Heat Release CapacityRichard N. Walters

Galaxy Scientific Corporation2500 English Creek AvenueEgg Harbor Township, NJ 08234

Richard E. LyonFederal Aviation

AdministrationFire Safety Section AAR-422

W.J. Hughes Tech CenterA.C Int’l Airport, NJ 08405

Outline

Background– Flammability Testing & Fire Processes– PCFC Instrument Description

Theory– PCFC– Heat Release Capacity– Fire Test Theory & Correlations– Molar Group Contributions

Conclusions

Flammability Testing

Extrinsic quantities resulting from the reaction of a macroscopic polymer sample to a severe thermal exposure– Burning Rate Ignitability– Flammability Fire Performance

Fire test results are dependent on– Orientation Geometry Sample Size– Color Ambient Conditions Operator

Thermal and mass diffusion effects dominate the fire behavior in large sample tests

An intrinsic material property is needed to correlate fire performance

Fuel Generation Process

Pyrolysis Combustion

Combustible & Non-Combustible Gases

Polymer

Thermal feedback from flame

Liquid Products & Tar

Solid Charred Residue

+ O2 Flame∆

Endothermic Exothermic

Char Formation

Charring in fires is an anaerobic process

0

0.2

0.4

0.6

0.8

1.0

0 0.2 0.4 0.6 0.8 1.0

Fire

Cha

r Yie

ld, g

/g

TGA N2 Pyrolysis Residue, g/g

Microcalorimeter Schematic

Pyrolysis Combustion

Sample

Pyrolyzer Combustor Scrubbers FlowMeter

O2Analyzer

O2N2

DAQ

Exhaust

Pyrolysis-Combustion Flow Calorimeter

Forced Non-Flaming Combustion Test

Small Sample Size: ~1mg

Rapid Screening of Materials

Measured Values:Heat Release RateTotal Heat ReleaseChar Yield

Calculated Values:Heat Release CapacityGlobal Ea

Measured Heat Release Rates

0

500

1000

1500

2000

2500

3000

3500

4000

0 50 100 150 200 250 300 350

Mic

rosc

ale

Hea

t Rel

ease

Rat

e (W

/g)

Time (seconds)

Polyethylene

Polypropylene

Polystyrene

ABS

PMMA

PET

PEEK

PBI

Microcalorimeter Theory

Heat release rate by oxygen consumption normalized to initial sample mass

Qc(t) = E∆O2

mo= h c,v

o (t) –1mo

dm(t)dt

Peak rate of heat release

Qcmax = E∆O2

mo max= h c,v

o –1mo

dmdt max

= hc,vo

β 1 – µ Ea

eR T p2

Heat Release Capacity

η c = Qcmax(β)β =

h co 1 – µ Ea

eR T p2

… dependent only on material properties

Material Property

Must satisfy the following conditions:

1. Independent of sample mass

2. Independent of heating rate

3. Measurable by different methods

Heat Release Capacity by TGA-GC/MS

0

500

1000

1500

2000

0 500 1000 1500 2000

Hea

t Rel

ease

Cap

acity

(PC

FC),

J/g-

K

β =

260

K/m

in

Heat Release Capacity (TGA-GC/MS), J/g-K β = 10 K/min

PP

PS

poly(α-methyl)styrene

PPONylon 66

PCPET

POM

PEEKKEVLAR

PPS

PI

PE1:1 Line

T.V. Inguizilian, Correlating Polymer Flammability Using Measured Pyrolysis Kinetics, MS Thesis, University of Massachusetts, Amherst, January 1999.

Flammability Character - UL-94V vs. LOI UL-94 - Vertical Bunsen Burner test based on operator observation

where materials are classified by their burning behavior

Oxygen Index - Minimum amount of oxygen needed to sustaina candle-like flame for 3 minutes based on operator observation

UL-94 LOIXXXXX

X

X

Vertical TestPreheating of sampleRadiant componentConvective componentConductive component

Upward versus Downward flame spread

UL-94 LOI

Fire Test Correlations - UL-94 & LOI

0 200 400 600 800 1000 1200

UL-

94 R

atin

g

Heat Release Capacity (J/g-K)

V-0

V-1

V-2

HB

Self

Extin

guis

hing

TransitionRegion

Burns Readily

10

15

20

25

30

35

40

45

50

0 200 400 600 800 1000 1200

Oxy

gen

Inde

x

Heat Release Capacity (J/g-K)

TransitionRegion

Burns Readily

20.9% O2

33% O2

Transition Region from self-extinguishing to readily burning: 200-400 J/g-K

Transition Region corresponds to 21-33 %O2 in LOI

Cone Calorimeter

Bench Scale Flaming Combustion Test

Sample Dimensions:10cm x 10 cm x 0.3cm

0-100 kW/m2 incident heat flux

Measured Values:heat release rate - O2 consumptionmass loss ratetime to ignitionCO & CO2smoke generation

Calculated Values:critical heat fluxheat of gasificationignition temperaturethermal inertia (κρc)

Fire Test Correlations - Cone Calorimeter

0

200

400

600

800

1000

1200

1400A

vera

ge F

lam

ing

HR

R (k

W/m

2 )

Heat Release Capacity (J/g-K)

250 15000 500 750 1000 1250

POM

PEPS

ABS

PBTPMMA

PESF

PEI

Nylon 66

Epoxy

PAI

PC

PEEK

PTPET

PPqext = 50 kW/m• 2

Cone Calorimetry - Bench scale flaming combustion test measuringheat release by oxygen consumption

Fire Test Summary

Fire Tests– Flaming combustion tests where physical attributes

influence the flammability– Samples can be designed to give false passes in the tests– Tests are hard to run and do not give quantitative results– Multiple tests and rigorous calculations needed to obtain

material properties

PCFC– Provides a material property directly– Quantitative results that represent the total fuel value of a

sample– Quick & easy to run

Group Contribution BackgroundMethods for predicting thermochemical data from molecular structure (Bensen 1968)

Interactions of several atoms summed and approximated by structural groups (VanKrevelen 1972)

Atomic-level bond topology used to predict molecular properties using connectivity indices (Bicerano, 1996)

Structural group contribution method for flammability (Walters 2000)

Group Contribution calculations can predict thermodynamic quantities such as:

– Heat of Combustion - Thermal Decomposition Temperature– Char Yield - Glass Transition Temperature– Heat Capacity

Theories of group contributions based on empirical correlations

Molar Group Contributions to ηcHeat Release Capacity

η c = Qc

β =hc

o 1 – µ Ea

eR T p2

Approach: Write the heat release capacity terms as additive molar quantities

Ψ = H V EeR T 2 =

niHiΣi

niViΣi

niEiΣi

eR niTiΣi

2

Expand summations over chemical groups, i, j, k… and neglect terms with mixed indices

Ψ = niHi V i Ei

eR Ti2 = niΨiΣ

iΣi

Obtain ηc in correct units from molar mass of component groups

η c = ΨMo

=niΨiΣ

i

niMiΣi

=NiΨiΣ

i

NiMiΣi

Calculating Heat Release Capacity

If ηc is a material property it should be calculable from additive molar group contributions like other polymer properties (e.g., heat capacity, refractive index, solubility parameter, etc.)

CH2CH CH2 O C O

OCH2 CH

CH2 O

CH3

CH3

Example: Bisphenol A Epoxy has 6 distinct chemical groupscomprising the polymer repeat unit.

Calculating Heat Release Capacity

CH2CH CH2 O C

CH3

CH3

OO

CH2 CHCH2 O* *

ηc = Ψ

M =niΨ iΣ

i

niMiΣi

=NiΨ iΣ

i

NiMiΣi

= 204.5 kJ / mole–K340 g / mole = 601 J g–KJ g–K

Chem ica lGro u p, i

N Mi (g /mo le) Ψ(k J /mo le-K)

Ni Mi(g /mo le)

Ni Ψ(k J /mo le-K)

C 12 28.3 12 28.3

CH 2 13 26.6 26 53.2

CH2 4 14 16.7 56 66.8

CH3 2 15 22.5 30 45.0

2 76 28.8 152 57.6

O 4 16 –11.6 64 –46.4T otal: 340 204.5

Molar Group Contributions to ηc

η c = Ψ

Mo=

niΨiΣi

niMiΣi

=NiΨiΣ

i

NiMiΣi

0

300

600

900

1200

1500

1800

0 300 600 900 1200 1500 1800

Cal

cula

ted

Hea

t Rel

ease

C

apac

ity (J

/g-K

)

Measured Heat Release Capacity (J/g-K)

Group Co ntributio n(k J /mo l-K)

Group Co ntributio n(k J /mo l-K)

118*

O -1 1 . 6 P

O

77.0

PN

O

O -1 3 . 8

69.5

NH2 -1 3 . 9 *

30.6

CF3 -1 4 . 8 C

CH3

CH3 29.5

C N -1 7 . 6

28.8 O

NN

O -1 8 . 9 * C

28.3

SO

O -1 9 . 2 CH

26.6

OH -1 9 . 8

CH3 22.5

Br -2 2 . 0

19.0

C

O

-2 2 . 0 N

18.7

C O -2 3 . 2 *

CH216.7

N

O

O -2 5 . 5

15.1

Cl -3 4 . 7 C C

9.7

N

O

O

N

O

O-3 6 . 4 *

H 8.1

C

O

OPendan t :-3 9 . 5

Backbone:-1 3 . 7 NH

7.6

N

-4 3 . 0 *

1:1 Line

Conclusions

Heat Release Capacity is:

– A rate independent flammability parameter– An intensive quantity (independent of sample mass)– Measurable by different laboratory techniques– A good predictor of fire and flammability behavior– Calculable from chemical structure– A material property: dynamic combustion potential

Acknowledgements

FAA

Dr. Stanislav Stoliarov