SCREENING FLAME RETARDANT ADDITIVESSCREENING FLAME RETARDANT ADDITIVESFOR PLASTICS USING MICROSCALE COMBUSTION FOR PLASTICS USING MICROSCALE COMBUSTION

CALORIMETRYCALORIMETRY

FEDERAL AVIATION ADMINISTRATIONAirport and Aircraft Safety R&D Division

William J. Hughes Technical CenterAtlantic City International Airport, NJ 08405

The Fifth TriennialInternational Aircraft Fire and Cabin Safety Research Conference

Atlantic City, New Jersey USAOctober 29-November 1, 2007

WEB: www.fire.tc.faa.gov E-MAIL: richard.e.lyon@faa.gov

Richard E. Lyon and Richard N. Walters

PROBLEM

Need Small Scale (milligram) Screening Test for FR Additives to

Reduce Development Costs and Accelerate Discovery.

APPROACH

• Measure Properties of Complete Combustion using Microscale

Combustion Calorimetry.

• Use a “Burning Efficiency” to Account for Incompleteness of

Flaming Combustion.

• Account for Uncertainty Using Probability.

RESULTS

CONCLUSIONS

SCREENING PLASTICS FOR FLAMMABILITYSCREENING PLASTICS FOR FLAMMABILITY

PP PEPS

PET

ABSPMMA

Epoxy

POM

NYLON 66

PCPPO

PUPAR

PPSU

PPSPES

PEI LCP's

PAI

PEKPEEK

0.1 1 10 100Materials Cost* ($/lb)

Phenolic

KAPTON

1000

10

100PVF

Heat Resistant Plastics

PSU

PVC (rigid)

Commodity Plastics

*Truckload Prices, 2001

Engineering Plastics

IntrinsicallyFlame

Resistant

Flam

mab

ility

(ηc)

, J/g

-K

THE GOAL OFTHE GOAL OF THE FR ADDITIVE APPROACHTHE FR ADDITIVE APPROACH

Flame Retardant Additives

Flame Retardant Additives

Plastic

Heat fromFlame

Heat LossFrom Surface

O2

Fuel Gases From Plastic

= Heat = Heat ResistanceResistance

FLAME RETARDANTS WORK INFLAME RETARDANTS WORK IN TWO WAYSTWO WAYS

Gas Gas PhasePhaseActivity Activity

Condensed Condensed PhasePhaseActivityActivity

Need to quantify the efficiency of these modes of action

APPROACHAPPROACH

Pyrolysis

Sample

Gases analyzed for residual oxygento compute heat release rate

O2

N2

Combustion

Flaming Combustion

MicroscaleCombustionCalorimeter

1) Reproduce elements of flaming combustion in non-flaming test

3) Relate thermal combustionproperties to fire and flametests using deterministic and probabilistic models

2) Measure thermal combustion properties of materials

MICROSCALE COMBUSTION CALORIMETER (MCC)MICROSCALE COMBUSTION CALORIMETER (MCC)

Combustor

Pyrolyzer

Mixing Section

Sample Cup

ScrubberFlow MeterO2 Sensor

Exhaust

Purge Gas Inlet

Sample Post & Thermocouple

Oxygen Inlet

Heat release rate by oxygen consumption

AUTOMATED / HIGH THROUGHPUT MCCAUTOMATED / HIGH THROUGHPUT MCC

Combustor with oxygen consumptionattached to TGA with automated sampling

Capable of ≈ 50 tests/day

MCC becameASTM Standard

April 1, 2007

MCC becameASTM Standard

April 1, 2007

THERMAL COMBUSTION PROPERTIES (MCC)THERMAL COMBUSTION PROPERTIES (MCC)

250 300 350 400 450 500 5500

100

200

300

400

500

Hea

t R

elea

se R

ate

(J/g

-K)

Temperature (°C)

φ = Pyrolysis residue (Char fraction)

Heat Release Capacity, ηc (J/g-K)

Total Heat Release, HR (J/g)

Temperature at Max HRR, Tmax (°K)

FORCED AND UNFORCED COMBUSTIONFORCED AND UNFORCED COMBUSTION

Cone Calorimeter (ASTM E 1354)

OSU Calorimeter (14 CFR Part 25)

ForcedCombustion

UnforcedCombustion

Vertical Flame Test (ASTM D 3801)

Oxygen Index (ASTM D 2863)

FLAMING COMBUSTION: FLAMING COMBUSTION: Gas Phase ChemistryGas Phase Chemistry

Complete Combustion (Oxidation) of Fuel Gases

CcHhOmNnXx + m O2

900 °CCO2 + H2O + N2 + HX

10 secFuel Gases

Incomplete Combustion of Fuel Gases in Diffusion Flame

CcHhOmNnXx + n O2

Flame

Fuel Gases + CO + CxHy + soot

n O2

m O2< 1χ =Flaming Combustion Efficiency,

MicroscaleCombustionCalorimetry

CO2 + H2O + N2 + HX Fires and Flames

FLAMING COMBUSTION: FLAMING COMBUSTION: Condensed PhaseCondensed Phase PhysicsPhysics

Heat Transfer Efficiency, θ = =q″net− κ(ΔT/Δx) char layer

− κ(ΔT/Δx) resin

Tem

pera

ture

Tsurface

Tdecomp

ΔT

Δx Δx Δx

VirginResin

GasChar Slope = Heat Transfer RateΔT

Δx∝

Char Layer Depth =

q″net0

Heat Release Rate (HRR) for Steady Burning:

Hc = Heat of CombustionOf Fuel Gases

0

χ = Combustion Efficiencyin Flame

θ = Heat Transfer Efficiencyat Surface

HRR = χ Hc0

Lgθ ′ ′ q flame − ′ ′ q rerad( ) + χ Hc

0

Lgθ ′ ′ q ext

HRR0

HR

R (k

W/m

2 )

q″ext (kW/m2)0

Cone Calorimeter

Data

STEADY BURNING MODEL: MacroSTEADY BURNING MODEL: Macro

HRP

HRP = = ηg

HRR = ηcηg

( ′ ′ q flame − ′ ′ q rerad ) + ηcηg

′ ′ q extHeat Release

Rate:

Heat Release Capacity

Flaming Combustion Efficiency

HcLg

HRR AND THERMAL COMBUSTION PROPERTIESHRR AND THERMAL COMBUSTION PROPERTIES

χθ0 ηcχθ

hg

ΔTp

Fire Response

FORCED FLAMING COMBUSTIONFORCED FLAMING COMBUSTION

HEAT RELEASE RATE:HEAT RELEASE RATE: Macro Vs. MicroMacro Vs. Micro

ConeCalorimeter

Natural PlasticsMicro-compositesNano-compositesGFRP (PA6, PBT, PC, PPS)

0

500

1000

1500

2000

0 500 1000 1500 2000

Heat Release Capacity ηc (J/g-K)

R = 0.93

PVC

PEIPPS

POM

PET

PVDF

PC

PMMA

ABSPA66

HIPSPP

HDPE

50 kW/m2

FEPPeak

HR

R in

Con

e C

alor

imet

er

(kW

/m2 )

PC/ABS BlendsFR Compounds

At large external heat flux, HRR ∝ ηc

ASTM E 906/OSURate of Heat

Release Apparatus

1

10

100

1000

1 10 100 1000Heat Release Capacity, J/g-K

Thermoplastic Sheet (1.6 mm) Thermoset Composites (single ply)

FAR 25.853(a-1) Maximum

OSU

Pea

k H

RR

, kW

/m2

OSU HEAT RELEASE RATE:OSU HEAT RELEASE RATE: Macro Vs.Macro Vs. MicroMicro

Flame Resistance

UNFORCED FLAMING COMBUSTIONUNFORCED FLAMING COMBUSTION

FLAME EXTINCTION CRITERIAFLAME EXTINCTION CRITERIA

HRP ≤

MACRO Extinction Criterion for Flame Tests

ηcηg HRR*

Flame Extinction Occurs at Critical HRR:

MICRO Extinction Criterion for Flame Tests

HRR* ≈100 kW/m2 (Downward Burning)

60 kW/m2 (Upward Burning)

(q″flame- q″rerad)≤

(q″flame- q″loss)HRR*

= ηc*

[O2∗ ] = ′ ′ q rerad

a+

HRR∗ ηg / aηc

≈ 12% + 2.8 kJ / g − Kθ χ ηc

(%)

Limiting Oxygen Index (LOI) = [O2*]

q″flame ∝ [O2] = a [O2]

a = 1.4 kW/m2-%O2

ηg =hg / ΔTp

χθ= ( 2 kJ / g ) /(50K )

χθ= 40 J / g − K

χθ

HRR* = 100 kW/m2 (downward burning)

FLAME RESISTANCE TEST (ASTM D2863)FLAME RESISTANCE TEST (ASTM D2863)

LOI EXTINCTION CONDITION:

′ ′ q rerad = σTmax4 ≈17kW/m2

EFFECTEFFECT OF BURNING EFFICIENCY ON L.O.I.OF BURNING EFFICIENCY ON L.O.I.

Halogen containing

0

20

40

60

80

100

10 100 1,000 10,000

Heat Release Capacity ηc, J/g-K

Lim

iting

Oxy

gen

Inde

x, %

v/v

Flamm

ability

0.3

w/ FR Additives

0.6LOI =12% + 2.8 kJ / g − K

θ χ ηc(%)

θχ = 1

q″flame ≈ 30 kW/m2

ηg =hg / ΔTp

χθ= ( 2 kJ / g ) /(50K )

χθ= 40 J / g − K

χθ

HRR* = 60 kW/m2

FLAME RESISTANCE: UPWARD BURNINGFLAME RESISTANCE: UPWARD BURNING

UL 94 V EXTINCTION CONDITION:

ηc ≤HRR∗ ηg

χθ( ′ ′ q flame − εσTmax4 )

≈ 2.4 MW J m−2 g−1K −1

χθ( ′ ′ q flame − εσTmax4 )

Comparative Burning Characteristics of Plastics in Vertical Position (UL 94 V or ASTM D 3801):

q″rerad = CHF ≈ εσTmax 4

0

200

400

600

800

1000

1200

1400

200 250 300 350 400 450 500 550 600

χθ0.2

0.40.60.6

Extin

ctio

n η C

, J/g

-K

Decomposition Temperature Tmax, °C

0.81

UL 94 VFlame Test

0.30.3

Burning Efficiency

0.60.6

EFFECT OFEFFECT OF χθχθ ANDAND TTmaxmax ON EXTINCTION ON EXTINCTION ηηc c

100 1000Heat Release Capacity, J/g-K

NR / Burns

V-2

V-1

V-0 / 5V

30 300 3000

1111

0.40.40.9

θ =χ =

UL

94 V

Rat

ing

Bromine FR Natural Plastics

Phosphorus FR 0.90.40.4

Flamm

ability

EFFECT OF BURNING EFFICIENCY ON UL 94EFFECT OF BURNING EFFICIENCY ON UL 94

Gas Phase FRCondensed Phase FRUNCERTAINTY DUE TO BURNING EFFICIENCY

OODDS OF BURNINGDDS OF BURNING AND AND HRRHRR

Assume that Odds of Burning is related to HRRand probability of burning, pB

HRRHRR*

pB

1 - pB=

HRRHRR*

pB =(HRR/HRR*)

1 + (HRR/HRR*)

Then,

pB

10 0.9995 0.99

1/2 0.111/5 0.0081/10 0.001

2 0.891 0.5

3

3

3

Odds ofBurning =

OODDS OF BURNINGDDS OF BURNING ANDAND THERMAL COMBUSTION PROPERTIESTHERMAL COMBUSTION PROPERTIES

Then,

= χθ ηc

ηg

And:

′ ′ q netHRR HRR* ≈ 60 kW/m2=;ηc

ηg′ ′ q critical

*

pB = (χθ ηc / ηc )3 / χθ

1+ (χθ ηc / ηc )3 / χθ

*

*

Depends only on burning efficiency (χθ)and heat release capacity (ηc )

pB

1− pB= HRR

HRR *⎛ ⎝ ⎜

⎞ ⎠ ⎟ 3

= χθ ηc

ηc

⎛

⎝ ⎜

⎞

⎠ ⎟ 3

′ ′ q net

′ ′ q critical

⎛

⎝ ⎜

⎞

⎠ ⎟ 3

≈ χθ ηc

ηc

⎛

⎝ ⎜

⎞

⎠ ⎟ 3 / χθ

* *

Heat Release Capacity, ηc (J/g-K)

NR / HB

V-2

V-1

V-0 / 5V

UL 94 V Rating

= Burn (HB / NR / V-2 / V-1) = No Burn (V-0 / 5V)

pB = Fraction of BurnResults in HRC Bin

MEASURE PROBABILITY OF BURNING,MEASURE PROBABILITY OF BURNING, ppBB

0 200 400 600 800 1000 1200 1400 1600

HRC bin width of ≈ 50 J/g-K gives statistically valid sample (n ≥ 5)

184 test results

Heat Release Capacity, ηc (J/g-K)

Prob

abili

ty o

f Bur

ning

, pB

pB =χθ ηc / 325( )3 / χθ

1+ χθ ηc / 325( )3 / χθ

EFFECT OF BURNING EFFICIENCY ON PROBABILITYEFFECT OF BURNING EFFICIENCY ON PROBABILITY OF BURNINGOF BURNING

0

0.2

0.4

0.6

0.8

1

10 100 100030 300 3000

χθ = 1

χθ = 0.6(FR Compounds)

χθ = 0.8Best Overall Fit

(pB ±0.11)

CONCLUSIONSCONCLUSIONS

Probabilistic Models are required to reconcile MCC data with flame resistance tests because

• Intumescence, charring (θ)• Gas phase inhibition (χ)• and Dripping

are comparable in magnitude and effect to thermal combustion properties at extinction.

Incompletecombustion in flame

Deterministic Models using thermal combustion properties appear adequate for forced flaming combustion.