Point/Counterpoint - The Airbus Position

Air Safety Week, March 29, 2004

'Negative Training'

Probable cause:

The probable cause of the accident involving AAL [American Airlines] 587 was the structural overload of the vertical stabilizer induced by the inappropriate and unnecessary application of cyclic, stop-to-stop inputs to the rudder pedals by the first officer in anticipation of what he mistakenly believed would be the aircraft reaction to an encounter with a wake vortex.

This mistaken belief and the consequent inappropriate and unnecessary pilot actions were conditioned by elements of American Airlines' AAMP that advocated the aggressive use of rudder for roll control. This was reinforced by negative training generated by American Airlines' modifications of the A300-600 training simulator that resulted in temporary inhibition of normal roll control functions such that pilots were forced to use rudder as a primary means of roll control to recover from simulated wake vortex encounters.

Contributing to the accident was the failure of American Airlines to make timely corrections to the AAMP in response to information provided to them by the manufacturers and FAA shortly after this specialized training program was introduced.

The rudder control system design was not at fault (again, sub-heads by ASW)

The A300B4-605R aircraft model, its flight control systems and its structure meet or exceed all certification requirements at the time of certification.

Neither during testing nor in 16 million flight hours of operator service experience did Airbus receive even one complaint or criticism of the handling qualities aspect of its design.

The rudder system characteristics (pedal forces and displacements) comply with certification requirements and were evaluated by the certification authorities (including FAA) in particular during the aircraft flight handling qualities evaluation where necessary rudder pedal inputs are performed to demonstrate the adequacy of the rudder pedal system for its intended use.

The rudder travel limiter ensures that the loads developed by a single, full rudder pedal input followed by a return to the neutral position will remain inside limit loads as prescribed by the certification requirement.

There are no possible flight control system failures that could have caused the large rudder and rudder pedals movements recorded ... during the 30 seconds prior to departure of the vertical fin.

It is clear ... that the copilot applied forces (up to 139 pounds) on the rudder pedals far above the maximum value required to reach the rudder stops for any commercial air transport category airplane ... From these data, it can be concluded that, should the aircraft be equipped with a VLA [variable lever arm] design, the pilot would have similarly ... reached the pedal stops. It might have taken a bit more time to get there. But because of his AAMP training, it is believed that a VLA design would have made no difference in the outcome.

Maximum design maneuvering speed [VA] does not include reversals

There could be some major misconceptions regarding design maneuvering speed.

The FAA-mandated wording in the AFM [airplane flight manual] states: "Maximum Design Maneuvering Speed (VA): Full application of rudder and aileron controls, as well as maneuvers that involve angles of attack near the stall, should be confined to speeds below VA."

This ... wording ... could lead some pilots to conclude that there are no restrictions to manipulating the flight controls (including the use of rudder reversals when operating at or below VA).

A300-600 high loading events are restricted to one carrier

Including the AA 587 accident, the A300-600 fleet has experienced a total of four "high lateral loads" events. They all occurred on the American Airlines fleet.

On the A310 fleet, which shares the same rudder system design, there are [three] "high lateral loads" events ... where the most probable cause is a crew rudder input after a full rudder trim action in the opposite direction (see ASW, Nov. 25, 2002).

Look to the carrier's AAMP

To introduce a simulated aircraft upset, American Airlines modified [its] training simulator by temporarily inhibiting roll and yaw controls ... instead of having one pilot close his eyes while the upset is introduced by the nonflying pilot. As a consequence, while trying to recover from the developing upset, pilot inputs on the control wheel and on the rudder pedals have no effect, thus leading the pilot to make even larger inputs. During the [2002] public hearing, American Airlines testified that it had not consulted the airframe manufacturers regarding this simulator modification. It is important to note that after the accident, American Airlines stopped using this method of inducing upsets in simulator [training].

It is ... clear that American Airlines' management had been made aware of the limitations of simulators before the AA 587 accident. This is clearly shown by ... an American Airlines internal memo from the Managing Director of Flight Operations Technical to the Chief Pilot and Vice President of Flight: