Air turbulence eyed as possible cause of plane crash

Shortly after takeoff, American Airlines flight 587 started to shake and rattle. Within seconds, the pilots experienced a sudden loss of control before the plane plummeted to the ground, killing 260 people.

Now, investigators are focusing on whether the American Airbus 300 was hit by air turbulence from a Japan Airlines jumbo jet that had left just before it and was some four miles upwind.

However, aviation experts caution that accidents are almost always caused by a chain of events. In this case, the National Transportation Safety Board (NTSB) is also looking closely at the A300's tail fin. The fin, which is made of composite materials such as graphite and plastics, looks as if it was snapped off.

"If that is what's going on, this is going to be a really significant accident," says one aviation expert. "Composites are very, very strong, and they're very, very light, and they're used more and more in thousands of airplanes."

Although the investigation into the crash is expected to take months - the NTSB has not yet ruled out other causes - it could eventually have deep ramifications. For example, the Federal Aviation Administration (FAA) might look more closely at the spacing between planes, especially 747s, which leave a lot of turbulence in their wake. This could add to take-off times at busy international airports. Moreover, Airbus Industries, which made about 430 of the A300s, might have to reexamine its design of the tail fin.

The tail-fin issue could be particularly difficult. To save weight, Airbus made the three main spars in the tail out of lighter material. However, experts say it's very difficult to detect when those composites start to "delaminate" - that is, when the very thin pieces start to lose their adhesion.

In fact, one of the pins on flight 587 connecting the tail fin to the fuselage suffered delamination immediately after construction and was repaired. The NTSB says it is now part of the investigation.

"The challenge has always been how to inspect composite pieces while they're in service to insure you're catching delamination if it occurs," says Clint Oster, an aviation expert at Indiana University in Bloomington.

Yesterday, American Airlines held meetings in Miami and Tulsa, Okla., to work out protocol for inspecting the tail area in planes. "You just don't raise the hood here," says John Hotard, a spokesman for American.

The Airbus that crashed on Monday had been subjected to another form of turbulence in 1994 - clear-air turbulence, which occurs in skies naturally.

Usually, pilots have time to warn passengers to buckle up before a plane starts bouncing. But that didn't happen in 1994, and 46 people were injured. The plane was given a "post turbulence" inspection.

"You could get a situation where a part

is weakened, and then it's subjected to some additional turbulence that normally wouldn't create a problem," says Mr. Oster. "But the combination of that weakened part with the turbulence could lead into a problem. And, once you get some sort of a breakup in an airplane, all kinds of unusual stresses get put on the aircraft, and lots of other perfectly good parts may fail."

Many of those parts will be tested by the NTSB, which is shipping the parts to its inspection facilities. The NTSB also hopes to find out more when it analyzes the flight-data recorder. This will give it an idea of the amount of stress the plane was subjected to, which could help determine if wake turbulence was the culprit.

On Wednesday, the NTSB demonstrated how serious it considers the issue when it showed reporters blow-ups of flight paths of the American jet and the JAL 747. The planes were basically on parallel courses, with the American flight slightly lower. "This would be consistent with a wake vortex encounter," said NTSB chairman Marion Blakey, adding that "We don't know if that actually contributed to the accident."

Outside airline safety experts warn that crashes are often difficult to explain until more information is developed - often months afterwards. "There's nothing really obvious here, says Gordon Bakken, an aviation engineer. "It reminds me of the TWA crash off of New York (in 1996). They had such a miserable time figuring out what in the world really caused that."