OK465

I would add 2 things to the FD discussion...

1. They had most probably never seen a situation where the FD bars disappeared and then returned

2. As A33Zab says, the bars flash for a short period and then go steady (like the flashing lights on an ambulance can draw your attention before you hear the siren)

Most of the returns weren't long enough for the bars to go steady, but the return when pitch had finally been reduced to 5.6 degrees would have flashed AND then gone steady, possibly a falsely reassuring appearance of overall system validity...

Three more things...

1. At the apex, it would only take a couple of seconds of ill-advised following of what could have been a fairly subtle initial FD command to stall the aircraft, the subsequent initial aircraft reaction and physical flight indications not necessarily being a 'hit you over the head' indication of a valid accompanying SW audio.

2. The FD's should not have remained selected

3. We'll probably never know

Old Carthusian

Franzl
The evidence there is that the PF induced that particular nose down attitude 'There, I've taken it down a bit' to quote. The human element is still the major issue here. Bringing in the FD obfuscates the circumstances and qualifies as a red herring in my book. There is no getting away from the impression that this particular flight crew was thoroughly incompetent.

jcjeant

Hi,

Old Carthusian
This impression (if it is every bit true) leads to several questions:
Is that the system of training and selection of Air France and the french civil aviation formation should be questioned?
Is what the instructor pilots of Air France are qualified ?
Is the management of Air France has sufficient attention to these formations and selections ?
Are the aircraft manufacturer informations out dated for the conduct of flight ?
Is that the training and selection criteria are out of date and have not be adapted to contemporary needs by regulators ?
Will the BEA put in light those questions in his final report ?

CONF iture

Do not see evidence where there is none : The PNF is talking about the mach reduction in order to conform to the speed for turbulence, not the attitude.
A speed reduction induces a higher pitch, not lower.

Old Carthusian

CONF iture
It can be read both ways and this is what I wanted to bring out. Selective reading of the evidence is never a wise thing and the truth no matter how painful it is needs to be faced. We will not know exactly what happened but the information we have points to human elements and has done ever since the CVR was decoded. Everything else is just fluff
jcjeant
Those are very pertinent questions and I think they do hit at the core of this accident. I have always thought that the Air France culture was part of the issue and the background.

Lyman

OLD CARTHUSIAN:

No. And No. Retired F4 refers to the Nose Down at 2:10:00, when the aircraft was on autopilot. Hence the pilots had nothing to do with the Nose Down Pitch.

Are you now claiming the autopilot was incompetent??

If so, let's begin...

Quote (from BEA pp 88):

"...02:10:00 Pitch attitude decreases from 1.8° to 0° in 3 seconds.
Also visible in the FDR readout on BEA IR3 page 111..."


Response?

bubbers44

We always should consider the autopilot incompetent. Never trust it, just use it to reduce workload but never trust it. You, the pilot,are the only thing to trust presuming you can hand fly an aircraft when automation fails.

The old guys know this.

HazelNuts39

The AP just did the job assigned to it: maintain altitude while traversing an upward gust.
Image posted earlier:

Lyman

HN39

For refernce only, what headwind would have been necessary to trigger Overspeed, at the 2:10:02?

HazelNuts39

Lyman,

The steady headwind doesn't matter. At ISA+15.6°C a sudden head-on gust of 30 kt would have increased Mach from 0.815 at 02:10:02 to 0.866 (Overspeed warning threshold Mmo+0.006).

Please note that in two zoom-climb incidents overspeed warning was triggered without triggering the overspeed protection.

RetiredF4

OC, there is no evidence at all that this phrase at 02:09:54 is related to giving up height. It could be related to the change of radar scale or the reduction of speed (most probable) but not to the altitude.

You are entiteled to make the assumption concerning FD display being a fact or a nonevent, but it´s not correct to use incorrect statements with it and disqualify stated facts of other posters (in this case the attitude of the aircraft) as being a red herring.


Unfortunately most of us have to agree on that statement.

AlphaZuluRomeo

This seems to be a very reasonnable explanation indeed.

But, given the situation:

• 360ft too low
• pitch 0° when it should be ~2.5°
• V/S negative when it should be 0.

How is it logical/reasonnable to engage such a climb that you get:

• 3000ft+ too high ultimately
• pitch more than 10°
• V/S far too much positive

CONF iture

Sure, none is reasonnable.

In the meantime, it has to be mentioned that the initial action is suggested by the horizontal bar for vertical navigation. At this time the AP/FD vertical mode is still most probably in ALT CRZ. The negative values must have sent the horizontal bar pretty high on the PFD. What I don't get is why the system decides to dump the AP but thinks it smart to still display the FD for 3 more seconds.

In this short period, with warnings and turbulence, there is absolutely no chance for the crew to identify a case of UAS and apply its memory items.

Machinbird

I am not ready to declare the AF447 crew to be incompetent, only their performance as incompetent. I can understand being a bit brain dead at 2AM.
(Seems to happens to me at work at 2AM routinely nowadays)

A paragraph taken from the NTSB report on the AA landing incident in Jackson Hole seems to fit AF447 pretty well:
.........and lose track of what is happening with the aircraft.

Clandestino

Because it is simple: CM2 pulled the aeroplane into stall and kept her there. That's why it fell.

Because alternate laws are must-have fail-safe, mitigating the risk of untimely activation of overspeed and alpha protections! Computers are unintelligent and get more easily confused than humans, that's why they refrained from intervention when the aeroplane was stalled. Wet dream of many a PPRuNer asking for more direct control in Airbus turned out to be a hellish nightmare.

Yes, but instrument scan, understanding the aeroplane's position & behaviour and manual dexterity in manipulating the controls are not inseparably connected. True, you have to be very good to in all of them, all of the time to be a good enough instrument pilot but there's no difference between instrument scan when flying manually or when AP is on (except you can get away with more easily with lazy or inexistent scan when George is in charge) and procedures can be reviewed in one's head while serenely cruising, or standbying at home or whatever. As any technical discipline, flying can only be mastered in crawl-walk-run sequence. People struggling with basic aerodynamics will have no idea what is behind the memory items, will learn them by rote just to pass the checkride and stand good chance of misremembering or misapplying them at 4:00 AM, no matter how many time they repeat the mantra.

Being realist, I believe whether you, I, anyone else and his dog believe or not, it will not have a slightest effect on what has already happened. If anyone in cockpit recognized a stall warning for what it was and acted accordingly, we wouldn't be having this thread. Eighth.

Rote response to stall warning would have saved him and everyone on board.

Shock, horror, surprise, followed by panic and disorientation.

Yes. Rote application of wrong procedure did. Rote application of right procedure didn't, even if its application was by sheer chance.

No. Few examples of canary in coal mine were Pinnacle, Armavia, Gulf Air, Colgan, Birgenair, Aeroflot Nord, Alitalia 46 at mt Crezzo... it's just we disregarded them with a bit of applied jingoism: those killed were Russian/drunk/Arab/regional jet jockeys/Italian/turboprop drivers/ex-military/etc. Now when it happened to western built and operated widebody, amount of irrelevant and plain wrong theories put forward on PPRuNE to compensate for our inability to face the facts is amazing thing to behold.

It's a joke, right? Some slight subjectiveness can never be totally eliminated but pilots whose estimations stray too far from objective are bound to get hurt.

It is not wrong and it is not obligatory.

It will never stall an aeroplane as long as there is sufficient power available. With 5° pitch maintained, aeroplane climbs, power drops, AoA goes up until level-off at 5° alpha is achieved. Any aeroplane.

Accident reports indulge in neither. It is a statement claiming following the memory items is definitively not necessary for successfully dealing with UAS.

>sigh< Let me try it this way: was AF447 really the only case of UAS in cruise on A330/340 fleet so we should base all our judgments about Airbus FBW cockpit design solely on it or do we perchance have some other incidents to check how other crews behaved and see whether the UAS on the Bus is really bound to be lethal by virtue of sidestick design?

To find where the neutral is, let go of stick. Were springs broken? Extremely probably not; no mention of it on CVR and CM2 successfully reduced pitch from 12° nose up to 6° before stall warning went off second time and inane pull was repeated. So no traces loss of control in pitch. Aeroplane did as commanded.

That's very selective reasoning. Once aeroplane was back at FL350 indicated, there was no reason to pull anymore, yet he pulled and pulled and pulled, reaching an apogee of FL379, 2900 ft above cleared level. If there was a reason to keep pulling, it was not on altimeter anymore.

In real life there are: turbulence, updrafts and downdrafts. 0° pitch can be perfectly acceptable transient cruise pitch.

A33Zab

between 02:10:05 and 02:10:07 FD2! (RH PFD) was not engaged.

one second remain, before both FDs were not displayed @02:10:08.
for that second the Nz delta was ~ -.15g....(Nzdemand 1g - Nzactual 1.15g).
IMO FD bar position would have been ND and minor.

Quote:

AMM: 22-11-00

(1) Generation of FD bar commands

(a) Pitch FD bar command

The pitch FD bar command is computed by using the measured vertical acceleration (NZ) and the NZFD command (pitch outer loop).

OK465

.....except an aeroplane with a stall 'angle of attack' of less than 5 degrees.

PJ2

Clandestino;

Thank you for your responses to my comments.
In reverse order... First, we're arguing the same point but for different reasons. I have always argued that the initial memory items (regarding the 5deg pitch attitude above FL100) were not obligatory but you continue to misunderstand the point of all my posts on the topic. The drill was indeed viewed as obligatory as far as the BEA was concerned (in their press conference) and as far as many on this board were/are concerned.

Note - edited to clarify the notion of "obligatory" and the execution of this drill:
The drill is obligatory in the sense that there is an abnormality and there are memory items associated with the abnormality. In this case, the first memory items are bypassed because they don't apply, because the airplane is above circuit altitude and above MSA. The memory item is, "level off and troubleshoot". It is not obligatory to pitch up to 5deg. That has always been my argument.

The point of my early and ongoing interventions was to provide reason to re-consider the assumed-obligatory nature of the drill and (see above) to question or at least think about why anyone would pitch a transport up while in stable cruise. The FCTM and various Airbus documents, some of which I have posted links to, indicate that automatically pitching up is not the correct response.

For the longest time, no one here agreed with that view and kept reaffirming that the correct response in all circumstances was first, pitch up, then re-stabilize the airplane.

The fact that a 5deg pitch isn't as harmful as a 15deg pitch-up is beside the point: Why pitch-up at all when in cruise flight just because the pilot considers that there is "immediate risk to the safety of the flight"? Where is the "immediate risk"? In my view, there is apparently far greater risk in destabilizing the airplane in cruise flight than in keeping it level, for troubleshooting. Thirty-odd other crews seem to have agreed with this view.

It is in this manner that I consider the UAS drill "wrong"...perhaps too strong a word, but the drill is, clearly, poorly-written. While others may not think so, I think that that requires an examination.

The memory drill's first question is, "is the safety of the flight at risk?" That is a crew decision which directs their response one way or another.

That is an important decision and I submit that the question, "is the flight at immediate risk" is more subjective than a decision based upon, say, flight phase. I'm trying to consider a way of making the response more clear. I considered flight phase to be a natural way to do this and gave some thought to a re-designed drill as per a recent post.

You will agree will you not, that loss of airspeed indications during the flight phase AF447 was in, is not nearly as serious as losing airspeed information during the low-altitude takeoff-initial climb phase? The intent of the pitch attitudes stated in the memorized portion of the drill was to provide immediate, safe numbers for such failure in a flight phase where looking up the numbers isn't possible. In cruise, the airplane is already established and in a stable flight phase and the crew has time to respond differently, as per the last memorized item, "when above circuit altitude or MSA, level off and troubleshoot".

Regardless, the main point I have always made and which you continue to miss is, Why destabilize a transport aircraft in cruise flight when a better course of action is to keep the pitch and power settings which existed prior to the failure? I made the mistake of stating that a pitch up to 5deg would lead to essentially the same result as AF447 and I was wrong and have corrected and stated that view a number of times. Please, move beyond my original mistake and argue if you will, from the present point being made.

Lyman

In layman's terms, but no less accurate:

A decision to do nothing... is still a decision. That is the point being missed, imho.

Perhaps a bit existential, but... The stage is set for doing the wrong thing when after a career of doing nothing, it feels as though, "something must be done..."

Lack of familiarity through "hands off" can lead to danger when necessary to "hands on"....rote, memory, feel, all are crusty from disuse, and not readily available in an emergent situation..

Machinbird, step in anytime....

Flyinheavy

@PJ2

[QUOTE]The memory drill's first question is, "is the safety of the flight at risk?"
That is a crew decision which directs their response one way or another.[QUOTE]

I think, they never applied this procedure anyway. They lost A/P at 2:10:05 and PF says at 2:10:14 "we haven't got a good display" to complete the phrase at 2:10:18 "of speed", while PM says at 2:10:17 "We've lost the speed etc.."
without any mentioning of UAS procedure.
If the transcript is real it shows that there was a lot of confusion and in fact no action only reaction of the PF to whatever situation he thought they were in.

As to the "There I've taken it down a bit" comment of PM at 2:09:54 it obviously refers to >Copilot's ND scale changes from 80NM to 40NM< at 2:09:53, so clearly no reference to pitch or speed as some were posting here.