The effects of non-premixed hydrogen enrichment on the flow field, fuel evaporation, flame topology, soot formation, and exhaust emissions of a swirl-stabilized Jet A-1 spray flame were investigated. Diagnostics included particle image velocimetry, polycyclic aromatic hydrocarbon laser-induced fluorescence, OH∗ chemiluminescence, soot luminescence, laser-induced incandescence, and exhaust gas analysis. Tests were conducted at a fixed thermal power of 8.83 kW, global equivalence ratios (ϕ) of 0.50 and 0.64, and hydrogen power fractions (H2) of 0%–50%. The stagnation point location exhibited non-monotonic and monotonic trends at ϕ=0.64 and ϕ=0.50, respectively, linked to variations in spray size and flame anchoring. At ϕ=0.64 and H2=0%, the flame was lifted, with soot concentrated along the shear region and peak mean soot volume fractions (f¯v) ∼50 ppb. At H2=10%, an attached non-premixed hydrogen flame front formed, reducing the peak f¯v to 8 ppb, likely due to enhanced fuel evaporation. At H2=20%–30%, increased droplet size and flame anchoring near the fuel nozzle potentially led to a high-temperature, fuel-rich zone that raised peak f¯v to 24–40 ppb. At H2=40%–50%, despite increased droplet size, the peak f¯v reduced to <5 ppb, below the detection limit, due to hydrodynamic and possibly chemical effects. CO emissions became negligible at H2=20%, relative to H2=0%, coinciding with flame anchoring closer to fuel nozzle, but increased at larger H2 as droplet size increased. NO emissions reached a minimum of 28 ppm at H2=10% and increased at larger H2, which may be associated with flames anchoring near the fuel nozzle and expected higher temperatures. At ϕ=0.50, soot and NO decreased, consistent with leaner operations and hydrodynamic effects, while CO increased, likely due to pronounced flame lift-off. Overall, for this configuration, soot and emissions are governed by competing effects of hydrodynamics, spray, flame topology, and hydrogen chemistry, depending on ϕ and H2.Novelty and significance statementHydrogen addition to liquid-fuel spray flames is often viewed mainly as a change in fuel chemistry. This study shows that it also reorganizes the spray, flow, and flame structures. In non-premixed hydrogen/Jet A-1 combustion, hydrogen enrichment can change the H/OH radical pool, while the reduced liquid-fuel flow limits carbon availability and alters droplet size, evaporation, and spray momentum. The attached hydrogen flame front also changes spray flame anchoring and topology. Together, these effects modify recirculation, residence time, local fuel–air mixing, and the spatial overlap between fuel-rich regions and reaction zones. This coupled behavior explains why soot, CO, and NOx do not vary monotonically with hydrogen fraction. The broader contribution is that hydrogen addition alone is insufficient to predict pollutant formation in dual-fuel spray combustion. Instead, soot formation and combustion emissions depend on the competition between spray evolution, flow hydrodynamics, spray flame anchoring, and the chemical effects of hydrogen addition.
Micron-sized, densely packed spherical aggregates of aluminum with a narrow particle size distribution were prepared by Emulsion-Assisted Milling (EAM). Combustion of this powder was compared to that of similarly sized atomized aluminum. Single particles of each powder were fed into an air-acetylene flame. The ignited particles continued to burn in room air. Each burning particle produced an optical emission pulse. Such pulses were measured using two photomultipliers filtered at 700 and 800 nm. The measured signals were processed to recover the pulse emission intensities, durations, and color temperatures for each particle. A statistically significant dataset of descriptors for the optical emission pulses produced by burning particles was obtained and processed for each powder. In addition, burned-out particles were collected and examined using electron microscopy. The results show that the EAM-prepared aluminum powder particles generate a stronger emission than atomized particles. The particle burn times are comparable for both powders. The optical signatures of the burning particles are successfully represented as a sum of contributions from both the particle surface and the standoff flame. In both cases, the flame is assumed to be at the boiling point of alumina. The particle surface temperature appears to be lower for the milled powder particles. Despite that, the emission from the standoff flame is markedly stronger for the EAM-prepared aluminum. Consistent with their lower surface temperatures, the EAM-prepared aluminum particles do not coalesce into single droplets during combustion, retaining at least partially their initial morphology. This altered surface morphology is likely responsible for the observed differences in their combustion characteristics.
This Brief Communication presents a novel bluff-body burner that exploits the well-known concept of compositional inhomogeneity and mixture stratification to extend turbulent flame stability toward leaner operating conditions. Fuel is injected orthogonally into an annular air stream, and the fuel injection ports can be recessed upstream of the burner exit plane by a distance Lr. Long recess distances allow almost complete mixing of fuel and air. However, at some intermediate Lr, mixture stratification within the recirculation zone that forms on the face of the bluff-body enables the stabilization of turbulent yet very lean flames. Experiments with propane-rich liquefied petroleum gas (LPG) show that reducing Lr from 105 mm to 15 mm shifts the peak-stability condition from a global equivalence ratio of φ ∼ 1.1 to φ ∼ 0.7, while maintaining a comparable maximum stabilization velocity. The lean stability limit is similarly extended from φ ∼ 0.8 to φ ∼ 0.5. Line Rayleigh-imaging of mixture fraction in non-reacting flows confirms that, at maximum stability, composition gradients are evident at the exit plane of the annulus and burnable mixtures dominate the recirculation zone downstream of the bluff-body.
