AF447 wreckage found

[xcitation]

Christiaan

I know about overide button on SS.
However I also read that the FO in left seat lacked the authority, unless he declares FO in right seat incapable. Difficult in this situation as there was no obvious medical problem with PF. Il voudrait un coup d'état, non?

[BOAC]

Out of interest, what happens in this 'computer game' if the one who has 'lost' control then presses the button for the requisite time?

[Lonewolf_50]

JJFFC

2/ Don't forget that neither the PNF nor the Capt. ever knew that the PF made an initial huge nose up => how many of us could have imagine that ?

From CVR release, early in the event PNF was nagging PF to get his nose down, or to stop climbing.

Am I right in saying that on this type if inputs are made on both sidesticks the system will take the algebraic sum of those inputs?
In which case if one pilot was maintaining full and up and the other full down this would equal neutral - not what you need for stall recovery.

Back to "take the controls" if you have to come on to the controls ... CRM issue.
MountainBear:

I have held that
(1) the logic of the stall warning system is flawed
(2) that this accident illustrates the nature of those flaws
(3) that these flaws are one possible explanation for the pilots behavior doing one specific phase of the accident
(4) that the professionals who designed the system should be held accountable for those flaws to the extent they played any role in this accident.

Rational position to take. I'm on board with most of it.
Dozy:

I'd say the extreme AoA was probably fouling the static data from 02:11:47 onwards.

More likely fouling pitot data. Static doesn't tend to be in the airstream, unless they've changed that recently. You don't want to contaminate static air with dynamic pressure, do you?

Originally Posted by houlett

When a french pilot said GAUCHISSEMENT he speak 's about ROLL

Merci, monsieur.
Three miles

If they felt they were fast and descending perhaps they thought this was a high speed stall warning? In which case relaxing the back stick but still mantaining some might make sense.
IMHO the ignorance {me - Ignoring?} of the audible stall warning could only be caused by the pilots assuming, that the stall warning would be triggered by too low speed (instead of AoA).
Speed incorrect -> Stall warning -> ignore it
When the speed came back they had stalled the plane and the speed stayed low => assumption: still incorrect.
Why should they believe now, that the speed was more correct than before?
Why should they believe with all this confusing stuff around them, that any other indication was more correct?
The speed even became lower and lower and the stall warning disappeared at some point. -> continue to ignore all this

Well said.
Kalium:
I wonder what he {captain} would have done if he'd heard nearly a minute's worth of stall warning.
Might have changed his SA, that.

Mimpe:

When the birds hit the 737 over Manhattan, pretty much the first action fronm Sullenberger was.." My Aircraft".

Good point, but Sully was the Captain. There was no role ambiguity for him. PNF had that to deal with.

...my concern is that PF may well have been more than just worried about overpeed. Its clear he had lost his initial crucial scan, was overcontrolling, and I feel as the aircraft slowed with zoom climb, may well have crossed the rubicon of lapsing into a sensory interpretation of what the aircraft was doing.

Possible. How often do AF train hand flying on instruments at high altitude? How comfortable was he in that scan and that task?

I feel the decleration he was experiencing in the late stages of the climb, heightened by marked anxiety, probably ended in a somatogravic illusion of acceleration, and persistent nose up inputs unto death ... situation required immediate transfer of control to PNF, whose spatial orientation and understanding of the situation appears clearly superior throughout. No time for egalitarian social norms.

Yes, early on. By the time the Captain got there, he confessed that he wasn't sure what was wrong. His SA hadn't quite deciphered the SW when it was correct.
GerardC

In all, 4 stall warning activations. Would not YOU be confused : nose down -> stall ; nose up -> NO stall.

Was this sort of thing ever practiced in the sim among A330 pilots? Among AF pilots?

This whole scenario reminds me how dependent we are on airspeed as a crosscheck for performance.
ross M:

About auto switching back to normal law protections. (That puts the robot into a struggle with the pilot for control deflection).

As I understand law changes, latching Alt 2 requires substantial effort to revert to normal law. Auto switch is probably not a good idea, from a design perspective, and for any pilot flying.

If you keep getting spurious inputs, auto re-engage would kick you back to Norm Law from Alt Law even if the problem didn't go away.

You could get stuck into the following cycle: Alt Law kick Norm kick Alt kick Norm kick Flicker -- talk about a pain to fly.

Imagine flying with the electric trim kicking on and off, intermittently. Eventually, you secure the trim so it stops doing that to you. (Yes, I've had trim weirdness like that in the dim and distant past. Ended up being a touch of hydraulic contamination. )

RWA: nicely said, thanks. So many good minds at work today!

carlosgustavo:

LONEWOLF_50. I dont know if you are right about the sound of stall inside a stall. All Airbus says on the O. Manual is: WHEN STALL WARNING STOPS PILOT MAY INCREASE BACK PRESS AGAIN IN ORTHER TO RECOVER TRAJECTORY.
I dont for you, but seems To me that if it works as you say Anywhere in the Operations Manual should be a Caution warning that at high AOA stall warning stops. BUT IT DOESNT.
Thanks for tour answer. Is good To learn from others and To share thoughts. Good flights.

The stall warning alpha looks (from my tech info) to trigger stall warning at a value of alpha less than stall alpha. That tells me you get the warning before you actually hit that critical alpha and stall.
Does that make sense to you? (I may be missing something). It makes sense to me.

For your kind words: Muchas gracias, carlos, y vaya con Dios.

Originally Posted by somebody

To that point the PNF did not brief CAP that the stall warning had been ringing for a minute.

Aye, he seems to have filtered that noise out.

BOAC:

Out of interest, what happens in this 'computer game' if the one who has 'lost' control then presses the button for the requisite time?

He gets to keep "flying," eh?

[carlosgustavo]

Excellent link To the matrix of the accidente. Looks real To me.
They were more focus on bank than in pich.
Anybody else think of a SPIN induced by stall plus turbucence that they didnt recognice because of the IMC condicions?

[DozyWannabe]

Originally Posted by Lonewolf_50

More likely fouling pitot data. Static doesn't tend to be in the airstream, unless they've changed that recently. You don't want to contaminate static air with dynamic pressure, do you?

I'm trying to make sense of the VSI trace after 2:11:45. Do those look like viable readings or is something messing with them? AFAIK VSI does not use pitot data as an input...

[Lonewolf_50]

Dozy, your comment on AoA fouling VSI implies to me that you think AoA had an impact on what was being read/sensed in the static ports. (But if that isn't what you were thinking, sorry about that). We have had considerable discussion on high AoA causing some problems with pitot tubes and the pressure they are sensing.

I'd need to look at the trace on a screen that doesn't have issues with images. Not all of the ones posted here come through on this machine/system.

carlos:
Don't think there was a spin. I wandered off the reservation on that score some months ago, before the data had been analyzed and some of it released.

The BEA reconstruction of the descent indicates a single "turn" (over three or four minutes) that changed heading of AF447 about 270 degrees to the right.

That doesn't look like a spin, but more of a stall with some waffling and rolling going on, mostly to the right.

[xcitation]

Originally Posted by Lonewolf_50

carlos: the reconstruction of the descent indicates a single "turn" (over three or four minutes) that changed heading of AF447 about 270 degrees to the right.

So, no, not a spin, but a stall with some waffling and rolling going on, mostly to the right.

+1
When I have rec flown in stormy weather in a light a/c I was forced down on a grass airfield. On landing I have been spun around 180 deg to wind once speed slows and wet. Very unsettling as a/c was still on runway. The a/c has lost force on its flying/control surfaces and then becomes a weather vane. With less force on AF447's flying and control surfaces the influence of wind/gusts is greater. Did AF447 right turn into the wind direction?

[ross_M]

Any lawsuits already been filed? I smell lots ahead. Plenty of blame to pass around and all defendants have deep pockets. Air France, Thales and Airbus must already be scampering for their lawyer teams.

[carlosgustavo]

You are probably right Wolf, 270 degrees turn is not much, but between
2.11.35z 2.12.15z they are stalling and turning right between 16 and 40 bank and the VSI gets its higher values, even unreliable. At that monent IDLE thrust is selected producing a pich down moment that stops the stall of the right wing stoping the rotation but mantaining the Stall.

Worse of all Wolf is that no investigation will solve this.

[jcjeant]

Hi,

Any lawsuits already been filed?

If you had return early from your travel to the Moon (the dark side of) ... or read the press you can know that two lawsuits are already filed in France
One against Airbus and one against Air France ....

Le Figaro - Flash Actu : Af447:familles veulent accs aux donnes

L'association Entraide et Solidarité AF447, qui représente les familles des victimes du crash du vol Air France Rio-Paris en juin 2009, tente d'obtenir des magistrats instructeurs l'accčs ŕ l'ensemble des données des enregistreurs de vol, a indiqué mercredi son avocat.

La juge Sylvie Zimmermann a cependant refusé en juillet cette requęte, au motif que l'enquęte technique du Bureau d'enquętes et d'analyses (BEA) n'était pas terminée et l'association a fait appel de cette décision, a déclaré Me Alain Jakubowicz.
En tant que partie civile, Entraide et Solidarité AF447 peut consulter les pičces du dossier de l'instruction menée par les juges Sylvie Zimmermann et Yann Daurelle, une enquęte dans laquelle Air France et Airbus sont mis en examen pour homicides involontaires.

"Le problčme est que les données ne figurent pas au dossier. Tout le monde en parle, notamment au travers du BEA, mais nous n'y avons pas accčs car la juge dit qu'elle ne les a pas", a déclaré Me Jakubowicz.
"Il y a cependant un déséquilibre puisque Air France y a accčs au travers de l'enquęte du BEA", a-t-il ajouté.

L'association, qui a plusieurs fois dénoncé la partialité des informations du BEA et leur orientation vers la faute de pilotage, a fait le 13 juillet une demande formelle pour avoir accčs ŕ l'ensemble des données. Mais Sylvie Zimmermann a rejeté cette requęte dans une ordonnance rendue le 18, selon Me Jakubowicz.
Entraide et Solidarité AF447 a fait appel, mais aucune date n'a pour l'heure été fixée pour ce débat qui se déroulera devant la chambre de l'instruction.

Réagissant au fait que le BEA a retiré de son dernier rapport une recommandation concernant les alarmes de décrochage, Robert Soulas, président de l'association, a jugé mercredi dans un communiqué que "ce triste épisode jette définitivement le discrédit sur l'investigation technique" et "génčre une crise de confiance sans précédent envers les autorités d'enquętes".
La catastrophe avait fait 228 victimes. Dans son dernier rapport d'étape, le BEA a notamment mis en cause la formation et les réactions de l'équipage aprčs le décrochage de l'avion.

Mutual Aid and Solidarity Association 447, which represents the families of the victims of the crash of Air France flight from Rio to Paris in June 2009, trying to get magistrates access to all the data flight recorder, said Wednesday's lawyer.

Judge Sylvie Zimmermann, however, refused this request in July on the grounds that the technical investigation of the Bureau of Investigation and Analysis (BEA) was not completed and the association has appealed the decision, said Mr. Alain Jakubowicz.
As a civil party, Solidarity and Mutual 447 may consult the documents of the investigation carried out by the judges Sylvie and Yann Zimmermann Daurelle, a survey in which Air France and Airbus are under investigation for manslaughter.

"The problem is that data are not included in the file. Everyone talks about it, particularly through the BEA, but we do not have access because the judge said she did not," said Mr. Jakubowicz .
"But there is an imbalance, since Air France has access through the investigation of BEA," he added.

The association, which has repeatedly denounced the BEA bias information and orientation error in piloting, July 13 made a formal request for access to all data. But Sylvie Zimmermann rejected that request in an order made on 18, according to Mr. Jakubowicz.
Mutual aid and solidarity called 447, but no date has so far been set for this debate to be held before the Board of Education.

Reacting to the fact that BEA has withdrawn its last report a recommendation on the stall warning, Robert Soulas, president of the association, said Wednesday in a statement that "this sad episode definitely throw discredit on the technical investigation" and "generates an unprecedented crisis of confidence towards the authorities investigated."
The disaster had caused 228 victims. In its latest progress report, including the BEA has challenged the training and the reactions of the crew after dropping out of the plane.

[Lonewolf_50]

The disaster had caused 228 victims. In its latest progress report, including the BEA has challenged the training and the reactions of the crew after dropping out of the plane.

Gotta love Le Figaro ... they start a rumor about pilots dropping out of the plane --, wait, what's this, skydiving at night in a thunderstorm over the open ocean? I know pilots are adrenaline junkies, but this is something else again! How the heck did they get back into the plane before it hit the water?

Perhaps google translate is to blame for this?

[SeenItAll]

To me the issue seems to be whether we want pilots to fly an aircraft by playing a computer game, or whether we want them to fly an aircraft according to the laws of physics.

It seems pretty clear that the AF pilots were doing the former rather than the latter. The basic difference between the two is that with the computer game, it is the computer's instructions and feedback that are your "law." The stall warning starts to sound, this means that the aircraft is approaching stall. Placate the computer by pulling nose up and the stall warning stops -- the computer responds by telling you that you are playing the game "correctly." Note that although these pilots were likely trained to understand that the computer game has other "rules," such as keep an appropriate speed and attitude, the computer states that the stall warning has priority over all these other rules -- so you don't address them until you've dealt with job #1, silencing the stall warning.

In contrast, flying the aircraft according to the laws of physics tells you that if you are in a severe nose-up attitude with less than 60 knots of airspeed at FL 370 (or even any two of these three conditions), you are in stall. These laws also tell you that in an A330 weighing over 400,000 lbs. with a maximum of 140,000 lbs. of thrust, you cannot climb out of a stall as you might with an F-22. While AB software engineers have done wonderous things with FBW, they have not repealed Sir Isaac Newton.

Now I don't know enough about A330 maneuverabilty to know whether once the stall was entered, it was recoverable, but shouldn't these guys have known that nose-up had no dynamic hope? Or were they just expecting that the FBW computer would find some deus ex machina way of extracting them from this dire situation?

In fairness, AB should certainly change its stall warning protocol to make more clear when it is inhibited because of unreliable airspeed. And I guess it may also be possible that the accelerations in the cockpit made it impossible for the pilots to gain any situational awareness. But the tapes seem to suggest a relatively nonviolent descent into the ocean. In the end, it all comes down to what you can rely on. These pilots relied totally on the computer, and were willing to suspend any belief in physical law.

[grity]

hi jcj risc a look at side 44 of the 3. report there are the values around 2:10:10 with much more details shown the purple line not the blue......

[jcjeant]

Hi,

Gotta love Le Figaro ... they start a rumor about pilots dropping out of the plane --, wait, what's this, skydiving at night in a thunderstorm over the open ocean? I know pilots are adrenaline junkies, but this is something else again! How the heck did they get back into the plane before it hit the water?

Perhaps google translate is to blame for this?

Drop out is to be understand as STALL .. I'm sure you known that
Must be certainly better jokes in the BEA offices ..
Don't forget that NTSB and a Brazilian party are also members (maybe as observers ?? but ..) of the investigation team .....
I wonder how all those people different by culture and language are communicating ...

[exeng]

BOAC

Out of interest, what happens in this 'computer game' if the one who has 'lost' control then presses the button for the requisite time?

Well BOAC the one who had 'lost control' now regains control - simples really - I have control -no I have control - no really I have control etc etc.

About as sensible as if no overide is pressed an input by the same amount in the same direction doubles the input, or if in the opposite direction negates the original input. Remember Hamburg.

I flew with one of our colleagues back in LGW days on the 737 who twice in two trips on a xwind landing put aileron in the wrong way. Frightening yes - but at least I could see what was happening and correct it before the inevitable - don't think that would have been possible on any Airbii.

Still the bus does have some good points like terrain escape which in my opinion is better than any Boeing, 777 included.

[DozyWannabe]

Originally Posted by exeng

About as sensible as if no overide is pressed an input by the same amount in the same direction doubles the input, or if in the opposite direction negates the original input. Remember Hamburg.

Aye, but if used and co-ordinated correctly you can get double the roll rate, which could be useful in an escape maneouvre, or if you see your opposite number doing something that doesn't look right you can negate their input before taking control.

I flew with one of our colleagues back in LGW days on the 737 who twice in two trips on a xwind landing put aileron in the wrong way. Frightening yes - but at least I could see what was happening and correct it before the inevitable - don't think that would have been possible on any Airbii.

Well, you could *see* it happening on a FBW Airbus (you have an ADI in front of you after all), that's never been the problem - the complaint has always been about *tactile* feedback. The thing is that tactile feedback used to serve multiple purposes when controls were directly connected to flight surfaces - you could feel how the surfaces were responding, you could feel what the other pilot was doing and assist if necessary (indeed you could also put more welly behind the maneouvre). But in todays modern world of hydraulically-controlled surfaces with no manual reversion you're down to being able to feel the other guys inputs, plus whatever the q-feel system is telling you. In a lot of cases this is nothing to be sneezed at, but in some cases it can give a false impression of having an effect on the flightpath when in fact you are not. UA232 was an exemplary display of CRM and aircraft control, but one of the interesting factors was the captain and co-pilot forcing their yokes forward and to the left, long after it was obvious that it was having absolutely no effect.

Airbus's philosophy was based on the idea that when going by the book, only one pilot should be manipulating the flight controls at any one time *except for extraneous circumstances* in which they designed the sidesticks to act in the same manner as the yoke, but by summing the inputs algebraically rather than by respective force. Backdrive was considered and ruled out because it added extra weight and complexity (complexity being the main issue), as well as causing extra problems in the event of systems failure or maintenance errors (e.g a cross-wired sidestick on one side).

Still the bus does have some good points like terrain escape which in my opinion is better than any Boeing, 777 included.

As always, there are usually positives and negatives to any differences in design philosophy, and to say that one is objectively better than another in all aspects is not only foolish, but tiresome after a while.

[mm43]

BEA Interim Report No.3 (English) Released.

English Version

[JJFFC]

@DozyWannebe

Hello DozyWannebe,




What do you mean by:

1. "this is the only correlation" at 02:12:35 ?
2. "correct behavior" at 02:12:40 and 02:12:48 ?

I understand that the stall alarm stoped after "stick forward" ?

Thank you,

[DozyWannabe]

@JJFFC

Simple. AF's argument is that the stall warning sounded at the same time as the pilots fed in nose-down inputs (i.e the correct way to recover from a stall) on every occasion. They are using this to argue that the stall warning gave a false impression to the pilots that they were doing the wrong thing by putting the nose down.

By matching up the trace graphics to the best of my ability it appears that this is not the case. On one occasion, the stall warning comes on at around the same time PF puts in a nose down input, but after then there appears to be no match whatsoever, and in my opinion appears to be in response to the pitch angle of the aircraft (which is stalling and out of control) as opposed to any sidestick input. In fact, if I've lined things up correctly, it appears in a couple of cases that the stall warning stops as nose-down is input, which would reinforce the correct behaviour by the terms that AF are putting forward.

[TioPablo]

A protection to me means a limit, something is prevented that could conceivably cause a serious problem

Dear Bear, sorry for my late answer... Was busy with other things...
What I meant to say all the time with the faulty switch of state (imo), is that the "reminder" given by Direct Law to use manual trim could have given a hint to the pilots about their attitude... And therfor: act as a protection...
If you don´t have any indication about your speed... What is the best action?
TO/GA?
Well, I don´t think so...
For me would be IDLE and trim-out to best gliding speed (AoA). Then pay attention to the blaring show...
If that switch of state (Direct Law thus), would have helped the pilots or not is at this stage irrelevant (or not). I think everybody were very busy during this years trying to solve this event and I´m almost certain we all will do their best to improve whatever went wrong. And as I said before.. I hope pilot unions will be there to see that safety recommendations (as issued by BEA), are implemented in a very near future.