The three pilots of AF447 had about 20,000 hours between them. Didn't make much difference. There's more to this game than just hours, I think.

Hi,

Other guys trying to figure how to fly a jet
Interesting similarity in this report (from page 17 of the report) with AF447 cockpit events
other jet .. other company .. other culture .. other era .. other original cause (but pitots anyways) .. but pratically .. same guys
http://www.airdisaster.com/reports/ntsb/AAR75-13.pdf

Sorry - hasty reading on my part. It was the autothrottle he disconnected, but it has the same effect - the automatics had no control over thrust; it was all in the hands of the pilot. (Section 1.11.3 of the accident report)

I think the 727 stall and the same stick back behaviour from the crew finally puts the technology argument to bed. Thanks jcjeant

jcjeant,

That 727 report does read similar to AF447, I agree.... Lessons not learned?

It seems that the 727 also did not have an angle of attack indicator. Why this instrument is omitted so often? It would have prevented both accidents.

Just to be clear here, there is a subtle difference. If the pilot chooses to disconnect the autothrust by use of the instinctive button on the thrust lever (to manually control the thrust) then Alpha Floor IS still available provided the aircraft is not below 100' RA. At Habshiem, the aircraft was below 100' RA so Alpha Floor would not have activated......

If however the autothrust has been deactivated by the pilot (hold instinctive button for >15 seconds) or is u/s, then Alpha Floor is not available. If the pilot has deactivated it, it will only be available again once the aircraft has completed an air-ground cycle i.e. It's lost for the remainder of the flight.

Alpha Floor is an autothrust protection not a FBW protection.

Despite it's detractors the Airbus/FBW/Alpha Floor system can get you out of a whole heap of trouble..... But it is imperative that you understand what it's doing, why it's doing it, it's limitations and how to recover back to normal flight.

Know your steed!

bubber44

You do know that LH, BA and others have been doing just that, quite successfully for decades? The USN and USAF put 200+ hour pilots into all kinds of situations far more demanding that level cruise all the time. Think of shipboard ops.

Flying time, if it is the same hour on autopilot over and over again, is just numbers in a book. The training environment is key to safety, not a whole bunch of hours. 20,000 combined hours didn't help AF 447.

I did say "theoretically". And as for what I do/don't know...

It doesn't matter - the PNF's verbal reactions to the PF's handling are all in the CVR traces - down there in black and white (along with red, blue and green in this case):

"You're going up, you should be going down"
"Above all, don't make lateral inputs so large"

- to give but two examples.

You're being selective with semantics. What he did was permanently disable the autothrottles in order to get the aircraft into position for the flypast despite the fact he was too high and off course (because AF's briefing sent him to the wrong runway) - by the time he crossed the threshold of the grass strip he was too low and too slow (and getting slower!). Disabling the autothrottles meant that his thrust setting commanded the engines to spool down and also that the automatics could no longer apply TOGA power if the aircraft got too far into the low/slow corner of the envelope (in effect *partially* disabling the alpha-floor protection). To my mind this could be explained in two ways - either he didn't fully understand the parameters of the very safety feature he was supposed to be demonstrating, or he had complete faith in his ability to execute the manouevre with the safety feature partially disabled. Please feel free to PM me with your thoughts on that.

One last thing - reading the actual BEA report in that case reveals that it was not simply the PF and crew that came in for censure - AF's p*ss-poor preparation of the flight plan, and the discrepancy between AF's guidelines for airshow displays and the national regulations are also explicitly referred to, but the press chose to ignore that, because the PF's Quixotic war of words with Airbus made for more sensational copy (and also arguably because said PF became their primary source for leaking information on the investigation so far - crossing him would mean losing their "exclusive"). As such, it has become received "wisdom" that the BEA focused on pilot error to the exclusion of all else when this was not in fact the case.

Using the search function (thread titles with term "Habsheim") reveals two threads dedicated to Habsheim, one of which was in the last year. Doing a search for threads and posts with the terms "A320 trees" reveals about a hundred threads that are partially or completely dedicated to Habsheim over the past few years, including one called "Airbus technology defects" by "the shrimp", who was either one of Jacquet's devotees, or even the man himself - you even posted in that thread, so you're being completely disingenuous.

Anyway, no more discussion of Habsheim here - we're already outside the main thread of discussion anyway.

Not at all, and I have explicitly stated that several times. I slacked off too much at university to be accepted into the safety-critical real-time module which is necessary to do the kind of work required in aeronautical informatics.

I find it slightly bemusing that you believe in order to have the viewpoints I have I must therefore have some part of my livelihood connected to the manufacturer (and indeed consider me to be a "shill", which under most circumstances I would take as a deep personal insult), so let me just be clear - I do not, I am about as neutral as it is possible to be on this subject, and even if I did, I would still be neutral because I take pride in my work as an engineer and as such I believe it is of the utmost importance that anything which is universally perceived as a problem must be fixed - and I have in fact done my career more harm than good on more than one occasion by speaking up about policies that I believed would lead to an inferior product at the end of the day.

I argue that is more to do with the fact that manufacturers have the deepest pockets, and as such it is in the interests of the lawyers acting on behalf of the affected to go after them primarily. I'm not saying that the manufacturers are or have ever been blameless - they've all done things for which they should be less than proud.

Let me be clear again, I'm not interested in blame and never have been. I'm interested in finding as many factors as possible that *actually* led to the loss of this aircraft and want to discuss what it is possible to do to prevent it.

It's not the first time that has happened and sadly it probably won't be the last. A poor repair by Boeing engineers killed 520+ people in one go in Japan, yet you don't hear arguments on this board that the idea of repairing the pressure bulkhead, or indeed pressurising airliners in the first place was a bad one. I could make the argument that is is in part because the introduction of pressurised airliners didn't bring with it dark accusations of trying to replace pilots in the way that the introduction of automatics going all the way back to the stick pusher devices of the '50s did*.

I'm not going to make that argument though, because it is counterproductive to what I'm trying to get out of this discussion.

I disagree - it is an accepted fact that technology has it's limitations - the continued weary complaints about home and business computers crashing at inopportune times and being difficult to fathom a lot of the rest of the time are but one example of that.

I believe that the attempt to blame automation in general (and that of the Airbus FBW philosophy in particular) is also counterproductive. The analogy I'd make would be akin to blaming the manufacturer of the Stanley/utility/boxcutter knife for all the crimes committed with it over the years. Both aircraft automation and the Stanley knife are simply tools. They have legitimate uses and in such cases are very good at what they do. The problems occur when they are abused for purposes for which they were not really designed (cutting aircraft handling training to the bone in the case of automation, use as an offensive weapon in the case of the Stanley knife).

[* - This information comes from Davies' HTBJ, which is one of the bibles of airliner literature.]

PULKOVO 612 Tupolev Tu-154 (85185) seems to have had an AOA indicator and it didn't help them, very unfortunately. Also available, is an official simulator video on Youtube.

Quoting Safety Concerns:-

Yes indeed, thanks jc.

There's one particularly close parallel. The factor that started the 'accident chain' in the 727 case appears to have been a misunderstanding prior to takeoff, resulting in the pitot heaters not being switched on. And we all know that a similar pitot problem initiated AF447's problems - except that the cause in that case was outdated/inferior design.

The 727 pitots froze up in the climb - and apparently relatively high-pressure air trapped inside them, while the unaffected static ports went on measuring the drop in outside pressure, resulted in very high speeds and a very high rate of climb being shown by the instruments. The pilots appear to have concluded that they were in a dangerous overspeed situation and reacted accordingly - and, tragically, fatally.

The AF447 pilots initially lost all airspeed indications. Later on they seem to have received some 'valid indications,' but they were so low (like 60 knots) that they (knowing that the pitos were stuffed anyway) probably felt that they could disregard them. Meanwihle the altimeter was unwinding at over 10,000 feet per minute and the wind noise (with the aeroplane standing on its tail and falling fast) would have been unlike anything they had heard before. They then lost the vertical speed indications as well.....

No way, IMO, that AF447 can just be put down to 'bad flying' alone. Various instrument malfunctions and lost displays (mostly caused, apparently, by the malfunctioning pitot-tubes, which were known to be sub-standard and should have been replaced much earlier) must have played a big part as well?

I know it's only a preliminary report but the BEA makes no reference to what the crew got up to in Rio, and whether they were properly rested in accordance with the rules.

Dozy Dozy ...
 

So Dozy, now you even go to the extend to literally modify the CVR translation to make your point ...

Page 92 EN or 96 FR

Must try harder - the actual phrases from the English version were:

"Go back down "

"According to the three you’re going up so go back down"

"You’re at... Go back down"

"Above all try to touch the lateral controls as little as possible eh"

I don't see any substantive difference between the phrases I pulled from memory versus the ones I just wasted 5 minutes re-downloading just to prove a point.

I am very sorry but I do not quite agree, IMHO.
In the case of that B727, the pilots were erroneously misinterpreting a "consistently" high-speed information given on all speed indicators. They were "tunnel visioned" for the same interpretation of a "high speed situation". And they were all dealing with that false high speed situation. This is consistent with a "there's the high speed mach buffet" statement of the co-pilot.
Later on, the captain "understood" they were on a stall and asked for flaps 2º,but failed to properly use ATT indicators.
In the case of the AF447, they had no clue of what speed they were flying. That's when you turn on to the "feed-back" of your flight controls. When there is nothing more one can rely on, one would tend to feel the aircraft. I do concede however that there are a lot of common errors in both cases. The first ones being not to follow ATT indicators and SOPs.

But when the control column feel at overspeed (vibration, unresponsive controls) is so similar to that of stall (stick shaker, unresponsive controls), one could argue that the benefit of active artificial feel is limited.

@Intruder, below - I was referring to the NWA 727 incident linked a page or two ago, where the pilots indeed mistook the stick shaker for Mach buffet.

You would have to get significantly over the VMO/MMO to get the vibration similar to a stick shaker. Also, control feel, even artificial, is WAY different at high (over-) speed than at low (stall) speed. Artificial feel is generally designed to be "heavier" at high speed to help prevent overcontrol.

alph2z
That's a different accident. Pulkovo 612 did not have any failed instruments. The official report states that the pilots deliberately tried to climb over a thunderstorm, lost a lot of airspeed in the process, encountered heavy turbulence and hail at FL380 and lost control at FL390. The indicator is there to allow pilots to avoid dangerous AOA well in advance (unlike stall warning/stick shaker). Obviously, it will not help if the pilot insists on flying at a dangerously slow speed, at the aircraft's maximum ceiling, in hail and in turbulence - all at once!!

And what does it change to the fact that your following statement is plain wrong ?
Please quote your FCOM reference ... ?


Can you make your mind at least ?

Initially you state it is :
"all in the CVR traces - down there in black and white (along with red, blue and green in this case)"
but when caught out it is suddenly only :
"pulled from memory" ?

But the point is, whatever your CVR quotes are, they show one thing :
2 PNF had no idea what PF was doing with its sidestick, at best they were guessing.

And it is dishonnest from you to substitute :
"Above all try to touch the lateral controls as little as possible eh"
by :
"Above all, don't make lateral inputs so large"

Very different meaning !


No he did not.
Or quote the report reference … ?

There is no such thing as :
"partially disabling the alpha-floor protection".
As you don’t understand the system, read A4 post above if you’re ready to learn, maybe you will accept if it’s not from me … ?

If this a/c had a yoke instead of a sidestick, then what was happening would have been clearer to the other pilots.

I dont think the Captain and PNF realised that PF was holding so much 'nose up' input.