- while I fully agree that, as with the PGF A320, power reduction would have aided recovery, I think you are confusing pitch attitude with AoA? - it is AoA that stalls a wing, not pitch. Pitch attitude is not in reality relevant to 447 from 38000' down. Even when the pitch attitude fell below zero the wing, I think you will find, was still fully stalled and being kept stalled. It is extremely straightforward to stall an aeroplane at a very low power setting if enough nose-up force is applied - otherwise airtest stalls could never be done.

AF don't fly AoA, Pitch was not inconsistent with GA, and there was no sense of a/c orientation other than that. It seems to me, they were unaware of the AoA disaster. Captain returned walking up the aisle to the Cockpit door, and he makes no comment re: deck angle, so why do we assume it was an issue, to them, the only ones who mattered. And, if not, ample evidence exists to show how they lulled into "other than STALL".

Inertial attitude from FDR, as well as Inertial speeds and accelerations were not available to the flight crew. The VS confusions are an example. It is a difficulty to resist a simplistic opinion, I will admit. Vertical speed, airstream noise, and an unsussable deck angle (due comfy g) make an either/or seductive. Rapid nose down descent, or mushy NU descent?

CHOOSE WISELY?

For that matter, the record of annunciations in the FDR do not prove they were displayed in the cockpit.

At some point, the realization must be that there may have been failures of electrical nature and mechanical, and that the crew were not privy to the nature of the failings. To assume that all evidence is fair, and balanced, as presented, obviously allows for some conclusions that cannot be supported. As yet, anyway.

Lyman - I am forced to join the club and say I don't understand most of that:{

BOAC-
Point well taken. Indeed, despite the 25 degree reduction in pitch the wing was still stalled so you could say it was the pilot inputs (and resulting THS setting) that kept the aircraft stalled even without the thrust. Thank you for the explanation....

Just a question
 

In the graph on page 111 of the english version on Interim report no.3, just below the V/S 'zipper' trace, is a trace indicating the source of the V/S parameter as being alternately IR and ADR. Although the IR part of the ADIRU receives the baro V/S from the ADR part, I assume that here IR means inertial V/S and ADR means barometric V/S. My question is: Does this trace only indicate what source the DFDR/DFDAU is using for the recorded V/S, or does this have implications for other systems, such as the calculation of FPA/FPV (note FPA trace lower down on same chart)?

I got a bit ot in the other, so I'll bring this up here.

A question regarding V/S sel.

In autoflight, what is the purpose of a 5k fpm descent? The controls are sheltered by the FCM, so why a rate? Is a rate the way altitude is selected and input by the computer? Even in ALTERNATE LAW 2, later, the PITCH is doled out in bits, the RoA is acquired in pieces, not in DIRECT fashion? Is 5k an increment?

Also, re: AL2 PITCH. Not being DIRECT, we see the time it took to establish the climb.

So, After the STALL, is the PITCH still being incremental, vice "brisk"?

As in "Briskly NOSE DOWN, please." Or is the ship in DIRECT at this point, and if so, what are some control surface rates we can look for in a potential recovery?

Right, then. After a STALL, the computer input PITCH commands are lost, and the ship has DIRECT from SIDESTICK? Makes sense, one would want severe rate. To avoid, SECONDARY STALL?

I don't have the answer to the question, but as long as the VS takes its source on the IR, the recorded VS and FPA make sense. When the source is the ADR, both values seem to be erratic and unreliable.

http://i45.servimg.com/u/f45/11/75/17/84/af447_25.png


The BEA is not that affirmative :

CONFiture,

Thanks for correcting me. If I had been writing an official report, I would probably have used BEA's words.

CONFiture,
Sorry for being lazy.
What's the bottom purple trace in the graph in your latest post (the one degenerating into another 'zipper')?
Seems to say "Angle de Pente" (pitch flight path vector), but difficult to decipher with the lack of resolution.

Since CJ doesn't read me, so he won't have a cow, isn't the Pitch FPV oscillating twixt 0 degrees and 30 degrees NU?

I'm reading you, but electing not to engage directly for the foreseeable, however for general purposes the pitch scale is oscillating between the bottom of the graph (i.e. an unreadable value) and 0 from the time the AoA becomes unusable, and between unreadable and max during the last few seconds, which indicates that the graph becomes meaningless after the AoA becomes unusable.

Correct CJ, that's the Flight Path Angle.

HN39,
I thought it was important to mention that the BEA wrote that word, especially as they did not mention the particularity of the SEL VS trace and they did not publish the AP/FD VERT MODE trace.

Unreadable value? How's that, the x says -35 degrees at the bottom of the spikes. Matter of fact, it is at this point BEA claim the AoA at 35 degrees.

Lyman,
Minus 35 is the bottom of the graph, the bottom of the spikes are below that, not shown, hence their value is unreadable on the graph. The actual data points are shown as little dots (better seen in the original) when they are 'readable'.

Lyman: Small point but facts Sir!
 

Couple of pages back 10/14/11
The PF was NOT the 2nd copilot recently back from rest, but the 1st copilot and had been in the cockpit since take-off. This was discussed at some length earlier in the thread... While it may not be earth-shatteringly important, these small details resonate with me, at least, and lend a lack of credibility to much of what you post - if the FACTS cannot be kept straight, why on earth should your hypothesis be any more accurate? This is not the first fact that has been stated as such, only to really be shown to be incorrect. I appreciate the "thinking outside the box" mentality, but just for creditabilities sake, please keep the facts straight!

Memory is not great since the stroke, so I apologize, and since retirement, I have no editor. Use what you can, if you like, or not.

Wish they'd give you the CVR, I'd be interested in what your thoughts are.

The audio record is greatly important, and without its presence, I don't think anyone can have a complete understanding.

Do you think Captain might better have kept at the controls, and taken third rest?

Yes, the audio signature background noise on the CVR would be very significant. Having heard 100's of hours of cockpit recordings, I find unusual cue sounds are pretty easy to identify. Hopefully the BEA have someone clued in on this.

Personally I am surprised the Captain choose to take his rest when he did - knowing the ITCZ was coming up and the weather less than optimal. However I have no knowledge whether this was Air France normal practice nor do I know what other major carriers do.

This is absolute normal practice in every western airline because every senior FO/cruise relief pilot/cruise captain can fully substitute a captain in all his tasks. Otherwise he wouldn't be cleared by training or the captain couldn't go to rest. Captains most of the time take the middle rest because that's the best one and he is enough tired to have a good sleep and he has enough time to prepare for the arrival. Well relaxed captain means good crew performance.

Ice
 

Dear Experts,

Could ice-build-up, at tail and/or wings, have caused the climb to FL380 or was it all due to PF inputs?

If you read BEA report No. 3, you find that parameters of steering overlay flight parameters nearly completly. This means that steering inputs have lead (mainly) to the outcome of the flight, not influences from outside. They conclude that the aircraft did not stall because of turbulence, failure of systems (other than the pitot tubes) or any other factor, including ice on the controls.

Garage Years wrote:
To which Dani replied:
Well I flew for 35 years at a western airline and this was certainly not "absolute normal practice" because there was no "absolute normal practice". Every operation was unique and rest periods were organized by each crew on every given day (or night).

As for the ITCZ, my mob only got to penetrate it late in my career but after discussions with several pilots with extensive experience in it, my rule (and that of many peers), was to ALWAYS be in the seat during that transition. If we were going to tackle with the top of a "dry thunderstorm" I wanted to be in charge.

My opinion only.

That is exactly the meaning of normal practise: Most of them do, some don't. In even more organized airlines, it's the airline that puts rest times into the duty roster. Of course the captain is always allowed to deviate from this if he deems it necessary.

Let me assure that I do it the same way as you do it: I stay in the cockpit if it makes sense, after carefully considering the capabilities of the rest of my crew (i.e. my two FOs). When you fly e.g. from Australia to the Middle East, you have ITCZ almost on all your 14 hrs of flight. So you have to get your rest somewhere in tropical weather. It needs some kind of courage to trust in your crew, you have to sleep, otherwise you will be underperforming during the approach, but you have always your doubts, when you lay in your bunk and cannot sleep because of the Monsoon over the Indian Ocean - especially after we know about AF447.

What also helps is a proper crew briefing before leaving your seat: Discuss the items you think there are weak points among your FOs. See if their memory items are fresh enough, if they have the correct mind set. If they have read AF447 accident reports. But don't forget all the other risks that are somewhat been forgotten: diversion, emergency descent, fire and smoke.

To the contrary, there are some reasons to think that something was not right regarding the attitude in the PFD(s?)

• 30 sec after AP disconnection another source for the PF's attitude was selected.
• Again the selector was moved 90 sec later.
• The Captain called : The wings to flat horizon the standby horizon

No, F/O air data source was switched to ADR3, not attitude (IR).

Both were ...

http://i45.servimg.com/u/f45/11/75/17/84/af447_26.png

Yes, thank you CONF iture

I would not take that as an indication that the PFD(s?) was/were wrong. Indeed, if the CPT was seated in the 3rd seat, the standby horizon was the most "accessible" for him.

Switching to "#3" source fits in the logic to try to get rid of the initial ADR fault (speed indication).

There are many ways of interpreting that - one I quite liked was that he was pointing to all three ADIs, i.e. both PFDs and the standby.

"Horizon, Horizon, Level the wings - *Standby Horizon*"

"The wings to flat horizon the standby horizon" are the words.
Not "Horizon, Horizon, Level the wings - *Standby Horizon*"
Not the first time you change the words Dozy ... but if you 'quite like' it that way ...

AZR,
Actually, the stdby horizon is not as attractive than the big PFDs.
ADR switch for speed indication makes sense, but something else could be at stake when IR switch is performed ...

In the french original:
EDIT: Earlier, the PNF referred to the three (displays?):

@Max Jack:
Probably not. Ice tends to make an airfoil less aerodynamically efficient, and thus lose lift.

If we're going to be completely picky about it, the English translation reads:

and the French original :

So it is said three times (once in fact by the PNF trying to draw attention to something) and my ordering was off, my bad - but it is really such a big deal?

this sequence on the tape is completly irrelevant. The captain tries to show the PF that he is not wings level, which is very difficult in a stalled state (he did in fact a 180 turn in his fall from the skies). The captain might have thought that he was a bit lazy on the ailerons, but of course in real it was the pitch that needed the biggest attention.

The exchange immediately following that one is interesting, because both the PNF and Captain say that the PF is climbing, and yet while the VS trace has gone unreliable, it never indicates above 0 until after 2:14:00, and the altimeter is rapidly and consistently unwinding throughout.

So where do they get "climbing" from?

A simple answer will be "from the horizon indicator" .. provided they don't check the V/S and the altimeter (or they not rely more to those two instruments)

But they were (for the most part) pitched up, but descending. Is there something in the translation from the French I'm missing here? The ADI can tell you how you are pitched and banked, but it can't tell you if you're climbing or descending without reference to the altimeter or VS indicator.

Just tossing one more herring (probably a red one) into the bin.

How many pilots would ever have seen - in real lifre or in the sim - a V/S 'needle' (yes, I've seen what it looks like on the display) pegged at well over 10k f/min ? And accepted it as real ? And drawn the 'right' conclusions ?

Just my herring....

Yes, and they don't understand how that is possible, so either the ADI or the altimeter/variometer must be wrong. The ADI shows an attitude that they associate with 'climb' ("I'm in TOGA, am I not?"). O.T.O.H, if the airplane is descending at a terrible rate, then the speed must be 'terrific'. The possibility that the airplane might be stalled never enters their mind.

Aircraft deficiencies are already in the current reports and recommendations, including the pitots and the lack of AOA indicator. We may see more in future reports on the a/c, and I think more on systemic training and procedural deficiencies. Not just about the crew.

If it happens to you, and I sincerely hope it does not, then there is abundant evidence on here that it wasn't due to lack of knowledge (or desire for that knowledge) of your a/c. [Edit: and that isn't intended as a criticism of this crew, in case it reads that way]. And I would hope you would get a fair hearing based on the facts, and that we all learn and improve. The goal should not be to assign fault or blame, but to understand why. Sadly, there are plenty of places in the world much worse than France for getting that to happen, even when a bus is involved.

One other thing. Anything we build that's as complex as an a/c has flaws. All of them are flawed in some way, and always will be. Even if we might disagree from a design perspective on where those flaws are, I'd far far rather be flown by someone like you who is actually interested in researching and understanding where his machine is or may be flawed, versus a pilot who thinks it is perfect and looks no further than the minimum training. If you are stuck flying a plane you don't actually like (or even actively dislike), then I'm sorry for that (but no design will please everyone), on a routine day you'll be the less happy pilot, but if things go wrong you might well be the better pilot as a result.

Am I the only one here who thinks there is little or no point in an AoA indicator on the panel of an 'ordinary' subsonic airliner?

We've seen arguments for it on carrier-based aircraft, and yes, Concorde had one, but that was hardly an 'ordinary' airliner.

Does anybody here have any suggestions on how you would "fly" an AoA indicator in normal operations? And how you would include it in your scan?

To me, the suggestion of adding an 'unfamiliar, new' instrument (requiring additional training, too) because, once upon a time, people were incapable of inferring the state of their aircraft from the available data, makes no sense.

I hope I made myself clear.