Black boxes on commercial airplanes have always had three main requirements: to be tough, smart, and lightweight. But the next generation has one more demand—to fly.
The classic black box is composed of three parts: a cockpit voice recorder and flight data recorder, and the flight data acquisition unit (FDAU), which feeds them information. The data is stored in the two recorders, which are designed to withstand the horrific forces of a plane crash. Mixed in the twisted wreckage, often on the ocean floor, they emit audible "pings," so rescue crews can locate the valuable information.
But they have such a weak homing signal that rescuers must already know about where a box is.
Now there's another way to preserve the same data, and make it easier to find: A black box called a DFIRS (deployable flight incident recorder set) is designed to eject from a crashing plane and flutter to earth on its own, like a leaf falling from a tree. The unit lands near the wreckage, floating if the crash was over water, and emits a strong radio locator signal for rescue crews to track.
In fact, such units have been in use for decades, mostly in military planes and helicopters. They've even been optional for commercial planes. But higher cost has restricted their application to about 4,000 units in use today.
Now DRS Flight Safety and Communications Corp. (Ottawa, Canada), is making a new generation of its DFIRS for U.S. Air Force electronic surveillance aircraft. And this new model is designed with the price and safety regulations for commercial planes.
"The idea of an emergency locator with a transmitter goes back 30-odd years, so it's an idea that's been around an awfully long time," says division president David Stapley. "Now we're bringing the technology into the 21st century."
Materials. "The key challenge to date is that the airfoil can't be too large or too heavy, because it has to react to aerodynamic forces such that it achieves a low terminal velocity," says Robert Austin, a systems engineer at DRS. "And because it's on an aircraft it must be light."
|
Once it deploys, the wedge-shaped DFIRS airfoil is designed to be clumsy in flight, so it impacts with a low terminal velocity.
|
The testing of flight data recorders is legendary. It puts boot camp to shame. Heck, it puts crash-test dummies to shame. First the unit has to survive an impact of 3,500 g within 6.5 msec., delivered by firing it from a 75-ft cannon with a 10-inch barrel into a wooden wall. Then it's got to deal with the impact of a 500-lb weight dropped on a 1/4-in spike from 10 feet. Next, it must withstand the crushing pressure of a 5,000-lb weight. Then it is baked at 1,100C for 60 minutes. And finally, the testers pressurize it as if were sitting 20,000 ft deep in the ocean for 30 days.
The DFIRS withstands this abuse with a three-layered shell. The outer layer includes carbon fiber and Kevlar, to ensure the overall shear strength of the airfoil. Then, a foam and plastic layer provides impact and crash strength, decelerating the sensitive electronic contents. And the shell's third layer contains microporous solids and aerogel, to provide thermal insulation.
Why not just make it out of stainless steel and be done with it? The answer is electronics—
Electronics. "There are conflicting design challenges, such as packaging the memory in a nonmetallic enclosure that does not interfere with the sensitive antennas of the ELT (emergency location transmitter)," Austin says. "You have to make sure it meets those requirements, and that it can sit on the outside of an airplane for 20 years, then work perfectly under the worst possible conditions."
So the box will not muffle the ELT transmissions, the entire package must have minimal metal, and yet be as strong as a bank vault.
|
|
Pending changes in flight data recorder safety standards
|
|
| Change |
Planned Implementation Date |
| Recording of digital communications |
1/1/05 |
| Self-contained 10 min. backup power supply for CVR area channel |
est. 2005 |
| Two hours CVR standard (of data) for new aircraft |
1/1/03 |
| Video recording capability |
under review |
| Magnetic tape recorders phased out |
2005 |
| Use of Dual Combi recorders on new aircraft (less video) |
1/1/03 |
| Retrofit of all aircraft for Dual Combi recorders (less video) |
1/1/05 |
|
Source: DRS
|
Also, new regulations require a great increase in the amount of data stored, plus a combined voice and data recorder, instead of the current system of separate boxes. Impossible? Until recently, it was.
"Just 10 years ago, flash memory was too large," Austin says. "But the advancement in memory modules has allowed us to do it. It's getting to the point where you can store the required data on four or five chips."
DRS designers now use 128 Mbit Intel chips, allowing them to radically shrink the size of the electronics package. Another way they achieve this laundry list of demands is by having the boxes draw data directly from the central data bus, instead of filtering the data through an FDAU.
Now that the electronics can be shrunk down, and still emit nearly no EMI, you would think the engineers' job was done. But increasingly strict safety regulations keep raising the bar:
the electronic requirements demand that it record 25 hours of data covering 88 aircraft functions, often resulting in a storage rate of several hundred parameters per sec.
the new generation includes emergency locator beacons, whose 406 MHz transmission allows rescue crews to pinpoint the unit within 30m, compared to the 25-mile diameter circle defined by the old-fashioned 121.5 MHz units (and the "pingers" transmit underwater at just 37.5 KHz).
the plane's registration "tail number" is included in that position message, which cuts down on false alarms and allows crews to confirm crashes sooner.
boxes must be compliant with current aircraft, so they can be integrated into their systems without rewiring the entire plane.
The FAA has always required that commercial planes carry two boxes—the cockpit voice recorder that records radio transmissions, pilots' voices, and engine noises; and the flight data recorder that tracks parameters such as altitude, airspeed, and heading, according to the National Transportation Safety Board (NTSB). But the DFIRS could re-cord all this data in a single box, so wiring two units to each plane (known as a "dual combi") would provide a backup.
Sensors. The next step is the deployment—how does the unit "know" when to flip off a crashing plane? On current models, there are four sensors on the skin of the aircraft (at the nose, tail, and in each wingtip), as well as a hydrostatic sensor in the tail, to detect water landings. When a plane flies straight into the side of a mountain, it takes about 25 msec for the tail to reach the spot where the nose first impacted, Austin explains. So the four sensors are set to release the unit within 15 msec.
|
|
Comparison of flight data recorders– deployable vs. fixed on-board
|
|
| Deployables |
Fixed On-board |
| Radio beacon locator compatibility |
Underwater pinger only; no locator on land |
Ease of recovery on land; survives impact away from wreckage |
Requires more time to remove from wreckage |
| Ease of recovery on water; airfoil floats |
High cost of underwater recovery, if located |
| Weight advantage; lighter |
Weight disadvantage; heavier |
|
Source: DRS
|
The time demands are fixed, but DRS
has evolved its deployment-trigger technology three times. The original sensor
used in the 60s is similar to that used in an automotive airbag—"It's just a
ball bearing on a spring," says Austin. "It's a very simple sensor, but it could
cause inadvertent deployment."
So in the mid-70s, DRS moved to frangible switches, which are programmed to monitor the plane's walls, so they deploy when an aircraft's skin crumples during a crash. And for the next model, DRS plans to use accelerometers.
|
Ejected 50 to 70 mm off the plane's tail section by a small spring, the DFIRS airfoil is carried to a soft landing by its 'falling leaf' design, causing it to flutter through the airflow.
|
When it actually deploys, the unit is designed to tumble like a leaf from a tree, achieving a terminal velocity of 75 ft/sec, about 50 mph. By comparison, an F-18 flying straight down can go into the ground at several hundred mph, says Austin.
The last problem for DRS engineers who are trying to design the perfect sensor is waiting for the opportunity to see if they work.
"We can design them," says Austin. "But it's very difficult and expensive to find someone who routinely crashes aircraft, so we can test them out."
Aerodynamics.
The company makes two types of airfoil for the DFIRS. One is an 18-inch square, 3-inch deep wedge-shaped module. The other type is round and bowl-shaped, and about the same size. The trick to their performance is that they're both de-signed to avoid flying with the stability of a spinning frisbee, but rather as a tumbling airfoil in a clumsy flop around the "wrong" axis.
The DRS engineers balance all these design demands in 3D CAD through PTC's Pro/ENGINEER Foundation program. They send the CAD files to Algor's Professional Mech Engineer to run a vibration analysis—for shock and impact performance—and then a thermal analysis. And finally, they transfer the results back to Pro/ENGINEER, and then to CAM for CNC milling.
But the power of all this simulation is useless in trying to model the aerodynamics.
"And fighter aircraft are even harder to model because there's a more dynamic airflow over the wing than the consistent, laminar flow on a commercial jet," Austin says. "You just have to crash some airplanes, like they crash cars to test airbags."
The modeling isn't made any easier since the deployable boxes are located in a different place on each type of plane. On Air Force One, a modified 747, a DRS ELT is located at the base of the tail, which is difficult to analyze because that's where the air splits around the aircraft. On an F-18 fighter jet, it's on the engine deck between the vertical tails, so it actually deploys upwards. And on a helicopter, it's on the stem of the tail, on the opposite side from the tail rotor.
The units are also attached differently to these different aircraft. On Air Force One and the AWACS E-3 (similar to a 707), it's mounted flush to the skin. And on high-speed fighter jets, it's actually mounted beneath the skin, below a very thin aluminum patch.
Future use. But how about the bottom line—do they work? DRS says their units have a 95% rate of successful deployment (where the boxes fly off the planes when they should), and a 100% rate of recovered data regardless.
Other applications DRS is looking at applications in voyage data recorders for ships, says David Stapley, president of DRS Flight Safety and Communications Corp. "Deployable recorder technology can be used on any platform from which data survival and recovery are essential."
|
The current version, the EAS3000, hit the market in 1997 and is currently used on E.H. 101 military helicopters in Canada, Britain, and Italy, and in Japan by the Tokyo Metropolitan Police. And an earlier version for the F-18 has been used since the mid '80s. Deployable ELTs such as this are a safety requirement on all helicopters flying passengers over the North Atlantic.
The next generation is the DFR2100, due on the market by 2002, Stapley says. A lead-in version is being developed for use on the U.S. Air Force's RC-135 aircraft. DRS also has orders to supply them for use on the Navy's new F-18E/F Super Hornet. And DFIRS are already in service on German Tornados and the F-18C/Ds of the U.S. Navy, Kuwait, Finland, Switzerland, and Malaysia.