Airbus has revealed that an A300-600 tail fin survived in-flight lateral loads that exceeded the structure's design limitations before the fatal American Airlines flight AA587 A300-600R accident on 12 November 2001. But the earlier case - in May 1997 and also involving an American Airlines aircraft - did not involve detachment of the fin as in the AA587 accident during climb-out from New York's Kennedy airport. Visual inspection immediately after the earlier incident indicated no damage.

American Airlines and US National Transportation Safety Board (NTSB) investigators have removed the fin involved in the 1997 incident and subjected it to ultrasonic inspection. This has shown "an indication of damage, possibly the pin bushing of the right rear [attachment] lug", according to the NTSB.

The aircraft involved had suffered a stall followed by "a period of oscillations" during which the aircraft lost 3,000ft (915m) in 34s. During that 34s, Airbus reports, the fin was subjected to "sustained, very high lateral loads" (Flight International, 5-11 March).

Since the AA587 accident, Airbus and the NTSB have revisited the data from the incident to work out the loads generated. Although the investigation at the time established what happened as a result of pilot control inputs following a stall caused by inattention to decaying airspeed during a level-out, they did not fully review the stress levels since there appeared to have been no ill effects. Airbus says the fin is designed to be able to withstand "flight load levels" with one of its six attachment points completely disconnected.

Industry sources say the NTSB, which since the AA587 accident has published two reminders to pilots about the dangers of heavy use of rudder in large transport aircraft, is looking again at the way American Airlines pilots are trained.

Airbus says it has been searching for other events in which high fin loads were generated, but says the closest incident involved a FedEx A300-600F in 1995. But in that case, the NTSB says, there was no rudder reversal, just "large rudder deflections" which were a result of "a rudder trim/autopilot interaction".

Source: Flight International