Why did Air France 447 go down?

[deicer]

Don't know if this was posted yet,,,

http://www.spiegel.de/international/world/0,1518,766148,00.html

The Hanover legal practice of Ulrich von Jeinsen, which composed the letter, and the Berlin aviation law expert Elmar Geimulla made mention in the letter that there could be "criminal consequences" should indications of a software error not be thoroughly investigated and another Airbus crashes for the same reason.

Von Jeinsen's motion is primarily based on the expert opinion of Gerhard Hüttig, a professor at the Institute of Aeronautics and Astronautics at the Technical University in Berlin. Just over a year ago, Hüttig recreated the Air France crash in a flight simulator. In the course of the exercise, Hüttig noticed a strange anomaly in the plane's reaction once it goes into a stall. The trimmable horizontal stabilizer, a flap instrumental in keeping the plane on an even keel, automatically adjusted to push the nose of the plane skyward.

Hüttig, a former Airbus pilot himself, and other pilots present for the test were unable to push the nose of the airplane down and thereby escape the stall.

When the BEA released its preliminary report last Friday, Hüttig immediately zeroed in on data relating to the trimmable horizontal stabilizer. During the final minutes of flight AF 447 as it plunged toward the Atlantic, the flap moved from a 3 degree deflection to a 13 degree deflection, almost the maximum possible. "The phenomenon is startlingly similar," he told SPIEGEL.

A Quiet Reaction

Hüttig passed along his simulator findings to Airbus, the European Aviation Safety Agency (EASA) and to BEA. On Oct. 27, 2010, Hüttig received a response from EASA which said that Hüttig's theory was inconsistent with the "current state of knowledge." "We suspect that the anomaly you found originated with the simulator you used in the study rather than with the airplane model A330," the response read.

Hüttig and Jeinsen told SPIEGEL that the data recovered from the wreck of flight AF 447 would now seem to have corroborated the simulator findings. Furthermore, Airbus has quietly reacted to the safety loophole. In a communiqué to airlines, Airbus provided a new version of pilot instructions for dealing with a stall. Furthermore, in the January issue of its internal safety magazine, there is a mention of manually trimming the horizontal stabilizers.

[Don Hudson]

Don't know if this was posted yet,,,

http://www.spiegel.de/international/world/0,1518,766148,00.html

The Hanover legal practice of Ulrich von Jeinsen, which composed the letter, and the Berlin aviation law expert Elmar Geimulla made mention in the letter that there could be "criminal consequences" should indications of a software error not be thoroughly investigated and another Airbus crashes for the same reason.

Von Jeinsen's motion is primarily based on the expert opinion of Gerhard Hüttig, a professor at the Institute of Aeronautics and Astronautics at the Technical University in Berlin. Just over a year ago, Hüttig recreated the Air France crash in a flight simulator. In the course of the exercise, Hüttig noticed a strange anomaly in the plane's reaction once it goes into a stall. The trimmable horizontal stabilizer, a flap instrumental in keeping the plane on an even keel, automatically adjusted to push the nose of the plane skyward.

Hüttig, a former Airbus pilot himself, and other pilots present for the test were unable to push the nose of the airplane down and thereby escape the stall.

. . . .

etc etc

We are not informed about Professor Hüttig’s background, his “Airbus” experience, his knowledge/experience of/on the A330, nor his capabilities in terms of analyzing the A330. We do not know if he understands the flight data process. I make this observation because everyone is looking for the so-called “smoking gun” but to me those who choose to go public with their theories and such input, claiming special status for their views and information, seem to have more on their agenda than merely “finding things out”. For these reasons I am taking the professor’s input with great caution and even greater reservations. A lot of people are claiming to have "flown the Airbus" but from what I read on an almost-daily basis, their notions and their work prove otherwise.

I think going public with information which claims to be serious and accurate, in advance of the full availability of the DFDR and CVR information, is not the right thing to do. There are others quietly working on the issue who have no need to “go public” to substantiate the work being done. The Spiegel articles have been sensationalist and in my opinion, poorly-written. Instead of a dispassionate examination of what we have before us, I am sensing a feeding-frenzy and a “positioning” or a “hardening” of theories which may very well be incomplete and wrong. In my view once again, none of us are in a position to be “pronouncing” on this accident.

WRT simulator work, I have done such simulator work myself. I've never been convinced that the simulator can and will tell us how the aircraft would perform at and beyond the boundaries of controlled flight. In subsequent discussions with the manufacturer of the simulator (Level D A330, which I flew) I learned more about this question. In my view, while these simulators are programmed to explore the boundaries of controlled flight, (low speed/high speed buffet, and stall characteristics, it is my understanding that the further away one explores from “normal”, the less there is of a correspondence between real aircraft behaviour/performance and simulator behaviour/performance, and, further, that extrapolations and pilot input, not flight data, form the basis of such computerized performance.

What I learned was in agreement with a paper written by Airbus's William Wainwright in a paper presented at the conference mentioned above, (link to the paper provided). Essentially, the simulator DOES NOT replicate aircraft behaviour at and in the stall or in other areas beyond the boundaries of controlled flight. It is as simple as that.

From the presentation:

Excerpt from Airbus Industrie presentation at l0thh Performance and Operations Conference

8. USE OF SIMULATORS

We manufacturers were very concerned over the types of manoeuvres being flown in

simulators and the conclusions that were being drawn from them. Simulators, like ai by

computer system, are only as good as the data that goes into them. That means ti-e

data package that is given to the simulator manufacturer. And we test pilors do not

deliberately lose control of our aircraft just to get data for the simulator. And even

when that happens, one isolated incident does not provide much information 3ecause

of the very complicated equations that govern dynamic manoeuvres involving now

linear aerodynamic and inertia effects.

The complete data package includes a part that is drawn from actual flight tests, a part

that uses wind tunnel data, and the rest which is pure extrapolation. If should be

obvious that conclusions about aircraft behaviour can only be drawn from the parts of

the flight envelope that are based on hard data. This in fact means being nol far from

the centre of the flight envelope; the pari that is used in normal service. It does not

cover the edges of the envelope. I should also add that most of the data actually

collected in flight is from quasi-static manoeuvres. Thus, dynamic manoeuvring is not

very well represented.

. . . .

Simulators can be used for upset training, but the training should be confined to the

normal flight onvelope; For example, training should stop at the stall warning. They are

"virtual" aircraft and they should not be used to develop techniques at the edges of the

flight envelope. lhis is work for test pilots and flight test engineers using their

knowledge gained from flight testing the "real" aircraft.

The Airbus publication FAST also reproduces Captain Wainwright's remarks. The link is, http://www.skybrary.aero/bookshelf/books/435.pdf

Stall response was updated at the 17th Performance and Operations Conference in Dubai. The link to the Airbus presentation is, http://xa.yimg.com/kq/groups/10483604/2002517561/name/Updated%20Stall%20Procedure.pdf

No news media I have read on this appears to be intent on "finding out", seems more interested in sales. Because "sources" who may be providing "background information" are almost certainly advancing one opinion over another (as the stakes are huge!), the media cannot be thought of as serious comment about what actually happened.

These views are a personal opinion, for what it’s worth, in the larger discussion on why AF 447 crashed. I pay absolutely no attention to any media for information or understanding on these particular matters.

Don

[boestar]

Don:Good Post. I would add the te reason that the Sim does not replicate conditions well beyond the envelope is because there is no real flight data from which to program the characteristics. There is of course theoretical behavior that is undemonstrated but that is theoretical and as you stated the margin for error increases the farther you get from normal.

Aircraft flight Testing for flight envelope exploration do go all the way to a stall ( I know this for smaller airframes but not A330 sized) and demonstration of the stall characteristics is a required demonstration for certification. Stalls performed for these tests are entered and immediately corrected. Canadair lost 2 airframes during stall limit testing during the challenger / Crj flight test program because they entered an en recoverable stall and the chutes subsequently failed resulting is fatal accidents both times.

Until we recieve the full analysis of the CVR FDR we will never know how the aircraft behaved once the boundaries were passed from normal flight to "WTF" Flight.

[RGT2.0]

By Robert Wall wall@aviationweek.com

LONDON

French air accident investigators looking into the crash of Air France flight AF447 have identified serious pilot training shortcomings that may have contributed to the loss of the Airbus A330-200 two years ago.

Among the findings are that “the copilots had received no high altitude training and manual aircraft handling” and also that there is no crew resource management “for a crew made up of two copilots in a situation with a relief Captain.”

In the case of AF447, the pilot was resting when the sequence of events began that led to the crash of the aircraft on June 1, 2009, killing all 238 persons onboard. The aircraft was flying from Rio de Janeiro to Paris.

The report also raises questions about how the pilots flying the aircraft at the time interacted. “No standard callouts regarding the differences in pitch attitude and vertical speed were made,” the report says. What is more, the report states that “neither of the pilots made any reference to the stall warning” and that “neither of the pilots formally identified the stall situation.”

A final report is expected next year. The third interim report, published July 29, highlights again that the aircraft stalled at high altitude and the pilots never performed the nose down inputs to recover. Throughout the descent, the crew maintained a nose up attitude of the airliner. In fact, the pilot flying made nose-up inputs and set thrust to takeoff/go around. The BEA notes that “in less than one minute after the disconnection of the autopilot, the airplane was outside its flight envelope following the manual inputs that were mainly nose-up.”

There was no indication of engine malfunction, with the report noting flight control surfaces matched inputs.

http://www.aviationweek.com/aw/generic/story.jsp?id=news/awx/2011/07/29/awx_07_29_2011_p0-353142.xml&headline=AF447%20Pilots%20Never%20Formally%20Identified%20Stall&channel=comm

[DEFCON]

Crew complement & training are certainly issues in this crash, but ‘magic’ aircraft being manually operated, regardless of altitude, should be capable of providing 'meaningful' information on the status of the aircraft to its crew.

Remember the Airbus claim; ‘we’ve built the first “pilot proof” aircraft’. Perhaps they did and that’s where the true problem resides?

[Trader]

I would say the report itself is suspect!!! The comment that he crew did not notify the pax is absolutely ridiculous!!!!!!!!! While they struggle to regain control there is an expectation to inform the pax?????????

[Moon The Loon]

Crew complement & training are certainly issues in this crash, but ‘magic’ aircraft being manually operated, regardless of altitude, should be capable of providing 'meaningful' information on the status of the aircraft to its crew.

I think perhaps it did give meaningful information. The two at the controls just didn't know what they were looking at, small picture (instrumentation) or big picture (weather and aircraft handling).

[J.O.]

I have a hard time getting past the situational awareness aspects. Going to the rest bunk just before entering one of the most active severe weather zones on the planet just doesn't sit well with me, especially since the Captain clearly knew they were approaching some big cells. I know it's not the only factor but it certainly set the accident chain in motion.

[Moon The Loon]

I have a hard time getting past the situational awareness aspects. Going to the rest bunk just before entering one of the most active severe weather zones on the planet just doesn't sit well with me, especially since the Captain clearly knew they were approaching some big cells. I know it's not the only factor but it certainly set the accident chain in motion.

[Kip Powick]

Air France will defend their pilots to the end.

In my opinion it was a total screw-up by all three members of the flight deck. Their inability to recognize the situation they were in, (which again, in my opinion), is basic aircraft aerodynamics, speaks volumes of the inadequate training many new "electric/glass" pilots are receiving.

[Trader]

Of course everyones assumption is that the pilots KNEW the situation they were in!!!! I havn't seen ANYTHING that explains exactly what the pilots saw in front of them. What were the screens really showing them????? Do you believe the screens or the various aural warnings?? Were they contradictory?

[Kip Powick]

Of course everyones assumption is that the pilots KNEW the situation they were in!!!! I havn't seen ANYTHING that explains exactly what the pilots saw in front of them. What were the screens really showing them????? Do you believe the screens or the various aural warnings?? Were they contradictory?

Well not exactly......................If everyone makes the asumption the pilots KNEW what they were in, it follows that they did not know how to recify the situation they were in........If they KNEW that they were in a stall, they should have KNOWN how to recover.

I don't think they KNEW that they were stalled, I don't think they KNEW how to recognize a stall...ergo.....no procedure to recify the situation.

[DEFCON]

Are the terms, ‘angle of attack’ & ‘pitch attitude’ being interchanged and applied improperly in these articles? If so, the ‘confusion and debate respecting the aircraft’s ‘normal’ or alternate’ law status is understandable?

[Trader]

Kip--I agree. My point is that I am not sure what they saw and heard would lead them to believe they were in a stall! I am guessing that the information they had was contradictory or they were unable to make sense of the barrage of information coming their way. The PF actually hands control over the other pilot around 10,000' because he can't get it under control. Maybe they didn't have proper training and didn't know what to do. Or perhaps, the information they had led them the wrong way!! The fact that he held full aft stick the entire time seems to indicate either lack of stall understanding or that they had other information (that was wrong).

[RGT2.0]

All this talk about loss of of control (LOC-I). I really wonder if these guys could have avoided all of this if they had of avoided the bad weather in the first place.

Did they know the weather was there and did they know how to use the airborne wx radar and avoid the storms?

...not to take away from the lack of training these guys had.

[Kip Powick]

Kip--I agree. My point is that I am not sure what they saw and heard would lead them to believe they were in a stall! I am guessing that the information they had was contradictory or they were unable to make sense of the barrage of information coming their way. The PF actually hands control over the other pilot around 10,000' because he can't get it under control. Maybe they didn't have proper training and didn't know what to do. Or perhaps, the information they had led them the wrong way!! The fact that he held full aft stick the entire time seems to indicate either lack of stall understanding or that they had other information (that was wrong).

I see what you are saying and understand your POV. The thing I don't understand is why any pilot would hold the stick back while looking at the altitude unwinding. The recovery from a deep stall can be done with minimum instruments, (glass or steam driven). So many pilots get fixated on the magic that they forget the basics of aviation and aerodynamics.

I remember one of my very first instrument approaches in the A310.....the damned aircraft was above the G/P and it wouldn't go down (I thought is should have and had forgotten that to "capture the G/P it had to do so from underneath the G/P....as a result I was on the localizer but not going down.......back to basics......disconnected all the magic, pullled the power off and pushed over to get the G/P and hand bombed the aircraft to the runway.......................my point is that when all else fails revert to very basic flying procedures.

Remember that 757, (I think it was ), that took off at night with the pitot static ports taped over.....? Again if they had totally disregarded all the misinformation they were getting and reverted to basic flying...maybe they would have all survived.

Any pilot that can not recognize a STALL, with very limited instrumentation, and know the appropriate recovery action severely lacks the basic understanding of aerodynamics.....and probably should not be flying.

[Trader]

Kip--I guess where I am going is that we don't know WHAT They saw. Did they see the alt winding down? DId they see it but not believe it!?? There was obvious confusion. It has been a while since I flew the 330 but I do remember practicing something similar in the sim. It was a scenario that is know with the Airbus where the computer compares the 3 ADR readings and tosses out the one that one that does not match the other two. The issue becomes when it tosses out the 1 CORRECT ADR because the 2 erroneous ones are similar!! So you have ONE bad ADR giving the wrong info--maybe someone current on teh airplane can comment or correct me.

In any case, I have my doubts about exactly what they saw in front of them. Holding back the stick, as they did, does not seem logical even for a poorly trained pilot.

[Kip Powick]

Kip--I guess where I am going is that we don't know WHAT They saw. Did they see the alt winding down? DId they see it but not believe it!?? There was obvious confusion. It has been a while since I flew the 330 but I do remember practicing something similar in the sim. It was a scenario that is know with the Airbus where the computer compares the 3 ADR readings and tosses out the one that one that does not match the other two. The issue becomes when it tosses out the 1 CORRECT ADR because the 2 erroneous ones are similar!! So you have ONE bad ADR giving the wrong info--maybe someone current on teh airplane can comment or correct me.

In any case, I have my doubts about exactly what they saw in front of them. Holding back the stick, as they did, does not seem logical even for a poorly trained pilot.

In my mind it really doesn't matter what was presented to them on the glass................................CONFUSED ?????????????????? go back to basics

Let's see......the altitude is winding down...........disregard all the glass magic.........look at the old standby AI....indicates nose up...... possibly above horizon....stick is full back..........hmmmmmmI think we have a pitch-up attitude and are in a stall.....ANSWER.................use standard stall recovery..................reduce power....push nose over..............watch IAS build and recover in wings level and add a suitable power setting to keep the plane aloft.

PS............A similar accident over the jungles of SA..........B737..........no one checked the Standby AI which gave all the correct info...the pilots were glued onto the magic and the aircraft went out of control and disintegrated at 10,000 ft on the way down. Accident report ascertained that the electronics were in error but the standby (steam driven/basic AI) gave all the correct info.

Anyhow, I appreciate your points but in my opinion, based on limited facts I might add, the pilots did NOT recognize their problem and the failure to recognize a simple problem resulted in castrophe....

There has not been an aircraft that I have been qualified on where I was not shown how to recognize a stall in that particular aircraft and I think the AF crew did not get that training, or if they did, it was insufficent and, perhaps, glossed over.

[Moon The Loon]

Air France will defend their pilots to the end.

In my opinion it was a total screw-up by all three members of the flight deck. Their inability to recognize the situation they were in, (which again, in my opinion), is basic aircraft aerodynamics, speaks volumes of the inadequate training many new "electric/glass" pilots are receiving.

:tu:

[Moon The Loon]

Of course everyones assumption is that the pilots KNEW the situation they were in!!!! I havn't seen ANYTHING that explains exactly what the pilots saw in front of them. What were the screens really showing them????? Do you believe the screens or the various aural warnings?? Were they contradictory?

The gist of the report seems to highlight the pilots had no idea of the situation they were in. So what they saw/heard/imagined is immaterial. They were incapable of interpreting not only the fact that the Captain had left the flight deck at a critical time of the flight, but that they had no idea how to both interpret or avoid what was displayed on the weather radar. That, of course, assuming they knew how to work the weather radar.

I could go on, but I already have.

Edited to remove in incorrect remark

[Moon The Loon]

All this talk about loss of of control (LOC-I). I really wonder if these guys could have avoided all of this if they had of avoided the bad weather in the first place.

Did they know the weather was there and did they know how to use the airborne wx radar and avoid the storms?

...not to take away from the lack of training these guys had.

x 10

[Trader]

Moon - am a pilot on the 777 and flew the 330/340 but that was over 3 years ago (think I met you on the crew computers in Heathrow at the Park Inn one time). As I mentioned, I know the Airbus computers deselect the ADRs that it THINKS is incorrect. If that was the case it may be possible that the pilots had the wrong information displayed on their screens (maybe a current 330 pilot can correct me or add to this). So, if I remember correctly, it is possible that the information on the screens (including the stby Kip) was wrong.

So yes, att + pwr = Perf but that assumes yo have reliable information. On a clear night in smooth air you could turn everything off and maybe keep it straight it and level with neutral stick and a known power setting. Toss in a T-storm, and airplane that was likely out straight and level when the autopilot disconnected, confusing and likely contradictory screens etc etc and you have a serious situation. My only point is that is seems 'easy' to simply say the pilots were incompetant!

[Moon The Loon]

Trader: Sent you a PM. Point: Why were they in a thunderstorm in the first place? The worst weather on the planet is in the ITCZ.

[Trader]

I agree , they shouldn't have been in the T-storm. Though from the radar pic I seen the wx was widespread and I would guess they were picking their way through. That is, however, only one factor.

[Trader]

I agree they shouldn't have been in that situation in the first place. But I do remember flying teh 330 at Gulf Air and having that scenario in the sim. The problem was it was a demonstration of what happens when the computer can't recognize which ADR is inaccurate and throws it back to the pilots by disengaging the autopilot and limiting screen info, so we didn't actually fly it. I recall that the airspeed went red, that we had simultaneous stall warnings and overspeeds etc etc. The point of the demonstration was to show how the instruments were unreliable and it was up to the PILOTS to determine what to believe and to determine which ADR's were faulty. Worse the ECAM did not show a triple ADR failure, but a dual ADR failure in this case. In this situation, the pilots have to know to go to the paper checklist (may be different now) to deal with it. I remember thinking that this was 'another' Airbus gotcha and I was happy to have seen it in the sim and , oddly enough, is one of the few things I remember from the aircraft.

I pulled out the 330 QRH and the one thing that it does note is that the STBY INST should be used. Looking at the 2 relevant checklists (ADR 1+2+3 FAULT and UNRELIABLE SPEED) it appears that it should have been an event that could have been handled with pitch and pwr settings via the QRH table (or better yet by memory for level flight). I am just not sure about what they would have seen in front of them and if the STBY would still be accurate!!!

In any case, this is one of the more complicated failures on the 330 which requires the use of 2 different checklists, immediate actions and VERY good knowledge of the system to determine that when the ECAM shows a DUAL ADR failure that it may in fact be a triple failure. I would say that if they had never seen this in the sim (been trained) then the result is not a suprise!!! If this was indeed the case then, in my opinion, it is pilot error but due to lack of training.

Anyone current on the 330 care to share their thoughts or correct an error I made?