This study employs three-dimensional (3D) numerical simulations to determine the minimum equivalence ratio for which a premixed hydrogen-air flame can propagate within a 4 mm gap bounded by cold isothermal walls. This limiting condition provides a quantitative measure of practical interest in hydrogen safety. Using a previously validated one-step reduced chemical-kinetic mechanism with a simplified transport model, the complex flame dynamics occurring near this propagation limit is investigated and compared with recent experimental measurements. The simulations confirm that the interplay between conductive heat losses to the walls and diffusive–thermal instabilities is responsible for fragmenting the continuous reactive front into isolated flame kernels capable of sustaining flame propagation under conditions that would otherwise be impossible. The simulations identify the critical equivalence ratio that triggers the transition between two different isolated flame morphologies: the circular flame and the double-cell flame. Circular flames, which propagate slowly as an oblate-spheroidal surface, are sustained primarily by diffusion, which supplies fuel from all directions. In contrast, for double-cell flames, comparable diffusive and convective fluxes increase sensitivity to far-field perturbations, which can lead to an unstable motion or a fractal-like propagation via successive flame splitting. Notably, both flame morphologies maintain symmetry relative to the gap centerplane. By mapping these regimes as a function of the equivalence ratio, regions of stable coexistence are identified and characterized. While the simulations accurately capture observed regime transitions, some quantitative discrepancies remain in the predicted velocity and size of these isolated flames when compared with reported experimental measurements.Novelty and significance statementThe novelty of the present work lies in the detailed 3D numerical characterization of the two different isolated flame-cell regimes that emerge near the extinction limit of hydrogen–air mixtures, providing insights into the internal cell structure that remains inaccessible to current experimental analysis. Crucially, we show that these isolated cell structures survive well below the standard flammability limit predicted by planar flame theory, challenging established safety assumptions for hydrogen-based systems.
We assess the use of broad-band (BB) and narrow-band (NB) forcing signals to retrieve the Flame Transfer Functions (FTF) from experimental measurements in premixed turbulent flames where the latter uses monochromatic and Chirp signals and the former dichromatic and white/coloured noise signals. At low excitation amplitudes, all forcing methods provide similar estimates of the FTF. Although broad-band excitation provides information over a range of frequencies simultaneously, it also promotes nonlinear saturation effects over the whole frequency range. We show that these effects occur due to interactions between the response at high and low frequencies that cause significant amplitude saturation, even for relatively low excitation amplitudes (<2%). High-frequency modes were found to modify the phase distribution at lower frequencies leading to an increase in spatial interference and a reduced global HRR fluctuation amplitude. These results show that saturation must be avoided at all frequencies which significantly limits the use of BB signals without substantial a priori knowledge of the flame response. To circumvent this limitation, we utilise an amplitude modulated Chirp signal to obtain a time-series that sweeps the frequency range over the same total time duration as the applied broad-band signal. We show that this method reliably captures the FTF by avoiding nonlinear frequency interactions and only uses a single time-series measurement. This significantly reduces experimental effort required to retrieve FTFs over a wide range of operating conditions.Novelty and significance statementThis work presents the first systematic experimental comparison of broad-band and narrow-band forcing strategies for measuring flame transfer functions (FTFs) in turbulent premixed flames. The findings highlight a fundamental limitation of broad-band approaches for reliable FTF identification in turbulent flames, as nonlinear cross-frequency interactions cause gain reduction across the entire frequency range. We propose an amplitude-modulated Chirp methodology that preserves narrow-band behaviour while sweeping the full frequency range within a single time-series. This approach enables accurate FTF retrieval without cross-frequency contamination and substantially reduces experimental time, facilitating efficient mapping of FTFs across operating conditions.
This study presents simultaneous time-resolved measurements of NH3, N2O, and NO concentration profiles in NH3/N2O/Ar mixtures behind reflected shock waves using tunable diode laser absorption spectroscopy (TDLAS). Spectrally resolved N2O absorption cross-sections were measured near 2193 cm−1 over 909–2113 K. Three equivalence ratios (φ = 0.5, 1.0, and 2.0) with an argon dilution ratio of 95% were investigated at temperatures of 1687–2235 K and pressures of 1.0–1.3 bar. Building on these data, the PTB-NH3/C2 mechanism was upgraded from v1.1 to v1.2. The key update is the adoption of the HNNO sub-mechanism for the H + N2O system, together with updated N2O + O = N2 + O2 and N2H2 + NO = N2O + NH2 rates consistent with recent ab initio determinations. A bounded multi-target genetic algorithm then refined nine sensitive pre-exponential factors within their uncertainty intervals. This upgrade reduced the mean absolute relative discrepancy in peak NO mole fraction from 38.8% to 10.7%, with consistent gains for the NH3 and N2O half-consumption times and the NO half-rise time. Sensitivity and pathway analyses show that the corrected N2O + O and N2H2 + NO rates remove spurious NO formation and resolve the systematic NO over-prediction, while the H + N2O channel, now described via the HNNO chemistry, together with thermal N2O decomposition governs the N2O consumption budget. Cross-validation against independent datasets confirmed that v1.2 is physically consistent and transferable beyond the training conditions.
Titanium hydride (TiH₂) particle-cloud combustion was investigated using ignition experiments, a zero-dimensional single-particle model, and an Eulerian–Lagrangian simulation framework. The single-particle model coupled endothermic TiH₂ decomposition with a two-branch apparent Ti-oxidation kinetic limit and a Stefan-flow-corrected external oxygen-transport limit, whereas detailed gas-phase H₂ chemistry was included in the particle-cloud simulations. For a 15 μm particle at 1800 K and an oxygen mass fraction of 0.23, the calculated peak temperature was approximately 1855 K, 55 K above the gas temperature. Across 1200–2100 K and oxygen mass fractions of 0.10–0.50, the heating time generally decreased while the peak particle–gas temperature difference increased; all standalone cases remained predominantly kinetically controlled. Experimentally, the maximum flame height increased from approximately 230 to 1450 mm as the particle concentration increased from 125 to 500 g·m⁻³, and the shortest venting time, approximately 12 ms, occurred at 375 g·m⁻³. The Eulerian–Lagrangian simulation reproduced the principal flame propagation trends and suggested that shear-layer vortices, ambient-air entrainment, and continued combustion of residual H₂ and partially reacted particles contributed to the external mushroom-shaped flame.
Ammonia is a carbon-free fuel with poor ignitability due to its low chemical reactivity, while hydrogen enrichment offers an effective route to promote ammonia ignition. However, practical ignition events typically occur within strained flow environments, where successful ignition depends on whether the flame kernel can be sustained under strong dissipation before developing into a self-sustaining flame. In this study, direct numerical simulations were conducted to investigate the forced ignition of premixed ammonia/hydrogen/air mixtures in laminar counterflow, with emphasis on strain rate and hydrogen enrichment. The results show that strain rate deforms the flame kernel into an asymmetric elliptical structure, with radial stretching and severe axial compression. The axially compressed flame front exhibits a narrowed reaction zone and intensified scalar dissipation, thereby controlling successful ignition or global extinction. Increasing strain rate suppresses ignition by enhancing dissipative loss at this axial flame front. Hydrogen enrichment simultaneously enhances the chemical heat release and dissipation but shifts flame kernel evolution from dissipation-dominated decay to chemistry-dominated survival by enhancing H/OH radical chemistry through H2 + OH = H + H2O, shortening the chemical timescale, and maintaining Da > 1.0. The ignition boundary is represented by the strain-dependent minimum ignition energy (MIE), with MIEc as the static ignition limit and ac as the extinction strain limit. Hydrogen enrichment shifts the ignition boundary toward lower ignition energies and higher strain rates, thereby expanding the ignitable region. After normalization by MIEc and ac, the ignition boundaries can be represented by an empirical normalized power-law correlation, providing an empirical representation of the ignition boundaries. These findings reveal a strain-controlled flame kernel development mechanism and provide insight into ammonia/hydrogen ignition under strained conditions. Novelty and significance statement The novelty of this work lies in characterizing ammonia/hydrogen/air forced ignition under well-defined strained-flow conditions, where flame kernel survival cannot be understood solely from quiescent ignition behavior. It identifies the asymmetric flame kernel response, clarifies the critical role of the axially compressed flame front, and reveals the competition between hydrogen-induced chemical enhancement and strain-induced dissipation during hydrogen-assisted ignition. This work is significant because it extends the understanding of ammonia/hydrogen/air forced ignition from quiescent conditions to strained flows and provides a quantitative basis for assessing flame kernel survival and extinction. The resulting ignition boundaries and normalized correlation provide guidance for interpreting strain-controlled ignition in practical carbon-free combustion systems.
The ignition delay time of aluminum-based alloys is a key parameter for evaluating their combustion stability, reactivity, and energy release efficiency. The traditional single-sensor measurement method is limited by its respective physical principles and operational characteristics, often resulting in inconsistent results. This is partly due to the fact that different methods respond to different physical stages of the ignition process, and thus the differences are inherently physical and difficult to reliably characterize the ignition behavior of different alloy systems. To compare the ignition delay times of three materials: micron-sized aluminum (μAl), aluminum-lithium alloy (Al-2.4Li), and aluminum-magnesium alloy (Al-3.0Mg), this paper conducts systematic research using single-modal measurement methods (image method, temperature method, spectral method) and a multi-modal data fusion method based on Dempster-Shafer (D-S) evidence theory. The results of single-modal experiments show that the measurement results of the image method and the temperature method are relatively consistent, while the results of the spectral method are significantly higher, highlighting the insufficient repeatability of a single method in different alloy systems. After applying the D-S evidence theory fusion framework, the measurement consistency is significantly improved, and the stability of the spectroscopy method increases by 57.2% - 68.9%, and the intra-group dispersion of multiple methods decreases by 62.9% - 88.1%. The normalized ignition delay time after fusion is: micron-sized aluminum 4.30 - 5.01 ms, aluminum-lithium alloy 4.37 - 4.77 ms, and aluminum-magnesium alloy 4.19 - 4.80 ms. Among the three samples, the ignition delay time of aluminum-magnesium alloy is the shortest, followed by aluminum-lithium alloy, and the longest is micron-sized aluminum. Mechanism analysis indicates that the "internal heating" effect produced by magnesium evaporation combustion shifts the ignition time of aluminum-magnesium alloy to the lower end, while lithium's preferential oxidation consumes part of the initial energy, resulting in a smaller improvement in aluminum-lithium alloy compared to aluminum-magnesium alloy. This study demonstrates that the multi-modal data fusion technology can effectively coordinate the physical differences of single-modal measurements and provide a unified data benchmark for accurately comparing the ignition performance of different aluminum-based alloys, which is of great significance for improving the accuracy of ignition analysis and the evaluation of combustion characteristics of aluminum-based alloys.Novelty and significance statement: This work resolves a critical barrier in energetic material characterization: single-modal methods (image, temperature, spectroscopy) yield inconsistent ignition delay measurements across Al-based alloys due to element-specific oxidation/evaporation behaviors. Applying D-S evidence theory-based multi-modal fusion reduces spectral method dispersion by 57.2–68.9% and intra-group variance by 62.9–88.1%, establishing a unified time reference. The fusion reveals unambiguous ignition ranking: Al-Mg < Al-Li < μAl. Magnesium's "internal heating" via evaporative combustion promotes ignition, while lithium's preferential oxidation consumes energy. This fusion framework transforms measurement noise into physically interpretable variability, capturing inherent ignition stochasticity rather than merely smoothing data. It provides a high-fidelity benchmark for elucidating alloy-controlled ignition dynamics and designing advanced reactive materials.
This work investigates the oxidation behaviors of neat ammonia, neat iso-octane, and their blends in a jet-stirred reactor over a temperature range of 650–1100 K and equivalence ratios of 0.50–0.56 at atmospheric pressure. Reactants, intermediates and products are quantified using synchrotron vacuum ultraviolet photoionization mass spectrometer (SVUV-PIMS) and gas chromatograph (GC) technologies. An updated chemical kinetic model is developed to interpret the kinetic interactions between ammonia and iso-octane oxidation. Rate constants for the reactions between IC8H18 and NH2 are computed using the transition state theory (TST) under the rigid-rotor/harmonic-oscillator (RRHO) approximation over 300–1500 K. The results show that the dual-fuel interactions significantly enhance oxidation reactivity at both intermediate and low temperatures. In neat NH3 oxidation, the reactivity of NH3 is extremely low below 1000 K due to the slow formation of HO2 and OH radicals. Blending with iso-octane substantially increases the radical pool, thereby accelerating NH3 oxidation via NH3 + OH = NH2 + H2O pathway. Compared with O and OH radical production, HO2 formation via H + O2(+M) = HO2(+M) is less effective in promoting NH3 oxidation because of competition for H radicals. The sensitivity analysis reveals that the newly-induced pathways, including CH3 + NO2 = CH3O + NO, IC4H7 + HO2 = IC4H7O + OH and CH3 + HO2 = CH3O + OH, play important roles in enhancing NH3 oxidation at intermediate temperatures. Meanwhile, the H-abstraction of IC8H18, CH2O and C3H6 by NH2 via IC8H18 + NH2 = C8H17 + NH3, CH2O + NH2 = HCO + NH3 and C3H6 + NH2 = C3H5 + NH3 exhibits the kinetic interaction between NH3 and IC8H18. NOx formation by NH3 oxidation accelerates iso-octane oxidation by inducing additional OH production pathways. Reactions between NO2 and fuel radicals (CH3, NH2, H) continuously produce active intermediates such as CH3O, H2NO and OH. The strong catalytic effects of NO-alkylperoxy radical and NO2-fuel radical interactions not only accelerate low-temperature iso-octane oxidation by sustaining the radical pool, but also modify the oxidation pathways by inhibiting the oxidation pathway from CH3O2 to CH3O2H. This work provides a valuable experimental database for kinetic mechanism development and advances the understanding of co-oxidation mechanisms for ammonia and large hydrocarbons.Novelty and significance statement: Large hydrocarbons with low-temperature chemistry have the potential to enhance the oxidation reactivity of ammonia in dual-fuel internal combustion engines. As a reference fuel of gasoline, the blends of iso-octane with ammonia offer a balance for carbon-free combustion and high energy content. However, the oxidation characteristics of ammonia/iso-octane blends and their kinetic interaction are still not fully understood. This work presents the oxidation characteristics of ammonia/iso-octane blends at low and intermediate temperature conditions by using SVUV-PIMS and GC techniques. The mutual reinforcement in the oxidation reactivity between ammonia and iso-octane is studied by an updated kinetic mechanism. Especially, the H-abstraction reactions between iso-octane/n-alkanes/n-alkenes/CH3/CH2O and NH2 are fully incorporated. The rate constant between IC8H18 and NH2 is computed using the transition state theory (TST) under the RRHO (rigid-rotor/harmonic-oscillator) approximation. This work provides insights into the co-oxidation mechanism of ammonia and iso-alkanes. The experimental data and kinetic mechanism are valuable for faithful kinetic model development with C-N chemistry interactions.
The combustion kinetics of 1-methylnaphthalene, a representative diesel surrogate component, were investigated by combining detailed speciation measurements in a quartz flow reactor with comprehensive kinetic modeling. Experiments were conducted at 940–1344 K, 0.04 atm, and an equivalence ratio of 0.5 using synchrotron photoionization mass spectrometry. Concentration profiles of >40 intermediates, including radicals, isomers, polycyclic aromatic hydrocarbons (PAHs), and oxygenated PAHs (OPAHs), were obtained as a function of temperature. Primary fuel consumption proceeds via both methyl-side-chain and aromatic-ring H-abstraction reactions, with comparable contributions above 1050 K. Benzobenzyl and benzofulvenallenyl radicals were identified as key intermediates governing PAH growth and oxidation. Naphthaldehyde and methylnaphthol are the major oxygenated products, formed via oxidation at the methyl side chain and aromatic ring, respectively, with the former exhibiting a much higher yield due to the greater abundance of naphthylmethyl radicals. Comparison with indene oxidation reveals substantially lower PAH growth propensity for 1-methylnaphthalene, underscoring the mechanistic differences between methyl-substituted six-membered and five-membered aromatic systems. A detailed kinetic model incorporating extended aromatic and oxygenated aromatic chemistry was developed and validated against the present data as well as literature results from jet-stirred reactor, flow reactor, ignition delay, and laminar flame studies. The combined experimental and modeling results provide new mechanistic insight into the oxidation pathways of 1-methylnaphthalene and the formation of OPAHs during aromatic fuel combustion. Novelty and significance statement The novelty of this study lies in the first comprehensive kinetic model for 1-methylnaphthalene combustion. Experiments identify previously unreported bicyclic aromatic radicals and debated oxygenated PAHs (OPAHs). A new mechanism, incorporating elementary decomposition pathways of naphthylmethyl radicals and OPAHs, elucidates their roles in shaping product distributions during oxidation. Comparison with indene oxidation clarifies distinct kinetic and thermodynamic behaviors in five- and six-membered radical recombination, providing insight into larger PAH growth. This work advances the fundamental understanding of 1-methylnaphthalene combustion as a representative bicyclic aromatic in diesel and jet fuels as well as PAH and soot evolution kinetics.
