Xcitation
 

Yes and no.

Not easy, not by any means, but troubling that they did not appear to diagnose the stall to the very end (unless you interpret the CDB's "impossible" remark). There were plenty of hurdles, e.g. the 60kt alarm "cut off" and no AoA indication (and the PF and PNF were probably receiving very different information at times), but also plenty of clues - pitch/attitude, trim (THS position), buffet and the VSI plummeting downwards. And the audible stall alarm, intermittently.

What worries me more is whether, unless the stall had been prevented in the period after A/P disconnect and reversion to ALT law, this crew were trained to recover from such a stall by hand flying and even with FLT380 to play with? As I have said previously, I believe PNF was close to diagnosis. From what has so far been disclosed, I still wouldn't bet my mortgage (if I still had one) on their recovering and that poses a lot of questions for the industry.....

Re: QF72. I just blame the US Navy! I remember flying near "Harold Holt" on the run down from Kuala Lumpur to Perth in the early-80s and strange things happening to my avionics (that was a 747, not an AB).......

Page 45 again
 

For me the previous two graphs (thanks for your comments) were just 'stepping stones' to the graph I was working on: AF447 cL-alpha prior to and post-stall.

Attitude
 

Something like this stuck to a flat surface would tell you if the plane was pointing up/down and whether the wings were level or not:



There are bigger/more expensive versions too, with digital readouts and 0.1degree accuracy. No computer needed. How about 2 of these:

Really? As far as I can see, only the instruments related to pitot data could be considered unreliable, that is until the point that the AoA became so extreme that various anomalies that would have messed up any display reliant on air data, mechanical or digital.


Where do you think the data in a glass cockpit comes from? The glass cockpit instruments are *driven* by gyros and inertial sensors, and they were telling the truth the whole way down.

From post #1625:

To set the record straight, some "Bus pilots" have not made this statement.

How is it possible?
 

It seems to me that the PNF was giving instructions to the PF based on indications he had, and PF didn't.
It also seems to me that the PF had his own indications, but was not trusting them, they were too crazy, confusing.
It also seems to me that the Captain was puzzled by what was going on with both side indications.

On the PF's side one thing (or nothing sometimes), on the PNF maybe just the altimeter worked kinda properly.

So it makes sense to me that the PNF did not have all the information necessary to hold the plane (i.e. Attitude Indicator), so he kept telling PF what to do.
On the other hand, PF had other kind of information, which PNF didn't.

Imagine the Captain's situation, he can't believe in his eyes. What he is seeing is not possible.

What is he seeing?
We don't know, many parts of the CVR are missing (to us).

I am not suggesting that a "coverup" is going on, but lets suppose BEA has an "inclination" to suggest that the pilots could have saved the situation. If the conversation between the 3 pilots would show they were confused, why show just part of the conversation?
Why show only the "confused" part?

All three pilots didn't have the slightest idea of what was going on, and is very hard to believe they only lost airspeed and AP.

Maybe when we see the full CVR trancript...

All the talk of what is a stall etc. and how us pilots 'know' exactly what it is like to be in a stall....What we don't know is if we would be able to decide if we were in a stall or a dive should we be in a cockpit with no outside visibility or working instruments and all we knew was that we were descending at a rate where we could be almost in zero g.
That may well be the difference between our experiences of stalling and their experiences on the day - experiences that no simulator training would have been able to reproduce.

Well sure. Only native anglophones /[insert your own culture], preferably those raised near [insert your place], should be recruited to fly modern jets. Many languages and cultures are unable to provide the mandatory level of efficiency required. People tends to say words that, once translated into English /[ or whatever], doesn't make enough sense to me. Consequently, such talks should be banned from cockpits and considered unsuited for high-technological environements due to their intrinsic flaw at being fully understandable by me.

Such statement do not surprise me. I could say exactly the same about English/[insert your own langage] forms but I'm not so ethnocentric.

What cause such a distorsion about the exact meaning of those words is primary due to several factors [many that a non native, whatever his proficiency level, would never catch in a lifetime]:
1. We are not listening at what they said, it is not complete and we are reading it from a poorly made transcription. Add the fact that, for the majority of the readers here, you are reading it from a poorly made translation of this poorly made transcription.

2. French talk is quite different from French written forms. Without its prosodie - rythm, tone, emphasis, etc. - it may be clearly ambiguous from the direct transcription. Hence, without listening the original conversation in integrality, I'm a bit clueless about their talks referential. Some of the referential is also lost by not seeing them, because there is many "shortcuts" used for talking. In fact, the ammount of informations they had, when talking to each others, is not part of this transcription. I'm sure that accessing to the original record would clear many ambiguities.

3. I suppose that BEA, at this stage, released the minimal work on CVR voluntarily; only people which had heard the CVR would have a right idea of what happened in the cockpit. There was strictly no effort made to make it really meaningful in English version -it is called a litteral translation: emphasis is put on words, without bothering about sense.

4. Last but not least, making sense is not a question of langage, it is a question of discipline. What I'm able to infer from the first reading is a serious lack of crew discipline. No further comment is needed about it.

Takara, do you mean that flight deck crew indisipline was a factor in your interpretation of what you have read (in your native tounge I would hope?) It would not have been the first accident that ever happened with such an enviroment (the Korean 747 crash at Stansted springs to mind), and it wouldn't be the last..

I have to put my ten cents worth in here. Takata is right IMHO, about French. The same applies to all languages that I have encountered, including English. No way can anyone, who has not heard the CVR, can arrive at an accurate interpretation of what was said. Also people gesture with their hands, arms ,face and head, when talking. The "Continentals" being the greater users of their bodies when communicating.

takata,
That's what I tried to say, but a lot less clearly....

TurboTed . . . Spirit Levels
 

Good thought, but unfortunately these are also driven by "G" forces. They are not accurate in a turn, during acceleration, etc.

vanHorck / Takata
 

Other forums can deal with the Seven Years War of 1756-63 and the rise and hegemony of the English language (even as spoken by our American cousins "over the pond") as the "global" language(!), so let's focus on Takata's "discipline" good point (well made). AF had 3 hull losses in 9 years, each where there have been not insubstantial CRM issues to varying degrees. But I cannot see where the use of the French tongue per se (as opposed to the over colloquial use of any language) has ever been a contributing factor to the hull loss (as opposed to a nuisance to non-French speakers in air accident reports!). If either the PF or PNF on AF447 had diagnosed the problem, it would have been conveyed to the other and to the CDB. There is nothing in the CVR so far disclosed, any more than anything in their respective flying history and experience, to suggest other than F/O 37 should have been PIC in CDB's absence (as PNF in LHS). I am really struggling to get "my head around" the entire "relief" pilot regime that was in force. I can see the benefit of a CPT having the discretion to select a PIC, e.g. a longserving F/O unsuitable to ever become a CPT or a less experienced but still experienced "high flyer" of a F/O heading upwards, but I can see nothing to indicate why F/O 32 should not have been under F/O 37. The latter should not have been "nagging", as one post (I think accurately) described his instructions to the PF.

After my first reading (yes, it's my mother tongue), I was quite astonished that a supposed "professional" crew would behave like that in such a critical situation. Good professionnals may use sometime a "relaxed" way of communication, but if the situation would become critical, they should instinctively revert to their discipline of communication.

I'm also quite sure that the CVR transcript was expurgated from many slang forms which should have been added to several sentences heard. PF sounds like a teenager, not a 32 years old fully qualified A330 driver. Captain said almost nothing, or it was cut in report.

BEA also inserted numerous leads (flight related) that, once put together, are showing a very same pattern all along this flight. Going further into this chapter, at this point, would not help the BEA as full cooperation with the company is needed in order to conduct this investigation to its end.

@CONF iture:
 

Hi Conf,

This discrete represents the source of the vertical speed indication on PFD.

From the report: page 94 (English version)

and confirmed by FCOM.

That will raise another question, why was IR V/S not available?
was it related to:
02:13:14 - .1/FLR/FR0906010211 34123406IR2 1,EFCS1X,IR1,IR3,,,,ADIRU2 (1FP2),HARD
and this fault message related to disapperance of FPV?

Quote:

---

The presence of the “FLAG FPV ON PFD CAPT (F/O)” message indicates that TRK-FPA
(Flight Mode Annunciator) mode was selected by the crew during minute 2 h 11, but that the
FPV was unavailable (see interim report 2 for details on the conditions of availability). Based
on a study of the other relevant parameters it may be concluded that the FPV was selected between
2 h 11 min 48 and 2 h 11 min 54.

---

fluid level meters
 

Hi,

TurboTed
Think again :)
At 2min10

"c'est pas possible"
 

2 h 12 min 44:
CDB: (…) C’est pas possible

There has been some discussion concerning what the captain was referring to when he said this. As a further note on language (this is becoming quite a theme, here), please note that "c'est pas possible" could also quite easily be a generalized expression of frustration. For example, if you arrive at the movie theater only to be told that the movie is sold out, you may well exclaim "c'est pas possible!" It doesn't mean you think that it's actually impossible that the movie is sold out, or that you don't understand how it came to be sold out.

I'm not saying this is the case here. Without hearing how the words were said, it's impossible to conclude anything either way. I just wanted to raise this as a possible option before people get too far into speculating which instrument he might have been looking at.

FPV ECAMs fault are due to their selection and rejection (they were unselected before this point) because airspeed was < 60 kt.

It looks like CONF-iture doesn't read (or believe in) BEA reports; this fault was explained and very understandable considering that pressure altitude, barometric vertical speed and true airspeed were altogether affected by high alpha and unvalid dynamic pressure resulting. In fact, it was a triple IR rejection of ADRs - this fault was reported three times, one by each IR, but consolidated into one single ACARS:

Page 38, report #2.

ADIRU2 (1FP2) (2 h 11)
ATA: 341234
Source: IR2
Identifiers: *EFCS1, IR1, IR3
Class 1, HARD

This message was generated by IR 2. For an ADIRU of this standard, it means that the IR considered that the three ADRs were invalid, that is to say that at least one of the three parameters was invalid (SSM status not NO) amongst pressure altitude, barometric vertical speed and true airspeed. As soon as the third ADR is rejected, the IR generates a message pointing to its ADIRU. If one of the IRs considers the three ADRs as being invalid, this must also be the case for the other IRs. It is therefore logical that, in parallel with this ADIRU 2 message generated by IR 2, an ADIRU 1 message was generated by IR 1 and an ADIRU 3 message by IR 3, which would explain the presence of the latter amongst the identifiers.

The fact that EFCS1 was present amongst the identifiers preceded by an asterisk indicates that EFCS1 had at least generated one class 2 message, perhaps followed by a class 1 message. There are too few elements available to determine precisely what the presence of EFCS1 amongst the identifiers means. Nevertheless, it is possible to state that it concerns a rejection of ADR by at least two PRIMs. It has not been possible at this stage to understand why EFCS2, the clone of EFCS1, is not an identifier.

Spagiola
 

It's just the most tantalising limited-release CVR comment, at FLT200. The reality is that the BEA, even with access to the full FDR/CVR, may never know what the CDB meant by that remark (and the possibilities are endless to speculate over).:uhoh:

As the exact quote is:
CDB: (…) C’est pas possible
One knowing what was "(...)", added to his tone, would have a very good idea of what the captain really expressed. But, I'll refrain from proposing to change this expression into something meaningful... as some could complain to JT about my langage.
Just note that Spagiola remark is by far the best proposal with all clues on hands.

Vertical speed.
 

Hi Takata,

Well in fact this question was raised by me.......and I did read the full reports but maybe I did missed this one.

Ok, try to formulate it another way.......for what reason could Inertial Data be not available?
IR part of ADIRU can operate normal even if ADR is providing unreliable information.



http://i474.photobucket.com/albums/r...999/ADR_IR.jpg

Takana, thank you for your views in regards to you native understanding of the CVR Transcript. You have (as the English quotation puts it) hit the nail on the head. Regardless of faults in the design philosophy of the aircraft (and the BEA have hinted them in their recomendations as well as AD's that EASA have introduced since), a servicable aircraft seems to have been lost with all souls due to a massive failure of CRM IMHO. While I will not take the view that the flight crew were not totally professional (because I did not know them and cannot comment as such), I would wager most Captains in such a situation would be have been more self preserving that what the CVR transcript leads to have been the actions of Captain Dubois in not taking control, especially when the PF was unsure of his actions. He did have the time (and height) to do so. Air France may continue to defend their crews most vigiously and that is only to be expected and conmmended as an employer. However the accountants heading the company must realise that AF's standing in the world is only upheld if they are fully free to ensuring every facet that caused AF447 to smash into the ocean is investigated, independantly reviewed and actions taken to prevent such events ever occuring in this manner again, to anybody.

Takata
 

Hi Takata

"(...)" itself may never be known (as opposed to deliberately omitted, and even then it could easily be ambiguous).

There is a widespread misconception that CVRs are easy to decipher. They most definitely are not. The hardest day's work that I ever did (and certainly the most emotionally draining), was not landing my Buccaneer on a flight deck or landing a 747 on various runways where ideally wide bodied jets should not go anywhere near (thankfully fewer in number these days), but a day at the AAIB helping investigators to make sense of what a F/O that I knew had said shortly before his death on duty.

It's hard work for the BEA, even with input from close colleagues of the deceased flight crew.

The cockpit issues you refer to are troubling, border line the other extreme to the CRM Van Zanten/Meurs issues.

What reason, what reason... would you like a +/- 10,000 ft treshold exceeded for inertial vertical speed? It is always the same reason, a designed parameter exceedance.

A baro inertial loop is integrated in the inertial V/S computation to bring the stability in time of the air data (no drift like inertial data). Degraded data: When the inertial vertical speed is not available, the baro vertical speed is automatically displayed. Failure : In case of vertical speed failure, the scale is replaced by a vertical speed flag.

In case of excessive vertical speed, the digital and analog indications
become amber. Excessive V/S when :
- V/S > 6000 ft/mn or V/S < - 6000 ft/mn,
- V/S < -2000 ft/mn below 2500 ft Radio Altimeter,
- V/S < -1200 ft/mn below 1000 ft Radio Altimeter.

But It looks to be set at +/- 10,000 ft in order to show a degraded status.

Sure, but this one was understood and deliberately omitted... and I may understand why.:ooh:
See legend:
(…) Words or groups of words not relevant to the conduct of the flight
(*) Words or groups of words not understood

IMO, it was relevant for understanding what the captain expressed.

I do think I am right in saying that the captain words should be an exclamation indicating disbelief or surprise to see something beyond his understanding or completely abnormal or completely wrong

So in the (...) can be the words
Mais ....
Putain ....
Merde ...
Bon dieu

So .. full speculation .. so many open possibilities ..

Hi jcjeant,
Spagiola's explanation of expressing a sentiment of extreme frustration close to anger, was much better than yours. That's the reason why it is usually introduced by a volley of profanities; some of those words you quoted are sometime all used at once for good effect : n... de d... ! de p... de b... de m... ! c'est pas possible !

Disbelief is not the primary sentiment, as in fact, one will clearly acknowledge that the situation is real but very frustrating.

Hi,

When you don't understand something ... you can be frustrated a max .. certainly if you (or think) know many things ...
So .. c'est pas possible can be an exclamation caused by something beyond his understanding
Anyways all is open .. and again .. this must be not important in the cause of the accident
The most important words are "j'ai les commandes" and the actions immediately following.
If those actions can be explained by other thing than a "pilot error" ... maybe this accident will bring some (new) progress for the aviation safety

ADR Switching
 

ADR Switching:

From my experience, I've noticed the information isn't presented instantaneously but rather 'scrolls' to the particular value.

At least this is what I've noticed with normal working ADRs, when using the ADR switch on the 320 family.

Now for their situation, having the AIR DATA switch placed to 'CAPT on 3' after the captain entered the cockpit, would revert the PF's altimetry information back to ADR2 - which, if in the (highly unlikely but possible?) event that one or both the F/O Static Ports were blocked - I imagine would start 'scrolling' UP indicating a climb (to the original altitude at the time of blockage). This of course would be highly confusing, especially coupled with the lack of V/S information on the PF side, and would certainly hinder their progress in recovering the situation.

(Once again - we do not have the ADIRU 2 information as it is not recorded by the FDR. Unfortunately I think that information would be the key to understanding the PF's actions).

Anyone have more experience with this ADR switching 'scrolling' aspect?

confusion, leadership
 

Some people here critizise, that no leadership was taken by either PNF or CDB.
I would say CRM still worked because as we say: If you dont have any better idea, just keep quiet. PNF had some good ideas but he wasnt so sure because otherwise he would have insisted more or taken control of the aircraft (which he did from time to time, but without pressing the takeover P/B) So again, nobody really was in control because inputs add up or are neutralized.
Confusion was high but I am very sure CDB noticed the aircraft pitch attitude as he was walking into the cockpit because at that time (02:11:43) it varied between 10degrees up and even more and this is definitely different from what he experienced the other 100plus times when he entered the cockpit in normal flights. What he then saw when he entered the flight deck brought up the "c'est pas possible" because he (and nobody else of us) has never seen a combination of those readings before.
Old sentence but true:
In an emergency you will not rise to the occasion but fall back to your level of training. Were they trained for this situation?

@Neptunus Rex: ...inner ear pressure.
I would say if you are stressed like they were you dont feel anything

@takata: I have been trying to get info about VS indication on PFD(not data) and could not find any tests (AMM) were this system is checked. Maybe you have another source?

Are you kidding ?

Hi peefactor,

good point, but on the A330, the "new information" (in this case ADR3) is presented much more instantaneously than on the "rolling" A320

@Takata:
 

Thx, clear now.

The 'stable' inertial V/S was replaced by unreliable and fluctuating ADR V/S.

'Another slice and hole for the swiss cheese model'

Hmmm, where did you find this situation in the plots ????

Where was any indication the PF was trying to get the Nose down ??
In contrary the Nose descended a couple of times despite continued NU commands from the PF, albeit slightly reduced from full NU to half NU a couple of times.
In the last minute the PF even fought the ND commands of the PNF by himself applying Full NU.

Having a look at the traces I do not see anything there which would confirm your assertion that the PF was seriously trying to get the nose down.

takata posted the speed graph. It does not show something I expected. In climbing from 35000 to 37500 it should have lost just a whole lot more speed than the graph shows, which suggests the plane was in a rather significant updraft with most of its upward motion being provided by wind energy.

That ought to be food for thought here, too. It would likely be more conflicting or exaggerated data for the PF to digest.

I though the 0.5*(v1^2-v2^2)=gh equation had been looked at earlier and found to be broadly consistent ?

In the more general discussion, has consideration been given to maximum altitude limitations if turbulent conditions are expected and unreliable airspeed remains a credible fault event which has to be accounted for?

The BEA report indicates that the AoA margin between stable cruise and warner was only 1.5 degrees. Given the coarse nature of the manual inputs it would seem that loss of autopilot precision is at best a serious matter at those altitudes and related/concurrent loss of air speed data and protections cuts one safety rope, then throw in turbulence.....In those conditions it seems that excursions towards a stall are likley even if one could hope that transition to full stall was avoided.

3holelover types, " I'm presuming those of you asking it are not pilots, because I can't imagine a pilot who could ever not know if he's stalled, PDQ, after the fact."

Then the PF on AF447 was not a pilot?

Neptunus Rex says, "The reaction on their inner ears should have told them that they were in a rapid descent."

No. Their inner ears can only tell them they accelerated downwards at some point. Once they are in a constant speed drop their inner ears would not sense the drop. They ears might detect something when pressure changed. Ears are tilt and acceleration sensors. They also have pressure on their backs to tell them the nose is pitched up - about the same as the instrument in front of their face tells them the same thing.

Neptunus Rex also says, "Plus, the Captain should have noticed the aircraft deck angle as he walked, or rather climbed, to the cockpit."

That one's a gimme.

Now, just how much of the acceleration into the fall they'd sense is up for grabs if the plane is in violent turbulence.

(Has learned far more about inner ears than she ever intended to learn due to a malfunction in my left ear similar to what the astronaut Alan Shepard had.)

You seem to be up at an early hour JD-EE. At that rate of descent I would have thought the pressure change rate would have ears popping. then there would be the increase in temperature, however if all else failed there was always the altimeter.

Black Sheds, one might ask about what part of the plane being at fault for providing bad data in the period immediately after the aircraft went to ALT 2 mode. Airspeed was out. "Protections" were out. PF pulled stick up enough to cause a serious zoom climb that left the plane with an "apparent" airspeed of about 215kts. (Assumption here based on airspeed coming back briefly and the number reported being within reasonable bounds for the kinetic energy transferred into potential energy.)

What led the pilot to disregard stall warnings all the way down?

But most of all why did the pilot pull nose up on the plane right after (within a second or two) the AP and ATHR disengaged? He had accurate data up to that moment. And an average kid can extrapolate into the future.... The drill is "leave it alone, fly pitch and power until you have air speed back. Do not make any large sudden changes in either pitch (no nose up) or power (no TOGA)."