By Alan Levin

11-13-01

Veteran aviation accident investigators say they have never seen a scenario

like the one that played out in the final moments of American Airlines

Flight 587.

They have little doubt about what led the jet to crash Monday in New York

City: Large pieces came loose shortly after it took off.

But that hardly solves the mystery. Investigators are still searching for

what could have caused the Airbus A300 jet to break into pieces and slam

into homes in the Rockaway Beach neighborhood of Queens about 9:17 a.m.

The two General Electric CF6 engines landed separately in the neighborhood,

each several blocks from the fiery main crash site, federal sources say. The

Coast Guard also pulled a part of the A300's tail from Jamaica Bay.

''Right now, to me, it is extremely mysterious. I find the evidence

perplexing,'' says Bernard Loeb, a retired chief aviation investigator with

the National Transportation Safety Board.

NTSB investigators said late Monday that preliminary information from the

cockpit voice recorder suggests that the crash was an accident. Only the

captain and co-pilot can be heard on the recording, said NTSB Chairwoman

Marion Blakey.

But they offered few specifics. They said that the flight lasted less than 2

minutes from takeoff to crash. Investigators have not yet examined radar

data that will provide the precise altitude at which the jet encountered

problems.

One of the first things they will try to determine is whether the engines

failed. Such failures occur dozens of times each year. In extraordinary

circumstances, engines shoot damaging shrapnel into an aircraft.

However, it's unheard of that two engines would break loose at or about the

same time, aviation experts say. Jet and engine manufacturers go to great

lengths to ensure that catastrophic engine failures cannot cause a crash.

Similarly, jets are designed with so much strength that even when the wings

or tail fins are damaged, they almost never break loose before crashing,

investigators say.

The part of the tail -- a fin that rises vertically -- that was recovered

had no visible marks indicating it was struck by an object that could have

torn it from the jet.

Even the early evidence about how the jet broke apart puzzled investigators

and others. Some pilots who saw the jet after takeoff said it did not come

apart, while other aviators said they saw pieces coming from it, one source

says.

''It doesn't sound like the typical thing,'' says Kevin Darcy, a former lead

accident investigator at Boeing who is now an aviation consultant.

Loeb says it is possible that the key to the crash could be with the tail

section, which was hauled from the bay Monday afternoon. The vertical fin

keeps the jet pointed straight and allows pilots to turn the jet's nose left

or right.

If the small wings on the tail, which raise and lower the nose, also came

loose, Loeb says, that could cause the jet's nose to move downward with

great force. That, in turn, could theoretically shake the engines off the

wings.

Loeb cautions that such a scenario is highly speculative. ''I don't really

know exactly. But I think that's possible,'' he says.

Whenever engines break off a jet, investigators focus on possible failures

within the engine. The CF6 engines on American's A300 fleet have recently

drawn the attention of safety regulators.

The NTSB issued a recommendation last December urging that the Federal

Aviation Administration, which regulates the aviation industry, address

safety issues on the CF6 engines.

On Sept. 22, 2000, a CF6 engine on a US Airways Boeing 767 blew up as

mechanics tested it on the ground in Philadelphia. There was a loud

explosion, and a fire broke out under the left wing of the jet, the NTSB

reported.

''The incident raises serious safety concerns because, if it had occurred

during flight rather than on the ground during maintenance, the airplane

might not have been able to maintain safe flight,'' the NTSB wrote.

Other incidents involving the engine have also raised concerns among

investigators.

In April 2000, a Continental DC-10, which had three CF6 engines, had one

engine break apart as it took off from Newark International Airport. Pieces

from that engine damaged a second engine. The crew landed the jet, and no

one was injured.

Since then, the FAA has required additional inspections of all CF6 engines.

''My understanding is that the airplane was in compliance with all

airworthiness directives,'' says Al Becker, a spokesman for American

Airlines.

Birds are another possible cause of engine failure. The marshy area near New

York's John F. Kennedy International Airport is a haven for birds. Several

crashes, though not on A300 jets with GE engines, have been caused by

engines exploding after birds were sucked into them.

However, several accident investigators say it seems unlikely that such

failures struck Flight 587. Major failures on jet engines are often obvious

after a crash, but sources said the engines exhibited little evidence of

such a failure. The sources cautioned that on-scene observations are not

always reliable.

Furthermore, earlier failures of the CF6 engines have not caused them to

break off. On the US Airways 767, for example, the engine remained on the

wing.

In fact, one engine specialist who asked not to be identified says modern,

high-power engines such as the CF6 have failed so rarely in flight that

investigators have little idea what would happen to a jet if one came apart.

Investigators also say it seems unlikely that failures violent enough to

shake each engine loose from the aircraft could strike both about the same

time.

Though federal officials went out of their way to say no evidence suggested

sabotage, officials say privately that they cannot rule it out.

Copyright 2001 USA TODAY, a division of Gannett Co. Inc.