The software gave them what they asked for at all times, but what they asked for was dumb. Regrettably to me this boils down to the fact that if you lose your airspeed instruments at altitude, close to your performance ceiling, initiating a 7000fpm zoom climb is about as wrong a response as you can have. Having done that if you then find yourself descending rapidly with the nose high you should recognise the stall for what it is. Its no good blaming the software or the system. If you did the same things in a 737 you'd get the same result.

Edit: if you have dual inputs on the sticks it will tell you, and you can override the opposite stick with the priority button. This accident was not caused by confusing dual inputs but by inappropriate and sustained solo inputs. The computers don't make assumptions, they follow commands. What the THS is doing is largely immaterial so long as the aircraft retains the capacity to deliver the commanded load factor, which it did throughout. Pulling and holding full aft stick in an Airbus is akin to pulling full back stick and holding full aft trim on a Boeing. There is no mystery or confusion about it, it's Airbus FBW 101.

Someone mentioned a 1.6 g spike at the start of the climb. Doesn't sound like much but would feel quite violent. It also probably mean they had just clipped a quite strong updraft. Personally I believe that no pilot would maintain a pitch attitude of 16 deg at height intentionally it would look too weird.
Also if you go hands free from straight & level, neutral stick, the a/c tends to climb at high level due it trying to maintain 1 g.
Methinks too much blame is being metered out to the pilots. But then again it is ever so.

OK465 : When I say that the "average" line pilot cannot recover from a deep stall I actually mean that very few of us have any training and experience in recovery from a departure in a swept wing jet.

While the majority of my trainees will have to follow some (very loud) commands from the old man behind, there are a few ones around who has the aggressive, composed attitude to keep fighting and who knows what it takes.

No doubt an airline, who would only employ Imperial Flight Test School graduates, can expect that this particular field of operations is covered (and the salaries will again be "up, where they belong" :E ), but we live in the real world, and my question remains:

When is enough (training) enough?

As I see it, a good many operators have a long way to go before reaching "enough"; just the fact that in the Mach-region pitch attitudes outside +5 to -2 degrees is risky business comes as big news to quite a few fellas.

So if a majority of pilots in a given airline has inadequate knowledge of the environment in which they fly, is that to be blamed on the individual pilot or is it a "system failure"?
If the pilot corps cannot deal with a 1/10.000.000 probability failure, is that to be blamed on the pilot, the airline or the manufacturer?
If you add severe weather, night, black sea, does that make it "natural causes", pilot error, system failure or design failure?

Which one will be the cheapest verdict for the huge economical interests involved here? And how do you make interim reports pointing in that direction in a subtle way? :*

"Look"?

In darkness at altitude over an ocean, possibly in cloud, I suspect there was no visual reference to "look" weird. I suspect that was part of the reason for failing to realise the attitude of the aircraft during the stall.

- with your training experience, I'm sure you could could you explain this statement to we non-AB folk? It seems a little odd.

As I understand it the aircraft (if not the pilots) knew the AOA was too high. So perhaps changing the audible warning to announce the AOA would be benificial.

At the very least change it so that the warning isn't the same pre and post stall.

It seems they never realised they were fully stalled so can you really blame them for they way they responded?

The other possibility is that one or more did realise they were stalled but assumed the PF realised it and had tried to get the nose down. There is a lot less discussion of the situation than I expected.

BOAC, I believe it's because 1G is acceleration due to gravity at sea level. The same acceleration at 40000ft might be 0.9999G (I'm sorry, I'm not clever enough to work out what it is) therefore to maintain 1G the aircraft will slowly pitch up.

Hmm! So the clever software is not clever enough to know its altitude, and will eventually command a loop to maintain 1g?:) Don't forget it will also need GS in order to compute centripetal acceleration.

ICE PACK

Sorry where do you get this from So it will not go into a gentle climb trying to maintain 1G.

Kalium Chloride

Not quite, that is why we "pilots" have to fly on instruments, namely the PFD, on most new A/C (artificial horizon), with reference to an ATTITUDE. I would certainly never expect to see 14 degrees plus at 35,000'. You also start to see big red arrow heads which show you where the horizon is as you go further from level and they get BIGGER the further you move away from level.

That's an urban legend. Releasing stick to neutral in pitch in normal and alternate laws results in pitch hold, not 1G hold.

Most probably all 3 pitots were affected by icing. Everything between probes and instruments was in perfect working order until aeroplane was smashed against the ocean. Static pressure system was totally unaffected. There's no indication that any of four horizon references toppled during final minutes of flight.

Per definition of deep stall, no pilot can recover unless aeroplane is equipped with antispin parachute and it's used properly and timely.

Flight controls computers acted in accordance with sidestick inputs.

There were instrumental references, three of them to be precise. Two EADIs and ISIS. It is mighty certain that two of them agreed and so far there's no reason to believe that third toppled.
Page 109 of BEA 3rd interim report refers. Pitch is "assiette" in French. would you believe pilot attempting to take off without clearance? Trusting only instrument that has failed and stalling the airliner on climbout? Failing to control speed and stalling on short final?

Initially it was the youngest pilot, sitting in RH seat. Control was later transferred to older F/O in LH seat. Sidestick traces clearly show who made which input and when.

BEA has thoughtfully included sidestick and control position plots in its report so anyone able to read them should not depend on BEA's wording only. A330 is big aeroplane with powered controls. Both of it spell: i-n-e-r-t-i-a.

I am sorry sir, but I fail to see how replacing FWC shouting "STALL-STALL-STALL" with one saying "Your AoA is ten... it's twelve now... whoops there goes fifteen... boy, you've hit 25 degrees AoA" would be benefical.

What really upsets me is that everybody seemsto concentrate on what happened once the aircraft stalled : Who Cares !!!!!!!????

The central question is : Why on earth did they get themselves into this and how ?
BEA has pretty much nailed it : the least experience pilot pitched a whopping 10 deg up without applying power !!! and the most experienced, instead of keeping his eyes on attitude and whatever was left, chose to turn his gaze elsewhere to call the captain to the flight deck.
The failure was never recognized, therefore the adequate checklist never called for.
How could anyone of these 2 ever realize they had stalled when awareness was at its rock bottom from the start ?

Of course they did everything they could to get out of this.................not having a clue about the problem they had to solve.

All this comes to basic airmanship, plain and simple.
Wether this airmanship is training related is open to debate.
Once stalled, I too, would have never been able to recover.

Me myself:

Your post tends to illustrate the good sense stated in Two's In post:

In short, a basic misunderstanding of what a stall is.

My Boeing aircraft experience ended in the early ninety's. Since then I have flown A310/320's/330 and A340.
Must tell you that I still miss the yoke and the (artificial feeling) feed-back that I used to receive from it, as well as from the Auto-Throttles.

Let me ask you this:
If the PF co-pilot was "feeling" a sluggish yoke, would he ask his crew mates:
"I have the impression that we have a crazy speed, what do you think about that?"

Clandestino,

To be honest I don't know how the A330 differs but if you disconnect the autopilot in level flight at altitude the A320 will pitch up slowly. I know, I've done it.

From my FCOM

"With the sidestick at neutral, wings level, the system maintains 1 g in pitch"

I've never heard of 'pitch hold'

[Static pressure system was totally unaffected.]

Mhhh... I may be wrong, but if the static pressure system was totally unaffected, why does the vertical speed graph (page 109 of the report, second graph from top) looks so weird, with huge jumps?? They may also have excluded the vertical speed info because of that.

I think the responsibility of the pilots in this case is pretty limited. They failed to identify a stall, but the fact that the stall warning appears only when you put your stick would make me believe (if I was pilot) that the stall warning is also wrong, as everything else.

I don't know who got the idea of excluding AOA measurements based on speed. If I had designed the system, I would have kept this data all the time (except possibly during the take off and landing runs. For knowing that you just use the landing gear sensors as input and that's it) .While in flight, I don't see any configuration where the AOA measurement would be too bad because of low speed. Was the A330 designed to fly at 60 knots anyway?
I cannot understand the logic behind this design. If somebody can, that would help me.:ugh:
For me, this bad logic is a driving factor of the crash. Had the stall alarm sounded *permanently* during the fall, I would have expected the pilot decisions to be different.

1/ the report is a factual description of what was happening
2/ it never says it is a crew fault.
3/ as in many other domain, from facts, in view of your own experience, you make your mind.

1/ the crew never recognized they were stalling. (would I had ??) It is a fact, the report is clear in that matter.
2/ a fault is when you do purposely something you should not do. I do not see any crew fault in that report.
3/ some occurences air data references losses because of freezed probes were reported on the same aircraft type several times in Air France and other companies.
.
3/a Airbus was aware of the problem, and a probe replacement campaign was scheduled.

3/b in the time frame the probe replacement campaign was ongoing, they have never been trained to cope with that kind of major failure. in fact it looks like the training in "pure" aircraft handling is very poor in the training syllabus.


I hope this accident will help the aviation international authorities to rethink the pilot training syllabus, but also to improve the " field experience feedback" processing in aircraft safety.

I am afraid... it will lead to discharge the responsibility of all accumulated mistakes from all parties onto the three crew members.

That is because a lot of operators believe that with high speed you cannot stall thus no need for AOA indicator. These poor souls believe that the answer is ALWAYS to point the nose down which is the answer MOST of the time however not all the time. In reality an airplane can stall at any attitude and airspeed (within structural limitations of course.) The IAS and ATT are subordinate and supporting to an AOA indicator in relation to what the wing is actually producing. Even if the pilots were clueless the on board computers would know this (tongue in cheek)

So does the 330 have an AOA indicator and if not do you think that if it did and if the pilots were trained to interpret it would it have helped? After all it is giving a true condition into what the wing is doing.

Yes but this goes even beyond that... even if you don't want to display the AOA info to the pilot (which I can admit), I don't understand why you would start to exclude what one sensor is saying using the data from an unrelated sensor.
As long as the plane is in the air, the relative airspeed cannot be less than 60 knots, I think everybody will agree about this. So why was this logic implemented?

Takata, pardon ma inexpérience: is that a flap actuator or a stabilizer one? Merci
 

To POIRE

Its THS screw shaft
(Are you expat in Middle East)
Take care

Cockpit of the future
 

Based on the principle that automation came to minimize pilot error, I foresee the following flight crew formation:

One pilot and one computer specialist. The computer will have a new "law", the emergency law. The computer will be able to quickly identify the nature of the emergency and take proper action. Auto pilot will disengage only when the pilot decides to do so. Of course, he will check first with the SS (System Specialist) if the computer will be able to solve the problem.

If computers are able to fly a crosswind ILS, with gusts, I don't see why not the computer can't "read" the checklist on, for eg., procedures for unreliable airspeed.

To train a computer is cheaper.

The pilot will be a "flight consultant" to aid the computer with the "gut feeling".

With you until the "gut feeling" :ugh:

What would have happened if none of the pilots had been there?

I'm not willing to go into that one again in public so I'm putting together a PM for you, but I will tell you it wasn't in Paris, it was at Habsheim/Mulhouse airfield, which I think is in the Alsace region. I will also tell you that a lot of second and third-hand information on that incident came from poorly-translated, largely inaccurate and occasionally sensationalist press articles. :)


It's a quirk of the A320 which was corrected in the later models (A330, A340 and A380) from the outset. It may have even been corrected in later A320s.

Most of the time you're at cruise - i.e. at an altitude where this becomes noticeable - in an A320 which exhibits this quirk (or indeed any other modern airliner for that matter) will be spent under FMC (autopilot) control with altitude hold engaged anyway, so it rarely comes up.

Hi,

I think the answers to your questions are in the BEA N°3 report ...

The PF made a announcement to the cabin personnal about turbulences to come
The PF was not belted on his seat (nothing .. completely free)
Are this a normal or professional behavior of a responsible pilot ?
My answer is no
What you can expect from such seriousness pilot in case of an emergency situation ?
The AF447 tragedy

Apologies, me again.

"There's no indication that any of four horizon references toppled during final minutes of flight."

So why didn't they look at the attitude indicator/s, which would have shown them to be steeply pitched up?

Isn't that a primary flight instrument when you have no external horizon reference?

Speaking of Laws....
 

I think a little humor is needed right about now.


1. A flight computer may not injure a human being or, through inaction, allow a human being to come to harm.
2. A flight computer must obey any orders given to it by human beings, except where such orders would conflict with the First Law.
3. A flight computer must protect its own existence as long as such protection does not conflict with the First or Second Law.

Nice to see somebody else here likes Asimov and has been reading "I Robot".

But no, the Airbus FBW flight control systems are not 'robots' and don't 'think'. They were an evolution of the basic AFCS (automatic flight control systems) that existed well before the A320.
Whether the designers should not have taken more authority away from the pilots (as they did), is an interesting question.

I can't remember the "Three Laws" being pinned up in our design office... maybe they should have been.

But even Asimov's robots were not programmed to deal with every aspect of human stupidity, and neither is the Airbus system (Habsheim comes to mind, again....)

For the record, I am certain no one is inferring that one or more of the 447 drivers were stupid. For if so, it would be a blanket condemnation of Aviation as it exists today. Something's wrong, but I would hope no one is trying to polarize........ Condemnation prior to investigation is truly.....stupid.

Would any of you B drivers care to comment?
 

My Boeing aircraft experience ended in the early ninety's. Since then I have flown A310/320's/330 and A340.
Must tell you that I still miss the yoke and the (artificial feeling) feed-back that I used to receive from it, as well as from the Auto-Throttles.

Let me ask you this:
If the PF co-pilot was "feeling" a sluggish yoke, would he ask his crew mates:

"I have the impression that we have a crazy speed, what do you think about that?"

Hi,

This is the reaction of Air France on the report No. 3 of BEA
I remain astounded by the words of the spokesman for Air France
This is a position of denial concerning the recommendations of the BEA
In french of course ........

Rob21
However, the landing is a different matter. For the A330, max crosswind for Autoland is 20 kts. For a manual landing it is 32 gusting 40.

You may consider it humour, but the fact is that Asimov's laws were indeed considered as a baseline by the engineers. In fact I believe they were considered as a baseline by the engineers responsible for all the advances in flight automation made over the last half-century.

Well, that's Rule 1 bust then - Stall Warning logic?:ugh:

Shame the 'engineers responsible' didn't READ the wallpaper.

Although this question wasn't addressed to me I'm going to answer it.

The question you ask is disingenuous. The complaint is not that the software didn't do what it was supposed to do but that what the software was doing was not communicated to the crew in an intelligible way.

It's simple.

(a) When the airplane is within the flight envelope the stall warning provides an aural communication to the pilots: SILENCE.

(b) when the software decides the airplane is outside established parameters it provide an aural communication to the pilots: STALL STALL STALL

(c) when the software decides that the airplane has so drastically exceeded established parameters that the data should be considered invalid it provides an aural communication to the pilots: SILENCE.

Now, if you think that the pilots were confused and failed to recognize the stall for what it was then you have to ask yourself how the above represents good software design. Because it does not. It's horrible software design. It's horrible software design for the software to communicate the exact same aural message (and silence is an aural message) to the pilots both when things are 100% good and when things have gone totally to hell.

@MountainBear : I'm sure that will be addressed.

A question......

I generally associate OZONE aroma with electrical issues.....


Fair?


(WRG)

I believe the whole discussion about the stall warning inhibited below 60 kts is a red herring. The stall warning on AF447 sounded continuously for 54 seconds without any indication that the crew acted on it. In my opinion it would have made no difference if it had sounded 30 seconds longer.

Once the aircraft was falling down with an airspeed less than 60 kts they probably had no way to unstall the plane anyway, they were miles away from any reasonable flight envelope for a commercial airliner, even miles away from test pilot territory. Terra incognita.

(As far as I know the stall warning is inhibited because the AoA vanes do not work below 60 kts. You cannot sound a stall warning when you have no working sensor to detect a stall.)

Negative.... ozone is associated with (usually high-voltage) arcing.
This WRG issue was associated with an avionics bus communication between two computers.
The voltages on such a bus are in the order of about 10V, and the electronics are incapable of producing an electric arc (short-circuit currents inthe order of milli-amps)

keep in mind AF has their own checklists, own language (french) and all book are in french.

which is not in the Airbus philosophy.

example: the cockpit is in english, ECAM, FMGS are in english, and at airfrance they think , talk, and eat in french.

"vitesse de fou!!", did he mean high speed or low speed?instaed of "woow, we are fast", or "damn, why are we so slow?"

pure mistake from the begining for one of the best airline.
they should review their pilot training, it seems to me they have a bunch of incompetents pilot who think a stall is a "crazy speed".

any aeroclub pilot, know what to do when they have the little red warning light flashing...plus on the 330, it s printed 'STALL' on the PFD.


golden rule:
1:nose down
2:full power, or TOGA

if you can not get out of stall, there is nothing you can do!

only idiot keep nose up during a stall, flat stall, deep stall, whatever you call....by chance they won't kill more people!

wait for the next crash done by a low trained LCC P2F pilot... we will have a good laugh when one these planes will crash on a school or hospital in the middle of LONDON killing thousands.. :ugh::ugh::ugh::ugh:

Hi,

The stick of PF busy making mayonnaise ..

http://i.imgur.com/X3HIW.png