AF447

[surplus1]

Originally Posted by PJ2

If the standby instruments fail, (which there is question of in the 447 ISIS case), there is no other system, archaic or no, and the mechanical standbys in the 320, 330 and 340 were tiny, poorly lit and would be impossible to read in heavy turbulence. These instruments are certainly not going to help you inside a thunderstorm.

What I am pleading is the case for examining in detail the recorded issues to see the pattern or common thread if there is one, from which, I think you'd agree, the problem can be defined and if chronic, fixed, and if intermittent and statistically granular, addressed strategically. In part this has already been accomplished with the Unreliable Airspeed memory items and QRH checklist.

I think you need to understand that no bread-and-butter "archane" system presenting basic attitudes/speeds etc is going to right an aircraft or permit the crew to manually do same, that is badly upset.

What such a system must do, (and in my view it requires good flight conditions to do it for the reasons stated, "good" being in cloud, at night, no moon, no visible horizon, possible icing and moderate turbulence) is provide a horizon and a direct speed indication until the crew can stabilize the aircraft.

I can assure you that no instrumentation no matter how robust and no autoflight system will permit safe flight inside, or even offer a good chance of surviving penetration of, a large, developing thunderstorm at high altitude by a transport category aircraft. No pilot, no designer, no certification body and no regulator has a right to expect that any airliner should perform otherwise. If one enters a thunderstorm, one is, for all intents and purposes, in test-pilot territory with an unknown outcome.

Well said sir, and thank you for all your other posts as well. They have been informative, measured, well articulated and reflect your professionallism and experience. Kudos captain!

I'm an old retired dog and have never flown a FBW airliner, but I have flown many others, for many years, all over the world. Before this accident I knew little about Airbus products or airliner FBW systems. Now I've read ever single post in several forums and done a lot of independent research reading manuals and official accident/incident reports. I don't know enough to fly one of those things, but I do know enough to observe that I do not regret that I never did.

What you say above about thunderstorms is accurate and well said. What you say about these systems is far too conservative and generous, for me. I appreciate why you take that stance and respect your consevatism and restraint. I am less conservative.

When we have multiple incidents of 3, 5 or all 6 screens going blank due to the failure of a single Bus, not once but several times. When we have a series of upsets or partial upsets from which pilots were lucky enough to recover, sometimes with injury to passengers; when we have the mysterious disappearance of an airliner with a seasoned crew (AF447) and a plethora of automated messages indicating multiple failures and warnings; when we have standy attitude instruments that can share in these malfunctions and that have independent power for no more than 5 minutes, and are difficult if not impossible to read in heavy turbulence; when we have, repeatedly, apparent multiple sensor failures producing erroneous data that disables funcionality at least in part; when we have at least a dozen senior captains in type all debating about how these systems truly function - and not agreeing after a month - I would call these experiences/events more than chronic. There's something very wrong about all this and it is dangerous.

Lengthy and cryptic checklists buried in the QRH, and multiple warnings with ECAMS scrolling so fast as to be illegible, and multiple memory items for problems that pilots do not appear (from what I read here, elsewhere and in the official reports) to understad in depth are not a solution for an obvious problem, nor are they a "reason" for its repeated occurrence - they are an excuse. The excuses are too many and too inadequate. They must be replaced by solutions that avoid repetition, not more excuses or modified check lists.

All airplanes have their problems it is true, but these particular designs seem, to me, to have more than their share. The manufacturer has not so far - in my opinion - resolved the problem whatever it is. They need to do so and fast! I don't know enough to suggest a remedy, but I do know that one must be found.

We do not know what happened to AF447 and unless the FDR and CVR are recovered and readable, most probably we will never know. The crew may well have made an error by inadvertently flying into the mother of all cells. I don't know and I am not prepared to guess or to assume without evidence. Even if that is so, there is no doubt in my mind that the systems of this design were less than helpful in what ever situation/scenario presented itself to that crew. That is unacceptable to me.

Some airliners of the type that I have flown have come apart in extreme turbulence. Yes, their pilots made errors but their systems did not induce those errors and did nor create or contribute to them. Most have survived ecounters with severe turbulence and their pilots were never confused with spurious warnings and unreliable instruments and systems producing incomprehensible warnings or uncommanded nose-overs.

Pilot error is something we all have to acknowledge. But, design induced pilot error is just plain unacceptable. This manufacturer needs to take a very hard look at its products. I'm not at all against technology, but experimentation with 'what ifs' should not be a part of airliner designs or operationl- no matter who produces them.

My apologies to all of you; I've already said too much.

[ELAC]

Quote: Will Fraser

... this accident suggests the reason all prior incidents had acceptable outcomes was merely that the (447) flying pilots had no horizon, no a/s, etc. and were in weather at night.

Will,

How have you come to this suggested conclusion?

From the pattern of Failures and Warnings in the ACARS messages it is reasonable to conclude a loss of airspeed information, but nothing more.

Extracting from the last posted summary of ACARS messages (assuming it is correct) the failures reported only lead to 3 items of information: Airspeed; Flight Director, and; Flight Path Vector being removed from the PFDs. Likewise the failure message related to the ISIS is for the speed/mach function which itself should only produce a red SPD flag on the ISIS, not a failure of the entire unit (which would display an OUT OF ORDER flag and related maintenance code). There is no information available to us that would lead to the conclusion that attitude or altitude data was lost, either on the PFDs or on the ISIS.

The problem with this discussion is that, in the absence of any further facts, it has drifted so far off base that we've come to the point of discussing the merits of plumb-bobs of all things. The contributions now seem to fall into the categories of: Declarations regarding the certainty of some critical flaw in design or concept; Various home-brewed plumb-bob solutions to said flaws; Validation of same by anecdotes or experiments in a Cessna, and; Occasional efforts to bring things back to the known facts about the aircraft and the accident.

It would all be quite humorous were it not for the fact that it is actually setting back efforts to understand what did happen to AF447, which is a very important question to those of us who fly the aircraft as pilots or as passengers. All one has to do is read the articles appearing in various media sources (such as those posted here from The Times and The Australian) to see the reverberation effect of inaccurate conclusions stated here becoming magnified and reflect back to us as "consensus opinions of professionals in the field". That's a fault of the media as much as anything, but it does shape perception of what happened and it should be a reminder that this is a "Professional" forum where there's an expectation that what you suggest in your posts is at least a speculation informed by knowledge of the aircraft/crew/accident and supported by the facts available.

The suggestion that the AF crew lost all PFD and/or ISIS data and that this was a or the reason for the accident is not one that is supportable given the known information to date.

0210 22-83-00-2 WRN FLAG
LEFT PFD SPD LIMIT
Rejected ADR still feeding data to PFD
If there is valid ADR, it's not being selected for LEFT seat.

0210 22-83-01-2 WRN FLAG
RIGHT PFD SPD LIMIT
Rejected ADR still feeding data to PFD
If there is valid ADR, it's not being selected for RIGHT seat.

0210 22-83-00-1 WRN FLAG
LEFT PFD NO F/D
Automatic Flight System (AFS/FMGC) loss of 2 ADR sources.
Safety mechanism, prevents erroneous F/D for pilot to follow

0210 22-83-01-1 WRN FLAG
RIGHT PFD NO F/D
Automatic Flight System (AFS/FMGC) loss of 2 ADR sources.
Safety mechanism, prevents erroneous F/D for pilot to follow

0211 34-12-00-0 FLR ISIS
ISIS (22FN-10FC) SPEED OR MACH FUNCTION
SUSPECT LOSS OF ADIRU1 AND/OR ADIRU3 FOR ISIS MACH
Suspect Loss of ADIRU3
NAV DISAGREE CONCLUSION DELAYED - ADDITIONAL FAILURES - RECOMMENCE FAULT ISOL

0211 34-12-00-1 WRN FLAG
LEFT PFD NO FPV

0211 34-12-01-1 WRN FLAG
RIGHT PFD NO FPV

[Will Fraser]

ELAC

Glad you asked. I mean no disrespect, but you are overlooking the obvious. In the absence of suitable technical data, you call me to task because you don't believe in something that I do. (Ostensibly).

With its highly trained and experienced Flight crew on a flag carrier built to exacting specifications, why do you not believe that had the crew had a Horizon, the a/c would not have ended up in the Ocean?

With a Horizon, and power, AB's suggested solution, pitch and power, would be expected to have worked. I believe it too. Call me old fashioned.

rgds.

Will

[jcarlosgon]

surplus1 ...I've already said too much.

I don’t think you have said too much. Just check this:

About IFALPA

IFALPA Briefing Leaflet - 30.06.09

Headline, “IFALPA requirements regarding pilot authority and flight control architecture”. Goes on to:

"The aircraft commander shall be given the authority and capability to select the level of augmentation for the flight control system."

And, wrap up:

"Automatic reconfiguration of pilot authority must be clearly indicated by the system.
Based on established procedures, the level of pilot authority must be selectable by pilot action.
The level of pilot authority must be clearly indicated by the system without any ambiguity.
If it is possible to regain a higher level of augmentation, the procedures shall indicate the status of the system."

[Loose rivets]

Will. I ended up chatting on JB about this point. Paras 4 and 5 are germane


http://www.pprune.org/jet-blast/3759...ml#post4993312

[lomapaseo]

ELAC

The problem with this discussion is that, in the absence of any further facts, it has drifted so far off base that we've come to the point of discussing the merits of plumb-bobs of all things. The contributions now seem to fall into the categories of: Declarations regarding the certainty of some critical flaw in design or concept; Various home-brewed plumb-bob solutions to said flaws; Validation of same by anecdotes or experiments in a Cessna, and; Occasional efforts to bring things back to the known facts about the aircraft and the accident.

It would all be quite humorous were it not for the fact that it is actually setting back efforts to understand what did happen to AF447, which is a very important question to those of us who fly the aircraft as pilots or as passengers. All one has to do is read the articles appearing in various media sources (such as those posted here from The Times and The Australian) to see the reverberation effect of inaccurate conclusions stated here becoming magnified and reflect back to us as "consensus opinions of professionals in the field". That's a fault of the media as much as anything, but it does shape perception of what happened and it should be a reminder that this is a "Professional" forum where there's an expectation that what you suggest in your posts is at least a speculation informed by knowledge of the aircraft/crew/accident and supported by the facts available.

Bingo

I don't mind opinions but they belong in a technical thread not in an accident thread. Even the well advised among us sometimes get too far ahead of the investigation to the disservice of the public. I have no problem in making other pilots aware of possible issues, but there has to be a more delicate way.

[takata]

Will Fraser:
With its highly trained and experienced Flight crew on a flag carrier built to exacting specifications, why do you not believe that had the crew had a Horizon, the a/c would not have ended up in the Ocean?
With a Horizon, and power, AB's suggested solution, pitch and power, would be expected to have worked. I believe it too. Call me old fashioned.

You are right, and this is the conclusion based only on the few facts available (ACARS).
Hence, what happened after 02.14?
We just don't know neither when and where she crashed. F-GZCP ended in the Ocean but those ACARS are not saying why she did so.

Short factual summary:
June 1st, at 0210, she was in the middle of a big CB cluster and her Thales ice-sensible pitot-probes started to freeze. Between 0210 and 0214, the various flight systems related to air data stream were automaticaly disconnected (as expected in the A330 documentation). At 0214 (+/-30s), no more ACARS were recieved from F-GZCP. On June 6th, the first bodies and wreckage were spotted and recovered by the Brazilian SAR confirming she had crashed.
End.

Everything else is speculation based on the few clues we have seen from the wreckage, rumors about the investigation work, supposed Airbus design flaws or whatever agenda some people have. All we need now to go further in this case is more facts and less "noise".
S~
Olivier

[captainflame]

I have to agree with you.

[PJ2]

Surplus1;

My personal thanks for a fine and collegial post, which speaks, in my view, a very fine truth about our industry and about this aircraft design. As I posted to Will Fraser, I am not a convinced Airbus pilot and do not dispute the serious misgivings which some have intelligently expressed here from a basis of combined experience and knowledge.

I also am not uncomfortable in the airplane. it is a dream to fly automatically and manually - it is an extremely well-thought-out airplane and entirely trustworthy. It is in all ways a "pretty" engineering feat.

That said, like many crews I have had a few serious system abnormalities on the 320/330, (though none as serious as was faced by the AF crew) and have disregarded, where appropriate, both the autoflight system and the ECAM checklist when I thought it was seriously out to lunch, (and it was, as I discovered after talking with maintenance.

On the 320, hand-flew (autothrust OFF) every approach and, until my company prohibited it unless there were no other airplanes in the sky, hand-flew every approach in the 330/340, sometimes to the consternation of the other crew members who had never seen the airplane flown like an airplane but "managed" as though one were at one's desk in some darkened room and what we were looking at outside was a realistic video.

One either submits to the designer's and engineer's intentions or as a professional aviator one draws a firm line over which the engineer is not permitted to cross. Hand-flying isn't "practise" - it is the finest way to maintain situational awareness. It is a human-factors defence activity which necessarily requires thinking and the attention of everyone.

Towards the end of the career I left the automation connected - too much risk to the career to disconnect and have something go wrong. If you ask me, there is something deeply wrong with that attitude but there is something equally wrong with pilots and their associations accepting it without a fight. But unions these days are terribly weakened - such amorphous concerns are largely dismissed and difficult to argue in what has become a singularly-focussed industrial discourse.

Will, I hear you.

Although I came very close once, I haven't lost a good friend to an accident during the career but I've been scared a number of times by weather and what it can do. And in that vein, no matter how many hours they get and how well they memorize, I don't think any pilot graduating from these new MCPL schools is a real pilot until they experience the adrenaline levels that nearly killing oneself and one's passengers brings on, with those aftershocks in the days that follow that startle one into fretfull wakefuless at 3am. I'm not saying this for dramatic effect. It can't be taught but those who have had the experience will already know that.

To me, it isn't young pilots who are the problem; we all must get experience some way. But pilots with less than a couple of thousand hours shouldn't be sitting in the right seat of any airliner faced with all the challenges an airline pilot must contend with. Flying the airplane well and knowing how to program the FMC and understanding IFR and instrument flight isn't what an airline pilot primarily does; - certainly not the captain anyway. That's expected.

The difficulty with the way the industry has change comes more subtlely as years of economic and political pressure and wonderfully safe flying brought about in part by the very automation we are concerning ourselves over, have permitted executive and senior managements to let their standards slide to survive in this extremely difficult enterprise. "Cheaper, faster, better", used to be NASA's credo during the early Shuttle days (and perhaps earlier); now it is aviation's. Seen it, know it and felt it and so have many here, but it is impossible to convey one's concern to those who have lost touch with the business they're in, in the face of "obvious" success, until aviation itself provides the ultimate intervention and dose of harsh reality. Flight safety specialists are usually unpopular millstones who always bring bad tidings, interfere with "progress" and don't have a sense of humour. Kicking tin does that and when serious events occur in flight data but the data is either ignored as commercially inconvenient or used against crews (which kills a just safety culture and stops all safety reporting), one tries to defend oneself against the cynicism that inevitably follows, especially when familiar with the undercurrents of hypocrisy and absence of ethics.

I view the discussion on automation, pitot - static issues, a severely degrading autoflight system with cascading faults in severe weather, in this context. At what point are pilot "over-trained" such that cutting training is a legitimate economic goal; at what point are aircraft "over-engineered" such that pilots may be seen and treated as "managers" instead of aviators?

The notion of "watershed moments" is over-worn, exhausted in the hyperbole of the safety discussion, mine included sometimes. I sense there are extremely valuable lessons for our industry in the present issues under discussion but they are as much philosophical as they are technical or economical; those dialogues have always been viewed as strongly at odds with one another but unless we understand our discomforts with design priorities while still making enough money to stay in business while keeping ones' charges safe, we will see the fatal accident rate begin to slowly climb as the business recovers and grows.

Thanks for the dialogue Will. Strong views held, yet the willingness to engage.

PJ2

[eklawyer]

Came back from the crew bunk in the bowels of the aircraft several months ago. Stood behind my seat to look at all the flashes in front of the aircraft, looked at both ND's with no returns showing.

Leaned over the FO occupying the left seat and turned the radar dim to max, low and behold if it was not painting red it was magenta. Threw the pillock occupying my seat out and initiated the weather avoidance. Followed by a few home truths once clear.

I was horrified what might of happened if I had come back 15 mins later than I did.

Flying into horrendous wx is not beyond the imagination especially at night !

[PJ2]

eklawyer;

Threw the pillock occupying my seat out and initiated the weather avoidance. Followed by a few home truths once clear.

I was horrified what might of happened if I had come back 15 mins later than I did.

Flying into horrendous wx is not beyond the imagination especially at night !

That's pretty worrisome stuff, eh? Don't know about yours, but my airline never once, in 35 years, told us/taught us how to use radar. As a result, I have watched the same kind of lack of understanding about how to set calibration, how to use tilt, how to calculate height, what the beam width is, what does and what doesn't return signal, signal attenutation and shadowing, the use of range and so on.

A friend on the 767 informs me that his F/O actually argued with him (flying in the same general area as AF recently) that the "best" radar for thunderstorm penetration was "MAP". Needless to say he got the same lesson you probably gave. There are as many theories on how to use the gain as there are pilots.

The use of 330/340 radar was only recently placed in the AOM but it never, ever comes up in sim rides, enroute check rides, annual recurrent training or even informal discussions. Not good enough.

[Tree]

A friend on the 767 informs me that his F/O actually argued with him (flying in the same general area as AF recently) that the "best" radar for thunderstorm penetration was "MAP".

Don't knock it until you try it! Based upon my experience and experimentation with the 767 radar, that is the best for painting "dry" ITC buildups. MAP and one notch down from MAX GAIN. This procedure was developed by a group with many decades of ITC experience and in the past also used expensive NV goggles as another layer of detection. NV goggles are now cheap enough in HK that one can afford to do their own research. Our Tech-Safety group also provided a copy of Archie Trammell's VHS tape.

[p51guy]

EK lawyer said what I said a couple of days ago about the two FO's flying while the captain is taking his rest and how their experience would handle this massive instrument failure. I was deleted soon, so most here didn't see it. Even though people disagreed with me I notice here in the US copilots are not held to captain standards. I really don't think it is much different in other countries. We have an unknown command capability on this flight at the time of the encounter.

[Lightning6]

Originally Posted by p51guy

EK lawyer said what I said a couple of days ago about the two FO's flying while the captain is taking his rest and how their experience would handle this massive instrument failure. I was deleted soon, so most here didn't see it. Even though people disagreed with me I notice here in the US copilots are not held to captain standards. I really don't think it is much different in other countries. We have an unknown command capability on this flight at the time of the encounter.

I would have thought the experience that the FO's on this flight would have the experience to deal with whatever situation they were confronted with, but in the possibility of a catastrophic and sudden system or airframe failure, would the captain have done any better?

[amicus]

Suplus1, Will and PT2

All thanks for superb, thoughtful, balanced and insightful efforts throughout the course of this extended discussion. Your comments and insights are deeply appreciated by this aerospace engineer who worries about your and your passenger's safety daily and who greatly values and welcomes your highly experienced and most valuable commentaries and discussions at all times. I rely on wise mentors in the aviation community to guide we engineers and hopefully helps us avoid the traps of hubris, over reliance upon automation, overly complex systems, which lead to tempting and allowing airlines to cut corners on costs and training. Such design and system approaches dangerously and imprudently lead some airlines and training schools to advocate that pilots become system monitors rather than skilled airman. There are many questions and ongoing concerns regarding training, airmanship and safety, backup instrumentation and systems design and we engineers ignore your advice, guidance and wise counsel at our peril. Thank you all.

[Flight Detent]

Whilst I'm reading here about the piloting experience levels and radar operation expertise, and the various levels of indication anomalies...can I remind everyone that the vertical stabilizer was found some 30nm from the site of the main crash debris!

Take that as you will, but for my money that component didn't ride the fuselage all the way to the water.

Maybe that may have had some affect on the piloting capabilities!

Aren't we getting a little sidetracked here!

Cheers...FD...

[Xeque]

BOAC

I think you are right in what you say on the other thread. Unfortunately, two earlier posts I made in this thread where I expressed similar reservations were deleted.

[Will Fraser]

Flight Detent

Everything you mention in the opening sentence is connected. There are three VS loss accidents that come to mind. AA587, "pilot induced" over loading the rudder; UAL 232, uncontained turbine disintegration (N1?) which took the fin and rudder with it, and JAL 123, aft bulkhead crippled the Rudder/VS when it let go after a faulty repair of a tailstrike seven years previously. No accident is exactly like another. Nothing is unimportant, and no salient assembly tells all.

This thread is novel in that it allows comment by interested parties regarding information that is available. The Internet is a dynamic force that changes the game (or at least some of the 'rules'). I think it is a force for progress, utimately.

Whatever happened, happened quickly. Did the VS let go but allow for some control? 123 and 232 allowed the pilots to manouver with difficulty, did 447 have a shot after the VS separated? Did they descend in 'reasonable control' though steeply impact the sea some distance from the loss? Or did the empennage fail as a unit along with the aft bulkhead, the HS somewhat attached the rest of the way? Keep in mind the ACARS is a pirated document, not released by AF (to my knowledge). Who could discount the Weather? Other than some rather fantastic and discounted possibilities, anything could have happened.

One other thing. The three I mention lost their VS' in the climb, 447 was in cruise. That is important.

Will

[Captain Bob]

MAP Mode?

Quote:
A friend on the 767 informs me that his F/O actually argued with him (flying in the same general area as AF recently) that the "best" radar for thunderstorm penetration was "MAP".

Don't knock it till you try it! Based upon my experience and experimentation with the 767 radar, that is the best for painting "dry" ITC buildups. MAP and one notch down from MAX GAIN. This procedure was developed by a group with many decades of ITC experience and in the past also used expensive NV goggles as another layer of detection. NV goggles are now cheap enough in HK that one can afford to do their own research.

You have to be kidding me? Flying around WX with the MAP Mode on? I hope I am never on any flights you are on. I would never let any of my FO's ever try to do that. 11,000 + hrs on the 757/767 and I have to raise the BS Flag on this one.

[Graybeard]

Thanks, EKlawyer, for sharing your scare. It is relevant.

Quote:
"A friend on the 767 informs me that his F/O actually argued with him (flying in the same general area as AF recently) that the "best" radar for thunderstorm penetration was "MAP". "

Tree confessed: "Don't knock it till you try it! Based upon my experience and experimentation with the 767 radar, that is the best for painting "dry" ITC buildups. MAP and one notch down from MAX GAIN. This procedure was developed by a group with many decades of ITC experience and in the past also used expensive NV goggles as another layer of detection. NV goggles are now cheap enough in HK that one can afford to do their own research."

Dunno whether you guys are flying Collins or Bendix/AlliedSignal/Honeywell, but I can tell you this about Collins WXR:

Rainfall scintillates, which is detected by doppler shift. Ground return doesn't of course, so the Collins radar distinguishes the two, and optimizes the mode selected. There will be a "GCS, Ground Clutter Suppression", IDNT, (ground clutter) Ident, or some other such submode of WX on the control panel. This is especially effective at low altitudes where there is high terrain. This submode will wipe out the gentle rainfall along with the terrain returns, but you don't care about that, anyhow.

If you want to see the high altitude ice crystals after you have picked your route by scanning the wet part of the storm, it's a good idea to have variable gain in WX mode, or use the Turb mode. The Collins WXR operates at full gain in Turb mode, so it will display returns below the level of light rainfall, just like Max gain in WX mode.

Do AF pilots get good WX training?