AA's Full Submission
PROBABLE CAUSE AND CONTRIBUTING FACTORS
The probable cause of this accident was the onset of a design-induced, adverse aircraft pilot coupling (APC) event that led to rapid development of excessively high aerodynamic lateral loads resulting in the catastrophic structural failure of the vertical stabilizer and rudder in only six and one-half seconds.
The event was triggered by an unexpectedly sensitive response of the rudder to an initial, single pedal input by the pilot during a wake vortex encounter. Due to the unique characteristics in the aircraft's flight control system design, the pilot became caught in an adverse APC/pilot involved oscillation mode as he attempted to counter the effects of that input. Specifically, after making a control wheel input followed by a rudder input intended to achieve a desired aircraft response, the over-sensitivity of the rudder control system induced the pilot to make additional, essentially cyclic, corrective rudder inputs as he attempted to stabilize the aircraft. Unknown to the pilot, because of the sensitivity of the rudder controls and the powerful nature of the hydraulically driven rudder actuators, these corrective inputs rapidly generated rupture loads. The rudder travel limiter unit (RTLU) and yaw damper failed to protect against the build up of these loads due to deficiencies in the flight control architecture design.
Contributing factors to the accident included:
1. The manufacturer's failure to disclose information learned from prior in-service high-load events demonstrating the adverse APC characteristics of the A300-600 flight control system and the resulting risk of structural overload;
2. Extraordinary rudder sensitivity at increased airspeeds due to a high rudder pedal breakout force relative to the shallow (low) rudder pedal force gradient and a corresponding reduction in rudder pedal travel that makes the A300-600 uniquely susceptible to adverse APC/pilot involved oscillation;
3. The rudder travel limiter unit's inability to protect the aircraft from excessive lateral loads;
4. The inability of the yaw damper, when the rudder pedal is held at the stop, to damp out motions resulting from the adverse APC/pilot involved oscillation tendencies of the aircraft;
5. Industry-wide lack of awareness before the accident of the catastrophic potential of rudder reversals, even at speeds below design maneuvering speed;
6. Industry-common, but incorrect, pilot assumptions about aircraft maneuvering speed based upon prevailing definitions of the term; and
7. The lack of clear regulatory verification requirements to identify and correct adverse characteristics through flight-testing and evaluation of handling qualities of flight control systems during original, as well as subsequent, "derivative" model, aircraft certification.
The "best" part is that full available rudder deflection force and pedal travel decreases from 65 pounds and 4 inches at takeoff speed to 32 pounds (over a breakout force of 22 pounds) and 1.2 inches at the accident speed.
Also of note is that no pilot in the static tests was able to modulate rudder at 250 kt. It was either maximum available deflection or nothing
The Committee also found that "the precursor or trigger is pilot related" in many adverse APC events, and that "an environmental or vehicle trigger" often "precedes the pilot trigger." (NRC Study, p. 54.) According to Figure 4a below, reproduced from page 19 of the Group Chairman's Aircraft Performance Study Addendum #1, the initial right wheel input preceded the First Officer's first application of rudder by approximately0.6 seconds. (Time in seconds is listed on the horizontal line of the figures; 839 seconds corresponds with 09:15:47.2 EST and 850.3 with 09:15:58.5 EST (tail separation). The second wake encounter begins at 841.8 seconds/09:15:50 and ends at 845.8 seconds/09:15:54. Therefore, recorded data in these figures account for 11.3 seconds before stabilizer separation.) The initial flight control inputs were triggered by the
second wake turbulence encounter. The pilot's initial full right wheel input did not overpower the wake, and the aircraft roll attitude did not respond. Roll actually increased from about 23 degrees left angle of bank, at the onset of the wake turbulence, to a peak of approximately 25 degrees before reversing toward wings level. (Figures 2 and 4e of Group Chairman's Aircraft Performance Study Addendum #1, pp. 17 and 23, respectively.)
Here we see rudder being called upon to bring bank back under control
after full right wheel input was found insufficient to counter the wake induced roll. I'd find it personally difficult to stay off the rudder with the bank going past 25 degrees, especially in a transport a/c. Is the pilot better off throttling back an engine
Looks perhaps like he should on this a/c type
Pssst Shadow -- What's with the pic of Tafari Makonnen?