How safe is your airplane?

After the crash of American Airlines Flight 587, some pilots requested that all Airbus A300 planes be grounded. But they're still aloft.

Published March 8, 2002 8:30PM (EST)

A group of pilots at American Airlines, the world's largest air carrier, have been rallying for the grounding of American's fleet of Airbus A300s. Their concerns follow the mysterious crash of an American A300-600 after takeoff from John F. Kennedy Airport on Nov. 12, in which 265 people died. A letter of protest circulated in Miami, New York and Boston, the three stations at which American bases its 34-strong fleet of the European-built wide-body jet, and at least five dozen pilots added their signatures.

American Flight 587, en route from JFK to Santo Domingo, Dominican Republic, crashed moments after takeoff when, to put it coarsely but accurately, the tail fell off. Why this happened is not known, but the more investigators learn about Flight 587, the more its rudder is coming under scrutiny. The rudder is the large moveable surface attached to an aircraft's tail, controlling side-to-side movement, or yawing, along the plane's vertical axis.

The newest evidence suggests that extreme movements of the rudder, caused either by malfunction or induced inadvertently by the crew, may have stressed the entire tail to the point of failure, possibly with help from a small crack in the tail's composite structure, which had formed years ago but was heretofore undetected. Black box analysis shows marked oscillations of Flight 587's rudder prior to the crash. Whether these oscillations were the reason for the fatal tail separation, or merely a symptom of a greater malfunction, is under study, as is the theory that a structural weakness -- inherent or otherwise -- was exacerbated by an encounter with the wake of a 747 the Airbus was following.

The Allied Pilots Association (APA), which is the collective bargaining agent for American Airlines pilots, has not endorsed its members' outcry. After the initial petition, opinions have simmered but little progress has been made toward an actual grounding of the fleet. However, follow-up discoveries of rudder and tail problems on other A300s have rekindled unease.

An American A300 discontinued a flight from Miami to Caracas, Venezuela, when it experienced erratic fishtailing after takeoff. The captain returned to Miami and the airplane was examined. The following day, the fishtailing occurred again. American insists there is no connection between the Caracas flight and the crash of Flight 587, but nonetheless it took the jet out of service for more serious examination.

Another American A300 returned to Lima, Peru, just after departure, for a nearly identical problem. Control units from both rudders were sent to Europe for analysis by their manufacturer.

And most recently, a bent actuator rod was found in the rudder of a Federal Express A300 during a routine check at the carrier's hub in Memphis. This is considered a highly unusual discovery. Federal Aviation Administration officials were sent to determine if any connection could be made between the damaged actuator and what befell Flight 587.

But what does this seeming pathology really indicate? In the aftermath of air crashes, there is often a rush to judgment over what appear to be latent dangers in our flying machines.

"Are we completely comfortable putting our friends and family on an A300?" the petition at American asked. "If the answer to that question is not a resounding yes, then logic would lead a well-trained pilot to conclude that no one should be flying on them either."

That friends and family invocation is something pilots frequently employ when making a point about safety. If a pilot balks, the inference goes, at loading his mother or best friend aboard an airplane, then he must have a point.

Almost never, however, do the pilots put themselves into the equation. While we shouldn't slight such a chivalric and respectful gesture from our polyester-clad professionals, it might be a more jarring statement if the pilot, not just his loved one, were missing from the guinea pig seat. But examples of pilots -- at least en masse -- refusing to fly airplanes are extremely rare.

And, frankly, for good reason. We needn't launch into statistical rigmarole to illustrate the remarkable safety record of commercial flying, and even in these times of daredevil skyjackings and en route angst, not even the most brazen Vegas renegade would put so much as a nickel on the odds of a plane going down. Delving into the specifics of particular airplane types, not a whole lot changes. This or that model of airliner is occasionally cited as having "one of the best safety records." But when the stats are boiled down, every commercial airliner out there can make essentially the same boast, including, yes, the A300. Which is safest among all that are safe? When crashes are measured in the single digits, out of millions of departures each year, one or two unfortunate coincidences can throw a misleading spin on the résumé of a certain Airbus or Boeing.

But just because we are safe doesn't mean we cannot be safer. And while not outright refusing to fly their A300s, and perhaps giving Granny or a next-door neighbor the grudging go-ahead as well, the American pilots have legitimate concerns about November's disaster. Whether or not those concerns justify a grounding is another matter, but their nervousness is well-taken. As I'm sure the worried pilots are aware, the forced grounding of airplanes is not unprecedented.

On May 25, 1979, an American Airlines DC-10 crashed on takeoff at Chicago's O'Hare Airport. It was later determined that a crack in an engine pylon, where the plane's giant turbofans are bolted to the wing, had caused separation of the engine during the takeoff roll. Detached, the power plant bounced across the wing, causing serious damage to flight controls and subsequent loss of the aircraft. The 273 fatalities that sunny afternoon still represent the all-time worst aviation disaster on U.S. soil. And when, in the weeks that followed, additional cracks were discovered in more DC-10s, the FAA ordered the temporary grounding of the entire DC-10 fleet. American, United and Northwest were all major operators of the type. Even foreign carriers were not allowed to bring their DC-10s into the United States.

More recently, the crash of an Air France Concorde in July 2000 was linked to the layout of the aircraft's fuel tanks. The plane had struck debris on the runway, and an exploding tire caused a tank to rupture. European authorities -- the British Civil Aviation Authority and the Direction Generale de L'aviation Civile or DGAC, of France -- revoked Concorde's airworthiness certificate, and it remained grounded for 15 months.

It has been more than 20 years, however, since the DC-10 fiasco. And the entire Concorde fleet involved barely a dozen aircraft carrying an extremely limited -- and exclusive -- share of the world's passengers. Some will argue the system does not always react with the best interest of the traveling public in mind. The FAA has long been accused of employing a tombstone mentality, submitting to the economic concerns of airlines until some post-disaster fallout forces it to do otherwise. Critics -- and perhaps cynics -- will refer to the saga of the Boeing 737, an effectively ongoing case study of an allegedly defective aircraft allowed to fly.

There are almost 4,000 737s operating today throughout the world, and the aircraft -- a small, twin-engine narrow-body -- is the bestselling jetliner of all time. But the 737 was aloft for years with a known rudder problem. This problem is believed to have caused at least two fatal accidents: the 1994 crash of a USAir flight near Pittsburgh in which 132 people died, and a United flight at Colorado Springs in 1991 that killed 25. At least two nonfatal incidents occurred as well, in which aircraft became nearly uncontrollable when the rudder malfunctioned during flight.

Beginning in 1996, the National Transportation Safety Board (NTSB), which independently investigates accidents before forwarding recommendations to the FAA, requested no fewer than 22 changes be made to the 737 rudder. Eventually the FAA ordered the complete redesign of the rudder's control units, and today all U.S.-registered 737s have been modified. But the aircraft was never grounded. The safety enhancements were applied over time, progressively, while the FAA, Boeing and the airlines worked together to solve the problem.

The FAA, which postures as the altruistic vanguard of the skies, will be loath to admit it, but in truth there is not, nor will there be, a zero-tolerance policy when it comes to air safety. There is, uncomfortable as the flying public might be with the idea, an allowable level of jeopardy, an acceptable threshold of disaster. The risks are quite small, but they do exist.

In the story of the 737, a sensible, risk-evaluated approach seems to have worked. While the needed upgrades were being hashed out, there were no additional accidents. The FAA took that chance. And they will likely take it again in the matter of the A300.

Call it a gamble if you will, but such is the example in many aspects of life, where threats are calculated and plotted, and disasters, while avoided at great cost, are both anticipated and accepted. The FAA has dragged its feet at times (cargo compartment fire protection), wasted its efforts on smokescreen issues (pilot drug testing), and blatantly pandered to industry on others (crew rest/duty time regulations). But to its credit, the agency also works hard to keep the numbers strongly in your favor, albeit with some needed prodding, often by pilot unions.

American is the only current operator of passenger-carrying A300s in the United States. FedEx and UPS fly them as freighters, and DHL flies a cargo version of an older model. But the grounding of any aircraft type, whether common like the 737 or rare like the A300, is not as simple as recalling a Ford Pinto. Neither is it statistically warranted, usually.

The American pilots cite the lack of a known cause of the Flight 587 disaster in their challenge to ground the A300. We cannot fault their agitation, but a single unsolved (so far) accident does not warrant grounding. Further, there is inadequate evidence linking the more recent A300 incidents with November's crash. With regard to the Caracas and Lima flights, American blames engine surging, not an erratic rudder.

In a first step, regulators are urging airlines to enhance their pilot training programs, recommending they now include procedures to avoid rapid or extreme rudder deflections. "Rudder inputs by pilots can cause catastrophic failure," said NTSB Chairman Marion Blakey. "Full rudder inputs can jeopardize the safety of a vertical tail fin." Pilots will chuckle at the idea of being dragged back to class to learn what is already patently obvious, as a pilot pushing a rudder to full-scale deflection is no different from a driver on the highway suddenly yanking the steering wheel 90 degrees. What many pilots may not be aware of, however, is exactly how sensitive a particular rudder's control units are to a pilot's inputs in a specific regime of flight. For now, with little else to go on, a little extra work in the simulator certainly can't hurt.

But pilots and passengers alike should realize that the uncertainty surrounding the A300 does not, by itself, point to a likelihood of dangerous tails waiting to snap off. Airbus and its customers should -- and must -- devise a way to further ensure integrity of its aircraft, but remember that even more proven hazards have not raised such pilot ire in the past, nor were they cause for widespread concern or apprehension. Consider again the 737, or, for that matter, the explosions of empty fuel tanks aboard at least three aircraft, one of which caused the infamous destruction of TWA Flight 800 in 1996. Planes occasionally crash and will continue to do so.

The annals of commercial aviation -- more or less an 80-year history -- are full of inexplicable accidents, a fact, however frustrating, inherent in the evolution of technology and safety. We should learn to be more comfortable with this. For in spite of such mysteries, the statistics remain firmly on your side, by a wide enough margin that none of us should be dissuaded from taking to the skies in virtually any airliner.


By P. Smith

MORE FROM P. Smith


Related Topics ------------------------------------------

Air Travel