Home > Aircraft Cabin Water Mist Test Report
NAWCADLKE-MISC-435100-0011
NAWCADLKE-MISC-435100-0011
FULL-SCALE AIRCRAFT CABIN
TWIN FLUID, LOW PRESSURE WATER MIST
FIRE SUPPRESSION SYSTEM
PROOF-OF-CONCEPT
EVALUATION
FINAL TEST
REPORT
16 June 2000
TWIN FLUID, LOW PRESSURE WATER MIST NOZZLE
(Actual installation
between two (2) overhead stowage bins.)
DISTRIBUTION AUTHORIZED TO DOD AND DOD CONTRACTORS ONLY; CRITICAL TECHNOLOGY; 16 JUNE 2000. OTHER REQUESTS FOR THIS DOCUMENT SHALL BE REFERRED TO THE COMMANDING OFFICER, NAVAL AIR WARFARE CENTER AIRCRAFT DIVISION, LAKEHURST, NJ 08733-5000. |
WARNING - THIS DOCUMENT CONTAINS TECHNICAL DATA WHOSE EXPORT IS RESTRICTED BY THE ARMS EXPORT CONTROL ACT (TITLE 22, U.S.C SEC 2751 ET SEQ.) OR EXECUTIVE ORDER 12470. VIOLATIONS OF THESE EXPORT LAWS ARE SUBJECT TO SEVERE CRIMINAL PENALTIES. |
DESTRUCTION NOTICE, FOR UNCLASSIFIED, LIMITED DOCUMENTS, DESTROY BY ANY METHOD THAT WILL PREVENT DISCLOSURE OF CONTENTS OR RECONSTRUCTION OF THE DOCUMENT. |
FULL-SCALE AIRCRAFT CABIN
TWIN FLUID, LOW PRESSURE WATER MIST
FIRE SUPPRESSION SYSTEM
PROOF-OF-CONCEPT
EVALUATION
PREPARED & REVIEWED BY: _______________________________
JOSEPH WOLFE (Code 435100B)
APPROVED BY: ___________________________________
DOREEN L. BEHMKE (Code 435000B)
The Naval Air
Systems Command (NAVAIRSYSCOM) would like to acknowledge the following
team members for their individual efforts in making this a dynamic,
successful and safe test series:
In 1998, the
Director for the Office of the Assistant Secretary of the Navy, Installations
and Environment, Safety and Survivability (OASN(I&E)(S&S)),
established a Government and industry forum, named the Aircraft Wiring
and Inert Gas Generator (AWIGG) Working Group, to enhance aviation safety
in government and commercial aviation sectors by the sharing of information
relative to aircraft wiring threats and options for dealing with in-flight
fires. The number of AWIGG participants increased with each meeting,
and technologies at various stages of development were continually presented.
Technologies introduced to the AWIGG that were identified as potentially
viable types of fire suppression systems for use aboard aircraft included
water mist and inert gas/aerosol generator technologies. In order
to investigate the feasibility of such systems for use in suppressing/extinguishing
in-flight fires aboard aircraft, the Director, OASN(I&E)(S&S),
pursued an initiative for a combined DoD/Navy and industry proof-of-concept
evaluation of these type fire suppression systems.
Through liaison
with International Aero Inc., the Office of the Director, OASN(I&E)(S&S),
identified The Memphis Group, which has an aircraft salvage facility
in Greenwood, MS, as a possible source for an aircraft test platform.
Following discussions on the test objectives, The Memphis Group offered
an aircraft and onsite support at its Greenwood facility for the DoD/Navy
and industry proof-of-concept evaluation. The test article selected
was a B-737-200 aircraft, tail number N9022U, that had been removed
from service. The Memphis Group also coordinated the requisite
airfield/fire fighting support for the evaluation with the Greenwood-LeFlore
Airport Manager.
Two (2) water
mist systems were identified for evaluation – a high pressure (
1000psi) system developed by the Naval Research Laboratory for the Navy��s
new Amphibious Transport Dock, the LPD-17, and a twin fluid, low pressure
(
20psi) system developed by the Naval Air Systems Command (NAVAIRSYSCOM),
Aircraft Division, Lakehurst, NJ for aircraft applications. The
high pressure system setup/evaluation was completed by Naval Research
Laboratory, Hughes Associates Inc., and Geo-Centers Inc. personnel.
The low pressure system setup/evaluation was executed by the NAVAIRSYSCOM,
International Aero Inc., Quadelta Inc., and Lectromechanical Design
Company personnel. In addition, NAVAIRSYSCOM, International Aero
Inc., Pyrogen Inc., Quadelta Inc., and Lectromechanical Design Company
provided the requisite material and technical support for demonstrating
the performance of commercially available aerosol generators against
realistic in-flight fire threats in aircraft electronics bays and in
the passenger compartment overhead area between the ceiling and upper
exterior of the fuselage.
The water mist systems and aerosol generator system tests were conducted at The Memphis Group Greenwood-LeFlore Airport facility, near Greenwood, MS, from 1 to 12 May 2000. This report focuses on the twin fluid, low pressure water mist fire suppression system. Results for the high pressure water mist system are provided in a report by the Naval Research Laboratory, and results for the aerosol generator system are provided in a report by NAVAIRSYSCOM, International Aero Inc., Pyrogen Inc., and Quadelta Inc.
APPROACH
The series of twin fluid,
low pressure water mist tests conducted on the Boeing 737-200 aircraft
included five (5) individual threats: lavatory fire, galley fire, seat
cushion fire, overhead stowage bin luggage fire and ceiling wire arc
fire. Before each test for the lavatory, galley, and seat cushion
fires, the system was cycled against a 4-inch diameter methanol cup
fire to identify any desired refinements to the system setup.
Following any final adjustments, each test was conducted in accordance
with the respective prescribed scenario. Each test was structured
to maximize the realism of the fire threat within the constraints of
the test article (reference Figure 1, Test Aircraft and Figure 2, Test
Interior), available test equipment and requisite safety precautions.
In addition, before and after each test, the aircraft was restored to
its original test condition to provide an accurate comparison of tests.
For instance, the lavatory was cleaned out of all debris in between
tests and swept of excess water to prevent an accumulation on the floor.
Also, the seats were covered and the cushions were removed during pre-testing
to insure dry conditions for each test.
Figure 2 - Test Aircraft Interior
The twin fluid, low pressure water mist nozzle was developed by the NAVAIRSYSCOM Lakehurst, NJ, specifically for aircraft applications. Reliability, fire suppression effectiveness, low maintenance requirements, and low weight were all considered as priority factors during system development. The converging-diverging nozzle design utilizes compressed gas (air or nitrogen) to atomize the water stream, achieving highly effective water mist dispersion with high mist momentum at water and gas pressures of approximately 20psi. The simplicity of the nozzle design (large diameter orifices) ensures high reliability and low maintenance requirements, whereas, the low gas/water pressures required for system performance ensure the overall system weight is far less than weights of water mist systems employing higher pressures. This proof-of-concept evaluation proved that the twin fluid, low pressure system can effectively suppress the types of fires featured in the threat scenarios, with minimal impact on existent aircraft interior configurations and only a nominal increase in overall aircraft weight. Aircraft interior configuration constraints were considered for the location selection and mounting of all twin fluid, low pressure nozzles. No interference with any existing aircraft components or equipment occurred. As graphically represented above in Figure 3, there were a total of 13 twin fluid nozzle locations, ten (10) located on four (4) stowage bins, two (2) in the galley and one (1) in the lavatory; the nozzles were distributed throughout the prototype fire test area as they would likely be positioned for an actual aircraft application. Nozzles a through d were mounted on top and at the rear of the overhead stowage bins directed towards the overhead area of the ceiling for fire protection in the area between the interior ceiling and outer fuselage. Nozzles e through j were mounted in the existing space between, and near the bottom of, the overhead stowage bins directed horizontally towards the center aisle for fire protection in the passenger compartment including the overhead stowage bins. Nozzles k and l were mounted on the forward ceiling in the galley area directed slightly downward and to the rear for galley shelf and floor fire protection. Nozzle m was mounted in the lavatory ceiling aimed directly downward. The twin fluid, low pressure nozzles were manufactured out of light weight heat resistant plastics. Since the pressures used for this system are very low, thin wall plastic or rubber tubing was used for the water lines, and standard PVC piping was used for water distribution. The air lines for the system were standard rubber air hose. Compressed gas for the system was provided by an engine driven air compressor, via a pressure reducing valve. A pressurized water tank was used as the water source. Although these sources would be inappropriate for an actual aircraft installation, they proved satisfactory for the proof-of-concept evaluation.
Two (2) thermocouple
trees were used to measure the gas temperatures in the aircraft.
Each tree consisted of seven (7) thermocouples positioned at the following
heights above the floor: 0.3, 0.6, 0.9, 1.2, 1.5, 1.8 and 2.1 m (1,
2, 3, 4, 5, 6, and 7 ft). The trees were located in the aisle
on the starboard side of the plane. Inconel sheathed type K thermocouples
(3.2 mm diameter (Omega Model KMQIN-125G-600)) and Teflon insulated
type K thermocouple wire (Omega Model TT-K-24) were used for this application.
Carbon monoxide, carbon dioxide,
and oxygen concentrations were sampled at two (2) elevations in the
rear of the aircraft. These concentrations were measured in the
aisle adjacent to the aft thermocouple tree. These measurements
were taken 0.5 m above the floor and 0.5 m below the ceiling of the
aircraft. MSA Lira 3000 Analyzers with a full-scale range of 10
percent by volume were used to measure the carbon monoxide concentrations.
MSA Lira 303 Analyzers, with a full-scale range of 25 percent by volume,
were used to measure the carbon dioxide concentrations. Rosemont
755 Analyzers, with a full-scale range of 25 percent by volume, were
used to measure the oxygen concentrations.
The reduction
in visibility (obscuration) resulting from the smoke and mist was measured
at two (2) elevations in the rear of the aircraft. The obscuration
was measured in the aisle adjacent to the aft thermocouple tree.
These measurements were made 0.5 m above the floor and 0.5 m below the
ceiling. The meters were installed with a path length of 0.3 m.
The gas and
water pressures for the twin fluid, low pressure system were measured
at the air and water distribution manifolds with commercial pressure
gauges. Attempts to measure the water flow rate were made using
a Flow Technologies, Inc. paddle wheel type flow meter placed in the
water line leading from the pressurized water tank, which served as
the water source, to the distribution manifold inside the aircraft.
However, because the flow meter had a full-scale range of 0-100 lpm
(2-26 gpm) and an accuracy of 1.0 percent of the measured value, the
measurements were unreliable due to the low flow rates associated with
the twin fluid, low pressure system. However, the water flow rates
were recorded using the Naval Research Laboratory (NRL) data and noted
in the recommendation section of this report.
The lavatory
arson fire threat was the first to be tested in the aircraft.
A system check was completed using a 4-inch pipe cap filled with methanol.
The cap was placed in the middle of the floor and allowed to pre-burn
for 15 seconds before mist activation. Data acquisition was incomplete
because no water flow rates were recorded. However, the cup fire
was extinguished in approximately 1 second when the water mist system
was operating properly.
(note recessed, inconspicuous position)
High pressure system nozzle
Figure 5 - Lavatory Nozzle Locations
Figure 7 - Post Test
The arson threat consisted of ��crumpled�� sections of a newspaper piled on the floor (reference Figure 6). Six (6) ounces of acetone was poured over the newspaper and it was then set on fire. There was a 16-second pre-burn time before the mist was activated for AWIGG Test #9. The fire was immediately under control in 1 second and small pieces of newspaper continued to burn until full extinguishment, which occurred after 36 seconds. The flow rate was recorded at 0.64 gpm for a total of 49 ounces or 1.44 liters of water. The heat release is estimated to have been 1,700-2,000 kW initially, as the acetone (estimated from the equation Q=m*H, where Q is heat release, m is burning rate and H is the heat of combustion) burned off. Then at mist activation, a fire output of 100kW, is estimated as the paper was burning along with any remaining acetone. The results are recorded in Table 1 and shown in Figure 7.
As with the
lavatory test, system checks were completed for the galley fire using
a 4-inch pipe cap filled with methanol. The cap was first placed
on the middle of the shelf in the galley and then on the floor.
The cup was allowed to pre-burn for 15 seconds before mist activation
for each check. Figure 8 illustrates the galley twin fluid low
pressure nozzle locations.
Twin fluid, low pressure nozzles
(mounted on the ceiling next to the forward galley bulkhead, facing down and aft)
Following the system checks, the galley arson threat scenario was conducted. A plastic trash bag filled with paper napkins, styrofoam cups and plastic cups was used as the fuel for the test. The bag was placed on the floor in the galley area and methanol soaked paper kim-wipes were placed on top of the bag and ignited. The bag was allowed to pre-burn for 15 seconds before mist activation. It is estimated that the heat release at the time of mist activation was approximately 300kW. The fire was controlled immediately; however due to combustion occurring inside the plastic bag, full extinguishment took 3 minutes and 47 seconds. In addition to suppressing/extinguishing the fire, the water mist caused a notable decrease in temperature as detected by the aft thermocouple tree.
As illustrated
in Figure 9, a 4-inch pipe cap with methanol was placed on the center
seat and ignited to check the mist distribution throughout the cabin
for the seat cushion test series. After the cup fire was successfully
extinguished, the arson threat condition was initiated.
Figure 9 - Four-Inch Cap on Seat
Figure 10 - Center Seat Cushion Fire Test
The arson seat cushion test
consisted of pouring 6 ounces of acetone on the bottom and back of a
center seat (reference Figure 10). The acetone was poured over
an approximate area of 0.04 m2 for an estimated heat release
of 80kW. The acetone was allowed to soak for a short period of
time while the fire fighter lit the seat on fire. Pre-burn times
of 15 and 30 seconds were tested for AWIGG Tests #23 and #24 where extinguishment
times were recorded as 24 seconds and 17 seconds and total water flows
were 0.98 gallons (3.71 liters) and 0.59 gallons (2.23 liters) respectively.
The results are recorded in Table 1.
It is important to note
that the closest twin fluid nozzle was located across the aisle and
one (1) row forward of the fire (reference Figure 11). The twin
fluid nozzles were positioned slightly downward from the horizontal
for this test series.
For this test
series, an overhead stowage bin was removed from the aircraft and set-up
outside to determine the amount of time required for the fire to destroy
the test article. This was done to determine a fail-safe time
constraint for the safety of the fire fighters and the necessary time
to abort the test if the mist proved unsuccessful in suppressing the
fire. Pictured below in Figure 12 is the stowage bin at the start
of the test. The threat consisted of three (3) nylon gym bags
filled with miscellaneous clothing articles, a paper gift bag filled
with tissue paper, a newspaper, a plastic poncho and methanol soaked
kim-wipes placed on top of the center gym bag that were ignited by the
fire fighter.
Figure 12 - Overhead Stowage Bin Fire Threat Condition
The results of the fire threat are shown in Figure 13 after 2 minutes of burn time.
Figure 13 - Overhead Stowage Bin at the 2-Minute Mark
Total damage
to the unprotected stowage bin was substantial, as shown below in Figure
14, after the fire was extinguished. As can be seen, almost all
of the combustible material was consumed. Although the test team
was aware that outside burning rates would differ from those inside
the aircraft cabin, it was determined that a 15-second pre-burn would
be sufficient.
For this test
series, the twin fluid nozzles mounted on the stowage bins were positioned
slightly upward from the horizontal. The closest twin fluid nozzle
to the stowage bin fire threat was mounted across the aisle and 2.5
feet both forward and aft of the center. Therefore, as in the
seat fire threat, no direct path to the fire existed and no nozzles
were mounted inside the stowage bins (as in the case of the nine (9)
high pressure nozzles mounted in the bins). In this test series, the
fire was immediately controlled; however, it was not totally extinguished
as it was with the high pressure system. A small flame continued
to exist in the center gym bag, the source of the fire. However,
if one observes the damage to the stowage bins in each case, as pictured
below in Figures 15 and 16, it is evident that there is no noticeable
difference in the level of protection provided by the high pressure
and twin fluid, low pressure systems.
Figure
15 - High Pressure Stow Bin Result
Figure 16 - Low Pressure Stow Bin Result
For this test
series, the twin fluid, low pressure nozzle system was the only system
tested. Ceiling panels were removed and aromatic polyamide wire
bundles were taped to the backside of the panel as illustrated in Figures
17 and 18. The panel was replaced in the ceiling and a saltwater
drip was used to start the arc while current was flowing through the
bundle. For the first attempt at extinguishing the arc, the oscilloscope
spike was used to determine when to activate the mist. During
this test, AWIGG #28, no large arc was observed prior to mist activation.
Only small sparks were present before the mist was activated.
After two (2) more attempts (AWIGG Tests #29 and #30) with similar results
the decision was made to visually observe the large arc, rather than
rely on the current surge indicated by the oscilloscope, to determine
when to effect mist activation. Therefore, the bundle used for
Test #28 was reused for Test #33 since it was not damaged. For
this test, the arc was visually observed before the mist was activated.
The water mist kept the wire bundle from breaking apart (from the continuous
arcing) by cooling the arc while the wire bundle remained energized.
Appendix A contains the Lectromec full report.
Figure 17 - Wire Bundles on Backside of Ceiling Panels
FILE NAME and TEST # | MIST SYSTEM | FIRE LOCATION | FIRE TYPE | EXT TIME | FLOWRATE (gpm) | WATER PRESSURE (psi) | TOTAL WATER FLOW (gallons) | TOTAL WATER FLOW (liters) |
AWIGG 01 | HP | LAV | TELL TALE | 0:06 | n/a | n/a | ||
AWIGG 02 | TF | LAV | TELL TALE | 1:02 | n/a | n/a | ||
AWIGG 03 | TF | LAV | TELL TALE | 2:20 | 0.56 | 20 | n/a | |
AWIGG 04 | HP | LAV | TELL TALE | 0:05 | 0.85 | 900 | 0.07 | 0.26 |
AWIGG 05 | HP | LAV | ARSON | 0:08 | 0.85 | 900 | 0.11 | 0.42 |
AWIGG 06 | TF | LAV | TELL TALE | 0:04 | 0.64 | 22 | 0.04 | 0.15 |
AWIGG 07 | TF | LAV | ARSON | 0:58 Reign | 0.5 | 20 | n/a | |
AWIGG 08 | * * * * A B O R T E D * * * * | n/a | ||||||
AWIGG 09 | TF | LAV | ARSON | 0:36 | 0.64 | 22 | 0.38 | 1.44 |
AWIGG 10 | TF | GALLEY SHELF | TELL TALE | 2:16 | 0.875 | 20 | n/a | |
AWIGG 11 | TF | GALLEY FLOOR | TELL TALE | 2:15 | 0.878 | 20 | n/a | |
AWIGG 12 | HP | GALLEY SHELF | TELL TALE | 0:04 | 2.43 | 880 | 0.16 | 0.61 |
AWIGG 13 | HP | GALLEY FLOOR | TELL TALE | 0:21 | 2.14 | 850 | 0.75 | 2.84 |
AWIGG 14 | HP | GALLEY | TRASH BAG | 1:21 | 2.172 | 861 | 2.93 | 11.09 |
AWIGG 15 | TF | GALLEY FLOOR | TELL TALE | 0:30 | 0.715 | 16 | 0.36 | 1.36 |
AWIGG 16 | TF | GALLEY | TRASH BAG | 3:47 | 0.714 | 16 | 2.7 | 10.22 |
AWIGG 17 | HP | SEATS CENTER | TELL TALE | 0:20 | 3.81 | 872 | 1.27 | 4.81 |
AWIGG 18 | HP | SEATS AISLE | TELL TALE | NO | ||||
AWIGG 19 | HP | SEATS 15 SEC | ARSON | 0:05 | 4.149 | 997 | 0.35 | 1.32 |
AWIGG 20 | HP | SEATS 30 SEC | ARSON | 0:06 | 4.077 | 944 | 0.41 | 1.55 |
AWIGG 21 | HP | SEATS SLASHED | ARSON | CONTROLLED | n/a | |||
AWIGG 22 | TF | SEATS CENTER | TELL TALE | 0:22 | 2.023 | 8 | 0.74 | 2.80 |
AWIGG 23 | TF | SEATS 15 SEC | ARSON | 0:24 | 2.451 | 16 | 0.98 | 3.71 |
AWIGG 24 | TF | SEATS 30 SEC | ARSON | 0:17 | 2.086 | 8 | 0.59 | 2.23 |
AWIGG 25 | HP | BIN 15 SEC | LUGGAGE | 1:27 | 1.016 | 902 | 1.47 | 5.56 |
AWIGG 26 | TF | BIN 15 SEC | LUGGAGE | CONTROLLED | 2.428 | 16 | n/a | |
AWIGG 27 | HP | SEATS SLASHED | ARSON | CONTROLLED | n/a | |||
AWIGG 28 | TF | CONCEALED SP | ARC | 1.213 | 20 | n/a | ||
AWIGG 29 | TF | CONCEALED SP | ARC | 1.215 | 20 | n/a | ||
AWIGG 30 | TF | CONCEALED SP | ARC | n/a | 20 | n/a | ||
AWIGG 31 | HP | GALLEY PAPER | ARSON | 0:10 | 4.563 | 810 | 0.76 | 2.88 |
AWIGG 32 | HP | GALLEY | TRASH BAG | 0:15 | 4.633 | 823 | 1.16 | 4.39 |
AWIGG 33 | TF | CONCEALED SP | ARC | 1.15 | 20 | n/a | ||
AWIGG 34 | HP | BIN 120 SEC | LUGGAGE | CONTROLLED | 1.584 | n/a |
The twin fluid,
low pressure nozzles proved highly reliable and maintenance free.
None of the low pressure system components were damaged during shipment
and installation, nor did any damage occur from pre-burn fire, heat,
smoke, or wear. In addition, due to their unique design none of
the twin fluid nozzles became clogged, unlike the high pressure nozzles,
some of which had to be replaced after installation due to clogging.
The installation
of the twin fluid, low pressure water mist system can be accomplished
with minimal changes to existent aircraft interior configurations.
In the passenger compartment, air and water supply lines and nozzles
were installed with no change to the overhead stowage bin configuration
aside from mounting brackets connected to the exterior of the bins.
Nozzles and supply lines in the galley area were mounted/routed on the
forward area of the ceiling for demonstration purposes only. In
an actual installation, the nozzles and supply lines would be mounted
and routed in less conspicuous locations. The nozzle in the lavatory
was mounted in a recess in the overhead that could possibly be used
in actual installations. The greatest impact in interior configuration
would be the installation of the lightweight gas and water supply lines.
The twin fluid,
low pressure water mist system proved highly effective in suppressing
and extinguishing fires during all threat scenarios. Due to the
water droplet size, velocity and dispersion characteristics associated
with the twin fluid nozzles, they do not need to be aimed directly at
the fire to be effective as in the seat cushion and stowage bin fires
where no direct path (line-of-sight) for the nozzles occurred.
Even in cases where the fire was not completely extinguished, the twin
fluid system controlled the fire, limiting the damage, threat, and temperature
increase experienced in the aircraft interior.
The low pressure
associated with the twin fluid system accommodates the use of very lightweight
and flexible material for gas/water supply lines, and obviates the need
for rigid, high pressure metal tubing that not only increases overall
weight but also complicates installation/routing.
Upon review
of the NRL data for the twin fluid system, it was clear that most data
tests were good for flame temperature, operating pressure, and water
flow rate. For the lavatory and galley tests, a few tests did
exhibit slightly elevated temperatures at the 2.1 meter height on the
aft thermocouple tree for obvious reasons. Also, for these tests
there was a slight dip in the transmittance for the optical density
meter. All other data for the twin fluid nozzle tests are inconclusive,
due to the data field being blank, instrumentation not operating correctly,
or not registering any noticeable changes.
It is recommended
that this test series be repeated with the inclusion of airline industry,
Federal Aviation Administration (FAA) and National Transportation Safety
Board (NTSB) professionals, and representatives from any other entities
that could contribute to a follow-up test program leading to final design
and certification of an interior water mist fire suppression system
for all passenger and cargo transport aircraft.
It is recommended
that subsequent water mist fire suppression systems tests be structured
to more accurately reflect the operating parameters that a production
system will have to accommodate:
It is recommended
that the fidelity of instrumentation used for measuring and recording
data be comparable to the ranges of system components used. The
water pressure gage and flow meter used for this test series were acquired
for recording data associated with the high pressure (1,000-2,000psi)
system tested and proved inaccurate for the twin fluid, low pressure
(20psi) system.
It is recommended that overall system weight and viability for both retrofit and new construction applications be included in determining the potential applicability for any prototype system that is evaluated.
AWIGG 737 Fire Suppression Testing
Greenwood, MS
May 2000
Prepared by:
Lectromec Design Co.
Dulles, Virginia
703-481-1233
Introduction
As part of the Aircraft Wiring and Inert Gas Generator (AWIGG) Working Group��s effort to identify viable fire suppression systems for suppressing/extinguishing in-flight fires aboard passenger and transport type aircraft, the Naval Research Laboratory (NRL) and the Naval Air Systems Command (NAVAIRSYSCOM) conducted proof-of-concept Aircraft Fire Suppression Testing at The Memphis Group��s Facility in Greenwood, Mississippi, in early May of 2000. The testing was conducted onboard a retired Boeing 737-200 aircraft. There were three (3) types of fire suppression systems evaluated: high-pressure water mist, low pressure water mist and Pyrogen aerosol generators. Different types of fuel sources for fires were used, including kerosene, methanol, acetone, paper (and other Class A combustibles), and electrical arcing. Lectromechancial Design Company (Lectromec) supported the NAVAIRSYSCOM effort by providing an electrical arc in wire bundles that were mounted in various zones in the aircraft and, where appropriate, providing instrumentation (i.e., oscilloscope records of arcing events).
This report covers the results of these arcing events as observed by Lectromec personnel. Other observations of these tests by other members of the team have not been integrated into this report (other than a still photo taken from John Brooks' of International Aero, Inc. mpeg video). For a more complete understanding of these tests, this report should be integrated with the data and observations outlined in the test reports on the low pressure water mist and Pyrogen aerosol generator systems of the other team members.
FIGURE 1. AWIGG FIRE SUPPRESSION TEST HELD IN GREENWOOD, MS: BOEING 737
Experimental Detail
The test bundles were constructed using wire that was removed from US Coast Guard helicopters, which were being rewired, and was provided by Joel Walker of Quadelta, Inc. The wire type was MIL-W-81381 of various gauges between 16 and 24.
The bundles were made of 20+ wires 2 to 3 feet long. Approximately one-half of the wires were shorted together to be connected to the high side of the power supply and the other half were shorted together to be connected to the low side of the power supply. In one region of the bundle the insulation of several of the wires (both high side and low side wire) was damaged with a razor such that the conductor was exposed.
The power supply used was a single-phase 60 Hz, 4.5kVA arc welder type generator. This replaced a higher capacity generator that became unavailable due to malfunction.
In the experiments initiated with saline drip, an approximate 1 percent by weight salt solution was used.
Results
TEST #1:
Location: Galley module outside of the aircraft
Initiation: Saline drip on pre-damaged wire
Series Resistance: 1.0 Ohm
Number of Wires/Type: 20+/MIL-W-81381 wires
Mitigation system: None
The test bundle was mounted on the Galley module outside of the aircraft at a ~30-degree angle. There was a period of scintillation and then an arc event. The event lasted 3 to 4 seconds with arc appearing to travel toward the source. Examination of the bundle indicated that the event had, in fact, occurred at the cut end of the sample where the water had flowed before dripping to the ground. The arc appeared to have gone out because it had run into a region where the bundle was tightly wrapped with tape.
TEST #2:
Location: Galley module outside of the aircraft
Initiation: Saline drip/carbon-oil mixture on pre-damaged wire
Series Resistance: 1.0 Ohm
Number of Wires/Type: 20+/MIL-W-81381 wires
Mitigation system: None
Because the bundle used in Test # 1 had not arced in the pre-damaged area, the bundle was reused for this test. In this test, the bundle was hung above a pan of a flammable liquid (kerosene) in an attempt to see if the arc would ignite the liquid. Initially a saline drip was used but the sample had been dipped in kerosene, which is an excellent electrical insulator and water repellant. The arc therefore could not be initiated with the saline drip.
A mixture of graphite powder and 3-in-1 oil was then spread onto the pre-damaged area of the bundle and the sample replaced about 1 inch above the kerosene. When power was applied to the sample, smoke was generated for several minutes and then a full arc was developed. The kerosene in the pan was lit immediately. The arc lasted several seconds before extinguishing.
TEST #3:
Location: EE-Bay
Initiation: Carbon-oil mixture on pre-damaged wire
Series Resistance: 1.0 Ohm
Number of Wires/Type: 20+/MIL-W-81381 wires
Mitigation system: Pyrogen Aerosol Generator
The test bundle was hung vertically in the EE-bay with a pan of flammable liquid underneath it. There were several other pans of flammable liquid placed around the EE-bay. Two (2) of the pans were manually set on fire and left to burn for 60 seconds. The test bundle was then energized. The Carbon-oil mixture on the pre-damaged area, generated smoke and some sparking, but a full arc was not developed. After a few minutes, the generators were set off to extinguish the fire in the pan which was manually started.
TEST #4 (AWIGG Test #28):
Location: Ceiling of main cabin rear
Initiation: Saline drip on pre-damaged wire
Series Resistance: 1.0 Ohm
Number of Wires/Type: 20+/MIL-W-81381 wires
Mitigation system: Low-pressure water mist
In this test, the bundle was mounted horizontally on the topside of a ceiling panel (Figure 2A) in the main cabin. A saline drip onto the damaged area was begun and the bundle energized. There were several minutes of scintillations and flashing (sparking), but a full arc did not develop. Several flashes in a row were interpreted as the beginning of an arc event and the water mist system was triggered. No further flashing was observed for several minutes. It is possible the water mist did prevent the arc from developing as the saline drip normally results in the sequence of: scintillations, flashing and then arcing. The arcing stage did not occur for this sample. Examination of the damaged area revealed little or no further damage caused by the electrical activity.
FIGURE 2A: TEST BUNDLES MOUNTED ON CEILING PANEL OF MAIN CABIN PRIOR TO THE
PANEL REPLACEMENT
TEST # 5 (AWIGG Test #29):
Location: Ceiling of main cabin rear
Initiation: Carbon-oil mixture on pre-damaged wire
Series Resistance: 1.0 Ohm
Number of Wires/Type: 20+/MIL-W-81381 wires
Mitigation system: Low-pressure water mist
This wire bundle was mounted parallel to the bundle from TEST #5 on the topside of the cabin ceiling. A carbon-oil mixture on the pre-damaged wire was used for initiation. The current was monitored after the bundle was energized and the water mist triggered when the current began to increase. Unfortunately, a full arc had not developed and did not develop after the mist was deployed. Again, it is difficult to determine if the mist prevented development of the arc.
FIGURE 2B: PHOTO OF CEILING PANEL IN MAIN CABIN AFTER REPLACEMENT AND PRIOR TO TESTS
TEST #6 (AWIGG Test #33):
Location: Ceiling of main cabin rear
Initiation: Saline drip on pre-damaged wire
Series Resistance: 1.0 Ohm
Number of Wires/Type: 20+/MIL-W-81381 wires
Mitigation system: Low-pressure water mist
Because the bundle in TEST #4 had not been damaged, it was decided to rerun the test making sure that the arc was well established before triggering the mist. The saline drip was begun and the bundle energized. There was flashing and then a full arc developed. Figure 3 depicts the beginning of the arc with the voltage showing the classic arc flat top waveform and the current reaching 40 amp peaks. Figure 4 depicts the electrical current throughout the arcing event. Between 6 and 7 seconds (5 seconds after the arc began), the arc is extinguished and it appeared that the wires had shorted together producing a sinusoidal current of 20 amperes. The water mist was triggered sometime after the start of the arc, but the affect on the arc is uncertain at this time. The video and oscilloscope recordings can be correlated to shed further light on the situation. Figure 5 shows the damage done to the bundle during the arc event.
FIGURE 3. TEST #6 (AWIGG TEST #33): VOLTAGE AND CURRENT WAVEFORMS AT THE START
OF THE ARC EVENT
FIGURE 4. TEST #6 (AWIGG TEST #33): CURRENT DURING THE ARCING EVENT
FIGURE 5. TEST #6 (AWIGG TEST #33): DAMAGE DONE TO BUNDLE BY ARCING EVENT
TEST #7:
Location: Galley module outside of the aircraft
Initiation: Saline drip on pre-damaged wire
Series Resistance: 0.5 Ohm
Number of Wires/Type: 20+/MIL-W-81381 wires
Mitigation system: Pyrogen Aerosol Generator
This test was conducted with the sample mounted on the inside of the Galley module that was sealed closed with duct tape (see Figure 6). After the sample was energized, scintillations began and an arc developed. The generator was deployed and the arcing was extinguished. Periodically the sample began to scintillate, sometimes strongly, for several seconds, but a strong arc was not observed.
The oscilloscope recording of the voltage (Figure 8) agrees with the visual observations (the electrical current data file was corrupted and the data could not be recovered). From the shape of the waveform, one can see that the arc began a little after the 85-second mark. The generator was deployed ~85.75 seconds and the arc form changed to one resembling a short circuit by the 86-second mark. Perhaps as the arc extinguished the molten metal solidified, welding two (2) wires together. Further in the oscilloscope record, short periods of electrical activity such as scintillations or flashes were observed, but no strong arcing was evident.
FIGURE 6. TEST #7: SAMPLE MOUNTED IN THE GALLEY PRIOR TO THE TEST
FIGURE 7. TEST #7: INITIAL ARC BEFORE THE GENERATOR WAS DEPLOYED
(CAPTURED FROM MPEG VIDEO)
FIGURE 8. TEST #7: VOLTAGE WAVEFORM OF THE INITIAL ARCING EVENT WITH
GENERATOR TRIGGER VOLTAGE ALSO RECORDED
All Rights Reserved Powered by Free Document Search and Download
Copyright © 2011