NTSB investigation American Airlines Flight 587

national transportation safety board (ntsb) employee brian murphy (second right) updates ntsb chairman marion blakey (third right) on investigation of tail fin , rudder aa flight 587 (february 11, 2002)

the a300-600 took off after japan airlines boeing 747-400 on same runway. flew larger jet s wake, area of turbulent air. first officer attempted stabilize aircraft alternating aggressive rudder inputs. force of air flowing against moving rudder stressed aircraft s vertical stabilizer, , snapped off entirely, causing aircraft lose control , crash. national transportation safety board (ntsb) concluded enormous stress on vertical stabilizer due first officer s unnecessary , excessive rudder inputs, , not wake turbulence caused 747. ntsb further stated if first officer had stopped making additional inputs, aircraft have stabilized . contributing these rudder pedal inputs characteristics of airbus a300-600 sensitive rudder system design , elements of american airlines advanced aircraft maneuvering training program (aamp).

the manner in vertical stabilizer separated concerned investigators. vertical stabilizer connected fuselage 6 attaching points. each point has 2 sets of attachment lugs, 1 made of composite material, of aluminium, connected titanium bolt; damage analysis showed bolts , aluminium lugs intact, not composite lugs. this, coupled 2 events earlier in life of aircraft, namely delamination in part of vertical stabilizer prior delivery airbus s toulouse factory, , encounter heavy turbulence in 1994, caused investigators examine use of composites. possibility composite materials might not strong supposed cause of concern because used in other areas of plane, including engine mounting , wings. tests carried out on vertical stabilizers accident aircraft, , similar aircraft, found strength of composite material had not been compromised, , ntsb concluded material had failed because had been stressed beyond design limit.

the crash witnessed hundreds of people, 349 of whom gave accounts of saw ntsb. half (52%) reported fire or explosion before plane hit ground. others stated saw wing detach aircraft, when in fact vertical stabilizer. witnesses reported seeing 1 of engines burst flames , break off plane, , others reported hearing loud sound sonic boom.

after crash, floyd bennett field s empty hangars used makeshift morgue identification of crash victims.

findings

photo showing crash site

according official accident report, after first officer made initial rudder pedal input, made series of alternating full rudder inputs. led increasing sideslip angles. resulting hazardous sideslip angle led extremely high aerodynamic loads resulted in separation of vertical stabilizer. if first officer had stopped making these inputs @ time before vertical stabilizer separation, natural stability of airplane have returned sideslip angle near 0°, , accident have been avoided. airplane performance study indicated when vertical stabilizer separation began, aerodynamic loads 2 times loads defined design envelope. can determined vertical stabilizer s structural performance consistent design specifications , exceeded certification requirements.

contributing factors include following: first, first officer s predisposition overreact wake turbulence; second, training provided american airlines have encouraged pilots make large flight control inputs; third, first officer not understanding airplane s response large rudder inputs @ high airspeeds or mechanism rudder rolls transport-category airplane; finally, light rudder pedal forces , small pedal displacement of a300-600 rudder pedal system increased airplane s susceptibility rudder misuse.

most aircraft require increased pressure on rudder pedals achieve same amount of rudder control @ higher speed. airbus a300 , later airbus a310 not operate on fly-by-wire flight control system, instead use conventional mechanical flight controls. ntsb determined because of high sensitivity, a300-600 rudder control system susceptible potentially hazardous rudder pedal inputs @ higher speeds . allied pilots association, in submission ntsb, argued unusual sensitivity of rudder mechanism amounted design flaw airbus should have communicated airline. main rationale position came 1997 report referenced 10 incidents in a300 tail fins had been stressed beyond design limitation.

airbus charged crash american airlines fault arguing airline did not train pilots characteristics of rudder. aircraft tail fins designed withstand full rudder deflection in 1 direction when below maneuvering speed, not guarantee can withstand abrupt shift in rudder 1 direction other. ntsb indicated american airlines advanced aircraft maneuvering program (aamp) tended exaggerate effects of wake turbulence on large aircraft. therefore, pilots being trained react more aggressively necessary. according author amy fraher, led concerns of whether appropriate aamp placing such importance on role of flight simulators in teaching airplane upset recovery @ all. fraher states key understanding crash of flight 587 lay in how accident pilots expectations aircraft performance erroneously established through clumsy flight simulator training in american s aamp.

statement of probable cause

from ntsb report of accident:

national transportation safety board determines probable cause of accident in-flight separation of vertical stabilizer result of loads beyond ultimate design created first officer’s unnecessary , excessive rudder pedal inputs. contributing these rudder pedal inputs characteristics of airbus a300-600 rudder system design , elements of american airlines advanced aircraft maneuvering program (aamp).

since ntsb s report, american airlines has modified pilot training program. previous simulator training did not reflect actual large build-up in sideslip angle , sideloads accompany such rudder inputs in actual airplane , according ntsb final report.