It takes a lot more than a few puffs of cigarette smoke to verify proper operation of the smoke detectors used in the cargo bays of the Boeing 777 airplane. To verify smoke levels in the 2% to 4 % obscuration per foot (%/ft) range, it can be very difficult to devise a test solution that generates these precise levels of smoke at an airline’s maintenance facilities, where space is limited.
The rate of air flowing through the detector also must be controlled and monitored. This was the problem the BFGoodrich Test Systems Division faced when designing a general-purpose VXI-based test station, the IRIS 7000 ATE, to test numerous line replaceable units (LRU) used on the Boeing 777. Today, the system is operational at the Component Maintenance Center of All Nippon Airways (ANA) in Tokyo.
This LRU, the cargo smoke detector (CSD) that discovers smoke in the air, is manufactured by AlliedSignal Aerospace. The CSD discriminates smoke from other nonsmoke aerosols and triggers an alarm when it detects smoke in sufficient concentration.
At AlliedSignal Aerospace’s manufacturing facility in Tucson, AZ, an 80-ft3 box contains the smoke generated by burning 10 punks (rotted wood). The smoke from this box is fed in controlled amounts to a 50-gallon aquarium that contains a laser detection system to measure the obscuration level of smoke in the air injected into the CSD. AlliedSignal calibrates the smoke-generation source in accordance with UL268, Smoke Detectors for Fire Protective Signaling Systems.
While AlliedSignal requires a dedicated ATE to daily test CSDs to support production schedules, ANA needs a general-purpose ATE to test the CSDs much less frequently, perhaps only once a month. Replicating AlliedSignal’s smoke-generation source at ANA’s facilities is prohibitive for three reasons:
Space—this system requires approximately 100 ft2 of floor space.
Smoke—this system generates so much smoke that special vent hoods and air handlers are required.
Cost—the costs of duplicating the system and the air-handling equipment is outside the scope of this project.
Designing and building a compact smoke generator would consume time and money the project could not afford. The solution is integrating an off-the-shelf smoke generator.
The smoke generator has a small chamber (approximately 1 ft3) to generate and control smoke. Burning punk sticks are inserted into a small fire box on the side of the chamber. Then a solenoid smoke gate turns on and off so the proper amount of smoke is drawn into the chamber.
Figure 1 shows the smoke generator connected to the rest of the system.
Fans 1 and 2 in the CSD pull air in through the four inlet ports, through its internal detector circuitry and out its exhaust.
Airflow on the inlet ports is controlled by valves 1 through 4 to simulate use of the CSD in the airplane. Flow meters 1 through 4 are high-accuracy vortex-sensing flow meters that have low pressure drop. The meters output a voltage (0 to 5 VDC) proportional to the airflow rate (0.25 to 10 CFM).
When valve 7 is closed and valve 6 is open, a closed-loop condition recirculates smoke through the system. When valve 7 opens and valve 6 closes, the exhaust hose on the fans is manually removed and fresh air comes in through valve 7 and out through the fans on the rear of the CSD. The manifold is a machined block of aluminum that evenly distributes air and smoke to the four input ports of the CSD.
All of this equipment, including the copper tubing to connect the pieces together, mounts in a 3′ × 3′ × 4′ cabinet with heavy-duty casters on the bottom. Then the smoke-generation system can be rolled up next to the IRIS 7000 ATE to test the CSD. The front panel of the cabinet is at a 70º angle so the operator can easily access the valve controls and airflow-rate displays.
A cable for all of the electrical connections provides a conduit between the IRIS 7000 ATE and the CSD mounted in the smoke-generation system. Figure 2 illustrates the basic electrical diagram for testing the CSD.
Two redundant electronic channels in the CSD improve reliability. Test commands, responses and smoke-alarm indications transmit over the ARINC 429 serial data communications bus.
A slave LRU to the CSD is the remote smoke detector (RSD) located in the electronic equipment bay of the 777 airplane. The RSD receives its power and commands from the CSD. The RSD loads simulate an RSD connected to the CSD during test.
AlliedSignal’s test specification requires that the CSD triggers an alarm when the smoke density of the air reaches 3 ±1% obscuration/ft. It also needs an accuracy tolerance of ±0.25% obscuration/ft for the smoke injected into the CSD.
A calibrated smoke detector installed in the system acts as a transfer standard (Figure 1). It determines the offset of the system for that test run and adjusts the smoke controls appropriately. The system injects smoke at the 2% obscuration/ft level to verify that the CSD does not trigger the alarm. Then the system injects smoke at the 4% obscuration/ft level to verify that the CSD does trigger the alarm.
Although our system generates a smaller amount of smoke than the AlliedSignal’s system, the smell is annoying. To combat this problem, we use a five-way electronic air cleaner to purify the air inside the smoke-generation system cabinet. We also line the interior surfaces of the cabinet with sound-absorbing insulation that helps attenuate the noise generated by the loud fans in the CSD.
The end product is a BFGoodrich Test Systems Division smoke-generation system that supplies the precise levels of smoke needed to properly test the CSD. This portable, compact system minimizes the noise and smell produced when verifying that a CSD operates per AlliedSignal’s test specifications.
About the Author
Paul O. Herrmann IV is a principal engineer at BFGoodrich Test Systems Division, formerly JcAIR. He received a B.S.E.E. degree from the University of Missouri and has 14 years experience in designing test systems. BFGoodrich Test Systems Division, 400 New Century Parkway, New Century, KS 66031, (913) 764-2452.
Copyright 1997 Nelson Publishing Inc.
April 1997