| Where There's Smoke, There's Not Always Fire: A New Approach to Fire Detection
 |
| Fire tests were conducted in the
FAA Cargo Compartment Fire Test Facility. |
 |
| Fire detection tests were conducted
with this micro scale carbon monoxide sensor that uses
a nanocrystalline tin oxide detector (center). |
 |
| This work shows the advantage of
using very different types of sensor technology to
gain a better understanding of the environment. A
microscale classifier is held next to a traditional
macroscale classifier. |
NASA Glenn Research Center has successfully tested a new
design concept for fire detection in commercial airliners. These
aircraft are equipped with fire detectors and extinguishers
in the baggage and cargo compartments. The current
detectors sense the smoke particles coming from a fire. But,
when they sense smoke, there is not always fire. The
sensors can be fooled by dust and other aerosols that sometimes
form in the cargo compartments. The smoke detectors can mistakenly
signal the pilots to needlessly initiate emergency procedures
for a fire. NASA’s newly tested concept is a
multi-sensor package built to read a wider picture of fire
emissions, not just smoke. NASA’s detector looks
for increased concentrations of combustion gases along with
the smoke, thus reading a more complete fire signature. The
multi-sensor package compares various gas concentrations
and smoke particle sizes to those values characteristic of
an actual fire. The result is a fire detector system
that works to effectively recognize the presence of fire
while screening out false alarms.
NASA Glenn applied its expertise in MicroElectroMechanical Systems (MEMS)
chemical sensors to this new fire detection concept in response to the significant
number of false alarms generated by the current fire detectors used on aircraft. Current
fire detectors use either photoelectric or ionization sensor to detect the
smoke particles produced by a fire. These sensors work well at detecting
the presence of fire, but it has been estimated that as many as 200 false alarms
occur for every actual fire. In-flight fires are rare events. However,
with no way for the pilot or crew to visually verify the presence of a fire
in the under-floor cargo compartment, any alarm has to be treated as an actual
fire. The pilots must discharge on-board extinguishing equipment, declare
emergency priority over other air traffic, and, as soon as possible land the
aircraft at the nearest airport. Prior to takeoff, a false alarm will require
that the pilots return the aircraft to the gate and delay the flight until
mechanics can repair the problem.
Designing a fire detection system for an aircraft cargo
compartment is a challenge. Unlike the conditions
seen by the smoke detector in your house, aircraft cargo
compartments present an array of challenges from widely
varying humidity and air temperatures and pressures causing
condensation, to dust, dirt and aerosols from the cargo
ramp environment disturbed by turbulence and vibration.
Tests recently conducted at the Federal Aviation Administration
(FAA) Cargo Compartment Fire Testing Facility in Atlantic
City, NJ, successfully demonstrated the miniaturized, combined
chemical and particle sensors approach. First, the
multi-sensors were subjected to dust and high humidity
levels which would regularly cause conventional smoke detectors
used in cargo bays to false alarm. Over a series of exposures,
the NASA MEMS multi-sensor package had a zero false alarm
rate. Then, the MEMS multi-sensor package was mounted alongside
conventional smoke detectors in the facility’s Boeing
707 under-floor cargo bay. Over the entire test series,
the multi-sensor package sensed the onset of actual fire
equally as well as the conventional smoke detectors.
The MEMS sensor system developed by NASA Glenn includes
a miniaturized CO and CO2 sensors, a smoke particle detector,
and integrated software. The on-board processor evaluates
the response of the complete sensor system and gives an
indication of whether there was a fire condition. Interpreting
the fire signature has been shown to be an effective approach
to fire detection while minimizing false alarms.
The development and use of MEMS-based chemical sensing
and particulate technology for fire detection has been
on-going for a number of years as part of a joint NASA-FAA
program. NASA and the FAA have also funded and managed
development of analytic fire models by Sandia National
Labs. Sandia’s modeling tool is intended for
fire detection system designers to help determine effective
sensor locations and determine effective set points to
activate cockpit warnings. The tool is also intended to
replace, where possible, the need for destructive testing
using actual fire to support design and certification.
These activities have also been coordinated with NASA
developments related to safety of its International Space
Station. False alarms are also of concern in Space Station
operations and for future Exploration Missions. Possible
use of this new fire detection approach is being considered
for those applications.
These tests have shown the feasibility of an integrated
multisensor fire sensor system. There are still possible
improvements needed to the sensors, packaging and software/interface,
long term durability and drift effects of the sensors, and
FAA certification criteria for these new multi-sensor detectors.
But these tests have successfully demonstrated a new generation
of aircraft cargo compartment smoke detectors that look for
the fire’s comprehensive signature rather than just
its smoke.
For more information, contact:
Gary Hunter
Gary.W.Hunter@nasa.gov
|
)
