6/13/02IASFPWG – London, UK

Ongoing Fuel Flammability Work at the FAA Technical

Center

International Aircraft Systems Fire Protection Working GroupLondon, UKJune 13 – 14, 2002

Steve SummerProject EngineerFederal Aviation AdministrationFire Safety Branch, AAR-440

6/13/02IASFPWG – London, UK

Agenda

Fuel vaporization computer model – validation experiments

Theoretical flammability limits as a function of MIE, FP, and O2 content

Fuel vapor simulant for use in future ignition testing

Reports to be published

6/13/02IASFPWG – London, UK

Fuel Vaporization Model – Validation

Experiments

6/13/02IASFPWG – London, UK

Acknowledgements

Professor C. E. Polymeropolous of Rutgers University

David Adkins of the Boeing Company

6/13/02IASFPWG – London, UK

Introduction

The original model proved a good method of predicting the evolution of hydrocarbons.• Results were presented by Prof. Polymeropolous

(10/01 Fire Safety Conference) Could prove to be a key tool in performing fleet

flammability studies. Fortran code has been converted to a user-friendly

Excel spreadsheet by David Adkins of Boeing.

6/13/02IASFPWG – London, UK

Physical Considerations 3D natural convection heat and mass

transfer within tank• Fuel vaporization from the

tank floor which is completely covered with liquid

• Vapor condensation/vaporization from the tank walls and ceiling

Multi-component vaporization and condensation

Initial conditions are for an equilibrium mixture at a given initial temperature

Gas, Tg

Liquid, Tl

Walls and Ceiling, Ts

6/13/02IASFPWG – London, UK

Major AssumptionsWell mixed gas and liquid phases within the

tank • Uniform temperature and species concentrations in

the gas and within the evaporating and condensing liquid

• Rag ≈109, Ral ≈ 105-106

Externally supplied uniform liquid and wall temperatures. Gas temperature is then computed from an energy balance

Condensate layer is thin and its temperature equals the wall temperature.

6/13/02IASFPWG – London, UK

Major Assumptions (cont’d)Mass transport at the liquid–gas interfaces

was estimated using heat transfer correlations and the analogy between heat and mass transfer for estimating film mass transfer coefficients

Low evaporating species concentrations Liquid Jet A composition was based on

previous published data and and adjusted to reflect equilibrium vapor data (Polymeropoulos, 2000)

6/13/02IASFPWG – London, UK

Assumed Jet A Composition

Based on data by Clewell, 1983, and adjusted to reflect for the presence of lower than C8 components

Compound Volume,% Molecular Boiling Density,Weight Point, °C kg/m3

C5 parafins 0.01 72 309 630C6 paraffins 0.15 86 341 664C7 paraffins 0.5 100 371 690C8 parrafins 0.5 114.2 391 700C8 cycloparaffins 0.5 112.2 397 780C8 aromatics 0.54 106.2 412 870C9 paraffins 2.333 128.3 415 720C9 cycloparaffins 1.433 126.2 427 800C9 aromatics 0.933 120.2 438 880c10 paraffins 5.533 142.3 433 720c10 cycloparaffins 3.433 140.3 444 800c10 aromatics 2.233 134.2 450 860c11 paraffins 8.633 156.3 469 740c11 cycloparaffins 3.233 154.3 469 800dicycloparaffins 3.033 152.3 474 890c11 aromatic 3.533 148.2 478 860c12 paraffins 10.733 170.3 489 750c12 cycloparaffins 7.933 166.3 494 880c12 aromatics 4.533 162.3 489 860c13 paraffins 11.433 184.4 508 760c13 cycloparaffins 8.433 182.4 498 800c13 aromatics 4.833 176.3 507 870c14 paraffins 5.833 198.4 527 760c14 cycloparaffins 4.333 192.4 563 940c14 aromatics 2.433 186.3 568 1030c15 parafins 1.333 212.4 544 770c15 cycloparaffins 0.933 206.4 573 900c15 aromatics 0.533 200.4 578 950c16 hydrocarbons 0.133 226.4 560 770

6/13/02IASFPWG – London, UK

Assumed Jet A Composition

0

5

10

15

20

25

5 6 7 8 9 10 11 12 13 14 15 16

Number of Carbon Atoms

MW: 164

% b

y Vo

lum

e

6/13/02IASFPWG – London, UK

User Inputs

Equilibrium TemperatureFinal Wall and Liquid TemperaturesTime ConstantsMass LoadingTank Dimensions

Note: For comparison with experimental results, recorded wall and liquid temperature profiles were entered directly in lieu of the final temperatures and corresponding time

constants

6/13/02IASFPWG – London, UK

Program Outputs

Equilibrium gas & liquid concentrations/species fractionation

Species fractionation as a function of timeUllage, wall and liquid temperatures as a

function of timeUllage gas concentrations as a function of

time• FAR, ppm, ppmC3H8

6/13/02IASFPWG – London, UK

Experimental Setup

17 ft3 vented tank placed inside environmental chamber.

Thermocouples used to monitor ambient, ullage, surface and fuel temperatures.

Blanket heater attached to bottom of tank used to heat fuel.

Hydrocarbon analyzer used to monitor ullage fuel vapors.

6/13/02IASFPWG – London, UK

6/13/02IASFPWG – London, UK

6/13/02IASFPWG – London, UK

Future Testing

Future tests to consist of:• Constant surface temperature tests.• Various steady state pressure (cruise)

tests.• Varying pressure tests (flight profile).• Varying wall to wall temperature tests.• Varying fuel distribution tests.

6/13/02IASFPWG – London, UK

Future Model Improvements

Capability of varying tank pressure.Capability of varying wall to wall

temperature calculations.Capability of varying fuel distribution.

6/13/02IASFPWG – London, UK

Theoretical Flammability Limits as a Function of MIE, FP

& O2 Content

6/13/02IASFPWG – London, UK

BackgroundPresent thinking in fuel tank inerting is that

above x% O2, the tank is at risk throughout the entire flammability envelope, below x% O2 it is inert.

6/13/02IASFPWG – London, UK

Background

Previous work has shown how flammability limits vary as a function of ignition energy.

6/13/02IASFPWG – London, UK

Background

It follows intuitively that flammability limits will shift in a similar manner as inert gas is added to the fuel tank.

Thus, if your fuel tank is only partially inerted, the flammability exposure time has still been reduced by a significant amount.

How can this be quantified, validated and built into the flammability model?

6/13/02IASFPWG – London, UK

Computed Flammability Limits as a Function of O2

Similar methodology as that in DOT/FAA/AR-98/26 to compute flammability limits as a function of MIE.

6/13/02IASFPWG – London, UK

Computed Flammability Limits as a Function of O2

Correlation of the variation of LOC with altitude.• Previously determined with a large (~20 J)

spark source.

6/13/02IASFPWG – London, UK

Computed Flammability Limits as a Function of O2

, where

• Tmin is the minimum of the parabola given by Tmin = Tfp + 22 – 1.5Z.

• a is a constant, determined by matching the curve as best as possible to the calculated 21% O2 curve for the given ignition energy.

a

OLnOLnTT 2min2minFuel

6/13/02IASFPWG – London, UK

Resultant Curves for a 20 J Calculation

6/13/02IASFPWG – London, UK

Flammability Limits as a Function of MIE, O2 and FP

Combining this with the parabolic MIE calculations and LOC curves for various ignition energies, results in flammability limits which vary as a function of ignition energy, O2 concentration and flashpoint.• The sum of this work was put together into a

working MS Excel model by Ivor Thomas, using the following LOC curves.

6/13/02IASFPWG – London, UK

6/13/02IASFPWG – London, UK

Conclusions By a set of simple calculations, one can obtain

varying flammability limits as a function of ignition energy, O2 concentration and flashpoint.

Once validated, this data can be used in the flammability model to show reduction in fleetwide fuel tank flammability as a function of the amount of inert gas added to the tank.

Future tests to validate these calculations are planned at the technical center.

6/13/02IASFPWG – London, UK

Fuel Vapor Simulant for use in Future Ignition Testing

6/13/02IASFPWG – London, UK

Background

Our current method for ignition testing of Jet-A fuel vapors is extremely time consuming (up to as long as 2 hours per test).

If a gaseous mixture was available to simulate the flammability properties of Jet A, it would allow us to perform more tests quicker.

6/13/02IASFPWG – London, UK

Background

Availability of said mixture would also have applicability to other issues (e.g. explosion proof testing of pumps, etc.)

Subcommittee of SAE AE-5 currently being formed to look at this issue.

6/13/02IASFPWG – London, UK

Past Simulants - HexaneLFL UFL Stoich. AIT (F)

Hexane 0.033 0.22 0.065 437JP-8/Jet A 0.032 0.24 0.068 ~420

6/13/02IASFPWG – London, UK

Past Simulants – Caltech Mixture

NTSB Docket No. SA-516, Exhibit No. 20OVolumetric Ratio of H2:C3H8 of 5:1Examined the effect of fuel concentration,

vessel size and ignition source on pressure history.

6/13/02IASFPWG – London, UK

Past Simulants – Caltech Mixture

6/13/02IASFPWG – London, UK

Proposed Research Activity at Tech Center

20 L combustion vessel to be constructed within 4 – 6 weeks.

Variable energy (0.5 mJ – 5 J) spark source to be obtained within 8 – 10 weeks.

Tests to be conducted in a manner similar to the procedures layed out in ASTM flammability standards (e.g. E582, E2079, etc).

6/13/02IASFPWG – London, UK

Proposed Research Activity at Tech Center

6/13/02IASFPWG – London, UK

Reports to be Published

6/13/02IASFPWG – London, UK

Reports

Summer, S. M., Fuel Flammability Characteristics of JP-8 Fuel Vapors Existing Within a Typical Aircraft Fuel Tank, DOT/FAA/AR-01/54

Summer, S. M., JP-8 Ignition Testing at Reduced Oxygen Concentrations, DOT/FAA/AR-xx/xx