ADIRU

Significantly, in the paragraph quoted above, the NTSB states that not even the 24 aircraft that were inspected using nondestructive testing should be exempt; it is of interest to all what manner of nondestructive inspection method these 24 aircraft have undergone, as Airbus has been adamant that visual testing is all that is required on A-300/A-310 tail surfaces.

New Understandings on Composite Delamination, Effects of In-Flight Rudder Break-up

Also, the recent Safety Recommendation, the NTSB stated dramatic new findings on the Air Transat 961 incident:

Further examination of the vertical stabilizer determined that its two rearmost attachment lugs were damaged due to the high stresses associated with the rudder failure and separation.4 These high stresses may have been dangerously close in magnitude to those that caused the in-flight separation of the vertical stabilizer during the November 12, 2001,

This statement from the NTSB is extremely important to the AA 587 accident investigation, for the following reasons:

This is the first time that:

 the NTSB has publicly considered that rudder disintegration could cause damage to the vertical fin of an A-300;
 it has been revealed that the vertical fin nearly separated from the A-310 aircraft involved in the Air Transat 961 incident (previous reports indicated that the damage on Air Transat 961’s vertical fin attachment points were caused by a previous incident.)
 a partial rudder (as the rudder fell apart, only portions of it were placing loads on the vertical fin) could generate enough load to damage the tail to the point of failure.

Indeed, it is also of note that the NTSB is discussing “flutter” of the rudder as possibly destroying the vertical fin:

4 When the rudder separation began, the rudder started to flutter, or swing back and forth violently. This, in turn, led to the vertical stabilizer moving left and right and the stress in the lugs increasing to the point where the lugs became delaminated.

Experts Warn That Composites and Metals React Differently Under Load

In the original letter to the NTSB and the FAA, our pilot group asked Martin Aubury, formerly Head of Aircraft Structures at the Australian Civil Aviation Authority, about the danger of “flutter” in load-bearing composite materials like rudders or vertical fins. He responded, in part:

“…composites are more susceptible than metals to repetitive loads in the range between limit and ultimate. That is because composites do not deform plastically so there is no redistribution/alleviation of loads…”

Metal structures have the advantage that the material has approximately the same strength in all directions. So small unanticipated loads in an unanticipated direction are of no consequence. Composites are different. Fibers are oriented to carry expected load and where there are no fibers there is very little strength. An unanticipated load in unanticipated direction has far more serious consequences for composite structures than for metals.”

We also noted:

If the phenomenon explained above by Mr. Aubury…(is)true, then once again, in light of the lack of ductility of composites, the structural certification requirements for the vertical stabilizer and rudder must be re-examined.

What this means is: if a rudder begins to disintegrate, it then places unanticipated (and therefore “un-designed-for”) loads on the vertical fin. Due to these loads, is composite material as currently used on the A-300 /310 adequately-structured and tested to handle such loads? In the AA 587 NTSB hearing, on October 31, 2002, Airbus Composite Specialist Erhard Winkler explained the rudder and vertical fin destruction sequence and load factors on AA 587, documented on the transcript of that hearing on pages 941 through 952. Given the two damaged lugs on Air Transat 961, it is clear that assumptions were made in the AA 587 investigation that have not borne out in the Air Transat 961 instance.

Given the lack of foresight for predictable-hydraulic-fluid intrusion (after all, three hydraulic servos are placed inside the composite of the rudder…these servos leak on occasion)…as well as the complete failure to (first) require periodic, effective inspections that would seek such deterioration and (second) to advocate an inspection after the finding that (because it was only using the “tap” test on the exterior panels) was still insufficient to insure safety…with all of these new variables…is it reasonable to question Airbus’ insistence that the vertical fin on the A-300/A-310 is strong enough to withstand a disintegrating rudder?

And, given the proximity of hydraulic fluid around this aging composite rudder, is it not likely that hydraulic fluid will intrude into other A-300/A-310 rudders? What does this say about the design of the other tails—the A-320, A-330, A-340—that all use the same basic-designed, composite tail?

In 2002, our letter to the FAA and NTSB went on to say, regarding AA 587:

“…Did the lack of ductility of the vertical stabilizer lead to premature separation? Would a metal fin have remained on the aircraft, all things being equal? Was the flutter sound heard on the cockpit voice recorder caused by a split rudder, with one or more parts free-floating? There is an AD (enclosure 3, #25) which refers to “freeplay” of the elevator causing severe vibrations in the back of the airplane. Could “freeplay” in a rudder cause similar vibrations which in turn were picked up on the voice recorder as a “flutter” sound?

Uncommanded Rudder

Finally, we must address the issue of “un-commanded” rudder. In a variety of incidents, many documented in the A-300 pilot group’s letters to the NTSB, FAA, American Airlines and the Allied Pilot’s Association, nearly 30 episodes of the rudder moving without pilot input on A-300 or A-310 aircraft have occurred. In the Air Transat case it is clear that the pilots did not manipulate the rudder in any way—meaning, the rudder moved in an “un-commanded fashion.” In May, 1999, AA Flight 916 took off out of Miami and encountered a severe “un-commanded rudder” event. A bulletin that was later issued to American pilots warned that the “uncommanded rudder” was swinging “up to 12 degrees of deflection.” (It should be noted that the rudder swings on AA 587 were less than this degree…which, according to the NTSB, was enough to rip off the vertical fin.) The pilots were –and are—very concerned about “un-commanded” rudder, as the “rudder training” called-for in the NTSB Safety Recommendation A-02-01 and -02 for all pilots will be of little use if the aircraft rudder begins swinging in an un-commanded fashion. It should be remembered that the A-300 pilot group issued several recommendations in 2002—including this one, regarding un-commanded rudder and composite design and inspection issues:

8) Finally, based on the concerns listed below, serious consideration must be given to grounding the entire A300-600 fleet until its airworthiness can be assured.

(a) Certification requirements of the vertical stabilizer and rudder are insufficient to prevent structural separation of the tail assembly under certain “rudder reversal” conditions.

(b) There have been documented cases of delamination in the elevators, rudders and vertical stabilizers, compromising the integrity of these structures. Nondestructive technology is not being used and baseline date has not been established to ensure the integrity of composite structures on an ongoing basis.

(c) There have been a disproportionately high number of documented cases of uncommanded rudder inputs which have, at times, led to large rudder displacements.

(d) The rudder limiter/pedal design is ill-conceived and its limitations may contribute to, rather than prevent, inadvertent “rudder reversals”.

(e) The A300 vertical stabilizer design should be reevaluated as to its method of attachment to the fuselage, in light of its composite structure.

It is worthy of note that in a portion of the investigation, the NTSB studied the possibility of a rudder disintegration (“Airbus Report AAL 587 Crash: Study of Aeroelastic Scenarios; SA-522, Exhibit 7U.”) Although this report considered three possible scenarios, it did not consider the scenario that occurred on Air Transat 961; it also made assumptions about rudder damping and rudder limiting functions that have since been reconsidered. Nonetheless, it concluded that the effect would be “strong instability” in at least one rudder disintegration scenario they did consider.