You're a wise man.

When the airbus reverts to direct law, it is an airplane again, just like the DC-9 (I'm old Northwest, too).

The problem Air France had was that bad data from frozen pitot/static sensors drove the autothrust and normal law protections the wrong way.

I'd like to think I would sense the deck angle, check the ground speed on the data/gps page, determine my airspeed input is wrong, disable the automation, and recover from the stall in the middle of the night over the middle of the ocean in thunderstorms. I'm ready to face that challenge now, with help from the Air France experience.

The airbus method does a lot for you, but you always need to be ready to turn off the magic and fly the airplane. When you turn off the automation, she'll do anything a DC-9 would do.

BandAide, yep but it needs to be in DIRECT law.
Not alt 1 or 2 or "unusual attitude law."
TT has it right::D

TT, absolutely correct & controlled contribution when it might be tempting to be a bit more colourful! I flew 73,75,& 76 before going A340. Hated the latter. You do not fly the plane. All autopilots out, no, you are NOT connected to the controls. You are still connected to computors. Tough, high speed conditions gave us those lovely things called PCU's. Lose all hydraulics in the 73 & you really were connected to the contol surfaces. It was called Manual Reversion. Very tough to fly , especially with an engine out. 75 made me scared when I read the clear statement in Mr Boeing's manual..."There is no manual reversion". Three hydraulic systems and a RAT made it unnecessary. 76 was the same. Forced on to the A340 made me rethink my career. Horrible. It flew me. Even when esteemed Instructor told me once, "Knock it all out & fly it like any other plane" ! Oh yeah, side stick was kinda fun & the plane did go where I pointed it. But, no thrust levers ! I mean they are there but they do not move. Going way off thread here though. I am a bit late into this discussion but I did notice pure Airbus pilots think through the automatics. Not their fault, they are trained that way & recurrently tested that way. Boeing pilots think like stick & rudder men. The AF pilots were faced with very confusing indications but it disappoints me to note (like a previous contribution) that 'Flight with unreliable airspeed' manouvres were not followed or were not properly understood.Oh, glad to be back on the Boeing.

I hope you wait forever. CVR and DFDR data were NOT designed for public release!

I've never flown anything more complex than a DHC Chupmunk, and that was 15 years ago, not to mention heavily supervised! All opinions here are usually welcome, that's one of the reasons I like it.

Why do you think that? If you're referring to the reliance on autopilot from climb-out to finals then you've got the wrong target. The A300 series did pioneer the use of digital computers in flight management systems, but it was the B757 and B767 that made the technology mainstream. These systems are completely distinct and separate from the FBW aspect.

A question I'd like to ask you, and I throw it out in general too - and I'd like an honest answer. If the FBW technology used in the Airbus was developed in the USA, would you feel so strongly about disliking it?

Not at all, as long as you're willing to learn more about how the thing works (and possibly lose some of the preconceptions in the process), and the same is true about any line pilot on any airliner. The fact is that given the amount of off-base commentary on the Airbus FBW technology and how it relates to other airliners (slowjet's post is a good case in point), it seems that some B757 and B767 pilots (not to mention those on the B777) are completely unaware as to just how much the computers were tied into their aircraft from the outset.

I'm well aware, but as long as the commentary demonstrates a misunderstanding of the concepts and technology involved, as well as a knee-jerk dislike of computers in aviation (including the false idea that Airbus was alone in pushing the technology) then someone has to make it clear, and it might as well be me.

No they are not. The amount of control the sidestick gives you is *way* in excess of what the autopilot turn control in a conventional airliner gives you.

You're missing the point I made that the yokes of more conventional airliners do not directly control the control surfaces either, and that has been the case since the late '60s.

This is true, but you cannot make an aircraft the size of a DC-10, B767 or indeed an A330 without sacrificing that directness of control, simply because the surfaces are too large and powerful to be moved by muscle force alone. Slowjet is correct when he says that even the little B737 was an absolute nightmare to control under manual reversion - which was why you had hydraulic assist. When the widebodies came along, they were hydraulic only, because muscle power could not move those surfaces *at all*.

Of course there's a difference between commanding roll rate directly and controlling ailerons directly, but of those Airbus pilots without an axe to grind who have controlled the thing in Normal, Alternate and Direct, the vast majority state that the difference is not that pronounced, and very easy to get used to - I think someone a long time ago compared it to losing power steering in a car, the only difference being that it was slightly more sensitive as opposed to more heavy.

Would the addition "more-or-less" help there? The fact is that it is controlled by exactly the same flight surfaces as any other airliner, you have speedbrakes to slow down and disrupt lift, flaps and slats as high-lift devices and tricycle landing gear to get 'er on and off the ground. That the connection to the ailerons and elevators uses a system that's marginally smarter than the old Q-feel systems is a difference, but in the long run it's not really such a big one.

If you can point me to anywhere I've blamed the pilots I'd be happy to concede, but I'm pretty sure I've done no such thing. Given that of the 30-odd UAS incidents that occurred - not to mention the thousands of safe flights that transit the ITCZ every year (many of which are on FBW Airbuses), only one ended up in an accident, don't you think you're overgeneralising a bit there? If you're saying that airlines have skimped on training because of the presence of these new technologies - which, remember, are not solely Airbus's domain - then that is a question that the airline training departments must answer, and for what it's worth I think it is AF's training department that has the most questions to answer in the case of this accident.

As an aside - the DC-9 is one thing, but Douglas made the mistake of trying to keep things too simple when they made their first widebody and disregarded some basic mechanical and physics principles - with the result that the DC-10 became notorious and the company eventually ceased to exist in its own right. Plenty of pilots loved the DC-10 when she was working, but when she developed a fault she could, and often would bite you badly.

On the B757 you were connected to an electronic feel unit hooked up to the hydraulics. It may have felt like you were directly connected, but you weren't.

If you want to see what happens when the pitot tubes or static ports are blocked on a 757, I suggest you look up the Birgenair and Aeroperu 757 crashes respectively. The computers were *heavily* tied in to the warning systems on those aircraft.

Massive, *massive* overgeneralisation. You have only to look at the posts made by Airbus pilots on the Tech Log threads (though admittedly that would take a while) to see that you're incorrect there. Posters to look out for include PJ2, Lemurian, HazelNuts39 among others.

That kind of comment / comparison only underlines the lack of understanding of the difference between commanding roll rate and commanding deflection. On a car the failure of power steering doesn't affect at all wheel deflection per number of steering wheel turns.

If you wanted an actual comparison you'd have a car where your steering wheel would not be connected at all to the steering rack, and the "steering assist" would modulate your steering inputs according to speed. If it were to fail, probably 99% of the time it wouldn't make a difference and you wouldn't notice - except if you were to say drive on the autobahn at 250km/h and fail to adjust your steering wheel deflection to your speed and gallantly drive off the road into the side barrier thinking "what the hell just happened".

It matters quite a lot. Not all computers are same or serve same purpose. It is very different ballgame when you are connected direct to flight computers as opposed to via FMGC. It is basic Airbus systems architecture. Every bus rated pilot must know that.

Interim 2 lists 37 UAS events in 330/340 that ended completely uneventfully (pages 50-53 & 99-102). Nowhere is the experience of the crews involved listed so your statement is not based on BEA reports. Would you be so kind to provide your source(s)?

@Shorrick Mk2:

I believe it was someone describing the way it *feels*, not the way it *works*. It's easy to mistake one for the other, especially given the misunderstandings as to what constitutes "direct control" of the flying surfaces. I've even read articles where people assume that the controls on the B777 are directly connected when they most certainly are not.

Is that based on actual experience of flying a bus in roll direct?

Dozy,

I could care less where the aircraft is made and it's not the technology....it's the core concept. Conceptually the Airbus does everything possible to remove aviator from aviation....right down to the name....BUS. In effect it trivializes the pilots and equates them with being a bus driver. Accordingly many are now paid like one.

TTex,

This is the core of my question/concern. My understanding is that the stall itself is a direct result of the PF's initial actions. The plane was literally flown into a stall. The elevator trim was a direct result of the pilots continued back pressure on the SS. So this creates two issues for me...

1) Completely incorrect initial response to a known issue resulting in loss of control.

2) Continued application of SS appears to be a lack of system knowledge on two fronts. Once a nose up condition is achieved (regardless of the correctness) continued back pressure is incorrect. Beyond that the maximum deflection wound the trim up all the way. Should not the PF be aware of the effect of this pressure??

To me the "autotrim" in response to stick input is another horrific idea from AB. Under manual control the only thing trimming an airplane ought to be the person flying it.

My perspective is that the PF was attempting to abdicate his responsibility to actually fly the plane and mistakenly expecting the computers on the airbus to figure things out for him. A pilot "flys" the plane with control inputs...the moment the input achieves the desired effect on attitude the stick is returned to neutral until additional input is required. As a general rule any individual deflection is limited to a relatively small range of the sticks total movement range.

How does one pilot judge anothers "airmanship"....isn't it via the deftness of his control inputs. Generally newer pilots tend to over control an aircraft via larger control movements leading to the need to "correct" their own inputs and creating a self induced scenario where they are behind the airplane.

So how does a professional pilot with thousands of hours dial in such a massive sustained control input and for what reason???

Was your experience different?

So your belief stems from a knee-jerk emotive response rather than a well-reasoned one. I don't say this to denigrate what you're saying, but if this is all it's about then it's the honest truth.

The use of the term "Airbus" was in fact originally a design specification for BEA (meaning British European Airways, not the French accident investigation branch), and this goes all the way back to 1965. See DeHavilland's HS.134 design to meet this specification below (and in the process note how much it looks like the eventual configuration of the B757 - now remember this was drawn up in *1967*.)

British Airliners 'Nearly Get It Right' Shock!

The concept is not therefore about "remov[ing] the aviator from aviation" and turning him into a bus driver - although let's be honest, the glamour was fading from the airline profession even by the mid-'70s, and Ronnie's union-busting antics meant that the only way pay and conditions could go from the '80s onwards was down. In fact the aviators who were instrumental in designing the Airbus FBW system would probably be mortally offended by that insinuation, and I for one wouldn't blame them.

"Airbus" was simply (at the time) a convenient umbrella name for the collaboration of European aviation companies - nothing more to it than that.

Anyways, back to the topic at hand...

Intruder, please don't mistake my comments. I'm not requesting that the CVR recording be played in public. But in case you aren't aware, the CVR transcripts are released in the US and obviously (proven by present leaks) the French also release transcripts.

see, NTSB - National Transportation Safety Board and you'll find that every report includes a complete CVR transcript.

We have one UAS event that resulted in a airframe loss. AF447. If I read interim 3 correctly, the AF447 PF had no other significant non Airbus experience. Therefore the only UAS that resulted in an accident was piloted by Airbus only pilots. It is entirely possible that the other 37 were Airbus only crews, but that doesn't change the fact that the accident pilot was Airbus experienced only.


http://images.ibsrv.net/ibsrv/res/sr...er_offline.gif http://images.ibsrv.net/ibsrv/res/sr...ons/report.gif http://images.ibsrv.net/ibsrv/res/sr...eply_small.gif

Dozy, I'm through going counter point - point with you. You argue like a politician. You restate your opponents questions into the question you want to answer.

I never mentioned autopilot turn control in a conventional airliner. I stated this in post #613: "As long as the FBW is in normal law, the SS is nothing more than a autopilot input device. Adding a bank request through the SS is the same as turning the heading select knob on the flight control panel. Both actions result in the autopilot turning the aircraft". Nowhere in that statement did I in any way compare the SS to an conventional (your word - which I find quite ironic)autopilot turn control.

Dozy,

There is nothing emotive or knee-jerk in my reasoning. The simple reality is that aviation is unforgiving and 99.9% wont get it done on the day you need the other .01%. Military aviation training comes with a sustained and quantifiable loss rate. The bottom line is that real mission capability requires realistic training and that leads to fatalities.

"Old School" professional pilots tended to have a military pedigree and while that might entail other issues by and large they had a very well qualified "stick and rudder" skill set.

The philosophical move away from aviation skill set to a sophisticated software designed to minimize or eliminate "pilot error" has led to less and less qualified pilots overall IMO. Flying will always entail risk and anything that minimizes a pilots actual ability to handle that other .01% will eventually come into play. The moment that "automatics" went from being an aid to the pilot and became a replacement aviation went backward 50 years.

PF's Experience
 

TTex600:
Responding more to a previous post that I can't find right now, one that called PF "very experienced," from the first Interim Report at
http://www.bea.aero/docspa/2009/f-cp...90601e1.en.pdf
is an abridged paste of his experience:
 Flying hours:
 total: 2,936
 on type: 807
PNF had approximately 4.5 times as much type experience.

Not at all - I think I answered them pretty directly.

Well, that is fundamentally incorrect as the FCU is *not* an autopilot. It is a flight control unit and feel system which is slightly smarter than the previous generation (which went into the A300, B757, B767 and MD-11). You may choose to see it that way, but that doesn't make it true. If the sidestick simply input to the autopilot it would be routed through the FMS (which is a completely different and separate system manufactured by a third party, usually Honeywell), which it isn't.

Again, incorrect - because the sidestick is not routed through the FMS and never has been, and also incorrect in a more material sense because the heading select knob orders a manouevre that is limited by the FMS in terms of turn rate and bank angle. With the sidestick you can command any bank angle (and consequent turn rate) up to 67 degrees, which is the limit of the flight envelope as discovered in testing.

Well, you didn't specify make or model of aircraft, so I went with the most generic term - so sue me.

I'm not trying to annoy you or start an argument here, I'm just pointing out that you have a fundamental misconception as to how the Airbus control and autoflight systems are implemented.

Tell that to those poor buggers in that B-52. Also, if you have a look in AH&N, you'll find a number of posts that relate to the early days of BEA and BOAC - the ex-WW2 pilots were known as "the barons", and some of them literally scared their junior officers with the chances they'd take and their apparent lack of knowledge of their limitations. Military training has one more problem as well, and that is ingrained adherence to chain-of-command. The F/O on the ill-fated Palm 90 flight was a former Navy pilot, I believe, and the CVR in that case shows that he was aware something was not right, but never felt he could contradict his Captain even though his Captain came up via turboprops and small charters in FL, and his own experience was in fast jets.

But that was never the case - you show me a single trade-level article proving that automation was designed to minimise pilot ability or replace the pilot in the cockpit, and I'll give you that point, but the fact is even in the early days that was not the intent at Airbus or anywhere else. If the current breed of MBA jockeys running airlines have decided that should be the case then it is their fault, and not that of Airbus. People obviously forget the flipside of FBW development which was that the protections actually allow pilots to perform things like evasive maneouvres more safely. FBW has nothing to do with autoflight reliance either - those are two completely different aspects of modern airliners, and as I said, it was Boeing who popularised the latter with the B757 and B767 (the A300 had it first, but it's market share was too small to make much of a difference on it's own).

I see this has already been answered, but I would like to suggest that you also read the story of the A300 and A310 (neither were FBW).
IIRC (it was a long time back), "airbus" in those early days was a generic term we used for widebody-shorthaul-twin designs in general, until "Airbus" snapped up the term and made it a trade mark.
Sorry, but "autotrim" is not an AB idea. It's much older than that. Concorde had an auotrim system, and apart from having to do some clever tweaking during flight testing, it never really was a problem. But then the Concorde pitch trim had the "bicycle bell", so you knew when it was doing its job.

If you read all with a closed or rigid mind .. of course you will find not a word about automation was designed to replace the pilot in the cockpit
But if you made ​​a critical reading (reading between the lines) you will quickly inform you of the possibilities and prospects (proposed) has not said or written about the automation
If you've read (closed and rigid mind) about the beginning of automation in industries .. it was written that it would relieve the worker of the painful and repetitive work and more leisure was assured ... he was not told at the time that this would send workers to unemployement :8
A this time some had already foreseen the danger (read between lines)

In fact it was a narrowbody specification originally (in 1965 - see link in previous post), but I imagine that the design quickly became a widebody with the advent of all major US manufacturers introducing theirs. The UK had their own proposal in the form of the BAC Two-Eleven, but ultimately this was squelched politically in favour of buying US-made designs in.

Well, you can know when it's doing it's job on the FBW Airbus as well, all you have to do is glance inboard. I suspect that given the near-constant operation an audio warning would have quickly become a nuisance - which is not to say that an audio warning when the trim goes past a certain point would be a bad idea - if any change was to be made, I think that would be the one to go for.

CJ's well aware of this, but the initial sidestick development and testing was done on a Concorde airframe. Someone kindly provided me the documentation, but I lack the requisite French to read it thoroughly. Maybe one day...

@jcj - This is only my opinion, but if someone decided that by, as you say "reading between the lines", then they got it wrong. The only training cost that the FBW Airbus series was designed to reduce was *conversion between types* by keeping a virtually identical flight-deck layout throughout the range.

Which automation at what level ?

The higher level A/P flight-director etc. - maybe. But, they were never intended to be a safety-critical must-be-working item, if they have become a crutch for less well trained pilots who cannot fly without them, that is operational failure (and cost cutting at expense of safety).

The FBW is more interesting. Look at the AB control law and particularly the way it is path-stable - it was designed to reduce pilot workload in hand-flying. Why spend effort on that if the intention is to get rid of the pilot ?

The real irony of course is that since AB designed this system, it appears that higher level automation SOPs and regulation has all but eliminated hand flying. If designing today, why would you bother reducing workload on a task pilots hardly ever do ?

The PF had 61 hours over the last 30 days and 368 hours with 18 take-offs and 16 landings over the previous 6 months (more then either the captain or other FO). Obviously he was getting significant "grooming" {looking at take-offs/landings}.

I think in the case of this discussion, the phrase "direct control" is meant to refer to a mechanical, or more accurately, a hydro-mechanical connection between the cockpit controls and the aircraft's flight control surfaces.

These guys rarely fly.

How do you expect them to recover from a stall?

Especially when other weird stuff is introduded like tempo loss of airspeed.

Modern pilots DO NOT expect to see a stalled aircraft.

They are 'aircraft managers'.

Answering a question with another question
 

A better question would be, can't I expect a professional transport pilot to not cause a stall in the first place? After all, I paid my money and I'm trying to get some sleep back there.

Here's another better question: If this pilot(s) was confused, why should that have happened? Eh? :(

Organfreak, FWIW, I have recalled tentative conclusions drawn by others in my post no. 1646 on the “thread 6” in tech log. Briefly, some ATPL’s thought PF confused the status with overspeed.

I was reminded of other accidents, and also by somebody posting earlier, that stressed people can fix on a presumed problem and keep trying to correct it, even when their efforts are unsuccessful – it is a “how the brain works” thing.

Well .. reading some stances of the CVR (Otelli book) not published by the BEA is chilling and I understand why the BEA is not happy with this publication

Who can find in these dialogues a bit of professionalism ?

Genuine question.
Is it Mr Otelly who did this CVR trimming !? Sort of the other way around of the BEA, leaving only the crap talking in the CVR or is it you ?


No quotation for example of what the Captain is actually saying for is briefing when he leaves. Just a notation about it.
While making sure to leave for all to read "Bon allez je me casse"

:confused:

All from Otelli
Well seem's that MD (captain) tell no more during the briefing (words in Otelli book + words in BEA preliminary report N°3)
In fact the briefing is made by the two copilots .. the captain "assist" .. likely a spectator ...
In all those chatters between pilots I don't perceive (my feeling) any signs of authority from the captain MD

English translation
 

Could someone possibly translate the excerpt from this 'unofficial' CVR publication into english - for those philistines like myself who never had the privilege of learning French, please?

There is absolutley no need to translate this. It has nothing to do with the accident and the information value of the discussion is zero. They are just casually talking. Which is not uncommon during a flight of 12 hrs or more. What professionalism? Do you expect a crew to sit on their mouth for all this time? Obviously you have never visited a cockpit on a long haul flight.

I agree that the hand over briefing is somewhat minimal but in reality you don't need to talk a lot (except a CRM question: "Did you sleep?" - "mediocrly" he replies, which means not really). Specially if it is the captain who has to be briefed, and nothing has happened, there isn't anything to be briefed except "everything is normal". You see with one glance the point you are flying to, the state of the Satcom and of course also the technical condition of the aircraft.

This is again a useless discussion to find answer in areas where they aren't, for the reason that they don't understand the real ones.

The "real issues" here are readily apparent and surprisingly simple...

1) the PF completely failed in following SOP for the event. Instead of flying pitch and power when confronted with unreliable airspeed he commanded the aircraft to climb. Beyond that initial failure he maintained maximum control surface deflection (full back stick) for most of the descent.

2) The PM failed to exercise command authority even though he was the senior pilot on the flight deck and was well aware that the PF was not handling the situation correctly.

3) The Captain did not assume command upon return the the flight deck in any meaningful way.

I'm waiting a formal invitation from you
BTW .. you will be outlaw to allow me in the cockpit
Those chatters are maybe not DIRECTLY tied with the accident .. but the tone (if you read between the lines) used between the pilots show a real lack of authority and discernment of MD (captain)
Lack of CRM .. already before the event

SLF.

Sorry. You are quite out of line. You have a pov, yes, but let's look at it more closely.

1. You are a line pilot flying for AF? How did you become so adamantly conversant with SOPs?

2. Explain how you have conclusively linked "Senior" with "Command"?

3. Captain issued orders that were followed. Smells like "Command" to me.

Defense? Because of your powerful and irrefutable pronouncements, I am interested in the trail of thought that produced them.

Seriously.

Can't we just stop now ?

Can't say you're not doing your alias justice...

Obviously not...

Lyman,

Are you a pilot of any sort, meaning have you solo'd any airplane...

I think that the answer is no from what you've written. My observations here have nothing at all to do with any given airlines "SOP". Things like unreliable airspeed can happen on any type of plane. Learning how to deal with these types of issues is a fundamental aspect of learning to fly. From what I can recall no mention of any checklist or procedural conversation has been highlighted.

The immediate response to unreliable air speed is to fly pitch and power and then once the plane is stable the PM can begin to sort out the details. The PM knew the PF was not responding correctly but i'm unaware of him initiating a formal scan related dialog. Meaning at no point did he read off pitch and power to the PF...which may very well been all that was needed to put the PF's head back in the game.

Aircraft law is derived from admiralty law. On any plane or ship there is always a chain of command. with the captain off the flight deck the PM was the senior staff member on the flight deck and unless otherwise specifically stated command authority was delegated by the captain he legally was in command of the airplane until the captains return.

By the time the captain returned it is entirely possible that the airframe was no longer recoverable. However from what dialog has been released and conversations here the Captains actions were entirely incorrect. He was unable to clearly see the PF's inputs or instruments. By relieving the PF and having the PM assume control during the brief transition he would have accomplished two things. The PM seems much more capable and coherent so he would have gotten a much better situation update in the 15 or seconds it would take to strap in and do a scan. He could then assume control and have both the best experience level and most accurate information.

It is inconceivable to me that the airplane crashed without the captain at the controls.

My comments needed no defense simply because they are so self evident and fundamental. Any pilot candidate would be fully expected to know that you fly pitch and power under those initial circumstances as well as understanding that the 1st step in unusual attitude recovery is to unload the air frame. It was fundamentally incorrect for the PF to apply any back stick pressure beyond what was required to level the airplane. The moment he began to climb the PM knew it was a mistake and failed to exercise his legal obligation to take command of the aircraft until the captains return...period end of story.