The temperature distribution determines the product in nanoparticle synthesis in flames, as precursor reactions, flame dynamics, particle nucleation, and growth depend on the temperature history of the aerosol. Fields of gas-phase temperature and OH concentrations are measured by two-line OH laser-induced fluorescence (LIF) in the SpraySyn burner during iron oxide nanoparticle formation. A correction of the LIF images based on the Beer–Lambert law was applied to compensate for spatial laser beam attenuation within the flame. This correction enables quantitative evaluation of the OH-LIF signals, allowing the determination of OH concentrations together with instantaneous temperature fields by LIF-intensity ratioing of two selected transitions excited in the A2Σ–X2Π (1,0) electronic band system. We investigate the effect of varying oxygen dispersion gas flow rates on the temperature distribution in the spray flame operated with pure ethanol and with mixtures of ethanol and 2-ethylhexanoic acid. In the latter case, the effect of iron(III) nitrate nonahydrate (INN) addition as a precursor for iron-oxide nanoparticle synthesis is examined for 0.01 and 0.05 mol/l. Increasing the dispersion gas flows reduces the overall temperatures and narrows the high-temperature zone. INN addition causes a slight temperature increase (by 20 K) at 0.05 mol/l. Temperatures determined from instantaneous measurements are compared with measurements from multi-line OH-LIF thermometry (Karaminejad et al., Appl. Energy Combust. Sci. 2023) and differences are discussed. The measured temperature and OH datasets provide detailed insight into the thermo-chemical structure of SpraySyn flames and serve as valuable input for the development and validation of CFD models for particle formation. Novelty and significance statement This study demonstrates the first application of two-line OH laser-induced fluorescence (LIF) for single-shot gas-phase temperature imaging in the particle-laden environment of the SpraySyn burner used for nanoparticle synthesis. Unlike the previous multi-line LIF approach that provided only time-averaged data, this method resolves instantaneous temperature fields together with OH concentration (number density) distributions. A key innovation is the implementation of a spatially resolved correction for laser beam attenuation using the measured laser light transmission and Beer–Lambert analysis. This correction enables compensation of laser attenuation within the flame and the determination of OH concentration fields simultaneously with instantaneous temperature measurements. Consequently, accurate temperature retrievals and quantitative OH diagnostics become possible, enabling direct comparison between instantaneous and averaged data for CFD validation in turbulent nanoparticle-forming flames.
The effect of a stretching flow field on the behaviors of a single flame ball in a counterflow field with a low-Lewis number premixture was numerically investigated. Steady-state simulations revealed that the flammability limit of flame ball (κ=2.57) is significantly broader than that of a counterflow planar flame (κ=1.02) where κ is a heat loss parameter. Unsteady simulations, subjecting an initial flame ball to various non-dimensional stretch rates ξ, demonstrated highly non-linear transient behaviors, including a steady ball-like flame, an unsteady ring-shaped flame propagating outward, a steady counterflow planar flame, a disk-shaped flame with an expanding hole, and instant blow-off. Crucially, the initial flame ball failed to transition into a planar flame under several stretch rates even though stable planar flame solutions exist. This indicates that the initial flame ball was an insufficient ignition kernel for planar flame formation under certain stretch rate conditions. Furthermore, it was revealed that quasi-steady non-spherical ball-like flames can survive under extremely high heat loss conditions up to κ=6.10, significantly exceeding the flammability limit of flame balls. This extension is attributed to strong local flame enhancement near the stagnation plane due to curvature-induced Lewis number effects. These findings deepen the fundamental understanding of near-limit flame behaviors and provide essential practical implications for establishing robust ignition strategies in lean combustion. Novelty and significance statement This study provides the first comprehensive numerical analysis of the dynamic transition from flame balls to propagating deflagrations in weakly stretched counterflow fields using low-Lewis number premixtures. The novelty lies in elucidating the stretch-rate-dependent transient behaviors and identifying the critical conditions under which a spherical ignition kernel either successfully transforms into a planar flame, forms quasi-steady non-spherical ball-like flames, or extinguishes. Fundamentally, this work advances combustion theory by clarifying the role of diffusive-thermal instability and local flame enhancement mechanisms in near-limit flame survival, offering predictive insight for upcoming microgravity experiments on the International Space Station. Practically, unraveling the dynamic responses of ignition kernels to flow fields provides essential theoretical guidance and practical implications for establishing robust ignition strategies in ultra-lean combustion technologies.
Combustion characteristics of a single droplet of liquid oxidizer continuously heated in a lean fuel atmosphere were experimentally investigated under various ambient fuel concentrations. Approximately 90 wt.% hydrogen peroxide (H2O2) was employed as the liquid oxidizer and suspended in a quiescent nitrogen (N2)–balanced methane (CH4) atmosphere. Combustion experiments were conducted in a range of CH4 concentrations (XCH4, defined as a volume fraction) between 0.00 and 0.10. Burning behaviors, evaporation rate constants (K), and maximum temperatures around the droplet and flame were measured. The effects of XCH4 and extent of the potential contribution of the H2O2 thermal decomposition heat to the combustion characteristics were further discussed together with the Spalding heat transfer number (B). The results identified four distinctive burning modes: a luminous hot flame mode, a faint blue flame mode, a transitional mode between them, and an evaporation mode where only evaporation without ignition occurred. Each mode was clearly distinguished by the maximum temperature measured using a fine thermocouple. A comparison of the experimentally obtained Bex, derived from K measured in this study, with a modified theoretical B (Bth_m), which incorporates the effect of the H2O2 thermal decomposition heat into the classical theory of the droplet combustion, suggested that the faint blue flame and evaporation modes were influenced by relatively large H2O2 thermal decomposition heat. In contrast, under higher XCH4 conditions in the hot flame mode, the heat release from oxidative reaction of fuel and oxygen was found to dominate. Novelty and significance statement This paper presents the first experimental study clarifying the contribution of the H2O2 thermal decomposition heat to its combustion characteristics. Four distinctive burning modes were identified, and the present study characterizes them through measured maximum temperature, evaporation rate constant, and heat transfer number. The key finding is that the heat release by the H2O2 decomposition reaction influences combustion characteristics in the blue flame and evaporation modes, whereas its effect is small compared to the oxidative O2–CH4 reaction in a luminous hot flame mode, where the relatively small heat release from H2O2 is masked. These results provide new insights into the liquid H2O2 combustion behavior in a fuel-lean atmosphere. Given reported industrial accidents involving the H2O2 thermal decomposition heat, in which combustion in a fuel gas atmosphere may trigger secondary catastrophic events, this study offers fundamental data for predicting the potential risk related to H2O2.
Understanding flame-shock interactions in confined environments is important for both fundamental research and practical applications, including the development of high-speed propulsion systems and the prevention of accidental explosions in pipelines and mines. This study examines the interaction between a propagating flame and a shock wave in a slender channel closed at one end and partially open at the other, filled with a premixed hydrogen-methane-air mixture. Ignition at the closed end produces a wall-friction induced accelerating flame that propagates towards the partially open end. A shock wave generated at this end using the wire-explosion technique travels opposite to the direction of flame propagation and interacts with the flame. The interaction induces Richtmyer-Meshkov instability, corrugating the flame surface and momentarily driving the flame towards the closed end. The reflected shock subsequently re-accelerates the corrugated flame towards the open end, producing a turbulent flame whose mean burning velocity is nearly two orders of magnitude greater than the laminar unstretched flame speed. This enhancement is attributed to increased flame distortion caused by turbulence generated through successive interactions with the incident and reflected shocks. Experiments were conducted by varying the strength of the incident shock, which consequently modifies the strength of the reflected shock and the turbulence intensity generated following shock passage. An order-of-magnitude estimate of the Karlovitz number and the reaction Karlovitz number indicates that the observed turbulent flame lies in the thin-reaction-zone regime. Analysis of the turbulent flame further reveals that the normalised turbulent flame speed is inversely correlated with pressure ratio, in addition to its classical square-root dependence on the ratio of turbulent diffusivity to molecular diffusivity.Novelty and significance statement: This study presents the first experimental observations of shock–flame interactions in a confined slender channel (12 mm square cross-section, aspect ratio 1, fineness ratio (length-to-cross-sectional dimension) 21.67) under closed–partially open boundary conditions. Previous studies (Yang and Radulescu, 2021; Yhuel et al., 2026) have largely focused on quasi-two-dimensional configurations or channels with large aspect ratios, where confinement effects differ across cross-sectional dimensions. The flame dynamics observed here differ qualitatively from those reported in these studies. High-speed imaging shows that interaction of the incident shock with the flame induces Richtmyer–Meshkov instability, followed by transition to turbulence after interaction with the reflected shock. The work provides quantitative insights into the resulting flame dynamics and the mechanisms of shock-induced flame acceleration in confined geometries. These findings advance understanding of shock–flame interactions in elongated channels and provide a framework for interpreting their behaviour in practical confined environments, such as pipelines and mines.
Recent battery fire incidents have prompted discussions regarding the safety of lithium-ion batteries. Solid-state lithium-ion batteries (SSLBs) with polymer electrolytes are considered safer alternatives to traditional liquid electrolyte batteries. However, this assumption of the fire safety of SSLBs has recently been challenged. In this study, the combustion behaviour of SSLBs with polymer electrolytes was investigated under different atmospheres (air, N2, perfluorohexanone (C6F12O), bromotrifluoromethane (CF3Br), and heptafluoropropane (CF3CHFCF3)) in an energy storage cabin. The results indicated that after thermal runaway occurred, SSLBs at a low state of charge (SOC) exhibited prolonged smoke release without visible flames, whereas those at a high SOC (≥50%) demonstrated intense jet fires with peak temperatures exceeding 600 °C. Unlike liquid electrolyte batteries, SSLBs can release significant amounts of CO, H2, H2S, and hydrocarbons even at a SOC of 0%. The gas released by an SSLB with a SOC of 0% is more toxic than that released by the SSLB with a SOC of 50%. All three fluorinated suppressants chemically suppressed combustion and reduced the flame duration by 50–70%, but the burning of SSLBs could not be suppressed in a 99% N2 atmosphere. CF3Br and CF3CHFCF3 were most effective, nearly eliminating jet flames and maintaining flame temperatures below 200 °C. However, fluorinated suppressants decompose at high temperatures, resulting in a significant increase in the concentration of toxic CO and the production of HF. Combined with density functional theory calculations, this study elucidated the pyrolysis pathways of suppressants and their correlation with gas generation and toxicity. The findings could guide the development of fire safety strategies and support battery simulation. Novelty and significance statement Recent battery fire incidents have prompted discussions regarding the safety of lithium-ion batteries [[1], [2], [3], [4]]. One potential solution to increase battery safety involves the development of SSLBs, in which volatile and flammable liquid electrolytes have been replaced with solid electrolytes. However, the safety of high-energy-density SSLBs has not been rigorously verified. In this study, the combustion behaviour of SSLBs under different atmospheres (air, N2, C6F12O, CF3Br, and CF3CHFCF3) was investigated. We also evaluated the effectiveness of these suppressants in mitigating SSLB combustion and toxic gas generation, providing new insights for reducing the fire hazard of SSLBs.
The interaction between compressible isotropic turbulence and a planar detonation wave is investigated through direct numerical simulation (DNS) with detailed chemistry. The selected conditions are informed by a regime diagram recently proposed by Breer and Blanquart [“A regime diagram for detonation–turbulence interactions,” Proc. Combust. Inst., 41, 105799 (2025)]. Specifically, a stoichiometric hydrogen/oxygen mixture diluted with 85% argon is selected alongside an upstream turbulence with Mach number Mt=0.189 and Reynolds number Ret=363. The turbulent inflow is generated from a separate DNS of forced homogeneous isotropic turbulence (HIT) with a supersonic mean convective velocity. Planes of turbulent velocity are extracted from a fixed position over time and used as the time-varying Dirichlet boundary condition in the detonation domain. Instantaneous profiles reveal a mostly planar shock and a wrinkled reaction front. The probability density function (PDF) of the local propagation speed has a mean equal to the CJ speed with fluctuations of ±10%. Ensemble averages, computed in time and in the statistically homogeneous transverse directions, highlight a compression of the turbulent length scales in the shock normal direction. The anisotropy generated by shock compression relaxes within 10 convected von Neumann Kolmogorov time scales in agreement with the results of non-reacting shock–turbulence interaction. However, the turbulence remains approximately 15% anisotropic when it encounters the reaction front. The PDF of the turbulent induction length is Gaussian-distributed around the CJ value with fluctuations up to ±20%. The joint PDF of the fuel consumption rate (source term) and water mass fraction (progress variable) is computed in order to examine turbulent detonation structure in chemical space. While the conditional mean of the fuel source term on progress variable closely matches the CJ solution, variations in fuel consumption rate of more than ±30% are observed. Overall, the DNS results are consistent with the predictions of the wrinkled–reaction–zone regime.Novelty and significance statement: The present work represents the first direct numerical simulation (DNS) of detonation–turbulence interaction (DTI) with detailed chemistry that uses a realistic turbulent inflow boundary condition that does not rely on Taylor’s frozen-turbulence approximation. A key novelty of this work is the design of DTI simulations around controlling parameters not previously identified in prior DNS, with unburnt turbulence conditions purposefully selected to probe a specific regime of interaction. This research is significant because it enables a thorough validation of the predicted structure of turbulent detonations in the wrinkled–reaction–zone regime. Furthermore, this study provides a blueprint for designing and analyzing DNS of DTI in a targeted interaction regime.
