Tunnel fires involving multiple fire sources often result in elevated smoke temperatures, significantly endangering the safety of evacuees and rescue teams and potentially compromising tunnel structural integrity. Existing research on dual fire sources in tunnels has largely focused on scenarios where fire sources are positioned along the tunnel centerline. However, real-world fire incidents often involve vehicles located at random transverse positions. Therefore, this study conducted 96 experimental trials in a small-scale tunnel, examining the influence of fire source spacing and transverse positioning on ceiling temperature distribution. The findings reveal that as normalized fire spacing $s/D$s/D increases from 0 to 0.8, the ceiling temperature distribution transitions from a single-peak to a double-peak profile. Notably, the peak temperature in the single-peak configuration is substantially higher than that observed in the double-peak scenario. When fire sources are positioned near the tunnel sidewalls, the sidewalls impede air entrainment, increasing the likelihood of flame merging. Furthermore, thermal feedback from the heated sidewalls amplifies the heat release rate of the fire sources, resulting in a greater maximum ceiling temperature rise. Building on the longitudinal temperature decay model developed by Heskestad and Hamada, this study introduces a new model that adjusts the plume radius based on the free flame height ${H_c}$Hc and the normalized transverse position $L/\left({w/2} \right)$L/w/2 of the fire sources. Finally, a predictive correlation for maximum ceiling temperature rise is formulated and validated using experimental data from this study and prior research.

Inverse estimation of fire parameters is a fundamental aspect of metro fire safety monitoring. When a metro carriage ignites within an interval tunnel, smoke propagates into the tunnel through the carriage's "closed at both ends, multiple side openings" structure, interacting with the tunnel slope to produce a complex and dynamic longitudinal temperature profile. In such highly dynamic and strongly perturbed fire environments, temperature field data obtained solely from conventional temperature sensors or distributed optical fiber sensors are insufficient to fully characterize the evolving fire dynamics. Moreover, local sensor damage or failure can substantially reduce the accuracy of existing inversion methods for estimating fire position, heat release rate, and tunnel slope. At present, there is a lack of dynamic inversion approaches capable of integrating multi-task perception and multi-parameter information, which restricts high-precision estimation of critical fire parameters and limits effective fire situation assessment. In this study, full-scale fire simulations using FDS (Fire Dynamics Simulator) are conducted to establish a comprehensive temperature field database for confined "tunnel-side-opening carriage" fire scenarios under varying slopes, heat release rate, and positions. To overcome the limitations of conventional data augmentation approaches-namely, fixed patterns and limited generalization-this work proposes a fully randomized dynamic data augmentation method tailored for small-sample learning. By introducing random masking along temporal and spatial dimensions and superimposing Gaussian noise, the method effectively simulates complex scenarios involving missing early-stage fire data and random sensor failures. To further enhance the model's adaptability to masked data, a hidden-dimension encoding layer is incorporated into the Transformer, forming a WM-Block structure that enables precise masking at the feature dimension level and strengthens robustness against noise interference. To mitigate the training challenges associated with multi-task learning under data masking, a two-stage training framework combined with early stopping is implemented. The resulting multi-task dynamic inversion model can simultaneously estimate tunnel slope, fire position, and heat release rate, while maintaining strong real-time performance and fault tolerance. Experimental results demonstrate that incorporating random temporal masking, random sensor masking, and Gaussian noise augmentation during both training and inference substantially improves model generalization and stability. Furthermore, the enhanced Transformer block with the WM-Block architecture significantly improves handling of masked data. Under a sensor masking rate of 0.01 (Phid = 0.01) and when only 60 % of the complete temporal data are utilized (Pseq = 0.4), the proposed model achieves a simultaneous inversion accuracy of 0.943 across all three parameters. Ablation studies and comparative analysis with the GRU model further confirm the robustness of the proposed approach. This research provides crucial technical support for fire perception in metro tunnel environments, and offers a reliable basis for fire risk management, early warning, and emergency decision-making in process safety and risk engineering. Furthermore, this method contributes to optimizing ventilation and smoke

In recent years, fire accidents at electrochemical energy storage stations (EESS) have occurred frequently, raising serious concerns about operational safety. However, the characteristics of EESS fire accidents and the current status of fire safety systems have not yet been systematically investigated. Therefore, this study combines statistical analysis and questionnaire survey, systematically identify accident characteristics, clarify causative factors, and assess the current status of fire protection systems. A total of 102 typical global fire incidents (2016–2025) were analyzed across multidimensional analysis, including country, time, battery type, operational phase, and cause. The results indicate that incidents are temporally concentrated, regionally clustered, and most likely to occur during the operation and maintenance phase. Accidents involving ternary lithium-ion batteries remain dominant, while those related to lithium iron phosphate batteries are increasing. Battery faults and system defects are the primary causes, accounting for 21.2% and 54.5%, respectively. Additionally, a field investigation of 18 EESS in China reveals that fire protection systems are typically designed by general contractors based on national standards but often suffer from low investment levels. Approximately 77.8% of the stations use heptafluoropropane as the extinguishing agent, with some located >60 min away from local fire brigades. Moreover, 22.2% have not filed emergency response plans, and 38.9% lack dedicated or part-time fire rescue teams, exposing weaknesses in emergency response mechanisms.

The flammable and explosive gases released during the thermal runaway (TR) of lithium-ion batteries (LIBs) pose a significant explosion hazard in confined spaces, where venting is often the last line of defense. In this study, the effects of vent shape (circular and square) and vent area on vented jet behavior and pressure-wave characteristics during explosion venting of LIB TR gases were experimentally investigated. An 80 L confined chamber was employed, in which in-situ gases generated from a 26700-type LiFePO4 battery undergoing TR were ignited to induce vented explosions. The evolution of vented jets, internal overpressure, and external shock-wave responses was systematically analyzed. The results show that increasing vent area significantly reduces the maximum internal overpressure, with reductions exceeding 50% as the vent area increased from 50 cm2 to 300 cm2, while the influence of vent shape on internal overpressure was limited. Under identical vent areas, circular vents generated more concentrated vented jets with higher axial front-edge velocities, whereas square vents promoted stronger lateral expansion and faster momentum dissipation. Consequently, square vents achieved superior external overpressure mitigation, with peak pressure reductions of up to 23% compared with circular vents at the same area. In addition, pressure decay analysis revealed that larger vent areas accelerated shock-wave attenuation, and square vents exhibited higher decay rates (1 tau), indicating faster pressure dissipation and shorter effective action durations. These findings provide quantitative experimental evidence that vent shape plays a critical role in controlling external explosion hazards, and that square vents offer improved near-field safety performance for LIB TR venting. The results offer practical guidance for explosion vent design in confined LIB systems and energy storage applications.

Urban rail transit has been widely adopted to alleviate surface traffic congestion. However, subway fire safety remains a critical concern because of the confined carriage space, high occupant density, and complex ventilation conditions. Most previous studies have focused on fires located at a single or central position, whereas actual fire events may originate from various locations within a carriage. In this study, reduced-scale experiments and full-scale simulations were conducted to investigate the effects of fire-source location and heat release rate (HRR) on flame behavior and temperature distribution in subway carriages. The results demonstrate that flame inclination and plume structures are strongly governed by the fire-source location due to asymmetric air entrainment induced by side door openings and wall boundaries. Quantitatively, for end carriage scenarios (A1–A4), only 25% of fire locations (A1) exhibit a symmetric flame structure. For middle carriage scenarios (B1–B6), only 16.7% of fire locations (B5) display a symmetric flame structure. In end carriages, door-facing flames tend to deflect toward the end wall, resulting in localized heat accumulation. Fires near the end wall produce higher maximum temperatures than those near the side wall, indicating that stronger confinement enhances local heat accumulation and increases the peak ceiling temperature. As HRR increases, the temperature difference between the two regions gradually decreases. Based on the experimental and simulation results, predictive models for the maximum ceiling temperature rise were developed. Moreover, fire-source location has little effect on longitudinal temperature attenuation but significantly affects transverse and near-source vertical temperature distributions. These findings provide guidance for temperature prediction, fire detection, and fire protection design in subway carriages.

This study investigates the hydrophobic modification of silica aerogels using trimethylchlorosilane (TMCS) and characterises their morphology, wettability, and surface properties through SEM, FTIR, and contact angle measurements. Additionally, the foam stability, fire-extinguishing efficacy, and burn-back resistance of three foam formulations were analysed to investigate the impact of silica aerogel particles on the performance of fluorine-free firefighting foams. The results demonstrated that TMCS effectively modified the surface of the aerogel particles, imparting hydrophobicity. SEM analysis revealed the irregular polyhedral structure of the hydrophobic aerogel. FTIR analysis confirmed an enhanced C-H vibration peak in the hydrophobic aerogels and a higher hydroxyl content in the hydrophilic ones. Fluorine-free foams with added aerogel particles exhibited improved drainage times and outstanding fire-extinguishing performance. The hydrophobic particles increased the drainage time by 55 s and achieved a cooling rate of 35.27 degrees C/s, with a burn-back resistance time extended to 1123 s, 1.64 times that of the control. Similarly, the addition of hydrophilic aerogels increased the burn-back resistance time by 41 %. Furthermore, to deliver an in-depth atomistic description of the interactions between aerogel particles and foam components, molecular dynamics (MD) simulations using the ReaxFF force field were conducted. ReaxFF-MD revealed the key molecular-level interactions that influence foam stability and fire extinguishing, offering insights into the synergistic effects between aerogel particles and the foam matrix under high-temperature conditions.

The smoke mass flow rate (SMFR) of the ceiling jet is an important parameter in the research of smoke flow characteristics and exhaust efficiency, but it has always been difficult to accurately quantify it in practice due to the presence of the air entrainment. To this end, the key factors affecting the ceiling jet entrainment were determined through theoretical analysis, and the relationships were deduced in the current study. Subsequently, a series of scaled model experiments and full-scale numerical simulations were implemented to investigate the effects of the factors such as heat release rate (HRR) and tunnel cross-section size on SMFR and to determine the prediction model for the SMFR of the ceiling jet. The results confirmed that the air entrainment of the ceiling jet constitutes a non-negligible portion of the smoke mass flow. It is jointly determined by the tunnel cross-sectional dimensions and HRR, ultimately resulting in the smoke volume in the one-dimensional spread region being 1 similar to 3.5 times that of the plume. This ratio exhibits a positive linear correlation with the tunnel's aspect ratio. Furthermore, for the ceiling jet caused by weak plumes and strong plumes, these factors exhibit different patterns of the influence on the air entrainment. Predictions were compared with experimental results and previous models and good agreement was achieved. In addition, based on Alpert's assumption about ceiling jet, the spatially averaged entrainment coefficient is quantified by fitting the numerical simulation results to reflect the entrainment intensity of the ceiling jet. This work can be used to estimate the SMFR in tunnel fire and provide some guidance for the smoke control.

Objective Forest fires play a crucial role in the replacement of plant communities. However, climate change may significantly affect tree regeneration after severe wildfires, transform ecosystems and cause huge economic losses. Pneumatic fire extinguishers have become the primary portable firefighting equipment in China's mountainous and roadless regions due to their high mobility and simple operation. However, prolonged operation often leads to engine overheating, which reduces rotational speed, resulting in lower jet velocity, increased air outlet temperatures, and decreased firefighting efficiency. Extensive studies examine the interactions between air flow and gas-phase combustion, but mainly focus on the effects of ambient wind on pool fire combustion. The understanding of their dynamic mechanisms during engine overheating, which causes reduced air jet velocity and elevated air flow temperature, is still poorly understood. To address these gaps, this study numerically investigates the suppression effect of high-speed air jets from pneumatic fire extinguishers on gas-phase combustion, aiming to provide methodological references and data support for optimizing pneumatic fire extinguisher design and improving pneumatic firefighting strategies. Methods This study employed the scale-adaptive simulation (SAS) turbulence model and eddy dissipation model to simulate the gas-phase combustion of n-heptane pool fires under the influence of air jets. The SAS turbulence model, based on the modifications of the k-L turbulence equation, blends the advantages of the Reynolds-averaged Navier-Stokes (RANS) and large-eddy simulation (LES) models. It can dynamically adjust its turbulence length scale to balance the modeling and resolution of turbulence stress transport. In this study, the SAS turbulence model was used to simulate n-heptane pool fires in air jets. The oil pan height was increased by 100 mm to minimize the gap between the simulation and experimental results. The simulation of high-speed jet interaction with gas-phase combustion involves two methods: computational domain coupling of the pneumatic fire extinguisher and n-heptane combustion for data transfer, and boundary condition transfer from the fan outlet to serve as the jet inlet conditions in the n-heptane combustion computational domain. For computational efficiency, the second method was chosen. The stable n-heptane gas-phase combustion simulation results were used as the initial flow field, and the RANS results at the fan outlet (average velocity of approximately 80 m/s) were extracted and implanted into the interaction computational domain inlet. Results The height increases of the oil pan improved model accuracy, keeping the SAS temperature error within an acceptable range. Results showed that high-speed airflow effectively suppressed gas-phase combustion. As the jet velocity increased, the flame shape underwent remarkable changes. At low velocities, the flame maintained a stable and continuous structure within a concentrated combustion area. However, as the velocity of the jet exceeded a certain threshold, it strongly impacted and disrupted the flame, which was stretched, distorted and ultimately extinguished. Furthermore, within the studied range, changes in the jet temperature had minimal impact on the gas-phase combustion of n-heptane. Regardless of the temperature, the flame structure and combustion efficiency remained similar at the same jet velocity. This is likely because physical dilution of the jet masks the sensitivity of the gas-phase combustion rate to the temperature. When the jet velocity dominates, minor temperature fluctuations cannot significantly alter the combustion process. Conclusions This study presents a simulation approach for analyzing gas-phase combustion suppression of n-heptane pool fires using high-speed airflow. The key findings are as follows: (1) jet velocity is the dominant factor in flame suppression. As the velocity increases, the kinetic energy of a jet increases, effectively diluting combustible gases and disturbing the flow field. This hinders fuel-oxidizer mixing and leads to flame breakup and suppression. However, increased jet velocities may cause fuel splashing in actual firefighting, so an optimal jet velocity range needs to be determined. (2) Within the studied range, the jet temperature exerts minimal impact on combustion. At the same velocity, jets with different temperatures produce similar flame structures and efficiencies, thus indicating that temperature regulation is not critical for suppression under these conditions. (3) The SAS turbulence model is effective for simulating air jet-gas-phase combustion interactions as it balances computational accuracy and cost, outperforming the RANS and LES models. Thus, it is suitable for further firefighting simulation studies. Overall, this study provides simulation methods and data references for optimizing pneumatic fire extinguishers. Future studies should focus more on combustion suppression effects under complex conditions and refine simulation methods to better suit actual firefighting scenarios.

During a fire in a naturally ventilated tunnel with multiple vertical shafts, some shafts may induce supplementary airflow, leading to destabilization the smoke layer and potentially causing smoke to ingress back into the tunnel interior, thereby compromising the safety of trapped occupants. This study investigates the dynamics of smoke propagation in tunnel fires under natural ventilation conditions in multiple shafts using numerical simulations. Various auxiliary measures to mitigate the destabilization of the smoke layer in the tunnel as well as to suppress the smoke propagation are proposed and evaluated. The research findings indicate that installing auxiliary fans inside the shafts or incorporating coupling plates beneath them can effectively attenuate the extent of smoke ingress to varying degrees. While these measures have a small effect on smoke removal efficiency, the auxiliary fans mitigate the collision between smoke flow and supplementary air by regulating the flow field at the bottom of the shaft, maintaining a stable smoke layer flow state and suppressing smoke backflow. Coupling plates increase resistance to horizontal smoke spread. This constrains smoke spread and prevents smoke flow from colliding head-on with supplementary air, which could cause backflow. In summary, this research enhances understanding of the intricate dynamics of smoke propagation in tunnels with multiple shafts during fires and offers insights to refine the design of natural ventilation systems in tunnels.

The smoke propagation process in inclined tunnels during subway carriage fires can be regarded as a complex dynamical system undergoing continuous dynamic evolution. As such systems approach the critical threshold for state transitions, critical slowing down frequently occurs. This manifests as a significant reduction in the recovery rate following system disturbances, accompanied by enhanced autocorrelation in state fluctuations. This phenomenon provides crucial early warning information for critical transitions in the system. To explore the dynamics of smoke propagation and identify early-warning indicators of system instability in subway carriage fires within inclined tunnels, a series of full-scale fire scenarios were simulated using the Fire Dynamics Simulator (FDS). The study investigates the influence of key parameters, including tunnel slope, heat release rates (HRRs), and the presence of carriage obstructions, on smoke movement and temperature distribution. Leveraging the theoretical framework of critical slowing down, the analysis focuses on the temporal evolution and mass flow rates of smoke to detect precursors to stability transitions in tunnel environments. The results indicate that the increased tunnel slope reduces the extent of smoke backflow and leads to more uneven smoke distribution near the fire source. At lower slopes, ceiling temperatures along the tunnel exhibit a distinct “double-peak” pattern, which converges into a single peak with increasing slope. Carriage obstructions significantly alter local thermal environments, producing rapid temperature drops in blocked areas, while unblocked zones show more gradual thermal decay. Moreover, variations in window configuration and lag distances have a pronounced impact on the autocorrelation and variance of smoke mass flow rates, serving as sensitive indicators of approaching phase transitions. The transition from stable to unstable smoke behavior is characterized by critical slowing down, reflected in heightened variance and temporal autocorrelation. These findings offer important implications for the early detection of hazardous conditions and the optimization of fire safety strategies in underground transportation systems.

Fiber-reinforced magnesium oxysulfate (MOS) cement boards offer considerable potential for fire-resistant construction owing to their low density, eco-friendliness, and high thermal stability. This study investigates the fire resistance and thermal insulation performance of MOS boards with varying thicknesses (3, 5, and 8 mm) under different heating intensities. Thermogravimetric analysis (TGA) identified a three-stage decomposition process: free water evaporation (<100 degrees C), dehydration of the 517 phase (100-350 degrees C), and brucite decomposition (>350 degrees C). Fire tests revealed that the unexposed surface experiences distinct thermal response stages-an initial temperature plateau followed by gradual heating-both governed by decomposition kinetics. Increasing board thickness directly correlated with extended durations of these stages. Conversely, higher heating rates accelerated decomposition reactions, thus shortening the gradual heating phase and diminishing overall fire resistance. Notably, all boards maintained structural integrity for over 1.5 h under the ISO-834 heating curve, with SiO2 fiber reinforcement significantly enhancing thermal stability and preventing thermal stress-induced failure. A thermal conduction model confirmed that MOS boards exhibit stable thermal conductivity (similar to 0.17 W/ m center dot K) within the 400 degrees C to 1000 degrees C range. These findings demonstrate that the sequential endothermic decomposition stages underpin the thermal insulation mechanism and underscore the critical role of thickness optimization in enhancing fire resistance. This work provides a theoretical and engineering foundation for integrating MOS boards into fireproof building systems.

Vertical shaft smoke exhaust is a vital method for controlling smoke during tunnel fires. To address the limited efficiency of traditional natural smoke exhaust in shafts, previous studies have extensively explored ways to enhance smoke exhaust efficiency by modifying the tunnel structure to mitigate the suction-through effect and boundary layer separation phenomena. However, there has been relatively little previous research on how to efficiently and simply enhance the smoke exhaust efficiency of shafts in already constructed tunnels. This study proposes an improvement by installing auxiliary fan within the shaft, analyzes changes in airflow organization and smoke flow characteristics inside the shaft after fan installation, using numerical simulations to compare smoke exhaust performance before and after the modification. Impacts of varying fan installation heights, wind speeds, and longitudinal positions on natural smoke exhaust efficiency of the shaft are analyzed. Results indicate that auxiliary fan deployed at the shaft inlet can effectively enhance natural smoke exhaust efficiency. And the improvement in exhaust efficiency is influenced by fan arrangement parameters: Efficiency increases with wind speed, though changes in wind speed below 4 m/s exhibit no significant effect on shaft exhaust efficiency; when wind speed exceeds 4 M/S, exhaust efficiency rises with increasing speed, reaching a maximum at 8 M/S. Exhaust efficiency first increases then decreases with changes in fan installation height, with overall efficiency higher when the fan is positioned outside the shaft than inside. The optimal fan arrangement is identified at a height of 1 M from the inlet on the Shaft's central axis with a wind speed of 8 M/S, with mid-axis placement demonstrating the best performance. Meanwhile, the modified froude number is used to quantify the driving force variation in natural smoke exhaust after auxiliary fan installation, providing positive implications for tunnel smoke control and natural smoke exhaust design of vertical shafts

The transport of heat and smoke in tunnel fires is governed by complex interactions among the fire plume, longitudinal ventilation, and buoyancy-driven shaft exhaust. Within the heated region, thermal expansion and aerodynamic resistance reduce the effective longitudinal airflow, extending the smoke back-layering length. This study quantifies this fire-induced throttling effect under hybrid ventilation conditions. A one-dimensional momentum model is proposed, introducing an equivalent throttling velocity to characterize the reduced effective ventilation. The model was validated through 24 reduced-scale (1:20) tests, with heat release rates (HRRs) of 2.68–26.37 kW, longitudinal velocities of 0.31–0.69 m/s, and shaft-to-fire distances of 1.5 and 2.5 m (model-scale), complemented by full-scale Fire Dynamics Simulator (FDS) simulations. Results show that the equivalent throttling velocity varies roughly linearly with the dimensionless back-layering length. It grows with higher HRRs but weakens as the ventilation velocity rises. A non-monotonic response occurs when the smoke front bypasses the shaft, shifting its role from exhausting smoke to entraining ambient air. In terms of magnitudes, this ratio escalated from 4.7% to 15.8% with increasing HRR, whereas varying the shaft location had only a minor systemic influence. Without the throttling correction, the relative deviation in the predicted back-layering length reached 44.5%; after correction, the reported deviation was approximately 5.8%. Overall, this study provides a robust framework for estimating effective ventilation velocities and optimizing hybrid smoke control in tunnel fires.

The back-layering length is a key parameter for evaluating the effectiveness of smoke control in tunnel fires. To investigate the influence of mobile smoke exhaust equipment (MSEE) layout parameters under different deployment conditions, this study combines 1:20 scale tunnel model experiments with fire dynamics simulator (FDS) to conduct the research. The results indicate that the back-layering length is influenced by multiple factors, including the fan height, the distance to the fire source, and the fan flow rate. It is negatively correlated with the fan height within a certain range. At relatively low fan mass flow rates, positioning the fan away from the fire source helps shorten the back-layering length, and placing the fan outside the tunnel can more effectively suppress smoke back-layering. This is because the jet requires a certain distance to develop; if placed too close, the fan momentum tends to reach local equilibrium with the smoke layer, and the jet effect is limited, failing to fully cover the smoke layer. Placing the fan outside the tunnel allows entrainment of more air, enhancing the jet's effect on the smoke layer and significantly suppressing back-layering. In contrast, under high mass flow rate conditions, the optimal exhaust position gradually shifts toward the interior of the tunnel, as the strong jet can effectively drive airflow within the tunnel and more efficiently suppress the smoke layer. In addition, this study develops a dimensionless model for predicting the back-layering length, which is related to the layout parameters of the MSEE. The predicted results show good agreement with both numerical simulations and experimental data.

High-temperature smoke remains a primary threat in tunnel fire safety. The curved walls of small-radius Urban Traffic Link Tunnels (UTLTs) significantly alter smoke flow patterns and temperature distribution. Furthermore, no quantitative model exists to assess the impact of transverse fire location variation on temperature distribution in small-radius UTLTs. To address this gap, this study integrates experimental and numerical methods to specifically investigate the influence of curvature radius and transverse fire location on ceiling temperature distribution. Key findings demonstrate: (1) Wall-adjacent fires exhibit substantially higher temperatures than non-adjacent scenarios, resulting from restricted air entrainment (increasing flame height) combined with wall thermal constraint effects. (2) Competition between centrifugal and inertial forces consistently produces a higher maximum temperature rise beneath the convex ceiling versus the concave side in curved sections. (3) A novel dimensionless parameter Rcs is derived from smoke control volume force analysis. This parameter quantifies the coupled effect of curvature radius and ventilation velocity on convex-concave ceiling temperature difference, enabling a predictive regression equation. (4) Through dimensional analysis, key governing dimensionless parameters are identified. Incorporating the Richardson number (Ri), which characterizes inertial-to-buoyant force competition, a predictive model for maximum ceiling temperature rise in small-radius UTLTs is ultimately established.

Foam fire suppression is a key technique for preventing and controlling various types of fires, and foam stability is one of the critical factors determining extinguishing efficiency. Traditional burnback tests struggle to provide precise quantitative assessments of foam thermal stability. In this work, custom-built foaming and thermal stability testing devices were employed to investigate the synergistic effects of hydrophobic silica particles (namely, porous aerogel (AR) and nano-silica (NS)) with different structures and sizes, together with polymers, on the thermal stability of foams at high temperatures. Systematic experiments were conducted to evaluate foam height, foam drainage, evaporation behavior, and internal temperature distribution. The results demonstrated that under sustained external radiant heating, all foam systems, irrespective of the inclusion of silica particles or polymers, exhibited an initial volume expansion followed by gradual collapse. Accelerated foam drainage was also observed in the later stages of heating. The porous structure of AR particles improved foam thermal stability more than the destabilizing effect caused by accelerated drainage due to their micron-scale size. Polymers combined with hydrophobic silica particles exhibited a significant synergistic effect, and foams containing porous AR particles showed superior thermal stability and heat-blocking performance compared with those containing an equivalent concentration of NS particles. The stability duration of foams stabilized with AR par-ticles was extended by up to 11.8 % compared with foams stabilized with NS particles, and the time for the upper foam to reach maximum temperature was prolonged by up to 9 %. The addition of polymers and solid particles also reduced the extent of foam mass loss due to evaporation under thermal radiation. These findings provide valuable guidance for the development of high-performance fluorine-free firefighting foams.

Lithium ion batteries (LIBs) pose significant fire and explosion risks due to thermal runaway (TR), and extinguishing such fires remains a major challenge. This study investigates the fire extinguishing efficiency of water, aerogel extinguishing agent, and foam in suppressing the TR of LIB packs, and establishes an evaluation model to assess their performance. The TR process of the battery pack is triggered by localized short-circuiting using a saturated sodium chloride solution of equivalent volume. Subsequently, the fire extinguishing procedure is executed utilizing a spray gun with a fixed flow rate to effectively suppress the fire. Water and aerogel both demonstrate rapid fire suppression capabilities. However, aerogel requires continuous coverage to cooling the elevated temperatures of the battery pack. In contrast, foam exhibits relatively poor fire extinguishing and cooling performance but possesses strong resistance to re-ignition. Water can prevent TR propagation (TRP) to adjacent batteries, whereas aerogel may still allow TRP, and foam shows the weakest TRP suppression effect. By categorizing fire extinguishing indicators into hazard control and prevention, an evaluation model is established based on the upper/lower limits, average value and standard deviation of each indicator. The comprehensive scores for water, aerogel, and foam are 0.69, 0.46, and 0.44, respectively. Water exhibited the optimal fire extinguishing effect, featuring the strongest hazard prevention capacity. Aerogel had a moderate fire extinguishing effect, as it required prolonged continuous spraying to suppress high temperatures. Foam performed worst, residual localized high temperatures tended to trigger secondary disasters. This study provides a novel evaluation framework and reliable strategy for LIB fire extinguishment; future investigation should incorporate potential risks such as electrical conductivity and toxic gas emissions to better reflect practical application scenarios.

Current research on flame retardants in firefighting foams centers on fire suppression efficacy, with relatively limited mechanistic insights into how these agents regulate and optimize critical physicochemical properties such as foam stability. This study investigates the regulatory mechanisms of two flame retardants, ammonium pentaborate (APB) and magnesium hypophosphite (MHP), on the stability of PFAS-free (Per- and polyfluoroalkyl substances) foam, based on a blended system of silicone surfactants, hydrocarbon surfactants, and foam stabilizers. Experimental results reveal that the flame-retardant-regulated, PFAS-free foam system exhibits elevated viscosity and a prolonged drainage time. APB adversely affects the physicochemical properties of the foam, whereas MHP significantly enhances foam stability by strengthening the liquid film and delaying drainage. Further study revealed that within a certain concentration range, the addition of MHP consistently improves the stability of PFAS-free foam, with overall enhancement first increasing and then decreasing. The system achieved optimal stability at 0.2 wt% MHP, corresponding to the best fire suppression performance. This concentration-dependent effect provides an important theoretical basis for the optimal dosage of flame retardants in PFAS-free foams, addressing a critical research gap in the regulation and optimization of foam stability by flame retardants.

To investigate the transport characteristics and discharge patterns of a novel extinguishing agent in the complex pipeline network of lithium-ion battery energy storage power stations, a visualized experimental platform was built in this work. Systematic discharge experiments were conducted under different pipe diameters and initial pressures. Combining real-time data from pressure sensors and high-speed camera recordings, the stage evolution, pressure variation characteristics, and decay behavior were analyzed. The results indicate that the extinguishing agent discharge process consistently exhibits three physically distinct stages: the liquid discharge, the transitional gas-liquid mixing discharge, and the gas release. The differential pressure ΔP1-2 serves as an effective criterion for identifying the transition between stages. Higher initial pressures and larger pipe diameters significantly shorten the duration of the liquid discharge stage, while smaller pipe diameters substantially prolong it due to reduced flow rates. Based on average parameter analysis during the stable discharge stage, a frictional resistance loss model combining the generalized Reynolds number and the Darcy–Weisbach equation is established and agrees well with experimental data, validating its effectiveness in predicting the pressure drop of the non-Newtonian extinguishing agent in horizontal straight pipes. In addition, the gas-release stage follows a three-parameter exponential decay model. Specifically, the decay time constant t1 decreases approximately with D−2, while the release-rate constant k=1/t1 increases approximately with D2. This study reveals the discharge evolution law of the new extinguishing agent under different working conditions, and provides a theoretical basis and data support for the optimal design of fire extinguishing systems in energy storage power stations.

Tunnel fire research typically assumes that the internal fire plume is symmetrically distributed, overlooking the asymmetric effects of changes in the longitudinal fire source position on the flow field structure. To investigate the impact of the longitudinal fire source position on smoke propagation characteristics and smoke back-layering length, a 1:20 scaled tunnel model was constructed. Twenty-four experimental cases were conducted to examine the effects of fire source location, heat release rate, and ventilation velocity on smoke back-layering length, and the results were compared with numerical simulations. The study indicates that the longitudinal fire source position significantly influences the flow field structure within the tunnel by altering the distribution of longitudinal resistance upstream and downstream, as well as the direction of the total induced airflow. As the fire source moves from the inlet toward the outlet, the effect of induced airflow on longitudinal ventilation shifts from synergistic enhancement to reverse blocking, causing a significant reduction in the effective wind speed within the tunnel. Consequently, under identical ventilation conditions, the closer the fire source is to the downstream portal, the longer the smoke back-layering length. Furthermore, a dimensionless prediction model for smoke back-layering length, incorporating the longitudinal fire source position, was developed. The pre-diction results agree well with numerical simulations and experimental data, with relative errors controlled within 620%.