-
-
- Home Page || Flight 587 Page || Letters from U.S.Readers || Send E-Mail to the Editor || Search U.S.Read or the Web ||
-
A few weeks ago, I sent an informal email to the Editor of The Wave, Howard Schwach, describing briefly our analysis of what the NTSB investigation is likely to conclude as the probable cause of the crash of American Airlines Flight 587 (FL587). Unfortunately, Howard and I did not have the opportunity to discuss my email, and Howard misinterpreted what I wrote as "AA587 Pilot Shocked by Forces on Rudder". Howard has graciously offered me some space here to set the record straight and express my true thoughts on Flight 587. My purpose was to briefly lay out a quick synopsis of the American Airlines (AA) and Airbus positions, what the NTSB is likely to conclude, and our reaction to that.
For all the bickering that has been going on these last two and a
half years between AA and Airbus, their latest
reports to the NTSB have highlighted an issue they actually agree
on
–– they both believe that when the Pilot Flying (PF) the aircraft made
his first of five rudder movements, he was so spooked by the sideways
(lateral) jolt created by that first movement that he tried to counter
that jolt by swinging the rudder to the other side. These
movements begat other movements which kept making the situation
worse.
Where AA and Airbus part ways is that AA holds the position that the
PF was spooked by his first rudder movement because the rudder pedal
system is the most sensitive in the industry. Airbus contends the
Pilot was shocked because he did not have a proper understanding of the
power of the rudder, had been inappropriately trained by AA to use the
rudder in response to wake turbulence encounters, and had received more
improper training in AA's simulators which emphasized the rudder as a
primary control to counter a severe, unexpected turbulence-induced roll
of the aircraft's
wings.
No one really knows if the Pilot of Flight 587 was "shocked" by the aircraft's movements in response to his first rudder movement but our analysis of all the data has led us to conclude that what did shock the Pilot was the effect of the "initiating event" (explosion and subsequent fire) which he may have initially interpreted as a wake turbulence e
Obvisously, the NTSB will have a tough job. They may settle somewhere in between -- something like this:
The Pilot in all likelihood left the AAMP training program with the
impression that rudder could provide the most reliable roll control
authority in wake upsets. He therefore was aggressive with the
rudder in responding to what he thought was a wake encounter when he
was in a 25 degree controlled left turn. In all likelihood, the
Pilot feared the wake encounter might roll the aircraft well past 25
degrees. The Pilot had publicly expressed his understanding that
AAMP encouraged these quick rudder inputs to Captain John Lavelle
several years earlier. There is no evidence the Pilot’s
understanding of rudder use was ever adjusted.
The Pilot was in all likelihood not aware of the limited pedal travel
at 250 knots which would provide him with full available rudder
travel. The sharp lateral responses of the aircraft to the quick,
stop-to-stop rudder movements, were in all likelihood interpreted by
the Pilot as a strong wake encounter.
The Pilot very quickly became caught in an APC/PIO (Aircraft Pilot
Coupling/Pilot Induced Oscillations) event given that he was not aware
that the small rudder pedal movements could produce such strong yaw
forces.
This lack of knowledge was a direct result of Airbus not communicating
the re-design of the rudder system to the airlines (where the rudder
pedal travel was significantly reduced compared with the original
A300).
The AAMP training program, coupled with the rudder sensitivity, created
a dangerous mix.
Also, significant contributing factors were:
1. Airbus failure to disclose the tail loads FL903 experienced in
1997; had they done so .... the industry would have learned several
years ago that rudder reversals could be catastrophic.
2. The failure to explicitly state, in 1997, that rudder reversals on FL903 nearly broke the tail off.
3. The failure to disclose the conditions under which the Rudder Limiter would fail to limit rudder travel (for example, when accelerating and experiencing rudder reversals).
4. AA’s over-emphasis (in their AAMP training program) on the ability of wake vortices to significantly upset large aircraft like the A300-600 model. NTSB studies show that had the Pilot not responded to the 2nd wake encounter, the aircraft would not have rolled beyond 34 degrees.
5. American Airlines modification of the A300 simulator which led to an emphasis of rudder use for roll control. This created negative training. This simulator change was not approved by Airbus.
Of course what most stood out to me from all these reports was the
complete, total absence, of ANY discussion on these factual issues:
1. Dozens upon dozens of eyewitness reports of fires, explosions,
“boom” sounds, BEFORE the tail separated.
2. No discussion of the Pilot’s ATC transmission at the start of the 2nd alleged wake event.
3. No discussion on why the CVR did not record that ATC transmission.
4. No discussion on why it was corrupted and affected TWO ATC channels.
5. No CVR analysis of the mysterious sound spectrum readings during the time of the rudder movements, which suggest there were power interruptions on board.
6. No CVR analysis of the two periods of “airframe rattles”.
7. no explanation for the premature cessation of DFDR data, 13 seconds before the CVR quit (impact), while the engines were still attached and running.
8. no ballistic trajectory analysis on when/why/how the engines separated.
9. No real discussion of the debris field.
10. no discussion of the lack of a careful sonar scan of the Bay for debris.
11. No CVR analysis of the Loud Bang.
12. No discussion of why Spoilers 2, 3, 5 also failed at the same moment of the Loud Bang when the rudder also failed (rudder data went to zero).
13. No discussion of the fact that the Pilots ATC transmission 13 seconds before impact was not recorded by the CVR, exhibiting the same CVR malfunction as his “try escape” transmission, 7.5 seconds before the assumed tail sep. moment.
14. No discussion of the tollbooth video and why there’s a smoke trail before tail sep., and why there is no “event” notable when the tail allegedly separated, and why/how the plane flew level for 8 seconds after the assumed tail sep. moment.
15. No discussion that the cockpit warning system (ECAM) does NOT indicate tail sep. or engine seps (Level 3 events) until 9:16:07.55 (9 seconds after the NTSB assumed time of tail sep.) -- just after the large flash and smoke seen on the video.
16. No discussion that the aircraft had stalled before the tail separated (as indicated by the ECAM and witness John Power).
17. no discussion that the radar data supports the interpretation that the major aircraft in-flight break-up occurred near the Jamaica Bay seawall, just after the flash and smoke on the tollboth video. Well after the NTSB assumed time of tail sep.
For all the technical debates in the five aforementioned reports, my
reaction to all that stuff is, “are we talking about the same crash?”
Victor Trombettas
Home Page || Flight 587 Page || Letters from U.S.Readers || Send E-Mail to the Editor || Search U.S.Read or the Web ||