AIRBUS encourages pilots to ignore the STALL warning :confused:
This is the procedure in force at the day AF447 crash
http://i.imgur.com/raVo1.jpg

At the end ... under TRAIN
Respecter les alarmes décrochage
Respect (take in account) the stall alarms

Hello mm43;

I've seen words to that effect in past documents and I'm still searching through my archive. I found one item that hints that the stall warning may be false, but all other documents I've viewed thus far, (2002 to the present) having to do with UAS, state that the stall warning should be respected. But I know I have read somewhere that the warning could be false. The following may hint that the stall warning is/could be false but nothing in the document, (1997 & 2005) states that the warning should be ignored:
On a related matter, the link to this Airbus presentation may have been posted previously but it is worth taking a look at it. The presentation was created in September, 2006.

Unreliable Speed - Latest Improvements

Hi jcjeant, PJ2;

Thanks for your responses.

Well I've done a Google search and found the culprit article for which a Google page translation can be found here.

It is on an Airbus 'bashing' blog dated 20 August 2011, and I believe that it has previously been commented on in the PPRuNe threads, though I can't find a particular reference to it. The name Henri Marnet-Cornus is associated with it!

Now I've found it back in the AF447 Thread. Tilting at shadows?:}

Legal experts?
 

@mm43
At the site that you cited, I found an interesting blog:

Il appartient aux seuls experts judiciaires* de faire enfin éclater la vérité (ce n’est pas la peine de compter sur le BEA) qui est :

Le 31 mai 2009lorsque le vol AF 447 décolle de Rio pour Paris, l’A330 F-GZCP n’avait pas un niveau de sécurité acceptable à cause du défaut des sondes Pitot Thales AA qui équipaient cet avion. La condition compromettant la sécurité qui en résultait, entretenue par l'absence d'une alarme spécifique "BLOCAGE PITOT" et par un système d'alarme décrochage non conforme au JAR 25, réduisait la capacité des pilotes à gérer une situation dégradée et a provoqué la destruction de l’avion et la mort de 228 personnes.

Alain de Valence, Hubert Arnould, Charles Magne, Eric Brodbeck, Michel Beyris.
Hope, I did not forget my French completely:

It is only up to legal experts* to finally reveal the truth (it's not worth counting on BEA), which is:

May 31st, 2009, when AF447 took off from Rio to Paris, the A330 F-GZCP did not have an acceptable level of safety because of the problems with
the Thales AA Pitot probes that equipped the aircraft. The resulting unsafe condition, sustained by the absence of a specific "BLOCKED PITOT" alert and by a
STALL WARNING SYSTEM not complying with JAR-25, reduced the pilot's ability to handle an abnormal situation and caused the destruction of the aircraft and the deaths of 228 people.

* Alain de Valence, Hubert Arnould, Charles Magne, Eric Brodbeck, Michel Beyris

Looks like a very bold statement "...NOT COMPLYING FAR-25", but on the other side, who knows....

Here is the link:
[URL="http://henrimarnetcornus.20minutes-blogs.fr/"]Les dossiers noirs du transport aérien

This can be part of the F/CTL ADR DISAGREE when the RISK OF UNDUE STALL WARN STATUS MSG may mainly occur in case of AOA discrepancy and let's remember how the AOA data were different for AF447. Again, still not a single word from the BEA on this point ...

Also that information is important regarding the stall warning and the way to deal with it.

@Flyinheavy:

I'd take *anything* that site has to say with a shovelful of salt. It is not remotely objective in any sense of the word and simply a repository for a disgruntled ex-AF captain to vent about his dislike for Airbus.

@CONF iture:

This might be a long shot, but is it not even slightly possible that the BEA factored that into the tests and found it to be irrelevant?

Thank you CONF iture, yes, have seen this caution in various places both historical and recent. Also, I do recall coming across the OSV document more than a year ago and thought it prescient. The puzzle is, with so much information available prior to the accident, why did the accident occur at all? Getting information out to crews is always a challenge and doesn't happen in a few days or even weeks, and the sense of urgency is always contextual, so something perceived to not be an immediate threat receives less emphasis and dedication of what has become today, minimal resources, than other, "more immediate" concerns.

I have more or less assumed that the crew rejected the STALLWARN as an anomaly in the procedure (mechanical), and that they came by that info through the Airline, via meter, Bulletin, or other......perhaps even the grapevine. Was AF so in the weeds they launched hundreds of flights with a cobbled together procedure that amounted to nothing more than "Scuttlebutt"?

Which of course explains clearly why the STALLWARN was ignored....

In the document, it specifically states the a/c can relay to the instruments
"NOSEUP/HIGH airspeeds" NOSELOW/SLOW airspeed, etc. IF Bonin was in NOSEHIGH, HIGH AIRSPEED (he obviously thought so, crazy speed) then NOSEHIGH/LOW SPEED (HE WAS), what does he do? Reject IAS? That makes the re-STALLSTALL partcularly poignant, just at the point when things were going the right way, etc.....His cues were there for NU/high speed, and did he accept them due the Bulletin as false? They seemed real....So, how to recover? Reference to VSI? It's pinned, how can he trust that, has anyone ever descended that way?

Any upset, a/c or pilotage, may interrupt the flight path, and once interrupted, it cannot be regained if systems become unreadable, and no trained response works to realign it. No going back, once lost. No trail of corn, no experience.

If true, it certainly explains things in a different light.

Worse, though, is what it says about the guesswork and negligence present around UAS in its 30plus iterations.

Were the crews and pax to be 'observers" on dozens (hundreds) of "test flights" whilst each occurrence involved "ad lib" recovery? Or no recovery (447)?

Challenge. Given the presence of STALL warnings, and unreadable/unreliable airspeeds, more information from the CVR is required. I cannot believe in the midst of this problem, the only conversation we see is what BEA have trickled out...

Was it? He immediately went to TOGA/15° pitch ...

@PJ2

If you read the AirCaraibe report, there they mention exactly the discrepancy of respecting S/W or not to.
ECAM status page stating:

"RISK OF UNDUE STALL WARNING" and
"UNDUE STALL WARNING MAY MAINLY OCCUR IN CASE OF AN AOA
DISCREPANCY"

Could it be, that the AF447 PF thought of Air Caraibe crew ignoring the S/W? Still would not justify pitching up to >10° at FL350.
What if he simply mixed up values of memory items below and above FL100 as he did UAS in the simulator not long before the accident.

Anyway, PNF would have had to at least call out "PITCH" and they probably would have broken the chain of events. Were they both so stunned, that neither of them thought of calling for the QRH?

May be BEA report to come will give more insight.........

HazelNuts39

My only response can be that in his confusion, he was randomly selective about his response/not to STALLWARN. I apologize if I seemed narrow in my post, but his behaviour to me, shows a lack of recognition, as the others displayed, and emanated from an event that was not well known, had never been trained, and was a mere part of the LOC chain.....

how about.... "was variously ignored, and taken as fact..."

Klopfstein Head-Up Display
 

The French Gilbert Klopfstein invented the HUD, :ok:and showed it in USA with MIT's congratulations. The French Armée de l'Air nor Zieglers (Aerospatiale, Airbus) did not like nor trust Klopstein and his HUD as he was Jewish. Klopfstein's HUD has been used on many aircrafts, i.e. F16, F18, Space Shuttle, etc. :rolleyes:

I have seen a HUD in a IT flying A310 (Air Inter, today included in AIR FRANCE !) around 30 years ago . :cool:

Something is sure, AF had that HUD and don't use it :ugh: not a question of price !

Would AF put it in the AF447, ... and crew trained, could you explain what would had been different ? :p

Isn't HUD simply a panel "annex" with the most important cues displayed so all the pilot has to do is change focus (if even that) not field, with his scan? So the AoA would assuredly be displayed. "Eyes out" doesn't mean alot in 447's situation?

Because it is conflicting information - Nothing is clear.

Following the Air Caraibes events and the subsequent Memo, it was a known and accepted fact by Airbus that unreliable airspeed indication in cruise do happen and could be tricky. Airbus had to publish a clear and unambiguous note to all crews how to properly deal with such scenario that triggers a load of ECAMs but can be easily mastered if you do this this and that.
They knew that something was coming, hopefully not as bad as AF447, but something has to be done to inform crews before they face the situation.

As a start Henri Marnet-Cornus has never worked for airfrance, so I would suggest anyone to take anything DozyWannabe has to say with a shovelful of salt if they don't want to be misguided or confused especially when the guy is talking about Alpha Max.

That the word STALL in a way or the other has not be pronounced once in 4 minutes time ... I also think we are served only part of the story by the BEA.

The Truth requires no Seasoning.

And it may taste foul, granted. The Pilots are not Gods, and they don't have much to say about the operation of the business that provides their livelihood.

If the argument devolves to what did they know, and when did they know it? the personal histories of the various messages is irrelevant. Without a guideline for procedure, who would wager on the outcome of a situation that one time out of 30 was fatal? No one. Did Airbus, POST 447, issue a primer to pilots re: Stall? YES, Also, High speed handling? also YES.

If that is a finding, to my way of thinking, the pilots might not have responsibility for this accident at all.....

RetiredF4:I totally agree with you RetiredF4.

Further to the need of an AoA display unit, I believe this would be better than nothing but, unfortunately, that feature only addresses the stall issue. The question is wider than that IMHO. It has to do with the need (or not) for the PNF to "know" exactly what the other pilot is doing in all phases of flight and circumstances. How can I "see" if my copilot is over-controlling the aircraft? (Especially in the final stages of the approach?) How can I confirm that he is using all available performance of the aircraft to (lets say) perform an EGPWS terrain avoidance maneuver?
The actual (side-stick) system seems to leave one of the pilots out of the loop, throwing him into a situation in-which he is just watching the unfolding of events, once he can only take part, after his (delayed) interpretation of what was already "done" by PF. How can we interpret a "reaction" without knowing the "action" that led to it? What we "see" on instruments is the reaction of the aircraft to certain actions/inputs (whether done manually or via automation) but, as you say, there is an expectation created upon those inputs. We manage expectations along our flights. Long term (strategic/FMC/FMGC), medium term (FG/FCU) and short term (manual). Just like the 3 levels of automation... And when in doubt: click click...click click.

But, like Dozy says, the Yoke has faults. There were accidents where the Yoke did not save? So SS has faults?

So with the Yoke shown to be fallible, the examination of the SS is:

1. Not necessary

2. Not relevant to this accident

3. Not sufficiently alarming so as to warrant a Grounding

4. All of the above.

5. None of the above.

Likewise, though the A0A would seem to have been helpful, (and it was on the a/c, but INVISIBLE), should the AoA be eliminated from the discussion? After all, how many accidents were not prevented by the presence of the AoA indicator?

Inquiring minds want to know.

With the inexperienced new hires coming to the right seat I think of my instructing days and how the instructor had to use his eyes to see if the student was going to be able to land without breaking anything until he approached soloing. It would be very difficult in a side stick airplane.

I think having the yoke as a reference in your sight made it easy to monitor. Even in the 757 I had an FO yank and bank to touchdown every landing and he said it was turbulence. Watching the yoke twitching I knew it wasn't. SS's might weigh less but they sure take you out of the real loop on what is going on.

I don't think it would take an ace to know at 35,000 ft you can not have the yoke in your gut. The PNF would have reacted but with the SS he didn't.

No it doesn't but if you are talking about visual scan I think the yoke being full back would make you react immediately leaving 35,000 for the impossible climb to 38,000 at the rate he was climbing before noticing the attitude indicator, don't you? They were too heavy to climb to FL370 because of weight and temp in turbulence so 1.3 buffet was their chart to climb.

Even if they used the AB 5 degrees nose up climb power memory items at their altitude it wouldn't have worked because unless they leveled out immediately they would have been too high and too slow and violated their RVSM separation.

I was responding to a momentary reply that is now removed about why the attitude indicator and 38,000 ft wouldn't have clued the PNF.

And it is interesting to note that there is still no trainer in flying school equipped with such independant SS, probably for that very reason.
But now guys arrive on the RHS of a 320 with virtually no flight time at all ...

Question for you or aguadalte or RF4 or anyone who seems to share my view how the Airbus concept suppresses valuable information : What made Airbus to embrace such philosophy ?

Cost is my guess. I was able to bid away from AB so didn't have to leave Boeing planes my whole career. Loved Boeings because you always had control. Two button pushes and you were a DC3 style airplane. You were 100% in control. It made flying easy.

You could still have given a try and like most of us you would have loved it. With time and perseverance we learn step by step what’s inside and how to master it and recognize the weak points. Of course you can still fly it like a DC3 except you’re not allowed to force direct law, so you’re still vulnerable to a Qantas mishap ... but what are the odds after all.

I am still left with the question - would a yoke or visible side stick have helped in this accident? The conclusion given the information we have is no. The PNF had all the information he needed and didn't need to guess what was happening from seeing the position of anything. Gentlemen this is just clutching at straws and displaying your prejudices not objective examination of the situation. Look to the psychological and 'soft' features of this accident not whether you can see a side stick or not. Instruments, gentlemen, instruments.

Sidestick concerns
 

The lack of visual cues or "feel" of the other pilot's control position is a bogey man.

You see what the jet is doing and question the other dude/gal and then get in the loop.

I have soloed dozens of troops in the Viper with the sidestick and zero feedback and zero stick movement ( actually about 1/8 inch after first few came on line). That's the family model. In the SLUF we had no two-seat trainers so first flight was solo. Flew close chase and talked to the troop.

As several here have stated over and over, it's the instrument crosscheck and determining if the jet is doing what it is being commanded to do, or not.

Ask any normal flight instructor how he would like to teach basic flying and see what they say. You need visual feedback of what the controls are doing. Once you have an advanced student you might get by with no visual feedback.

However, this is an advanced flight situation - not a basic training exercise. No visual cues outside the aircraft at all.

I think bubbers44 meant cues direct from controls, not outside.

Surprising that the cmd didn't command the pf as to his action and ask, repeatedly, what he was doing. On the evidence, could an assertion that all the crew were aware of the pf inputs be refuted ie there was a consensus in the fd? All the ss discussion seems to be that the pf was a secretive incompetent renegade and if only the others had known and acted!

And if the pnf had to get up to let the cpt in, why didn't the cpt immediately take his seat ? The mystery of the undone straps was never resolved was it?

CONF iture:
The provocative answer to that question would be: "because Airbuses were designed by engineers but unfortunately flown by pilots"...
My personal guess is that they wanted to brake with the "old" concepts. Starting with a blank sheet, they wanted to show the world that they were able to design an all new aircraft reflecting the European potential for innovation. "An aircraft made against pilot's errors"...
I had the opportunity to meet Pierre Baud, when I was invited to Toulouse in the early ninety's, to fly one of their A330 testbed aircraft. That was my first experience with a FBW aircraft (before that, only Boeing and A310's) and I personally had the chance to verbalize my worries in this regard (no feed-back on SS, added inputs on SS, lack of need for trimming, ATS in step of Auto-Throttles) and he explained to me that this was a new concept of aircraft and that the FBW represented not only an evolution on the handling characteristics but also the chance to save on weight and that the final goal would be MFF and communality.
I do understand that at certain point of the design work, AI engineers had to take options. They have opted for this system and there is no go-back, at this time. It is not yet clear for the majority of the pilot's community that the system is wrong. There's a great deal of pilots feeling quite comfortable flying SS without feed-back, but as Donald A. Norman wrote on his:

@aguadalte & CONF iture:
 

The automation was made less intelligent with ALT LAW (PROT LOST),
the human (pilot) was in control...but failed.

Conclusion?......green light for MORE automation.

It's always distressing when everyone's talking and no-one takes a second to listen:

"It is not yet clear for the majority of the pilot's community that the system is wrong."

Oh really?

If the accident/incident rate were different between AB and Boeing, then you would have a point, but as we know, it isn't.

I hate to ask, but I surmise you do have time on the bus?

Thanks for the reply aguadalte. Interesting.

That kind of aircraft is so dependent of the quality of the inputs received through the numerous probes. If a doubt exits on any of those probes or sensors, the aircraft should reverse straight to the most simple law : the direct one and get rid of any kind of automation. To have quietly autotrimmed 13 deg NU under those conditions has been a killer.

That's the advantage of fully visible yokes, no need to wait or ask, you are continuously in the loop.

Again, it has never been question to doubt what the instruments are telling.
It is question to know what has been commanded, or not. How do you know if you can't be aware of the command inputs ? Maybe the aircraft is doing something that has not even been commanded by the PF.

Fully visible flight control commands place you one step ahead in the thinking process.

@CONF iture


I share some of your views, but this statement seems a little strong, as it has been showed by BEA.3 that PF did use N/U SS inputs nearly all of the time for whatever (wrong) reason.

Means I would concur if he would have tried to overcome a NU trimm with N/D inputs...

I don't recall many incidences where the SS bus has not responded appropriately to sick movement, one that comes to mind is the cross wired stick, that reversed roll. F/O took command and landed safely.

As for AF447, the PF almost continuously applied nose up, when he applied nose down the A/C responded.

After all this discussion, I still cannot figure out why the pilots managed to stall this aircraft, it seems it was a simple case of loss of airspeed?

If - and *only* if - you see them. Many haven't.

CONF, if you only ask people who already agree with you, you're less likely to learn anything new. Most of the useful stuff I've learned has come from people who either agree partially or don't agree with me at all.

Pretend this isn't coming from me if it helps, but at least one of the reasons for the switch was that the yoke was designed around cable control. Before hydraulic assistance you needed a control column that size to have the leverage to move the flight surfaces. The 737 still requires it because it's manual reversion mode is cable-operated. None of the widebodies did, as they were fully-hydraulic, but it was kept anyway with a complex electro-mechanical feedback system in place. Same with the 757. The 777 and 787 use a software-controlled feedback system, the code for which is in all likelihood more complex than every system on the A320 put together. The A320, like the 757, was a narrowbody that was fully-hydraulic, and therefore one of the main reasons for having a yoke was no longer there.

With a fully-hydraulic system, one of the main reasons for having the controls interconnected - i.e. the need for both pilots to exert leverage when control cables are damaged - also goes away, so Airbus developed a system that would attempt to enforce one pilot in control at all times. Being an airliner and not a trainer, the need for one pilot to feel what the other is doing was greatly reduced, and losing the interconnection also removed the possibility of pilots fighting over the controls or having to work against the force exerted by an incapacitated pilot's body interfering with the yoke.

Being able to see the primary flight control movement buys some time, but only a matter of a second or two, if that. And if a PNF is really unsure about what their colleague is doing, they can take control at the press of a button and lock them out by holding it down (although would only be recommended in extreme circumstances - e.g. EgyptAir 990).

If you come at the problem having already concluded that the yoke is a superior control method in every sense, then you don't find this stuff out because you don't want to. Similarly if someone were to come at it from a conclusion that the SS is better, then they'd close their ears to what you're saying. Believe it or not, I do take on board the advantages the yoke has - but I don't think it's enough to say it's better or safer in every respect, because in some scenarios it clearly isn't.