the fuel tank flammability assessment method – flammability analysis federal aviation...

1The Fuel Tank Flammability Assessment Method – Flammability Analysis

Federal AviationAdministration 1

The Fuel Tank Flammability Assessment Method

Flammability Analysis

Federal AviationAdministration



The need to the limit flammability creates a need

to define and assess flammability exposure

assessdefine



What is Flammability?

• In order to define flammability, we first need to define:

• Flash point: The flash point of a flammable fluid is the lowest temperature at which the application of a flame to a heated sample causes the vapor to ignite momentarily, or “flash”, and is determined by a simple, standardized test.

• Lower Flammability Limit (LFL): At any point below the LFL, the fuel vapor/air mixture is too lean to burn.

• Upper Flammability Limit (UFL): At any point above the UFL, the fuel vapor/air mixture is too rich to burn.



• Previous work has shown that the LFL and UFL can be defined, in terms of temperature as:• LFL= (Flash Point - 10) - Altitude/808, • UFL=(Flash Point + 63.5) - Altitude/512

(where temperature is in Deg F and altitude is in ft.)

What is Flammability?



• Add a slide here that shows a plot of fuel temp bounded by LFL/UFL, showing flammable range



The need to the limit flammability creates a need

to define and assess flammability exposure

assessdefine



Background

• The Fuel Tank Flammability Assessment Method (FTFAM) is an Excel© based macro based on work originally performed by the 1998 ARAC Fuel Tank Harmonization Working Group.

• It is a comparative analysis tool to examine airplane fuel tank flammability.

• The program utilizes Monte Carlo statistical methods to determine several unknown variables, using standardized distributions in order to calculate the fleet average flammability exposure time of a given fuel tank.

• From 1998 – Present, the FAA has utilized input from industry and information gained from various research activities to help refine and improve the model’s capabilities.



Basics of the Monte Carlo Method

• The Monte Carlo method is based on the use of random numbers and the statistics of probability to help solve problems that don’t yield to simple mathematics.

• Modeling of a problem is done by assigning random values, based on known distributions, to each unknown variable and calculating the results for that case.

• Computing the average results or range of results over a significantly large number of cases, then reduces any errors associated with the calculation.



Basics of the Monte Carlo Method (cont.)

• As a simple example of the Monte Carlo method, consider somebody rolling a six-sided die. • What is the probability of each of the numbers coming up, if the

die is rolled just one time?• If the die is rolled 6 times, what will the distribution of numbers

be?• How about if it’s rolled 100 times, 1,000 times, or 1,000,000?

• In theory, the more times the die is rolled, the less error that is associated with the resulting distribution of numbers.

To Try This and Other MC Examples Click on the Dice



Background (cont.)

• The FTFAM utilizes these techniques to generate values for several unknown variables, utilizing standardized distributions.• Fuel flashpoint temperature• Ambient ground temperature• Ambient cruise temperature• Flight mission length

• Additional functionality of the program:• Single flight analysis (for troubleshooting)• Random Number Freeze (for troubleshooting)• Warm day analysis• Flammability Reduction Method (FRM) effectiveness analysis



Program Overview – Flammability Analysis

Determination of Fuel Tank Flammability Exposure

Surrounding Environment

Mission Data

Fuel Properties

Fuel Tank Thermal Characteristics

Mission Length

Fuel Management

Mach Number

Flashpoint Temp.

Ambient Temp.

Total Air Temp.

Temp. Differential

Exp. Time Constants

Calculated by Monte Carlo

Method


Method


Method


MethodUser Input User Input User Input User Input

Ambient Pressure

Calculated by Model

Number of

Engines

User Input



Program Overview – Surrounding Environment



Surrounding Environment – Ambient Temperatures

• Ground and Cruise Ambient temperatures are calculated through Monte Carlo Method using distribution data generated by the 1998 ARAC

• Ground Ambient• Mean Temperature = 59.95 °F

• Standard Deviation Below 50% = 20.14 °F

• Standard Deviation Above 50% = 17.28 °F

• Cruise Ambient• Mean Temperature = -70 °F

• Standard Deviation = 8 °F




• With Ground and Cruise temperatures defined, we now need a method for computing the transition from ground to cruise

• To do so, several things must be understood and taken into account:• In general, the ambient temperature decreases with increasing

altitude. The rate of this decrease is referred to as the Temperature Lapse Rate.

• As altitude is increased, a point known as the Tropopause is reached. This is the boundary between the troposphere and stratosphere, and at this point there is no variation in ambient temperature. The tropopause ranges in height from 26,400 ft to 58,080 ft.

• When a cold ground ambient temperature exists, the initial temperature lapse rate can be significantly different than on warmer days. At times, it can even go positive, causing an increase in ambient temperature with altitude. This effect is referred to as a Cold Day Inversion.




• The program uses the following calculations to define the standard temperature lapse rate:• When the altitude is less than 10,000 ft

• When the altitude is greater than 10,000 ft

• The model defines the tropopause as the point at which the ambient temperature reaches the pre-selected cruise temperature, and at this point cuts off the lapse rate and holds the ambient temperature constant.

altTT grndamb 57.3

75.3 ambamb TT




• To deal with the issue of the cold day inversion, the program sets a different initial temperature lapse rate for flights with a ground ambient temperature less than 40 °F

• The lapse rate for these flights when the altitude is less than 10,000 ft is:

AltT

T gndamb

10

3.4




• For the landing portion of the flight, the program uses the same temperature lapse rates, however selects a new ground ambient temperature to ramp towards.

• In addition, the program takes into account that on long duration flights, an aircraft is flying into a new climate, thus causing a change in cruise temperatures.

• To handle this, for flights where the flight time is greater than 2 hours, the program ramps to a new cruise ambient temperature over a 45 minute period, starting just after the midpoint of the cruise cycle.



Surrounding Environment – Ambient Pressure

• Ambient pressure is calculated through direct relationship with altitude



Program Overview – Mission Data



Mission Data – Number of Engines

• Number of Engines is a user input that is used by the program to calculate the climb rate of the aircraft.

• This climb rate also depends on the percentage of the flight time (randomly generated) to the maximum flight time of the aircraft.

# of Engines <20% <40% <60% <80% <100%

2 20 20 30 30 353 25 30 35 35 404 25 35 40 40 45

% of Maximum Flight Time

Time (minutes) To Reach Initial Cruise Altitude



Mission Data – Mission Length

• Mission Length for each flight is selected at random through the Monte Carlo method

• The distribution data that is used in the Monte Carlo selection process comes from data determined by the 1998 ARAC and varies depending on the maximum mission length of the aircraft.

• The data is generated as a percentage of total flights existing in each 200 nm block up to the maximum mission length of the aircraft.



Sample Mission Range Distributions Based on

Maximum Range



Mission Data – Fuel Management

• The quantity of fuel in the tank at any point during flight has a direct impact on the thermal effects which determine fuel temperature, and thereby tank flammability.

• The model uses the quantity of fuel in the tank as part of the thermal model to determine the change in tank thermal time constants throughout the flight.

• To do this, the model assumes a constant rate of fuel usage between the time at which the tank starts to burn fuel and the point at which it is empty.

• The model uses this assumption of a linear rate of fuel usage to generate a linear change in tank thermal time constants.



Mission Data – Mach Number

• The mach number at cruise is given by a user input• Mach number changes with altitude are defined as:

• Altitude < 10,000 ftMach = 0.4

• 10,000 ft Altitude < 30,000 ft

• Altitude 30,000 ftMach = Cruise mach number

4.020

4.010

cruiseMach

AltMach



Mission Data – Miscellaneous

• There are a number of additional assumptions that the program makes in order to develop the full mission profile:

• The descent rate used is the same for all flights and is 2500 ft/min down to 4000 ft and 500 ft/min below 4000 ft.

• Any changes in cruise altitude are treated by the program to be instantaneous.

• For very short duration flights of less than 50 minutes, the climb rate is set to be 1750 ft/min.

• Preflight ground time is defined by the following table

• Postflight ground time is 30 mins for all flights.

< 3 hrs

Between 3 and 4

hrs > 4 hrsPreflight Ground

Time30 mins 45 mins 90 mins

Flight Time



Mission Data – Miscellaneous

• Three different cruise altitude steps are allowed to be defined by the user. These are treated differently depending on the length of each flight as follows:

• Flight Times < 50 min – This is a short up/down flight and the first cruise altitude step is never reached. 40% of the flight time is allocated for climb, and 60% for descent.

• Flight Times Between 50 and 100 minutes – The flight cruises at the first altitude step and does not step to the other levels.

• Flight Times Between 100 and 200 minutes – Two altitude steps are used with the step increase occurring midway through the cruise time.

• Flight Times > 200 min – All three altitude steps are used with the cruise time equally split between them.



Program Overview – Fuel Properties



Fuel Properties – Flashpoint

• The Standard Specification for Aviation Turbine Fuels (ASTM D 1655), specifies a minimum flashpoint value of 100 °F for Jet A fuel.

• Similar standards for other aviation fuels also only specify a minimum value.

• In attempt to determine the actual flashpoint of jet fuel as used in service, the FAA conducted a study in which 293 samples were taken from both domestic and international flights. Results of this study are published in FAA report DOT/FAA/AR-07/30.

• The results of that study are used by the program to develop the standardized distribution of flashpoints from which a Monte Carlo analysis can be performed.



Mean = 120°F

Std Dev. = 8°F



Program Overview – Fuel Tank Thermal Characteristics



Program Overview - Fuel Tank Thermal Characteristics

• As the fuel tank is heated or cooled by a change in outside air temperature or by heat input from various engines/systems, the fuel temperature will increase or decrease respectively.

• The program assumes that the response of the fuel temperature to these changes follows an exponential decay law.

• This exponential trend is represented in the program by utilizing several exponential decay time constants and equilibrium temperatures that the fuel will eventually reach if given sufficient time.

• In order to determine these inputs, sufficient flight test data and analysis of the thermal behavior of the fuel tank is required.




• For Ground Conditions, the input data required is:• Fuel temp., relative to ground ambient temp. that the fuel will

reach if given sufficient time.• Exponential time constant for a near empty fuel tank.• Exponential time constant for a near full fuel tank.

These inputs are required both for an engine on and engine off condition

• For Flight Conditions, the input data required is:• Fuel temp., relative to TAT that the fuel will reach if given

sufficient time.• Exponential time constant for a near empty fuel tank.• Exponential time constant for a near full fuel tank.




• As discussed previously, the program assumes a linear rate of fuel usage, and therefore the time constant values move in a linear fashion from the tank full to tank empty values.




• Using the tank thermal time constant and equilibrium temperature values, the program calculates the fuel temperature at each time step utilizing the following exponential decay law:

• This equation can also be used to solve for time constant input values, provided sufficient flight test data is obtained.

t

ifuelequil

ifuelifuel eTT

TT

11,

1,,




• From this example, the following inputs are determined:• Equilibrium Fuel Temp. on the Ground = Ambient Temp.• Equilibrium Fuel Temp. in Flight = TAT – 25°F• Exponential Time Constant, Tank Near Full = 149.8• Exponential Time Constant, Tank Near Empty = 29.9

• A comparison then of the model outputs and flight test data must be made to ensure accuracy of these inputs.



Program Overview – Flammability Analysis

Determination of Fuel Tank Flammability Exposure

Surrounding Environment

Mission Data

Fuel Properties

Fuel Tank Thermal Characteristics

Mission Length

Fuel Management

Mach Number

Flashpoint Temp.

Ambient Temp.

Total Air Temp.

Temp. Differential

Exp. Time Constants


Method


Method


Method


MethodUser Input User Input User Input User Input

Ambient Pressure

Calculated by Model

Number of

Engines

User Input



Answers to Questions From Tuesday

1. What are the major differences between version 9a and 10 of the FTFAM? Are there user’s manuals available from each of the previous versions?• The major differences are the way in which the time constants are calculated throughout

the flight, but there were numerous other changes. A file is available which lists all of the differences, and this will be made available. Either way, the only acceptable version is 10.

2. What is the basis for the time constant of 3500 for calculating oxygen evolution?

• Change in Partial Pressure (PP) of O2 in fuel = (PP of O2 in Ullage - PP of O2 in Fuel) *(1-e^-t/tau)

• Based on gradients between ullage and liquid. So, by setting tau at 3500 you are saying that under quiescent conditions, the O2 is going to come out of the fuel very very very slowly. Setting it to 100 during climb creates a much more rapid evolution of O2 out of the fuel. This is what you would expect as the fuel now is being agitated. There is a report of O2 evolution and some tests ran (DOT/FAA/AR-05/25) These time constants were derived from comparing some of these calculations to the test data, doing some curve fitting. Also, one is not required to use these time constants, or this method for taking the O2 evolution effect into account. This is merely one method that is approved. If one is able to perform tests and validate some other way of modeling it, then they certainly can. Also, the GBI report had some additional data, that report # is DOT/FAA/AR-01/63.



Answers to Questions From Tuesday

3. Although both the ground and cruise ambient temperatures are chosen from a standardized distribution, are the two linked? For instance, if a ground ambient is a very hot temperature, would the cruise ambient be chosen as a warmer than usual temperature?• No, they are not linked...each is randomly generated

independent of the other. Remember, the temperature decays according to the defined lapse rate, so depending on the temperatures generated, the cruise temperature may never be reached.

4. What is the rational behind the discretization of the mission length histograms into 200 knot blocks?• No specific rationale for the 200 nm blocks...at least not that

we know of. Just how the data was broken down at ARAC, and then carried over to this version



Program Overview

• The program itself is split into several separate worksheets, which can be categorized as shown below.

• In all of these worksheets, a yellow cell denotes a user input cell.• Any cell not shaded yellow must remain unchanged by the user

unless approved by the FAA



Program Overview – Worksheet Descriptions

• Intro

Provides a brief statement of the model’s intended purpose as well as notes and limitations of its use.

• User Inputs & Results

Main interface of the FTFAM. This worksheet contains all user inputs necessary for performing a Monte Carlo flammability analysis as well as results from the analysis.

• Single Flight

Allows the user to simulate and evaluate a particular flight scenario, either by entering specific data or by selecting a flight from the MC analysis. Results from this single flight are also displayed here.



Program Overview – Worksheet Descriptions

• FRMAllows the user to evaluate the effectiveness of an FRM. It contains all necessary user inputs and the results for an FRM analysis. This worksheet is only needed if an FRM analysis is being conducted

• Summary of n CasesDisplays the results of each flight in tabular format, sorted by the percentage of flammable time.

• Internal Calculations WorksheetsThese four worksheets contain all of the essential information processed by the model. All data inputs, claculated values, and results are stored here for use by the program and/or user. These worksheets should not be modified by the user in any way, and information contained in them is provided to the user for troubleshooting purposes only.



User Inputs & Results Worksheet



FTFAM Usage - User Inputs

• User inputs for the program are divided into six categories• Airplane Data• Flight Data• Fuel Tank Usage Data• Body Tank Input Data• Tank Thermal Data• Multi-Flight Monte Carlo Data




• User inputs for the program are divided into six categories• Airplane Data

• Maximum range of aircraft

• Number of engines

• OAT cutoff temperature

• Flight Data• Fuel Tank Usage Data• Body Tank Input Data• Tank Thermal Data• Multi-Flight Monte Carlo Data




• User inputs for the program are divided into six categories• Airplane Data• Flight Data

• Cruise Mach number

• Tank ram recovery

• Cruise altitude steps

• Fuel Tank Usage Data• Body Tank Input Data• Tank Thermal Data• Multi-Flight Monte Carlo Data




• User inputs for the program are divided into six categories• Airplane Data• Flight Data• Fuel Tank Usage Data

• Tank full/empty times

• Engine/equipment start time

• Body Tank Input Data• Tank Thermal Data• Multi-Flight Monte Carlo Data




• User inputs for the program are divided into six categories• Airplane Data• Flight Data• Fuel Tank Usage Data• Body Tank Input Data (This refers to tanks completely enclosed in the

fuselage, or similar container with no direct cooling to ambient air)

• Is the tank in the fuselage?

• If yes, then what is the temperature of compartment surrounding tank?

• Is the tank pressurized?

• If yes, then what is the pressure differential?

• Tank Thermal Data• Multi-Flight Monte Carlo Data




• User inputs for the program are divided into six categories• Airplane Data• Flight Data• Fuel Tank Usage Data• Body Tank Input Data• Tank Thermal Data

• Fuel temperature differentials relative to ambient• Exponential time constants (define how fuel heats/cools in

response to heat input)

• Multi-Flight Monte Carlo Data




• User inputs for the program are divided into six categories• Airplane Data• Flight Data• Fuel Tank Usage Data• Body Tank Input Data• Tank Thermal Data• Multi-Flight Monte Carlo Data

• Number of Flights

• Random number freeze? (y/n)

• Warm day analysis? (y/n)



Program Overview - Main Calculations



FTFAM Usage - Outputs

• Monte Carlo Flammability Analysis• Table displaying the following

data for each flight (<5000):• Preflight ground time

• Flight time

• Ambient temperature

• Cruise temperature

• Fuel flashpoint temperature

• Amount of time that the tank was flammable

• % of flight time that the tank was flammable

• Table displaying warm day (ground ambient temperature > 80F) results

• Chart showing a summary of the Multi-Flight Monte Carlo Analysis

Fleet Average Flammability Exposure

0

20

40

60

80

100

0 200 400 600 800 1000

Mission NumberF

lam

mab

ilit

y- p

erce

nt

Fleet Average Flammability Exposure %

22.62%



FTFAM Usage - Outputs

• Single Flight Analysis• Time-based and altitude-based plots of fuel temperature, TAT,

LFL and UFL

Flammability studies

-40

-20

0

20

40

60

80

100

120

140

160

180

200

0 100 200 300 400 500 600 700 800 900 1000

Time- minutes

Tem

per

atu

re-

Deg

F

0

20

40

60

80

100

tfuel

tat

LFL

UFL

Flammability %

Tank Flammable

Ambient and TAT Temperatures vs Altitude

0

5000

10000

15000

20000

25000

30000

35000

40000

45000

-40 -20 0 20 40 60 80 100 120 140 160 180 200

Temperature Deg F

Alt

itu

de

100

0's

ft TAT

FuelTemp

LFL

UFLTank Flammable



FTFAM Usage – Modifications to the Program

• There are certain aspects of the program’s code that may need to be modified by the user based on acquired aircraft data.

• Fuel Management• As discussed previously, the program assumes a constant rate

of fuel usage, thereby resulting in a linear rate of change in the tank thermal time constants.

• It does not account for unique fuel management techniques, such as fuel transfer systems.

• If the linear decay model that the program uses does not adequately represent the fuel burn process, this portion of the program code can be modified.

• Any code modification must be shown, by flight test and a detailed analysis of the tank’s usage of fuel backed up by data, to provide an accurate representation of fuel usage by the tank.




• Tank Thermal Characteristics• As discussed previously, the thermal behavior of the fuel tank

due to its surroundings is based on calculations using the fuel temperature differential relative to ambient temperature and TAT, as well as the exponential tank thermal time constants.

• If flight test data and a detailed analysis of the tank’s thermal behavior shows that this method cannot yield an accurate representation of the actual fuel temperature profile, then this portion of the program code can be modified.

• Any code modification must be shown, by flight test and a detailed analysis of the tank’s thermal behavior backed up by data, to provide an accurate representation of the fuel temperature profile.



The latest version of the FTFAM and its associated

User’s Manual can be downloaded at the

FAA’s Fire Safety Section Website at:

http://www.fire.tc.faa.gov/systems/fueltank/FTFAM.stm

Any updates to either the FTFAM or its associated User’s Manual will be posted in the Federal Register



Example* – Heated Center Wing Tank

• Consider an example aircraft with a heated center wing tank (CWT). The aircraft has 4 engines, a maximum range of 10,000 nm, and is designed to fly with a typical cruise speed of mach 0.81. A flight test is flown to determine the CWT’s thermal characteristics.

Click Here to Open a Spreadsheet with Sample Data and Calculations to Determine Inputs to the FTFAM

• The following slides show screenshots of the FTFAM displaying inputs to the program based on this data as well as a comparison of the calculated results to the sample flight data results.

*note that this example, or any other contained within this presentation, do not represent any particular aircraft or fuel tank configuration, and any data provided is for

demonstration purposes only.



Single-Flight Conditions1320 max

Flight time 398.0 MinutesT.O. Ambient Temp 80.0 Deg F Flammability ExposureEarly Cruise Ambient -63.6 Deg F 80.31 %Late Cruise Ambient* -63.6 Deg FLanding Ambient* 80.0 Deg FFlash Point 115.0 Deg F

Run selected flight from Monte Carlo as a single flightEnter Case number

Flammability studies

-40-20

020406080

100120140160180200

0 100 200 300 400 500 600

Time- minutes

Tem

pera

ture

- Deg

F

0

20

40

60

80

100

Fuel Temp

TAT

LFL

UFL

Flammability %

Ambient and TAT Temperatures vs Altitude

0

5000

10000

15000

20000

25000

30000

35000

40000

45000

-60 -40 -20 0 20 40 60 80 100 120 140 160 180 200

Temperature Deg F

Alti

tude

100

0's

ft

TAT

FuelTemp

LFL

UFL

Run Single Flight

Run Selected Flight

Return to Input page

the fuel tank flammability assessment method – flammability analysis federal aviation...

Documents