NH3/CH4 co-firing is a promising low-carbon fuel strategy, yet its application in micro gas turbines is hindered by poor flame stability and high NOx emissions, especially in compact combustion chambers. Porous medium combustion (PMC) offers a potential solution, but the micro-scale combustion mechanisms within different ordered cell topologies remain unclear. To overcome this issue, three topological cell structures including tetrakaidecahedron (Ter), regular hexahedron (Cube), and biclinic hexahedron (Bi-Cube) were designed and fabricated with identical porosity (epsilon = 97%). By combining experiments and numerical simulations, a topology-mediated multi-field (flow-thermal-chemistry) analysis was established to reveal how cell geometry governs flow recirculation, heat redistribution, and reaction pathways. The results show that the Bi-Cube structure achieves the widest stable combustion range (4h = 0.6-1.6), a higher and more uniform temperature field (Tmax = 1315 degrees C), and reduces the NO peak by 450 ppm compared to the Ter structure (ENH3 = 50%, 4h = 0.9). This is attributed to the high-shear and high-recirculation flow field enabled by its rotational symmetry design, which allows efficient mixing and heat recirculation at low flow resistance, thereby promoting a uniform temperature distribution that suppresses localized hot spots and shifts the NO formation/reduction balance toward N2. As ENH3 increases, the dominant NO formation pathway shifts from CH3 oxidation/recombination competition to NH oxidation/reduction kinetics competition, with the transition at ENH3 approximate to 50%. The Bi-Cube further favors the NH -> N2 reduction pathway under this regime. This study elucidates topology-mediated fluid-thermal coupling effects on NH3/CH4 combustion, providing a rational design basis for efficient, low-emission co-firing technology in micro gas turbines.

This study investigates the vertical fire spread characteristics on the external walls of high-rise buildings with overhead floors, focusing on the coupled effects of overhead floor height, reference wind speed, and heat release rate per unit area (HRRPUA). A numerical model of a 16-story building was established using Fire Dynamics Simulator (FDS), considering three overhead floor heights (3.0 m, 4.5 m, 6.0 m), four reference wind speeds (0 m/s, 3.2 m/s, 10.7 m/s, 20.6 m/s), and three HRRPUA levels (250 kW/m2, 500 kW/m2, 1000 kW/m2) representing a range from typical to extreme electric bicycle fire scenarios. A key metric, termed the “flame fusion height” (defined as the vertical height where flame plumes from openings merge, identified by the 540 °C isotherm), was analyzed to quantify fire spread severity. The results reveal a complex, nonlinear interaction among the factors. The flame fusion height increased significantly with HRRPUA, but this effect was strongly modulated by wind speed and overhead floor height. A critical reference wind speed of approximately 10.7 m/s was identified. Below this threshold, wind could enhance flame attachment and increase fusion height compared to calm conditions. Above it, wind exerted a dominant suppressing effect, drastically reducing the fusion height by disrupting flame stability. Increasing the overhead floor height overall led to a reduction in flame fusion height by 20–50

NH3 is a carbon-free fuel with potential for decarbonizing high-temperature industries, but its low reactivity and high NO emissions remain major challenges. Blending with CH4 provides a practical pathway considering CH4 not only improves flame stability but mitigates costs with possibility of partial fuel substitution. In terms of NO formation, spatially resolved NO speciation data in canonical counterflow diffusion flames (CDFs)-a configuration representative of industrial non-premixed combustion-are still scarce. In this work, spatially resolved NO measurements were performed in NH3-CH4 CDFs using a self-developed calibration-free technique that combines ceramic microprobe sampling with mid-infrared laser absorption spectroscopy. The effects of NH3 fraction, oxygen enrichment, and strain rate were systematically investigated, supported by detailed kinetic modeling and pathway analyses. Results show that NO formation in NH3-CH4 CDFs increases monotonically with NH3 fraction, in contrast to the non-monotonic trends reported in premixed flames. This behaviour is explained by enhanced HNO-mediated fuel-N pathways and the spatial segregation of important species that is unique in CDFs. Furthermore, unlike CH4 CDFs where NO decreases with strain rate, pure NH3 flames exhibit higher NO at elevated strain rates due to flame stretching that increases OH-NH2 overlap. These findings fill a critical data gap and reveal fundamental differences between premixed and counterflow diffusion flames; they also provide mechanistic insights for burner design and NOx control in NH3-fueled systems.

Understanding the complex combustion dynamics within scramjet engines is critical for advancing high-speed propulsion technologies. However, the large-scale, high-dimensional nature of temporal flow field data from numerical simulations presents significant challenges for visual interpretation, feature differentiation, and cross-case comparison. In this paper, we present TemporalFlowViz, a parameter-aware visual analytics workflow and system designed to support expert-driven clustering, visualization, and interpretation of temporal flow fields derived from scramjet combustion simulations. Our approach leverages hundreds of simulated combustion cases with varying initial conditions, each yielding temporal sequences of flow field images. We employ pretrained Vision Transformers to extract high-dimensional embeddings from flow field frames, apply dimensionality reduction and density-based clustering to uncover latent combustion modes, and construct temporal trajectories in the embedding space to track the evolution of each simulation over time. To bridge the gap between latent representations and domain knowledge, specialists annotate representative cluster centroids with descriptive labels. These annotations serve as contextual prompts for a vision-language model, which generates natural language summaries for individual frames or complete simulation cases. Furthermore, the system supports parameter-based case filtering, similarity-based case retrieval, and coordinated multi-view exploration to facilitate in-depth analysis. We demonstrate the effectiveness of TemporalFlowViz through two expert-informed case studies and domain expert feedback, showing that TemporalFlowViz facilitates hypothesis generation, supports interpretable pattern discovery, and accelerates knowledge discovery in large-scale scramjet combustion analysis.

Ammonia, a hydrogen-rich fuel with easy liquefaction and mature infrastructure, is a promising solution for the challenges of hydrogen storage/transportation via decomposition. Autothermal microchannel reactors, which utilize ammonia oxidation heat and feature low cost and high integration, were evaluated herein through simulation. The influences of reactor material and channel architecture were uncovered. It reveals that the catalyst temperature increases sharply near the inlet then decreases gradually downstream, causing local ammonia conversion rate increasing rapidly initially and then more slowly. Only a fraction of energy is utilized for decomposition, and the remaining is lost mainly through reactor walls. Using aluminum as the material reduces the high-temperature catalyst region, leading to ammonia conversion rate CRdecom,out and reactor energy efficiency eta reactor 9.5% and 9.3% respectively lower than stainless steel 310. Decreasing channel number or length lowers reactor wall loss and expands high-temperature region, improving CRdecom,out and eta reactor. Nevertheless, excessive reduction risks catalyst agglomeration and simultaneously decreases residence time, deteriorating performance. Consequently, selecting appropriate channel number and length is crucial. Based on this study, a reactor configuration using stainless steel 310, a channel number of 1200 and a channel length of 80 mm is recommended, which improves CRdecom,out and eta reactor by 5.6% and 5.8% versus baseline. However, the performance remains suboptimal. Parameter analysis thus informed the development of an improvement approach that extends high-temperature catalyst region while eliminating catalyst overheating, and reduces heat loss. This study is expected to provide valuable insights for designing high-performance autothermal microchannel reactors.

Partially cracking ammonia into hydrogen offers a practical route to overcome its inherently low reactivity while leveraging existing ammonia infrastructure for zero-carbon combustion. In this study, we designed and tested a similar to 10 kW premixed swirl burner to systematically investigate the influence of swirler geometry, ammonia cracking ratio, and excess air ratio on flame stability and NOx emissions. Complementary numerical simulations were conducted to elucidate the principal elementary reactions and pathways governing nitrogen oxides (NOx) formation. The results indicate that under lean premixed conditions, nitrous oxide (N2O) concentrations remain at single-digit parts per million (ppm) levels, while residual ammonia and nitric oxide (NO) concentrations in the exhaust exhibit a cracking-ratio-dependent trade-off: stronger swirl over-stretches the flame and elevates NH3 slip at low cracking ratios, whereas at higher cracking ratios it promotes complete oxidation at the cost of increased NOx. In contrast, weaker swirl suppresses NO but leaves elevated NH3 slip ratios due to incomplete burnout. Increasing the cracking ratio shortens the visible flame, shifts the flame morphology from V-shaped to straight, and promotes NOx via enhanced radical chemistry. A higher excess air ratio further accelerates NO production due to intensified OH-driven pathways. Overall, this study provides direct experimental and numerical evidence that swirl design and controlled cracking are critical levers for optimizing ammonia-hydrogen premixed burners. The insights contribute to the rational design of ammonia-hydrogen burners for high-temperature industrial applications.

Towards high-efficiency energy conversion using bulk aluminum, this work presents an investigation on the generation, ignition, and combustion of molten aluminum sprays. A novel dual-gas atomization method integrated with inductive heating was developed to produce aluminum sprays at pressures below 0.5 MPa, yielding droplets with a mean size of 180 µm. Finer droplets within the spray were successfully ignited by the post-flame flow of a premixed fuel-rich hydrogen/oxygen (H₂/O₂) flame under ambient conditions, exhibiting strong luminosity and combustion temperatures exceeding 2700 K. In-situ optical diagnostics and ex-situ microscopic analyses of the intermediate products revealed key features of the spray combustion process, including the formation of a core-shell structure of aluminum encapsulated in alumina, aluminum vapor ejection, and the coalescence and breakup of burning droplets. These findings provide insights into low-pressure aluminum spray formation and multi-droplet spray combustion, offering guidance for the design of metal-fuel-based power systems. Novelty and significance statement This study presents the development of the first controlled spray combustion system converting bulk aluminum into ignitable sprays, which is a significant advance beyond prior studies focused on powders or isolated droplets, as a critical step toward practical aluminum-fueled power systems. A novel, low-cost, and efficient aluminum atomization method is developed at the lab scale. Through combined in-situ optical diagnostics and ex-situ microscopic characterization, this study reports new results on the formation and combustion of molten aluminum sprays. The findings provide critical knowledge for developing practical high-efficiency metal-fuel energy conversion technologies and advancing the aluminum-based energy cycle.

Staged burners have been widely used in industry to reduce NOx emissions. However, the overall effectiveness depends on its operation conditions. Traditional monitoring methods of NOx and CO emissions, such as gas analyzers, are too slow and impractical for real-time optimization. This study investigates flame chemiluminescence as a practical, non-intrusive indicator for combustion control. A staged-burner experimental platform was developed allowing independent control of fuel and air supply. Pollutant concentrations (CO, NOx) and flame radiative emission spectra (OH*, CH*, C2*) were measured under varied excess air ratios. Results reveal that CH*/OH* ratios decrease linearly, while C2*/CH* and C2*/OH* exhibit unimodal behaviors with increasing air ratio. These spectral signatures strongly correlate with emission levels, enabling indirect yet reliable indicators for low-emission operation. Spatially resolved measurements with ICCD imaging further validate optimal detection zones and filter bandwidth effects, ensuring applicability with cost-effective sensors. Overall, this work demonstrates the feasibility of chemiluminescence-guided closed-loop control, offering a pathway toward cleaner and more efficient industry burners.

The thermochemical nonequilibrium effect on hydrogen- and ethylene-fueled supersonic combustion was modeled in the framework of a two-temperature-based Dynamic Zone Flamelet Model (DZFM) with up to 221.21 million cells. The numerical predictions of wall pressure match well with the experimental data for both the cold and reacting cases. Thermal nonequilibrium is first introduced by the inlet compression, then intensified by the jet-excited shock wave and expansion wave, and eventually emerges again in the divergent nozzle after a temporal equilibrium status. Delayed ignition was observed under nonequilibrium for both hydrogen and ethylene combustion. The relative importance of Tt and Tv on elementary reactions under different degrees of thermal nonequilibrium (Tv/Tt) was analyzed using reaction path flux analysis over the typical temperature range in scramjets. Inverse reaction path directions were observed in the hydrogen ignition process between equilibrium and nonequilibrium. The reaction rate of the chain branching reaction H+O2=O+OH is dominated by Tv under intense thermal nonequilibrium (Tv/Tt<0.5). Inhibition of H+O2=O+OH results in delayed hydrogen ignition under nonequilibrium. For the ethylene case, the inhibition of the dissociation reactions C2H4=C2H3+H and C2H3=C2H+H2 dramatically decreases the concentration of active radicals, leading to delayed ignition under nonequilibrium. The thermochemical nonequilibrium effect overall leads to a lower air capture rate, lower combustion efficiency, and lower net thrust.

To overcome the limitations of disordered porous ceramics in micro gas turbines (MGTs) combustors, this study systematically evaluates the mechanical and thermal performance of disordered foams against three ordered SiC topologies (regular hexahedral RH, regular octahedral RO, and regular Kelvin RK). Coupled convection-radiation analysis demonstrates that structural periodicity significantly enhances multi-functional properties. Specifically, the RK topology exhibits superior heat transfer and temperature uniformity, while the RH and RO excel in mechanical strength and flow resistance, respectively. Mathematical models were established to quantify correlations between porosity, mechanical properties and heat transfer efficiency. Results indicate that while porosity dictates mechanical strength, heat transfer is synergistically controlled by both porosity and characteristic size. Notably, the RK achieves optimal comprehensive efficiency at 75.0% porosity and 6.0 mm edge length, providing a quantitative framework for the performance-driven design of SiC ceramic components in MGTs combustors.

Low-temperatures and low-pressures at high-altitudes (Plateau) pose serious challenges for aero-engine combustor ignition. Plasma-assisted ignition has emerged as a promising technology to achieve reliable ignition in extreme conditions. Experiments were conducted in a dual-dome aero-engine combustor sector using RP-3 aviation kerosene, under high-altitude ground-starting conditions (altitude: 0 -> 6 km, inlet pressure: 101.32 -> 47.22 kPa, inlet temperature: 288.15 -> 249.19 K). The ignition performances of conventional electric spark ignition, oxygen-enriched spark ignition, and gliding arc plasma ignition were systematically compared. Results showed that gliding arc ignition expanded the ignition boundary by 20.00% and 10.20% at 4 km altitude compared with electric spark ignition and oxygen-enriched ignition, respectively. Analyses of flow field, spray distribution, and cracking products provide direct evidence of the mechanisms by which gliding arc plasma improves atomization, evaporation, and ignition reliability. CH* chemiluminescence visualization revealed a distinct flame kernel propagation pathway unique to gliding arc ignition. In contrast, for electric spark ignition and oxygen-enriched ignition, the flame kernel exhibited a long propagation pathway and short duration time, resulting in high energy dissipation and low ignition success probability. The gliding arc plasma not only enhanced atomization, mixing, and evaporation at low-temperatures, but also enabled continuous heat release accumulation and high energy utilization efficiency, thereby facilitating multi-point ignition inside the combustion chamber. At 6 km altitude, gliding arc ignition reduced the ignition delay time by factors of approximately 33 and 13 times compared with electric spark and oxygen-enriched ignition, respectively. These findings demonstrate that gliding arc plasma ignition substantially elevates the effective ground ignition altitude of aero-engines and provides a reliable pathway for ignition under low-temperature, low-pressure conditions.

Cycloalkanes constitute a crucial component of real-world transportation fuels. Understanding their combustion chemistry, particularly the pathways towards harmful soot emissions, is critical for optimizing practical combustion systems. In this study, we experimentally investigated the fuel decomposition and aromatic formation behaviors of selected cycloalkanes, namely cyclopentane (CPT), cyclohexane (CHX), and methylcyclopentane (MCPT), in laminar counterflow diffusion flames. These model fuels are selected to focus on the effects of ring size and methyl substitution on aromatic species formation. Our experimental findings reveal that C5-ring fuels (CPT and MCPT) demonstrate significantly enhanced aromatics yields compared to six-membered CHX, and this trend was further validated through comparative studies of analogous cyclic alkenes (cyclopentene vs. cyclohexene). Measurement and modeling results highlight the advantages of C5-ring fuels in generating odd-carbon species, whose interaction reactions (e.g., C5H5 + CH3, C5H5 + C5H5) serve as efficient aromatic formation pathways. Methyl substitution on the ring causes MCPT to produce more benzene and two-ring aromatics than CPT, as the presence of CH3 enhances the fuel decomposition pathways forming odd-carbon cyclopentyl and C5H5 radicals. This study also identifies certain limitations in current models' ability to accurately predict cycloalkane combustion chemistry. In this regard, the novel experimental data herein may serve as valuable validation benchmarks for the future development of predictive models.

CO and NO are major gaseous pollutants emitted during municipal solid waste combustion, and gas emission models based on component ratios enable rapid prediction of gaseous emissions under variable waste compositions, supporting component-level emission assessment and control. This study investigates five representative MSW components, namely vegetable leaves, fruit peels, rice, paper, and plastic. CO and NO emissions are measured at various furnace temperatures (700–1100 ℃) and air flow rates (800–1600 mL/min). Emission characteristics at various mixing ratios (0.25–0.75) are examined at 1000 ℃ and 1200 mL/min. A mixture design is employed to develop predictive models for CO and NO emissions based on component ratios. The results show that CO emissions from all components and NO emissions from food waste reach minimum values at specific furnace temperatures. Increasing the air flow rate promotes CO and NO production; however, CO emissions from rice reach the peak at 1400 mL/min. Component mixing inhibits CO formation but enhances NO formation. The quadratic models predict a minimum CO emission at 0.082 mg and a maximum NO emission at 0.323 mg in the mixture excluding plastic and rice. These findings provide a model reference for the rapid estimation of CO and NO emissions from mixture.

Ammonia/methane cofiring is a practical route for decarbonizing high-temperature industrial combustion, but NO control remains challenging because NH3 introduces strong fuel-NO chemistry. H2O addition is an established NO-mitigation strategy in hydrocarbon combustion; however, its role in NH3-containing flames remains unclear, especially in nonpremixed flames for which spatially resolved data are still scarce. In this work, spatially resolved NO measurements were performed in pure CH4 and NH3-CH4 nonpremixed flames using a counterflow configuration by a self-developed calibration-free diagnostic based on microprobe sampling and tunable diode laser absorption spectroscopy. Premixed flames using the same counterflow configuration were analyzed numerically for comparison. The effects of H2O fraction, NH3 blending ratio, and equivalence ratio on NO formation were evaluated, and sensitivity and pathway analyses were used to separate the thermal-dilution and chemical effects of H2O addition. In pure CH4 flames, H2O addition suppresses NO in both configurations, mainly through thermal dilution, although this suppression weakens under rich premixed conditions. In NH3-CH4 nonpremixed flames, however, H2O addition increases NO because H2O-induced radical changes offset the temperature reduction. By contrast, in NH3-CH4 premixed flames,H2O addition reduces NO, and the suppression becomes stronger at higher equivalence ratio because the thermal-dilution effect remains dominant. These results show that the NO response to H2O addition in NH3-CH4 flames is strongly configuration-dependent and that NO-control strategies effective for hydrocarbon flames cannot be transferred directly to ammonia-containing systems.

In counterflow diffusion flames, traditional one-dimensional simulations often yield inaccuracies due to the presence of two-dimensional effects, which arise from the actual burner geometry as well as buoyancy effects. Direct Numerical Simulations (DNS) can accurately resolve these effects but require significant computational resources. This study introduces a flamelet model to address this gap. The model is built using three tabulated scalars: two mixture fractions and one stoichiometric scalar dissipation rate. The results demonstrate that the flamelet model effectively captures two-dimensional effects while maintaining computational efficiency comparable to one-dimensional simulations. Hence, it has the potential to serve as an alternative to traditional onedimensional models and could provide the community with a useful tool for assessing the performance of kinetic mechanisms in counterflow diffusion flames. Parametric studies reveal that increasing strain rates and asymmetries in curtain flow rates amplify the two-dimensional effects. Furthermore, the buoyancy effect was investigated, showing its increasing significance as the Richardson number increases. The flamelet model provides an efficient and accurate tool for simulating counterflow flames with realistic burner geometries and buoyancy effects.

Light-matter interactions provide versatile routes for probing and controlling chemical reactivity, charge transport, and material properties. Time-periodic external fields can reshape electronic states and open new dynamical pathways beyond the field-free Born-Oppenheimer (BO) picture. Floquet nonadiabatic dynamics has consequently emerged as an important framework for describing coupled electron-nuclear dynamics under periodic driving. In this Perspective, we first discuss recent developments in Floquet nonadiabatic dynamics methods for closed and open quantum systems. We then highlight how this framework provides mechanistic insights into electron transfer at molecule-metal interfaces, quantum transport in molecular junctions, carrier dynamics in crystalline solids, and multicolor Floquet engineering. Finally, we outline key conceptual and computational challenges that must be addressed to transform Floquet nonadiabatic dynamics from model-based demonstrations into predictive, first-principles simulations of realistic light-driven processes.