AF 447 transmitted failure sequence (from one of my airline pilot buds)

[DemoTex]

There is some information getting out about the mechanical failures
that were automatically transmitted to the company by the aircraft
ACARS systems (a continuously operating communication system that is
sending messages). Every malfunction of the plane is automatically
transmitted to the company at the same time the pilots get the warning.

Here is the time line that I have found along with the failures. All
time is GMT or Zulu time. East coast time is 4 hours less.

1. 0210Z Autopilot failure. This would be 2210 or 1010PM east coast time.
The pilot is now hand flying the plane in possibly severe turbulence
due to the storms.

2. 0211Z ADIRU failure (air data inertial reference unit) The pilots
primary attitude, airspeed, and altitude displays are starting to
fail. There are backups and I don't know how much has failed.

3. 0213Z SEC 1 Fault, Alternate flight control laws: This is one of
the computers amongst several the pilot uses to control the elevator
and spoilers (the roll axis). He still has backups, but now the
computer systems for flight controls are degraded and the pilot has
to be careful not to lose control in the turbulence. The flight
control has degraded to what we call alternate law. Time to
concentrate on some good hand flying at night in turbulence.

4. NAV warnings and Flight control warnings: Nav warnings are the
least of his concerns at this point and I still don't know what he
had left for flight instruments, but he has lost some and maybe most
primary flight instruments. It's time to ignore warnings and failures
and just concentrate on what is working. Maintain control of the
plane, wings level, proper pitch attitude and power settings until we
get through the turbulence.

5. Last ACARS message at 0214Z 35,000 feet, failure of cabin
altitude, electrical problems, pressurization problems. The pilot was
unable to safely hand fly the plane. I don't know why. The airplane
is breaking up. The pilot has lost control.

Notice that 4 minutes have elapsed since the autopilot failure. At
0214z the multiple warnings and the pressurization problem make me
wonder if the pilots were unable to keep the airplane under control,
overstressed it at high altitude, high airspeed, in severe turbulence
and the overstressed airframe came apart. The pressurization warning
may mean the fuselage has broken open and this last transmission
indicates that after 0214Z no electrical systems were working because
the plane is coming apart as it falls out of the sky and the
structural integrity of the wings, fuselage etc. is destroyed. The
wreckage for Air France will be scattered over a very wide area if
this is true: i.e., at least several miles as seat cushions are
falling from the sky starting at 35,000 feet.

That is the text of my friends message. I haven't studied it yet. I promised DU that I'd post anything I get on AF447 that seems plausible .

I'm headed back out into the wilderness. I'll be back on DU tonight.

[thereismore]

1. The corporate f**kers would not bother tell this to anybody until now? nt

[anigbrowl]

3. What are you talking about? this info was available last week

The details of the ACARS messages was released very shortly after the crash. The interpretation is (necessarily) speculative.

[DemoTex]

18. Good points.

Since I am out here in the Deschutes (Oregon) wilderness, I get little information. The guy who sent me this email is not a trained accident investigator. Like me and a number of my contacts.

[MADem]

2. Those poor people.... how are things in the fire tower, BTW?

[Stinky The Clown]

4. The concept of "hand flying the airplane" is troubling to me seeing as how it is an Airbus

There is no actual 'hand flying'. It is all fly by wire, with controls actuated by electrical signals and no analog input.

If the plane's problems started with an electrical anomaly (various system failures as mentioned in the ACARS notes in the e-mail above), why would it not be a reasonable follow-on that the 'manual' controls would also suffer some degree of failure?

I don't know beans about airplanes, but the foregoing jumps to mind as I read this missive.





The other thing that's been nagging at me is the structure of the vertical stabilizer and rudder. Didn't we have an Airbus suffer a catastrophic failure of that on climb-out from JFK just a week or so after TWA 800, just off the South Shore of Long Island? The plan was, I think, on its way to South America. If the tail is as it was then, violent turbulence and tail failure seem also a possibility, although everything seems to be pointing elsewhere.

[RamboLiberal]

5. I think you're thinking of the American Airlines flight 587 that lost its tail

November 2001 - when terrorism fears were so high. They claim the over correction of rudder by copilot when he encountered wake turbulence from a previous plane was the cause. I never completely bought that a trained pilot would use the rudder in a manner that would cause the tail to tear off the plane. I'm suspicious of composite in this instance and how they are adhered to the fuselage. I'd be curious if the tail of the Air France plane may not have been torn off in turbulence and contributed to the crash.

AA 587 http://en.wikipedia.org/wiki/American_Airlines_Flight_587

[Stinky The Clown]

9. Yeah, that's the one

No matter what the investigation showed, when major parts fall off a plane because a pilot is attempting regain control, something just isn't right.

Yeah ... composites, in this instance, worried me, too.

[RamboLiberal]

12. Interesting snippet here on the recovered tail section-AF447

From images of the recovered tail section, the damage looks like a lateral fracture, said William Waldock, who teaches air crash investigation at Embry-Riddle Aeronautical University in Prescott, Arizona.

"That would reinforce the idea that the plane broke up in flight," Waldock said. "If it hits intact, everything shatters in tiny pieces."

Goelz said the faulty airspeed readings and the fact the vertical stabilizer was sheared from the jet could be related.

The Airbus A330-200 has a "rudder limiter" which constricts how much the rudder can move at high speeds. If it were to move too far while traveling fast, it could shear off and take the vertical stabilizer with it.

"If you had a wrong speed being fed to the computer by the Pitot tube, it might allow the rudder to over travel," Goelz said.

http://www.cbsnews.com/stories/2009/06/09/world/main5074961.shtml?tag=topHome;topStories

One U.S. aviation consultant said the almost perfectly intact vertical stabilizer provides evidence that the jetliner broke apart before hitting the water and said it resembles the condition of a tail that was torn off an Airbus that crashed in New York City in November 2001.

In the 2001 accident, U.S. investigators found that pilots for the American Airlines A300 jet had encountered turbulence in the wake of another jet and that they applied so much force on the rudder that the tail tore off the plane, sending the craft into a sharp dive.

Robert Ditchey, an aeronautical engineer, former pilot and retired airline executive, said he was struck by the similar conditions of the Air France tail and the American Airlines tail. Ditchey said he doubted the tail would have remained so well preserved if it had crashed into the water while still attached to the fuselage.

Investigators have learned that the Air France jetliner's autopilot was disengaged for unknown reasons about four minutes before the crash.

Ditchey said the new evidence opens the possibility that after the autopilot was disengaged, the plane encountered weather-related turbulence. The pilots may have overcompensated, as pilots did in 2001, and caused the tail to rip off, he said.

http://www.latimes.com/news/nationworld/world/la-fg-brazil-crash9-2009jun09,0,1886961.story

Airbus is only large passenger plane I've ever heard of that overcompensating had torn the tail off. Love for DemoTex and our other DU aviators weigh in if they know of.

[Stinky The Clown]

17. Jaybus! That's really freaky.

[leveymg]

20. Tails and rudders have torn off A300 series three previous times, B737 also had rudder problems

On March 6, 2005, Air Transat Flight 961, an Airbus A310-308, took off from Cuba's Veradero Airport. Crew was able to return and land aircraft safely. Rudder sheared off.

In 2002 a Fed-Ex A300 cargo plane also experienced the loss of a rudder. In FedEx's own test on the rudder on the ground, engineers claimed the actuators, the hydraulic system which actuates the rudder, tore a large hole around the hinges, in exactly the spot where the rudders of both flight 961 and flight 587 parted company from the rest of the aircraft. See, http://www.yachtsurvey.com/composite_troubles_in_aircraft.htm

November 2001, American Airlines Flight 587 departed from Kennedy Airport. Plane crashed after the entire tail section snapped off, blaimed by NTSB on rudder overcorrection to wake turbulence.

OTHER A300 SERIES RUDDER PROBLEMS: http://www.ntsb.gov/Pressrel/2002/020311.htm

NTSB Advisory
National Transportation Safety Board
Washington, DC 20594
March 11, 2002
SIXTH UPDATE ON NTSB INVESTIGATION INTO CRASH OF AMERICAN AIRLINES FLIGHT 587


--------------------------------------------------------------------------------

The National Transportation Safety Board today released the following updated information on its investigation of the November 12, 2001, crash of American Airlines flight 587, an Airbus A300-600, in Belle Harbor, New York, which resulted in the deaths of all 260 persons aboard and 5 persons on the ground.

Vertical Stabilizer and Rudder

Testing has begun on small composite samples that have been removed from the accident aircraft's vertical stabilizer and rudder - which separated from the aircraft during the accident - at NASA's facility in Hampton, Virginia. Three of the six lugs (the attachment points between the stabilizer and the empennage) will be taken to the U.S. Army's Research Laboratory at Aberdeen Proving Ground in Maryland to undergo a CT-Scan (commonly referred to as a CAT-Scan) to further define any sub-surface damage. Two other lugs are being taken to the Lawrence Livermore National Laboratory in California because they are too large to be accommodated at Aberdeen. The Board has not made a decision yet on whether to take the sixth lug for further testing.

Previous Airbus Events

The vertical stabilizer of an American Airlines Airbus A300-600 that experienced an upset event in 1997 (see February 25 advisory for details) was removed from the aircraft and has undergone ultrasonic examination at American Airlines' maintenance facility in Tulsa, Oklahoma. This aircraft experienced high lateral loads associated with reversing rudder movements during the 1997 accident and Safety Board, Airbus, American Airlines and BEA (the NTSB counterpart agency in France) investigators wanted to see if any damage might have resulted from that event that wasn't detected in the post-accident inspection. This event was classified as an accident because a flight attendant received a serious injury.

Investigators report that an indication of damage (possibly delamination) has been found that apparently was not present at the time of manufacture. The indication is at the pin bushing of the right rear lug. Airbus has informed the Board that the stabilizer will be removed from service. The Safety Board will conduct further examinations of this component once the current testing in Tulsa is complete.

Airbus and the FAA are evaluating service history data to identify other upset or maneuver events, if any, that may warrant further investigation. Airbus has advised the NTSB that in May 1995, a FedEx Airbus experienced large rudder deflections, but not rudder reversals. The deflections were the result of a rudder trim/autopilot interaction. A course of action will be taken once the evaluation is complete (This is not to be confused with the FedEx Airbus A300-600 that was found during maintenance in February to have a damaged rudder . That rudder is being taken to Germany for further examination).

"Filtering" of Flight Recorder Data

As previously reported, some of the information dealing with flight control movements contained on flight 587's flight data recorder was filtered (see January 15 media advisory for details). The Board has made good progress in defining the original filter characteristics, and in designing an inverse filter to better define actual rudder movement. The Board still has some on-aircraft testing to complete to finalize its work in that area. Airplane motion and loads will be further refined through simulation and analysis.

To be fair, however, there were seven fatal and near-fatal accidents attributed to problems during the 1990s with rudder actuation in the Boeing 737. See, http://www.airlinesafety.com/faq/B-737Rudder.htm

Houston, we have a problem.

[Stinky The Clown]

23. It seems to me that three repeats of the same ***massive*** failure ought to raise a few alarms

And now ..... maybe ..... a fourth repeat.

[leveymg]

25. You would think. Read the article linked re: moisture problems with composites

But, I think there are two related problems with the A300 tail. The composite tail is bolted into a bracket on the aluminium fuselage. The failure point seems to be a shearing where those two disperate materials meet. A conventional aluminium tail wouldn't suddenly shear at the mounting points, but would bend and stretch along its length under excess load, like a palm tree in a storm.

If Airbus wants to solve the problem, there are two possible fixes - an aluminium tail or a conformal style composite tail.

Conformal structures wrap part of the way around the fuselage. The material needs to laid out so that it allows a greater amount of "stretch" distributed over a larger area along the the sides of the fuselage so the verticle stabilizer has some additional support to account for side-to-side loads.

The other problem has to do with rudder attachment points shearing off, possibly due to moisture buildup, repeated freezing and unfreezing, resulting in deterioration of the carbon fiber and de-lamination of structures inside the verticle stabilizer. Perhaps, heating might help, but this may be a basic problem with composites that hasn't been addressed. See, the link in my last post.

The problem with composites are showing up on tails rather than wings, and rudders rather than flaps and ailerons. Again, I think this is due to the fact that the mounting points of a verticle stabilizer are more highly stressed than wings. The tail is attached into a narrow slot at the rear of the fuselage, and the rudder has a relatively larger area compared to ailerons. Deployment of flaps are limited by airspeed, otherwise they would tear off - that's not a common problem. In addition, wings are designed to flex to a far great degree than are tail surfaces in many aircraft, which is a fundamental design problem in extreme turbulence.

The B737 problem appears to have been with the hydraulic servo mechanism that actuates the rudder. The 737 has a very large verticle stabilizer and rudder area to its overall length to prevent side-slipping. It handles like a very short wheelbase car that oversteers - like an early Porsche. Pilots sometimes used too much rudder to recover from excessive yaw angles, and faulty servos momentarily locked up or wouldn't reverse fast enough (see other link) - again, just like power steering units in some cars. The result was repeatedly fatal, but I haven't heard of this problem recurring in recent years - the redesigned 737 with winglets is much more stable in yaw than the older versions. You can see the difference from the ground in the way the newer ones rotate smoothly around the wings rather than turning sharply on the tail, as did the old 737s.

[RamboLiberal]

35. Wow - that's a fantastic explanation of the problem

that's easy for us laymen to understand. Makes sense. I was a little more familiar with the servo mechanism problem on the 737 since I read amd watched everything I found on it because of the interest the US Airways flight 427 crash near my hometown of Pittsburgh.

[RamboLiberal]

I'm well aware of the rudder problem on 737 craft

since I'm from Pittsburgh and US Airways Flight 427 crashed 10 miles or so from Pittsburgh airport due to that rudder failure. I didn't like to fly on a 737 till they got that resolved. The rudders didn't come off the aircraft on the 737 though.

[hootinholler]

10. Didn't they find the tail section?

I though I saw that briefly on the news yesterday.

:shrug:

-Hoot

[RamboLiberal]

11. Yes they did - pictured

[FightingIrish]

6. On an unrelated topic, there is stuff headed your way.

There's a cell showing hail just north of La Pine with the top at 24,000 ft. and a line of building storms south of it. Stay safe.

[DemoTex]

16. No shit! It kicked my ass in Bend. Separate post coming up.

I'm back at the lookout tower now.

[high density]

7. Some facts mixed in with a lot of wild speculation NM

[jmowreader]

14. It's speculation, but not all that wild

DemoTex is a retired airline pilot who has proven to be a good resource on aviation-related matters.

[high density]

15. That doesn't change the fact that we have no idea what the pilots were doing

We know the automated messages sent from the plane, but we don't necessarily know what caused them. ACARS messages are likely telling us only part of the story. This information also do not show us how the pilots responded to the problems. We can say "the plane is under alternate law now because X failed" but talking about the pilots in that system is inserting assumptions into the situation.

"Last ACARS message at 0214Z 35,000 feet" -- I don't think we have any evidence showing the airplane was at 35,000ft at this point.

[malaise]

8. Thanks DemoTex

The debris is also suggesting that the aircraft fell apart.

[The Velveteen Ocelot]

13. Interesting.

Did the autopilot fail or was it disconnected? The A330 has two autopilot systems, so it seems odd that both would fail. Only one can be used at a time, so if one had a fault they could have used the other one. If an autopilot is manually disengaged there would be an aural warning in the cockpit (which should turn up on the CVR, if they ever find it). Also, if a pilot moves a rudder pedal or a sidestick to a significant degree that autopilot (whichever one is engaged) senses this as a fault, which gives you a warning message and a continuous repetitive chime. So -- did an autopilot fail as the result of some other malfunction, or was it manually disconnected? Did both fail?

One possibility for an autopilot fault could be the failure of at least two ADIRUs and/or the failure of at least one PRIM and one SEC computer. It does appear that one of the ADIRUs and one of the SEC computers failed for unknown reasons. Lightning strike? The 330 has about 40-some static wicks, but a bad lightning strike possibly could do it. So if you lose some of these systems you could have problems, but there are three ADIRUs and two SECs -- you'd have to lose a lot of independent components. Also, if normal AC power is lost only #1 is powered from an essential bus, and air data reference information would still be available on the captain's flight instruments.

The failure of SEC #1 shouldn't have been a big deal by itself, since the SECs redundantly control rudder trim and limiting, yaw damping, and flight in direct law. If all 3 PRIMs were working they should have remained in normal law instead of degrading to alternate law (which still retains a lot of flight control protections other than pitch protection). However, the failure of more than one ADIRU could have caused a reduction to alternate law. Even so, fying in alternate law is not really a very significant degradation, although if you go to ALT 2 (requiring the loss of more systems) you lose bank angle protection.

There was obviously a lot more going on than we know yet. These failures should not have caused the accident; I suspect there was a more catastrophic occurrence that caused the failures. The loss of pressurization suggests a structural failure, and there's no help for that. Even so, the 330 is a very good, safe aircraft with a lot of redundant systems, and I wouldn't hesitate getting on one for a minute.

Thanks for the information. I am geeky enough to be fascinated by these mysteries, as tragic as they are. If you get any more info please post it.

[chicago legal pro]

19. Why didn't the crew radio any emergency messages?

[Crabby Appleton]

22. Aviate - navigate - communicate

that's the priority for the pilot. If you notice the plane landing in the Hudson river, the pilot was very terse and not talking much - very short answers - the pilot is busy.

[chicago legal pro]

26. According to the time scenario above several minutes went by

The Hudson river pilot gave several messages from the start of the problem. These pilots didn't utter a word.

[The Velveteen Ocelot]

27. Maybe they were in an even bigger crisis than the Hudson River crew.

It sounds like they were in the middle of a catastrophic event or series of events (Fire? Rapid decompression? Total electrical failure?), which were so serious that the last thing on their minds was calling ATC. A basic rule of flying is "aviate, navigate, communicate," in that order. I suspect that they had their hands full just trying to aviate.

[chicago legal pro]

28. So in a four minute period no one on the crew had time

to push a button and declare an emergency. I'm glad I don't fly Air France.

[The Velveteen Ocelot]

29. What good would it have done if they had?

They were far out over the ocean; declaring an emergency and a buck fifty would have gotten them a double skim latte. All pilots -- not just Air France pilots -- are taught to handle the emergency first, then communicate only when the aircraft is under control. If they had something like a fire or a structural failure or a loss of control, there's no way they'd have had any time to bother with a radio communication that wouldn't have helped them. And if they had a total electrical failure they might not have been able to use their radios anyhow.

[chicago legal pro]

36. No indication of electrical failure until four minutes in

Pilots should always think of possible rescue no matter where they are at. Hopefully the pilots that fly me can think and chew gum at the same time. They can do more just attempt to fly. There is more than one person in the cockpit last time I looked.

[OnTheOtherHand]

30. your point being?

If I'm flying a crippled plane over NYC, probably I should give ATC some clue where I'm trying to take it. If I'm over the Atlantic...?

I don't think we even know what the crew did or tried to do. Maybe at some point we'll have some grounds for second-guessing them -- but over their failure to "declare an emergency"? Seems thin.

[chicago legal pro]

34. If you are flying over the Atlantic where there is no radar

is exactly where you should give the ground some heads up.

[konnichi wa]

31. The crew is 2 people trying to keep their vehicle from crashing into the sea.

Assuming they could find a few spare seconds to chat with someone on the ground, how would that have helped?
You must have some idea since you've formulated a conclusion based on zero experience in the situation...
I'm glad I'm not depending on you for ...like...anything. :eyes:

[chicago legal pro]

37. And I'm glad I don't depend on you for anything ever.

He helps to give people on the ground some idea of where you are at if you are in an emergency. I was a crew member on a KC-135 in the air force -- not a pilot -- but I do know our guys could do more than one thing at a time. Our pilots could "aviate, navigate, communicate," all at the same time.

[konnichi wa]

40. Those things are to be done consecutively, not concurrently.

Stick to briefs, legal beagle.

[chicago legal pro]

43. Maybe you should join the USAF and tell them what they are doing wrong.

[Hassin Bin Sober]

41. Your ill informed opinion doesn't jibe with other serious fatal airline accidents.

In accident after accident one of the common characteristics is lack of communication while the pilots are trying to fly the airplane.

The first example that comes to mind is AA191 - The controllers where asking specific questions but the pilots ignored them as they were rolling in to uncontrolled flight. Admittedly, the problem to crash time frame is shorter but it's the first accident that comes to mind since it hit close to home for my family.

The control tower voice recorder recorded a controller contacting the airliner when he witnessed the engine separation just after take-off, but the crew didn't answer as they were too busy trying to save the aircraft. The recording begins with the controller talking without transmitting on the frequency: "Look at this, look at this, he blew off an engine. Equipment, I need equipment, he blew an engine. Oh, shit...!!" The controller then transmitted, "And American one, uh, ninety-one heavy, you wanna come back and to what runway?" Without keying the mic, the controller can be heard: "He's not talkin' to me ...yeah, he's gonna lose a wing. There he goes, there he goes..."

For more, look at AA587 Long Island or Dallas Delta191 or the Pittsburgh US Air #427 737 crash or the plane that went in to the Potomac.

Perhaps you should look up your old AF buddies and ask them if they would be radioing their position if they were in a stall spin or inverted. I think I know the answer.

BTW, I've been a pilot for twenty years.

[chicago legal pro]

42. Left out the flight that went into the Hudson

For some reason. I don't have access to the transcripts and time frames to the other flights you mention and you provided no links.

[Hassin Bin Sober]

44. WTF?

The Hudson incident was mentioned up-thread. The point was already made Scully used very minimal and abrupt conversation .... i.e. to get a vector to the nearest airport - The "N" part of the A.N.C. doctrine that you can't seem to grasp. Notice how once the tower and their vectoring ability became of no use to the Captain he ceased communication?

You are welcome to look those flight numbers up on wiki. There are numerous links to CVRs, tower comm. and accident reporting sites.

Add another couple to your research - The Continental/Colgan 3407 - not a peep from the crew. Probably because they were trying to fight a spin. It tends to occupy your time. Or American Eagle 4184 .... two minutes the crew fought and regained control at least once and not a peep from the crew to the tower.

Clearly the Air France pilots were in a world of shit. The cockpit probably sounded like a 4th of July parade with all the alarms going off.

.....but you rode in the back of a KC-135 so you are an authority. :eyes:

[chicago legal pro]

45. Didn't say I was "an authority" did I?

But "riding in the back" of a KC-135 I have been on a hella of lot more flights than most who post here. And no, I don't bow down to pilots just because they are pilots. Your 20 years doesn't mean shit to me by itself.

[Hassin Bin Sober]

47. "been on a hella of lot more flights than most who post here"

So has a flight attendant.

[chicago legal pro]

51. Typical patronizing comment about "flight attendants"

Of course they know nothing, right? Typical comment from a know it all internet "pilot".

[Hassin Bin Sober]

52. I would rather trust a flight attendant behind the controls than some ....

...ex-airforce honey wagon attendant.

[chicago legal pro]

54. The closest you have come to a cockpit

is when your mother bought you a flight simulator game. Watch out for the turbulence around your desk!

[Crabby Appleton]

48. They should put an internet know-it-all such as yourself on all flights

so you can instruct the pilots in case of an emergency.

[JBoy]

50. Cheaper to just put his book in every seat back:

[DaveinJapan]

21. Perhaps this is a very dumb question but...

...is it really *better* for the plane to be on autopilot if in fact it was extremely turbulent and/or bad weather?

Seems like you'd rather have the plane be in the pilots' hands at times like those. (??)

[ThomWV]

24. Don't airplanes fly through hurricanes without breaking up?

I do not understand what about turbulence, even severe turbulence, would cause a well maintained commercial aircraft to 'break up' while still in the air. Propeller driven aircraft (old ones at that) flying at lower altitudes up to, into, and through hurricanes, flying through the eye-wall where turbulence is extreme, do not break up even after hundreds of episodes.

So while the sequence of events may be accurately portrayed it still does not answer the basic question of what caused the plane and flight to fail. Yes, it got shook, but shaking should not break a plane.

Is there an aeronautical equivalent of the Rogue Wave? Is there some phenomena so powerful it can swat an aircraft from the sky but so rare that we have no good knowledge of it?

[Trekologer]

33. It likely wasn't a single cause but multiple ones that came together

If you're purposely flying into a storm and are accepting the plane being roughed up a bit, then the plane will probably be fine. If you're flying at a high speed, unexpectedly into a storm, and fighting against the turbulence, its highly possible to put significant stress onto control surfaces, wings, the vertical stabilizer (tail fin) to the point of damaging or breaking them off the plane.

Here's some complete speculation... the plane flies into the storm and experiences extreme turbulence. Due to the problem pitot tube, the pilots and/or autopilot think that the plane is slowing down so it speeds up beyond a safe speed. In order to attempt to maintain control of the airplane during turbulence, the pilots and/or autopilot move the rudder beyond the maximum that would usually have been allowed at the speed the plane was at (due to the pitot tube). The stress to the rudder to too much and the vertical stabilizer is sheared off of the plane. Control is nonrecoverable and the aircraft continues to break up.

[groundloop]

38. Hurrican Hunter Aircraft...

Are beefed up to be able to handle the stresses. They also undergo extensive inspections on a far more frequent interval than would be possible for airliners. I've seen presentations on these, and talked to one of the pilots a few years ago. It's my understanding that while yes, there are thunderstorms and even tornadoes embedded in hurricanes, they're not as severe as the thunderstorms we're talking about in this case. 100 mph updrafts would be pure violence.

The timeline information provided is very telling. In most crashes there isn't any ONE event that caused it, but rather a chain of events. In this case it's sounding like it's POSSIBLE that electrical/computer failures led to a situation that overstressed the tail.

Those are good comments about concentrated stresses encountered in composite structures with the use of fasteners and joining to adjacent structures. It's generally understood that when properly implemented composites are superior to aluminum. Composites aren't subject to fatigue the way aluminum is, with a virtually unlimited lifetime. Just the fact that composites are used in a plane is no reason to condemn it.

[crimsonblue]

39. I understand that composites can be superior to metal...

and I have a decent background in materials engineering. The weakest point will always be at the interaction between two unlike materials. Composites have a much lower ductility and resilience than metals, and very strong forces that may only cause deflection and bending in metals may cause cracking in composites. Since there is a documented history of catastrophic failure of planes utilizing a similar tail design such as this, it is a valid area to look at.

[Hassin Bin Sober]

46. There was an interesting episode of Air Emergency where they ...

... discussed a chopper crash in the North Sea caused by a failure, due to lightning strike, of the composite/metal connection on the tail rotor. One of the investigators/experts closed out the episode with an ominous warning regarding these connections and their inability to resist direct strikes .... kinda makes one wonder.

[BlueCollar]

56. As an aircraft structural mechanic I'd tend to agree

I've been a structures mechanic for over thirty years. Composites have many desirable qualities, but rapidly lose their integrity if damaged. A minor de-lamination caused by impact or corrosion is very significant when it comes to composite integrity.

[Trekologer]

32. That's an interesting question

Unfortunately, I forgot where I read this, but the answer to your question differs between the two major airplane manufacturers: Boeing and Airbus. Boeing planes will turn the control over to the pilots (autopilot disengages). Airbus planes will take over control from the pilots (autopilot remains on).

[trof]

49. Turbulence penetration

It's been a while, but...
I think I remember:
1. Autopilot/autothrottle: OFF
2. EPR (Engine Pressure Ratio) or RPM at 80%.
(Airspeed approx 250 KIAS (Knots Indicated AirSpeed)?)

Do NOT chase airspeed or altitude.
Maintain a constant attitude (level) as much as possible.

[rschop]

53. re: failure of flight AF 447

I hate to ask a dump question but why in Christ's sake were they flying in a thunder storm?

Their radar could have spotted this storm 30 miles away, and they could have flown around it. Any thunderstorm at their flight level of 35-40,000 feet has to be a monster that should have been avoided at all costs.

[The Velveteen Ocelot]

55. Sometimes onboard radar does a thing called attenuation.

Basically, what happens is that an area of a cell is so intense that the radar can't interpret it and it reads black (instead of red, meaning severe precip), and it can be mistaken for a hole in the storm. Also, radar can detect only precipitation at a distance, not conditions like windshear. While it might display a hook echo, which indicates rotation (possible tornado) that could be avoided, radar isn't foolproof and it doesn't detect all possible weather threats.