material fire properties and predictions for thermoplastics

ELSEVIER

Fire Safety Journal 26 (1996) 241-268 Copyright ~) 1996 Elsevier Science Limited

Printed in Northern Ireland. All rights reserved 0379-7112/96/$15.00

P I I : S 0 3 7 9 - 7 1 1 2 1 9 6 ) 0 0 0 3 3 - 1

Material Fire Properties and Predictions for Thermoplastics

D. Hopkins Jr & J. G. Quintiere*

Department of Fire Protection Engineering, University of Maryland, College Park, MD 20742, USA

(Received 4 January 1996; revised version received 23 April 1996; accepted 28 April 1996)

ABSTRACT

Ignition and burning rate data are developed for nylon 6/6, polyethylene, polypropylene and black polycast PMMA in a cone calorfmeter 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 /m 2 for nylon 6/6, 19kW/m 2 for polyethylene, l l k W / m 2 for polypropylene and 28 k W / m 2 for black PMMA is obtained for irradiation levels ranging from 0 to 90 k W / m 2. The burning rate model is shown to yield good accuracy in comparison to measured transient burning in the cone assembly. Copyright © 1996 Elsevier Science Ltd.

NOTATION

c Specific heat hc Convective heat transfer coefficient k Thermal conductivity L Heat of gasification m Mass q Heat flow

*Author to whom correspondence should be addressed.

241

242

T t

Y

D. Hopkins Jr, J. G. Quintiere

Temperature Time Space coordinate

Greek symbols

o~

6

E

P o"

Thermal diffusivity Thermal penetrat ion depth Heat of vaporization Heat of combustion Emissivity Density Stefan-Bol tzmann constant

Subscripts

c Convective cr Critical ext External fl Flame g Gas ig Ignition 0 Initial, ambient m Mean r Radiative s Steady v Vaporization

Superscripts

(") Per unit time ()" Per unit area

1 I N T R O D U C T I O N

Predicting the way a fire will behave under realistic conditions can be a challenging task. In order to examine the effects of fire on its surroundings and the hazard posed to occupants, information about the heat release rate, total energy and flame spread characteristics of the fire need to be known. Subsequently, it is desirable to have a means of determining fire growth and spread in terms of measurable material properties. A general model for predicting the burning rate of materials is needed to accomplish this.

Quintiere 1 developed a one-dimensional model which includes charring, vaporization, extinction, flame and heat conduction effects.

Material fire properties and predictions 243

However, unsteady solutions for the burning rate were not determined. Quintiere and Iqbal 2 developed a model that solves the one- dimensional unsteady heat transfer equations during the preheating and gasificatJion periods using an integral method. They assume a polyno- mial profile for the temperature within the solid to satisfy the heat transfer boundary conditions. However, the effects of flaming are not included in the solution.

The objective of this research is to develop practical transient burning rate moclels which utilize data obtained from the cone calorimeter. 3 The models will be dependent on the class of material, namely thermoplastic, charring, dripping and laminated. It is intended to first succeed for thermoplastic-like materials: materials approximated by constant surface temperature vaporization, which are large enough to be considered one-dimensional in behavior.

Mass loss data and surface temperatures have been measured for four commercial products as described in Table 1: nylon, polyethylene, polypropylene and black PMMA. Measurements were conducted under a cone calorimeter heating assembly for irradiances ranging from 0 to 90 k W / m 2. A schematic layout of the apparatus is shown in Fig. 1. The heater and sample holder configuration are identical to those prescribed by the ASTM standard, 3 with the exception that a small flame was used as a piloted ignition source as opposed to an electric spark. The gas pilot flame applies no heat to the sample, and data with an ASTM

TABLE 1 Identification of Materials Studied

Material Description Thermal Density b p (kg/m ~) diffusioity a a

(m2/s)

PMIVIA Polycast, black, 8.8 × 10 ~ 1190 polymethylmethacrylate, 2.5 cm

Nylon Polypepco, extruded, 1.2 X 10 -7 1169 unfilled, type 6/6, 2.5 cm

Polyethylene Allied Resinous 2.2 × 10 7 955 Products, Inc., type '0', 2.5 cm

Polypropylene Poly-Hi, clear, high- 6.7 x 10 -s 900 density, 2.5 cm

"Literature values.'2'~ 3 h Measured values.

244 D. Hopkins Jr, J. G. Quintiere

TYPICAL FLAME SHAPE

m m m m m m m m m m m m m

SAMPLE

-'LOT

KAOWOOL

Fig. 1. Schematic layout of cone heater assembly.

standard cone calorimeter showed no significant differences with ignition or burning rate. 4'5 Ignition and burning rate data are used to calculate effective material properties which can be used to estimate the transient burning rate of the materials under various exposures.


Calculated transient mass loss rates are compared to experimental data to assess the utility and accuracy of the model.

The ultimate goal of the research is to incorporate the burning rate models into fire growth simulations 6-9 to provide an assessment of hazard. It is first necessary to succeed at developing a technique which utilizes cone calorimeter data to predict the transient burning rate in the cone configuration, at least for thick thermoplastics and charring materials. The technique presented herein has been examined by Quintiere and Rhodes 4's for black polycast polymethylmethacrylate (PMMA). The model is based on the formulation outlined by Quintiere ~ and implemented by Quintiere and Iqbal 2 for nonflaming pyrolysis of a thermoplastic. Further validation of the technique using nylon 6/6, polyethylene and polypropylene is demonstrated in the present study. The primary objective is to see if the method used by Quintiere and Rhodes 4"s for PMMA is general for thermoplastic-like materials burning in the cone calorimeter.

2 THEORY

The model is one-dimensional and assumes that surface vaporization occurs at a specified temperature, Tv. The model can be extended to include an analysis for charring materials, but at the present time only thermoplastic (noncharring) materials will be considered. The model has been previously described 4"5 and only an overview will be presented here. The theory, approximate and intended for use with test data, should yield physically meaningful results for real materials.

2.1 Preheat ing to ignition

The one-dimensional unsteady heat conduction equation applies to the preheating period. Constant properties are assumed such that the governing equation is

a T c~2T - - = a (1) 0t c)y 2

with a constant initial temperature:

T = T~, at t = 0 (2)

Furthermore, considering convection at the surface (y = 0) and radiative heat loss,

aT q " - - k - - = = eCl"xt - h c ( T - T o ) - EorT 4 (3)

O y


A standard approximate integral solution is applied to formulate a solution to the above problem. Therefore, the solution to eqn (1) comes from integrating the equation between zero and some penetration depth, 6, where at y = ~,

T=T0

and OT - - = 0 ( 4 ) ay

These conditions define the penetration depth beyond which no heat transfer occurs. A quadratic profile is assumed for T such that the three conditions given by eqns (3) and (4) are satisfied. If q" is assumed to be constant, which is a reasonable assumption for large external irradiances, then the penetration depth 6 is approximately given as

a ---- 6v~t (5)

A more complete solution to the above problem has been shown by Abu-Zaid and Atreya m to yield

6 = a/12oa (6) e4

where e4 was shown to vary between 1.6 at 15kW/m 2 and 1.9 at 50kW/m z. Taking the asymptote as 2.0 yields eqn (5). Since the solution presented here is approximate, the accuracy sacrificed by using eqn (5) is assumed to be acceptable, although the error will be greatest at low external irradiances. Our motivation at these approximations is to achieve analytical results to ease data interpretations and practical solutions for real materials.

From the theory it can be shown that the time to ignition is

2 (Tig- To) 2 t,g =~(kpc) (0,) 2 (7a)

If the net heat flux to the surface is constant, this becomes

~" (Tig- T°)Z (7b) tig = ~- (koc) ((1,,) z

which is an exact result for this limiting condition. This approach also gives a method for determining the surface

temperature as a function of time. This is done implicitly by selecting T,, calculating the corresponding net heat flux, and using eqn (7a) to determine the time. This result also allows for the determination of a


critical flux for ignition, q'c'r, by extrapolating ignition data for (tig) -I/2 to zero. At this intercept,

1 4~txt = - [hc(Tig - To) -4- ~orT4g] ~ q"~ (8)

E

The critical flux for ignition is determined using eqns (3) and (7a). Once the critical flux is found, the ignition temperature can be determined using eqn (8). This process of using ignition data to match eqn (7a) will yield effective properties: kpc and Tig, and not necessarily 'real' properties. Also, we use eqn (7a) because it is consistent with the transient burning rate solution to follow.

2.2 Burning rate

The governing equations for the gasification period can be derived in integral form by selecting the appropriate control volumes for the vaporization plane and the solid. The governing equations for the burning rate follow from eqn (1) which governs conduction to the material below the vaporizing plane at a fixed temperature. Conse- quently, at y = 0, the vaporization plane:

and T = T~ (9)

_ / , o r = 4" - m"±n (lO) ay

where:

m" is the mass loss rate per unit area; AHv is the heat of vaporization; and 4" is the net surface heat flux.

It shotdd be noted that the net surface heat flux, 4", is different for the gasification period in eqn (10) than for the preheating period as defined in eqn (3). The net surface heat flux for the gasification period is

4 "= {4"x, + 4~ - {o'T{ (11)

where q"n is the flame heat flux, composed of incident radiant and convective heat fluxes. Assuming a quadratic profile for the temperature, which satisfies the boundary conditions supplied by eqns (2) and (9), and substituting the profile into eqn (10) yields

rh"AHv = c~" 2k T, - T ( - to) (12)


where the last term represents the transient conduction heat loss into the solid. If the total flame heat flux is assumed to be constant, the differential equation for mass loss rate can be solved exactly. 2 It can be shown that

/ - - / ig (~2 AHv [ (~ig - (~ ( ~s-- ~11 (13) - 6a L 6~ ln\6s - - ( ~ i g ] - I

where

and

2k L 6s = - - ~ , , , the steady value for 6 under burning

C q

L = AH, , + c(T,, - To), is the heat of gasification.

It follows that the steady-state mass loss rate is given by

(14)

(15)

rh~'= q" (16) L

It should be noted that in addition to the flame heat flux, other properties are needed to obtain a solution. These properties need to be derivable in a convenient manner consistent with the burning rate and ignition models. The properties are E, p, c, k, Tv, or Tig, and AHv or L.

2.3 Flame heat flux

For the present experiments, only the steady mass loss rate measurements are used to determine the flame heat flux. Rhodes ~ measured surface heat flux in his burning rate study with PMMA under the cone heater. In addition, a methane gas burner was used to simulate a fuel sample in the cone to more carefully examine the effect of the flame heat flux. It has been shown by Quintiere and Rhodes 4'5 that the flame heat flux for thermoplastic-like materials burning in the cone calorimeter can be constant. This was shown for PMMA and the methane simulator as long as the flame height is above the cone heater. This has good utility in aiding in the analysis of cone calorimeter data.

The constancy of the flame heat flux may be attributed to the shape of the flame for materials burning in the cone calorimeter. It can be shown that a long columnar homogeneous flame of height (L) to diameter (D) of greater than approximately 2 has a constant emissivity. 4"5 If the flame has a constant flame temperature and the convective heat flux is relatively small or constant, then the total flame heat flux for that material in the cone calorimeter is constant for all


cases whe re L / D > 2 . Also, the small flame width in the cone arrangement ensures that nearly all of the heater external flux reaches the sample surface ( > 95% for PMMA). 4"5 This is fortuitous for the cone arrangement. It makes the analysis of such cone data relatively simple. The main purpose of this study was to examine the procedure for three additional thermoplastics: nylon, polyethylene and polypropylene. Measurements of the flame heat flux were not at- tempted for these thermoplastics.

3 E X P E R I M E N T A L SET-UP AND P R O C E D U R E

Ignition and burning rate experiments for the thermoplastic materials were performed using a radiant cone heater assembly. The apparatus, shown in Fig. 1, consisted of a cone heater, a load cell, a methane pilot ignitor and a data acquisition system.

Three thermoplastic-like materials were selected for evaluation in this study based on availability from commercial retailers (see Table 1). The samples were nominally 10 cm (4 in.) x 10 cm (4 in.) x 2.5 cm (1 in.) thick.

Samples were placed on a standard cone metal holder in the horizontal orientation on a bed of Kaowool. The Kaowool was used to insulate', the back side of the specimen to minimize heat loss effects. In order to maintain only one-dimensional burning, cardboard was bonded to the sides of the samples to inhibit edge burning effects. Furthermore, aluminum foil was wrapped around the edges and the back of the sample to prevent dripping.

The experimental procedure consisted of exposing a sample (in the horizontal orientation) to a constant external irradiance from the cone heater assembly. The time to piloted ignition was measured and mass loss dwta recorded for each test.

The arrangement of the assembly was such that the top of the sample was initially located 1 in. below the base of the cone heater as is done for standard cone calorimeter tests? The load cell was oriented between two guide bars to ensure proper placement of the sample underneath the center of the cone heater assembly. A 1 in. methane flame, located on one edge approximately 0.5 in. above the surface of the sample, was used a's a pilot ignition source. Figure 1 shows the arrangement of the pilot and the assembly.

Once the cone was set to the desired constant external irradiance and the sample was prepared and situated on the load cell, the load cell was shifted into position underneath the cone. Simultaneously, the data


acquisition system was initiated. It should be noted that sliding the sample into place caused some bouncing of the load cell. This affects the mass loss recordings for the first few seconds of the experiment. This is not, however, expected to have adverse effects on the results since ignition nominally took longer that 10 s.

The ignition time is defined as the time at which a continuous flame is supported on the material surface. In some instances flashing occurred on the surface of the sample prior to sustained flaming. However, in all cases the ignition time was taken as the time at which flaming was sustained over the entire surface of the specimen.

In some experiments, a fine wire type K thermocouple (0.003 in. diameter) was mounted on the thermoplastic material surface to at tempt to measure the surface temperature at ignition. The thermocouple wire was placed on the sample surface by heating the wire so that it recessed into the surface of the material. Measurements of the surface temperature were recorded until the onset of burning. Figure 2 shows a typical result of the measured surface temperature for the

800

P o l y e t h y l e n e - 36 kW/m 2

Igni t ion at 126 s

Fig. 2.

600

o o

.=

: 400'

o

E

I - 200

El° B m

El El El

0 i I i

0 5 0 100 150 2 0 0

Time (s) Surface temperature results for polyethylene with a 3 6 k W / m 2 external

irradiance.


preheatiing period for polyethylene exposed to a 36 kW/m 2 external irradiance. The temperature at the onset of the sudden rise in temperature is defined as the ignition temperature (T~g). These results will be reported, but sufficient work was not done to ensure their high accuracy.

Experiments to determine the mass loss rate of the materials were performed concurrently with the ignition experiments. The mass loss of the samples was recorded using a load cell. The mass loss readings were recorded every 2 s for approximately a 1200s period. The transient mass loss rate is found using a five-point least-squares fit of the mass loss data.

4 ANALYSIS OF RESULTS

The following sections outline the procedure for utilizing test data to determine effective material properties for use with the theory to predict ignition times and transient burning rates. The analysis of the results is done in a manner consistent with the protocol outlined by Quintiere and Rhodes 4'5 for PMMA. The intent is to see if the protocol is general for all thermoplastic-like materials. Results obtained from the thermoplastic experiments are used to obtain the required properties to utilize the ignition and burning rate models. While most of the necessary properties are deduced from the experimental data, the density is determined by measurement, and the thermal diffusivity, which is assumed to be approximately constant for a given material, is obtained from the literature. 12'13 The density and the thermal diffusivity for the thermoplastics tested are shown in Table 1.

The approach used to obtain the needed properties will be demonstrated for nylon 6/6. The approach is identical for the analysis of the other materials. The properties for all of the materials can be seen in Table 2.

4.1 Ignition

In order to predict the ignition time and the critical flux for ignition, the ignition temperature and the thermal inertia must be determined. Once this is accomplished it is possible to determine the critical flux and the ignition time as a function of the external irradiance.

4.1.1 Ignition temperature and thermal inertia Figures 3(a), (b), (c) and (d) show plots of (ignition time) -'/2 as a


TABLE 2 Deduced Properties

Property Nylon Polyethylene Polypropylene PMMA Units

kpc 0.87 1.8 2.2 2.1 kj2/m 4 s K 2 kpc (li terature) t4 - - 0.638 0.367 0.339, 0.365 kj2/m 4 s K 2 k 3.3)<10 -4 6.4)<10 -4 3.8)<10 -4 4.3)<10 -4 k W / m K

c 2.3 3.0 6.3 4.1 kJ /kg K 0", 14 9 5 4 k W / m 2 Tig (model) 380 300 210 180 °C L 3.8 3.6 3.1 2.8 kJ /g L (literature) Is 2.4 2.3 2.0 1.6 kJ /g q"~,. n 20 19 11 28 k W / m 2 4~ 30 25 14 37 k W / m 2

The values of L from Tewarson are shown not to necessarily suggest differences due to accuracy, but to suggest differences due to commercial product variations. This has been illustrated for P M M A . 4"5

function of external irradiance for nylon 6/6, polyethylene, polypropylene and PMMA, respectively. The line through the data represents the best fit of the data, favoring data, in some cases, where the results are more accurate when ignition takes longer. Moreover, the protocol requires an extrapolation to the critical flux. The best-fit line in Fig. 3(a) gives an intercept of approximately 14 kW/m 2, which represents the critical flux for ignition for nylon 6/6. The values for the critical flux for ignition, 0'c'r, for the other materials are shown in Table 2. Using the value for the critical flux, eqn (8) can be used to determine the ignition temperature, Tig, given To = 20°C, hc---10 W/m2K (determined from a standard analysis for natural convection over a flat plate), and E = 1.0. This yields an ignition temperature for nylon 6/6 of T~g = 380°C. We emphasize that this is a 'model' temperature.

Using the slope of the line in Fig. 3(a), along with eqns (3) and (7a), the thermal inertia can be calculated as follows. From eqns (3) and (7a),

_ _ _

/igm ~/~ pc(Tig-To)_] [ ~kpc(Tig-To)] (17,

Subsequently, from eqn (17), the slope of the line in Fig. 3(a) is

1 Slope = 2 (18) kpc (Tig - To)

I

L

Fig. 3.

(a)

Material fire properties and predictions

Nylon 253

0.3

T 0t~

L,

O.2-

0.1

0.0 0

I 10

[]

m m

" I ~ 1 ~ i I i i i 20 30 40 50 60 70 80 90

External f lux (kW/m 2)

0.3

0.2

0.1

0.o

(b) Polyethylene

0 100

m

j 323. 20 40 60 80


Ignition data for: (a) nylon 6/6; (b) polyethylene, (c) polypropylene: (d) PMMA.

(Continued overleaf)

254

(c) 0.3


Polypropylene

I ( l )

v

L

0.2"

0.1

0.0 0

m m

[]

• i I I i

10 20 30 40 50 60 70

0.3

0.2

0.1

0.0 0


P M M A

(d)

O NPrisT~tandard 7

Slope = 0.005053

20 40 60 80

External flux (kW/m 2)

Fig. 3. (Continued).


Equation (18) can be solved to determine the thermal inertia:

koc = ~ Slope (T~ - To)

Therefore, for nylon 6/6 the slope from Fig. subsequently the thermal inertia is

kJ 2 kpc = 0.874 m%K2

Again, this is a 'model' deduced property.

(19)

3(a) is 0.00364 and

4.1.2 Thermal conductivity and specific heat Knowing the thermal inertia, kpc, as was determined above, the density, p, as measured and the thermal diffusivity, a = k/pc (Table 1), which waries only slightly with temperature, the thermal conductivity, k, and the specific heat, c, can be determined.

For nylon 6/6, with p = l169kg/m 3 and ot = 1.24× 10-Tm2/s, the thermal conductivity and the specific heat are

k = 3.29 × 10 -4 kW/m K

and

c = 2.27 kJ/kg K

The thermal diffusivity is the only property that must be indepen- dently determined from the test protocol. However, this is usually given for commercial plastics. It should be recalled that k- and c-values are needed for the burning rate formula, eqns (12) and (13).

4.1.3 Property results Table 2 contains the deduced property values for the three materials of this study and the PMMA from the previous related study. 4"5 In addition, we list values for kpc taken from Thomson and Drysdale 14 as generic literature values. The 'model '-deduced values for kpc are higher in all cases. This is a model characteristic, but also is consistent with large increases in temperature above ambient. It is expected that kpc for solids will increase with temperature, i.e. A(kpc)~ (AT). 2

Table; 3 compares the deduced 'model' ignition temperatures and compares them to measured values. The results for similar materials are also included from Thomson and Drysdale. ~4 In general, ignition temperature measurements vary with the heat flux level and the wavelength distribution, and tend to produce lower ignition temperatures at lower heat fluxes. 4'5"14 The model temperatures are lower than measured results in all cases, and model dependent, i.e. they make the


TABLE 3 Model and Measured Ignition Temperatures

Material Model ignition temperature (°C)

Measured ignition Measured ignition temperature (°C) this temperature (°C) from study and Quintiere Thomson and

and Rhodes 4 Drysdale j4

Nylon 6/6 380 ~500 - - Polyethylene 300 315-330 360-367 Polypropylene 210 250-360 331-340 PMMA 180 250-3554 288-341

model work. Since the burning rate approximate solution is dependent on the ignition model, we will use these ignition parameters in the prediction of ignition and burning rate to follow.

4.2 Burning rate

Since most thermoplastic-like materials approximate vaporizing solids, it is possible to represent the steady-state mass loss data using eqn (16). Therefore if the flame heat flux is assumed to be constant, which has been shown to be the case for thermoplastic-like materials burning in the cone calorimeter, 4'5 then eqn (16) can be written as

rh" { 1 ~.,, (4,'~- Eo-T~) (20) = ~ ~)q~x, + L

Subsequently, a plot of the steady-state mass loss rate data as a function of the external flux has utility in determining the heat of gasification and the total flame heat flux. Figures 4(a), (b), (c) and(d) show plots of steady state mass loss rate vs external irradiance for nylon 6/6, polyethylene, polypropylene and PMMA, respectively.

4.2.1 Effective heat of gasification and total flame heat flux Equation (20) suggests that the slope of the fit to the data in Figs 4(a), (b), (c) and (d) represents the inverse of the heat of gasification, 1/L. Subsequently, for nylon 6/6,

= 0.26456 kJ

Therefore

kJ L = 3.78 - -

g

Fig. 4.

E

¢,

_c

t ~

I E

_=

¢/)

O

¢/} 0 ' }

30

20

10

(a)


N y l o n

257

l I i i i i i I

0 10 20 30 40 50 60 70 80 90

Heat f lux (kW/m 2) Polyethylene

(b)

30

i n

20 '

10 ¸ 27916 = . g/kJ

0 0

u i m m u u I m

10 20 30 40 50 60 70 80 90

Heat f lux (kW/m 2)

Steady-state mass loss rate as a function of external irradiance for: (a) nylon 6/6: (b) polyethylene: (c) polypropylene: (d) PMMA.

(Continued overleaf)

258

I

E

¢)

0

if)

50

4O

I E

30

~ 20 ._o

ffl

10

30

20'

10

(c)


P o l y p r o p y l e n e

I O / I I ! ! |

10 20 30 40 50 60

(d)

Heat f lux (kW/m 2)

P M M A

7 0

I I I I I I I

10 20 30 40 50 60 70

Heat flux (kW/m 2)


80


Furthermore, the intercept of the fit to the data in Fig. 4(a) is

qg - ~ rT~

L

For nylon 6/6 this is,

qa - Eo-T { g L - 5.32 mS s

(21)

Therefore, for an average vaporization temperature taken to be the deduced value of the ignition temperature (380°C) and an emissivity, ~, assumed to be 1.0, the flame heat flux is estimated using eqn (21). For nylon the flame heat flux is

kW qg = 30.4 m2

Table 2 shows the heat of gasification and the estimated flame heat flux for the four materials. In addition, the net flame heat flux is shown. The net flame heat flux is q f " -~ t rT 4, and consequently it does not depend on the model ignition temperature. The net flame heat flux is the mallerial's contribution to burning in the cone. It is seen that PMMA is the strongest, and polypropylene is the weakest.

4.2.2 Transient mass loss rate The mass loss rate per unit area can be calculated using the properties in Tables 1 and 2 along with eqns (12) and (13). The convective heat transfer coefficient, he, was taken to be 0.01 kW/m 2 K.

Figures 5(a), (b) and (c) show the calculated and experimental mass loss rates for nylon exposed to external irradiances of 31, 50 and 75 kW ra 2, respectively. It should be noted that at low external heat fluxes there are discrepancies in the calculated vs the experimental ignition times for nylon 6/6. This can be seen for the exposure to a 31 kW/m 2 irradiance, shown in Fig. 5(a). The reason for the discrepancy is that at low heat fluxes (q~x, < 35 kW/m 2) a bubble formed on the surface of the nylon sample, thus preventing the release of pyrolysates. Ignitio~L did not occur in these cases until the bubble over the surface broke. This phenomenon was only see in the nylon experiments and only at low external heat flux exposures. Once the nylon samples ignited they quickly reached a steady-state burning rate. While the nylon was burning small bubbles from the surface could be seen bursting and burning in the air near the sample. With the exception of the erroneous ignition time shown in Fig. 5(a), the calculated mass loss rate results for nylon are in good agreement with the experimental data. Note that the ignition times are also predicted.

40

i Od

E

q}

0

q~ t~

30

20

10

0

(a)


N y l o n 31 k W / m 2

o Experiment ] Calculated

260

0 200 400 600 800 1000 1200

T ime (s)

N y l o n (b) 50 k W / m 2

40

Fig. 5.

I Od E

o oO

t~

30

2°t 0 200 400 600 800 1000 1200

T ime (s) Transient mass loss rates for nylon 6/6 exposed to: (a) 31 kW/m2: (b)

50 kW/m2: and (c) 75 kW/m 2 cone irradiance. (Continued opposite)


Nylon 75 kW/m 2

(c )

40

~ ' 30

E a ,

20

o ==

lo o Experiment

0 I I I I I I

0 2 0 0 4 0 0 6 0 0 8 0 0 1 0 0 0 1 2 0 0

Time (s)

Fig. 5. (Connnued).

The quick rise to a steady burning rate after ignition as well as an approximately constant flame height (which is much taller than the top of the cone heater) is indicative of the burning behavior of all of the thermoplastic materials tested. However, the model is too approximate to pred:ict the critical mass flux at ignition ( - 1 - 5 g/m 2 s). Polyethylene and polypropylene both seemed to become a pool of liquid inside the aluminum foil while the samples were undergoing steady burning. In some instances the material could be seen boiling.

Calculated and experimental mass loss rates for polyethylene exposed to external irradiances of 36, 70 and 87 kW/m 2 are shown in Figs 6(a), (b) and (c ) , respectively. Scatter can be seen in some of the polyethylene experimental data. The exact reason for the fluctuations is not clear. Despite the scatter, the calculated mass loss rate results for polyethylene appear to be in good agreement with the experimental data. "['he increase at 70kW/m 2 after 900 s is a substrate insulation effect not taken into account in the model.

Calculated and experimental mass loss rates for polypropylene

40

Fig. 6.

J

E

0

o9

30 ¸

20

10

0

(a)


Polye thy lene 36 kW/m 2

0 0

0 200 400 600 800 1000 1200

T ime (s)

P o l y e t h y l e r ~ e (b) 70 k W / m 2

40

~ 30

%

20

0

Calculated I

i i i i

262

0 200 400 600 800 1000

T ime (s) Transient mass loss rate for polyethylene exposed to: (a) 36 kW/m2: (b)

70 kW/m 2" and (c) 87 kW/m 2 cone irradiance. (Continued opposite)


m

O9 o9 o~

(e)

0

Polyethylene 87 kWlm 2

10'

40

ff 30

ao

0

0

8 0 0 0

o Experiment

" - - - Calculated

0 I i i

0 2 0 0 4 0 0 6 0 0 8 0 0

T i m e (s)


exposed to external irradiances of 27, 50 and 61 kW/m 2 are shown in Figs 7(a), (b) and (c), respectively. The calculated mass loss rate results appear to be in good agreement with the experimental data.

Calculated and experimental mass loss rates for PMMA exposed to external irradiances of 25, 50 and 75 kW/m 2 are shown in Figs 8(a), (b) and (c), lrespectively.

It should be noted that in Figs 5(c), 6(b), 6(c) and 7(b) the mass loss rate rises above the steady value after some duration. This rise in mass loss rate is caused by an increase in temperature of the material associated with the fact that the steady-state penetration depth is greater than the material is being heated. Because the sample is insulated with Kaowool on the back side, little or no heat is lost through the back, and consequently the mass loss rate increases. This effect is shown by eqn (12), where for a perfect insulator the second term goes to zero. This phenomenon is expected to have been present in all of the experiments, prior to burnout, should the tests have been allowed to proceed long enough.

264

40

(a)


Polypropylene 27 kW/m 2

I

E v

t~ O

O'3

Fig. 7.

30

20

"~" 30 I

E

~ 2o

o

~ lO

o, 0 200 1000 1200

(b) 4o

0 0

400 600 800

T ime (s)

Polypropylene 50 kW/m a

0

o Experiment m" - -Ca lc .

I I I I I

2 0 0 4 0 0 6 0 0 8 0 0 1 0 0 0 1 2 0 0

• T i m e ~s) Transient mass loss rate for polypropyLene exposed to: (a) 27 kW/m 2, (b)

50 kW/m2; and (c) 61 kW/m 2 cone irradiance. (Continued opposite)


I

E

o

( / } ¢/}

o n

¢0

40

30

(c) Polypropylene 61 kW/m 2

o Experiment

Calculated [

0 i i i i

0 200 400 600 800 1000

T ime (s)


5 CONCLUSIONS

The modeling prescription developed by Quintiere and Rhodes, 4'5 which c~n utilize cone calorimeter data to derive useful properties needed to predict ignition and transient burning rates for thermoplastic- like materials, has been shown to yield good results. The level of accuracy has been demonstrated for nylon 6/6, polyethylene, polypropylene and PMMA. The simplicity of the ignition model is advantageous and can be used to infer the critical flux for ignition without direct measurement.

The inference of a constant flame heat flux to the surface for thermoplastic-like materials burning in the cone calorimeter has been shown to yield good results. Values for the net flame heat flux were found to be 20 kW/m 2 for nylon, 19 kW/m 2 for polyethylene, 11 kW/m 2 for polypropylene and 28 kW/m 2 for PMMA. The constancy of the flame heat flux appears to be an attribute of all thermoplastics burning in the cone calorimeter due to the long column-like shape of the flame.

For all of the thermoplastic materials examined thus far, the model ignition temperature has been found to be less than the measured

40

Fig. 8.

E

0

0

30

20

10

(b)

(a)


P M M A 25 k W / m 2

- 8

0

9 o Experiment

Calculated

i i i I

O0 2 0 0 3 0 0 4 0 0 5 0 0

0 0

o o e o 8 o °

40

30

20

10

0 0

T ime (s)

P M M A 50 k W / m 2

6 0 0

266

8

I o [ o Experiment

I I o ~ Calculated

! ! ! ! !

1 0 0 2 0 0 3 0 0 4 0 0 5 0 0

I 6 0 0

Time (s) Transient mass loss rate for black P M M A exposed to: (a) 25 kW/m2; (b)

50 kW/m2; and (c) 75 k W / m 2 cone irradiance. (Continued opposite)

e~ I

o

Q

40

30

20

10


P M M A 75 k W / m 2

(c)

o _ _ _ o oo_ = =. °

¢ 0

0

o

J f [ o Experiment - " " Calculated

! i i I I

0 100 2 0 0 3 0 0 4 0 0 5 0 0

Time (s) 6 0 0

267


ignition temperature. This is attributed to the nature of the approximate solutions and underscores the fact that these properties are effective modeling parameters. Use of the calculated ignition temperatures in the model has been shown to yield good results, and should be applicable in more general calculations provided the same approximate formulation is used. An application to PMMA pool fires was used to assess the flame radiation heat flux as a function of pool diameter. 16

R E F E R E N C E S

1. Quintiere, J.G., A semi-quantitative model for the burning of solid materials. NIST-4840, National Institute of Standards and Technology, June, 1992.

2. Quintiere, J.G. & Iqbal, N., An approximate integral model for the burning rate of thermoplastic-like materials. Fire Mater., 18 (1993) 89-98.

3. Standard test method for heat and visible smoke release rates for materials and products using oxygen depletion. ASTM E 1354, A S T M Fire Test Standards, 4th edn. American Society for Testing and Materials, Philadel- phia, PA, 1993, pp. 968-9.


4. Quintiere, J.G. & Rhodes, B.T., Fire growth models for materials. M.S. Thesis, Department of Fire Protection Engineering, University of Mary- land at College Park, January, 1994.

5. Rhodes, B.T. & Quintiere, J.G., Burning rate and flame heat flux for PMMA in a cone calorimeter. Fire Safety J. 26(3) (t996) 221-240.

6. Wickstrom, U. & Goransson, U., Prediction of heat release rates of surface materials in large-scale fire tests based on cone calorimeter results. ASTM J. Testing Evaluat., 15(6) (1987).

7. Karlsson, B., Modeling fire growth on combustible lining materials in enclosures. Report TVBB-1009, Department of Fire Safety Engineering, Lund University, Lund, Sweden, 1992.

8. Quintiere, J.G., A simulation model for fire growth on materials subject to a room-corner test. Fire Safety J., 18 (1992) 313-41.

9. Quintiere, J.G., Haynes, G. & Rhodes, B.T., Applications of a model to predict flame spread over interior finish materials in a compartment. In Int. Conf. for the Promotion of Advanced Fire Resistant Aircraft Interior Materials, FAA Technical Center, Atlantic City, N J, March 1993.

10. Abu-Zaid, M. & Atreya, A., Effect of water on piloted ignition of cellulosic materials. NIST-GCR-89-561, National Institute of Standards and Technology, February 1989.

11. Rhodes, B.T., Burning rate and flame heat flux for PMMA in the cone calorimeter. Department of Fire Protection Engineering, University of Maryland at College Park, May 1994.

12. Modern Plastics Encyclopedia, Vol. 48, No. 10A, ed. S. Gross. McGraw- Hill, New York, 1971.

13. Steckler, K.D., Kashiwagi, T., Baum, H.R. & Kanemaru, K., Analytical model for transient gasification of non-charring thermoplastic materials. In Proc. 3rd Int. Symp. on Fire Safety Science, ed. G. Cox & B. Landlord. Elsevier, Oxford, 1991, pp. 895-904.

14. Thomson, H.E. & Drysdale, D.D., Flammability of plastics I: ignition temperatures. Fire Mater., 11 (1987) 163-172.

15. Tewarson, A., Generation of heat and chemical compounds in fires. In SFPE Handbook of Fire Protection Engineering, 2nd ed., ed. P.J. DiNenno. NFPA, SFPE, Boston, MA, 1995, pp. 3-69.

16. Quintiere, J.G., Hopkins, D., Iqbal, N. & Rhodes, B., Thermoplastic pool fires. In Symp. on Thermal Science and Engineering in Honor of Chancellor Chang-Lin-Tiem, University of California, Berkeley, CA, 1995.

material fire properties and predictions for thermoplastics

Documents