AF 447 Thread No. 6 [Archive] - Page 3

airtren

26th Aug 2011, 23:11

Hello QwainGlyndwr,

What I had in mind is the 'g load factor' involved in the pitch control (elevators and auto-trim) that may be an impediment to a very quick, but very short (time wise) set of actions, that involve also Elevators and possibly (tbd) THS, which a successful Stall Recovery may require.

"rudderrudderrat" had a quite similar perspective:
Hi Henra,

I'm quite confident this aircraft would have dropped the Nose upon stall if this hadn't been so vigorously countered by the massive NU inputs and subsequent THS following.

In Direct Law - I agree.
In ALT LAW - when stick free, the FBW computers would maintain the same attitude and probably apply more nose up elevator and trim in an attempt to maintain 1g during the stall.

airtren

Without having a Normal Acceleration in the picture, is there any indication that in order to being able to quickly and effectively react and have full control of the control surfaces at Stall, would have been better achieved by being in Direct Law?

Again, I would have to defer to proper pilots, but from a strictly aerodynamic point of view I can't see any advantage in being in Direct Law other than the obvious fact that there would not have been any autotrim deflection of the THS - something that IMHO needs sorting anyway.
It depends a bit on where you consider 'Stall' to be. By the time the stall was fully developed the actual details of how the system would respond to control inputs are far less important than the fact that there were no control inputs. If you are considering the actual entry to stall in the first place then the better pitch damping given by Alternate might have been a factor for good.

I don't think much of the harmonisation of roll and pitch given by Alt2, but I suspect that a 'double whammy' of direct law in both axes might have led to even worse results. But I trespass into pilot territoryhttp://images.ibsrv.net/ibsrv/res/src:www.pprune.org/get/images/smilies/wink2.gif

HarryMann

26th Aug 2011, 23:19

Taking this one stage further from Owain's analysis etc. isn't it now likely that Airbus (on behalf of BEA?) have simulated the 330's recovery from such a fully developed stall attitudes and trajectories.

If the speculation that Airbus really intends to test-fly such a profile at some point holds water,.then they must be very sure that it's doable. A tail chute might be a consideration, as some of the early T-tailled test aircraft had to start using here.

I would think a fairly good dynamic model of stall recovery has been attempted and the likely recovery profile, loads and altitudes known - which will hopefully feature in a report one day.

DozyWannabe

26th Aug 2011, 23:33

@airtren - Speaking purely theoretically and from the technical evidence I have, if the manual trim wheel is moved then there isn't a damn thing any protection can do to stop it, and if you hold it in position, there's nothing it can do to change it either. It's way outside normal procedure, but let's face it - this flight profile was nothing if not way outside anything that went into the book.

airtren

26th Aug 2011, 23:41

jcjeant,

Someone mentioned at one point the Challenger Shuttle accident happening in front of so many watching it, and the O-rings problem that was known for quite some time.

This was mentioned before perhaps....There are many components in the lessons of AF 447, but at least 3 of them are related to flying/pilots: one is flying with lack of speed indications, second is avoidance of High Altitude Stall, and the third is High Altitude Stall Recovery (in night conditions).

Like after the Challenger accident, there are many signs that this may be somewhat of a turning point in training and practicing, as I am sure talent is there, or was always there.

A new generation of Flight Simulators for training is in order, better curriculum, including a wider span of theoretical physics and basic principle of flying, better documentation, better procedures... and the list can get very long.....

These are just a few...
Happy to read that.
But is not this exactly what is expected of professional pilots?
Or professional pilots position is now .. only be able to pilot planes when all is smooth but not when problem arise ?

Lyman

26th Aug 2011, 23:47

Tried to make the point in #494, and quite frequently before; I see you bring it up above (#499): Relative g demand Law. If in DIRECT LAW, wouldn't the PITCH have been non reliant on this "slowpoke", and the PF would have STALLED far sooner than the AL2 "normed" ascent allowed?

Not a BAD thing, and perhaps with a higher level of energy, the NOSE drop and Buffet may have contributed to a more straightforward STALL ID and recovery?

airtren

27th Aug 2011, 00:03

Dozywannabe,

It's in the realm of talent, training and practicing.

I am aware of two Airbus 310 day light Successful Stall Recoveries in Europe - Interflug and Tarom, both A310 - in which pilots, in spite of the mistakes that created the Stall situations, were able to prove their airmanship, and mastering of their profession and tools in the cockpit, including the THS wheel, by recovering from what would have been unrecoverable situations, and landed safely, with no injuries, or damage to the airplanes.

It was not some corporate procedure document they've read at Stall that saved them.

From what I understand, those pilots, were from a generation in which training was not different than that of the military pilots.

So they had it in their reflexes, didn't have to think much, what to do right away in a stall.

From what I've read, at 4100ft, and 30kt AoA NU at 60 degrees, - yes forty one hundred feet - from where the Tarom A310 recovered, there was not much time. The Captain and FO leveled the plane at 800ft, from nose dive to regain speed. The captain used the THS Wheel, like the Stick, with quick, soft, short (time-wise) commands....

@airtren - Speaking purely theoretically and from the technical evidence I have, if the manual trim wheel is moved then there isn't a damn thing any protection can do to stop it, and if you hold it in position, there's nothing it can do to change it either. It's way outside normal procedure, but let's face it - this flight profile was nothing if not way outside anything that went into the book.

ChrisJ800

27th Aug 2011, 00:30

If loss of airspeed was the only underlying fault in AF447, can anyone explain why the Autopilot needs to disengage? Isnt the AP taking inputs from pitch, roll and yaw only? I can understand that the autothrottle needs to disengage as obviously that needs a reliable airspeed input. But Ive flown plenty of less sophisticated types where the AP will happily keep functioning despite airspeed and many other system failures. This would allow you to focus on the problem whilst AP aviates and you manually set throttle to an appropriate N1 setting.

DozyWannabe

27th Aug 2011, 01:23

@ChrisJ800 - depends on the type of autopilot. Older types would be limted in the amount of roll, pitch and yaw they could command (and as such should theoretically not induce LOC as long as you keep an eye on them). Modern types are "Full authority", and can theoretically command anything.

For an example of what pitot failure can do to an autopilot, check the Birgenair 757 case. During climb, the blocked pitot indicated overspeed and the A/P commanded nose-up to the A/P's limit of control - the overspeed warning continued, so the Captain reduced thrust manually and almost immediately stalled and never recovered.

bubbers44

27th Aug 2011, 02:09

Welsh Wingman

27th Aug 2011, 02:16

Following link re: Interflug/Tarom (not sure the latter Captain was ex-military, although he was certainly ex-BAC 111 with all that entails re: stall alert....)

Video Wanted - Interflug A310 Stall Incident - JetPhotos.Net Forums - The Friendly Way to Fly (http://forums.jetphotos.net/showthread.php?t=49509)

Useful pre-AF447 Discovery Channel "automation" documentary, now with an added resonance.

Interview with ex-Lufthansa safety pilot, which draws the battle lines.

Chill Lyman, chill......!

DozyWannabe

27th Aug 2011, 02:40

Useful pre-AF447 Discovery Channel "automation" documentary, now with an added resonance.

That's a Discovery Channel screening, but I had most of the series (including that episode) on VHS from its original Channel 4 broadcast in 1996. (17 years old, the guitarist in an indie band and still a closet nerd... :) )

Interview with ex-Lufthansa safety pilot, which draws the battle lines.

Yeah, Heino Caesar. The problem was that he drew the battle lines without fully understanding what it was he was fighting. His interview in the accompanying book goes into more detail, and it seems most of his information came from inaccurate contemporary news articles and taking Bernard Ziegler at face value. As such a lot of the misconceptions about the Airbus systems that percolated into the piloting community come from him.

Welsh Wingman

27th Aug 2011, 03:41

And your opponents (of which I am not one in the wider context) will argue he was spot on.....

To the extent that I have an issue with "Airbus", if I am to be dragged in to such a debate, it is limited to the "leap" from Concorde/A300 to the A320. More than incremental, and it should have been limited (at that stage) to the FMS move forward and not concurrently removed artificial feedback from the cockpit.

I cannot stress enough, as a pilot to a software engineer, the Pandora's box that was opened up in relation to line management perceptions on flight operations and training and which is at the very heart of what went spectacularly wrong with AF447. Airmanship plus - forget it.

VGCM66

27th Aug 2011, 06:53

jcjeant

Personally I do not think the AF447 accident has any relation (or could be explained) by trigonometry .. geometry or other formulas .. academic or by the curvature of THS
The accident was related to the actions of the pilot(s)
This aircraft was built according to specifications and apparently he has met these requirements since it was authorized to make commercial flights.
And since the AF447 accident he continue to perform commercial flights
This aircraft is equipped with up to date flight and navigation aids
So .. it indicates to pilots (in the case of 447) that there may stall ... or if this warning is not followed by appropriate reactions of the pilots .. he will warn that the aircraft is experiencing stall
The pilot did not have appropriate responses to deal with the event and did not have appropriate responses to correct his first mistake
The aircraft systems were all put in place to prevent the pilot from his mistake
The stall alarm has been running for 54 seconds continuously
The artificial horizon presented a large blue area
The altimeter showed a rapid loss of altitude
The origin of the event is the icing three pitot probes at the same time (same manufacturer .. same type .. no luck for redundancy)
There was also 3 pilots icing (brain icing) ....
The three pilots, however, were of different types .. but of the same brand (AF)
Even there .. redundancy has not worked ...

And that was that. I agree. The rest is just a wasteful and pointless exercise. But do not mind me. Knock yourselves out.

I still recommend to watch Mayday (or whatever the name is in your country) Season 11 Episode 2.

Mayday | Episode List (http://www.discoverychannel.ca/episodeList.aspx?sid=12966)

Note: The final report was written in Spanish.

rudderrudderrat

27th Aug 2011, 07:00

Hi Aitren,
So they had it in their reflexes, didn't have to think much, what to do right away in a stall.
A300 / A310 is conventional controls like a Boeing. It's like Direct Law.
The pilot's did not have to go through the "W.t.f. is it doing now?"
It is simply - "what you feel on the elevator is what you get."

Hi bubbers44,
Letting the pilots fly it manually to keep it in proper speed range was their solution.
In Alt Law, the pilots are flying it manually in Roll, but the FBW computers are still flying it in pitch (confusing?). It is exactly like having the autopilot engaged in a pitch mode in CWS, but not having the roll channel engaged.

Hi jcjeant,
The accident was related to the actions of the pilot(s)
I agree.
The only question is why 3 qualified pilots could not recognise that they were in a stalled condition.
Edit.
The stall alarm has been running for 54 seconds continuously
I know of one event where "Dual Input" sounded for over 90 seconds, and neither pilot "heard" it.
When overloaded, the brain processes auditory senses last or not at all.

jcjeant

27th Aug 2011, 08:10

Hi,

I agree.
The only question is why 3 qualified pilots could not recognise that they were in a stalled condition. The first question is :
What is a qualified pilot ?
On which basis (in AF) they are named "qualified" ?
On which basis are they recruited ?
What are the level of qualification and how is the recruitment in other companies ? different .. more .. less ?
What is the following of their qualification .. rate of refreshing courses ... etc ...
What is the security policy of the company .. the general atmosphere on such safety within the company ..
I want to remind the Colin report about Air France

Owain Glyndwr

27th Aug 2011, 08:34

So, the Elevator to THS surface ratio is 3:7, which is 1: 2.33. ....

No, the THS area, by definition, includes the elevator area, so the ratio is 3.33:1

But as the THS moves itself, the Elevators move/angle is relative to the THS, and absolute to the a/c body, and thus "complex" - it has two components: its own, and that of the THS.

At this quick writing I am not sure if I am clear.

As the pitching momentum depends on the angle, the complex angle has a complex effect on the pitching momentum.

See my reply to Chris Scott

Imagine cruise AoA, and THS and Elevators at corresponding position.
What would Manual Trim NU of THS do?

WTH do you think it would do?? The THS would follow the manual input!

It considers only:
max NU, with both THS and Elevators at max NU, and
max ND with both THS and Elevators at max ND,
and leaves out the other two combinations, which are part of the set of possible positions see my original post bellow:

You have the numbers and you can do the arithmetic just as well as I.

question:
are there any other forces (with their respective arms) involved? at that AoA?

Well I did inadvertently leave out the basic wing/body moments. The wing would be giving a ND pitch and the body a variable NU pitch that doesn't get very large until the AOA is around 30 deg. Overall a ND pitch.

JD-EE

27th Aug 2011, 09:13

HazelNuts39,

I am rather wondering what the performance of BUSS might be in "unusual attitudes". It is intuitively obvious to me that if the plane can be expected to stay flying "normally" on pitch and power one could work backwards from altitude, pitch, and power to get an airspeed indication when the plane is more or less in a steady state. At reduced accuracy it might even be able to make good guesses while altitude is varying. What's not obvious is what could have been derived from the plane's movements and power settings once the upset had begun. At that time the velocity vector becomes more interesting than mere speed.

BUSS would be good for getting through an upset period if you don't do anything stupid.

JD-EE

27th Aug 2011, 09:19

rudderrudderrat, "Pilots can only sense accelerations - not speed. The sensation of high speed was due to the unusual air noise around the cockpit sounding like they were going very fast."

Indeed, and they're not as good at it as they think they are. He was leaning back. That felt like acceleration to him. The comment made makes it seem like he was pulling back on the stick to get the nose up to slow down and it wasn't happening. The acceleration was continuing. His brain froze solid.

A full motion simulator can show you this effect happening. And if the training program includes a module to explain how such simulators work maybe then the idea that you cannot trust even the pressure on your back or seat when flying with no visibility at all outside the plane. That brain freeze prevented him from recognizing the stall. (Darnedifiknow what led the PNF to be so meek about breaking through PF's fog. It cost him his life, whatever the reason.)

JD-EE

27th Aug 2011, 09:23

The one question I have on this then, is about the elevator which clearly had aerodynamic authority all the way down. If, after the stall was fully-developed and even with the THS at -13.5deg, if the SS had been placed in the full ND position and held there, depending upon when this was done, (earlier the better of course!), and given the ND pitching moment afforded by the THS, would such elevator position be sufficient to eventually get the nose down or would it partially/fully stall given the already-positive AoA of the THS, and lose all effectiveness thereby?

Now ask the important question. Could the THS deliver ENOUGH lift at any setting to eventually produce an ND attitude? It's not sufficient that the THS and elevator produce lift. It must be sufficient lift to materially affect the orientation of the plane.

As a thought experiment a stiff sheet of cardboard held outside the car window tilted at 60 degrees produces some lift. But it's very small compared to the lift when it is nearly horizontal. And it depends highly on the velocity of the car. Memory if playing with such stuff when I was MUCH younger suggests the curve involved is not a simple lift proportional to some constant times the cosine of the angle relative to horizontal. The upward force needed is, however, proportional to the cosine of the angle of the plane to horizontal.

Can the THS/Elevator combo win this race? I don't have the data to know one way or another. I am sure simply generating lift "ain't enough."

rudderrudderrat

27th Aug 2011, 09:31

Hi JD-EE,
He was leaning back. That felt like acceleration to him.
I disagree.
He would know that with a nose up attitude around 15 degs that the sensation on his back felt normal. Most of the time he held around 15 degs pitch - they didn't exceed 20 degs of pitch at any time. If he was confused about acceleration and pitch attitude he would have pitched up even higher.

Pilots recognise the difference between horizontal accelerations felt on take off and the sensation at some nose up attitude - by using the PFD.
This crew simply used the wrong pitch attitude.

I think he's used the pitch attitude he's associated with TOGA power. OK at low level - but completely inappropriate at cruise FLs and they never recognised the fact they were stalled.

JD-EE

27th Aug 2011, 09:50

jcjeant, you're annoying. You're starting to ask my kind of questions.

(grinning ducking and running ---->>>>>> that way!)

JD-EE

27th Aug 2011, 09:52

Really ---

I wrote and posted 520 BEFORE I wrote and posted 519. Something glitched!

Spooky!

JD-EE

27th Aug 2011, 09:53

Chris, I think the autopilot engineers didn't want the AP flying the airplane if there were two competent pilots to take over if the inputs to the computer were lost. The AP if programmed to keep flying without airspeed input would have held 3 degrees nose up until altitude hold required a different attitude, and the last N1. The AT would eventually have no way to know if it needed to change with no AS. I think that is why they wouldn't let it keep flying level. It could get too fast or too slow with no AT help. Letting the pilots fly it manually to keep it in proper speed range was their solution.

What do the pilots know that the computers don't already know? In fact everything the pilots see is filtered by computers. The AF447 pilots were just as blind and bereft of real data as the pilots. So there really is a valid question here. Why DID the pilots have to take over?

The only valid excuse that comes to mind is that the designers did not know this particular failure mode existed. Until this crash the doctrine was that such errors could not happen.

Now, the computer actually has more information than the pilots saw, it appears. At the very least it has AoA. It also has GPS, raw and processed and the inertial system that it can use to work with. It HAS altitude. It never lost it. So it can monitor its pitch and thrust and fiddle both to maintain altitude. It might be a "lumpy" process. But it would keep the plane flying. The pilots may have had the GPS/INS data displayed somewhere. It appears it was as little in their scan pattern as the trim wheels. So it might as well not have been there if it was. The computer would scan these things if the software design had it doing so.

JD-EE

27th Aug 2011, 10:08

rudderrudderrat - that's the theory. His actions and words suggest he plain forgot, fell back on seat of the pants flying, and died.

02:11:40 "I have the impression (we have) the speed."

How did he get that impression?

rudderrudderrat

27th Aug 2011, 10:15

Hi JD-EE,
Why DID the pilots have to take over?
Good question.

If the Autopilot could have remained engaged controlling both pitch and roll - but with a Health Warning such as "ATT", then it would have freed up more brain cells.

What is the point of playing "Keep the Wings Level" with roll direct when faced with UAS?

fell back on seat of the pants flying, and died. ...
02:11:40 "I have the impression (we have) the speed."
How did he get that impression?
I would guess from the unusually loud air noise. If they thought they were accelerating, they would have closed the Thrust Levers.

Lyman

27th Aug 2011, 12:25

JD-EE, rudderrat. He wanted evidence for overspeed, he found it, and acted accordingly? Brain Lock, dangerous, lethal, at times.

Sometimes, and I have been guilty of it here at times, The human Brain makes up its mind, then looks for evidence in support of conclusion.

Doubt everything, especially one's own opinions. Advice to self.

"I have the Impression...." Linguistically, that is a statement in want of a supporting opinion. If supported by either of the other two pilots, he is affirmed. Then the group can act. PF is too fond of the plebiscite, the election. "Wrong, My Airplane....."

Like the AD's Three, a vote, then?

Chu Chu

27th Aug 2011, 13:55

JD-EE,

You're the expert on this, but trying to program the computer to recognize every possible failure mode (and to avoid confusing them with each other or the failure modes you inevitably failed to anticipate) seems like a much taller order than just progressively shutting down the automation when it becomes apparent something's wrong. Maybe something you could do if starting from scratch today, but not with the technology the A-330 was designed around?

PJ2

27th Aug 2011, 14:36

I agree with PJ2 and others that going to 5 degrees pitch is not what one would expect a seasoned pilot to do, and is not the right thing to do in the circumstances. Depending on how quickly the airplane is rotated to 5 degrees pitch, the AoA would probably temporarily exceed the stall warning threshold of about 4 degrees, and in any case the 'stabilized' AoA in still air would move closer to the stall than with 2.5 degrees pitch. On the other hand I believe, based on 'gut feeling' rather than a numerical analysis, that if CLB power had been set and attitude had increased to but not exceeded 5 degrees pitch, that the airplane would not have stalled.Yes, I agree with that - a nominal 5 degrees pitch, held fast and gently achieved with tiny stick movements, and thrust levers in the CLB detent would not stall the aircraft.

PJ2

27th Aug 2011, 15:31

I would not disagree as the altitude margin was confortable this time, but nevertheless, I see AF447 as a compulsory trigger to modify the UAS procedure to something that makes sense in any circumstances. One way of perhaps making a change which retains the essential and necessary actions in the original drill is to modify the qualifying condition at the start of the memorized drill, which is, "If the safe conduct of the flight impacted". There is little guidance on how to make this decision in a hurry. In some explanations, the "safety of the flight is impacted" if all three airspeed indications fail". While such an assessment is probably a wise and accurate one, I disagree that the automatic response to this qualifier in the drill should be a pitch-up to 5deg.

The drill was originally conceived to deal with pitot or static system failures close to the ground, the takeoff phase being the obvious and most critical phase. The memorized items, as I have argued, cater to the takeoff phase, and, I have argued, the "Above FL100" caters to the climb, perhaps even out of high altitude airports, but not to cruise levels where the aircraft is already in stable, level flight. I submit that it is a mistake of interpretation and lack of clarity in the drill, to execute the 5deg pitch-up using this qualifier. This qualifier is followed by three items and the statement, "When at, or above MSA or circuit altitude, level off for troubleshooting". Well, one is already well above MSA and is already levelled off. Why divert from that qualifying condition, only to return to it "quickly", as the "How to do this drill" notes require?

If the memorized items are to be correctly placed and the subjective, very individual assessment as to whether the "safety of the flight is impacted" avoided if the memory actions are qualified not by "the safety of the flight" but by flight phase and by altitude. Then the drill is much more clear as to what actions to take. Qualifying by flight phase and altitude are the first memory items after the "safety of flight" qualifier anyway.

So it might look like this:

DRAFT - DO NOT USE - FOR DISCUSSION ONLY:
http://www.smugmug.com/photos/i-58PX24p/1/M/i-58PX24p-M.jpg

As Clandestino has pointed out, these drills and checklists have been created carefully by test pilots and others who know this work thoroughly. The suggestion for change will have implications not imagined or considered and clearly needs examination by many others. But the basic notion is removing the initial qualifier to define more precisely, where different behaviours, (crew responses) will occur. I think that confusion and the training resulting from this confusion is at the heart of the initial pitch-up - what cannot be explained in this view is the immediacy of the response and the absence of SOPs and CRM.

Changing drills and checklists occurs all the time as new knowledge and experience emerge. This is normally done through meetings between checkpilots, standards pilots, flight safety/flight data people and so on and is otherwise a process with heavy vetting of such changes.

In fact, the UAS memory items and ADR Checklist have perhaps half a dozen iterations from around 2001 on - the first drill did not mention "levelling off" at all and was a straight response to the Birgenair and Aeroperu accidents. However, the B727 pitot-icing event pointed out by a poster recently, was not considered, nor, I believe, were pitot failures at cruise altitude until after 2002 or so even though such failures had been occurring since 1996.

There are other serious matters at hand which this accident has made clear. But the accident sequence began with the instant pitch-up, with no SOPs followed, and it is important to establish why this occurred and to take measures to examine and change those reasons.

HazelNuts39

27th Aug 2011, 15:45

Now, the computer actually has more information than the pilots saw, it appears. At the very least it has AoA. It also has GPS, raw and processed and the inertial system that it can use to work with. It HAS altitude. It never lost it. So it can monitor its pitch and thrust and fiddle both to maintain altitude. It might be a "lumpy" process. But it would keep the plane flying. Which computer? The AP and A/THR disconnected, and handed 'control' to the FCPC, which reconfigured to Alternate law. The purpose of 'pitch and power' is to "keep the plane flying", not to maintain altitude or airspeed (in UAS and in turbulence). The FCPC does just that - maintain pitch and power. Why can't it maintain bank angle?

Ian W

27th Aug 2011, 16:04

JcJeant
The first question is :
What is a qualified pilot ?
On which basis (in AF) they are named "qualified" ?
On which basis are they recruited ?
What are the level of qualification and how is the recruitment in other companies ? different .. more .. less ?
What is the following of their qualification .. rate of refreshing courses ... etc ...
What is the security policy of the company .. the general atmosphere on such safety within the company ..
I want to remind the Colin report about Air France

There is another question - to what extent do simulator exercises provide sufficiently real training for emergencies?

There is a considerable psychological impact that the human factors people seem to have ignored. That is that transferring computer game simulator learning into handling 'real world' emergencies does not work well with all people. In the 'old days' before clever simulators all such training was in the air and if a trainee repeatedly 'lost it' under pressure it became immediately apparent. This weeding out will not happen with sim rides where trainees learn to pass the sim ride.

It could be said that the only reason that pilots are required is for when the automatics fail.

Several posts here have statements like: "The simulator does not perform like the real aircraft after a stall" or "the simulator cannot behave as the real aircraft does outside the normal flight envelope" and even "the simulator does not fly like the real aircraft in Alternate Law"

So the simulator is being used to train pilots in how to handle emergencies, upsets and LOC, despite actually not behaving like the real aircraft in emergencies, upsets and LOC? :ugh:

The current approach of only using simulators for training pilots to fly simulators does not seem extremely logical. I am sure that the accountants love it.

BOAC

27th Aug 2011, 16:08

Zorin_75

27th Aug 2011, 16:20

PJ2

27th Aug 2011, 20:08

Thanks BOAC. Yes, I knew there were "holes" and perhaps that's one. The intent is to examine the drill, for all reasons given.

PJ2

27th Aug 2011, 21:29

Zorin_75;

Yes, my thought was to keep it as close as possible to the original. Thinking about it, BOAC's critique applies equally to the original drill.

gums

27th Aug 2011, 21:58

I am disturbed by some here that believe you can handle an emergency situation or an "upset" ( love that term, and cracks me up and guess it really means "WTF?, over") by following rote, memorized procedures for more than a few seconds.

I am a "professional pilot" or would not have joined the fray here. Granted, folks like me and RF4, Wolf, 'bird, Smilin', et al flew way beyond the envelope I would expect for a pilot hauling SLF and trying to get the zillion $$$ jet safely to the destination. Nevertheless, we had our "boldface" procedures to rely upon for the first few seconds. Then we tried to figure out exactly what was going on and what we could do about it. The ol' OODA loop.

Make no mistake, great cockpit displays and unambiguous caution/warning indications are extremely important. I don't believe the AF447 crew had such.

The stall warning implementation deeply disturbs me. No "stick shaker" gizmo, no clear display of AoA, etc. AF447 was above that 60 knot speed that disregards the AoA, and even if the pitot system was FUBAR and output was zero, the AoA should have been provided in a clear manner to help the crew. Some jets have very benign stall entry characteristics, preceded by mild buffet sans wing rock, little yaw moments, etc. hence the 'stick shaker/rudder shaker" and other gizmos. They told you, GET THE NOSE DOWN!!!

A few here have noted that at high speed the AoA only needs to be changed slightly to get really large changes in gee and pitch changes. True, but the 'bus FBW major pitch law is mainly gee, so big deal. OTOH, once getting slower and approaching or actually in a stall, the AoA is the best gauge, and friend, you have!! It's also super when flying an approach and something is awry with your weight versus approach speed calculations. in other words, it tells you something is wrong. And remember that the plane flies on AoA and not airspeed ( which can go Tango Uniform, as has happened numerous times in the 'bus"). i might add that the Viper AoA versus gee command function reduced available gee command from 9 to 1 as our AoA increased. So at max AoA we could only command one steenkeeng gee!

So I shall still not let Airbus off the hook. Neither shall I let the crew off the hook.

I would pray that current and future crews have a requirement to fly the jet in the alternate laws under strict supervision. Not depend upon simulators. Actually "feel" the plane and maybe even the buffet associated with an approach to stall.

Sorry if I have ranted.

doppeldecker

27th Aug 2011, 23:19

why did the autopilot disengage ?

gums

28th Aug 2011, 00:01

Yep, Ventus, we've all heard about the "deep stall", and went thru the litany a thousand posts ago.

I even posted a graphic of the pitch moments of the Viper to show the problem, and no T-tail or downwash on the HS or....

We have looked at some unofficial charts of the Airbus and it does not appear to to have the classic deep stall combinations of downwash or lack of nose down pitch moments due to c.g. as with the Viper In other words, it appears that if the crew had applied nose down stick and maybe nose down trim that the jet would have "flown out" of the stall.

This is not unusual for many military fighters built since the late 60's. The hornet can duplicate the AF447 profile quite easily. To recover? let go of the back pressure and push the stick forward, just like a Cessna 150. Duhhhh?

Problem is recognition of the stall and the approach to stall on a well-designed, modern jet. Buffet can be confused with mach problems. No horrendous yaw excursions or abrupt pitch changes. The sucker will settle into a decent stall and not spin due to great yaw control laws and such. May have some roll tendencies, but not like +/- 20 or 30 degrees back and forth.

Apparently, the crew couldn't hear or disregarded the aural stall warning sounds. So I would think a simple stick "vibrator" could get the pilot's attention to look at the "new, improved" AoA indicator that should be installed. No timy bar graph, but something as big as speed and such. Red at top, yellow in middle and green at bottom. Pull and get into red, push and get into green. Real simple, and the Navy pilots have used the equivalent for 50 years when landing on the rolling deck of a carrier.

DozyWannabe

28th Aug 2011, 03:01

What do the pilots know that the computers don't already know? In fact everything the pilots see is filtered by computers. The AF447 pilots were just as blind and bereft of real data as the pilots.

Whoa there, milady! For a start, the FMC/FCUs do not have a different set of reactions programmed based on the time of day for a given location. The specifications were laid down on the understanding that during daylight hours, the pilots have an outside visual reference that the computers not only do not have, but cannot comprehend. The design assumptions also factored in that the computers cannot match the human ability to problem-solve and compensate for situational changes in real-time.

What this boils down to is that in night-time IMC, the human pilots have less of an advantage over the computers than they do in daylight hours in terms of perception, but again, the human ability to problem solve "on the fly" is still better than that of any computer system of the vintage used in modern airliners - as a reminder :

Aircraft Computer Chip H/W Development S/W Development
-------- -------- ----- --------------- ---------------

A320: ELAC Motorola Thompson-CSF Thompson-CSF
68000

SEC Intel SFENA Aerospatiale
80186 Atelier Logiciel

A3[3/4]0: FCPC Intel Aerospatiale Aerospatiale
80386 ADL Atelier Logiciel

FCSC Intel Sextant Aerospatiale
80186 Avionique Atelier Logiciel

Yes the computers have raw AoA information, which right now the pilots do not (this may change). However, I believe there are still too many variables in play to build the kind of system you're suggesting using the available technology and still get it into a certifiable state.

Fundamentally, while I understand your desire to engineer a way out of this situation, I think following that path is well-intentioned, but misguided. History has shown that nature (and the human psyche in particular) can outwit the best efforts of engineers almost every time.

grity

28th Aug 2011, 04:10

gums So I would think a simple stick "vibrator" could get the pilot's attention to look at the "new, improved" AoA indicator that should be installed. No timy bar graph, but something as big as speed and such. Red at top, yellow in middle and green at bottom. Pull and get into red, push and get into green. Real simple, and the Navy pilots have used the equivalent for 50 years when landing on the rolling deck of a carrier.

hy gums and other real captains, who will test the simple side-window-wool-yaw-string in the real world of a jet?

http://www.dg-flugzeugbau.de/uploads/RTEmagicC_seitenfaden-4.jpg.jpghttp://www.dg-flugzeugbau.de/uploads/RTEmagicC_seitenfaden-4.jpg.jpg

gums

28th Aug 2011, 05:09

Yeah, Grity, if you could see the AoA string sucker, then you might/could use it! Night could be a problem.

We had a "yaw string" on the T-33, and that sucker worked up to 400 knots or so. Couldn't see it at night, but in daylight it helped one to learn to coordinate rudder with aileron.

A thorough training curriculum seems best recommendation from the accident board at this time, IMHO. Real planes at real conditions and with an experienced instructor. More value than with any simulator ever built.

I also want to take exploit this post to challenge one contributor that asserts the best way to stop on a slick runway is to watch the speed indications on the gauges. Beam me up! If I hit the brakes and I don't "feel" some effect, but "feel" I am actually going faster, then I don't need to look at some steenkeeeng gauge. I know that I am sliding on the ice or water. If anti-skid is working, I'll see some effect, and feel it as well. I'll guarantee most pilots here can sense the lack of decelleration faster than the airspeed/groundspeed gauge can reflect. The same applies to when a good crosswind is present when rolling out on a slick runway. Ain't no gauge invented except a HUD with a good flight path vector that is still working with weight-on-wheels that shows you that you are sliding off the runway. You look out the windshield and feel your skid. Simply holding a heading won't help.

On the other hand, relying too much, if at all, on your human senses while in bad weather most often leads to disaster. Gotta trust the instruments ( except 'bus airspeed in icing conditions). Our human sensors are great for detecting initial body rates, but degrade very quickly after a few seconds. Our human gee sensing is fairly close to what a machine sensor would detect, but that is not good for survival after the initial pitch input on the controls. It simply lets you know that the plane is reacting to your initial input. After that you had better trust the instruments.

loxosceles

28th Aug 2011, 05:19

Interested non-pilot observer here. I've read much of the AF447 threads here and on airliners.net, and was fascinated by this accident when it first occurred.

One thing struck me though in the last thread here. In post 1346 in thread 5, "xcitation" brought up airbus manual specs stating that during high speed protection the flight computer adds "positive nose up g demand to the sidestick order." I would like to know, definitively: would such a computer-added input be included in the input traces sent to the FDR? If so, it seems possible there was a logic error that caused added nose up g demand during an UAS situation, and that nose up was included in the recorded data from the sidestick. In that case, the PF may not have been actually holding stick back input, and there would be no way to tell from the ADR traces.

I would like assurance that the input data traces from the FDR are raw inputs, and are never altered at all by the computer, in any situation. If they can be altered, then the PF's input traces are useless, because it could be PF error, or it could be computer error that is indistinguishable from PF error.

PJ2

28th Aug 2011, 06:47

loxosceles;
I would like assurance that the input data traces from the FDR are raw inputs, and are never altered at all by the computer, in any situation. If they can be altered, then the PF's input traces are useless, because it could be PF error, or it could be computer error that is indistinguishable from PF error.The parameter for side stick position reports the side stick position, unmodified by any other inputs.

A slight confusion may occur if one considers that, for example, an aileron parameter will indicate the position of that surface, as controlled by the FCPCs [flight control primary computers] which are controlled in turn either by the side stick in manual flight or the EFCS [electronic flight control system] in autoflight. But the parameter still indicates actual "position" of the control surface.

Aircraft system, aircraft state and environmental parameters' design and mode of operating are thoroughly understood such that any untoward behaviours of a system are almost certainly going to be known. This was evident in the QF72 pitch incident.

PJ2

28th Aug 2011, 06:58

Fine posts, gums, thank you.
We have looked at some unofficial charts of the Airbus and it does not appear to to have the classic deep stall combinations of downwash or lack of nose down pitch moments due to c.g. as with the Viper In other words, it appears that if the crew had applied nose down stick and maybe nose down trim that the jet would have "flown out" of the stall.IIRC there is the opinion that the aircraft could have been recovered from the stall, but opinions have varied regarding the altitude from which recovery could have been made. From what has been said, the elevators remained effective throughout the descent.

Mr Optimistic

28th Aug 2011, 09:55

JD-EE

28th Aug 2011, 11:07

Chu Chu, with a wry grin I note that in this case the meatware proved faulty, probably for the same reason, faulty or incomplete programming, as the software might fail. I think that's an interesting if off beat way to look at the problem.

BOAC

28th Aug 2011, 11:07

I am disturbed by some here that believe you can handle an emergency situation or an "upset" ( love that term, and cracks me up and guess it really means "WTF?, over") by following rote, memorized procedures for more than a few seconds. - I am equally disturbed that you are 'disturbed':)

What EXACTLY do you propose a pilot should do for those 'few seconds' when all is 'WTF' confusion? What do you suggest that stall recovery should be? 'By rote' reduce AoA or some other gums method involving what process? Nose high, speed falling - 'by rote' lower the nose, use power with care and bank if necessary or 'by gums' - let's see, I have 25deg nose up and 4000fpm up and speed decreasing through Vstall+20, so out with the spin-wheel/QRH and I need......................?

I'm afraid I cannot see your point, sir. Indeed, on the many occasions when I was, like you, 'outside the envelope' in my miltary days it was the 'by rote' that saved me..

JD-EE

28th Aug 2011, 11:12

HN 39 asks, "Which computer?"

At the very least the wee little computer that drives the FDR surely knows. If it knows then its source can also route that data where ever else it is needed.

Chu Chu

28th Aug 2011, 14:24

I guess the general idea is that the pilots are as smart as the programmers/procedure writers, and therefore can fill in any holes on the spot. I'm sure that hasn't actually been the case since the Wright Brothers. But maybe the AF 447 pilots could have filled in the particular hole here if they'd been better equipped.

gums

28th Aug 2011, 14:42

Sorry, BOAC, I worded my concern poorly.

You are spot on as far as the initial procedure and then "plan B" and so forth. That process has certainly saved many of us here. Lest anyone here think otherwise, I fully support your position.

I was trying to point out to some folks that simply executing the tried and proven procedures/techniques may not be enough to recover from some "upsets" or cope with unusual system malfunctions. In other words, there are a very few conditions that are not covered by the known procedures. The two incidents that strike home are Sully's ditch and the Sioux City DC-10 landing attempt. After their initial actions, and not getting results, both crews went on to new territory and did not sit there for 3 minutes doing the same thing.

Hope that makes my position clear.

Lyman

28th Aug 2011, 14:49

I first used "ROTE" as a pejorative; and to expand, 'rote' (the word) is not in and of itself unhelpful. As JD-EE eloquently puts it, 'the meatware had a software glitch'.

In America, 'rote learning' is seen as dispassionate, and insensate. Unfortunately, in a cultural setting, a word frequently conjurs all manner of misunderstanding, here, what confuses is the unsaid.

The "clipped" sense of 'rote' means more or less, "Instead of using your thinker".

Common usage is, well, common.

Rote.

In this misunderstanding here, perhaps an opportunity to close the consensus gap.

ROTE is 'bitchin'. ROTE is 'sick'. ROTE is 'memory items', for purposes here?

Language, so boring w/o it. 'Fly'.

gums. that is it, in suede. It is the 'Between' SOPS, and (UAS Drill), and Sully's test flight (w/o the brief), that the solution lies.

It is a no-man's land, a Maginot Line of suspicion, ignorance, and Pride, with globs of ego in there as well.

Take one of our engineer friends for a hop in your family model, and see if they grok what it can be like, flying with one's hair on fire.

Time is GOD's way of making certain everything does not happen at once.

'Transition' is what we do, or DIE. "I have the Controls' hmmm.......

really? because........

ChristiaanJ

28th Aug 2011, 15:07

HN 39 asks, "Which computer?"
At the very least the wee little computer that drives the FDR surely knows. If it knows then its source can also route that data where ever else it is needed.
I suppose you are referring to the FDIU (flight data interface unit).
It's barely a 'computer'... the French term 'ordinateur' would be better.
It doesn't 'know' anything, and it doesn't 'think', it just converts and formats the multiple original AC and DC signals and digital info to a single formatted digital data stream (ARINC 717) which is then recorded by the FDR.
The link below is litle more than advertising, but may give you a few hints.
Teledyne FDIU (http://www.teledynecontrols.com/productsolution/fdiu/overview.asp)

Lyman

28th Aug 2011, 15:14

PJ2 this is in re: your following POST (don't ask me).......

(your #553)

Nothing personal, and certain to be flamed as PARSE.

They did not ADAPT, they CREATED. Adapt means to conform, and I am sure that is not what one meant, because in that sense it is remarkably disrespectful to some extremely talented people.

Creation is what is at stake here. Mostly Man destroys, the rest ADAPT, and the brilliant CREATE.

That is the essence of thus entire kerfuffle. Shall we CREATE, OR ADAPT?

ADAPTATION is the default, To create is to control. Automation is the ultimate adaptation, and clearly, there is a point of view at work here.

PJ2

28th Aug 2011, 15:17

gums;
The two incidents that strike home are Sully's ditch and the Sioux City DC-10 landing attempt. After their initial actions, and not getting results, both crews went on to new territory and did not sit there for 3 minutes doing the same thing....and the British Airways B747 Flight 9 encounter with volcanic ash at Jakarta. Your point is well taken. In other words, these crews knew their aircraft and flew it. If something didn't work, they adapted through knowledge and experience.

Lyman

28th Aug 2011, 15:39

Sorry if I seem to bracket one.

Automation is an adaptive environment, by definition. The upside is that man created the environment, for once. Then he fell immediately into what is demeaning, an adaptive posture.

On the one hand brilliantly creating his own environment, "Can you imagine?". Then, dumbly acquiescing into its power.

One has suggested Human Factors is the proper venue for this discussion, there you go.

grity

28th Aug 2011, 17:34

gums We had a "yaw string" on the T-33, and that sucker worked up to 400 knots or so. Couldn't see it at night, but in daylight it helped one to learn to coordinate rudder with aileron. are you shure you have a side window "yaw string" on the T-33?, you had it in the middle!?!, god for control not to much side slide...
but the side window yaw string is not for the side slide he is a good AoA indicator!, and I am shure it is easy possible to enlight it at night..... just open the window like nico and hang it out ....
http://www.spox.com/de/sport/formel1/0803/Bilder/514/rosberg-nico-hut-514.jpg

OK465

28th Aug 2011, 18:05

Brief off topic:

Way in the past, on a day we were flying the local DC-9, they gave us a recorder box with a microphone on a long cord attached.

The purpose was to sample cockpit noise levels.

On taxi out the guy in the right seat gave me a sly grin, opened his sliding window and hung the microphone outside.

When we got to the runway we got one of those "no delay please, keep it rolling, traffic 2 mile final" clearances to go. He quickly closed his window and off we went.

The banging noise on the right side of the fuselage prompted an "Oh $%#" from him and some momentary concern for the right engine, but it stopped after the flaps were up. Engine fine.

We figured the mic was gone, and we had a whole flight to come up with a story. However after we landed, he opened his window and reeled the still attached mic back in.

The folks who analyzed the recording were somehow unable to see the humorous aspect of this.

Be careful what you hang out. :)

(The F-4 had a yaw string for a time but it only served to scare you at high AOA so it was removed.)

Back to JC's hamsterwheel.

grity

28th Aug 2011, 18:50

OK, funny story

mayby I better mean a "side window pitch string" and not a yaw-string

Mr Optimistic

28th Aug 2011, 19:36

Perhaps it's worth bearing in mind that when computers crash it is human beings who reset, repair and reboot them.

3holelover

28th Aug 2011, 20:01

I think what's happened is that, as the geek squad (sorry guys and gals) began to be the ones setting up airplanes, the old guard/pilot crowd had to defer to them regarding training requirements.... The geek squad then inundated them with a whole bunch of "if then else" and logic flow style training, and the piloting was forgotten.

This tragic accident has likely(hopefully!) served as a catalyst to correct that.

HazelNuts39

28th Aug 2011, 20:08

Grity,

I liked your picture (happy memories), but just to be precise, as noted by gums ... "AoA string" would be more appropriate.

Clandestino

28th Aug 2011, 21:04

PJ - the problem with your suggestion is that it does not cover all bases, in that with a total baro data failure one could see a crew 'floundering' for a pitch attitude, whereas 5 is SAFE, will not cause a stall from steady cruise and will then give time to enter the trouble shooting phase with pitch/power from the QRH.Exactly! Thank you for sparring me the effort of explaining.

I am disturbed by some here that believe you can handle an emergency situation or an "upset" ( love that term, and cracks me up and guess it really means "WTF?, over") by following rote, memorized procedures for more than a few seconds.In context of AF447, only significant upset was pilot induced. That what you call "rote" are memory items for unreliable airspeed procedure. It is not upset but can turn out to be, if handled inappropriately.

Make no mistake, great cockpit displays and unambiguous caution/warning indications are extremely important. I don't believe the AF447 crew had such.I am sorry sir, but ambiguity of "STALL STALL STALL STALL STALL STALL STALL STALL STALL STALL STALL STALL STALL STALL" aural warning is lost upon me but it is for BEA HF group to discover why crew went selectively deaf on stall warning. Granted, stickshaker and big red flashing "STALL" in the middle of PFD or as separate light somewhere on glareshield might have helped a bit but questions remain: why did the RH pilot pulled the aeroplane in the stall in the first place? Why didn't the LH pilots survival instinct kick in?

the sucker will settle into a decent stall and not spin due to great yaw control lawsThere are no control laws affecting the A330 rudder. It is mechanically controlled, with electrical: trim, yaw damping and turn coordination. Rudder is totally conventional (in the context of heavy jet) and can not be blamed for A330 resistance to spin.

I also want to take exploit this post to challenge one contributor that asserts the best way to stop on a slick runway is to watch the speed indications on the gauges. (...) Ain't no gauge invented except a HUD with a good flight path vector that is still working with weight-on-wheels that shows you that you are sliding off the runway. You look out the windshield and feel your skid. Simply holding a heading won't help.No kidding... you might as well name me, sir.

You might find interesting that in my regular airline life, I have absolutely no need to look down to see any of those aforementioned parameters. They are all nicely and neatly presented in front of me on Flight Dynamics (TM) Head-up Guidance System. Inertial deceleration. Air data derived deceleration. GPS groundspeed. Inertial flightpath vector. Runway remaining in low visibility roll-out mode.

I find all of those a bit more reliable and precise than my own nether regions, sir.

However, during my A320 days, we had no HGS and crew coordination during tricky landings was even more important than nowadays, when it's bloomingly high. It was expected that PF would keep an eye on the world outside, trying to keep the aeroplane on the centerline so PNF would keep head down and monitor spoiler deployment, reverser operation and deceleration. We believed that one's sense of deceleration is not to be trusted, especially when one is struggling to maintain the control so we task shared. I don't think we were wrong.

gums

28th Aug 2011, 22:13

No problem, Clandestino.

I have been having a problem expressing myself lately.

Still want folks to know that the human "body rates" can help for the first second when something unusual happens. Also, you cannot depend upon gauges when you have to make very quick control inputs flying a few feet on the wing of your leader. Besides, you can't do much except take an occasional glance at the HUD or ADI to let you know you have not been in a sustained bank or whatever.

The HUD's I flew quit showing the aircraft vector once weight-on-wheels switch was activated. I agree that the speed indication is very important once you know/feel that you are slowing down. But I never needed to look at the speed "thermometer" in the first second I tapped the brakes on an icy runway. Most of my early birds did not have an "inertial" groundspeed indication, and the pneumatic pitot reading went to zero around 40 knots.

Same observation as far as skid is concerned. I'll take my butt for the first half a second, then depend upon my eyeballs to keep tracking down the centerline. And BTW, when I landed the Viper with that leading edge flap up, the flight path marker was almost useless ( I musta had 15 degrees of yaw), so I flew the ILS gauges even tho I broke outta the overcast about 1200 or so feet AGL. I simply tracked my butt down the centerline until "impact", heh heh. Next guy to try it almost pranged, as I had not yet briefed all the pilots on the 'technique". Additionally, the preferred technique ( actually "mandatory") was to land in a crab when crosswinds were high. The aileron-rudder interconnect cut out with weight-on-wheels, so trying to land "wing low" in a slip resulted in a huge yaw moment upon touchdown. Not good, to say the least, and another instructor tried it one day when messing around.

My view of the 'bus rudder is not that the FBW system has a "spin resistance" feature as we had in the Viper once our AoA was above 30 degrees. Seems to me that the basic yaw damper function and a great directional stability that is inherent in the jet's design did the trick. i would also postulate that spoilers could have compensated for adverse yaw once the crew got into uncharted territory.

I agree that the true "upset" was not loss of the pitot system or A/P disconnect. We are in the same "zone" in that regard.

The stall warning that seems to be audible in the cockpit of AF447 still puzzles me. I am not sure if the crew ignored it due to the unreliable airspeed or what. A mystery to me as well as many here.

Finally, I am overjoyed that commercial jets are finally getting good HUD's. The Shuttle didn't get one until just before Challenger, or even about the same time. My 'nam roomie was the lead astronaut for the Shuttle's HUD, and had to overcome some resistance from Crippen, Young and others of the "old guard". I was a techno-geek, and used the HUD more than most in the old days. The "inertial" vertical velocity" was especially handy, and later HUD's had inertial ground speed. Still had an easy cross-check with the steam gauges, plus the AoA indexers on each side of the HUD ( ask 'bird and RF4 and others that flew Navy jets).

Lonewolf_50

28th Aug 2011, 22:35

If someone who knows has the time and patience, could they explain what actions would be necessary for the crew to get the AP back and what dynamic state the a/c must be in for the AP to capture control ? Reason for asking is that before the 'details' became known I rather expected to find the crew heads-down dealing with the computers but this seems not to have been any sort of factor.
Mieklour pointed out that if you get ALt 2 latched at cruise, you fly the rest of the flight in Alt 2 and you land in Alt 2. What that means, I think, is that you don't get A/P back, but I may not be reading the block diagrams correctly.

DozyWannabe

29th Aug 2011, 00:58

LW_50 :

According to the documents I've got, inhibition of A/P is not necessarily the case as a result of Alternate Law latching, in fact the only control law where the manuals specifically state A/P is unavailable is Direct. However, the A/P disconnect in this case was due to ADR disagree. If I recall correctly, it was possible in some cases to re-engage A/P once stability was restored, but according to the Flight article below, EASA are mandating a change to inhibit A/P for as long as ADR disagree is active. They don't say if this behaviour should be latched.

A330/340 change to inhibit autopilot if airspeed unreliable (http://www.flightglobal.com/articles/2011/08/02/360261/a330340-change-to-inhibit-autopilot-if-airspeed-unreliable.html)

It's curious in a way, because it appears the AF447 crew did not try to re-engage A/P, but maybe as part of the simulator testing they've discovered anomalous behaviour if you try.

Ian W

29th Aug 2011, 01:10

Clandestino posted:
I am sorry sir, but ambiguity of "STALL STALL STALL STALL STALL STALL STALL STALL STALL STALL STALL STALL STALL STALL" aural warning is lost upon me but it is for BEA HF group to discover why crew went selectively deaf on stall warning. Granted, stickshaker and big red flashing "STALL" in the middle of PFD or as separate light somewhere on glareshield might have helped a bit but questions remain: why did the RH pilot pulled the aeroplane in the stall in the first place? Why didn't the LH pilots survival instinct kick in?

A little exercise for those that are interested.
Go back a page or two and read one of the more technical posts - say by Owain, while doing that try to recite a well known child's rhyme to yourself at the same time have someone read a paragraph of different text for you to to write down. You will find that you cannot do all of these - in fact if you concentrate on the reading YOU WILL NOT EVEN HEAR the person talking to you.

This is not selective deafness - it is because the human brain has a limited number of cognitive channels and they can only handle ONE input at a time. So aural verbal, visual verbal and speaking verbal activities all use the same single verbal cognitive channel. (see the work of Christopher Wickens).

Some people can if trained and practiced rapidly switch between various verbal inputs and outputs - but if something important happens on one input they WILL NOT HEAR the others.

Think about how many times when you are driving and there are complex lane changes and road signs... that you may have had to ask a passenger to repeat themselves.

The reason that 'steam gauges with needles' seem to be easier to read is that they are a spatial cognitive load and form patterns that can be recognized without much cognitive effort. All the glass cockpit tapes with numbers and the ECAM require visual verbal analysis; and no-one can read one thing and fully understand it while saying something else and listening and comprehending something else again. The human brain cannot do it so will just 'drop' any input that is the overload - it is perfectly possible that the pilots literally did not hear the stall warning as their verbal processing cognitive channel was already overloaded. A stick shaker or other haptic input, like someone tapping you on the shoulder when you are busy, can have an immediate attention getting effect that a voice alarm or flashing words may not.

One of the aspects I expect the BEA Human Factors investigators to look at is the cognitive workload that the ECAM and failure messages put on the pilots. Especially the aspect of overloading particular cognitive channels. Perhaps every potential emergency scenario should be subject to what is called a 'cognitive walk-through' that actually assesses the cognitive loads and identifies likely overloads.

Older pilots may well have followed a rather older but repeatedly successful dictum - disregarding all the cacophony - aviate (i.e. pitch and power), navigate, then communicate.

It mightn't pass the sim check ride - but it may have had a better outcome in the real world.

ChrisJ800

29th Aug 2011, 01:32

Ian W, interesting post. I read somewhere that Vietnam war pilots could get saturated with aural inputs and sometimes failed to hear an important tone or alarm such as incoming missile...

DozyWannabe

29th Aug 2011, 02:59

One of the aspects I expect the BEA Human Factors investigators to look at is the cognitive workload that the ECAM and failure messages put on the pilots. Especially the aspect of overloading particular cognitive channels. Perhaps every potential emergency scenario should be subject to what is called a 'cognitive walk-through' that actually assesses the cognitive loads and identifies likely overloads.

Hi Ian,

I think you're definitely on to something with your second and third sentences there, but I'd like to see what you think about an alternative theory/explanation that I posted on the R&N thread. The particular links are below.

http://www.pprune.org/rumours-news/447730-af447-wreckage-found-166.html#post6665388 (1)

http://www.pprune.org/rumours-news/447730-af447-wreckage-found-166.html#post6665478 (2)

http://www.pprune.org/rumours-news/447730-af447-wreckage-found-166.html#post6665527 (3)

The meat of the theory is in the third one, but the first two provide background notes as I was working through it. The first in particular lists data points in the interim report that I hadn't put together before.

In a nutshell, I'm not convinced that the ECAM messages provided that much of a distraction, except initially - the PNF calls out the most important ones at any rate. Disregard the PF for now, and focus on the PNF - he's clearly getting jumpy as soon as the PF takes manual control, but because of the informal handover, he doesn't know his boundaries of authority, so he summons the Captain to clarify. The stall warning goes off shortly afterwards and after approx. 50 seconds of letting the PF handle the aircraft, the PNF has had enough and takes control. The inputs he makes are correct, but not enough - a couple of seconds later the PF takes control back unannounced (i.e the PNF still thinks he has control) and starts pulling the nose up again. A second or two after that, the stall warning stops, with both the PNF and the PF thinking they have control. The PNF's inputs were correct (lower the nose, stabilise roll), but again - not enough. Did he think he had resolved the problem with the end of stall warning? Complicating matters even further is the arrival of the Captain - 2 seconds before the SW stops, when the PNF presumably turns round to talk to him, and as such he loses all focus on the flight controls and the PF is still pulling back.

I'm not sure if human/machine cognitive overload is the problem here so much as the level of pressure and stress on the PNF. For all my opinion's worth as a layman with an interest, could it be the human/human interface that is causing the issue?

Old Carthusian

29th Aug 2011, 04:56

Iain W
There are many examples of 'operatives' of machinery with 'steam gauges' ignoring them with disastrous consequences. The idea that a spatial cognitive load would work better than a visual verbal stimulus is not actually correct. We possess the ability to filter these out as well. The same goes for a stick shaker which given how the PF reacted might have been just as useless as all the other warnings. Recognising the actuality and responding appropriately is the key to dealing with a situation. This involves knowing your machine and good training. The latter aspect is vital in overcoming the psychological shock (and shock is the operative factor here) of a sudden unexpected event. A chain of command could have helped here and more decisive action by the PNF. DozyWannabee raises an important point about human/human interface. Once again we have to ask is AF's flight deck management procedure inadequate?

grizzled

29th Aug 2011, 05:03

Dozy...

RE: I'm not sure if human/machine cognitive overload is the problem here so much as the level of pressure and stress on the PNF. For all my opinion's worth as a layman with an interest, could it be the human/human interface that is causing the issue?

IanW's comments and yours are not mutually exclusive. Both factors were at play here and I believe, in terms of "cause", the final report will reflect and expand upon the points expressed in IanW's excellent post, as well as your insightful observations re the PNF /PF interactions. (Just because you're a "layman" in terms of piloting, doesn't mean you have this wrong.)

grity

29th Aug 2011, 08:47

HN, AoA- or side- or pitch-string a correct name is not vitalli for a 20" wool-fathom

how exact will it work outside the side window of an A330??

between 2 and 3 deg is my expectation,

so mayby it is not exact enough to deside if AoA is 4 or 5 deg, but il will react very fast every time and for sure even with slow speed (<<60kn) and you can absolute shure distinguish between highspeed and stall,

it is so easy to check out, I hope that we find a jet-captain for a honest try

wilyflier

29th Aug 2011, 09:07

Probably pretty hard to see in the dark from inside??

rudderrudderrat

29th Aug 2011, 09:48

Hi DozzyWannabe & IanW,

I agree with both your posts and believe it points to another couple of holes in the cheese lining up.

The auditory overload of ALT Alert, ECAM Dings, "Read ECAM" etc. possibly led to the mental exclusion of "Stall Stall". They simply didn't hear it. In a previous post (512), I mentioned another crew who didn't hear "Dual Input" for over 90 seconds.

Both pilots are provided with the FD/AP modes on their PFDs so they can monitor what the AP/FD is attempting to do - yet when manual flying without the FD, PNF has absolutely no idea what input PF is making (due Side Stick design) or what he is aiming for. PNF probably did not realise how much over controlling input was being made by PF.

JD-EE

29th Aug 2011, 10:28

Chu Chu and others - this is an interesting video from Rockwell Collins.

This is where pilots may become obsolete. Computers doing it right can recover from singularly amazing upsets. Sure, this is a model. And with enough power you can fly a lawnmower and control it. This one is a DARPA project. So I suspect the cheats have been turned off. Only the Collins adaptive technology remains to save the plane.

Sufficient data existed on AF447. The meatware failed. It failed partly because data needed was not present in the cockpit in a form the pilots appreciated. It failed partly because the pilots were not trained to fly the plane in unusual configuration where dial-a-flight didn't work and most of the "protections" had vanished. Nor were they encouraged to practice flying the plane. Finally the software design did not permit or make feasible flying in a simulated ALT2 environment for practice.

If the airlines are not going to train pilots and provide pilots with all the equipment that might be needed when going gets tough and instruments are all you have, I for one would not want to fly the plane under those circumstances even given the otherwise sterling flight safety records they have.

JD-EE

29th Aug 2011, 10:34

ChristiaanJ, know in context was "has the data present for use". The data certainly was present on the plane. It equally certainly was not adequately presented to the crew.

JD-EE

29th Aug 2011, 10:49

Dozy, in light of Ian's fine presentation, a repeat performance indicating that the noise here obscured his message the first time, perhaps icons would be better than words. I suspect an image of a plane hanging by its nose to indicate stall might get some attention.

Only a small fraction kidding.

Old Carthusian

29th Aug 2011, 11:06

JD-EE
I am not sure what you are driving at - nothing connected with the plane failed after the initial pitot tube problems. The pilots were inadequately trained, their CRM was poor, they made no attempt to follow SOPs - the PF seems to have been suffering from acute stress reaction (shock). Nothing in fact to do with the machine but everything to do with the people and the airline.

The airplane was perfectly flyable and should have been flyable on instruments alone. Too many people are grasping at straws and thinking that somehow the machine or the warnings had a big impact on the accident. It didn't, the PF was the one who made the NU inputs and it wasn't because of cognitive overload and not hearing the stall warnings - there were three pilots. The PF panicked and went into shock pulling back on the stick. That's all there is to it. Air France training and cockpit procedures didn't give the PNF the authority or confidence to override the PF. By the time sufficient authority arrived (the Captain) it was too late.

The maxim that all pilots should follow - know your machine and these guys didn't. This is not the responsibility of the machine. It is the responsibility of the pilots and the airline.

CONF iture

29th Aug 2011, 11:52

I mentioned another crew who didn't hear "Dual Input" for over 90 seconds
I'm interested in that one, would you have a report ?

Computers doing it right can recover from singularly amazing upsets
They knew exactly where the wing was going to break and all the probes of any kind would be undamaged.
Have not we seen already a F15 losing big part of the wing and a pilot bring it back for a safe landing too ?

CONF iture

29th Aug 2011, 12:11

It is mechanically controlled, with electrical: trim, yaw damping and turn coordination.
Captain my great Captain, you may as well forget about any mechanical link for the AF447's rudder.

DozyWannabe

29th Aug 2011, 12:52

Dozy... IanW's comments and yours are not mutually exclusive. Both factors were at play here...

Agreed - I intended the cross post to suggest an "as well as" rather than "instead of", though I remain to be convinced that the PNF had much trouble understanding what the aircraft was telling him. The PF... not so much.

Old Carthusian - I don't know if it was "panic" causing the PF to make the inputs he did, but his actions certainly suggest him developing a degree of tunnel vision from the start of the sequence. We can't know if he heard the PF call out "Alternate Law", he certainly didn't interpret "Controls on the left" as the PNF saying "I have control".

GarageYears

29th Aug 2011, 13:35

Just an idle thought - how could you implement a stick-shaker on an Airbus-type side-stick?

That doesn't seem that easy, compared to a conventional yoke, at least.

Clearly the vibrational aspect of the 'shaker' is significantly important, since the aural aspects appear to be relatively easily dismissed. However, unlike a yoke, there is no significant mechanical structure to attach the shaker drive to.

If the SS were back-driven then it would be comparatively easy to implement I suppose... but it isn't.

HarryMann

29th Aug 2011, 14:16

The discussion seems to be oscillating between flight systems and instrumentation
.. and human factors.

On the latter, both Ian and Dozy make good points...and no, they're not mutually exclusive so a good contribution

Could I remind everyone, that if I remember correctly, PF had recently arrived in the cockpit from a rest period.
Do we know what this rest period consisted of?
I am suggesting that cockpit crew recently back from rest (sleep even) are not the ideal choice for immediate command, when another is up to date with the flight.
Perhaps it is rather naive or let's say simplistic thinking that freshly rested pilot is the more alert and certainly not up to speed as much as an exisiting crew member.

I think we all know that some people wake up and brighten up quickly, othjers take a lot longer - that is, those that can 'cat nap' and revive almost instantly well and those that can't.
Being one of the latter, I can imagine the lack of considered thought and action when confronted with a sudden shock - especially if not anticipating the responsibility of being placed in charge of the flight.

== Yaw strings and pitch strings at night ===

No problem, miniature lightweight LED ribbons/tapes are now being used in many applications, the batteries being small enough to be encompassed within the string I should think... go on, give it a go ;)

3holelover

29th Aug 2011, 14:30

HarryMann.... It was the PNF who'd recently returned from rest. PF had been in his seat since departure.

DozyWannabe

29th Aug 2011, 14:31

Could I remind everyone, that if I remember correctly, PF had recently arrived in the cockpit from a rest period.

Sorry, that was the PNF - we discussed it on the R&N thread, and he seems to have been fairly alert, to judge by the CVR. The PF had been in the right-hand seat since pushback. There was a suggestion that the PF had actually been on holiday in Brazil with his wife, so your suggestion could have merit in another way - should someone who has had several days off be put in a "relief pilot" role over the guy who flew the outbound leg?

Of course if neither of them had had manual handling training, that question could be academic, but I'm struck by the difference in handling approach by the two F/Os. The PNF appears to be doing as one would expect, gingerly handling the controls until the aircraft starts pointing the way he wants it to go, whereas the PF seems to be handling them with a frankly astonishing coarseness from the outset.

SRMman

29th Aug 2011, 14:58

The gist of many posts has been that 'modern' airline pilots aren't like they used to be, skill levels and understanding of the aircraft are lower because transport aircraft fly themselves, FBW, etc etc. Given that it seems to me unlikely that things will ever revert to the 'old days' in terms of longer training and acquired experience before new pilots reach the flight deck, maybe a different approach is needed, reflecting the "computer games" world that a lot of younger people (and some oldies!) live in today.

It has occurred to me, and apologies if this has already been suggested, that maybe a display could be made available on the PFD of a 3D model of the aircraft, as seen from the outside showing its attitude and direction of flight through the air relative to the ground. In fact something pretty similar to one of the views available in MS Flt Sim. The model, viewed either from behind or side, would be driven by the normal sensors and would replicate the actual aircraft, thus providing an easy and quick-to-assimilate visual understanding of the aircraft's behaviour. Thus, unusual attitudes in IMC or at night could be more readily recognised and countered.

On a different subject, it has been said that the flight condition AF447 was in during the final few minutes was far outside the certified flight envelope, and would never have been part of any test programme. Also the modern flight simulator does not or can not provide accurate simulations of this behaviour. But now we have actual data, traces, etc from the FDR of what happened during an A330 developed stall, could not this data be used or extrapolated by the simulator manufacturers to give more realistic aircraft pre and post stall training?

oldchina

29th Aug 2011, 15:18

Dozy, it was posted earlier but no-one picked up on it: is it not strange that the latest BEA report makes no mention of the crew's rest status? I'm not suggesting they got up to exhausting fun and games in Rio, just asking were they properly rested in accordance with the rules?

I think the poor PF deserves to be remembered as one who underperformed on the night, for whatever reason, compared with his usual self.

jcjeant

29th Aug 2011, 15:21

Hi,

]It has occurred to me, and apologies if this has already been suggested, that maybe a display could be made available on the PFD of a 3D model of the aircraft, as seen from the outside showing its attitude and direction of flight through the air relative to the groundNot a big problem .. and anyways already old process
More up to date will be a holographic representation .. even with vectors .. and a nice woman voice describing all the necessary actions to be performed for recover from a stall .. a spin .. etc .:cool:
Some emergency sequences can even be programed .. so it's just to activate .. and read the newspaper .. waiting end of process
Of course .. no more joystick .. buttons .. etc .. .. but instead .. some nice touch screens
All is possible today .. :)
But ... is this very useful for pilots ?

Lonewolf_50

29th Aug 2011, 15:49

Lone, I noticed this, "Alt 2 latched does not mean "flight control locked and no longer able to be moved"" before I noticed it was to bear and ignored the message. I have an idle question that does not bear on this crash.
I saw what you did there. ;)
Why might Alt 2 latch in and not be able to be escaped back to normal once the aircraft is flying normally again? Permanently latching out the aircraft safety systems for half of a long (10 hours?) flight seems counter productive.
I think Dozy had a better answer than I do, at post 564 (http://www.pprune.org/tech-log/460625-af-447-thread-no-6-a-29.html#post6668487).
According to the documents I've got, inhibition of A/P is not necessarily the case as a result of Alternate Law latching, in fact the only control law where the manuals specifically state A/P is unavailable is Direct. However, the A/P disconnect in this case was due to ADR disagree. If I recall correctly, it was possible in some cases to re-engage A/P once stability was restored, but according to the Flight article below, EASA are mandating a change to inhibit A/P for as long as ADR disagree is active. They don't say if this behaviour should be latched.
A330/340 change to inhibit autopilot if airspeed unreliable
It's curious in a way, because it appears the AF447 crew did not try to re-engage A/P, but maybe as part of the simulator testing they've discovered anomalous behaviour if you try.

Ian: post #565. Well said. :)
Old Cathrusian, good points in response.
Comment: You addressed scan breakdown, Ian addressed "data channels" in the human brain. These are both matters of interest in the human engineering side of this equation.
Likewise, Old Carthusian, the following: The maxim that all pilots should follow - know your machine and these guys didn't. This is not the responsibility of the machine. It is the responsibility of the pilots and the airline. The balance between self initiated systems knowledge growth, and training in its application to the mission seems to be a sticking point.
Dozy:
Did you get this impression from reviewing the traces?

Of course if neither of them had had manual handling training, that question could be academic, but I'm struck by the difference in handling approach by the two F/Os. The PNF appears to be doing as one would expect, gingerly handling the controls until the aircraft starts pointing the way he wants it to go, whereas the PF seems to be handling them with a frankly astonishing coarseness from the outset.

grity

29th Aug 2011, 16:34

Harry Mann== Yaw strings and pitch strings at night ===
No problem, miniature lightweight LED ribbons/tapes are now being used in many applications, the batteries being small enough to be encompassed within the string I should think... go on, give it a go braidet wire fibre optic or persistend material can work also, but first please the simple 2ct test with wool and daylight!

DozyWannabe

29th Aug 2011, 16:39

Dozy:
Did you get this impression from reviewing the traces?

Yes. The CDB (Captain's) sidestick trace is deflected in the pitch axis no more than halfway nose down (which would be appropriate to regaining level flight or stall recovery) during the "confusing" period I refer to. The F/O's trace tells a very different story, averaging halfway back or more with a single sustained application of full back stick. The same is apparent in the lateral SS traces, stop-to-stop in the case of the F/O stick.

Hi DozzyWannabe & IanW,

I agree with both your posts and believe it points to another couple of holes in the cheese lining up.

Thanks! Nice to agree on something... :)

The auditory overload of ALT Alert, ECAM Dings, "Read ECAM" etc. possibly led to the mental exclusion of "Stall Stall". They simply didn't hear it.

I wouldn't be surprised if the PNF did, given his apparent mounting alarm until he calls "Controls on the left". The PF appears to be a definite candidate for tunnel-vision, however.

PNF has absolutely no idea what input PF is making (due Side Stick design) or what he is aiming for. PNF probably did not realise how much over controlling input was being made by PF.

Again, a lot of people who don't like the sidestick design are making that assumption, but if that is the case then why do we have the references to going up when they should be going down and to touch the lateral controls as gently as possible by the PNF?

Lyman

29th Aug 2011, 18:14

Hey GY.

Saying my goodbyes, and wanted to wish you the best. I think you knew my purpose was to get some passion flowing in an otherwise "watching Paint Dry" venue.

Anyways, for my money, the vibrating Stick is a non starter, except maybe for the girls.

The way to a Pilot's attention is through his spatial awareness. SHAKE the AIRFRAME.........

Buffet underperformed here, and why not shake the part of the airframe that the pilot has directly in tune?

His ASS. SEAT SHAKER. There's your "tactile".........Make it annoying, and DON"T turn it OFF until he gets it flying propers.

Hope we meet up some day.

Lonewolf_50

29th Aug 2011, 18:19

Bear perhaps, with such a device, those whose erotic proclivities are via excitation in certain areas might too frequently fly to the edge of buffet because it feels interesting ... :} (Confession: This is me recycling an old joke about the attachments we assumed the ladies flying MiGs for the Soviet Air Force in the 80's had attached to their g-suits as add-ons to the leg bladders ... punch line being "and they always seem to come back from training sessions with an over G indication on the g-meter!" ) :}

All kidding aside, a seat shaker might be something to explore as a feature for SW physical sensation in SS equipped aircraft. (Perhaps AB looked into it and the test pilots gave it a thumbs down?)

I recall a couple of decades ago some USAF flight simulators we visited as we worked to upgrade a Navy helicopter motion simulators. Some used seat vibration to induce the feeling of being in flight.

This is fairly old tech. Put another way, the tech is there, but as with most tech, the fine points of application is the real tricky part.

GarageYears

29th Aug 2011, 19:29

Well, if you're really off, then happy skies to you, and yes, poking the 'bear' (no pun intended) is often a good way of getting the creative juices flowing. I'm no fan of paint drying, grass growing or any other similar waste of time.

I would be honored to shoot the breeze with you any time.

I like the idea though - I have worked simulators with seat-shakers in the (recent) past, so that technology is readily available. Given a benign airframe, tepid into and past the stall, the idea of artificially giving the pilot buffet, that won't go away until out of the stall is a pretty good solution.

HarryMann

29th Aug 2011, 19:57

HarryMann.... It was the PNF who'd recently returned from rest. PF had been in his seat since departure. Duh! :O Thanks, I should have double-checked, going back a long time now.

At least some other suggestions about fitness for control have come out of it :D

Something was up with the PF for sure, and I agree the distinction in handling the stick seems starkly contrasted.

All these human factors are open for speculation, such as whether PNF's confidence and subsequent assertiveness was knocked a little by PF being given 'Command' ?

But it is speculation that we can never prove, in 4 minutes of extremely sparse exchanges... Not being a native French speaker I personally find the 'go up'; 'go down' comments somewhat strange for pilots*... in English it would almost certainly be 'slow down' or 'faster', or get the 'Nose down' or up. Strange as there is not an up and down control in an aircraft, that is a secondary and dependent motion, not to be taken for granted - hence another subtle sign that there was little comprehension of a stalled (and unresponding state).

*Has it been shown that the true meaning was '...(we go up) Indicative as opposed to Imperative (or even Subjunctive)

Ian W

29th Aug 2011, 20:39

From Old Carthusian
Iain W
There are many examples of 'operatives' of machinery with 'steam gauges' ignoring them with disastrous consequences. The idea that a spatial cognitive load would work better than a visual verbal stimulus is not actually correct. We possess the ability to filter these out as well. The same goes for a stick shaker which given how the PF reacted might have been just as useless as all the other warnings. Recognising the actuality and responding appropriately is the key to dealing with a situation. This involves knowing your machine and good training. The latter aspect is vital in overcoming the psychological shock (and shock is the operative factor here) of a sudden unexpected event. A chain of command could have helped here and more decisive action by the PNF. DozyWannabee raises an important point about human/human interface. Once again we have to ask is AF's flight deck management procedure inadequate?

I should have made it plainer in my wording.
When the verbal cognitive channels are overloaded you can add more to them and all you do is make the effect worse. However, the analogue gauges do not require any verbal cognitive analysis so that channel is unloaded. So you can see a VSI needle move hard down and understand what it means even if the verbal cognitive channel is in overload - numbers changing though won't work as that requires a level of verbal analysis.

And I agree that does not mean that analogue gauges are always right - I lost a friend due to a misread of an old 3 needle analogue altimeter.

mm43

29th Aug 2011, 22:15

Originally posted by HarryMann ...
...I personally find the 'go up'; 'go down' comments somewhat strange for pilots ...

The French version uses words such as 'monte' (climb) and 'descend' (descend). I suggest that the BEA were under time constraints and the translation wasn't peer reviewed by AAIB / NTSB. The PF / PNF obviously knew what they were talking about as the odd English word was inserted into a French string.

HarryMann

29th Aug 2011, 22:18

And I agree that does not mean that analogue gauges are always right - I lost a friend due to a misread of an old 3 needle analogue altimeter.

Ah, but that would be quite another matter surely... training and familiarity rather than sensory overload ?

ChristiaanJ

29th Aug 2011, 22:39

mm43,
I think you're right.... too many people read too much into the translations.

HarryMann,
"an old 3 needle analogue altimeter".
IIRC, they killed several people....
IMHO, the combined 'drum' (for the 'real figure') and one 'needle' (for the 'trend') instrument was the best solution.
I'm not sure whether the current PFD implementation is "as good".
I would like to see pilots' opinions on that, but if anything it's slightly O/T.

rudderrudderrat

29th Aug 2011, 23:04

Hi DozyWannabe,

... but if that is the case then why do we have the references to going up when they should be going down and to touch the lateral controls as gently as possible by the PNF?
PF takes manual control at 2.10.06.
From 2.10.7 - 18
"The copilot sidestick is positioned:
- nose-up between neutral and ¾ of the stop position
- to the left in half-travel position then to the right in half-travel position and twice, alternatively left to the stop position then right to the half-travel position (Period of 4 seconds)."

2.11.00 PNF says "Above all try to touch the lateral controls as little as possible eh"

PNF was either happy with PF's inputs for nearly 1 minute (doubtful) or unaware of the over controlling inputs.
That is one of the major disadvantages of separate Side Stick design.

HarryMann

29th Aug 2011, 23:25

That first minute, I imagine PNF just was not watching PF's inputs, or the result thereof... head-down flippig pages, maybe?

DozyWannabe

29th Aug 2011, 23:36

He starts giving guidance to the PF ("Watch your speed") at 2:10:27, so if he had his head down in the ECAM it wasn't for any more than 22 seconds, certainly less than a minute.

He starts making comments on the PF's flight path ("Stabilise, Go back down (descend)") between 2:10:28 and 2:10:31 - so he's starting to sound unhappy about 25 seconds after the PF takes control. He makes a direct reference to lateral controls at 2:11:00, as you say - but he's clearly aware that something's not right 30 seconds before that.

Remember that the rules state that the non-relief pilot is not to take control without the Captain's say-so under most conditions. This wasn't specific to AF's Airbus crews, it was fleet-wide.

PuraVidaTransport

30th Aug 2011, 01:14

Hope this is a helpful chart...made it for my use but figure others may benefit. I cut out the traces I wanted to analyze together and put them into one larger file. The yellow highlight is time when the stall alarm was active. Made it by superimposing the BEA trace of the stall alarm. I eyeballed the times on the quotes with larger lines indicating multiple seconds of conversation as detailed in the BEA report. Any errors, let me know.

lomapaseo

30th Aug 2011, 02:48

PuraVidaTransport

Hope this is a helpful chart...made it for my use but figure others may benefit. I cut out the traces I wanted to analyze together and put them into one larger file. The yellow highlight is time when the stall alarm was active. Made it by superimposing the BEA trace of the stall alarm. I eyeballed the times on the quotes with larger lines indicating multiple seconds of conversation as detailed in the BEA report. Any errors, let me know.

It looks like the biggest error was the image hosting service you are using. I got some interesting re-directs to XXXXXX sites

But was finally able to capture your hard work so here it is for anybody else

http://fromtheflightdeck.com/MEL/PPRune/af447trace3.jpg

grizzled

30th Aug 2011, 02:49

RE: Too many people are grasping at straws and thinking that somehow the machine or the warnings had a big impact on the accident. It didn't, the PF was the one who made the NU inputs and it wasn't because of cognitive overload and not hearing the stall warnings - there were three pilots. The PF panicked and went into shock pulling back on the stick. That's all there is to it

Though I agree with much of what you say in the post that contains the above quote (and other posts you've made), I have some difficulty with your assertions -- based on the facts as we know them so far -- that: A. The aircraft's instrumentation (or lack thereof) was not a contributing factor, and B. The source of the PF's actions was "panic".

If we are investigating accidents today for the purpose of learning lessons so as to avoid, or minimise the chances of, making the same mistakes again, we have to be very careful not to over simplify what happened and why; especially when it comes to the human. Of course the actions of the PF (primarily), the PNF (a close second), and even the Captain, are the prime "causes" (in terms of specific actions at the time) of the loss of the aircraft. But, if you are saying that there is nothing to be learned here about instrumentation, display presentation, and human behaviour, in addition to your points about AF procedures and training, I think you're missing out on the notion that the ultimate goal of investigators is to try to get as close as possible to the "why" (actually "why's"). In that light I believe there is more we can learn from this event, including in the areas of displays, warnings, controls, etc; as well as the obvious ones such as CRM, procedures and training (both in terms of AF specifically and industry wide).

PuraVidaTransport

30th Aug 2011, 02:58

Sorry about that Lomo, the three or four times I checked it (and even now), never saw any porn or links to porn. Perhaps the link I used just works for me and my cookies OR you have some "interesting" cookies of your own ;)

If you will PM me (or perhaps a moderator) with the url that you used to fix the problem, I will edit my original post.

Be honest, I just did a Google search for an image host after noticing I couldn't just attach a photo file here. Found that one (postimage.org) first and admit, know nothing about them...all my fault and my apologies to any offended.

Old Carthusian

30th Aug 2011, 03:27

Grizzled
The perceived existence of a problem with instrumentation or displays misses the point. Humans have been misinterpreting instrumentation since it started being used and with any redesign (which may indeed be welcome) you will find that someone will find a way to misinterpret it. Can it be improved? - well, yes, everything can always be improved. Will it eliminate accidents like this one? - highly unlikely. It's just this, know what you are using and what it tells you.

One must also be careful not to attribute any negative connotation to the term 'panic' which was used in a purely technical sense. The nature of the state of panic and of acute stress disorder (shock - which are what I believe happened to the PF) include a form of tunnel vision in which almost every other stimulus or input is excluded except that which the person in the situation chooses to focus on. This kind of reaction shuts down verbal cognitive channels pretty effectively. It can be countered by reliance on procedures - following the procedures is a very good way of getting out of the state. The other option is training. If you are well trained you are less likely to be prone to this type of reaction (Hard training, easy execution).

With the PF though I am very sceptical that improved or different displays would have changed his reaction though this can never be known for certain. Knowing his machine more thoroughly and following the SOPs coupled with sound CRM would have been more likely to help the crew avoid the accident. Flight deck management procedure strikes me as a bit too casual. The cultural issues that arise though are much more complex - the physics or aviation issues are fairly easy to understand. However, the human factors need much more careful consideration than I believe they have been given by AF (I do believe we are on the same page on this).

Whatever the level of automation, the pilot is the person responsible for his aircraft and needs to understand and react appropriately. The airline needs to ensure that the pilot is given the support and training to do so and the CPD (Continuing Professional Development) throughout their career. The automation logic employed by Airbus is to my mind perfectly logical and suited for the environment it operates in. In certain circumstances it does remove itself but those are conditions for which it would be inappropriate. It is designed as an aid not as a replacement but it does look as if some pilots and some airlines are treating it as the latter not the former. This I need hardly mention is dangerous.

grizzled

30th Aug 2011, 03:53

Well said; as you say, we are on the same page on most points I think.

I too believe that "culture" in this case played a most significant role. And I use that word in all the senses it is used in today's Safety Management writings: national, corporate and professional.

Re your last point: It is designed as an aid not as a replacement but it does look as if some pilots and some airlines are treating it as the latter not the former. This I need hardly mention is dangerous.

On that, I could not agree more. Many aviation safety professionals are suggesting (rightfully, IMHO) that we are just now seeing the beginning of the next major "trend" in accidents, which can generally be described as: LOC incidents resulting from a (widening) gap between what is required from the crew in an emergency situation involving loss of some significant aspect or component of computerization or FBW, and what the crew is adequately / appropriately trained for and experienced in doing.

mm43

30th Aug 2011, 04:14

The following is a combination of FDR traces posted in a format used to convey the data that HazelNuts39 had originally constructed from the BEA's initial report on the FDR data.

These copied traces have used the highest resolutions available and also have a Smoothed Normal Acceleration curve added to give an indication where the +/- 'g' forces were. From the commencement at 02:10:51.4 of the Stall Warning, thin dotted lines showing the 'smoothed' traces for both the AoA and Pitch Attitude angles are included.

CLMAX occurred at 02:10:56.5, and the 'g' break within a further second. Some post stall pitching oscillations matched by the AoA values took place for a few more seconds before the AoA rapidly moved toward 30 degrees and the FPA moved into negative territory.

Back earlier at the AP and ATHR disconnect(s), the traces are rather 'lumpy' - an artifact of the turbulence occurring as the aircraft commenced the initial climb.
http://oi55.tinypic.com/vr55aq.jpg
The Initial Stall Analysis is here (http://oi56.tinypic.com/9pt3ds.jpg), and HN39's description of how the initial data was constructed is in AF447 - Thread No.5 - post #70 (http://www.pprune.org/tech-log/456874-af-447-thread-no-5-a-4.html#post6564645)

rudderrudderrat

30th Aug 2011, 07:38

Hi DozyWannabe,

Remember that the rules state that the non-relief pilot is not to take control without the Captain's say-so under most conditions.
I've never heard of that.
PF / PNF can swap roles at any time - usually one crew member is allocated the pleasure of doing the take off and landing for his records. Unusually with this crew was their relative seat positions. The more experienced senior Co-pilot was not in his natural seat (Co-pilots occupy the RHS), and may have felt some anxiety about taking control and using his non natural hand on the side stick.

DozyWannabe

30th Aug 2011, 08:14

@rudderrudderrat

Read the 3rd interim report. This was a very specific procedure within AF for a long-haul flight whereby the crew consists of a Captain, F/O and a relief pilot who is not a Captain. The Captain is supposed to brief the rest of the crew as to roles and responsibilities (including parameters for taking control) prior to the flight or at handover. The record suggests that this was not done.

I honestly don't think you're going to find any technical reason for this - handedness is not an issue, and as it seems relief pilots can sit in either seat it's not like F/Os get no practice with it on that side. I know some people want it to be about the sidestick, autotrim and lack of feedback but the odds are stacking up against it.

rudderrudderrat

30th Aug 2011, 08:27

Good Morning DW,
...handedness is not an issue, and as it seems relief pilots can sit in either seat it's not like F/Os get no practice with it on that side.
All the hand flying practice will have been done from the co-pilot's natural position (RHS). During cruise relief, when seated in the LHS - all inputs will have done through the auto-pilot.

Occasionally, in the simulator the co-pilot will get an opportunity to hand fly from the LHS - and from what I've observed, some find it much more difficult. Try playing tennis with your non natural hand and let me know how you get on.

The side stick is a perfect device for controlling the aircraft in Normal Law (nudge to attitude and let go), but is not ideal for controlling Roll in Alternate Law (it's too sensitive and needs constant pressure until the aircraft is perfectly trimmed).

The conventional control column is able to accept both hands and is more natural than the side stick from either seat.

Edit. If there was so little difference between the handedness of the side sticks, seat positions etc., why do you think it is mandatory for Training Captains (who can occupy either seat) to additionally perform EFTO & GA procedures from the co-pilot seat in the simulator? If there was so little difference - we wouldn't bother.

DozyWannabe

30th Aug 2011, 08:47

So what you mean is "I prefer a conventional control column, so I'm going to keep blaming the sidestick design wherever possible".

That's fine, we're all entitled to our opinions - just don't pretend you're being objective when you've already decided you're going to fit this accident around your personal bugbears with the system.

rudderrudderrat

30th Aug 2011, 09:49

Hi DozyWannabe,

So what you mean is "I prefer a conventional control column"
under certain circumstances like direct law - yes. Otherwise conventional aircraft would have adopted side stick controls for weight saving and reduced engineering complexity years ago.

This AC series is designed around Normal Law - and the side stick works brilliantly well when used in this law. If the autopilot had simply disengaged with Normal Law still active - PF would not have had any difficulty keeping it straight and level.

I don't have any personal bugbears - I'm trying to explain why I think the PF had so much difficulty in hand flying in ALT LAW at high altitude.

RetiredF4

30th Aug 2011, 11:39

mm43
The following is a combination of FDR traces posted in a format used to convey the data that HazelNuts39 had originally constructed from the BEA's initial report on the FDR data.

Very interesting indeed.

Pay attention to the smoothed normal acceleration curve:

02:10:05 - 02:10:12, 7 seconds, load increase from 1,0 g to 1.4 g
02:10:12 - 02:10:40, 28 seconds, load decrease from 1.4 g to 0.8 g
02:10:40 - 02:10:55, 15 seconds, load increase from 0,8 g to 1.1 g
02:10:55 - 02:12:00, 65 seconds, load decrease from 1,1 g to 0.9 g

What the crew really felt despite their SS actions, was the initial overshooting in g in the first 7 seconds, then the correction phase for 28 seconds, then a increase again to 1,1 g in 15 seconds, followed by going below 1 g in the following 65 seconds.

Caught of guard by speed loss and AP/ATHR dropout combined with lack of training and stress might have increases the danger to follow the feel in the pants instead stick to the instruments, like we climbed too much, but now it feels like leveling.... even descending when below 1 g and do not disturb the cabin too much.....

Loose straps with impaired seating position (we do not know, wether the seatbelts of PF separated due to impact or if they had been donned inproperly). might have influenced the resistance to unload further. Best seen here also, that the erratic stick inputs of PF (after initial pitch) does not influence the big picture concerning g-loading that much. When we talk about too much backstick pressure after the initial pullup and compare it to the g loading, then the resulting aircraft response does not represent the SS inputs that clearly.

Talking about traces, the post from lampaeso (http://www.pprune.org/tech-log/460625-af-447-thread-no-6-a-31.html#post6670269) is an interesting compilation of traces with cockpit communication, thanks very much for that work.
What struck me most is the fact, that the pitch came down below the horizon in several timeframes, and mostly in combination not to SS pitch ND inputs, but to big bank angles to the right in combination with full oposite SS input and even SS full NU inputs.

02:11:55, 25° RH bank, SS full left, SS full NU, pitch -10°
02:12:10, 20° RH bank, SS full left, SS full NU, pitch -10°
02:12:50, 40° RH bank, SS full left, SS half NU, pitch - 5°
02:13:00, 40° RH bank, SS full left, SS half NU, pitch -5°
02:13:55, 30° RH bank, SS full left, SS full NU, pitch -5° (double input)

Watch also the stall warning and VVI during these occasions as well as the CVR.

RatherBeFlying

30th Aug 2011, 16:50

We have been critical of PF for stalling the thing, but as RF4 has pointed out he was trying for a couple minutes to stay right side up with the a/c seemingly not responding to left stick.

A couple weeks ago I got a ride in the back seat of a Romanian Lark glider with a long thin wing. Thermaling the beast past 30° bank found me with the stick in the opposite rear corner.

Had to unload a bit to get some more aileron authority. Many gliders lose aileron authority before they stall from a steep bank in a thermal and I am at a loss that an experienced glider pilot would not have been familiar with this.

As mentioned many times before the transition from an apparent overspeed problem to a stall problem was not recognised.

HarryMann

30th Aug 2011, 18:11

I know some people want it to be about the sidestick, autotrim and lack of feedback but the odds are stacking up against it. I don't think you should polarise this so much

Few, if any, are denying that PF was way below par from manual takeover to end of story... with general crew shortcomings becoming apparent.

But really, you have to stop this black and white routine, any engineer will tell you very few solutions, much in nature and certainly not accidents are so black & white

All those 3 things did play some part (notwithstanding similar accidents in the past happening without them - that is a false negative), I believe.

To start slamming a control column about like that (just think of the equivalent), would definitely have woken up the Captain upon entry, even if PNF had still stayed subdued (this accident will no doubt turn into a study case for at least pilot shock & stress behaviour as much as CRM itself).

Indeed, AF should be consulting the RAF crew selection centre (as was) about some help?

There is almost no argument that Captain would not have responded to visually seeing a full aft stick in PFs hands - end of.
No doubt he would then have looked straight at the Trimwheel, and PFD

My guess Captain had no idea that PF (had been and) was still obsessed with an UP/ DOWN mentality, and at night, was totally removed from any sense of wind over wings and AoA (something even crude Q-pot feedback could provide) :ugh:

I think PF having flown sailplanes is very much a red herring.. he was no Sully.
.

grity

30th Aug 2011, 18:20

mm43, your nice graphic.... what can be the reason of the escalate pitch- (and AoA-) change, frequency whith up and down within 5 sec ... (in the beginning of stall after 2:11:00.....)is it turbulenz from outside or is it a typical self-induced pitching movement of an A330 with higher AoA?

if one will learn to find a way out of stall..... one need to understand this.... and learn to use it....

mm43

30th Aug 2011, 19:00

Originally posted by grity ...

.... or is it a typical self-induced pitching movement of an A330 with higher AoA?Hello grity,

They were the test pilots taking an A330 into its first stall, and the period you have referred to is what I have called the 'PRE STALL'. I believe that pitching oscillation occurs as the wing starts to lose lift and the nose drops, then momentarily recovers lift and the nose pitches up. The fact that the elevator and THS were max NU may also have had a bearing on this oscillation.

IMO the pre stall buffet the BEA tried to define is not very clear, but the pitching oscillation in hindsight appears to be the very last clue.

grity

30th Aug 2011, 19:51

mm43, ok but
the elevator is not at max NU in the "pre stall" it is just oscillating between -2 and -7 deg, there is also much SS with pull and push, but I am not shure if we can decide if the oscillation is caused by the elevator or by the move of the center of lift back and forth with the losing lift or all together with turbulence ...... it is realy not easy to discriminate cause and effect....

HazelNuts39

30th Aug 2011, 20:16

mm43;
Thanks for another great graphic! I'm wondering what you mean by 'pre stall'. At about 2:10:56 the "gee" starts to fall while AoA is increasing through nine degrees. Isn't that an indication of stall?

mm43

30th Aug 2011, 20:50

grity;

Ok. The center of lift is moving, and the the cause of the pitching may be the 0.3g peak to peak shown in the Normal Acceleration trace. I suspect it will be a point of further discussion.

HN39;

Rather difficult to determine exactly what was going on. Your point is a good one, though it also seems with the synchronized oscillation of the pitch attitude and AoA angles that lift hadn't been lost.

What do you believe the Alpha Max was at 02:10:56?

I'm quite happy to change the indications on the graphic if a consensus can be reached.

DozyWannabe

30th Aug 2011, 21:43

I don't think you should polarise this so much

Harry, I'm not polarising anything - in fact I'm being neutral to a fault. You and others are saying that feedback and/or yokes would definitely have changed the situation, but I'm saying that isn't necessarily the case. I'm not saying it definitely isn't, or that feedback is purely a throwback. I'm saying there's not enough evidence to say that feedback would have made a difference.

Few, if any, are denying that PF was way below par from manual takeover to end of story... with general crew shortcomings becoming apparent.

I'm not even saying that! Everyone has bad days at the office, it's just that a bad day at the office for an airline pilot has an almost unique potential to become a bad day for everyone on their aircraft.

To say that lack of feedback may not have been an issue is not blaming the pilots, and to say that one necessarily implies the other is a strawman argument that is beneath a lot of the more reasonable people on here.

But really, you have to stop this black and white routine

Harry, can't you see that I'm the one trying to keep the shades of grey here by saying that the sidestick/passive feedback combo may not have (not didn't - *may not* have) played a causative role here.

All those 3 things did play some part (notwithstanding similar accidents in the past happening without them - that is a false negative), I believe.

That is your opinion, shared with some others - and as an opinion it is valid. My opinion (that it's a possibility, but equally that it could be a red herring), is also shared by some others and IMO equally valid. I'm not trying to argue here, I'm just trying to make sure that the parameters of debate remain open.

To start slamming a control column about like that (just think of the equivalent), would definitely have woken up the Captain upon entry, even if PNF had still stayed subdued (this accident will no doubt turn into a study case for at least pilot shock & stress behaviour as much as CRM itself).

There is almost no argument that Captain would not have responded to visually seeing a full aft stick in PFs hands - end of.
No doubt he would then have looked straight at the Trimwheel, and PFD

I disagree. See the posts on the Stony Point NWA 727 crash in the other thread. A three man crew *did* misdiagnose a stall as overspeed, *did* move the control column to its full extent, *did* stall the plane and *did* crash and die.

*You* may have picked it up in that situation, but there were and are trained pilots who aren't as good or as knowledgeable as you.

HarryMann

30th Aug 2011, 21:59

DW

Right, :ok: apologies then...

I wonder if yourself or anyone else could explain this to me:

I think I'm overspeeding, so I'm holding the stick back, I've tried a wedge load of thrust (oddly) and also now reduced it somewhat (more sensibly)

I'm also holding the stick almost fully to one side, then the other to keep the wings level, as they have a habit of rolling around in a rather airy fairy floppy fashion

Why in an overspeed situation, would this be happening, when the norm is a spiral dive (banked one way or the other and increasing) or just laterally diverging (again banked one way) ?

DozyWannabe

30th Aug 2011, 22:35

Right, :ok: apologies then...

Don't mention it, and likewise, I'm sorry if I gave you the impression that I was being disrespectful at any point.

I think I'm overspeeding, so I'm holding the stick back, I've tried a wedge load of thrust (oddly) and also now reduced it somewhat (more sensibly)

I'm also holding the stick almost fully to one side, then the other to keep the wings level, as they have a habit of rolling around in a rather airy fairy floppy fashion

Why in an overspeed situation, would this be happening, when the norm is a spiral dive (banked one way or the other and increasing) or just laterally diverging (again banked one way) ?

I don't think the PF was ever completely convinced he was overspeeding, as I think he would have argued the toss with the PNF when the latter practically bit the PF's hand off when he reached for the speedbrake lever if he did. I suspect he didn't really know what was going on and was trying to throw out ideas (which is fairly common in a dynamic problem-solving environment). I also suspect that he was suffering encroaching tunnel vision from the outset and heard neither the "Alternate Law" callout from the PNF, nor the stall warning. As such, there's a little nagging suspicion in the back of my head that he was convinced that as long as he had TOGA selected, no amount of back stick could cause him to stall.

Another thing I'm not convinced of is that he was fighting a right-roll tendency that was entirely down to external forces. After the aircraft was stalled, yes the clues were all there, but at the apex of the climb and just prior he was still slamming the stick around. I wonder whether the initial right-roll was induced by an unfortunate combination of a pocket of turbulence at the same time A/P disengaged - and because of the magnitude of the inputs what we are actually seeing prior to the stall is PIO. While I'm not going to join the chorus of those saying that failure to release the raw FDR data is tantamount to wrongdoing, I do hope that with the final report will come some traces that are of a higher resolution in the time axis - especially around the time of A/P disconnect to the apogee. That way it should be possible to determine whether the PF stick inputs were excessive inputs to combat a serious roll issue or whether PIO had a part to play.

Again, remember you're clearly a pilot with some considerable technical and aeronautical knowledge under your belt. This is no slur on them, but a lot of guys on the line do not.

Linktrained

30th Aug 2011, 22:35

PF may have had very much less hand flying at cruising level than either PNF or the Captain. And on this occasion it would have been sudden, not premeditated - and with a limited panel. (Only the PNF's instrument panel is recorded. it is assumed that the PF's panel would have been the same.)

When I asked how often pilots hand flew an A330, (let alone at cruising level, where some replies suggested that it requires more concentration) the responses were somewhere between "not if I can help it" to " my Company does not allow me to ..." ( Interesting to know the logic.)

All that I have seen of PF's gliding experience was that he had held a gliding licence some years ago. That could be quite irrelevant - or not.

Someone may know just how accurately Simulators reproduce an A330 's performance at cruising level. How easy is it to hand fly when some of your instruments failed, even before you have started ? Is the Simulator stable in rough air, or do you have to struggle to keep the wings level ? How often is this practiced ?

This kind of training was usual using a Link Trainer, fifty or more years ago, even in rather minor airlines.

One of the American trunk carriers in about 1950, was in the middle of a pay dispute with their pilots and nearly ordered their next batch of aircraft to be made without autopilots... "Make them work..." Somebody said. (They had bean counters, too !)

xcitation

30th Aug 2011, 22:52

I don't think the PF was ever completely convinced he was overspeeding, as I think he would have argued the toss with the PNF when the latter practically bit the PF's hand off when he reached for the speedbrake lever if he did.

Yes PF did take the extreme step of deploying the airbrake. Hence PF actions were consistent with him being convinced of overspeed at that point.

2 h 12 min 04
-> 2 h 12 min 07 The airbrakes are controlled and deployed. I have the impression that we have some crazy speed no what do you think?

2 h 12 min 07 29,736 The angle of attack 2 is temporarily valid at 41°.The stall warning is triggered. No above all don’t extend (the) SV : “Stall, stall”

Clandestino

30th Aug 2011, 23:01

Still want folks to know that the human "body rates" can help for the first second when something unusual happens.Absolutely! Feelings of acceleration or turning are not to be taken lightly. Without any doubt they have to be crosschecked against instruments to see whether they are in accordance with the aeroplane's maneuvers or they're just illusions.

Captain my great Captain, you may as well forget about any mechanical link for the AF447's rudder. Because you say so? Because Airbus has radically redesigned rudder architecture on 330s since a couple of years ago? Because someone is unable to appreciate the difference between "controlled" and "operated" when discussing the flight controls as in "manually controlled and hydraulically operated"?

It could be said that the only reason that pilots are required is for when the automatics fail.Of course it could be said but it would be a lie. There's no automatic decision making. There is no automatic take-off. More correct version would be: pilots must be able to manually fly the aeroplane out of any situation automatics put it into.
Several posts here have statements like: "The simulator does not perform like the real aircraft after a stall" or "the simulator cannot behave as the real aircraft does outside the normal flight envelope" and even "the simulator does not fly like the real aircraft in Alternate Law"

So the simulator is being used to train pilots in how to handle emergencies, upsets and LOC, despite actually not behaving like the real aircraft in emergencies, upsets and LOC?

The current approach of only using simulators for training pilots to fly simulators does not seem extremely logical. I am sure that the accountants love it.
What a slippery slope! Simulators don't simulate the areas that we know nothing about because no flight testing was done (or required to be done) in them. They have to very realistically simulate stall behaviour but to check what the transport aeroplane would do if someone stubbornly insisted in pulling into stall would be asking far too much from the test pilots. Statement that simulator doesn't realistically simulate aeroplane's behaviour within normal flight envelope or different control laws is also untrue and that leaves your idea that we train in simulators just to fly simulators completely baseless.

It's not just accountants that love using simulators for training. Pilots, insurers and anyone living under the training zone appreciate it too. Face it: you want to go far, high, fast and with plenty of payload. There's no way in our Newtonian world to make an aeroplane that would do all that and still have the stall behaviour of Cessna 172. People get killed by stalling 172s, anyway.

(...) it is because the human brain has a limited number of cognitive channels and they can only handle ONE input at a time.That's the best explanation of someone's inability to chew gum and simultaneously tie his laces I've heard in a long time. Attention distribution, proper scanning and avoiding the fixation are basic pilot's survival tools, taught and checked from day one at the controls. If you suck at them, tough luck, you can not be a pilot. What AF447 crew did was the sign of the severe incapacitation, not usual and ordinary human behaviour.

The reason that 'steam gauges with needles' seem to be easier to read is that they are a spatial cognitive load and form patterns that can be recognized without much cognitive effort.
(...)
I should have made it plainer in my wording.
When the verbal cognitive channels are overloaded you can add more to them and all you do is make the effect worse. However, the analogue gauges do not require any verbal cognitive analysis so that channel is unloaded. So you can see a VSI needle move hard down and understand what it means even if the verbal cognitive channel is in overload - numbers changing though won't work as that requires a level of verbal analysis.
Speed and altitude tapes might be digitally driven but their presentation is analogue! There might be a digital readout added but there's no need to read it while performing normal instrument scan. Patterns on modern EFIS can be recognized without much cognitive effort and you might be surprised that Airbus EFIS does include VSI needle.

I have never heard about real life pilot that had problems switching from round dials to tapes (me included). I have came across those who made painless transition from sidestick and tapes to round dials and yoke, so when reading complaints about modern aeroplanes' cockpits, I am left with the distinct what-the-heck-are-they-talking-about feeling.

One of the aspects I expect the BEA Human Factors investigators to look at is the cognitive workload that the ECAM and failure messages put on the pilots. It is very basic task sharing: one pilots flies, the other takes care of ECAM. It was very similar before ECAM, since we started putting two guys up front. One would fly, the other would execute the checklist. Eastern learnt the hard way what happens when everyone goes troubleshooting and no one is minding the store. So Crossair.

Perhaps every potential emergency scenario should be subject to what is called a 'cognitive walk-through' that actually assesses the cognitive loads and identifies likely overloads.And then what? Prohibit by decree emergencies that overload the crew? There were thirty two cases of UAS on A330/340 fleet before AF447. Question is what made the AF447 crew overloaded and not other 32.

Older pilots may well have followed a rather older but repeatedly successful dictum - disregarding all the cacophony - aviate (i.e. pitch and power), navigate, then communicate.New pilots too - not following that old dicta is guaranteed to shorten your operational life even on modern jets with all the electronic bells and whistles. You are very unfairly pitting the ideal pilot of yesterday against the single modern crew which tragically underperformed.

ChristiaanJ

30th Aug 2011, 23:06

Someone may know just how accurately Simulators reproduce an A330 's performance at cruising level. How easy is it to hand fly when some of your instruments failed, even before you have started ? Is the Simulator stable in rough air, or do you have to struggle to keep the wings level ? How often is this practiced ?That question has been asked before (I've asked it) and so far I've not seen a clear answer. The consensus (from NASA reports etc.) seems to be that even FFS systems just extrapolate from the limits of the flight envelope. To what extent those extrapolations are 'real' is still an open question.

This kind of training was usual using a Link Trainer, fifty or more years ago, even in rather minor airlines. Let's be honest.... A Link Trainer (and yes, I've "flown" one) did not yet really attempt to fully simulate the flight characteristics of a specific aircraft type, neither would it have been capable of doing so.

DozyWannabe

30th Aug 2011, 23:14

Yes PF did take the extreme step of deploying the airbrake. Hence PF actions were consistent with him being convinced of overspeed at that point.

Agreed, but he doesn't seem convinced enough to press the point with the PNF when he chews him out for it and shuts down that avenue. The PNF's immediate following action is to defer to the Captain.

xcitation

30th Aug 2011, 23:29

Originally Posted by Ian W
Older pilots may well have followed a rather older but repeatedly successful dictum - disregarding all the cacophony - aviate (i.e. pitch and power), navigate, then communicate.
New pilots too - not following that old dicta is guaranteed to shorten your operational life even on modern jets with all the electronic bells and whistles. You are very unfairly pitting the ideal pilot of yesterday against the single modern crew which tragically underperformed.

PF/PNF did fly pitch and power but it was for the low altitude UAS as they had practiced in the sim, Rio departure, 3 months prior to the incident. The key difference being the high 15 deg pitch at low altitude UAS compared with 5 deg pitch high altitude (now revised 2.5 degree). No high altitude UAS was trained for in the sim. Therefore it would appear they both applied the nearest training they had in the abscence of specific training.

JD-EE

30th Aug 2011, 23:33

Olr Carthusian

Most likely their training did fail them, very badly and fatally. In addition the displays lead to confusion. And data that might have resolved that confusion was either not presented to them at all (AoA) or was presented on a display that was not a part of their usual instrument scan - pure inertially derived speed, attitude, and so forth.

The primary cause of the whole chain of events was PF pulling NU on the stick when he should have very carefully stabilized roll without touching pitch. After that he seems to have become fixated on the concept of the plane going too fast rather than that it was actually stalled. He had no experience in training or real life with the situation into which he was thrust. And communications between crew members never cleared up the situation. Contributions by PNF were ignored. PNF tried to take over. PF grabbed back control and the plane and its freight died.

Communications were inadequate to absent.
Cockpit Resource Management was sloppy.
Data was inadequate.
Training was inadequate.

That set of inadequacies teamed up as a perfect storm of inadequacies to send AF447 and its passengers to the bottom of the ocean.

I say inadequacies rather than other words because on the whole all three factors are sufficient that the vast majority of flights take place uneventfully. And I suspect fixing even one of the problems might have saved AF447. All four should probably be fixed for best results. (Fix 4 and both 1 and 2 should end up repaired in the process. Fixing 3 is a major change to get it right. But they should be able to at least show an icon if a plane in a stall to get the message across through a non-verbal channel as well as the verbal warnings.)

chrisN

30th Aug 2011, 23:37

There have been several references in these threads to the PF’s previous gliding experience, most recently post #625 by Linktrained, and also several suggesting that gliding training would be good for all pilots.

I cannot claim any knowledge of CAT flying, but I do know something about glider flying and gliding training, and also something about gliding accidents and what surviving pilots have said on occasion.

For the record, not all glider pilots take to their training as ideally one should. One pilot I talked to, who crashed following a cable break and spun, luckily surviving, told me he knew the tail had fallen off so that there was nothing he could do. (IMHO, like PF in AF447, he became fixated on a problem he didn’t have, and never diagnosed what he was actually dealing with.) When the nose went down, he “knew” that trying to raise it by pulling back on the stick would not work, because of his “missing tail”, but he tried anyway. When it did not succeed, and he saw the ground spinning round (his words – still did not connect it with a spin), he let go and the glider recovered from the spin itself. Hence his survival, in spite of failing to recognise what he should have been trained to understand.

Good basic training in flying, including stall and spin awareness, is no doubt an ideal that would stand any pilot of fixed wing aircraft in good stead. Unfortunately, there are some pilots who do not learn it well enough. To paraphrase Dozy, I am not saying that this was or was not the case with PF; only that his prior gliding experience (How long ago? How extensive? How good was he?) is not necessarily a useful piece of information.

JD-EE

30th Aug 2011, 23:43

HarryMann said, "No problem, miniature lightweight LED ribbons/tapes are now being used in many applications, the batteries being small enough to be encompassed within the string I should think... go on, give it a go"

Why not get a small little LED flashlight and shine it at the window when you want to see the string. It should reflect back quite nicely. Shine the light mostly upwards at the string to avoid direct flashlight reflections into your eyes.

Now, what kind of chewing gum will you use to affix the front edge of the string to the window? I suspect common transparent tape would not do to well, nor would common duct tape. Umpty degrees below zero is not compatible with most adhesives I know that won't become a permanent part of the window.

DozyWannabe

30th Aug 2011, 23:46

No, there are plenty of other reasons for not doing that... one being that the aircraft simply 'could not be put in that situation'

Harry, I don't mean to nag like an old fishwife, but what is your source for that assertion, and if you do have one, how old is it?

Airbus themselves have been perfectly willing to accept the possibility of their FBW aircraft being "stall-capable" in degraded control laws, and have been quite open about it since at least the mid-'90s.

HarryMann

30th Aug 2011, 23:50

Let us be clear...

EASA Airbus knows closely what a very high AoA A330 behaves like, did and does... assertions that a decent simulation of beyond the stall behaviour, couldn't be made... Pah!

Wouldn't but not couldn't... lateral and directional might have had to be a bit of a fudge, but plenty of high AoA data exists even there. CFD tunnels as well as work in many different (real) tunnels exists and a pretty reasonable model could have been built well beyond the stall, very accurate longitudinally including simulated buffet and no doubt predicted nodding or other type specific behaviour.

No, there are plenty of other reasons for not doing that... one being that the aircraft simply 'could not be put in that situation'
Another? Airlines wouldn't be interested in it...(read, pay for the extended sim or the time in it)

JD-EE

31st Aug 2011, 00:04

Be honest, I just did a Google search for an image host after noticing I couldn't just attach a photo file here. Found that one (postimage.org) first and admit, know nothing about them...all my fault and my apologies to any offended.

Just a general note - spammers and malware authors are getting depressingly good at manipulating Google results. So sometimes the first one or even two results can lead to bad sites. Doing a quick check with the Linux "whois" tool shows that postimage.com seems to be the real site. Postimage.org has a hidden registration. That is as often as not an indication of something questionable being hosted at the site. If you play with "http://www.dnsstuff.com/" over time you can develop some intuition about "good" vs "bad" sites.

Linktrained

31st Aug 2011, 00:17

Link Trainers did not attempt to simulate any ONE kind of aircraft. It was used to train many thousands of pilots in what it COULD do, teach and then PRACTICE some aspects of flying on instruments.

At that time, there was nothing better.

The Stratocruiser simulator had a model landscape with a "small" television camera, which then moved in sympathy with the pilots inputs to give a visual image, initially for the Captain, probably in black and white !

The point that I was trying to make, was that PF, (any PF ?) might have made a better show, if he had had practice. I do not know how many hours a year Simulators are used, 3,000, 4,000 or 5,000 hours - or more ?

If some bean counter queries the cost, just say 447.

xcitation

31st Aug 2011, 00:20

Quote:
Originally Posted by xcitation http://images.ibsrv.net/ibsrv/res/src:www.pprune.org/get/images/buttons/viewpost.gif (http://www.pprune.org/tech-log/460625-af-447-thread-no-6-a-32.html#post6672077)
Yes PF did take the extreme step of deploying the airbrake. Hence PF actions were consistent with him being convinced of overspeed at that point.

Agreed, but he doesn't seem convinced enough to press the point with the PNF when he chews him out for it and shuts down that avenue. The PNF's immediate following action is to defer to the Captain.

Yes. I agree that PF was confused at best. What concerned me was that he could have hit his head as stated in the BEA report, which has never commented on in this forum (unless I missed it).
Someone did mention Laural and Hardy as far as communications. I am incredulous/perplexed by their sloppy communications in such a grave situation. It looks like they got into some very bad verbal habits and could not switch into professional comms mode when things went wrong.

PNF was not much better at his role. He should be pulling out the correct high altitude UAS checklist and making some direction to the PF. None of them ever mentioned the word stall / décrochage even if only to dismiss it as being false. What more important warning do you have on an aircraft when it tells you that it is now a falling lump of metal and no longer a flying machine. At least someone should happen to mention it amongst all the bickering and confusion. If the house is on fire do you ignore it and carry on fixing a leaky faucet/tap?

Machinbird

31st Aug 2011, 00:26

CONF iture

31st Aug 2011, 00:56

Because you say so? Because Airbus has radically redesigned rudder architecture on 330s since a couple of years ago? Because someone is unable to appreciate the difference between "controlled" and "operated" when discussing the flight controls as in "manually controlled and hydraulically operated"?
'Arrogant' takes double 'R' I think but I’m not sure about 'Pretentious' is it double 'T' ?
Let me see, if rudder control is electrical does it mean it can be mechanically-controlled too … ?

grizzled

31st Aug 2011, 01:03

JD-EE,

Your post (#636) about spammers and malware, though it will be considered OT by some, contains very helpful insights for those of us less knowledgeable in these things. Thanks!

OK465

31st Aug 2011, 01:08

...lateral and directional might have had to be a bit of a fudge, but plenty of high AoA data exists even there.

HM:

High fidelity simulation of post stall 'beta' effects (in all configurations) is an area of concern that is currently being aggressively addressed...

(by people much smarter than me. Like yourself. :))

Ian W

31st Aug 2011, 02:06

Dozy..
I'm not sure if human/machine cognitive overload is the problem here so much as the level of pressure and stress on the PNF. For all my opinion's worth as a layman with an interest, could it be the human/human interface that is causing the issue?

I think there is a mix of overload impacts that affect each crew member differently.

First there is what is called the 'classic inverted U curve (Yerkes and Dodson 1908) which balances arousal and performance. The crew were at the lowest point for attention and arousal at 2am in a relatively boring and benign cockpit environment. This is the worst level of attention and alertness the bottom of the inverted U and poor performance. Ideally, the crew need to be engaged so that their arousal/attention is at the top of the curve of the inverted U and performance is best. However, instead, what happened was a sudden burst of emergency activity both mentally and physically demanding with instruments and techniques routinely used in Normal Law not working. Learned sim ride behaviors seemed to make things worse. This moved the PF at least, right to the other end of the Inverted U where the level of arousal was so high that performance drops off. At that level the cognitive overload that we had discussed could lead to 'attentional tunneling' (or cognitive tunneling) where the PF concentrates on a few aspects of the task and is not conscious of anything outside those aspects.

The PNF has a similar jump to the wrong side of the U but as you state has a problem as he finds that what he says is not being listened to. This is probably due to the PF attentional tunneling, but PNF may think that it is some lack of authority on his part - the authority gradient. He may also be aware that he is overloaded with all the various aural verbal analysis, visual verbal analysis trying to read ECAM listen see what PF is doing now. So PNF wants the captain back. It is noticeable, that PNF is not making it plain shouting that the aircraft has stalled get the >deleted< nose down!! - no, PNF is muttering about the delay getting the captain.

When the captain does arrive, instead of a peaceful cockpit that he left, it is a cacophony of alerts and wind noise. PF is 'making mayonnaise' with the SS and presumably staring fixedly at the instruments. I did not see any transcript of a succinct briefing from PNF to the captain giving an order of events detailing the problems - only the captain being asked what did he think is going on? Was the captain even told that the aircraft was in Alternate Law? (is it immediately obvious if you are not told?) Would the captain have identified the stall if he had just a three sentence briefing on events?

I have said elsewhere that this turned into a human factors nightmare. Books will be written about this incident. Procedures and training both live and simulated will be changed I hope for the better.

gums

31st Aug 2011, 02:34

Not so sure the crew was just sitting back and enjoying the ride, Ian.

They had just told the flight attendants to be ready for some bumps, and they were turning a bit to left of course.

The human factors should play a large part of any recommendations, and the crew coordination/procedures should also be addressed.

GarageYears

31st Aug 2011, 02:36

The point that I was trying to make, was that PF, (any PF ?) might have made a better show, if he had had practice. I do not know how many hours a year Simulators are used, 3,000, 4,000 or 5,000 hours - or more ?

Most simulators are in near 24/7 use - the ones I typically work on once in service are most often on a 20 on, 4 maintenance cycle, so are running 24/7, generating revenue for 20 hours a day. Many times I have tried to book a slot to perform an update and found the sim has been booked full-up 24/7 for several weeks. There is a massive demand for sim time and not enough simulators. The problem is often there is time on the wrong simulator....

Ian W

31st Aug 2011, 02:53

Clandestino

Originally Posted by Ian W
(...) it is because the human brain has a limited number of cognitive channels and they can only handle ONE input at a time.
That's the best explanation of someone's inability to chew gum and simultaneously tie his laces I've heard in a long time. Attention distribution, proper scanning and avoiding the fixation are basic pilot's survival tools, taught and checked from day one at the controls. If you suck at them, tough luck, you can not be a pilot. What AF447 crew did was the sign of the severe incapacitation, not usual and ordinary human behaviour.

Each of the cognitive channels in YOUR brain can only handle one input at a time. There are discussions about how many cognitive channels there are but if you really think you can concurrently read a text, while repeating a well known rhyme correctly, listening to someone talking and accurately writing down what they are saying - then you are NOT human.

I suggest that you google 'Wickens cognitive channel' and read the research there much of which was carried out with pilots. You will see how easy it is to initiate cognitive tunneling in professional pilots.

It is the poor designs and procedures that do not take account of human cognitive limitations that can lead to accidents.

Machinbird

31st Aug 2011, 03:33

Ian W knows of what he speaks.

Any of you who think you can always multitask in a highly demanding environment have forgotten your early days of instrument training where it was easily possible to break down your scan.

We try to plan ahead to prevent overloads, but occasionally stuff can snowball.
There is no pilot alive who's scan cannot be broken by the right circumstances.

It is how you recover from this mess that separates the men from the boys.

Judging from how little actual stick time airline pilots seem to be getting, I am amazed that loss of control incidents are not more prevalent.

The term scan as I am using it refers to the multi-sensory process of acquiring information regarding your flight from all sources. It is very close in meaning to situational awareness.

grity

31st Aug 2011, 06:16

JD-EE Now, what kind of chewing gum will you use to affix the front edge of the string to the window? I suspect common transparent tape would not do to well, nor would common duct tape. Umpty degrees below zero is not compatible with most adhesives I know that won't become a permanent part of the window. airbus once had solved the problem to fix a vertical stabilizar on a fuselage.....

tape will work very well, but also this: open the window, hold a lace out, close the window....... will fix it for a test

yes please

RetiredF4

31st Aug 2011, 07:12

Clandestino
Patterns on modern EFIS can be recognized without much cognitive effort and you might be surprised that Airbus EFIS does include VSI needle.

What are the limits, within those the VSI needle displays vertical speed?
And what is the picture, when vertical speed is greater then the display available?
Could somebody post a picture with VS pegged?

02:11:50 PF: ........"I have a problem, it´s that i don´t have vertical speed indications"
..............Captain: ."OK"

HazelNuts39

31st Aug 2011, 07:20

What do you believe the Alpha Max was at 02:10:56?Alpha Max for M=0.65, the Mach number at 02:10:56, is 7.6 degrees. The lift coefficient and AoA versus time are shown in this graph (https://docs.google.com/leaf?id=0B0CqniOzW0rjMDU1M2MyODgtMGVhOS00NGJkLTg5YmMtYjBmZmM wMzczMmI4&hl=en_GB). The lift coefficient is calculated for the normal acceleration and airspeed calculated from ground speed and vertical speed in still air, which fits the recorded airspeed prior to stall. I can't explain the strange variations of cL and AoA after 02:10:58,4 when the airplane is evidently stalled.

RetiredF4

31st Aug 2011, 07:27

HazelNuts39
I can't explain the strange variations of cL and AoA after 02:10:56 when the airplane is evidently stalled.

Look at the bank angle. The aircraft was rolling and probably yawing and the ailerons had been deflected at the reversals of the roll from PF SS input. It is normal, that overall lift is reduced, and the wings will produce different amounts of lift.

HazelNuts39

31st Aug 2011, 09:48

RetiredF4;

That doesn't explain that the sum of two wings, plus tail and fuselage looks like this.

RetiredF4

31st Aug 2011, 10:10

That doesn't explain that the sum of two wings, plus tail and fuselage looks like this.

At least that what it feels like in real life (with different aircraft though).

In bank we loose lift on both wings, in a stalled environment with disturbed airflows anyway. Same with yaw. The bank reversal points correlate with the decrease in CL in that timeframe.

But i´m open for a better explanation.

HarryMann

31st Aug 2011, 10:18

What are the limits, within those the VSI needle displays vertical speed?
And what is the picture, when vertical speed is greater then the display available?
Could somebody post a picture with VS pegged?

02:11:50 PF: ........"I have a problem, it´s that i don´t have vertical speed indications"
..............Captain: ."OK" A very good question RF4

Let's hope for a good answer... whatver it is

Diagnostic

31st Aug 2011, 10:30

@RetiredF4:
Thanks for your perspective on things...
What are the limits, within those the VSI needle displays vertical speed?
And what is the picture, when vertical speed is greater then the display available?
Could somebody post a picture with VS pegged?

02:11:50 PF: ........"I have a problem, it´s that i don´t have vertical speed indications"
..............Captain: ."OK"
I was catching-up with the thread yesterday after some time away, and I saw you mentioned the VSI / VVI a few days ago also. I've searched for the info on the VSI display, but can't find it yet.

I remember reading (I think in a BEA report?) that when some barometric parameters are exceeded (including CAS < some value???) that VSI display is inhibited and a flag was shown instead of a value. So there were times when the PF had no valid VSI shown (my memory is that NCD was set in the SSM of the ARINC data, to indicate it was not valid, but I'm happy to be corrected when the actual reference is found). I'll look again when I have a bit more time.

This might also be what is shown on the graphs in BEA report 3 (English) pages 106 and 111 (but that is not the presentation which I remember reading) - we can see that at times, the recorded VSI is shown switching to zero, then back again to a high value. At the times when the FDR trace records zero that will be NCD in the ARINC stream since, of course, the VSI was not really switching rapidly to zero and back to a high value.

Personally, I think this intermittent VSI indication also played a part in the PF's behaviour. As well as intermittent airspeed indication, he also had this intermittent VSI and, at some times, intermittent (counter-intuitive) stall warnings e.g. ND = stall warning, NU = no stall warning.

We now know that these behaviours all had a single cause (low CAS inhibits some data & warnings), but he/they may not have made that "connection" (and may not have been trained to know that). With 3 different sources of data all behaving "strangely" to him, (not only UAS) perhaps he simply did not know that he could (and should!) trust the artifical horizon?

At the moment, I'm not sure whether the PF (and PNF) didn't hear the repeated "Stall Stall" due to cognitive overload, or whether they heard it but chose to not believe it, thinking it to be spurious due to having other (apparently) incomprehensible indications as I mention above?

@Ian W:
Just wanted to also say thanks for the human factors comments, very interesting and fits with my experience about cognitive overload. I do also see a significant variation between people in this respect (I overload more quickly than some other people I know - I can't give commentary to an instructor in full sentences without the flying suffering :ugh: - but other people manage it fine!).

HazelNuts39

31st Aug 2011, 10:35

What are the limits, within those the VSI needle displays vertical speed? Maybe this picture from a sim (https://docs.google.com/leaf?id=0B0CqniOzW0rjZjU2MThiZDMtZTYyNy00NGYwLTg1ODAtODgzN2Q zMTUyODE2&hl=en_GB) gives a rough idea?

AlphaZuluRomeo

31st Aug 2011, 10:37

PF/PNF did fly pitch and power but it was for the low altitude UAS as they had practiced in the sim, Rio departure, 3 months prior to the incident. The key difference being the high 15 deg pitch at low altitude UAS compared with 5 deg pitch high altitude (now revised 2.5 degree). No high altitude UAS was trained for in the sim. Therefore it would appear they both applied the nearest training they had in the abscence of specific training.
I disagree, here.
Had they followed the same drill as in their sim check (Rio depature) they would have :
- indeed seeked for 15° pitch attitude
- but also set TOGA at the same time.

TOGA was selected around 02:10:50, i.e. 40" after the NU first input, and at a time where the PF was lowering the nose, and just after the stall warning begun.

RetiredF4

31st Aug 2011, 10:44

HazelNuts39
Quote:
Originally Posted by RetiredF4
What are the limits, within those the VSI needle displays vertical speed?
Maybe this picture from a sim gives a rough idea?

Thank you, that´s what i´m looking for.

If i read it correct, it shows a descent rate of 1.800 fpm (number 18 besides the needle)

Would it look the same and would there be the number 100 for a 10.000 fpm descent rate?

Clandestino

31st Aug 2011, 10:55

What are the limits, within those the VSI needle displays vertical speed?
And what is the picture, when vertical speed is greater then the display available?
VSI needle is green and it beautifully simulates looking at the needle hinged to the right of the scale and being looked at through narrow slit - it appears to thicken with deflection from zero. Scale limits are 6000 fpm up and down, when they are exceeded needle is stuck at maximum deflection but it turns yellow as the digital readout adjacent to it. I don't recall any limitation to digital VSI value.

It's possible, though I can't say how probable it is, that VSIs blanked out sometimes during the descent. Someone with engineering expertize of A330 ADIRS and displays would be better suited to answer that than me.

Let me see, if rudder control is electrical does it mean it can be mechanically-controlled too … ? Rudder actuators are mechanically connected to pedals via mech mixer. Yaw damping, turn coordination, trimming and travel limit are achieved electrically. THS actuator is mechanically linked to trimwheels. FCOM chapter 1.27.10 and 1.27.30, part that describes mechanical backup refer.
'Arrogant' takes double 'R' I think but I’m not sure about 'Pretentious' is it double 'T' ?I couldn't have put it better myself, sir.

PF/PNF did fly pitch and power but it was for the low altitude UAS as they had practiced in the sim, Rio departure, 3 months prior to the incident. Seemingly they forgot everything they ever learnt about high altitude aerodynamics and performance. Actually, they needed not to know anything about it, it would be enough just to follow the procedure in order to survive. If your notion is proved to be true (which I mildly doubt) crew applied part of UAS drill that was not to be used above acceleration altitude (most of the time 1500 ft above runway) at FL330. I say it was incap.

Ian W knows of what he speaks. I prefer to have the opinion that he does not. It is much more comforting to think his biased views stem from the ignorance of the matters aeronautical, rather than from agenda. Basically he claims that the cognitive overload of AF447 crew makes the whole Airbus man/machine interface wrong and disregards tens of thousands normal flights made by FBW Airbi crews every day. He disregards 32 UAS that ended as very minor incidents. Applying such a hasty generalization is something that we have used to expect of politicians every day, yet it doesn't make it a lesser logical fallacy.

Any of you who think you can always multitask in a highly demanding environment have forgotten your early days of instrument training where it was easily possible to break down your scanMultitasking, whether it involves human or electronic brain, is not about doing many things at once. It is about properly managing cognitive/computational resources in sequential manner. Of course I can do just one thing at the time but in proper instrument scan, that one thing can be done a couple of times a second.

Pilots who don't show improvement of their instrument skills during their initial training to the point where they meet minimum applicable standards, don't get IR and are therefore unable to fly airliners. In a perfect world, anyway. Curves of pilot skill required vs pilot skill available are as important to flying as power required vs available. When skill required exceeds available, trouble is guaranteed.

Books will be written about this incidentTo borrow the phrase from Peter Garrison: most of them will be written by the folks with "movie rights" stenciled on their glasses.

Old Carthusian

31st Aug 2011, 11:38

JD-EE
There is no evidence that data presented to the pilots was inadequate and even if it was this is another training issue. There are procedures for dealing with loss of information which should be followed. It is not a problem with the aircraft or its displays.

I am interested in the idea that humans can only handle one cognative input at a time. Under certain circumstances this is true but only during times of stress. Outside such situations the human brain can handle several. The point of proper training, CRM and well defined procedures is that they put the pilot into a situation where the cognative input is controlled and directed towards solving the situation. All of these factors give the pilot the ability to overcome the narrowing of inputs and continue to function normally. Iain W's reference to Wickens MRT Model though could be thought of as misleading - it is a much more complex theory than the element presented and in fact as Wickens talks about people having problems when handling more than one task which requires the use of the SAME resource. This is slightly different to a statement that we are able to handle only one cognitive input at a time. Visual perception and auditory perception would in fact constitute two different resources. Thus conducting a proper instrument scan and hearing the stall warning would not be mutually exclusive. I do not feel that this can be an adequate explanation for the accident and in fact would even go as far as to doubt it especially as Wickens also notes that as workload increases performance can also increase.

Drmedic

31st Aug 2011, 12:52

:D
Dear Ian W!

I take your comment as one of the most significant in the af 447 case and on this forum. This what it all boils down to, a human error with huge consequences. Lost in 90 seconds. What really scares me is how close you are to catastrophe as a defenseless passenger when those ones who are paid for, respected, carefully selected and appointed to deal with just this type of situations, are so near their limits of capacity. Bear in mind also those "Close call", similar cases referred to earlier.

rudderrudderrat

31st Aug 2011, 13:18

Hi Clandestino,

You will have had cognitive overload on many occasions - but probably have not realised it. Have you never missed an RT call?

These pilots are suffering from cognitive overload - and all they are doing is flying an approach:
Whats that beeping sound? - YouTube

Those who don't appreciate that it is a human limitation should not be in aviation.

AlphaZuluRomeo

31st Aug 2011, 13:32

If i read it correct, it shows a descent rate of 1.800 fpm (number 18 besides the needle)
Yes.
From the FCOM:
The displayed vertical speed information is normally based on both inertial and barometric data. If inertial data is not available, it is automatically replaced by barometric information. In this case, the window around the numerical value becomes amber.
No mention of what is to be displayed when the barometric value is unavailable in the FCOM. I thought BEA wrote something about that in one of its interim reports, but cannot find it at the moment.

Would it look the same and would there be the number 100 for a 10.000 fpm descent rate?
Look the same : almost.
As V/S is > 6000 ft/min (climb or descent):
- the needle would turn amber
- the needle would stay at the end of the scale
As V/S is > 6000 ft/min (climb or descent):
- the analog display would turn amber, too
- no mention (FCOM) of a change of value/scale with this display for 10K ft/min, so I assume it would read 100.

jcjeant

31st Aug 2011, 13:42

Hi,

Hi Clandestino,

You will have had cognitive overload on many occasions - but probably have not realised it. Have you never missed an RT call?

These pilots are suffering from cognitive overload - and all they are doing is flying an approach:Same .. but with a tragic end at take off (at 6min10)
Accidente LAPA 3142 LV-WRZ - YouTube

CONF iture

31st Aug 2011, 14:18

Seems to be a fascinating movie. Thanks for the link jcjeant.
I haven't read the report, so I don't know, but it looks that the crew decided to voluntary ignore the warning, or link that warning to something else they could deal with without rejecting ?

CONF iture

31st Aug 2011, 14:28

Rudder actuators are mechanically connected to pedals via mech mixer. Yaw damping, turn coordination, trimming and travel limit are achieved electrically. THS actuator is mechanically linked to trimwheels. FCOM chapter 1.27.10 and 1.27.30, part that describes mechanical backup refer.
As I told Dozy, SmartCockpit (http://www.smartcockpit.com/) is not necessarily updated. I have well tried to put doubt in your mind … but what could we teach to a guy with the head of your size after all.

http://i45.servimg.com/u/f45/11/75/17/84/af447_17.png (http://www.servimg.com/image_preview.php?i=111&u=11751784)

Now, give a call to Airbus, they might reinstall a mechanic link on the rudder, just for you. Who knows ?

I make fun of you, Captain, but by reading your replies to others ... you fully deserve it. Reserve, modesty, humility … can be part of the picture too you know ?

Lonewolf_50

31st Aug 2011, 15:12

Clandestino:
More correct version would be: pilots must be able to manually fly the aeroplane out of any situation automatics put it into.

Cannot agree more. If one could get that written into an FAA training standard, and into design standards, and I believe three generations of airline pilots will thank you profusely. :)
It would be interesting to hear what this might do to training costs ... and thus ticket costs. Ya get what ya pay for, eh? :E
Ian W knows of what he speaks.
I agree in general, in terms of task load and cognitive channels.

I learned well how any crew can be eventually task saturated. It was done to me, and by me, in some of the more rigorous sim training events we used to do in my squadron, a lifetime ago ... some of the best training I got.

Any of you who think you can always multitask in a highly demanding environment have forgotten your early days of instrument training where it was easily possible to break down your scan.

Yep.
There is no pilot alive who's scan cannot be broken by the right circumstances. It is how you recover from this mess that separates the men from the boys.
Yep
The term scan as I am using it refers to the multi-sensory process of acquiring information regarding your flight from all sources. It is very close in meaning to situational awareness.
Machinbird, I respectfully submit that scan (instrument / vfr / combined) is a subset of situational awareness. It is a critical part of what informs your situational awareness, but not all of it.

A33Zab

31st Aug 2011, 15:23

A330 has 2 versions of the ruddercontrol system.

1/ original system with mech cable loop (like F-GZCP) is called - mechanical rudder - altough the cable loop is a backup system.

3/ the enhanced version without mech cable loop is called - electrical rudder - the backup is here electrical by means of the BCM (Backup Control Module).

Clandestino

31st Aug 2011, 18:51

You will have had cognitive overload on many occasions - but probably have not realised it. Have you never missed an RT call?I certainly did, but as a firm believer that aeroplane flies thanks to principles discovered by Bernoulli and not Marconi, I have never dropped the aeroplane to fly the mike.

Those who don't appreciate that it is a human limitation should not be in aviation. And what bearing on AF447 does it have? That it was inevitable? That we have to change instruments to fight it? That the 32 crews that didn't put themselves into stall when faced with UAS were just lucky? Of course anyone can be overloaded but simple loss of IAS resulting in cognitive overload (provided it's proven) is not something anyone reasonably informed of airline flying would be expecting to see. Besides, if cognitive overload resulted in stupor, aeroplane would probably fly itself to the other side of the ice clouds and chances are everything would be just fine afterwards.

As I told Dozy, SmartCockpit is not necessarily updated. I have well tried to put doubt in your mind … but what could we teach to a guy with the head of your size after all.

Now, give a call to Airbus, they might reinstall a mechanic link on the rudder, just for you. Who knows ?

I make fun of you, Captain, but by reading your replies to others ... you fully deserve it. Reserve, modesty, humility … can be part of the picture too you know ?

1/ original system with mech cable loop (like F-GZCP) is called - mechanical rudder - altough the cable loop is a backup system.
Thank you very much, gentlemen, for kindly answering my question whether Airbus has recently modified the A330 rudder. So it did, but seemingly not before F-GZCP left the Toulouse on her maiden flight. Seemingly we can not blame whatever happened to AF447 on alleged idiosyncrasies of the new electric rudder after all.

CONF iture, reservedly, modestly, and humbly claiming that e.g. Habsheim crash was caused by software protecting the aeroplane was never my cup of tea. I prefer to have my opinions as closely aligned to facts as it is possible and am not particularly caring about sugarcoating. I accept that many time I will be wrong but I also expect that I will be corrected by someone more knowledgeable than me. Way the corrections are delivered is of no particular interest to me. FWIW, I have no problem with them being extremely derisive, as long as I get facts straight. I have always preferred substance to style.

Perhaps a better way to put it would be:
"An aircraft that can not be demonstrated to be average line pilot worthy, shall not be certified as airworthy".That's one of the main points of airworthiness certificate since the first one was issued.

Diagnostic

31st Aug 2011, 19:21

mm43

31st Aug 2011, 19:26

Originally posted by HazelNuts39 ...

Alpha Max for M=0.65, the Mach number at 02:10:56, is 7.6 degrees.Thanks for the reply and the attached s/s graphic. Will have a good look at it shortly.

Lonewolf_50

31st Aug 2011, 19:45

Perhaps a better way to put it would be:

"An aircraft that can not be demonstrated to be average line pilot worthy, shall not be certified as airworthy".http://images.ibsrv.net/ibsrv/res/src:www.pprune.org/get/images/smilies/evil.gifhttp://images.ibsrv.net/ibsrv/res/src:www.pprune.org/get/images/smilies/smile.gif
Ventus, since most of the people who fly the 330 seem to like flying it, and seem to manage flying it, odds are that in this case, it meets that standard. :ok:

Zorin_75

31st Aug 2011, 20:18

That the 32 crews that didn't put themselves into stall when faced with UAS were just lucky?
Basically, yes.

Those previous 32 UAS events (which you've mentioned a few times now, in an apparent attempt to show that AF447 was the exception), were not all identical in their effects, and many of those crews also made various mistakes in handling them (some of which were the same as AF447 e.g. not treating it as a UAS event in the first place!). Look at BEA Interim Report 2, page 51 onwards (English version of the PDF) for the evidence.

Of course they aren't all the same, but one can't help noting a few similarities between the 13 incidences they examined more closely: IMC, cruise level, slight to strong turbulence, a/p disconnect, alternate law (in 12 cases). Sounds quite familiar so far...

"The variations in altitude stayed within a range of more or less one thousand
feet. Five cases of a voluntary descent were observed, of which one was of
3,500 feet. These descents followed a stall warning;"

That indeed makes AF447 sort of an exception, having quite involuntarily descended 48000 ft from FL350.
Whatever the other crews did - they didn't point the nose into orbit and kept it there for minutes a couple of seconds after they had to start flying their plane themselves. Please tell me that isn't considered a matter of luck these days?

Diagnostic

31st Aug 2011, 20:38

Whatever the other crews did - they didn't point the nose into orbit a couple of seconds and kept it there for minutes after they had to start flying their plane themselves.
Agreed, of course :)

The point I was trying to make, is that to lump the other 32 UAS together as some kind of non-events, and not the near disasaters which some could have been due to poor / incorrect handling in some cases (e.g. re-engaging A/P immediately after the disconnection, not recognising that it was a UAS at all etc.) is ignoring that some of them could have turned out differently IMHO.

As I agreed above, other crews didn't make the same major mistake as AF447 which you mention, but potentially disasterous mistakes in handling some other UAS events were made, as described in the report. That none of those resulted in crashes does have an element of luck IMHO (swiss cheese model).

Or to put it another way - I doubt that the AF447 crew were the only crew at that time, with the same lack of training/practice in handling/recognising UAS and hand-flying at cruise altitudes. If one of the other crews mentioned in the report, who didn't handle their one of the 32 other UAS well, had been in the AF447 situation at that height, time of day, flight deck situation (2 x F/O) etc. etc., would they have done better? I wouldn't bet on it. That's all I'm saying.

Zorin_75

31st Aug 2011, 23:11

If one of the other crews mentioned in the report, who didn't handle their one of the 32 other UAS well, had been in the AF447 situation at that height, time of day, flight deck situation (2 x F/O) etc. etc., would they have done better? I wouldn't bet on it. That's all I'm saying.I, perhaps naively, somehow hope that pilots being competent to fly a plane are still the norm rather than the exception, but maybe you're right and the odds are decreasing. If so, we can only hope that AF447 will be a wakeup call, the price was high enough.

Diagnostic

31st Aug 2011, 23:41

We can only hope that AF447 will be a wakeup call
Agreed completely.

From what I read on here, training (which is where I see one of the main deficiencies in the set of "holes" leading to the tragedy) has already been improved. If the A/P can disengage at FL350 and dump the plane into the hands of the flight crew, in Alt2 law, in some turbulance, at 2am, then they had better be trained and allowed to practice to cope with that event. It may not be a common event (and that may have driven budget decisions about training), but the criticality of hand-flying at that altitude (as explained by the ATPL pilots here) makes me think that the AF447 crew were put into a situation they were ill-prepared for. Of course there were other factors too.

I also think that luck (that word again!) played a part with AF447, in the sense of bad luck, with the timing of the Captain's departure to go on his break. If only that had been delayed by a few mins, he would have been on the flight deck when the problems started, and so the less experienced F/O wouldn't have been PF. Also the Captain would have been "caught-up" with events because he would have been there as the situation developed, so he wouldn't have been left with that "information void" which he had upon re-entering the flight deck that night etc. etc.

Anyway, just to summarise, I only posted earlier as the comparison seemed to be being made of: 32 other crews performance with UAS = good; AF447 crew performance with UAS = bad. IMHO, as explained in that 2nd Interim Report, the comparison is not so clear-cut.

thermostat

1st Sep 2011, 03:01

3holelover, I agree with most of what you say. I will go a step further by saying that part of airmanship is not to fly a heavy airplane at high altitude near the coffin corner, through Cb's. I get a kick out of the entire discussion about different "laws" and how to recover from a high altitude stall etc. etc. when this whole thing could have been avoided by diverting around the storm. That simple !! It's unbelievable how much this thread has gone off on a tangent to somehow blame the 'plane or the lack of knowledge in high altitude stall recovery. And one other thing. It wasn't ice crystals that caused the pitot blockage, it was supercooled water which froze on contact with the pitots. That same SC water would have made the AOA vane inoperative also. It would be helpful if we could stick to the real facts in all discussions. Just stay away from CBs at high altitude. End of story.

Mad (Flt) Scientist

1st Sep 2011, 03:17

...And one other thing. It wasn't ice crystals that caused the pitot blockage, it was supercooled water which froze on contact with the pitots. That same SC water would have made the AOA vane inoperative also. It would be helpful if we could stick to the real facts in all discussions.

Since AF447 HAD AoA vanes, and they actually continued to provide good data except when invalidated by the system due to low airspeed, how do you conclude that the AoA vanes would have been inoperative? They were in the same atmospheric conditions as the pitots. The pitots couldn't cope, the vanes did.

mm43

1st Sep 2011, 04:23

Originally posted by Mad (Flt) Scientist ...

The pitots couldn't cope, the vanes did. No! AoA Vane #1 was having a 'rest' well before the AP disconnect.

http://oi51.tinypic.com/sfw9ic.jpg

TJHarwood

1st Sep 2011, 05:33

I respectfully submit that the very length of these various threads on AF447 in itself indicates a widespread enough fear in many quarters that other high altitude UAS events on various aircraft could all too easily end equally fatally (under circumstances identical or very similar to AF447).......

rudderrudderrat

1st Sep 2011, 08:06

Hi Clandestino,

And what bearing on AF447 does it have?
The most important warning on that flight deck was "Stall Stall" - which appears to have been ignored / not acknowledged / not heard. (Similar to the You Tube "What's that Beeping?")

If their human auditory channel was filled to capacity with previous flight deck audio noises, (Cavalry Charge, ECAM Dings, ALT Alert etc), and the crew are attempting to verbally communicate whilst they trouble shoot - what is the point of having the only stall warning alert simply another verbal "Stall Stall"?
Even my iPhone has both audio ring plus Vibrator to get my attention.

Do most other aircraft have a stick shaker as a stall warning? If Apple can fit one into my phone, why was it not possible to fit something similar into the side sticks?

Before AF447 event, there was NO Stall Warning QRH procedure. The previous procedure was buried in FCOM.
The new QRH procedure emphasises reducing angle of attack, and does NOT recommend the use of TOGA power. (as the old FCOM procedure did)

BOAC

1st Sep 2011, 08:47

The most important warning on that flight deck was "Stall Stall" - which appears to have been ignored / not acknowledged / not heard. - not a complete picture since the 'cessation' of the warning was recognised? Time to think again.

Clandestino

1st Sep 2011, 10:22

Do most other aircraft have a stick shaker as a stall warning?Shakers are installed if natural pre-stall buffet is too weak to be recognized (or non-existent). I hope final report addresses whether the Airbus got through certification too easily or FBW Airbi indeed have sufficient natural buffet to warrant omitting the stickshaker. I agree that with sidestick, effective shaker needn't be much larger than those used on modern cellphones.

(Cavalry Charge, ECAM Dings, ALT Alert etc)All transient and not going off simultaneously with stall warning. IIRC even if they were trigered simultaneously with stall, stall warning overrides all other audio alerts.

how to recover from a high altitude stall Same way that you recover from mid or low alt stall. Hint: therr's difference between approach to stall and stall itself, which is seldom appreciated on PPRuNe.

'cessation' of the warning was recognisedIf this theory is congfirmed by HF group, that would show complete incap on the PF side. Unstalling by pull? Only when flying inverted or you're doing some very serious aerobatics.

Before AF447 event, there was NO Stall Warning QRH procedureBecause it was assumed that pilots understood stall & recovery from day one and never forgot it. Whooops.....

I respectfully submit that the very length of these various threads on AF447 in itself indicates a widespread enough fear in many quarters that other high altitude UAS events on various aircraft could all too easily end equally fatally (under circumstances identical or very similar to AF447).......Little less respectfully, I offer alternative explanation: as usual, fatal accident wakes many up to the reality of aviation being a very hazardous game. Length of the thread is due to inability to lay our fears to sleep with usual lithany of "It was only loco/Africa/Russia/Asia/Latin America/cargo anyway"

rudderrudderrat

1st Sep 2011, 10:29

Hi Clandestino,

Quote:
Originally Posted by rudderrudderrat
Before AF447 event, there was NO Stall Warning QRH procedure.
Quote:
Originally Posted by Clandestino
Because it was assumed that pilots understood stall & recovery from day one and never forgot it. Whooops.....

Then why did the QRH have "EGPWS ALERTS" in the Emergency Procedures?
Using your logic - you would think that "Pull Up" before hitting the ground and apply TOGA power would be instinctive - so why include that one?

I think Airbus believed they had designed an aircraft that was so well protected and that the chance of stalling was so remote - that a Stall Warning QRH procedure was deemed not necessary.

With the benefit of hindsight - that has now been fixed.

Shakers are installed if natural pre-stall buffet is too weak to be recognized (or non-existent).
You'll find that every Boeing & Lockheed has a dedicated stall warning stick shaker.

Old Carthusian

1st Sep 2011, 11:33

You'll find that every Boeing & Lockheed has a dedicated stall warning stick shaker.

Which can and has been ignored by pilots. It is important to remember that this is not an accident related to the aircraft configuration. It is an accident related to aircrew behaviour.

Ian W

1st Sep 2011, 14:27

Which can and has been ignored by pilots. It is important to remember that this is not an accident related to the aircraft configuration. It is an accident related to aircrew behaviour.

Which would appear to be a trained and ingrained 'by rote' behavior with little actual practical follow-up training.

Diagnostic

1st Sep 2011, 15:26

@Old Carthusian:
You'll find that every Boeing & Lockheed has a dedicated stall warning stick shaker.

Which can and has been ignored by pilots. It is important to remember that this is not an accident related to the aircraft configuration. It is an accident related to aircrew behaviour.
All very true, sir.

However, in the spirit of trying to close as many holes in the proverbial swiss cheese as possible, I see no harm in at least considering whether stall warnings via a different cognitive route (i.e. non-verbal, non-text) might be helpful - although as you say, also not perfect.

In the case of AF447, the question in my mind is whether the stall warnings were (a) not heard (e.g. cognitive overload or similar) - perhaps stick shaker would help here; (b) heard but not believed (due to multiple apparent problems with the aircraft behavior, leading especially the PF to ignore all warnings?) - stick shaker might be ignored in this case too!; (c) some other reason.

Personally, I fear that the answer is (b) but of course, I don't know. I think that rudderrudderrat has put forward a very good case for auditory overload (also Ian W) - I know that when I'm trying to speak to someone, I do tune-out all extraneous sounds; I can't tune out only selected sounds unless one sound is especially loud. So perhaps the answer is (a), and for that reason shouldn't a stick shaker be considered, for those occasions when it would help?

@mm43:
AoA Vane #1 was having a 'rest' well before the AP disconnect.
I mentioned this a few weeks back, and sensor_validation kindly pointed out a very plausible potential cause (link below) unrelated to ice (I don't know if that's what you were suggesting?):

http://www.pprune.org/tech-log/456874-af-447-thread-no-5-a-89.html#post6628318

@TJHarwood:
I respectfully submit that the very length of these various threads on AF447 in itself indicates a widespread enough fear in many quarters that other high altitude UAS events on various aircraft could all too easily end equally fatally (under circumstances identical or very similar to AF447).......
I also thought you were right. I was only replying to (what I believe) Clandestino was saying, where he seemed to be saying that if 32 other crews could manage clogged pitots successfully, then AF447 was the exception. The reason I posted was that this assertion seemed to conflict with the data & analysis presented in the 2nd Interim BEA Report (which I found very interesting, as it touches on some aspects of my professional life).

If I misinterpreted what Clandestino was saying, and if everyone agrees that some of the other 32 incidents were also (or could, in other circumstances like in turbulance, at night, at FL350, have developed into) "close shaves", then I happily withdraw from the conversation for the moment. :)

Mad (Flt) Scientist

1st Sep 2011, 15:31

I think Airbus believed they had designed an aircraft that was so well protected and that the chance of stalling was so remote - that a Stall Warning QRH procedure was deemed not necessary.

With the benefit of hindsight - that has now been fixed.

Actually, I think they (and others) took the position that not only was it not necessary, it was actually prohibited, at least from the flight manual.

CS 25.1585 Operating procedures

(b) Information or procedures not directly
related to airworthiness or not under the control of
the crew, must not be included, nor must any
procedure that is accepted as basic airmanship.

The idea that you recover from a stall by pushing the nose down was felt, at least by some, to be such basic airmanship that it was actually prohibited by regulation from inclusion. Airbus were not alone in having no such procedures.

steamchicken

1st Sep 2011, 15:47

That's really impressive data visualisation - brings out the difference between the PF and PNF's responses and also the dual input events.

(A thought - when we want to focus on the trend in a particular metric, we plot it on a chart like the DFDR traces. Strangely, aircraft displays show scrolling numbers or else multiple needles. Would a "value plus trend-line" be better? Medical monitoring displays do this.)

DozyWannabe

1st Sep 2011, 19:10

Then why did the QRH have "EGPWS ALERTS" in the Emergency Procedures?
Using your logic - you would think that "Pull Up" before hitting the ground and apply TOGA power would be instinctive - so why include that one?

My wild-assed guess in this situation would be that stall recovery is taught in single-engined trainers from very early on in basic aircraft handling training. Most single engined trainers are not equipped with EGPWS.

I think Airbus believed they had designed an aircraft that was so well protected and that the chance of stalling was so remote - that a Stall Warning QRH procedure was deemed not necessary.

Well, it's an opinion with merit in a theoretical sense, but I'd be surprised if that tured out to be the case. I remember reading an interview with a pilot in "The Tombstone Imperative" in which he advanced the opinion that things like stall recovery and microburst recovery have been mitigated by the arrival of the large high-bybass turbofans that were introduced with the widebodies and were installed on every new jetliner from the mid-70s onwards - you could just apply TOGA and keep the wings level and the AoA within a certain range and power out of it. It now appears that this is not always the case.

What makes your assertion doubtful for me is that if Airbus truly ever felt that way then they would not have put a stall warning in the aircraft in the first place - yet they did. So from an engineering perspective at least, Airbus always knew it was possible.

You'll find that every Boeing & Lockheed has a dedicated stall warning stick shaker.

So did the VC-10, BAC 1-11 and HS Trident - in fact they pioneered the technology. However, as D.P. Davies said in the third and final edition of HTBJ, pilots had trouble trusting them and adjusting to their presence. Stick shakers in the Trident, 727, 757 and MD-80 have all been ignored on the line with fatal consequences.

As to the other discussion, the presence of supercooled droplets in the area AF447 flew through is debated. Even the BEA won't put the pitot failures down to anything more specific than "ice crystals".

jcjeant

1st Sep 2011, 19:38

Hi,

DW
What makes your assertion doubtful for me is that if Airbus truly ever felt that way then they would not have put a stall warning in the aircraft in the first place - yet they did. So from an engineering perspective at least, Airbus always knew it was possible.Incorrect
This is a mandatory specification from regulators .. nothing to do with any Airbus decision.
As for a Airbus be able to not stall .. it's true .. for Airbus Industrie (or EADS)
This is black on white from Airbus
No training necessary for stall condition
This is one of the point in their commercial incentive ... low cost training

Note (from BEA report N°3 page 63)

 Airbus

The procedures were modified by Airbus in May 2010: replacement of the “Stall warning”
additional abnormal procedure by the “Stall recovery” and “Stall warning at lift-off”
procedures.
Before this date .. Airbus don't write about "Stall recovery" .. but instead .. "Stall warning"

And this is the change made after the AF447 accident (BEA report N°3 page 82) for the simulator training

5.1.3 Crew training

Training in a flight simulator

Additional session entitled “Unreliable IAS”:

 Summer 2009 (A320, A330/340)
 Booklet and briefing from the session: key technical points, HF and TEM (Threat and
Error Management) considerations
 Revision of emergency manoeuvres, at take-off and in cruise.
 High altitude flight in alternate law
 Approach to stall, with triggering of the STALL warning
 Landing without airspeed measurement information
 Associated briefings (all cockpit crew):
o Weather radar
o Ice crystals

Note1: This information has been integrated into the type ratings.

Note 2: The stall procedures were modified following the modification of the STALL procedures by the
manufacturer, as indicated in 1.18.

Yet .. nothing about training for stall recovery ..........

Ice crystals ? .. well I wonder what can be a training in flight simulator for "ice crystals" ....
A employee of the training center trowing ice crystals on the simulator casing ? :eek:

I am almost certain that these new workouts are not going to make more significant progress in terms of flight safety.
Only the real flight training will provide pilots with the mastery of certain events
More flight simulator training for this kind of event it's just like put a additional (redundant) fire alarm horn near another instead construct more fire escapes
You can't save lifes with a supplementary horn .. you can just spare money for short term .. but problem not solved (I was there)

CONF iture

1st Sep 2011, 20:04

I have always preferred substance to style.
Then stick to your words, concentrate on the facts and give a break to your snobbish and pompous style, that won't hurt.

CONF iture, reservedly, modestly, and humbly claiming that e.g. Habsheim crash was caused by software protecting the aeroplane was never my cup of tea.
Basic rule : Do not put words in my mouth or assume by quoting me if you pretend I have said so ...

A330 has 2 versions of the ruddercontrol system.
1/ original system with mech cable loop (like F-GZCP) is called - mechanical rudder - altough the cable loop is a backup system.
3/ the enhanced version without mech cable loop is called - electrical rudder - the backup is here electrical by means of the BCM (Backup Control Module).
I would rather think, according to the serial number, that F-GZCP was the Enhanced version of the 330.
I believe airfrance has or had 16 A330, 4 of them being A330E.
F-GZCP was also the newest in the fleet.

As you may have more accurate information ... please, let me know.

mm43

1st Sep 2011, 20:08

Originally posted by Diagnostic ...
I mentioned this a few weeks back, and sensor_validation kindly pointed out a very plausible potential cause (link below) unrelated to ice (I don't know if that's what you were suggesting?)Well it could be related to ice, or it could be related to the EASA AD, and there is no way of proving it now. Though looking at the traces (when valid), it was tracking again prior to 2:12:20.

In any case, it wasn't an issue in ALT 2B as the SW used the highest AoA value.

DozyWannabe

1st Sep 2011, 20:29

I would rather think, according to the serial number, that F-GZCP was the Enhanced version of the 330.
I believe airfrance has or had 16 A330, 4 of them being A330E.
F-GZCP was also the newest in the fleet.

Air France F-GZCP (Airbus A330 - MSN 660) | Airfleets aviation (http://www.airfleets.net/ficheapp/plane-a330-660.htm)

F-GZCP was an A330-203. As far as I know, they didn't retrofit design changes from later models to later production runs of earlier models because of the training issues that would ensue (pilots reported a significant difference in feel between the old system and the new). So for the moment let's assume that F-GZCP was of the old design when it came to the rudder controls. Either way I don't think it matters a great deal regarding this accident because the rudder barely came into it.

Still waiting on your reply on the other thread by the way...

Diagnostic

1st Sep 2011, 21:19

@mm43:
Well it could be related to ice, or it could be related to the EASA AD, and there is no way of proving it now.
Agreed, sir - my point is just that we do have another plausible potential reason, and so I don't think we can say the AoA vanes (specifically #1) were definitely affected by ice.

I'm interested - other than the Perpignan incident (where the AoA sensors had been "abused" outside their specs before the flight), is ice affecting AoA sensors as much of a problem as pitot icing?

Svarin

1st Sep 2011, 21:22

DozyWannabe dared writing :

the rudder barely came into itRead the report in full, including page 1xx where you will find the rudder traces. Just because BEA did not expand on the subject does not mean it is irrelevant.

You yourself have used the expression PIO (which stands for Pilot Induced Oscillation). Experts now prefer APC event (Aircraft Pilot Coupling), which better emphasizes the interface issue. There is a definite APC event in this instance, in roll and yaw. Rudder played a huge part, as a fast-moving, oscillating control surface with zero feedback to the pilots.

Regardless of what actually caused these oscillations, and why did PF had such trouble keeping the ship's wings level, rudder certainly played a major part in the whole sequence.

That the system governing this particular control surface was subjected to a large update is in itself source of questioning.

Yes, looking at the traces is dull work. But as long as we take these recorders for the truth of the matter, the traces are all we have as unbiased data, along with the ACARS messages. Unless one is prepared to use and study them carefully, one can only follow pre-made conclusions written under careful supervision by interested parties. Such should not satisfy an inquiring mind.

DozyWannabe

1st Sep 2011, 21:41

@Svarin:

Question - if there is no feedback through the rudder pedals of an A330, how come pilots reported a significant difference in feel between the earlier A330/340 (200-300) models and the later A340 models (500-600)?

And by "barely came into it", I mean that it came into it after control was lost, and was very much secondary to the pitch and roll inputs that went through the sidesticks.

One thing that I find interesting in the traces is that the sidestick traces are based on a graph where left stick input relates to values > 0, whereas roll traces say that values > 0 relate to right wing down. If this is that case then PIO/APC needs to be looked into, as the traces of roll vs. sidestick show the precise opposite.

PS. I like the term "dared" - it makes me sound like a far more "caution-to-the-wind" guy than I actually am. :)

RetiredF4

1st Sep 2011, 22:30

DozyWannabe
One thing that I find interesting in the traces is that the sidestick traces are based on a graph where left stick input relates to values > 0, whereas roll traces say that values > 0 relate to right wing down. If this is that case then PIO/APC needs to be looked into, as the traces of roll vs. sidestick show the precise opposite.

It´s a reaction, i´m quite familiar with from big ugly F4, caused by adverse yaw in combination with swept back wings.

Might be the same here. That would also explain the nosedrop in addition with the max bank angle points and the nose rise when by use of rudder the bank reversed.

As i don´t know it, others comments more than welcome.

F4 flight characteristics High AOA
With high AOA the down going aileron (in our case on the right wing) produces a lot of drag (the down gong aileron stands nearly 90° to the FPA). That produces a yaw into the low right wing. We had the spoiler on the other side go up, but that one was blanked out at high AOA. The wings are swept back as we know. Due to the yaw to the right the right low wing looses lift, as it is now more swept back in relation to the relative wind, wheras the left wing is moved forward and is less swept back in relation to relative wind. The right low wing therefore produces less lift than the left high wing, minimizing, equalizing or even overcoming (dependent on AOA) the upforce of the deflected aileron on the right wing. Live expierience: Full left aileron and the ship rolls to the right. Only help is use of rudder in the same direction as rolling is intended. In our F4 best was to roll with rudder alone and keep the ailerons at neutral.

The laws called ”Alternate” are engaged when the protections related to the normal laws (ALPHA 1, VM01) are lost. The laws called ”Direct” are engaged when the Nz law is lost.
The other functions available are :
- speedbrake function
- ground spoiler function
- MLA (except in direct laws)
- rudder travel limitation
- sideslip estimation (except in alternate 2 or direct laws)
- computation of characteristic speeds (except in direct laws)
(bolding by me)

Yaw alternate law
This law, elaborated in the FCPCs and FCSCs, is engaged in the event of
loss of the lateral normal law. Its characteristics are as follows:
- The roll control is direct, an order on the side stick directly commands a
deflection, according to a kinematic.
- The yaw control is achieved from the pedals, through a mechanical linkage.
- A Dutch roll damping function using limited yaw rate data is introduced
via the yaw damper servoactuators.
In the event of loss of the inertial data from the ADIRUs, the yaw rate data
for Dutch roll damping are provided to the FCPCs via a unit comprising 2
rate gyros. In case of 3 FCPCs loss, the Dutch roll damping is achieved by
the FCSCs with yaw rate data coming from the rate gyro unit.
(bolding by me)

If i read the TTM correct, automatic countering of yaw/sideslip would not be available in ALt2.

Mr Optimistic

1st Sep 2011, 22:41

Are you discussing the control of the aircraft when stalled ? If so, surely the control characteristics are markedly different from what anyone is used to in normal flight, and the aircraft response ditto. Aircraft seems to have been very well behaved given the situation.

Clandestino

1st Sep 2011, 22:42

I was only replying to (what I believe) Clandestino was saying, where he seemed to be saying that if 32 other crews could manage clogged pitots successfully, then AF447 was the exception.You have interpreted me kind of correctly, sir. I was pretty serious when I mentioned curves of crew performance available vs required. It is possible that some of the crews exposed to the pitot icing under the circumstances of AF447 crew would have gone through the same motions and ended up similarly. It is possible that AF447 crew would have made much better job under more favourable circumstances, however such speculations are only useful if they bring us closer to answer why the crew did what they did. Only when this is answered we can proceed to "How do we prevent it in the future?"

Number 32 was often quoted in threads referring to AF447. I was able to trace it as far as PBS show "Crash of flight 447". Interim report #2 refers to 33 cases preceding AF447 and four occurring afterwards. In those cases that could be further analyzed, crew pushed forward when faced with stall warning, CM2 of AF447 pulled and kept pulling. My entirely subjective criterion of exception is thus met.

Read the report in full, including page 1xx where you will find the rudder tracesActually it is Interim report 3, page 112, Lateral parameters.

There is a definite APC event in this instance, in roll and yaw. Rudder played a huge part, as a fast-moving, oscillating control surface with zero feedback to the pilots.There is not in yaw.

To have PIO/APC you have to start with pilot's oscillatory input. Rudder starts to oscillate without pilot's input at about 2:11:45 (following yaw damper commands) when the aeroplane is so deeply stalled that high AoA and low speed make it pretty inefficient. There's fast pedal movement to the left at approximately 2:13:05, followed by sharp reversal to the right and return to neutral over next minute. No oscillatory inputs. No APC. Just an aeroplane gently mushing in stall.

Yes, looking at the traces is dull work. But as long as we take these recorders for the truth of the matter, the traces are all we have as unbiased data, along with the ACARS messages. Unless one is prepared to use and study them carefully, one can only follow pre-made conclusions written under careful supervision by interested parties. Such should not satisfy an inquiring mind. I could not put it better myself, sir.

xcitation

1st Sep 2011, 22:46

Hi Clandestino,

Quote:
And what bearing on AF447 does it have?
The most important warning on that flight deck was "Stall Stall" - which appears to have been ignored / not acknowledged / not heard. (Similar to the You Tube "What's that Beeping?")

If their human auditory channel was filled to capacity with previous flight deck audio noises, (Cavalry Charge, ECAM Dings, ALT Alert etc), and the crew are attempting to verbally communicate whilst they trouble shoot - what is the point of having the only stall warning alert simply another verbal "Stall Stall"?
Even my iPhone has both audio ring plus Vibrator to get my attention.

Do most other aircraft have a stick shaker as a stall warning? If Apple can fit one into my phone, why was it not possible to fit something similar into the side sticks?

Before AF447 event, there was NO Stall Warning QRH procedure. The previous procedure was buried in FCOM.
The new QRH procedure emphasises reducing angle of attack, and does NOT recommend the use of TOGA power. (as the old FCOM procedure did)

Indeed, a major puzzle why ignore the stall warnings for 3 or 4 minutes? In the entire transcript of the incident not once was "stall" or "décrochage" ever mentioned by any of the 3 pilots.
Reasons
1. Both FOs overloaded for 4 minutes and did not hear it.
2. Both FOs heard it and ignored it as being false (as per recent QF72 etc malfunction.) But why not verbalize that?
3. Both FOs didn't understand it/unfamiliar "<Stall><Stall>. What's that?". Would the french décrochage or flat line tone been better warning? Were they verbalizing that they did not understand it. Perhaps confused in the heat of the moment. Or did "What's that?" refer to something else.
4. Both FOs heard it and accepted it and did not know what to do. Again why not verbalize that?

Old Carthusian

1st Sep 2011, 22:52

The case for cognitive overload is by no means proven I am afraid. Iain W's reference to Wickens was somewhat simplified and didn't really offer a suitable explanation. It does sound seductive but is only a small part of what is a complex theory. Once more parts of the theory are included cognitive overload falls by the wayside. I highlight the states of panic and shock as more likely causes for the PF acting as he did because the characteristics of those states better fit his actions than cognitive overload. This also ties in much more effectively with the apparent lack of professionalism in the aircrew.

It might be that stall warnings via a different cognitive route do offer an aid in the future but we have to think about the situation as it is and as Diagnostic mentions it is possible that a stick shaker might not have saved the day. We have to look at the situation as it is and try to understand why the crew acted as they did. Then we can start to look at possible solutions which might indeed include a stick shaker. But this accident is about professionalism, training, culture, CRM and SOPs all mixed togther and above all knowing your machine. It is not about failings in the machine. I would say, a big shake up in how AF and probably other airlines train their pilots would seem to be necessary.

Chu Chu

1st Sep 2011, 23:18

The examples of cognitive overload discussed seem to focus on someone who's focused on a specific task they believe is going normally, to the exclusion of an input that would have warned them it wasn't.

I wonder how well the theory applies to someone who realizes something's wrong but doesn't know what it is. It seems to me that when you go into troubleshooting mode, as the AF 447 pilots apparently did, you're more likely to step back and look/listen for clues that might explain what's wrong.

Ian W

1st Sep 2011, 23:49

ChuChu
The examples of cognitive overload discussed seem to focus on someone who's focused on a specific task they believe is going normally, to the exclusion of an input that would have warned them it wasn't.

I wonder how well the theory applies to someone who realizes something's wrong but doesn't know what it is. It seems to me that when you go into troubleshooting mode, as the AF 447 pilots apparently did, you're more likely to step back and look/listen for clues that might explain what's wrong.

The effect that is repeatedly observed is that the human finding something difficult starts to enter 'attentional tunneling'. This may be called several things 'the pilot's scan broke down', he had 'tunnel vision' etc.

The pilot with cognitive tunneling may concentrate on entirely the incorrect stimulus - it is not something that lends itself to logical decomposition.

Ideally, simulation rides should be set up to initiate attentional tunneling then the ride frozen or rerun in playback and the tunneling pointed out to the 'subject'. It was found that using this approach with military pilots reduced susceptibility to tunneling.

Yet again it is back to training.

Diagnostic

1st Sep 2011, 23:58

@Clandestino:
Thanks for the clarification.
Number 32 was often quoted in threads referring to AF447.
Indeed - I was just quoting your mention of that number, without detailing the variations around it.

In those cases that could be further analyzed, crew pushed forward when faced with stall warning, CM2 of AF447 pulled and kept pulling. My entirely subjective criterion of exception is thus met.
But that only addresses one reaction - to the stall warning. Other crews made different errors, which the crew of AF447 did not make (e.g. re-engaging the A/P during the UAS event). So depending on what specific criterion is considered, other crews were exceptions and not AF447. :) That's exactly why I agree with you about the subjectivity - thanks again for your clarification.

@xcitation:
Indeed, a major puzzle why ignore the stall warnings for 3 or 4 minutes?
Absolutely - I said something very similar earlier. :) The lack of apparent recognition of the stall warnings, but also the lack of explicit verbal reasons for ignoring them, is a puzzle.

How are we ever going to know the reason(s)? I can only think that the BEA Human Factors folks will eventually give us their best guess, which will be based on more info than we have e.g. intonation used on the CVR recording.

@Old Carthusian:
Thanks for the summary.
But this accident is about professionalism, training, culture, CRM and SOPs all mixed togther and above all knowing your machine.
Agreed - many factors indeed. I also believe that there is a large dose of the "Ironies of Automation" [Bainbridge] in there too - humans monitoring machines (any) don't do very well when they have to take over completely and suddenly, with no prior warning. That seems to be a human "failing", and not something that we can (should?) expect to "train out of people".

That's why I've been following with interest the comments here about the advantages of designing some kind of "ATT Hold" autopilot in the event of UAS - something to gradually ease the transition from A/P to manual control, and lessen that "shock factor" (which is my currently favoured reason, from everything I've read so far, for the NU inputs from the PF - just muscular tension from the shock of having the plane "dumped in his lap". That's why he didn't know he was causing the climb, as it wasn't a conscious input).

I've learned here some of the difficulties of implementing something like that, especially if (as in this case), the plane is in some turbulance when it is engaged, and we may not want the current attitude (which could be momentarily significantly NU or ND in turbulance) to be the one which is held by the ATT Hold system. That requirement for the PF to control the plane in the turbulance immediately, prevented PJ2's lovely phrase of "don't just do something, sit there" (IIRC) after the A/P disconnect. :(

If the onboard systems had given the crew time to think & get up to speed, instead of immediately forcing them to react, in a state of shock from their "cruise state", I wonder if we would have had a better outcome. The Human Factors analysis of the final accident report will make facinating reading, I expect.

CONF iture

2nd Sep 2011, 00:05

In any case, it wasn't an issue in ALT 2B as the SW used the highest AoA value.
To me, the issue is at the different level, and could be very significant.
Due to the AoA discrepancy, it is very probable that the following STATUS message was displayed on ECAM to the AF447 crew :

RISK OF UNDUE STALL WARNING
I believe blue to be the color but cannot positively find that information for now.

But once again, the real issue is that BEA didn't say yet a single word about it ...

DozyWannabe

2nd Sep 2011, 00:33

@CONF iture

The QRH would have told them to disregard that message though - if indeed it appeared (the PNF did not call it out). At any rate, I suspect that message referred to the short-term SW that could be expected while the systems sorted out what sensors were valid and which ones weren't. I'd be surprised if any of the designers in the early '90s thought maintained application of full back stick would be the response to UAS.

WRT the BEA, give them time. They're a very different organisation these days from what they were 23 years ago.

infrequentflyer789

2nd Sep 2011, 00:44

3. Both FOs didn't understand it/unfamiliar "<Stall><Stall>. What's that?". Would the french décrochage or flat line tone been better warning? Were they verbalizing that they did not understand it. Perhaps confused in the heat of the moment. Or did "What's that?" refer to something else.

Or, as a variation on that, they understood, but ignored or distrusted the warning preciseley (or in part) because it was in english. Already overloaded, they focused, what attention they gave to it, on the messenger and not the message.

The exclamation "shut up gringo" might ring a bell for some readers... (in that case, clearly time for the brain to process the race/language of the speaker and prepare and verbalise corresponding insult - but not to process the message content).

CONF iture

2nd Sep 2011, 01:00

F-GZCP was an A330-203. As far as I know, they didn't retrofit design changes from later models to later production runs of earlier models because of the training issues that would ensue (pilots reported a significant difference in feel between the old system and the new). So for the moment let's assume that F-GZCP was of the old design when it came to the rudder controls.
This page won't directly specify the status Enhanced or not of the aircraft. Only the serial number indirectly will.

http://i45.servimg.com/u/f45/11/75/17/84/af447_18.png (http://www.servimg.com/image_preview.php?i=112&u=11751784)

There is no training issue between both versions, at most a PDF document to log on.
You simply don't retrofit an earlier model in newer Enhanced one. Enhanced is much much more than a design change in the rudder engineering, starting with the size of rudder + fin.

Linktrained

2nd Sep 2011, 01:07

PF had practiced the UAS out of Rio on the Simulator three months ago (with this PNF ?) which was a 15 degree climb and TOGA, ( for lower levels). He may have tried to do just this. He was slow in increasing his thrust to TOGA on AF447. But he may have decided that this was the correct drill. His mind was set.

I have mentioned earlier one fleet in the sixties, where, in the Crew Room it was joked "that it did not matter which Drill you did... so long as you did it quickly."

More training ? Garage Years #644, says that his simulators are in use 20 hours a day with 4 hours for maintenance. One of AB's sales lines was that conversion from one AB to another AB was simpler..

mm43

2nd Sep 2011, 01:14

Originally posted by RetiredF4 ...
If I read the TTM correct, automatic countering of yaw/sideslip would not be available in ALt2.
Have another look at the traces and you will find the yaw damper was working and contributed more and more as the IAS decreased. The pedals were only used from 2:13:05, and all other rudder movements were from the yaw damper.

Check Post #373 (http://www.pprune.org/tech-log/460625-af-447-thread-no-6-a-19.html#post6658728) and you will note that the second group of traces is a compilation of pitch, roll and rudder input. I don't think the yaw damper was helping with the roll as lateral stability decreased.

@CONF iture

Re AOA Vane #1
RISK OF UNDUE STALL WARNING

As with the STALL STALL, the lack of verbalization between the PNF and PF is a problem. No wonder the BEA want a cockpit camera.

CONF iture

2nd Sep 2011, 01:19

The QRH would have told them to disregard that message though
Once again Dozy, you may not be in the best position to appreciate how a technical recommendation figuring only on the second page of a QRH procedure will catch your attention when action is taking place.

Regarding the BEA, the only thing that has changed is that corporations are getting more powerfull than ever ...

DozyWannabe

2nd Sep 2011, 01:30

@CONF iture:

With all due respect, I'm well aware of how long it takes to pull out a QRH in an emergency situation, don't think I havent borne that in mind. However, right now we know neither

whether that ECAM message appeared
whether it was noted by either pilot

What we do know is that it was never called out, and that the PNF seems to have been increasingly concerned about the way the aircraft was being handled. Unfortunately there's nothing on the CVR excerpts to confirm or deny "Hey - that's a Stall Warning" - "ECAM says disregard it", but I'm pretty sure that if such a thing was said then it would have been included (if the BEA are so big on protecting corporations like Airbus, why have the DGAC already mandated a software change?). You're clearly an intelligent guy - if you stopped fighting 24-year-old windmills and applied yourself fully to the facts we could really use you.

ChrisJ800

2nd Sep 2011, 04:37

Quote:Xcitation:

Indeed, a major puzzle why ignore the stall warnings for 3 or 4 minutes? In the entire transcript of the incident not once was "stall" or "décrochage" ever mentioned by any of the 3 pilots.
Reasons
1. Both FOs overloaded for 4 minutes and did not hear it.
2. Both FOs heard it and ignored it as being false (as per recent QF72 etc malfunction.) But why not verbalize that?
3. Both FOs didn't understand it/unfamiliar "<Stall><Stall>. What's that?". Would the french décrochage or flat line tone been better warning? Were they verbalizing that they did not understand it. Perhaps confused in the heat of the moment. Or did "What's that?" refer to something else.
4. Both FOs heard it and accepted it and did not know what to do. Again why not verbalize that?

5. PF believed he was applying a correct procedure of stick back, apply TOGA and airbus protections especially alpha prot would prevent the bus from stalling with minimal height loss.

A33Zab

2nd Sep 2011, 06:46

Hi CONF iture,

I would rather think, according to the serial number, that F-GZCP was the Enhanced version of the 330.
I believe airfrance has or had 16 A330, 4 of them being A330E.
F-GZCP was also the newest in the fleet.

As you may have more accurate information ... please, let me know.

An *e-rudder* version is not equipped with a rudder travel limiter unit.
Since a RTLU is found with the VS, the F-GZCP was a *m-rudder* version.

jcjeant

2nd Sep 2011, 06:58

Hi,

WRT the BEA, give them time. They're a very different organisation these days from what they were 23 years ago. Interesting
Do you have any information about the change from 23 years ago :confused:
A historic of the change .. the new status ?
So far I do not find any change in the BEA organization along those 23 years...

What we do know is that it was never called outIndeed it was only some "professional" call out on the beginning of the event AF447
From the PF
I have the conrols
Ignition start
From PNF
Alternate law protection
Wing anti-ice
The other CVR words .. are more like a discussion between Laurel and Hardy .. as I write sometime ago

glad rag

2nd Sep 2011, 09:00

Agreed, sir - my point is just that we do have another plausible potential reason, and so I don't think we can say the AoA vanes (specifically #1) were definitely affected by ice.

I'm interested - other than the Perpignan incident (where the AoA sensors had been "abused" outside their specs before the flight), is ice affecting AoA sensors as much of a problem as pitot icing?

The AD certainly makes one stop and think, that's for sure.

Svarin

2nd Sep 2011, 09:14

Clandestino wrote (about PIO/APC) :

There is not in yaw.

To have PIO/APC you have to start with pilot's oscillatory input. Rudder starts to oscillate without pilot's input at about 2:11:45 (following yaw damper commands) when the aeroplane is so deeply stalled that high AoA and low speed make it pretty inefficient. Indeed, when the aircraft is fully stalled and mushing, and PF is holding full deflection left-up on his sidestick, and he is not touching the rudder pedals, yaw damper does give orders. After some time looking at these traces, I still cannot make sense of them. This is an area of interest which I hope will be investigated.

However, more akin to the APC definition which you suggest is the start of the sequence, right when PF takes control. Roll and yaw are mixed up in an oscillation, which appears to the pilots as a pure roll movement, but where in fact rudder is actuated and lateral accelerations are measured.

Surprisingly, no trace of gyros is provided. Yaw rate gyro would be interesting to look at. Heading traces are insufficiently detailed to extrapolate yaw rate from them.

Moreover, I am greatly interested in aileron authority and response to SS inputs. It appears that right inboard aileron maximum deflection remains limited to a modest amount. This maximum available deflection decreases significantly within the first seconds after A/P quit. Such limitation is surprising in the context of a Direct Roll flight law, which is part of Alternate 2B global flight law.

Interesting elements of comparison are available with the D-AXLA accident report. FDR traces show a large number of fascinating bits of information. Such are not provided in the 3rd interim report which is our subject matter. Maximum aileron deflection appears different in this report. However, the exact type of aircraft is different of course. Expert input welcome.

AlphaZuluRomeo

2nd Sep 2011, 12:11

Due to the AoA discrepancy, it is very probable that the following STATUS message was displayed on ECAM to the AF447 crew :

RISK OF UNDUE STALL WARNING

@CONF iture:

With all due respect, I'm well aware of how long it takes to pull out a QRH in an emergency situation, don't think I havent borne that in mind. However, right now we know neither

whether that ECAM message appeared
whether it was noted by either pilot

Well, perhaps A33Zab (or other knowledgable people) may help here. I would have thought that that kind of ECAM would also trigger an ACARS message, for the maintenance teams to check the vanes after the flight. Am I mistaken, here?
If not, given that we know there wasn't such an ACARS message, we may safely assume that the ECAM/warning was not triggered.

RetiredF4

2nd Sep 2011, 12:25

Mr. Optimistic
Are you discussing the control of the aircraft when stalled ? If so, surely the control characteristics are markedly different from what anyone is used to in normal flight, and the aircraft response ditto. Aircraft seems to have been very well behaved given the situation.

I was answering upon DozyWannabes remark

One thing that I find interesting in the traces is that the sidestick traces are based on a graph where left stick input relates to values > 0, whereas roll traces say that values > 0 relate to right wing down. If this is that case then PIO/APC needs to be looked into, as the traces of roll vs. sidestick show the precise opposite.

and tried to find an explanation for it. It is not relevant for the start of the upset and stall, but it is relevant information, as it is a typical stall behaviour for a swept wing aircraft. Known to the crew it should have helped to recognize the stall (if they missed everything else) and to react more adequate to the rolling tendency by using manual rudder instead of aileron to roll wings level.

mm43
Have another look at the traces and you will find the yaw damper was working and contributed more and more as the IAS decreased. The pedals were only used from 2:13:05, and all other rudder movements were from the yaw damper.

Yaw damping was active, also according to manual, but not side slip estimation.

- sideslip estimation (except in alternate 2 or direct laws)

Honestly i don´t know how this missing sideslip estimation influences the behaviour of the rudder in a stall, or how the functioning yaw damping influences it in a positive or negative way. But hopefully somebody can explain.

Check Post #373 and you will note that the second group of traces is a compilation of pitch, roll and rudder input. I don't think the yaw damper was helping with the roll as lateral stability decreased.

That compilation does not help, as it reflects the pitch attitude, the rudder position and the bank angle, but not the input from stick or rudder. Would be interesting to compile those with the input from SS and Rudder, the difference then would be the FCPC input (would do it myself, but am too dumb for it).

CONF iture

2nd Sep 2011, 14:27

(if the BEA are so big on protecting corporations like Airbus, why have the DGAC already mandated a software change?). You're clearly an intelligent guy - if you stopped fighting 24-year-old windmills and applied yourself fully to the facts we could really use you.
You’re not without noting that ADs are usually and conveniently not 'linked' to any accident …

Fully applying to the facts is actually my main purpose, but what do you imply by stating the BEA has changed ?
Is it how finally the awaited independency and transparency are upon them, so how do you see how things were managed 20 years back ?
I can’t remember you ever publicly doubting the official line …

An *e-rudder* version is not equipped with a rudder travel limiter unit.
Since a RTLU is found with the VS, the F-GZCP was a *m-rudder* version.
This, is a relevant piece of observation that I could hardly refute.
Thanks A33Zab.

Mr Optimistic

2nd Sep 2011, 15:49

Re 'verbalisation', is it still correct to think that BEA have not released all the cvr transcript so more may yet come ?

lomapaseo

2nd Sep 2011, 16:21

CONF iture etal

I'm puzzled by the arguments involving ADs, DGAC and the BEA :confused:

ADs are specific and the responsibility of the regulator to enforce

Accident causes are general (multiple layers of swiss cheese) and the responsibility of the BEA to dissect

What am I missing in understanding the arguments ?

lack of ADs don't cause accidents, but the presence of ADs may minimize future accidents

xcitation

2nd Sep 2011, 16:28

Indeed, a major puzzle why ignore the stall warnings for 3 or 4 minutes? In the entire transcript of the incident not once was "stall" or "décrochage" ever mentioned by any of the 3 pilots.
Reasons
1. Both FOs overloaded for 4 minutes and did not hear it.
2. Both FOs heard it and ignored it as being false (as per recent QF72 etc malfunction.) But why not verbalize that?
3. Both FOs didn't understand it/unfamiliar "<Stall><Stall>. What's that?". Would the french décrochage or flat line tone been better warning? Were they verbalizing that they did not understand it. Perhaps confused in the heat of the moment. Or did "What's that?" refer to something else.
4. Both FOs heard it and accepted it and did not know what to do. Again why not verbalize that?

5. PF believed he was applying a correct procedure of stick back, apply TOGA and airbus protections especially alpha prot would prevent the bus from stalling with minimal height loss.

Good call. I lost sight of that. ALT LAW occurred at the outset when AP off.
PF immediately 3/4 stick back for no known reason. After 20 seconds PNF states "ALT LAW, Protections Lo"(st). By then the speed had decayed and they were dancing the edge of the flight envelope.
PF was over controlling from the get go. Large scale stick inputs, thrust set at TOGA and then idle, Air Brakes deployed. Ignoring PNF directions. Did he panic?
Sadly if PF had left the controls alone and sat on his hands it could have been a different story. Near the end he verbalizes his desire for max nose up thus confirming it was intentional SS input. This would fit him not being aware of ALT LAW or not understanding it.

But I’ve been at maxi
nose-up for a while
No no no don’t climb
So go down

xcitation

2nd Sep 2011, 16:44

Can a french aviator clarify the extent of usage of the word "stall" in their training.
Is the english "stall" widely used in French aviation training or is the native "décrochage" used?
Is hearing "stall" a routine part of stall avoidance training in the sim?

rudderrudderrat

2nd Sep 2011, 16:52

Hi xcitation,
Sadly if PF had left the controls alone and sat on his hands it could have been a different story.
If he'd done that - it would have continued to roll right wing down until it was inverted.
They were struggling to keep the wings level in ALT LAW from the outset.

Clandestino

2nd Sep 2011, 17:54

Due to the AoA discrepancy, it is very probable that the following STATUS message was displayed on ECAM to the AF447 crew :

RISK OF UNDUE STALL WARNINGI think that BEA should be informed about your concerns. There's a faint possibility they believe that two agreeing AoA probes outvoted the third. It is of utmost importance to discover which version is correct.
I was just quoting your mention of that numberYup, my mistake. It wasn't 32, it was 33 before and four between the accident and publication of interim 2. Thank you for pointing me towards the more reliable source.

that only addresses one reaction - to the stall warning
Yup, because that made the difference between 37 happy landings and one crash. Other crews may not followed procedures, but they kept the aeroplane aloft.

Moreover, I am greatly interested in aileron authority and response to SS inputs. It appears that right inboard aileron maximum deflection remains limited to a modest amount. If you look closer it seems that it's not just right inboard aileron. Left inboard aileron also goes only to about 5° up and no more. Is it normal to have ailerons with different maximum up and down deflection? Is it usual? Is it only Airbus specialty? We have the right to know!

Regarding the aileron authority: do you remember what aileron authority you had when stalling your first trainer? Widebody plummeting oceanwards at AoA higher than 35° is bound to have it even worse than that.

xcitation

2nd Sep 2011, 18:15

Hi xcitation,

Quote:
Sadly if PF had left the controls alone and sat on his hands it could have been a different story.
If he'd done that - it would have continued to roll right wing down until it was inverted.
They were struggling to keep the wings level in ALT LAW from the outset.

Well maybe you hit the nail on the head. That is about the only thing they all focussed on. Forget speed, stall, attitude, THS and just focus on the roll. Amazing that they had such roll control once they stopped flying and became a falling 200 tonne piece of metal. If anything from all this the airbus is an awesome forgiving airplane at low speed. Would you expect to have control of liner at speed as low as 30kts in chop? If you asked me I would say it tumbles and break up.

PinkHarrier

2nd Sep 2011, 18:48

Have I missed something? What would cause three pilots to know that the last thing they should do is put the nose down?

DozyWannabe

2nd Sep 2011, 19:11

If he'd done that - it would have continued to roll right wing down until it was inverted.
They were struggling to keep the wings level in ALT LAW from the outset.

Are you sure? The initial roll could have just been a nasty bit of turbulence - everything after that could be PIO/ADC - the lateral control inputs from the PF certainly seem in excess of those required at cruise altitude.

HazelNuts39

2nd Sep 2011, 19:36

The airplane rolled through seven degrees in two seconds, not really an alarming rate (the 'second' hand on your analogue wrist watch 'rolls' twice as fast). Upward gusts alternated with downward gusts about every five seconds.

OK465

2nd Sep 2011, 20:15

I would not mind if someone would explain to me how releasing the SS in roll direct results in the aircraft rolling inverted, but more importantly explain this to the RAF who just sub-leased 2 of these aircraft and are probably not aware of this.

henra

2nd Sep 2011, 20:46

Hi xcitation,

If he'd done that - it would have continued to roll right wing down until it was inverted.

Lookng at the likelyhood of the reasons for that tendency to roll it seems pretty obvious to me that we can assume that turbulence played a or the major role in this.
If that was the case the probability of a stable continued role to one side until inverted is rather low.
Looking at the outcome I'm with excitation on this one.

grity

2nd Sep 2011, 22:25

steamchicken A thought - when we want to focus on the trend in a particular metric, we plot it on a chart like the DFDR traces. Strangely, aircraft displays show scrolling numbers or else multiple needles. Would a "value plus trend-line" be better? Medical monitoring displays do this. even the interpretation of a medical monitor needs big experience, hehe we discuss since many weeks the lines of less minutos..... but in this case a trendline of the altitude, wouled helped the captain a lot to spot the situation (maby with a future-line calculation for different pitch or powers....?)

rudderrudderrat

2nd Sep 2011, 23:20

Hi OK465, DW, henra,

Unless the aircraft was trimmed exactly to balance any asymmetric thrust (due TL moved in response to thrust lock), it would tend to roll one way and continue to roll, unless counteracted by pilot or Auto Pilot input.
Since it was in Alternate Law, there was no AP or FBW roll input - only pilot input.

At what attitude do you think it would stop rolling if left by itself?

bubbers44

3rd Sep 2011, 01:24

I, by choice never flew the Airbus but friends that do say if you leave the plane in a bank it maintains it forever. Even in alternate law. If that is true it wouldn't roll inverted but I am sure a pilot would correct it if this information is wrong. I am so happy I only flew Boeings. They always do what you want.

OK465

3rd Sep 2011, 02:46

Hi Triple R:

Thanks for the explanation. :)

I see now that releasing the SS in this situation while continuing to manipulate the thrust levers is fraught with potential peril.

Ian W

3rd Sep 2011, 12:12

Rudderrat
Hi OK465, DW, henra,

Unless the aircraft was trimmed exactly to balance any asymmetric thrust (due TL moved in response to thrust lock), it would tend to roll one way and continue to roll, unless counteracted by pilot or Auto Pilot input.
Since it was in Alternate Law, there was no AP or FBW roll input - only pilot input.

At what attitude do you think it would stop rolling if left by itself?

With an aircraft stalled the roll would lead to yaw in the direction of the roll. As the 330 is swept wing the wing on the downside would then be more directly into what airflow there was and generate slightly more lift and the aircraft would nod up and start rolling back the other way to repeat the yaw and roll on the other side. As I believe Dozy has said this is relatively standard swept wing behavior. Like a long period dutch-roll

However, into this action PF was putting in as much aileron as he could, often apparently full SS. The result of this would be to make the low wing AOA higher so less lift so a tendency for the aircraft to apparently do the opposite of what was wanted until the wing sweep effect was larger than the aileron AOA effect.

Perhaps someone with spare time could check - but if the total time with the SS held to the left is larger than to the right then that could explain the overall right turn in the descent track as the right wing was being held deeper in the stall by aileron AOA.

Speculation now - but for a pilot sitting in the right hand seat, just pulling the stick directly toward the pilot would be a Nose Up - Left Roll input. This would maintain the stall and result in a track that turned to the right.

rudderrudderrat

3rd Sep 2011, 13:22

Hi Ian W
However, into this action PF was putting in as much aileron as he could, often apparently full SS. The result of this would be to make the low wing AOA higher so less lift so a tendency for the aircraft to apparently do the opposite of what was wanted until the wing sweep effect was larger than the aileron AOA effect.

I don't know the physics of stalled aircraft behaviour, but I think the stability was a result of dihedral rather than sweep.

When stalled, if one wing is lower, it presents a greater plan surface area to the airflow than the higher wing, resulting in a levelling couple.
I think PF's roll input would help level the wings because the lowered aileron would present more drag. (edit in the same way that down elevator lowered the nose and reactivated the stall warning)

However, I will probably agree with Owain's analysis.

BOAC

3rd Sep 2011, 14:53

Quite a few folk here just a bit confused about aircraft, and in particular swept wing, stability, but I don't actually think any of this is relevant to this thread.

What is a fact relevant to this thread is that attempts to lift ANY stalled wing with aileron will result in the wing going further downwards. This can be seen on the PGF AB320 traces as well.

VGCM66

3rd Sep 2011, 16:13

Absolutely unbelievable but true...:ugh::

AF447's initial altitude drift went virtually unchallenged (http://www.flightglobal.com/articles/2011/08/12/360432/af447s-initial-altitude-drift-went-virtually-unchallenged.html)

RetiredF4

3rd Sep 2011, 16:27

Quote: Ian W
However, into this action PF was putting in as much aileron as he could, often apparently full SS. The result of this would be to make the low wing AOA higher so less lift so a tendency for the aircraft to apparently do the opposite of what was wanted until the wing sweep effect was larger than the aileron AOA effect.

I think it was the increased drag and not the higher AOA, i agree though that it looks like it did not help in leveling the wings.

rudderrudderrat
i don't know the physics of stalled aircraft behaviour, but I think the stability was a result of dihedral rather than sweep.

When stalled, if one wing is lower, it presents a greater plan surface area to the airflow than the higher wing, resulting in a levelling couple.

I tend to agree. Dihedral helped, sweep back did not help.

I think PF's roll input would help level the wings because the lowered aileron would present more drag. (edit in the same way that down elevator lowered the nose and reactivated the stall warning)

Here i disagree again. See my post from few days ago Adverse yaw (http://www.pprune.org/tech-log/460625-af-447-thread-no-6-a-35.html#post6675997)

From follow up post (http://www.pprune.org/tech-log/460625-af-447-thread-no-6-a-36.html#post6677023)
Honestly i don´t know how this missing sideslip estimation influences the behaviour of the rudder in a stall, or how the functioning yaw damping influences it in a positive or negative way. But hopefully somebody can explain.

Would somebody be able to explain on my followup question concerning the function of yaw dampening and missing side slip estimation in Alt2?

BOAC
Quite a few folk here just a bit confused about aircraft, and in particular swept wing, stability, but I don't actually think any of this is relevant to this thread.

I agree inthe way, that it did not play a role causing this upset.

But i strongly disagree in general.
The aerodynamic behaviour of the aircraft seems to be a blanc card in nowadays training and knowledge base. We wonder, wy the pilot pulled the stick despite being in a stall. We might also wonder, why he applied full left aileron and not using rudder to level the wings. It might have the same reason: Not knowing, what the outcome in an aerodynamic point of view) theses actions would be. He wanted to stop the descent and climb up again, therefore pull back on the stick. That worked probably more than 99% of his flying career. Tell the aircraft what to do, point the nose up and apply sufficient power, and the aircraft will follow the orders. Same with bank. If right wing is down, put the SS to the left. When the wing is level, put the stick to neutral and it will stay there. That worked again more than 99% of his flying career.

Unfortunately there are situations, when you have to switch from normal behaviour to "trouble shooting mode", and for that you have to know the performance envelope of your aircraft and how it behaves, if you touch that envelope or go outside of it.

Military fast jet pilots do that on a day to day basis, therefore they know exactly how the aircraft will behave and and how to react by flight control input to get the desired result.

This crew had no clue what might have helped to get out of the situation they themselves flew (stalled) into. Agreed, they should not have gotten into this situation first hand, but that´s what ´s happening in our non perfect technical and human world.

Therefore again, i tend to say, that it is also very important to look into those issues after the stall happened instead of exchanging arguments about tactile feedback and yoke Vs.SS the 50th. time. We placed those statements and they stand from either side.