

DOI: 10.1007/s44369-025-00003-1

Keywords: trapped vortex combustor, hydrogen fuel, ammonia fuel, propane, gas turbine combustion, fuel-air mixing, air injection angle, turbulent kinetic energy, RANS simulation, alternative fuels, aeroengine idle condition, combustion instability