A. ACCIDENT: DCA-96-MA-070
Location: East Moriches, New York
Date: July 17, 1996
Time: 2031 Eastern Daylight Time
Airplane: Boeing 747-131, N93119


B. GROUP IDENTIFICATION
The group met at JFK Airport on July 11, 1997, through July 20, 1997. The
following group members participated in the investigation.

Chairman: Daniel R. Bower, Ph.D.
NTSB

Test Director: Robert Benzon
NTSB

Member: Michael Collins
Federal Aviation Administration

Member: Roland Johnson
Boeing Commercial Airplane Company

Member: Steve Green
Air Line Pilots Association

Member: Terry Stacey
Trans World Airlines

Additional participants were involved for the implementation of
the Flight Test. These participants were:

Member: Kevin Rickard
Evergreen Airlines

Member: Steven A. Bongardt
FBI/DOD Coordinator

Member: Robert L. Swaim
NTSB Systems Group Chairman

Member: Merritt Birky, Ph.D.
NTSB Fire and Explosion Group Chairman

FlightCrew:

Captain: Dale M. Ranz
Boeing

First Officer: Jamie C. Loesch
Boeing

Flight Engineer: George E. Kegebein
Boeing


C. SUMMARY
On July 17, 1996, at 2031 EDT, a Boeing 747-131, N93119, crashed into the
Atlantic Ocean, about 8 miles south of East Moriches, New York, after taking off from
John F. Kennedy International Airport (JFK). The airplane was being operated on an
instrument flight rules (IFR) flight plan under the provisions of Title 14, Code of Federal
Regulation (CFR), Part 121, on a regularly scheduled flight to Charles De Gaulle
International Airport (CDG), Paris, France, as Trans World Airlines (TWA) Flight 800.
The airplane was destroyed by explosion, fire, and impact forces with the ocean. All 230
people aboard were killed.

In support of the investigation into the TWA Flight 800 accident, a series of nine
flight tests were performed to obtain time/temperature histories within a 747-100 series
airplane. The tests followed specific preflight, taxi, takeoff, and climb flight profiles.
Data was collected from center wing tank (CWT) surface temperatures, CWT air
temperatures, and pressure within the several bays of the CWT and the wing tip surge
tanks. Also obtained were the air temperature time history of the environmental control
system (ECS) air-conditioning pack bay beneath the CWT, air conditioning pack
component surface temperatures, and vibration measurements, CWT ullage vapor
samples, and some electromagnetic interference data.

Data was acquired with selected combinations of air-conditioning packs in operation,
and with three levels of fuel in the CWT. The pre-flight conditions, operations, weight,
taxi, takeoff, and flight path of TWA 800 were determined in detail and these conditions
were approximated as closely as practical in one of the flight tests. CWT ullage fuel/air
vapor samples were obtained on several flights at taxi, takeoff, and at altitude, and liquid
fuel samples were obtained from the CWT before and after several flights.
This report documents the formulation of the flight test plan, and
implementation of the flight test. Temperature measurement locations are detailed, and
the conditions for each individual flight test are described. Composite plots will show the
temperature time history of several measurement locations when key events in each flight
test occurred. In the text of this report, several documents, drawings, and plots are
referred to. These documents are contained in the following attached exhibits:

Exhibit 23B - Flight Test plan
Exhibit 23C - FAA Comments on Flight Test Plan
Exhibit 23D - Component Drawings
Exhibit 23E - Test Item Requirements List (TIRL), Instrumentation Location, Flight
Test Schedule
Exhibit 23F Ð Flight Test Results: TWA800 Emulation Flight


D. DETAILS OF THE INVESTIGATION

Section I Ð Flight Test Series
The flight tests consisted of a total of nine flights. Three of the nine flights were
performed for Boeing Aircraft. Data from the flights performed for Boeing is considered
to be proprietary information. A portion of one NTSB flight was used to obtain data for
the FBI, and a portion of another NTSB flight followed fuel management procedures
suggested by the FAA. The flight test procedure and the data obtained in the six NTSB
flights is summarized in this document. Details of the flight test planning,
instrumentation requirements, and each flight profile requirement are included in the
original Flight Test Plan given in exhibit 23B, pages 1 through 50.

Section I-A - Test Aircraft
The flight tests were performed on a Boeing 747-121 series airplane. The
airplane was leased from Evergreen Airlines, and Evergreen Airlines staff was utilized
for mechanical dispatch, maintenance, operations, and ground support of the flight test
series. The aircraft, N480EV, was built as Boeing line number 106 (Serial Number
20348), and was configured as a freighter at the time of the flight test. Prior to flight test,
the airframe had accumulated 92,504 flight hours and had 19,803 cycles. The aircraft was
flown on an experimental certificate for the flight tests. Evergreen provided ballast
weight and arranged fueling such that the airplane weight and balance, dispatch fuel load,
and takeoff fuel load for all the flights matched TWA800 as closely as practical.

Section I-B - Flight Test Plan
The Flight Test Plan given in exhibit 23B was developed by the Safety Board
staff and forwarded to the parties for comment. The flight test plan was developed such
that modification of some procedures in the flight test plan could be accomplished during
the flight test program. Hence, the flight test plan developed was modular in nature for
each flight, and the order of flights was determined as necessary throughout the course of
the flight test series.

In responding to the request for comments about the flight test plan, the Federal
Aviation Administration (FAA) requested additional temperature measurements to be
made, and a crossfeed procedure to be followed on one of the test flights. These requests
were implemented, and the additional temperature measurements and procedure were
included in the flight test program. Copies of the FAA comments and requests are
included in exhibit 23C, pages 1 through 6.

The flight tests were designed to examine possible combinations of ECS pack
operations prior to pushback and taxi. The combinations of pack operations were used to
provide possible differing heat loads and heat distributions to the CWT.
The general parameters for each of the flights in the flight test series are given in
table 1 below. Details of the specific pre-flight procedures and operations during each
flight are given in the attached Flight Test Plan, and are described briefly in later sections
of this document. For the flight in which ECS packs 1 and 3 were used, the ECS pack
operation was extended to accurately depict the pre-flight operations of TWA800. All of
the flight tests utilized an ascent flight profile similar to that of TWA flight 800 up to
17,500 feet altitude 1 . Several flights continued along a typical climb profile to reach a
cruise altitude of 35,000 feet altitude.


Section II Ð Airplane Areas of Interest

Center Wing Section Fuel Tank
The center wing tank of the 747-100 2 series of airplanes is illustrated in exhibit
23D, pp. 2 - 3. As seen in these illustrations, the entire 20Õx21Õx(4.5-6.0)Õ center wing
section consists of a forward dry bay, and several wet 3 bays. The lateral separations are
labeled from fore to aft as spanwise beam 3 (front wall of the first wet bay), spanwise
beam 2, the mid spar, spanwise beam 1, and the rear spar (rear of tank). The bays aft of
the mid spar also have a partition (butt line 4 zero [BL 0] rib), separating the mid and aft
bays into left and right sides. Each bay communicates 5 with the adjacent bays via small
holes located near the top and bottom corners of each bay partition to allow fuel and
vapor to flow between bays. In some of the lateral partitions are openings for tubing such
as the pick-up tube for the scavenge pump. The opening where the tube passes through
the partition is a larger diameter than that of the tubing, creating additional small
communication area exists in those locations also.

ECS Units
The ECS units are located in an enclosed area directly below the CWT. The
ECS packs and related pneumatic system are arranged schematically on the airplane as
shown in exhibit 23D, page 1. Hamilton Standard Inc. manufactured the ECS units on
the test aircraft and on the TWA800 aircraft. Each pack receives regulated bleed air from
the engine compressors, removes heat from the bleed air with a primary and a secondary
heat exchanger, and exhausts the excess heat beneath the airplane. The cooled bleed air
is routed to the cabin to provide a pressurized interior climate with comfortable
temperature. Drawings of the ECS units, with appropriate station locations, are shown
schematically in exhibit 23D, page 4.

Wing Tip Surge Tanks and Vent Stringers
The CWT and the wing fuel tanks maintain relative pressure equilibrium with
the atmosphere. The tanks are vented via enclosed stringers 6 in the wings, which lead to
a surge tank at each wing tip. The vents from the CWT and from all of the tanks in that
wing are collected into each wing tip surge tank. A single tube connects each surge tank
with the outside atmosphere on the lower wing surface. These vents allow the tanks to
equalize the internal tank pressure with the atmosphere during aircraft climb and descent.
According to Boeing, the vents are sized to permit passage of fuel if a fault occurs during
ground refueling. This design criterion requires a larger vent cross sectional area than
required to provide air pressure equalization alone.


Section III - Instrumentation

The flight test instrumentation locations were designed to obtain air
temperatures and pressures within the CWT and temperatures on surfaces within the
CWT. Particular attention was made to obtain the surface temperatures on the CWT
lower tank skin external surface. Air temperatures were also obtained in the ECS pack
bays below the CWT, and several ECS pack component surface temperatures were
recorded. Also obtained was air temperature and pressure measurements were made
within the wing tip surge tanks and in some vent stringers leading to the wing tip surge
tanks. Descriptions of each sensor location are described in detail in the Flight Test Plan.
Instrumentation locations are additionally described in the Test Item Requirement List
(TIRL), developed by Boeing and shown in exhibit 23E, pp. 1 - 40. Overviews of the
sensor locations are given in the sections below. Each of the sensor locations and
installations were photographed.

Section III-A Temperature Instrumentation
The temperature thermocouples used in this experimental test program were
constructed of bimetallic, type K, Chromel/Alumel. The thermocouples were constructed
of 20-gage wire (USWG) for use in the tank, and 24 USWG for use in the ECS
equipment bay. All thermocouples were designed to introduce no more than 0.02 milli-
joules into the CWT in a normal or failed condition. The thermocouples were either
mounted at their prescribed locations for air temperature measurement, or bonded to the
surface for surface measurements.

CWT Ullage 7 Air Temperatures
As described in previous sections, relatively small communication is possible
between adjacent bays of the CWT. Each bay was instrumented individually to
determine any spatial temperature gradients between the bays. Additionally, each bay
was instrumented to determine any vertical (top to bottom) temperature gradients. The
air temperature instrumentation locations are shown schematically in exhibit 23E, page
41. Each of the rear bays was equipped with a vertical array of three thermocouples. In
the forward two bays, vertical arrays of three thermocouples were placed on each side of
the bay. In addition, placed in the center laterally in the forward two bays was a larger
array of thermocouples. The center arrays consisted of five thermocouples near the top
and bottom of the CWT, placed at one-inch intervals. Thermocouples were also placed in
the center of each vent inlet.

CWT Fuel Temperatures
Instrumentation intended to measure liquid fuel temperature was placed at four
locations in the CWT. The locations were used with the recognition that the small
amount of fuel present in the CWT during the flights with 50 gallons of fuel in the CWT
would migrate to different portions of the tank with changes in aircraft attitude. Hence,
thermocouples were centered laterally in each rear bay, ½ inch from the CWT lower tank
skin internal surface. Additionally, thermocouples were placed at the calculated lowest
location in the CWT as the airplane sits on the ground, ½ inch from the CWT internal
lower skin surface.

CWT Internal Surface Temperatures
Several locations inside the CWT were instrumented to obtain internal surface
temperatures. Thermocouples were mounted in the center of the side of body rib in each
bay, on both left and right sides. The centers of the CWT internal upper skin of the two
forward bays, and the left aft bay were instrumented with surface thermocouples. The
rear spar, in the center (vertically and horizontally) of the left and right rear bays was also
instrumented with surface thermocouples. Additionally, a thermocouple was placed
where the crossfeed tube intersected the Butt Line 0 rib.

CWT Lower Tank Skin External Surface Temperatures
The external surface of the CWT lower skin, which forms the top surface of the
ECS pack bay, was instrumented in several locations. On left and right sides of the
CWT, thermocouples were centered under each CWT fuel bay (including the forward dry
bay) at the left and right butt line 22 location. At the left and right butt line 58 position,
10 thermocouples were spaced approximately every two feet, starting 1 foot aft of the
front spar (the front spar is the forward wall of the dry bay). Additionally, surface
thermocouples were placed on the CWT external lower skin above the left and right
pneumatic bleed ducts at the front and rear of the tank. These external locations and the
locations on the internal side ribs are shown schematically in exhibit 23E, page 42.

ECS Pack Bay Air Temperatures
The ECS pack bay was instrumented with air temperature thermocouples located
in several locations to determine any spatial air temperature gradients within the pack
bay. On the right and left butt line 22, thermocouples were place 4 inches below the
surface thermocouples at those locations. At right and left butt line 58, thermocouples
were placed 4 inches below every other surface thermocouple, starting at the rearmost
surface thermocouple location. The locations relative to the CWT lower skin are shown
in exhibit 23E, page 43.

ECS Component Temperatures
Thermocouples were placed on some components of each ECS unit. On each
ECS unit, thermocouples were placed on the pneumatic bleed duct near the flow control
valve, at the outlet of the water separator, on the upper surface of the inlet to the heat
exchanger, and on the upper surface of the compressor outlet. In the Flight Test Plan,
thermocouples were to be placed on one of the exhaust louvers of the heat exchanger.
However, installation difficulties prevented a thermocouple placement directly on the
louver, and the thermocouple was placed on the side support housing of the louvers.
Wing Tip Vent Surge Tanks and Tank 3

In both wing tips, air temperature thermocouples were placed in the center of the
surge tanks where the flow from all of the tanks combine to flow out the wing vent.
Thermocouples were placed at each (left and right) overboard vent duct collector can,
inside each surge tank. In the right surge tank, thermocouples were placed at the surge
tank end of the vent stringer channel coming from the CWT, tank number 3, and tank
number 4. Vent outlet thermocouples (2) were placed in the left wing surge tank
overboard vent duct. One was placed inside the duct near the bottom surface of the wing,
and one was placed inside the horizontal section of the duct, approximately 8 inches
inboard from the outboard end of the horizontal duct. The thermocouple in the horizontal
section was installed such that there was an unobstructed path to the surge tank flame
suppression sensor as installed on the TWA800 accident aircraft 8 . A surface
thermocouple was also placed on the vent stringer exit from the number 3 tank, on the
inside surface of the top wing skin.

At the request of the FAA, air temperature thermocouples were placed inside
main fuel tank number 3 (right inboard main tank). A thermocouple was placed at each
vent inlet, and a thermocouple was placed near the high point of each fuel pump power
conduit. The details of the locations of the thermocouples are included in the FAA
response to the Flight Test Plan. These locations are shown schematically in exhibit 23E,
page 44. This drawing shows the CWT, tank 3, and the vent stringers leading to the wing
tip. The air temperature measurement locations at the CWT and tank 3 vent inlets are
shown. Also shown are the measurement locations in tank 3 near the fuel pump power
conduits.

Section III-B Pressure Instrumentation
The pressure gages used in the flight tests were manufactured by Rosemount
Inc., (part # 1332A16EP2), and measured total pressure. Pressure was measured near the
vent inlets in the forward two bays and near the vent inlet in the rear bay. Pressure gages
were also placed in each (left and right) of the wing tip surge tanks.

Section III-C Acceleration Instrumentation
Acceleration of the CWT lower skin surface was measured using three
accelerometers, one aligned along each axis of the aircraft (vertical, longitudinal, and
lateral). Endevco Inc. manufactured the accelerometers (part # 7290A-10M41A) used for
all axes. The acceleration data was acquired by the recording system (described in a later
section) maximum sampling rate of 800 Hz.

The accelerometer block was to be mounted on the external surface of the CWT
lower skin, centered underneath the right rear bay. However, a structural support
member was located at the point mid way between the keel beam and the side of the tank.
Hence, the accelerometer block was located midway between the support beam and the
keel beam, at butt line 31. The acceleration block was centered longitudinally between
the right rear spar and spanwise beam 1.

Section III-D Vapor Sampling Equipment
A device to sample fuel/air vapors from the CWT ullage was designed and
constructed by Boeing with design input from the Desert Research Institute (DRI). The
sampling unit consisted of six one-liter evacuated stainless steel canisters. The unit was
connected via copper tubing to the forward CWT wet bay (between spanwise beam 2 and
spanwise beam 3). The tubing was installed through a replacement access door, and the
inlet to the sampling unit was approximately 2.5 feet from the CWT internal lower skin
as installed.

The sampling unit consisted of a six port manifold within an aluminum case,
equipped with a main cutoff valve, and six individual cutoff valves, one for each of the
evacuated sampling canisters. Two samples were required at each vapor acquisition; one
sample canister was used to purge the lines leading to the unit, and one sample canister
collected the actual sample to be analyzed. Thus, three valid samples could be collected
on the desired flights. The entire unit was installed in the main cargo compartment, and
was operated manually during the flight tests at the desired collection times. Details of
the vapor sampling equipment are included in the vapor sample factual report 9 developed
by DRI.

Section III-E FQIS Parameters
During the System Group investigation, the fuel quantity indication system
(FQIS) was found to have two unshielded wires and one sheilded wire routred between
the rear win spar and the cockpit. One of the wires, known as the LO Z 10 wire, was
instrumented with shielded wiring to record the system voltage and current. Prior to one
flight, the FQIS wiring was exposed to various external sources of electromagnetic
induction (EMI). Sources such as a hand held transceiver, laptop computer, cellular
telephone, electric shaver and airplane systems were operated in the cockpit and cabin
compartments. A description of the procedure used to test the energy induced by each
device is described in a later section.

Section III-F Flight Parameters
Several flight parameters were recorded on the data acquisition unit. Parameters
such as airspeed, altitude, roll angle, pitch angle, heading, total air temperature, were
recorded. Flight parameter data was obtained from the aircraft air data computer, inertial
reference unit, and global positioning system (GPS) unit. The fight data recorder was
secured following the first three flights as a backup data source.
Section III-G Data Acquisition

The Boeing Commercial Aircraft Company provided the flight test
instrumentation, flight data acquisition equipment, and performed the installation and
removal of flight test instrumentation. Safety Board personnel provided installation
requirements, and provided oversight of the installation procedure. Boeing furnished
instrumentation operation personnel and the flightcrew as detailed in the Flight Test Plan.
All flight test data was recorded using a Loral Portable Airborne Digital Data System
(PADDS II) system. The PADDS unit consisted of an analog to digital converter,
Remote Multiplex unit (RMUX), and the Central Multiplex unit (CMUX) which
amplified the signals from the individual thermocouples and converted the signals into
numbered counts 11 . The signal counts were then passed into a laptop computer for
conversion into engineering units. The data time history, in engineering units, were
recorded on magnetic tape at the respective sampling rate.


Section IV Ð Flight Tests Description

Members of the flight test group, the flight crew, and instrumentation
technicians attended a flight readiness review on July 8-10, 1997, at JFK Airport. The
procedures and the overall safety of the flight test series was discussed and agreed upon
by all parties, and approval of the flight test plan was obtained from all parties.
Installation of the instrumentation, vapor sampling equipment, data acquisition
equipment, and maintenance repairs to the aircraft continued through July 14.

Jet A fuel was procured and placed in the CWT prior to the first flight of the
flight test series. The fuel, purchased from Olympic Airways, was originally loaded onto
an Olympic Airways 747 in Athens, Greece, and flown on a regular service flight from
Athens to JFK. The route was similar to the last leg completed by the accident airplane.
The fuel was unloaded from the Olympic Airways 747 wing tank into a fuel truck,
transported, and 50 gallons of fuel was loaded into the test airplane CWT prior to the first
flight. This fuel remained in the CWT for the first four flights.

During the flight tests, the occupants of the aircraft were limited to the flight crew
and personnel required for conducting the test. In addition to the flightcrew, the NTSB
Program Test Director, the NTSB Flight Test Group Chairman, Boeing Test director,
Flight Analysis engineer, and Flight Instrumentation engineer were on board the aircraft
for all flights. For the flights in which vapor sampling was performed, an additional
vapor-sampling operator was on board.

This portion of the document describes the flights performed for the NTSB only
(six total flights), and describes procedures as performed in the flight tests. All times in
this report are Eastern Daylight Time (EDT), and are given in 24-hour format HHMM or
HHMM:SS. The flight test series, showing flights as performed, are given in the
schedule shown in exhibit 23D, page 45.

Monday, July 14, 1997
Packs 2 and 3
The first flight of the test series began on Monday, July 14. The data
acquisition unit started recording data before the auxiliary power unit (APU) or any ECS
packs were in operation. The data recording commenced at 0920 Eastern Daylight Time
(EDT). ECS packs 2 and 3 were started at close to 0950, and operated at full cold. Fuel
began to be loaded into the main tanks at approximately 0930, and the fuel truck delivery
temperature of the fuel was 80° F. The outside air temperature (OAT) at ECS startup was
86° F. The ECS packs ran continuously after initial startup. By the time pushback from
the blocks occurred and initial engine start was accomplished at 1210, the outside air
temperature had risen to 91°F.

The TWA procedure for engines start (as included in the Flight Test Plan)
during taxi was utilized. This procedure entails shutting down the ECS packs during all
engine starts. The aircraft proceeded to taxi, and liftoff from JFK occurred at
approximately 1237. The TWA800 ascent profile was used up to 14,000 feet altitude,
and a similar climb rate was used up to 17,500 feet altitude. The aircraft remained at
17,500 feet altitude, 250 knots indicated airspeed (KIAS) from 1253 until approximately
1522. The airplane then continued to the Atlantic City, NJ airport to perform additional
flight testing for the FBI. During this series of tests, data was continually recorded.
After the testing for the FBI was completed, the aircraft returned to JFK airport at
approximately 1910.

Tuesday, July 15
All three packs
On the second day of testing, the airplane fuel was loaded in the main tanks
prior to the APU startup at 0830. The data acquisition unit started recording data 15
minutes after APU start. The temperature of the fuel at delivery was 82° F as measured
on the fuel truck. The ECS packs 1, 2, and 3 were started at approximately 0845, and all
were operated at full cold. The outside air temperature at ECS startup was 82° F. At
1128, a problem with the bleed air isolation valve occurred during the engine start
procedure. The aircraft was pushed back into position, and the valve was manually
opened. The ECS packs remained off during this procedure. After the valve was closed,
pushback was performed a second time. When all four engine starts were accomplished
by 1202, the outside air temperature had risen to 90° F and the fuel temperature in the
main wing tanks had risen to 91° F.

During the taxi, an ullage vapor sample was collected in the vapor-sampling unit
at 1206. Takeoff roll and rotation occurred at 1211. The TWA800 ascent profile was
used up to 14,000 feet altitude, and a similar climb rate was used up to 17,500 feet
altitude. During the ascent, a fuel/air vapor sample was acquired at 1218 when the
aircraft was at 10,300 feet altitude. The third vapor sample was acquired at 1224 as the
aircraft passed 14,200 feet altitude. At 1226, the aircraft leveled at 17,500 feet altitude,
and remained at that altitude for approximately 1 hour. At 1330, the airplane began to
ascend to 35,000 feet altitude, and the ECS packs operation was switched to automatic.
The aircraft remained at 35,000 feet altitude for approximately 2 hours, and started to
descend at approximately 1600. The aircraft touched down at JFK at 1628. Upon
landing, pack 2 was switched off, and packs 1 and 3 remained running on full cold. The
remaining fuel in the main tanks was 53° F; the outside air temperature at touchdown was
90° F. The engines were shut down at 1637. The APU and ECS packs 1 and 3 remained
running for the next flight test.

Packs 1 and 3 (TWA800 Emulation Flight)
The airplane was taxied into position for the preparation of the TWA800
emulation flight. The airplane main tanks were defueled to the proper level as
determined by the Operations group to be representative of the fuel load of TWA 800
prior to fueling. ECS packs 1 and 3 continued in operation during the entire on ground
phase. The fuel load for the TWA800 emulation flight was loaded in the aircraft main
tanks beginning at 1800. The same amount of fuel that was loaded onto TWA800 was
loaded into the wing main tanks of the test airplane. Four trucks were used to fuel the
airplane, and the temperature of the fuel delivered ranged from 88° F to 91° F. The
outside air temperature at fuel loading was 88° F, and the fuel loading was completed at
1908.

Pushback occurred at 1940, and engine number 4 was started five minutes later.
Engines one and two were started approximately 10 minutes after engine number 4. The
ECS packs were turned off for the engine starts. At 1957, the first vapor sample was
taken during the start of taxi. Engine three was then started at 2012. The TWA takeoff
procedure (as given in the Flight Test plan) was followed, and the packs were turned off
for takeoff. Liftoff from JFK runway 22R occurred at 2021. The landing gear was
raised, and the TWA procedure for ECS pack restart was followed. ECS pack 1 was
started at 400 feet altitude, pack 2 at 600 feet altitude, and pack 3 at 800 feet altitude. All
three ECS packs remained in operation for the rest of the flight.

At 2028, the second vapor sample was obtained as the airplane passed through
10,000 feet altitude. Also at that altitude, the main tank cross-feed procedure as defined
in the TWA Flight Handbook for fuel management (and given in the appendix of the
Flight Test Plan) was performed. The emulation flight reached the TWA800 event
altitude at 2032. At 2033, when the airplane passed through 14,200 feet altitude, the third
vapor sample was obtained. The airplane continued to climb, and leveled at 19,000 feet
altitude at 2036. The airplane remained at 19,000 altitude and started descent at 2239.
Touch down occurred at 2257, and the airplane was parked at 2313. The data recording
equipment was operated for approximately 20 minutes after engine shutdown.

Wednesday, July 16
Boeing Flight 1
On the morning of July 16, the first flight for the Boeing Commercial Aircraft
Company was performed. ECS Packs 1 and 3 were started at 0750, and fuel was loaded
in the main wing tanks at 0815. Pushback occurred at 1011, and taxi began at 1031.
Takeoff from JFK occurred at 1044, and the same climb profile as the previous tests was
used up to 18,000 feet altitude. Starting at 1100, the aircraft remained at 18,000 feet
altitude and 300 KIAS. At 1328, the aircraft started to climb, and reaching flight level
(FL) 350 12 at 1345. The airplane remained at this altitude for two hours. Descent from
FL 350 began at 1545, and the aircraft landed at JFK at 1628.

Packs 1 and 2
The data recording for this flight test began at 1636, immediately after the
previous test. Packs 1 and 2 were switched to full cold at that time. At 1704, the fueling
of the main tanks was started, and was completed approximately 45 minutes later. The
fuel temperature upon delivery was 84° F, and the outside air temperature was 81°. Push
back occurred at 1918, and after engines start, taxi commenced at 1932. The first ullage
vapor sample was obtained two minutes later.

Liftoff from JFK occurred at 1955. The TWA pack restart procedure was used on
climbout, and all three packs were operated for the duration of the flight. At 2001, as the
aircraft passed through 6000 feet altitude, the crossfeed procedure as specified by the
FAA was initiated (see the Flight test Plan, appendix II for details). Over the next five
minutes, tank 3 supplied the fuel for engines 1, 2, and 3 with both number 3 boost pumps
operating. Tank 4 supplied the fuel for engine 4, with both number 4 boost pumps
operating. After five minutes had elapsed, the TWA Flight Handbook fuel management
procedures were resumed.

Fuel vapor samples were obtained as the aircraft passed through 10,400 feet
altitude at 2006. The third vapor sample was obtained at 2012, as the aircraft passed
through 14,700 feet altitude. The aircraft leveled off at 17,500 feet altitude and remained
for approximately two hours. The aircraft touched down at JFK at 2241. The data
recording units continued to record data for approximately 20 minutes after engine
shutdown.

Thursday, July 17
6000 Pounds Fuel in CWT
Refueling of the main wing tanks and CWT with two trucks began at 0726. The
temperature of the fuel at the beginning of fueling was 80° F from one truck, and 82° F
from the other. The outside air temperature was 80° F. At 0732, the data recording
commenced, and all three packs were turned on and placed in full cold mode. By 0751,
the fueling was completed, and 6000 pounds of fuel had been placed in the CWT.

During the ground portion of this flight test, the EMI from several personal
electronic devices was tested. A cellular phone, pager, ham radio, and electric shaver
were operated in the cabin areas. A laptop computer was operated, employing the file
save operation and a CD-ROM operation. These devices were operated as they were
walked along the upper deck wall, near the FQIS CWT wiring. The devices were then
walked along the left side of the main deck, near the CWT FQIS wiring. Additionally,
several airplane systems were operated from the flight engineerÕs panel. The external and
internal lights were cycled on and off, each radio was transmitted, and the radar altimeter
and radar transponder was operated. The pitot heat, window heat, stall warning, and
electric trim actuator were activated in the cockpit. During these personal electronic
device and airplane system operations, the FQIS system LO Z voltage and current were
monitored real time and also recorded on the data recording system.

Pushback occurred at 1017, and taxi commenced at 1030. At 1043, the airplane
lifted off, and performed the same ascent profile as the previous flights. By 1059, the
aircraft had leveled at 17,500 feet altitude. Ascent to FL350 started at 1200, and reached
FL350 at 1217. Descent began at 1427, and the CWT was used to fuel all engines during
the descent. The aircraft touched down at 1452. Engine shutdown occurred at 1501, and
the flight engineer panel AC power switches were cycled on and off. Data recording was
stopped approximately 15 minutes after engine shutdown.

Friday, July 18
12,000 Pounds Fuel in CWT
Upon completion of the previous flight and review of the preliminary data, it
was decided a flight in addition to those outlined in the Flight Test Plan should be
performed. After consultation with the instrumentation crew and the flightcrew, an
additional flight was planned for Friday morning. It was determined that the addition of
an extra flight at this point in the flight test program would not greatly affect the overall
flight test schedule and completion date.

On the morning of the 18 th , ECS packs 2 and 3 were turned on and the data
recording units were started at 0652. The ECS packs were placed in auto mode; the
outside air temperature was 79° F at pack startup. At 0720, fuel was loaded into the main
tanks and CWT. The CWT was loaded with 12,000 pounds of fuel. The fuel temperature
upon delivery was 80° F, and fuel loading was completed by 0749.
Pushback occurred at 0800. By that time, the outside air temperature had risen
to 81° F. As in the previous tests, the packs were turned off for engine startup. Taxi was
initiated at 0809, and liftoff from JFK occurred at 0832. The same ascent profile was
used during the climb, and level off at 17,500 feet altitude occurred by 0847. At 0916, a
further ascent was initiated, but the airplane was asked to remain at FL190 by air Traffic
Control (ATC). The airplane continued its climb at 0923. By 0938, the aircraft had
leveled at FL350. Fuel from the CWT began to be used for all engines at 0941. By 1010,
the fuel level in the CWT had been reduced to 2,000 pounds, and the flight engineer
determined that the pumps probably could draw no more fuel from the CWT until
descent, and engine fuel source was switched back to the main wing tanks. The fuel level
remained at close to 2,000 pounds until descent.

Descent from FL350 started at 1038, and the CWT fuel pumps were placed back
on, and the remaining CWT fuel was used for the engines. At 1043, the right CWT pump
was turned off. The left CWT fuel pump was turned off at 1046, and the scavenge pump
was turned on. The scavenge pump remained on until 1106, when the CWT fuel quantity
gage read nearly zero. Touchdown at JFK occurred at 1113. Engine shutdown occurred
at 1127, and the data recording remained in operation for an additional 15 minutes.

Friday, July 18 (PM) and Saturday, July 19
The final two flights of the flight test series were performed for Boeing. The
first flight on Friday afternoon utilized a modified pack bay inlet, and the flight on
Saturday was made following replacement of the ECS pack bay seals with new hardware.
Both flights were performed with 50 gallons of fuel in the CWT, and with packs 1 and 3
running 3 hours before takeoff. The same flight profile as the first Boeing flight was
followed for both flights.

The flight test program was completed on Saturday, July 19. The aircraft was
refurbished and returned to Evergreen Airlines over the next two days. During the flight
test series, a total of 39 hours, 43 minutes of flight time was accumulated. A total of
close to 70 hours of data consisting of temperatures, pressures, accelerations, and flight
parameters were recorded during the flight tests.


Section V Ð Flight Test Data

Temperature and flight parameter data collected during the flight test series is
presented in this section. However, due to the large volume of data collected, much of
the data will be presented in separate documents 13 . This section of the report will
discuss pertinent results from the TWA800 emulation flight (Tuesday, July 16 PM) only.
Upon completion each flight test, the Flight Test Group Chairman took the
recorded data into custody, and delivered the data tape to Tom Jacky of the NTSB. Mr.
Jacky supervised the readout of the data at Boeing facilities in Seattle, WA. Once the
13 All of the flight test data will be included in separate Addendums to the Flight Test Group ChairmanÕs
Factual Report of Investigation. The ullage fuel/air vapor sampling data and liquid fuel sampling data are
reported in separate documents within the Fire and Explosion Group ChairmanÕs factual report.
Acceleration data will be included in a separate addendum document to the Flight Test Group Report.
data was read out, the data was transferred electronically to the Safety Board
headquarters in Washington, DC. The length of each flight test, the number of recorded
parameters, and the sampling rate of each parameter resulted in several gigabytes of data
for each flight test.

The plots described in this section contain the temperature time histories of
several CWT ullage and surface locations, ECS pack bay temperatures, and wing-tip
surge tanks for the entire flight. Examination of the data as recorded (one sample per
second) showed that the temperatures for much of the CWT vary relatively slowly with
time. Displaying the data at one sample per minute still accurately captures the
temperature change with time, and allowed the creation of smaller size data files. Hence,
the data for an entire flight have been sub-sampled 14 to one sample every 60 seconds.
For portions of the data with larger temperature gradients over time, such as the ascent
portions of the flight, the data is presented in one sample per second format. The data in
this section is displayed as a function of elapsed time, which represents the elapsed time
of data recording during each respective test flight.


ection V - A TWA800 Emulation

Flight Parameters
As described in the previous section, in the flight test program efforts were made
to duplicate as accurately as possible the preflight operations, takeoff and ascent of
TWA800. The time of liftoff of the test flight was within 1 minute of the time of day of
TWA800, and two days before the anniversary of the accident flight. Exhibit 23F, page 1
compares the altitude and airspeed of this test flight with the FDR data from TWA800.
As shown in this plot, the test flight elapsed time from rotation to the accident altitude
was within ten seconds as compared to TWA800. Slight variations evident during the
ascent of the flight test aircraft were necessary to comply with Air Traffic Control (ATC)
instructions during the flight tests. However, the overall climb profile matched the
accident flight within 1000 feet, and elapsed times to the brief level off at 6000 feet and
12,800 feet during the climb matched within one minute. Additionally, the airspeed
throughout the ascent matched the speeds recorded on the TWA800 FDR to within 20
knots for the majority of the FDR recording.

CWT Ullage
Exhibit 23F, page 2 shows the temperature time history in the center (laterally
and longitudinally) of the forward bay, between spanwise beams 2 and 3 15 . The
temperatures shown are arranged vertically, and consist of an upper thermocouple located
3 inches below the upper skin surface, a thermocouple centered vertically, and a
thermocouple 3 inches above the internal lower skin surface of the CWT. Exhibit 23F,
page 3 shows the temperature time history in the center of the bay between spanwise
beam 2 and the mid spar 16 . Whereas there were measurements made at one inch vertical
spacing near the top and bottom surfaces of the center of the two forward bays, only one
measurement each near the top and bottom are shown for consistency with measurements
at other locations 17 . The temperature measurements shown on pages 2 and 3 are located
3 inches below the internal upper skin surface, 3 inches above the internal lower skin
surface, and centered vertically between.

Vertical measurements were also obtained at the left and right sides of the
forward two bays. Shown in exhibit 23F, pp. 4 -7 are measurements from the
thermocouples placed 4 inches laterally from the respective side walls. The
thermocouples installed near the side walls utilized the same vertical spacing as the aft
and mid bays.

Shown in exhibit 23F pages 8 and 9 are the ullage temperatures time history in
the left mid bay and left rear bay of the CWT, and shown in exhibit 23F pages 10 and 11
are similar plots for the right rear bay and the right mid bay. The position of the
measurement locations are centered laterally between the side body rib and the BL 0 rib,
and centered between the respective front and rear partition. The measurements in the
mid and rear bays are in the same relative vertical positions (3 inches from upper and
lower skin surfaces, and centered between) as those shown for the forward bays.
Pertinent events of the flight test, such as start of taxi and time of reaching TWA800
event altitude, are noted on the plots.

Evident in all the bays is a steady increase in temperature while the aircraft is
stationary on the ground with the ECS packs in operation. A vertical gradient of
temperatures exists in all of the CWT bays, and most predominantly in the aft and rear
bays, is. For all bays, the temperature in the lower part of the tank is warmer than the
temperature near the upper skin surface when the aircraft is stationary. However, as the
aircraft is pushed back and begins to taxi, there is a change of slope in the temperatures,
i.e. the temperatures begin to either level off or decrease slightly. Each bay exhibits a
slightly different trend, and the difference between the upper and lower measurement
changes in different manners depending on the phase of the flight.

Exhibit 23F, pp. 12 - 14 show a wire frame representation of the CWT, and
temperatures within the CWT at certain times during the TWA800 emulation flight test.
These plots represent the vertical distributions of temperature at their respective
measurement location in each bay. Exhibit 23F, page 12 shows the conditions at start of
taxi, page 13 shows the distribution at takeoff, and page 14 shows the temperature when
the test aircraft reached the TWA800 event altitude. For both aft and mid bays, the left
side of the tank has considerably higher temperature at all vertical levels.
As evident in the temperature time history plots, the highest ullage temperatures
observed when the aircraft was on the ground and in flight occurred in the left mid and
aft bays. The maximum temperature observed in the tank is in the left mid bay, where a
temperature of 145° F near the internal lower surface of the tank is observed before taxi
commences. In addition, the largest temperature gradient of over -20° from bottom to top
is noted in the left mid bay. The left rear bay reaches a maximum temperature of 138° F
near the tank internal lower surface, with a -9° F degree vertical gradient evident before
taxi.

The forward two bays also exhibit a vertical and spatial gradient of
temperatures. Exhibit 23F, pages 2 and 3 showed the time histories of the center array in
bay number one and bay number two, respectively. The maximum temperature measured
in forward bay #1 is 124° at the internal lower surface, with a -5° gradient from bottom to
top. A similar gradient is evident in the forward bay #2 between the mid spar and
spanwise beam 2. When the test airplane reached the same altitude as the TWA800
event, forward bay #1 shows a maximum temperature of 117° near the internal lower
surface, and the vertical gradient to the upper skin surface has reduced -3°. Bay 2 has a
maximum temperature of 120° at the center of the internal lower surface, and a -3°
gradient to the upper measurement. The sides of the two forward bays show a lower
temperature as compared to the center, and the maximum vertical temperature occurred at
the center thermocouple.

The temperature time history for the lower internal tank surface of forward bay
#1 shows large, short duration increases in the temperatures near the start of taxi. The
temperature of the short duration, higher spiked values are consistent with the
temperature of the fuel in forward bay #1, measured during the same portion of the flight
test. The fuel temperature, measured ½ inch from the lower internal tank surface of
forward bay #1 is shown in exhibit 23F page 15, and shows a similar sharp rise in
temperature during taxi, when the probe is immersed in liquid fuel.

The liquid fuel temperature measurement ½ inch from the lower internal tank
surface at the rear spar is shown in exhibit 23F, page 16. As demonstrated in these plots,
the liquid fuel can migrate between bays as the aircraft changes pitch attitude, or as the
aircraft brakes or turns during taxi. If liquid fuel is heated in one bay of the CWT and
then migrates to another portion of the tank, the liquid fuel temperature can be
considerably higher than the ullage air temperature in a particular bay.

The temperature time histories in the left and right mid bay ullage, with specific
focus on the ascent portion of the TWA800 emulation test flight is shown in exhibit 23F
page 17. The data in this exhibit show the data as collected, at one sample per second.
During taxi, takeoff, and ascent, all of the ullage temperatures exhibit a decline in
temperature. Each vertical position of each bay shows a decrease in temperature as the
aircraft ascends. By the time the aircraft reaches the same altitude as the TWA800 event,
the temperatures in the warmest part of the CWT (i.e. the left mid bay) have reduced to
128° F in the lower location, with a -14° gradient to the upper location. The temperature
in the right mid bay is at 114° F near the lower internal tank surface, and a -7° gradient to
the top.

Exhibit 23F, page 18 shows a similar examination of the temperatures in the left
and right aft bays during ascent. The aft bays show a similar behavior, with the left aft
bay lower location reducing to 120° F and a -6° gradient to the upper location at the event
altitude. The right aft bay lower location is 112° F at the TWA800 event altitude, with an
-8° gradient to the top location. The temperatures measured in the ullage bays at the
TWA800 event altitude are summarized in Table 2 below. As evident in these
measurements, the warmest part of the CWT at the TWA800 event altitude is the left mid
bay, and the coolest part of the tank is at the side of the forward bays.

CWT Lower Tank Skin External Surface Temperatures
Exhibit 23F, pages 19 and 20 show the time histories of the CWT external lower
skin surface temperature measurements on the left and right side of the keel beam, at the
BL22 18 location. As described in a previous section, each of the surface thermocouples is
centered longitudinally below each bay at the BL22 lateral location. Similar to what was
measured in the ullage, the surface below the left mid bay maintains the warmest
temperature throughout the rampÐhold portion of the flight test. As shown in exhibit
23F, page 19 for the left side of the tank, the maximum external surface temperature
obtained below the left mid bay was 203° F. Immediately before start of taxi, the
temperature was steady at 200° F. The surface below the left aft bay maintained an
approximate 9° F lower temperature than the left mid bay during the ramp hold.
A substantial spatial gradient from rear to front is evident from the left side
surface measurements. A 60° F spatial gradient between the left mid bay and the dry bay
BL22 location is evident during the ramp-hold. On the right side of the keel beam, at
BL22, the surface temperature is significantly lower than the corresponding position on
the left side of the keel beam. The maximum temperature observed on the right side of
the keel beam at the BL 22 location is below the dry bay, where a temperature of close to
150° F is observed immediately before taxi. However, at taxi start, a much smaller
surface spatial temperature gradient (-16° F) existed between the maximum and
minimum temperature at the right BL22 location.

The BL58 location surface temperatures on the left side of the keel beam are
shown in exhibit 23F, page 21 and 22. The number label on each time history trace
represent the thermocouple locations as described in the previous instrumentation section,
and are numbered sequentially from front of the tank to the rear (i.e. 1 is the most
forward, 10 is the most aft). In a similar fashion to the BL22 location, the left side BL58
is the warmest near the rear of the tank. Near the time of taxi start, the rear of the tank at
BL58 had reached over 180° F, and spatial gradients of over -40° F existed from rear to
front. On the right side of the keel beam, shown in exhibit 23F, page 23 and 24, the
temperatures near the center (i.e. near the mid spar) of the tank reach a maximum of close
to 150° F before taxi. When taxi commenced, all of the bay surfaces reduced
temperatures, although at significantly different rates and magnitudes. The rates of
temperature change also varied considerably upon liftoff. At the TWA800 event
altitude, all of the CWT lower skin external surface temperatures are in the range of 120°
to 145° F.

ECS Pack Bay Air temperatures
The air temperatures measured 4 inches below the CWT lower skin surface (i.e.
in the ECS pack bay) are shown in exhibit 23F, pp. 25 - 28. The air temperatures display
a similar trend as the surface temperatures, with large spatial gradients evident. The
warmest measurements in the ECS pack bay occurs on the left side of the keel beam,
below the left mid and left aft bays at the BL 22 location. Below these left bays, the air
temperature reached to close to 238° F before taxi. The measurements show cooler air
temperatures on the right side of the keel beam in the ECS pack bays, with the maximum
air temperature measured of close to 190° F. On both sides, the BL22 location air
temperatures are warmer than the BL58 air temperatures, and show higher temperature
measurements towards the rear of the ECS pack bays.

Upon taxi and liftoff, changes in the trend and distribution of air temperatures in
the ECS bays are apparent. During the ascent, all of the ECS pack bay air temperatures
reduced as the altitude increased. Additionally, the spatial gradients diminished during
the climb, and a different distribution of temperatures is evident after the test aircraft
leveled off at 19,000 feet-msl.

ECS Pack Component Surface Temperatures
As described previously, surface temperatures were measured on several
components of the ECS components. Measurements were made to determine the upper
(facing towards the CWT lower skin) surface temperature of some of the hotter
components of the ECS units. Exhibit 23F, page 29 shows the surface measurements on
the top surface of the inlet to the heat exchanger, and at the compressor outlet for all three
packs. During the on ground portion of the test, the inlet to the heat exchanger for the
packs in operation (packs 1 and 3) reached temperatures between 330° F and 350° F. The
compressor outlets of the packs in operation were approximately 280° F. For the pack
not in operation, the heat exchanger inlet still reached a temperature of 270° F before
pushback. During taxi and ascent, the temperatures of these components varied
considerably, especially during portions of the taxi when the units were turned off for
engine start and liftoff. Moderately steady temperatures for the heat exchangers and
compressor outlets are observed after the aircraft leveled at 19,000 feet and had all three
packs in operation.

Exhibit 23F, page 30 shows the top surface temperature measurements of the
pneumatic bleed control valves, which control the bleed air from the compressor section
of the engines into the ECS units. The bleed control valves for packs 1 and 3 warmed to
360° F and 340° F respectively before pushback. The control valve for Pack 2, which
was not in operation prior to liftoff, reached approximately 310° F before pushback. All
control valve temperatures fluctuated during taxi and takeoff. When the airplane leveled
at 19,000 feet and all three packs were in operation, the valve components for all packs
ranged between 220° F and 250° F.

Exhibit 23F, page 31 shows the temperatures measured at the exhaust louvers of
the ECS units. As described in the instrumentation section, the exhaust louvers are
located on the bottom of the aircraft, at the location where the heat removed from the
bleed air is expelled to the atmosphere. However, instrumentation installment difficulties
precluded the measurements being directly in the exhaust flow, and were mounted to the
side louver support. Hence, the potential existed for the temperature measured at this
location not being the desired temperature, or influenced by other factors. As seen in this
exhibit, the temperature traces for the packs in operation do not match during the ground
portion of the test. When the aircraft reached altitude, the other component time histories
imply the three pack operating well within comparable temperature ranges. However, in
this plot the temperature at the exhaust louver of pack 3 shows an approximate -70° to -80°
offset. This offset first appears within the first 60 minutes of the test, and continues
during portions of the flight test when all three packs were operating and other
component temperature measurements were consistent.

Shown in exhibit 23F, page 32 are the surface temperatures of the ECS water
separators. The water separators remove water from the cooled air before it is ducted into
the cabin compartment. The surface temperatures of the water separators for packs 1 and
3 remain relatively low for the duration of the entire flight. The surface of the water
separator for pack 2 follows closely the ambient air temperature in the ECS pack bay
(pack 2 is on the left side of the keel beam) until lift off, when pack 2 is turned on. After
liftoff, all three water separators have the similar surface temperatures for the rest of the
flight until touchdown, when the units are turned off.

Wing Tip Vent Surge Tanks and Tank 3
The temperature measurements in the right wing tip surge tanks and vent
stringers are shown in exhibit 23F, pp. 33 - 35. Exhibit 23F, page 33 shows the
temperature centered in the CWT vent stringer, at the point where the CWT vent joins the
collector can 19 . As demonstrated in these plots, the temperatures in the vent stringer and
collector reach quasi-equilibrium with the outside ambient temperature during the ramp-hold
portion of the test. Soon after liftoff occurred, the temperature in the CWT vent exit
and collector reduced considerably as the aircraft increased altitude. This is also shown
in exhibit 23F page 34, which shows the ascent portion of the trajectory in detail. A
similar trend is shown in the vent exits from main tank number 3 and 4, shown in exhibit
23F page 35. At the TWA800 event altitude, the temperatures in the all the vent stringers
and in the surge tank have reduced to approximately 68° F.

Main wing tank #3 ullage temperature is shown in exhibit 23F, pages 36 and 37.
Temperature measurements were made at the vent inlets and near the fuel pump power
conduits as described in the instrumentation section. Tank 3 exhibits a similar behavior
as that of the wing vents, in that a relative constant temperature of approximately 88° F is
obtained, with the exception of the temperature in the outboard vent opening. This
temperature exhibits a different trend during ramp hold, but obtains a similar quasi-equilibrium
before taxi commenced. After lift-off, all of the measured tank 3 ullage
temperatures show a decrease with increasing altitude. When the test airplane reached
the TWA800 event altitude, the ullage temperatures in Tank 3 had reduced to the range
75° F to 85° F.

Daniel R. Bower, Ph.D.
Aerospace Engineer
Flight Test Group Chairman

---

 

---

Sound off on our message boards Tell us what you think! You said it...