Characterization of thermal properties and analysis of combustion behavior of PMMA in a cone calorimeter

Fire Safety Journal 46 (2011) 451–461

Contents lists available at ScienceDirect

Fire Safety Journal journal homepage: www.elsevier.com/locate/firesaf

Characterization of thermal properties and analysis of combustion behavior of PMMA in a cone calorimeter Jocelyn Luche a,n, Thomas Rogaume a, Franck Richard a, Eric Guillaume b a b

Institut Pprime, UPR 3346 CNRS, De´partement Fluides, Thermique, Combustion, ENSMA, BP 40109, 86961 Futuroscope, France Laboratoire national de me´trologie et d’essais (LNE), Test Direction, 78197 Trappes Cedex, France

a r t i c l e i n f o

abstract

Article history: Received 10 January 2011 Received in revised form 14 July 2011 Accepted 18 July 2011 Available online 9 August 2011

This paper deals with the thermal degradation of a black poly(methyl)methacrylate (PMMA) in a cone calorimeter (CC) in air with a piloted ignition. The influence of several heat fluxes (11 kW m  2 and 12 kW m  2, and ten values from 15 to 60 kW m  2 in steps of 5 kW m  2) on PMMA sample degradation and the decomposition chemistry has been studied. Thus, thermal properties have been deduced and calculated from ignition time and mass loss rate (MLR) curves. During our experiments, among compounds quantified simultaneously by a Fourier transformed infrared (FTIR) or gas analyzer, five main species (CO2, CO, H2O, NO and O2) have been encountered, regardless of the external heat flux considered. The main product concentrations allow calculation of the corresponding emission yields. Thus, mass balances of C and H atoms contained in these exhaust gases were able to be compared with those included in the initial PMMA sample. Using the standard oxygen consumption method, heat release rate (HRR), total heat release (THR) and effective heat of combustion (EHC) have been calculated for each irradiance level. Therefore, these different results (thermal properties, emission yields, HRR, THR and EHC) are in quite good accordance (same order of magnitude) with those found in previous studies. & 2011 Elsevier Ltd. All rights reserved.

Keywords: PMMA Cone calorimeter FTIR Heat release rate Mass loss rate Emission yields Effective heat of combustion Thermal properties ISO 5660 standard

1. Introduction Plastics and polymers, such as poly(methyl)methacrylate (PMMA), are used as replacements for conventional materials (e.g. wood, metals, glass, etc.) in numerous domains such as electrical appliances, light diffusers, optical fibers, furniture, transport, hygiene and health, etc. Indeed, these widespread kinds of synthetic compounds, with good technical, mechanical, chemical, optical performances, etc., are easily manufactured for a moderate cost and can be used in a wide range of applications. Thus, compounds made of PMMA can be used in numerous applications as signs and signboards (illuminated panels, 3D lettering, indicator panels, etc.), POS (point-of-sale) advertising (display stands, testers, notice-boards, etc.), interior design (shopfitting, furniture, projection screens, glazing, etc.), transport (deflectors, sun visors, registration plates, ship portholes and windows, etc.) or industrial (machine guards, dials, precision parts, etc.) devices. In spite of these various uses linked to numerous advantages, plastics (highly combustible materials) constitute a grave danger (human injuries and deaths) in case of

n

Corresponding author. E-mail address: [email protected] (J. Luche).

0379-7112/$ - see front matter & 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.firesaf.2011.07.005

fire, as under the right conditions, they readily ignite and burn vigorously. As a reference fuel material used in a cone calorimeter (CC), solid acrylic PMMA polymer has been widely used – with or without filler – during previous studies for assessing polymer flammability and characterizing the mass loss rate (MLR) during combustion processes [1–14]. Only some of these previous studies and other specific studies propose the chemical analysis and quantification of the effluents emitted during PMMA thermal degradation, and so include the calculation of the heat release rate (HRR), which is the most important parameter in fire characterization studies [1–5,15–22]. Finally, a few studies propose a chemical kinetic mechanism for predicting the PMMA degradation in a CC [23,24]. Generally, only the assumption of a single-step degradation corresponding to depolymerization and formation of MMA is carried out. The present paper deals with a complete and detailed study of thermal degradation of a black non-charring PMMA in a CC, including the characterization of thermal properties and the quantification of exhaust gas concentrations of this typical polymer. Thus, the main purpose of this work is to characterize the influence of a widespread range of irradiance levels (external heat flux) on the MLR of a PMMA sample and on the product amounts released during this combustion process. Next, the HRR was determined based on the O2-consumption principle in a CC.

452

J. Luche et al. / Fire Safety Journal 46 (2011) 451–461

Nomenclature Greek notations

e f krc

r s

emissivity oxygen depletion factor thermal effusivity, kJ2 m  4 K  2 s  1 density, kg m  3 Stefan–Boltzmann constant 5.6704  10  8 W m  2 K  4

Notations A c CC CHF DHg E ECO EHC EHC FHF FHFnet FID FTIR hc HRR

sample area exposed to cone calorimeter heat flux, cm2 specific heat, kJ kg  1 K  1 cone calorimeter, – critical heat flux, kW m  2 latent heat of gasification, kJ g  1 net heat release per unit mass of oxygen consumed, MJ kg  1 of O2 net heat release per unit mass of oxygen for CO, MJ kg  1 of O2 effective heat of combustion, kJ g  1 averaged effective heat of combustion, kJ g  1 flame heat flux, kW m  2 net flame heat flux, kW m  2 flame ionization detector, – Fourier transformed infrared, – convective heat transfer coefficient, W m  2 K  1 heat release rate, kW m  2

Finally, the MLR and HRR parameters have been correlated with main gaseous compound evolutions and their emission yields to propose a description of physical and chemical phenomena occurring during PMMA thermal degradation.

2. Experiments 2.1. Materials The material used in this study is a black non-charring PMMA, commonly known as Altuglas, supplied by the company VACOUR and synthesized via radical polymerization. The PMMA elementary analysis was conducted by a combination of catharometry and non-dispersive infrared (NDIR) detection. Table 1 presents the averaged composition along with the analysis methods that were repeated three times and provided a dispersion of 70.3 wt%. Table 1 Elemental analysis methacrylate).

of

black

poly(methyl

Elements

Composition (wt%)

Carbon (C) Hydrogen (H) Oxygen (O) Nitrogen (N) Sulfur (S) Chlorine (Cl) Water (H2O) Total

59.1 7.9 31.9 o0.3 o0.2 0.1 0.6 o100.1

HRR k m MAir _ Air m Mi _i m _e m MLR NDIR P PMMA q_ 00ext R SMLR SMLR t T tig Tig T0 THR V_ e Vm Xi0 Xi Yi

averaged heat release rate, kW m  2 thermal conductivity, kW m  1 K  1 mass, g molar mass of air, g mol  1 mass flow rate of the incoming air, g s  1 molar mass of species i, g mol  1 mass flow rate of species i in the exhaust duct of CC, g s1 mass flow rate in the exhaust duct of CC, g s  1 mass loss rate, g s  1 non-dispersive infrared, – pressure, Pa poly(methyl)methacrylate, – external heat flux, kW m  2 universal constant of perfect gas 8.314472 J mol  1 K  1 s mass loss rate, g m  2 s  1 averaged specific mass loss rate, g m  2 s  1 time, min or s temperature, K ignition time, min or s ignition temperature, K initial (ambient) temperature, K total heat release, MJ m  2 volumetric flow rate in the exhaust duct, L s  1 molar volume, L mol  1 mole fraction of species i in the incoming air, – mole fraction of species i in the exhaust gas measured by analyzer, – emission yield of species i, gi/gsample

Elementary analysis results show that no inert load, flame retardants or fillers were used during the manufacturing of the PMMA sample; neither chlorine- nor sulfur-based additives were found. Indeed, 100 wt% of the total sample mass was composed of C, H, O, N and S atoms. This result is in good accordance with various compositions of such PMMA reported in [25]. Based on this elementary analysis composition, the raw chemical formula of the virgin PMMA was determined to be (C4.9H7.8O2.0)n (with n¼PMMA polymerization degree). Its molecular weights and molecular weight distributions were obtained from an Agilent Technologies Size Exclusion Chromatography (SEC) calibrated with a polystyrene standard. Samples of 25 and 70 mg (1 and 2 in Table 2) were dissolved in chloroform and then injected into the SEC. Table 2 shows the molecular weights (Mn and Mw), molecular weight distributions (Ip) and polymerization degree (n) of each sample analyzed. Molecular weight distribution (Ip), with a value around 3.7, shows that PMMA is a polymolecular compound obtained by radical polymerization (IpZ2) according to the definition given by some authors in the literature [26,27]. From these results, the polymerization degree (n) of the PMMA sample can be determined as equal to value ranging from 1500 to 1800.

Table 2 Characterization of black poly(methyl)methacrylate mass (SEC analysis). Elements

Mn

Mw

Ip¼ Mw/Mn

n

1 2

178,670 150,440

639,340 578,480

3.6 3.8

1785 1503

J. Luche et al. / Fire Safety Journal 46 (2011) 451–461

2.2. Cone calorimeter The PMMA reaction-to-fire characterization was carried out in a CC (Standard ISO 5 660 [28]) made by Fire Testing Technology Limited, under fully ventilated conditions. The sample dimensions were 99 71 mm long, 9971 mm wide and 14 71 mm thickness, with a mass of 169.6 72.9 g. Thus, mass density (r) calculated from these data is found to be equal to 1213.3719.1 kg/m3. Fig. 1 presents a schematic view of the CC with a solid PMMA matrix in the sample holder exposed to an irradiance level from an electric heater. Several heat flux levels were used: 11 and 12 kW m-2, and ten heat flux levels from 15 to 60 kW m-2 in steps of 5 kW m-2. Tests were carried out with a piloted ignition in air, and were repeated at least five times for each condition. Moreover, the ignition spark was positioned above the sample up to ignition (and removed thereafter), so all the tests were performed in a flaming condition. According to the standard ISO 17554 [29], the experiments were stopped manually if no ignition occurred after 30 or 32 min after ignition or when the mass loss became zero. During these experiments, the CC fan flow rate was taken to be equal to 0.024 70.002 m3 s-1. At the end of all experiments and, regardless of the external heat flux chosen, all the PMMA sample was entirely degraded, and no solid or liquid residue was found inside the sample holder. 2.3. Gas analysis Exhaust gases produced by the thermal degradation of the PMMA sample in the CC were sampled from the exhaust duct using two kinds of on-line gas analyzers:

453

HORIBA apparatus during our experiments. Moreover, this FTIR spectrometer allows the identification and quantification of other combustion products such as NO2, NH3, HCN, N2O, CH4, C2H2, C2H4, C2H6, C3H6, C3H8 and H2O (not quantified by the HORIBA gas analyzer). Furthermore, before all the experiments, the two devices used for gas analysis were calibrated with well-known concentration standards to quantify the five gaseous combustion products for the HORIBA apparatus and the 15 products for the FTIR spectrometer measured simultaneously during the combustion process. The entire transport line (from the sampling point to the quantification apparatus) was heated to 170 1C and was not dried before passing through the FTIR measurement gas cell. This allowed the transport of combustion products while avoiding water vapor condensation and the trapping of water-soluble compounds. It also enabled quantification of H2O (vapor).The sampling line was connected to a filtration box where a cellulosic filter and a stainless steel filter were used to retain heavy products and soot particles. After filtration, gases were transported up to the gas analyzers (FTIR and HORIBA). The FTIR analysis technique used (including sampling and filtering device) was validated during the SAFIR project [30], which constituted the basis for toxicity analysis carried out following the guidelines of the standard ISO 19702 [31–35].

3. Thermal property results 3.1. Parameters of inflammation and combustibility

 A gas analyzer HORIBA PG 250 equipped with three units to



simultaneously quantify gaseous products (NO, O2, CO2, CO and SO2). The NOx analysis unit uses a cross-modulation ordinary pressure chemiluminescence method, the SO2, CO, and CO2 unit uses an NDIR absorption method and the O2 unit uses an electrochemical Zirconia method using a galvanized cell. The sampling flow rate was 0.4 NL min  1. A Fourier transformed infrared (FTIR) spectrometer (ThermoNicolet 6700 equipped with a MCT-A detector and a measurement gas cell of volume¼0.2 L and optical path-length¼2 m). The sampling flow rate was equal to 2 NL min  1. The FTIR analyzer was also used to quantify the NO, CO2, CO and SO2 species, which allows the control and comparison of these concentration values with those values measured using the

For the PMMA studied in the CC under the conditions outlined in the previous part, sample ignition can be achieved when an external heat flux q_ 00ext (with a relative uncertainty of 75%), higher than a critical heat flux (CHF), is applied during a time interval tig . These different parameters can be defined and calculated by using Eqs. 1 and 2, which are described and fully detailed in Refs. [1–4,14]: tig ¼

  Tig T0 2 2 krc 00 3 q_ ext

q_ 00ext ¼

1h

e

ð1Þ

i hc ðTig T0 Þ þ esTig4  CHF

Cone calorimeter fan Spark ignition Temperature and differential pressure measurements Soot filter

Cone calorimeter hood

GAS ANALYZERS: -FTIR -HORIBA PG250 Heated sampling line (170°C) Cone calorimeter (Electrical heater) Weighting device

Sample holder

Fig. 1. Schematic layout of the coupling of cone calorimeter and gas analyzers.

ð2Þ

454

J. Luche et al. / Fire Safety Journal 46 (2011) 451–461

" # " # hc ðTig T0 Þ þ esTig4 1 e 00 _ pffiffiffiffiffi ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Uq ext  pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi tig ð2=3ÞkrcðTig T0 Þ ð2=3ÞkrcðTig T0 Þ

ð3Þ

1=2 ¼ f ðq_ 00ext Þ curve The slope of a plot obtained from the tig (Eq. (3)) allows computation of the different thermal properties such as the theoretical critical heat flux (CHF), the specific heat (c), the thermal conductivity (k) and the theoretical ignition temperature (Tig) by using the following equations: " # " # 4 hc ðTig T0 Þ þ esTig e Slope ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ; yintercept ¼  pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð2=3ÞkrcðTig T0 Þ ð2=3ÞkrcðTig T0 Þ

ð4Þ   yintercept CHF ¼  Slope

ð5Þ

Fig. 2 shows plots of the curve evolution of the square root of the inverse of ignition time as a linear function of external irradiance (Eq. (3)), with a straight line used to fit the data (slope¼0.00509 and yintercept ¼  0.02434) up to the minimum experimental heat flux corresponding to critical ignition (11 kW m  2) represented by a vertical dashed line. Moreover, in Fig. 2, the fit line (the curve slope) reaches an intersection point with the x-axis of approximately 4.8 kW m  2. This value represents the theoretical CHF for PMMA ignition computed from Eqs. (4) and (5). This CHF theoretical result is in good accordance with previous results found in the literature where the CHF (obtained by the same way) is equal to 4 kW m  2 [11] or 5 kW m  2 [1–5]. The effective CHF determined experimentally in this study (11 kW m  2) is in good accordance with results found in the literature (9–13 kW m  2 [5,11,14,36]). The difference between the experimental and extrapolated CHF is due to the non-linearity of the ignition time leading to the non-linearity of the square root of the inverse of ignition time (Fig. 2) for the lowest external heat flux, especially near the CHF [14] (where the solid material is thermally thin compared to the time of thermal transfer). Another explanation [5] of this

difference between theoretical and experimental values of CHF has suggested the influence of the surface reradiation of the PMMA sample, which causes a higher experimental CHF than the value extrapolated from Fig. 2. Moreover, different methods described in previous studies [37–39] have shown that the theoretical CHF value can be used to calculate a more realistic CHF. Indeed, for example, some intercept values (CHF) are defined by the authors as being only 0.64 [37] or 0.76 [38] fraction of the CHF. 3.2. Transient burning rate Fig. 3 shows the experimental transient evolution of the MLR at the different external irradiance levels or external heat flux used during the PMMA thermal degradation experiments. Time t¼0 marks the beginning of the exposure to the desired irradiance level rather than the moment of ignition. Regardless of the heat flux chosen, three or four – more or less significant – decomposition time ranges (corresponding to three or four peaks or slope variations, which reach a maximum MLR value as summarized in Table 3) can be observed in Fig. 3 with a major one that has the most significant peak intensity (corresponding to the 3rd stage in Table 3). Moreover, an increase of external heat flux moves the different stages towards the first one while increasing curve intensity. Indeed, when the external heat flux increases from 11 to 60 kW m  2, all the PMMA sample mass is lost on a shorter time (e.g. around 10 min at 60 kW m  2 against 45 min at 11 kW m  2), with a higher maximal intensity of the MLR peak (e.g. around 0.5 g s  1 at 60 kW m  2 compared to 0.15 g s  1 at 11 kW m  2). As shown, the MLR shapes and their intensities depend strongly on and change with the irradiance level value. After a latency time, the first stage starts with the ignition of the PMMA sample and the very quick rise of MLR after ignition. Next, the PMMA decomposition continues during the 2nd and 3rd time ranges, which correspond to the MLR evolution following a plateau (2nd column of Table 3) and reach a significant peak (3rd column of Table 3), respectively. Finally, the last decomposition time range (4th column of Table 3) corresponds to an MLR decrease and the consumption of the remaining PMMA amount in the sample holder. At the end of the different experiments, no liquid or solid residue was found in the sample holder. The specific mass loss rate (SMLR) is determined as the ratio between the MLR and the sample surface exposed to the CC external irradiance level (i.e. A¼88.4 cm2, as explained in the paragraph 6.7 of the ISO 5660 standard [28], where the opening for the specimen face exposed to heat flux coming from the CC heater is equal to 9.470.5 cm  9.470.5 cm). Moreover, the SMLR is defined (Eq. (6)) as the sum of the heat flux from the flame (FHF) and the external heat flux (q_ 00ext ), less the radiative heat flux loss (esTig4 ), divided by the latent heat of gasification (DHg ). According to a hypothesis carried out in previous studies [1–4], black PMMA approximates a vaporizing solid, and the flame heat flux (FHF) is constant. The FHF [1–4] or total FHF [9] is defined as the sum of the radiative (FHFr) and convective (FHFc) components of the FHF (FHF ¼ e FHFr þFHFc , where e is emissivity). The net FHF [1–4,9] is defined as the material’s contribution to burning in the CC and is the difference between (total) FHF and radiative heat flux loss from the sample surface (FHFnet ¼ FHFesTig4 ). SMLR ¼

Fig. 2. Square root of the inverse of ignition time (s  0.5) as a function of heat flux (kW m  2).



 4 FHFesTig 1 q_ 00ext þ DHg DHg

ð6Þ

Experimental averaged SMLRs are plotted in Fig. 4 as functions of external incident heat flux. The SMLR (or MLR) is accelerated

J. Luche et al. / Fire Safety Journal 46 (2011) 451–461

455

Table 3 Time range of the four stages of PMMA thermal degradation. Time range (min) Heat flux (kW m  2)

1st stage

2nd stage

3rd stage

4th stage

11 12 15 20 25 30 35 40 45 50 55 60

23  25 13  15 57 23 12 0.5  1.5 0.5  1 0.5  1 0.2  0.4 0.2  0.4 0.2  0.4 0.2  0.4

25  35 15  24 7  15 3  12 29 1.5  8 18 1  7.5 0.4  7 0.4  6.5 0.4  5 0.4  5

35  45 24  34 15  25 12  18 9  15 8  14 8  11.5 7.5  10.5 7  10 6.5  9 5  8.5 58

445 434 425 418 415 414 411.5 410.5 410 49 48.5 48

Fig. 4. Averaged specific mass loss rate SMLR (g m  2 s  1) vs. heat flux (kW m  2).

The slope ( ¼0.4268) and yintercept ( ¼6.3606) of the best fit line to flaming black PMMA data obtained from the SMLR ¼ f ðq_ 00ext Þ curve (Eq. (6)) allows the computation of other thermal properties such as gasification heat (DHg ¼2.34 kJ g  1), flame heat flux (FHF¼17.67 kW m  2) and net flame heat flux (FHFnet ¼ 14.90 kW m  2) by using Eq. (7), which is deduced from Eq. (6): Slope ¼

Fig. 3. Mass loss rate (g s  1) as a function of experimental time (min).

when the irradiance level is increased up to a high value. Indeed, the SMLR increases from 11 to 32 g m  2 s  1 when incident heat flux increases from 11 to 60 kW m  2.

1 ; DHg

yintercept ¼

FHFesTig4

DHg

ð7Þ

The black PMMA gasification heat (DHg ¼2.34 kJ g  1) found in this study is in good accordance with previous results from the literature, as presented in Table 4. As seen for thermal properties and CHF, theoretical results calculated or experimental ones measured during this black PMMA thermal degradation study in a CC are of the same magnitude order as those found previously in literature. Slight differences can be observed between the different results; this is probably due to the wide disparity between PMMA samples used, indicating a real difference due to elementary composition and

456

J. Luche et al. / Fire Safety Journal 46 (2011) 451–461

density as well as experimental measurement uncertainties. The difference between results can also be explained by the difference due to the cone designs used or due to result interpretations. In the latter case, in the literature for theoretical CHF determination, only a part of the results (up to 40 kW m  2 in [1–4] or between 20 and 50 kW m  2 in [5]) are fitted by a straight line, which has a more or less significant influence on the CHF determination, and so influences the computation of other parameters (c, k, DHg).

3.3. Heat release rate (HRR) The HRR is the most significant parameter for the fire hazard material evaluation [40,41] since it controls the rate of growth in fire, including heat and ultimately the amount of smoke and toxic gas generated. This parameter is measured by using the oxygen consumption calorimetry technique [42] based on the ISO 5660 standard [28]. The O2, CO2, CO, H2O and N2 species account for approximately 99% of the exhaust gases in the majority of fullscale fire tests [43,44]. Thus, these gaseous compounds, identified as the major species in our experimental results (Fig. 7 and Table 5), will be considered as the only species in the exhaust gas flow. As H2O concentrations were measured during our experiments, the HRR value can be computed by the following equations (described in detail in Refs. [43–49]):   1 1f XCO MO2 0 _ ð1XH0 O XCO EfðECO EÞ HRR ¼ m ÞXO0 2 ð8Þ 2 2 A 2 XO2 MAir Air

f¼

0 XO0 2 ð1XCO XCO2 ÞXO2 ð1XCO Þ 2

Me ¼ 18 þ 4ð1XH2 O ÞðXO2 þ 4XcO2 þ 2:5Þ

ð11Þ

where A¼88.4 cm2 (sample area exposed to CC heat flux), 1 E¼13.1 MJ kg  1 of O2; ECO ¼17.6 MJ kg  1 of O2; MO2 ¼ 32 g mol , _ Air is the mass flow rate of the incoming air, m _e MAir ¼29 g mol  1, m is the mass flow rate in the exhaust duct of the CC, Xi is the mole fraction of species i in the exhaust gas measured by the analyzer, Xi0 is the mole fraction of species i in the incoming air and f is the oxygen factor. Fig. 5 shows the transient evolution of the SMLR and the HRR at four irradiance levels (15, 30, 45 and 60 kW m  2). As seen previously for MLR, HRR and SMLR in this paper, transient evolutions of these parameters depend strongly on the irradiance level, and both parameters have the same curve shapes compared to the MLR one for a given external heat flux. As for the MLR transient evolution in Fig. 3, the four PMMA thermal decomposition stages summarized in Table 3 are also found from HRR and SMLR curves in Fig. 5, regardless of the incident heat flux applied to the sample.

ð9Þ

XO0 2 ð1XO2 XCO XCO2 Þ

_ Air _e ð1XH2 O Þð1XO2 XCO XCO2 Þ m m ¼ 0 Þ MAir Me ð1XH0 2 O Þð1XO0 2 XCO 2

ð10Þ

Table 4 Comparison of PMMA gasification heat of this study with values from literature.

DHg (kJ g  1)

References

2.34 1.34–1.48 1.6 1.96 2.2–2.7 2.6 2.77

This study [12] [5,6,10,11–13] [8] [6] [9] [1–4]

Fig. 5. HRR and SMLR curves as functions of time at different external heat flux.

Table 5 HRR, SMLR , EHC and THR obtained during this study and from literature. Results from this work Heat flux (kW m 11 15 20 25 30 35 40 45 50 55 60 Mean

2

)

HRR (kW m

2

274.3 7 116.1 317.4 7 111.1 349.4 7 138.5 412.0 7 166.0 463.8 7 222.0 499.9 7 238.7 577.7 7 267.4 633.4 7 303.4 691.4 7 346.9 735.3 7 315.4 780.8 7 365.7 521.4 7 174.8

)

Results from literature SMLR (g m 11.2 74.5 12.8 74.4 15.0 75.3 17.4 76.3 18.6 77.9 20.8 79.0 23.2 79.9 25.9 710.5 27.7 711.4 30.5 712.2 31.6 713.0 21.3 77.0

2

s

1

)

EHC (kJ g

1

24.0 72.3 25.0 73.0 23.7 76.7 24.0 76.2 24.8 74.0 24.7 79.1 24.9 72.3 24.0 74.1 24.6 74.4 25.3 77.7 24.3 72.8 24.5 70.5

)

THR (MJ m

2

445.6 450.1 452.9 450.4 466.4 451.1 466.4 471.1 492.6 459.4 468.5 461.3 7 13.6

)

HRR (kW m)  2

EHC (kJ g  1)

21.8 [23] 401.56 [1–4]

24.0 [1–4] 23.2 [15,16]

642.21 [1–4]

24.1 [1–4]

24–25 [5,14,17–21,45]

J. Luche et al. / Fire Safety Journal 46 (2011) 451–461

The effective heat of combustion (EHC) value is a time- and irradiance-level-dependent parameter that corresponds to the heat released from the volatile portion during solid material combustion. From the HRR and SMLR, the EHC can be computed using the following equation: EHC ¼

HRR SMLR

ð12Þ

Averaged results for SMLR, HRR, EHC and total heat release (THR) are presented in Table 5 for the set of irradiance levels studied (from 11 to 60 kW m  2). Globally, HRR increases from 275 to 780 kW m  2, SMLR increases from 11 to 32 g m  2 s  1, EHC values are around to 24.570.5 kJ g  1 and THR values are around 460714 MJ m  2 with increasing external heat flux (from 11 to 60 kW m  2, respectively). These averaged HRR and/or EHC results are compared to some black PMMA thermal degradation data found in the literature. As seen in Table 5, results found during this black PMMA thermal degradation study in the CC are in quite good accordance (same order of magnitude) with those given by previous studies.

4. Exhaust gas concentrations and emission yields In Fig. 6, CO2 concentrations obtained simultaneously from the gas analyzer (HORIBA) and the FTIR spectrometer are plotted as functions of time for external heat fluxes equal to 15, 30, 45 and 60 kW m  2. As can be seen in Fig. 6, regardless of the heat flux considered, good accordance is observed between the CO2 concentrations measured with the two types of gas analysis devices. The same observations can be carried out for the CO and NO concentration comparison between the HORIBA gas analyzer and the FTIR spectrometer. Therefore, regardless of the CO or CO2 concentrations chosen (measured with the HORIBA or the FTIR apparatus), this choice has a minor impact (o2%) on the different calculation results, i.e. on HRR, EHC, THR or emission yield results. External heat flux influence on PMMA sample thermal degradation has been studied with concentration evolution of major products. Evolution curves of CO, CO2, O2, H2O, NO, NO2 and HCN concentrations are plotted as functions of time for an external

457

heat flux equal to 15, 30, 45 and 60 kW m  2 (Fig. 7a–d, respectively). As can be seen in Fig. 7, an influence of the external heat flux value can be noticed between the different curve sets for each exhaust gas. Indeed, when the irradiance level increases from 11 to 60 kW m-2, CO, CO2, H2O, NO, NO2 and HCN concentration levels increase, and the O2 one decreases. During these experiments, concentrations of CH4, C2H2, C2H4, C2H6, C3H6 and C3H8 have been measured with maximum values that never exceed 10 ppm. No significant amounts ( o1 ppm) of SO2, N2O or NH3 have been quantified (i.e. the maximum value measured for each concentration is near the detection limit, but lower than the quantification limit of the gas analysis apparatus for these gaseous compounds). Moreover, curve evolutions of main exhaust gas production (and thus O2 consumption) are linked to the MLR curve evolution. Indeed, when the PMMA sample is degraded (i.e. MLR increases), we can observe an oxygen consumption (O2 concentration decreases) and simultaneously an exhaust gas production (concentrations of volatile gases, such as CO, CO2, H2O, NO, NO2 and HCN, increase). Conversely, when the PMMA thermal degradation becomes less important (i.e. MLR decreases), O2 amount rises (i.e. reaches the initial ambient value of O2 concentration) and other major product concentrations decrease. Thus, the curve shapes of these different products (and the oxygen one by the mirror effect) correspond perfectly to the MLR ones. So, the physical and the chemical events arising during the PMMA thermal degradation process can be identified by correlating the MLR and gaseous compound curves with Table 3 given in Section 3. The product emission yields quantified during the different experiments are useful tools to extrapolate exhaust gaseous species production obtained from bench-scale (CC) to full-scale scenarios [42]. Table 6 gives the emission yields of the main products released during the PMMA decomposition process. The gaseous species emission yields (Yi), calculated as the ratio _ i ) and the PMMA between the exhaust product i mass flow rate (m MLR, are expressed by the following equation: Yi ¼

_i m MLR

ð13Þ

The mass flow rate of species i released during the combustion process (Eq. (14)) is represented by the product of the species mole fraction (Xi), the volumetric flow rate in the exhaust duct (V_ e ), the molar mass of species i (Mi) divided by the molar volume (Vm) given by Eq. (15), where R is the universal constant of a perfect gas, and T and P are the temperature and the pressure of the gas mixture in the CC exhaust duct, respectively;

Fig. 6. Comparison of CO2 concentrations (ppm) from gas analyzer and FTIR as functions of time (min) at 15, 30, 45 and 60 kW m  2.

_i¼ m

Xi V_ e Mi Vm

ð14Þ

Vm ¼

RT P

ð15Þ

During the burning process, only three species (CO, CO2 and H2O) present emission yields of high consistency. Emission yields of these species are found to be quite constant (taking into account uncertainties on measurements and calculations), regardless of the initial irradiance level used during the experiments. Indeed, in Table 6, CO and H2O emission yields decrease slightly from 0.008 to 0.006 gCO/gsample for CO and from 0.7 to 0:6 gH2 O =g for H2O when the irradiance level increases from 11 to 60 kW m  2, respectively. Conversely, CO2 emission yields increase slightly from 2.0 to 2:2 gCO2 =g when the external heat flux increases from 11 to 60 kW m  2, respectively. Moreover, for the irradiance level range studied (from 11 to 60 kW m  2), mean values are found to be equal to 0.007 gCO/gsample, 2:148 gCO2 =g and 0:658 gH2 O =g . Other chemical compounds (e.g. NOx, HCN or

458

J. Luche et al. / Fire Safety Journal 46 (2011) 451–461

Fig. 7. Major product concentrations (ppm) and mass loss rate (g s  1) as functions of time (min) at 15 (a), 30 (b), 45 (c) and 60 kW m  2 (d).

Table 6 Emission yields of main exhaust products released during PMMA combustion. External heat flux (kW m  2)

11 12 15 20 25 30 35 40 45 50 55 60 Mean

Chemical compound yields (gspecies/gsample)

CO/CO2 (%)

CO

CO2

H2O

0.008 70.001 0.008 70.002 0.008 70.002 0.008 70.002 0.006 70.001 0.006 70.001 0.006 70.001 0.006 70.001 0.006 70.001 0.006 70.001 0.006 70.001 0.006 70.001 0.007 70.001

2.034 70.407 1.977 7 0.395 2.126 7 0.425 2.169 7 0.434 2.069 70.414 2.160 70.432 2.182 7 0.436 2.206 70.441 2.216 7 0.443 2.217 7 0.443 2.201 70.440 2.221 7 0.444 2.148 7 0.430

0.720 7 0.144 0.758 7 0.152 0.748 7 0.150 0.694 7 0.139 0.655 7 0.131 0.633 7 0.127 0.640 7 0.128 0.582 7 0.116 0.649 7 0.130 0.611 7 0.122 0.614 7 0.123 0.587 7 0.117 0.658 7 0.132

light hydrocarbons), which have been measured during these experiments, have low emission yield values (i.e. o1 m g/gsample), regardless of the irradiance level considered. During this study in a CC, the averaged emission yield results found for CO2 and CO species (i.e. 2:148 gCO2 =gPMMA and 0.007 gCO/gPMMA) are comparable with those found in previous studies [1–4] (i.e.  2:54 gCO2 =gPMMA and  0.008 gCO/gPMMA), [14] (i.e.  2:12 gCO2 =gPMMA and  0.010 gCO/gPMMA), [15,16] (i.e.  3:23 gCO2 =gPMMA and  0.016 gCO/gPMMA) or [40–42] (i.e.  2:05 gCO2 =gPMMA and  0.010 gCO/gPMMA).

0.362 0.383 0.380 0.361 0.309 0.293 0.275 0.266 0.262 0.250 0.262 0.257 0.305

5. Discussion The thermal degradation study of virgin PMMA has allowed the computation of several thermal properties (Section 3), of this polymer, which are in good accordance with previous studies: theoretical CHF 5 kW m  2, experimental CHF¼ 11 kW m  2, DHg ¼2.34 kJ g  1, averaged heat release rate HRR  275–780 kW m  2 for external heat fluxes ranging from 11 to 60 kW m  2, respectively, and EHC 24.5 kJ g  1, regardless of the external heat flux chosen.

J. Luche et al. / Fire Safety Journal 46 (2011) 451–461

The chemical analysis (Section 4) of the exhaust gases by FTIR, NDIR and flame ionization detection (FID) has shown that only four products (CO, CO2, H2O and NO) and O2 were measured as main species. Indeed, low concentrations were measured for the other compounds, such as NO2, light hydrocarbons and HCN. Emission yields of all species quantified during this work have been calculated and only CO, CO2, H2O, NO and O2 have values of high consistency (i.e. higher than 1 mg/gsample). This was confirmed by the atomic balances for C and H presented in Table 7, which compares the C and H atom masses contained in the initial PMMA and those contained in the main gaseous products (i.e. CO, CO2 and H2O). Indeed, C atom masses from CO and CO2, and H atom masses from H2O are, in the majority of cases (taking into account measurement uncertainties), equal to the initial C and H atom masses of the PMMA sample, respectively. The results in Table 7 show that all initial C and H amounts (taking into account measurement uncertainties) are converted into H2O, CO and CO2 species, and no other compound (or in very small amounts, i.e. o1 mg/gsample, as measured for light hydrocarbons or HCN) containing high quantities of C and/or H atoms (e.g. heavy hydrocarbons, polycyclic aromatic hydrocarbons, soot, etc.) seems to be produced during thermal degradation of PMMA. Thus, the same amount of exhaust gas (i.e. the same value of exhaust gas emission yield) is produced during the thermal degradation process, regardless of the external heat flux applied to the PMMA sample (no residue found in the sample holder at the end of the different experiments). However, the MLR or HRR correlated with species concentration evolutions have shown that external heat flux has an influence on the intensities of sample degradation (transient MLR curves) or product emission/consumption (transient HRR and concentration curves). Indeed, when external heat flux increases from 11 to 60 kW m  2, the SMLR increases from 11 to 32 g s  1 m  2 and the HRR increases from 275 to 780 kW m  2, respectively. Moreover, external heat flux has an influence on thermal degradation and product evolution durations, which are faster for higher external heat (  10 min at 60 kW m  2 vs.  45 min at 11 kW m  2). As seen previously (Sections 3.3), in our experimental conditions in the CC, the PMMA sample thermal decomposition seems to take place according to four time ranges – more and less significant and separated, depending on incident heat flux – which are deduced from Fig. 3 and summarized in Table 3. From transient evolutions of the main released product concentrations, MLRs and HRRs, the four main time ranges where thermal decomposition of black PMMA takes

459

place and the associated physical and chemical phenomena can be identified as follows: – A first one (1st column in Table 3) corresponding to the start of thermal degradation with the beginning of ignition. During this first time range, a significant bubbling phenomenon at the PMMA sample surface occurred, and all concentrations (except for the O2 one, which decreased) and MLR increased strongly. – A second one (2nd column in Table 3) corresponding to a thick PMMA sample thermal degradation. During this second stage, a swelling phenomenon of the PMMA sample occurred and all exhaust gas amounts and MLR continued increasing (except for the O2 concentration, which decreased), but more slowly than previously. – A third one (3rd column in Table 3) corresponding to a thin PMMA sample thermal degradation. During this stage, the swelling phenomenon ended, and PMMA matrix surface cracking occurred. Thus, O2 consumption and effluent production, as well as the MLR value, reach an optimum value, which corresponds to the main peak (maximum for exhaust gases and minimum for oxygen) observed in Fig. 7. – A fourth one (4th column in Table 3) corresponding to the oxidation of the remaining amount of the PMMA sample up to the end of the degradation process (end of the ignition). During this time range, small CO, CO2 and other gas production (exhaust product concentrations and MLR values decrease) and small O2 consumption (oxygen concentration increases up to the initial value) were observed. At the end of this step, no residue was found in the sample holder.

6. Conclusion This study has dealt with the experimental characterization of thermal degradation and the combustion of a non-fire-retarded black PMMA. Experiments were carried out using a CC device under piloted ignition in order to characterize the external heat flux influence on the degradation process and on gaseous emissions. The theoretical thermal properties of PMMA degradation have been calculated. In addition, the exhaust gas analysis was performed using an FTIR spectrometer and a HORIBA gas analyzer for heat fluxes ranging from 10 to 60 kW/m2 (with an experimental CHF equal to 11 kW m  2). The gaseous species with the main concentrations and main emission yields were

Table 7 Atomic balance between the solid (source) and the exhaust gas (emissions) measured for the 11 irradiance levels studied. External heat flux (kW m  2)

11 12 15 20 25 30 35 40 45 50 55 60 Mean

PMMA mass (g)

168.7 72.5 171.6 171.6 72.5 169.2 73.6 173.3 72.5 167.5 72.6 168.0 72.1 168.2 71.1 170.4 72.6 171.6 73.0 168.8 72.5 168.7 72.5 169.6 72.9

Initial atomic mass (g)

Atomic balance in exhaust gases (g)

C (from PMMA)

H (from PMMA)

C (from CO and CO2)

H (from H2O)

99.7 7 1.5 101.4 101.4 7 1.5 100.07 2.1 102.4 7 1.5 99.07 1.5 99.3 7 1.2 99.4 7 0.7 100.77 1.5 101.4 7 1.8 99.8 7 1.5 99.7 7 1.5 100.47 1.5

13.3 7 0.2 13.6 13.6 7 0.2 13.4 7 0.3 13.7 7 0.2 13.2 7 0.2 13.3 7 0.2 13.3 7 0.1 13.5 7 0.2 13.6 7 0.2 13.3 7 0.2 13.3 7 0.2 13.4 7 0.2

94.1 718.8 93.1 718.6 100.1 720.0 100.7 720.1 98.3 719.7 99.1 719.8 100.4 720.1 101.6 720.3 103.4 720.7 104.2 720.8 101.8 720.4 102.6 720.5 99.9 720.0

13.5 7 4.5 14.5 7 4.8 14.3 7 4.7 13.1 7 4.3 12.6 7 4.2 11.8 7 3.9 12.0 7 3.9 10.9 7 3.6 12.3 7 4.1 11.6 7 3.8 11.5 7 3.8 11.0 7 3.6 12.4 7 4.1

460

J. Luche et al. / Fire Safety Journal 46 (2011) 451–461

CO2, CO, H2O, NO and O2, regardless of the irradiance level studied. The results have shown that emission yields presented quasi-constant values (measurement and calculation uncertainties taken into account) and were slightly influenced by an external heat flux increase. The main product evolutions have been correlated to the evolutions of the HRR and the SMLR. From the HRR and SMLR, EHC  24.5 kJ g  1 was computed for each irradiance level. Since the experiments were carried out in an atmospheric area without external ventilation, the oxygen supplied to the flame was only piloted by convection. Then, oxidative reactions were mainly piloted by flame temperature. As expected, the results obtained confirm the influence of irradiance levels on the degradation of the PMMA sample (i.e. on SMLR, HRR and product amounts in the gaseous phase). Using Thermogravimetric Analysis or a Tubular Furnace, where the reactor temperature was increased at a given heating rate, some previous studies [23,24,50–56] have shown that the non-charring black PMMA polymer was thermally degraded according to the following four-step reaction mechanism (where the term ‘‘gas’’ represents CO, CO2, H2O): – Depolymerization of PMMA to form MMA-derived monomers [23,24,50–56] by pyrolysis or oxidation reactions, Ox:=Py:

PMMA!MMA-derived compoundsþ Gas – Oxidation of the MMA-derived compounds and their transformations into oxygenated compounds (alcohol, carboxylic acid, etc.) [23,24,47–53] inside the solid matrix, O2

MMAderived compounds!Tar=Oxygenated compounds þ Gas

– Oxidation of the oxygenated compounds to form exhaust gases and small gaseous compounds (formaldehyde, methanol, acetone, acetylene, etc.), O2

Tar=Oxygenated compounds!Small gaseous compoundsþ Gas

– Small gaseous compound oxidation into gas, O2

Small gaseous compounds!Gas

This mechanism probably takes place during the entire PMMA thermal degradation process, and the four steps occur continuously on the whole duration of the CC experiments as soon as ignition starts. Using numerical simulation, the next step of this work will be to calculate the Arrhenius parameters of each reaction by using a parameter optimization method (e.g. genetic algorithm [57–59]) to demonstrate the validity of this PMMA thermal degradation mechanism in CC experiments under wellventilated air atmospheres.

Acknowledgments The authors are grateful for the financial support of the French MAIF foundation (‘‘Fondation MAIF’’) and the COMPFEU program to this work. References [1] B.T. Rhodes, J.G. Quintiere, Burning rate and flame heat flux for PMMA in a cone calorimeter, Fire Safety Journal 26 (1996) 221–240.

[2] D. Hopkins Jr, J.G. Quintiere, Material fire properties and predictions for thermoplastics, Fire Safety Journal 26 (1996) 241–268. [3] B.T. Rhodes, Burning Rate and Flame Heat Flux for PMMA in the Cone Calorimeter, NIST-GCR-95-664, National Institute of Standards and Technology, 1994. [4] J.G. Quintiere, B.T. Rhodes, Fire Growth Models for Materials, NIST-GCR-94647, National Institute of Standards and Technology, 1994. [5] T.-H. Tsai, M.-J. Li, I.-Y. Shih, R. Jih, S.-C. Wong, Experimental and numerical study of autoignition and pilot ignition of PMMA plates in a cone calorimeter, Combustion and Flame 124 (2001) 466–480. [6] S. Agrawal, A. Atreya, Wind-aided flame spread over an unsteady vaporizing solid, in: 24th Symposium (International) on Combustion, The Combustion Institute, Pittsburgh, PA, 1992, vol. 24, pp. 1685–1693. [7] C. Vovelle, J.-L. Delfau, M. Reuillon, J. Bransier, N. Laraqui, Experimental and numerical study of the thermal degradation of PMMA, Combustion Science and Technology 53 (1987) 187–201. [8] J.L. Jackson, Direct Measurement of Heat of Gasification for Polymethylmethacrylate, National Institute of Standards and Technology, 1986 NISTIR 88-3809, September. [9] P.A. Beaulieu, Flammability characteristics at heat flux levels up to 200 kW/ m2 and the effect of oxygen on flame heat flux, Ph.D. Thesis. Worcester Polytechnic Institute, 2005. [10] R. Magee and R. Reitz, Extinguishment of radiation augmented fires by water sprays, in: 15th Symposium (International) on Combustion, The Combustion Institute, Pittsburgh, PA, vol. 15, 1974, pp. 237–247. [11] C. Vovelle, R. Akrich, J.L. Delfau, Mass loss rate measurements on solid materials under radiative heating, Combustion Science and Technology 36 (1984) 1–18. [12] J.E.J. Staggs, The heat of gasification of polymers, Fire Safety Journal 39 (2004) 711–720. [13] A. Tewarson, R.F. Pion, Flammability of plastics—I. Burning intensity, Combustion and Flame 26 (1976) 85–103. [14] A. Tewarson, SFPE Handbook of Fire Protection Engineering, NFPA, 2002, Generation of Heat and Chemical Compounds in Fire, pp. 3–82 [Section 3, Chapter 4]. [15] J.R. Ebdon, B.J. Hunt, P. Joseph, C.S. Konkel, D. Pric, K. Pyrah, T.R. Hull, G.J. Milnes, S.B. Hill, C.I. Lindsay, J. McCluskey, I. Robinson, Thermal degradation and flame retardance in copolymers of methyl methacrylate with diethyl(methacryloyloxymethyl)-phosphonate, Polymer Degradation and Stability 70 (2000) 425–436. [16] D. Price, K. Pyrah, T.R. Hull, G.J. Milnes, J.R. Ebdon, B.J. Hunt, P. Joseph, Flame retardance of poly(methyl methacrylate) modified with phosphorus-containing compounds, Polymer Degradation and Stability 77 (2002) 227–233. [17] A. Tewarson, F.H. Jiang, T. Morikawa, Ventilation-controlled combustion of polymers, Combustion and Flame 95 (1993) 151–169. [18] J.L. Consalvi, Y. Pizzo, B. Porterie, Numerical analysis of the heating process in upward flame spread over thick PMMA slabs, Fire Safety Journal 43 (2008) 351–362. [19] R.E. Lyon, J.G. Quintiere, Criteria for piloted ignition of combustible solids, Combustion and Flame 151 (2007) 551–559. [20] J. Zhang, X. Wang, F. Zhang, A.R. Horrocks, Estimation of heat release rate for polymer–filler composites by cone calorimetry, Polymer Testing 23 (2004) 225–230. [21] D.J. Irvine, J.A. McCluskey, I.M. Robinson, Fire hazards and some common polymers, Polymer Degradation and Stability 67 (2000) 383–396. [22] V. Babrauskas, in: V. Babrauskas, S.J.. Grayson (Eds.), The Effects of FR Agents on Polymer Performance, Plastics, Part B. Chapter 12: Heat Release in Fires, Elsevier Applied Science, NY, 1992, pp. 423–446. [23] W.R. Zeng, S.F. Li, W.K. Chow, Preliminary studies on burning behavior of polymethylmethacrylate (PMMA), Journal of Fire Sciences 20 (2002) 297–317. [24] W.R. Zeng, S.F. Li, W.K. Chow, Review on chemical reactions of burning poly(methyl methacrylate) PMMA, Journal of Fire Sciences 20 (2002) 401–433. [25] B.S. Kang, S.G. Kim, J.S. Kim, Thermal degradation of poly(methyl methacrylate) polymers: kinetics and recovery of monomers using a fluidized bed reactor, Journal of Analytical and Applied Pyrolysis 81 (2008) 7–13. [26] Y.H. Hu, C.Y. Chen, Study of the thermal behaviour of poly(methyl methacrylate) initiated by lactams and thiols, Polymer Degradation and Stability 80 (2003) 1–10. [27] Y.H. Hu, C.Y. Chen, The effect of end groups on the thermal degradation of poly(methyl methacrylate), Polymer Degradation and Stability 82 (2003) 81–88. [28] ISO 5 660, Reaction to fire tests—heat release, smoke production and mass loss rate—Part 1 Heat release rate (cone calorimeter method), 2002. [29] ISO 17554:1998 International Standard – Fire Tests – Reaction to Fire – Mass loss measurement. [30] T. Hakkarainen, E. Mikkola, J. Laperre, F. Gensous, P. Fardell, Y. Le Tallec, C. Baiocchi, K. Paul, M. Simonson, C. Deleu, E. Metcalfe, Smoke gas analysis by Fourier transform infrared spectroscopy—summary of the SAFIR project results, Fire and Materials 24 (2000) 101–112. [31] ISO 19702, Toxicity testing of fire effluents—guidance for analysis of gases and vapours in fire effluents using FTIR gas analysis, ISC 13.220.01, 2006. [32] T. Hakkarainen, E. Mikkola, J. Laperre, F. Gensous, P. Fardell, Y. Le Tallec, C. Baiocchi, K. Paul, M. Simonson, C. Deleu, E. Metcalfe, Smoke Gas Analysis by Fourier Transform Infrared Spectroscopy, The SAFIR Project, European

J. Luche et al. / Fire Safety Journal 46 (2011) 451–461

[33]

[34] [35] [36] [37]

[38]

[39]

[40] [41] [42] [43] [44] [45] [46]

Commission Research Program Contract Number SMT 4-CT96-2136— Technical Research Centre of Finland, Copyright Valtron Teknillinen Tutkimuskesus, 1999. /http://www.vtt.fi/rte/firetech/research/htmlS Smoke analysis by Fourier transform infra-red spectroscopy (SAFIR). P. Fardell, E. Guillaume, Sampling and measurement of toxic fire effluents, in: Fire Toxicity [Chapter 11], ISBN 1 84569 502 X–ISBN-13: 978 1 84569 502 6, March 2010. L. Speitel, Fourier transform infrared analysis of combustion gases, Journal of Fire Sciences 20 (2002) 349–371. L. Speitel, DOT/FAA/AR-01/88 Fourier transform infrared analysis of combustion gases, Final Report, October 2001. A. Tewarson, D. Ogden, Fire behavior of polymethylmethacrylate, Combustion and Flame 89 (1992) 237–259. M.A. Delichatsios, Piloted ignition times, critical heat fluxes and mass loss rates at reduced oxygen atmospheres, Fire Safety Journal 40 (2005) 197–212. M.J. Spearpoint, J.G. Quintiere, Predicting the piloted ignition of wood in the cone calorimeter using an integral model—effect of species, grain orientation and heat flux, Fire Safety Journal 36 (2001) 391–415. Janssens, M.L.; Kimble, J.; Murphy, D., Computer tools to determine material properties for fire growth modeling from cone calorimeter data, in: Proceedings of Fire and Materials 2003, 8th International Conference, Conference Papers, Organised by Interscience Communications Limited. January 27–28, 2003, San Francisco, CA, pp. 377–387. V. Babrauskas, R. Peacock, Heat release rate: the single most important variable in fire hazard, Fire Safety Journal 18 (1992) 255–272. V. Babrauskas, Ignition Handbook, Fire Sciences Publishers, 0-9728111-3-3, 2003. V. Babrauskas, S. Grayson, Heat Release in Fires, Elsevier Science Publishers Ltd, Essex, England, 1992. M.L. Janssens, Measuring Rate of Heat Release by Oxygen Consumption, Fire technology 27 (1991) 234–249. P.A. Enright, Heat release and the combustion behaviour of upholstered furniture, Ph.D. Thesis, University of Canterbury, 1999. M. Janssens, SFPE Handbook of Fire Protection Engineering, NFPA, 2002 [Section 3, Chapter 4, Calorimetry], pp. 38–62. W. Thornton, The relation of oxygen to the heat of combustion of organic compounds, Philosophical Magazine and Journal of Science 33 (1917) 196–203.

461

[47] C. Hugett, Estimation of rate of heat release by means of oxygen consumption measurements, Fire and Materials 12 (2) (1980) 61–65. [48] V. Babrauskas, Development of the Cone Calorimeter—A Bench-Scale Heat Release Rate Apparatus Based on Oxygen Consumption.NBSIR82-2611, National Bureau of Standard, Gaithersburg, USA, 1982. [49] V. Babrauskas, Development of the cone calorimeter—a bench-scale heat release rate apparatus based on oxygen consumption, Fire and Materials 8 (1984) 81–95. [50] M. Comuce, T. Rogaume, F. Richard, T. Fateh, J. Luche, P. Rousseaux, Kinetics and mechanisms of the thermal degradation of poly(methyl methacrylate by thermogravimetry and Fourier-transform infrared spectroscopy analysis, in: Sixth International Seminar on Fire and Explosion Hazards, Weetwood Hall, Leeds, UK, April 11–16, 2010. [51] H. Arisawa, T.B. Brill, Kinetics and mechanisms of flash pyrolysis of poly (methyl methacrylate) (PMMA), Combustion and Flame 109 (1997) 415–426. [52] G. Lopez, M. Artetxe, M. Amutio, G. Elordi, R. Aguado, M. Olazar, J. Bilbao, Recycling poly-(methyl methacrylate) by pyrolysis in a conical spouted bed reactor, Chemical Engineering and Processing: Process Intensification 49 (10) (2010) 1089–1094. [53] J.E. Brown, T. Kashiwagi, Gas phase oxygen effect on chain scission and monomer content in bulk poly(methy1methacrylate) degraded by external thermal radiation, Polymer Degradation and Stability 52 (1996) 1–10. [54] Z. Jiang, W.K. Chow, J. Tang, S.F. Li, Preliminary study on the suppression chemistry of water mists on poly(methyl methacrylate) flames, Polymer Degradation and Stability 86 (2004) 293–300. [55] W. Kaminsky, M. Predel, A. Sadiki, Feedstock recycling of polymers by pyrolysis in a fluidised bed, Polymer Degradation and Stability 85 (2004) 1045–1050. [56] K. Smolders, J. Baeyens, Thermal degradation of PMMA in fluidised beds, Waste Management 24 (2004) 849–857. [57] C. Lautenberger, G. Rein, C. Fernandez-Pello, The application of a genetic algorithm to estimate material properties for fire modeling from bench-scale fire test data, Fire Safety Journal 41 (2006) 204–214. [58] C. Lautenberger, C. Fernandez-Pello, A model for the oxidative pyrolysis of wood, Combustion and Flame 156 (2009) 1503–1513. [59] C. Lautenberger, C. Fernandez-Pello, Generalized pyrolysis model for combustible solids, Fire Safety Journal 44 (2009) 819–839.