Go Back  PPRuNe Forums > Flight Deck Forums > Tech Log
Reload this Page >

AF447 final crew conversation - Thread No. 1

Tech Log The very best in practical technical discussion on the web

AF447 final crew conversation - Thread No. 1

Old 23rd Oct 2011, 11:26
  #321 (permalink)  
 
Join Date: Sep 1999
Posts: 542
The big issue with Airbus was and still remains the lack of feedback on controls. They let their aircraft use one single channel to communicate with the pilot, the one through the eye, meaning an intellectual, a serial input to the brain. To a small degree they use the audio channel as well, however in this particular case shows how small: The THS movement is not audio connected, you canít hear it moving, you need to look at it again, with your single serial channel.
This was from Gretchenfrage(a very fine contributor) on the old thread and in my view is critical to the argument.Of course,its shouted down by all the Airbus apologists but its fundamental to understanding the lethal traps that this machine can land the unwary in.Same can be said of spoiler activation on landing..you see,hear and feel the activation of this important retardation device in a conventional a/c.In Airbus,you look for it on the ECAM.

Add the deactivation of the stall warning when the a/c remained in the stall and you can see how the trap was set.But I agree this was pilot error from the moment the Captain left his 2 FO's in charge of wx avoidance at night in the ITCZ to the time that the pitots blocked and the UAS procedure wasn't followed.How very sad.How much Airbus design will figure in the final report's causal factors will depend on BEA's impartiality and integrity.I dont hold out much hope.
Rananim is offline  
Old 23rd Oct 2011, 12:55
  #322 (permalink)  
 
Join Date: Oct 2011
Location: Lions Peak
Posts: 11
I was discussing AF447 with a friend of mine yesterday. He has a master of science in psychology. Reading the final crew conversation and asking me a lot of questions about flight AF447 itself, he came to this conclusion:

The crew was probably in a state of "cognitive bias". They selected and interpreted only those pieces of information, which did fulfill their own theory of what was happening. This mental state blocked all the "easy way outs" many of you described as "nose down / pitch and power". This is one step further from a simple "misinterpretation" of data. It takes a very strong signal from the outside to open up a way out of this mental state. And it takes time to realize, that the "reality" you fully believe you are in, is not the reality happening all around you. The strong signal was obviously missing and time was something, the crew did not have at the end...
Class_Y is offline  
Old 23rd Oct 2011, 13:16
  #323 (permalink)  
 
Join Date: May 2002
Location: GC Paradise
Posts: 1,047
A very interesting first post from Class Y. Thank you for your useful contribution.
FlexibleResponse is offline  
Old 23rd Oct 2011, 13:52
  #324 (permalink)  
 
Join Date: Mar 2002
Location: Florida
Posts: 5,296
I was discussing AF447 with a friend of mine yesterday. He has a master of science in psychology. Reading the final crew conversation and asking me a lot of questions about flight AF447 itself, he came to this conclusion:

The crew was probably in a state of "cognitive bias". They selected and interpreted only those pieces of information, which did fulfill their own theory of what was happening. This mental state blocked all the "easy way outs" many of you described as "nose down / pitch and power". This is one step further from a simple "misinterpretation" of data. It takes a very strong signal from the outside to open up a way out of this mental state. And it takes time to realize, that the "reality" you fully believe you are in, is not the reality happening all around you. The strong signal was obviously missing and time was something, the crew did not have at the end...

translation:

they were confused
lomapaseo is offline  
Old 23rd Oct 2011, 14:01
  #325 (permalink)  
Per Ardua ad Astraeus
 
Join Date: Mar 2000
Location: UK
Posts: 18,582
The 'fly in the ointment' (prefered to references to woodpiles....) in the Psycho's theory is that the Captain arrived on the flight deck fresh to the situation and WITHOUT the 'cognative bias' and should therefore have been able to freely analyse the situation. Otherwise, regarding the PF, spot on, but PNF?? You have two independent 'mental states' (before the third arrived) that need to be 'confused' and 'biased' I do not see evidence that PNF suffered from the same bias. The training is (should be/was in my day) to IGNORE the 'confusion' and focus on the instruments which, as far as we know, were 'telling the truth'.

Your friend does, of course, observe another issue which needs to be looked at either from the presentation of information to the crew or in training.
BOAC is offline  
Old 23rd Oct 2011, 14:48
  #326 (permalink)  
 
Join Date: Oct 2011
Location: Lions Peak
Posts: 11
The 'fly in the ointment' (prefered to references to woodpiles....) in the Psycho's theory is that the Captain arrived on the flight deck fresh to the situation and WITHOUT the 'cognative bias' and should therefore have been able to freely analyse the situation.
That is exactly what I thought and the way I objected the "bias-approach" at first. But then he (the psycho) told me about a series of experiments they did in the army: A commanding officer (CO) is called by his NCOs to handle a specific situation. One group of COs was told to enter the command and control center, call for an elaborate report by his NCOs and then quickly decide on the situation. The second group was told to enter the COC, and gather the required the information by having a "question-and-answer-game" and - when the CO is sure to have all the information he needs - decide quickly on the situation.

If the investigating psychologists fed erroneous information into the process, group one had a significantly higher percentage of COs who made wrong decisions because they somehow trusted the NCOs report and often did not double-check the information given to them.

Now back to AF447: the Captain arrives on the flight deck, everything has gone awry, he asks for a report, time is running fast. He was not on the flight deck when things started. He listens to the report. His training should have made him "step back and reconsider", but his mind is still trying to make sense of the report. Before he is able to come to his own conclusions, the PF (and maybe the PNF) overwhelm him with more (erroneous) information. It may have been very easy to slip into the same "reality" as the PF (and the PNF). Once in, there is no easy way out...

But you are right BOAC: the "man-machine interface" and the specific training should be examined very closely.
Class_Y is offline  
Old 23rd Oct 2011, 15:16
  #327 (permalink)  
 
Join Date: Aug 2011
Location: Grassy Valley
Posts: 2,123
The steady state in this cockpit is data/solution. Normally, it functions quite well, but as we see, it can rapidly degrade into chaos. The pilots are TRAINED to this model, and thinking is discouraged, actively, in favor of memory (rote), and reference (QRH).

To expect pilots to function to one model for virtually their entire career in flying, yet retain a facile thought process when data is missing, is preposterous. As we see, it can be deadly.

With all respect, there are ample exhibits of this regime on thread. Looking for the solution immediately, disregarding evidence that doesn't sustain one's bias, is clear, here.

Then, when there is "consensus" actively defeating any challenge to the dogma? Is that not what the "psycho" describes? Solution Bias.

Speed over quality. Activity before decision. These are deadly artifacts of a philosophy that actively discourages using one's head.

The deadliest symptom of rote behaviour is how quickly and terminally it defeats success, when it is interrupted. "Start over".

It is called a "train of thought", for a very good reason, and when derailed, there is no further chance for rescue.

May as well call for the tug. In mid-Ocean, that is obviously not possible.
Lyman is offline  
Old 23rd Oct 2011, 15:18
  #328 (permalink)  
 
Join Date: Aug 2002
Location: Switzerland, Singapore
Posts: 1,306
The discussion goes in a very interesting direction and I see plently to learn from. One has to consider that not only the stalled aircraft was difficult to recover, but also the crew's state of mind. The closer they came to impact, the more difficult it was to get out of this funnel, this narrowing tunnel.

Like in most accidents, it would have been easiest to escape the spiral (literally and metaphorically) at the beginning (just keep pitch, wings level and thrust), and the more away from the fair way they got, the more difficult it became to save the plane and her inhabitants.

When the captain came to the cockpit, it was already impossible, aerodynamically and CRM-wise.
Dani is offline  
Old 23rd Oct 2011, 15:42
  #329 (permalink)  
 
Join Date: Jun 2009
Location: Bedford, UK
Age: 66
Posts: 1,245
Sounds like 'brain freeze'. Difficult not to suspect an ingrained belief that the a/c 'couldn't stall' prevent proper root cause analysis.
Mr Optimistic is offline  
Old 23rd Oct 2011, 15:58
  #330 (permalink)  
 
Join Date: Jun 2001
Location: UK
Posts: 3,959
The first step in solving a problem is the acquisition of the correct information.

It's no good giving the "right" answer to the "wrong" problem.

I think Lyman has some good points in his last post.

Generalising, much of pilot training is procedural these days (the "training" mindset) rather than original thought (the creative mindset).

The former works well when things are normal but breaks down when things become very abnormal.
fireflybob is offline  
Old 23rd Oct 2011, 16:10
  #331 (permalink)  
 
Join Date: May 2011
Location: BOQ
Age: 75
Posts: 489
...it was already impossible, aerodynamically...
Why is it 'a given' to some posters that the 'stall' was unrecoverable, yet to another group of posters it is 'a given' that all that was required for recovery was an actual nose down SS input of some magnitude & duration?
OK465 is offline  
Old 23rd Oct 2011, 16:21
  #332 (permalink)  
Per Ardua ad Astraeus
 
Join Date: Mar 2000
Location: UK
Posts: 18,582
OK - the answer is that we are in uncharted aerodynamic territory and thus don't know, and subjectivity rules. Even given that, I cannot understand Dani's "When the captain came to the cockpit, it was already impossible, aerodynamically and CRM-wise." since the a/c was above 35k - over 3 mins from impact..

The thing about 'cognative bias' is that it is very similar to instrument disorientation exercises - the teaching is to trust the instruments and IGNORE the 'bias'. Here, of course, the big unknown is what the instruments were telling.
BOAC is offline  
Old 23rd Oct 2011, 17:19
  #333 (permalink)  
 
Join Date: Aug 2011
Location: Grassy Valley
Posts: 2,123
BOAC. Yes. In this case, the instrumentation is not known. Dani (perhaps by mistake) condemns the plane to doom with seven miles to recover. Earlier I said recovery was possible much lower, and you called me a fool.

This was a hybrid Stall, the aircraft had tantalizingly good horizontal speed, and was perhaps close to re loading the wings, (more than just "partially"), but the Descent promised a long carefully controlled pull up to effect the solution.

Who knows what velocity was in store, but when the goose is fully cooked anyway.........

You emphasize instruments, but others adamantly demand Pitch and Power.

You may be confusing "conclusion bias" with Solution bias. The crew here were fully stuck on conclusions, and weren't even close to a solution. It is hard to be patient when you are about to die. Training crews to "Operate the machine".....Wait, let me be precise...."Drive the Machine", to the exclusion of ad hoc and intuitive mind power is (will always) cause tragedies.
Lyman is offline  
Old 23rd Oct 2011, 17:32
  #334 (permalink)  
 
Join Date: Dec 2001
Location: England
Posts: 1,387
After stalling, the angle of attack stayed above 35 degrees so was it really "close to re loading the wings" ?
cwatters is offline  
Old 23rd Oct 2011, 18:29
  #335 (permalink)  
 
Join Date: Oct 2010
Location: UK
Posts: 132
Lyman:

I have read all reported BEA information regarding this tragic accident and have come to the understanding (rightly or wrongly) that the only erroneous indication on the PF's instruments was the Air Speed with all other indications including altitude, pitch angle, roll angle, engine thrusts, heading, etc., all reading correctly. It is also my understanding that the PF predominantly held the a/c in a nose-up attitude manually using the Side Stick while the a/c was both flying and stalled.

When I read your posts I gain the understanding that it is your belief that pitch angle and altitude along with several other indications were all in error thus denying the PF the opportunity to safely fly the a/c by reference to the instruments.

Please explain to me where you have gained your information regarding any faulty indications other than to the Air Speed as I am unable to ascertain from where you have gained this understanding and thus am finding it difficult to agree with you views?
AV Flyer is offline  
Old 23rd Oct 2011, 20:05
  #336 (permalink)  
 
Join Date: Apr 2008
Location: Durham
Age: 58
Posts: 187
AH yes - human factors. I'm not a pilot but have a fair degree of experience in human factors (including the army - for which I will comment on)

Everyone has cognitive bias. It is what enables you to make a decision to cross a road or buy a pie. Mentally we prioritise inputs (not information, which is processed data) and select. Its not in itself a problem unless the parameters get squeezed, such as the need to achieve a goal against the lethality of getting it wrong. If I want to cross a road I am usually going to wait until the traffic has a gap long enough to cross safely. If my wife is being attacked on the other side of the street I may change my parameters rather drastically, and run in front of traffic I wouldnt have done if the need wasnt so pressing. That is not to say that cognitive bias isnt powerful- some people adhere to concepts way beyond dispassionate argument. I just dont think that to label a situation as "cognitive bias" gets anywhere near the state of mind of the flight crew.

I suspect that the complexity of the pressures of cognitive input squeezed by the absolute need to keep the aircraft flying makes the description of "cognitive bias" a little stretched.

I also feel that some physical effects would affect the crew. I imagine that the aircraft was being bounced around a bit, with its corresponding effect upon mental processes irrespective of bias. If I cannot see clearly an instrument because I am being shaken around I will be slower to assimilate the data and process it.

The example of the army briefing had its own bias. Essentially it had little to do with psychology but to information flow - any senior commander could reach a better decision if he had the time or resources to question a source of data. To rely on a single direction of a stream of information (ie a NCO briefing to a CO) is always going to be less effective than being able to engage in a two way exchange of information.
mercurydancer is offline  
Old 23rd Oct 2011, 20:18
  #337 (permalink)  
 
Join Date: Aug 2002
Location: Switzerland, Singapore
Posts: 1,306
"When the captain came to the cockpit, it was already impossible, aerodynamically and CRM-wise." since the a/c was above 35k - over 3 mins from impact..
Why is it 'a given' to some posters that the 'stall' was unrecoverable,
Because it was the very exact moment they stalled, at time 12:11:45, when the captain entered the cockpit.

"impossible" is maybe a bit to stark, I should have said "most probably". It's not only about "would it be theoretically possible" but "could they have find it out that they were stalled" and thus recover. If they knew for sure that they were stalled, it might have been possible of course. Probable, but unlikely.
Dani is offline  
Old 23rd Oct 2011, 20:30
  #338 (permalink)  
 
Join Date: Nov 2010
Location: FR
Posts: 478
Originally Posted by BOAC View Post
OK - the answer is that we are in uncharted aerodynamic territory and thus don't know, and subjectivity rules.
I agree on uncharted aerodynamic territory (for an A330).
On the subjectivity:
- many of the people putting forward that the stall was unrecoverable may have an agenda (i.e. "defending" the pilots, cf AF unions)
- many of the people putting forward that the stall was recoverable may have an agenda (i.e. "defending" the plane)

Known facts :
- some planes (particularly T-tail designs) may encounter a "deep" stall; as the definition of a "deep stall" isn't the same for everybody, let's describe it as a stall-in-which-you-cannot-unstall-the-aircraft-by-pushing-on-your-pitch-control-because-your-elevators-don't-have-any-authority-anymore.
- it is not proven (and, in my opinion, it is unlikely) that the A330 is prone to this kind of stall.
- sims tests cannot be representative (why? because uncharted aerodynamic territory = sim emulates, not simulates accurately)
- AF447 proved that pushing the stick and/or reducing thrust => less AoA, even on a stalled A330, cf FDR traces; it did not prove that the stall was (or wasn't) recoverable, because no sustained attempt was made.

Originally Posted by BOAC View Post
Even given that, I cannot understand Dani's "When the captain came to the cockpit, it was already impossible, aerodynamically and CRM-wise." since the a/c was above 35k - over 3 mins from impact..
+1
AlphaZuluRomeo is offline  
Old 23rd Oct 2011, 20:42
  #339 (permalink)  
 
Join Date: Nov 2010
Location: FR
Posts: 478
Originally Posted by Dani View Post
Because it was the very exact moment they stalled, at time 12:11:45, when the captain entered the cockpit.
Huh?
The aircraft stalled around 02:11:10 as shown by:
- negative V/S
- AoA > 15į
The captain entered the cockpit at 02:11:42, that's ~30 seconds later.
AlphaZuluRomeo is offline  
Old 23rd Oct 2011, 20:48
  #340 (permalink)  
Per Ardua ad Astraeus
 
Join Date: Mar 2000
Location: UK
Posts: 18,582
Originally Posted by Dani
it might have been possible of course. Probable, but unlikely.
- I am just not understanding your posts right now. Recovery, in my opinion, could have been 'definite and likely'. To suggest otherwise is to imply that the elevators could not lower the nose, and heaven knows where that will take yet another 447 thread.

I still have two basic things I do not understand

1) Why the aircraft climbed/was climbed

2) Why not one of the three pilots was able to recognise nose high/huge rate of descent for the fact that it just was not 'flying'. I will give the crew the benefit of the doubt that if they had recognised it they would have taken the correct recovery action although the cock-eyed stall warning logic would not help.

1) is for the investigators

2) is for the training system - teaching crews they still need to be able to fly, to ignore bizarre and misleading audio messages and that the a/c will not protect them regardless.
BOAC is offline  

Thread Tools
Search this Thread

Contact Us - Archive - Advertising - Cookie Policy - Privacy Statement - Terms of Service - Do Not Sell My Personal Information

Copyright © 2018 MH Sub I, LLC dba Internet Brands. All rights reserved. Use of this site indicates your consent to the Terms of Use.