American Airlines Flight 587: The ATC Audio and the Tail Failure

This video plays the recorded radio from Kennedy tower and New York departure on the morning of 12 November 2001 and explains each stage with the NTSB's findings. It is a short, clear account of what controllers heard and what they could not see.

Taxi and take-off

Flight 587, an Airbus A300 with 260 people on board, taxied for runway 31L behind a Japan Air Lines Boeing 747. The tower cautioned the crew about wake turbulence, the rough air a heavy jet leaves behind it. Flight 587 lifted off about 1 minute 40 seconds after the 747 and was never closer than 4.3 nautical miles. The NTSB found that the tower met the FAA's wake turbulence spacing rules.

The last radio call

The crew checked in with New York departure at 1,300 feet, climbing. Departure cleared them higher and then turned them toward a fix named WAVEY. The crew's readback, "Direct WAVEY", was the last radio call from Flight 587.

What the flight data recorder showed

As the controller gave that turn, the flight data recorder showed a brief jolt, which the NTSB called typical of a minor wake encounter. About 15 seconds later came a second, similar jolt. This time the first officer moved the control wheel quickly and pushed the rudder pedals right, left and right again, with several more full pedal reversals. The NTSB found that these inputs, not the wake, built up an increasing sideslip. The load on the fin reached about twice its design limit load, and at 9:15:58 the fin broke away.

Confusion on the frequency

The aircraft hit the ground about 16 seconds after the fin separated. The first report of a crash did not come from the tower. A pilot on the frequency saw a fireball south of the field and told the controller. Controllers asked whether they were missing any departures and could not say what type of aircraft it was. Some thought it might be a 767. Departure called "American 587" repeatedly and got no answer. Another American crew described a large column of black smoke on the southern shore of Long Island. The fin and rudder had fallen into Jamaica Bay, and the main wreckage struck the corner of Newport Avenue and Beach 131st Street in Belle Harbor. Four homes were destroyed, and three more were substantially damaged.

Why the pilot did it

The NTSB's probable cause was the separation of the vertical stabilizer under loads beyond its ultimate design, created by the first officer's unnecessary and excessive rudder pedal inputs. It named three factors behind those inputs: his tendency to overreact to wake turbulence, his training, including a simulator exercise that portrayed wake turbulence as far more violent than reality, and a rudder system with the lightest pedal forces of any airliner compared. The presenter's closing view, which is his own, is that the report describes a pilot whose training pointed him to a rudder that needed very little force, at a time when many pilots misunderstood how much the tail could take.