EXPLORING AMERICAN FLIGHT 587


In the current time frame, there is precious little information available on the American Airlines flight 587 crash. This presentation is intended to provide a logical estimation of what probably happened in the crash. A major indicator of what happened is found in the remains of the aircraft - as reported, so far.

Looking at the debris pattern, it's suggested that the rudder and vertical stabilizer came off first. With that event, the center-of-gravity would go forward; the nose would pitch downward. The NTSB cited a 30 degree pitch, downward; a radical angle, under any circumstances.

Under normal circumstances, a pilot would apply aft pressure to the yoke, trying to raise the nose. However, the rudder separation damage would have taken out all the hydraulic systems (no flight controls remaining - including pitch-trim). Hence, engine power would have been the only remaining control - ala UA-232. This assumes that the vertical stabilizer and rudder combination was the exclusive damage in the tail.

To attempt a pitch-up effort, a pilot would use max power on both engines, trying to get the low-slung engines to cause a pitch-up. Thereafter, the snowball effect took hold, and the rest is history.

Now, "...about that rudder coming off..."

What caused such a catastrophic event?

If something flew back & hit the vertical stabilizer, there would be a big dent in the leading edge - there is none.

The NTSB described two events of airframe rattling, 13 seconds apart, with a call for "Max Power," 10 seconds after the first occurrence, nearly coincident with loss of control. The NTSB cited three events of rudder deflection, coincident with the reported lateral "G" loads. The NTSB indicates that there is no evidence of uncontained engine failure.

At 107 seconds after start of takeoff roll - airframe rattle.

At 114 seconds - comment about wake turbulence (pilot speculation?).

At 121 seconds - Second airframe rattle.

At 125 seconds - Call for "Max power.

At 127 seconds - Loss of control.

At 144 seconds - End of recording.

Thus, there would have to be an extraordinary set of forces to separate the engines.

With aerodynamic airframe oscillations and max power, there is the possibility of an engine motion getting started, called "auto-gyration," which is a circular vibration, about the engine support pylon. In an extreme, that could be of enough magnitude to break the mounting bolts and/or airframe (wing section).

That takes us back to the rudder, as the culprit - and the' why and how' of the failure.

The citation of the lateral "G" load cited isn't worth mentioning, under 'normal' conditions. The reading could easily be produced by the aerodynamics/motion of the rudder. The term to remember is "airframe rattle." Wake turbulence is a jolt, or a 'rolling' of the aircraft; not an airframe 'rattle.'

The only images of the rudder indicate severe damage. The rudder being torn up with the vertical stabilizer in one piece doesn't make sense.

There is also a history of rudder actuator problems on this model of aircraft, ala the B-737 rudder actuator problem. The difference is in the fact that the 737 had only one primary actuator, and a standby actuator. The Airbus has three primary actuators, acting in concert. Thus, it's possible to get one actuator fighting the other two. Presumably, the problem was fixed.

The Washington Post report of ten-degrees of left rudder trim indicates some kind of 'yawing' force to the right was encountered; but when?

The abruptness of the catastrophic events wouldn't lead a pilot to trim the aircraft, as though a power failure had occurred, with time available.

The rudder trim would have been double-checked at the zero setting, prior to takeoff. If, after takeoff, the aircraft tended to roll to the right - due to a rogue hydraulic actuator or loose vertical stabilizer - it would be reflex for a pilot to add left rudder trim, not necessarily questioning what required that trim. The normal departure attention requirements would have been sufficiently distracting, that a pilot could easily dismiss the trim issue until a later time.

A minute into the departure, the vertical stabilizer could have begun to 'work' or oscillate. With airspeed increasing, the oscillation could have increased to a high-frequency oscillation, hence the airframe vibration, or 'rattle.'

In all likelihood, it was the oscillation which snapped the lower vertical stabilizer retention structure, leading to a snow-balling of events. In the onset of the catastrophic events, everything would have been extremely fast moving.

If there was a failure of the vertical stabilizer mount, it is possible for there to be maintenance paper-trail, associated with any failure in the crash. A progressive failure should have a history of maintenance write-ups and 'fixes.' The write-ups should address the aircraft being difficult to 'trim.'

If one of the six attachment points failed, it's highly conceivable that a lateral oscillation got started, which became progressively worse, until all points failed. The fact that a forward attachment assembly, in particular, was repaired in the factory - prior to delivery - leaves Airbus in a defensive posture, hence a major debate may be forthcoming, illustrating more facts than normal.

Some of the attachments - left side - appear to be 'pulled-through' their bolt, as opposed to other attachments which appear to have snapped off in mid-section.

This could possibly come up as either the result of a single event, or a design / manufacturing error, with a lot of expensive re-grouping. Hence, Airbus, in particular is heavily invested in finding the truth.

While there is a lingering fear of terrorism, it is worth mentioning that a bomb would have left the lower attachment plates in place, still attached to the vertical stabilizer. That wasn't the case.

Again, the suggestion is that after the tail came off, with the power advanced to maximum, the airframe oscillations/vibrations caused a state of "auto-gyration" about the wing-to-pylon attachment points, causing the wing structure to fail. These types of forces would not have been reasonably considered in the design. The engines were designed to break-away from the lower pylon mounts, only for vertical loads, not for gyration or lateral loads.

The possibility of an unlocked or deployed engine reverser is diminished by the reported uniformity of engine noise & lack of an abrupt spin.

Again, the information is yet limited, hence the conclusions.