Wall flame heights with external radiation

An existing flame heat transfer fire testing apparatus was used to study the upward flame spread potential of two kinds of wall materials: (1) PMMA (Polymethylmethacrylate) and (2) Douglas Fir Particle Board. PMMA is noncharring whereas Douglas Fir Particle Board is a charring material. Various levels of external radiant heat flux ranging from 1.8 W/cm² to 3.4 W/cm² were imposed onto the wall samples in order to measure the flame heights as a function of energy release rate. Flame height measurements were established visually by a review of video recordings. The results for these wall flames correlate flame height to the 2/3 power of energy release rate per unit sample width. The wall results are generally higher than data from gas burner line fires against a wall for a range of 10 to 200 kW/m.Note: This paper is a contribution of NIST and is not subject to copyright.     

Experimental measurements of upward flame spread on a vertical wall with external radiation

The overall objective of the project is to gain an understanding of the flame spread phenomenon under simulated surrounding fire conditions. In this phase of the project, emphasis is placed on obtaining experimental data for upward flame spread with applied external radiation on practical wall materials. A second phase (not yet reported) is the development of a numerical flame spread model and the experimental results presented here will be used for comparison with model predictions. Flame height, and in some cases pyrolysis height, were recorded as functions of time for 120 cm×30 cm samples; and these data were used to quantitatively investigate the effect of external radiation on several materials. Infrared heating panels were used to supply radiant fluxes of up to 15 kW/m² to the sample. Many wood-based materials do not exhibit flame spread to the top of the sample when ignited without applied external flux. With moderate levels of external radiation (5–10 kW/m²), many of these materials sustained flame spread to the top of the sample. With increasing external radiation levels, flame spread was also more rapid. A comprehensive series of tests was run on particle board to investigate the effect of igniter strength, preheat of the sample, and sample thickness. Igniter strength was not a significant factor and did not cause the flame spread to be sustained; the effect of preheat, even at moderate levels of radiant flux, was important; and sample thickness had a slight effect, with thicker samples burning slower. Total heat feedback to the sample was measured and the maximum values for various samples are reported. Experimental data obtained in this project will be used to aid in the development and validation of a numerical flame spread model.     

Influence of sidewalls on width effects of upward flame spread

A previous study has reported width effects for turbulent diffusion upward flame spread on thermally thick materials with sidewalls. However, sidewalls are not realistic. The present study has revisited this topic by performing experiments without sidewalls using 9 mm thick and 1000 mm tall PMMA slabs with widths of 100, 200, 300, 500 and 700 mm and by providing a hypothesis of the sidewall effects. Experimental data have revealed that the width effects still exist when sidewalls are absent. Flame heights and spread rates were higher for wider flames, although heat feedback to the fuel did not vary much with flame width. Compared to flames without sidewalls, the existence of sidewalls lengthened flame heights and generally reduced heat feedback along the central lines of the flames, resulting in higher flame spread rates for narrower flames and lower flame spread rates for wider flames. In addition, the absence of sidewalls enhanced the delivery of pyrolyzate towards the central line of the flames throughout the whole flame width.     

Width effect on upward flame spread

One previous experimental study has reported a width effect for upward flame spread rate on thermally thin fuels. A similar effect is expected for thermally thick fuels. This study revisited this topic by developing a hypothesis and performing experiments with sidewalls using 18 mm thick, 1000 mm tall PMMA slabs of widths 100, 200, 300, 500 and 700 mm. In the hypothesis, a lateral diffusion throughout the flame width was proposed to cause thicker flame along its centerline for wider flames and enhance combustion efficiency. Higher heat release rate per unit width, larger flame height, higher flame temperature and more heat feedback to the surface were consequently present. The corresponding flame spread rate was also increased and a power value of 0.35 existed between the flame spread rate and width in this study. All the experimental results clearly supported the hypothesis.     

Flame Heights and Heat Transfer in Façade System Ventilation Cavities

The design of buildings using multilayer constructions poses a challenge for fire safety and needs to be understood. Narrow air gaps and cavities are common in many constructions, e.g. ventilated façade systems. In these construction systems flames can enter the cavities and fire can spread on the interior surfaces of the cavities. An experimental program was performed to investigate the influence of the cavity width on the flame heights, the fire driven upward flow and the incident heat fluxes to the inner surfaces of the cavity. The experimental setup consisted of two parallel facing non-combustible plates (0.8 × 1.8 m) and a propane gas burner placed at one of the inner surfaces. The cavity width between the plates ranged from 0.02 m to 0.1 m and the burner heat release rate was varied from 16.5 kW to 40.4 kW per m of the burner length. At least three repeated tests were performed for each scenario. In addition, tests with a single plate were performed. The flame heights did not significantly change for Q′/W < 300 kW/m² (where Q′ is the heat release rate per unit length of the burner and W is the cavity width). For higher Q′/W ratios flame extensions up to 2.2 times were observed. When the distance between the plates was reduced or the heat release rate was increased, the incident heat fluxes to the inner surface increased along the entire height of the test setup. The results can be used for analysing methodologies for predicting heat transfer and fire spread in narrow air cavities.     

Burning rate and flame heat flux for PMMA in a cone calorimeter

Ignition and burning rate data are developed for thick (25 mm) black Polycast PMMA in a cone calorimeter heating assembly. The objective is to establish a testing protocol that will lead to the prediction of ignition and burning rate from cone data. This is done for a thermoplastic like PMMA. The incident flame heat flux, for irradiation levels of 0–75 kW/m², is found to be approximately 37 kW/m² for black PMMA. Its constancy is shown due to the geometry of the cone flame. Also, this flame is shown to be nearly transparent for cone irradiance (>90%). The heat of gasification of the black PMMA used is found to be approximately 2.8 kJ/g; higher than values reported for other PMMA samples. This is believed to be due to differences in molecular structure or pigmentation of the PMMA tested. A burning rate model is demonstrated to yield good accuracy in comparison to measured transient values. An exact solution is found for constant heat flux conditions.     

Experimental investigation on transverse ceiling flame length and temperature distribution of sidewall confined tunnel fire

This paper presents an experimental investigation on the transverse ceiling flame length and the temperature distribution of a sidewall confined tunnel fire. The experiments were conducted in a 1/6th scale model tunnel with the fire source placed against the sidewall, 0 m, 0.17 m and 0.35 m above the floor, respectively. Experiments of fire against a wall without a ceiling, 0.35 m above the floor in a large space, were also conducted as a control group. Results shows that for small heat release rate (HRR), the flame is lower than the ceiling and extends along the sidewall. With the increase of HRR and elevation of burner height, the flame gradually impinges on the ceiling and spreads out radially along it. The flame impingement condition and the flame shapes of the wall fire with and without ceiling are presented. From the viewpoint of the physical meaning of flame impinging on the ceiling, the horizontal flame length should be a function of the unburned part of the fuel at the impinging point. Based on the proportional relation between the flame volume and HRR, the effective HRR (Q_ef) at the ceiling is determined and the effective dimensionless HRR, Q^*_ef is defined to correlate the horizontal ceiling flame length. Additionally, predictive correlations of transverse ceiling temperature distribution are proposed for the continuous flame region, the intermittent flame region and the buoyant plume region under the ceiling, respectively.     

Effects of particle size on flame spread over magnesium powder layer

The spreading of a flame over a layer of magnesium powder has been examined to clarify the mechanisms of flame spread over metal powder layers and to establish effective ways to extinguish metal fires. Four grades of magnesium powder were used, with average grain diameters of 60, 170, 360 and 500 μm. The flame spread rate for larger particle sizes (D>170 μm) increased slightly with increasing particle size. However, the flame spread rate for smaller particle sizes (D<63 μm) increased sharply. The detailed surface temperature history of the magnesium powder layer was measured using infrared thermography (IR). Based on the temperature distribution in the pre-heat zone, the characteristic length l, characteristic depth δ, and characteristic time τ were calculated. The characteristic scale ratio, L/δ, for the larger particle sizes (D>170 μm) is almost unity, which suggests that the dominant heat transfer mechanism is heat conduction through the magnesium powder layer (solid phase). However, the L/δ for the smaller particle sizes (D<63 μm) is about 2.7, suggesting that the dominant heat transfer mechanism is convection. The flame spread rate over the magnesium powder layer was calculated by the de Ris model, a one-dimensional flame spread model, and a surface flash model. For larger particle sizes (D>170 μm), there is good agreement between the experimental flame spread rate and the flame spread rate estimated by the one-dimensional model. However, the flame spread rate was underestimated by the de Ris model, apparently because the de Ris model only considers heat feedback from the gas phase. For small-particle sizes (D<63 μm), there is good agreement between the experimental flame spread rate and the flame spread rate estimated by the surface flash model. This suggests that the flame spread over a small-particle layer can be described by a mechanism rather similar to that of gas phase flame propagation.     

Chemical flame heights

A “chemical” flame height has been defined from the ratio of CO to CO₂ yields, y_CO/y_CO2, and has been shown to be functionally identical with previous results based on flame luminosity. The chemical flame heights have been determined for propane and acetylene data for fire Froude numbers, Q^*, ranging from 0.1 to 60,000. The functional dependence of Z_f/D on Q^* was found to be in excellent agreement with previous luminous flame height correlations. It was thus concluded that the present methodology can be used to accurately quantify the luminous flame height for well-ventilated diffusion flames of surface fires.     

Numerical model of upward flame spread on practical wall materials

The focus of this paper is the development of a thermal, finite difference numerical model to describe one-dimensional upward flame spread on practical wall materials. Practical materials include composite materials and those that char, in addition to clean burning, homogeneous materials. A set of equations used in the model is developed and the methods for obtaining necessary “fire properties” are discussed. Some of the particular features of the model include the use of a correlation for flame heat feedback and the use of an experimentally measured mass loss rate to incorporate the burning characteristics of practical materials. A comparison of the numerical predictions with the experimental results for flame heights and temperatures are shown for Douglas fir particle board. The model correctly predicts trends but underpredicts the flame heights and pyrolysis height in the cases tested. Two additional cases are shown for materials for which experimentally measured heat release rate data are used in place of the mass loss rate data. The flame and pyrolysis height predictions are in much better agreement for these cases. Further efforts to obtain material property data that is appropriate for flame spread modeling is indicated by this work.     

Modeling thermal behaviors of window flame ejected from a fire compartment

A series of reduced-scale experiments were carried out in order to investigate thermal behaviors of window flame, which exposes the upper floors as well as the adjacent buildings to potential risks of fire spread. A stainless pan filled with alcohol was used as the fire source and was placed inside a cubic compartment of 900 mm side. Temperatures and velocities at various points inside and outside of the compartment were measured. The compartment was pressurized during the experiment by mechanically supplying air at several mass inflow rates through an inlet duct set at the bottom part of the compartment. This was for simulating fire conditions under the effect of external wind pressure. On the basis of the experimental observation, line (i.e., two-dimensional) heat source assumption was adopted for developing a model of window flame behavior. A dimensionless parameter Q^′*${Q^{\prime }}^{*}$ was derived from the governing differential equations in order to generalize the measurement results. Expressions for temperature rise along the trajectory ΔT_m and characteristic flame width b_T were developed incorporating the parameter Q^′*${Q^{\prime }}^{*}$.     

Contribution to flashover modelling: Development of a validated numerical model for ignition of non-contiguous wood samples

A computational model of flashover is presented that closely follows the experimental setup at CNRS-ENSMA-Poitiers. A propane burner with thermal power of 55 kW is used as a primary source of fire and square beech wood samples (30 mm×30 mm×5 mm) as fire spread targets. The computational model describes the wood pyrolysis with a progress variable. Using the conservation of heat fluxes at the solid–gas interface, the thermal diffusion in the wood samples is coupled with the convective and the radiative heat transfer in the ambient gas phase. The incoming heat flux at the upper surface of the wood samples reaches values between 20 and 30 kW/m². With the ignition and subsequent combustion of the pyrolysis volatiles, the heat flux increases by approx. 12 kW/m². The results show that the ignition of the wood samples is triggered at an approx. surface temperature of 650 K. Due to large local variations in incident heat flux, significant differences in the ignition times of the wood samples are observed. The comparison of the calculated and the experimentally measured temperature shows a good agreement for the first wood sample and the model predicts the ignition time very well. But for the second and the third wood samples the model overpredicts the temperature, which leads to a premature ignition of these wood samples.     

Material fire properties and predictions for thermoplastics

Ignition and burning rate data are developed for nylon 6/6, polyethylene, polypropylene and black polycast PMMA in a cone calorimeter heating assembly. The objective is to examine a testing protocol that leads to the prediction of ignition and burning rate for thermoplastics from cone calorimeter data. The procedure consists of determining material properties, i.e. thermal inertia, specific heat, thermal conductivity, ignition temperature, heat of gasification and flame heat flux from cone data, and utilizing these properties in a model to predict the time to ignition and transient burning rate. The procedure is based on the incident flame heat flux being constant in the cone calorimeter which occurs for flames above the top of the cone heater. A constant net flame heat flux of approximately 20 kW/m² for nylon 6/6, 19 kW/m² for polyethylene, 11 kW/mP² for polypropylene and 28 kW/m² for black PMMA is obtained for irradiation levels ranging from 0 to 90 kW/m². The burning rate model is shown to yield good accuracy in comparison to measured transient burning in the cone assembly.     

Effect of oxygen on flame heat flux in horizontal and vertical orientations

A systematic empirical and analytical study was conducted to directly quantify the effect of enhanced ambient oxygen concentration on flame heat flux at bench scale and its ability to represent large-scale flame heat flux of well-ventilated fires. The Advanced Flammability Measurements Apparatus was used to conduct bench scale horizontal and single wall vertical orientation experiments for black polymethylmethacrylate, propylene gas and black polyoxymethylene. The key aspect of this study was direct experimental measurements of flame heat flux back to the burning surface for 20.9–40% ambient oxygen concentrations over a range of applied heat flux. The total flame heat flux, as well as the radiative and convective components, was experimentally measured with various gages. To gain more insight into the effects of oxygen, the flame emissivity, flame height and flame temperature were measured and used to calculate the radiative and convective components of the flame heat flux. Gas burner experiments were conducted to decouple the solid and gas phase effects of the ambient oxygen. Large scale tests of black polymethylmethacrylate were conducted in a horizontal orientation and literature data was used for single wall vertical orientations for comparison to the bench scale, enhanced oxygen results. The main conclusion is that the flame heat flux in enhanced ambient oxygen bench scale does not simulate large-scale flame heat flux in horizontal orientations but simulates a more severe large-scale geometry (parallel wall) in vertical orientations and is useful for evaluation of materials’ vertical flame spread potential.     

受限距离影响下竖向顺流火蔓延规律研究

以长45.0 cm、宽5.0 cm、厚2.0 mm的广告画布作为研究对象,开展了受限距离为3.0~9.0 cm的竖向顺流火蔓延实验,定量分析受限距离对火焰长度、火蔓延速度、点燃时间等参数的影响。结果表明：受限距离3.0~8.0 cm时,受限侧火焰长度明显大于未受限侧火焰长度;受限距离超过8.0 cm时,两侧火焰长度基本一致。随受限距离增加,火焰长度与火蔓延速度均表现为先增加后减小,最终趋于稳定,而点燃时间则表现为先减小后增加,最终趋于稳定值。建立了材料表面接收热流与点燃时间的关系式,分析了材料表面接收热流随受限距离的变化趋势,阐释了火蔓延速度随受限距离非单调变化的原因。     

Warehouse commodity classification from fundamental principles. Part II: Flame heights and flame spread

In warehouse storage applications, it is important to classify the burning behavior of commodities and rank them according to their material flammability for early fire detection and suppression operations. In this study, a preliminary approach towards commodity classification is presented that models the early stage of large-scale warehouse fires by decoupling the problem into separate processes of heat and mass transfer. Two existing nondimensional parameters are used to represent the physical phenomena at the large-scale: a mass transfer number that directly incorporates the material properties of a fuel, and the soot yield of the fuel that controls the radiation observed in the large-scale. To facilitate modeling, a mass transfer number (or B-number) was experimentally obtained using mass-loss (burning rate) measurements from bench-scale tests, following from a procedure that was developed in Part I of this paper.Two fuels are considered: corrugated cardboard and polystyrene. Corrugated cardboard provides a source of flaming combustion in a warehouse and is usually the first item to ignite and sustain flame spread. Polystyrene is typically used as the most hazardous product in large-scale fire testing. The nondimensional mass transfer number was then used to model in-rack flame heights on 6.1-9.1 m (20-30 ft) stacks of ‘C’ flute corrugated cardboard boxes on rack-storage during the initial period of flame spread (involving flame spread over the corrugated cardboard face only). Good agreement was observed between the model and large-scale experiments during the initial stages of fire growth, and a comparison to previous correlations for in-rack flame heights is included.     

A Physical Model for Flame Height Intermittency

In this paper, a simple physical model for flame height intermittency in buoyant diffusion flames is proposed. We use dimensional analysis to link existing relations for the pulsation period of diffusion flames and the burnout time of discrete volumes of gaseous fuel. From this analysis, relationships for flame height and intermittency as a function of the ratio of pulsation to burnout timescales is presented. To test these concepts, an automated method to measure flame height from digital images was developed. For a sequence of images, the flame is isolated from its surroundings using a popular histogram-based image segmentation algorithm. Spatial accuracy is obtained through automation of a common photogrammetric technique. The physical model and measurement methods are tested using four methane burners with equivalent diameters ranging from 23 cm to 81 cm and heat release rates ranging from 10 kW to 1500 kW. A total of 64 burner/HRR combinations were observed. 50th percentile flame heights ranging from 0.25 m to 3.5 m were measured. A statistical analysis of the automated flame height measurements showed total spatial error rates of less than 10% for even the least optimal camera setups. The experimental results show that flame intermittency (relative to 50th percentile flame length) scales with dimensionless heat release rate (Q* _D ) to the ? 1/5 power when the ratio of burnout to fire pulsation timescales is >1. When this ratio is less than one, the relative flame intermittency rapidly approaches unity since only a single flame pulse exists in the plume.