Realistic Fire Scenarios Research Articles

Study on Public Perceptions and Disaster Prevention Framework of Tunnel Fires Based on Social Media and Artificial Intelligence

To investigate public perceptions regarding tunnel fire disasters and optimize the tunnel fire disaster prevention framework, this study takes the emerging social media platform Douyin as a case study, conducting an in-depth analysis of 2133 short videos related to tunnel fires on the platform. A computational communication method was used for analysis, Latent Dirichlet Allocation was used to cluster the discussion topics of these tunnel fire short videos, and a spatiotemporal evolution analysis of the number of videos posted, user comments, and emotional inclinations across different topics was performed. The findings reveal that there is a noticeable divergence in public opinion regarding emergency decision making in tunnel fires, related to the complexity of tunnel fire incidents, ethical dilemmas in tunnel fire escape scenarios, and insufficient knowledge popularization of fire safety practices. The study elucidates the public’s actual needs during tunnel fire incidents, and a dynamic disaster prevention framework for tunnel fires based on social media and artificial intelligence is proposed on this basis to enhance emergency response capabilities. Utilizing short videos on social media, the study constructs a critical target dataset under real tunnel fire scenarios. It proposes a computer vision-based model for identifying critical targets in tunnel fires. This model can accurately and in real-time identify key targets such as fires, smoke, vehicles, emergency exits, and people in real tunnel fire environments, achieving an average detection precision of 77.3%. This research bridges the cognitive differences between the general public and professionally knowledgeable tunnel engineers regarding tunnel fire evacuation, guiding tunnel fire emergency responses and personnel evacuation.

Thermodynamic characterization on the suppression process of water mist under the interaction effect between two buoyant non-premixed flames

Realistic fire disaster scenarios always refer to multiple fires burning at adjacent positions at the same time (MF), leading to a higher risk compared with a single fire (SF). That is, fire suppression in realistic fire disasters differs from existing fire suppression studies focusing on a single flame. It is worthy to extend the fire suppression study to multiple fires. This study investigates the thermodynamic characteristics of applying the water mist on the double flames through FDS (Fire Dynamic Simulator v. 6.9.1). The fire suppression mechanism is discussed from two aspects (the cooling effect and the dynamic effect). The cooling effect of the water mist on the double flames is revealed based on the onset time of cooling. The droplet distribution is considered the main factor in the cooling effect of the water mist on the double flames (DF). The dynamic effect dominates in the two effects of the water mist. It is found that the interaction effect between double flames significantly improves the difficulty of penetrating the combustion area. This study can provide numerical data support and theoretical reference for water mist fire suppression systems applied in realistic scenarios.

Flexural behaviour and shear capacity of RC flat plate sections during fire exposure

PurposeFire safety is a pivotal requirement in building codes. Prescribed design criteria have been the norm to achieve it, which imposes limitations on engineers, including the inability to accommodate new solutions/materials. The shift towards performance-based design offers the potential to address shortcomings of the prescribed design. However, this shift also significantly increases the workload on structural engineers without a corresponding increase in their engineering fees. Simplified design tools are needed to assist engineers in this transition.Design/methodology/approachThe paper is divided into sections investigating equivalent standard fire duration, thermal deformations, flexural behaviour and shear capacity of flat slabs when exposed to fire. The first section conducts a parametric study correlating equivalent and realistic fire durations using the average internal temperature profile (AITP) method, resulting in statistical equations estimating equivalent fire duration. The second section evaluates thermal deformations and flexural behaviour through a parametric study considering various parameters. This section results in statistical equations estimating thermal deformations and flexural behaviour of flat slab sections during fire exposure. The final section focuses on shear capacity, developing simplified heat transfer formulas and statistical equations predicting compression zone depth reduction. The section presents methodologies predicting flat slab sections' one-way and two-way shear capacities during fire exposure.FindingsStructural engineers can use the proposed methods for daily design work without applying complex heat transfer calculations. When the equivalent standard fire duration is utilized, a flat slab’s thermal deformations, flexural behaviour and shear capacity under an actual fire condition can be calculated. As such, the methods would be highly beneficial in assessing the structural integrity of a building during an active fire incident.Originality/valueThe paper provides engineers with the tools required to evaluate the safety of flat slab sections during fire exposure. The methodologies presented in the paper enable engineers to use performance-based design for slab sections by (1) converting any real fire scenario to a standard fire with an equivalent duration, (2) assessing their thermal behaviour, (3) evaluating their flexural behaviour and (4) evaluating their flexural and shear capacities. The paper concludes with a case study example demonstrating the detailed application of the developed methods.

Characteristics of concurrent flame spread over convex charring materials

In real forest fire scenarios, there are a large number of convex and concave terrains at the junction of flat land and slopes. In this paper, a typical charring material, kraft paper, is selected as the experimental sample, and the variation law of key characteristic parameters of flame spread over convex solid surface are investigated. The experimental results show that all the flame spread processes have obvious deceleration behaviors, and the influence mechanism of flame tangent angle on the flame spread deceleration of convex surfaces is revealed. The characteristic length of the pyrolysis zone shows a tendency of increasing and then decreasing with time. The variation curves of radiant heat flux present two peaks. Combining with the characteristic length, the reason for the occurrence of two peaks is explained. The exponential relationship between the instantaneous mass loss rate and the characteristic length of the pyrolysis zone is obtained. Based on the range of power exponents, the heat transfer controlling mechanism in the pyrolysis zone is classified into three categories for all the working conditions. Moreover, a piecewise power-law relationship between the dimensionless maximum flame height and the dimensionless heat release rate is established. The results of this study will fill the research gap of flame spread of solid materials with curved surfaces, and will be of great academic value for the improvement and development of theories on building and forest fire safety.

Residual strength and time to failure of bonded wooden lap joints at high temperatures and superimposed constant loading

ABSTRACT The study focused on the effect of high temperatures on wooden bond lines of single lap tensile shear specimens according to EN 302-1 made from beech wood at superimposed sustained mechanical loading. The quasi-duration of load (DOL) tests encompassed temperature levels of 160–232°C with specimens loaded at 30% of their mean shear strength at ambient temperature for 40 min. The investigations comprised one phenol resorcinol formaldehyde (PRF), two melamine urea formaldehyde (MUF) and three one-component polyurethane (1C-PUR) adhesives as well as solid wood reference specimens. Complementary ramp load tests were performed with specimens heated in unloaded state up to 250°C. The investigations revealed three highly differentiated adhesive DOL behaviours although showing a well comparable pure temperature-bound strength reduction effect at 200°C. PRF and one MUF showed little to no additional influence of applied load, while for two 1C-PUR brands a drastic impact on survival times and residual strength was revealed. One MUF and a special 1C-PUR forwarded results between both extremes. The investigations provide evidence that the assessment of the temperature behaviour of structural adhesives can be biased when relying exclusively on ramp load tests with specimens heated in unloaded state, and thus necessitates a superimposed load as present in realistic fire scenarios.

Machine learning-driven real-time identification of large-space building fires and forecast of temperature development

In firefighting, real-time knowledge of fire dynamics is crucial yet often unavailable. The building geometry and measured gas temperatures are integrated to ascertain fire states and predict temperature evolution, utilizing a machine learning framework that combines long short-term memory networks and transfer learning. The model is pre-trained on datasets with 1000 samples from parametric fire models and 200 from field simulations, enabling real-time predictions with on-site data. Numerical simulations in portal frame buildings demonstrate over 95% accuracy in identifying fire states and 90% in predicting gas temperatures. A method using correlation coefficients and standard deviations effectively detects damaged thermocouples with over 96% accuracy, given a damage ratio below 30%. Validation in two real fire scenarios showed more than 92% accuracy in locating fires and over 89% in forecasting gas temperatures 20 min ahead, which costs 2.14 s and 1.83 s, respectively. The machine learning framework also exhibits robustness to changes in ventilation conditions by making temperature predictions using reliability theory. The proposed framework provides clear information about the state and development of the fire to firefighters and is useful for promoting smart firefighting.

Quantitative assessment of the extent of damage to timber structures during the full course of a real fire using the energy equivalence method

In this study, an innovative method, the Energy Equivalence Method (EEM), is proposed in the context of the Performance-Based Fire Engineering (PBFE) framework for quantitatively assessing the degree of damage of timber structures during the full process of a real fire. To verify the validity of the method, glued laminated timber (GLT) specimens and GLT columns were selected as the subjects of fire action, and a series of equivalent fire experiments were conducted by using a parametric fire model (referred to as the DNI fire model), which is currently used in Germany, to simulate real fire scenarios. By comparing the equivalence of charring depth (CD), pyrolysis zone temperature, and residual load-carrying capacity of the test specimens after the action of the DNI fire and its equivalent standard fire, the validity of the EEM in quantifying the degree of damage to timber members after the DNI fire was fully verified. Subsequently, based on the fire experimental data, accurate thermal and structural models were developed, and these models were utilized to further validate the effectiveness of the EEM in quantifying the extent of damage to timber members under the action of a DNI fire over an arbitrary time period. The results of the study show that the proposed EEM not only can be used to assess the extent of structural damage during fire action and the residual load carrying capacity of the structure after fire action, but also promotes the transition from standard fire safety design to the PBFE design.

Modeling the performance of multilayer insulation in cryogenic tanks undergoing external fire scenarios

Multilayer Insulation (MLI) is frequently used in vacuum conditions for the thermal insulation of cryogenic storage tanks. The severe consequences of the degradation of such materials in engulfing fire scenarios were recently evidenced by several large-scale experimental tests. In the present study, an innovative modelling approach was developed to assess the performance of heat transfer in polyester-based MLI materials for cryogenic applications under fire conditions. A specific layer-by-layer approach was integrated with an apparent kinetic thermal degradation model based on thermogravimetric analysis results. The modeling results provided a realistic simulation of the experimental data obtained by High-Temperature Thermal Vacuum Chamber tests reproducing fire exposure conditions. The model was then applied to assess the behavior of MLI systems for liquid hydrogen tanks in realistic fire scenarios. The results show that in intense fire scenarios degradation occurs rapidly, compromising the thermal insulation performances of the system within a few minutes.

Precision-Boosted Forest Fire Target Detection via Enhanced YOLOv8 Model

Forest fires present a significant challenge to ecosystems, particularly due to factors like tree cover that complicate fire detection tasks. While fire detection technologies, like YOLO, are widely used in forest protection, capturing diverse and complex flame features remains challenging. Therefore, we propose an enhanced YOLOv8 multiscale forest fire detection method. This involves adjusting the network structure and integrating Deformable Convolution and SCConv modules to better adapt to forest fire complexities. Additionally, we introduce the Coordinate Attention mechanism in the Detection module to more effectively capture feature information and enhance model accuracy. We adopt the WIoU v3 loss function and implement a dynamically non-monotonic mechanism to optimize gradient allocation strategies. Our experimental results demonstrate that our model achieves a mAP of 90.02%, approximately 5.9% higher than the baseline YOLOv8 network. This method significantly improves forest fire detection accuracy, reduces False Positive rates, and demonstrates excellent applicability in real forest fire scenarios.

Experimental and numerical investigation into burning rate and fire plume characteristic of pool fire with plate obstacle

Pool fire is a common type of fire accident in the petrochemical industry, which often causes significant casualties and property damage. In most real fire scenarios involving the interaction between pool fire and obstacles, the surrounding obstacles would impact the development of the pool fire significantly, altering the results of fire risk assessment. To address these knowledge gaps, a comprehensive experimental and numerical analysis was conducted to explore the pool fire behavior with plate obstacle. The results show that as the obstacle height decreases or diameter increases gradually, the average burning rate increases obviously. The temperature and flow fields were obtained through numerical simulation, and indentified as three distinct patterns. It was found that the vortices formed near the obstacles facilitate the mixing of air and fuel vapor, thereby changing the fire plume behavior. In addition, the classical model for the plume centerline temperature was modified for pool fire with plate obstacle. This work enhances understanding of pool fire behavior with plate obstacles, which directly contribute to mitigating fire damage and improving energy application safety.

Fuel loads and their composition, and compartment characteristics in educational, office and library buildings

AbstractThis paper presents the fire load data in educational, office and library buildings, obtained through an extensive inventory survey. This data collection effort is prompted by a growing need to simulate compartment fires, wherein estimating realistic fire scenarios is essential to assess the level of fire severity in a structure, and consequently the strength of the various structural members at elevated temperatures. The attributes of compartment fires primarily depend upon the fuel load and its composition, compartment dimensions, ventilation characteristics, and construction materials. Despite an acute need, fire load data across the world is scarce and outdated, and does not reflect the change in the type of materials in‐use today. The survey data presented in this paper is collected from 108 rooms in 10 educational buildings, 51 rooms in three office buildings, and 13 rooms in a library building. This paper also presents the composition of fire loads and the levels of ventilation in these buildings. The studies show that fire loads can vary significantly depending on the room‐use; thereby basing fire load values solely on the overall category of a building may result in either conservative or unsafe design. This study also finds that certain room types (e.g., computer labs) have significant plastic‐based fuels, indicating that typical modelling recommendations based on cellulosic fuels for heat release rate, combustion heat, etc. may not always be appropriate. The paper finally examines the statistical distribution that best describes the measured values of the fire load densities.

Investigation of the fire mass loss rate in confined and mechanically ventilated enclosures on the basis of a large-scale under-ventilated fire test

This study concerns the under-ventilated combustion regimes of a fire scenario in mechanically ventilated enclosures. Using the theoretical approach of the well-stirred reactor model and the data-base of large-scale pool fire tests, the study analyses the mass loss rate as a function of the environmental conditions. The novelties lie in the analysis of a large set of realistic scenarios involving complex geometries (several compartment connected) and new ventilation configurations and the applicability of the well-stirred reactor model to interpret these scenarios. The variables of interest are the combustion regimes (stationary, transient or with rapid quenching), the duration of the combustion phase and the mass loss rate. The results show a satisfactory prediction of the well-stirred reactor model to interpret the experimental results of complex and realistic fire scenario and the robustness of this model to deal with the under-ventilated regimes. The analysis also highlights the interest of two new parameters rarely used in the literature, the ventilation factor and the mass factor, to characterize under-ventilated scenarios. The analysis points out the importance of the relationship expressing the burning rate as a function of environmental conditions (oxygen level and external flux) and extinction conditions.

Development of two-layer zone model BRI2-CRIEPI applicable to fire PRA for nuclear power plants

In implementing practical fire Probabilistic Risk Assessment (PRA) for nuclear power plants, a practical fire model is required for more realistic fire scenarios. The fire simulation zone model BRI2-CRIEPI has been continuously improved to allow analysis considering the characteristics of compartment fires under mechanical ventilation conditions.In this study, the specific characteristics of the fire model under mechanical ventilation conditions to estimate inner pressure and flow rate fluctuation were newly applied, such as the relationship between compartment pressure and flow rate of a ventilating fan, i.e., the fan curve, the vitiation effect of the mass loss rate (MLR) due to the depletion of oxygen owing to fire progress, the representative oxygen concentration corresponding to the elevation of a fire source, and a thermal radiation feedback model from the flame and surrounding environment in case of flame impingement beneath the ceiling.The analytical results with the BRI2-CRIEPI were validated by full-scale compartment fire test results performed by the international Organisation for Economic Co-operation and Development (OECD) Nuclear Energy Agency (NEA) project PRISME3 (Propagation d'un incendie pour des scénarios multi-locaux élémentaires in French). Especially, validation result for HGL temperature, which is very important metric in the fire PRA, shows that there is no obvious bias (the model bias factor of 1.03) and no particular systematic trend (the model relative standard deviation of 0.14) in the model prediction.

Burnout resistance of concrete‐filled steel tubular (CFST) columns under realistic fire conditions

AbstractIn this paper, we investigate the structural behaviour of concrete‐filled steel tubular (CFST) columns under various fires. Our contribution is based on two fire scenarios in modelled FDS with subsequent 3D FE numerical analysis (Abaqus), where material properties are defined for heating and cooling using user subroutines. Our goal is to examine the burnout resistance of CFST columns under different realistic fire scenarios involving varying number of burning cars.The loss of compressive strength of concrete due to high temperature is irreversible. Hence, the burnout resistance is affected by the field of the highest temperatures reached inside the column. Moreover, the pre‐existing load affects the deformation of the column during heating and influences the post‐fire behaviour. The analysis is performed in several steps: first, results from the CFD models are extracted and appropriately mapped, then the heat transfer analysis between fire and solid is performed, and the resulting temperature history of the CFST column is calculated. Subsequently, a mechanical analysis is performed, where (1) the column is loaded, (2) the analysis of the column's structural fire behaviour is carried out, (3) if the column sur‐vives burnout, then in the last step, it is loaded until failure.

BACSSOC: A novel clustering method for mobile forest protection using wireless sensor network with lower energy consumption and lower latency

The utilization of Mobile Forest Protection Using Wireless Sensor Network (Mobile-FPWSN) has exhibited notable advantages in mitigating economic losses in forestry and enhancing the efficiency of fire prevention, thereby leading to its widespread adoption in the forestry protection industry. However, in the current developmental stage of Mobile-FPWSN, the impact of sensor nodes on the system is often disregarded in forest fire prevention and control. Latest studies have indicated that clustering techniques exhibit noteworthy benefits, including a decrease in energy consumption and an extension of system life. However, current clustering methodologies do not comprehensively address the concerns of latency, energy consumption, and delay while operating under Mobile-FPWSN conditions. In order to tackle these challenges, a novel clustering method called Boltzmann adaptive chaotic salp swarm optimization clustering (BACSSOC) has been proposed for Mobile-FPWSN. This method aims to significantly prolong the system’s lifetime, efficiently reduce energy consumption, and minimize system latency to the greatest extent possible. BACSSOC dramatically reduces the energy consumption of cluster heads during the adaptive clustering phase of Mobile-FPWSN, thereby decelerating the depletion of node energy and improving the forest fire monitoring time in comparison with other state-of-the-art algorithms. To evaluate the efficacy of BACSSOC in comparison to ML-AEFA, PCFMO, and SOA-EACR algorithms in mobile FPWSNs, we have conducted a unique forest fire simulation experiment based on a realistic forest fire scenario. The results have unequivocally demonstrated that BACSSOC surpasses ML-AEFA, PCFMO, and SOA-EACR algorithms, exhibiting advantages of at least 15% in reducing node energy consumption, 7% in decreasing system latency, and 9% in increasing system lifespan. These findings highlight the remarkable potential of BACSSOC as an effective network clustering solution in mobile FPWSNs.

An open-source software framework for the integrated simulation of structures in fire

The traditional methods to understand the development of elevated temperature in a structure, and also the associated structural response, are not representative of realistic fire scenarios. To provide a more accurate and realistic reflection of the fire development, the current paper develops a generic middleware which interfaces between the computational fluid dynamics (CFD) software Fire Dynamics Simulator (FDS) and the finite element (FE) analysis software OpenSees. This framework enables a fully integrated simulation of a realistic fire scenario including the heat transfer through the structure and the resulting thermo-mechanical response. The proposed framework is open-source and freely available and therefore can be used and further developed by researchers and practicing engineers and customised to their requirements. This paper shows validation against two sets of experimental results and one real fire incident. A number of different types of thermal boundary conditions such as gas temperatures and heat fluxes, are obtained from the CFD analysis and are then used in the subsequent heat transfer and thermo-mechanical analysis. The primary advantage of this computational tool is that it provides consultants and designers with the means to undertake large-scale projects requiring performance-based fire engineering solutions.

The collapse mechanism of steel framed structure exposed to parametric design fire and travelling fire scenarios

The structural steel elements exposed to fire is susceptible to collapse owing to the temperature sensitivity of steel sections. The technical advances in the field of fire dynamics have enabled to simulate realistic fire scenarios to determine the performance of structural elements. However, a detailed parametric study is necessary to increase the knowledge on the performance of structural steel elements to various design fires. In this study, the behaviour of three-dimensional steel framed structures subjected to the action of standard fire and parametric fire as well as the travelling fire scenarios are analysed. In this work, the general-purpose finite element software ABAQUS is used to perform the analytical study. The material models calibrated with experiments were adopted for conducting parametric study. The results of the analytical study show, the uniform exposure of temperature to compartment is the most critical scenario as the compartments exposed to uniform fire is causing total collapse of structure. Uniform exposure of parametric fire to compartments is found to be less critical as the cooling phase of fire enabling more resistance against total collapse of the structure. Also, the uniform exposure of standard fire, a small resistance is observed from the unexposed columns. However, the structure exposed to parametric fire was sustained large deflections. The structure exposed to travelling fire is found to receive greater resistance before collapse. However, the structure found to collapse as the fire reaches final compartment.

Common OECD/NEA FIRE and PRISME Cable Benchmark Exercise

This paper deals with the common OECD/NEA FIRE and PRISME Cable Benchmark Exercise conducted for a realistic cable fire scenario in an electrical system of a nuclear power plant. The major goal was to simulate a real cable fire event in order to assess the behaviour of fire models for a complex and real fire scenario with the current state of knowledge. Since a numerical Benchmark Exercise on a real fire event is quite challenging, a three-step methodology was adopted: (1) a calibration phase of the cable fire models, (2) a blind simulation of a cable fire test, and (3) the real cable fire event simulation. The selected fire event from the FIRE Database occurred in 2014 in a heater bay of a nuclear power plant. Thirteen institutions from nine NEA member countries participated in this Benchmark Exercise. Simulations were performed using computational fluid dynamics codes, a lumped parameter code, and zone codes. The simulations have provided a wide range of results which reflect the difficulty for fire simulation codes to correctly predict the development of the fire heat release rate over time of a cable tray fire, even under simple, open atmosphere conditions.

Mechanical properties of CFRP bar with different elevated-temperature experiences

Material properties of carbon fiber reinforced polymer (CFRP) at elevated temperatures are of great importance to promote its application in civil engineering. Previous literatures focused mainly on steady-state temperature tests, neglecting the thermal experiences that CFRP may undergo in real fire scenarios. To address this gap, it conducted four types of tensile tests on CFRP bars to simulate different elevated-temperature experiences ranging from 25 to 300 °C. This study investigated properties of CFRP under tensile tests at room temperature and steady-state tests, as well as properties at elevated temperatures with sustained load and after exposure to elevated temperatures. The tests showed that CFRP bars without sustained load remain about 85%, 75%, and 52% of their original strength at 100 °C, 150 °C and 300 °C, after 0.5 h thermal experience, respectively. In addition, the ultimate stress decreased by 11% when heating duration increased from 0.5 h to an hour and thereafter remained stable at 300 °C. Meanwhile, the CFRP bars with sustained load (stress level of 0.3) at 300 °C presented a 7% reduction of tensile strength compared to those without sustained load. For the CFRP bars exposed to 300 °C for an hour and cooled naturally to room temperature, the tensile strength and modulus decreased by 17% and 5%, respectively, but remained higher than those tested at 300 °C due to the solidification of softened resin. This investigation provides insights into the behavior at elevated temperature of CFRP.

Simulation Study on Suppressing Shielded Fires by Water Mist Systems

This article presents a numerical analysis of the performance of three water mist fire suppression systems, with different characteristics, on shielded fires as representing more realistic fire scenarios in an enclosure. A diesel pool fire with a peak heat release rate (HRR) value of 75 kW is covered by an obstacle above it with different shielding conditions to investigate the influence of the obstacle size and the vertical distance between the obstacle and the nozzle on the efficiency of the water mist systems. The obstacle prevents a large number of droplets from directly reaching the fuel surface and flames. The modeling and numerical analysis of this study were carried out by the fire dynamics simulator (FDS) programming tool, and the designed model was validated against the experimental data for both dry and wet tests. The results show that two of the analyzed mist sprays could successfully extinguish the shielded fires in a short time with an obstacle size of 25 cm × 25 cm and 50 cm × 50 cm, placed at two heights. However, the third mist system had a longer extinguishing time compared to the first two nozzles. All three nozzles failed to suppress the fire covered by the largest obstacle (1 m × 1 m). The shielding conditions and nozzle characteristics can affect the performance of water mist systems to some extent.