SaturnV

TraderAl, this is a picture of a Stinger missile hitting an F-14 on a static stand. Note the smoke trail and the bright flash. No witness to AA587 has reported a similar bright flash nor any smoke trail from burning propellant. (Given the altitude and the visibility that day, a smoke trail would have been visible from the launch site up to impact point)

A Stinger's 2.2 lb. warhead is puny. The weight is about five times that of the explosive charge in an ordinary concussion grenade. A 2.2lb warhead detonating in a proximity airburst would likely not be lethal to an aircraft the size of an A-300. By way of comparison, the SA-5 surface to air missile that struck the Sibir Airlines Tupolov over the Black Sea several months ago carries a 450 pound HE warhead. The SA-5 explodes in a proximity airburst. As I recall, the Sibir pilots were aware they had been hit and had communication with ATC to that effect.

You stated that you know that a Stinger uses a proximity airburst to destroy its target (rather than an impact and warhead explosion mode). Can you please give a factual source for this knowledge. If you cannot, I suggest this prolonged sidetrack to an otherwise informative discussion thread be put to rest.

Mirkin About

Trader Al, Janes lists all variants of the Stinger as having a 1KG HE blast smooth case fragmentation head with a time delay CONTACT fuse . Is that clear enough , can we please get back to the "merits" of composite construction as the posts so far have been fascinating.

[ 06 December 2001: Message edited by: Mirkin About ]

Evo7

Mirkin About

I've tried quoting Janes already...

FFFlyer

SystemsGuy - surely if there was lose or damaged wiring making intermittent contact then this scenario would not have been specifically simulated or tested.

'If a function, mode, condition, etc. hasn't been specifically simulated or tested, one cannot know with any confidence what will really happen in that case.'
Isn't this then unpredictable and random behaviour?

rubecula

FYI

[ 06 December 2001: Message edited by: rubecula ]

Dagger Dirk

I've reproduced some facts between the lines [below] in order to make it easier to refer to the data.

A learned aerodynamicist said: "I can see how a wiring fault could change persistent state settings, such as rudder travel limitations, but I do not see how a wiring fault could generate oscillatory commands."

The question and answer session [below at bottom] was probably not quite clear enough (in its basic premise) - so I'll try and re-state the proposition:

a. Ockham's Razor tells me that the second wake encounter was the initiator. What did it initiate? A significant yaw around the vertical axis. What did the FCS think of that? It quickly decided that that yaw had to be corrected and fed in a restorative yawing moment via a rudder displacement signal. In a serviceable airplane that would have been the end of that i.e. the aircraft's natural directional stability plus the rudder movement would have dampened down that yaw quite quickly with little or no oscillation. But what if the rudder limiter's variable stops were not where they should have been? The FCS computer software would be saying "a yaw of that magnitude requires 15% of the presently allowable rudder deflection (at this speed)" and so it applies that. However (see the limiter's variability directly below), instead of applying 15% of the (say) 20 degrees of travel allowed by the variable stops at 250kts it applies 15% of the full 30 degree travel (to the hard stops). The result is an overcompensation (which the FCS understandably misreads as another externally inspired yaw) and so the FCS feeds in another restorative rudder correction which begets yet another overtravel. So instead of a dampening you have an excitation process underway. By this time we have set up a destructive harmonic (the CVR "rattle") and a fin-rudder interaction that has the fin into an out-of-phase flexure (an FCS-induced flutter mode if you will). Perhaps we should see it as the mirror image of aerodynamically-inspired flutter. If the software thinks in percentages of available travel, then the theory is sound.

b. Now why would/could this be? If a wiring fault existed, which caused the variable (computer-driven) stops to be reset to the hard stops, would the computer know? i.e. is there some feedback mechanism that would indicate that this was the case? Possibly/probably not....why should there be? Few systems make provision for wiring flaws. Why wouldn't pilot pedal-inputs be excessive also? Normally a rudder sits quiescently and neutrally in trail. Large rudder displacements are only used in (say) low speed landing or take-offs in crosswinds, low-speed (practise) asymmetric work (and most of that in the simulator) or, at higher IAS, unusual attitude recoveries (wake or CAT - both quite rare). At the lower speeds the variable limiter stops are not in play. Minor disturbances or out-of-trim/unbalanced minor asymmetric thrust situations are usually soaked up quickly by the FCS running the rudder-trim to softly bias the rudder. It is only when a wake upset hits with a large lateral gust that the FCS would be stimulated to respond with a rapid corrective and restorative rudder movement of any magnitude. The latent flaw needs to be "kick-started".

c. So whether any wiring (or software) induced flaw in the rudder limiter's variable stops was permanent or intermittent, I doubt whether there is an FCS BITE function to ascertain its existence. It would simply lurk and await its (yawing) moment. Its moment for AA587 may have been a lateral gust sufficiently great to have kick-started a process that was self-sustaining and divergent (i.e. one that exceeded the dampening capabilities of the airplane's inherent directional stability at that speed). Liken it to the 737's "cross-over" speed.

d. If you examine the other five incidents in light of this, you may also be able to identify that they may have luckily been in a speed regime where the inherent directional damping was sufficient to cause the overall process NOT to be divergent, and/or without a sufficiently strong lateral gust to get the process "kick-started". (other five: http://www.iasa-intl.com/PDF/AA587yawSurvey.rtf )

I hope that's clear and comprehensible. I don't have the facts on the A300 nor AA587 to say more than that I think it's an analysis worthy of follow-up - simply because it fits the bill and accords with a logical dissection of the scenario. There's no doubt that postulating a divergent phugoid around the vertical axis might inspire some scoffing in academic quarters - but Airbus have managed some amazing things with their designs.

I retreat to the notion that a computer is a computer, and the supposition that an FCS can handle an unknown loss of a major variable is likely wishful thinking. Garbage, in, garbage out.

Consider all the modal problems we've seen with FBW and near-FBW aircraft as they encounter unexpected flight conditions. The A320 with spoilers stuck out after a maintenance check, F-100 with a locked air/ground sensor, the A320 Warsaw crash (another air/ground modality). etc etc (you know of many more).
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ +
The A300 has a rudder limiter that places variable stops on the actuator input, depending upon airspeed. Two computers electrically drive the stops. Full 30-deg. travel is allowed up to 165 kt., then it is cut back to only 3.5 deg. at 395 kt.
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ +

Quest: What makes the rudder move? Answer: hydraulics (in response to electrical signals from the electronics (the FCS computers) or pilots' pedals or yaw-damper signals or gusts)

Quest: What would cause fin-destructive rudder movement? Answer: Excessive throw (in spite of or beyond the rudder limiter's variable stops) or a flutter-inducing excessive rate

Quest: Why would you get excessive rate? Answer: an FCS failure (hydraulic chatter caused by either hyd-valve cross-talk or an electrical wiring glitch)

Quest: Why would you get excessive rudder-throw? Answer: If the variable stops on the rudder limiter were inappropriately set for the speed. FCS calls for a % of what should be available, but gets instead a % of what is available.

Quest: How could the variable stops become mis-set? Answer: Software patch fault / intermittent or permanent wiring fault / latent software anomaly

Quest: Why would this fault turn up infrequently? Answer: Unique combination of circumstances required:
a. Aircraft at higher IAS with limiter inop (or variable stops mis-set) and rudder able to be deflected beyond its design stops for that IAS (but would ONLY do so if stimulated by a requirement to counter yaw)
b. What would cause the rudder to need to counter yaw to the extent that it might exceed the (now absent) limit of the missing rudder limiter stops? Answer: A wake turbulence encounter (i.e. an external stimuli of some significant magnitude). Usually the rudder sits peaceably in trail with no real task. Slight out-of-trim is usually constantly and continually accommodated by a rudder-trim imposed rudder-bias.

Quest: So what would happen after a rudder moves excessively in response to an FCS signal? Answer: The FCS would identify a need to again immediately deflect the rudder (simply because the resultant yaw would have been far in excess of the reaction to the initial yaw that it directed - and expected). => react/over-react/react/over-react => divergent phugoid

Quest: If this confusion continues (between the FCS reasonable request and the rudder's resultant excessive travel), what might happen? Answer: The "rattle" heard on the AA587 CVR may have been that rapido argument between the FCS and the rudder, manifested initially by a chattering of hyd valves, latterly (or simultaneously) by a flailing rudder-induced fin-flutter.

TraderAl

Heh, about to give up.

My knowledge of the proximity burst comes from an ex-US Ranger who was in charge of deploying stingers in various overseas conditions for several years. As far as using stingers and being on the ground he is an expert. If you are an American your tax dollars spent mucho on his education. If you read the content from China Lake that comes with the pic of the stinger strike, you will find it destroys aircraft by detaching appendages and that rarely is a large explosion seen as the white fireball in daytime is not easily detectable. The white flash goes off like a photog flash bulb. In fact that si why the China Lake test was done at dusk, as in daylight the stinger strike is hard to discern. Reports of using stingers in Afghanistan in the 80s states that what is discerned from the ground during daylight is no explosion, but a wing or tail falls off and then the aircraft drops. The trail is not easily seen due to the speed and the effieciency of the stinger. The destruction of aircraft by dettachment of appendages is done through a proximity blast, not contact. The 2.2 pounds is more than enough to detach a wing or a tail in proximity. Larger amounts of explosive is required if you depend on contact or penetration. I do not know what "smooth casement fragmentation" means. Does it mean that shotgun like pattern would be seen on AC surface or does it mean it is only containing and shaping the burst? The above is all I know for sure.

I can read Janes. I can do "China Lake" in Google. I can read all the wild TWA 800 stuff. What I am appealing for is a qualified response from someone reading this who has first hand knowledge about MANPADS forensics. NOt those who have already made up their mind and then start hitting Google.Com. Why is the thesis of a stinger, known to be in the possesion of those who dropped the WTC and considering all of the above, not a very calm and non-emotional consideration? Why does it bring up hysteria and disdain and even anger?

If I were a pilot, I would prefer to know that the approaches are not vulnerable to such an attack, which the first step would be to ascertain whether such an attack took place with AA 587.

It is easy to see the "stretch" being made with flutter, hydraulic lock chatter, composite material - all looking internal in the AC. Why not calcualte the load required to sheer the tail off of a healthy AC and then look for something that provides such a load? Why not first assume the AC was healthy (which years of operation would suggest) and the load was extraordinary.

Obviously this was what was first considered in terms of the wake consideration. In that context, this is the most pragmatic posting in regards to the question of the topic heading.

I find it far more likely that some extraordinary load sheered off a healthy tail rather than consider the total gutting of the second largest airplane manufacturer due to a flawed core design approach, the use of composites. This might be the case, but why not consider the simple answer first?

Few Cloudy

Well Trader Al, I guess that one reason some people don't like the missile theory, apart from not having detected missile damage so far, is that it is out of sync. with the modus operandi on 11.September.

Hallmarks then were,

No import of weapons - use of what was around and improvisation (in this case aircraft full of fuel and plastic knives)

Suicide method in order to get right to the objective (the WTT and Pentagon)

No military hardware and no random crash site.

The AA crash, if caused by missile, would be a new tactic for bin Ladin.

sky9

Trader Al

Listen very carefully I'll say his only once:
Foxtrot Oscar

and start your own thread somewhere else: Better still start your own BB.

arcniz

FFFlyer - re Surely... not tested. Where I come from, as a designer you test it IF IT COULD happen. That's what separates the beginners from the finishers in the technology game - and the folks who design airframe control systems are surely not lightweights or careless. Prudent care would require testing critical control system behavior for every failure imaginable before releasing the design. That this was done should be ascertainable as a matter of record. How do you say subpoena in French?

Probably the structural flaw in the system, if one exists, is that the most careful top to bottom system-level scrutiny occurs once at original design time and is not repeated with anything resembling such intensity when changes are made, despite the acknowledged propensity of software 'fixes' for causing unintended consequences. Removing this long-term management problem is the rationale behind my suggestion of 'perpetual re-simulation' during the airframe product life.

From the POV of a computer looking at the universe in slices of a few millionths or billionths of a second, everything is lumpy. An ordinary gold-plated aircraft toggle switch, for example, has to be checked over a long span of time - say a few milliseconds, to see if it is really open or closed, because the contacts bounce a little when the switch reassuringly thunks into position, and instantaneous sample checks during the bounce interval may say open, may say closed. Electronic and especially digital controls have to be designed to tolerate ambiguity and noise as normal qualities of the universe. But intermittent shorts or opens and other non-characteristic noise may be interpreted in ways different from what one would want. One typical effect of noise on sense lines is that decisions may be deferred - or at least greatly delayed, while waiting for it to go away.

With appropriate software design, any/all events of observed transients on signal and sensor and control wires can and should be recorded as abnormal events for maintenance review. This provides red-flags before hard failure. How many aircraft have this type of diagnostic recordkeeping for the wiring web?


Difference between 'unpredictable and random' vs 'not tested' in this case is (sadly) that you can reconstruct the facts AFTER the event when the process is deterministic instead of random -- just as you could have constructed them BEFORE the event, if sufficient analysis were applied.

If events are truly random, even luck will not help you much in prediction or reconstruction.

arcniz

DaggerDirk - It is interesting, in passing, that Ozzies and others of the British persuasion fall back to the logical Mr. Occam, while Yanks tend to cite the darkly capricious Mr. Murphy.

A thought for your learned aerodynamicist: Electronic, digital, and human control processes all take information from one or more events happening in the past and use that to determine events or processes happening in the future. The control decision is basically: Where, how hard, and for how long?

Time delay and the potential for oscillation are inherent in control sytems. Mass and velocity retard the effects of control, so intended time spans for issued control actions in heavy jet aircraft may be quite long. The trend for modern control systems is to make them heuristic - able to learn from the environment and from consequences of control actions. So, just like a pilot who kicks the pedal harder when he thinks or knows it's jammed, 'smart' control systems will naturally tend to amplify the intensity of their actions if the measured results from control actions are less than expected. Very abnormal outside stimulus (wake) at just the wrong time might conceivably provoke this kind of behavior - on a bad day - but it would probably take a series of precursor events to predispose the control system to excercise very aggressive control authority. In normal and exceptional circumstances, the control systems should have the same objective as the pilot - to push the airframe back into an acceptable envelope.

We do not have sufficient detail or knowledge for any other than armchair analysis, but I strongly concurr with the idea that there is merit in reviewing the class of possibilities suggested by your logic and neatly clarified explanations: the concept that the automatic controls reacting improperly to an uncommon set of circumstances could have scuttled the aircraft by overcontrolling the tail enough to rip it off. Your dialectic of queries is an entirely appropriate way to move the topic forward. Ta.

Ignition Override

Do people really believe that a missile caused the AA 587 tragedy? I'm surprised. Now as for TWA 800, several witnesses saw a smoke/flame trail before the plane exploded. That is quite different.

Dagger Dirk has an interesting AD link, which concerns the A-300 tail.

Good topic guys/gals. I plan to read the whole thing when my days off begin after two trips on a 30 year-old plane.

TraderAl

Regarding missile strike: it is not an issue of believe or not believe, nor whether there is a "conspiracy" afoot. [Perhaps what is obviously a pre-arranged deal between the FBI and the NTSB to have the NTSB take over the investigation is a "conspiracy", but that does make sense in light of the possible panic and considering the screw-up of TWA-800.]

What is strange and alarming, though, is the the uncritical view of almost all, especially it seems professionals and those who usually leap into a dispute, to consider an extraneous force being present large enough to sheer a sound tail. This was the first view most people pursued, which shows how reasonable this approach is, when considring the wake vortex - but as people figured out that could not be enough of a force we are immediately plunging into surmise.

I would, in time of war, eliminate a missile with 100% certainty and provide proof of that certainty. But to even consider it brings such loathing and panic that clear heads do not prevail - just read previous responds to my questions, they are all emotional and those that offer facts are only shallow and based basically on what are popular tenets easily dredged up on the web. Yet people will go into great detail about composite failure on a AC with over a decade of good performance history. (By the way, a MANPAD would fill both Murphy and Occum's law)

If you consider the record to date on the AA 587 investigation, you will find not one attempt to address this issue, with NTSB allowing surface and flippant considerations to grow and dominate this possible thesis. NTSB has almost categorily eliminated the possibility of a bomb, or knife wielding terorists able to claw their way into the cockpit during take-off, but has been studied in not mentioning a missile. Yet with even far thinner evidence, weeks were spent to prove TWA-800 was a missile. Go figure. The NTSB has not mentioned nor examined the possibility of a MANPAD.

I am sure, with such easily available evidence in the wreckage, a MANPAD thesis can be easily refuted or considered with certainty. Why have no qualified experts eliminated this thesis?

Is it conspiracy or being "nuts" to ask a government agency to release such proof during wartime, after our enemy dropped 3 planes and destroyed the WTC and has suffered us to decale another general threat warning?

I. M. Esperto

New York Post

NTSB BIRDBRAINS STICKING HEADS IN ROCKAWAY SAND

By STEVE DUNLEAVY


December 2, 2001 -- LITTLE wonder the National Transportation Safety Board
has bleated for help from NASA to help them out in the tragic crash of
American Airlines Flight 587. The NTSB has shown in the past that it is run
by a bunch of bumbling bureaucrats who couldn't find a needle in a thimble.

Here they were with 265 dead, and God knows how many mourners, giving us
this claptrap that the tail fell off mysteriously.

"No tail fell off, not before the explosion. I swear to that," said retired
firefighter Tom Lynch, who was doing his exercise march along Rockaway
Beach
Boulevard on Nov. 12.

"I had my head up taking in that beautiful, clear day and was staring
straight at the plane.

"It made a bank turn and suddenly there was an explosion, orange and black,
on the righthand side of the fuselage. It was a small explosion, about half
the size of a car.

"The plane kept on going straight for about two or three seconds as if
nothing had happened, then 'vwoof' - the second, big explosion on the right
wing, orange and black.

"It was only then that the plane fell apart. It was after the explosion and
I'm telling you, the tail was there until the second explosion."

Lynch, who lives near the crash site in Belle Harbor, claims he has 13
people who saw the plane on fire before the breakup. Until the explosion
the
tail was intact.

He contacted the FBI, NTSB, Rep. Anthony Weiner, and Sens. Chuck Schumer
and
Hillary Clinton.

"I got no response from anyone," said Lynch, "Sabotage? That's for other
people to decide. At first, we hear there were seagulls in the engine, the
plane was caught in a jet stream and the tail fell off. No damn tail fell
off until after the second explosion."

Jim Conrad, who retired last month as a police lieutenant after 34 years,
accidentally met Lynch in a dentist's office one week after the crash.

"I saw exactly what Tom saw. I was near a stop light at the Marine Parkway
Bridge. First, the small explosion. The plane kept on going, tail intact,
then the big explosion and the plane nose-dived. The first thing I said
was:
'The bastards did it again.'"

For the NTSB to seriously speculate that the bloody tail fell off in the
face of so much evidence that it didn't happen is arrogant and treating us
all like a bunch of morons.

But NTSB spokesman Ted Lopatkiewicz said: "We don't have any evidence of an
explosion [after searching] the wreckage or from the cockpit recorder. It
doesn't mean it didn't happen."

[END]

innuendo

The Dec 10 issue of Time magazine carried an article stating that the FAA and Airbus, (I believe it was those two bodies, lost the copy), had written to American Airlines to express their concern about the companies procedures on use of rudder during turbulence. Other than obviously harsh use of the system I can't imagine how you would misuse the rudder during turbulence. Most yaw dampers do the job unless it is extreme.
Any one from AA know what they are talking about?

stuartbaker

I've been trying to follow up on a Letter in last week's (4-10 Dec) Flight International which talks about "Aircraft-Pilot Coupling" where pilot inputs in-phase with the aircraft yaw movements upset the stability of the control system. In the example given, of an S3 Viking on test flights out of Pax river, the sideslip forces generated were large enough to break off the vertical stabilizer.

I've found a Master's thesis (http://scholar.lib.vt.edu/theses/available/etd-062199-123258) on the subject but does anybody know anything more about this particular event or have a simplified explanation of the phenomenon? (Preferably one that doesn't include the words Nyquist or Bode 0r any 3 dimensional diagrams)

null

Dagger Dirk

Possibilities and Plausibilities

1. A software bug in the frequency-time-domain of yaw-axis control response?

2. Rudder-trim switch fouling its panel and able to drive rudder to full throw (3 AD's = a number of goes to get it right => a possible choice [for mis-installation] of two wrong switches)[see previous post this thread]. "interference between the 408VU panel and the rudder trim control knob, which could prevent the self-recentering of the switch to the neutral position when released, thus causing a rudder movement up to the maximum deflection, and which could lead to critical flight situations."
There is actually another almost identical AD covering the aileron trim switch (which sits alongside on that same 408VU panel). Anyone want to theorize whether someone managed to "Murphy" an aileron-trim switch into the rudder-trim switch position? Above the panel they look identical. Below the panel the wiring has been modified for the lengthier rudder-trim switch #3.

3. and now we add PIO's (Pilot Induced Oscillations)

I've reviewed the PIO academic paper (URL in Technobabble's post) and dismiss that as irrelevant. It is EASILY possible to induce PIO's by changing that F14 pilot's stick to stabilator ratio and simulating a hyd-fail with a low-pressure rate-limited actuator. I severely doubt that that is what happened to AA587. There is a world of difference between plugging into a tanker's refuel basket with a crippled FCS and simply riding out a wake turbulence encounter. One is a precision task requiring maximum concentration / fine control technique / many rapid inputs (and is thereby prone to overcontrol) and the other is simply NOT. My money is on there having been an external stimuli (the wake encounter) excitating a latent flaw in an unserviceable system (the rudder limiter - as per my previous post on this thread). Just as perversely, you could rationally theorise about it having been an undetected FCS software glitch - as Systemsguy has put forward. But a fleet-common glitch would perhaps have been more likely to announce its presence over the long history of the early non-FBW Airbusses - than would a latterly developed hard-flaw in a wiring bundle. If NASA could manage to mix British Imperial and metric in a Mars Mission, then it's perhaps possible that a flight control system could contain some flawed logic - the type that doesn't readily accommodate another failure (the rudder limiter) within the same system. It may not seem as silly as an Imperial versus metric fiasco, but Percentage travels can be a rough game -if unintentionally misapplied inside a malfunctioning closed system.

There's no doubt that it is possible to set up a rudder/fin interaction (courtesy of a flawed FCS input) that could lead to fin flutter, composite failure and fin departure. Stepping ahead a little here, I think that the real challenge is in detecting that flight control system failure mode, not in working out more esoteric inspection methods for composites.

PIO expertise: http://www.piofree.com/white_papers_and_essays.htm

pelican larry

Any Thoughts ???
From: http://www.newsmax.com/archives/arti...7/200515.shtml

Sabotage Explains Flight 587 Crash, Says Expert
Dave Eberhart, NewsMax.com
Tuesday, Dec. 18, 2001
Federal investigators still have no evidence indicating that a benign structural failure played a role in the tail breaking off of Flight 587 last month, sending the plane tumbling into Queens, N.Y., according to AviationNow.com.
But as National Transportation Safety Board and other safety experts wrestle to solve the mystery of the powerful forces that ripped the plane’s fin off and then cast the engines from their mountings, one aviation expert said sabotage of the aircraft’s left engine while still on the ground could explain what shook the aircraft to pieces.

Expert Marshall Smith opined, "A single point failure, the in-flight actuation of the left engine thrust reverser, can account for all three observed phenomena of the clean breaking off of the tail and the failure of both pylons holding the engines.

"If the left engine thrust reverser had either partially or completely actuated during flight, it would cause the plane to go into a flat spin to the left. The airplane would spin something like a flat Frisbee with the right engine pushing forward and the left engine pushing backwards,” Marshall explained.

"Within a second of the flat spin occurring, the sideways windblast would rip off the tail assembly, since it was never designed to take such a side blast of air.

"As soon as the tail assembly broke off, there is now very little wind resistance to the flat spin. At this point the engines would cause the aircraft to spin even faster with the g-forces away from the center of the spin becoming so great that both engines would be violently ripped off the wings and thrown outward away from the plane,” Marshall said.

Marshall’s opinion is that the spin accounted for why the engines were found so far away from the crash site and why the tail came off first.

Terrorist Scenario

The mechanical engineer, aviation ground school instructor and former NASA adviser painted this scenario:


During the night, a terrorist saboteur disguised as a ground crew mechanic reached up in the back of the left jet engine of the American Airlines Airbus and cut the hydraulic line going to the thrust reverser actuator and the control safety sensor lines.

The next morning after the jet engines were started, the hydraulic fluid began dripping from the cut line.

When the aircraft was about 3,000 feet in the air, the sound of an "airframe rattle” was heard in the cockpit voice recorder (CVR) record. Cause: the tampered-with left thrust reverser had started to close, causing the plane to turn to the left.

The pilot compensated by applying right rudder to bring the nose back to straight flight by turning to the right.

The aircraft commenced a "side slip.” During this condition, the burbling air flowing over the extended control surfaces made the plane shake, rattle and roll, accounting for the airframe rattle noise heard on the CVR at 107 seconds into the flight.

The pilot thought he had overcompensated, worried about losing too much airspeed, and returned the controls to normal. The rattling momentarily stopped as indicated on the CVR.

The plane continued to turn back to the left.

Seven seconds later, one of the flight crew commented about "air turbulence.”

The pilot again tried to compensate for the plane's strong drift to the left caused by the partially closing thrust reverser by again applying right rudder and opposite aileron. The same rattling sound is heard at 121 seconds into the flight.

Four seconds later, at 125 seconds into the flight, the first officer calls for "full power,” presumably to compensate for the side slip, which had slowed the plane down to dangerously low speed.

As soon as the power went to full, the spinning effect caused by the partially or fully actuated thrust reverser caused the plane to spin out of control in a flat spin.

Two seconds later, at 127 seconds, the CVR indicated the flight crew making a comment about being out of control. No more comments are made after that, and the recording ends 17 seconds later when the plane hits the ground.

Fighting to control the aircraft, the pilot held full right rudder and hard left aileron just as the left thrust reverser came into the full-on position. The application of full power greatly increased the turn to the left, created a huge side force on the tail and rudder assembly, and snapped them off cleanly.

Within another second, without the vertical tail assembly to slow the spin, the plane spun violently to the left about the center of gravity of the airplane. The plane spun horizontally with the full power from both engines increasing the spin faster and faster until both engines broke off.

The flight crew at the front was thrown violently forward with such g-force they were instantly rendered unconscious or killed, explaining why no more comments from the flight crew are heard after applying full power.

With the plane completely out of control and the engines still running at full power, the engines broke away ripping the fuel tanks in both wings and igniting the plane.
Wake Turbulence Discounted

Marshall created his saboteur scenario because he concluded early on that it is not possible for any type of wake turbulence from a preceding jet to rip off the tail of an airplane. Furthermore, he concluded, even with the vertical stabilizer gone, Flight 587 would not have gone out of control in such a way that both engines would also fall off.

He pointed to a 1985 incident where a Japanese Boeing 747 with the vertical tail assembly completely torn away continued to fly in large circles for over half an hour before hitting a mountain.

According to Marshall, Flight 587, an Airbus A300, used a modern "fly-by-wire” computer system and could fly quite easily with complete loss of the vertical fin and rudder.

"Most air accident investigators would easily conclude that the chances of three simultaneous airframe failures all occurring at the same time is not probable. It must be one or the other but not all three. It would be much easier to conclude that something else actually caused all three failures,” Marshall said.

Marshall pointed to a statement by New York City Mayor Rudolph Guiliani at a news conference Nov. 14 that the rescue workers recovered 262 bodies including "a man still holding a baby.”

"Certainly no man can be strong enough to hold on to a baby through that force, unless instead the plane was in a flat spin. For the passengers in the center of the plane, the force would have been downward [not forward] as the plane hit the ground, and the baby would be simply forced deeper into the man’s lap as he sat in the passenger seat.

Further clues pointing to his theory, said Marshall: news videos of the crash scene as firemen put out the flames. A large section of the central part of the plane is lying on the ground almost intact but in flames.

Midnight Blue

Quote Marshall:
He pointed to a 1985 incident where a Japanese Boeing 747 with the vertical tail assembly completely torn away continued to fly in large circles for over half an hour before hitting a mountain.
End Quote.

He better doesn´t call himself an "expert".
The vertical tail assembly of the Japanese 747 never was torn away!
They lost control, because they had no more hydraulic pressure to move the control surfaces due to rupture of hydraulic system in the tailsection. (Source http://aviation-safety.net/database/type/index.htm)

And he should calculate how much "rpm" of the A300 would be necessary, to create a force, which could kill the cockpitcrew in an instant (I doubt this). Must have been several rotations in less than 2 minutes???
Why didn`t anybody mention a "frisbeeing" airplane...

PaperTiger

The vertical tail assembly of the Japanese 747 never was torn away!

Sure looks like it was to me.
http://www.airdisaster.com/special/special-jal123.shtml