f mpv.

I FE
Fh

N AFWAL-TR-88-2101

HOT SURFACE IGNITION TESTS OF AIRCRAFT FLUIDS

A. H. Johnson
A. J. Roth N. Albert Moussa
BOEING ADVANCED SYSTEMS BLAZETECH CORPORATION
Propulsion Technology 145 Highland Avenue
P.O. Box 3707 WinChester
Seattle, WA 98124-2207 MA 01890-1435

November 1988

FINAL REPORT for period May 1987 to May 1988

Approved foiý public release; distribution is unllimitedfl DI

A" ELECTE n
APR 2 8 1•98
AERO PROPULSION LABORATORY
AIR FORCE WRIGHT AERONAUTICAL LABORATORY
AIR FORCE SYSTEMS COMMAND
V
WRIGHT-PATTERSON AIR FORCE BASE, Oi 45433- 6523 ..

A 2 0
AOý -
97-
 I ,I

NO•ICE

Vlhen Government drawings, specifications, or other data are used for any
purpose other than in connection with a definitely Government-related
procurement, the United States Governmer.t incurs no responsibility or any
obligation whatsoever. The fact that the Government may have formulated or in
any way supplied the said drawings, specifications, or other data, is not to
be regarded by implication, or otherwise as in any manner construed, as
licensing the holder or any other person or c.rporation; or as conveying any
rights or permission to manufacture, use, or sell any patented invention that
may in a way be related thereto.

This report has been reviewed bj the Office of Public Affairs (ASD/PA)
and is releasable to the National Technical Information Service (NTIS). At
NTIS, it will be available to the general pitblic, including foreign nations.

This technical report has been reviewed and is approved for publication.

.POBERT-. CLODFEL•; Projct Engineer CHARLES L. DELANEY, ChiefV

Fuels BranchO' Fuels Branch
Fuels and Lubrication Division

FOR THE COMMANDER

BENITO P. BOIT-ERI, Assistant Chlef

Fuels and Lubrication Division
Aero Propulsion e: Power Laboratoly

if your address has changed, if you wish to be removed from our mailing
list or if the addressee is no longer employed by your organization please
notify WRDC/POSF, WPAFB, OH 45433- 6523 to help us maintain a current mailing
list.

Copies of this report should not be returned unless return is required by

security considerations, contractual obligations, or notice on a specific
document.
 UHCLASSIEIED
SECURITY CLASSIFICATION OF THIS PAGE

REPORT DOCUMENTATION PAGE

is REOTE RT kg ICAT ION It. RESTRICTIVE MARKINGS

2.. SECURITY CLASSIFICATION AUTHORITY 3. DIST RIUBUT!ON/AVAI LADBILITY OF RF:'0RT

2b, ECL3311 DNG

IN/OWNGA CT CHEULEApproved for public rele~ase, distribution
D NC/A S S F C T O / I W G A I GS H D L lted
is unli mim

4. PERFORIMING ORGANIZATION REPORT NUMBER(S) 5. MONITORING ORGANIZATION REPORT NUMBER(S)

AFIA2L-.TR-88-2 101
G&.NAME OF PERFORMING ORGANIZATION M. OFFICE SYMBOL 7a. NAME OF MONITORING ORGANIZATION

* Boeinq Advanced Systems Wright Research and Development Center

6c. ADORESS (City. State and ZIP Code) 713.ADUAIESS (City. Statz and ZIP Code)

Seatle,
9124-207Wright
A Patterson AFB, OH 45433-6523
Be. NAME OF FUNDING/SPONSORING Bb. OFF ICE SYMBOL 2. PROCUREMENT INSTRUMENT IDENTIFICATION NUMBER
ORGANIZATION 1If oFpIicabIA) F31-4C2
Aero Propulsion & Power Lab tWROC/POSF _______________________________
Sc. ADDRESS IWity. State and ZIP Code) 10. SOURCE OF FUNDING NOS. _______

Wright Research and Development Center PROGRAM PROJECT TASK WORK~ UNIT
Wright-Patterson AFB OH 45433-6523 ELEMENT NO. NO. NO No.

1.TTE(nldSeuiyCwfcto)62203F 3046 07 94
12. PERSONAL AUTHOR IS)

A. M. Johnson, A. J. Roth and N.A. Moussa

14 OFREPORT 3b. TIME COVERED 14. D)ATE OF REPORT (Yr., Mo., Day) 15. PAGE COUNT
RA F FOM Ma 87ý TOj.j November 1988 233
10. SUPPLEMENTARY NOTATION

17. COSATI CODES 18. SUBJECT TERMS (Con gin&., on reverse if necessary anj1 ident~fy by block numIberl
FIED
GOUPSUB OR Hot surface ignition, MiI-H-5606, Mil-H-83282 ,MilL-7808
15. ABS !AC I f IeesrJP - 4, *IP - 8, engine bleed duct, UNV fire~ detection
19 ASRAT(Conginur on murfe ifnee.rua identify by block numhar)

Five fluids commionly f~und in aircraft engine components, JP-4 and JP-8 fuels, Mil-H-5606
and Mil-H-83282 hydraulic fluids and Mil-L-7808 lubricating oil, were tested in the
Aircraft Enqine Nacelle Fire Test Simulator (AENFTS) to define their Minimum Hot Surface
lanition Temperature (MHSIT's) when introduced as a spray or stream onto a hot
engine bleed duct. The test employed a simple, uncluttered test section and a
realistically simulated portion of the F-16 cngine compartment. MHSIT's for all but
* Mil-H-83282 were consistently found to be higher than the fluids autoignition temperature.

2M. DISTRIBVTION/AVAkILABILITY OF ABSTRACT t21. ABSTRACT~ SECURITY CLASSIFICATION

UN C LASS I FI E D/UNLIMITED (I SAME AS RIPT. (J OTIC VSERJ C0 UNCLASSI FIED

'QA 226- NAME OF RESPONSIBLE INDIVIDUAL 2b. TELEPHONE NUJMBEA 22c. OFFtCE SYfMBOL
____________________________(5_ ~ 53) 255-4208 IIRDC/POSF

DDFORjM 1i3 8 P EDITION ')r IJAN i3IS OBSOLETE. UNCLASSIFIED

SLCURITY CLASSIFICATION OF THIS PAGE
 SUMMARY

Hot Surface Ignition Temperature (HSIT) testing was performed in the

Aircraft Engine Nacelle Fire Test Simulator (AENFTS) located in building
71-B at Wright-Pattersor Air Force Base. The objective of this test
program was to measure Minimum Hot Surface Ignition Temperatures (MESIT's)
of five common aircraft fluids ;Mil-H-5606 and tlil-H-83282 hydraulic
fluids, JP-4 and JP-8 fuels and Nil-L-7808 lubricating oil) using an air-
heated bleed-air du:t in a high realism test article.

First, tests were conducted on a Tirrle piece of Inconel bleed duct in a

bare engine compartment with spray, of JP-4 fuel and Mil-H-5606 hydraulic
fluid. The purpose of these tast was to checkout systems and techniques
and acquire data which could be Orectly compared with data from previous
tests. (Results of these ttst& are presented in Table 4, page 59.)

Next, a simulation of a portion of the F-16 engine compartment was inserted

into the AENFTS and the fivi aL•craft fluids, JP-4 and JP-8 fuels, Mil-H.-
5606 and Mil-H-83282 hydraulic "J.,tids and Mil-L-7806 lubricating oil were
injected as spray or streuins clto various locations on the hot bleed-air
duct. Variables including vetilation air pressure, temperatlure, velocity
and fluid flowrate were varitd to study their effct on the MHSIT of these
fluids, (Results oi these tests are presented in Tables 7, 8 and 9, pages
73, 74 and 75.)

The lowest MHSIT's identified for each of the fluids at 14.4 psia and 120OF
were:

0 5606's MHSIT was found to be 700°F when it was streamed onto the
hot bleed duct at 2 ml/second with 1 ft/setond ventilLt iop
airflow.

0 83282's MHSIT was found to be 750°F when it"'wis spravp d9n, the.'
hot bleed duct at 8 ml/second with both 0 tand,4. ft/second.
ventilation airflow.

i •
 " 7808's !HSIT was found to be 990°F when it was streamed onto the
hot bleed duct at 2 ml/second with 1 and 2 ft/second ventilaticn
airflow.

" JP-4's MHSIT was found to be 1150OF when it was sprayed onto the
hot bleed duct at 8 ml/second with 1 and 2 ft/second ventilation
airflow.

"o JP-8's MHSIT was found to be 1100OF when it was sprayed onto the
hot bleed duct at 8 ml/second with 2 ft/second ventilation
airflow.

Heating the ventilation airflbw was found to reduce the MHSIT's for all
five fluids. The effect of varying the ventilation airflow pressure on
MHSIT's was complicated by the AENFTS's requirement for different airflows
fo liuea: ri iuaintesting. in general, higher pressures •

produced lover MHSIT's, although none were identified at the highest

pressure investigated (20 psia) which were lover than those noted above,
probably because a higher airflow velocity was required to obtain this
pressure.

Finally, the results were examined in light of simplified analyses of the

key p..ce.ses i, -J chemical kinetics and droplet atomization,
-s
dynamics and heating/evaporation (for spray) and nucleate versus film
boiling (fur streams). Thus, the observed MRSIT differences and
similarities between the various fluids and between sprays and streams were
interpreted.
aooeanion 70!.•
IGORO Fos
•GRA&I-
X IS
I WC I
DTIC TABc
uatllvawN ý.PECTF.D)

-A Avellabiltity eCogO
Uy and/o r
'ii
 PREFACE

This is a final report of work conducted under F33615-84-C-2431 by the

Boeing Military Airplane Company, Seattle, Washington, during the period
May 1987 through August 1988. Program sponsorship and guidance are
provided by the Fire Protection Branch of the Aero Propulsion Laboratory
(AFWAL/POSF), Air Force Wright Aeronautical Laboratories, Air Force Systems
Command, Wright-Patterson Air Force Base, Ohio, under Project 3048, Task 07
and Work Unit 94. Robert G. Clodfelter was the program manager.

The contents of this report cover a portion of the worh defined under Task
III of the contract, AEN (Aircraft Engine Nacelle) Test Requirements. In
general, the task requires utilization of the AEN fire test simulator to
establish the fire initiation, propagation, and damage effects exhibited by
aircraft combustible fluids under representative dynamic operation
environmental conditions, followed by the evaluation and development of
protection measures.

Other documentation which has been submitted, as part of this contract,

includes!

AFWAL-TR-87-2004 Effects of Aircraft Engine Bleed Air Duct Failures on

Surrounding Aircraft Structure, April 1987.

AFWAL-TR-87-2060 Development and Evaluation of an Airplane Fuel Tank

Ullage Composition Model:
Volume 1: Airplane Fuel Tank Ullage Computer Program, Oct. 1987.
Volume I!: Experimental Determination of Airplane Fuel Tank Ullage
Composition, Oct. 1987.

AFWAL-TR-87-2089 Optical Fire Detector Testing in the Aircraft Engine

Nacelle Fire Test Simulator, March 1988.

AFVAL-TR-88-2022 Fire Extinguishing Agent Evaluation in the Aircraft

Engine Nacelle Fire Test Simulator, June 1988.
Iii
AFWAL-TR-88-2031 Advanced Air Separation Module Performance Evaluation,
July 1988.

AFWAL-TR-88-2123 OBIGGS Preliminary Design Studies for A-6, P-3 and F-18
Aircraft (to be released about 28 Feb. 1989).

Boeing wishes to acknowledge with appreciation the contributions of N.

Albert Moussa of BlazeTech Corporation, who joined this program in
September of 1987 and helped with test planning, analysis and
interpretation of data and documentation, and the technical personnel of
SelectTech Services, Inc., in particular, A.J. Roth, who planned and
performed the test and documented the test results.

A key Boeing contributor was C. L. Anderson, who provided technical

guidance.

iv
 TABLE OF CONTENTS
PAGE
1.0 INTRODUCTION 1
1.1 Background 1
1.2 objective 9
1.3 Approach 10

2.0 TEST FACILITIES 13

2.1 AENFTS Facility 13
2.1.1 F-16 Nacelle Simulator 16
2.1.2 Fluid Delivery System. 21
2.2 Test Article 25
2.2.1 Simple Duct TesL Apparatus 25
2.2.2 High Realism Test Article 34
2.3 Data Collection and Reduction 42

3.0 TEST PROCEDURE 49

4.0 TEST RESULTS 53

4.1 Simple Duct Tests 53
4.2 High Realism Tests 65
4.2.1 Conditions of Fluid Injection 67
4.2.2 The Effect of Ventilation Air Conditions
on aAmWT 72

4.2.2.1 Spray 72
4.2.2.2 Stream 84
4.2.2.3 Comparison of High Realism Stream vs.
Spray 92
4.2.2.4 Comparison of High Realism and Simple
Duct Results 96
4.3 Summary of Results 102
4.3.1 Simple Duct 102
4.3.2 High Realism 103
4.3.3 Test Article, Facility and Technique 105

V
 TABLE OF CONTENTS

PAGE
5.0 ANALYSIS AND INTERPRETATION OF THE RESULTS 106
5.1 Spray Analysis 106
5.1.1 Predicted Droplet Behaviors for the
Five Fluids 107
5.1.2 Measured vs. Predicted Trends in MHSIT's
for the Jet Fuels 107
5.2 Stream Considerations 113
5.3 Discussion of Spray vs. Stream Results for the
Five Fluids 115

6.0 CONCLUSIONS AND RECOMMENDATIONS 124

6.1 Conclusiois 124
6.2 Recommendations 125

REFERENCES 127

APPENDICES
APPENDIX A: Summary of Hot Surface Ignition Test
Data A-i
APPENDIX B: Temperature Data Uncertainty Aualysis B-i
APPENDIX C: Pertinent Aircraft Fluid Properties C-i
S. .. Spray Analysis
APPENDIX E: Determination of Boiling Regimes for
Hot Surface ignition Tests E-i

vi
 LIST OF FIGURES

FIGURE DESCRIPTION PAGE

1. Summary of Results of Previous Hot Surface

Ignition Studies 3
2 Simple Duct Test Article 11
3 High Realism Test Article 12
4 AENFTS Facility 14
5 AENFTS Test Section 15
6 F-100 Engine Showing Accessories for F-16
Simulator 17
7 F-16 Simulator Engine Side 18
8 Rear View of F-16 Simulator Prior to Installation 19
9 Front View of F-16 Simulator Prior to installation 19
10 Closeup of Fire Zone of F-16 Nacelle Simulator 20
13, F-16 Nacelle Simulator Installed in AENFTS Test
Section 20
12 Installation of Baffle to Redirect Airflow in F-16
Simulator 22
13 Schematic Diagram of Fuel Injection System 23
14 Details of the Hot Air and Electrically Heated
Simple Duct Test Articles 46
15 Electrically Heated Simple Duct Test Article 27
16 Cutaway Drawing of Simple Duct Test Article 28
17 Simple Duct Thermocouple Locations 29
18 Temperature Variation Along Air-Heated Simple Duct 31
19 Temperature Variation Along Resistance Heated
Simple Duct 32
20 High Realism Test Article 35
21 Fluid Injection and Thermocouple Locations on High
-Realism Test Article 37

Vii
 LIST OF FIGURES (Continued)

FIGURE DESCRIPTION PAGE

22 Temperature Variation Along High-Realism Duct 39

23 Video Camera Placement 41
24 Duct Temperature Variation With Time During Simple
Duct Test 43
25 Schematic Diagram of Computer Data Acquisition
System 46
26 Sample Plot of Fire/No Fire Test Data 51
27 Effect of Va on Simple Duct MHSIT for 5606 55
28 Effect of Va on Simple Duct MHSIT for JP-4 56
29 Effect of Va on Air Heated Simple Duct MHSIT 57
30 Effect of Va on Resistance Heahed Simple Duct
MHSIT 58
31 Variation in Duct Temperature with Time During
Air-Heated and Resistance Heated Simple Duct Tests 62
32 Effect of Spray and Stream Location on MHSIT with
High-Realism Test Article 68
33 Effect of 5606 Spray Flowrate on MHSIT 70
34 Effect of Pa on MHSIT with Flaid Spray 76
35 Effect of Ta on MHSIT with Fluid Spray 78
36 ffect. of Ta ony 83282' MUSIT' with.l V1-4A Spray an

37 Effect of Va on MHSIT with Fluid Spray 83

38 Effect of a Baffle on MHSIT with Fluid Spray 85
39 Effect of Va on MHSIT for 5606; Upstream vs
Downstream Spray 86
40 Effect of Pa on MHSIT with Fluid Stream 87
41 Effect of Ta on MHSIT with Fluid Stream 90
42 Effect of Va on MKSIT with rluid Stream 91
43 Effect of a Baffle on MHSIT with Fluid Stream 93

viii

~~!
 LIST OF FIGURES (Continued)

FIGURE DESCRIPTION PAGE

44 Spray vs. Stream at Location 3; Effect of Pa on

MHSIT 94
45 Spray vs. Stream at Location 5; Effect of Pa on
MHSIT 95
46 Spray vs. Stream at Location 3; Effect of Ta on
MHSIT 97
47 Spray vs. Stream at Location 5; Effect of Ta on
MHSIT 98
48 Spray vs. Stream at Location 3; Effect of Va on
MHSIT 09
49 Spray vs. Stream at Location 5; Effect of Va on
MHSIT 100
50 Comparison of Simple Duct and Sigh Realism MHSIT 101
51 Effect of Air Pressure on MHSIT; Theory and
Experiment 110
52 Effect of Air Temperature on MHSIT; Theory and
Experiment ill
53 Effect of Air Velocity on MHSIT; theory and
Experiment 112
54 JP-4 Correlation Based on a Simplified Expression
of Ignition Delay and Transit Time 118
55 JP-8 Correlation Based on a Simplified Expression
of Ignition Delay and Transit Time 119
56 5606 Correlation Based on a Simplified Expression
of Ignition Delay and Transit Time 120
57 83282 Correlation Based on a Simplified Expression
of Ignition Delay and Transit Time 121
58 7808 Correlation Based on a Simplified Expression
of ignition Delay and Transit Time 122

ix
 LIST OF TABLES

TABLE PAGE

1 AENFTS Temperature Instrumentation 47

2 AENFTS P:essure Instrumentation 48
Simple Duct Test Matrix 54
4 Summary of Results of Simple Duct Tests 59
5 High Realism Test Matrix 66
6 Effect of Stream Flowrate and Injection Time on
MHSIT 71
7 Summary of the Effect of Air Pressure on MHSIT 73
8 Summary of the Effect of Air Temperature on MHSIT 74
9 Summary of the Effect of Air Velocity on MHSIT 75
10 Autoignition Temperature Test Results for Fluids
Used in AENFTS Hot Surface Ignition
Test Program 81
11 illustrative Results for Spray Analysis 108
 1.0 IT•rODUMIONI

This report describes tests performed to define Hot Surface Ignition

Temperature (HSIT's) for five fluids commonly found in an aircraft engine
compartment, Mil-H-5606 and Mil-H-83282 hydraulic fluids, Hil-L-7808
lubricating oil and JP-4 and JP-8 fuels, when injected as sprays and
steams onto a hot Inconel engine bleed duct. For simplicity, these fluids
will henceforth be referred to as 5606, 83282, 7808, JP-4 and JP-8. These
tests were performed to provide a better understanding of the mechanism
and risk of hot surface ignition in an aircraft engine compartment and to
improve the existing data base available to the aircraft designer.

The hot surface ignition tests were conducted with two test articles:

o a short section of bleed duct mounted in an uncluttered test

section, heated alternately by electrical resistance heaters
and by hot high-pressure air

o a FlOO-PW-200 engine right-side bleed duct mounted in a test

section cluttered by actual engine components and simulated
aircraft structure

The test facility employed, the Aircraft Engine Nacelle Fire Test
Simulator (AENFTS) is equipped so that the velocity, pressure and
temperature of its airflow, simulating engine compartment ventilation air,
could be varied to represent a variety of aircraft flight conditions.

1.1 Background

The Auto-Ignition Temperature (AIT) of the fluids present in an aircraft

engint. compartment is often used as a guideline when determining the
maximum surface temperatures which should be allowed in the compartment.
A fluid's AIT is that temperature at which its vapors will ignite in air
at atmospheric pressure without an external source of ignition. Most
published AlT data has been acquired in accordance with the method of ASTM
D 2155. ASTM D 2155 was recently replaced by ASTH E 659, which employs a
larger flask and is considered to provide more reliable data.

AIT test conditions are much different than those in an aircraft engine
compartment, so attempts have been made in the past to measure Minimum Hot
Surface Ignition Temperatures (MHSIT's) that would more closely represent
aircraft engine compartment component limit temperatures. HUSIT's avove
the AIT of the fluid, sometimes hundreds of degrees Fahrenheit higher,
have been• determined during past test programs.

A summary if the results of previouo ho': surface ignition studies

(References 1 to 5) is given on Figure 1. The plotted temperatures should
be reduced as noted for each study in order to estimate the temperature at
which ignition would not occur. This is due to test methodology and
0
measurement errors. A 5 F reduction in the ASTH D 2155 AIT value is
acceptable due to the known precision of this test method.

Much of the past HSIT test work has been done with greatly simplified test
articles vhicub were intended to simulate hot surfaces, primarily bleed-air
ducts, in aircraft engine compartments. The HSIT's determined in these
tests varied considerably, probably becausa of differences in test methods
and equipment. None of these tests employed a test article which closely
simulated an actual aircraft engine compartment. Hence the aircraft
engine compartment design information available from documentation of
these programs has been difficult to use. A summary of prior test efforts
follows:

Rolls Royce

In the mid-1960's, G. Beardsley of the Rolls-Royce Ltd. Aero Engine

Division investigated the hot surface ignition of kerosene in a wind
tunnel having an electrically heated floor (Ref. i). The effect on hot
surface ignition temperature of variation in the tunnel velocity and the
presence of assorted turbulence causing obstructions on the tunnel floor
was also investigated.

2
 w
z :10

00 U. -
a.~U w LLI
+ to~

WWC
zC3 CD
Go5~ ( 0 CDC
AC C.g. 0 '

&Z w 3 EICL =

OL6

-L'
04,

0 " Z 0-4C

.j cC 0I-'U (

W a
co I -

e 03 ^ ^
EZ IL C 4 ..0 - V).4.

N~~Ie % 0 f ~ W-J cc
ix V M*+CL*ad

3W 0 0vvý
The tunnel employed in these test- w-is a rectangular low speed wind
tunnel, about 3.9 inches high and 6.9 inches wide. The tunnel floor had
an electrically heated working section 18 inches long. Ventilation air at
room temperature and ambient pressure was supplied to the wind tunnel at
velocities up to 10 ft/sec. Kttrosene was sprayed onto the hot rectangular
tunnel floor at 500 psig (3.6 - 3.8 ml total volume).

The test procedure that was followed was to select a tunnel floor
temperature and a ventilation air velocity that would produce ignition and
then increase the ventilation air velocity in increments of 0.1 ft/sec
while the tunnel floor temperature was held constant until no ignition
occurred. This was repeated and the lowest air velocity which did not
produce ignition in 10 attempts was recorded. Air velocities suffi"4Ient
to prevent ignition of the kerosene were determined for a range of tunnel
floor temperatures from 700°F to 1472 0 F.

The effect on thesne hot surface ignition temperatures of the installatioi,

of a variety ot eistructions on the tunnel floor was also investigated.
These obstructions were made from steel, were 1 inch thick, varled in
height (from 0.25 to 2.5 inches), width (from 1 inch to the full tunnel
width) and distance from the tunnel exit (4, 7, 10 and 13 inches).

The effect of these obstructions on the hot surface ignition temperature

varied with the size and location of the obstruction and with the tunnel
floor temperature being tested. Installing the .mnalleL QUSLIuctIioh
generally resulted in higher airflow velocities being required to prevent
ignition at a given tunnel floor temperature.

This study, while limited to kerosene, was one of earliest hot surface
ignition test programs applicable to aircraft engine compartment design
because, with its variety of flow obstructions, it did address the fact
that engine compartments also have a variety of flow obstructions which
create local velocity regions both above and below the average compartment
velocity.
Myronuk

Hyronuk (Ref. 2) studied the hot surface ignition of aircraft fluids in

the late 1970's. A laboratory scale engine compartment fire simulation
system at the Ames Research Center was used. The objective of this test
program was to determine the MHSIT for a variety of fluids and to
investigate the effect of the heated surface material, ventilation
velocity, fluid flowrate, injection method and heated surface
irregularities and obstructions on these MHSIT's. Aircraft fluids that
yere tested inoluded JP-4, JP-5, 5606 and 83282. The test article used
was a 30 iWchl section of the outside surface of stainless steel or
titanium alloy pipe, 3 inches in diameter and 39,4 inches long. The test
article was heated by a premixed propane-air flame inside the test
section.

Conclusion-, reached during this program included:

o HSIT's increased for all fluids tested as ventilation velocity

was increased

0 HSIT's were higher for more volatile fluids

o HSIT's may not be inferred from the AIT's of the fluids

Measured MHSIT's for test fluids at 2.6 ft/sec velocity and 15 ml/sec
fluid spray for 1 second (in 68OF air at ambient pressure), 1058°F for JP-
4, 878°F for 5606 and 797°F for 8328, were all well above the AIT's for
those fluids.

The nature of the test article and technique may have compromised the
application of these results to aircraft engine compartment design: (1)
4
The simulator used flame heated stainless steel or titanium pipe as ts
hot target. Since the surface material can atfect the MHSIT, and inconel
bleed ducts are frequently the hottest component in and aircraft engine
compartment, this choice may introduced bias into the MHSIT data. (2) The
ttst procedure called for increasing the surface temperature from one test
5
to another until ignition was achieved. In the present program it was
found that this technique led to falsely high MBSIT's. (3) The effects
of air temperature and pressure on MHSIT's were not addressed.

Parts

Parts performed a serieg of aircraft fluid flammability tests in the late

1970's (Ref 3). Using tquipment built at Monsanto for the test program,
hydraulic fluid and lubricating oil AIT's, heats of combustion and MHSIT's
were measured. A semi-automatic autoignition test apparatus comprised of
a crucible furnace, a temperature controller and a Vycor flask was used to
measure fluid AIT's and a oxygen bomb calorimeter was used to determine
the heats of combustion of the test fluids.

The determination of hot surface ignition temperatures was performed on a

hot manifold ignition test apparatus (Federal Test Method Standard No.
7915, Hethod 6053, 1.
5 January, 1969)1. The• t-es
---- -t consisted of a
stainless steel box with the hot target, a resistance heated 24 inch long,
3 inch diameter stainless steel pipe, suspended inside the box. Fluid was
sprayed through an oil burner nozzle (hollow cone, 80 degree spray-angle,
rated at 1.5 gph for oil at 100 psig). All test injections were made at a
pressure of 1000 psi for 1 second. Fluid streams were supplied by
burettes of three different sizes (flovrate = 0.35 ml/sec, 1.0 ml/sec, 1.7
ml/sec) and up to 25 ml were Injected during each test.

The lcjest minimum hot surface ignition temperatures recorded were 730°F
for 5606, 630°F for 83292 (below the fluid's AIT) and 1300OF for 7808 and
JP-4 and 1200OF for JP-8. Streams of fluid appeared to ignite at lower
surface temperatures than sprays.

As there were no provisions for ventilating the t-st article in a

controlled manner, the Parts test program did not address the effect of
ventilation air velocity, temperature or pressure on MHSIT. Because
ventilation airflow velocity has been found to have a major effect on
MHSIT's, these data cannot be directly applied to ventilated aircraft
engine compartment design. Because the test article wao fabricated of

E
6
stainless steel and heated electrically, some additional deviation from
the actual hot surface Ignition situation found in a normal aircraft
engine compartment could be expected.

Strasser

Strasser performed hot surface ignition tests using JP-4, JP-8, 5606,
83282 and 7808 (Uet. 4). The fhlids were injected (in a stream at various
flowrates) onto the hot surface, a cylindrical target 1, 2 or 4 inches in
diameter or e flat rectangular target. The target was electrically heated
and was suspended in an 8 inch diameter tube which could be ventilated
with air from a compressor. The ventilation air could be heated to 350OF
at 200 scfm.

It was found that the MHSIT's of the test fluids in,.;reased with increasing
ventilation velocity. Increasing the ventilation aii temperature,
however, produced lower MHSIT's. The lowest IHSIT's in 80°F air were
920°F for JP-4, 9001F for JP-8, 930°F for 5606, 880°F for 83282 and 1010°F
for 7808 for ventilation airtlow up tG 2 ft/secand.

Application of these results to engine compartment deslan is complicated

by the use of an unobstructed 8 inch diameter tube for ventilation, the
airflow within an aircraft engine compartment normally being complicated
by the presence of a jumble of engine components and aircraft structure.
in, ana ii oun, the Use oLf±Cectricl resistanceL2&I Ilsaetin and~ stanll

targets provides further complication.

Foose

In 1982, J. G. Foose of General Dynamics performed hot surface ignition

temperature tests in the same test facility employed for the current
program. This work is documented in Appendix A of Reference 5. The
objective of this test program -as to examine the feasibility of removing
insulation from FIOO-PW-200 eng,.ne bleed-air ducts in the F-16 engine
compartment.
7
An electricaliy heated test article composed of two six ivich segments of
the 13th stage bleed.-air duct from an F-16 engine was used to determine
the MHSIT's for sprays and streams of 5606, 83282, JP-4 and 7808, directed
onto the hot bleed-air duct. To enhance ventilation airflow simulation, a
cushion loop clamp for the bleed-air duct and engine compartment
obsttu.-tions, including the augmentor fuel pump controller and a portion
of the oil-air heat exchanger tank, were added to the test article for
some test conditions. This test article was mounted in the AENFTS
alloving simulation of ventilation airflow velocities from 0 to 10 ft/sec.

The minimum HHSITs measured during this program et a ventilation velocity

of 6 ft/sec in air, at ambient pressure and temperature, were:
o 1100°F for JP-4 (streamed onto duct with cushion clamp and

onto duct with flow obstructions)

o 1025°F for 5606 (streamed onto duct with clamp)

o 950°F for 83282 (streavied onto duct with obstructions)

o 1325°F tor 7808 (spray onto bare duct).

After analysis of the data acquired in these tests, it was covi±luded that
th* risk associated with the omission of the bleed duct insulation was
ar:ceptable because of the infrequent occurrence of the maximunm bleed duct
temperatures during flight, even though estimated bleed duct operating
temperatures could occasionally exceed the measured ignition temperatures
of the test f].ttds.

The General Dynamics program addressed a number of important aircraft

engine compartment hot surface ignition variables. These includcd fluid
In'sction method, the effect of engine compartment clutter (such as the
cushion Oamp and heat exchanger oil tank) and the effect of ventilation
velocity

8
While the engine compartment simulation employed in these tests was much
mnore realistic than in earlier pregrams, there were some compromises:

o Tvo short lengths of electrically heated inconel tubing were

used to simulate the F-16's bleed duct. Fluids impinging on
the surfsce may have behaved differently if the duct had been
air heated

o Only two airflow obstructions were used while the F-16 engine
compartment is crowded with obstructions

o Variation in ventilation air pressure and temperature

representing complete F-16 flight conditions was not included

TM. current program was planned to enhance the data base developed in

these tests.

1.2 Objective

The objective of the first part of the program, the simple duct tests, was
to investigate the phenomenon of hot surface ignition of flammable fluids
within an aircraft engine compartment with a test article thr.t was simple
enough to allow control of most of the test variables. This part of the

program was planned to allow:

o comparison to past data, especially the General Dynamics data

(Appendix A of Reference 5)

o determining the differences between an electrically heated

duct and an air heated duct

o determining the differences between selected aircraft fluids

o investigating the effect of a cushion clamp

o investigating the effect of duct orientation (horizontal or

vertical)

The objective of the second part of the program, the high realism tests,
was to determine the minimum hot surface ignition temperatures for each

aircraft fluid of interest over a range of severe but realistic aircraft

operating conditions. These tests were intended to provide design
information that had previously been unavailable concerning safe surface
temperature limits within aircraft engine compartments based on the actual
9
flammable fluids present and the temperature, pressure and velocity of the
compartment ventilation airflow,

1.3 Approach

For the first part, a seven inch long section of 1.5 inch OD inconel
tubing, 0.036 inch thick, was employed as a hot target. It could be
heated, either electrically or, with hot air (Fig. 2). This duct and a
spray nozzle were placed in the otherwise empty test section of the
AENFTS. HHSIT tests were run with JP-4 and 5606. The same cushion clamp
as used in the earlier General Dynamics tests was added to the duct for
some of these tests. These tests were run to:

o checkout the facility and test article and provide a data

baseline

0 compare HHSIT data acquired with this simple test article with
the data acquired in the Reference 5 (General Dynamics) tests.

o compare data acquired with electrical and hot-air heating

For the second part of the program, a more complex test article was
installed in the AENFTS test section. A simulation of the right side of
the F-16 engine compartment, which had been built for earlier AENFTS
testing (Refs. 5, 6 and 7),
was reworked so that hot-air from the AENFTS
bleed-air heating system was routed through an actual F-16 engine bleed-

adiL ductIri taIlld.JU Jina rcrccn, .,.-- ju.........umb....*c nt--imnztnk,

and clamps (Fig. 3). A spray nozzle and drip lines were added so that
the flammable fluids of interest could be sprayed or introduced as a
stream to various locations on the heated bleed duct.

Experiments were conducted with 5606, 83282, 7808, JP-4 and JP-8 entrained
onto the hot bleed duct to define realistic MHSIT values for the fluids.
Initially tests were conducted to establish the method of flammable fluid
delivery (spray or drip, fluid flowrate and duration) and delivery
location which produced the lowest MHSIT's for each fluid. The
ventilation airflow velocity, pressure and temperature were varied to
define HHSIT sensitivity to these parameters.
10
 Pd.

i0R

_ _ _ .3 4.o'

sg~ ,
--
ill IV

- LiL

it t C

3 .o
12,
 2.0 TEST FACILITIES

2.1 AENFTS Facility

The AENFTS is a ground test facility designed to simulate the fire hazards
which exist in the annular compartment around an aircraft engine. The
AENFTS is insialled in 31-Bay of Building 7i-B, Area B, Wright-Patterson
Air Force Base, Ohio. This facility (Fig. 4) includes air delivery and
conditioning equipment design.2d to simulate engine compartment ventilation
airflow, a test section within which fire testing can safely be conducted,
and an exhaust system which can cool the combustion products and scrub
them sufficiently to allow their release into the atmosphere. In
addition, it includes a gas fired heating system to provide simulated
engine bleed-air to the test section.

The test section of the AENFTS (Fig. 5) is a two radian (114 degree)

engine case, and a 24 inch radius duct, which simulates the engine
compartment outer wall. The test section is approximately 14 feet long
and is equipped with access ports and viewing windows that are provided
for access to test equipment and instrumentation and for observation of
the test activities taking place vithin.

As shown in Figure 4, the AENFTS ventilation airflow conditioning systems

Include a blower that provides air at atmospheric pressure (to simulate
low speed sea level flight conditions), a high pressure compressor and air
storage bottle farm to provide ventilation airflow simulating ram pressure
in low altitude supersonic flight conditions and an air driven ejector (to
evacuate the test section to simulate high altitude flight conditions).
The airflow from the atmospheric blower or from the air storage bottle
farm can be heated (without vitiation of the ventilation airflow) using 5
100 KW immersion heaters. The shorter curved test section wall, which
simulates the case of a turbojet or turbofan engine, can be heated with
radiant heaters. These heaters were not used in the current pzogram,
however.
13
 'AIR

ISCRUBBING A

HIGHPR~qo 4-DRYER

ATMOSPHERIC
c-BLOWER
GLYCOL

RADIANTi EATERS 00UC0HATR

(ARROWS INDICATE AIRFLOW DIRECTION)

Figure 4. AEN.TTS Facility

14
 20
z ý<- a OR
0 g 7-
LU -4-

UJ LL
ý- -1-j
LU WU0. cc

0 Z; Lou w

.:I -j j A
LL fg 0 D
z WNWru w IN z
i-n ct > ?o

-j uj
U)

LE <
U) z
U) <
LL
LU 0
cr. uj
LL LL.
CC
Z 00-
(C) LLi
ILU
Lr
<LL, W
W
in LUL,
Z m ý-
ZZ
I 6-4ul uj a a0
4
0 m z cc

LU
cc
La US CL

15
Simulation of the hazards associatcd with hot engine bleed ducts and the
leakage that might result from damage to bleed ducts or the engine case is
provided by the AENFTS bleed-air heating system. (See Reference 8 for
experimental determination of damage results from bleed air leakage.)

This system consists of a natural gas fired heater with automated

flowrate, pressure and temperature control systems which heated incoming
high-pressure air from the air-storage bottle farm. Up to 1500OF and 220
psia could be provided at flowrates up to 1 pound per second. An
insulated flex duct delivers this heated simulated engine bleed to the
AENFTS test section

2.1.1 F-16 Nacelle Simulator

In an actual aircraft engine compartment, the ventilation airflow does not

flow uniformly as in the clean AENFTS test section. Regions of reverse
LJ.LW c1lu LJL.W Oi atl ul
on hlall•e ! t•ll,
r, t -Ill
F-a11 UtIg Ltv L L ilt
Ued
Federal Aviation Administration's Technical Center (Ref. 6) and the F-ill
engine compartment is cleaner and designed for higher ventilation airflow
rates than the F-16 engine compartment. To simulate a more realistic
environment, having the complex of tubes, ribs, clamps, wires, and other
flow disturbances of a real aircraft engine compartment, a portion of the
F-16 nacelle was simulated for earlier testing in the AENFTS (Refs. 5, 6
and 7). This test article was found to be suitable for the present
program.

The forward right side of the F-100 engine, as it exists in the portion of
the F-16 engine compartment selected for simulation, is shown in Figure
6. A scrap early prototype F-100 engine was obtained and the components
in this region were removed and installed on a 5 foot-long simulated
engine side stainless steel base plate constructed to fit the engine side
of the AENFTS test section (Fig. 7). Intrusion into this region of the F-
16's glove tank and structural ribs was simulated in sheet metal (Figs. 8,
9 and 10) and fitted into the AENFTS test section over the engine side
base plate (Fig. 11). The final assembly represented one-third of the
engine compartment annulus. The remaining AE fTS test section length,
16
 bo

17
6118
 v.. Yiw
maear f FP-19 .5i{mtut~ar Prior to Installation

F-'.ure 9 Front View of F-i 6 Simulcxtor Prior to InstMIGU~On

19
200
approxcimately 60 inches, simulated the less cluttered annulus around the
afterburner.

Fused quartz viewing windows were provided in the 15 inch square access
ports on the nacelle side of the AENFTS. One of these opened onto the
forward "arch" of the F-16 bleed duct.

The distribution of the ventilation airflow through the F-16 nacelle

simulator was modified, for one series of tests, by the installation of a
baffle plate at its upstream end. As shown In Figure 12, the baf fle
redirected the airflow upward so that local velocities around the hot
bleed duct would be lower.

2.1.2 Fluid Delivery System.

Two fluid injection methods were used in the hot surface ignition test
PLUgtaiii1. Popzay vas uscd in both the simple duct and high realism tests
and additionally stream injection was used in the high realism tests.
Both spray and stream were part of a fluid delivery system shown in Figure
13.

Nitrogen gas was used to pressurize the stainless steel 2 and 3 liter
Hoke cylinders that were used as fluid reservoirs. The pressure of the
fluid reservoirs was monitored with a pressure transducer and the pressure
was displayed on the AENFTS console in the control room. The pressure was
controlled by operating a three position switch on the AENFTS console.
The fluid reservoirs were pressurized to 100 psig for stream injection,
and from 105 to 135 psig for 8 mi/sec spray, depending on the fluid being
sprayed. Lower reservoir pressures were required for fluids such as JP-4
and JP-8 while higher pressures were needed to achieve the proper spray
flowrate for 5606 and 83282 and 7808, fluids with higher specific
gravities and viscosities.

From the fluid reservoirs the fluid moved through 0.25 inch diameter
stainless tubing to a fiilter and then to an air actuated ball valve that
was connected to a switch and a timer located on the AENFTS console in the
21
 ,aa~tRIDIRECTED
-SPACE BETWEEN
*NOZR -SIPE BASE
4 fLATE AND SIMdULATED
CLOVE .TANK

Figur'e 12. Installation of Raffle to Redirect

Airflow in F-16 Simulator
22
 So"T - _

a WNU A1UVAWE

ana

FLO

Figutre 13. Schematic Diagram of Ftuel InJection System

23
control room. Spray was normally injected at 8 ml/sec for 5 seconds and
stream was normally injected at 2 ml/sec for 10 seconds. These values
were chosen because they were in the middle of the range used for the
General Dynamics tests (Appendix A of Ref. 5) but variation in their
magnitude was subsequently found to have little effect on the MHSIT's.

The two fluid introduction methods utilized slightly different hardware.

During the simple duct testing, the spray nozzle was always located 12
inches upstream of the leading edge of the simple duct. For the high
realism tests, two nozzle positions were used, one upstream, about 12
inches ahead of the F-16 bleed duct (the spray being directed in the same
direction as the ventilation airflow), and the other downstream, about 12
inches aft of the aft end of the F--16 bleed duct (the spray being directed
against the ventilation airflow). In both cases the fluid line passed
directly from the ball valve to the spray nozzle, a Wagner 621 nozzle with
a 0.021 inch equivalent diameter orifice and a 6 inch flat spray halfwidth
at 12 inches from the nozzle.

When a stream of fluid was employed, the fluid passed through a

micrometering valve after leaving the ball valve. This valve controlled
fluid flowrate from I to 3 ml/sec and the normal setting was 2 ml/sec.
One of six fluid stream injection lines (0.070 inch I.D.) was connected to
the output of the micrometering valve to select the injection location
desired. The fluid stream tube exit was normal to the bleed duct surface
and about 1/2 inch from the duct surface. A "T" was installed in each
stream injection line outside the AENFTS so that compressed air could be
used to clear the line of fluid after the completion of a test. Hence the
next injection used room temperature fluid from outside the AENFTS, rather
than preheated fluid that had been sitting in the stream injection line in
the hot nacelle. The ball valve timer then was adjusted to allow for the
filling of the empty stream injection line so that the net stream
Injection time for each stream location was 10 seconds.

24
2.2 Test Article

2.2.1 Simple Duct Test Apparatus

The simple duct was a seven inch long straight piece of uninsulated
Inconel tubing (Fig. 14, 15 and 16). It had an outside diameter of 1.5
inches and a wall thickness of 0.036 inch. On either side of the seven
inch bare surface of Inconel, the duct was insulated with Fiberfax ceramic
putty and ceramic cloth and covered with fiberglass tape in order to
restrict ignitions to occurring on the portion of the Inconel duct that
was instrumented with thermocouples. During testing it was observed that
fires were ignited only on the bare metal portion of the duct.

Figure 17 shows the placement and numbering of the thermocouples.

Thermocouples 1 through 4 were placed radially around the duct 3.5 inches
from the insulated edge, starting with thermocouple 4 on the leading edge
of the duct. Thermocouple i was tack-welded on the top, 2 o, the rear
edge of the duct and thermocouple 3 on the bottom of the duct. The
remainder of the thermocouples, number 5 through 8, were placed on the
leading edge of the duct spaced one inch apart. All thermocouples were
Type K Chromel-Alumel.

Two methods of heating the duct were available (Fig. 14). The electrical
resistance heating method used three 1 KY Watlow Firerod resistance
heaters mounted every 1200 in a 6.5 inch steel cylinder that slid snugly
into the Inconel duct. These were the same heaters and steel cylinder as
had been used in the General Dynamics (Ref. 5) tests. Voltage to the
resistance heaters was manually controlled from the control room using a
simple 3-position switch and the duct temperature was observed on console
monitor. The duct could also be heated by air from the AENFTS bleed-air
heating system. The hot air was then mixed with cold high-pressure air to
control its temperature and then piped to the test article. The mixing
valve was also controlled manually while observing the duct temperature
indicated for location 3 (Fig. 17) on the the console monitor. Duct
temperature for both the simple duct and high realism tests was controlled
prior to injection but not during injection. After flowing through the
25
 IV

cn2.3 .

0ii
.S2

C
4!

0 U
.- 3: -%
<
26
 -7 INCH LONG, 1.5 INCH OD, 0.36

STEEL CORE DRILLED TO

ACCEPT "FIREROD" UNITS
THREE "FIREROD" 1-KW -

RESISTANCE HEATERS

VEVfILATION AIR1,1W

t417,11

-7.5

"vp0

ELEC'HICALLY, HEATED DUCT ASSEMBLY AS INSTALLED IN THlE AIFNFTS

Figure 15. Elertticaly Hea~ed Simple Duc-t Tesi Art 1ce

 -a

0--

u cc

w 0W

"A 0 A7 LJ -
lo IT-c
UU 6

ZZN

7--

z 4 L

L-J

(4a

(zi - -2- D

LU~
I--
CK-

F-L

D &iD W

F-I- QI
0.7 _jI~L

U2
z U.

0 0
3: Ll R

0 c 5: L

CI) N~I ;i:

> 4 4 c
5C
Lfl
C\J It
CV, I t2
(0-K-
IU
HU>-

CV, 4

92
simulated bleed air duct, the heated air was piped outside the AENFTS test
section and exhausted near the ceiling of the room.

The thermocouples read uniformly (within + 20 0 F) when the test section was
air heated (Fig. 18) but there were larger temperature va rations (+..
0 were used (Fig.
150 F) on the duct surface when the resistance heaters
19). Similar variation was seen during the earlier General Dynamics
AENFTS testing. As a result, the ignition temperature for the resistance
heated simple duct tests depended on which thermocouple was chosen to
report the duct temperature. During this program it was felt that the
hottest point on the test section was most likely responsible for ignition
(when it occurred) and therefore the thermocouple at position 3, which
read the highest, was chosen as the reference thermocouple. This position
(Fig. 17) was at the mid-point of the duct and at the side with respect to
the approach of the ventilation airflow. For convenience, the
thermocouple at position 3 was also used in the air heated simple luct,
though ditue was little .ariation. between it and the other. 7 ,afl--1---

The criteria for selecting the reference thermocouple had been different
during the General Dynamics testing. During that program, to employ the
most conservative interpretation of the data, the thermocouple chosen was
the one which consistently read the lowest. The General Dynamics test
article consisted of two parallel heated sections of duct. The
thermocouple employed was at the mid-point of the upstream duct facing the
oncoming ventilation airflow. Other thermocouples were placed backside of
the upstream and front and backside of the downstream duct. The
variations between this thermocouple and the hottest points observed
0 0
(normally the backsides of boh ducts) ranged from 25 F to 75 F.

While there was no temperature data reported from the General Dynamics
test for a thermocouple located at the side of the duct, comparison can be
made between the data acquired for this position (position 3) and for the
position facing the oncoming flow (position 4) from the present program.
It can be concluded (Figure 19) that the General Dynamics HHSIT's would be
lower than those obtained in the current program by about 50°F at 1
ft/second and by almost 2001F at 8 ft/second.
30
 UL I-

u I

El :Dw

(A a.-

0 u
F-J
< 5-I

LIf) C* Er)
\."\\

(~pU- FO41
(A~~L L2.I UJG
L.I"fld'1

31
u LU
4

z -j
Q O~LU

U 0~

F- LU
0 00
z
'z

F- w . 44

LU

U Ld L

Cr)z
LU2

>--

Z
uLu
LU
w w
V)J .
UQ-

00iQ

(A 0301) 3uflIvu~fldYg lorla

32
An uncertainty analysis of the duct temperature measurement data was made,
based on analysis of the hardware installation and test procedures
(Appendix B). It was concluded that the duct temperature data would have
an uncertainty of + 250 F.

The simple duct was mounted for most of its testing in a horizontal
position with AENFIS test section also in a horizontal position (Fig. 16).
In addition, some of the simple duct testing was performed with the duct
and AENFTS test section in a vertical posidion. The high realism testing
was performed with the AENPTS test section in the vertical position. For
the simple duct testing the test section was completely free of
obstructions such as tanks, piping and ribs. The effect of the presence
of a cushion loop clamp on H11SIT was examined during the simple duct tests
for boith rasistance heated and air heated ducts. This was the same clamp
that had been used during the General Dynamics testing and was also
installed at stream location 5 on the high realism test article. During
the simple duct testing it was mounted at the center of the seven inch
incoriel duct.

Prior to beginning hot surface ignition testing, with the AENFTS test
section in the horizontal position, the ventilation air velocity was
surveyed using a pitot probe. The pitot probe was installed temporarily,
about 0.75 inch upstream and 0.5 inch above the mid point of the simple
duct. The ventilation airflow was calibrated to give airflow velocity as
a function of measured mass flow for the range of 1 to 10 ft/sec. For the
simple duct ttst, the atmospheric blower was used to supply the
ventilation air. The pitot tube was not Installed during hot surface
ignition testing. T'iis calibration was not repeated with the test section
in the vertical position because no significant difference was
anticipated.

Spray was the only meEhod of fluid injection used in the simple duct
tests. 5606 and JP-4 were sprayed at 8 ml/sec for 5 seconds from a Wagner
621 flat spray nozzle (0 inch spray fan half-width at 12 inches from the
spray nozzle exit and a nozzle equivalent diameter of 0.021 inch). The
flat spwey was parallel to the long axis of the test section and the

33
nozzle was placed 12 inch upstream of the lead edge of the duct. Viewing
the simple duct through the rig viewing window, it was clear that the
spray completely covered the uninsulated area of the duct. Spray flowrate
was controlled by pressurization of the fluid reservoirs and spray
flowrate vs. pressure curves for the fluids were generated prior to
testing.

The simple duct test section was visible through a quartz window in the
AENFTS wall. A video camera was aimed at the test section and made it
possible to remotely view injections and ignitions during testing.

2.2.2 High Realism Test Article

Actual 11-16 engine nacelle parts were used in the construction of the high
realism hot surface ignition test article. Among its principal features
were the 13th stage bleed-air duct, air-oil heat exchanger tank, augmentor
fuel pump and various tubes and clamps (Fig. 20). The AENFTS was oriented
vertically throughout the high realism test so that the position of the F-
16 components was the same as if they were installed in the aircraft. The
bleed-air duct was made of the same material as the simple duct, 1.5 inch
O.D. and 0.036 inch wall thickness inconel tubing. Hot high pressure air
was supplied to the duct through the outer wall of the AENFTS test
section. Bleed-air traveled from left to right into the augmentor fuel
pump and thence into insulated piping to exit the test section. On the
actual engine, bleed-air flows through the duct in the same direction
though it enters from the engine and it reenters the engine through a
perforation at the augmentor fuel pump. Ventilation velocities were
measured by a pitot probe installed at a point 1 inch in front of and I
inch above the most upstream edge of the bleed duct along the center line
of the AENFTS (Fig. 20). This position was not directly behind any major
obstructions and hence did not represent an abnormally low local
ventilation velocity.

Ventilation air was supplied from the facility air storage bottle farm to
the high realism test article using the high-pressure, low-flow airflow
system except for those test conditions at 11/ft/sec velocity where the

34
IIp

iIi

35
high-pressure high-flow system was employed because the low-flow system
would not provide enough airflow to reach this velocity.

The ventilation air temperature was measured by a thermocouple (TAIR-1)

suspended in the ventilation air stream about halfway between the nacelle
side wall and the engine side wall. The thermocouple was located about 4
inches upstream of the most upstream edge of the bleed duct on the
centerline of the test section (Figures 20 and 21). A standard commercial
1/8 inch stainless steel sheathed thermocouple assembly was employed. A
shield, constructed from 1/2 inch diameter 0.008 wall stainless steel
tubing, was placed between the thermocouple and the hot bleed duct to
shield the thermocouple from radiation from the bleed duct. The tubing
was cut away around the thermocouple bead to allow it to be fully immersed
in the airflow and only a flap extended above the bead. The outside of
the shield tubing was covered with fiberfax insulation to minimize its
heating from the bleed duct radiation.

A pressure transducer (PNACIN) was used to measure the ventilation air

pressure. This transducer was installed in the AENFTS about 6 feet
upstream ot the test article (Fig. 20).

Due to the effect of tVe various clamps and obstructions in the high
realism test article it was anticipated that hot surface ignition
conditions would be different at different locations on the bleed--air
duct. Fcr this reazon a total of six stream and two spray fluid injection
locations were used in the high realism test phase (Fig. 21.). The spray
injection method used the same nozzle used in the simple duct tests, a
Wagner 621 flat spray nozzle, placed either upstream or dounstream of the
bleed duct. %tren the spray came from downstream, the nozzle was placed
six inches downustream of the aft end of the bleed duct. When the spray
came from upsturiam, the nozzle was placed six inches upstream of the its
upstream end. For both positions, the nozzle was located on the test
section centerline and the spray fan was aligned along the duct.
Preliminary tests made with the viewing window removed and the duct
unheated indicated thzt spray from either direction thoroughly wetted the
entire bleed duct. No attempt was made to determine to what degree the

36
 * U

37 N

_--

II A
ventilation airflow reduced this though the occurrence of ignitions at all
velocities demonstrated that fluid continued to strike some part of the
duct at all velocities. The fluids were generally sprayed at 8 ml/sec for
5 seconds.

There were six fluid stream locations. Fluid traveled from the fluid
reservoir where it was maintained at 100 psig through a micrometering
valve and then through one of six stainless steel tubes (0.070 inch I.D.)
to the stream location. The fluids streams were generally supplied at 2
ml/sec for 10 seconds through the stream tubes. This flowrate produced a
solid stream of fluid out of Zhe injection tube. The effect of
ventilation airflow velocity on the streams of fluid's ability to reach
the duct was not investigated but the momentum of the streams and the
narrow gaps (0.5 inch) between the stream tubes and the bleed duct made
any change caused by velocity unlikely. All tubes ended normal to the
bleed duct surface except for location 4 which was angled forward and

the engine side o' the test article, as far as possible from the hot bleed
duct. After each fluid stream injection, compressed air was blown through
the stainless steel tube so that the next stream injection would use fluid
at nacelle room temperature, rather than uncontrollably heated fluid in
the stainless steel tube.

Eight type K thermocouples were also tack-welded along the bleed-air duct
(Fig. 21) of the high realism test article. A plot of duct temperature at
the various thermocouple locations is shown in Figure 22. The bleed duct
the.mocouples were generally associated with fluid injection locations and
the MHSIT is reported by using the associated thermocouple temperature for
the stream injection location:

Stream location 1: The fluid was introduced onto a Harmon clamp that
joined the bleed duct to the bleed-air penetration fitting on the nacelle
side of the AENFTS. The associated thermocouple, THSI-l, was tack-welded
to the thick walled fitting. As the test progressed, it was found that
due to the thickness of the fitting, the surface temperature of the
fitting was relatively low and the temperature response was extremely
38
 -JJ

0 0I

< vi

C LO

I-

w
U
-LL.
co - - -
K. 'i.

(A 9310) 12ll~~jNfl.
XG
39
slow. Since the injection location was the least realistic and did not
exist in an actual F-16 nacelle, testing at this location was terminated.

Stream location 2: The fluid was introduced onto the bell shaped inlet to
the Inconel bleed duct. THSI-2 was attached to tlhe bell shape-d inlet as
well. THSI-3, tack-welded to the bleed duct several inches above bell-
shaped inlet, was not associated with any stre.m lccation. It was
employed reference duct temperature thermocouple throughout the test since
it was the hottest location. It was also used to report the MISIT for
spray since a fluid spray was location non-specific and it vas felt that
the hottest location on the duct most accurately represented the ignition
temperature for spray injection.

Stream location 3: The fluid was introduced onto the top of a horizontal
section of bare duct. THSI-4, tack-welded to the underside of the duct at
the same location was the associated thermocouple.

Stream location 4: The fluid was introduced onto the aft (downstream)
face of a vertical section of bare duct. THSI-4 was the thermocouple
closest to this location also.

Stream location 5: The fluid was introduced onto a horizontal duct

portion of the duct at the point where the the cushion clamp was
installed. THSI-5 was tack-welded just upstream of the clamp. THSI-6 was
tack-welded to the bottom of the following bend, a location similar to
THSI-5 but lacking the clamp and was not associated with any stream
injection location.

Stream location 6: The fluid was introduced onto a Harmon clamnp between
the bleed duct and the augmentor fuel pump controller butterfly valve.
THSI-7 was located on the duct upstream of the change in duct diameters
for this clamp. THSI-8 was tack-welded to the inlet of the augmentor fuel
pump just downstream of the stream tube.

Three different video cameras were used to view the high realism test
article during testing (Fig. 23). It was possible to detect that ignition
40
p I

- U
-1
I 1)
had occurred but not all parts of the the bleed duct were visible and it
was not possible to determine the exact ignition location. Also asaisting
in the detection of a fire was an ultraviolet fire detector made by
Graviner, Ltd. (Ref. 9). This detector was mounted next to the video
camera placed in the test section side window downstream of the bleed duct
where it consistently detected the presence of fires in the test secion.
Another camera viewed the bleed duct through a test section side window on
the upstream side of the high realism test article. The third camera
viewed the inside of the AENFTS through an edge window on the top of the
test section. These three cameras and the UV detector made it possible to
determine whether or not there was ignition from the control room.

Extensive leak checks were performed on the high realism test article in
order to ensur': that no bleed-air would leak from the bleed-air duct and
alter the normal ventilation airflow patterns and change local ventilation
velocities. To ensure that there were no leaks in the bleed duct
installation; the duct was nre.qssrizPed with cold and hot air. smoke
generators were employed and visual checks were made with yarn tufts.

2.3 Data Collection and Reduction

Critical variables in this test were the duct temperature (THSI-1 through
THSI-8), ventilation air velocity, ventilation air temperature (TAIR-1),
ventilation air pressure (PNACIN) and test fluid type, injection method,
flowrate and duration of injection. Duct temperatures and ventilation air
conditions vere logged to disk on the facility computer while test fluid
variables were set prior to the test and were logged on a hand log data
sheet. The computer acquired data was logged just before the injection of
the test fluid, and the MHSIT measured in this test program was an initial
duct temperature and not the temperature of the duct at the time of
ignition. This does not compromise the value of the test data, however,
as this initial temperature is comparable to the engine compartment
temperatures which can bf- specified by an aircraft designer.

Duct temperature vs. time data for selezted simple duct tests were
acquired using a Honeywell Visicorder strip chart recorder. Figure 24
42
 o ~C-
:E' L )~

oI z

LL) o c,

(IA (Iý )a,mr(

I',,,0
u
V)

/ Z

to( - OIl to ) C

6 d 6d

(spuDsnotLj±)

43
shows the temperature data recorded for each of the 8 duct thermocouples
during the 5 seconds when 5606 was being sprayed onto the duct, during the
next 1.5 second when the hot air was bringing the duct back up to
temperature and briefly following ignition (the point of Ignition having
been determined from visual observation of the test along with review of
the video tape record following the test). The th rmocouples most
directly in the path of the spray, those at positions 4, 5 and 7, show the
greatest temperature reduction during the 5 seconds of 5606 spray and
those on the sides and back face of the duct are relatively unaffected by
the spray.

Also logged by hand was whether or not there was ignition, the ignition
time delay (time elapsed between beginning of injection and the ignition
of the fluid) and the video tape time, a record of the location of the
test on the tape. A closed circuit TV camera with a zoom lens was mounted
on a tilt and pan mechanism on the top of the fuel cart. During fire
tests, the camera was focused on the viewing window in the test section
adjacent to the test fire zone. Its output signal was observed ou a TV
monitor on the AENFTS control panel to allow the test operator to observe
ignitions and assure safe conduct of the test. A video cassette recorder
(VCR) received and recorded the signal from the TV camera. The video of
the simple duct tests would show the spray injection pattern, smoke
furmation arid the ignition of the fluid. The video tape of the high
realism tests only revealed the reflection of the occurrence of a fire,
the view beinig ibIII Ot vI'IIJp tfL yL o -U utLLMsLt by th tanks and1 rib, nV th...

F-16 simulator.

The video system was fairly low speed (30 frames/second, 2 fields/frame)
and ignitions occurred too fast to make precise visual observati.,ns of the
beginning of ignition. The video tape did provide a backup of the
ignition delay timer.

Ignition delay time was measured by the fluid injection system timer that
started when an electrical signal was sent to the injection system ball
valve and ended when the UV fire detectoL unit detected a fire or the test
operator observed a fire on the video displays and pressed a switch.
44
These were subsequently manually corrected for the delay between the
signal being sent to the ball valve and the fluid's first contact with the
duct based on earlier manual calibration for each drip location. These
ignition delay data are tabulated in Appendix A.

AENFTS test data consisted of temperatures and pressures which were

measured by thermocouples and pressure transducers in the test cell and
sampled, digitized, averaged and calibrated by the facility computer

system (Fig. 25). Lists of the temperature variable names and ModComp
data channels (Table I) and the pressure variable names and data channels
(Table 2) are included in this section. These millivolt values were
converted to engineering unit data for the temperature and pressure at the
AENFTS flowmeters and used to calculate flowrates and ventilation
velocities in the test section. The airflow equations for the venturis
are based on the Compressed Gas Handbook (Ref. 10) and those for the sonic
nozzles are based on data from their manufacturer. This information was
immediately used to update the AENFTS display terminals (approximately
once every 10 seconds) and by activating a data log switch the operator
could send this data to a line printer and also log it on disk. The
AENFTS facility computer is a 16 bit, general purpose digital computer for
real time multi--programming applications with 64K RAM memory manufactured
by Modular Computer Systems Inc. (ModComp) of Ft. Lauderdale, Fl.

45
 CAL

4:f

M IL

-J7h

62
................
 T9rable 1. AENFT*S Temperature Instrumenta lin
NERMOCOUI.. *GOWCIP SOFTWNARE
MUNUERi CHNA5I.E SYMBOL OF11CRIPTIO6N TYPE ACCURACY

TC-28 1 TINGIh Engine side skin tamp zone K 1 4 degrees F.

T-92 TENG19 Engine aide skin tamp zone 1
TC-30 3 TENGZA foglooe side skin temp moaae2
TC-31 4 TENG20 Engine, aide skin tomp zone 2
TC-32 S TENG3A Engine side skin tomp soneR 3
TC-33 6 TER936 ::ngine aide skin temp zoes 3
TC-34 I TEN04A Engine, side skin temp zone 4
YC-36 TfNG4S Engine side skin tamp sone, 4
TC-36 9 TINGSA Engine, side skin temp some
TC-37 10 SENGS seeging side skin tamp vones,
TC-39 it TEIIOGA Engine side skin tamp Zone 6
TC-39 12 TENG60 Engine side skin temp Rone 6
TC-40 13 lAIR-1 Nacelle air temp 20on 1
TC-41 14 TAIR-2 Recall* air temp zone 2
YC-42 i5 TAIR-3 Nacelle air temp zRep 3
TC-43 16 TAJR-4 Neaells, air toemp Rone 4
IC-44 17 lAIR-S Nacelle sir temp zone, 5
TC-45 is TAIR-6 Nacelle sir temp zone 6
TC-46 it TNACIA Nacelle side skiR tomp zone 1
TC-47 20 INACID Necelle side akin temp zone 1
IC-de 21 THAC2A Nacelle side skin temp zone k
TC-49 22 T"AC28 Nacelle side akin temp zone 2
IC-50 23 TNAC3A NeCalle *i140 Ski" temfp RON* 3
TC-S1 24 THAC3S Nacelle side skin temp 8*ne 3
TC-I 105 TH3I-1 Test article temp 01
IC-2 106 11431-2 Test article temp 02
TC-3 107 liiS1-3 Test article temp 03
TC-4 108 THSI-4 Test article temp 04
IC-S 109 TIISI-S, Test article toev 05
TC-6 110 T11SX-6 Test article temp 06
TC-7 III T14SI-7 Teit article temp 07
IC-8 112 TN4SX-9 Test article temp 08
TC-SD 31 TOUTLS Na4celle outlet air temp (long)
IC-59 32 TOUlTS" Nacelle outlet air temp (short)
TC-60 33 TWACIN Reaclle Inlet air temp
7C-61 34 TOL-08 Low flow venturi temp T
C-235 YO1-21 31@wtr outlet tamp K
TC-t63 3S T-NIFL Ni fia/Ni press temp
7C-64 37 TSTKLO Lower exhaust stack tamp
Iub UiiiKur hast stack tamp
upper ox I
TC-70 39
0
40
A1
ATIPAO
OAT-AV
TNACRN
Pad outside air temp
Roof outside
Nacelle room air tomp
air temp I
43 T-NPAO North pod teajp K
44 RTONEF Reference roomo temp T
TC-72 45 7GLYCOD Cold glycol temp .1
IC-74 47 T--FIYD Hyd. reservoir temp J
IC-7S 46 I-FUFL Fuel Injection reservol' temp J
YC-91 94 110110 Lo flo/NI press temp T
IC-201 97 TSHiOUT Bleed 1-tr. Outlet tom K
IC-202 99 TSHI4AI bleed air temp at nacelle inlet K
TC-20S 9, T~r1NOI Bleed sir "Ozzle Inlet temp K t. 4 degrees F

47
 o snu
on minnmnin o inuo m mn muou kn tn
"an C~N
" m C4C4
v C1.4fr 4 C-4N C4 C'4
10 a CD Q 0 C) CD CD a C. 2 0 0 ý (D CO

+1T,+1++1 +1 +1+1+1H+1+1+1+1~ +1+1+1 +1 +1+1 +1+1

U

00 C 0
C4IN C14 fat
'Ut tn Ut 0 N e
-4 -A at w - m * ~
4 '-to trU t .4
t* -4 W4 W Os 4 Ck-v W. W..
* *0.~04 fi m wg 4 I 4
C34 W *04 Wa
".4W404".4 0 MO 02
Q 1.~4: D
2 C0~ 04a -4 C
cDDoD 0m ~0 Q in0 0
"
0
~0 a 0
0a,An C0DC
in ~~ '1n'Nan
~ ~ ~ ~Sl
Op4. U
I In 0ejl
400 C m
C 00 o 0 0 C C4
0 0 0+400 00 0

ItC4
U~ ~ C9 1p I'r%
4
C40 %0C40 U
IM4 C45Q G4o~
t-- t-
4 t'
M qw o CO 4 - 64 %

3: C4
0211

ti% $4 040

I.-o

0w14 o J4 t4 4):"C4
4)~ id
U) zr- r-4r-I 9) 1 C
4001-4 WO1wM
qK H 4110 HW
14 m
4 -v
%W
" 0
-4 4 CI
2, H0 0
. j w 1 .
1-4 00 J4 m"Ck Cpo w 4)to , 4

V)r.-0 V- H000
WOrr.-.4-4 C: = A 0 v C 41
" .0t4 4w 0t45, a a 4E')0)4E-'.4) E-.E-4 aE('
X =~ 0444
all Q, - N>>>4 40 Vq0Ar
0 z1ý0 0 = w' w41 0 ) 4 0448r
 3.0 TEST PROCEDURE

A standard test procedure was followed throughout the hot surface ignition
test program.

1. The ModComp computer data acquisition system was started enabling

real time monitoring and logging to disk of hot surface ignition variables
such as duct temperatures, ventilation velocity, ventilation air
temperature and ventilation air pressure.

2. The video system was started allowing visual observation, from

the control room, of fluid injections and ignitions (including the fire
warning light triggered by the UV fire detector) in the AENFTS as well as
video taping of the hot surface ignition test.

3. The test fluid was prepared. Injection method, flowrate,

injection duration and fluid type are selected at this point.

4. Ventilation air conditions including velocity, temperature and

pressure were established.

5. The test article was heated to the planned test temperature.

Either electrical resistance heaters or 1 lb/sec air from the bleed-air
heating system was used to heat the simple duct and only hot air was used
to heat the high realism bleed-air duct.

At the beginning of the day, approximately I hour was required to heat the
duct to test temperatures. Between injections, however, reasonable steady
temperature conditions were reached in 2 to 5 minutes.

6. At this time, the ModComp data displays were checked to ensure

test conditions were met. Next, data was logged to the ModComp disk, the
video tape machine was started and the fluid was injectcd. The video
displays were observed for injection of the fluid and ignition.

49
In this test program, the duct temperature was started from a high
temperature and was lowered as ignitions were obtained. Once the duct
temperature was high enough to ignite the test fluid, the duct temperature
was reduced 50°F and the fluid was reinjected with all other variables
such as ventilation velocity, fluid flowrate, etc. held constant. This
process was repeated until ignition did not occur. Two additional
injections were then performed at that duct temperature. When a bleed
duct temperature was reached at which three tests could be conducted
without ignition occurring, testing at those ventilation air conditions
and test fluid conditions considered to be completed.

THE MINIMUM HOT SURFACE IGNITION TEMPERATURE (MHSIT) WAS THEN DEFINED TO
BE THE LOWEST BLEED DUCT TEMPERATURE THAT HAD PRODUCED IGNITION. BECAUSE
THIS TEMPERATURE WAS GENERALLY 50°F ABOVE THE TEMPERATURE WHERE 3 TESTS
WITHOUT IGNITION HAD OCCURRED AND THE DUCT THERMOCOUPLE TEMPERATURE
MEASUREMENT ERROR WAS ESTIMATED TO BE _25°F (SECTION 2.2.1), THE
UNCERTAINTY OF THIS MEASUREMENT (N!HTT) VAS CrON!Tf!RF' TOA A-0 AJn
hoP
0
-75 F. Ventilation air variables were normally held constant for 30
seconds at the end of each injection to ensure that there would he no
ignition in that particular test. This allowed for any normal ignition
delays. Data from a typical series of these tests (83282 stream onto high
realism test article at location 3, duct temperature measured at location
5, variation is velocity) is shown in Figure 26.

At the completion of a day of testing, test data on the hand log data
sheets and ModComp disk were combined. It was then possible to plot
fire/no fire data using duct temperature on the Y-axis and ventilation air
velocity (or temperature or pressure) on the X-axis as shown in Figure 26.
The symbols used in this plot are plus sign for ignition, and an open
square for no ignition.

Data of this type are presented in Appendix A for all the tests conducted
in this program. Ignition delay data are also tabulated in Appendix A.

50
 + 0
++

0 o9

< V)

2 oA

/00

LL>-I

u.j

LO_ __ _ _ C) r ___ 00

030)
(A PuBflOL!CRI I)fc
51
For this test program, the minimum hot surface ignition temperature
(MHSIT) is defined as the lowest temperature to produce igrkition that was
above the temperature where three tests occurred without ignition. For
example, at 2 ft/sec on Figure 26, the MHSIT was approximately 8500F.
MHSIT data for the five test fluids was obtained in this way throughout
the test program.

52
 4.0 TEST RESULTS

4.1 Simple Duct Tests

All simple duct tests were performed prior to installing the high realism
test article. The simple duct tests enabled comparison to past data and
the study of hot suLface ignition test data in a uncomplicated
environment. In the clean AEWFTS test section the ventilation velocity
was measured and assumed to be uniform at the duct compared to the
obstruction filled high realism test article which would have areas with
velocity both above and below the average. The simple duct was also
visible through an AENFTS window and spray patterns, nacelle ventilation
flow patterns and ignitions were directly observable.

A test matrix (Table 3), summarizes the three groups of simple duct tests
that were performed during this program. Two airplane fluids, JP-4 and
5606 were sprayed on the simple duct from upstream in the test section at
8 ml/sec for 5 seconds. Ventilation air pressure and temperature were
held as constant as possible at 14.4 psia and 1200F, respectively.
Ventilation air velocity was varied from 0 to 8 ft/sec.

The Simple Duct tests were organized into three groups in the test matrix.
The majority of the tests were performed in the first two groups, (1) the
ef.c of the cuho ±v. loo p cl.samp and' 1\1 thic efL tc. 't
of r
theA% duc hetn

method on the MHSIT. A cushion loop clamp was placed on the simple duct
where it was contacted by the fluid spray so that the effect of such a
device (acting as a flow obstruction and/or fuel vapor trap) on HHSIT
could be evaluated. The simple duct was built so that it could be heated
either with electrical resistance heaters or with hot-air at I lb/sec from
the AENFTS bleed-air heating system.

All four configurations for the simple duct (air or resistance heating,
bare duct or duct with clamp) were tested with both JP-4 and 5606. The
results of these tests are shown in Figures 27, 28, 29 and 30 and
summarized in Table 4. For these tests, the AENFTS test section was

53
 Table 3. Simple Duct Test Matrix

(ALL tests employed spray from upsteam at duct centerline)

II I I
TEST CONDITION I CONDITIONS OF VENTILATION AIR I FLUIDS
Ii I
-----------------------------------
IPressure ITemperature IVeLocity I JP-4 I 5606
I (psia) I (deg. F) I (ft/sec) I I
I I I I I I
IEFFECT OF THE PRESENCE OF CLAMP ON MHSII - GROUP 1

I(1) Bare Duct 14.4 120o1 1--8 X X

1(2) Duct with CLarrp 14.4 1201 1 - 8 X X
I 8ml/sez spray for 5 sec
I *ir hont•a and renistance heated

I horizontal naceLLe/duct
I r
IEFFECT OF THE DUCT HEATING METHOD ON MHSIT - GROUP 2

ISame tests as above:

I1) Resistance Heated 1 14.4 1 120 1 1 - 8 X X

1(2) Air Heated I 14.4 120 11 - 8 X X
I 8mt/sec spray for 5 sec I
I with and without clamp II
I horizontal naceLLe/duct II
I I
IEFFECT OF NACELLE/DUCT ORIENTATION ON MHSIT - GROUP 3 I
1(2) Horizontal Nacelke/Duct (same tests as ak~ve) 14.4 120 1 - 8 I X
1() Vertical Nacelle/Duct: Test Section Rotated 1 14.4 120 1 - 8 I X
Bmt/sec spray for5sec I I
bare duct I I
air heated I -

54
 BARE BLEED DUCT

D AIR HEATED BARE DUCT

1.5 0 RESISTANCE HEATED BARE DUCT / •

V GENERAL DYNAMICS BARE DUCT

(RESISTANCE HEATED WiTH AENFTS
TEST SECTION HORIZONTAL SEE
1.4 SECTION 2.2.1 FOR G.D.
w TEST CRITERIA)

-U
wof 1.3 - j /

90.
_ 1.2 5606 SPRAY FROM UPSTREAM
8 ML/SEC FOR 5 SECONDS
Pa - 14,4 PSIA; To - 120 DEG. F

1 / DUCT TEMPERATURES MEASURED AT LOCATION 3

MHSIT MEASUREMENT UNCERTAINTY: APPROXIMATELY +25 DEG. F. -75 DEG. F

0.9 - T- T I I I

0 2 4 6 IB

VENTILATION VELOCITY (F'T/SEC)

DUCT WITH C'JSHION CLAMP

1.6 ---

* AIR HEATED DUCT WITH CLAMP

1.5 - * RESISTANCE HEATED WITH CLAMP

* GENERAL DYNAMICS DUCT WITH CLAMP
I(QLG-IQTAKir 1-WATF-l-n wiT"- AFJFTS TF'T RFrTlnWil 2
' 1.4 HORIZONTAL - SEE SECTION 2.2.1 FOR
. C,.D. TEST CRITERIA) .

1.3-
,, /.

&I-0
(L 1.1

.1-/ 5606 SPRAY FROM UPSTREAM

"8ML/SEC FOR 5 SECONDS
Pa - 14.4 PS'A; Ta - 120 DEG. F
DUCT TEMPERATURES MEASURED AT LOCATION 3

MHSIT MEASUREMENT UNCERTAINTY: APPROXIMATELY +25 DEG. F, -75 DEG. F

0.9 1 1 1 --- - 1- T
0 2 4 6 8

VENTILArION VELOCITY (FT/SEC)

Figure27. Effect of V. on Simple Duct MHSIT for 5606

55
 BARE BLEED DUCT
1.6 ....... -- --- - ____....

1.5

1.4

1.3JP-4 SPRAY FROM UPSTREAM

,,., • 8 ML/SE'C FOR 5 SECONDS

Pa - 14.4 PSLA; Tq - 120 DEG F

.. 1.2 DUCT TEMPERATURES M.ASURED AT LOCATION 3

:1.1 NO FlRES EXPERWENCED WITH AIR HEATED SARE

BLEED DUCT AT DUCT TEMPERATURES UP 70
MAXIMUM AVAILABLE (1350 DEG F)

0 RESISTANCE HEATED BARE DUCT

V GENERAL DYNAMICS BARE DUCT (RESISTANCE HEATED WITH AENFTS
TEST SECTION HORIZONTAL - SEE SECTION 2.2.1 FOR G.D. TEST CRITERIA)
0.9 . IIT 1

VENTILATION VELOCIIY (FT/SEC)

DUCT WITH CUSHION CLAMP

MHSIT MEASUREMENT UNCER~TAINTY. APPROX:14ATELY +2-5 DEC. F. --76 DEG. F

1.5

If-EI
1.3-I
St.

F-
O~ .2JP--A, SPRAY FROM UpSrYREAM
I ~8 NM./SE.2 FOR 5 SECONDS
I- Pa - 14A4 PSIA; Ta - 120 DEG F
"-3
1.1I
DUCT TEMPERATURES MEASURED AT LOCATION 3

1 1[ AIR HEATED Ducr WITH CLAMP

* RESISTANCE HEATEra DUCT WITH CLMP
v GENERAL D.TIAMICS DUCT WITH CLAMP (RESISIANCE HEATED WITH AENFTS
TF'ST SrC't1tjt4 HOPIZONTAL. - ýLE SECTION 2.2.1 FOR G.{D. =EST CRITERIA)
0.9 .]T•tl"

0 2_ 4 6 8i

VE"NTILATION VELOCIYý (I-TI/9Eý.)

Figure 28. Effect of V. on Sim;e Duct MJ-HSIT (b Jr -

56
 JP-4 SPRAYED FRUM UPSTREAM
1.6 T ...
4NOJF-4 FIRES WITH BARE DUCT AT TEMPERATURES UP TO
THE MAXIMUM AVAILABLE (1350 DEG. F) WITH BARE DUCT

1350 DEG. F MAXIMUM _ _..___

fZ. 1.3 -. '

0. 13--b-•
1. AIR HEATED DUCT WITH CLAMP
1.2

1.1 (3Pq 14.4 PSIA; Tra 120 DEG F

8 ML/SEC FOR 5 SECONDS
DUCT TEMPERATUJRE MEASURED AT LOCATION 3

MHSIT MEASUREMENT UNCERTAINTY: APPROXIMATELY +25 DEG. F, -- 75 DEG. F

0.9 -4 ... .. . . . ...
-- .1. ... "- ........... ..1 T 1 T
0 2 4 6 8
VENTILATION VELOCITY (FT/SEC)

5606 SPRAYED FROM UPSTREAM

B ML/SEC FOR 5 SECONDS

1.5 Pa - 14.4 PSIA; Ta - 120 DEG F
DUCT TEMPERATURE MEASURED AT LOCATION 3

1.4A

t3

S 1,3 .,Zl AIR HEATED BARE DUCT

U)
r
0
5 I 1.2
.
- -I
t3AIR HEATED DUCT WITH CLAMP
I)I
1.

.. - -9MHSIT MEASUREMENT UNCERTAINTY:

.
APPROXIMATELY +25 DEG. F, -75
I
DEG. F
--
4
I
0 2 4 6 8
VENTILATION VELOCITY (Fr/SE;)

Figure 29. Effect of Va on Air HeatedSimple Duct Mi ISIT

57
 JP-4 SPRAY FROM UPSTREAM

1.-- RESISTANCE HEATED1

1.5 -' DUCT WITH CLAMP

1.4
0INRESISTANCE HEATED 13ARE DUCT

1.2-
-SIMPLE DUCT TEST ARTICLE
8 ML/SEC FOR 5 SECONDS
1 Pa - 14.4 PSIA; To - 120 DEG F
DUCT TEMPERATURE MEASURED AT LOCATION 3

MHSIT MEASUREMENT UNCERTAJNTY: APPROXIMATELY +25 DEG. F, -75 DEG. F

0.9 .. -. .....--- I ' ....
-.. . .........
- . . T-- - Tr-.. .
0 2 4 6 8

VENTILATION VELOCITY (Fr/SEC)

5606 SPRAYED FROM UPSTREAM

1.6/
1.5 -,

1.4
14 RESISTANCE HEATED)
BARE DUCT lI

M..1.3

o
ri• ."
RESISTANCE HEATED
DUCT WITH CLAMP
S 1.2

119

1.1 SIMPLE DUCT TEST ARTICLE

8 ML/SEC FOR 5 SECONDS
Pa - 14.4 PSIA; To - 120 DEG F
1 DUCT TEMPERATURE MEASURED AT LOCATION 3

MIISIT MEASUREMENT UNCERTAINTY: APPROXIMATELY +25 DEG. F, -75 DEG. F

0.9 ------ . . .. l . . .. l . ... . . 11
02 4 6 8

VEt titATION VELOCITY (FI/SEC)

Figure 30. Effect of Va on Resistance Heated Simple Duct MHSIT

58
 Table 4. Summary of iesults of Simple Duct Tests

ALL TEMPERATURES IN DEGREES F.

Pa z 14.4 PSIA (APPROXIMATE)

To = 120 DEG. F (APPROXIMATE)
ALL Ftuids Sprayed frbm Upstream
AENFTS Test Section Horizontal Except as Noted

I. III
TEST CONFIGURATION I T/C I APPROXIMATE VENTILATION AIRFLOW VELOCITY (rT/SEC)
POSITION
I 0I 1 2 4 6 8

5606 - SIMPLE DUCT

lAIR HEATED SIMPLE DUCT 3 NO TEST 1300 1200 1300 NO FIRE NO FIRE

JAIR HEATED SIMPLE DUCT, AENFTS TEST 3 NO TEST NO FIRE NO FIRE NO FIRE NO TEST NO TEST
SECTION IN VERTICAL POSITION

IRESISTANCE HEATED SIMPLE DUCT 3 NO TEST 1150 1170 1360 1510 1550

IGENERAL DYNAMICS (RES. HTD. BARE DUCT) 4 1110 1010 1210 1310 1420 1430

I 5606 - DUCT WITH CUSHION CLAMP I

I I I
!AIR HEATED DUCT WITH CLAMP I 3 I NO TEST 1100 1100 1150 1200 1300
II I
IRESISTANCE HEATED DUCT WITH CLAMP I 3 NO TEST 1150 1170 1190 1320 1380
I I I
IGENERAL DYNAMICS (RES. HTD. DUCT W/CLA.P)I 4 1 t140 1050 1020 1320 1420 NO TEST

I JP-4 - SIMPLE DUCT

I
lAIR HEATED SIMPLE DUCT

IRESISTANCE HEATED SIMPLE DUCT

3

3
NO TEST

NO TEST
I NO FIRE

1370
NO FIRE

137(F
I
NO FIRE

1500
NO FIRE

1520
NO FIRE

1540

IGENERAL DYNAMICS (RES. HTD. BARE DUCT) 4 1340 NO TEST 1320 1330 1330 1340

JP-4 - DUCT WITH CUSHION CLAMP I

pAIR HEATED DUCT WITH CLAMP 3 NO TEST 1250 1250 1350 Nu TEST 1340

IRESISTANCE HEATED DUCT WITH CLAMP 3 NO TEST 1360 1430 1510 1520 1540

IGENERAL DYNAMICS (RES. HID. DUCT W/CLAMP)I 4 1320 NO TEST 1310 1380 NO TEST NO TEST

T/C POSITION: LOCATION OF THERMOCOUPIE EMPLOYED TO MEASURE

MHSIT - SEE FIGURE 17 FCR DETAILS
NO TEST: NO HOT SURFACE IGNITION TEST PERFORMED AT THESE CONDITIONS
NO FIRE: IGNITION DID NOT OCUR AT 1350 DEG. F MAXIMUM AVAILABLE
DUCT TEMPERATURE; MHSIT ASSUMED TO BE GREATER THAN 1350 DEGo F

MEASUREMENT UNCERTAINTY: APPROXIMATELY +25 DET. F, -75 DEG. F

59
located in the horizontal position (Fig. 16) so that it represented the
lower third of an aircraft engine compartment.

The third group of tests addressed (3) the effect of duct orientation on
the MHSIT. Here the AENFTS test section was rotated to the vertical
position, representing one-third of the right side of an aircraft engine
compartment. Only 5606 was used in this group of tests. The data
acquired with this configuration was compared to the 5606 air-.heated bare-
duct data from previous tests.

The effect of ventilation air velocity on MHSIT for 5606, for the air--
heated bare duct, air-heated duct with clamp, resistance heated bare duct
and resistance heated duct with clamp is shown in Figures 27. Similar
data for JP-4 is shown in Figure 28. General Dynamics MHSIT data (Ref. 4)
for 5606 and JP-4 for their resistance heated bare duct and resistance
heated duct with clamp are also included on these figures.

With the resistance heated duct, thermocouple 3 was chosen to report duct
temperature in all cases because it was consistently the hottest location
monitored. There was little temperature variation on the air-heated
simple duct so thermocouple 3 was again selected. It was felt that the
highest temperature on the duct was closest to the MHSIT because the
sprayed fluid teached all parts of the hot duct completely and ignition
was likely to occur at the hottest point. As noted in Section 2.2.1, a
different criteria was employed to select the reference thermocouple for
the General Dynamics test data. The thermocouple with the consistently
lowest reading was employed. As also noted in Section 2.2.1, it could be
anticipated that this difference in reference thermocouple selection
criteria would result in the General Dynamics MHSIT's being lower than
those mecsured in the current program (from 50°F lower at 1 ft/sec to
nearly 200°F lower at 8 ft/sec).

A trend for MHSIT to increase with ventilation velocity is apparent with

5606 Spray From Upstream (Fig. 27) . The General Dynamics data exhibixs a
dip in MHSIT at around 1 to 2 ft/sec, The dip is less pronounced in the
current data though there is a similar dip in the air-heated bare-duct
60
data. The facility maximum for air-heated bleed-duct temperatures was
about 1350OF and there were no fires with the air-heated bare-duct at 6
and 8 ft/sec, probably because the duct was not hot enough. Hot surface
ignition tires did occur with the air-heated bare-duct at 1 , 2 and 4
ft/sec. This suggests that the MHSIT's might continue to increase with
ventilation velocity for the air heated bare duct as well, though they
would have occurred at temperatures too high to observe in this program.

The presence of the cushion clamp lovered the MHSIT for 5606 for the air-
0 0
heated and the resistance-heated cases, from 100 F to 200 F. This was
probably because the clamp held the fluid next to the duct longer,
allowing the fluid more time to heat than with the bare duct where it was
blown away before it could ignite. The General Dynamics data for 5606,
however, indicates very little effect of the clamp on MHSIT, except at 2
ft/sec where the MHSIT decreased by 200OF when the clamp was present.

Differences between resistance-heating and air-heating the simple duct caft

also be noted in Figurv. 27. The air-heated duct generally produced
ignition at lower temperatures than the resistance-hea~ted duct. Both the
air-heated bare-duct at 4 ft/sec and the air-heated duct with clamp at 1
to 8 ft/sec ignited thc fluids at temperatures 50OF to 100OF lower than
the resistance-heated counterparts.

Differences in the heating rates of the resistance-heated and air-heated

simple duct test articles is illustrated in the temperature history of the
duct thermocouples directly impacted by the spray (position 4), as shown
in Figure 31. These data were obtained during similar tests with 5606
sprayed onto the duct with cushion Jlamp at a ventilation air velocity of
6 ft/sec. The initial duct temperature was set at 1300OF and ignition was
observed about 6 seconds after the 5606 spray was initiated. Little
change in the heating rate at position 4 is seen following ignition,
probably indicating that the fluid is burning elsewhere on the duct.
Subsequent tests were run at 1250OF and 1200OF before the MHSIT for this
condition was established to be 1200 0F. As was shown in F~igure 24, the
thermocouple at location 3 (at the side of the duct), which was used as
61
 0~

C!;~~ w '0*'-
V) ýNho

0 bV1

P 0

I.-.-

el''J

(r~pU~rno~i-)
'00 / NvL~dGlin 1S.

62
the reference temperature to define the MHSIT's, did not experience the
ubstantial cooling and reheating noted at position 4.

Figure 31 shows that the air-heated duct had a higher heating rate than
the resistance-heated duct which led to a smaller temperature drop at
position 4 when the relatively cold fluid was sprayed onto the duct and
also to a faster temperature recovery. Because it was generally based on
a manual record of a visual observation, accuracy of the ignition delay
measurement, estimated at +0.25 seconds, is probably inadequate to
conclude anything from the ignitions times noted on the plot.

There was a recurring problem with leakage of bleed-air during the simple
duct testing. Leaks in the bleed duct assembly were discovered when the
test section viewing window was opened on three different occasions.
These leaks occurred only at the welds on the duct about 12 inches
downstream from the target area (the location is noted in Detail 2 of
F i amu rnro 16 U4 t1
U. th
t -. r
4
nA -. , rn ~and 5 1 lktrsnn of ambln te-nhnatre

bleed air flowing through the duct, the leakage could be manually felt no
further than 3 inches from the leak. On each occasion that leakage was
discovered, repairs were made before testing resumed. Because it is
impossible to identify when the leakage recurred, however, this must be
interpreted as a minor source of uncertainty in the quality of all of the
simple duct data. (This problem was permanently corrected prior to the
high realism phase of hot surface ignition testing.)

Some of the same trends seen with 5606 can be seen with JP-4 (Fig. 28).
No data for the air-heated bare-duct is shown because there were no
ignitions of JP-4 spray, up to the facility maximum duct temperature of
0
1350 F. Hence, JP-4 spray on an air-heated bare-duct has a MHSIT greater
than 1350°F at ventilation velocities less than 8 ft/sec.

The cushion clamp had less effect on JP-4's MHSIT than on 5606's MHSIT.
As with the General Dynamics tests, the resistance heated duct data shows
little or no effect of the clamp on MHSIT. Above 4 ft/sec the current
resistance-heated duct MHSIT continues to increase as the velocity
increases. The General Dynamics MHSIT remained constant. The lowest
63
MHSIT was again found vith the air-heated duct with clamp. No ignition
was observed at velocities greater than 4 ft/sec with the air heated duct,
again probably because the MHSIT was above the 1350°F maximum temperature
available.

The MHSIT's determined when JP-4 and 5606 were sprayed onto the air-heated
duct are compared an Figure 29. JP-4 ignited at higher temperatures than
5606, probably becauae of its higher volatility. This allowed it to
vaporize faster and hence required less time in contact with the hot duct
for ignition to occur. This general trend in MHSIT was demonstrated
throughout the test program for other fluids as well as with JP-4 and
5606; the more volatile the fluid, the higher its MHSIT.

Figure 30 shows a similar comparison for the resistance-heated simple-

duct. While the MHS1T's are generally higher for JP-4 than 5606,
ventilation velocity seemed to have a greater effect on the MHSIT of 5606
thaJ-/4. '"h. XHITP for IP-4~~ inr-e 00 0 F from I to A ft/qpr whilp

the MHSIT for 5606 increase about 250°F for the duct with clamp and 400°F
for the bare duct. This large rise in the 5606 bare duct MHSIT actually
causes the ignition temperatures for 5606 bare duct to be equal to the JP-
4 MHSIT at 6 and 8 ft/sec. The MHSIT for JP-4 appears to depend less on
the ventilation velocity than 5606.

The effect of test section and duct orientation on MHSIT was also examined
in the simple duct test. 5606 was sprayed on the vertical air-heated
bare-duct and the MHSIT's were to be compared to the horizontal duct data
that was taken earlier in the test program. While fires were ignited at
1, 2 and 4 ft/sec in the horizontal test, no ignitions at all were
recorded in the vertical test.

As will be discussed in detail in Section 4.2, it was found during the

high-realism tests that fluid stream ignition temperatures for a vertical
length of bare duct were much higher than the MHSIT for a fluid stream on
a horizontal length of bleed duct. Natural convection patterns were
observed around the simple duct caused by smoke rising and swirling from
the fluid being heated on the duct. These patterns were observed to be
64
much more active when the duct was in the vertical position than when it
was installed in the horizontal position suggesting that there were
greater variations in local velocity for the same ventilation airflow
rate. This higher effective ventilation velocity may have been
responsible for raising the vertical duct MHSIT for 5606 above the
available facility maximum temperature.

4.2 High Realism Tests

The High Realism Tests were intended to simulate hot surface ignition
scenarios in an F-16 engine compartment. To accomplish this the simple
duct test article was removed and the F-16 nacelle simulator, containing
an actual F-16 bleed air duct and other engine components was installed in
the AENFTS test section. As illustrated in Table 5, the effect of many
engine compartment variables on the MHSIT of the five test fluids (5606,
83282, 7808, JP-4 and JP-8) were studied in the high realism tests.

The high realism tests vere classified into two general groups: (1)
conditions of fluid injection and (2) the effects of ventilation air
conditions on MHSIT. The first dealt with the effect of injection
location, flowrate and duration on MHSIT. The second group of tests
concerned thj effect of ventilation air pressure, temperature and velocity
on the MHSIT of each fluid at stream and spray locations selected from
those tried during the first group of tests.

Because the objective of this test series was realistic stimulation of the
F-16 engine compartment, the AENFTS test section was installed in its
vertical position. The results obtained with the simple duct test
article, however, had indicated that horizontal placement of the AENFTS
test section resulted in the lowest MHSIT's than vertical placement.
Hence, although the bleed air duct of the realistic test article had both
vertical and horizontal sections, lower MHSIT values may have been
observed had the AENFTS test section been installed in the horizontal
position for these tests.

65
 TABLE 5. High Realism Test Matrix

lCONDITONS OF VENTILATION AIR I FLUIDS I

Pressure Teaperature Velocity j JP-4 IJP-8 I 5606 103282 1 7808

I (psia) ! Cdg. F) I(ft/, ec)I I I I I I

11.0 CONDITIONS OF FLUID INJECTION

1.1 EFFECT OF INJECTION LOCATION ON MISIT
Spray from Upstreva 1 14.41 1201 1 X X X X X
8mt/sec for 5 sec I I
Spray from Downstr. 14.4 120 I X X X X X
I lm/sec for 5 sec
Stream toc 1 - 6 14.4 120 1 x X X X x
Zmt/sec for 10 sec
1.2 EFFECT OF INJECTION FLOmJA7E/DURATION ON NHSIT
Spray from uipstreamI
ftowrate 14.4 120 0- I X
4,8,12 mt/s for 5sec
Stream t location 3
ftowrate 14.4 120 1 X I
1,2,3 ml/s for lOsec
duration 14.4 120 1 x
lOsec,4Osec a 2mt/s
12.0 EFFECTS OF VENTILATION AIR CONDITIONS ON MHSIT
2.1. SPRAT, (3-,,'/we, ',fr 5 b•)

FroM DownstreamI
Air Pressure
Rem simulation 114.4, 20 120 11 X X X X X
Attitude simulation i5,10,14.4 1 120 2I X X X X X
Air Te.vperature 14.4 1120,300,600 1 2 X IX X X X
Air Velocity 14.4 120I 0 - 8 I X
Baffle 14.4 120 11 1 X IX X X X
No bleed air flow 14.4 1480 - 600 1 2 1 I X
From Up•tream I I I I I
Air Velocity 1',.4 120 0 - 8 X
2.2 STREAM (2mL/sec for 10 sec)
Location 3 I I I I i
Air Pressure
Rom sin-ulation 114.4, 20 1 170 11 X X X
Attitude simulation 15,10,14.41 1201 21 X X X
Air Temperature 14.4 1120,300,600 2 I X X X
Air Velocity 14.4 120 0 -8 X X X
Baffle 14.4 120 11 X X X
Location 5
Air Pressure
Ram simulation 114.4, 20 I 120 j 11 1 X IX
Altitude simulation 15,10,14.41 120 2 X X
Air Temperature 14.4 1120,300•600 2 X X
Air Velocity 14.4 120 0 - 8 X X
Baffle 14.4 120 11 X X
----------------------------------------------------------------------------------------------------
NOTE: AlR tests with vertical nacelle and air heated duc-t

66
4.2.1 Conditions of Fluid Injection

There were two fluid spray locations (from upstream of the bleed duct and
from downstream of the bleed duct) and six fluid stream locations where
the five fluids were injected. The fluid injection locations are
identified in Figure 21. The initial tests were conducted at a
ventilation airflow velocity of 1 ft/sec, temperature of 120OF and
pressure of 14.4 psia. The purpose of these tests was to determine the
injection location at which the fluid ignited at the lowest temperature.
The fluids were then injected at these locations during subsequent hot
surface ignition testing.

The results of these tests are summarized in Figure 32. The MHSIT for
each fluid is plotted versus injection location. Usually, 5606 and 83282
had the lowest MHSIT's, especially at stream location 3 (DL-3). Stream
location 3 was at the top of a bend in the bleed-air duct, a horizontal
bare duct surface similar to that used in the bare-duct simple duct tests.
The thermocouple at position 3 (THSI-4) was the closest to stream location
3 and was used to report the duct temperature at this location. 7808 also
had a minimum MHSIT at stream location 3. HencE, later testing in this
program for 5606, 83282 and 7808 occurred at stream location 3.

83282 spray from downstream also ignited at a lower temperature than the
othpr fluids. llence. further testing was performed using 83282 spray from
downstream. This spray location was also used for the spray comparison
for all fluids in later tests. Thermocouple location 3 generally read the
highest temperature on the bleed duct and since the spray contacted many
locations on the duct, it was felt that the hottest temperature would most
closely represent the MHSIT. For this reason, thermocouple location 3 was
used to report duct temperature on the high realism bleed duct for all
spray tests.

Stream location 5 was used as the fluid stream testing location for JP-4
and JP-8. While it was not the minimum igniting location for these tv-c
fluids, location 1 having provided lower ignition temperatures for JP-4,
it was felt that the thick bleed duct fitting existing there (which was
67
 HIGH REALISM TEST ARTICLE
1 .5 . . . . . .. . .. . .... . . . . . . .

1 Fr/SEC Mrw ugV.imW uwr M. =

,Ma,"0aM Y +-2S. ,
1.4 -Va -
Ta - 120 DEG F
1.3 Pa - 14.4 PSIA

b. 1.2 - 4 A

21,-
(- 0.9
0560

•
8,3282322 JP-8
jp8SPRAY FROM DOWNSTREAMs
0.7 in: STREAM AT DRIP LOCATION "n"

SFU: SPRAY FROM UPSTREAM

0.6 (SEE FIGURE 19 FOR IDE'.NTIFCA11ON OF LOCATIONS)

0.5 - -------- r---r- ---

i Ii1
SF0D DL1 DL2 DI.3 DL4 DL5 DL6 SFU

S/LOCATION
/ ///7/7'
~//

/I/ //
/* //Ar* AIR77{\

/02
I-oval

LY:

Figure 32, Effect of Spray and Stream location on MHSIT with Hlgh-Raiaism Test Article
68 -
not a part of the F-16 nacelle) caused this to be an unrealistic location.
Stream location 5 was at the bottom of a duct bend where a cushion loop
clamp has been placed. This was similar to the horizontal duct with clamp
in the simple duct tests. Thermocouple 5 (THSI-5), the closest
thermocouple to stream location 5, was used to report the duct temperature
at this location.

Figure 32 indicates that 5606 had the lowest MHSIT of the five fluids at
most locations. In order of increasing MHSIT, 83282, JP-8, 7808 and JP-4
followed. In all these tests except for spray from downstream, 83282
ignited at higher MHSIT's than 5606. Generally the fluid stream ignited
at lower temperatures than the fluid spray, an exception being 83282 spray
from downstream which ignited at the same MHSIT as 83282 stream at
location 3.

The test the effect of injection flowrate and duLation vu

to determine
MHSIT was performed only with 5606. 5606 was sprayec from upstream at 4,
8, 12 ml/sec for 5 seconds at various ventilation velocities ranging from
0 to 8 ft/sec. Figure 33 shows 5606's MHSIT for the three spray flowrates
as the ventilation airflow rate changes and shows that the spray flowrate
has little effect on the MHSTT.

The effect of stream flowrate on MHSIT was also examined. 5606 was
injected on location 3 at 1, 2, 3 ml/sec for 10 seconds at a ventilation
velocity of 1 ft/sec. Table 6 shows the MHSIT for those stream flowrates.
Also included in Table 6 is MHSIT data for 2 ml/sec for 40 seconds.
Neither the stream flowrates evaluated nor the injection duration seemed
to effect the MUSIT significantly.

The flowrate and the injection duration seem only to effect air/fluid
mixing characteristics at these fairly low flowrates. They appear to have
had little effect on the MHSIT, probably because the duct temperature
response did not differ as the flowratk• were changed these small amounts.
If the flowrates were increased enocfh to quench the entire duct
significantly, the MHSIT probably would have been higher.
69
 UIU

wiw

LU U

LUL1h
1 i)

iiz

Lrj (0 r_

0< ~ C5

U 1()31JflLV Jdt l -z.l

70
 ~00_
C--

r-, C'D
Co L~i

-uCC

:~
Ctf

4::1La

71
4.2.2 Th- Effect of Ventilation Air Conditions on MHSIT

The first part of this group of tests dealt with the MHSIT's determined
for the 5 fluids with spray injection when the ventilation air pressure,
tempu-ature and velocity were varied. In addition, the effect of a
baffle, installed to change the airflow dynamics in the test article, was
investigated. A test where no heat was supplied to the bleed duct and the
effect of elevated ventilation airflow temperature on the MHSIT of 83282
was also conducted. A similar group of tests was run next using a fluid
stream at location 3 (5606, 83282, 7808) or location 5 (JP-4, JP-8)
instead of a spray (See Figure 21 for location identification). A summary
of the effects of ventilation air pressure, temperature and velocity on
MBSIT is provided in Tables 7, 8 and 9.

4.2.2.1 Spray

The first ventilation air variable tested during the spray test was
ventilation air pressure (Fig. 34). All fluids were sprayed at 8 ml/sec
for 5 seconds. Design limitation of the AENFTS prevented employment of
the same ventilation airflow velocities for the high pressure (ram
simulation) and low pressure (altitude simulation) tests, complicating
direct comparison between high and low pressure MUSIT data. High pressure
test conditions were obtained by closing the 24 inch valve downstream of
the diffuser section and throttling the airflow exiting the test section
with the 8 inch valve leading to the AENFTS ejector. There was sufficient
leakage around these valves that it required a minimum airflow velocity of
11 ft/sec to reach 20 psia in the test section. The low pressure test
conditions employed the AENFTS ejector system, again with the 24 inch
valve closed and the 8 inch valve employed for downstream throttling. To
reach 5 psia in the test section, a maximum airflow of 2 ft/sec could be
handled by the ejector.

As the ventilation air pressure decreased, the MHSIT was expected to

increase. Actually, none of the sprayed fluids were ignited up to the
0
facility maximum bleed duct temperature (1350 F) at 5 or 10 psia.
However, a trend in the W1SIT can be seen in the data taken at 14.4 and 20
72
 Table 7. Summary of the Effect of Air Pivaure on MHSIT
ALL TEhPERATURES IN DEGREES F.

iI III
FLUID T/C ALTITUDE SIMULATION RAM SIPRLATION
IINTRODUCTIONI POSITION FLUID
I I I - 2 FT/SEC (APPnOXIMATE)
aV Vo = 11 FT/SEC (APPROXIMATE)
I Ta a 120 DEG. F (APPROXIMATE) Ta = 120 DEG. F (APPPOXIMATE)I

I I j AAPPROX1MATE AIR PRESSURE (PSIA)

I I I I II
I I I t 10 14.4 14.4 201

STREAM I
INJECTION
LOCATION

3 4 5606 1320 1100 740 1140 1080

3 4 83282 1S5 1150 40 1180 840

3 4 rout wu FiRE 1340 .... I ? ... ,,

5 5 JP-4 NO mtý 140 1200 1320 12401

5 5 JP-8 H1UFIR2 NO FIPE 115; 1220 1240

SPRAY 3 5606 NO FIRE NO FIRE 750 1300 1200

FROM
DOWNSTREAM 3 83282 NO FIRE NO FIRE 800 1220 820
III I

3 7800 NO FIRE NO FIRE 1060 1270 1190

3 JP-4 NO FIRE NO FIRE 1160 1330 1240

3 JP-8 NO FIRE NO FIRE 1100 1290 1250

I I I I I

SPRAY 3 5606 NO TEST NO TEST 1210 f NO TEST NO TEST

I r. I I I II
UPSTREAM I I I
I I I I

TIC POS1lION: LOCATION OF THERMOCOUPLE EMPLOYED TO MEASURE

MHSIT - SEE F IGURE 21 FOR DETAILS
NO FIRE: IGNITION DID NOT OCCUR AT 1150 DEG. F KA;'IMU AVAILABLE
DUCT TEMPERATUkE; MSHIT ASSUMED TO BE GREATER THAN 1350 DEG. F
NO TEST: NO HOT SURFACE IGNITION TEST PERFORMED AT THESE CONDITIONS

NHSIT MEASUREMENI UNCERTAINTY: APPROXIMATElY +25 DEG. F, -75 LEG. F

3
Table 8. Summary of die Effect of Air Temperature on MHSIT

Pas 14.4 PSIA (APPROXIMATELY), Va = 2 FT/SEC (APPROXIMATELY)

ALL TEMPERATURES IN DEGREES F.

I I I i I
I STREAM T/C I FLUID JAPPROXIMATE VENTILATION AIRFLOW TEMPERATURE (DEG. F)I
INJECTION POSITION I I
LOCATION I ! 120 300 600
! I I II _

3 4 5606 740 640 600

3 4 83282 840 750 600

3 4 7808 990 1040 850

5 5 JP-4 1180 1210 1180

5 5 JP-8 1130 940 1040

ISPRAY 3 5606 750 700 600

IFROM I
IDOUNSTREAM 3 83282 800 650

1 3 7808 1060 1060 950

3 JP-4 1160 1050 750

3 JP-8 1100 950 600

* 83282 WOULD IGNITE WITH AIR TEMPERATURE

AT 600 DEG. F EVEN WITHOUT DUCT HEATING;
SEE TEXT FOR EXPLANATION

T/C FOSITION: LOCATION OF THERMOCOUPLE EMPLOYED TO MEASURE MHSIT - SEE FIG 21 FOR DETAILS

MHSIT MEASUREMENT UNCERTAINTY: APPROXIMATELY 425 DEG. F, -75 DEG. F

74
 Table 9. Summary of the Effect of Air Velocity on MH/SIT

I Pa x 14.4 PSIA (APPROXIMATE) I

I Ta x 120 DEG. F (APPROXIMATE)
I I
I ALL TEMPERATURES INDEGREES F. I
II I II
I I T/C FLUID J APPROXIMATE VENTILATION AIRFLOW VELOCITY (FT/SEC)
POSITION I
I I I I 1 2 4 6 8 11 11 I
I IWITHOUT I WITH
STREAM BAFFLE IBAFFLE
IINJECtIONI
ILOCATIO I

3 1 3 5606 NO TEST 700 740 840 990 1040 1140 700 i

3 3183282 790 790 840 800 840 650 1180 730 i

3 3 7805 NO TEST 990 m9 1090 1090 1130 1230 1120

I 5 5 JP-4 1250 1200 1200 1250 1260 1310 1320 1210

I ' 5 5 1 JP-8 1160 1150 1150 1200 1250 1260 1220 1160

3 1 5606 NO TEST 1100 750 NO TEST NO TEST NO TEST 1300 1250

SPRAY 3 83282 750 800 800 750 800 1010 1220 1170
FROM I I
IDOWNSTREAI 3 7808 NO TEST 1100 1060 NO TEST NO TEST NO TEST 1270 1210

3 JP-4 NO TEST 1150 1160 NO fEST NO TEST NO TEST 1330 1270

3 JP-8 j No W ST 150 1100iNV TENT

Itj itTEST TT 12"90 .Q, I
I I I I I I

I SPRAY I I I I I
FROm i 3 1 5606 1050 1000 1210 1180 1200 1250 NO TEST NO TEST I
IUPSTREAM I
I II II I I
I i I

NJ TEST: NG HOT SURFACE IGNITION TEST PERFORMED AT THESE CONDITIONS

T/C POSITIONt: LOCATION OF THERMOCOUPLE EMPLOYED TO MEASURE MHSIT - SEE FIGURE 21 FOR DETAILS

MHSIT MEASUREMENT UNCERTAINTY: APPROXIMtATELY +25 DEG. F, -75 DEG. F

75
75
 _ z

En 0

LL-J

sLL
d _ i~Z
GOa

~t

0~

W~)CJQn I
I ~ QI Ir~n- .,

00

V)I

76L
psia where the MHSIT's for all fluids decreased as the pressure increased.
The IHSIT for 83282 spray from downstream decreased nearly 400°F as the
ventilation air pressure was increased from 14.4 to 20 psia, in contrast
to the other fluid's behavior where the differences were closer to 100oF.

The next variable that was examined in these tests was the effect of
ventilation air temperature on MHSIT for all five fluids (Fig. 35). The
MHSIT was determined for each of the fluids at ventilation air
temperatures of 120 0 F, 300OF and 600 0 F, measured at location 3, just
upstream of the heated bleed duct (Fig. 21). In all cases, the MHSIT's
decreased as the ventilation air temperature was increased. This was
probably largely due to the ventilation air-heating the fluid prior to its
contacting thE duct. This pre-heating of the air/combustible-fluid
mixture reduced the heat transfer required from the duct to reach its
ignition temperature. Hence ignition occurred with lower duct
tpmnpratiurps. The five fluids seemed to be affected differently; 7808
showed relatively little change (about 1000 F) in MHSIT as ventilation air
temperature was increased, as did 5606 (about 1501F). Conversely, JP-4
and JP-8 spray displayed a large decrease (400°F to 500°F) in MHSIT as the
ventilation air temperature increased from 120OF to 600 0 F.

The HHSIT for 83282 at an air temperature of 300OF was 150OF lower than
its MHSIT at an air temperature of 120 0 F. When the air temperature was
increased to 6000F, however, the 83282 would consistently ignite even when
the bleed air duct was cooler than the ventilation airflow. Since the AIT
0
of 83282 is 670 F, it appeared that a situation had been encountered where
a hot surface ignition temperature was lower than the fluid's AIT. Hence
a second series of tests was performed that used hot ventilation air but
no bleed air duct heating.

These tests were begun with the indicated ventilation air temperature at
600t4'. Again, a 5 second spray of 83282 at 8 ml/sec ignited with no hot
air being supplied to the bleed air duct. Next the air temperature was
decreas2d incrementally until three fluid sprays tests had occurred
without ignition at a ventilation air temperature of 480 0 F.

77
 00
m (f~ Na
-0

0 U,

<U /
LLU

00 o 0
000

(1
9til 030 Vd 31J-~-~

/78
The duct temperature measured at location 3 provided erratic data during
this test. Substituting data from the thermocouple at location 6, which
normally provided similar duct temperature information to that at location
3, it appeared that the duct was normally only about 80F to 10OF cooler
than the airflow surrounding it (Fig. 36).

The duct temperature measured at the last test condition where an ignition
had occurred, 5101F, was concluded to be 83282's MHSIT under these
conditions. Before reaching specific conclusions about 83282's behavior
under these conditions, however, some additional uncertainties must be
considered:

1. Thermocouple error analysis, as documented in Appendix B,

suggests that the duct thermocouples would generally indicate a
temperature which was about 25 0 F lower than what actually existed on the
duct. The last ignition occurred at an indicated duct temperature Cf
0
501 F, hence the actual duct temperature was probably about 526 0 F.
Similar analysis of the air temperature thermocouples suggests that
radiation error would be much larger and that the indicated temperature
would be about 60°F lower than the actual air temperature with the
specific conditions which existed during this test. Hence the air
temperature had been about 570°F when the last ignition occurred.

2. Because it initially appeared that the 83282 did have a

HHSIT
b~lw At AILT, ,,uater-al propert-le tests. were -- dut-_U by.
• •
AFWAL/MLBT at the completion of the hot surface igpition testing on the
actual fluids which had been used in these tests. The results of these
tests, as shown in Table 10 indicate that (1) the fluids employed in these
tests still met appropriate specifications after the tests had been
completed and (2) the specific batch of 83282 used had an AIT of 700 0 F,
per ASTM D 2155 and of 690 0 F, per ASTM E 659.

Hetice The test results indicate that 83282 actually ignited in a situation
where the air temperature was about 570°F and the duct temperature was
about 530°F and that this specific batch of fluid had an AIT (per both
ASTH D 2155 and E 659) more that 100°F higher than existed, either in the
79
 0
-- o
to

+0
00
LOl

0
U')

+ L
+

0 Id

+ eq

U' LiJ
CL

Oww

0 0
0 t

CL LJ

ly-0 0 0 0 0

(j _n) K-IHJ a CAf

-f1~~V1

800
 *j 4 In a

*z
1 4 LU 0 0
U) 4.is I I- I- l- I.- I-

I 4In

P- -. %I. - -A a -

4 5 W. 4jLf
3t -C 4x C.-Uic

I-s 4 LU toL

U W U* L I

a:~~~rU * 0..
V)- 0.j 0.
E

uSr

,2 C9 L
0* -- -. --

-C *A Q
-U za, s -
I.- & u

*i LUJ
- - - - - -

- Z LU S &.

* o
it I- . a

L
4 L

40 S 4* 0

E s s U5 t I .

:3 50 r- Nl

8-f
*j :-
~ :8 * r-N: r- LA 2
L

-j a. * I.-

-K0
La s n Is r w nV. s

:---------W
- x - - - - - - - -
~~u
-. 3 5

-a
-- -- --
13
air or on the duct during this test. The Reference 3 tests also found
83282 to have a hot surface ignition temperature below its AIT, 6301F
(Figure 1) compared to an ASTM D 2155 AIT of 657 0 F.

Figure 37 shows the effect of ventilation velocity on the MHSIT for all
five fluids when they were sprayed from downstream. This test was
performed at a ventilation air pressure of 14.4 psia and a ventilation air
temperature of 120 0 F. MHSIT data was acquired with 83282 at velocities of
0, 1, 2, 4, 6, 8 and 11 ft/sec while the MHSIT data for the other fluids
were acquired only at 1, 2 and 11 ft/sec. 83282's MHSIT remained fairly
constant at about 750°F to 8000P for ventilation velocities of 0 to 6
ft/sec and increased to 1000 0 F at 8 ft/sec and 1220OF at 11 ft/sec. The
MHSIT's found for 5606 at 2 and 11 ft/sec are similar to those found for
83282. As was found during the simple duct testing (Fig 27), all five
fluids show the same general trend of having a "dip" in their MHSIT at
some intermediate velocity and having a substantially higher MHSIT at the
highest velocities tested. Insufficient data was acquired to establish
exactly where the "dip" occurs with 5606, 7808, JP-4 and JP-8. As with
the simple duct testing, however, the 83282 MHSIT's are fairly constant
from 0 to 4 ft/sec and begin to increase significantly beyond this point.

Caution is necessary if these test results are to be applied to an

aircraft design problem. While the effect of increasing the velocity was,
as anticipated, to generally increase the MHSIT's, these velocities were
measured in a single location within the test article. As in an aircraft
engine compartment, there were also regions of higher velocity and regions
of stagnation. Unless the designer is confident of uniform airflow, the
minimum MHSIT's found in these tests should be applied with the
understanding that low local velocities may well exist in the vicinity of
an aircraft bleed duct.

To examine the changes in the MHSIT's that could be caused by reduction in

local ventilation velocities such as might be caused by blocking the
ventilation airflow with an engine accessory, a baffle was installcd in
the test article (Fig. 12) to route the ventilation air away from the hot
bleed duct where the fluids were sprayed. For these tests, the indicated
82
 cV.

NN

ooo
r-o
-< L.

ED+

n-o

:D C)o

w: Co W-

00
0 z
w V)C)1 n
< L- z

V)jV
V) 0 C0

(a pu a ano~j
J~'03( '2--ISHI 3.-fJivzndJ3i LcriCG
83
ventilation air velocity was 11 ft/sec. The actual velocity at the hot
bleed duct was proLably significantly lower but a pitot probe measurement
was not made. The temperature was 120OF and the pressure was 14.4 psia.
The MHSIT for each fluid, sprayed from downstream, both vith and without
the baffle, is shown in Figure 38. The baffle reduced the local
ventilatiun velocity at the duct enough to lower the MHSIT's from 50°F to
0
200 F.

A test was also performed to observe the effect of ventilation velocity

variation on the MHSIT of 5606 when sprayed from upstream. The MHSIT's
that were measured are contrasted with the MHSIT's determined whet 5606
was sprayed from downstream in Figure 39. While the MHSIT meaoured during
the upstream spray test remained fairly constant from 2 to 8 ft/sec, the
MHSIT at 8 ft/sec is about 200OF higher than the MHSIT at 0 ft/sec. The
0
lowest MHSIT for 5606 (750 F) remained that measured at 2 ft/sec with
spray from downstream.

4.2.2.2 Stream

Two stream locations were used for testing in this phase of the program:
(1) Stream location 3, which resembled the horizontal bare duct in the
simple duct tests, the location that produced ignitions at the lowest
temperature for 5606. 83282 and 7808 in earlier testing. This was again
used with 5606, 83282 and 7808. (2) Stream location 5, which resembled
the horizontal duct with clamp in the simple duct tests, (along with
Stream location 1, which was not considered to be representative of the
duct installation in the aircraft) had produced ignition a' the lowest
temperature for JP-4 in earlier tests (Fig. 32). This location was used
for JP-4 and JP-8 testing. At both of these locations, fluid streams were
injezted at 2 ml/sec for 10 seconds.

The effect of ventilation air pressure on MHSIT for all the fluids at
these stream locations is shown in Figure 40. The hollow symbols are

MHSIT data for 14.4, 10 and 5 psia at a velocity of 2 ft/sec. The solid
symbols represent the MHSIT for the fluids at 14.4 and 20 )sia ventilation
84
L. ow
LL.J 0 0 I

z C)-

0 0

0 tO

V)~

0 C) 0 0 0 OD

(GpuD o2: ±
j.u-S 0~VEcdJ .100
C 85
 F7T
**~j 0

U))

uii

LU -'

LO- N L

(<L V 4:_L
(A '0 0 :IHL@Zn-" ,l~ -,)c
800
 5606. 63282 AND 7808
1. • r UWCUI"INV. WPROXM LLY +25
MIW MISURDMI5 MW.r. -75 DM. I

..... 1350 DEG. F FACIWY TLMPERATURE UMrrI

1.2 -

1.2 -',

"To - 120 DEG. F

0.9 R2
M./SEC STREAM AT

LOCATION 3 POR 10 SECONDS

S~ti
0.7 - E 560• Va - 2 FT/S U 5606. Va - 11 FT/S
A 63282. Va - 2 FT/S A 83282. Va - 11 FT/S
0.6- * 7808, Va - 2 FT/S + 7808. Va - 11 FT/S
tNO
FIRE WITHIN TEMPERATURE ULMIT OF FACi1'2TY
0.5 +---- r,-,r---'T-
0 4. a 1 16 20 24
VENTILATION AiP Pf MV,4RE (PSIA)

JP-4 AN/,j JP-11

1.5 9W MEASURVENT LWWMTAJrtA MMOAAWTYLY +25 KE& F. -75 MiES.F

1.4 -. 1~1350 DEG. F FACI~LIY TEMPERATURE LiMT __

C' .

1.2
1.1 HIGH REALISM TERT ARTICLE

2 U./SC
SREA ATLOCATION 5 FOR 10 3EC

0.' 2JP-8 F~IRE WrTHIN ~.NO

I/ECSTEA TEMPERATURE LIMIT OF FACIUlY
0. NO JP-4 FIRE WITH-IN TEMPERATURE LimrT OF FACIulY
S0.
S6NOJP8FIREWTHNTEPRTUR MT OF TU-

1*
06: A JP-4. Va -2 FT/5
JP-8. Va - 2 FT/S
A JP-4. Va - II fl/s
* JP-8. Va - I11Fr/S

0 4 U 12 16 20 24

W.~'flLATION AIR PRESSURE (PSIA)

Figure 40. Effect of Pa on MHSIT with Fluid Stream
57
air pressure with a ventilation velocity of 11 ft/sec. The ventilation
air temperature was held constant at 120OF for all of chese tests,

With 5606, 83282 and 7808, unlike the tests employing spray from
downstream, ignitions were obtained at air pressures below ambient as well
as at ambient pressure and 20 psia. The MHSIT's of 5606 and 7808 were
more sensitive to the pressure change below than above ambient pressure.
The decrease in MHSIT for 5606 as ventilation air pressure was increased
from 14.4 tO 20 psia was only about 50°F while decreasing the ventilation
air pressure below ambient affected the MHSIT much more strongly. Between
ambient and 10 psia the MHSIT was increased 350°F; between 10 psia and 5
psia it was increased an additional 20C°F. Similarly, for 7808 the change
in MHSIT when ventilation air pressure was raised from 14.4 to 20 psia was
not as largf (100 0 F) as the increase in MHSIT 'hen the ventilation air
pressure was lowered from 14.4 to 10 psia (250 F). 0

The MHSIT of 83282 was affected by changes in pressure both above and
below ambient pressure. Raising the pressure from 14.4 to 20 psia lowred
the MHSIT from 1200°F to 850°F while lowering the pressure from 14.4 to 10
psia increased the HHSIT 300OF and decreasing the pressurE further to 5
psia increased the MHSIr 0
another 200 F. Since sprays of these fluids
could not be ignited at pressures lower than ambient, these data indicate
that the chance of hot surface ignition of fluids is greatly reduced at
altitude when low pressures exist in a ventilated engine compartment.

4
The effect of increasing the ventilation air pressure on the MHSIT of J3k-
and JP-8 was not as pronounced as the effect seen on the hydraulic fluids
and the lubricating oil. The MHSIT for JP-3 actually increased slightly
when the ventilation air pressure was increased from 14.4 to 20 psia but
the increase was within the uncertainty of the temperature measurements
0
(425 F). When the ventilation air pressure was decreased to simulate
altitude, JP-4 fires occurred at 10 psia, at a slightly higher duct
temperature than at ambient, but did not occur up to the facility maximum
of 1350°F at 5 psia. With JP-8, no fires occurred at either 10 or 5 psia,
up to the 1350OF maximum available temperature.

88
Neyt, the effect of air temperature on the MHSIT of 5606, 83282 and 7808
stream at location 3 and JP-4 and JP-8 at location 5 was examined (Fig.
41). Ventilation velocity was set at 2 ft/sec and the ventilation air
pressure was ambient (14.4 psia). The MHSIT of both 5606 and 83282
decreased with increasing ventilation air temperature, as anticipated, but
the MHSIT for 7808 increased slightly as the air temperature was increased
from 130OF to 300OF before being reduced substantially as the air
temperature was increased to 600 0 F. This test at 300OF was repeated
because these results were unexpected. The same results were obtained on
the second try.

In this test 83282 was observed to ignite at a MHSIT of 600°F when the
ventilation air temperature was 600 0 F. As in the earlier spray tests,
this MHSIT was below the reported AIT for 83282. Again, this MHSIT was
the last target temperature where ignition did occur prior to three tests
where Ignition did not occur. As with the spray tests with 600°F air,
radiation losses led to thermocouple errors of about 60°F in these tests.
Hence the ait t!mperature waS ab.-o t .. th1 fl1i{d, ATT at the
time of this last ignition.

Variation of the ventilation air temperature did not greatly affect the
MHSIT of JP-4 and JP-8 streams. The JP-4 MHUIT remained almost constant
for all air temperatures tested. The JP-8 MHSIT decreased and then
increased as the air temperature was increased.

The effect of ventilation air velocity on the MHSIT's of the fluids

injected as streams is shown in Figure 42. These tests were conducted at
ambient pressure and at an air temperature of 120 0 F. Between 0 and 8
ft/sec, decreasing the ventilation air velocity seemed to have almost no
effit on the MHSIT of 83282 while it did decrease the MHSIT's of 5606 and
7808. At a ventilation velocity of 8 ft/sec, the 83282 stream ignited at
an MHSIT about 200OF lower than the 5606 stream. At 1 and 2 ft/sec the
5606 stream ignited at a MHSIT about 100OF lower than the 83282 stream.
While the MHSIT's of JP-4 and .JP-8 did increase as the airflow velocity
was increased the effect was minimal.
89
 5606, 83282 AND 7608
1.5 --. . . . . .

1.4 - HIGH REALISM TEST ARTICLE

Pa - 14.4 PSIA. Va - 2 FT/SEC
1.3 - 2 ML/SEC STREAM AT
O LOCATiON 3 FOR 10 SECONDS
1.2-

7508
71.1

0.9-

83282
0. -.

0.7 5606-"-- -

0.6 . -:t- f

MM45T MEASURAMWT UNCETAiN1M. APPROXMA1LY +25 DEG. F. -75 DEC. F

0 200 400 600

VENTILATION AIR TEMPERATURE (DEG. F)

:I:HW
Ij JP-4 AND JP-8
L4W APPROXMAVLY +23 OCO. F. -75 DEG. F

d .JP-4

,, 1.1

0.9,- sP-8

0.8 - HIGH REALISM TEST ARTICLE

.u Pa - 14,4 PSIA, Va - 2 FT/SEC
o 0.7 - 2 ML/SEC STREAM AT
LOCATION 5 FOR 10 SECONDS
0.6

0.5 I
0 200 400 600

VENTILATION AIR TEMPERATURE (DEC. F)

Figure 41. Effefct of T. on MH$IT with Fluid Str#4m
90
 5606, 83282 AND 7808
1.5

1.4 HIGH REAULSM TEST ARTICLE

Po - 14.4 PSIA. Ta - 120 DEG. F
1.3 2 ML/SEC STREAM AT
0bJ LOCATION 3 FOR 10 SECONDS
1.2

S 1.1 ---4,-

a 7808 ,-

Li

0..

0.8 83282

0.7
0 5606
0.6-

M"srr MEASURCMDIT UNCERTINNTY- APPROXIMA11LY +25 OM'. F*,-75 DEG. F

0.5 1 i I 1 i i1 1 1 1
0 2 4 6 8 10 12

VENTILATION VELOCITY (FT/SEC)

JP-4 AND JP-8

1.5

1.4

t • I.,;, -IJP-4.i -

i 2

0.9 HIGH REALISM TEST ARTICLE

I Pa - 14.4 PSIA. Ta - 120 DEG. F
0.6 2 ML/SEC STREAM AT

1- LOCATION 5 FOR 10 SECONDS

C.7

0.61

0 .5
S]
"
iMSIT M1EASURIBAENT UNCIEWTAWY':
1 1 '1i I
APPROXMAATELY +25 DCG. F. -75 DEG. F
I I I I I I

0 2 4 6 8 10 12
VENTILATION VELOCITY (FT/SEC)
Filure 42. Effect of Va on MHSIT with Flud Stream
91
The effect of installing the baffle in the test article to reduce local
airflow velocities around the bleed duct on the MHSIT's of all five fluids
is shown in Figure 43. The NHSIT's with the baffle installed are
significantly lower for 5606 and 83282 consistent with these fluid's
MHSIT's sensitivity to velocity (Fig. 42). A much smaller reduction in
MHSIT is seen with the baffle installed for 7808, JP-4 and JP-8, again
consistent with the insensitivity of these fluids's MHSIT's to change in
ventilation airflow velocity. Installation of the baffle reduced the
airflow around the bleed duct enough that the hot bleed duct often heated
the ventilation air (probably from 120OF to about 300 0 F) before injection
could take place. Hence part of the cause of the lower MHSIT seen with
the baffle In place was probably the elevation of the ventilation air
temperatures.

4.2.2.3 Comparison of High Realism Stream vs. Spray

Test data acquired with stream and spray fluid introduction to the high
realism test artic'le Wd5 COFpre andt:
CLIffrc in M.HSIT duc to the
method of fluid introduction were examined. Stream injection of the fluid
consisted of a solid stream from an 0.070 inch ID tube and traveled only
about 0.5 inch before the stream contacted the bleed duct. In contrast,
spray injection utilized a flat spray nozzle placed about 6 inches from
the bleed duct.

The effect of ventilation air nressure on the MHSIT of the test fluids is
shown in Figures 44 and 45. The MHSIT of 83282 for both spray and stream
appears to be strongly affected by pressure. MHSIT's of 5606 and 7808
seem to be affected less. For these fluids, however, the MHSIT for stream
injection at both 14.4 psia and 20 psia is lower than the MHSIT for spray
injection at those pressures. This was consistent with the fact that
sprays of neither fluid would ignite at 0
the maximum (1350 F) duct
temiperature available at pressures below ambient while streams of both
would. The MHSIT's of JP-4 and JP-8 were less affected by pressure.
Again the MHSIT found with stream injection were lower than those found
with spray.

92
on V

(nV

/4 U-J

IIIC,4z
""oJ
COO) _j
'-u

a. EL

I Lo
0- 6 < C , *

093c
 Z3~ .

00 V)
wa O w
CjI I

LOo

>ot to 00O

0 LLL C

LJI

Ld I-f, 17o

C1 00 0j n 0

?I,
 Z3 LU~

LL

LUL

~V) LL

a. I'

-, t
I- CL LO !LC

V) LO (n fLOV

~ I z
C4 Z(I)Z,

0 0
0> U02
0~ 0 0-HO'
LL

w 0 I- -a- CL

E ) nt)V

oi J06f0 'ia 47l~dd~

 Comparing the effects of air temperature varlition on the MHSIT's of 5606,
83282 and 7808 (Fig. 46), data for both spray and stream shor a tendency
for the IHSIT to decrease with ircreasing temperature (excepting the 7808
NHSIT's at 120OF and 3000 F) and no substantial differences are seen
between spray and stream injection. In contrast, with JP-4 and JP-8 (Fig.
47), a large decrease is seen in HHSIT's for sprays as ventilation air
temperature is increased and the MHSIT's obtained with streams are less
affected. The MHSIT for JP-4 streams does not change at all as air
temperature is increased, but the MHSIT for JP-4 spray decreases from
about 115 0 OF when the air temperature is 120OF to about 750 0 F the air
0
temperature is 600 F.

This difference between the two methads of flu4 tintroduction is probably

due to the small droplets of JP-4 spray being heated by the ventilation
air and requiring less heat transfer from the bleed duct to reach ignition
temperature. The fluid stream, however, is affected by the hot
vauiiaio ai -ybiel rfK oteigteUu A"A eble dc

provides most of the heat to raise the fluid streams's temperature to its
ignition temperature. The JP-8 data is very similar, showing a large
effect of air temperature on NHSIT for spray while a much smaller affect
is shown for a flixid stream.

The effects of ventilation airflov velocity on MHSIT's for 5606 and 83282
with the two methods of fluid delivery are shoin in Figur. 4q. HHSIT data
for 82282 show minor differences betwaen stream and spray. With 5606 the
differences between stream and spray from downstream are also relatively
minor, ti,,ugh the MHSIT's for spray from upstream are higher at all
veiocities tested. The effect of the velocity on the MHSIT for both
injection methods is similar. A similaL comparison of the two methods of
fluid delivery for JP-4 and JP-8 is shown in Figure 49.

4.2.2.4 Comparison of High Realism and Simple Duet Results

A compariscn of the MHSIT's measured for a spray delivery of 5606 and JP-4
acquired with the simple duct &nd the high realism test articles is made
in ?iure 50. The simple duct test article data includes KHSIT's for the

96

4
iCi

.~N 0

q)N Ono (D

a6 6 6
*~Q) Gpuonoa.
iLV~3~N~-L.L ~Ml
97z
 -- L

V) V

ot4 I
a_

0-in 0~/
~II /

a-L
Ln ir. V) rf)

n~ U)
-. O .-
I I
/
/Z /
/
+
tflIZAOl
P 1W :

Laa
0i!
b "Nb

01,/
42
0!7 //
re. l/
(If
LX.
I II

0 +

(opuosnoLt.L)
(-q oja) mji"3Vd~N31 -Lono
98
 u4l

00
AN~

~t 00
0r 11 Lf

V)
0k

I~ o_)ý-3

n
0
0
v 0

CL a
C4 C4

n o- O L 00
to N 0o

0 O
~0 a)l 0~1

(s pucrsno L41)
±000
(A4 03(3) 3ýfin-Vý;13ýN3-
 LL.

a-,

.4-4

IA.I

D p

-t4

00

a-
EnI
.4! 06 +

100.
 8606 SPRAY
MHSIT MEASUFVEENT UNCERTAINTY: +25 DEG. F. -75 D*. F

I
1.5 Po - 14.4 PS"A
To - 120 DEG. F

"1,2f J ..-- 1
,d," 1.3 - " -.". . ..- .- -, "

c 1.1 -

0.9
0.8I- +- 1

0.7 - AIR HEATED SIMPLE DUCT; SPRAY FROM UPSTREAM

+ AIR HEATED DUCT WITH CLAMP; SPRAY FROM UPSTREAM
* RESISTANCE HEATE SIMPLE DUCTT; SPRAY FROM UPSTREAM
0.6 - A RESISTANCE HEATED DUCT WITH CLAMP; SPRAY FROM UPSTREAM
X HIGH REAISM TEST ARTICLE; SPRAY FROM UPSTREAM
v HIGH REALSM TEST ARTICLE; SPRAY FROM DOWNSTREAM

0 2 4 6 a! 10 12
VENTILATION VELOCITY (FT/SEC)

.JP-4 SPRAY

1.5

1.4-

=Po 1.2

F-14.4,
a1 To-120 DEG. F

13 AIR HEATED SIMPLE DUCT. SPRAY FROM UPSTREAM - NO, FIRES

0.5 UP TO 1350 DEG. F MAXIMUM AVAILABLE DUCT TEMPERATURE
+ AIR HEATED DUCT WITH CLAMP; SPRAY FROM UPSTREAM
0.7 - * RESISTANCE HEATED SIMPLE DUCT: SPRAY FROM UPSTEAM
A HEATED DUCT W)TH CLAMP; SPRAY FROM UPSTREAM
,ESISTANCE
0.6 X HIGH REALISM 'lEST ARTICLE: SPRAY FROM DOWNSTREAM
0.5 _ S T.....
....
MHSIT,...... T
MEASUREMENT , UNCERTAINTY: +25
, -T_ ,-,F. -75 DEG. F
DEG. --
0 2 4 8
a 10 12

VENTILATION VELOCrrT (FT/SEC)

Flgut* 50. Comparson of Simple Duct and High Reawism MH$IT

101
bare duct and the duct with clamp with both air and resistance heating.
MHSIT data for a similar spray onto the high realism test article is
included for comparison. With the 5606 spray, the upstream spray high-
realism test article data agrees best with the air-heated simple duct with
the cushion clamp data. The downstream spray high-realism data acquired
at 2 tt/sec, however, indicates a MHSIT about 250°F lower than any of the
other tests.

With the JP-4 comparison shown on Figure 50, the downstream spray high-
realism test article data again indicates lower MHSIT's than any of the
simple duct tests. The difference is smaller than with 5606, however,
0
about 100 F.

4.3 Summary of Results

4.3.1 Simple Duct

1. The bare duct ignited 5606 at higher temperatures than when a cushion
clamp was added. This was probably due to the clamp holding the fluid
next to the hot duct for a longer time, allowing more heating of the
fluid. The addition of a the clamp did not affect the MHSIT for JP-4
spray onto the resistance heated duct, however. JP-4 being more volatile,
probably vaporized almost immediately with or without the clamp and was
not held against the duct longer by the clamp. Because no ignition
occurred when the JP-4 was sprayed onto the air-heated bare-duct at the
.......
m•~m, , terMper.ture .. •la but did wheii thv Clamp was added, it
appears that the addition of the clamp to the air-heated duct lowered the
MHSIT for JP-4.

2. Lower IHSIT's for both JP-4 and 5606 were generally measured with the
air-heated duct than with the resistance heated duct. This was probably
mostly due to the higher heating rate available with the air-heated duct.

3. The MHSIT of 5606 showed a strong dependence on the ventilation

velocity, increasing 200OF to 400°F when the ventilation velocity was

102
increased from 1 to 8 ft/sec. JP-4's NHSIT was not as affected by
velocity.

4. At low ventilation velocities, the MHSIT for 5606 was about 200°F
lover thai, the MHSIT for JP-4 whether measured with the resistance or air-
heated duct. At ventilation velocities of 6 to 8 ft/sec the MHSIT's of
the two fluids were about the same.

5. Whether the simple duct was oriented vertically or horizontaJly had a

signiticant effect on the MHSIT for 5606. When the duct was placed in a
vertical position no fires were ignited with the duct heated to the
maximum temperature available, at any ventilation airflow velocity.

4.3.2 High Realism

it The injection location was found to strongly affect the MHSIT's in the
high realism tests. A variety of factors, including the local ventilation
air velocity and temperature and the heat transfer coefficients of the
particular fluid contact site on the duct affected what MHSIT was
determined for the fluid. It was found that a stream onto a horizontal
bare section of the duct ignited 5606, 83282 and 7808 at the lowest
temper-tures. It was found that a stream onto a horizontal section of the
duct, where a clamp was located, ignited JP-4 and JP-8 at the lowest
temperatures. It was also found that spray from downstream also igna ed
83282 at a relatively low temperature. For the range of spray and stream
flovrates that were Investigated, little effect of injection flowrate or
duration was observed.

2. The MHSIT's of all five test fluids, both spray and stream, increased
dramatically as ventilation air pressure was levered. Hence, MHSIT'S are
significantly increased for aircraft at altitude,

3. The MHSIT's of all five fluids generally wa.- decreased as the

ventilation air temperature was increased. The MHSIT Ai 7808, however,
0
was affected only slightly. With an air temperature of OO°F, the MHSIT
of 33282 (spray and stream) was below the fluid's AIT.
103
4. For all fluids the MHSIT for both spray and stream was higher at a
velocity of 8 ft/sec than at a vclocity of 1 ft/sec. The MHSIT's of JP-4,
JP-8 (spray and .ream) were effected only slightly by velocity, however.

5. The effect of ventilation air temperature on the MHSIT of JP-4 and JP-
8 was different for spray and stream fluid introduction. High ventilation
air temperatures dramatically decreased the MHSIT of JP-4 and JP-8 spray
while effecting the MHSIT's for stream introduction only slightly. This
was probably because the spray droplets were preheated in heated air
before they made contact with the hot duct while the fluid stream had less
time for preheating before it struck the hot surface.

6. In general, the hydraulic fluids, 5606 and 83282, tended to ignite at

lower MHSIT's than the JP-4 and .JP-8 fuels. Lubricate 7808 was somewhere
in between for the majority of the test conditions.

7. Both fluid injection modes, spray and stream, are important in

determining the lowest MHSIT depending on test conditions and the type of
fluid.

8. The high realism test article with its associated clutter gave lower
values of MHSIT than the simple duct test article. This difference may
have been even greater if the AENFTS test section had been horizontal
during testing with the high realism test article.

9. The reported MHSIT's may be up to 75 0 F higher than actual values due

to test technique and measurement error and the MHSIT's may have been
lower if the AENFTS test section had been installed in its horizontal
position.

10. To determine the maximum safe design temperature, the highest

operational compartment temperature and pressure should first be
established. At these conditions, the lowest MHSIT, independent of
ventilation airflow, should be noted for each fluid -f interest after
examining experimental uncertainties and both injection modes, spray and
stream. The lowest value of MHSIT resulting from the above procedure

104
should then be reduced by at least 150OF to arrive at the maximum safe
design temperature. (Note: Elevated fluid temperatures and large hot
surfaces (engine case) were not considered in the above suggested
0
reduction of at least 150 F)

4.3.3 Test Article, Facility and Technique

1. The AENFTS provides a realistic simulation of the airflow conditions

affecting hot surface ignition in and aircraft engine compartment. The
facility's inability to simulate a full range of altitude conditions is of
relatively little importance because the altitude conditions were found to
reduce hot surface ignition risk.

2. The F-16 nacelle simulator, with its realistic clutter (providing

regions of stagnation as well as regions of high velocity) and its real
aircraft engine bleed duct, allowed tests to be conducted which closely
duplicated real aircraft engine compartment conditions without the cost
and safety problems which would be encountered trying to conduct fire
tests in a real aircraft.

3. The Graviner UV optical fire detector unit which was employed with
both the simple duct and the high realism test articles continued to
perform throughout the entire test program without problems. Because its
intended 1 second response time tended to be slower than the test
operator's visual response to the fire on the console video monitor it was
not normally the first indication that a fire had ignited. During some
situations, however, where the fire was small, or the viewing window was
obscured by soot, it was the first indication observed. No change in its
sensitivity was observed during this program--even when compared to its
performance when first installed in the AENFTS in 1984.

105
 5.0 ANALYSIS AND INTERPRETATION OF THE RESULTS

In this section the main findings presented in Section 4 are analyzed and
interpreted in light of the key processes involved in hot surface
ignition. To do this, reference is made to the preliminary and
approximate analyses of sprays and streams, presented in Appendices D and
E, respectively. These analyses were formulated specifically to help
identify the relative importance of various concurrent processes and their
contributions to the test results. Also, they form a basis on which
modeling effort can be carried out in the future.

5.1 Spray Analysis

Appendix D presents a simplified characteristic time analysis for the

major processes involved in the hot surface ignition of sprays. These
procecses include atomizationn droplet translation and deceleration, gas

phase heat transfer in the duct boundary layer, droplet heating and
vaporization and chemical kinetics of the ignition reactions.

For simplicity, the analysis focuses on the case of a single bare duct
instead of the complex geometry of the high realism tests. Also, no
attempt was made to calibrate the analysis using test data. Thus, the
analysis is very crude, but still, it allows approximate predictions of
minimum hot surface ignition te.2mperatures, MISIT's. As expected, the
numerical predictions are not accurate, but the predicted trends are in
reasonable agreement with test data in most cases. Furthermore, the
analysis identifies the relative importance of the various processes thus
enabling interpretation of the results.

106
5.1.1 Predicted Droplet Behaviors for the Five Fluids

Illustrative results comparing the droplet behaviors for sprays of the 5

fluids are showi in Table 11. Note that:

o for JP-4 and JP-8, the (volume to surface or Sauter) mean

diameters of the droplets produced in the spray are of the order
of 50 microns. These small droplets decelerate quickly before
reaching the duct and then are entrained by the ventilation air.

0 for 5606, 83282 and 7808, the droplets are much larger (in the
range of 180 to 250 microns) mainly because of the higher
viscosities of these fluids. These larger droplets decelerate
more slowly and impact the duct at a high velocity and high Weber
number (suggesting that these droplets may shatter upon impact.)

o concomitantly, the time required for droplet evaporation (due to

gas-phase heat transfer) is much shorter for the lighter fluids
(of the order of 10 ms versus 100 to 1000 ms for the heavier
fluids).

Thus, there is theoretical evidence that droplets from sprays of the

heavier fluids may accumulate on Olhe duct formng a thjlLqU,d film, that
boils on the surface, approaching the situation of a stream. Thus, in so
for as proccsses in the liquid phase, difference between the behavior of
spray vs. stream should be less accentuated for heavier fluids than for
lighter fluids (other things being equal).

5.1.2 Measured vs. Predicted Trends in MHSITs for the Jet Fuels

In Section 4, the results of a systematic study of the effects of

ventilation air parameters (pressure, Pa, velocity, Va and temperature,
Ta) on the MHSIT's of the five fluids under the high realism test
conditions were presented. These results are compared below with our
predictions for JP-4 and JP-8 (for which a complete set of fluid
107
 Table 11. Illustrative Results for Spray Analysis

ITEN JP-4 JP-8 5606 7808 8328'

DROPLET FORMATIUN -- Nozzle 621, 0.021" diem., 8 cc/s, 117 psig

Liquid density, g/cc 0.76 0.81 0.88 0.95 0.85

Surface tension, dyne/cm 22 23 32 30 30
Absolute viscosity, g/cm.s 0.01 0.01 1.15 2.30 4.57
Sauter diameter, micron 48 54 177 206 248
Initial Droplet Reynolds No. 89 101 330 384 462

DROPLET DECELERATION -- Ventilation air z 4 ft/s a 120"F, 14.4 psis

nozzle 0 1 ft from duct

Deceleration Length , cm 6 8 54 73 86
Deceleration time, ms 2 2 15 21 25
Time of duct impact, ma 204 189 21 16 14
Velocity at impact, cO/S 122 122 613 1145 1425
% Deceleration at i"pact 100 100 86 71 63
Transit tim near duct, m 31 31 6 3 3
End of heating time, me 235 220 28 19 17
Weber nmber at Impact 2 3 185 855 1428

HEATING/EVAPOKATION ANALYSIS -- Duct Temperature - 1135 099. F

Boiling Pt., Deg. C 99 192 244 293 432

MoLecuLar weight, g/gote 125 167 266 425 400
Latent heat, NO, J/g 211 229 204 357 105
Specific heat J/g(Deg.C) 2.06 1.95 2.19 2.19 2.19
Hvi eff., J/9 375 565 694 955 1007
NMst transfer (6) number 1.06 0.52 0.35 0.23 0.11
Droplet Reynolds in dt#t BL 0 0 13 31 48
Nusseot No 2 2 4 5 6
Evaporation time,ms 7 16 254 488 1343
(a different heat transfer
mechanism may apply here)
KINETICS OF IGNITION

Activation Energy cat/mole 43060 37780

Frequency factor ms/atm**n 1.17E-09 1.68E-08 not found in literature
Reaction order, n 2 2
Chemical time, me 24 19

IGNITION CRITERIA

(Evap*Chem)/transit time 0.97 1.10

Is there sufficient time? yes no

108
properties was found in the literature. Similar comparisons for the other
fluids cannot be presented because kinetic data for their ignition
reactions could not be located. See Appendix C).

The high realism tests are more complicated than this simplified analysis.
Trends in the dependance of the MHSITs on ventilation air parameters can
be observed (Figs. 50 to 52). Note that:

o In Figure 51, MRSIT decreases with increases in Pa mainly due to

the reduction of the time required for completion of the chemical
reaction associated with ignition. Although, the trend is
correctly predicted for both JP-4 and JP-8, the agreement between
predicted temperatures and measurements is much less for JP-8 than
JP-4. Clearly, the analysis needs improvement.

0 In Figure 52, MHSITs decreases with increase in Ta, which is also

due mainly to a decrease in the chemical reaction time. To
account for this effect, it is assumed for simplicity that the
reaction temperature in the duct boundary layer increases by one
half of the air preheat temperature (over ambient). This is a
standard assumption in the literature. Similar trends are
observed for ooth test data and theory.

o In Figure 53, MHSIT increases with increase in V- mainly due to a

reduction in the transit time available for heating, and a need
for a higher temperature so that ignition occurs within the
shorter time. (The data shown in this figure are for JP-4 sprayed
on a simple duct with clamp. At the time the analysis was carried
out, this was the only test conditions for which there was data
that could be compared with the analysis on a Va,-HSIT plot.)

Note: an initial dip in the MHSiT-Va relationship was observed at a low Va

(0 to 1 ft/sec) in the early tests on a simple duct with 5660. It may be
attributed to the trajectory of the droplets near the duct as a function-
of the relative magnitude of the ventilation velocity (Va) and the natural
convection velocity (Vn) around the hot duct. When Vn dominates over Va,
109
 X 0
topA

>

IfL

_~ It I
V_ 1 o

II1
6 q~ z/

CL Q

It <
K LU)

0 <t

110
 00 <

II~

(IC) C

A.-ANAINOWN1
3VA~fS IH I~lINNI6
 ~00

<2I-..

N 4
L±J

lif z

0
(I) L C5
Clj 00 l

(Spu""OqI-
J."dNII 3DA~fI
OWN1 IO HMNNII
112a
only larger droplets will reach the surface (while the smaller ones will
be entrained above and away from the duct in the natural convection plume
around the hot duct). Since Vn is of the order of I to 3 ft/sec, this
effect is anticipated at Va = 0 to 2 ft/sec and should disappear at higher
Va•

Albc note: The analysis predicts that higher MHSITs are required with
larger drv'plets, although the effects of Vn is not included (explicitly)
in the analysis such inclusion is recommended in future work.

5.2 Strepn Considerations

In the stream tests, the fluid Is injected by a drip tube onto a

horizontal section of the hot duct. The fluid stream impacts the duct and
spreads radially while boiling due to heat transfer from the duct surface.
After a very short distance, the spreading fluid breaks up into rivulets.
The rivulets continue to flow over the duct and further break up to
smaller sizes. (See Plate E-1 in Appendix E) Once vapors are formed,
the remaining processes (mixing with air, further heat up of the mixture
and chemical reactions) are essentially similar to the case of a spray.

Key unknowns in analyzing the above situation are the details of the
boiling phenomenon, a very complex subject even for non-reacting fluids.

Key questions include:

o is boiling in the nucleate or film regime?

0 are the liquid and vapor saturated or not?

o what are the effects of fluid properties and nacelle pressure?

Depending on the answers to these questions, large differences are

expected in the rates of fluid evaporation and the attained temperatures,
which in turn would effect the MHSIT results of this study.

113
To answer the above questions, first, an attempt was made to measure the
temperature time history at the duct surface Immediately below the
injection point of a stream. The measurement was not successful because
of the clutter in the nacelle. (This should be done in the future on
simple duct-type experiments).

Second, the literature was searched for data on the boiling behaviors of
aircraft fluids of interest. No directly pertinent data for either the
above configuration or even the sinpler case of a boiling liquid pool was
found. Accordingly, estimates were made of the boiling regimes for each
fluid and for the test conditions in this study (by indirect means as
presented in Appendix E).

The results are very approximate and require direct verification in future
work. Still, they suggest the following physical picture for the case at
hand. As liquid flows over the hot plate, Its temperature rises but the
fluid remains subcooled. Vapors are formed only very near the surface
where a very thin layer of fluid reaches the saturation temperature. The
vapors rise through and condense In the liquid. The applicable regimes
appear to be mainly subcooled film boiling for all the fluids of interest
except for 83282, where subcooled nucleate boiling and liquid wetting of
the duct also occurs. (See Figs. E-3 to E-7 in Appendix E).

At the edge of the spreading liquid film, the produced vapors exit the
diirt/!inid interface and are available for mixing with air and for
ignition. Because of proximity to the hot duct, the exit temperature of
these vapors (from the interface) may be larger than the saturation
temperature of the fluid (i.e., the vapor superheats under the liquid).
Various correlations were tried based on this physical picture, but the
results were not successful. Clearly, this subject deserves further study
in the future.

114
5.3 Discussion of Spray Vs. Stream Results for the Five Fluids

Driving upor the point of view presented in Sections 5.1 and 5.2, the
major mechanrsti.2 differences and similarities between sprays and streams
includ"z

o differences in the mechanism of heat transfer from the duct to the

liquid, namely, gas phase conduction, and boiling at a hot surface
for sprays and streams, respectively. These mechanisms have
different dependencies on the ventilation parameters that can be
reflected in the MHSIT data.

0 similar gas phase kinetics reactions regardless of spray or

stream. Thus, the effects of ventilation parameters via chemical
kinetics arŽ expected to be similar. Consequently, the discussion
of the effects oZ these parameters on sprays in Section 5.1 also
annlile nnalitntivelv to streams.

c under some conditions, a spray may produce a thin liquid film on

the du:t surface, approaching the situation of a stream as
descrilbd in 8;ection 5.1.1.

Against these thIcoLetical considerations, the results of the systematic

study of the effects of ventilation parameters on the MHSITs for the five
fluids for both 3prays and streams is discussed below. Note that there
are other ýroccsi-s which may come into play that are too numerous to
mention. A1z'-, ,v. three items listed above may take place concurrently
in a givet, test vith reinforcing and/or competing effects. Thus, the
discussion can unly bo 3uggestive and qualitative. More detailed testing
and analysis is reqired to support the discussion on firmer ground.

Effect of Ventiljt;Qn A.r Temperature, Ta

For the heavier fluids (5606, 83282 and 7808), the effects of Ta on MHSITs
are about the same (within 1000F) whether the fluids are injected as a
stream or a spray (from downstream). Also, MHSIT decreases moderately
with increase in ,. (Fig. 46).

On the ether hand, for the lighter fluids (JP-4 and JP-8) the
corresponding effects differ significantly 0
(up to 450 F) for a spray
versus a stream: MHSIT decreases strongly with increase in Ta for sprays
and moderately to negligibly for streams. (Fig. 47).

The above effects can be explained as follows: a higher Ta promotes

faster gas-phase heat transfer but has little effect on boiling heat
transfer, with a concomitant greater reduction of the MHSIT for sprays
than for streams as observed for the lighter fluids. In the sprays of
heavier fluid•,, a liquid film is produced on the duct and hence their
behavior is more like a stream than spray.

Effect of Ventilation Air Pressure, Pa

The effect of decreasing the pressure, to simulate increasing altitude, is

generally an increase of the MHSIT as shown in Figures 44 and 45. Also,
the effect of :Irreasing the pressure to simulate ram is to decrease the
MIISIT. loli is consistent with a decrease of the rate of chemical
reactic.•- at lowet pressure as shown in the calculations presented In
Section 5.1.2.

Furthermore, pressure can affect heat transferLn,,e effect,

(1) negligible for gas-phase heat transfer in sprays, and for film
boiling in stream involving JP-4 and JP-8.

(2) significant tor boiling in the nucleate regime, ,'.ere increasing

the pressure can activate nucleation sites at lower superheat
temperatute, thereby greatly enhancing boiling heat transfer
rates. Such a boiling regime is believed to occur for sprays and
streams of 83282 as found in Appendix E. (The boiling regime for
5606 and 7808 is somewhat intermediate between that of 83282 and
that of JP-4 and JP-8. See Appendix E.)
116
These effects (items 1 and 2) are consistent with the data in Figures 44

and 45 where sprays and streams of 83282 show a much stronger pressure
dependence than the other fluids in either spray or stream.

Effect of Ventilation Air Velocity, V.

As shown in Figure 48 for 5606 and 83282, MHSIT increases with increasing

Va for both sprays and streams. This trend is consistent with the

analysis in Section 5.1.2.

The difference between the results for the spray from upstream and

downstream for 5606 is greater than expected. The spray from upstream
data were acquired with an older nozzle, although of the same type as the
rest of the data. The older nozzle may have been more worn, producing
larger droplets. However, it is unclear whether the differences in these

results can be attributed to the difference in the nozzles.

Data Correlation

The above discussions and the results of Section 5.1.2 suggest that the

MHSIT tempecature might be correlated with the ratio of ignition delay

time to transit time, for both sprays and streams. This ratio depends on
many variables as discussed in Appendix D and is difficult to determine
narticularly for the fluids for which we have no kinetic data.
Alternatively, the parameter Va/LPa2, which is an approximation for this

ratio under the conditions of a kinetic-limited regime and a fixed

chemistry (i.e., same fluid), was used. (Here L denotes a heating length
and was taken to be the duct diameter for the simple duct data and the
projected length of the hot duct along the ventilation air flow direction
for the high realism tests).

The correlation between MHSIT and this parameter was tested and the

results are shown in Figures 54 to 58 for each of the five fluids. Delta
T, on the Y-axis, is the difference between the preheated air temperature
and room temperature, 80 0 F. As noted in Section 5.1.2, the effect of air
preheating can be approximated by increasing the duct temperature

117
 JP-4 CORRELATION
ALL TESTS; SPRAY AND STREAM
1.5 --
_ _-T

1.4 - +

1.3 + +
v
1.2

- 1.1
* 1

0.9

0.8 + AIR HEATED SIMPLE DUCT W/CLAMP

* ALTITUDE (SPRAYM FROM DOWNSTREAM)
* ALT ITUDE (SREA )STREAM)
0.7 I A RAM (SPRAY FROM DOWNSTREAM)
ff(SREAM)
m m_
'e
*, PREHEAT
To {STREM FROM DOWNSTREAM)
PREHEAT (SPRAY
0.6 A, VELOCITY (SPRAY 6ROM DOWNSTREAM)

0.5

1
V VELOCITY (STREAM)
-- I - "-- .

3
......
5I
,JP-4 MHSIT
ALL TES-)TS: SPRAY AND STREAM
1.5

1.3-
+

++V+ S
1.2

1.1

-Uo 11
in"0.1
019

0.8 + AJR HEATED SIMPLE DUCT W/CLAMP

0 ATIJTUDE (SPRAY FROM DOWNSTREAM)
4 ALTITUDE (STREAM)
0.7 A RAM (SPRAY FROM DOWNSTREAM)
v RAM STREAM,
O PREHEAT (SPR-LAY FROM DOWNSYREAM)
0 PREHEAT (STREAM)
0.6 - A VELOCITY USPRAY FROM DOWNSTREAM)
v VELOCIY (STREAM)
0.5 I - - TI - . - I
0 2 4 6 a 10 12

VENTILATION AIR VELOCITY (FT/SEC)

Figure 54. JP-4 Correlation Ba&ed on a Simptftad Expression
of IgnUion Delay and Transit Tim
:- 118
 JP-8 CORRELATION
ALL TESTS; SPRAY AND STREAM
1.5----

1.4-
*v v A
1,3-*~*.
V
V

. ,, 1.1

0
N

0.9

0.8 *t ALTITUDE (SPRAY FROM DOWNSTREAM)

* ALTITUDE (STREM)
A RAM (SPRAY FROM DOWNSTREAM)
o.7 - RAM STREAM)
m PREHEAT (SP•AY FROM DOWNSTREAM)
* PREHEAT (STREAM)
0.6 tA VELOCITY (SPRAY FROM DOWNSTREAM)
V VELOCITY (STREAM)

0.5 1 1
0 1 2 3
)2
Ln(V./Lp

JP-8 MHSIT
ALL TESTS; SPRAY AND STREAM
1.5-

1.4

1.3

1.2 V
V U

I.-o

0.9

0.8 m ALTITUDE (SPRAY FROM DOWNSTREAM)

* ALTITUDE(STREAM)
A RAM (SPRAY FROM DOWNSTREAM)
0.7 - 1o RAIL
PREH TREAM)
T (SPAY FROM DOWNSTREAM)
* PREHEAT (STREAM)
* VELOCITY (SPRAY FROM DOWNSTREAM)
0.6 -3 V VELOCITY (STREAM)

0.5 - I
0 2 4 6 0 10 12

VENTILATION AIR VELOCITY (FT/SEC)

Figure 55. JP-8 Corelation Based on a SimpUfted Expression
of Ignition Delay and Transit Time
119
 5606 CORRELATION
ALL TESTS; SPRAY AND STREAM

1.4 - ,
7.-x +
1.i.

4 x +
1.2-'V +
14 A + v +

+ a

I
a X AIR HFA1EO SIMPLE DUCT
CA - +MAR HIZAW.D SIMPL.E DUCT W CAMP
M ALTITUDE (SPRAY FROM DOANREAMM)
* ALTITUDE (STREAM)
IL7 - A RAM (SPRAY FROM DOWNSTREAM)
V RAM (STREAM)
O3 PREET (SPRAY' FROM DOWNSTREAM)
* PREhEAT (STREAM) DONTEM
A VELOCTIY (WPAY F DOWNSTREAM)
V VELOCrTY (S ) ----
; VELOCITY (SPRAY M UPSTREAM)
,.5 -
-1 3, -

5606 MHSIT
ALL TEST WAY
M A14 STREAM
'.5-"

t.4-

.3 + + x A

1.1 A x*x
01: V V=•
1 V

.• o.g •
V + AIR HEATED SIMPLE DUCT
0.8 X AIR HEATED SIMPLE DUCT W/CLAMP
0 AL71TUDE (SPY MYIOM DONSTREAM)
# ALTITUDE (SOREAM)
A. RAM (SPRAY FOM DOWNSTREAM)
0.7 v T RAM (STREAM)
SOP T (SPRAY FROM DOWNSTREAM)
0 PREHEAT (STREAM)
La a A-VELOCITY SPRAY FROM DOWNST-EAM)
v VELOCIRTY "m)
3W VELOCITY (SPRAY FROM UPSTRIEAM)
0.. .. 1 -1 1 I
0 2 4 6 a 10 2
VErnLAXON A•R VELOCIT (FT/SEC)
Figure 56. 560 4wm Baf.d on a Simpfkd Bxpresson
of Igamn D.oay and Tran4I Tbme
120
 83282 CORRELATION
ALL TESTS; SPRAY AND STREAM

1.2 -

1.3
C44

.-. 1.1
V

4,

0.9 - ,

0.8- v

* ALTITUDE (SPRAY FROM DOWNSTREAM)

0.7 - + ALTITUDE (STREAM)
* RAM (SPRAY FROM DOWNSTREAM)
Y RAM (STREAM)
0.6 - 13 PREH-.AT (SPRAY FROM DOWNSTREAM)
0 PREHEAT (STREAM)
0 VELOCTIY (SPRAY FROM DOWNSTREAM)
0.5 _ _V VELOCITY STREAM)

-1 1 2 2
Lf(Vo/LJa )2

83282 MHSIT
ALL TESTS; SPRAY AND S""EAM
1.5

1.4-

1.3

1.2

Q.--. 1.1I

SIIJ

0.9

0.8-
0> A I ALTITUDE (SPRAY FROM DOWNSTREAM)
07* ALIlTUDE (STREAM)
0.7 A RAM (SPRAY FROM DOWNSTREAM)
1V
) RAM (STRFAM)
0 PREHEAT (SPRAY FROM DOWNSTREAM)
0.6 - * PREHEAT (STREAM)
A VELOCITY SSPRAY FROM DOWNSIREAM)
V VELOCITY STREAM)
0 .5 • I I - I I I I '1

0 2 4 6 8 10 12
VENTILATION AIR VELOCITY (FT/SEC)
Figure 57. 83282 Correfatton &sad on a Simplifted Expression
of Ignitim Delay and Transit Time
121
 7808 CORRELATION
ALL TESTS; SPRAY AND STREAM

1.4

.AA

A
S1.2 9 V
o A v
"-
t. 1.1 U• --

- 1

0.9-

U ALTITUDE (SPRAY FROM DOWNSTREAM)

* AL1TIU! 7 (STREAM)
A RAM (S-RAY FROM DOWNSTREAM)
0.7 -- RA C(TREAII)
0. EPR- HiT (SPRAY FROM DOWNSTREAM)
0.8 * PREHEAT (STREAM
A VELOCITY SPRAY OM DOWNSTREAM)
V VELOCITY STREAM)

0.5 I
0 1 2
)2
La/Lp

7808 MHSIT
ALL TESTS; SPRAY AND STREAM

1.4 ,

1.3

v V
1.2 A

V v
c,- 11 A V v

N 0,9

0.8 U] ALTITUDE (SPRAY FROM DOWNSTREAM)

* ALTITUDE fSTREAM)
A RAM (SPRAY FROM DOWNSTREAM)
0.7 V RAM (STREAM)
O PREHE.AT (SPRA.AY FROM DOWNSTREAM)
0 PREHEAT (STREAM)
A VELOCITY (SPRAY FROM DOWNSTREAM)
0.6 V VEIOCITY (STREAM)

0.5 i 1 1 I
0 2 4 6 8 10 12
VENTILATION AIR VELOCITY (FT/SEC)
Fig•-e 58. 7808 Core4on Bocd on a SimplUled Expression
of Ignklon Dday and Transk Tim-
122
(ijeaction temperature) by half the preheat temperature. The top and
bottom halves of each Figure include the same data shown with the proposed
,:orrelation at the top and MHSIT versus Va at the bottom. This permits
evaluation of the use of this correlation.

Despite the large scatter in all these plots, note that the correlation is
,3ijperior to the use of MISIT. Furthermore, the correlation can be
considered particularly successful for 83282 (except for one data point)
The success of the correlation for 83282 fluid suggests that its ignition
delay is mainly limited by chemical kinetics; while it is not for the
other fluids, where other processes are also important. This finding is
consistent with the low MHSITs found for 83282 and the inference that it
boils in a nucleate iegime. However, more work is required to determine
with certainty the relationships between such findings.

123
 6.0 CONCL.USIONS AND RECOHMENDATIONS

6.1 Conclusions

The results of tiie present study add significantly to the data base-
available on hot surface ignition temperature, particula-ly as it applies
to aircraft engine compartmeut design. The most important features of the
new data are:

o their collection on a simulated portion of an F-16 nacelle using

real components and systems

0 a systematic variation of ventilation air pressuLe, tempeiature

and velocity covering a range of realistic conditions simulating
aircraft operation tinder various ram and altitude conditions

o use of the five flammable fluids of most interest in aircraft

aoplications injected as sprays or streams and determination of
their relative performance tinder identical test condition

o ignitions of 83282 at temperatures below its AlT per ASTM D 2155

Key findings of this study inclide:

The difference in NISIT between sprays and streams is significant for .P-J 1

and JP-8 but less so for 5606, 7808 and 83282.

GeneLally, the MHSIT occurted above the autoignition temperature of the

fluids except for the case of an 83282 (spray or stream) with preheated
600°F ait, where ignition occurred at a temperature below the autoignitlon
temperature of that fluid.

The results of this study were interpreted in light of simplified analyses

of the key pricesses involved in hot surtace ignition. These include
chemical kinetics and droplet atomization, dynamics, heating and
vaporization (for sprays) and regime of boiling heat t':ansfer (nucleate

124
versus film) for streams. Thus the observed differences and similarities
between the various fluids and between sprays and streams were
interpreted. Also a correlation is presented to describe the effects of
the ventilation air test conditions on the MHSIT ior each fluid.

The aircraft engine compartment designer should attempt to employ real

data to establish safe design temperatures in new aircraft engine
compartments. Extreme caution should be employed in any extrapolation of
such data to untested conditions.

6.2 Recommendations

In letrospect, a number of recommendations about this test program can be

made. Modifications of test method and instrumentation as well as new
tests to be performed are suggested.

1. A test using the stream injection of the test fluids on the air heated
simple duct should be performed. No data on the MHSIT for stream on the
simple duct was taken in this test program. A fluid stream directly onto
a thermocouple tack-welded onto the simple ducý would provide temperature
vs. time response of a wetted suriace on the duct and this data would
piovide insight into the boiling regimes and neat transfer coefficients of
the vatious test fluids. This data is important ti allow correlation of
hot surfac.e ignition variables with MHSIT.

2. In the future, installation of a high speed motion picture camera to

film the simple duct during injection and ignition would be a useful tool.
This would provide information as to the exact location of ignition on the
duct as well as the velocity of the flame front.

3. Absolutely scrupulous leak checking and leak repairing is a

requirement for using air to heat a bleed air duct for hot surface
ignition testing. Some simple duct data was called into question due to
the existence of air leaks during the simple duct phase of the IISIT test
prOg0 anl.

125
4,. Since it was found that the local ventilation velocity was more
itspoctant thnm the average ventilation velocity in the obstruction filli
high, realis•m test article, it is desirable in the future to characteri2,
the ventilation velocities at the stream location sites. Measurement o.
liacal ventilation air temperatures may also prove to be more meaningful
than in upstream ventilation air temperature measuremert.

S.. In thfrs test progLam, nitrogen was used to pressurize the fluid
reservoirs in order to inject the fluid at the proper tlowrate. However,
as the fluid was streamed onto thle duct, the observation waz- often, made
that gas would bubble out of the fluid. Assuming that this gas was
nitrogen, it may have affected the local oxygen concentration, driving the
MUSIT higher. To lessen this effect, and to better simulate the aircraft
engine compartment (where fluids are generally pressurized by pumps) air
would be a more suitable gas with which to pressurize the f luid
reservoirs.

6 • efc
,,h of fluid. temperature on MHS!T was not tormal•y studied in

this test program. In spite cf the difficulties of measuring and

controlling fluid temperature, it would be a worthwhile variable to study
in the context of NHSIT.

7. Dlata concerning the effect of air pressure on the fluid AIT's was
sought and not found. Additional testing directed at defining this effect
might help to interpret KHSIT phenomenc-a.

8. Additional research should be diiected to interpret the finding that

for 83282 (only) the MHSIT occurred at a temperature below the AlT.

9. The ignition delay time was measured during these tests and the
results are reported in Appendix A. They should be analyzed in future
work.

126
REFERENCES

1. Beardsley, G., "Investigation of the Effect of Air Velocity on

the Spontaneous Ignition Temperature of Kerosine Sprayed into a
Heated Tunnel Having a Rectangular Obstruction (Flame-holder) In
the Airstream." Rolls-Royce Ltd. Aero Engine Division Technical
Report Rbts/RT. (FEW.10), February 1967.

2. Myronuk, D. J., "Dynamic, Hot Surface Ignition of Aircraft Fuels

and Hydraulic Fluids," Technical Report AFAPL-TR-79-2095, October
1980.

3. Parts, L., "Assessnment of the Flammability of Aircraft Hydraulic

Fluids," Technical Report AFAPL-TR-79-2055, July 1979.

4. Strasser, W4aters and Kuchta, J. M., "Ignition of Aircraft Fluids

by Hot Surfaces Under Dynamic Conditions." Technical Report
AFAPL-TR-71-86, November 1971.

5. Johniron, A. M. and Gtenich, A. F., "Vulnerability Methodology and

Protective Measture for Aircraft Fire and Explosion Hazards, Volume
II, Part 1 - Fire Detection, Fire Extinguishment and Surface
Ignition Studies." AFWAL-TR--85-2060, January 1986.

6. Johnson, A. M., "Fire Extinguishing Agent Evaluation in the

Aitrciatt Engine Nacelle Fire Test Simulator," AFWAL-TR-88-2022,
June 1988.

7. Johunson, A, M., "Optical Flire Detector 'resting in the Aircraft

Engine Nacelle Fire Test Simulator," AFWAL-TR-87-2089, , March
1 988•

8. Johnson, A. M., "'IIft ct:; of Aircraft Engine Bleed Air Duct

Failuie!s on Suiiounditig Aircraft Structure," AFWAL--TR-87-2004,
April 1987.
127
9. Springer, R. J. et al., "Advanced Ultra-Violet (UV) Airctaft Fire
Detection System", AFWAL-TR-82-2062, August 1982.

10, Kunkle, J. S., et al., "Compressed Gas Handbook," NASA GSP-3045,

1969.

~2e
 APPENDIX A: SUMMARY OF HOT SURFACE IGNITION TEST DATA

VENT VENT VENT DUCT FIRE IGNITION

AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
N = no start of injection)

SIMPLE DUCT TEST ARTICLE

5606 SPRAY, HOT AIR HEATED BARE DUCT 8/26/87
0.91 133 14.13 1352 F 6,1
1.81 134 14.13 1351 F 6.3
4.00 126 14.13 1309 N
3.93 124 14.13 1328 N
4.00 124 14.13 1336 F 6.2
3.99 122 14.13 1305 F 6.4
6.00 130 14.12 1303 N
6.06 130 14.12 1331 N
8.06 130 14.12 1332 N
8.02 123 14.12 1333 N
8.02 128 14.12 1330 N
6.28 129 14.12 1341 N
6.22 128 14.12 1338 N
1.91 127 14.12 1301 F 6.2
0.75 134 14.12 1295 F 6.1
0.89 142 14.12 1248 N
0.99 144 14.12 1250 N
0.70 147 14.12 1251 N
1.96 137 14.12 1251 F 6.8
3.94 133 14.12 1249 N
3.95 133 14.12 1253 N
3.99 131 14.12 1252 N
1.95 127 14.12 1205 N
2.19 127 14.12 1198 F 7.6
1.88 128 14.12 1150 N
i.75 i27 14.1i2 libO
2.00 127 14.12 1150 N
SIMPLE DUCT TEST ARTICLE
JP-4 SPRAY, HOT A•TI HEAITE BARE PUCT E/27/88
0.90 129 14.42 1333 N
1.09 136 14.43 1337 N
1.95 123 14.43 i358 N
1.94 125 14.43 0359 N
1.82 1?5 14.43 1363 N
3.97 113 14.43 1370 N
4.04 120 14.43 1368 N
4.01 117 14.43 1366 W
5.73 117 14.43 1370 N
5.95 116 14.43 1374 N

A -I
 VENT VENT VENT DUCT FIRE IGNITION,
ALR AIR AIR TEMP IGNITE? DELAY
KELQCI1TY TEMP PRESS
('ft/sect) (deg. F) (psia) (deg. F) F = yes (seconds after
N - no start of injection)
6.01 117 14.43 1377 N
8.0 11.7 14.43 1370
6
8.07 118 14.43 1369 N
8.11 117 14.43 1370 N
1.68 124 14.43 1373 N
1.36 143 14.43 1357 N

SIMPLE DUCT TEST ARTICLE

JP-4: SPRAY, HOT AIR HEATED DUCT WITH CLAMP 8/28/87

1.1Z 134. 14.25 1344- F 5.9

1.79 133 14.25 1363 F 5.9
3.68 l1" 14.28 1372 F 6.3
a.w 122 14.24 1364 N
5.99 118 14.28 1369 N
5.99 117 14.25 1372 N
8.12 117 i4.28 1373 N
8.14 116 14.24 1376 F 6.9
1.15 126 14.28 1304 F 6.4

4.05 124 14.29 1316 N

4,04 124: 14.28 1Z17 N
4 n2 124 14.23 1,' N
8.12 118 14.29 1330 N
8.17 119 14.28 1337 N
8.11 118 14.28 1337 N
1.36 126 14.28 1257 F 6.4
2.07 127 14.28 1254 N
1.98 128 1.4.28 1255 F 6.4
1.10 138 14.28 1205 N
1.12 13q 14-26 1201 N
1.29 144 14.Z6 1204 N
1.94 133 14.28 1209 N
2.05 130 14.28 1202 N
1.82 130 14.26 1205 N

SIMPLE DUCT TEST ARTICLE

5606 SPRAY, HOT AIR HEATED DUCT WITH CLAMP 8/31/87

1,27 142 14.21 1287 F 6.1

1.91. 127 14.21 1301 F 6.3
4.01 114 14.21 1316 F 6.2
5.95 112 14.21. 1323 F 6.5
8.10 112 14.21 1330 F 18.9
).31 131 14.20 1243 F 6.3

A -Z
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F - yes (seconds after
N - no start of injection)
1.96 129 14.20 1255 F 6.3
4.05 122 14.20 1266 F 6.2
5.98 116 14.20 1274 F 13.1
8.03 116 14.20 1301 N
8.06 116 14.20 1303 N
8.07 115 14.20 1307 N
1.28 129 14.20 1205 F 6.4
1.87 132 14.20 1207 F 6.4
4.05 126 14.20 1217 F 6.4
5.91 121 14.20 1227 N
5.96 121 14.20 1250 F 7.5
1.13 128 14.20 1155 N
1.43 134 14.20 1165 F 6.4
2.05 135 14.20 1160 F 6.5
4.00 128 14.20 1173 F 6.8
5.97 124 14.20 1182 N
6.00 122 14.20 1198 N
5.97 121 14.20 1198 N
1.15 131 14.20 1103 N
1.29 131 14.20 1110 F 25.3
1.97 134 14.20 1111 F 6.6
3.96 129 14.20 1122 F 6,6
1.24 135 14.20 1048 F 7.4
2.12 135 14.20 1064 F 7.4
1.30 134 14.20 1085 F 6.8
1.77 128 14.19 1103 F 7.2
3.91 125 14.19 1131 N
4.17 122 14.19 1120 N
4.05 1i9 i4.20 1123 N
4.01 113 14.18 1156 F 6.5
1.37 125 14.17 1041 N
1.24 128 14.14 1056 N
1.23 131 14.12 1052 N
1.94 122 14.1E 1060 N
1.99 120 14.18 1057 N
2.00 117 14.18 1063 N

SIMPLE DUCT TEST ARTICLE

5606 SPRAY, HOT AIR HEATED DUCT WITH CLAMP (VERTICAL) 9/9/87

1.34 119 14.37 1318 N

1.08 123 14.37 1352 N
0.98 122 14.37 1351 N
1.28 124 14.37 1352 N
1.90 125 14.37 1357 N

A- 3
 VENT VENT 'VENT DUCT fIRE INITION
SAIR AIR TEMP IGNITE? MELAY
*ELLCAT¥ TEMP PRESS
If t/sýec) ideg. F) r1psia) (deg. F) F yes (seconds after
N Flo start 'of injectior)
1.96 124 14. 17 1360 '1
3.17 125 14.37 1353 N
3.69 123 14.37 1356 IN
-3.94 122 14.37 13M2
4.03 121 14.37 153 14
li.89 120 14.37 1346
5.97 120 14.37 1350 t
5.96 119 14.37 1349
3.01 119 14.37 1345 14
6.04 118 143S 1343 IN
6.09 118 1't.36 1347 IN

SMI1E BUCT 'TEST ARTICLE

S606 WPRAY, 4iESISTANCE $LEATEO RARE DUCT 9/15/87

1.13 92 '11.39 .1169 N

1.25 98 14.38 1235 "N
,0.99 10i 14.40 1308 T 8.1
1.'7 103 a-F
14.40 1251
1.19 Hi 14.40 A200 N
1.00 115 11..40 101 F 6.7
1.27 120 14.40 11 f5 F 7.8
1.18 120 14.39 .110i N
1.26 118 14,39 196 N
0.93 117 14.39 10 N
2.0? 115 14.39 1123 N
2.00 113 14.39 1121 h
1.83 112 14.39 1123 'N
2.04 111 14.39 11T Ul
1.90 111 14.39 1181 N
1.92 ill 14.39 1174
1.81 112 14.39 1230 C 7.4
1.97 114 1.4.39 1179 'N
2.11 112 14.39 1172 F 7.4
1.86 112 14.40 117 N
1.95 111 14.39 1123 N
1.89 111 14.39 1125 N
3.96 114 14.39 1.411 N
3.94 116 14.39 1455 F 5.9
4.02 116 14.39 1418 f 7.6
3.99 117 14.39 1362 F 6.6
4.04 117 14-39 1316 N
3.97 117 14,39 1312 N
4.05 117 14.39 1313 N
6.00 118 14.39 1506 f 6.8

A -4
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
N - no start of injection)
6.05 120 14.39 1472 N
6.06 120 14.39 .1472 N
6.08 120 14.40 1471 N
8.08 120 14.41 1544 F 6.4
8.09 120 14.41 1493 N
8.11 120 14.41 1489 N
8.07 119 14.41 1488 N

SIMPLE DUCT TEST ARTICLE

JP-4 SPRAY, RESISTANCE HEATED SIMPLE DUCT 9/15/87

8.08 116 14.34 1533 N

8.09 117 14.34 155q N
8.10 117 14.34 1544 F 6.4
8.14 119 14.33 1488 N
8.08 120 14.32 1484 N
8.12 120 14.32 1488 N

6.20 122 14.31 1523 F 6.2

6.10 123 14.34 1474 N
6.09 121 14.34 1467 N
6.05 122 14.34 1470 N
4.02 123 14.34 1496 F 6.4
4.11 124 14.34 1454 N
4.03 124 14.34 1448 N
4.08 124 14.34 1449 N
2.03 125 14.34 1474 F
2.22 127 14.34 1424 N
1.99 127 14.34 1424 F 6.2
2.15 127 14.34 1381 N
2.20 127 14.34 1379 F 6.6
2.16 125 14.34 1332 N
2.15 124 14.33 1333 N
2.22 123 14.32 1333 N
1.22 127 14.33 1412 F 6.2
0.85 133 14.32 1367 N
1.01 132 14.32 1371 F 6.4
1.02 133 14.32 1329 N
1.29 132 14.32 1326 N
1.16 130 14.32 1323 N

SIMPLE DUCT TEST ARTICLE

5606 SPRAY, RESISTANCE HEATED DUCT WITH CLAMP 9/17/87

8.12 13 14.21 1487 F 6.4

A -5
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
N = no start of injection)
8.08 113 14.21 1435 F 6.2
8.14 113 14.23 1385 N
8.12 114 14.20 1386 F 7.3
8.11 114 14.25 1336 N
8.13 115 14.20 1334 N
8.13 115 14.25 1335 N
6.15 116 14.23 1473 F 6.5
6.12 116 14.24 1420 F 6.1
6.08 117 14.24 1368 F 6.9
6.06 116 14.25 1319 F 6.7
6.01 116 14.25 1272 N
5.99 116 14.23 1269 N
6.02 117 14.25 1270 N
4.07 117 14.24 1353 F 6.4
4.09 118 14.23 1307 F 6.4
4.12 117 14.19 1254 F 9.6
4.05 118 14.22 1203 N
4.05 117 14.24 1203 F 10.5
4.10 117 14.25 1159 N
4.07 117 14.Z3 1i5i Nf
4.05 118 14.21 1153 N
2.14 118 14.20 1223 F 8.3
2.11 117 14.24 1176 F 10.4
2.12 117 14.20 1128 N
2.20 117 14.21 1123 N
2.10 117 14.13 1122 N
0.89 118 14.10 1205 F 6.7
1.16 117 14.15 1158 N
1.07 117 14.17 1156 N
1.27 116 14.10 1158 F 7.7
' in 1117 11 1111 nI
1.10 Il, 14.14 1,11w
1.14 115 14.17 1106 N
1.28 116 14.18 1106 N

SIMPLE DUCT TEST ARTICLE

JP-4 SPRAY, RESISTANCE HEATED DUCT WITH CLAMP 9/18/87

8.14 106 14.14 1536 F 6.3

8.12 108 14.06 1488 N
8.11 108 14.09 1483 N
8.15 108 14.02 1486 N
6.07 110 14.10 1521 N
6.04 110 14.08 1522 r 6.2
6.08 110 14,11 1471 N
6.06 111 14.11 1468 N

A ~6
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
N = no start of injection)
6.02 111 14.03 1467 N
4.13 112 14.08 1506 F 6.3
4.04 112 14.15 1453 N
4.02 113 14.18 1449 N
4.03 113 14.18 1449 N
2.06 115 14.19 1425 N
2.07 115 14.18 1426 F 6.4
1.92 115 14.18 1374 N
2.07 116 14.17 1372 N
2.06 114 14.18 1375 N
0.98 116 14.19 1407 F 5.9
0.88 118 14.18 1360 F 6.7
1.24 119 14.18 1309 N
1.14 117 14.19 1306 N
1.33 117 14.18 1306 N

A-7
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
fELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) f = yes '(seconds after
N = no start of injection)
HIGH REALISM TEST ARTICLE
5606D, SPRAY FROM UPSTREAM 11/17/87

1.08 100 14.29 1364 F 1.3

2.06 127 14.29 1295 1
1.8
.14 117 14.29 1279 F 1.7
6.17 103 14.29 1299 F 6.2
7.90 126 14.29 1336 F 2.3
1.07 133 14.29 1249 F 2.5
1.95 133 14.28 1207 F 2.5
4.03 129 14.28 ],215 F 4.5
6.04 122 14.28 1200 N
5.88 119 14.28 1249 N
5.95 113 14.29 1302 F 4
7.85 112 14.28 1302 F 5
8.11 113 14.29 1256 N
8.07 125 14.29 1252 F 5.3
5.94 131 14.29 1252 F 0.8
6.00 119 14.28 1265 F 6.6
ifl 11 IA ) 0117 9 q

1.93 136 14.28 1160 N

1.90 138 14.28 1156 N
1.88 143 14.28 1156 N
4.07 159 14.28 1172 N
3.94 159 14.28 1185 F 5.3
3.95 162 14.28 1163 N
3.89 162 14.28 1157 N
3.89 161 14.28 1154 N
5.63 155 14.28 1212 F 2.7
5.65 146 14.28 1169 N
5.! 144 14,28 1165 N
6.04 137 14.28 1162 N
8.03 123 14.28 1216 N
7.66 118 14.28 1223 N
7.61 119 14.28 1212 N

HIGH REALISM TEST ARTICLE

f606, SPRAY FROM UPSTREAM 11/18/87

1.04 93 14.47 1247 F 3.3

1.03 97 14.47 1201 F 6.1
1.23 215 14.47 1158 F 4.7
1.04 118 14.47 1090 N
1.03 118 14.47 1096 N
1.00 121 14.47 1101 F 5.7

A -8
 •I
VENT VENT VENi DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
N = no start of injection)
0.98 123 14.47 1056 F 5.1
0.97 130 14.47 1004 N
0.96 134 14.47 996 N
0.95 126 14.47 1001 N
1.00 136 14.47 1099 F 4.4
1.02 147 14.47 1105 F 5.7
5.80 117 14.47 1252 F 5.7
5.85 119 14.48 1201 F 6.2
5.91 117 14.48 1153 N
5.80 113 14.48 1150 N
6.09 i11 14.48 1151 N

20fps, all no fire

0.00 125 14.48 1335 N
0.00 113 14.47 1363 N
0.00 117 14.47 1373 N

4ml/sec spray flowrate

1.00 157 14.48 1199 F 3.1
1.01 131 14.48 1150 N
1.00 143 14.48 1147 F 2.6
1.01 131 14.48 1102 N
1.01 140 14.48 1098 N
1.01 147 14.48 1097 N

HIGH REALISM TEST ARTICLE

5606 SPRAY FROM UPSTREAM, 4ml/sec, 11/19/87
.As
In va r 1
1.94 122 14.50 15'1 r 1.J
1.91 123 14.49 1199 F 2.3
1.93 124 14.50 1154 N
2.03 124 14.49 1149 N
1.96 123 14.50 1151 N
4.03 131 14.49 1249 F 1.9
3.98 130 14.50 1205 N
3.97 128 14.49 1198 F 4
3.99 125 14.49 1154 N
3.85 123 14.49 1149 N
3.95 122 14.49 1150 N
5.94 113 14.49 1256 N
5.99 136 14.49 1345 F 1.2
6.06 136 14.49 1305 F 1.7
5.94 134 14.49 1251 F 2.2
5.98 132 14.48 1205 N
5.95 131 14.48 1200 N

A -9
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
Jft/sec) (deg. F) (psib) (deg. F) F = yes (seconds after
N = no start of injection)
5.83 128 14.48 1199 N
7.97 120 14.48 1294 F 7.5
7.76 118 14.48 1247 N
7.92 119 14.48 1248 N
7.93 122 14.48 1251 N
0.00 148 14.46 1245 F 2.6
0.00 165 14.46 1203 F 3.1
0.00 147 14.45 1148 F 6.1
0.00 143 14.45 i102 K
0.00 147 14.45 1096 N
0.00 118 14.45 1097 N
0.93 117 14.45 1249 F 1.4
1.00 110 1..45 1203 F 1.6
1.00 104 14,45 1154 F 1.4
1.00 10 14.44 1105 N
1.01 116 14.45 1099 N
0.99 122 14.45 1097 N
!.0z 136 14.45 1247 F 2.9
f •l 1ll IA A 1gni F 3.1
1.01 124 14.44 1153 F 4.8
0.99 118 14.44 1101 F 3.5
1.00 110 14.44 1050 F 4.4
0.99 108 14.44 1007 N
1.02 115 14.44 996 N
1.00 118 14.44 997 F 3.1
0,99 110 14.44 958 N
1.00 114 14.44 949 N
1.00 117 14.44 949 N

HIGH fEALISM TEST ARTICLE

5606, SPRAY FROM UPSTREAM, 300 PNZFUL, 11/20/87

0.00 131 14.35 1193 N

0.00 131 14.35 1241 F 2
0.00 128 14.34 1206 F 3.3
0.00 131 14.33 1153 F 5
0.00 125 14.33 1103 F 4.7
0.00 131 14.32 1052 N
0.00 114 14.32 1053 F 4.9
0.00 131 14.33 1003 N
0.00 98 14.35 999 N
0.00 113 14.26 997 N
7.63 130 14.35 1302 N
7.82 124 14.35 1353 f 2.1
7.62 116 14.35 1296 F 5.7

A -10
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
N = no start of injection)
7.66 110 14.35 1250 N
8.00 113 14.35 1248 N
7.S1 114 14.34 1246 N
4.01 122 14.36 1248 F 4.4
3.91 114 14.36 1204 N
3.94 115 14.37 1198 N
4.38 115 14.37 1198 N
1.09 127 14.37 1249 F 2.3
1.10 146 14.38 1197 F 4.8
1.02 115 14.37 1153 N
0.98 110 14.38 1149 F 7,5
1.01 123 14.38 1102 F 5,7
1.00 128 14.37 1053 N
0.99 117 14.38 1047 N
0.99 115 14.37 1050 F 5.9
1.04 123 14.37 1004 N
1.00 109 14.37 999 N
0.96 107 14.37 1000 N

HIGH REALISM TEST ARTICLE

5606 DRiP, LOCATIONS 1 AND 2, 1/20/88
LOCATION 1, 11.8 SECOND INJECTION

1.09 144 14.17 1079 F 1.9

1.12 182 14.17 1130 F 1.6
1.09 169 14.16 1102 F 1.9
1.02 136 14.15 1034 F 15.4
1.00 140 14.15 1024 F 15.2
1.00 i23 14.14 954 F 16
1.02 103 14.14 924 N
i.01 114 14.14 932 N
1.01 120 14.13 939 N
7.90 109 14.13 1031 F 1.8
7.97 115 1f.13 1043 F 15.4
8.04 119 14.13 1007 N
7.f5 122 14.13 961 N
7.93 122 14.13 1009 N

LOCATION 2,

1.01 134 14.13 1256 F 1.2

1.01 144 14.13 1205 F 3.7
1.01 127 14.12 1167 F 1.7
1.00 142 14.13 1082 F 1.6
1.00 122 14.13 1063 F 1.8

A - 11
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F - yes (seconds after
N = no start of injection)
1.02 147 14.13 1029 F 1.6
0.99 127 14.13 983 F 1.7
1.02 117 14.13 938 F 2
0.99 110 14.13 855 N
1.00 129 14.13 879 N
0.98 134 14.13 885 N 4

HIGH REALISM TEST ARTICLE

5606 DRIP, LOCATIONS 2 AND 3, 1/21/88
location 2,

8.15 116 14.36 1199 F 2.1

7.85 120 14.35 1134 N
7.92 120 14.35 1140 N
7.97 120 14,36 1129 N
7.98 117 14.34 1190 F 4.9

location 3, 11.1 sec injection

1.06 165 14.33 1187 F 1.5

1.01 150 14.33 1136 F 2.4
0.97 132 14.33 1088 F 2.4
0.99 139 14.33 1U46 F 2.2
1.00 134 14.33 998 F 2.2
0.98 129 14.33 948 F 3.6
1.00 121 14.33 898 F 2.6
1.00 122 14.33 847 F 2.6
1.01 112 14.33 799 F 2.9
0.99 93 14.33 743 F 2.9
0.99 97 14.33 699 N
0.99 108 i4.33 697 F
1.01 Jf1 14.33 645 N
1.02 112 14.33 646 N
1.02 109 14.32 647 N
1.00 111 14.32 702 N
l•O0 116 14.32 749 F 4.2
1.00 121 14.32 698 N
0.99 122 14.32 696 N
1.02 118 14.32 699 N

HIGH REALISM TEST ARTICLE

5606 STREAM, LOCATION 3, 1/22/88

8.02 128 14.36 996 N

7.88 130 14.36 1091 F 5.1

A - 12
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
N = no start of injection)
7.95 124 14.36 1040 N
7.87 121 14.36 1039 N
7.97 120 14.36 1039 F 9.4
7.75 117 14.36 994 N
7.96 114 14.36 991 N
7.64 112 14.36 993 N
HIGH REALISM TEST ARTICLE
5606 STREAM, LOCATION 4 AND 5, 1/26/88
location 4

1.14 151 14ý43 1152 N

0.97 182 14.44 1248 F 3.7
0.72 200 14.44 1193 F 4.7
1.07 166 14.43 1151 F 15
1.02 154 14.43 1101 F 10.5
1.03 150 14.42 1053 F 13
1.00 153 14.42 1002 N
1.0? 172 14.4Z 1002 F 4.5
0.94 145 14.42 951 N
0.98 159 14.42 951 N
0.99 169 14.43 949 N
7.77 128 14.42 1241 N
8.13 144 14.42 1288 F 3.4
7.86 150 14.42 1240 N
8.04 126 14.42 1234 N
8.06 125 14.42 1237 N
8.09 117 14.42 1286 F 2.9
location 5, str.ai.... on

1.02 142 14.42 1240 F 2

1.03 167 14.43 1188 F 1.9
1.02 172 14.43 I146 F 2.1
0.99 146 14.43 1090 F 2.1
1.00 130 14.43 1042 F 2.6
0.99 119 14 42 993 F 2.6
0.99 108 14.43 930 F 3.3
0.96 109 14.43 890 F 4.6
0.94 89 14.43 841 N
0.98 93 14.43 838 F 4.6
1.00 73 14.43 794 N
0.99 88 14.43 792 N
0.98 92 14.44 796 N

A - 13
 VEWT VENT VENT DUCT FIRE IGNJTIOft
AMi AIR AIR TEMP IGNITE? DELAY
WfuxC ITY TEMP PUESS
(ft/sec) (deg . F) (psia) (deg. F) F = yes (stecoWds after
N= no start of ij~ectiq) .)

HIGH REALISM TEST ARTICLE

5606 STREAM, LOCATION 5 AND 6, 1/27/88
Ioc 5 straight on clamp

8.05 92 14.59 1180 F 5.6

8.12 94 14.59 1129 N
7.95 95 14.59 11W9 K
8.03 103 i4.58 1129 F 7.8
8.08 116 14.58 1081 N
8.07 126 14.%& 1081 F 6.7
7.86 133 14.58 1031 N
7.91 131 14.58 1031 N
7.98 127 14.58 1033 N

location 6, drip 5606, 2ml/sec

0.97 131 14.58 1298 F 3

0.98 138 14.58 1254 F 3.8
0.91 P. ~Y 114 14.
Ii DO'
. C In .1f.i
1.03 137 14.58 1207 N
1.03 147 14.58 1Z07 F 8.7
0.99 109 14.58 1166 N
1.00 130 14.57 1166 N
1.01 144 14.57 1166 N
8.05 117 14.57 1286 F 6.1
7.89 126 14.57 1230 F 3.6
7.80 127 14.57 1189 F 4.5
7.58 125 14.57 1146 N
7.18 120 14.57 1144
IA
N
N
!17
1._ Iq.57

HIGH REALISM TEST ARTICLE

5606 STREAM, LOCATION 3, 1/28/88

0.98 101 14.62 1085 F 3.6

0.99 128 14.62 1036 F 2.6
1.00 117 14.62 988 F 3.9
1.00 114 14.62 935 F 6.3
0.99 115 14.61 887 F 8.6
0.99 115 14.61 838 F 6.3
0.99 117 14.61 788 F 9.3
1.00 119 14.61 738 N
0.99 118 14.61 738 F 6.9
1.00 120 14.60 691 N

A - 14
 VENT VF'Nf VENT DUCT FIRE IGNITION
AIR Ali% AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
N = no start of injection)
1.00 118 14.60 689 N
0.98 115 14.60 688 N

1.00 117 14.61 688 N

1.00 117 14.60 689 N
1.01 118 14.59 741 N

HIGH REALISM TEST ARTICLE

5606 STREAM, LOCATION 3, 1/29/88
2ml/sec, 40sec

1.05 112 14.51 988 F 3.9

1.01 114 14.51 932 F 4.4
1.02 127 14.52 885 N
1.01 131 14.51 936 F 3.7
0.99 135 14.51 886 F 4.3
0.99 136 14.51 838 F 5.2
1.00 143 14.51 792 N
0.98 146 14.51 r42
r
0.99 144 14.51 792 N
0.99 141 14.51 788 N
1.00 137 14.51 787 F 4.1
1.00 132 14.51 741 N
0.99 128 14.50 739 F 10.3
1.00 133 14.50 694 N
1.00 129 14.50 693 N
1.01 128 14.50 693 N

4ml/sec, 40sec

1.02 151 14.50 843 F 39

1.01 142 14.49 789 N
1.00 143 14.50 791 N
1.00 129 14.49 791 F 41.8
1.00 152 14.49 739 N
0.99 131 14.49 739 N
1.02 126 14.49 739 N

ImI/sec, 40 sec

1.00 98 14.48 845 N

1.01 112 14.48 892 F 4.7
1.00 114 14.48 833 N
1.02 116 14,48 830 N
1.02 116 14.48 834 F 3.8

A- 15
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F - yes (seconds after
N no start of injection)

HIGH REALISM TEST ARTICLE

5606 STREAM, LOCATION 3 AND 5, 2/1/88
location 3
3ML/SEC
1.02 110 14.39 987 F 3
1.01 126 14.39 917 F 3.2
1.00 129 14.39 886 N
0.99 132 14.38 886 F 2.7
1.00 131 14.38 835 F 2.7
1.00 127 14.38 786 F 3.7
1.00 124 14.38 734 N
0.99 119 14.38 734 N
1.01 119 14.37 735 N

2ML/SEC
1.01 116 14.37 787 N
1.00 122 14.37 840 F 2.6
0.99 124 14.3, 784 N
1.00 121 14.37 782 F 5.4
1.02 120 14.37 734 N
0.99 118 14.37 731 N
1.00 115 14.37 736 N

IML/SEC
1.01 123 14.37 845 F 4.9
1.01 130 14.37 787 N
1.02 127 14.37 785 N
i.00 126 14.37 10 Fr
1.00 126 14.37 734 N
1.02 124 14.37 736 N
1.01 122 14.37 737 N
HIGH REALISM TEST ARTICLE
5606 drip, location 5, oblique onto clamp, 2 ml/sec
1.02 124 14.36 901 N
1.00 125 14.37 953 N
1.02 125 14.36 1001 N
1.03 146 14.37 1196 F 1.6
HIGH REALISM TEST ARTICLE
5606 STREAM, LOCATION 5 AND 3, 2/2/88

A - 16
 VENT VENT VENT DUJCT FIRE IGNITION
AIR AIR AIR ,EMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
N = no start of injection)
location 5 oblique on clamp
1.00 147 14.50 1142 F 2.2
1.06 161 14.51 1087 F 2
0.98 137 14.51 1039 N
1.01 154 14.50 1042 F 2.5
0.98 126 14.50 995 N
1.00 134 14.50 996 N
1.00 143 14o50 996 F 4.9
0.99 132 14.50 944 N
1.00 131 14.50 945 N
1.00 131 14.50 946 N

location 3
5.85 113 14.49 1092 F 2.4
6.00 114 14.50 1045 F 3.2
5.97 112 14.49 995 F 3.6
5.92 110 14.49 942 N
6.05 112 14.49 942 H
6.01 113 14.49 944 N
4.10 115 14.49 946 F 12.1
4.03 113 14.49 897 F 2.6
3.91 112 14.49 843 N
3.97 112 14.49 841 F 4.7
4.00 109 14.49 796 N
3.98 ]12 14.49 795 N
4.05 113 14.49 793 N
2.01 117 14.49 852 F 2.8
2.00 121 14.50 793 N
3.n) 11n IA AD Ica 3.2
2.01 119 14.49 743 N
2.00 118 14.50 741 N
2.02 117 14.50 741 F 7
1.99 112 14.50 690 N
2.01 114 14.50 685 N
2.00 113 14.50 695 N

HIGIH REALISM TEST ARTICLE

5606 SPRAY FROM DOWNSTREAM, 2/3/88

1.00 112 14.45 1057 N

1.00 119 14.45 1100 F 5.8
1.00 138 14.45 1055 N
1.00 135 14.44 1053 N
0.99 135 14.44 1053 N
0.99 136 14.44 1099 F 5.9

A - 17
 VENT VENT VENT DUCT FIRE I.GlQT Lft
AIR AIR AIR TEMP IGNITE? DKLXY
'JELKILTY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F ý yes (seconds after
N.= no start of injeýKtioi)
8.24 102 14.43 1152 N
7.96, 126 14.42 1199 N
7.87 125 14.41 1256 F 5.8
7.84 124 14.41 1206 N
8.06 124 14.41 1204 F 5.8
7.95 121 14.41 1158 N
718, 119 14.41 1154 N
7.-91 117 14.41 1153 F 6.3
7.98 113 14.41 1106 F 6.2
I.02 118 14.40 1047 N
7.92 113 14.41 1049 N
7.91 113 14.41 1051 N

HIGH REALISM TEST ARTICLE

JP-A SPRAY FROM DOWNSTREAM, 2/3/88

7.98 112 14.40 1253 N

7.90 109 14.40 1281 N
8.05 109 14.40 1281 F 6
7.97 106 14.40 1207 N
8.00 80 14.40 1211 N
8.01 80 14.40 1211 N
7.92 79 14.39 1213 N
1.17 133 14.40 1202 N
1.05 140 14.40 1249 F 5.7
1.00 125 14.39 1I07 F 5.7
1.00 124 14.39 1157 N
1.01 136 14.39 1154 F 6.5
0.96 115 14.39 1105 N
1.00 122 14.39 1100 N
1.00 129 14.39 1099 N

HIGH REALISM TEST ARTICLE

JP-4 SPRAY FROM UPSTREAM, 2/4/88

1.08 177 14.48 1354 F 1.3

1.03 171 14.49 1306 F 1.5
0.95 122 14.48 1251 F 2.2
0.97 134 14.48 1205 F 5.9
0.95 120 14.48 1153 N
0.99 135 14.48 1148 N
1.00 150 14.48 1147 N
8.03 97 14.48 1301 N
8.03 108 14.48 1359 F 1.3
8.00 110 14.48 1306 F 1.7

A - 18
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
N = no start of injection)
7.84 113 14.48 1256 N
8.07 113 14.47 1247 N
7.92 112 14.47 1248 N

HIGH REALISM TEST ARTICLE

JP-4 STREAM, LOCATION 1,2,3,4 , 2/5/88

LOCATION 1
1.05 163 14.56 1130 F 2.7
1.01 144 14.56 1073 N
0.98 132 14.56 1069 N
0.98 127 14.56 1068 F 8.2
1.02 145 14.56 1042 N
0.98 122 14.55 1021 N
0.98 124 14.54 1015 N

LOCATION 2
AUJ
1 Al
114
1A3
c 14.54 1L
1A )1
LJ rF 7
1.02 140 14.54 1163 F 3.5
1.00 136 14.54 1112 F 6.3
1.00 125 14.54 1071 N
0.98 104 14.54 1070 N
1.01 107 14.54 1072 N

LOCATION 3
0.99 101 14.54 ].088 N
1.02 128 14.54 1140 N
1.03 128 14.54 1187 F 18.7
1.0! II 14-54 111R N
1.03 133 14.53 1139 N
1.01 109 14.53 1136 N
1.05 128 14.53 1191 F 12.1

LOCATION 4
1.00 108 14.54 1237 F 9.6
1.00 111 14.54 1183 N
1.01 132 14.54 1187 N
1.00 145 14.54 1186 N

HIGH REALISM TEST ARTICLE

JP-4 STREAM, LOCATION 5 (VAR. FLOW) AND 6, 2/8/88

LOCATION 5, 1.5 ML/S

0.89 164 14.55 1198 F 3.3
0.97 152 14.55 1145 F 3.1

A - 19
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TE14P IGNITE? DEtAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
N = no start of injection)
1.02 116 14.55 1088 N
1.04 136 14.55 1091 N
1.01 151 14.55 1092 F 2.6
0.96 104 14.54 1045 N
1.01 122 14.54 1044 N
1.00 131 14.54 1045 N

LOCATION 6
1.00 134 14.53 1168 N
1.02 142 14.53 1203 N
1.03 176 14.53 1250 F 6.7
1.01 144 14.52 1211 F 10
1.00 126 14.52 1161 N
1.02 132 14.52 1160 N
1.02 134 14.51 1161 N

LOCATION 5, 3 ML/S
0.99 95 14.51 1198 F 1.8
0.98 87 14.50 1145 F 3.9
0.99 88 14.50 1090 N
1.00 103 14.50 1089 N
1.00 107 14.50 1091 N

HIGH REALISM TEST ARTICLE

JP-4 STREAM, LOCATION 5 STRAIGHT ON CLAMP, 2/9/88

1.02 136 14.47 1200 F 2.2

1.03 133 14.47 1144 F 4.6
1.03 123 14.47 1089 N
1.03 127 14.47 1091 N
1.03 136 14.47 1091 N
1.01 154 14.46 1090 N
1.00 96 14.46 846 N
0.99 106 14.46 844 N
0.99 110 14.46 844 N

HIGH REALISM TEST ARTICLE

7808 SPRAY FROM UPSTREAM

1.01 116 14.45 1250 no

1.03 126 14.45 1302 yes 6.2
1.05 149 14.45 1333 yes 3.4
1.04 139 14.45 1303 yes 4.9
1.00 105 14.45 1253 no
1.01 118 14.45 1253 yes 4.6

A -20
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F - yes (seconds after
N - no start of injection)
1.01 99 14.44 1205 no
1.02 114 14.44 1203 no
1.03 122 14.44 1202 no
1.04 131 14.45 1203 no
HIGH REALISM TEST ARTICLE
7808 STREAM, LOCATION 1,2,3,4,5 , 2/12/88
position I PNACIN
1.01 144 14.37 1124 N
1.02 168 14.37 1132 N
1.03 175 14.37 1097 N
position 2, 11.8 sec
1.02 159 14.37 1214 F 2.3
1.01 134 14.37 1175 F 31.9
1.01 121 14.37 1129 N
1.01 146 14.37 1128 N
i.ue 1b9 14.37 1127 N
position 3
1.01 160 14.37 1237 F 2.5
0.99 140 14.37 1188 F 2.4
1.01 140 14.37 1136 F 2.3
1.00 126 14.37 1089 F 3.3
1.00 112 14.37 1039 N
1.00 122 14.37 1038 F 7.5
0.99 137 14.36 993 F 9.2
1.01 107 14.36 936 N
1in 14.37 939 N
1.01 138 14.36 937 N
position 4
1.01 155 14.36 1132 N
1.04 157 14.36 1236 F 2.2
1.03 127 14.36 1191 N
1.03 152 14.36 1191 N
1.03 164 14.36 1192 N
position 5
0.98 132 14.36 1236 F 2.1
1.01 127 14.36 1188 F 4.2
1.02 106 14.36 1140 N
1.01 118 14.36 1138 N
1.00 133 14.36 1137 N

A - 21
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
N = no start of injectior)

HIGH REALISM TEST ARTICLE

7808 STREAM, LOCATION 6 AND 3, 2/17/88

location 6
1.03 154 14.42 1240 N
0.88 165 14.42 1246 N
1.04 173 14.42 1235 N

location 3
2.00 109 14.41 1088 F 3.1
2.04 115 14.41 1037 F 7.8
2.00 117 14.41 991 F 14.3
2.02 116 14.41 938 N
1.99 112 14.41 936 N
1.99 111 14.41 939 N
4.02 106 14.41 1150 F 2
4.08 114 14.41 1088 F 8.3
3.99 120 14.41 1041 N
4.05 124 14.4l 1039 N
4.04 128 14.41 1041 N
5.96 132 14.41 1198 F 1.6
5.97 133 14.41 1134 F 1.8
5.91 131 14.40 1091 N
6.00 132 14.40 1092 F 7.6
6.10 131 14.40 1040 N
5.89 129 14.40 1043 N
5.97 128 14.40 1039 N
7.98 133 14.39 1205 F 1.7
7.89 129 14.39 1132 F 2.5
8.05 126 14.39 1085 N
8.04 124 14.38 1088 N
7.91 121 14.39 1090 N

HIGH REALISM TEST ARTICLE

7808 AND 83282 SPRAY FROM DOWNSTREAM, 2/19/88

7808 SPRAY FROM DOWNSfREAM

1,08 166 14.15 1310 F 2.6
1.01 133 14.15 1256 F 3.6
1.00 138 14.15 1207 F 5.8
0.98 121 14.15 1149 N
1.02 142 14.15 1152 F 6.5
0.98 118 14.15 1107 N
1.00 131 14.15 1104 N

A - 22
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
N - no start of injection)
1.01 136 14.15 1104 F 7 f/
0.98 120 14.14 1054 N
0.99 128 14.14 1052 N
0.99 136 14.14 1052 N
83282 spray from downstream
1.01 140 14.13 1308 F 1.2
0.99 134 14.13 1253 F 4.1
0.97 124 14.13 1205 F 5.7
0.95 il1 14.13 1156 F 5.8
0.98 120 14.13 1106 F 6.1
0.96 105 14.13 1056 F 5.7
0.96 97 14.13 1003 F 6
0.99 89 14.13 945 F 6.1
0.98 81 14.13 902 F 5.9
0.98 68 14.13 852 F 5.8
1.01 88 14.13 802 N
1.01 101 14.13 796 N
1.00 111 14.13 798 F 6.1
Io01 116 14.12 751 N
1.00 12U 14.12 749 N
1.00 128 14.12 753 N

83282 spray from upstream

1.00 120 14.11 1107 N
0.98 120 14.12 1202 F 2.6
1.01 127 14.11 1155 N
1.02 136 14.11 1153 F 6.1
i.0i 140 14.11 1106 N
1.00 148 14.11 1103 N
1.02 156 14.11 1106 N
1.01 171 14.11 1204 F 2.9
0.99 142 14.11 900 N
0.99 147 14.10 907 N
0.99 149 14.10 906 N
HIGH REALISM TEST ARTICLE
83282 STREAM, LOCATION 1,2,3,4 2/22/88

location 1
1.25 147 14.41 994 N
1.13 185 14.41 1083 N
1.27 192 14.41 1122 N
1.19 172 14.41 1131 N
1.13 165 14.41 1153 F 2.5

A - 23
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
N - no start of injection)
1.10 154 14.41 1122 N
1.07 148 14.41 1122 F 2.3
1.04 138 14.41 1087 N
1.01 147 14.41 1077 N
0.99 136 14.41 1075 F 14.3
1.01 150 14.41 1048 F 15
0.98 141 14.09 1003 N
"1.01 173 14.41 980 N
1.02 193 14.40 982 N
location 2
0.93 131 14.40 1167 F 2.1
0.96 134 14.40 1124 F 1.9
1.00 166 14.40 1070 N
1.00 184 14.40 1070 N
1.00 191 14.39 1069 N

1.02 154 14.39 983 F 2.2

1.01 167 14.39 935 F 2.1
A0.99 150 14.38 886 F 1.8
1.01 138 14.39 836 F 2
1.02 138 14.37 787 F 2.3
1.02 135 14.376 737 N
1.02 154 14.37 739 N
1.02 161 14.37 741 N

location 5
1.02 134 14.37 1104 N
1.00 162 14.37 1201 F 12.2
1.00 169 14,36 1203 F 12.2
1.00 170 14.36 1143 N
1.00 196 14.37 1143 N
1.02 187 14.36 1141 N
HIGH REALISM lEST ARTICLE
I83;'82 STREAM, LOCATIOM 5, 6 AND 3, 2/23/88

location 5
1.01 118 14.36 1197 F 1.6
1.1132 1.4.36 1093 F 1.6
1.05 156 14.36 1142 F 19.5
G,910 1.692F 37,8

A -24
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
N = no start of injection)
0.99 116 14.36 944 N
1.00 130 14.36 945 N

location 6
1.03 192 14.36 1247 F 43
1.00 202 14.35 1247 F 2
0.98 158 14.35 1211 N
0.99 177 14.35 1211 N
1.02 188 14.35 1214 N
location 3
0.00 188 14.35 894 F 1.9
0.00 146 14.34 839 F 2
0.00 144 14.35 793 F 2.8
0.00 120 14.34 742 N
0.00 127 14.34 743 N
0.00 113 14.34 743 N
!90 11OA A 4A pal Fv

2.04 129 14.34 839 F 2

2.01 99 14.33 792 N
2.03 116 14.34 790 N
2.02 126 14.34 793 N
4.03 140 14.34 955 F 2
4.10 113 14.33 894 F 2
4.08 121 14.33 837 F 1.9
4.05 111 14.33 795 F 2.8
4.06 117 14.33 750 N
3.91 128 14.33 747 N
3.95 127 14.33 747 N
6.14 138 14.33 957 F 2
6.12 132 14.33 892 F 2
6.10 132 14.33 841 F 2.4
6.05 131 14.33 797 N
5.97 132 14.32 793 N
5.99 132 14.32 795 N

HIGH REALISM TEST ARTICLE

83282 STREAM, LOCATION 3, SPRAY FROM UPSTREAM AND SPRAY FROM DOWNSTREAM
2/24/88

location 3
8.25 103 14.44 942 N
7.76 114 14.44 1051 F 2.1
8.03 116 14.44 994 F 1.8
7.93 117 14.44 946 N

A - 25
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F - yes (seconds after
N = no start of injection)
7.98 118 14,44 944 F 6.4
8.01 116 14.44 895 F 2.6
8.06 116 14.43 846 F 2.4
7.89 115 14.43 797 N
7.91 115 14.43 793 N
7.80 116 14.43 796 N
spray from upstream
1.01 99 14.43 1203 F 3
1.01 119 14.42 1156 F 5.2
1.01 122 14.43 1104 N
1.01 128 14.43 1099 N
1.01 130 14.43 1099 N
1.01 135 14.43 1150 F 5.0
spray from downstream
2.04 82 14.41 1002 F 6.2
1- 0on 1A A9 7 I

2.00 83 14.42 954 N

1.99 90 14.42 954 N
3.99 99 14.42 1108 F 5.5
4.04 103 14.42 1056 F 6
4.04 100 14.42 1007 F 5.7
4.00 100 14.42 953 F 6.9
4.04 94 14.42 902 F 6
4.02 92 14.42 853 F 6.5
3.94 87 14.42 805 F 6.8
3.88 86 14.42 753 N
3.97 100 14.42 756 N
4.07 103 14.41 746 F 6.5
3.92 84 14.41 707 N
3.97 99 14.42 699 N
3.99 106 14.41 695 N
2.00 111 14.42 957 F 5.9
2.03 103 14.41 902 F 7
1.91 80 14.42 852 N
1.97 89 14.42 851 N
1.98 100 14.42 850 F 2.8
1.94 82 14.41 805 N
1.98 91 14.42 806 N
?,03 97 14.41 805 N
1.98 103 14.42 804 N
2.02 108 14.42 805 F 5.9
1.97 88 14.42 754 N
1.97 95 14.42 751 N

A - 26
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
N - no start of injection)
1.97 100 14.41 750 N
2.01 102 14.41 749 N
2.00 107 14.41 749 N
1.98 108 14.41 749 N

HIGH REALISM TEST ARTICLE

83282 SPRAY FROM DOWNSTREAM, 2/25/88

0.00 114 14.44 903 F 6.1

0.00 118 14.44 853 F 1.2
0.00 132 14.44 798 F 6.6
0.00 96 14.44 754 F 4.2
0.00 100 14.44 703 N
0.00 119 14.44 696 N
0.00 97 14.43 701 N
0.95 97 14.43 858 F 3.4
0.99 104 14.43 804 N
1.00 108 14.43 794 F 6.2
1.01 106 14.43 752 N
0.99 109 14.43 743 N
1.00 110 14.43 748 N
6.08 96 14.43 855 N
6.00 111 14.43 907 N
5.93 120 14.43 950 F 6.8
6.00 122 14.42 901 F 5.9
5.94 120 14.43 848 F 5.8
6.04 123 14.43 799 N
6.02 129 14.42 796 N
5.98 129 14.42 805 F 6.5
5.84 125 14.42 751 N
5.80 127 14.42 753 N
5.97 127 14.42 750 N
8.20 14.42 1008 F 6
7.97 121 14.42 955 N
7.94 123 14.42 949 N
8.10 123 14.42 946 N
26.19 110 14.42 1228 N
20.84 110 14.41 1289 F 5.7
20.12 105 14.41 1242 F 5.9
19.10 104 14.41 1215 N
19.84 105 14.41 1201 N
20.37 105 14.41 1199 N

HIGH REALISM TEST ARTICLE

JP-8 SPRAY FROM DOWNSTREAM AND SPRAY FROM UPSTREAM, 2/26/88

A - 27
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F - yes (seconds after
N = no start of injection)

spray from downstream

1.08 142 14.38 1209 F 6.1
1.0] 153 14.37 1155 F 5.8
1.02 161 14.38 1106 N
1.04 160 14.37 1105 N
1.03 167 14.37 1105 N
1.05 186 14.37 1153 N
1.03 156 14.36 1198 F 5.7
1.01 150 14.36 1159 F 5.7
0.98 145 14.36 1105 N
0.98 150 14.36 1099 N
0 98 156 14.36 1099 N
0.00 170 14.36 1097 N
0.00 158 14.36 1100 N
0.00 131 14.35 1101 N
2.02 132 14.35 1105 F 5.7
2.04 142 14.34 1054 N
)An lAf. IA Ot1I~l

1.99 147 14.34 1051 N

3.92 130 14.34 1051 N
4.01 121 14.34 1050 N
3.97 117 14.34 1051 N

spray from upstream

1.01 125 14.33 1306 F 1.5
1.01 120 14.33 1256 F 3.2
1.00 109 14.33 1205 F 2.9
1.00 101 14.33 1155 N
1.01 120 -!53
14A N
1.02 132 14.33 1152 N
1.01 108 14.33 1150 N
2.00 9ý 14.32 1154 N
1.96 100 14.32 1153 N
2.00 105 14.32 1154 F 6.1
2.01 106 14.32 1106 N
2.02 111 14.32 1102 N
2.02 114 14.32 1102 N
0.00 158 14.32 1098 N
0.00 123 14.31 1099 N
0.00 114 14.31 1098 N

HIGH REALISM TEST ARTICLE

JP-8 STREAM, LOCATION 1,2,3,4 , 2/29/88

A- 28
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) 'deg. F) F = yes (seconds after
LOCATION 1 N = no start of injection)
1.04 176 14.45 1174 F 1.2
1.02 159 14.45 1119 F 1.4
1.02 164 14.44 1090 F 1.2
1.01 162 14.44 1040 F 1.2
1.01 157 14.44 1005 F 1.7
0.99 135 14.44 946 F 3.2
0.98 122 14.43 903 F 3.7
1.01 119 14.43 866 N
1.02 134 14.43 848 F 6
0.99 121 14.43 820 N
1.00 133 14.43 799 F 9
0.99 119 14.42 783 N
1.00 131 14.43 773 N
0.99 138 14.42 769 N
LOCATION 2
1.03 131 14.42 976 N
1 Al 1ro IA Alb I A"I7 r- ni r
A .u AI~t It GIV// r IC .0
1.00 144 14.42 1022 F 1.2
1.02 129 14.42 973 N
1.01 146 14.42 976 N
1.00 153 14.41 976 N
0.98 157 14.41 975 N
1.00 165 74.41 1073 N
1.00 186 14.41 1161 F 2.6
LOCATION 3
1.00 144 14.40 992 F 1.6
0.9q 119 14.41 941 F 2.4
1.03 119 14.40 891 F 3
1.03 109 14.40 845 N
1.01 133 14.40 843 F 4.5
1.01 121 14.40 799 N
1.00 132 14.41 798 N
1.01 146 14.40 796 N
LOCATION 4
1.00 131 14.40 992 N
1.01 159 14.41 1103 N
1.02 192 14.40 1204 F 1.4
0.99 152 14.40 1145 F 2.3
1.02 136 14.40 1096 F 3.4
1.01 125 14.40 1052 F 13

A -2
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
N = no start of injection)
ARTICLE
HIGH REALISM TEST
JP-8 STREAP, LOCATION 2,4,5,6 , 3/1/88

LOCATION 2
1.03 131 14.50 1173 F 1.9
1.04 147 14.50 1121 F 2.8
1.04 141 14.50 1078 N
1.03 174 14.48 1077 F 3.2
1.03 159 14.48 1029 F 3.1
1.01 139 14.48 980 N
1.02 163 14.47 979 N
1.01 173 14.47 979 F 3.2
0.98 115 14.47 930 N
1.00 134 14.46 930 N
1.01 147 14.47 931 N
1.01 159 14.46 735 N
1.01 166 14.46 734 N
1.00 168 14.46 735 N

LOCATION 4
1.00 160 14.46 1209 N
1.01 190 14.46 1254 F 3.1
1.00 171 14.46 1191 F 4.7
1.00 158 14.46 1142 N
1.02 187 14.45 1143 N
1.01 199 14.46 1146 F 4.6
1.01 158 14.45 1095 N
1.02 192 14.45 1095 F 3.5
1 .Lfli • 15
i~u j l
14
A
IE
. 45
-r .J
J f lhIL
4 7,,fv
10%9
1.02 182 14.45 1045 N
1.00 193 14.45 1047 N

LOCATION 5
0.99 155 14.44 1193 F 2.4
0.99 157 14.44 1137 F 5
1.04 154 14.45 1092 F 6.5
1.04 14U 14.44 1041 N
1.00 177 14.44 1041 N
0.99 190 14.44 1044 N

LOCATION 6
0.99 123 14.44 1171 N
1.00 144 14.43 1263 F 7.9
1.01 155 14.43 1206 N
1.00 170 14.43 1207 F 6.8

A - 30
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
N = no start of injection)
1.00 144 14.43 1164 N
1.02 152 14.43 1158 N
1.02 152 14.43 1158 N
1.00 150 14.42 1154 F 9.6

HIGH REALISM TEST ARTICLE

JP-8 STREAM, LOCATION 6 AND 1, 3/2/88

LOCATION 6
1.00 128 14.36 1177 N
1.02 148 14.36 1236 N
1.03 118 14.35 1259 N
1.02 169 14.34 1305 F 1.9
1.01 163 14.34 1250 F 3.2
1.03 146 14.34 1207 N
1.01 176 14.34 1209 F 3.7
1.01 150 14.34 1162 F 4.4
i.01 143 14.34 lii• N-•
1.00 172 14.35 1113 N
0.97 184 14.34 1110 N

LOCATION 1
0.00 138 14.34 896 F 2.2
0.00 142 14.34 860 F 2.1
0.00 151 14.34 829 N
0.00 168 14.33 815 N
0.00 174 14.33 815 N
2.00 124 14.34 912 F 3.1
) AI1 1n 1A "A 07r r 7

1.95 99 14.34 834 N

1.97 109 14.35 816 F 2.7
1.98 87 14.35 784 F 3.4
2.03 86 14.36 749 N
1.99 91 14.35 723 N
1.98 96 14.36 716 N
4.08 110 14.36 906 N
4.01 122 14.35 940 N
3.96 127 14.35 955 F 11.2
3.99 129 14.35 965 N
3.97 130 14.35 951 N
4.02 131 14.35 952 N
6.07 129 14.35 1162 F 5.4
6.02 128 14.35 1135 F 7.8
5.81 129 14.35 1097 N
6.00 126 14.35 1080 N

A - 31
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F - yes (seconds after
N - no start of injection)
5.89 126 14.34 1079 N

HIGH REALISM TEST ARTICLE

JP-8 STREAM, LOCATION 1, JP-4 STREAM, LOCATION 1, 3/3/88
JP-8, LOCATION 1
8.23 120 14.34 1161 F 1.5
8.05 121 14.34 1130 F 2.6
7.94 120 14.34 1097 F 2.2
8.01 120 14.34 1054 F 3.6
8.15 122 14.34 1010 N
8.11 122 14.34 997 N
7.91 120 14,33 992 N
6.01 124 14.33 1122 F 2
6.00 123 14.33 1099 F 2.3
5.91 120 14.33 1058 F 3.7
6.01
r. an 119
Ioý) 14.33
A 'J 1020
•n•
N

6.04 118 14.33 992 N

JP-4, LOCATION I
0.00 132 14.32 1125 F 3.5
0.00 123 14.32 1090 F 4.5
0.00 133 14.32 1082 F 3
0.00 141 14.32 1037 F 4.2
0.00 133 14.32 991 N
0.00 159 14.32 981 N
0.00 176 14.32 979 N
1.96 145R 4o 1164 F 4
2.04 130 14.32 1123 F 3.2
2.01 121 14.32 1085 F 4.8
2.01 118 14.32 1038 N
2.04 131 14.32 1029 N
2.02 142 14.31 1033 N
4.06 111 14.31 1140 N
4.03 126 14.31 1164 F 3
3.98 133 14.31 1137 N
3.99 131 14.31 1124 F 10.4
3.95 130 14.31 1104 N
4.08 130 14.31 1086 N
4.04 132 14.31 1080 N

HIGH REALISM TEST ARTICLE

JP-4 STREAM, LOCATION 1, 3/4/88

A - 32
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
N = no start of injection)
6°05 i11 14.35 1155 F 3.8
6.00 113 14.36 1136 F 7.2
5.98 118 14.36 1102 F 9.4
5.94 120 14.35 1060 F 12.4
6.01 115 14.36 1014 N
5.96 116 14.36 1001 N
5.99 118 14.35 999 N
7.97 120 14,35 1167 F 9.1
8.21 125 14.35 1143 N
8.01 126 14.34 1125 N
8.10 125 14.34 1125 F 12
8.16 125 14.34 1104 N
7.92 128 14.34 1089 N
8.02 126 14.34 1090 N
6.05 112 14.34 1167 F 10.8
5.97 117 14.34 1134 N
5.94 117 14.34 1119 N
6.04 1
115 4.33 1137 N
5.97 114 14.34 1136 N
6.06 113 14.34 1130 N
6.08 114 14.34 1135 N

HIGH REALISM TEST ARTICLE

JP-4 STREAM, LOCATION 1, 3/7/88

4.03 114 14.47 1169 F 1.6

4.04 123 14.49 1146 F 8.1
4.02 133 14.49 1110 F 10.8
4.02 140 14.50 1067 N
4.00 144 14.50 1054 F 13.5
4.03 144 14.50 1021 N
3.99 143 14.49 1005 N
3.99 139 14.47 1010 N
6.06 124 14.45 1157 F 10
5.97 126 14.42 1142 N
6.01 129 14.42 1127 N
5.97 129 14.43 1137 N
5.88 131 14.41 1177 F 11
8.16 124 14.41 1162 N
8.01 125 14.41 1177 F 13.9
8.11 121 14.41 1156 N
7.87 119 14.41 1138 N
8.01 117 14.41 1136 N
4.02 151 14.40 1178 F 1.9
3.99 141 14.40 1149 F 2.4

A - 33
 VENT VEN4T VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F - yes (seconds after,
N - no start of injection)
3.96 129 14.41 1100 F 13.1
3.93 124 14.40 1064 N
4.01 129 14.40 1050 N
3.99 127 14.40 1047 N
2.05 185 14.40 1134 F 2.8
2.01 194 14.39 1113 F 9.7
2.00 197 14.40 1071 F 13.3
1.99 187 14.39 1026 N
2.00 184 14.40 1013 F 14.4
1.96 178 14.38 968 N
1.98 171 14.38 956 N
2.00 169 14.38 962 N
HIGH REALISM TEST ARTICLE
JP-4 STREAM, LOCATION 1, 83282 STREAM, LOCATION 3, 3/8/88
0.00 170 14.42 1076 F 13.2
0.00 161 14.42 1038 N
0.U0 180 14.4? 1040 N
0.00 195 14.42 1039 N
1.01 174 14.42 1080 N
1.03 207 14.42 1125 F 11.4
1.01 170 14.42 1072 N
1.03 194 14.42 1082 N
1.01 210 14.42 1082 N
83282, LOCATION 3
1.83 583 14.40 1000 F 0.4
2.03 621 14.39 959 F 0.4

1.95 622 14.39 860 F 0.4

1.94 624 14.39 805 F 0.4
1.91 624 14.38 748 F 0.4
1.95 627 14.38 697 F 0.8
1.97 631 14.39 655 F 0.6
1.97 627 14.38 609 F 0.7
1.93 585 14.39 547 N
1.98 502 14.39 552 N
2.08 481 14.39 557 N
2.25 631 14.38 558 N
2.02 612 14.38 555 N
2.05 650 14.38 599 N
1.98 645 14.37 650 F 1.8
2.03 671 14.37 615 F 2.9
2.02 656 14.37 555 N

A - 34
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
N = no start of injection)
1.98 599 14.38 540 N
2.12 658 14.37 553 N

HIGH REALISM TEST ARTICLE

83282 STREAM, LOCATION 3, 83282 SPRAY FROM DOWNSTREAM, 3/15/88

LOCATION 3
1.97 313 14.34 917 F 2.3
1.95 314 14.34 855 F 2.3
1.97 316 14.34 804 F 2.4
1.98 315 14.34 751 F 2.6
1.99 315 14.34 702 N
2.04 322 14.34 698 N
2.03 329 14.33 702 N
line full
2.08 331 14.34 703 F 2.3
2.04 331 14.33 702 N
2.05 332 14.33 702 N
2.02335 14.133 0
2.00 316 14.33 646 N
2.03 322 14.33 648 N
2.11 332 14.33 647 F 3
2.08 32] 14.33 597 N
2.12 329 14.33 596 N
2.11 332 14.32 596 N

SPRAY FROM DOWNSTREAM

2.02 301 14.32 914 F 1.1
1.97 288 14.32 853 F 4.1
_-98 _83 1432 803 F 3.1
2.00 284 14.32 755 F 2.2
2.01 295 14.32 706 F 2.2
2.05 314 14.32 652 F 6.6
2.08 322 14.32 599 N
2.10 331 14.32 601 N
2.09 341 14.32 603 N

HIGH REALISM TEST ARTICLE

JP-4, JP-8 STREAM AT LOCATION 5 AND SPRAY FROM DOWNSTREAM, 3/31/88

JP-4, LOCATION 5
2.00 290 14.51 1208 F 6.7
2.02 296 14.51 1152 N
2.07 307 14.51 1147 N
2.07 316 14.51 1144 N

A -35
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F - yes (seconds after
full line N - no start of injection)
2.07 323 14.50 1145 F 8.5
2.02 318 14.51 1090 N
2.03 312 14.50 1086 N
2.03 308 14.50 1083 N
JP-4, SPRAY FROM DOWNSTREAM
2.0? 283 14.50 1207 F
2.02 290 14.50 1154 F
2.00 288 14.50 1105 F 6.8
1.99 294 14.49 1059 N
2.08 301 14.49 1051 F 6.6
2.01 296 14.49 1007 N
2.00 285 14.49 1001 N
2.01 293 14.50 1004 N
JP-8, LOCATION 5
2.03 286 14.49 1194 F 4.6
2.04 294 14.49
1A45 F 6.6
2.02 297 14.50 1097 F 7.8
2.03 299 14.50 1043 F 8.8
2.00 298 14.50 997 N
2.01 301 14.49 988 F 11.6
2.03 295 14.49 946 N
2.00 296 14.48 947 N
2.04 299 14.49 944 F 8.9
2.02 302 14.49 899 N
1.97 299 14.49 890 N
1.99 303 14.49 890 N
line full
1.99 304 14.49 892 N
1.99 308 14.49 895 N
1.97 312 14.49 892 N
JP-8, SPRAY FROM DOWNSTREAM
2.05 280 14.49 1216 F 2.4
2.03 286 14.49 1161 F 5.7
2.04 295 14.49 1108 F 6.2
1.99 300 14.49 1054 F 6.7
2.01 300 14.49 1002 F 6.9
2.04 294 14.49 952 F 7
2.00 295 14.49 902 N
2.07 293 14.49 895 N
1.99 306 14.49 899 N

A - 36
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F - yes (seconds after
N = no start of injection)
HIGH REALISM TEST ARTICLE
7808, SPRAY FROM DOWNSTREAM, 3/31/88

2.00 282 14.49 1210 F 2.2

2.03 292 14.48 1154 F 4.1
2.03 296 14.49 1100 N
1.98 294 14.48 1101 F 6.1
1.92 289 14.48 1056 F 7.1
2.01 287 14.48 1008 N
2.01 289 14.49 1002 N
2.08 298 14.48 1000 N

HIGH REALISM TEST ARTICLE

7808 AND 5606 STREAM, LOCATION 3, 5606 SPRAY FROM POWNSTREAM, 4/1/88

7808, LOCATION 3
2.04 289 14.43 1097 F 1.7
2.03 4196 14.43 1047 N
2.01 294 14.43 1044 F 9
1.99 303 14.43 1001 N
1.98 201 14.43 998 N
1.99 303 14.43 993 N
line full
1.97 303 14.43 993 N
2.00 307 14.43 995 N
2.02 307 14.43 997 N

5606, LOCATION 3
1.99 293 14.43 1000 F 1.7
2.00 294 14.43 945 F 1.6
1.98 294 14.43 901 F 3.5
2.03 296 14.43 849 F 1.4
2.01 292 14.42 796 F 2
1.99 286 14.43 749 F 1.7
2.01 289 14.42 695 F 13.9
1.95 293 14.43 644 F 18.6
1.97 293 14.42 596 N
1.99 293 1.4.42 592 N
1.98 301 14.42 598 N
line full
2.00 303 14.42 600 N
1.97 294 14.42 599 N
1.99 301 14.42 598 N

5606 SPRAY FROM DOWNSTREAM

A - 37
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? hLAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F - yes (secondF after
N ro start of injectilop)
2.01 287 14.41 1211 F 4.2
2.01 290 14.42 1158 F 5.8 -
2.03 295 i4.42 11.04 F 6
2.04 294 14.41 1054 F 6.a
Z.05 ý96 14.41 1002 F 6.3
2.03 296 14,41 952 F 6.4
2.00 293 14.41 901 F 6.3
2.01 293 14.41 853 F 6.1
2.02 296 14.11 802 F 6.4
2.01 296 14.40 749 F 7.3
2.03 300 14.41 697 F 6.5
2.01 297 14.40 b47 N
1.92 e89 14.41 646 N
2.00 300 i4.40 647 N
5606 SPRAY FROM DOWNSTREAM
1.97 559 14,40 798 F 6.2
f0i 55 A A( 7A') A~

2.C5 552 14.39 698 F 6.5

2.02 557 14.40 648 F 6.7
2.03 568 14.40 597 F 7.8
2.01 580 14.40 b4l N
1,95 593 14,39 650 N
2.00 576 14.40 550 N
1.95 572 14-$9 549
0.00 599 14.40 549 F
0.00 601 14.39 497 v
0.00 593 14.39 445 F
I.co 595 14.9 401 F
0.00 574 14.S8 416 F

HIGH REALISM TEST ARTICLE

5606 AND 7808 STREAM ,T LOCATION 108 AND JP-8 SPRAY FROM
DOWNSTREAM !4/4/88)
5606. 1OCATION 3
1.98 581 14.31 794 F 2
1.96 554 14.31 739 F 2.3
1.99 566 14.31 695 F 2.8
2.01 572 11.30 646 F 12
1.99 569 14.30 598 F 2.6
1.99 576 14.30 552 N
1.97 579 14.30 551 N
1.98 584 14.29 550 N
fu' 1 1in

A- 36
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DEI.AY
VELOCITY TEMP PRESS
(ft/ses) (deg. F) (psia) (deg. F) F = yes (seconds after
N = no start of injection)
2.00 600 14.30 550 N
2.02 626 14.29 551 N
2.01 611 14.29 551 N
0.00 614 14.29 550 N

7808, LOCATION 3
1.99 58? 14.29 994 F 1.2
1.98 591 14.29 949 F 1.3
1.99 605 14.29 903 F 1.9
1.q9 596 '4,29 853 F 1.9
1.98 577 14.29 800 N
1.97 576 14,29 799 N
2.05 589 14.29 798 N
full linc
2.G9 598 14.29 798 N

7808 1)SPRAY
7 FROM DOWNSTRIAIN
CC, 1A .10 r
Fin
2.05 546 14.28 1054 F 6.2
2.02 546 14.28 1002 F 6.8
2L07 560 14.28 P50 N
?,06 575 14,28 950 F 7.2
2.05 584 14.28 900 N
2.03 585 14.28 899 N
2.07 594 14.28 903 N

JP-8 SPRAY FROM DOWNSTREAM

2.00 549 14.27 1210 2
1.99 537 14.27 P•57 F 2.1
2.02 537 14.28 1105 F 2.2
2.00 526 14.28 1049 F 2.4
i.99 522 14.27 999 F 2.9
2.00 534 14.27 949 F 3.1
2.04 547 14,27 895 F 3.8
2.05 564 14.27 847 F 5.2
2.0' 55b 14.27 798 F 5.9
2.06 565 14.27 748 F 6.6
2.03 564 14.27 697 F 7.1
1.99 572 14.27 649 N
2.04 582 11.27 651 F
2.01 585 14.27 595 N
2.01 595 14.27 597 F 7.2
1.99 590 14.27 548 N
1.99 591 14.27 547 N
2.00 598 14.27 553 N

A - 39
 VENT I-NT VENT DUCT FIRE IGNITION
AIR ,IR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
N = no start of injection)
0.00 615 14.27 550 F

HIGH REALISM TEST ARTICLE

JP-8 STREAM, LOCATION 5, 83282 STREAM AT LOCATION 3 AND SPRAY
FROM DOWNSTREAM AND JP-4 SPRAY FROM DOWNSTREAM, 4/5/88

JP-8, LOCATION 5
2.06 571 14.30 991 N
2.05 5E8 14.30 1090 F 7.9
2.04 574 14.30 1043 F 8,7
2.02 572 14.29 995 N
2.02 565 14.29 989 N
2.00 562 14.29 998 N
full line
2.04 575 14.29 997 N

83282, LOCATION 3
2.02 557 14.27 789 F 2.3
2.03 558 14.28 746 F 2.4
2.01 567 14.28 691 F 2.6
2.06 578 14.28 646 F 3.9
2.08 591 14.28 597 F 4.4
2.05 601 14.27 545 N
2.02 596 14.28 548 N
2.02 608 14.26 549 N
full line
2.09 621 14.27 549 N

83282, SPRAY FROM DOWNSTREAM

i.99 582 14.27 698 F 2.4
2.01 582 14.26 645 F 2.5
2.00 581 14.27 597 F 2.9
2.06 582 14.26 548 F 3.3
2.04 588 14.26 493 F 3.6

JP-4, SPRAY FROM DOWNSTREAM

2.00 550 14.26 1101 F 2.9
2.03 540 14.26 1056 F 3.1
2.03 545 14.26 1006 F 4.5
2.00 556 14.26 947 F 5.8
2.10 561 14.26 905 F 6.4
2.06 551 14.26 849 F 6.7
2.08 552 14.26 802 F 7.6
2.07 567 14.25 752 N
2.07 571 14.25 750 N

A - 40
 VENT VENt VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
N = no start of injection)
2.15 593 14.?5 753 F
2,03 575 14.25 700 N
2.02 575 14.24 700 N
2.03 594 14.24 701 N
0.00 604 14.24 700 F 7.1

HIGH REALISM TEST ARTICLE

JP-4, JP-8 STREAM A! LOCATION 5 AND SPRAY FROM DOWNSTREAM, 4/7/88

JP-4, LOCATION 5
2110 119 14.25 1174 N
2.16 137 14.25 1296 F 5.2
2.10 137 14.25 1246 F 3.7
2.06 132 14.25 1195 N
2.09 130 14.25 1181 F
2.04 127 14.25 1137 N
2.05 129 14.25 1131 N
2.03 125 14.24 1131 N

JP-4 SPRAY FROM DONWSTREAM

2.15 117 14.24 1319 F 2.4
2.11 121 14.24 1264 F 2.7
2.00 111 14.24 1208 F 5.5
1.96 110 14.24 1162 F 6.5
1.93 105 14.23 1106 N
1.99 108 14.24 1098 N
1.96 105 14.23 1097 N

JP-8 SPRAY FROM DOWNSTREAM

2.04 29 i4.23 13i3 F 2
2.01 124 14.23 1256 F 2.4
1.96 118 14.23 1205 F 1.9
1.99 117 14.23 1153 F 5.9
1.99 113 14.23 1101 F 6
2.00 115 14.23 1056 N
1.99 117 14.23 1048 N
1.99 117 14.24 1051 N

JP-8, LOCATION 5
1.93 101 14.23 1188 F 4.8
1.97 101 14.23 1138 N
2.05 109 14.23 1130 N
2.04 113 14.23 1131 F
2.00 103 14.24 1089 N
2.02 109 14.23 1082 N

A - 41
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (ps:ia) (deg. F) F = yes (seconds after
N = no start of injiection)
2.00 10q 14,24 1033 N
HIGH REALISM TEST ARTICLE
5606, 7808 SPRAY FROM DOWNSTREAM, 7808 STREAM AT LOCATION 3
JP-4, JP-8 STREAM AT LOCATION 5, 4/8/88

5606 SPRAY FROM DOWNSTREAM

2.03 107 1A.41 1214 F 4
2.07 112 14.10 11.62 F 2.9
2.06 107 14.40 1100 F 5.9
2.01 103 14.40 1052 F 6.2
2.02 102 14.40 1003 N
2.02 109 14.40 998 F 6.3
1.96 102 14.39 953 F 6.3
1.97 97 14.39 90c F 6.8
1.95 95 i4.18 851 F 6.7
1.96 101 14.38 800 N
1.98 112 14.38 793 F 6.5
1i fl 11% Ii F
1.93 104 14.37 702 N
1.97 107 14.37 700 N
1.96 108 14.37 703 N

7808 SPRAY FROM DOWNSTREAM

1.95 121 14.36 i108 F 6.6
1.95 101 14.36 1057 N
1.96 110 14.36 1060 N
1.99 115 14.36 i058 F
1.97 110 14.36 1008 N
2.11 A6 14.36 1003 Pd
2.01 119 14.36 1002 i
7808, LOCATION 3
2.02 316 14.36 1092 F 3.7
2.04 316 14.36 1042 N
2.08 323 14.36 1044 F 8.1
2.02 326 14.36 993 N
1.96 302 14.36 978 N
1.94 305 14.36 991 N
JP-8, LOCATION 5
1.95 573 14.36 1092 F 5.6
2.05 569 14.36 1041 N
1.93 554 14.36 1039 N
2.00 572 14.36 1043 N

A - 42
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deqo F) F yes (seconds after
N no stirt of injection)
,JP-4, LOCATION 5
1.99 5F5 14.36 1136 14
1.9s 560 14.36 12!4 F 3-1
"2.CC 571 14.35 1182 F 7
2.03 580 14.36 1138 N
2.05 583 14.36 1140 N
2.05 589 14.36 1141 H
HIGH REALISM TEST .ARTICLF
5E65. 83282 STREAM AT LOCATION 3, JP-4 STREAM AT LOCAlION b,
VEmgilATION AIR PRESSURE TESTS 4,'22/88
5606, LOCATIOn 3
2,18 !18 9.83 999 N
?.08 119 10.49 1201 F 1.5
2.18 137 10.15 1147 N
2.22 153 10.06 1150 F 2,3
r- . u It.f2 1 uA3
2.10 157 10.05 1092 N
2.04 166 10.06 109e N
1.95 118 10.2b 1185 2.7
i.98 121 10.25 1144 F 8.6
2.02, 120 10.17 1095 F 7.5
1.94 116 10.15 1042 N
2.06 138 10.06 1039 N
2.04 iA9 10.02 10433 N
1.99 133 5.22 1327 N
2.11 143 4.76 1317 N
!55
1. 517 !324 F 2.5
2.04 120 5.38 1298
I.S9 137 5.37 1293 N
2.0 152 5.17 1290 N
JP-4, LOCATION 5
2.13 98 4.95 1324 N
2.05 116 5.C9 1309 N
2.03 127 5.04 1308 N
2.00 134 5.04 1303 N
2.00 133 10.14 133] F 3.5
2.04 125 10.i8 127? 1" 3.4
2.05 121 10.19 1237 N
1.99 130 10.13 1213 u 7.5
2.03 123 10,02 1i9u N
2.08 i43 9.97 1173 N

A - 43
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F yes (seconds after
N = no start of injection)
2.02 1A9 10.03 1177 N

83282, LOCATION 3
2.12 95 9.71 1193 F 2.6
1.97 95 10.24 1147 F 12.6
1.97 96 10.31 1096 N
2.02 110 10.?3 1093 N
2.05 119 10.08 1094 N
2.01 1?6 5.22 1350 F 2.9
1.99 123 5.28 129e N
2.02 142 5.17 1286 N
1.97 155 5.07 1262 N

HIGH REALISM TEST ARTICLE

7808 AND 5606 STREAM AT LOCATION 3;
7808, JP-4, 83282, JP-8, 5606 SPRAY FROM DOWNSTREAM;
JP-8 STREAM AT LOCATION 5,
VENTILATION AIR PRESSURE TESTS 4/25/88
7808, LOCATION 3
2.11 126 10.18 1294 N
2.09 136 10.11 1341 N
1.99 181 10.30 1341 F 2.1
1.97 171 10.29 1290 1
1.89 160 10.18 1281 N
1.91 155 10.11 1287 N
2.29 127 5.03 1335 N
2.02 160 5.29 1332 N
2.01 175 5.23 1333 N

7808 SPRAY FROM DOWNSIREAM

1.97 i14 10.07 1359 N
2.12 144 10.0: 3355 N
2.05 i65 I0.,i 1347 N
2.56 133 4.98 1254 N
2.C5 165 r.06 1344 N
1.97 137 1.]2 1349 l

J9- 4 SnRAV FROM DOWNSiREAM

2.13 145 10.06 1352 N
2.17 222 '0.02 1347 N
1.94 260 10.09 1349 N
4-.• 155 5.01 1349 N
2.l5 229 .;.i1 P345 N
2.,0 2G6 S.09 i34Z N

A 44
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
N = no start of injection)
83282 SPRAY FROM DOWNSTREAM
1.98 102 9.98 1340 N
2.08 161 10.00 1348 N
2.03 188 10.15 1353 N
1.86 137 14.38 1362 F ---
2.07 147 5.05 1341 N
2.00 169 5.13 1349 N
1.91 187 5.22 1353 N
JP-8, LOCATION 5
1.88 83 10.08 1346 N
2.04 113 10.01 1340 N
2.04 133 10.20 1334 N
2.07 161 14.37 1340 F ---
2.18 123 5.01 1343 N
2.00 141 5.11 1317 N
2.03 159 5.04 1317 N
JP-8 SPRAY FROM DOWNSTREAM
1.81 129 14.37 1364 F
1.93 133 10.05 1362 N
2.07 187 10.06 1364 N
2.08 213 10.13 1361 N
2.07 122 5.04 1362 N
2.02 147 5.07 1360 N
2.08 180 5.05 1358 N
rcnic ennipy rnnM nnukic-1nrp A
JVVV OF\fUF I I RVjI UUMIIJ I n fI'|L
2.11 125 10.08 1367 N
2.12 148 10.05 1359 N
2.03 164 10.16 1358 N
2.19 127 5.10 1352 N
2.10 163 5,02 1351 1
2.02 183 4.93 1353 N
5606 SPRAY FROM DOWNSTREAM (LOW INJECTION PRESSURE)
1.89 142 10.02 1366 N
!.99 175 10.13 1365 N
2.01 188 10.12 1364 N
S54 236 14.26 1346 F ---
5606 SPRAY FROM DOWNSTREAM
10.83 76 19.71 1196 F 5.4
I0,4 83 19.88 147 N

A - 45
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
N = no start of injection)
10.15 75 19.83 1156 N
10.17 74 19.83 1155 N

JP-8, LOCATION 5
11.67 77 20.18 1290 F 5.7
11.53 79 20.16 1239 F 8.5
10.97 75 20.32 1192 N
10.82 74 20.16 1169 N
10.74 72 20.14 1160 N

JP-8 SPRAY FROM DOWNSTREAM

11.43 73 20.58 1200 N
11.25 71 20.18 1310 N
11.05 71 20.07 1345 F 5.5
10.94 73 20.24 1312 F 3.2
10.82 73 20.15 1247 F 5.5
10.87 74 20.01 1208 N
in 7A ~ 7 )
r-IUI
)A
LU.
V
I C
AJ
1lfE
ILr-v.J
10.68 71 20.14 1204 N

5606, LOCATION 3
11.05 77 19.99 1170 F 5.8
10.93 75 20.01 1129 F 5.1
11.06 78 19.89 1079 F 6
10.98 79 19.99 1033 N
10.68 74 19.95 1029 N
10.66 74 19.91 1029 N

HIGH RFAIISM TFST ARTICIF

5606, 7808, 83282 STREAM AT LOCATION 3;
5606, JP-8, JP-4, 7808, 83282 SPRAY FROM DOWNSTREAM;
JP-8, JP-4 STREAM AT LOCATION 5; 4/26/88

5606, LOCATION 3
11.81 70 14.50 1329 F 2.5
10.53 71 14.46 1291 F 3.3
11.58 70 14.47 1234 F 4.2
11.22 69 14.36 1192 F 7.6
11.78 70 14.35 1137 F 5.4
10.69 69 14.34 1091 N
10.98 69 14.34 1085 N
11.30 60 14.35 1083 N

5606 SPRAY FROM DOWNSTREAM

11.25 69 14.35 1300 F 5.7

A - 46
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
N = no start of injection)
11.54 70 14.35 1257 N
9.80 69 14.34 1254 N
10.77 70 14.34 1251 N

JP-8, LOCATION 5
11.44 79 14.34 1285 F 9.8
11.33 74 14.34 1224 F 5.3
12.37 74 14.35 1195 N
11.39 73 14.34 1178 N
10.70 73 14.33 1180 N

JP-8 SPRAY FROM DOWNSTREAM

10.93 76 14.34 1349 F 3.8
]1.28 75 14.34 1303 N
11.11 75 14.34 1293 F 5.9
10.58 74 14.33 1258 N
9.82 75 14.33 1252 N
9.00 74 I4,32 125! N

JP-4, LOCATION 5
10.97 83 14.32 1319 F 12.7
10.01 80 14.32 1291 N
10.81 80 14.32 1264 N
11.29 79 14.33 1263 N
10.92 79 20.15 1281 F 6.4
10.81 78 20.03 1h37 F 11.7
10.77 79 20.04 1192 N
10.68 74 19.99 1171 N
10.83 72 20.04 1165 N
JP-4 SPRAY FROM DOWNSTREAM
11.28 77 14.32 1344 F 2.3
11.47 77 14.31 1296 F 5.8
9.87 76 14.30 1247 N
10.37 77 14.31 1239 N
10.08 75 14.30 1237 N
11.49 79 20.29 1229 N
11.17 73 20.25 1290 F 4.1
11.29 89 20.16 1239 F 5.7
11.10 39 20.03 1193 N
10.76 76 20.04 1190 N
10.61 73 19.98 119 N

7808, LOCATION 3
11.64 84 14.30 1293 F 2

A - 47
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
N = no start of injection)
12.27 92 14.30 1232 F 10
11.22 84 14.29 1189 N
10.56 83 14.29 1186 N
11H98 82 14.30 1184 N
11.52 84 20.21 1183 F 2.3
11.50 86 20.10 1135 F 3.3
11.30 82 20.14 1081 N
11.18 78 19.96 1091 N
11.07 76 19.87 1099 N
7808 SPRAY FROM DOWNSTREAM
9.68 84 14.29 1290 F 6
9.75 79 14.29 1240 N
11.72 78 14.30 1228 N
10.50 78 14.30 1228 F 5.9
11.84 83 14.30 1186 N
9.96 80 14.29 1183 N
11.79 78 14.30 1180 N
11.66 80 19.8i 1178 N
10.94 74 20.06 1237 F 5.6
11.24 80 19.93 1188 F 5.8
10.95 77 20.01 1131 N
10.81 75 19.93 1140 N
10.88 75 20.05 1144 N
83282, LOCATION 3
11.10 98 14.29 1176 F 2.4
10.27 95 14.29 1145 N
11.06 88 14.29 1127 N
I..LIV ou 14.291 1133 N
11.17 85 19.77 1133 F 2.2
11.10 87 19.91 1089 F 3.6
11.11 87 19.87 1043 F 3
11.06 87 19.91 988 F 2.3
10.96 86 19.89 938 F 1.9
10.92 85 19.91 891 F 2.3
10.91 85 19.82 835 F 2.3
10.83 83 19.85 780 N
10.69 81 19.71 791 N
10.82 81 20.54 796 N
S3282 SPRAY FROM DOWNSTREAM
11.31 94 14.29 1196 F 5.9
11.38 86 14.29 1139 N
10.79 83 1.4.29 1133 N

A - 48
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
N = no start of injection)
11.11 81 14.29 1136 N
11.30 83 19,94 1137 F 5.7
11.07 81 20.03 1090 F 5.9
11.03 82 20.00 1042 F 6.2
10.99 82 19.92 991 F 6.2
10.94 84 19.98 943 F 5.9
11.06 85 19.87 893 F 6.3
10.95 84 19.88 833 N
11.11 80 19.94 824 F 6.4
11.08 81 19.87 780 N
11.27 78 19.79 790 N
11.38 77 20.00 793 N
11.25 77 19.93 794 N

HIGH REALISM TEST ARTICLE

83282 SPRAY FROM DOWNSTREAM, NO BLEED AIR FLOW; 4/27/88
VENTILATION AIR TEMPERATURE EFFECT ON MHSI

2.U8 593 14.25 480 F 2.6

1.96 560 14.25 453 F 3.9
1.90 538 14.25 434 F 5.9
2.01 531 14.25 433 F 6.8
1.96 510 14.25 412 F
1,95 486 14.25 393 N
1ý99 482 14.25 386 N
2.00 483 14.26 383 N

83282, 7808, 5606 STREAM AT LOCATION 3;

83282, 7808, 5606, JP-4, JP-8 SPRAY FROM DOWNSTREAM;
JP-4, JP-o8 STREAMI AT LOCATION 5;
BAFFLE TESTS 4/28/88

83282 SPRAY FROM DOWNSTREAM

10.62 223 14.37 999 N
11.34 255 14.37 1105 N
14.15 310 14.40 1210 F 3.3
11.41 339 14.36 1161 F 4.4
12.80 359 14.37 1107 N
13.66 360 14.38 1099 F 6.4
9.93 361 14.35 1053 N
12.27 362 14.37 1048 N
12.36 376 14.36 1048 N

83282, LOCATION 3
11.60 364 14.35 1188 F 1.8

A - 49
 VENT VENT VENT Duct FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F ý yes (seconds after
N = no start of injection)
10.15 366 14.35 1159 F 1.3
11.98 370 14.36 1105 F 1.9
12.64 385 14.37 1056 F 1.2
10.53 381 14.35 1003 F 4.2
11.67 385 14.36 953 F 2.3
11.83 404 14.36 904 F 2
10.21 379 14.35 849 F 2
11.94 370 14.36 797 F 2.3
11.57 351 14.36 755 F 4
10.27 335 14.35 703 N
10.68 340 14.36 698 N
10.45 338 14.35 699 N

7808, LOCATION 3
13.05 247 14.41 1299 F 1.6
10.26 272 14.38 1263 F 1.9
11.74 289 14.40 1208 F 2.5
10.87 274 14.39 1159 N
11.91 311 14.40 1152 N
13.22 314 14.40 1151 F 2.4
10.26 315 14.37 1110 N
12.43 336 14.40 1099 N
13.69 359 14.40 1097 N

7808 SPRAY FROM DOWNSTREAM

12.46 302 14.40 1310 F 0.6
11.23 338 14.38 1259 F 0.8
11.30 337 14.38 1202 F 5.7
10.02 375 i4.36 1142 N
12.17 375 14.39 1147 N
12.23 373 14.39 1151 N

5606, LOCATION 3
11.15 254 14.40 1207 F 2.6
11.77 272 14.39 1158 F 3.4
12.43 288 14.40 1106 F 3.3
11.61 299 14.39 1054 F 2.3
12.46 324 14.40 995 F 2.8
11.84 336 14.39 951 F 3.9
11.76 310 14.39 901 F 6
10.97 302 14.38 855 F 3.2
11.49 283 14.39 800 F 2.9
10.61 263 14.39 755 F 5.2
10.88 250 14.39 702 F 14.4
8.78 237 14.37 652 N

A - 50
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
9.47 N = no start of injection)
239 14.38 639 N
11.84 248 14.40 645 N
5606 SPRAY FROM DOWNSTREAM
13.06 271 14.41 1197 N
12.40 309 14.40 1304 F 3.7
11.63 306 14.39 1257 F 3.3
13.33 338 14.40 1201 F 4.5
11.10 330 14.38 1151 N
10.65 326 14.38 1148 N
11.57 336 14.38 1152 N

JP-4, LOCATION 5
13.20 301 14.40 1341 F 6.7
13.77 306 14.41 1306 F 6.9
11.32 331 14.38 1257 F 8
12.31 362 14.39 1206 N
i2.31 3bZ 14.39 1196 N 'dU
11.59 336 14.38 1199 N
JP-4 SPRAY FROM DOWNSTREAM
14.91 357 14.41 1356 F 0.9
11.70 336 14.39 1306 F 1.5
11.59 351 14.38 1244 F 2.2
12.46 412 14.39 1197 F 5.9
11.57 405 14.38 1153 F 5.7
12.13 398 14.38 1105 N
12.41 404 14.38 1099 N
11.01 A n7 IA nfl lnnfl v eI

12 .55 414 14.38 1056 N

13.55 419 14.39 1049 N
11.71 435 14.37 1047 N
JP-8, LOCATION 5
13.19 270 14.41 1300 F 3.1
10.73 253 14.39 1255 F 4.3
13.34 285 14.41 1204 F 4.9
13.60 326 14.41 1157 N
11.07 336 14.39 114G N
12.11 374 14.38 113n N

JP-8 SPRAY FROM DOWNSTREAM

12.36 273 14.40 1260 F 2.2
10.62 108 14.38 1198 N
10.94 98 14.38 1191 N

A - 51
 VENT VENT VENT DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) r - yes (seconds after
N = no start of injection)
9.38 98 14.37 1193 N

HIGH REALISM TEST ARTICLE

JP-4, JP-8 STREAM AT LOCATION 5, 4/29/88

JP-8, LOCATION 5
0.00 131 14.36 1202 F 2.2
0.00 128 14.36 1151 N
0.00 148 14.37 1140 N
0.00 172 14.37 1157 F 3.1
0.00 133 14.37 1105 N
0.00 163 14.37 1101 N
0.00 183 14.38 1098 N
1.08 144 14.37 1202 F 3.2
1.08 140 14.37 1151 F 4.3
1.07 133 14.37 1105 N
1.14 168 14.37 1103 N
0. 1 3 11Gi N
2.09 132 14.36 1208 F 7.4
2.07 i25 14.37 1154 F 12.3
2.08 131 14.36 1106 N
2.22 166 14.37 1099 N
2.19 185 14.37 1097 N
4.04 127 14.37 1302 F 2.4
4.10 134 14.36 1258 F 2.4
4.07 125 14.36 1203 F 9.6
A.07 127 14.3G 1152 N
4.15 154 14.37 1149 N
4.23 165 1i.37 1148 N m
6.08 118 14.37 1252 F 7.9
6.09 126 14.36 1212 N
6.31 148 14.37 1199 N
6.14 157 14.36 1200 N
8.13 119 14.38 1307 F 6.6
8.08 119 14.37 1261 F 8.3
7.99 116 14.37 1261 F 1.5
7.89 114 14.37 1211 N
8.04 127 14.38 1203 N
8.08 133 14.37 1199 N

JP-4, LOCATION 5
0.00 151 14.35 1300 F 5
0.00 135 14.35 1248 F 7
0.00 137 14.35 1206 N
O.OG 169 14.35 12C3 N

A -52

i.
 VENT VENT VENI DUCT FIRE IGNITION
AIR AIR AIR TEMP IGNITE? DELAY
VELOCITY TEMP PRESS
(ft/sec) (deg. F) (psia) (deg. F) F = yes (seconds after
0.00 190 14.35 1201 N = no N start of injection)
1.06 147 14.35 1302 F 2.5
1.05 143 14.35 1254 F 6.6
1.06 140 14.35 1204 F 9.2
1.05 134 14.35 1154 N
1.11 168 14.35 1151 N
1.13 189 14.35 1152 N
1.99 127 14.35 1306 F 2.3
2.07 136 14.35 1255 F 9.3
2.06 136 14.35 1205 N
2.15 173 14.35 1198 F 10.5
2.03 140 14.35 1155 N
2.15 173 14.35 1148 N
2.22 191 14.35 1148 N
4.06 122 14.35 1306 F 3.8
4.06 122 14.35 1251 F 4.8
4.02 123 14.35 1204 N
• 11 e 14. 3 5 1197 N
4.28 169 14.35 1200 N
5.95 110 14.35 1306 N
6.10 110 14.35 1360 F 6.9
6.08 119 14.34 1300 N
6.22 151 14.35 1301 F 9.8
6.12 147 14.35 1262 F 11.9
6.33 168 14,35 1251 N
6.47 184 14.35 1251 N
6.10 199 14.35 1252 N
8.15 137 14.35 1365 F 2.8
7.98 119 14.35 1309 N
8.02 125 14.35 1306 F 11.3
8.19 124 14.35 ;259 N
8.10 130 14.35 1259 N
8.08 130 14.35 1251 N

A - 53
 APPENDIX B: Temperature Data Uncertainty Analysis

This Appendix is assembled from analyses that were

performed following the completion of the hot surface
ignition testing. It Contains the following:

Item Page

1.) Error Accumulation ROM Estimate B-2

2.) Boeing Advanced Systems Coordination

Sheet L8327-088-PJC-057, Determination
of the Intrinsic Thermocuople
Measurement Error for
the Hot Surface Ignition Test, 28 July,
1988. B-3

3.) Extrapolation of Error Analysis to Air

Temperature Thermocouples B-10

B-i
 1.) Error Accumulation ROM Estimate

TEMPERATURE; 900OF 1400OF

ERROR (ERROR) 2 2
ERROR (ERROR)

T/C WTRE +3.6 12.96 5.6 31L26

(Spccial grade)

EXTENSION WIRE +4.0 16.00 4.0 16.00

(Assumed std. wire)

REFERENCE JUNCTION •0.2 0.04 0.2 0.04

(ElectriA.al MV insertion
type aszurmied)

MODCOM`W AIS ÷!.B 3.24 2.25 5.06

(Based -inMV insertion
data Dzovided)

TABfl[ 'O0K UP 4-0.2 0.04 0.2 0.04

(Vorst case assumed)

49.8 32.28 12.25 52.5

A,_2 + + .. =+ 5.68 +7.25

ADD KNOWN AIS BIAS FACTOR +15.18 0 F +24.85OF

+ 3.82OF +10.35°F

This represents a ROM estimate of potentially significant error sources in

the temperature measurements system. Numbers are based on standard
industry information or typical numbers from past similar work.

B-2
 COORDINATION SHEET

TO: A.M. Johnson 33-14 L8347-088-RJC-057

CC: J.L. Howard 86-12 July 28, 1988
A.L. da Costa 33-18

SUTJECT: Determination of the Intrinsic Thermocouple Measurement

Error for the Hot Service Ignition Test

REFERENCES

[1] S.V. Patankar, Numerical Heat Transfer and Fluid Flow,

Hemisphere Publishing Company, 1980
(2] D.R. Pitts and L.E. Sissom, Heat Transfer, Schaurs
Outline Series, McGraw-Hill Book Company, 1977

SUMMARY
An analysis of an intrinsic thermocouple installation for- the hot
service ignition test was done to determine the error between the
actual surface temperature of the bleed air duct and the
temperature measured by the thermocouple. Results show that for
a wide range of conditions during the test a max'mum error of
approximately -27 F could be expected, for surface temperatures
in the range of 450 to 1350 F. Note that the error produces a
lower thermocouple reading than the actual surface temperature of
the duct, due to the fin effect of the thermocouple leads. The
minimum error for the range of test conditions was -9 F. For the
test conditions where the air temperature within the enclcsure
was greater than the bleed duct temperature the thermocouple
created a positive error in the reading.

INTRODUCTION
A set of tests had been completed in order to determine the
ignition conditions for a variety of fuels and lubricants in a
simulated engine compartment. As part of the post data analysis
for the test, an analysis of the "fin effect" error of the
thermocouple leads has been completed. When a thermocouple is
used to measure the surface temperature of an object, it will
inherently change the surface temperature at the point of
measurement, due to its presence. This error is due to the
convective and radiative heat losses of the thermocouple leads to
the surrounding air flow and walls and hence, its action as a fin
attached to the surface.
A wide range of air temperature and flow velocities were present
during

B-3
IM
the tests and it is required that an error correction be
estimated for the surface temperature measurement due to this fin
effect error. The bleed air duct is made of Inconel 625 and the
thermocouples were a type K (chromel/alumel), 24 AWG and
insulated with a fiberglass braid.

MODEL DESCRIPTION
The thermal mathi model used to determine the thermocouple error
consisted of a steady-state thermal network model of a 12 inch
section of the bleed duct and one of the thermocouples attached
to it. The thermocouple lead length was set at about 6 inches,
which was the length exposed to the air flow in the enclosure.
The solution technique for the thermal network is based on the
presentation in [1]. The thermal math model consisted of 36
nodes for the bleed air duct, 11 nodes for the thermocouple and 3
boundary nodes, see Figure 1. One boundary node represented the
temperature of the air flow through the bleed air duct, while the
other two represented the air flow through the enclosure and the
enclosure surface temperature. These boundary values (see Tables
1 and 2 in results) were taken from the test conditions recorded
during the test.
The correlations used for the forced convection air flow over the
bleed air duct and the air flow inside the bleed air duct can be
found in [2]. For flow over the duct and leads the correlation
used was for cross flow ove a •,• -der
3 Ren
Nu =cPrl/
Where,

Nu - The Nusselt number

Re - The Reynolds number, VD/v
Pr - The Prandtl number, v/k
V - Fluid flow velocity
D - Characteristic length, outside diameter of the cylinder
k - Thermal conductivity
v - kXinematic viscosity

The constants are given by the range of the Reynolds number,

Re C n

p 0.4 to 4
4 to 40
40 to 4,000
4,000 to 40,000
40,000 to 400,000
0.989
0.911
0.683
0.193
0.0266
0.330
0.385
0.466
0.618
0.805
Then the heat transfer coefficient is given by,
h Nu*k/D

K B-4
For the forced convection on the inside of the bleed air duct the
following correlation was used,
8 Pr1/3
Nu = (0.023)Re0.
Where the characteristic length is the inside diameter of the
duct. The heat transfer coefficient is determined as shown
above.

The radiation heat transfer to the enclosure was assumed to be

from a gray body, the outside of the duct and the thermocouple
leads, to a black enclosure. The enclosure was reported to be
covered with a coating of soot from the testing and hence its
emissivity should be close to 1.0. The emissivity of the bleed
air duct and thermocouple leads was assumed to be 0.80 for this
analysis.
The thermal properties for the model were fixed at,

Thermal Conductivity
(Btu/hr-ft-F)
Inconel 625 10.0
Chromel 10.0
Alumel 10.0
Fiberglass Tape G. 1
Note that the chromel and alumel are estimated values at this
time. It is expected that the thermal conductivity for these two
materials will be very close to the values used. The air
velocity in the bleed duct was calculated using a mass flow rate
of 1.0 lbm/sec at a pressure of 130 psia and the given
temperature of the test condition.

RESULTS
Results are presented for a total of 3 basic sets of conditions
that were run in the test section. These sets of conditions were
chosen to represent the minimum and maximum errors expected for
the test configurations that -ere r•n. Table 1 shows the results
for the lowest enclosure air temperature, over a range of flow
velocities and bleed air duct temperatures. The maximum error
occurs for the highest difference between bleed duct temperature
and the enclosure air temperature, as would be expected. Vote
that the enclosure wall temperature is expected to be lower than
the air temperature, due to losses to the surrounding
environment.
Table 2 contains the results for the elevated enclosure air
temperature cases. This also resulted in the ernlosure wall
temperatures to be higher. Only the extreme enclosure air
velocities (minimum and maximum) were run for these cases. As
would be expected the errors in the thermocouple readings are
lower

B-5
 than the cases for which the temperature difference between the
bleed duct and enclosure are higher. It is interesting to note
that for case 3a, in which the air teaperature in the enclosure
exceeds the bleed duct air temperature, that the effect of the
thermocouple was to locally heat the duct. This condition gave
the lowest predicted error in the thermocouple reading.

Prepared By: _ _ _ _

655-0790407

Approved By:
F.M. Rafch ie~

Boeing Advanced Systems

Propulsion Technology

Attachments:

1.) Figure B-i. Modeled Section of Bleed Duct with Thermocouples

2.) Table B-1. Uncertainty Analysis for Ambient Air Temperature Tests

3.) Table B-2. Uncertainty Analysis for Elevated Air Temverature Tests

B-6
 Attachment 1

-
C

00

Q(3
i II-I

0( I (10

I-
o
I
0 (D r

-2j

B-7
 Attachment 2

Table B-1. UncertainlyAnalysis for Ambient Air Temperature Tests

Case Number
la lb Ic id
Air
Temperature (F) 120.0 120.0 120.0 120.0

Enclosure Wall
Temperature (F) 110.0 110.0 110.0 110.0

Enclosure Air
Velocity (ft/sec) 0.25 2.0 11.0 20.0

Enclosure
Pressure (psia) 14.7 14.7 14.7 14.7

Bleed Duct
Temperature (F) 800.0 500.0 1350.0 1600.0

Bleed Duct Air

Velocity (ft/sec) 322.0 245.0 462.0 526.0

Bleed Duct
Pressure (psia) 130.0 130.0 130.0 130.0

Duct Surface
Temperature (F) 704.0 446.0 1010.0 1119.0
Thermocouple
Temperature (F) 686.0 434.0 985.0 1092.0

Calculated Error
(Ttc - Tsurf) (F) -18.0 -12.0 -25.0 -27.0

B-8
 Attachmen( 3

Table B-2. UncertaintyAnalysis for Elevated Air Temperature Tests

Case Number
2a 2b 3a 3b

Air
Temperature (F) 300.0 300.0 600.0 600.0
Enclosure Wall
Temperature (F) 200.0 200.0 350.0 350.0
Enclosure Air
Velocity (ft/sec) 0.25 20.0 0.25 20.0
Enclosure
Pressure (psia) 14.7 14.7 14.7 14.7
Bleed Duct
Temperature (F) 600.0 1350.0 500.0 1350.0
Bleed Duct Lir
Velocity (ft/sec) 271.0 462.0 245.0 462.0

Bleed Duct
Pressure (psia) 130.0 130.0 130.0 130.0

Duct Surface
Temperature (F) 551.0 1002.0 485.0 1053.0
Thermocouple
Temperature (F) 542.0 982.0 487.0 1039.0
Calculated Error
(Ttc - Tsurf) (F) -9.0 -20.0 2.0 -14.0

B-9
3.) Extrapolation of Error Analysis to AENFTS Air
Temperature Thermocouples

for the following te't conditions

AIRFLOW

T = 600OF T/C Bead

V = 2 FT/SEC
4 1/2"
P = 14.4 PSIA

'-1/4" CRES wall

Various wall temperatures from 100OF to 500OF

If we assume that the conduction loss down the lead is negligable, then we
have:
Radiation Heat Transer = Convective Heat Transfer
or
4
A. . IT T _.4\ hA T - T
"c'1C '1c N'tc wail - "tc \ air - tc

where it has been assumed that the shape factor between the bead and the
wall is equal to 1.0, and:
c7" = Stefan-Boltzmann Constant
0.1713 X 10-8 Btu/hr-ft
2 - OR4

•tc = emissivity of the thermocouple

h = heat transfer coefficient between the

thermocouple bead and the airflow

B-i0
 this leads to:
h
4 4
(Ttc - Twall ) = _ (Tair - Ttc )
v,- et c

Then, for a given Tair and Twal!, this must be Solved for Ttc.

and:
etc = 0.8, which is the usual assumption for an unknown
material

Nu k
h = Nu = Nusselt Number
D k = thermal conductivity
D = characteristic length

and N.. = f37a 0.6 for a ga5 AL, VVL a sphere (the Tic bead)

and U D
Re =

then U = 1 ft/sec

D = 0.125" = 0.0104'

v = 40.8 X 10-5 ft 2 /sec (440 0 F)

0
k = 0.02333 Btu/hr-ft-°F (440 F)

(1 ft/sec)(0.0104 ft)
Re = = 25.5
2
40.80 X 10-5 ft /sec

Nu = 0.27 (25.5)0.6 = 2.6

( 2.6 )( 0.02333 Btu/hr-ft-°F )

h = 5.8 Btn;/r-ft-OF

( 0.0104 ft )
B-11

t... i i 'i It - I
Hence for 600°F air at 2 ft/sec and 14.4 psia:

Thermocouple indicated temperature and error for a given wall temperature

are tabulated below based on the preceeding approach and assumptions:

Twall Ttc T/C

error
(OF) (OF) (OF)

100 457 143

200 469 131
300 488 112
400 515 85
500 552 48

...
ilar analysis for "ov-r aki at 2 fL/sec and 14.4 psia and a wall
temperature of 423 0 F:

Twall Ttc T/C

error
(OF) (OF) (OF)

423 457 23

Finally, when it was established that 83282 was igniting in air with an
indicated temperature of 510 0 F, v'th a wall temperature observed to be
425 0 F, extrapolation between these two tables allowed estimation of the air
temperature measurement error in this case to be 60 0 F.

B-12
 APPENDIX C. PERTINENT AIRCRAFT FLUID PROPERTIES

Properties are needed for the five aircraft fluids of interest kJP-4, JP-8,
5606, 83282 and 7808) te carry out calculations on the basis of whica one can
interpret the test results. The properties of interest include:

o basic properties such as molecular weight and vapor pressures

o thermophysica! properties such is surface tension, viscusity, and

thermal conductivity

o kinetic properties for the ignition reactions

o boiling heat transfer coefficients

Most of these properties (except for the heat transfer coefficients) were
readily available for JP-4 and JP-8. However, they were not readily available
for the heavier fluids (5606, 83282 and 7808) because:

- each of the above fluid designations represent a family of fluids

that satisfy military speqifications. For example, 5606 consists
of a heavy kerosene as a base stock. thickened with additives.
Also, properties will vary for different petroleum base stocks.

- these properties are not specified in military specs (unlike

viscosity, lubricity, thermal stability, etc.)

- little modeli..g effort has been devoted to these fluids in the past
(unlike the case of the JP fuels)

C-i
 To obtain even rough values or estimates for such properties, three approaches
vere followed:

o a manual search of the MIT and the AFWAL library including the MIL
standards

o a computerized search of 4 databases: NTIS, Compendex, NASA and

Chem Abstracts.

o telephone calls and selected meetings with key researches in the

field and a supplier of these fluids.

A summary of the found thermophysical properties is given in Table C-i.

Kinetic data for the ignition reaction were found in the literature for JP-4
and JP-8 (Ref. C-i), but not for the other fluids. They are given in Table C-
1. We recomnend that such basic data be measure in future work.

Vapor Pressure Data

We collected vapor pressure (Pvap) data at various saturation temperatures

(Tsat) from a variety of sources. The data are shown in an Arhenlus plot in
Figure C-1. Note that the relationship between log Pvap and the inverse
a bs o l u t e t e mp e r a t u r e T •.- i s l in e a r P x c e p t a t h fg h q -• . . . . . . . . . .. . . . . ..'

I
5606. This is not surprising for these multicomponent fluids.

We used the Clayperon equation to calculate the latent heat of vaporization

(Hfg) froin the vapor pressure data:
S- Tsat2 drvap
1
H fg =- p

M.W. I Pvap dTsat

where,
R
P Universal gas constant
M.W.= Molecul.ar weight

C-2
 TablUe C-1. Fluid Properties Used In Caculatlons

PROPERTIES JP-4 JP-8 5606 83282 7808

Molecula Weight, kg/kmole 125 166 266 400 425

Avg. Latent Heat*, HfS, kJ/kg 211 229 204 105 357
Effective Saturation Temperature*, OF
at 20 psia 236 405 496 865 570
at 14.4 psia 209 376 470 805 559
at 10 psia 182 346 444 744 548
at 5.5 psia 141 302 403 656 529

LIQUIDS AT 20 0 C (from vario-s sources)

Specific Gravity 0.76 0.81 0.88 0.85 0.95
Absolute Viscosity, g/cm's 0.0081 n.013 1.15 4.57 2.30
Kinematic Viscosity, mm-sq/s 0.95 1.65 131 536 242
Surface Tension, dyne/cm 21.- 23.3 31.5 30 30
Specific Heat, kJ/kgoK 2.06 1.95 2.19 2.19 2.19
Thermal Conductivity, W/m°K 0.115 0.11) 0.!i[ 0.115 0.115
Autoignition Temperature, OF 447-475 46C 435 650 735

SATURATED VAPORS (taken as air at various temperatures)

Absolute viscosity#, gm/cm-s 2.4f'-04 2.4E-04 2.4E-04 2.4E-04 2.4E-04
Thermai Cnnir-iti,-4t , To /O,/O I
.1 A3
003 O 3 0.03 0.03

KINETICS OF OVERALL IGNITION REACTIONS (from Reference C-1)

JP-4 JP-8

Pre-exponential Fact, ms/atm2 1.17F-09 1.68E-08

Activation Energy, kcal/mole°K 43.1
Order of Reaction 2

* estimated based on vapor pressurie ,iata

# = at 100 0 C

C-3
 CID

_ _ _ ___ _ __ _ _ __ _ _ __ __ -

N
00
0

00

Fo

co

Ld O L

I-n

NON W

- 13
DdM~~~Y ld '3fIS-. 0 "
We calculated the slope of the log Pvap-Inverse Absolute Tsat
relationship (Fig. C-1) at various temperatures and averaged the
values (for simplicity). Also, we used reasonable average
molecular weights for these multicomponent fluids as given in
Table C-1. Thus, we calculated an estimate for an average latent
heat of each fluid as given in Table C-1.

It should be noted that our data collection/analysis was very

rough. For example, we found only two data points for the vapor
pressure of 7808. Although, the results arc adequate ftr our
proposes, we recommend that effort be devoted in the future to
further collect and document fluid property data.

REFERENCES

C-i. Spadaccini, L. j. and J. A. TeVelde, "Autoignizion

Characteristics of Aircraft-type Fluids", NASA CR-159886,
June 1980.

C-5
 APPENDIX D. SPRAY ANALYSIS

In the spray tests, liquid flows at high velocity from the nozzle- The liquid
breaks up into droplets that decelerate, heat up and evaporate as they
approach the duct. The vapors, mix with air, heat up further and ignition
occurs when the appropriate temperature-composition-time are attained.

In this Appendix, we present a simplified classical analysis for these key

processes in spray ignition, adapted to the conditions of the simple duct
tests in this study. Our objective is to determine characteristics times and
relative importance of various processes that can help us interpret the
results (and not to develop any extensive mathematical models). For
simplicity, each key process is treated separately. The results are then
combined as appropriate.

We discuss below these key processes along with their governing equations.
The key processes are:

o atomization

o droplet dynamics

o droplet heating/evaporation in duct boundary layer

o chemical kinetics

o ignition criterion

o drcplet impingement and heating at duct surface

Atomization

Consider the atomization of a fluid through a pressurized nozzle into stagnant

air. As the liquid exits the nozzle, it acquires a high velocity. The liquid

D-1
 I!
surface becomes unstable and breaks tip into ligaments and eventually into
droplets. Fron. dimensional reasoning, one expects the resulting average
droplet diameter (da) to be given by:

da/Dn, . function (Ren r Wen)

where, Dn = nozzle diameter

Ren - Reynolds number at nozzle exit
Ven = Weber number at nozzle exit

Correlations have been proposed by various investigators for pressure

atomization (Refs. D-1 and D-2) such as:
2 8 . j1-1/8 -0-1/5
d 3 2 - 150,000 Dn1/ . AP-3/ v 1/4 •11/4

where: d3 2 . Volume to surface mean diameter, micron

2
P Pressure across nozzle, dyne/cm
1 Liquid surface tension, dyne/cm
3
Ai1 Liquid density, g/cm
'0 - Liquid absolute viscosity, g/cm's
3
a= Air density, g/cm

Dolt Dynamics

Consider a droplet of initial diameter (di) propelled with an initial velocity

of (VO) into a moving vas stream with a gas (Vcty
). The gas stream is
the ventilation air. It exerts a drag force on the droplet that decelerates
it.

The conditions of spray injection, the droplet Reynolds numbers (Red) is of

the order of 10 to 1000. For this range of Re, the drag coefficient (Cd)) of
the droplet was obtained (within +30%) by curve fitting the data of McAdams
with the following expression:
Cd = 13 / (Red)1/2

where: Red Reynolds number based on di and the droplet local velocity
relative to the gas (Vi - Vg)
D-2
A force balance on the droplet yields the instantaneous droplet velocity (V)

and its trajectory (x) as a function of time (t):

2
V/Vi = 1/(1 + 0.5,6iff t)
where = 9.8 (Pg -"Ug)l/ / (ol " d3/2)

- velocity decay parameter

', g = densities of liquid and gas, respectively

g = gas viscosity at the ventilation air temperature

and x/(xd-x) = 0.5Vi " "t for t < time to impact

where xd = O.5 fV /p

= deceleration distance required for V to decay to Vg

If xd is greater than the spacing between the nozzle and the duct, the
tiajec:too-y equatlon above yields directly the time to impact, with the

duct; otherwise, first V decays to Vg according to the above equations,

and then, the droplet moves with ,:he gas stream at Vg until impact with
the duct.

The transit time near the duct (tt) is obtained by dividing a heating
length by the droplet velocity at the duct leading edge. For the simple
duct experiments, the heating length is assumed to be the duct diameter
(1.5 inches). For the high realism tests, the heating length is taken
to be 22.5 inches (roughly the projected length oe the hot duct along
the spray direction).

Droplet Heating/Evaporation in a Hot Stream

The results of the above analysis are used as initial conditions for the
droplet heating/evaporation analysis.

D-3
Consider a spherical droplet of fuel at an initial temperature (To), engulfed
in a hot air stream at a higher temperature (Th). The hot air is in the
boundary layer of the duct. The relative velocity between droplet and air is
governed by the droplet initial velocity and by drag from the much slower
ventilation air.

The heat transfer from the air to the droplet Is given by the Nusselt number
(Nu) as follows:

1
+ 0.6 (Red) 1 /
2 /3
Nu =2 - (Pr)

where Nu h • d/k
d instantaneous droplet diameter
h heat transfer coefficient
k thermal conductivity of air
Pr Prandtl number . 0.7

and gas properties are taken at a mean temperature in the boundary layer.

"The vapor mass flux (i") from the droplet surface due to heat/mass transfer is
given by:

;" - Nu (k/d Cg) In (B + 1)

where Cg - specific heat of air

B - mass transfer number = cg'(Th-Tb)/(Hfgcl ('b-To))

Tb = boiling point of liquid at ambient pressure

Hfg = latent heat of vaporization at Tb

cI = specific heat of liquid

D-4
A mass balance on the droplrt yicl~dF 'he droplet evraporation time (te) w.iirh
is given by one of two equations depending on the flow field;

For nearly stagnant flow, Nu = 2 and:

2 8
te = di P, . Cg/( kg.ln (B+1))

where di = initial droplet diameter

for forced convection flow,

te
te = 0.6d9
0.56 d 2
• qe cg/[kg " (Red) 1
/ 2
. (Pr)1/ 2
in (B+1)1

Kinetics of Ignition

Following Reference D-3, a one-step overall second order reaction was used to
model the kinetic portion of the ignition delay time:

tc = f - exp (E/RT) / pn

where: tc = induction time, ms

f = pre-exponential factor, ms/atm2

E = activation energy, kcal/mole°K

P = nrpresire atm
R = universal gas constant
T = absolute temperature,°K
n = order of reaction

Ignition Criterion

Ignition is assumed to occur when the sum of the evaporation and chemical
times is less than the transit time near the hot duct. The temperature is
increased parametrically until this criteria is satisfied. This was done only
for JP-4 and JP-8 for which we found kinetic data in the literature.
D-5
Droplet Impingement Against a Hot Surface

From the droplet dynamic analysis presented above, one predicts that small
droplets decelerate befoie reaching the duct; are entrained by the
ventilation air; and may flow around the duct in its hot boundary layer. On
the other hand, large droplets %ill decelerate more slowly and impinge on the
duct at high velocity and large Weber number. The dynamic of such an
Impingement and its effect on heat transfer from the surface to the droplet is
a complicated process that is not completely understood.

Accordingly, from the literature we identified the types of reported behaviors

and governing parameters that may be used in Interpreting our test data. Key
experimental results found for water are:

1. Effect of Weber Number (We): Figures D-1 and D-2 (from Ref. D-4) show the
dynamic behavior as a function of time of a 2.3 mm water drop Impacting a hot
polished gold surface (at 40 00 pt Nners
-
Neie of 15 and 40, .......

velocity = 0.7 and 4 m/s) respectively. (While these conditions are different
from those of our tests, this study presented photographs that are quite
revealing concerning the key processes.) Note that:

o at low We, the droplet spreads on the surface, then rebounds from
the surface "intact" (at 15ms)

o at high We, it spreads, and disintegrates into a large number of

smaller droplets as a fine spray (in 8.8 ms). Subsequent
vaporization may now occur much more rapidly because of the smaller
diameters of the droplets.

D-6
A6!

6 x5 ' 9. K 10 '

('Curse of the impact of a waLer drop on a hot polished gold surface (surfac" tPmperature 4(0)"C, drop diameter
2"3 mm, We.. 15). Time is measured from the moment of initial contact.

Figure /-1. Effect of Weber Numver on Boiling

D-7
 x _
05! x ss0i
S.-S

?6 x0'in X C s 2.a XC
;Q

-04 ' 01A

2 3 mim, Wt:,, -
COtTSC of (lie impact of a waler drop oi a lhot pohs.hed gold surf'ace SIsurface eicnpcrature 44X)C, drop diameter
mewred from the moymn i of in i ail eon tact.
184). lime hinca

Figure D-2. Effect of Weber Number on DoiMQn

D-8
2. Effect of surtace temperature: Figure D-3 shows the water droplet
impacting the same surface at 4 surface temperatures, and for 4 times (from
impact) at each temperature. The droplet conditions were: diameter = 2.17
mm, velocity = 1.25 m/s, 300 angle normal to the surface, We = 41, T - 20°C.
Note that the effect of increasing surface temperature (Tw) is highly
nonlinear:

o up to about Tw 0 2000C (Tw - Tsat = 100 0 C), the contact between

liquid and vapor is such that a great deal of the vapor produced at
the bottom of the drop is trapped by the liquid and breaks through
during the collision, Thus, a spray is produced as shown for T =

1780C.

0 at about Tw = 2200C (Tw - Tsat = 1200C), the droplet behaves vary

calmly with no spray formation.

o at Tw = 2500C and above (Tw - Tsat > 150%C), vapors are formed
explosively inside the drop. When they rise and break out through
the liquid, they produce a very fine spray.

c when Tw is increased to about 4000C (Tw - Tsat - 3000C), a vapor

film is formed underneath the drop. No vapor bubbles were seen in
•huerp. I ,i uvii Is said to have reacne he fuly spheroidal
(or Leidenfrost) state.

Clearly, the boiling behavior depends on where vapor bubbles are formed and on
whether or not the±y are trapped by the liquid film. We recommend that this
subject be investigated in future work for the conditions of interest to hot
surface ignition.

D-9
 b.4

f,. 168q"

Cokurse oi' OIC inolpt of' a witer drop of' 20 C Upon a hot. poliShCd gold Nurlatx (dro.p diameter
2-17 nmm. r,, 1-~25 rnisec, t-30 ,'so that We,~,,- 41). Thc pictures aie in all four cases taken at rc.
spectively 0-63, 1-88, 4-84, arid 9-27 x 10 3sec after initial contact.

Figure D-3. Effect of Weber Number on Boiling of Heated Fluid

D-180
REFERENCES:

D-1. Elkotb, H. M., "Fuel Atomization for Spray Modelling", Progress in

Energy and Combustion Science, 1982, Vol. 8, pp. 61-91.

D-2. Dorman, British J. of Applied Physics, Vol. 3, June 52, p. 189.

D-3. Spadaccini, L. J. and J. A TeVelde, "Autoignitlon Characteristics

of Aircraft-type Fluids", NASA CR-159886, June 1980.

D-4 Wachters, L. H. J. and N. A. J. Westerling, "The heat transfer from

a hot wall to impinging water drops in the spheioidal state",

Chemical Engineering Science, 1966, Vol. 21, pp, 1047-1056.

D-11
APPENDIX E. DETERMINATION OF BOILING REGIMES FOR HOT SURFACE IG•!TICN
TESTS

In the stream tests, the fluid is injected by a drip tube onto a horizontal
hot duct. The fluid jet impacts the duct and spreads as a thin film outwardly
from the impact point in all radial directions. This liquid flow produces e
stagnation-like flow field. After a very short distance, the spreading tluid
breaks up into rivulets. This distance depends on a number of variables and
is of the order of 0.3 inch based on photographs obtained in -1 similar
experiment. The rivulets continue to flow ovvr the duct and further break up
into ligaments and droplets (See Plate E-'.)-

A key unknown in this study is whether boiling occurs in the nucleate or film
boiling regime. Depending on the regime, large differences are expected in
the rates of fluid evaporation and the attained temperatures--which in turn
would affect the ignition results in this study. Identification of the
boiling regime was deemed needed to help interpret the results.

We searched the literature for data on boiling behaviors of aircraft fluids of

interest. We found no directly pertinent data under either the above
configuration or the simpler case of a boiling liquid pool. We set out to
determine the boiling regimes for the test cor.ditions of this study using
standard correlations from the literature. The results are presented in the
Appendix.

Boiling Heat Transfer Correlations

In boiling, the driving force for heat transfer is the excess wall temperature
(Tw) cver the saturation temperature (Tsat) of the liquid. Thus. the heat
transfer coefficient (h) in boiling is usually defined as

h (q/A) / (Tw - Tsat)

E-1
 Plat
606hydrull
E 1. F~dStnamedont Ho

f.l.Test.Ri
i~............ .....

74t,_I
Photo
aquiredduringexperient perormed o Hortanto tAtli aet3(eeilTs

Methd
711,Sandad
ethd N.
6~3, 5 Jn, 6) ist~led t VAFD

E-2
Correlations for h are given ini the literature (Ref. E-l) for static, pool
boiling of single-component pure substances. For simplicity, we applied these
correlations to our flowing, multi-component fluids using the collected
thermophysical properties and estimated "average" thermodynamic properties
from Appendix C. Also, we simplified these correlations to the case where the
ratio of vapor density (p ) to liquid density (pl) is much less than unity.
Such a case applies to the conditions of this study.

These simplified correlations are presented below:

Nucleate boiling

In this regime, bubbles form (nucleation), grow and move away from the
surface. Their motion stirs the liquid near the hot surface and produces very
high heat transfer coefficients (hn). The dependence of hn on the excess
temperature is give by:

hn Ii

n
"hn,max
-h t(TV
~(TV
-

-
Tsat)
sta
Tsat~maxJ
12

Unfortunately, Reference E-1 did not define (TV - Tsat)max. Therefore, we

assume (based on the shape of boiling curves for a variety of fluids) that:

( T w - Tsat)max 0.3 (T- Tsat)min

where (Tw - Tsat)min is defined in the next section.

hn,max was obtained from th,ý equation given in Reference E-1 for the maximum
(or burnout) heat flux in the nucleate boiling regime (q/A)n,max as:

H
Pl 0.6
(q/A)n,max . 143

E-3
' Hfg •

06
-
Pv
Filz Boiling Regime

In film boiling, the vapor generated at the liquid-vapor interface produces a

vapor film that separates the liquid from the hot surface. Heat is
transferred by conduction and radiation across the vapor film. For the case
of liquid droplets (as in a spray), the droplets will dance around the
surface, and the phenomenon is referred to as "Leidenfrost".

The film boiling regime starts at a specific value of excess wall temperature
over saturation temperature given as:
(gP1/3 11fg-P / /
(TV - Tsat)min 0.127
kvff

2
where: g M acceleration of gravity, ft/hr
kvf = thermal conductivity of vapor in the film
between the wall and liquid, Btu/hr ft'°F
- surface tension of liquid, lbf/ft
Avf viscosity of the vapor in the film between
wall and liquid, Ibm/hr-ft

The heat transfer coefficient (hf) in this regime is given by:

k vf 3 .' Hf • P,, "Pl 1/4

hf- 0.425(g)11 4
(Tw - T sa t) v vi
*AP j

The boiling heat traitsfer coefficients and heat fluxes are plotted as a
function of (Tw - Tsat) at ambient pressure (14.4 psia in the nacelle)
according the the above correlation in Figure E-1 and E-2, respectively.
Since the transition between these two regimes is unstable, we simply
illustrate it by connecting (for each fluid) the point of maximum heat flux to
that of minimum heat flux by a dashed straight curve (on the log-log plot).
Note that the transition from nucleate to film regimes occurs at a different
excess temperature for each fluid.
E-4
 OD

ucc
Ix V

V....
cso

4.

ILI

o 0 0C0
0t0

o 0

00
0

~s*±N:31014-4OD NRASNVýI. JLV3H

 OD
00~
w
-0
o- i3 tA

op CK)

N LO

uj.5.D

OD

F 14)Lo

00'~

00

ZUWj4/nl~e oxfL-j jvoH

E-6
Applicable Boiling Regime for Hot S,,rface Ignition Tests

It should be noted that the excess wall temperatures at which boiling changes
from a nucleate to a film regime are functions of pressure according to the
above equations. This is illustrated in Figs. E-3 to E-7 which show these
excess temperatures as a function of pressure for the five fluids of interest.
These plots delineate (by straight lines) the conditions at which the nucleate
regime ends and the film regime begins.

In addition, In Figs. E-3 to E-7, we indicate as data points, the minimum

excess temperature at which ignition was observed in this test program, i.e.,
(MHSIT - Tsat). Here, we assume that MESIT is "similar" to the wall
temperature (Twall) measured in boiling-type studies. (This is not exact
since the duct temperature may drop during the current test while Twall can be
maintained constant in boiling-type studies.) The data are for stream
injection and for all tested ventilation air conditions. The pressure
dpnndPn~Pno nf hnth HSTQT anA T arc i1cIudGd 4,- tLe data.• ote tl-t
ýsar AL

o all the JP-4 and JP-8 tests fall well into the film boiling regime.
For these fluids, the required ignition temperature is high and the
saturation temperature is low, yielding a large excess temperature.

o the 83282 tests may fall in either regime and are mostly in the
nucleate regime. For this fluid, the required ignition temperature
is low and the saturation temperature is high, yielding a small
excess temperature. (It may even go negative, i.e., for this heavy
fluid, the lighter ends might ignite while the heavier ends,
remaining in the fluid, wet the duct surface.)

E-7
 LU

!2

LJ~

CL o

(00

E-83
 L-C"

C-,
z 0
00

F--

000

I "I

C0.

LLLLJ

L- L~
<

0 0

C) 0 0 0 0 0
C0 0 0
a) 00 0 0D f fn N

]J')OsI )U3AO IISHIA4 SS3DX3

E-9
 U -

c~cc

0 0

00 M
zO <
z
'
<

LUI

LO

zz

J. lusl U]IAO licHPN SS1gZXg

E-.10
 F--

LUU

u J
O
:D

~0

E--
 LUO

000

C)C

00.
 o the 5606 and 7808 tests are intermediate between the above two
behaviors, falling mainly in the film regime.

Overall Heat Transfer Coefficients

Figure E-1 indicates that the heat transfer coefficients are of the order of:

2
o 100 to 3000 Btu/hr'ft '°F for the nucleate boiling regime

o 20 to 40 Btu/hr'ft 2 .OF the film boiling regime

Furthermore, as the excess temperature increases, h will Increase in the

nucleate boiling regime while it will decrease in the film boiling regime.
However, the heat flux will increase in either case as shown in Figure E-2.

-lo, n.t....at the convective heat transfer coefficient from the hot air
(inside the duct) to the duct wall (at 1 lb/sec) is about 200 Btu/hr'ft 2 "oF.
Since this convective heat transfer step is in series with the boiling step,
the slower of the two will limit the overall heat transfer. In the case at
hand, the overall heat transfer will be limited by the internal resistance on
the air side for the case of nucleate boiling (for 83282); and by the external
resistance on the vapor side for film boiling (for the other fluids).

Temperature of Bulk Fluid

As described previously, when the fluid stream impacts the duct, it spreads as
a thin film outwardly from the impact point in all radial directions. After a
very short distance, the spreading fluid breaks up into rivulets. This
distance depends on a number of variables and is of the order of 0.3 inch
based on photographs obtained in a similar experiment. It is reasonable to
rssume that most of the direct heat transfer (from duct to film) occurs in
this small area (with a 0.3 inch radius).

3|
From an energy balance en this flowing stream, one predicts that bulk fluid
temperature will not reach saturation near the duct surface. Even with the
maximum possible overall heat transfer coefficient (200 Btu/hr'ft 2 "OF), the
bulk temperature rises only to 160°F due to the high fluid flow rate and the
small contact area between the fluid and duct. In reality, a much lower h and
temperature will be obtained because of the external thermal resistance in the
film regime.

Thus, the bulk liquid temperature is much lower than its saturation
temperature. In other words, boiling occurs in a subcooled regime, i.e.,
vapors are formed near the hot duct surface and rise through a cooler liquid
where they condense releasing their latent heat.

Physical Model of Hot Surface Boiling/Ignition

Based on the discussion above, one can develop the following picture for the
case at hand, as ill'istrated in Fig. E-8. As the liquid flows over the hot
plate, its temperature rises but the fluid remains subcooled. Vapors are
formed only very near the surface where a very thin layer of fluid reaches the
saturation temperature. The vapors rise through and condense in the liquid.
The applicable regimes are mainly subcooled film boiling for all the fluids of
interest except for 83202 where subcooled nucleate boiling also occurs.

At the edge of the spreading liquid film, the produced vapors exit the
duct/liquid interface and are available for mixing with air and for ignition.
A key parameter is the exit temperature of these vapors. This exit
temperature can be modeled based on the physical description described above.
Such a model is recommended for future work. As a rough approximation, we
estimate that this exit temperature will be the arithmetic mean value between
the temperature of the hot duct aid the saturation temperature of the fluid.
We suspect that the exit temperature may correlate with the autoignition
temperature (AIT) of the fluids. Since we had no AIT data at various
pressures, we could not test this correlation. Such work is recommended in
the future.

E-14
 0

L I QU I D .• _• ,•/O F S T R E AM
TEMPERATURE- .

So o BUBBLES

•'79• F. P7/"/
/ / ""/ //zI ' / / ,- TEMPERATUR•E AXI S
TINITIAL TSATURATION ATWALL DITNEAOGUT
z

(a) SUBCOOLED NUCLEATE BOILING -

LIQUID SURFACE

SLM
TEN PERATUiE-\
6 IO
BBUBLES

OF STREAM
LIQUID

i0 -

..--
. • .. • . . •• -" --.,,-- -,, LI QUI D/VAPOR
"-INTERFACE
TEMPERATURE
AXIS

VAPOR FILM TINITIAL TSATURATION TWALL DISTANCE ALONG DUCT

(b) SUBCOOLED FILM BOILING

NOTEV THlE TEMPERATURE PROFILE IS ILLUSTRATED IN THESE DIAGRAMS (a
a b) FOR A GIVEN LOCATION
ALONG THiE DUCT

Figure E-8. Boiling Regimes of Interest

E- 15
Rwcomended Future Work

Ybe discussions in Appendices D and E highlight the need for a basic sdvy 1C
elucidate the key processes involved in the ignition of aircraft fluids when
im contact with a hot surface. Such a study may be carried out under the
simplest conditions that are amenable to control and measurements. For
example, experiments may consist of placing small droplets on a hot surface
with a hypodermic needle. A flat surface would be used with a small
depression at its center to stabilize the droplets. The test variables would
include:

"o fuel type, droplet size and velocity

"o single v.s. multiple droplets

"o plate temperature from 100 0 C to 400 0 C over fluid boiling points

"o selected plate materials such as stainless steel and titanium

Heasurements/observations (from video records) would inzlude:

o plate temperature as a function of time

o droplet spreading behavior and size as function of time

o location and time of ignition

Similar experiments can be conducted simulating a small stream instead of a

spray, The results t-om such experiments would permit estimation of
evaporation rates and heat transfer coefficients and identification of boiling

regime under the range of conditions of interest. One can then use these
results to interpret test data under more practicality conditions.

E-16
Reference

E-1 Rohsenow, W. M. and H. Y. Choi, "Heat, Mass and Momentum Transfer",

Prentice-Hall, Edglewood Cliffs, N. J. 1961.

E-17

U.S. GOVERNENNT PRINTIG OFFICE: 1989-648-056