Home Page || Flight 587 Page || Letters from U.S.Readers || Send E-Mail to the Editor || Search U.S.Read or the Web ||
 
August 19, 2002
 
 
The Five Rudder Movements - A Pilot's Reaction to Wake Turbulence?
 
The Rudder Movements Timeline and What It Shows
 
by Victor Trombettas (with Arthur Doucette)
 
 
On November 20th, the following comments came from the NTSB's first web-based Update on the Flight 587 Investigation:
 
The FDR (Flight Data Recorder) indicates that flight 587 encountered two wake vortices generated by JAL flight 47. The second wake encounter occurs about 8 seconds before the end of the FDR data. For the first few seconds after the second wake encounter, the aircraft responded to flight control inputs. Both wake encounters averaged about 0.1 G lateral (side to side) movement. During the last 8 seconds of FDR data, the plane experienced three stronger lateral movements, two to the right of 0.3 and 0.4 Gs, and then one to the left of 0.3 Gs. These lateral forces corresponded in time with rudder movements. The NTSB continues to investigate the cause of the rudder movements. The FDR's rudder data becomes unreliable about 2.5 seconds before the end of the recording.
 
We get some pretty specific numbers here--"The second wake encounter occurs about 8 seconds before the end of the FDR data", "The FDR's rudder data becomes unreliable about 2.5 seconds before the end of the recording." Note that the NTSB was able to pinpoint the "rudder data becomes unreliable" point down to the half second. That indicates a fair degree of certainty. This 8 second sequence quickly grabbed the attention of the investigators when Tom Haueter, the NTSB's Deputy Director of Aviation Safety, stated in November, ""Obviously, the whole time we're talking about is the last eight seconds ... We have eight seconds we're going to be looking at in extreme detail."
 
It becomes very clear that the NTSB believed (and we assume, continues to believe, since they have not updated, or commented on, this data since November):
 
a. the second wake encounter occurred 8 seconds before the FDR stopped recording.
b. Only 2.5 seconds before the FDR stopped recording, the rudder position data in the FDR became "unreliable".
 
Therefore, we only have 5.5 seconds of reliable rudder data during those last 8 seconds of FDR recording. This is important. Let's look at this in simpler terms. We have a 5.5 second period which begins with a one-tenth g-force lateral movement, a.k.a., the 2nd wake encounter. The NTSB has never revealed what effect this wake encounter had on the plane. Did it create a sideslip or yaw? A roll? No clues. After this wake vortex has hit the plane we must allow some time for the pilot to mentally process that encounter and to decide to use the rudder. All of this, from the wake encounter, to a pilot's decision to react, and then to the five rudder swings, must occur in a 5.5 second period. We'll come back to this 5.5 second period later.
 
On Feb. 8th 2002, the NTSB issued a Safety Recommendation in which they warned against "sequential full opposite rudder inputs". In layman's terms--"don't swing the rudder from side to side--you might break the tail".
 
Please note this text from that Safety Recommendation:
 
"During (emphasis mine) and shortly after the second (wake) encounter, the flight data recorder (FDR) on the accident aircraft recorded several large rudder movements (and corresponding pedal movements) to full or nearly full available rudder deflection in one direction followed by full or nearly full available rudder deflection in the opposite direction."
 
"During the second wake encounter". In other words, "As the second wake encounter occurred ... the rudder was already moving.
 
That requires an incredibly quick response time from the Pilot. It is difficult to imagine this because pilots:
 
a. never react with rudder as their first response to wake turbulence
b. Usually have their feet on the floor; not on the rudder pedals
c. might employ ailerons before they use rudder
d. usually let the plane fly through most turbulence and return itself to normal attitude without trying to "control" the plane back.
 
The NTSB's Feb. 8th Safety Recommendation implies that Flight 587 hit the second wake vortex, and as it hit, during the encounter, the Pilot exercised zero patience and decided, contrary to his training, to abandon restraint and went straight for the rudder. We are not given any attitude information on 587. What happened to 587 that caused the Pilot to react with the rudder? We assume Flight 587 was in normal climb attitude just before the second wake vortex hit. Flight 587 had encountered the first alleged wake vortex 20 seconds earlier (the first 0.1g lateral movement), but that was an insignificant encounter because Marion Blakey stated last November that the "attitude of the plane was unchanged" after that encounter. What made the 2nd alleged wake encounter so unique as to solicit almost superhuman response times from the First Officer? There was nothing unique about that 2nd wake vortex as it was, in most likelihood, at least 8 football fields away and much higher and never reached Flight 587. If there was no 2nd wake encounter, why did the rudder start moving at all?? There have been leaks to the press over the last several weeks that Investigators are becoming increasingly convinced the First Officer was responsible for the rudder movements. If that is true, he was using the rudder for the same reason he had called for max power a second earlier ... because they were in recovery mode even before the 2nd alleged wake encounter. The crew had declared loss of control when only 2 of the 5 rudder movements had completed, when they were still 4.5 seconds from the "rudder data unreliable" point, the point the NTSB believes the rudder and/or tail broke free. Since the crew felt it necessary to institute a recovery procedure before the alleged 2nd wake encounter ... the simple question arises ... why? What put the crew into a situation where they were worried about the safety of the plane if, according to the NTSB, the plane's attitude is fine and they haven't yet hit the 2nd wake vortex? The two periods of airframe rattles (the second period of rattles precede the 2nd wake encounter by 5 seconds), have to loom as very large leads in this case. The condition of the rudder also looms as a large lead--why was it ripped apart so badly and discovered in the Bay at least 600 feet from the vertical stabilizer in multiple pieces? Why did so many people see explosive flashes, raging fires, heavy smoke, before the tail came off? It appears that many people are content in assigning the rattles to wake turbulence.
 
More interesting information from the NTSB's Safety Recommendation:
 
"within about 7 seconds, the rudder traveled 11° right for 0.5 second, 10.5° left for 0.3 second, between 11° and 10.5° right for about 2 seconds, 10° left for about 1 second, and, finally, 9.5° right before the data became unreliable."
 
Earlier I showed how the NTSB had identified a 5.5 second period which began with the 2nd 0.1g lateral acceleration, was followed by rudder movements (we discovered in the February Safety Recommendation that it was five rudder movements), and ended with the rudder data in the FDR becoming unreliable. But in February, the NTSB stated, "within about 7 seconds, the rudder ..." made it's five movements. Seven seconds? We only have 5.5 seconds with which to work with. Not seven. How does a 5.5 second period, which had a fixed start and end point (the second 0.1g lateral movement as the start point and the loss of "reliable rudder data" as the end point), become an almost 7 second period? The data released by the NTSB on February 8th points to only one answer ... the rudder was moving before the alleged 2nd wake encounter. The implications of this are significant. The NTSB has not stated this as a matter of fact. What they stated in November was that the rudder moved after the wake encounter. In February, they said the rudder moved during the wake encounter. The data, however, indicates the rudder was moving before the wake encounter.
 
Based on information from the Feb. 8th Safety Recommendation, and an estimate of the maximum maneuverability speed of the rudder at 39 degrees per second, it appears the rudder started the first of the five rudder movements and ended the fifth one in a total 6.21 seconds (this calculation is consistent with the NTSB's stating the rudder moved five times "within ... 7 seconds").
 
 
The Rudder Movements Timeline
 
I'm also including in the rudder timeline any other data available to us including FAA ATC (Air Traffic Control) tape events.
 
At 9:15:58 AM, the First Officer calls for "max power" approximately a quarter second before the rudder movements begin. This, as a standalone item, is significant. Why do we have the pilot instituting an "escape" maneuver when the 2nd wake encounter hasn't hit yet?
 
1st rudder movement starts at 9:15:58.29 AM (about a quarter second after the "max power" call), moves 11 degrees to the right (from neutral) in .28 seconds, and arrives at 11 degrees right at 9:15:58.57 and stays there for .50 seconds until 9:15:59.07.
 
The alleged 2nd wake encounter occurs at 9:15:59. In other words, the first rudder movement begins .71 seconds before the 2nd wake encounter, and is already at 11 degrees right at the time of the 2nd 0.1g lateral acceleration. Any chance this lateral acceleration was caused by, or corresponded to, this first rudder movement?
 
Also, at 9:15:59, there is the first of several interference-riddled and garbled transmissions on both the Local Control and Departure Control Air Traffic Frequencies.
 
2nd rudder movement starts at 9:15:59.07, moves 21.5 degrees to the left (from right) in .55 seconds, and arrives at 10.5 degrees left at 9:15:59.62 and stays there for .30 seconds until 9:15:59.92.
 
As the 3rd rudder movement is about to start, the "losing control" comments begin on the CVR (Cockpit Voice Recorder) and the losing-control-sequence noises also have begun on the FAA tapes even though the tail does not separate from the plane, according to official sources, until 4.5 seconds later.
 
Also, the transponder sends an erroneous altitude figure at 9:16:00 of 3,019 feet to the JFK radar. According to Ted Lopatkiewicz of the NTSB, the plane is at approximately 2400 feet.
 
Something vicious is already at play here with the pilots already having gone to max power, the rudder moving wildly (either in recovery mode by the Pilot or system induced or both), the transponder is malfunctioning, and the radios are sending poor quality, interference-riddled transmissions, yet the FDR is still recording reliable rudder data.
 
3rd rudder movement starts at 9:15:59.92, moves 21.25 degrees to the right (from left) in .54 seconds, and arrives at 10.75 degrees right at 9:16:00.47 and stays there for 2.00 seconds until 9:16:02.47.
 
The garbled phrase, "Losing Control" is discerned on the Departure Control tape at 9:16:02. The word "Control" is clearly discernible. Loss of control is being declared less than 2 seconds after the call for max power.
 
4th rudder movement starts at 9:16:02.47, moves 20.75 degrees to the left (from right) in .53 seconds, and arrives at 10 degrees left at 9:16:03 and stays there for 1.00 second until 9:16:04.
 
5th rudder movement starts at 9:16:04, moves 19.5 degrees to the right (from left) in .50 seconds, and arrives at 9.5 degrees right at 9:16:04.5 and the rudder position data in the FDR now becomes "unreliable".
 
This analysis raises serious questions about the sequence of events on board Flight 587. If the rudder movements were pilot induced, was he responding to wake turbulence? It is very difficult to see this as a realistic scenario given the analysis above. It cannot be considered a response if it came before, or "during", the trigger event (the wake turbulence). What was the attitude of the plane before and after the alleged 2nd wake encounter? Would it not have been appropriate for the NTSB to release attitude information along with the Feb. 8th Safety Recommendation? If you're going to warn pilots to not use the rudder in certain situations or upsets, wouldn't it be helpful to pilots to provide them with a complete case study and reveal as much as was then known about 587's flight profile?
 
The NTSB states the information this analysis is based on is "preliminary". Therefore, no conclusions can be drawn by anyone using this data. However, Investigators appear to have behaved in a manner contradictory with that belief since Safety Recommendations have been issued as a result of analyzing this preliminary data. And, Investigators have leaked information to the St. Petersburg Times, and the LA Times, implying there is increasing evidence the pilot was responsible for the rudder movements and, therefore, the pilot is indirectly responsible for the crash. If such actions can be taken by parties to the Investigation using preliminary data, then is it unprofessional for parties outside the official investigation to analyze that same data and come up with their own conclusions? The real issue is ... why do we have such a paltry amount of information 10 months after America's second worst aviation disaster? We're not asking for declaration of cause. Just as much factual information as possible. Hopefully, October will change all that. See illustrations below.
 
Rudder Movements Timeline Picture 1 (see Picture 2 below)
 
 
Rudder Movements Timeline Picture 2 (includes critical CVR events, transponder problem, and Waveforms of FAA Air Traffic Control Tape Noises)
 
Home Page || Flight 587 Page || Letters from U.S.Readers || Send E-Mail to the Editor || Search U.S.Read or the Web ||