Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Fire Properties of Materials Archie Tewarson* ABSTRACT Fire properties of materials associated with the pyrolysis, ignition, combustion, fire propagation, and flame extinction processes are discussed. The relationships between the fire- hardening of materials and fire properties are enumerated. Fire-hardening is defined as a process where resistance to pyrolysis, ignition, combustion, and fire propagation is increased, and release rates of heat and fire products are decreased. INTRODUCTION Flammability is an interaction of pyrolysis, ignition, combustion, fire propagation, and flame extinction processes. The first four processes are brought about by the heat exposure of the material. The heat exposure must be of sufficient strength to satisfy the requirements of the pyrolysis process. Pyrolysis is generally an endothermic process, characterized by the softening, melting, discoloration, cracking, decomposition, and vaporization and so forth of the material and release of products (i.e., smoke, toxic, and corrosive pyrolysis products). The boundary of the pyrolysis process is defined as the pyrolysis front. Ignition is a process in which the pyrolysis products mix with air and form a combustible mixture, and the mixture ignites by itself (auto-ignition) or is ignited by a flame, a hot object, an electrical spark, or similar means (piloted ignition). Combustion is a process in which the pyrolysis products react with oxygen from air, with a visible flame (flaming combustion). Heat and products (i.e., smoke and toxic and corrosive combustion products) are released in this process. Fire propagation is a process in which the pyrolysis front, accompanied by the flaming or nonflaming combustion process, moves beyond the point of origin at a certain rate, defined as the fire-propagation rate. Heat and products (i.e., smoke and toxic and corrosive combustion products) are released at an increasing rate during the propagation process. Flame extinction is a process in which the pyrolysis, ignition, combustion, and fire- propagation processes are interrupted by external agents such as water, Malone, or alternatives. Heat and products are released at a decreasing rate until flame extinction. Pyrolysis products continue to be released past the flame extinction as long as the heat within the material continues to satisfier the requirements of the pyrolysis process. Flammability Section, Factory Mutual Research Corporation, Norwood, Massachusetts. 61
62 Improved Fire- arm Smoke-Resistant Materials The release of heat and products (i.e., smoke and toxic and corrosive pyrolysis and combustion products) are hazardous to life and property. Hazard due to smoke and toxic and corrosive products is deemed as nonthermal hazard ~ewarson, 1992). Hazard due to heat (i.e., high temperature and radiation) is defined as thermal hazard (Tewarson, 1992~. For protection of life and property from fires, materials need to be fire-hardened, and active and passive fire projections need to be provided. Fire-hardening is defined as a process in which resistance to pyrolysis, ignition, combustion, and fire propagation is increased, and release rates of heat and fire products are decreased. The fire-hardening requirements for the materials are considered in terms of the fire properties listed in Table I. Fire-hardening can be achieved by several techniques of modifying the fire properties. TABLE 1 Fire Properties of Materials to Assess Degree of Fire-Hardening Fire Property Description of the Fire Property Pyrolysis Heat of gasification (AH') Surface re-radiation loss (q ',,) Yield of a product Product generation parameter Critical heat flux, (q ',,) Thermal response parameter Flame heat flux (q f3 Net heat of complete combustion (AHT) Chemical heat of combustion HAHN Convective heat of combustion (AHOY) Radiative heat of combustion (AH,~ Yield of a product Energy required to pyrolyze a unit mass of a material origimally at ambient temperature Heat lost to the environment from the hot surface Amount of a product generated per unit mass of a material pyrolyzed Amount of a product generated in pyrolysis per unit; amount of energy required to pyrolyze a unit mass of a material Ignition Process Minimum heat flux at or below which a flammable vapor-air mixture is not created Ease of in-depth penetration of the thermal wave and time delay to reach the ignition temperature Combustion Process Heat flux transferred from the flame back to the surface Amount of energy released in the complete combustion of a unit mass of a material pyrolyzed with water as gas Amount of energy actually released in a fire from the combustion of a unit mass of a material pyrolyzed Component of the chemical heat of combustion carried away from the flame by flowing combustion product-air mixture Component of the chemical heat of combustion transmitted away from the flame by radiation Amount of a product generated in the combustion per unit mass of a material pyrolyzed
Archie Tewarson J TABLE 1 (continued) 63 Fire Property Description of the Fire Property Heat release parameter Product generation parameter Fire-propagation index Visibility through smoke (not defined) Smoke damage (not defined) Toxic effects of products (not defined) Corrosion damage by products- · · . corrosion index Amount of energy generated in combustion per unit amount of energy required to pyrolyze a unit mass of a material Amount of a product generated in combustion per unit; amount of energy required to pyrolyze a unit mass of a material Fire Propagation Extent and rate of fire propagation beyond the ignition zone Nonthermal Damage Maximum distance over which an observer can see Smoke damage due to discoloration, smell, or electrical malfunction Toxic effects of products on humans Rate of corrosion per unit mass concentration of a material pyrolyzed PYROLYSIS When a material is exposed to heat flux, pyrolysis products are generated. The rate of generation of the pyrolysis products is defined as the mass pyrolysis rate (Tewarson, 1988, 1994) m"p = ~ or . ~ (1) where m p is the mass pyrolysis rate in (kg/m2 s), it e is the external heat flux (kW/m2), if ,~ is the surface re-radiation loss (kW/m2), and AH5 is the heat of gasification (MJ/kg). The fire-hardening of materials requires that the values of surface re-radiation loss and heat of gasification be as high as possible. Heat of Gasification For a meldng type of material, the heat of gasification is expressed as: T m AHg = |CpradT~ Him + T., T v |cp,ldT + AHv T nit (2)
64 Improved Fire- arm Smoke-Resistant Materials where IBM add AHv are the heats of melting and vaporization at the respective melting and vaporization temperatures in MI/kg; Cp a, and Cp ~ are the specific heats of the solid and molten solids in MJ/kg, respectively; and Ta, To, and Tv are the ambient, melting, and vaporization temperatures in K, respectively. For materials that do not melt, but sublime, decompose, or char, Equation 2 is motived accordingly. Table 2 lists examples of the heat of gasification values taken from Tewarson 198S, 1994. The values are measured by differential scanning calorimetry (DSC) and by the mass pyrolysis technique using the Factory Mutual Research Corporation (FMRC) Flammability Apparatus shown in Figure 1. Modifications in the pyrolysis behavior of the materials to increase the Cp, AHm, and AHv values and the melting and vaporization temperatures would increase the heat of gasification and reduce Me mass pyrolysis rate (Equation I) and other related fire properties. Surface Re-Radiation Loss Surface re-radiation loss is proportional to the fourth power of the pyrolysis temperature of the material. Stronger chemical bonds and pyrolysis mechanisms favoring retention of carbon in the solid phase (charring) would result in higher pyrolysis temperature and surface re- radiation. Mass pyrolysis rate decreases with increase in the surface re-radiation loss (Equation I). Table 2 lists examples of the surface re-radiation loss taken from Tewarson 1988, 1994. The values are quantified by the mass pyrolysis technique in the FMRC Flammability Apparatus (Figure I). COMBUSTION In the combustion process the pyrolysis products burn with air; a flame is established over the surface; and heat transferred from the flame back to the surface sustains the combustion process, with or without the external heat flux. For the combustion process, Equation ~ is expressed as (Tewarson, 1988, 19941: n _ d/ e+ ~ f ~ or OHS (3) where 7h p is the mass pyrolysis rate in the combustion process (kg/m2 s), and Of is the flame heat flux transferred back to the surface (kW/m21. The f~re-hardening of materials requires that the flame heat transferred back to the surface be reduced as much as possible. Results from numerous small- and large-scale fires show that, as the surface area of the burning material increases, the flame radiative heat flux increases and reaches an asymptotic limit, whereas the flame convective heat flux decreases and becomes much smaller than the flame radiative heat flux at the asymptotic limit (Hottel, 19591. In small-scale experiments with fixed surface area, flame radiative heat flux increases and flame convective heat flux decreases with increase in the oxygen mass fraction (Ye), as shown in Figure 2 (Tewarson et al., 1981~.
Archie Tewarson TABLE 2 Surface Re-Radiation Loss and Heat of Gasification of Polymers 65 Heat of Gasification (MJ/kg) Surface Re-Radiation Mass Pyrolysis Polymer Loss (kW/m2) Technique. DSC Polypropylene 15 2.0 2.0 Polyethylene (low density) 15 1.8 1.9 Polyethylene (high density) 15 2.3 2.2 Plasticized polyvinylchoride (PVC), LOIb = 0.20 10 2.5 Plasticized PVC, LOI = 0.30 - 2.1 - Plasticized PVC, LOI = 0.35 - 2.4 - Rigid PVC, LOI = 0.50 - 2.3 - Polyoxymethylene 13 2.4 2.4 Polymethylmethacrylate 11 1.6 1.6 Polystyrene (granular3 13 1.7 1.8 Expanded polyurethane (flexible) 16-19 1.2-2.7 1.4 From FbIRC Flammability Apparatus (Figure 1). See Tewarson (1988, 1994) for other materials. bLOI: Limiting Oxygen Index. SOURCE: Data from Tewarson (1988, 19941. For YO > 0.30, the flame radiative heat flux reaches an asymptotic limit comparable to the limit for normal air burning in large-scale fires (Tewarson et al., 1981; Tewarson, 1988, 19941. Thus, large-scale flame radiative heat flux conditions can be simulated in small-scale experiments. The technique to simulate large-scale flame radiative heat flux conditions in small-scale flammability experiments by the oxygen mass fraction variations is defined as the Flame Radiation Scaling Technique (Tewarson, 1988, 1994). Table 3 compares the results from the flame radiation scaling technique used in the small- scale experiments in the FMRC Flammability Apparatus and results from large-scale fires. The data show that the asymptotic flame heat flux values from the FMRC Flammability Apparatus are in good agreement with the values derived from the mass pyrolysis rate in large-scale fires. The asymptotic flame heat flux values vary from 22 kW/m2 to 77 kW/m2. dependent cr~mar~lv ~ ~ , on the pyrolysis mode rather than on the chemical structures. For example, for liquids, which vaporize primarily as monomers, the asymptotic flame heat flux values are in the range of 22 kW/m2 to 44 kW/m2, irrespective of their chemical structures. For polymers, which vaporize as high molecular weight oligomers, the asymptotic flame heat flux values increase substantially to the range of 49 kW/m2 to 71 kW/m2, irrespective of their chemical structures. The independence of the asymptotic flame heat value from the chemical structure is consistent with the dependence of the flame radiation on optical thickness, soot concentration, and flame temperature. Modifications in the pyrolysis behavior to enhance release of higher monomer fraction relative to oligomer fraction and reduction in the carbon atom fraction relative to other atoms in the pyrolysis products (enhanced surface charring) would reduce the flame heat flux transferred back to the surface and the mass pyrolysis rate (Equation 3~.
66 In 0~8 i, o' i- O ° i-O 3 ~ 1 - NOI103S d~ddn - WtO'Z Ct A: 111 z o J ~ Z o <: Ct ~ o J ~ m<: Improved Fire- aru] Smoke-Resistant Materials · ~ / ~~ \ z ~ ~ g O2 'JJ ~ Z O ~ ~ ~ ~ O ~ ~ A LLI A ilk' < O Q a) o ~ O Tic O ~ ~ , AS 3 z fir o ~ c TIC - Q in 3 C= \ o ~ C) o J LL In In C`c ~ Cat c o C_ - o NOI103S U3~0 w9,'l <,1 ~ /! ~ ~ , ~ ~ o L$J | ~ Z 2 5E CN L, tY '= J ca ~ c, _- _~ ~ _~ 7 ,~ £3 E ~ , 1 ~ i ~ Z a I CY - - ~1 <: o: LL `:Y ~ ~z a,) ~ ~ Q Q() E a, g .. 3 _ 1 / o ~: o 3 ._ ._ l l ~ 1 3 - ~ .5 3 .~: ~ ~ t 3 O '4: 3 ~n .~ o (o w ~q C~ o ~Q - .s 3 o w 3 ~4 o In - ~n w - - · - - - ~ ce 3m o~ ~ ~ o C) ?` ~ o ~ ~Q d ~ := ~: =-_ ~ O
Archie Tewarson 60 50 x - cot I ~ 30 3 _ ~0 20 10 o 67 Aft Radiative [///~ Convective l =~ ~ : 0 , ...._... , ........ _ ~ ,~..~....~... ........ _ _ ; P; . ..... _: .... i. Do i // ~ ~ FIGURE 2 Flame radiative and convective heat fluxes at various oxygen mass fractions for the stP~dy-state combustion of 100 x 100 x 25-mm-thick slab of polypropylene. Data are from the FMRC Flammability Apparatus. Numbers are the oxygen mass fractions. The mass pyrolysis rate is directly proportional to the heat release rate and the generation rates of products. Decrease in the mass pyrolysis rate, thus, would reduce the thermal and nonthermal hazards. IGNITION Ignition is a process in which the pyrolysis products are generated at a certain rate, mix with air, and form a combustible mixture that ignites by itself (auto-ignition) or is ignited by a flame, a hot object, or similar means (piloted ignition). The rate of generation of the pyrolysis products leading to ignition is defined as the critical mass pyrolysis rate. Minimum heat flux at or below which the critical mass pyrolysis rate is not achieved and there is no ignition is defined as the critical heat flux (CHF). The CHF value is very close to the surface re-radiation loss. Relationships have been developed between the time to ignition and external heat flux (Tewarson, 1988). These relationships are as follows: it) for thermally thick materials, the surface is at the ignition temperature and the back is close to the ambient temperature at the ignition condition; and (2) for thermally thin materials, the surface is at the ignition temperature and the back is close to the ignition temperature at the ignition condition.
68 Improved Fire- Art Smoke-Resistant Materials TABLE 3 Asymptotic Mass Pyrolysis Rate and Flame Heat Flux in Combustion Mass Pyrolysis Rate Flame Heat Flux (kg/m2.s) x 103 (kW/m2) Flame Radiation Large- Flame Radiation Large Polymers/Liquidsa Scaling Techniques Scale Scaling Techniques Scale Aliphatic Carbon-Hydrogen Atoms Polyethylene 26 - 61 Polypropylene 24 - 67 Heavy fuel oil (2.6-23 m)b 36 29 Kerosene (30-80 m) - 65 - 29 Crude oil (6.5-31 m) - 56 - 44 n-Dodecane (0.94 m) - 36 - 30 Gasoline (1.5-223 m) - 62 - 30 JP-4(1.0-5.3m) - 67 - 40 JP-5 (10.60-17 m) - 55 - 39 n-Heptane (1.2-10 m) ~66 75 32 37 n-Hexane (0.75-10 m) - 77 - 37 Transformer fluids (2.37 m) 27-30 25-29 23-25 22-25 Aromatic Carbon-Hydrogen-Oxygen Atom Polystyrene (0.93 m) 36 34 75 71 Xylene (1.22 m) - 67 - 37 Benzene (0.75-6.0 m) - 81 - 44 Al iphatic Carbon-Hydrogen- Oxygen Atoms Polyoxymethylene 16 - 50 Polymethylmethacrylate 28 30 57 60 (2.37 m) Methanol (1.2-2.4 m) 20 25 22 27 Acetone (1.52 m) - 38 24 - Aliphatic Carbon-Hydrogen- Oxygen-Nitrogen Atoms Expanded polyurethanes 21-27 - 64-76 (flexible) Expanded polyurethanes 22-25 - 49-53 (rigid) Aliphatic Carbon-Hydrogen-Halogen Atom Polyvinylchloride (PVC) 16 - 50 Ethylenetetrafluoroethylene (ETFE) 14 - 50 (Tefzel) Fluannated ethylene-propylene (FEP) 7 - 52 (Teflon) Aflame Radiation Scaling Technique: Pool diameter fixed at 0.10 m, Yo20.30. bNumbers in parentheses are the pool diameters in meters. SOURCE: Data from Tewarson (1988, 1994~.
Archie Tewarsor~ 69 Ignition of Thermally Thick Materials - _ ~ 4 _ d/ e ~ or ~ Tig~P up (4) where ti' is the time to ignition (s), q `~ is the critical heat flux (kW/m2), ATig is the ignition temperature of the material above the ambient temperature (K), k is the thermal conductivity of the material (kW/m K), p is the density of the material (kg/m3), and cp is the specific heat of the material - /kg K). Still ~kpcp is defined as the thermal response parameter (TRP) for the thermally thick material (kW s''2/m2). For thermally thick materials, the square root of time to ignition is directly proportional to TRP and inversely proportional to the external heat flux. Figure 3 shows a typical example of the data for a thermally thick polymethy~methacrylate (PMMA) slab at various velocities (I) of the co-flowing air. 0.35 0.30 0.25 0.20 o · _ ·F~ c 0.15 o T 0.10 ._ ~ 0.05 Gym 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Natural Flow -I Co-Flow; v' = 0.~S m/s -I- Co-Flow; v' = 0.09 m/s - ~ Co-Flow; v' = 0.05 m/s 0.00 ~ ~ 1 ~ 1 ~ 1 ~ 1 1 ~ 0 1 0 20 30 40 50 60 - 70 80 90 1 00 External Heat Flux (kW/m2) FIGURE 3 Ignition data for 100 x 100 x 25-mm thick polymethylmethacrylate (PMMA) slab with blackened surface. Data measured in the FMRC Flammability Apparatus.
70 Improved Fire- aM Smoke-Resistant Materials Ignition of Thermally Thin Materials 1 ~ <Ye ~ Car ~= _ big 4 pcp6ATig (5) where ~ is the actual thickness of the material (m). The TRP for the thermally thin material (kJ/m2) is defined as pcp6ATi`. For thermally thin materials, the time to ignition is directly proportional to TRP and inversely proportional to the external heat flux. The fire-hardening of materials requires that the values of the CHF and the TRP be as high as possible. The CHF and TRP values have been reported for numerous natural and synthetic materials (Tewarson, 1988, 1994~. Thermal Response Parameter (TRP) The TRP values depend on the physical and chemical characteristics of the materials. An example is shown in Figure 4, where the TRP value is plotted against the resin fractions of the 2000 0 E 0 ~ 1 USA c: - u ce Q I 0 5 lY 1 600 1 200 800 E ,._ 400 | Kevlar Fibers 0.OO 0.20 Polyester-FG -0 Epoxy-FG -~ PPS-FG {) Phenolic-FG I Vinyl-FG Epoxy-FG-Ph¢nolic --I-- Cygnets or Epoxy-Oraphite ~ Phanolic-Kevlar \~ rune r mere 1 'I 'it' -- 1 0.40 0.60 0.80 Resin Fraction FIGURE 4 Thermal response parameter (TRP) versus the resin fraction for the composite systems. Data are from the FMRC Flammability Apparatus.
Archie Tewarson 71 composite systems. The TRP values increase with decrease in the resin fraction and increase in the fiber fraction. For the same resin fraction, the TRP value is highest for the graphite fiber systems, intermediate for the glass fiber systems and lowest for the KevIar0 fiber system, following the trends in the thermal conductivities of the fibers, that is, graphite > glass ~ KevIai.. For higher thermal conductivity fibers, a larger fraction of the heat applied to the surface is transferred to the interior, and time required to reach the ignition temperature is longer, resulting in the higher TRP value. The residual flexural strength retained (RF SR) is one of the parameters used to assess the structural performance of the composite systems (Sorathia et al., 19931. The dependency of RFSR on the properties of the composite systems is very similar to the dependency of TRP. A relationship between the RFSR and the TRP has thus been postulated (Tewarson and Haskell, 19941. Variations in the chemical bonds within similar generic resins and additives also play a major role in the ignition behavior of the composite systems by affecting the TRP values, as indicated by the data in Table 4 for f~berglass-reinforced polyester and epoxy composite systems. FIRE PROPAGATION Fire propagation is a process in which the pyrolysis front moves beyond the ignition zone, accompanied by the sustained combustion process. The rate of the movement of the pyrolysis front is defined as the fire-propagation rate. For a sustained fire-propagation process, flame or external heat sources need to transfer heat flux ahead of the pyrolysis front to satisfy the CHF and TRP values. The upward fire-propagation rate in the direction of air flow for thermally thick materials is expressed as (Sibulkin and Kim, 19771: l/2 _ 61/2 ~ U - ~ Tig~kp cp) (6) where u is the fire-propagation rate in m/s; of is an effective flame heat transfer distance (m), assumed to be constant; Of is the flame heat flux transferred ahead of the pyrolysis front (kW/m21; and ~T,,:~kpcp is the TRP for the thermally thick materials in kW · s''2/m2 (Equation 41. The flame heat flux transferred ahead of the pyrolysis front is a function of the rate of heat actually released in the fire-propagation process, defined as the chemical heat release rate. Figure 5 shows an example of the chemical heat release rate for the downward fire propagation for a 300-mm long, 100-mm wide, and 25-mm-thick vertical slab of PMMA in an oxygen mass fraction of 0.446 (Tewarson and Ogden, 1992~. The slope of the curve is the fire-propagation rate. The figure also shows the combustion of the entire slab in normal air and in reduced oxygen mass fractions. The flame extinction occurs at an oxygen mass fraction of 0. 178.
72 Improved Fire- and Smoke-Resistant Materials TABLE 4 CHF and TRP for Thermally Thick Composites with Fiberglass Reinforcement _ Fiberglass (weight %) CHF (kW/m2) TRP (kW.s"21m2) Polyester 0 ~ 296 30 ~ 256 70 (1) 10 275 70 (2) 10 382 70 (3) 15 406 70 (4) 10 338 77 - 426 Epoxy 0 - 257 65 (1) 10 420 65 (2) 10 410 65 (3) 10 400 76 15 667 SOURCE: Data from Tewarson (1988, 1994). ............... 0.233 . 2 o =1: Cal 0 200 0.178 ......... ~ ~l < ,. < l ~ 0.233 400 600 800 1 000 1 200 1 400 Time (s) FIGURE 5 Chemical heat release rate versus time for the downward fire propagation, steady combustion, and flame extinction for 300-mm long, 100-mm wide, and 25-mm-thick PMMA vertical slab under opposed air flow condition in the FMRC Flammability Apparatus. Air flow velocity = 0.09 m/s. Numbers are oxygen mass fractions. .............. 0 - ~ _ _ ~ ..... ` 0 E ....i~...... ......
Archie Tewarson 73 The chemical heat released in the fire-propagation process has a convective and a radiative component. In larger-scale fires the flame heat flux transferred ahead of the pyrolysis front is mainly a function of the radiative component of the chemical heat release rate or the radiative heat release rate. Numerous correlations have been developed for the relationship between the flame heat flux transferred ahead of the pyrolysis front and the radiative heat-release rate, one of which is ~ewarson and Khan, 19881: Off a' tirade 2/3 (7) where Q',~ is the radiative heat release rate per unit width or circumference of a slab or a cylinder of a material respectively (kW/m). From the definition of the radiative heat release rate (Tewarson, 1988, 1994~: Q red %radQ T (Xrad/Xch) Q ch (6Hra~/68ch) Q ah (8) where X,ad is the radiative component of the combustion efficiency, Xch, Q or is the heat release rate for complete combustion per unit width or circumference of the slab or cylinder of the material, respectively (kW/m); tech iS the chemical heat release rate per unit width or circumference of the slab or cylinder of the material, respectively (kW/m); math iS the chemical heat of combustion (MJ/kg); and AH,, d is the radiative heat of combustion (MI/kg). The average value of math 0.42 + 0.13 ~ewarson, 198S, 1994) and, from equations 7 and 8: ~'f ~ (0 . 42QCh ~ / and, from equations 6 and 9: u1/2 (a' (9) l/~ ~ Tig~P Cap (10) The righthand side of Equation 10, with a proportionality constant assumed to be 1,000, 0tCh in kW/m and AT'g`/kpcp in kW s''2/m2, is defined as thefire-propagation index (FPl) (Tewarson, 198S, 19941: F - ~ ~ COO (O . 42Q!CA) l/3 PI TRP (11)
74 Improved Fire- arm Smoke-Resistant Materials Fir - Propagation Index The following FPI values, based on the data from small- and large-scale fires, have been found to characterize the general fire-propagation behavior of materials under high flame radiation conditions: · FPT ~ 7: No fire propagation beyond the ignition zone. Polymers are identified as nonpropagating group N-! polymers. Flame is at critical extinction condition. 7 < FPT ~ 10: Decelerating fire propagation beyond the ignition zone. Polymers are identified as group D-! polymers. Fire propagates beyond the ignition zone although in a decelerating fashion. Fire propagation beyond the ignition zone is limited. 10 < FPI < 20: Fire propagates slowly beyond the ignition zone. Polymers are identified as propagating group P-2 polymers. FPT 2 20: Fire propagates rapidly beyond the ignition zone. Polymers are identified as propagating group P-3 polymers. Thefire-hardening of materials requires that the FP! values of the materials be less than or equal to 7. Examples of some typical FPI values are listed in Table 5. TABLE 5 Fire-Propagation Index of Materials Fire Polymers Thickness (mm) FPI Group Propagationa Polymers Polymethylmethacrylate 2530 P-3 P Fire-retarded polypropylene 25~ >10 P-3 P PVC/PVF cable 5 (diameter)7 N-1 N FEP/FEP cable 10 (diameter)5 N-1 N Composite Systems Polyesterl 70% FGb 4.8 13 P-2 P 4.8 10 P-2 P Polyester2 70% FG 19 8 D-1 D 45 7 N-1 N Epoxyl 65%FG 4.4 9 D-1 D Epoxy2- 65% FG 4.8 11 P-2 P Epoxy3 65% FG 4.4 10 P-2 P Phenolic 80~o Kevlar ~3.2 3 N-1 N Phenolic 84% Kevlar ~4.8 8 D-1 D 'P: propagating; D: decelerating propagation; N: nonpropagating. bFG: fiberglass. SOURCE: Data from Tewarson (1988, 1994~.
Archie Tewarson 75 The FPI is inversely proportional to the first power of the TRP value and directly proportional to the one-third power of the chemical heat release rate. Thus, the TRP value has a much stronger effect on the FPI value than the chemical heat release rate. An increase in the TRP value and a decrease in the chemical heat release rate by the various techniques discussed in the previous sections would lead to FPI values that are ~ 7. GENERATION OF HEAT AND FIRE PRODUCTS As a material is exposed to heat in the ignition zone, the first step is the generation of the pyrolysis products. The second step is the mixing of the pyrolysis products and air and the ignition of the mixture. The third step is the establishment of sustained combustion as a result of the burning of the pyrolysis products with air. The fourth step is the movement of the pyrolysis front beyond the ignition zone as a result of the heat transfer from the flame or external heat source beyond the ignition zone. Heat and fire products (i.e., smoke, toxic, and corrosives are generated in each of the above steps. Heat Release Rate The chemical heat release rate is directly proportional to the mass pyrolysis rate in the combustion process, defined in Equation 3: Pi' h = AH In ( 12 ) where Q'tCh is the chemical heat release rate (kW/m2) and Hash iS the proportionality constant defined as the chemical heat of combustion - /kg). Arch iS always less than the net heat of complete combustion, IT, because, as in fires, the combustion process always remains incomplete. The ratio ~HCh/~HT iS defined as the combustion efficiency, xch. The convective and radiative components of the chemical heat release rate and chemical heat of combustion are defined as the convective and radiative heat release rates, Q"co'' and Q"raa' and convective and radiative heat of combustion, AHco, and Brads respectively. The convective and radiative components of the combustion efficiency are expressed as Go,, and Xrad, respectively. Data for the heats of combustion for a variety of materials have been reported (Tewarson, 1988, 1994~. The chemical heat of combustion decreases, but combustion efficiency increases with the introduction of the oxygen atom into the carbon and hydrogen atom containing chemical structures of the materials. The combustion efficiency decreases, but its radiative component increases with the increase in the chemical bond un saturation and aromaticity and with introduction of the halogen, sulfur, and nitrogen atoms in the structures of the materials. Thefire-hardening of materials requires that the heats of combustion be reduced to values as low as possible. Techniques to retain large fractions of the carbon atoms in the solid phase (charring), introduction of the oxygen atoms in the chemical structures, and enhancement of the
76 Imp roved Fire- arm Smoke-Resistant Materials chemical bond saturation may be helpful, similar to the techniques to increase the heat of gasification discussed previously. Heat Release Parameter From equations 3 and 12: Q., = [ Batch] By' ~ I' - if' (13) The ratio of the chemical heat of combustion to heat of gasification, AHC,,/AH:, is defined as the heat release parameter CARP) in MI/MT. The HRP defines the amount of energy generated in combustion per unit amount of energy required to pyrolyze a unit mass of the material. HRP has a convective and a radiative component. The HRP value is independent of the fire size but depends on the fire ventilation. Me fire-hardening of materials requires that the HRP be reduced to values as low as possible. HRP values for numerous materials have been published (Tewarson, 1988, 19941. Selective dam from this tabulation are list in Table 6. The data suggest that the nonpropagating fire condition is satisfied when HRP C 2. TABLE 6 Fire-Propagation Index and Heat Release Parameter for Selected Materials Polymers Thickness (mm) FPI Group HRP Polymers Polymethylmethacrylate 2530 P-3 15 Fire-retarded polypropylene 25~ >10 P-3 19 PVC/PVF cable 5 (diameter) 7 N-1 1 FEP/FEP cable 10 (diameter) 5 N-1 2 Composite Systems Polyesterl 70% FGa 4.8 13 P-2 5 Polyester2 70% FG 4.8 10 P-2 19 8 D-1 - 45 7 N-1 2 Epaxyl-65% FG 4.4 9 D-1 6 Epoxy2 65% FG 4.8 11 P-2 5 Epaxy3 65% FG 4.4 10 P-2 6 Phenolic 80% FG 3.2 3 N-1 1 Phenolic 84% Kevlar. 4.8 8 D-1 4 aFG: fiberglass. SOURCE: Data from Tewarson (1988, 1994).
Archie Tewarson He Ef37ect of Ventilation' on the HRP 77 The effect of ventilation on the HRP values of the nonhalogenated materials is expressed as ~ewarson et al., 1993~: (HRP) V = (HRP) ~ [ exP <~/2 . 2) -1~2] where (HRP)V is the ventilation-controlled combustion value, and (HRP)oo is the well-ventilated combustion value; ~ is the equivalence ratio: sob"> mate (15) where S is the stoichiometnc mass air-to-fuel ratio (kg/kg), A is the exposed surface area of the burning material (m2), and ffl~`r is the mass flow rate of air (kg/s). Figure 6 shows the ratio of (HRP3V to (HRP)oo as a function of the equivalence ratio. For 4} 2 1.0, (HRP3V I (HRP)oo ~ 1.0, an indication of ventilation-controlled combustion. For ~ 2 4, the HRP value becomes less than about 40 percent of the HRP value for well-ventilated combustion, and flame is extinguished. Equation 14 provides a relation strip to determine the HRP values for various fire ventilation conditions. By knowing or predicting the net heat flux for a fire scenario, the chemical heat release rate can be calculated from Equation 13. 1 .2 1.04 8 0.8 I `, 0.6 - ~L I 0.4 0.2 o.o I I I I i I I'I I I 10-1 10° 1 1 ~_b ~ _- _- _ - _ orb q it's + Wood · PMMA Nylon O PE PP O PS 1 01 Equivalence Ratio 1o2 FIGURE 6 Ratio of the heat release parameter for the ventilation-controlled to well-ventilated combustion of materials versus the equivalence ratio.
78 Improved Fire- am Smoke-Resistar~t Materials Mass Generation Rate of a Product The mass generation rate of a product in pyrolysis or combustion is directly proportional to the mass pyrolysis rate: By, = y m" (16) where G"' is the mass generation rate of product j (kg/m2 s), and yj, the proportionality constant, is defined as the yield of the product (kg/kg). Yields of CO, CO2, hydrocarbons, and smoke for a vanely of materials have been report ~ewarson, 198S, 19941. The yields of products of incomplete combustion, that is, CO, hydrocarbons, and smoke, increase with increase in the chain length and chemical bond unsaturation and aromaticity and with introduction of the halogen, sulfur, and nitrogen atoms in the structures of the materials. In fires the yield of a product is significantly less than its stoichiometr~c yield, defined as the maximum possible conversion of the material to the prompt: v jM MF (~17) where tj is the stoichiometric yield of product j, Mj is the molecular weight of product (kg/mole), and My is the molecular weight of the pyrolyzed material assumed to be a monomer (kg/mole). The stoichiome~ic yields of products depend on the relative numbers of hydrogen, oxygen, nitrogen, sulfur, halogen, and other atoms relative to the carbon atom. The stoichiometric yields of the products provide an insight into the nature of the products and the maximum possible mass generation rates of products expected in pyrolysis and combustion processes in fires. The ratio Y.11tj is defined as the generation efficiency of the product j, A. The fire-hardening of materials requires that the yields of products of complete and incomplete combustion and pyrolysis be reduced to values as low as possible. Techniques to reduce the heats of combustion and gasification, discussed previously, would also help to reduce the yields of the products. Product Generation Parameter For the pyrolysis process, from equations ~ and 15 Gail = [ YJ ~ ~ (I/ e (Y or) (18)
Archie Tewarson For the combustion process, from equations 3 and 15 j {L AHG ~ ~ f At rr) ~ L9 79 The ratio of the yield of a product to heat of gasification, y,JAHg, is deemed as the product generation parameter (POP) in kg/MT. The PGP defines the amount of a product generated per unit amount of energy required to pyrolyze the material. PGP is independent of the hire size but depends on the fire ventilation. Thefire-hardenir~g of materials requires that the PGP valuesforproducts of complete aru' incomplete combustion aM pyrolysis be reduced to values as low as possible. Techniques to reduce the HRP values, discussed previously, would also help to reduce the PGP values. The CO and smoke PGP values for selected materials are listed in Table 7, along with the FPI values and the group classification. Within each fire-propagation group, fire-hardening requires that the CO aru] smoke PGP be reduced to as low values as possible through various techniques. TABLE 7 Fire-Propagation Index and Product Generation Parameter for Selected Materials Thickness PGP, PGP, Polymers (mm) FPI Group CO Smoke Synthetic Polymers Polymethylmethac~rlate 25 30 P-3 0.0062 0.014 Fire-retarded polypropylene 25 ~ > 10 P-3 0.012 0.029 PVC/PVF 5 (diameter) 7 N-1 FEP/FEP cable 10 (diameter) 5 N-1 0.085 0.002 Composite Systens Polyesterl-70% FGa 4.8 13 P-2 0.017 0.021 4.8 13 P-2 0.056 0.037 Polyester2- 70% FG 19 8 D-1 45 7 N-1 - Epoxyl 65% FG 4.4 9 D-1 0.088 0.068 Epaxy2 65% FG 4.8 11 P-2 0.053 0.088 Epaxy3-65% FG 4.4 10 P-2 0.072 0.052 Phenolic 80% FG 3.2 3 N-1 0.007 0.002 Phenolic 84% Kevlar. 4.8 8 D-1 0.002 0.003 aFG: fiberglass. SOURCE: Data from Tewarson (1988, 1994~.
80 The Effect of Ventilation on the PGP Imp roved Fire- and Smoke-Resistant Materials The effect of ventilation on the HRP values of the nonhalogenated materials is expressed as (Tewarson et al., 19931: (POP) j,v = (PGP)Yj,- [ exp(4t/~) `] (20) where (PGPjjV is the product j PGP value for ventilation-controlled combustion (kg/MT), (PGPjj so is the product j PGP value for well-ventilated combustion (kg/MJ), and c', it, and ~ are the ventilation correlation coefficients. The values of cr. ,8 and ~ for CO, CO2, 02, hydrocarbons, and smoke for venous materials have been reported (Tewarson, 198B, 1994~. The values of c' and ~ are strongly dependent; whereas, the value of `S is weakly dependent on the chemical structures of the nonhalogenated materials. The ventilation correlation coefficient c' primarily reflects the magnitude of the fire properties in nonflaming fires (high ~ values). The ventilation correlation coefficients reflects the magnitude of the fire properties in the transition region between the well-ventilated and ventilation-controlled combustion of the materials. The ventilation correlation coefficient ax reflects the range of ~ values for the transition region. High value of ax is indicative of strong effect of ventilation on the combustion of the materials. High values of ,B and ~ are indicative of rapid change of flaming combustion to nonflaming combustion by a small change in the equivalence ratio, such as for the halogenated polymer (e.g., PVC [polyviny~choride]), for which combustion in normal air itself is unstable. Equation 20 suggests the following three conditions for the ventilation-controlled combustion of the polymers: (~) for . ~ > 6, Yj v = yj " (l + c`), (2) for . < < it, Yj v = Yj-, and (3) for ~ z if, Yjv as yjOO (! + c'/2.71. Figures 7 and ~ show the ratios of the CO and smoke PGP values for the ventilation- controlled to well-ventilated combustion as functions of the equivalence ratio. With increase in the equivalence ratio, the CO PGP value increases to as high as 60 times the value for the well- ventilated combustion. The smoke PGP value increases to only about 2.6 times the value for the well-ventilated combustion with increase in the equivalence ratio. Equation 20 provides a relationship to determine PGP values for various fire ventilation conditions. By knowing or predicting the net heat flux for a fire scenario. the generation rates of the products can be calculated from equations IS and 19. Preferential Conversion of Carbon in the Material to CO with Decrease in Fire Ventilation With decrease in fire ventilation during the combustion of the nonhalogenated materials, the preferential conversion of the carbon in the material to CO follows the order: wood (C-H-O aliphatic structure) > PMMA (C-H-O aliphatic structure) > nylon (C-H-O-N aliphatic structure) > PE (polyethylene) (C-H aliphatic linear unsaturated structure) ~ PP (polypropylene) (C-H aliphatic branched unsaturated structured > PS (polystyrene3 (C-H aro
Archie Tewarson Cod 8 o a_ CL - - o AL - 10° , ................... 1 ........................................ 4::::: :::::::::::::::: Fl ~ mi n ~ ::::::::::: ::::::::::::::::.::::~: to-do-' rid I......... I...0~......... it............ For em, .. . do. . . ., ./. A. . ,/ · ~ A! :/ ~ ·~. ~ /~. ~ ·~. · ~/· ~ 'I' 'i'' Ad' A'' '/ ·/' /' i / ·y· ·~·~ 'art Nonflaming bitt/ - ~ ~/~/~/~< ~~ a/// + Wood ~/////~///; · PUMA //// ~ Nylon /// ~ PP ~ A/// /////; O PS > - ~/~/~ //~/~; ~/~/~ ~ ~/~ ~///~. /~ ///// ~/~ ~ + Wood · PUMA Nylon Cl PE ~ PP O PS Off ~ ////// ~/~ 1 o1 Equivalence Ratio 1o2 81 FIGURE 7 Ratio of the CO generation parameter for ventilation-controlled to well-ventilated combustion of materials versus the equivalence ratio. 2.8 8 2.4 03 Q A_ ~ 2.0 - CL C, 1.6 1.2 1 2 ! Fleming I . ~ . . . . . . . , , , ~ . . . . . . . . . . ................................. ..................................... ..... . . . . . . . . . . ................................. ....................................... ..... . . . . . . . . . ~A . ~. I 2 ~ ~ ////.//. /~/ Nonfloming ~ · PUMA , Nylon , PE iO PP ~ 10° Equivalence Ratio 10' FIGURE 8 Ratio of the smoke generation parameter for ventilation-controlled to well-ventilated combustion of materials versus the equivalence ratio.
82 Imp roved Fire- and Smoke-Resistant Materials matic structure). A similar trend is found for the liquids and gases. The presence of O and N atoms in the chemical structures of the materials with aliphatic C-H structure appears to enhance the preferential carbon atom conversion to CO. The order could be due to preferential pyrolysis of the material to CO and/or preference for the reactions between OH and CO compared to the reactions between OH and C. A decrease in the OH concentration with increase in the equivalence ratio is also suggested by the order. Preferential Conversion of Carbon ire the Material to Smoke with Decrease in Fire Ventilation With decrease in fire ventilation during the combustion of the nonhalogenated polymers, the preferential conversion of the carbon in the material to smoke follows the order: PS > wood ~ PE = PP ~ nylon > PMMA. The order for the preferential conversion of the carbon atom to smoke is opposite to the order for the conversion of the carbon to CO, except for wood. The order could be clue to preferential pyrolysis of the material to carbon and/or preference for the reactions between OH and CO compared to the reactions between OH and C, and/or decrease in the concentration of OH. THERMAL AND NONTHERMAL DAMAGE Damage due to heat is defined as thermal damage; and damage due to smoke, toxic, and corrosive products is defined as nonthermal damage ~ewarson, 1992~. Nonthermal damage depends on the chemical nature and deposition of products on the walls, ceilings, building furnishings, equipment, and components, and so forth, and on the environmental conditions. The seventy of the nonthermal damage increases with time. Some examples of nonthermal damage to property are corrosion damage, electrical malfunctions, and damage due to discoloration and odors. Toxic effects of fire products on the human body that result in an injury or loss of life are examples of nonthermal damage to life. The subject of toxicity has been discussed (NRC, 1986~. This paper deals with the subject of nonthermal damage in industrial and commercial occupancies due to smoke and corrosive fire products. The subject of corrosion for commercial and industrial occupancies has been reviewed based on the knowledge denved from the telephone central office (TCO) experience for the deposition of atmospheric pollutants and fire products on equipment, severity of corrosion damage, and ease of cleaning the equipment (Reagor, 1992; FCC, 1993~. In TCO fires involving PVC-based electrical cables, contamination levels in the range of about 5 ,ug/cm2 to 900 ~g/cm2 have been observed (Reagor, 1992; FCC, 19931. In general, an electronic switch would be expected to accumulate zinc chloride levels in the range of about 5 ~g/cm2 to 9 ,ug/cm2 from the interaction with the environment over its expected lifetime of 20 or more years. Clean equipment is expected to have less than about 2 ~g/cm2 of chloride contamination; whereas, contaminated equipment can have as high as 900 ~g/cm2. Thus, equipment contamination levels due to chloride ions and ease of restoration have been classified into four levels (Reagor, 1992), which are listed in Table 8.
Archie Tewarson TABLE 8 Contamination Levels for the Surface Deposition of Chloride Ions for Electronic Equipment 83 Chloride Ion (,ug/cm2) Level Damage/Cleaning/Restoration 2 One No damage expected. No cleaning and restoration required. < 30 Two Equipment can be easily restored to service by cleaning without little impact on long-term reliability. 30 to 90 Three Equipment can be restored to service by cleaning, as long as no unusual corrosion problems arise, and the environment is strictly controlled soon after the fire. < 90 Four The effectiveness of cleaning the equipment dwindles and the cost of cleaning quickly approaches the replacement cost. Equipment contaminated with high chloride levels may require severe environmental controls even after cleaning in order to provide potentially long-term reliable operation. SOURCE: Data from Reagor (1992~. CORROSION Corrosion is defined as an unwanted chemical reaction and/or destruction or deterioration of a material because of reaction with its environment. Most of the knowledge on corrosion damage has been based on air pollution, for example, that due to acid rain. and on laboratorv- scale pyrolysis and combustion experiments. _ _ , _ _ _ _ , In fires mew surfaces are exposed to fire products that include water (generated in the combustion process and present in the ambient airs. The exposure is of short duration, a few minutes to a few days. Figure 9 shows an example of corrosion of a thin copper film (5,000 A' exposed to the combustion products of PVC homopolymer and commercial materials as measured in the FMRC Flammability Apparatus (Figure I). The slopes of the lines represent the corrosion rate. The corrosion rate from the PVC homopolymer is significantly higher than the rate from the PVC commercial materials, indicating dilution and/or partial neutralization of hydrogen chloride (MCI) by the pyrolysis products of nonhalogenated additives in commercial materials. The corrosion is faster in the initial stages and becomes slower in later stages due to protective oxide film formation on the surface. The corrosion rate of a metal exposed to the pyrolysis and combustion products is found to satisfy the following relationship (Tewarson, 19941: z? = I[ - -'corr r ~ core (21 ) where Renoir is the corrosion rate (A/min), ,u is corrosion constant [(A/min)/(kglnI)], and CCorr is the average concentration of the corrosive product (kg/m3). In the gas phase the average concentration of the corrosive product is equal to the ratio of the total mass of the product in kg to the total volume of water in the gas phase in of.
84 Improved Fire- am Smoke-Resistant Materials 500 400 300 200 100 a . ~ ! , . , . , . ~ ., 1 ,., ~ 1! , l l ................... ~- ................... ~ .. 0 250 500 : : : : ... .., ~- , PVC Homopolymer . . I . ! _. ~- i _. I'.= an_ ~ - ................... 1 ,... 1 . . i ! PVC Commercial Materials 1 ' 1 . 1 . 750 1 000 1 250 1 500 Time (second) - FIGURE 9 Gas-phase corrosion from the combustion products of PVC homopolymer and commercial materials. Data from the FMRC Flammability Apparatus. The total mass of the corrosive product is equal to YCO,7WT, where YCO!T iS the yield of the corrosive product (kg/kg) and WT is the total mass of the material pyrolyzed (kg). If VT is the total volume of the fire product-air mixture, then the volume of water isfwVT, where fw is the volume fraction of water in the fire product-air mixture. The concentration of the corrosive product then becomes YCo~,WT/fw VT, and from Equation 21: . _ IlYcorr WT R _ corr fwVT Rearranging Equation 22: (22 ~ Corrosi on Index = 11 tore = RCorz/ ( WT/ VT) ( 2 3 ) w The corrosion index (CI) is the rate of corrosion per unit average mass concentration of the material pyrolyzed (A/min)/(kg/m). The CT values have been reported (Tewarson, 1994~. The typical CT value for gas-phase corrosion for a highly halogenated polymer with hydrogen atoms
Archie Tewarson 85 in the structure, such as PVC, is 4 x 103 (A/min)/(kg/nf). For a highly halogenated polymer with no hydrogen atoms in the structure, such as tetrafluoroethylene (TEE), the CI value is 0.6 x 103 (A/min)/(kg/m3), indicating the importance of the formation of water in the combustion and the inefficiency of the hydrolysis process with water from the ambient air to generate acids. The CT values suggest that: . For significant gas-phase corrosion it is necessary to have hydrogen atoms in the structure of the halogenated polymers. For example, the C! values for PVC (hydrogen atoms in the structure) and TFE (no hydrogen atoms in structure) differ by factor of seven. The difference is probably due to (~) the inefficiency of the hydrolysis process in the gas phase during the conversion of fluorocarbon products generated from TEE to hydrogenfluoride; and (2) the high water solubility of hydrogen chloride generated from PVC. Fire retardation of nonhalogenated polymers by halogenated compounds increases the CT values. Presence of water in the environment is not necessary for the gas-phase corrosion from the products of halogenated polymers with hydrogen atom in the structure as water is generated in the combustion process. Increase in the oxygen concentration of the environment increases the C! values. Fire-hardening requires that within each fire-propagation group, the C be reduced to values as low as possible. SMOKE DAMAGE Smoke is a mixture of black carbon (soot) and aerosol (Siegla and Smith, 1981; GoIciberg, 19851. It has been suggested that soot nucleation and growth occur near the highly ionized regions of the flames in combustion processes and that some of the charges are transferred to smoke particles. Smoke damage in industrial and commercial occupancies is considered in terms of discoloration and odor of the property exposed to smoke; interference in the electric conduction path and corrosion of the parts exposed to smoke is a carrier of the corrosive products. FLAME EXTINCTION Flame extinction is achieved by applying fire extinguishing agents, such as water, Halon@, or alternates, which interrupt the pyrolysis, combustion, and f~re-propagation processes by: (~) interacting with the burning material in the solid phase (mainly removal of heat), (2) reducing the availability of oxygen to the fire (creation of nonflammable mixture), and (3) removing the heat from the flame and interfering with the chemical reactions within the flame.
86 Improved Fire- arm Smoke-Resistant Materials The~?ame extinction requirements are lowerformatenals with a higher degree of fire-hardening. For example, group N-] materials with FP! < 7 do not require fire protection. When the extinguishing agents, active in the gas phase, are applied to a flame, the HRP values decrease; the PGP values of the products of incomplete combustion, such as CO, smoke, and, mixture of hydrocarbons, increase. These results are very similar to the results for the ventilation-controlled fires (figures 6-~. Figures 10 and ~ ~ show the ratios of the HRP and PGP values for well-ventilated combustion of a polyester-70 percent glass composite system in the presence and absence of Malone 1301. In Figure 10, the PGP ratio for hydrocarbons increases to Il5, for CO it increases to 10, and for smoke it increases to 2. These results suggest a possible interruption by Malone 1301 of the reactionts) in which CO and hydrocarbons are consumed, rather than the reactions in which smoke is consumed. This type of behavior is also found for the ventilation-controlled combustion of materials with oxygen atoms in the structure (figures 7 and 8~. As shown in Figure ~ I, the HRP ratio decreases to 0.5S, below which the flame becomes unstable, leading to flame extinction. This is similar to the behavior shown in Figure 6, which results from the increase of the equivalence ratio. SUMMARY Fire-hardened materials offer resistance to pyrolysis, ignition, combustion, and fire propagation and would be materials of choice for commercial aircraft interiors to reduce hazards due to heat (thermal hazard) and smoke, toxic, and corrosive products (nonthermal hazard). The resistance to pyrolysis and ignition would be increased by increasing the values of (~) the gasification temperature or surface re-radiation loss and heat of gasification to r~luce the mass pyrolysis rate, and (Z) the ignition temperature or the critical heat flux (CHF) and the thermal response parameter ~RP) to delay ignition and increase removal of heat from the surface to the interior. Stronger chemical bonds, pyrolysis mechanisms favoring retention of carbon in the solic! phase (charring), enhancement of thermal conductivity, density, and specific heat of the materials are some of the factors expected to be effective in this endeavor. Some of the commercial materials introduced recently satisfy these requirements. The resistance to combustion would be enhanced for materials with high resistance to pyrolysis and ignition. In addition, the Fame treat flunk transferee back to the surface aru] the heat of combustion need to be decreased to reduce the mass pyrolysis rate in the combustion and the heat release rate. These two fire properties could be reduced by (~) modification of the pyrolysis behavior to enhance release of higher monomer fraction relative to the oligomer fraction, (2) reduction in the carbon atom fraction relative to other atoms in the pyrolysis products (enhancing the char formation), (3) introduction of the oxygen atoms in the structure, and (4) decrease in the chemical bond unsaturation, aromaticity, and others. Initially the processes of pyrolysis, ignition, and combustion occur within the area where the material is heated. The area is defined as the ignition zone. If the heat flux transferred beyond the ignition zone satisfies the CHF and TRP values, fire propagation beyond the ignition would be initiated. The thermal and nonthermal hazards depend on the rate and extent of fire propagation beyond the ignition zone and are characterized by the fire-propagation index (FPl).
Archie Tewarson 8 ,o2 avdlrocaIbon!------ l----- ' 1 :~ 1 : . ........................ . ~......................... .......................... , . . i . ~......................... ......................... ......................... .. .- .i. ~.. ~ vl , ., ~-~ is - ............ ~ ............ , .............. - . ~...................... ....................... . i ......................... ,~. ~i ........................... ........................ ............ + ............ 0 ............ | .. . ~. ~. ~. A ............ _ ............ _ . ~........................ . ~ ........................ . .r ~ . ', I ~ ~ _ A _ .t i ;;;;;;;,r ~ L ,. . . ... 7 ........ ~ . ~;; ~_ _ ; .......... ; r ~ o ~ ~ ~ .. ....................... ........ 0 . k e ' ... ....... ........................ ..' ,. . . ~ , . ..' 10°! ~ _ it__ 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Halon (Volume 5~) ........ ....................... COMER ' = ........... . .. . . '' , i ;j I - .. .::::: ::Jr:: ~.:.L v ~ ,, ~ ~ A. - -, ~;....................... ............ 4 ....................... ........................ ............ + . A ............ _ ............ _ . ~ .. .e ............ ; ......................... _ ..~. ,...~" ''''/ L6/./ /// ,Y/r, / ~ // 7~ ~ of;: ' ;j,! A, ,j< ;~. f74 ·~- i= ~/ ~ ~ of, a/ ~ ~ of/ ~ L/? ~- // // Asia '/, 1/, ~ ~ 7,,: ~ ~ f/;~ 7/,~7 ~7, Em. :/7// '/./ Em/ 4.0 4.5 FIGURE 10 Ratio of the POP in the presence and absence of Halon0 1301 for the well-ventilated combustion of polyester-70 percent glass composite system exposed to 60 kW/m2 of external heat flux. Data are from the FMRC Flammability Apparatus. 0.9 8 - P~ _ 0.8 ° 0.7 hi: 8 ~ 0.6 - 0.5 0.4` I T . ~ . \. N , t , ................. . ~ . ~-..T~ I. ~-~ 1 -1 ~1 - ~ .0 1.5 2.0 2.5 3.0 3.5 Halon (Volume I) FIGURE 11 Ratio of the HRP in the presence and absence of Halon0 1301 for the well-ventilated combustion of polyester-70 percent glass composite system exposed to 60 kW/m2 of external heat flux. Data are from the FMRC Flammability Apparatus. S_~E Le~E, 4.0 4.5
88 Improved Fire- arm Smoke-Resistant Materials Under high flame-radiation conditions, that is, large-scale fires, materials with FPI values <7 are nonpropagating, group N-1 materials. Materials with FPI values > 7 and < 10 show decelerating propagation and are identified as group D-1 materials. Materials with FPI values 2 10, but ~ 20, show slowly propagating fire beyond the ignition zone and are identified as group P-2 materials. Materials with FPI values 2 20 show rapidly propagating fire beyond the ignition zone and are identified as group P-3 materials. Fire-hardening requires materials to be group N-] materials. The FPT values would be reduced by increasing the CHF and TRP values and decreasing the heat release rate. Within each fire-orocaaation group. it is necessary that the heat release rate and the generation rates of fire _# _ 4, . ~ ~ . , ~ ~ .' ~ ~ ~ . ~ . .... products be reduced to values as low as possible. rule neat release rate wlthm each ilre- propagation group is characterized by the heat release parameter (HRP) (or the ratio of the heat of combustion to heat of gasification). For group N-1 materials, HRP is <2. The generation rates of products within each fire-propagation group are characterized by the product generation parameter (POP) (or the ratio of the yield of the product to heat of gasification). The POP values within each fire-propagation group need to be reduced to as low values as possible. Parameters to characterize smoke and toxic damage have not been defined; for corrosion damage, a corrosion index (CI) has been identified as the corrosion rate of a metal per unit concentration of the material pyrolyzed. NOMENCLATURE CHF critical heat flux (kW/m2) cco" average concentration of a corrosive product (kg/m3) C} corrosion index cp specific heat (MI/kg K) ETFE ethylenetetrafluoroethylene (Tefzel) fw volume fraction water (-) FG fiberglass reinforced FEP fluorinated ethylene-propylene Teflon) FPT fire propagation index {10~ (0.42 Q'c0''3 / BATS ( - p)~/2~} G"j mass generation rate of product j (kg/m2 s) dHi heat of combustion, gasification, melting, or vaporization per unit mass of material pyrolyzed (MI/kg) HRP heat release parameter (AHCb/AHg) k thermal conductivity (kW/m K) mair mass flow rate of air (kg/s) m"p mass pyrolysis rate (kg/m2 s) M molecular weight (kg/mole) PE polyethylene POP product generation parameter (yj/AHg)(kg/MI) PMMA polymethylmethacrylate PP polypropylene PS polystyrene
Archie Tewarson PVC q't Q''i Q'i Qcorr s AT,g TRIP TOT u vg VT WT Yj Yo Greek c' f. Of Xch SCOT Grad ~j p Hi Subscript a ch con corr cr e f fc fr g 89 polyvinylchloride heat flux (kW/m2) heat release rate per unit sample surface area (m"AHch) (kW/m2) heat release rate per unit sample width (kW/m) corrosion rate (A/min) stoichiometric mass air-to-fuel ratio (-) time (s) temperature (K) ignition temperature above ambient (K) thermal response parameter, thermally thick [AT,g (kocp)t'2] (kW s~'21m2) thermal response parameter, thermally thin (AT,g bpcp) (LJ/m) fire-propagation rate (m/s) co-flow air velocity (m/s) total volume of fire product-air mixture (m3) total mass of material pyrolyzed (kg) yield of product j (Wj/Wf) (kg/kg) mass fraction of oxygen (-) ventilation correlation coefficient for nonflaming region (-) ventilation correlation coefficient for transition region (-) ventilation correlation coefficient for the equivalence ratio (-) equivalence ratio (Sm"pA/m,,,) thickness or depth (m) effective flame heat transfer distance (m) combustion efficiency (Q"ch / m"~HT) convective component of the combustion efficiency (Q"con /m"~HT3 (-) radiative component of the combustion efficiency (Q"rad /m,'~HT) (~) generation or consumption efficiency of a product (yj / I) (-) corrosion constant (A/min)(kg/m3) density (kg/m3) stoichiometric yield for the maximum conversion of fuel to product j (-) air or ambient chemical convective corrosion critical external flame flame convective flame radiative gas or gasification
90 Improved Fire- arm Smoke-Resistant Materials i chemical, convective, radiative · · · ~ lg 1gmtlon j fire product m melting n net o initial red radiation stoich stoichiometric for the maximum possible conversion of the fuel to the product rr surface re-radiation s surface th depth v ventila~cion-controlled fire w water ~well-ventilat~ Superscript . 11 per unit time (s-l) per unit width (m~l) per unit area (mu) REFERENCES FCC (Federal Communications Commission). 1993. Network Reliability: A Report to the Nation. Section G in Compendium of Technical Papers. Presentation by the Federal Communications Commission's Network Reliability Council at the National Engineering Consortium, Chicago, Illinois. Goldberg, E.D. 1985. Black Carbon in the Environment-Properties and Distribution. New York: John Wiley & Sons. Hottel, H.C. 1959. Review: Certain laws governing the diffusive burning of liquids by Blinov and Khudiakov (1957) (DokI Akad), Nauk SSSR, Vol. ~ 13, 1094, 1957. Fire Research Abstract and Reviews (~:41-45. NRC (National Research Council). 1986. Fire and Smoke: Understanding the Hazards. Board on Environmental Studies and Toxicology, NRC. Washington, D.C.: National Academy Press. Reagor, B.T. 1992. Smoke corrosivity: Generation, impact, detection, and protection. Journal of Fire Sciences (10~: 169-179. Sibulkin. M., and I. Kim. 1977. The dependence of flame propagation on surface heat transfer. , , as. ~ _ _ ~ T ~ ~ · ~ ~ . · ~ · ~ ~ ~ ~ ~ _ ~ ~ ~ ~ 11: Upward burmng. Combustion science and technology 1 1:;~-4Y. Siegla, D.C., and G.W. Smith, eds. 1981. Particulate Carbon Formation During Combustion. New York: Plenum Press. Sorathia, U., C. Beck, and T. Dapp. 1993. Residual strength of composites during and after fire exposure. Journal of Fire Science ~ ~ :255-270.
Archie Tewarson 91 Tewarson, A. 1988. Generation of heat and chemical compounds in fires. Chapter I-13, Pp. I- 179 to I-199 in The SFPE Handbook of Fire Protection Engineering. Quincy, Massachusetts: The National Fire Protection Association Press. Tewarson, A. 1992. Non-thermal damage. Journal of Fire Science 10: ISS-241. Tewarson, A. 1994. Flammability parameters of materials: Ignition, combustion, and fire propagation. Journal of Fire Science 12:329-356. Tewarson, A., en c! W.E. Haskell. 1994. Fire Hardening of Composite Systems. Presentation at the Annual Conference on Fire Research at Gaithersburg, Maryland, the National Institute of Standards and Technology, October 17-20. Tewarson, A., and M.M. Khan. 1988. Flame propagation for polymers in cylindrical configuration and vertical orientation. Pp. 1231-1240 in 22nd Symposium (international) on Combustion. Pittsburgh, Pennsylvania: The Combustion Institute. Tewarson, A., and D. Macaroni. 1993. Polymers and composites An examination of fire spread and generation of heat and fire products. Journal of Fire Sciences Il:421-441. Tewarson, A., and S.D. Ogden. 1992. Fire behavior of polymethy~methacrylate. Combustion and Flame 89:237-259. Tewarson, A., I.~. Lee, and R.F. Pion. 1981. The influence of oxygen concentration on fuel parameters for fire modeling. Pp. 563-570 in I8th Symposium (International) on Combustion. Pittsburgh, Pennsylvania: The Combustion Institute. Tewarson, A., F.H. liang, and T. Morikawa. 1993. Ventilation-controlled combustion of polymers. Combustion and Flame 95: 15 1-169.
