This study compares the effects on a smoke layer of water sprays injected downward, upward or according to an inclined counter-flow configuration. The impact is analyzed considering stratification, mixing and cooling effects upstream (fire side) and downstream (opening side) the position of the spray. The experiments were conducted in a 1/5th scale model reproducing a room connected to a corridor. The injection of the poly-dispersed spray was carried out in the corridor where a layer of smoke was flowing in the upper part. Thanks to the experimental configuration, there is no direct impact of the spray on the fire source and the production of smoke, but only on the hot flow of smoke. The effect of the spray was evaluated for the different directions of injection and two water feeding pressures. The measurements have shown that effective cooling of the upper layer is observed downstream of the spray. The efficiency of the cooling is dependent on the injection angle. A more or less significant heating of the lower layer is measured upstream for all the injection angles. The injection angle has an influence on the smoke mixing and cooling, an upward spray injection—either vertical or inclined—being more impactful. The strongest interaction is observed for an inclined counter-flow injection, similar to the configuration of firefighters cooling a smoke layer while moving forward in a corridor toward a fire source. Moreover, two water injection pressures were investigated: 4 and 8 bars. Increasing this pressure reduces the droplet diameter and increases the water flow rate. In the present experimental configuration, modifying the water injection pressure showed an effect, yet limited because the droplet size distribution was not strongly impacted. All experimental data are available in an open-access database for further uses.

For many engineering and fire investigation applications, understanding the behaviour of materials exposed to fire often relies on fairly simple measurements (e.g. char thickness) to deduce the quantities of interest (e.g. fire size or duration) by trying to account for a number of complex phenomena (e.g. influence of wood species, dimensions or moisture content). The aim of the work presented here is to analyse the charring process as a function of different experimental parameters, mainly a varying heat flux. The robustness lies in the large number of cone calorimeter experiments, as the char characteristics of 783 samples were measured. The results show promising correlations between the char depth (combination of missing and remaining char thicknesses) and the cumulative exposure, which represents the energy received by the sample, for a single or a double exposure and independently of the other experimental conditions.

Char depth can be a useful measurement to help fire investigators in their work, but the conclusions that can be drawn from it are still debated. While its use to highlight fire spread patterns is usually recognised, there is no consensus on its further interpretation to determine heat exposure or fire duration. The present paper aims at investigating the relationship between char depth and cumulative exposure (heat flux density integral over the exposure duration, in [MJ/m2]). A cone calorimeter was used to expose samples of four wood species (oak, beech, fir, and particleboard) to controlled heat fluxes for varying durations. This proved that, for a given cumulative exposure, the measured char depths are comparable, even if this cumulative exposure is obtained in very different ways. This quantitative analysis (e.g. predicting the cumulative exposure based on a char depth measurement) is an important addition to the typical qualitative use of char depth measurements (i.e. the estimation of fire spread direction). Moreover, the layer of wood that might have burned away, which is often difficult to estimate in the field, could be easily related to the char depth, making this type of measurement even more useable by fire investigators.

The aim of this study is to investigate the radiative protection provided by various fire hose nozzles used by several Fire Rescue Services in France and to propose an experimental set-up to quantify it. This study combined the use of radiative sources (a radiant panel or a fire inside a standard shipping container) and metrological devices (radiative heat sensors, IR camera and spectrometer) to estimate the radiative attenuation of water sprays used to protect the firefighters against thermal effects which occur during a fire. For all the fire hose nozzles tested in this work, the maximal effective attenuation reaches 75%. For most of them, an increase of the flow rate improves the radiative attenuation. However, this study shows that a similar attenuation can be reached for different flow rates, suggesting that the droplet size diameter and the droplet volume fraction also play a significant role in the efficiency of the provided spray.

This article aims at assessing the thermal action received by an unprotected steel structure during the combustion of alternative fuel vehicles. Five tests were realised in the underground of an airport hangar. To obtain a confinement effect and increase the probability of observing a thermal runaway of the traction battery, a second car was parked next to the alternative fuel vehicle. Five motorisations were tested: diesel fuel, H-2 fuel cell, natural gas, electric, and liquefied petroleum gas. The diesel fuel car test served as reference case. The cars were located under a steel structure, representative of a car park, next to a wall. The wall served here to increase the confinement effect. Thirty-nine sections of the structure were instrumented for a total of 200 thermocouples. An extensive analysis including fire behaviour for each test and comparisons between the tests is presented here.

A detailed understanding of flame radiation is useful in numerous fire applications, and can be improved through experiments and modelling: the work presented here is a contribution to both aspects. Data were gathered for kerosene pool fires (30?250 cm wide), in order to characterise the flame geometry and emission. It was shown that the average visible flame can be well approximated by a conical shape. Radiation of soot in the flame was described with an equivalent absorption coefficient and temperature. The results appear to be only weakly dependent on the fire size, which is a valuable information for upscaling approaches. Experimental data were used to compute heat fluxes received by targets outside the flame with four models, based on different assumptions regarding radiation absorption in the flame. Numerical results were compared to experimental ones in order to assess the various approaches and determine their applicability domain. The two most detailed models give good results for all fire sizes, thus showing that the key parameters of flame radiation have been identified and taken into account. The simplified models are less satisfactory, and seem to be applicable only in specific conditions which are rarely met in the case of fires.

This paper focuses on the calculation of the heat release rate from two different fires: 1) a load composed of building materials and 2) a light commercial vehicle. A methodology is developed which uses Computational Fluid Dynamic (CFD) modeling, machine learning method and three radiative heat flux measurements to determine the temporal evolution of the heat release rate. Once this heat release rate is rebuilt, the method provides by numerical simulation many other physical quantities.

We present an equivalent medium model that considers emission from the flame volume to compute spectral radiation intensities. This is considered a more detailed description than common radiation models used in fire safety applications, which often treat the flame as a grey emitting surface. Given its importance in fire radiation, the present study focuses on emission by soot. The required parameters for the model are an equivalent absorption coefficient, and an equivalent temperature. They were determined using two independent measurements, namely multispectral opacimetry, and infrared spectrometry. Results are presented for kerosene pool fires (widths from 30 cm to 250 cm). It is shown that the equivalent temperature can be considered independent of flame size, and thus an intrinsic parameter of the fuel. These results also make clear that radiation emitted from the flame is not grey, and can only be considered black for very large flames (i.e. wider than 1.75 m for present tests).

In order to provide experimental data to develop fire spread models, numerous tests were performed for a wide range of fire sizes and fuels. Here we focus on large scale kerosene tests (approx. 0.5 to 6 m2), with an emphasis on radiative flame properties and present a novel multispectral approach composed of an opacimetry setup and a FTIR spectrometer. This allowed to compute emittance values (based on transmittance measurements) and equivalent soot temperatures (by fitting the calculated intensities to the measured ones). These results, as well as mass loss rates and flame heights, show good repeatability and agree well with the literature, including with correlations predicting the burning rate, flame height and spectral dependence of the extinction coefficient. Generally it is shown that radiation emitted by the flame is not gray, and can only be considered black for very large flames (i.e. wider than 2.5 m for these tests). Additional data (smaller fires, other fuels, complementary parameters - e.g. flame shape) are still to be analyzed and used to model flame heat radiation.

This paper presents a complete methodology for the assessment and modelling of the flammability and fire resistance of carbon fibre (CF) reinforced thermosets (three different types of epoxy) and a thermoplastic resin (PEEK) used for the fuselage of modern aircrafts. A global ranking of the composites is presented for thermally thin conditions (1 mm thick) using thermogravimetric analysis (TGA) and cone calorimeter measurements and four parameters for thermally intermediate conditions (4 mm thick) including a fire growth parameter, a smoke parameter, a toxicity parameter and the final mass residue. In addition, the shielding/charring effect of CF layers was characterised by modelling the reduction in the imposed heat flux due to this layer together with the previously determined thermal properties. By measuring the temperature at the back insulated surface of the composite in the cone calorimeter, we can also assess the fire resistance (integrity) of the composite and the heat transferred to the insulation behind this fuselage composite material in the aircraft application.

This paper investigates the effects of brominated and halogen-free fire retardants on the thermal stability and fire performance of glass-fibre reinforced poly(butylene terephthalate). Brominated polystyrene was used as the brominated fire retardant (BFR), whereas aluminium diethylphosphinate (Alpi) with/without nanoclay as halogen-free fire retardants (HFFRs). Tests were conducted using thermogravimetric analysis (TGA), limiting oxygen index (LOI), UL94 and the cone calorimeter. TGA results show that decomposition of glass-fibre plus PBT (PBT+GF) starts earlier in the presence of fire retardants (FRs). In the cone calorimeter, all FRs reduce significantly the heat release rate (HRR) compared to PBT+GF, with brominated polystyrene achieving lower HRR primarily because bromine released in the pyrolysis gases inhibits combustion, without however changing the mass loss rate (MLR). Alpi alone has significant effects on reduction of both HRR and MLR, which become considerably more when combined with nanoclay. The efficiency of combustion of the brominated polystyrene compound is much lower than that of HFFRs, which indicates that unburned pyrolysing gases are released during the combustion of brominated fire retardants.