AF 447 Thread No. 12 [Archive] - Page 2

DozyWannabe

14th Aug 2014, 15:17

You have changed the subject. I have not studied the A320 development enough to be able to spout off datum's about it.

I haven't changed the subject, I was simply responding to this :
In industries that produce equipment that can kill people if it goes wrong, those who are careful about adopting new non validated technologies pose far fewer risks than those who accept new technologies without proper evaluation.

with the assertion that the fourth-gen Airbus flight deck layout and FBW systems were very carefully evaluated before adoption. :ok:

Who is disputing its fine safety record?

Roulis, it would appear.

The subject of our original thread drift was about posters’ comments that you nullified using your dubious statistical evidence based on a very small sample base.

Not at all - I wasn't trying to prove anything conclusively via stats, merely backing up my own assertion with multiple anecdotal incidences where connected "conventional" controls did not make a difference, in spite of what those who prefer that layout would claim.

The only conclusion one can draw from your comments is you think these posters are reactionaries. We have differing views.

I think there's a reactionary tendency with some of those posters, yes - but I was speaking more generally. When the A320 was introduced, there was a distinct group of the piloting fraternity who made various claims about how the passive sidesticks and FBW technology were designed specifically to "take pilots out of the loop", and would eventually cause serious safety issues by that same design. Not only was the first assertion untrue, but also neither has come to pass. That's all I'm saying.

jcjeant

14th Aug 2014, 22:05

Not only was the first assertion untrue, but also neither has come to pass. That's all I'm saying.

Methink that removing pilot etc .. will not come because FBW or any automation ... but instead as a logical continuation of statistics studies
Indeed the old statistics indicate that percentage of accidents occurred by structural defects ... then it came the time of mechanical problems (engines and others) and now the statistics show an increasing number of accidents due to "pilot error"
The statistics show that it should therefore (for the better) replace the pilot with something else to solve this problem like it was performed for structures and engines :)

AirRabbit

14th Aug 2014, 23:19

Statistics Studied

Your suggestion might be a very logical progression of thought if ONLY the “old statistics” that showed the defects that occurred in engines, structures, or instruments where corrected by doing away with engines, structures, and instruments. However, airplanes of today have not replaced engines, structures, and instruments with “something else!” We still HAVE engines, structures, and instruments – but the ones that we have now have been evaluated and substantially improved. So … the logic that you have recommended really is the following: We need to evaluate how we train today’s pilots and then substantially improve that training, and thereby, improve the competency of those pilots. Only when we can replace engines with something else, replace structures with something else, and replace instruments with something else, will we even approach a time when we might consider replacing pilots with something else.

Cool Guys

15th Aug 2014, 05:00

I think there's a reactionary tendency with some of those posters, yes....

Hi Dozy, I understand where you are coming from.
There are some seemly very experienced pilots here who post their opinions based on many hours flying with their bums in the pilot’s seat. You think they are just being reactionary and your not so humble opinion is based on “anecdotal incidences”

roulishollandais

15th Aug 2014, 08:03

safety record
We could hope much better records than AF447 and such wonderful flights. : Who is disputing its fine safety record?
Roulis, it would appear. We need to evaluate how we train today’s pilots and then substantially improve that training, and thereby, improve the competency of those pilots.All the problems of automation and systems, decreasing safety statistics by scandalous accidents where pilots lost their airworthyness (AF447, Habsheim, Asiana, aso) in FBW are the result of the choice of regulators, airlines, and manufacturors to hire, select, train ignorant pilots -not fighting fake licenses , forgetting the role of the third man in the cockpit, using less corporate test pilots and test engineers independant from private industry i.e.) aswell in flight (no more aerobatics i.e.), in maths (not increased to dynamic systems used by FBW), and in engineering (former pilots made human teams with mecaniks and aircraft conceptors). That science was open, but today copying and counterfacts of systems have build a world of secrecy and paranoia which destroys spread of knowledge to pilots' community.
That is why I like PPRuNe! who allows us to share better:)) but it is not enough to stop that nightmare of new generation accidents who are so sad, and decrease systems' credibility and statistics which should be better. Remind I'm both a pilot and IT guy, pluridisciplinarity is the limit I trust.

DozyWannabe

15th Aug 2014, 21:51

There are some seemly very experienced pilots here who post their opinions based on many hours flying with their bums in the pilot’s seat. You think they are just being reactionary and your not so humble opinion is based on “anecdotal incidences”

Not at all. To suggest an extreme hypothetical example, if you took a group of people who had years of experience driving a Ford Model T (which had a completely different pedal and gear layout to today's de facto standard) and put them in a modern car, they would have difficulty operating it and would no doubt tell you that the modern car is inferior and less intuitive to operate than the Model T they are used to driving. And from their point of view that position would be absolutely true.

To bring that example closer to the subject at hand, another "crux" point in aviation, particularly in the US, was around the introduction of the Boeing 727 - which made jet operations out of smaller fields (and therefore previously the preserve of propliners) possible for the first time. There were quite a few nasty crashes early in the life of the B727 caused by crews bleeding off too much speed on the approach and not taking into account the extra time required for a jet engine to spool up and provide thrust compared to a prop-driven airliner. This did not mean that propliners were inherently safer - in fact within a decade or so it would become apparent that the introduction of the jet engine was one of the single most significant improvements to safety there had been in civil aviation, but it took time for the necessary changes in approach and attitude to "bed in".

Regarding Airbus, there's a whole political dimension to the subject which is mind-numbingly boring, but the overriding fact is that - just as before - after an initial period of getting used to the technology, the setup has been proven to work well, and as safely as any other setup currently flying.

All the problems of automation and systems, decreasing safety statistics by scandalous accidents where pilots lost their airworthyness (AF447, Habsheim, Asiana, aso)

Oh, please - NWA 6231 (http://en.wikipedia.org/wiki/Northwest_Airlines_Flight_6231) was more-or-less circumstantially identical to AF447, and pre-dated the introduction of civil FBW by 14 years. Habsheim was flown manually, and the Asiana accident had a whole slew of additional contributory factors.

in FBW are the result of the choice of regulators, airlines, and manufacturors to hire, select, train ignorant pilots

Rubbish. FBW refers to an electronic connection between the flight controls and flight surfaces - nothing more.

using less corporate test pilots and test engineers independant from private industry i.e.)

As I said before, Gordon Corps (chief pilot engineer on the A320 project and chief safety engineer) had been a certification pilot for the UK ARB (one of the most respected testing and certification bodies in the world) for around two decades prior to being hired by Airbus - it was that experience that they wanted. And they wanted him to make the product safe, not toe any company line.

and in engineering (former pilots made human teams with mecaniks and aircraft conceptors).

Which was exactly what Airbus did with their FBW line, and I'm sure Boeing did too.

That science was open, but today copying and counterfacts of systems have build a world of secrecy and paranoia which destroys spread of knowledge to pilots' community.

Again, rubbish. The block diagrams in Airbus FCOMs are at the same level of detail as those available to pilots and FEs of prior designs, if not more so.

roulishollandais

15th Aug 2014, 22:52

Divers DozyWanabee
All the problems of automation and systems {...}in FBW are the result of the choice of regulators, airlines, and manufacturors to hire, select, train ignorant pilots "problems" is the subject of that sentence, not "FBW" !!!

Pilots are so ignorant of systems that they are unable to draw or understand a block diagram ; remember the discussions about feedback, THS or SS !!!

I fear BZ used only Gordon Corps's name as alibi for insiders. The book of BZ shows the point of view of BZ. Franz Joseph Strauss was much more important in BZ's mind... behind himself and Baud.

Unions felt -with some realism- to be let outside the human teams. That means the most of the pilots were considered as strangers.

NWA6231 is typically a misjudgement, very different from the FOG in the brains of the crew of AF447, Habsheim or Asiana.

PJ2

15th Aug 2014, 23:33

Hello roulis;

We used to have the DC8's systems memorized. We had to be able to draw the terminal and airport charts of Chicago and New York as part of the route check-out on these terminals. We also had to know Morse Code at 5wpm and draw models of frontal weather from the forecasts and SA's, (now METARS), prior to graduating from our initial airline groundschool of two months.

Gradually, as more complex aircraft entered service, the "Need-to-Know" concept took "knowing"'s place. It wasn't possible to know all there was to know about the airplane one flew. So training priorites had to be adjusted and everyone was making it up as they went, particularly when the Airbus was introduced. Even reading the books didn't fulfill the need to know however, because they were very thin on information. So, (personally) I went to the AMMs to learn, (as I'm sure others did).

Most of the time it worked well. When things are going well, flying is easy, but that's not the reason we're there, (and it's the reason why there will never be a pilotless commercial air transport system that replaces the scale of aviation services we have today).

Today, much of this knowledge is handled by software. Also, we now have an entire generation of pilots in the cockpit who grew up interacting with computers and know them intuitively.

The difficulty arises when the assumptions are made that computers "are" airplanes and "know" aerodynamics, etc, etc.

I think you are expressing some important notions regarding competency, knowledge, capacity and their polar opposites. The mistakes are made by both management and the pilot group, the first in assuming that because the airplanes they purchase have sophisticated software autoflight systems which reduce the need for training and competence, and the second group in assuming a relaxed and cold-soup-accepting stance when it comes to automation and both the requirements and the inviolable rules of aviation regarding personal knowledge, ability, skill and competency. That is the essence that makes a pilot a "professional" vice merely a journeyman in the cockpit.

It is up to both groups to get their act together. By and large, from what I see at conferences and in those associations I continue to maintain with aviation after retirement, that is happening. But we are indeed in a major transition phase between analog and digital flight even as the fatal accident rate remains remarkably, enviously low.

Turbine D

16th Aug 2014, 00:40

Hi Dozy,

Originally posted by Dozy: Oh, please - NWA 6231 was more-or-less circumstantially identical to AF447, and pre-dated the introduction of civil FBW by 14 years.
Would you explain exactly why, in your opinion, NWA 6231 is circumstantially identical to AF447 other than the fact they both stalled and crashed? Exactly what caused the demise of NWA 6231?

DozyWannabe

16th Aug 2014, 01:53

Exactly what caused the demise of NWA 6231?

Summary from WikiP:

Investigators found that the pitot heads had ice damage which caused the crew to receive incorrect airspeed readings. The crew, believing the readings were true, pulled back on the control column and raised the nose, which caused the aircraft to stall.

I figure that's fairly close to AF447 (and Birgenair 301 for that matter)...

Pilots are so ignorant of systems that they are unable to draw or understand a block diagram ;
On what are you basing this assertion?

remember the discussions about feedback, THS or SS !!!
Which discussions? There have been more than a few!

I fear BZ used only Gordon Corps's name as alibi for insiders.
You fear incorrectly. A cursory search of Flight International's archives will tell you that GC was very much the lead pilot engineer.

The book of BZ shows the point of view of BZ.
Which one - "Airbus Cow Boys"?

Franz Joseph Strauss was much more important in BZ's mind... behind himself and Baud.
Doesn't matter what BZ thought - GC was the lead engineering test pilot for the A320 project for the very reason that he understood safety and certification needs inside-out. I have it on very good authority that while BZ was engineering lead in name, he had very little to do with the day-to-day work.

Unions felt -with some realism- to be let outside the human teams.
No, the SNPL had a long-standing grievance with Airbus over the deletion of the FE station from the original A300 back in 1972, and as such sought to thwart Airbus at every turn ever since.

Cool Guys

16th Aug 2014, 10:37

after an initial period of getting used to the technology, the setup has been proven to work well, and as safely as any other setup currently flying.

I don't doubt that.
There were many new technologies implemented in the 320. Obviously most of these new technologies were improvements and overall the end result was a fine machine. However that does not mean all those new technologies were improvements. There is a chance some were in fact a step backwards, even if by a small amount. How can you pick out some anecdotal incidences about the 320 in general and conclude that one of those many new technologies (side sticks with no tactile feed back) is in fact an improvement? Particularly when there are a bunch of experienced pilots voicing their concerns?

Turbine D

16th Aug 2014, 12:37

Original Quote by Dozy Summary from WikiP:

Quote:
Investigators found that the pitot heads had ice damage which caused the crew to receive incorrect airspeed readings. The crew, believing the readings were true, pulled back on the control column and raised the nose, which caused the aircraft to stall.
I figure that's fairly close to AF447
You figured wrong. The NWA crash was caused by the crew forgetting to turn on the pitot heaters before departing KJFK on a cold rainy winter late afternoon. As they climbed out, the pitots froze creating a false high speed instrument reading in the cockpit. They responded to this erroneous reading by increasing the climb angle to the point the aircraft stalled. So how is that fairly close to AF447 that had no speed readings except brief intermittent ones on the way down after stalling from cruise altitude?:(
I think you are taking poetic liberties to support your assertions that don't fit well in the Tech Log, maybe in Rumors and News…

http://www.airdisaster.com/reports/ntsb/AAR75-13.pdf

DozyWannabe

16th Aug 2014, 18:14

@Turbine D - I did say "fairly close", by which I meant basic aspects only - namely frozen pitot tubes -> crew (mistakenly) pulls up -> aircraft stalls and crashes. Obviously not every aspect is the same.

roulishollandais

18th Aug 2014, 11:51

Hello PJ2,

Apologize for my late answer to you post #307. You were doing an acurate description of evolution of pilots' knowledge, functions and actual work.

-Intuitive knowledge of new generation helps to learn but not to overcome every aerodynamic situation like degenerating dutch rolls (I'm seeing that USAF decided to fight them forever if possible) or actual stalls in limited airliners and intuitive knowledge does not protect against the drift of airworthyness due to abusive complexity to manage in some mishap cockpits.

The younger generation of pilots and engineers and managers grew up with software : but we know (at least some of us in IT know) that software grew bad, accepting many bugs appearing in subit countrywide pans of electricity, trains, files with mistakes and losses, subprimes, and money scandals (list is unlimited).
I want to point strongly the fact that software's quality is mostly bad and generally accepts an uncomputed rate of reliability and what it means in real life for victims of accidents.

- The description of the both point of view of management teams and pilots (and pilots'unions) points an ordinary human behaviour : to be safer we set two (or more) levels of responsibility to double the verifications, but quickly, to avoid competition for exemple, every "responsible" man/organisation watches only the half of the chain... and a little later the verifications are no more doubled but divided by two, and if an accident happens you will no more find any "reponsible" person.

Exemple 1 about that responsibility dilution :
After the Ariane 501 crash and his wonderful J.L.Lions report, the engineer Ducrocq commented :"Everyone was responsible, so nobody is responsible" ; final cost was 8 billions French Franks ; in the 12 pages report I noted 99 mistakes had been done, from around 80 types of failures, not far of all the failures you can do in software...

Exemple 2 : In the 1992 Air Inter Ste Odile crash I discovered that the height of the mountain the Bloss on the official Instrument approach chart was false. The 2710 FT summit (IGN) of the Bloss against which the wreckage was found at 2620 FT (BEA report) was inside a highest level line of 2500 FT. It appeared to me that the mistakes was years old, and around 7-8 organisations had shared the responsibility of drawing, publishing, using that chart during these years. More, in the same highest line level of 2500 FT was the very popular well known statue of Ste Odile seen by any and many ATCs every day for departure or arrival (depending of the wind) inside of the arrival or departure cone, but no ATC asked himself how that statue could figure inside the 2500 FT line of their IAC which was the always present before their eyes on their table, main tool. (764m Mont Sainte-Odile ? Wikipédia (http://fr.m.wikipedia.org/wiki/Mont_Sainte-Odile) ).
Happily that official VORTAC approach procedure respected the margins.
The BEA did not strickly lie about that -so that the most of you discover these facts now!- but wrote few very difficult understandable sentences if any about theses differences and read the report without noting the faulty height.

Not only that, Air France drawed their own charts (with ATLAS). Surely they used the wrong chart, rectified the height, but deleted the ICAO level lines, and deleted the mention "FAF" as they did on all their approach charts (1992) before that accident. Again more than 10 organisations drawed, published and used the chart without the "FAF" position ; that loss allowed the plane to descend at 1340 FT after 9 NM to STR VOR instead waiting at 3660 FT until 7NM to STR. Experts said nothing, BEA no-commented turning around again too. Two derogations had been given (BEA report) to draw and use a CDA and everyone forgot it was a VOR-DME (or VORTAC) approach. The expert Belotti said me he ignored the PROMIN ICAO and French law's rule of 15% obstacle clearance and had honesty to no more came to the trial...
Since that trial the list of persons who signed the crash reports (around 15 in that crash) is no more public...

We need to have multiple levels of verifications and cross-checks for some critical points, it is important to don't transform them in crash swisscheeze.

PJ2 we are used to read how seriously you are studying hard the folders where you were and still are involved, and in connection with others in the safety chain, which happily is more than 99.9 or 99.99 ;-) or "enviously low". I wanted to say that, despite working together as you are doing, management and pilots are not in a common balance. The latter pay with their lifes the mistakes -as do the passengers-.

And my obligation, being aware of the high unvisible criticity of flying softwares (which may be analogic too) is to move the eventual danger flag existing in that world, decreasing -in my conviction- the safety level under its possibility.

Going in that new world doesn't allow to leave historical airworthyness in abusive complexity that brain is not sure to masterize in the short minutes, or seconds or less allowed for decision in flight.

Thanks for your always appreciated very professional posts.

DozyWannabe

18th Aug 2014, 19:46

@Winnerhofer:

From the paper you just linked -

“Go ahead, you have the controls. We are still in TOGO eh.” TOGO, short for ‘touch and go’ refers to a process of regaining lift at low altitude, something that occurs when a pilot suddenly abandons a landing and reinitiates ascent of the aircraft at the last possible moment.

Ouch. More research required.

PJ2

18th Aug 2014, 21:48

roulis;

I wanted to say that, despite working together as you are doing, management and pilots are not in a common balance. The latter pay with their lifes the mistakes -as do the passengers-.As I have just finished watching ITVs examination of British aviation and thence the story of Shipwrecks which I found fascinating in the quality and quantity of the parallels with the aviation industry (in terms of private power, invisible accountability and unbridled profit at the expense of thousands), I have to say that aviation has taken/learned a great deal from 19th Century shipping.

It has to be acknowledged that if aviation doesn't make money for it's owners/investors/employees then it will fail no matter how accountable, moral or socially-responsible it is. That's a principle of human endeavour - dreams, imagination and invention may be the seeds and even the energy for such enterprises but it has to survive on it's own. This realization changed the shipping industry, (just read the history of Plimsoll and recall how he was treated by the British Parliament), and I think for many obvious reasons, aviation learned faster.

Today, Safety is no longer a backwater appointment or even a career-ending management position as it was even two or three decades ago, within an airline's bureacracy, and programs that were seen as interfering with profitability and share price are today widely accepted as necessary partners in the enterprise.

Having been involved in management - pilot (union) work for many years, I have seen the positive changes at least in two companies, particularly where data analysis and the larger principles of SMS are "in play". While there is very much a need for the regulator's presence and oversight, SMS is now a far better way to do aviation's business.

One aspect of this question you might examine is the question of what changes a culture. In Canada we had a such a turning point with the takeoff accident at Dryden. The result was a public inquiry into aviation in Canada, led by Mr. Justice Virgil Moshansky called the Commission of Inquiry into the Air Ontario Crash at Dryden, Ontario (Canada) (http://epe.lac-bac.gc.ca/100/200/301/pco-bcp/commissions-ef/moshansky1992-eng/moshansky1992-eng.htm), popularly known as the Dryden Inquiry. It is well worth reading, particularly Vol's II & III. The Dryden Inquiry was a game-changer in Canada; the reverberations from Moshansky's work haven't stopped yet.

I'm not going to dwell on the details here, but the question of change itself is an important one for other country's regulators to examine, if only to reify their own approaches to this partnership of financial success and flight safety, in which pilots and their representatives are full partners and play an important role, even as some who would quietly seek otherwise dismiss pilots' views as "unionism" and "playing the safety card".

roulishollandais

19th Aug 2014, 09:19

Discovering Samuel Plimsoll! Perhaps he is not so far in today's aviation !!!

PJ2

19th Aug 2014, 09:59

I think he has everything to do with aviation, but of course he woudn't have known it! It shows that the human and corporate behaviours and the principles that govern success and failure are the same.

One of the most significant insights developed in aviation safety came from Charles Perrow, when he wrote "Normal Accidents" in 1984. The process of changing our notions about "pilot error" and our notions of cause and blame began with Perrow and today are an integral part of SMS.

The other very significant change that arose out of the Dryden Inquiry, (as well as out of other industries such as coal mining - ...yes, there is a connection!), and which became a crucial part of SMS from the start, was the notion of the "Accountable Executive". In Canada it was called the "Westray Law" after a mining accident (https://en.wikipedia.org/wiki/Westray_Mine)killed 26 miners and the courts held the executives accountable because they knew of the dangers and ignored them.

The notion of normal accidents is in contrast to the notion of "pilot error". The understanding that accidents occurred when people thought they were doing exactly the right thing opened the door to new understandings of why accidents occur. The concept of "organizational accidents" arose out of Perrow's work.

Although preceded by Bill Starbuck's paper, "Challenger: Fine-tuning the Odds Until Something Breaks (http://pages.stern.nyu.edu/~wstarbuc/mob/challenge.pdf)", the finest explication of these notions is, in my view, Diane Vaughan's seminal sociological work, "The Challenger Launch Decision".

There is a tremendous history behind the spectacular drop in fatal accidents and hull losses seen in the Airbus paper posted by A33Zab, (it should be noted that Boeing has been doing this same kind of study for a long time; the latest issue can be found here (http://www.boeing.com/news/techissues/pdf/statsum.pdf)), began to change with the emergence of these important, broader notions about why machines and people fail. It is only the lawyers who seem to continue to focus on those who arrived at the accident site first...the pilots.

It is getting easier to make the business case so that shareholders, management and lawyers alike see that spending money on "nothing happening", (i.e., no accidents), is a good thing. These enlightened approaches, still resisted by corporate interests and their lawyers, are gradually showing that there is profit in operating safely, mainly because these days accidents do indeed cost far more than their prevention.

roulishollandais

19th Aug 2014, 10:37

I don"t think that "safety increases profit" ! Accidents increase PIB, GDP much more. The benefit of safety is only marginal and false money of bankers jumps in the safety closed loop.

SMS is surely a great improvement, but must be seen as a procedure not a social procedure, but a brain understanding unless the lawyers jump in the loop with worlds and worlds !

I still don't see the tremendous improvement of safety in Airbus statistics posted by A33Zab as due to FBW, but the pages 9 and 10 showing the excessive long time 10 years to stabilize the FBW accidents (not enough testing of flying software) and the decreasing level of pilot (accidents' rate x3 from 1994 to 2013) in not limited A/C.

I don't want to consider safety issue in aviation as culture problem. It is not the ICAO concept after a World War. Accidents result of actual facts. "Culture" the best way to mix culpability to gestures,isn't?

Nearly everybody is convicted inside of himself that he is doing the right thing...

DozyWannabe

20th Aug 2014, 00:03

but the pages 9 and 10 showing the excessive long time 10 years to stabilize the FBW accidents (not enough testing of flying software) and the decreasing level of pilot (accidents' rate x3 from 1994 to 2013) in not limited A/C.

As I said before, you're reading that wrong. For one thing there have been zero accidents in which the FBW software was a major factor since the A320's introduction over 25 years ago. Every new generation of jet has had a greater number of accidents and incidents in the early years of operation.

I'll say it again - the dotted line in the graphs indicates that less than a million flights of that type were made, so the ratio will appear higher if the number of accidents is divided by that lesser number. This is why the line is dotted, because the statistical baseline for the graph has not yet been met.

This also applies to the dotted line referring to second-gen jets - the ratio is skewed by the minimum required number of flights (a million) not being reached - not because piloting abilities without assistance have been reduced as you seem to think.

Again - the other main reason the second-gen chart climbs slightly before dropping below a million flights is because older aircraft start to have more technical issues as they age, and this is compounded by those older aircraft being sold on from major airlines to smaller operators, and to developing countries, where the safety infrastructure is not as rigorous.

[EDIT: @roulis - I've already explained how the points you make about software aren't really valid when talking about the Airbus FBW system. I'm not going to repeat meself, but here's a link to the first post : http://www.pprune.org/tech-log/528034-habsheim-10.html#post8237990 ]

_Phoenix_

21st Aug 2014, 00:39

Dozy, please revise the FDR graphics of XL888 and AF447. How is it possible that THS is maintained at max nose up position while AoA > AoAmax = stall warning on, for a prolong period of time? Obviously, this situation is not an accidental/false stall warning due gust.

Why the code doesn't implement a line:

If THS<-10 AND AoA>AoAmax Then MOVE(THS=0, rate X(AoA))

...otherwise “Say a prayer”

roulishollandais

21st Aug 2014, 15:57

I had to reread Bernard Ziegler-Rizzi's book in detail, about Gordon Corps. He is mentioned page 78 testing C* on the A300 via A/P , and page 80 and picture p.80 the first A320's flight.
Page 78:
"Le joystick vola d'abord sur Concorde, mais ne fut pas retenu sur un avion qui en était encore pour l'essentiel aux calculateurs analogiques. C'est aussi sur Concorde que fut inventée la loi de commande de vol appelée «C star», qui, seule, permettait d'utiliser un manche aux débattements limités. Elle consiste à piloter non pas une position de gouverne mais directement la trajectoire : sans ordre, l'avion reste aussi stable que vous lorsque vous êtes immobile sur vos pieds (à condition de ne pas avoir bu, bien sûr, mais les calculateurs ne boivent pas). Si vous tirez vous sautez, si vous poussez vous plongez. Il fut ensuite essayé sur un A300, à travers le pilote automatique, le copilote gardant un contrôle par le manche. «normal» à titre de sécurité. Ce fut l'excellent Gordon Corps qui mena une grande part de ces essais. Nous avions même dressé un rideau entre le pilote testeur à gauche et le pilote de sécurité qui, à droite, gardait la commande classique, pour que les réactions du volant n'influent pas le jugement. Devenu plus tard officier de sécurité sous la direction d'Yves Benoist, venu là exercer son étonnante connaissance de l'avion acquise au service de la réception, Gordon mourut en 1992 dans l'Himalaya à plus de quatre mille mètres, en se rendant à pied sur les traces d'un accident."

Page 80 with a picture (from left to right : Jürgen Hammer, Gordon Corps, Gérard Guyot, Pierre Baud, Bernard Ziegler, Jean-Marie Mathios -mécanicien navigant-).
"Pierre Baud était commandant de bord et il y avait deux copilotes, le regretté Gordon Corps, qui avait beaucoup participé à la définition et aux essais préliminaires sur l'A300, et votre serviteur, qui tenait à être à bord d'un avion qu'il avait porté sur don berceau. Jürgen Hammer, ingénieur navigant et pilote de motoplaneur, était de la fête, pour assister Gérard Guyot devant l'installation d'essais".

DozyWanabee, do you know who was sitting in the right seat and in the jump seat from the two copilots ?

About your last questions it seems that our interpretations are different. No problem.

DozyWannabe

21st Aug 2014, 19:12

Dozy, please revise the FDR graphics of XL888 and AF447. How is it possible that THS is maintained at max nose up position while AoA > AoAmax = stall warning on, for a prolong period of time?

Why the code doesn't implement a line:

If THS<-10 AND AoA>AoAmax Then MOVE(THS=0, rate X(AoA))

Two reasons.

The code is not keyed in in that manner, as I said in the linked post - the logic paths are mapped graphically in the development software, and the actual code is generated by machine from pre-tested blocks of logic. These blocks are then further tested as a module, again in conjunction with the other modules and finally regression tested as a whole - a process that at the time could take several weeks per module, and could well end up being revised based on what the test pilots brought back from flight testing. The A320's systems took 6 years (plus a further 2 years beforehand on the A300 testbed) to develop for a reason! :ok:

(Perhaps more importantly) In Alternate Law (AF447), the computers defer to the input of the pilot - in this case the THS moved and stayed put because the PF was pretty much consistently pulling up. If he'd let go, then the "soft" protections would have started returning the THS towards the centre, and if the pilot had pushed forward for long enough, the THS would have moved forward significantly faster. In Direct/Abnormal Attitude Law (XL888), autotrim is off and all THS movement must be ordered by the pilot via the trim wheels.

DozyWanabee, do you know who was sitting in the right seat and in the jump seat from the two copilots ?

Well, we're talking about the A300 test bed from what I can make out, not the A320.

But regardless, I'm sure BZ would have been present for the "photo-op" flights, as befitted his title. But the actual grunt work of doing the thousands of hours of flight testing, as well as refereeing the discussions between the engineering pilots and the aero/software engineers would have been done with GC in charge. Baud was chief test pilot, so of course he'd be in the LHS a lot of the time - and for the earlier test flights it would make sense to have GC in the RHS as monitoring pilot, as it was his job to evaluate the systems and how well they were working, so he could take reports back to the engineering teams and fine-tune them.

About your last questions it seems that our interpretations are different. No problem.
With respect, it's a little more complicated than that. It seems to me that you've come up with a conclusion based on your own beliefs (namely that FBW technology has eroded piloting skill), and are reading the graphs in such a way that fits that belief. What should be happening is that the data should be used to draw conclusions, not the other way round. The interpretation I presented (with regard to ageing airframes and older jets being sold to less safety-conscious markets) is not actually my interpretation, but those of many articles, papers and books I've read on the subject over the years - most of which I hope were written by people who had access to the supporting data needed to draw those conclusions.

RetiredF4

21st Aug 2014, 20:03

DW (coloring by me)
(Perhaps more importantly) In Alternate Law (AF447), the computers defer to the input of the pilot - in this case the THS moved and stayed put because the PF was pretty much consistently pulling up. If he'd let go, then the "soft" protections would have started returning the THS towards the centre, and if the pilot had pushed forward for long enough, the THS would have moved forward significantly faster. In Direct/Abnormal Attitude Law (XL888), autotrim is off and all THS movement must be ordered by the pilot via the trim wheels.

Dozy, where do you have that (colored part) conclusion from? From year long discussions on several AF447 threads I remember, that the FBW system in ALT 2b, which was the mode AF 447 defaulted to, still strives for flight path stability in the pitch axis and thus in the absence of most all protections in the mentioned flight LAW would try to maintain the selected flightpath (which was most of the time an unhealthy climb) despite the unavailability of the required speed, although when the stick would have been released to neutral position. Remember, in normal LAW you change the flight path to a specific climb rate or bank angle, release the stick to neutral and the FBW logic would maintain this bank angle or climb rate until the pilot intervenes by a new command or the protections cut in. The same happens in ALT2b, except that there are no suitable protections active.

Your description would fit the pitch direct mode, (because it it would not have trimmed nose up anyway) which never latched.

DozyWannabe

21st Aug 2014, 20:30

Hullo F,

Yes, you're right about "striv[ing] for flight path stability in the pitch axis", but from http://www.airbusdriver.net/airbus_fltlaws.htm:

ALTERNATE LAW
A low speed stability function replaces the normal angle-of-attack protection

System introduces a progressive nose down command which attempts to prevent the speed from decaying further.
This command CAN be overridden by sidestick input.
The airplane CAN be stalled in Alternate Law.
An audio stall warning consisting of "crickets" and a "STALL" aural message is activated.
The Alpha Floor function is inoperative.

This low-speed stability function operates in both major modes of Alternate Law as I understand it. The important points with regard to AF447 (bolded) are 1 and 2 - namely that the mode will bring the nose down when low speed is detected (as would most probably have been the case when the pitot tubes shed the ice), and that because this is a "soft" protection in a degraded control law (and therefore the possibility exists that the human pilot may know something the degraded systems don't), the mode can be overridden by sidestick input. Because the THS trims to follow the nose-down input, it should have come forward if the PF let go of the stick and allowed the protection mode to operate.

I'm pretty sure I mentioned the existence of "soft" protections in the original mega-thread...

Disregard the following EDIT section, people, it appears that if Alt2B was latched, Low Speed Stability never returned...

[EDIT: Taking a look at the DFDR traces, the speeds came back online from approx. 02:10:38 until approx. 02:11:45 - at which point the airflow generated by the stall starts fouling the pitot tubes. The latter point is around the same time the THS hits the NU stop. The THS starts its roll NU at approx. 02:10:55, which leaves a window of about 40-45 seconds in which the low-speed stability function could have arrested and/or reversed the THS movement before the stall rendered the air data unusable again. Furthermore, releasing the stick between 02:10:38 and 02:10:55 presents a 17 second window in which the THS roll could have been avoided entirely. ]

DozyWannabe

21st Aug 2014, 22:30

@Winnerhofer:

Doesn't need to cover sub-modes, Low Speed Stability is available in all modes of Alternate Law.

Yes the A330 series is architected differently (PRIMs and SEC[ondarie]s replace ELACs, SEC[spoiler/elevator]s and FACs in the A320 series), but the overall logical setup in this case is more-or-less the same.

Sorry I couldn't find a better source at short notice but:

A330 Systems Parameters flashcards | Quizlet (http://quizlet.com/4934635/a330-systems-parameters-flash-cards/)

Alternate Law Protections

Amber lines on AOB and Pitch bars showing no protection in those areas
Load Factor
High and Low speed stability only and can be overidden

See also:
http://tinyurl.com/mbjpdtj
http://www.skybrary.aero/index.php/Flight_Control_Laws

etc. :ok:

RetiredF4

21st Aug 2014, 22:37

Come on Dozy,
you can do better than that.

From FCOM A 330/340

In certain failure cases, such as the loss of VS1g computation or the loss of two ADRs, the longitudinal static stability cannot be restored at low speed. In the case of a loss of three ADRs, it cannot be restored at high speed.

AF447 Final report:
The flight control law changed from normal to alternate at about 2 h 10 min 05. The alternate law was 2B and it did not change again thereafter. Due to the rejection of the three ADR by the flight control computers (PRIM), the abnormal attitudes law could only have been triggered for criteria relating to inertial parameters, but these conditions were never met.

There was no low speed stability function available in Alt 2b. A neutral stick would not have led to a reduction in trim. With the kind of tripple ADR failure Direct Law should have latched and thus disabling autotrim, but this did not happen as all three ADR's were rejected simultaneously.
But i'm not going to go through this again after all these years.

DozyWannabe

21st Aug 2014, 22:53

Are "longitudinal static stability" and "low speed stability" one and the same thing then? It may be, and you may be right, but I'm not 100% sure.

The wording is a little vague - it does not state what happens when the three ADRs come back online. I was pretty sure that it would bring the soft protections back, but I could be wrong.

[EDIT : This link : http://webcache.googleusercontent.com/search?q=cache:GJKBom42EaUJ:www.smartcockpit.com/download.php%3Fpath%3Ddocs/%26file%3DA_0-Flight_Laws.pdf+&cd=1&hl=en&ct=clnk&gl=uk

seems to suggest that Low Speed Stability is lost in the event of double or triple ADR failure, but I'm having trouble finding material confirming what happens when the ADRs come back online. I'm *guessing* that "longitudinal static stability" in the context of the FCTM is referring to the requirements of restoring High AoA Protection mode (and thus Normal Law), but I admit it'd definitely be handy to know for sure. ]

[EDIT 2 :
but this did not happen as all three ADR's were rejected simultaneously.

At 2 h 10 min 08, the two flight directors disappeared and auto-thrust was lost. This time interval was probably due to the fact that the three ADR were not rejected at the same time in the two FMGEC.

(I must say that the Final Report seems to drop some useful background info from the 3rd Interim, and the wording regarding Alt 2B is a bit sketchy in places - though if it is correct in that 2B denotes loss of low speed stability, and that the law never changed from 2B, then you're right - and it was missing, my apologies.)
]

_Phoenix_

22nd Aug 2014, 03:29

if the pilot had pushed forward for long enough

How much is long enough?

see FDR at 2:12:32 to 2:12:42. Bonin pitched down +4 deg but elevator never passed -15 deg. Commonsense says that 10 sec. is more that enough for a crew in distress.

(XL888), autotrim is off and all THS movement must be ordered by the pilot via the trim wheels.

see http://www.icas.org/ICAS_ARCHIVE/ICAS2012/PAPERS/605.PDF

***quote
When the airplane enters a stall, the load factor and the pitch rate feedback responses tend to fall off, due to lift loss and nose-down pitching moment. Consequently, the elevator is more and more driven nose-up, due to integration of the increasingly large C error, in particular, if a nose-up stick input is maintained. Operation of the automatic stabilizer trim function will exacerbate this condition to the point where both the elevator and the horizontal stabilizer can end up on the nose-up stop, unless the pilot reverses his stick input early enough to overpower the adverse C feedback error and drive the elevator to a nose-down position.

these characteristics may have played a role in the two accidents, making recovery from any full blown stall for a C or CU based FBW design difficult, even for experienced test pilot

The pilot’s awareness of the horizontal stabilizer position is of crucial importance to flight safety, especially when the automatic stabilizer function is suddenly lost in flight conditions close to the edge of the envelope***

Therefore, the options are:
- more and more recurrent training for pilots, make them supermen
- redesign of the trim wheels for more visual awareness, maybe color spectrum from up to down pitch, with blinking amber for MAN Pitch mode or extreme positions

However, I still believe that the implementation of the line of code posted previously is a feasible option for the issue.
Sometimes, the cowboy style is the best option.

roulishollandais

22nd Aug 2014, 05:08

The interpretation I presented (with regard to ageing airframes and older jets being sold to less safety-conscious markets) is not actually my interpretation, but those of many articles, papers and books I've read on the subject over the years - most of which I hope were written by people who had access to the supporting data needed to draw those conclusions.I'm writing from my experience not from what I'm reading here and there without having to prove the facts, writing black when they know it is white, shouting up if the publisher requests to shut up, and not present at the Court to testify, or rewriting again and again the wrong buzz like about the failures of the both Ste Odile approach charts.
I'm amazed to see these new refs after years of threads and official investigations.

RetiredF4

22nd Aug 2014, 07:18

DW,
We discussed the matter in great detail 2 years ago,

http://www.pprune.org/tech-log/489774-af-447-thread-no-9-a-29.html#post7303835

and my post there still holds.

http://www.pprune.org/7303835-post578.html

jcjeant

22nd Aug 2014, 13:19

I still believe that simplicity is the solution
Full automatic when all is well
Full manual when something goes wrong (my aircraft)
All Intermediate laws are a labyrinth in which pilots often get lost
But of course why make it simple when you can complicate all !
Of course it was certainly gratifying for Airbus programmers and technicians to show their technological know-how ......

DozyWannabe

22nd Aug 2014, 16:38

@RetiredF4:

Looks like that post came in the middle of one of my self-imposed "sanity breaks", and I must have skimmed over it when I came back. Apologies for missing it, as I said.

@_Phoenix_:

If you look, the "Low Speed Stability" basically does what your proposed "line of code" suggests, except it pitches the aircraft down rather than moving the THS directly (which makes sense given that the autotrim follows pitch commands). Now, as RF4 states, it appears that Alternate 2B loses all but the load factor protections due to the rejection of all three ADRs (a double failure will keep High Speed Stability) - looking at the Flight Law table I linked to earlier, it does not seem to preclude the stability protections from returning after a time, it may just be that there was not enough time in which the airspeed data came back to do so in this case.

This seemed odd to me at first, but having thought about it for a bit I'm guessing that the design team considered an inappropriate automatic pitch down in the event of all 3 ADRs agreeing on bad data more of a risk than eliminating the low speed stability protection entirely. Again, this makes sense if you bear in mind that they'd assume the pilots to be making appropriate inputs, and not consistently pitching the aircraft up into and beyond the stall regime.

@jcj - The pilots didn't get "lost" in the modes, they simply never properly assessed the situation or came up with a solution. For whatever reason, the PF seemed to have become fixated on pulling the nose up - going as far as to hit the sidestick priority button several times (and thus shutting out the PNF), and at the same time told neither the PNF nor the Captain what he was doing and why.

The only time flight laws ever came up on the CVR was the PNF's call of "Alternate Law" at the start of the sequence.

A33Zab

22nd Aug 2014, 17:04

Terminology
ADR = Airbus
ADIRU = Boeing

Airbus & Boeing = ADIRS (Air Data Inertial Reference System)

Airbus ADIRS = 3x ADIRU
Boeing ADIRS = 1 x Dual Channel ADIRU + 1x Dual Channel SAARU

DozyWannabe

22nd Aug 2014, 17:26

@A33Zab - right.

@Winnerhofer - ADR in this instance refers to the Air Data reference module of the combined ADIRU unit (I in this case referring to the Inertial reference module).

I'm writing from my experience not from what I'm reading here and there without having to prove the facts, writing black when they know it is white, shouting up if the publisher requests to shut up, and not present at the Court to testify, or rewriting again and again the wrong buzz like about the failures of the both Ste Odile approach charts.
I'm amazed to see these new refs after years of threads and official investigations.

Sorry - didn't see this earlier. I don't think there's that much new here in terms of data, just a bit more background info. To be honest, a lot of what's here was probably in the older threads as well, but I don't think many people are nuts enough to want to wade through thousands of posts in their entirety.

I must confess I don't share your cynicism regarding all reporting (those written by those that know aviation, anyway). Sure, some of the books I read got things wrong, and the way I've found that out over the years has been to do even more reading - many people on here have helped with that as well. That said, I think some of the errors in those books were either genuine mistakes, or simply that they were writing about events that had not yet been thoroughly analysed. In today's world of self-publishing, there are certainly likely to be more published works that present an agenda rather than simple reportage, but you can generally spot those when comparing other works on the subject. Hell, M. Asseline's book is probably one of the most agenda-driven works in modern aviation history!

Regarding the approach charts, I think there are several reasons why that aspect became a footnote - firstly that the FCU input mechanism was a glaring issue that needed to be fixed, and that took up many column inches at the time, secondly that all the SNPL hoopla about Airbus FBW was being pushed hard at the time, and also because if I recall correctly, the flight was being vectored by radar from the ground, and given the high workload in the flight deck regarding changing approach procedure it is unlikely that they based their flightpath on the charts. Significantly I believe the radar controller actually vectored them incorrectly (turned them onto finals a little too early), so they were laterally adrift of the localiser as well as descending too rapidly.

jcjeant

22nd Aug 2014, 19:39

The only time flight laws ever came up on the CVR was the PNF's call of "Alternate Law" at the start of the sequence.

From memory ... all the FBW Airbus accidents or incidents held in one of the alternates laws .. not in normal law
For all those accidents incidents it is commented "pilots error"
The price paid was victims
Memory again .. the only time alternate law saves lives is the crash of the Hudson

DozyWannabe

22nd Aug 2014, 20:11

From memory ... all the FBW Airbus accidents or incidents held in one of the alternates laws .. not in normal law

For all those accidents incidents it is commented "pilots error"

Sorry, I'm not sure what you're describing, are you referring to just the FBW Airbus accidents that were outside of Normal Law ore are you saying that all FBW Airbus accidents happened outside of Normal Law? (if so, they didn't)

In any case, please give us a couple of examples, because I know the investigative agencies don't tend to use "pilot error" anymore in their reports - in fact in the case of BEA investigations, they're not allowed to attribute responsibility at all.

If you're referring to how the press report these things, then that's just down to lazy journalism.

Memory again .. the only time alternate law saves lives is the crash of the Hudson

Alternate Law doesn't save or take lives, it's just a reconfiguration mode for degraded systems, similar to what the Flight Engineer used to do in the days of 3-crew cockpits.

roulishollandais

22nd Aug 2014, 20:24

@DozyWanabee
Many things to say about wrong books, reports, analysis, SteOdile, aso.
We need some cynism to keep the eyes open and the mind clear when we see people dying for nothing.

Just one word about Ste Odile flight's ATC , as I would not like that the reader of our thread would get misinformed again : The flight of Ste Odile was NOT RADAR VECTORED. After the trial there is no more doubt about that for anybody. The approach was a VORTAC /VOR-DME approach. The controller did a immediately corrected mistake and showed as not being considered by anybody as connected to the crash (sense of turn, the BEA had to note the fact and noted it, ask question, but nothing followed).

During the trial we discovered that military police searched the Controller immediately without he was asked anything to help to find the plane still missing.
The military lawyer of the controller said me they didn't want to attack Air France for the wrong chart (probably because the IAC was wrong too on that military airfield open to civilian planes).

"Une injustice faite à un seul est une menace faite à tous" - Montesquieu

DozyWannabe

23rd Aug 2014, 01:42

We need some cynism to keep the eyes open and the mind clear
Healthy scepticism is a good thing, but cynicism not so much in my opinion.

Just one word about Ste Odile flight's ATC , as I would not like that the reader of our thread would get misinformed again : The flight of Ste Odile was NOT RADAR VECTORED. After the trial there is no more doubt about that for anybody. The approach was a VORTAC /VOR-DME approach. The controller did a immediately corrected mistake and showed as not being considered by anybody as connected to the crash (sense of turn, the BEA had to note the fact and noted it, ask question, but nothing followed).
OK, we're off-topic here, so I don't want to get too much into the subject, but from what I understand (based on what I've read and the portions of the report I've translated), the controller offered radar vectors to the ANDLO beacon (from which the VOR-DME approach could be made), but turned them slightly too early, and then used non-standard language when asked to confirm their position. This is in everything I've ever read on the subject.

During the trial we discovered that military police searched the Controller immediately without he was asked anything to help to find the plane still missing.
Right, and you can thank the SNPL and M. Asseline for that, because the gendarmerie were absolutely determined that there would be no margin for error this time.

The military lawyer of the controller said me they didn't want to attack Air France for the wrong chart...
So let me get this straight - you're asking me to disregard pretty much everything I've read on the subject on the basis of your word that a military lawyer told you something.

Forgive me if I'm a little sceptical... ;)

roulishollandais

23rd Aug 2014, 17:19

@DozyWanabee
Today the trial and appeal are completed. You should find the Final Judgment somewhere. It is public. I myself have attended the both trials of Colmar. A point was even discussed on "judicial truth" and "scientific truth" but not the "journalistic truth"! The trial and judgment have addressed all of these things with bright and talented lawyers whose names are now well known.

I understand your scepticism : that is the way it happened . I thought the military Lawyer would talk from the Air France approach chart for defense, ans days passing I finally asked him when ... His answer was "We don't want attack Air France". I was amazed how he would defend the military Controller : He did that perfectly, answering to all the questions of the Court and other Lawyers, but not more and finally said that Military had to decide too. (Me Soulez-Lariviere defended Mr Frantzen (SGAC) in a similar way with administration).

These Lawyers did their job very well.... As did the Airbus Lawyers team and the Court who has been able to find a way of Justice threw that big garbage of things written or hidden.
Victims were less happy and let me add that they took a wrong way :
1. Creating an association of victims after a crash cannot be used in French Law to get damage (we don't have Class Action)
2. Trying to manipulate the Court "to get more money" (from Airbus) was a very very bad idea.
Not only they didn't show a criminal act from the five persons, but they missed a better damage...

Bpalmer

24th Aug 2014, 05:48

The simulators have no accurate model to duplicate the behavior of the airplane post stall. Sure, they will exhibit a some sort of behavior (and I have personally flown it in several A-330 simulators), but if it reflects the actual airplane is unknown.
There are however, efforts to design at least a generic stall response for inclusion in flight simulators. Recent US FAR changes (that go into effect in 5 years) will require training in recovery that depends on this and other simulator enhancements just now being hashed out.

AF447 was a wake-up call in many ways. Many parts of the industry have woken up, but the fix is not complete.

For now, practice by turning off the flight directors and autothrust and flying some good old basic attitude instrument flying.

The flight directors are doing 85% of the brain work if you're hand flying with them on. Don't be a flight director addict!

Bpalmer

24th Aug 2014, 05:55

The flash of light contributes to the theory that an explosion is what brought down AF447, which was carrying 228 people from Rio de Janeiro to Paris.”

There is absolutely no evidence in the flight data to support this theory. Where was this aircraft that they could have possibly seen them?

Check out this animation that shows the other aircraft in the area deviating around the weather that helped bring AF447 down. (note the animation does not load with Chrome - works good in IE
Flight Paths of Flight AF 447 and of the flights that crossed the zone around the same time (http://www.bea.aero/en/enquetes/flight.af.447/trajectoires/trajectoires010609.html)

noske

25th Aug 2014, 09:05

OK, we're off-topic here, so I don't want to get too much into the subject, but from what I understand (based on what I've read and the portions of the report I've translated), the controller offered radar vectors to the ANDLO beacon (from which the VOR-DME approach could be made), but turned them slightly too early, and then used non-standard language when asked to confirm their position. This is in everything I've ever read on the subject.
Off-topic indeed, but this might be a good opportunity to mention that that accident (and Bangalore, too, btw.) is featured on the FAA "lessons learned" web site. They have an animation of the flight path and even an English translation of the full report (from which I'd recommend section 21.5, "Reconstruction of the most probable scenario").

Lessons Learned (http://lessonslearned.faa.gov/ll_main.cfm?TabID=1&LLID=57)
Air Inter Flight 148 Flight Path Animation (http://lessonslearned.faa.gov/AirInter148/AirInter_FlightPath_pop_up.htm)
http://lessonslearned.faa.gov/AirInter148/Accident_Report_Eng.pdf

roulishollandais

25th Aug 2014, 14:41

The investigation concluded that, based on conversation on the cockpit voice recorder, the descent was intended to be performed at a flight path angle of 3.3 degrees. However, the investigators noted that the actual commanded descent was 3,300 feet per minute, and may have been the result of having inadvertently selected the vertical speed mode on the autopilot control panel. In this case, minus 33 would have appeared on the autopilot display, implying a 3,300 foot per minute descent, while in the flight path angle mode, minus 3.3 would have appeared in the same window.That is the anti Airbus brigade theory! We don't CFIT due to high vertical speed but descending under the bottom of the approach !

Flying at 220 Kts a 15% slope requests 3300 Ft/minute, as we pilots all know (220 x 15 = 3300)
The crew was not aware that , descending step by step the VOR-DME approach and not the derogatory CDA, altitude 3660 FT was mandatory at 7 NM to STR VOR as "FAF" was not indicated.
The crew respected "don't descend 4300 FT before 9 NM to STR due to possible false GPWS warning", and the next mandatory step seemed to be 1340 FT passing STR , without the three magic letters "FAF". Descending the 15% slope from 4300 FT to 1340 FT they met the ground.

Sorry for off topics.

Chris Scott

25th Aug 2014, 16:20

Bon soir roulishollandais,

I am not following this current AF447 discussion - just occasionally dipping in - and I certainly do not want to get involved in a latterday argument about Mont St. Odile and the ergonomics of the FPA/VS switching and displays on the original A320 FCU and FMAs. Suffice it to say that they were much improved by modifications resulting from the investigation into that accident, as I remember only too well.

But I think you might have another look at your sentence:
Flying at 220 Kts a 15% slope requests 3300 Ft/minute, as we pilots all know (220 x 15 = 3300).

I do not think so. You may remember that the rule of thumb for a 3-degree slope is that the required VS (in ft/min) is five times the groundspeed in knots. (I think most of us mentally divide the speed by ten, and then halve the answer to obtain the VS in hundreds of feet-per-minute.) At 220 kt that gives 1100 ft/min (add a tenth of that for a 3.3-deg slope.) Using the one-in-sixty "rule" (approximation), a 15-deg slope would demand about about five times as much, i.e., 5500 ft/min.**

** Non instrument-rated readers please note, the above are approximations only.

roulishollandais

25th Aug 2014, 18:52

Dear Chris,

220 kts. 3° = 5% (standard ILS) 220 * 5 = 1100 Ft/men

220 kts. 15% 220 * 15 = 3300 Ft/min

The protection /limit slope is 15%, not 15° (PROMIN)
The crew displayed that VOR-DME 15% limit as Vs 3300 FT/Min, not the CDA 3.3° they did not want to use.

Are we together now?

Of course they improved the ergonomics and displays, that does not prove it was a factor or cause in the CFIT.A question had emerged, why not do better ? That is the logic of annex 13 recommendations.

Chris Scott

25th Aug 2014, 22:49

Apologies, roulis,

Yes, I misread your 15% slope as 15 degrees. (Maybe I need new reading spectacles!) Must admit I am not accustomed to seeing descent profiles or glide-slopes expressed as percentages... Climb gradients always have been, of course.

As already stated, I am not commenting here on the ergonomics of the A320 FCU, but do not disagree with your last paragraph. Having flown similar approaches on the type before and after the modifications, I'm broadly familiar with the accident scenario and the main hypothesis.

DozyWannabe

26th Aug 2014, 17:15

That is the anti Airbus brigade theory!

Actually, I'd disagree with that. It was a very valid criticism of the FCU display, and one which had occurred before - but on those previous occasions the mistake had been caught in time. For another thing, I think the same issue was present on other mfrs types which had Honeywell hardware with a FPA/VS descent mode on the glareshield interface.

Of course there were other issues as well, but from a technical perspective this non-intuitive aspect of the FCU interface was probably the most glaring issue that needed to be fixed.

The crew respected "don't descend 4300 FT before 9 NM to STR due to possible false GPWS warning"

Except, of course, that (IIRC) due to a regulatory loophole regarding domestic commercial flights, the Air Inter A320s were not fitted with GPWS, and that was another *massive* issue that needed to be fixed.

Victims were less happy and let me add that they took a wrong way :
1. Creating an association of victims after a crash cannot be used in French Law to get damage (we don't have Class Action)

That appears to be changing now (article from Feb 2014):
France Finally Embraces Class Actions | Publications | McDermott Will & Emery (http://www.mwe.com/France-Finally-Embraces-Class-Actions-02-27-2014/)

[EDIT : It appears the court did nevertheless find AF and Airbus liable for damages in 2006 : http://www.aviationpros.com/news/10395428/french-court-says-airbus-air-france-liable-for-damages-in-1992-crash ]

2. Trying to manipulate the Court "to get more money" (from Airbus) was a very very bad idea.

Right, and I have to wonder whether the SNPL were involved in the decision to pursue Airbus but not AF (Capt. Asseline's legal representatives did the same in the Habsheim court case and in doing so missed an open goal - to this day I can't fathom why).

@Winnerhofer - re: contaminated blood, it didn't just happen in France, but also the UK and other places. In fact a good friend of mine (he's a fair bit older than me) sadly now carries Hep C as a result of a transfusion of contaminated blood in the early '80s.

Anyways - I think that's about as far off-topic as I dare to go! Back to the gallery...

Chris Scott

27th Aug 2014, 13:11

Actually, Winnerhofer, it seems that Air France's imposition of English on its pilots is only for communications with ATC at Roissy (CDG). I don't think there are any existing or proposed French auto-callouts in AF cockpits? On the other hand, the AF checklists, including their QRH, seem still to be in French, and (as the AF447 CVR demonstrated) AF cockpit business is conducted in the native tongue. Far be it from me to offer an opinion on that.

Those of us who have observed some of the actions of SNPL over the decades can understand that pilots forming the residue within AF of the ancien Air Inter, the former French domestic airline, might find it distasteful to adapt to ATC comms in English. It's almost unknown at any French airfield to hear a Francophone pilot speaking English for ATC comms, so this French anomaly is not specific to CDG. As the Liberation article indicates, Air France has been the only non-Anglophone flag-carrier outside the former Soviet bloc refusing to speak English for ATC comms at its home airfields.

roulishollandais

3rd Sep 2014, 02:35

Which is the Air France UAS procedure today?

Chris Scott

4th Sep 2014, 15:04

Hello Winnerhofer,

That would all benefit from a definitive translation into English.

Quote from Stéphane Gicquel, secrétaire général de la Fédération nationale des victimes d'attentats et d'accidents collectifs:
" « Les acteurs de l'aérien ont une culture de l'entre-soi. Les familles de victimes ont une vraie difficulté à croire à leur indépendance. » "

By "l'aérien", is he referring to ICAO, the regulators, the manufacturers, the airline industry, or individual airliner crews? Whichever it is. no doubt it could provoke some lively debate.

jcjeant

4th Sep 2014, 15:18

Some will like .. some not :)
Spoiler:
French
http://henrimarnetcornus.20minutes-blogs.fr/media/00/02/1875577343.pdf

roulishollandais

4th Sep 2014, 17:49

Airbus et l’EASA ont demandé aux pilotes d’assumer la responsabilité du défaut d’un équipement de l’A 330 en sous-estimant le risque.Underestimating the risk to deepstall that plane or the risk to be unable to recover from the stall ?

roulishollandais

4th Sep 2014, 17:53

By "l'aérien", is he referring to ICAO, the regulators, the manufacturers, the airline industry, or individual airliner crews? Whichever it is. no doubt it could provoke some lively debate.it includes passengers too. "Vaste programme !"* of course.
* Charles de Gaulle

jcjeant

4th Sep 2014, 18:21

RH
Underestimating the risk to deepstall that plane or the risk to be unable to recover from the stall ?I think the two are linked ....
en sous-estimant le risque.
But I think the risk that mean Cornu is:
Décembre 2002 : La FAA rend obligatoire le remplacement de la sonde Rosemount par les sondes
Goodrich et Thalès AA en précisant le risque de sortie du domaine de vol et qu’il s’agit de la réponse
à une « unsafe condition » December 2002: The FAA mandates replacement of the Rosemount probe probes
Goodrich and Thales AA specifying the risk to go out of the flight domain and that is the answer to
an "unsafe condition"

DozyWannabe

5th Sep 2014, 00:30

Underestimating the risk to deepstall that plane or the risk to be unable to recover from the stall ?

If you'll excuse the Google translation of the original HMC PDF:

Airbus and EASA have asked drivers to take responsibility for default equipment of the A 330 underestimating risk.

For one thing, the Thales AA probes were *not* a default fit, they were an option.

For another, there's an interesting sideline here:
Pitot probes : Air France - Corporate (http://corporate.airfrance.com/en/press/af-447-rio-paris-cdg/pitot-probes/)

Following the fluctuations and/or inconsistency in aerodynamic speed indications on the long-haul A330 and A340 reported by some airlines, the French Civil Aviation Authorities published an Airworthiness Directive (AD) per type of aircraft to impose the replacement of the ROSEMOUNT P/N 0851GR Pitot probes, either with the GOODRICH P/N 0851HL probes, or by the SEXTANT (THALES) P/N C16195-AA probes. This operation had to be completed by 31 December 2003.

Even more interesting:
The authorities did in fact attribute these incidents to the presence of ice crystals and/or quantities of water which exceeded the specifications of the P/N 0851GR Pitot probes, manufactured by ROSEMOUNT, which was acquired in 1993 by GOODRICH.

So in effect, the Goodrich probes which were recommended to replace the Thales probes after the accident were in fact a development of the original Rosemount probes which were also known to be problematic.

As for the overriding premise of M. Marnet-Cornus's PDF, it's the standard SNPL modus operandi of indignant tone, selective presentation of information and questionable attempts at misdirection.

(again please excuse Google Translate)

..the Autopilot and auto-thrust are disconnected, flight controls move ALTERNATE into law.

While control of the aircraft is reduced, the crew must keep them in a very small area of flight and face a incredible number of alarms in a very short time. This is an excessive workload.

The ALTERNATE, which was imposed their law and using which they had never been trained at high altitude, is a mode degraded flight controls. The A 330 was no longer consistent with all regulations of the basis of his certification, he was just "Controllable" very unstable in roll. The loss of information speed entailed the risk of asking its flight.

The second paragraph is incorrect - in fact the likelihood is that the aircraft would have stayed more-or-less on course had the crew done absolutely nothing with the controls in response to AP disconnect.

The third paragraph makes reference to a claim that Alternate Law makes the aircraft unstable in roll. Again, this is a fabrication. Alternate 2 makes roll control direct - so the aircraft will be *more sensitive* in roll, but that's a long way from unstable.

In the space of 49 seconds, to 02h10.54, the A 330 wins and can not be recovered

He uses this phraseology a few times - implying that the aircraft's systems forced themselves into an uncontrollable state from which the pilots could not recover. This completely ignores the fact that it was the PF's control inputs that put the aircraft into that situation in the first place.

For over four years, the media relentlessly dogs custody of the politico-industrial system cowardly tried to make dead drivers solely responsible for this tragedy

Yet another fabrication. The BEA report implicitly criticises the airline, the regulator, the manufacturer and the system as a whole.

All these points of very questionable veracity occur in the introduction, three pages in. I have a very difficult time believing that the rest of the document is much better in that regard.

Turbine D

5th Sep 2014, 00:56

Dozy, Re:Pitot tubes

Could you explain to all of us why Goodrich pitot tubes, that worked very well on Boeing jets, were not considered as replacements in 2007? Now quite honestly, Airbus and Air France are experienced entities, not new-bees to this business of aviation and how speeds are measured and which devises are best. I think there is much more to this story from a technical point of view that isn't being revealed.

The Air France time line is nothing more than a talking point or "positioning", that is to say, "Who me? I didn't do anything wrong." I am sure Airbus has a similar talking point. Lawyers like talking points.

DozyWannabe

5th Sep 2014, 01:34

Could you explain to all of us why Goodrich pitot tubes, that worked very well on Boeing jets, were not considered as replacements in 2007?
Goodrich tubes were standard fit on Airbus types (including the A330) as well, and to the best of my knowledge still are. The Thales pitot tubes were an optional fit.

This is illustrated in M.Marnet-Cornus's document:
In 2008 175 Airbus A 330 of 32 companies Airlines are equipped with the [faulty] sensor.

A quick archive.org dive indicates that there were:
555 [A330s built] as of 31 July 2008

which in turn indicates that around 70% of A330s built were not equipped with the Thales probes (meaning they were standard Goodrich fit) as of 2008. Some of the remaining 30% would have been equipped with a mix of Goodrich and Thales units as well.

[EDIT : HM-C's focus on the pitot tubes is itself a misdirection of sorts. The aircraft did not crash simply as a result of the UAS condition, in fact there's a good chance that had there been no manual input, the aircraft would have bumped its way through the turbulence with the speed indications returning in a minute or less. The aircraft crashed because the PF's instinctive response of pulling back on the stick without any discussion or notification was inappropriate, and became more and more inappropriate as the sequence unfolded.]

DozyWannabe

5th Sep 2014, 21:36

Read the original threads. The manufacturer/certification status of the AoA vanes state that the readings given from the vanes are not sufficiently reliable below 60kts IAS. Certainly not a situation where you want inappropriate hard protections making things more difficult.

roulishollandais

8th Sep 2014, 20:00

I was taught early in my career as a fighter pilotThat expert and former USAF's aerobatics pilot is perhaps not the best to claim that aerobatics was of no use to emphasize hand flying skills and basics helping the B757 crew to survive.

roulishollandais

8th Sep 2014, 20:53

Mr Marnet-Cornus seemed to write again in a story telling what has been investigated and discussed, medias included.
He wrote that as a tribute to pilots' honor. But their survival would have saved their honor too.

Mr Marnet-Cornus seems to have misread our PPRuNe's threads :
- He seems to have missed what our specialist gums wrote about the inertial HUD and our following exchanges.

Why did French regulators, Air France heads, and important pilots' Unions like SNPL, UNAC, SNPNAC, SPIT, and younger Unions reject or at least forget the Klopfstein's HUD design ? (I have my idea ).

- He seems to have accepted that ACARS claim to be private informations , perhaps with Copyright, not to be used by pilots, Justice, investigations*, and pilots' feedback in training.

Not a word about HUD or not ECAM reported ACARS which should have saved the flight.

That report points other interesting questions but I feel it as partial. Dozy points the SNPL behind that report, but nothing shows that it is a SNPL report, it is Mr Marnet-Cornus' report.

*But the BEA did not follow that denegation :ok:

VNAV PATH

9th Sep 2014, 07:19

The 2 pilots of Habsheim accident were 2 instructors and freshly brewed in Toulouse A320 process. Concierge's job..

tdracer

9th Sep 2014, 16:34

Apologies if this is re-hashing old ground, but AF447 was the subject of "Air Disasters" last week on the Smithsonian channel (I think it's what's known as "Mayday" elsewhere).

At any rate, assuming the presentation was reasonably accurate, it noted that the head pilot, after returning to the cockpit, was initially confused about what was going on, but eventually figured it out and started telling the Pilot Flight to push the nose down. Which raises the question: Did the PF ignore him, was it already too late (not enough altitude to recover), or was the aircraft in such a deep stall that it couldn't recover (or at least not recover without exceptional pilot skill)?:confused:

Not trying to start another argument, just trying to understand.:)

Lonewolf_50

9th Sep 2014, 17:00

At any rate, assuming the presentation was reasonably accurate, it noted that the head pilot, Captain Dubois
after returning to the cockpit, was initially confused about what was going on, but eventually figured it out and started telling the Pilot Flight to push the nose down.
Before the Captain returned to the flight deck, the First Officer, Mr Robert, had on a couple of occasions (per the CVR transcript available), advised the flying pilot, Mr Bonin, to "go down" (which I think he meant as lower the nose). The transcript and FDR suggest that Bonin never held the nose down long enough to make enough difference in terms of getting the AoA away from "stall" and toward "flying." Did the PF ignore him, The PF was task saturated and well behind the aircraft by the time Captain Dubois arrived. That much seems clear based on the report's findings. Some of the Captain's initial directions to Mr Bonin included help in keeping the wings level ... see the CVR transcript. This makes some sense if what he thought (initially) was the problem was the pilot needing to recover from an unusual attitude ... leveling the wings is usually the first step in that recovery to level flight ...
was it already too late (not enough altitude to recover), or was the aircraft in such a deep stall that it couldn't recover (or at least not recover without exceptional pilot skill)?:confused: Hard to say, but probably true. That point in particular received pages and pages of treatment in the various AF 447 threads. You'll want to read the ones after the CVR transcript came out.

Given the AoA and the time Captain Dubois had to assess, and then try to correct, the flight condition when he arrived, every second that he was not actively directin recovery from stall inputs was a second lost to the eventual point of "no return." (At what point did he realize that the plane was stalled? Hard to say).
As CRM seems to have broken down between the other two pilots before he arrived, he was greeted with a confused situation that he tried to untangle ... and I'd say he ran out of time.

The poignant "it is not possible" (in Fr, impossible) remark points to him not believing something ... such as

This plane is stalled
and / or
These two geniuses were given a plane in straight and level flight and we are not flying straing level
and/ or
We are falling at what rate of descent????

There are any number of things that, having arrived on the flight deck and tried to sort out the problem, were so far beyond what he expected to find that it took some time/effort to overcome.

Time he didn't have.

DozyWannabe

9th Sep 2014, 18:30

The 2 pilots of Habsheim accident were 2 instructors and freshly brewed in Toulouse A320 process. Concierge's job..

Capt. Asseline was indeed senior Captain on AF's A320 programme - not so sure about the FO. But it had nothing to do with "concierges" or training - it had to do with poor planning, oversight and execution. In particular, AF never tried to verify Asseline's proposed flight plan with Airbus - if they had, then it would have become clear that Asseline was proposing a level of risk that Airbus's own test pilots would never allow (they would demonstrate Alpha tech at higher altitudes or perform low-level flypasts, but they would never combine the two, least of all with pax on board). In fact it's unknown if Asseline ever told AF ops exactly what he had in mind.

was it already too late (not enough altitude to recover)

By the time the Captain worked out what had happened (when the PF stated words to the effect of 'I've been at max nose-up for some time now'), it was far too late.

@Winnerhofer - your last post is very dubious. For one thing, Bonin's wife was on the flight, but would have been back in the cabin. The idea of playing "musical chairs" with the flight crew is ludicrous. Bonin had been in the RHS since pushback - no need whatsoever to move his seat. I don't know where to start really...

VNAV PATH

9th Sep 2014, 19:17

@ Dozzy wanabee

CPT Asseline's first officer was in fact a Captain involved in the same AF A320 program. They were good friends what can explain some compliancy.. But they were overconfident in airplane behaviour. Good schoolboys..

DozyWannabe

9th Sep 2014, 19:30

Disagree. This isn't the place for discussing that - there was a thread which covered it a while ago here:
http://www.pprune.org/tech-log/528034-habsheim.html

I don't think it was overconfidence in the aircraft. Once in the air, it was a series of repeated failures (particularly on Asseline's part) to properly assess the risks involved in what they were doing.

Lonewolf_50

9th Sep 2014, 21:26

Winnerhofer, I have no inside information.

My reply to tdracer was grounded in the information released to the public. If there is an omission there, I don't know how various bits of information will come to light. As you point out, court cases take a while and sometimes new info comes to light as part of the process.

In the case of AF 447, all I can say is "we shall see."

PJ2

10th Sep 2014, 05:33

Hi Lonewolf_50;

There are some (who know their aerodynamics) who have done some work on this and who believe that the airplane could be recovered, even from quite low altitudes, (thicker air). The horizontal stabilizer was never fully stalled and the airplane demonstrated the ability to pitch down when ND stick was held those rare times in the ND position. As you know I've done this scenario in the simulator a number of times and the result is always the same; the elevator is effective, the stab returns from +13deg to a more normal -2deg or so which all helps to unload the wing, reduces the AoA and unstalls the wing. I wouldn't say they were done by 15000ft but perhaps by 8000ft or so, (which I have not tried; high up, it takes about 15000ft on avg to unstall the airplane - lower down, less).

I think they could have made it with room to spare. Also, I disagree with the use of the term "deep stall", which only applies to the BAC-111 and DC9 configurations. Discussed at length and not about to be re-argued here.

alexd10

10th Sep 2014, 14:10

And more than that.. If the stab would have been manual trimmed to full ND position? how long would take to pitch the AC to a recovery AOA? Can be estimated, or at least guessed?

With the stabilizer not stalled, I suppose with a decent CFD soft it would be possible to estimate the pitching couple and pitching time too.. but again, demonstrating the fact that aircraft was recoverable until late in the event don't fit some very sensitive interests.

Lonewolf_50

10th Sep 2014, 15:02

PJ2, glad to "see" you again.

No argument with your points.

My response was more about the space between the ears of the three pilots involved, and an attempt to address why (since they were in the "surprise mode" for quite a while during this upset) there was a point at which the Captain could have begun to issue the directions, or take the controls himself, to return the aircraft to "flying" versus "falling," overcome the inertia downwards with flight control inputs that would have arrested the descent (after overcoming the stall) before the aircraft reached the surface of the ocean. I recall the assessments and estimates you mention.

Since he did not have the benefit of knowing the scenario he was in as he entered the cockpit, the "run out of time" issue boils down to establishing inside his decision making ability "what is the aircraft's flight condition?" and then deciding "what shall I do/direct to get it back into control?"

I raised the point on how he was coaching Bonin into controlling his roll control since I think his initial assessment was that he was dealing with an unusual attitude recovery problem ... which he sort of was, but I have no idea if he ever arrived at the conclusion that "this sumbitch is stalled!" I'd guess he did at some point, but where was that? No idea.

Any decision point reached before the estimaged point of no return (your point on the denser air as they fell is well made) would need to be acted on by a correct application of anti-stall control inputs and a stall recovery.

I recall the long discussions on the point of no return, which you so kindly summarized concisely. My intent in the reply hopefully does not contradict anything from that discussion. If it does so, then my error is acknowledged, and maybe ti addressed tdrader's question clumsily. I suppose my "time he didn't have" contradicts the idea that if he'd reached the decision sooner that "it is stalled" and "Do This To Unstall The Aircraft!" were his instructions, then there should have been enough time to recover before impact with the surface.

PJ2

10th Sep 2014, 18:39

Lonewolf_50;

I read but rarely enter something these days LW. The episode of the disappearing posts on R&N set my sails. However, I'm aware that others may be reading the AF447 work anew and I think a reminder of the significant effort put in by many, many contributors on "The AF447 Threads" is worth it as it is a veritable gold-mine of thinking and research if read with care, knowledge and an open mind.

I should point out for those looking to examine these threads the importance of the first post in almost all of the them. JT has done us all a favour and very kindly constructed and provided links to all the threads - I'll pick page one of the eleventh thread (http://www.pprune.org/tech-log/511119-af-447-thread-no-11-a.html) just as an example. One can search the entire PPRuNe archive using the simple search tools which JT and another prescient poster have provided us.

Insofar as "corrections" go, well, it's always a discussion and never a persuasion as far as I'm concerned because we weren't there an have only a tiny flashlight in a very large room that has tons of evidence upon which we simply cannot shine our little light. We can only make sense of the minute traces of the accident and take thinking as far as reasonableness and logic permit.

As far as court cases go and blame etc., I have absolutely no interest in it whatsoever. Though it understandably exists, that arena is a very foreign and hostile world to a flight safety person.

john_tullamarine

10th Sep 2014, 22:46

How can we work with imperfection?

The reality, I suggest, is that life is imperfect and we have to manage that regardless of the specifics of the day. Doesn't matter whether the topic be aeroplanes, cars, the state of one's health, wealth, or the attractive lass at the other end of the office, etc ..

Mitigation against the potentially adverse sequelae of such a reality involves a healthy and conservative mix of resources, knowledge, training, and practice, with a fortuitous portion of Lady Luck's smile.

DozyWannabe

11th Sep 2014, 01:10

I wouldn't say they were done by 15000ft but perhaps by 8000ft or so, (which I have not tried; high up, it takes about 15000ft on avg to unstall the airplane - lower down, less).
As it happens, around 9000ft is pretty much where they were at when the PF makes his "maxi nose up for some time" comment, and based on the assumption that the Captain did indeed figure out what had been happening from that (subtracting around 1000ft for the thought process to click), I think it's fair to say at that point it was too late.

Also, I disagree with the use of the term "deep stall", which only applies to the BAC-111 and DC9 configurations. Discussed at length and not about to be re-argued here.
Agreed (fervently).

Bernard Ziegler designed the Airbus to be pilot-proof.
No he did not. He headed the technical team that came up with the design, but the specification he had a hand in laying out was not due to his perception of piloting competence (or lack thereof), it was simply that the technology had reached a point where the control systems could be designed around the safe limits of the airframe. The intent was for the technology to *aid* pilots, not to patronise them (though BZ himself did have a couple of "foot-in-mouth" moments there).

_Phoenix_

11th Sep 2014, 03:03

@Lonewolf 50

The poignant "it is not possible" (in Fr, impossible) remark points to him not believing something ...

Capt. Dubois didn't comprehend why the aircraft was climbing while sidestick was forward

see below the situation dissected:

CVR // sidestick position // pitch attitude

2 h 12 min 32,4 Dubois: No you climb there // aft -8 deg // up 5 deg
2 h 12 min 32,8 Bonin: I’m climbing okay so we’re going down // full forward 16 deg //up 6 deg
2 h 12 min 34,2 Robert: You’re climbing // full forward 16 deg // up 8 deg
2 h 12 min 39,2 Bonin: okay we’re in TOGA // forward 10 deg // 0 deg
2 h 12 min 41,9 Bonin: On alti what do we have here? // forward 10 deg // down 3 deg
2 h 12 min 43,8 Dubois: (…) it’s impossible // aft -4 deg // down 3 deg

the reconstruction scene:
https://www.youtube.com/watch?v=XAom93qwoN0&feature=player_detailpage#t=2475

It looks like a dejavu of XL888 crash.
.. flight control system automatically “trims” the THS to max nose up. The moment arm overcomes the elevator authority. The increased speed only serves to force the aircraft in a perpetum stall, like a dart plane, see pitch attitude graph

Gretchenfrage

11th Sep 2014, 06:18

Incompetence Meets Hubris
Incompetence
These accidents all have something in common: pilot incompetence
The pilots in these cases may have known their airplane fairly well.
They may have memorized their company’s operating manual and their SOPs.
But in all cases they did not understand some of the basics of flying an airplane.
There is always a Captain who has a real license to supervise the others with lesser licenses.
But on closer inspection what we are really saying is that a pilot who cannot legally take a friend for a ride can occupy a cockpit seat while the captain is back in First Class.
Automation and Hubris
Bernard Ziegler designed the Airbus to be pilot-proof.
He is a good pilot, and he noticed that many pilots are less skilled than himself.
In the interest of safety, he designed an airplane that could not be stalled.
But it has been known for thousands of years that hubris is followed by nemesis, that pride goeth before destruction, and an haughty spirit before a fall. (Proverbs, 16:18)
Hubris is arrogance before the gods.
The goddess Nemesis alone can see the fine line between doing the best work you can and believing that your work is somehow superior.
Cross the line and she is ruthless, finding your fatal flaw and using it to bring you down.
AF 447 was the fall of the hero.
Pilot carelessness led the airplane into a line of thunderstorms.
Supercooled water drops overwhelmed the pitot heaters, temporarily removing all three sources of airspeed information.
The autopilot dropped off.
The flight control computers switched from Normal to Alternate Law.
The airplane can be stalled in Alternate Law.
Human or robot, there is always a fatal flaw.
How can we work with imperfection?
AF447 | formercaptain.ca

Winnerhofer

I just love your contribution!

This should be glued up the wall of every manufacturer, administration and airline management.

Simply to daily remind them of these their deadliest sins.

And i consent with PJ2. It's quite revealing to see which contributions are "deleted" and which ones are deemed worthy to "remain".

Basil

11th Sep 2014, 10:34

Why try to take advantage of (possible) alpha floor protection at any time instead of just flying the aeroplane?
Or am I, with no real Bus experience, missing something?

Bpalmer

11th Sep 2014, 18:09

Captain left the flight deck in the hands of such low-time copilots.
Well, that's not how I would put it. Robert had more Airbus time than the other two combined.
Robert is the one that turned up the gain on the radar as soon as he got up there. The captain was negligent of monitoring the ITCZ activity and adjusting the radar gain and tilt accordingly.

Coffin Corner: The A330 with it's supercritical wing makes "coffin corner" a slightly different animal than it was in the 727 days—where Mach buffet was a real threat. The A330 on the other hand is performance limited more than MMO limited. But, pilots have not been educated on the characteristics of the modern airfoils where going critical Mach is much more difficult. Perhaps that's why Bonin apparently confused the shaking of the airplane due to low-speed stall with "some crazy speed" (Mach buffet perhaps). This is covered in more detail in my book "Understanding Air France 447."

The other very enlightening thing about doing stalls at FL300 is how completely useless the engines are in the recovery. Let's face it we're pretty close to full power at cruise anyway. Throwing the thrust levers up to TOGA doesn't do a damn thing (try it at cruise some time and watch the airspeed not move). That's something that practicing stalls in a jet plane at 5000 feet has definitely given many pilots the wrong impression of—since the engines work to fix the stall problem very nicely at that altitude. Hence 447's cry: "but we've got the engines, what's happening?" Some claim that the pilots didn't do any stall recovery. I contend they did exactly the stall recovery they were taught in their A320 checkout (the only time they practiced them) : apply TOGA power and minimize loss of altitude. Stalls are a pitch problem, not a power problem.

PJ2

11th Sep 2014, 18:17

"Stalls are a pitch problem, not a power problem."

Stalls are an AoA problem, period.

DozyWannabe

11th Sep 2014, 19:49

.. flight control system automatically “trims” the THS to max nose up.
Only in response to sustained human input.

The moment arm overcomes the elevator authority. The increased speed only serves to force the aircraft in a perpetum stall, like a dart plane, see pitch attitude graph
No it doesn't. If you find the relevant sections of the original threads, it's made clear that the elevators *always* have sufficient authority to override the THS, with the caveat that they must be held in that position to do so. The reason the Pitch Attitude graph looks like it does is because there was no sustained effort on the part of either PF or PNF to get the elevators to nose down and hold them there.

I contend they did exactly the stall recovery they were taught in their A320 checkout (the only time they practiced them) : apply TOGA power and minimize loss of altitude.
As you know, I've read your book and was very impressed by it - but I have to query your point here - at no point did any of the crew (including the Captain) acknowledge that they were stalled, verbally or otherwise. I contend that there was no procedure being followed - just almost random improvisation.

Lonewolf_50

11th Sep 2014, 20:17

"Stalls are a pitch problem, not a power problem."

Stalls are an AoA problem, period.
Amen, deacon.

Mr Phoenix[/QUOTE] Capt. Dubois didn't comprehend why the aircraft was climbing while sidestick was forward
[quote]2 h 12 min 32,4 Dubois: No you climb there // aft -8 deg // up 5 deg
2 h 12 min 32,8 Bonin: I’m climbing okay so we’re going down // full forward 16 deg //up 6 deg
2 h 12 min 34,2 Robert: You’re climbing // full forward 16 deg // up 8 deg
2 h 12 min 39,2 Bonin: okay we’re in TOGA // forward 10 deg // 0 deg
2 h 12 min 41,9 Bonin: On alti what do we have here? // forward 10 deg // down 3 deg
2 h 12 min 43,8 Dubois: (…) it’s impossible // aft -4 deg // down 3 deg

I may recall incorrectly, but weren't they already falling by the time Captain Dubois arrived in the cockpit? I guess I need to go back to the thread and find the altitude traces to match up with your time data points there ... will get back to you when I have found it. What I recall is that around 1410 the event began, and they fell from FL 33-35 to the surface in the following four minutes. So, at 1412 the falling toward the sea surface, in a stall, was already in progress. However, I may have recalled that incorrectly.

Thanks.

DozyWannabe

11th Sep 2014, 20:58

I may recall incorrectly, but weren't they already falling by the time Captain Dubois arrived in the cockpit?

You recall correctly. They were established in the stall and the descent by the time the CDB arrived.

I suspect _Phoenix_ may have meant to refer to the aircraft being in a nose-up attitude rather than "climbing". In fact at the point Dubois arrived, the status included these aspects:

Bonin had his sidestick at full NU and would continue to hold it there for a further half a minute approximately
The nose was up at the point he arrived, but was about to swing back down
Unfortunately, he arrived almost at the exact point that the excessive AoA began fouling the pitot tubes and messing up the ASI readings again

[EDIT : In fact, two seconds after Dubois arrives, the VSI is indicating a descent rate of -10,000ft/min. 13 seconds later, Bonin reports that he's lost VS indication - I have to wonder if it's possible that it was indicating, but that his mental model didn't expect to see the "needle" pegged near the bottom of the scale.]

Schramm: Yes, the same well, uh, approach to stall means that one must never go beyond that. They were in unchartered waters.
Anchor: Well, yes because even Airbus had never envisaged such a situation.
Schramm: Indeed, in Normal Law, Airbus is unstallable.

The "anchor" is being disingenuous here. The Airbus FCOMs have always stated that it is possible to stall the aircraft outside of Normal Law. The AF representative (M. Schramm?) is, I suspect, doing a little bit of dancing around the point. In fact, "approach to stall" training and "stall recovery" are very different beasts - one thing that became apparent in the wake of this accident was that over the previous decade or so, the airlines as a whole started dropping stall recovery training from their recurrent syllabus in favour of stall avoidance ("approach to stall"). This was the case across almost every major airline and every type - from Airbus to Boeing and the remaining MDs.

This is why Boeing and Airbus subsequently collaborated on a programme to improve understanding of and training for stall scenarios.

FO Robert had the most time on type, this is true - and he certainly seems to have initially been the closest to understanding what was happening, but the likelihood is that the last time either he or Bonin actually trained for/performed a stall recovery in a powered aircraft was during their PPL training.

DozyWannabe

11th Sep 2014, 22:07

Ah, but there's no ELAC on the A330...
True, and that's an oversight - but they do get the nomenclature correct elsewhere as far as I can see. I'll give it a proper read when I have the time. A quick skim reveals a few conclusions that are at best conjecture - but this is TV after all...

Re: line check - that doesn't necessarily indicate anything on it's own. To illustrate the point, Capt. De Crespigny of the Qantas A380 which blew an engine out of Singapore (and subsequently commanded the flight to a safe landing and minimal injury) finishes his book on the subject by pointing out that he was on a checkride on that particular flight and was told by the Check Captain that he had in fact failed it (presumably because the book didn't take a situation of that gravity into account). Needless to say, he passed the next one with no problems!

infrequentflyer789

11th Sep 2014, 23:16

"Stalls are a pitch problem, not a power problem."

Stalls are an AoA problem, period.

For me, as a non-pilot who once knew a bit about making stuff that flies, the single most scary thing from all the 447 and other stall threads has been the number of apparent pilots who don't seem to know this. That and the realisation that many of them don't even have an AoA indicator in the cockpit (which is probably connected to previous point).

[Actually, for several years I might have said that stall is a boundary layer separation problem, it just happens at a particular AoA - but that depends on your point of view. By the time the pilot is in control, the engineering is done, and AoA is the difference between flying and falling. ]

DozyWannabe

11th Sep 2014, 23:48

That and the realisation that many of them don't even have an AoA indicator in the cockpit (which is probably connected to previous point).
To be honest, it's fairly reasonable to state that if the aircraft is descending while nose-up and/or apparently unstable in pitch and roll, the odds are good that it is in a stall.

Mr Optimistic

11th Sep 2014, 23:50

Agreed re boundary layer.

_Phoenix_

12th Sep 2014, 01:53

@ DozyWannabe and Lonewolf_50

I suspect _Phoenix_ may have meant to refer to the aircraft being in a nose-up attitude rather than "climbing".

No, WE know that aircraft was falling in a nose-up attitude.
But, Capt. Dubois mental picture was different: "why the aircraft was climbing while sidestick was forward"... HE was right!

2 h 12 min 32,4 Dubois: No you climb there // aft -8 deg // up 5 deg // No Stall Warning

Then, he might saw Bonin pushing sidestick at full forward stop, the stall warning goes off... "nice! as expected"

there was no sustained effort on the part of either PF or PNF to get the elevators to nose down and hold them there

Again, PF kept sidestick forward for 10 sec. He was a glider pilot too, imagine that input for a glider

..sometimes, I'm wondering why Airbus doesn't replace sidestick with a touch screen ADIRU, linked to a tablet, so the capt. can fly the plane from rest area or lavatory.

flyingchanges

12th Sep 2014, 02:11

Then, he might saw Bonin pushing sidestick at full forward stop, the stall warning goes off... "nice! as expected"

And the airplane gave he impression that the correct input was wrong.

I still contend that if they had not touched the airplane at all, there would have been a better outcome.

_Phoenix_

12th Sep 2014, 02:31

@flyingchanges

Actually, DozyWannabe is right, they should nailed the sidestick at forward stop, since for 10sec forward input the elevator rotated from -30deg NU to -15deg NU, THS remained at -15deg NU - constant as gravity

What is the ideal crew for AF447? A pilot and a dog.
The pilot is there to feed the dog, and the dog is there to bite the pilot in case he tries to touch anything

tdracer

12th Sep 2014, 04:09

There are some (who know their aerodynamics) who have done some work on this and who believe that the airplane could be recovered, even from quite low altitudes, (thicker air). The horizontal stabilizer was never fully stalled and the airplane demonstrated the ability to pitch down when ND stick was held those rare times in the ND position.

Thanks, that's the bit of information I was looking for. Clarification, when I say "deep stall", I don't mean to imply an unrecoverable stall condition (such as what happed on the BAC 111), but rather a sustained stall condition where the aircraft is going down as fast (or faster) than it's going forward. As I understand it, on the BAC 111 (and some other 'T' tail aircraft), if you get into a deep stall, the spoiled airflow from the wing effectively blanked the tail and literally prevented pushing the nose back down to recover.

I've been on flight tests where they did dozens of stalls and "wind up turns". I suspect most of you are unfamiliar with that term so I'll explain - for a 'normal stall test, the pilot holds a constant altitude while allowing airspeed to decay (throttles at or near idle) until the aircraft stalls. These stalls are pretty gentle (at least on the 767 I was on) - a little forward push and the airplane falls out of the stall and recovers (although if the pilot had to use a big rudder input to keep it straight and level, the rudder would tend to shear the line to the 'trailing cone' Pamb sensor - in which case we got to go home early :E). A 'windup' turn was used to test the inlet at very high angles of attack with the engines at takeoff power - hold constant altitude but pull the turn ever tighter until the aircraft stalled and fell out of the turn. It's basically a way to test the engine inlet reaction to a takeoff over rotation, without doing it a few feet off the ground. Three hours of windup turns is also the only time I've gotten airsick (straight stalls didn't bother me, but windup turns did).

Anyway, back to the point - in a deep stall, with forward airspeeds well below that necessary for level flight - even if the horizontal stab isn't stalled, it's not going to provide much authority - simply not enough Q. It's going to take a long time to push the nose down enough start gaining sufficient airspeed to regain control.
Back during the original 777 flight test, an unnamed FAA pilot somehow managed to get into a deep stall at high altitude - even though the right seat Boeing pilot immediately took control, they reportedly lost over 10,000 feet before he was able to regain control (second hand info - I wasn't on that flight and I'm pretty sure I'm glad I wasn't :eek:).

Again, PF kept sidestick forward for 10 sec. He was a glider pilot too, imagine that input for a glider
Yea, while 10 seconds probably felt like an eternity, in a deep stall I suspect at least 30 seconds just to get the nose down and get airspeed up to a reasonable, controllable level.

For me, as a non-pilot who once knew a bit about making stuff that flies, the single most scary thing from all the 447 and other stall threads has been the number of apparent pilots who don't seem to know this.
Ditto! And the idea that one can power out of stall by advancing the throttles has a potentially fatal flaw - engine inlets are not designed to provide high flow at those angles of attack. Inlet separation and engine stall is a distinct possibility (reportedly one engine stalled/surged during the Birgenair 757 plunge)

PJ2

12th Sep 2014, 13:51

tdracer;

Thanks for the elaboration on your meaning of "deep stall". Davies discusses the phenomena in Handling the Big Jets - what you've described re the BAC111 is how he describes it.

AF447 was descending a bit faster than it was going forward, (12,000fpm, or about 130kts) but although reduced, the elevator still had authority.

While the stick was never held in the ND position for "10 seconds", there is a time period when it oscillates between NU & ND but is mainly ND for about 10" at around 17000ft in the descent and the pitch can be seen to change to ND and the AoA moves from about 43deg to about 38deg; then the stick returns to NU, essentially until impact.

While the sim does not have stall algorithms from actual flights past the approach-to-stall, going from memory the sim took even longer than the time you've estimated, (~45" from beginning of recovery at around FL350 to recovery at FL220, approx). The stick was held in the full ND position for the entire time, the descent rate reached 20,000fpm. The flight path vector symbol moved up very slowly from it's initial indication of ~ -35deg; the THS began moving from its full -13deg (NU, tail down) position a few seconds after the stick was held in ND position - the pitch was controllable throughout, though very slow to respond.

http://www.smugmug.com/photos/i-Vc8RFmq/0/L/i-Vc8RFmq-L.jpg
Recovery is 30" away...

As you know, air is thicker lower down and control may have been better and recovery time a bit shorter. I'll bet they're worked on algorithms to simulate an actual full stall. Still, the sim was pretty good - shake, rattle and hum, with the "Stall, stall, stall" message blaring - oddly it never quit until the sim became unstalled so not sure the NCD loss of the warning situation is reproduced.

One last item - I had thought the same thing regarding the engines - that high incidence angles would upset the engine to the point of putting the fire out but they both remained operational (and at high power) throughout, until impact.

DozyWannabe

12th Sep 2014, 15:11

But, Capt. Dubois mental picture was different: "why the aircraft was climbing while sidestick was forward"... HE was right!

2 h 12 min 32,4 Dubois: No you climb there // aft -8 deg // up 5 deg // No Stall Warning
Hmm - I think there's a degree of doubt regarding translation - as far as I can tell, the distinction between "climb/descend" or "pitch up/down" can be a bit tough to decipher from the French without context.

The altimeter would have been unwinding at a significant rate, so unless he made that statement before he'd checked the altimeter display, I feel a little dubious about making the call.

they should nailed the sidestick at forward stop
Once in the stall yes, but if flyingchanges means leaving the controls alone from initial AP disconnect, then I'm inclined to agree with him.

To be sure, I'm not saying that pilots shouldn't use the controls in such situations, but a pilot should be sure what they intend to accomplish and how they intend to accomplish it before acting, unless in a split-second "life-or-death" situation.

Bpalmer

12th Sep 2014, 18:55

"Stalls are a pitch problem, not a power problem."

Stalls are an AoA problem, period.

Well, yes, of course. I didn't mean to imply they were a pitch attitude problem, but a problem resolved with pitch control, not thrust levers.

Bpalmer

12th Sep 2014, 19:01

but the likelihood is that the last time either he or Bonin actually trained for/performed a stall recovery in a powered aircraft was during their PPL training.

The accident report details that they each had stall training in their A320 initial training-as it was the full course for that airplane. However, since their A330/A340 training was a differences course, it did not include stalls.

DozyWannabe

12th Sep 2014, 20:18

@Bpalmer:
The stall-related training on the A320 initial course seems to have been approach to stall only.

From Final Report Section 1.18.3.5.1:
Stall phenomena are covered during the initial A320 type rating, according to the same philosophy of the manufacturer and the operator. They are not reviewed during the long haul passage, in CCQ 330, or during recurrent training.
At the time of the accident, the immediate actions were: simultaneously reducing angle of attack and applying TOGA thrust from the first signs of the stall (Stall warning / buffet onset). A minimal loss of altitude was expected.

(Note that it went to stall onset/warning - not recovery from a developed stall condition.)

and 1.18.3.6:
Note: Stall and stall recovery exercises are undertaken during initial pilot training (in particular basic training, private pilot, professional pilot, etc.) but not during type rating training.

flyingchanges

12th Sep 2014, 20:40

Once in the stall yes, but if flyingchanges means leaving the controls alone from initial AP disconnect, then I'm inclined to agree with him.Yes, my exact thought. The machine was trimmed and stable the last time there was valid information presented. I would be inclined to not do a whole lot more than keep the wings level until we had moment to ponder the situation. At that moment, it is important to not kill yourself before you have time to figure things out. I know exactly where the trim and throttles (physically) need to be on my machine to stay out of trouble long enough to get things sorted out. In this situation, time is your friend, figuring out how to make your time last is the true solution.

Bpalmer

12th Sep 2014, 21:26

Stall phenomena are covered during the initial A320 type rating,
and

Note: Stall and stall recovery exercises are undertaken during initial pilot training (in particular basic training, private pilot, professional pilot, etc.) but not during type rating training.

Seems they can't both be true.

Further more the tendency of the flight control law to want to pitch down at stall would be dependent on if the low-speed stabilities were active while in Alternate law, which in AF447's case, they were not (due to lack of IAS). So unlike the α-prot (aoa command) in normal law, or when the low-speed stabiliites are active, the pitch command was still centered on g-load/pitch rate (aka C*). I believe you can see the elevator response (up) to the g load going below 1 as the stall progresses and the airplane starts to accelerate downward. This requires pilot action - stick forward - for as long as it takes to get the nose to drop and the aoa to reduce. I also found that in the simulator manual forward trim was required, as even a full down elevator was not enough to counteract a full up stab as airspeed started to recover.

DozyWannabe

12th Sep 2014, 22:34

@Bpalmer:
I think we're looking into technical specifics a little deeply there in terms of the discussion we were having on training.

Reading between the lines, I think when the report is saying "Stall phenomena are covered during the initial A320 type rating", the important word is "phenomena". This covers a few areas, but I think the clue is in the following context - namely "(Stall warning / buffet onset)"

The second paragraph in full said (sorry for repost, but we're across the default page boundary):
At the time of the accident, the immediate actions were: simultaneously reducing angle of attack and applying TOGA thrust from the first signs of the stall (Stall warning / buffet onset). A minimal loss of altitude was expected.
This is an "approach to stall" action set, not that of recovery from a fully-developed stall, and the key phrase is "A minimal loss of altitude was expected". This refers implicitly to the industry's change of focus from stall recovery to avoidance, because recovery from a developed stall necessitates an appreciable loss of altitude.

Regarding your technical point, while the low-speed stability soft protection was not available in Alternate 2B - I think the pitch attitude readout from the DFDR implies that the airframe has a natural tendency to pitch nose-down as the speed drops off regardless of any extra impetus from the flight control system. Of course, a manual nose-down input will enhance that tendency and make a recovery happen more swiftly, but the FCOM (http://www.bea.aero/docspa/2009/f-cp090601.en/pdf/annexe.11.en.pdf) at the time seemed to cover this with the instruction "PITCH ATTITUDE...REDUCE".

PJ2

13th Sep 2014, 00:16

Bpalmer;

Thanks - read it differently; agree with your meaning as explained.

Machinbird

13th Sep 2014, 00:37

Some have discussed keeping hands off completely in the period immediately after the A/P disconnected and the switch to Alt 2b law.

Just a personal opinion, but that course would have a high probability of ending in a spiral dive. There is nothing to level the wings in Alt 2b but the aircraft's natural roll stability (which isn't all that good in most jets).

Bonin had to get on the stick, but only in roll, and not with such a heavy hand. If he had used a gentle touch in roll and stayed off the pitch axis, AF447 would have ended in Paris.

The problem was that Bonin had never seen that control configuration in training and so used a highly inappropriate control technique which appears to have generated a roll PIO.

DozyWannabe

13th Sep 2014, 01:21

@Machinbird:
IIRC the turbulence conditions were described as "MODERATE". Would that really have been enough to roll the aircraft on its own? I may be being somewhat naive in this supposition, but I would think it requires significant sustained force on an object of that kind of mass...

ItsMeFromEarth

13th Sep 2014, 02:44

What was the delta ISA?

_Phoenix_

13th Sep 2014, 03:18

stick was never held in the ND position for "10 seconds", there is a time period when it oscillates between NU & ND but is mainly ND for about 10" at around 17000ft

The period of time in discussion is between 2:12:33 to 2:12:43, around FL210. The ND position was dominant, the average value clearly on positive side of the graph. However, the elevator did not followed the input, it never passed half way from maximum of NU. Therefore, "though very slow to respond" is not due to the "developed" stall condition. But because the Dnz feedback response tends to fall off, due to lift loss and the airframe has a natural tendency to pitch ND. Then elevator is more and more driven NU, due to integration of the increasingly large CL error. Yes, there the pilot was right, but the aircraft response went badly wrong.
The automatic stabilizer trim function will only start moving ND after the elevator (not the sidestick) has moved from a NU to a ND position. I think even 30 seconds are not enough to straighten the THS from "braking chute" position.

I think there's a degree of doubt regarding translation

I think translation is accurate since it's copy/paste from Appednix 1 of the final report and the french verb monter = climb, nothing else.
I agree, Dubois warned Bonin to go down.

Bonin had to get on the stick, but only in roll, and not with such a heavy hand

He had a heavy hand or maybe the sidestick is not suitable for the hand of pilot in distress
I look to the roll command input vs the phenomenal NU input, between 11:40 to 12:15, I'm wondering what if he kept the sidestick inadvertently at max NU while fought to bring the wings to horizontal...

Owain Glyndwr

13th Sep 2014, 06:47

Machinbird

Some have discussed keeping hands off completely in the period immediately after the A/P disconnected and the switch to Alt 2b law.

Just a personal opinion, but that course would have a high probability of ending in a spiral dive. There is nothing to level the wings in Alt 2b but the aircraft's natural roll stability (which isn't all that good in most jets).

We don't have to speculate - Fig 64 of the Final Report shows a simulation of the aircraft response to the turbulence present both with and without pilot input.

With no pilot inputs 40 secs after loss of airspeed the aircraft would have ended up at 280 kts CAS, FL350, 3 deg pitch, a shallow climb and bank wandering around 10 deg right, but not divergent.

Just a thought, but there is nothing to say that Alt2b loses the stability augmentation terms (roll and yaw damping) in the laws.

tdracer

13th Sep 2014, 06:49

One last item - I had thought the same thing regarding the engines - that high incidence angles would upset the engine to the point of putting the fire out but they both remained operational (and at high power) throughout, until impact.

IIRC, AF447 was GE CF6-80E engines - the CF6-80C2/80E engines are impressively tolerant to inlet distortion. Some other big fan engines would not have reacted so gracefully.

As an aside, we propulsion engineers tend to take a somewhat perverse satisfaction when the engines are operating fine right up until impact in an event such as this (the Aero Peru 757 also comes to mind :( ). OTOH, Lauda 767 was our worst nightmare :uhoh:.

roulishollandais

13th Sep 2014, 09:37

but I would think it requires significant sustained force on an object of that kind of mass... All the normal planes love to roll, loop, stall, dive, spin, recover, hammerhead, in one word to fly.They are born to fly! Planes, and pilots too! So long nobody limited their pleasure to fly by any limiting/protecting/castrating design, software or system, every well born plane is easy to roll to well born pilot ;)
Going from normal law to alternate 2B law without enough training is more hasardous as a difference appears in rates of pitch and roll stick movements. Regulators have to verify pilots are prepared to that event. It is not a stability issue, but adaptation from pilots when trained to such modification in flight.
Thank you Machinbird and Owain Glyndwr to point the right things both.:)
Regulators must do their work better, faster.

_Phoenix_

13th Sep 2014, 12:53

but I would think it requires significant sustained force on an object of that kind of mass...
For that kind of mass, the aircraft has that kind of wings and vertical stabilizer...
Remember, the aircraft encountered weather with ice pellets.
See Figure 26 of final report. In only 2 seconds, after AP disconnect, the roll angle passes from 0 to 8.4 while roll input remained neutral. Next 15 seconds, PF continuously corrected the right tendency of the aircraft, 4 times, 2 times at left limit stop.
In the same figure, it is interesting to notice the AOA for "STALL STALL". Robert might saw 270 kts and asked "What is that?"

A pathetic tv reconstruction uses "What is that?" to emphasize the "idiot" scenario

Air Crash Investigation S12E13 - Vanished (Air France Flight 447) - YouTube (http://www.youtube.com/watch?v=-N9SOtgL4Ys#t=2302)

DozyWannabe

13th Sep 2014, 12:57

Then elevator is more and more driven NU, due to integration of the increasingly large CL error. Yes, there the pilot was right, but the aircraft response went badly wrong.
Sorry - not buying that - I've never been entirely convinced that the DFDR elevator readout doesn't take the THS position into account. I see OG has popped along - Owain, what's your take on this?

What I can tell you is that in the sim, the THS began rolling forward less than a second after the stick was held ND.

Also, the movement in the time period you state is two "blips" forward - while it trends that way, it's nowhere near enough to make a significant difference.

maybe the sidestick is not suitable for the hand of pilot in distress
Plenty of successfully and safely landed emergencies on Airbus FBW types say otherwise.

I'm wondering what if he kept the sidestick inadvertently at max NU while fought to bring the wings to horizontal...
No - the spring centralisation is very noticeable - you know when you're breaking it out of the central position. A deflection of that magnitude can only be deliberate.

_Phoenix_

13th Sep 2014, 13:37

@DozyWannabe
Final report: "The trimmable horizontal stabilizer (THS) began a nose-up movement and moved from 3 to 13 degrees pitch-up in about 1 minute and
remained in the latter position until the end of the flight."
The sidestick was held ND, at least in couple of occasions

Notice the neutral position of sidestick. The NU stop is close to vertical
The NU travel is much shorter than max roll command. If you have a heavy hand pilot in distress, my supposition "holds water"

AIRBUS SIDE-STICK OPERATION - YouTube (http://www.youtube.com/watch?v=Y4uHmvFbe7A#t=44)

DozyWannabe

13th Sep 2014, 13:58

The sidestick was held ND, at least in couple of occasions
No it wasn't. Your "10 seconds" consisted of two short "blips" - one to the full ND position for about 1s, then back to neutral, followed by a "blip" to the half-forward position for less than a second, then back to neutral before resuming the NU input.

Notice the neutral position of sidestick. The NU stop is close to vertical
The NU travel is much shorter than max roll command.
Having used one in the simulator, I can tell you that's not true. You're using a video in which the lens is distorting the apparent Z-axis.

This was not accidental - this was a pilot making a "haché" of controlling the aircraft when he didn't (immediately) need to.

_Phoenix_

13th Sep 2014, 14:40

DozyWannabe,
You have to zoom a bit in the 15 s interval, see there 2 triangles, in about 10 seconds interval. The area of those triangles is the quantity of ND input.
For Robert, the quantity of ND input is half of area under the graph, due to "smart" dual input

edit: the prolong left roll command came after 2:11:31, not after AP disconnect.

DozyWannabe

13th Sep 2014, 15:14

You can't "zoom in" much past 200% without the resolution becoming too coarse to tell much. Also, your second "triangle" has too many "blobs" of red/brown (representing data points) to make a call of that ilk. Robert was making no inputs during "2:12:33 to 2:12:43, around FL210". This is not a software flight control problem, this is a "pilot freaking out and playing whack-a-mole with the controls" problem.

The only time of appreciable "DUAL INPUT" was around 2:13:44 to 2:14:02, with Robert trying to get the nose down and Bonin sporadically pulling up.

At about 02:14:18, Bonin slams the stick back one last time, and hits the sidestick priority switch, locking Robert out. He makes no mention of doing this.

Owain Glyndwr

13th Sep 2014, 16:38

Dozy

Sorry - not buying that - I've never been entirely convinced that the DFDR elevator readout doesn't take the THS position into account. I see OG has popped along - Owain, what's your take on this?

I don't think I have a "take" and I'm not sure I understand what you mean by the readout not taking the THS position into account. The elevator deflection as recorded in the DFDR is measured relative to the THS chordline, so the elevator position relative to the fuselage datum is measured elevator angle plus THS deflection.

Does that answer your point?

Or did you mean the bit about increasing NU elevator?

In which case, although I wouldn't use the same words I tend to agree with Phoenix.

Remember that in both normal and alternate laws S/S movement commands a delta g not an elevator position. Neutral S/S is in effect a command to "keep me at the 'g' I have at this moment", so if the stick remains neutral and the 'g' falls because there isn't enough lift to sustain the g at the time the stick was released the system will apply up elevator in an attempt to recover the selected level. If this persists then the THS, following the integral of elevator deflection, will also move NU.

Coming back down, so far as I can see, one would need to apply a negative g command greater than the actual (less than 1g) situation at that time before the system would actually apply ND elevator. That command would have to be held for some considerable time before the integrated elevator signal caused the THS to start moving ND.

jcjeant

13th Sep 2014, 17:04

Obviously when it will be talks of Airbus at the trial .. discussions on behavior and the system of THS going to be warm while it operated in accordance with the specifications
The STALL alarm system and behavior will certainly be put on the grill also

Machinbird

13th Sep 2014, 17:51

Just a thought, but there is nothing to say that Alt2b loses the stability augmentation terms (roll and yaw damping) in the laws.
They would have to be on a fixed (safe) gain basis if so.

That means that roll stability terms might be a bit twitchy in cruise at altitude and thus might have helped contribute to the initial roll PIO. (sheer speculation at this point)

Owain Glyndwr

13th Sep 2014, 18:00

They would have to be on a fixed (safe) gain basis if so.

That means that roll stability terms might be a bit twitchy in cruise at altitude and thus might have helped contribute to the initial roll PIO. (sheer speculation at this point)

Yeah, they might well be fixed gains, but we don't know if those particular terms are scheduled anyway do we?

OK, it might be a bit less damped in alternate but I suspect that the major change was from stick movement commanding roll rate to stick movement commanding roll acceleration.

Machinbird

13th Sep 2014, 20:45

OK, it might be a bit less damped in alternate but I suspect that the major change was from stick movement commanding roll rate to stick movement commanding roll acceleration. Absolutely right. That difference was apparently never appreciated on AF447's flight deck.

The roll oscillation finally stabilizing was likely related to the decrease in roll acceleration available as a result of the substantial deceleration that the aircraft had experienced as was approaching stall speed.

Owain Glyndwr

14th Sep 2014, 07:05

Machinbird

The roll oscillation finally stabilizing was likely related to the decrease in roll acceleration available as a result of the substantial deceleration that the aircraft had experienced as was approaching stall speed.

Not sure about that. I take your point that system gain (roll acceleration/stick movement) was reduced as the airspeed fell, but against that you have to set worsening dutch roll damping and increased roll/sideslip ratio as AOA increases. And I don't think one can exclude Bonin learning how to fly it!

AdamFrisch

14th Sep 2014, 13:27

There's a pretty good in depth article about the accident and the culture of pilots in France in the latest issue of Vanity Fair. It's an interesting read.

roulishollandais

14th Sep 2014, 17:49

Vanity Fair ! Not very professionnal reference for p..p..RuNers ! You may find better anywhere :(

_Phoenix_

14th Sep 2014, 18:14

Owain Glyndwr,

system gain (roll acceleration/stick movement) was reduced as the airspeed fell, but against that you have to set worsening dutch roll damping and increased roll/sideslip ratio as AOA increases. And I don't think one can exclude Bonin learning how to fly it!

The slide slip and the yaw rates were not oscillatory before 2:12:15, the spiral mode is predominant. After that yes, at large AOA, the natural Dutch roll damping is slow, it requires a lot of effort and remarkable skills, to preserve stability...like standing on a ball, while holding a large umbrella in gusty wind. As far as I know, all the crashes due to control lost (stall) ended up in a spiral dive. Bonin kept the wing leveled, 38000ft stalled descent, i.e. see between 2:11:30 to 2:12:15, he used left rudder and sideslip, to increase the lift on right side of the wing, since roll command was insufficient, at left stop

RetiredF4

14th Sep 2014, 19:14

Phoenix
Bonin kept the "wing leveled, 38000ft stalled descent, i.e. see between 2:11:30 to 2:12:15, he used left rudder and sideslip, to increase the lift on right side of the wing, since roll command was insufficient, at left stop

That is a hell of a confidence in a guy, who stalled a perfectly flying aircraft without noticing.

Machaca

14th Sep 2014, 23:01

Graydon Carter, editor of Vanity Fair, describes the article (http://www.vanityfair.com/magazine/2014/10/graydon-carter-new-establishment):

...Vanity Fair International Correspondent William Langewiesche’s “The Human Factor,” on page 256—a gripping minute-by-minute account of the loss of Air France 447, which inexplicably plunged into the Atlantic in 2009, claiming the lives of all 228 people on board. You probably remember the event: a routine flight from Rio to Paris, no calls of distress, and then the plane just goes missing. It was a disappearance that bedeviled the aviation business. As Langewiesche discovered, the pilots had become confused and then frantic after a momentary loss of airspeed indications—this in a fully automated, state-of-the-art airplane that was otherwise flying straight and level, and experiencing no mechanical problems. What happened? The explanation has to do with the enormous advances in aircraft automation and what this has done to pilots. In this case, when the autopilot disengaged, the pilots essentially forgot how to fly.

_Phoenix_

14th Sep 2014, 23:47

RetiredF4,
That is a hell of a confidence in a guy, who stalled a perfectly flying aircraft without noticing.

How about hell of a confidence in that aircraft... capable of 2000ft/min climb with wing already stalled, see figure 28 of final report, 2:10:50 to 2:11:10.

roulishollandais

15th Sep 2014, 07:50

fully automatedAre we in Tech log?

Owain Glyndwr

15th Sep 2014, 08:39

Phoenix

We'll have to disagree on this.

As you say, there is very little sideslip in the early stages, but plenty of roll,which must say something about the roll/sidelip ratio
The yaw damper trace shows there was oscillatory behaviour in the yaw rate at a period that roughly equates to what one might expect for the dutch roll of an aircraft of this size in cruise, which contradicts your view that the motion was nonoscillatory
the spiral mode is never oscillatory so I don't see how you can claim that it dominated the initial rolling cyclic motion.

But anyway, all I was trying to say was that dutch roll characteristics get wprse as AOA increases, and your later remarks seem to confirm that.

infrequentflyer789

15th Sep 2014, 10:57

Obviously when it will be talks of Airbus at the trial .. discussions on behavior and the system of THS going to be warm while it operated in accordance with the specifications

Autotrim is a damned if you do damned if you don't scenario. If its purpose and accepted effect is to reduce pilot workload in normal flight, then the corollary is that turning it off will _increase_ pilot workload, and turning it off in an emergency when pilot workload is already increased will...

I hope any discussion will include whether or not turning autotrim _off_ was the "right" thing for the system to do in incidents such as G-THOF and D-AXLA, where the pilots needed to trim manually and did not do so. It should also include other similar incidents e.g.:

"At the moment the stall warning activated, the horizontal stabilizer trim began to increase its pitch-up action in a progressive manner until it reached the pitch-up stop."

- Name that type, and why did the trim go up ?

It would be good if the discussion included not just the specifications but the certification requirements and the behaviour of other a/c / mfrs in similar circumstances if the requirements are not tightly specified.

The Boeing engineers' comments (sorry, I lost the link) regarding FLCH "trap" are also worth raising. They knew about and discussed the exact scenario that later planted a 777 onto a sea wall, and whether or not the system should override the pilot commands. In the end, Boeing chose to go with their central philosophy - that the pilots must have the final say in controlling the plane. After decades of censure from some quarters over its implementation of hard-protections overriding the pilot, would it not be ironic if AB were to be censured in this case for a mode where (due to bad sensor inputs) its plane acted like a Boeing and slavishly followed the pilot commands all the way to the sea / sea-wall ?

The STALL alarm system and behavior will certainly be put on the grill also

I would hope we find out what the "modifications" to the SWC that were alleged earlier in this thread actually are, and how they impact the false alarm probability and the possible impact on _perceived_ false alarm probability. Is it not a reasonable conjecture that 447 crew believed the stall alarm was false, especially given the BEA note that other UAS crews ignored stall alarms believing them to be false (no one seems to have investigated _why_ they thought they were false - or at least it isn't in the report) ?

If the certification standards don't specify SW behaviour in absence of AoA data, I would want to know what other types/mfrs do - alarm or not - before attaching blame to AB. Similarly on the air data unit (not, I think, by AB, and maybe used elsewhere) that invalidated AoA data in the first place. Also relevant is why AF rejected the BUSS option - on cost ?

DozyWannabe

15th Sep 2014, 16:14

See Figure 26 of final report. In only 2 seconds, after AP disconnect, the roll angle passes from 0 to 8.4 while roll input remained neutral. Next 15 seconds, PF continuously corrected the right tendency of the aircraft, 4 times, 2 times at left limit stop.
Right, so let's look at that a bit more dispassionately.

AP disconnects at the red line, seemingly while damping a left bank, presumably from turbulence. At AP disconnect, for 0.2 seconds the pink "Roll Attitude" graph indicates a cessation of roll (stable at about 0.8 deg right), followed by a subsequent right roll, beginning at 02:10:06.4 over 0.8 seconds, peaking at 8.4deg right around 02:10:07.2. This seems consistent with a "bump" caused by turbulence - we don't know how the situation would have progressed, because it is countered with a left-bank control input.

The fact that the input graph is rendered at 180 degrees (i.e. opposite) to the attitude graph makes it a bit of a 'mare to read, but Bonin moves the stick to the half-left position to correct what is presumably a turbulence-induced roll. This seems to be an instance of overcontrolling, because the aircraft then rolls 6 degrees to the left.

The controls over the next 15 seconds or so look very much like a case of opposite PIO in roll until he begins to get a feel for what he's doing, and the roll oscillations begin to stabilise.

For this reason I'm not sold on the "incipient spiral" theory. Furthermore, Bonin's completely unnecessary pitch inputs begin simultaneously with the roll inputs at 02:10:07. No reasonable time is given to assessing what is actually happening in terms of the aircraft's status, and from that point on, every input he makes is reactionary.

After decades of censure from some quarters over its implementation of hard-protections overriding the pilot
The hard protections don't "override" the pilot, they give the pilot what he or she is asking for up to the safe limit of the airframe, and will continually monitor the aircraft's status to comply with the commands given to the best of the aircraft's ability.

noske

17th Sep 2014, 08:20

There's a pretty good in depth article about the accident and the culture of pilots in France in the latest issue of Vanity Fair. It's an interesting read.
I don't think the article was available online when you wrote this, but anyway, now it is:
Should Airplanes Be Flying Themselves? | Vanity Fair (http://www.vanityfair.com/business/2014/10/air-france-flight-447-crash)

jcjeant

17th Sep 2014, 11:52

The "Hard protection" are they even a paradox
Let me explain ...
When the aircraft is in "normal" law (that is to say when things are going well .. the aircraft is in an optimum state and in the specification) all "hard protection" are active (recall that they are there for the pilot to fly the plane in an optimum way .. that is to say keep it from flying out of the flight domain)
When it starts to go wrong .. often the normal law no longer applies and some "hard protection" are no longer in force
Ironically it is during the time when there are problems and when the pilot would be glad to have all help he can (and the "hard protection" are flying aid) that "hard protection" choice to partially stop .. or totally

Vanity Fair:
Nice article .. nothing new ..

roulishollandais

17th Sep 2014, 15:13

Noske, AdamFrisch, Winnerhofer
Vanity Fair is not such a document we may quote as a reference document, in thesis, investigation, safety agencies reports, uni-works, serious exchanges.

Are the facts told there true or not, I have no possibility of verification. But in any case I already posted sometimes on PPRuNe, for semantic reasons, that I could not agree the CVR transcription and I rejected it.

If you maintain Vanity Fair as the truth your question must be now : "Why did the BEA modify facts and words ? Who from the regulators (DGAC including BEA) organised the lies, and why ? Why do the victims' families accept these lies ? And why do the specialists of human factor agree to sign the BEA's report(s) they know misexplaining the facts in that case.

I'm sure that Vanity Fair will not answer to these questions ! They will not be able to improve air safety by telling only some points of view to sell more paper without harming their concerned advisers.

DozyWannabe

17th Sep 2014, 15:16

@jcj:
The vast majority of Airbus FBW pilots will never (on the line, at least) encounter a control law other than Normal (and briefly Flare) for the entirety of their careers. They'll likely never encounter the hard protections on the line either, but we know of at least a few where the hard protections did aid a positive outcome.

G0ULI

17th Sep 2014, 16:10

As a low hours private pilot I found the Vanity Fair article to be a good account of events with reasonable inferrences drawn to cover the gaps in the official reports. As a Systems Analyst, I personally believe that the side stick control systems used by Airbus were the major contributory factor to this incident.

Side stick controllers are fine in single pilot military aircraft where weight, space constraints and the ability of the pilot to maintain control inputs under high G forces are paramount. They have no place in multi-pilot commercial aircraft unless physically linked so that each pilot is aware of conflicting inputs.

Fly by wire controls have improved aircraft safety, but by allowing aircraft to be flown using a sidestick control safety has been reduced by removing the physical and visual cues given by linked conventional control columns. Had this aircraft been fitted with such controls, it would have been immediately obvious that the control column was being held back and that exaggerated flight control inputs were being made. The situation would have been recognised at an earlier stage.

Clearly it is impossible for Airbus to modify their cockpit design to incorporate dual conventional control columns, but surely a means could be devised to ensure that each pilots control sidestick mirrors the position of the other. The force feedback mechanisms of PC joystick controllers could be a good starting point.

I personally dislike side stick controls to the extent that I have made it a point to never fly on a commercial aircraft that uses them. Fly by wire is fine, just so long as the cockpit looks "normal"!

The current practice of flying aircraft at the edge of the performance envelope in order to minimise fuel use and maximise profit can only lead to more of these upsets in the future, although most will be resolved with a less tragic outcome.

DozyWannabe

17th Sep 2014, 16:18

@G0ULI:

If you check the original thread, you'll find that there are positive and negative aspects to both sides when it comes to connected versus non-connected controls. Remember also that even in a multi-crew airliner, only one of the crew is supposed to be using the controls to fly the aircraft at any one time.

The only reason they used to be physically connected was because prior to hydraulic controls being the norm, there were occasions where extra muscle power may be needed in the event of control failure - with hydraulic controls, that ceases to be an issue. Also note that there have been several similar accidents to this one in aircraft with connected controls.

G0ULI

17th Sep 2014, 16:31

DW
You are absolutely correct and there were many other issues that complicated matters in this incident particularly CRM and the apparent lack of recent manual flying experience by all the crew. While there is no need for physically coupled controls with the advent of FBW systems, it does provide an additional level of safety by giving immediate visual and physical cues in stressful situations such as flight in abnormal attitudes. No need to interpret a display, the evidence is immediately in front of you.

DozyWannabe

17th Sep 2014, 16:51

In theory yes. In practice it seems that whether the controls are connected or not, if the PNF lacks the confidence to assertively take control and keep hold of it then you're going to be in trouble.

Also, modern connected controls don't necessarily do what one might think - the NTSB found that in the EgyptAir 990 case, with one pilot pushing forward on the column and the other pulling back, the result was actually a split-elevator condition which seriously impacted the aircraft's stability.

G0ULI

17th Sep 2014, 19:07

Winnerhofer
With a conventional linked control column, each pilot would feel resistance from the control inputs of the other. So positive physical feedback. The control positions are clearly visible to each pilot and to anyone else on the flightdeck. So positive visual feedback.

No need to check the instruments or press override buttons.

If it is a fight for the controls, then the stronger person will win, but that is more of a CRM issue.

It would seem obvious that the aircraft should respond to only a single control input and not try to interpret conflicting inputs from opposite sides of the cockpit by summing or subtracting the different control positions. The easiest way of accomplishing this is to have physically coupled controls in front of each pilot.

infrequentflyer789

18th Sep 2014, 00:06

Vanity Fair:
Nice article .. nothing new ..

I don't know, this bit:

it was considered to be serious enough that Air France had decided to replace the probes with ones of an improved design and had sent out an advisory to warn pilots of the problem.

is a new one on me - I always thought it was Air BUS that advised that the probes should be replaced, and Air France that refused and then pushed back at the AB recommendation. See e.g. safetymatters: Shifting the Burden (http://www.safetymattersblog.com/2011/05/shifting-burden.html)

_Phoenix_

18th Sep 2014, 01:36

Owain,
...there is very little sideslip in the early stages, but plenty of roll,which must say something about the roll/sidelip ratio. The yaw damper trace shows there was oscillatory behaviour in the yaw rate

Are we looking to a Dutch roll mode or to an adverse yaw damping?

_Phoenix_

18th Sep 2014, 02:46

DozyWannabe,
This seems to be an instance of overcontrolling, because the aircraft then rolls 6 degrees to the left.
...The controls over the next 15 seconds or so look very much like a case of opposite PIO in roll...

I just agree with your presumption of overcontrolling, that's normal startle affect at AP disconnection. Watch again the magenta graph after 2:10:10, the next 15 seconds. There is not a roll oscillation, equal quantity left-right. The roll attitude is always much larger on right side, therefore Bonin constantly corrected the right roll tendency. I think he was quite busy... keeping the aircraft leveled was only one of the tasks he had to do in the same time

...Bonin's completely unnecessary pitch inputs...

He trusted some instruments that look valid. He followed the FD bars.
see final report Figure 69: Evolution of FD crossbars
Why FD didn't step out, as AP disconnected, by design?

noske

18th Sep 2014, 08:22

fully automated

Are we in Tech log?
I am aware that whatever Langewiesche writes cannot be a primary source of technical information. He's good at telling stories for a broad audience without making too many errors on the technical details. No more and no less.

But the bizarre summary quoted by Machaca ("As Langewiesche discovered, the pilots had become confused..."), the sensational conclusion suggested by the article's title, and Langewiesche's questionable decision to work for a magazine that many seem to see on the same level as "Vogue" make the article look sillier than it actually is.

rudderrudderrat

18th Sep 2014, 12:43

Hi _Phoenix_,
Why FD didn't step out, as AP disconnected, by design?
BEA asked the same question on Page 210 Final report.

"4.3.7 Ergonomics
The crew did not de-activate the flight directors and did not call out any changes in FMA mode. It is not sure that they noticed the appearances and disappearances of the flight director crossbars. It is likely that the crew did not know of the mode changes when the flight director became active again, reading and assimilating the displays on the FMA in dynamic and stressful conditions not being instinctive or natural. It seems that requiring an action from the crew to re-engage this automatic system would, on the one hand, lead to a consistency with the autopilot and the autothrust, and on the other hand stimulate a check on the modes and the consistency of the commands presented at the time of the re-engagement.

Consequently, the BEA recommends that:
EASA require a review of the re-display and reconnection logic of the flight directors after their disappearance, in particular to review the conditions in which an action by the crew would be necessary to re-engage them; [Recommendation FRAN-2012-047]

Further, even if it is not sure that the crew followed the orders from the flight director while the stall warning was active, the orders from the crossbars were in contradiction with the inputs to make in this situation and thus may have troubled the crew.

Consequently, the BEA recommends that:
EASA require a review of the functional or display logic of the flight director so that it disappears or presents appropriate orders when the stall warning is triggered. [Recommendation FRAN-2012-048]"

neila83

18th Sep 2014, 15:53

roulishollandais

18th Sep 2014, 15:59

Thank you rudderrudderrat for your acurate post :) - very far from the Vanity Fair/Langewiesche magma including many false assertions - .

roulishollandais

18th Sep 2014, 16:27

Reading Langewiesche - supposed expert "human factors" - in Vanity Fair in October 2014 and the Airliners.net forum commenting on the same ... vulgar article ... shows that the forum PPRuNe is still better than that of Airliners.net !
I am appalled that a "specialist" human factors as Langewiesche could write so much against facts. The manipulators write beautiful texts, with 90% true and 10% false.

Owain Glyndwr

18th Sep 2014, 18:22

Phoenix

Are we looking to a Dutch roll mode or to an adverse yaw damping?

Good question! but impossible to answer definitively from the published DFDR traces.

OTOH:

One cannot roll the aircraft without exciting all three lateral modes, so dutch roll behaviour must be in there somewhere!

IIRC the outer ailerons are locked out at high speeds and with much of the rolling moment coming from spoiler deflection on the downgoing wing I would think it quite difficult to generate any significant adverse yaw.

PS: Looking at the traces again with a critical eye I see there was a permanent small rudder trim offset ( 0.5 deg right rudder) which could have generated a slow spiral divergence to the right.

PPS: I am now reliably informed that this apparent indication of a standing trim offset is a quirk of the A330 design - something to do with dis-similar metals in the linkage and changes of temperature apparently. This would make my speculation a nonstarter, but to maintain thread continuity I'll leave it here.

DozyWannabe

18th Sep 2014, 19:03

PS: Looking at the traces again with a critical eye I see there was a permanent small rudder trim offset ( 0.5 deg right rudder) which could have generated a slow spiral divergence to the right.

Right, but the rudder position appears to be more-or-less centred correctly. Would the Yaw Damper not compensate? Unfortunately the graphs are imported at a low resolution, so it's harder to determine exactly what the Yaw Damper and TLU were doing exactly (0.5 degrees offset being relatively minor on those graphs).

mm43

18th Sep 2014, 22:03

@Owain, Dozy

I've been following this current discussion, and remembered something similar in Thread No.8, i.e.

Post #219 - AF447 - Thread No.8 (http://www.pprune.org/tech-log/482356-af-447-thread-no-8-a-11.html#post7158429)

DozyWannabe

18th Sep 2014, 22:19

@mm43 - Cheers for the reminder!:

However, (and begging your pardon if I'm mistaken) - in that case you were talking about the behaviour after the stall had developed, whereas here the discussion seems to be centred on the initial right roll immediately after AP disconnect - i.e. when the aircraft was still flying.

[EDIT : Sorry - missed this...]
I just agree with your presumption of overcontrolling, that's normal startle affect at AP disconnection. Watch again the magenta graph after 2:10:10, the next 15 seconds. There is not a roll oscillation, equal quantity left-right. The roll attitude is always much larger on right side
Right, but on the second roll input to the right (02:10:13), he has the stick over a little further for slightly longer than he had it to the left, which could also be another explanation for the right-roll tendency. Additionally, they were in turbulence, which makes detailed correllation of input to response significantly more difficult.

OG's comment regarding the Yaw Trim setting is very interesting, because that may indeed have initially caused a tendency to roll to the right - however (and despite the low resolution of the graph import), it appears that the Yaw Damper might have been compensating towards the left shortly after AP disconnect.

He trusted some instruments that look valid. He followed the FD bars.
Again - that's a possibility, but by no means a certainty.

Why FD didn't step out, as AP disconnected, by design?
It doesn't always, which is why the first memory action in the "safe conduct" Airbus UAS drill:
http://www.bea.aero/docspa/2009/f-cp090601.en/pdf/annexe.06.en.pdf

is to manually ensure that both AP and FDs are OFF.

Owain Glyndwr

19th Sep 2014, 07:19

Dozy

Right, but the rudder position appears to be more-or-less centred correctly. Would the Yaw Damper not compensate? Unfortunately the graphs are imported at a low resolution, so it's harder to determine exactly what the Yaw Damper and TLU were doing exactly (0.5 degrees offset being relatively minor on those graphs).

I agree it is very difficult to read the rudder deflection - after a couple of attempts at digitisation I conclude the actual rudder and trim deflection are both around 0.3 deg rather than 0.5.

The yaw damper reacts to rate of yaw which would be VERY low in the early stages of a spiral divergence so it would be unlikely to have any measurable effect on these few seconds after AP disconnect. Prior to that of course the AP was holding the aircraft steady.

PJ2

19th Sep 2014, 18:02

Dozy;

You've mentioned a previously discussed key point in all this and that is the complete absence of memory items employed to deal with an abnormality. Instead, the PF launches into something on his own, leaving the PNF out of the loop, bewildered and trying to catch up while guessing what was happening.

Had the memory items been called for and actioned as trained and the ECAM discipline executed as trained, the stall and all ancillary effects of the stall would almost certainly not have occurred.

Some may put the crew's response down to "startle effect" but I don't think so and I don't buy it, for that is what training is all about - to avoid or reduce the effect of surprise and momentary anxiety and to maintain cockpit discipline in the face of elevated risk to the aircraft. The point is to stabilize/maintain control of the aircraft, start the drills/checklists, secure the aircraft and decide the next course of action. That is absolutely standard trained/expected behaviour in an abnormality or emergency, (which this was not, until it was turned into one).

I think there are important lessons here including flight control systems, automation dependencies, FBW design, triage of warnings, loss of the stall signal, movement of the THS during an abnormal flight control computer condition and so on, but the price of such learning in exchange for the original set of behaviours which led to this outcome, for whatever reasons such occurred, is obviously extremely and needlessly high.

Machinbird

19th Sep 2014, 20:25

Instead, the PF launches into something on his own, leaving the PNF out of the loop, bewildered and trying to catch up while guessing what was happening.

It has been apparent to me for quite a while that AF447 encountered a roll PIO after the the autopilot dropped out. What I have been surprised at is the lack of understanding by other pilots, particularly some very experienced ones, of just what this means to the pilot who is attempting to fly during a PIO. In fact, from the low key recognition by BEA of what the roll oscillation implied, I suspect that they did not appreciate the impact of a PIO on the pilot flying.

As I have said in the past, I did briefly experience a divergent roll PIO in an early visual simulator, and I do understand the potential impact on the pilot flying and it is devastating in a number of ways. Apparently this experience makes me part of a very small group who had an encounter, so I feel obligated to try to pass on an understanding to others.

Supposing you were hand flying and your aircraft suddenly began to roll continuously in one direction. Would you be able to handle that using CRM?
The best you would probably do is have both pilots applying counter roll yoke/stick inputs. It would de-structure a crew in short order if you did not get it under control.

Having your aircraft roll in an oscillating manner contrary to your control inputs is equally distressing. I have described my experience as like beginning to fall and desperately looking for a hand hold. Does that sound like another day at the office?

PIO only occurs when a pilot is executing direct manual control. Since most of you are flying almost the entire flight on autopilot, you will have fewer potential encounters with PIO than was previously the case in my day.

There have been some PIO incidents with modern aircraft. Some that come to mind are a few roll PIOs on approach in the A320 series and AA587 with the yaw PIO, and of course AF447. They are infrequent and difficult to predict. They are very dependent on pilot control strategy. You can just about bet that Bonin routinely employed the mayonnaise stirring technique when he flew.

PIO incidents greatly increase the probability of accidents. There are a few things that can be done during training to decrease the likelihood of encountering PIO while hand flying.

Expose pilots to PIO events in simulators. Once in a lifetime should be sufficient.
Monitor pilots' control techniques during training and promote a minimalist style of flying. Just because the FBW system lets you get away with mayonnaise stirring does not mean that it will not bite you some day as it did with AF447.

wiggy

19th Sep 2014, 20:59

Is it just me or is there a lot in that Vanity Fair Report that wasn't in the official CVR transcript? And a lot that really helps understand the cockpit dynamics and some of what was happening? Some of the conversation/lack of it is slightly alarming, the Captain seems focussed on anything except managing the flight prior to going to the bunk.

Leaving aside the timing of bunk rest I think the translations aren't great, but we need someone who is both a French national and a French speaker to really join the dots on this. I live in France so I've got a bit of a handle on the shrugs and winks side of thing :sad: ...but that said my French language isn't great. Even so having read the BEA French language CVR transcripts a while back I think the Vanity Fair translations, whilst being correct at a school boy/girl level, are a bit "clunky" and don't really convey the full meaning or emotion.

jcjeant

19th Sep 2014, 22:46

wiggy
Even so having read the BEA French language CVR transcripts a while back I think the Vanity Fair translations, whilst being correct at a school boy/girl level, are a bit "clunky" and don't really convey the full meaning or emotion. The BEA French language CVR transcripts convey also not the full meaning or emotion
This is not in regard of each sentences .. explanation about the tone used by the speaker
The real sound record of the CVR is the evidence you need for have the full meaning or emotion (provided your mother language is french)
AFAIK

DozyWannabe

19th Sep 2014, 23:21

Some may put the crew's response down to "startle effect" but I don't think so and I don't buy it, for that is what training is all about - to avoid or reduce the effect of surprise and momentary anxiety and to maintain cockpit discipline in the face of elevated risk to the aircraft.
Right, but as you know, it doesn't always work that way - and startle effect is now very much a recognised phenomenon. I believe we were discussing the Orly Tarom investigation earlier in this thread in which 80% of pilots tested instinctively pulled up when they received a Stall Warning they weren't expecting.

My view (for what it's worth) is that it's a more nuanced scenario than saying the whole crew were subject to startle effect. I believe Bonin certainly was initially, whereas Robert - while he didn't break into the UAS drill - was definitely trying to be more analytical. I'm going to cross-post something I put in a PM a couple of days ago to illustrate what I see:

The way I see it, there have been enough investigations into accidents involving startle response that there doesn't necessarily need to have been a logical reason for an instinctive "pull-up" reaction. What is unusual is that it was effectively held for so long.

Psychologically speaking, it's as objectively fascinating as it is subjectively horrifying - pretty much the absolute worst-case scenario as far as CRM goes.

What I believe beyond doubt is that whatever "spooked" Bonin, it locked him into a panic and "tunnel vision" mindset for the duration. What makes me think that is that he used the SS priority button on no less than three occasions - and not only did he not verbally express that he was doing so, but on two of those occasions did so having previously verbally handed control to Robert.

Robert starts the sequence with a very solid handle on things as far as I can tell, but for some reason his self-confidence seems to evaporate rapidly after he elects to wait for Capt. Dubois to return. From that point on he seems to have been susceptible to an insidious and creeping sense of denial. He wouldn't be the first person this happened to, but I have to say that I've always been of the opinion that he should have trusted in his training and instinct and been more assertive, because of the three crew, he seemed to call it correctly the most.

So it seems to me what we have for the first minute or so is Bonin apparently completely spooked and acting in a totally reactive manner, with Robert trying to be analytical (which is the right thing to do) - however the thought never seems to occur to him that the worsening situation is being caused by Bonin's control inputs and that he should therefore immediately take control rather than try to assist verbally.

For his part, Capt. Dubois bears some responsibility for this breakdown in CRM - but he arrives at the worst possible moment through no fault of his own. I have to suspect that Robert's mounting alarm had him on the verge of taking control anyway just as Dubois arrived - especially with the cue that Bonin's attempt to deploy speedbrakes indicated a significant misinterpretation of the situation.

Dubois' arrival also coincides with the cessation of Stall Warning as the AoA vanes go outside their certified operating envelope. Of course this behaviour is something that the engineers need to look into and improve if possible, but in the roughly five years since the accident, no manufacturer has come forward to claim that this would not have happened with their design, which may possibly be rather telling - though I'm not sure.

From this point onwards, Robert and Bonin's higher brain functions seem to be taken up with trying to explain the situation to their captain, but all the time, Bonin's "hind-brain" is still fixated on "pull up at any cost".

For what little it's worth, I am convinced that instead of electing to call Dubois back, Robert had stuck to his guns and said words to the effect of

"Pierre, you must listen to me. I know what's wrong. I have control. Please put your hands in your lap, read me the ECAM, and keep an eye on the speeds to tell me when they come back."
then we wouldn't be having this conversation.

In fact, from the low key recognition by BEA of what the roll oscillation implied, I suspect that they did not appreciate the impact of a PIO on the pilot flying.

In fairness, I don't think this reference was particularly "low-key" - they certainly seem to have considered the effect on the PF (it's worth reading the whole section, but it's a long piece, so I won't quote it all here):
2.1.2.3 Control of the flight path
When the autopilot disconnected, the roll angle increased in two seconds from 0 to +8.4 degrees without any inputs on the sidesticks. The PF was immediately absorbed by dealing with roll, whose oscillations can be explained by:

A large initial input on the sidestick under the effect of surprise;
The continuation of the oscillations, in the time it took to adapt his piloting at high altitude, while subject to an unusual flight law in roll (direct law).

In addition, the deviation in roll may have been caused by the risk of turbulence that had preoccupied the PF in the minutes leading up the autopilot disconnection.

Going back to the discussion of the roll, I was clued in to the existence of Figure 64 in the final report, which is quite illuminating - particularly the 5th (Roll attitude), 7th (Lateral wind) and 8th (Vertical wind) graphs down.

Looking at the bottom two graphs (7 and 8), you can see that the prevailing lateral wind was a crosswind to the right which has a high peak of 35kts and a low peak of 5kts. In general, from 02:10:08 to 02:10:20 the right crosswind is around 30kts. The Vertical wind graph indicates an updraft of around 15-16kts at the point of AP disconnect, which if I'm not mistaken would, in combination with the right crosswind, initiate a tendency to roll right.

Comparing this data with Fig. 26, it appears that Bonin's first left roll input at 02:10:07 coincides with a drop in the prevailing right crosswind to 8kts, which I suspect goes some way to explaining the initial overcontrol. The crosswind subsequently picks up again to 30kts one second later, and remains around there for the following 12 seconds.

What is interesting to me about the 5th graph (Roll Attitude) in Fig. 64 is that there are three plots. One (blue) is the DFDR data, which we've seen in Fig.26. The second (orange) is data from a simulation with both the wind and pilot input reconstructed and, unsurprisingly, this matches the DFDR data closely.

Most interesting to me, though, is the third (pink) plot, which is a simulated reconstruction of the aircraft's response to the winds *without* the pilot input. Now - presuming all the other factors (rudder trim etc.) have been fed into the simulation, this should give an accurate picture of the roll condition had the controls been left alone.

As such, the pink plot indeed indicates a right roll - however, it maxes out at 15 degrees (which is significant, but still less roll than would be considered normal) at 02:10:25. This roll attitude reduces with the reduction in average crosswind and updraft - and by the time the speeds start to come back online at around 02:10:38, the roll has reduced to 7 degrees.

Now - *if* this simulation is accurate, it indicates that there was no immediate risk of a spiral condition that required immediate attention. Obviously there's no way of saying for certain, but to me it indicates that there is at least some degree of accuracy to the "should have left the controls alone" theory.

Sorry for the essay folks! I'm away off with work for the next few days, so if I don't respond it's not because I'm ignoring people. :ok:

Machinbird

20th Sep 2014, 02:01

Dozy, I totally agree with your analysis of Bonin's mental state, except you really do not understand the reason for Bonin's mental state. That was a direct result of the PIO.

Bonin was startled by the AP disconnect, yes, but that was not what sent him into PIO land. His past bad flying habits caught up with him. He did not initially appreciate that he was no longer in Normal law, however he made control inputs that were unexceptional (to him) and had the aircraft react in an entirely unexpected manner. From that point on, he was mentally fighting for his life. That is the effect of PIO on the pilot flying, and you might call that a startling result, but it was not the result of startlement.

If BEA had personnel with PIO experience on the team, they should have made that connection, but that is not the direction they took, ergo, they were not able to properly cover that base IMO. This is not to cast doubts upon their work however. It seems that hands-on PIO experience is very rare.

I am not sure Bonin actually adapted to the new AF447 flying characteristics. It is equally possible that as the aircraft slowed and Q decreased, the aircraft's response began to match his mental model and he was finally able to stabilize the roll oscillation.

jcjeant

20th Sep 2014, 13:11

AF447 : un accident d'une extraordinaire complexité - Page 2 (http://avia.superforum.fr/t1517p20-af447-un-accident-d-une-extraordinaire-complexite#53006)
« STALL ! STALL ! STALL ! »

Dès que la vitesse a été supérieure à 60 kt l’alarme de décrochage s’est déclenchée …

Par un geste réflexe les deux pilotes ont réagit et ont tiré à cabrer sur leurs Sidesticks.Once the speed was greater than 60 kt the stall warning was triggered ...

By a reflex gesture both pilots reacted and fired on their nose sidesticks

Even because the Stall alarm problem (- 60 knots no alarm ... + 60 knots alarm)
• L’alarme STALL ne répond pas à la CS 25 car elle doit être active tant que le phénomène qui l’a déclenché n’a pas été corrigé.

• The alarm STALL does not respond to the CS 25 because it must be active as long as the phenomenon that triggered it has not been fixed

the pilots reflex is not good
When stall alarm sound .. it's never good to pull (raise the nose)

PT6Driver

20th Sep 2014, 17:43

AdamFrisch

20th Sep 2014, 20:18

What the Langewische and Vanity Fair article highlights will not be found in any accident report. It's also what makes it so interesting. And that is that French pilots are predominantly from the upper classes, well educated and from "good" and respected families high up in academia. All the pilots on AF447 were graduates from prestigious schools. Something they share with many Southern European, Asian or Middle Eastern airlines. Rich kids, basically. Rich kids who might be there simply because it was expected of them, or it maintains their social status, not necessarily because they're great pilots or have a passion for flying.

This is not the case in the US, Canada, UK and northern Europe. There's a fundamental difference there. And it is possible that it could have played a big part in the outcome...

jcjeant

20th Sep 2014, 20:19

PT6Driver
We need to look at why he pitched up
At the early time .. maybe this ......
http://avia.superforum.fr/t1517-af447-un-accident-d-une-extraordinaire-complexite#49903C'est un fait très important, lors de givrage de sondes Pitot, les altimètres des A330-300 et A340 ne bougent quasiment pas.
Sur A330- 200, l'altitude diminue, le variomètre indique une descente.
C'est ce qu'il s'est passé à bord de l'A 330 - 200 / AF 447, les pilotes ont subitement vu l'avion se mettre en descente.

This is a very important fact, Icing pitot probes, altimeters A330-300s and A340 almost no move.
On A330 200, altitude decreases, the VSI indicates descent.
That's what happened on board the A 330 to 200/447, the pilots saw the plane suddenly put down.Dans le même temps, l’altimètre perd 350 pieds, descente confirmée par le variomètre qui indique un taux de descente de 600 ft/mn.
At the same time, the altimeter loses 350 feet downhill confirmed by the VSI indicating a descent rate of 600 ft / min.http://www.via-caesar.fr/PATRICK/PFD-3.jpg

Machinbird

21st Sep 2014, 15:57

Machinebird
Not often I post on here but..
The PIO is somthing of a side issue, because his imediate reaction is to pitch up. The slowing speed will surely then have compounded the roll control problems.
There was a complete and utter breakdown of the methodology in dealing with non normal events. No one called the correct failures, no one allocated tasks, no one had any command of the cockpit far less the aircraft.
PF's initial reaction to pitch is what killed him and everyone else.
We need to look at why he pitched up and why did the pair of them not follow any known procedure for dealing with problems.
PT6Driver, First, let me welcome you to the discussion.
Your response is about what I would expect from 99%+ of the pilot population, however it is evident to me that you do not understand/believe the nature of the PIO experience.

It appears to me that the initial pull up that Bonin experienced was largely an unintended consequence of his battle with stabilizing the aircraft in roll, and I'll try to explain why.

Ask yourself what your response would be to an aircraft that is not following your control inputs. Would you calmly begin going through the checklist, or would you stabilize the aircraft first?

As pilots, we are trained to control aircraft rather precisely. If a wing drops, we pick it up. If the nose gets a little high we ease it down. That is all that Bonin initially did. He tried to pick up a low wing the way he always did and it over corrected with extreme rapidity. As pilots we automatically correct for errors on the subconscious level, so he again corrected the overshoot in the manner he was accustomed to and the aircraft again overshot his roll target.

With our hindsight viewpoint, we know that 100% of the roll oscillation was being generated by Bonin, but he did not know that. In his mind, he could not stop trying to tame this strange flight characteristic, he was the pilot flying, and he knew the autopilot was definitely out.

From my past analysis of the AF447 roll information, I know that the rapidity of his inputs then increased. The only thing that makes sense is that he determined that he was just a little slow in responding to to the roll rate and he could stop it if he was a little quicker about it. But the roll oscillation continued........... About this time, the aircraft began pitching up significantly.

Do you still hold the view that PIO is a tangential issue?

PJ2

21st Sep 2014, 16:29

Hi Machinbird;

Re, "Do you still hold the view that PIO is a tangential issue?"

If I may enter a response to your question to PT6Driver, yes, I continue to hold that view because PIO doesn't explain the PF's simultaneous instant pitch-up, which continued to be held in until the stall.

Also, even if PIO occurred and even as quickly as the stick was moved, the stick movements and the airplane's excursions are not significant - at most 10deg of roll throughout until he got control of the roll, whereas the stick movement rearward and resultant pitch-up was within a second or two of the loss of airspeed information to the point of triggering the first stall warning, and continued upwards well past the time the roll was under control. In fact, the bank angle did not increase beyond about 10deg until almost 90 seconds after the initial loss of airspeed information and about 45 seconds after the stall warning had begun.

If as you say, pilots automatically correct for errors at the subconscious level and thus the PF corrected the roll, why was the pitch not similarly corrected, but actually increased, with an obvious, certain outcome for a transport aircraft at cruise altitudes?

I understand a slight delay while the airplane is stabilized and control maintained, but once control of the roll was achieved, SOPs for the abnormality would then be called for. However, by pitching up and thereby significantly destabilizing the aircraft, an abnormality became a loss of control and obviously rendered the calling for the original drill ineffective because they had another problem on their hands.

What, and where is the root-cause link between PIO and the loss of control less than a minute later?

PIO is not causal here, nor has the initial pitch-up been explained even though I had posited that the PF was "recalling" the UAS drill for right after takeoff, pitching the airplane to the memorized-15-deg attitude. It's simply a theory among others.

We're back to my original comment in July of 2009 on the second thread - if the PF had "done nothing", (meaning, just maintained pitch and power as the airplane doesn't care about loss of an airspeed indication), the stall would not have occurred and the roll would have been controlled.

RetiredF4

21st Sep 2014, 18:58

PJ2
We're back to my original comment in July of 2009 on the second thread - if the PF had "done nothing", (meaning, just maintained pitch and power as the airplane doesn't care about loss of an airspeed indication), the stall would not have occurred and the roll would have been controlled.

I think we all respect and share this point of view, but the discussion imho at the moment is what went wrong when Bonin did what he did and why did he do it.
Besides doing nothing an expierienced crew should be able to do something without diregarding basic aerodynamics and without crashing within 4 minutes from FL 350

Multiple reasons had been mentioned, the startle factor in the report, the PIO effect discussed here, the former mentioned wish by Bonin to climb to higher altidude, suspected uncorrect seat position, to name a few. Most probably it was a combination of different factors, and i would not exclude PIO as one part of the unsuspected behaviour which might have left Bonin without any confidence into the aircraft and its feedback and its imstruments indications. He was pretty fast out of the loop, and he managed to kick Robert out of the loop as well. Both together managed to express so little information to the later arriving captain Dubuois, that he remained out of the loop as well. There must be reason for this total breakdown of professionalism we expect from pilots, and i think it is vital to dig deeper than the official report goes. I have expierienced PIO myself (assume in the same type like machinbird) after Take off, therefore i can follow this line of arguments.

We discussed another point in former threads, which might have played an aditional role. Without airspeed indication the FBW system uses preset gaines for deflecting flightcontrols in relation to SS deflection. I do not remember wether we came to a conclusion for which speeds those would be optimized. But sure Bonin was not used to them as such training (loss of airspeed in high altitudes at high speeds) was not done in normal training. The control inputs might well have been inapropriate for the preset (but unknown) gains thus causing flightcontrol deflectins and thus aircraft reactions not encountered before in all three axis.

PJ2

21st Sep 2014, 19:51

Thank you for your response, RF4. Yes, I do recall the various theories as to why the continued pitch. I'm not sure it's possible to know for sure though I do agree that digging deeper is necessary.

I'm not averse to new notions and am open to theories which move possible cause(s) some distance away from pure human performance factors, but I'm still not clear for example on how PIO is connected to a sustained pitch-up and stall. It does no good to say 99% of pilots don't know what they're talking about as regards PIO, (I've experienced it in Alternate law in the A330 sim but one gets used to it and is careful with the stick), and if that's true and it may possibly lead to a LOC, then there's a problem! In the meantime, we both know even viscerally, (subconsciously) that such pitch attitudes are just never achieved in a transport aircraft because we know the airplane will rapidly lose energy because the engines are right very near their max delivery of thrust and there's just no reserve power for recovery, so how come this one is an exception?

To me the rapid loss of focus and discipline is a more significant series of causes than all other explanations including PIO, the point being that if we are to learn something from this accident, it is in the area of human performance, the clarity of interface design, (warnings, drills) and the presentation of complex situations to those who must assess quickly but who might face such events once in an otherwise quiet, full career.

The Vanity Fair article says something quite valuable in its conclusion:

This is another unintended consequence of designing airplanes that anyone can fly: anyone can take you up on the offer. Beyond the degradation of basic skills of people who may once have been competent pilots, the fourth-generation jets have enabled people who probably never had the skills to begin with and should not have been in the cockpit. As a result, the mental makeup of airline pilots has changed. On this there is nearly universal agreement—at Boeing and Airbus, and among accident investigators, regulators, flight-operations managers, instructors, and academics. A different crowd is flying now, and though excellent pilots still work the job, on average the knowledge base has become very thin.

It seems that we are locked into a spiral in which poor human performance begets automation, which worsens human performance, which begets increasing automation. The pattern is common to our time but is acute in aviation. Air France 447 was a case in point. In the aftermath of the accident, the pitot tubes were replaced on several Airbus models; Air France commissioned an independent safety review that highlighted the arrogance of some of the company’s pilots and suggested reforms; a number of experts called for angle-of-attack indicators in airliners, while others urged a new emphasis on high-altitude-stall training, upset recoveries, unusual attitudes, flying in Alternate Law, and basic aeronautical common sense. All of this was fine, but none of it will make much difference. At a time when accidents are extremely rare, each one becomes a one-off event, unlikely to be repeated in detail. Next time it will be some other airline, some other culture, and some other failure—but it will almost certainly involve automation and will perplex us when it occurs. Over time the automation will expand to handle in-flight failures and emergencies, and as the safety record improves, pilots will gradually be squeezed from the cockpit altogether. The dynamic has become inevitable. There will still be accidents, but at some point we will have only the machines to blame.

_Phoenix_

22nd Sep 2014, 02:42

Owain Glyndwr, Rudderrudderrat, DozyWannabe thank you for your replies and inputs.
I subscribe to “I'm away off with work.., so if I don't respond it's not because I'm ignoring people”

BEA doesn't remark any PIO because it wasn't.
Here is a quote from final report:
"When the autopilot disconnected, the roll angle increased in two seconds from 0 to +8.4 degrees without any inputs on the sidesticks. The PF was immediately absorbed by dealing with roll, whose oscillations can be explained by:
- A large initial input on the sidestick under the effect of surprise;
- The continuation of the oscillations, in the time it took to adapt his piloting at high altitude, while subject to an unusual flight law in roll (direct law)"

“Mayonnaise Stirring” is just normal and characteristic to any AB-FBW aircraft
https://www.youtube.com/watch?v=pJrLKLoJG_o#t=201

Simply, that type of input at cruise speed is overcontrol, not to be confused with PIO. Pitch input was too much either, then the aircraft ended up in coffin corner, at almost 38000 ft. THS moved at NU stop. So, at that point, they were already doomed. From there, any other excellent crew would end-up pancake on ocean surface.

See how the longitudinal stability works for a non FBW aircraft
“...et rapidement les choses revient normale”, "...and things quickly returned normal"
https://www.youtube.com/watch?v=WAiA6ep95Sc#t=1508

Also, it's interesting to revise an accurate tv reconstruction. Robert acknowledge verbally “STALL”, in CVR transcript is (!), that was at only 50 seconds after AP disconnect, 30 seconds later, nothing else matter, either with “mayonnaise stirring” or with an expert touch
Sometimes, we can not see the forest due to leaves:
https://www.youtube.com/watch?v=WAiA6ep95Sc#t=268

roulishollandais

22nd Sep 2014, 04:49

It seems that we are locked into a spiral in which poor human performance begets automation, which worsens human performance, which begets increasing automation. The pattern is common to our timeIf we are agreeing with that, we are loosing the highest part of humanity : FREEDOM and RESPONSIBILITY.
I'm sure some resistant Folks are able to fight with success again for these old values high in the sky, high on our flags around the World. That cannot be written on closed SOPs and algorithms. Human must be able to close the loop.

PJ2

22nd Sep 2014, 06:03

roulishollandais, it is such a temptation to engage the philosophical question implied!

RetiredF4

22nd Sep 2014, 06:30

maybe we all should read on the matter of PIO, or better on APC, as that's the term now used.

Pilot Induced Oscillation (http://www.jetcareers.com/pilot-induced-oscillations.html)

A few points from the article

Traditionally, the letters P.I.O. have been taken to mean "Pilot-Induced-Oscillation". Today there has been some movement towards redefining PIO to move away from the traditional "blame the pilot" mindset. In that vein, PIO came to be defined as Pilot-Involved-Oscillation. The USAF Fight Test Center prefers the term as "pilot-in-the-loop oscillation", and many in the field have replaced the letters PIO completely by a new acronym, "APC", or Aircraft-Pilot-Coupling, although APC can refer to either an open or closed loop event.1 The NRC states: "Aircraft-pilot-coupling (APC) events" are inadvertent, unwanted aircraft attitude and flight path motions that originate in an anomalous interaction between the aircraft and the pilot.

A PIO can occur in any axis, although pilots are most familiar with oscillation in pitch and roll.

There are many ways to manipulate the controls and still perform with the certification standards required to pass checkrides. Some pilots fly by making small inputs, predicting what is needed (known as "low-gain"), while at the other end of the spectrum we have pilots who tend to use relatively large control inputs to accomplish the same task (high-gain). While this is, to some extent, attributable to pilot experience and technique, there are events that can drive a "low gain pilot" towards the "high gain" side, and high-gain can precipitate APC.

The "startle factor" is actually a term used in the test flight community. It can be due to a system failure, ice accumulation, TCAS RA or an unexpected mode change in the flight control system. For example, a pilot is flying in cruise when the autopilot disconnects due to an out of trim condition or similar anomaly. The pilot now tries to bring the aircraft back to the assigned cruise altitude with a large control input. The sudden necessity for the pilot to intervene often results in a much bigger correction than needed, and if conditions are right, a PIO may result.

And now we are back to PJ2's point.

First, it should be recognized that by definition, APC/PIO cannot happen unless the pilot is making inputs that are sustaining the maneuver, i.e., the pilot is in the loop. Consequently, the first step is to get out of the loop.

Because a pilot may be highly focused on a task when PIO develops, it is important that the pilot-not-flying assist the pilot-flying in recognizing the situation. It may take forceful intervention to get the pilot to reduce his gain, freeze the controls or, in particular, release the controls altogether.

Gretchenfrage

22nd Sep 2014, 06:32

The most revealing part of the VF article is the following:

Over time the automation will expand to handle in-flight failures and emergencies, and as the safety record improves, pilots will gradually be squeezed from the cockpit altogether. The dynamic has become inevitable. There will still be accidents, but at some point we will have only the machines to blame.

Right now if you dare blaming any machine, you get grilled and shouted down by the usual crowd, or should i say lobbyists. The machine seems to get away with almost everything, it's always the failure of the human of not having memorised well enough the ways the machine wants to get interpreted and operated.

The last sentence points at a almost philosophical (yes, PJ2), albeit inevitably realistic fact, that at some point even all the manufacturers and regulators will have to get out of their denial mode, pull their heads out of the sand and they will be confronted with only their machine.

What will happen then? It's anyones guess, but i gather that there will be no reversal of the trend, there will simply be another machine invented and sold overpriced to supervise and control the former machine. And by then it most probably also be capable of making a selfie!

It's only the world we are craving for.

infrequentflyer789

22nd Sep 2014, 19:09

Right now if you dare blaming any machine, you get grilled and shouted down by the usual crowd, or should i say lobbyists. The machine seems to get away with almost everything, it's always the failure of the human of not having memorised well enough the ways the machine wants to get interpreted and operated.

The last sentence points at a almost philosophical (yes, PJ2), albeit inevitably realistic fact, that at some point even all the manufacturers and regulators will have to get out of their denial mode, pull their heads out of the sand and they will be confronted with only their machine.

Until the machines design build and maintain themselves (at which point it will likely be their world not ours), there will always be a human to blame.

But in this case, you are forgetting, the machine was blamed, judged and punished - straight after the accident. The AA summarily removed from service and the more capable BA equally damned by association, all long before anyone knew how the pilots had mishandled the situation. In some ways this is unfair on the machine - the limits to its capabilities were already known, previous incidents dutifully reported to its human masters, its designers had already created a more capable replacement, and the aircraft mfr had recommended replacement. All well over a year before the crash.

One airline (at least) however refused/delayed replacement, presumably because some human there thought that that airline's humans in the cockpit could cope just fine with the less capable machine. That turned out to be spectacularly wrong.

So, where does the fault lie in the end ? With the machine that suffered the fault, but was already due for replacement precisely because of that fault, or with the humans who failed to cope with the fault, or with the human(s) who decided the fault was not a risk and could be handled by the other humans ?

PJ2

22nd Sep 2014, 20:17

RetiredF4;

Many thanks for your post. Machinbird, many thanks as well - as I say, I'm open to "PIO" explanations. I'll read the work at the link you've provided RF4, and am also searching / examining the papers extant and downloadable on the subject*. I'm not an engineer but flew the A330 1999 'til retirement, (as I believe you know) so all writing is of course from that p.o.v.

*Many papers require membership or attendance at the institutions publishing the papers. In my view, the advancement of knowledge and available peer review of SMEs who may not qualify under such narrow restrictions limits knowledge in exchange for the $20 bucks or so one has to pay for each paper. My views on patents are similar...

alf5071h

22nd Sep 2014, 20:37

Because the two views – machine or human, are held by humans, then the human is held accountable – us, ourselves, how we think about the situation; the process of thinking.

This is not as philosophical as might first appear; consider the events preceding AF447.
A technical weakness was identified with the probes – a rare form of icing. Solution, modify the probes. Until this could be done, the risk of a multiple failure was mitigated by having at least one modified probe, but if this was not possible, the crew would be trained to manage the outcome of the failure - procedure.

This is a logical engineering (technical), machine view. However, there was a break-out point when considering the crew – not as a weakness which required additional training, but as an asset which could mitigate the threat.

The flaw in the engineering view (manufacturer and regulator) was the failure to realise that the threat was the ice crystal conditions, not the resultant of an icing encounter with deficient probes. The alternative realisation required a human, operational viewpoint.
Thence in true TEM fashion, avoidance should have been the primary human activity, supported by information about the threat and how to detect it. Avoiding Cbs involves a routine and already trained for-behaviour, with procedural backup in the event of mismanagement (error).

Re automation dependency; it’s not the pilots who are automation dependent, but all levels of the industry. Manufacturers and regulators have difficulty thinking outside of their numerical, probabilistic approach to aircraft design and certification, whereas the real world involves the indeterminate behaviour of humans interacting with a machine. If regulators become locked-in by a machine view then blaming the human is a logical outcome.

‘A problem to control, or a solution to harness’. (http://www.scribd.com/doc/240601926/Employees-a-Problem-to-Control-or-Solution-to-Harness)

jcjeant

23rd Sep 2014, 19:10

Info OSV

L'interruption de l'alarme stall ne signifie pas systématiquement la sortie du décrochage (sur Airbus)
The Interrupt of the the stall alarm does not automatically mean out of the stall (Airbus)
(On Airbus) how the pilot know that he is out of the stall ?
By putting his hand through the side window?

_Phoenix_

24th Sep 2014, 01:39

Winnerhofer,

The quote "by Bill Palmer" is very good. Basically, I have the same perspective , Bill has better wording. One remark though… Instead of a piano player comparison, I would choose a Toyota Prius driver. I mean: Even with brake fully applied, the airspeed still builds up… "The experts are right." The driver needs recurrent training and practice for manual shifting in neutral... Is this right? or… Toyota has to change design!
Already in 2010, Prius was the safer car regarding the sudden acceleration issue
https://www.youtube.com/watch?v=IqG4mZbM9rs

400drvr

24th Sep 2014, 01:46

There is an excellent article in Octobers issue of Vanity Fair called Human Factors. Written by a non-pilot but very well researched. A recommended read for all of us.

I hope you read it.
http://www.vanityfair.com/business/2014/10/air-france-flight-447-crash

roulishollandais

24th Sep 2014, 04:55

Not again please with VF ad , many contestable assertions, manipulation

Machinbird

24th Sep 2014, 06:10

The article posted by Retired F-4 draws heavily on the National Research Council book on PIO that I have recommended in the past. (National Research Council. Aviation Safety and Pilot Control, 1997) For one with an understanding of aviation and the concept of feedback loops, it is an easy read, and as a soft cover version, is very affordable.

I use the term PIO because that is what I am comfortable with, but I interpret it as Pilot in the Loop Oscillation. PIO cannot occur if you are not exercising direct control.

I recently had a message from a tactical jet pilot describing a PIO incident he experienced. He stated that he was too busy during the incident to communicate with his back seater in any meaningful way. The workload was so high that he had no spare capacity to communicate! Can you see the relevance to the AF447 cockpit scene?

Yes, AF447 was not in a PIO when it departed from controlled flight, however the PF had just been through 30+ seconds of hell that shook his faith in his aircraft's control system and fatigued his mind. At the same time, the PIO incident had set him up for a stall by stealing his airspeed and leaving him in a nose up attitude.

Why didn't he control the nose attitude you might ask. The simple answer is that he didn't have the attention to spare to adequately perform that task. He was concentrating almost exclusively on the roll channel and I believe was already tense and nervous before the autopilot dropped out which caused him to pull the stick back unconsciously.

In the main AF447 threads, a gent who goes by the handle "grity" prepared several charts of stick motion for the first few seconds after the autopilot dropped. The following chart is the best of those:
http://home.comcast.net/~shademaker/StickMoves.png
The numbers on the move lines correspond to the time in seconds after time 02:10. Bonin took control at time 02:10:07.

If you will notice, the majority of the first 11 seconds that the chart covers was spent in the 4 to 8 degree back stick range and only toward the latter part of the period does Bonin provide a couple of pulses in the nose downdirection. For the first 9 seconds he is just banging the stick from side to side while holding it fairly consistently in the 4 to 8 degree aft stick range.

Here is what these roll inputs did to the aircraft:
http://home.comcast.net/~shademaker/InitialLateralMoves.jpg
As expected, there is still much skepticism about what PIO had to do with the loss of AF447.
Simply, that type of input at cruise speed is overcontrol, not to be confused with PIO.There is a simple difference between overcontrol and PIO. If you know you are overcontroling you can stop.
If you are in a PIO, you see the situation as the aircraft fighting your control and you cannot stop because you must keep the aircraft under control.

noske

24th Sep 2014, 10:04

Human Factors. Written by a non-pilot
Aw, c'mon. You can say a lot of things about Langewiesche, and he's certainly never been an airline pilot, but he used to earn his living as a pilot for several years, before he became successful as a writer.

but very well researched.At least he's been talking to a lot of people and has picked up some interesting observations along the way.
For example, from his visit to the BEA:
With most of the weather still lying ahead and an anxious junior pilot at the controls, Dubois decided it was time to get some sleep. The chief French investigator, Alain Bouillard, later said to me, “If the captain had stayed in position through the Intertropical Convergence Zone, it would have delayed his sleep by no more than 15 minutes, and because of his experience, maybe the story would have ended differently. But I do not believe it was fatigue that caused him to leave. It was more like customary behavior, part of the piloting culture within Air France. And his leaving was not against the rules. Still, it is surprising. If you are responsible for the outcome, you do not go on vacation during the main event.”
And:
Bonin didn’t just ease the stick back—he hauled it back, three-fourths of the way to the stop, and then he kept on pulling. Alain Bouillard, the French investigator, equated the reaction to curling instinctively into a fetal position.

G0ULI

24th Sep 2014, 13:19

So we have full control inputs in a high altitude flight regime where the rule is to fly with a light touch. Had a conventional linked control column been fitted to the aircraft, the PNF would certainly have noticed the gross control movements and probably been prompted to intervene at a much earlier stage. A classic case of the left hand not knowing what the right hand is doing and vice versa. Sure, the crew weren't communicating with each other in identifying the problem, but the situation was greatly compounded by a control system and instrument layout that effectively hid each pilots' actions from the other. The cockpit design and control layout are not designed for instinctive responses in abnormal flight upsets. Flight control inputs can be summed or subtracted if both pilots grab the sidestick controllers at once without one of them pressing an override button. Either pilot can press the button without the other being aware in a high stress situation so you end up with an aircraft that has unpredictable responses and a crew that are fighting each other (and the aircraft) at the controls without being aware of the situation, which seems to be the case here for at least part of the descent.

CONF iture

24th Sep 2014, 18:03

classic case of the left hand not knowing what the right hand is doing and vice versaThat's the beauty of the SIDESTICK philosophy as Airbus elected to implement it ...

bubbers44

24th Sep 2014, 20:45

GOULI, you are correct. No pilot would allow the PF to pull back a control yoke to the stops and hold it there. Only if the control inputs are hidden from him would he allow this mishandling, especially at high altitude, to happen.

Maintaining cruise attitude and power would have made this pitot tube malfunction a non event. That is what we were taught early in our flight training.

infrequentflyer789

25th Sep 2014, 11:09

No pilot would allow the PF to pull back a control yoke to the stops and hold it there. Only if the control inputs are hidden from him would he allow this mishandling, especially at high altitude, to happen.

Presumably those people in the accident record who did exactly that, despite yokes, shakers, pushers, manual trim, non-fbw etc., are not actually pilots then - along the lines of "no true scotsman" ?

I don't sit at the pointy end, but one thing I have learned from the user interface work I have done is that no matter how easy you try and make it for the user to do the right thing with the interface, and how hard you try and make it to do the wrong thing, sooner or later you always encounter the user who is actively trying to do the wrong thing. And they will find a way to succeed.

The real question is not what interface changes can be made to help the pilots do what they want to do, it is working out why this, and a lot of other crews, pulled _back_ and wanted to pull back to try and recover a stall. And why PNFs either think the same and assist, just acquiesce, or never make an effective attempt to take control.

"Climb, climb, climb"
”But I’ve been at maxi nose-up for a while”

They were falling out of the sky, and they both wanted to resolve this by climbing (and they checked the engines are at full thrust as well I think). Stall is only ever mentioned by the automated warning, how many times, over 100 ?, never by any of the pilots.

DozyWannabe

25th Sep 2014, 19:53

The machine seems to get away with almost everything
Apart from where the report explicitly criticises the stall warning behaviour and recommends a review of the control and warning interfaces, right?

Right now if you dare blaming any machine, you get grilled and shouted down by the usual crowd, or should i say lobbyists.
Again, I disagree. From my perspective it seems that while there are indeed lobbyists for the manufacturers - and an all-pervading corporate tendency to "circle the wagons" on the part of all manufacturers and vendors, it's also fair to say that on the other side there exists a subset of pilots - and some pilots' unions in particular - who act almost blindly as "lobbyists" in the other direction.

Sadly, what this means is that members of both of these "lobby" groups have ended up metaphorically sticking their fingers in their ears while loudly and repeatedly stating that their own case is a fundamental truth. This perception of an immutable "other side" means that we end up with both parties trying to shout the other down - and pretty much the only beneficiaries of this situation are journalists, many of whom make a lucrative living from a "let's you and him fight" scenario.

The last sentence points at a almost philosophical (yes, PJ2), albeit inevitably realistic fact ... they will be confronted with only their machine.
Except for the point that, like a lot of general journalism on the subject, the article assumes that technological advances and increased automation capabilities are aimed at removing human pilots from the flight deck in the foreseeable future which, to the best of my knowledge, is absolutely not the case (though it does make for interesting copy and is a psychological trigger for knee-jerk responses).

At some point, it's going to be necessary for those with an entrenched viewpoint on both sides of the argument to re-evaluate that position. Of course the legal/marketing departments of the manufacturers are always going to be aggressive in defending their products - for better or worse in this day and age they wouldn't be doing their jobs if they didn't. But I must say that the engineering side of things has accepted in recent times that while the technology has solved a lot of the inherent safety questions (and the stats absolutely support that conclusion), it has also added a new set of challenges in terms of safety. That the human/technology interface will always be something of a "work-in-progress" also seems to be generally accepted. However, the tech-sceptics on the piloting side also need to accept that the worst-case scenarios they predicted back in the '80s have not come to pass.

Ultimately both sides need to let go of the past. For example, at this point in time you will have F/Os coming on to the line who weren't even born when BZ made his infamous "concierge" remarks, and as such bringing that up at this point in time is largely irrelevant.

Furthermore, there needs to be an acceptance that when it comes to aviation safety, "blame" is a counter-productive concept. And I'm sorry to say this, but from my point of view it appears that the "tech-sceptic" piloting lobby is very much behind on this. Various articles and discussion from that end maintain that the AF447 BEA report "blames" the crew, when in fact it does no such thing.

Until the machines design build and maintain themselves (at which point it will likely be their world not ours), there will always be a human to blame.

But in this case, you are forgetting, the machine was blamed, judged and punished - straight after the accident.
...other very good points...
+1-Yes-Thank you-Got it in one. :ok:

however the PF had just been through 30+ seconds of hell that shook his faith in his aircraft's control system and fatigued his mind.
With all due respect, I think that's a very valid theory (particularly given your experience in that area) - however due to the nature of investigation of fatal accidents it can never be any more than that.

Why didn't he control the nose attitude you might ask. The simple answer is that he didn't have the attention to spare to adequately perform that task. He was concentrating almost exclusively on the roll channel and I believe was already tense and nervous before the autopilot dropped out which caused him to pull the stick back unconsciously.
Again, with respect, I feel a little dubious on that point for two reasons. Firstly, if my experience in the sim was anything to go by as regards the real aircraft, the spring-centre of the sidestick was certainly positive enough such that I don't believe inadvertent pitch input would be an easy thing to accomplish. Secondly, he makes a very specific reference to the fact he was deliberately pulling up, which does not tally with the idea the pitch aspect was unintentional.

I must say I'm in PJ2's camp there.

tdracer

25th Sep 2014, 20:26

Being a non-pilot, it would seem to me there is one aspect of this accident that deserves further study (which Bubbers touches on) - why would any pilot not only put the aircraft into a stall, but hold it there despite the aural warnings and indications of an extreme angle of attack for several minutes until the aircraft crashed?

It's easy to point out an inexperienced pilot, taught you can't stall an A330 and not trained how to control the airplane in Alternate law. Further I do think it is valid to suggest that the sidestick may have contributed to the PNF not realizing what the PF was doing (unlike seeing the control yoke pulled full back). But I think that misses the bigger point that other pilots have done much the same thing - holding an airplane in a stall while dropping thousands of feet until impact - on non-FBW aircraft with cross linked controls. Some were highly experienced senior pilots that certainly should have known better, and should have been under no illusion regarding their ability to stall the aircraft (Bergenair comes to mind).

So what is going on mentally that is preventing these pilots from recognizing they are in a stall, despite aural and/or stick shaker indications telling them exactly that? If we can figure out the cognitive processes involved, perhaps we can figure out how to prevent future occurrences.

DozyWannabe

25th Sep 2014, 20:53

...taught you can't stall an A330...
In fairness, in order to believe that one must believe that he never read his FCOM - which explicitly states that the A330 can be stalled outside of Normal Law.

But I think that misses the bigger point that other pilots have done much the same thing - holding an airplane in a stall while dropping thousands of feet until impact - on non-FBW aircraft with cross linked controls. Some were highly experienced senior pilots that certainly should have known better, and should have been under no illusion regarding their ability to stall the aircraft (Bergenair comes to mind).
Thank you - I've been banging on about that one for years.

alf5071h

25th Sep 2014, 21:17

tdracer, you approach the point of assuming that no one would overlook all of the available cues. Furthermore, it would be inappropriate to assume that something prevented the crew recognising the situation. Thus as you conclude we need to understand the mental process, but, BUT, many very experienced researches still seek a usable answer.

The metal processes which create success or failure are the same – the way in which we think; it’s only circumstance (outcome) which distinguishes between them.
Many posters would be well served by reading articles by James Reason and heeding the risk of bias (#582). The errors which the crew suffered (only determined by hindsight) – the processes, inabilities, assumptions, are no different than the manner in which many of the views above are stated. We are our own worst enemy – or at least our thinking is; #523 attempts to show this.

In the absence of knowing the crew’s thoughts, any understanding this accident depends on supposition. There has been significant technical analysis above, part supposition and part unverified ‘fact’ more often based on assumption; we must take care to define and acknowledge the line between them.
There is still great opportunity for supposition about the crew’s thoughts and behaviour. The VF article touches on this, but I suspect that most people are uncomfortable with this approach – too personal, whereas it’s much more satisfying to blame someone (which the report avoids).

In search of a solution, the industry needs to improve individual understanding of behaviour (behaviour is triggered by thought) – HF training (not necessarily CRM). Also review situations to minimise opportunity for error, and if encountered, how to recover or reduce the effects of the outcome – not stall recovery, the situations which lead up to it, the flight path, the procedures, training. Each might be seen as small and insignificant (bias), but together at a critical time there is great hazard.
The roots of this accident might be found well before this flight started.

DozyWannabe

25th Sep 2014, 22:13

Care to provide a summary for those that are a little reluctant to click/download random links?

infrequentflyer789

25th Sep 2014, 23:49

Care to provide a summary for those that are a little reluctant to click/download random links?

<opens page in scratch VM for paranoia reasons, downloads file>

At the time I accessed it (not that I'm trying to make you more paranoid there, just covering all bases) it was a PDF file on a file sharing / dropbox type of site. The file seems ok, scans clean, simple PDF comprising scanned images of printed spiral-bound report - no PDF features that would need Adobe reader, so you can open it in something safer ;)

Trouble with these docs are that they are scans of printed originals and either the scans, the original prints, or both, are poor quality. The appendices in particular may be second or third generation physical copies - printed or copied elsewhere, collated into physical report, photocopied again for distribution and then scanned, at a guess.

Summary of content - not dissimilar to the previous one. 200+ pages. In French, and with not many pictures...

I am torn as to whether the quickest and easiest approach would be:
a) to try attacking it with some professional rescan and OCR software, followed by google translate and then attempt to decipher the result - but I think it is going to be a waste of time
b) to (re)learn French properly

I have scan read the contents and parts of some interesting looking sections / pages, using my level of French, which 20+ yrs ago was "foreigner who has clearly put a bit of effort in but will need to be spoken to slowly in small words", and is now that plus "and has now forgotten a lot of it". I can still order food and wine and follow directions to the WC, i.e. the important stuff, but that doesn't help a lot here...

There looks to be some interesting stuff in there, comparing with other UAS incidents and interviewing some of those pilots (BEA report says other crews ignored SW believing them spurious - but frustratingly did not cover _why_ the crews believed that, this report might). Seems they have also done test flights (real not sim) and tried to recreate the scenario (up to a point), various FDR comparisons with test flight and 447 appear to be shown (quality of traces about as bad as you'd expect).

Most of the report is responding to questions, those posed by the court, by Airbus (I think), and by other parties (victims counsel ?). Fair amount on human factors and procedures and FCTM.

List of causes and conclusion around p140, not sure the list aligns with BEA exactly. Conclusion says (I think) that appropriate actions by crew would have avoided accident, that UAS and Stall procedures were perfectly adequate and would have avoided accident if either was executed correctly.

Hope that helps a bit.

Note: sidestick does not appear to be mentioned :E Also, looks like they asked pilots (maybe the test flight pilots?) about the controllability in ALT2B and conclude it should not have been an issue.

jcjeant

26th Sep 2014, 01:05

Hi,

Why this second expertise ?

http://i.imgur.com/IbFHOQc.jpg

it was at the request of Airbus .. as they were not enterely satisfied by the results of the first expertise for some reasons ........

PJ2

26th Sep 2014, 17:38

alf5071h, infrequentflyer789, dozy;

Re, "The roots of this accident might be found well before this flight started." - alf5071h

I think it is reasonable to say that this is almost certainly the case.

Over time, many "streams" became the "rivers" which may legitimately be included in the examination of the accident.

The dividing line between proper examination of human factors and the notions of "blame" varies in width, (blurs), depending upon the context of the conversation; it is a useful dialog to have in its own right.

What separates the ample signs of hindsight bias and blame throughout these threads, from the reasonable examination of human factors including command-and-control issues in both the cockpit of AF447 and the wider "rivers" of organizational factors?

It is an important distinction because both must have a venue in which serious discussion can take place.

As with almost all accidents, informed supposition and circumstantial "conclusion" form part of the impressions of what actually occurred in the cockpit of AF447. None can know for sure of course but those who do the work can often reasonably surmise without concluding.

The various theories of why the PF handled the stick the way it was and why these actions continued beyond the point where standard training was to call for drills and checklists in the manner now widely understood cannot conclude anything as we do not have the required data - what was going on in the mind of the PF and PNF and why things came apart so swiftly and why the the captain was unable to take command and why the unstoppable descent rate regardless of pitch and power inputs did not register. How do the notions of "panic" or "freezing at the controls" advance understanding?

Fright and narrowed focus in sensory and thought processes are understandable human responses to rapidly-unfolding events which are both confusing and life-threatening. The industry has known about such factors since the eighties. It created the notions of organizational factors and created CRM techniques, putting Perrow's, Reason's, Helmreich's, Maurino's, Kern's, Dekker's et al. work into practise and using level D simulators to combine forces in reducing the effects of such human, visceral psychological and biological responses. The combination works, as the industry's experience has proven over and over.

But we must explain things: Where and why were the failures here? Unlike the modelling of mechanical failure, can there be no "cartesian" conclusions here but only a thorough surmising?

This does not seem good enough to set improvement and change by. The solution isn't just to do "more CRM" or "more training"; I am familiar with, and have seen superbly-trained, veteran pilots do silly things in airplanes that leave one (thankfully) merely shaking one's head.

Why does this continue, is a question for all human endeavours but is paramount in high-risk ventures. To what levels may our present sciences probe such "skews" in thought and action? What is the model by which we may understand and comprehend AF447; more crucially, what is the understanding that understands, and how do we get there?

I think these are not philosophical questions but material questions on "the critical path" to new solutions.

DozyWannabe

27th Sep 2014, 02:49

How do the notions of "panic" or "freezing at the controls" advance understanding?
...
This does not seem good enough to set improvement and change by.
Believe me, I understand your frustration.

But I think that accepting the notion that there is a level of psychological shock capable of activating the mammalian "fight-or-flight" response - which even training, process and drill may not counteract - is a useful thing to understand in itself.

Don't get me wrong, it's a bloody frightening concept - and one which potentially flies in the face of several tenets the piloting community holds as dear and sacrosanct. In particular it raises a spectre that implies that even if you, as a pilot, attain a level of skill and experience on the level of, say, a Yeager or a Beamont - there nevertheless potentially exists a set of circumstances in which all that skill and experience may go out the window and as a result may not only get you killed, but killed in a manner which could easily be interpreted/construed as being of your own doing. Now I'm not a line pilot, but I don't mind admitting that that thought process would scare the bejeezus out of me if I were.

On the other hand, the flipside of accepting this possibility could be positive. By which I mean that if it were to become a central tenet of CRM that any pilot, no matter how skilled or experienced, may fall victim to irrational behaviour in the heat of the moment and - crucially - that this should neither reflect on their perceived piloting qualities nor necessarily be anything more than a one-off aberration (after all - anyone can have a "bad day at the office"), then I can't help but think it may go some way to erasing the stigma involved in having to take control in such circumstances - both on the part of the pilot who takes control and the pilot who has been relieved of control.

In some ways, this ingrained perception that pilots who made mistakes must have been in some way deficient was negated as far back as "Fate Is The Hunter". The only pilot in the book that Gann explicitly refers to as being incompetent and having obtained his position by deception is eventually found out and loses his career, but is not killed as a result of his incompetence. On the other hand, several pilots who Gann states were absolutely competent end up losing their lives regardless. It's a thought-provoking point.

Cool Guys

27th Sep 2014, 03:22

For any accident/incident/problem there is a root cause, and if that root cause is found and fixed the issue will be resolved and will not happen again. Aircraft accident investigation is particularly good at this. In some cases the root cause cannot be found so we investigate the contributory factors and the chain of events (effects) resulting from the original issue and we implement counter measures for these. Contributory factors can be minimised or eliminated and any effects can be minimised by introducing containment/redundancy/contingency measures into the system. Depending on the severity and nature of the issue this may be a legitimate, complete handling. The 787 battery issue is a contentious example of this. In the case of AF 477 the root cause of the PFs decision to pull rather than “do nothing” has not been identified however some contributory factors have been identified and some counter measures have been introduced.

Recently in this thread there has been 2 lines of discussions, one regarding the root cause and the other contributory factors. A couple of days ago some pilots were discussing contributory factors such as the user interface, PIO and cognitive bias. A poster then got on his high horse and demanded the root cause be found and a perfect solution implemented. Since then the root cause has been the topic of discussion. This is a valid topic but rather than allowing demands for a perfect solution to drown all else, in the absence of a root cause being found we can and must address these contributory factors. It is a very challenging task to find the root cause of a human decision during a time of stress. With our present level of understanding of the human mind it may take us 20 years to understand why the PF pulled rather than “did nothing”, we may never understand. However the contributory factors and effects can be identified and understood better. There can be many possible effects and contributory factors so a full handling can get quite complex.

The posts a couple of days ago were suggesting (from my understanding) the PF was overloaded due to PIO and the PNF was confused because he did not know what the PF was doing. This would appear to be one possible scenario which has tangible and more easily obtainable resolutions but it seems to have been discounted in the process of a quest to discover the root cause.

PJ2

27th Sep 2014, 07:33

Dozy, sorry, no, you're misunderstanding my statements. I'm not the expressing frustration in the least.

I'm characterizing a state of affairs and wondering, positing the question. [added] By far, these are areas, (the cognitive, the 'psychological'), of investigation/examination that would bear most fruit, so to speak. The frustration if there is any at all, is knowing that "more automation" may be the answer some arrive at and that is just a variation on a current theme and won't solve a thing. There is a basic level of complexity and speed in cockpits of transports that must be "married" to appropriate capacity, ability, motivation, propensity towards a disciplined mind and an abiding curiosity of those hired. None of these human qualities are decent partners with automation or more automation. Airline managements of the eighties made the fundamental mistake of assuming automation could be the third (or second) pilot and saw a huge financial benefit in terms of training and hiring, particularly when the "best-and-brightest" candidates as Sully put it, were no longer choosing commercial aviation as a profession. It's a huge story as I know you and many here already know.

AF447-type accidents are going to continue, (in fact, have continued), until this mistake and the turn towards such research and the resulting changes in design take hold of industry priorities.

The "fright" is part of aviation and part of the business. There are ways to work with it; it isn't a sword of Damocles, it is just a fact. But some pilots haven't read Gann and think because its comfortable and really safe that it's a nice office job. Seen it, flown with it - THAT is what scares me - not the possibility of losing it - I know SOPs and CRM work - used it, but the industry is changing. Re-read the last paragraph of the VF article. This is what it has become, and is what needs coming to terms with.

400drvr

27th Sep 2014, 15:41

Of course it can be stalled once the computers give up the ghost. It's still an airplane after all. 2 wings, 2 engines and all the other plane parts that make it fly.

PJ2

27th Sep 2014, 20:14

400drvr;

I posted the graphic below several years ago on one of the AF447 threads. It's from an early CBT, (ca. 1999) training presentation:

http://www.smugmug.com/photos/i-KRNSqw4/0/L/i-KRNSqw4-L.jpg

400drvr

27th Sep 2014, 20:50

Having never flown the Bus I would have to go with "B"

Thanks for the post. I updated my post with a link to an article in Octobers Vanity Fair that was well written and researched. A real eye opener.

Regards

alf5071h

27th Sep 2014, 20:58

PJ, #548 “we must explain things” … and the other questions.
This is a normal human desire; however, if we are unable to describe the overall operational system due to its complexity (man, machine, and environment), nor define the interactions, we will always face an uncertain future.
In these terms AF447 is unexplainable. Self-satisfaction will find whatever we choose to look for (root cause), we only fix what we find, thus avoiding repetition of the last accident, but providing little benefit for avoiding other accidents.

At best we must continue to ask questions, particularly by considering the widest range of possible contributions. In an accident there are many contributory items (at that time, in that situation) all are required, yet the absence of any one could prevent it. Thus our thoughts about potential contributions provide opportunity for safety activity.
Some considerations relating to AF447 :-

Loss of Airspeed (LoA) simulator training; did the FOs handle the aircraft during this session or was it a Capt only exercise? Did the LoA simulation degrade the control laws?

Are FOs given opportunity to decide on a course of action for avoiding weather, or is this always deferred to the Capt? Not a CRM view of a Capt stating a course of action and then seeking FO confirmation, but mentoring – “look at this situation, what would you do”, then discuss the reasoning and judgement, thereby learning and exercising skills of thought.

There are several new safety initiatives which could supplement current safety activities.
Amalberti requires that we must avoid ‘the big one’, but without great detail of how to. He also notes that continuing improvements in the current safety methods must avoid upsetting the delicate balance and creating ‘new’ accidents; yet many recent accidents could have been created – automation, reduced training, commercial pressure, SOP culture.
‘Resilience’ provides some practical suggestions; a continuous process of adapting to changing situations with an alternative view of human performance. Although Resilience does not provide a solution (it cannot – as above), it does stress the need for activity – safety is what is done. By continuing to ask questions, reviewing normal operations we might find indicators as to why accidents have occurred, but of greater importance aspects which can be improved to avoid the next (big) one.

The flight operational side of our industry shows signs of lagging in these activities compared with Eurocontrol, or are we too constrained – regulatory or self inflicted.
‘Systems Thinking for Safety’. (www.skybrary.aero/bookshelf/books/2882.pdf)

G0ULI

27th Sep 2014, 21:21

Consider the driver of a modern, very up market car fitted with all the latest driver aids and stability assistance, cruising along an Autobahn at high speed in a rainstorm. The car has differential braking to all four wheels, distributed drive to all four wheels, rear wheel steering assistance and an electronic front wheel steering system.

The car turns into a bend, the inside wheels are braked slightly, the rear wheels turn in and the driver's steering input is harsh enough to cause the front wheels to start to lose traction, so the engine power is reduced and the sensitivity of the steering input reduced. The car continues around the bend, apparently completely in command of the driver, who notices nothing out of the ordinary.

[AF447 flying in moderate turbulence on autopilot]

On a straight stretch of motorway the car ploughs through some deep puddles and again the electronic stability controls activate to make this a non event for the driver. However due to a faulty seal or a design fault, some water enters the electronics compartment housing the stability systems.

[AF447 hits heavy icing conditions, ice starts to block pitot tubes]

At the next bend in the road, the driver turns the steering wheel and water sloshes up onto part of the circuit boards causing a fault condition. All stability controls are disabled and the car reverts to direct response to the accelerator, brakes and steering.

A light comes on the dashboard to warn the driver as the car starts to veer off course to the side of the road. The driver jerks the steering wheel and floors the accelerator to try and kick the back end of the car around. Perhaps it swings too far, so opposite lock is applied. That doesn't work, so the driver then brakes in a panic, causing all four wheels to lock, whereupon the car aquaplanes and fishtails or spins down the road into a crash barrier.

Drivers involved in aquaplaning and black ice accidents nearly all report feeling the vehicle accelerating as the skid begins. This is actually caused by a feeling of acceleration as braking force is lost, rather than any real increase in speed.

[AF447 Controls revert to alternate law (effectively direct input) and fail to protect the aircraft from stalling. The control column is held back to maintain altitude and reduce speed as the pilot flying seems convinced the aircraft is overspeeding. This feeling may have been induced by the extra noise of ice crystals hitting the aircraft giving a false audible impression that the aircraft was flying faster than normal.]

Now a highly experienced driver would have known to feather the throttle to keep the wheels spinning and to gently steer around the bend and maintain control of the vehicle until they could stop or sort out the situation. They might go wide around the bend, but they would probably avoid an accident.

High level automation in cars makes mediocre drivers feel like world class racing drivers by covering for their limitations. When all that assistance is suddenly removed, accidents happen.

In the case of AF447 a situation occured where a pilot with relatively limited experience was suddenly required to respond to a situation and gross control inputs were made in the same way that an inexperienced driver might respond to a sudden skid. Under normal circumstances, the aircraft would have modified the control inputs to maintain safe flight. Under alternate law, these safety measures were severely degraded or absent.

There were many other contributory factors, but this might help explain the initial actions of the pilot flying. It isn't necessary to seek out some complicated psychological reason for these responses. They are just normal human reactions to a sudden shock.

Unfortunately, once control is lost of either a car or an aircraft, it is very difficult for anyone other than the most experienced driver or pilot to retrieve the situation and only then by intervention at the earliest possible stage.

This might make it easier for non pilots to understand how the initial situation arose. The initial panic and unexpected responses from the aircraft also served to disorientate other crew members to the extent that identification of the problem and recovery was delayed until it was too late.

It is a situation that could happen to any one of us, particularly in a modern car.

PJ2

27th Sep 2014, 23:25

400drvr;

Re (b), QED.

This point regarding loss of protections including stall protection was stressed over and over again throughout the three courses I had on the Airbus A320, A340 & A330. Flight in the Alternate Laws was taught and practised. Both the A340 & A330 simulators were sensitive in roll but moreso, I am told by those who have flown it in Alternate law, than the airplane.

The VF article link has been posted earlier, above. Good article.

DozyWannabe

28th Sep 2014, 01:35

Dozy, sorry, no, you're misunderstanding my statements.
Apologies if that is the case...

The frustration if there is any at all, is knowing that "more automation" may be the answer some arrive at...
It's definitely a possibility, but it would appear that in at least one case, the thought process is going the other way. Remember this thread from a few months back? :
http://www.pprune.org/rumours-news/496702-airbus-takes-pilots-back-basics-a350.html

Airline managements of the eighties made the fundamental mistake of assuming automation could be the third (or second) pilot and saw a huge financial benefit in terms of training and hiring
I'd say the "brain drain" and the rise of MBAs in lieu of long-time airline experience in management really only took hold in the late '80s and early '90s. In day to day terms the state-of-the-art in FMS and automation hasn't really altered a great deal since the advent of the B757 and B767 in the late '70s/early '80s (and yes, I know Airbus pioneered a lot of it with the A300 in the early to mid-'70s, but as I understand things it was the then-new Boeing types which really cemented it as part of the norm). I think that when modern FMS/automation systems were introduced, it was still at a time when the builders assumed at least some airline management could be trusted to be sensible. Of course, with the drain in airline management experience towards the end of the decade, it could be reasonably argued that the monkeys had been given the keys to the banana plantation - and if the example of the "basics" push from Airbus is a sign of how things are going to go, then there's some hope that the balance will go some way to being redressed.

particularly when the "best-and-brightest" candidates as Sully put it, were no longer choosing commercial aviation as a profession. It's a huge story as I know you and many here already know.
Sure - however I would caution folks to beware the allure of the rose-tinted spectacles when looking at the past. In particular, blind pew's book contains some truly eye-popping examples of cadets being rushed onto the line before they were ready in the midst of a long-running industrial dispute. And this was at a well-known British airline back in the late '60s and early '70s, a period which quite a few of the old guard consider to be a golden time if reading stuff on here is anything to go by.

It's no secret that remuneration and benefits for pilots have been eroded considerably over the last few decades, but this is true of pretty much any profession (excluding banking and management) anyone would care to name. And with airline travel becoming such a commonplace thing, the truth is that what the airlines offer prospective pilots is a job with built-in expectations of unsociable hours, long periods away from home and family - and a pay packet which compares unfavourably to that of the branch manager in the local bank.

The "fright" is part of aviation and part of the business.
Of course - the point I was trying to make being that recognition of startle effect should be trained as part of CRM with a "no harm, no foul" understanding on the part of both parties. By which I mean pilots like Robert and the F/Os on the Birgenair and West Caribbean flights should not have to first consider the potential risk to their careers before being empowered to take remedial action.

Of course it can be stalled once the computers give up the ghost. It's still an airplane after all. 2 wings, 2 engines and all the other plane parts that make it fly.
I agree with the sentiment, but feel compelled to point out that the computers did not "give up the ghost" during the crucial phases* of this accident. The computers in fact functioned correctly and transitioned to a control configuration which gave the pilots increased control authority while maintaining as many normally-functioning systems as possible. As I've said before, the switch to Alternate Law is essentially only doing what the Flight Engineer would do on older types in the event of a component or partial system failure.

* - The ACARS readout indicates a transitory failure of PRIM 1 and SEC 1 very late in the sequence, but the DFDR data indicates that this didn't affect the behaviour of the aircraft or the system as a whole

@G0ULI - I'm afraid I'm going to have to pick a few holes in your car analogy. For one thing, even most of the current state-of-the-art driving aids only serve to alert the driver to potential hazards. Self-driving cars are still barely out of the experimental phase, and being at the controls of a self-driving car can be the only logical equivalent of the kind of automation experienced at the controls of an airliner in cruise.

The loss of traction control (which is a driving aid - not automation) could kinda/sorta be used as an illustrative example but for one salient fact - namely that a driver used to traction control who suddenly loses it is used to driving with the traction control on, and would therefore experience the loss as a deviation from the norm in terms of handling. However, *neither of the F/Os on that flight deck had ever received any training or experience in high-level manual handling at all* - and there was therefore no experienced "norm" from which the handling could deviate.

Also, the only manual handling experience they did have would have been on takeoff and approach/landing (i.e. at low level) where - I believe - the Normal Law roll rate is greater than it would be at high altitudes anyway.

In the case of AF447 a situation occured where a pilot with relatively limited experience was suddenly required to respond to a situation
Debatable. If the graphs in Fig. 64 of the report are accurate, then there was no immediate danger of departure from controlled flight and thus no "sudden [requirement]" to start manipulating the controls (though admittedly it would take some degree of nerve to cover and observe only).

It isn't necessary to seek out some complicated psychological reason for these responses. They are just normal human reactions to a sudden shock.
The interesting psychology isn't really applicable to the initial shock response - it has more to do with issues arising later in the sequence, particularly the PF's repeated (and silent) use of sidestick priority to take control away from the PNF (even having verbally handed over a few seconds prior) and the PNF's gradually increasing self-doubt over his interpretation of the situation which was - initially, at least - correct.

The initial panic and unexpected responses from the aircraft also served to disorientate other crew members to the extent that identification of the problem and recovery was delayed until it was too late.
Again, inaccurate. F/O Robert was trying to analyse the situation and appeared to be some way to figuring things out, but was likely distracted by Bonin's inputs making the situation worse.

Machinbird

28th Sep 2014, 03:54

G0ULI

28th Sep 2014, 12:15

DozyWannabe
I deliberately picked an example of a car which is available to purchase today rather than a self driving car, which I agree would have been a more accurate comparison.

The point being that automation covers many deficiencies in technique to the point where a less experienced pilot can be unaware of, or forgets to check trim settings and fails to use secondary instruments to infer missing information when the primary systems fail.

In the interests of fuel economy and cost saving, aircraft are being flown automatically at and beyond the ability of the best human pilots. When that automation fails, seeking to maintain extreme altitude for a given aircraft weight is not a wise course of action. It is however contrary to a pilot's instincts to deliberately fly to a lower altitude at night, in bad weather, over the middle of an ocean.

Machinbird
I suspect that pulsing, twitching the controls, or mayonnaise stirring is symptomatic of the PlayStation generation of pilots. It certainly is a popular technique among computer games players to speed up reaction times and maintain finer control. The technique works well so long as you have a computer to interpret and modify the inputs, less so when everything is hooked up directly to the controls.

jcjeant

28th Sep 2014, 16:01

The point being that automation covers many deficiencies in technique to the point where a less experienced pilot can be unaware of, or forgets to check trim settings and fails to use secondary instruments to infer missing information when the primary systems fail.We can automate most .. improve existing systems .. invent others .. make extensive training pilots ... all this is possible .. this is often done ... everything evolves
But one thing is certain .. the man is always the same .. with its strengths and weaknesses .. and that no automation or system will not be able to change this
There will always be people who have a particular momment do the opposite of what is required and action will succeed even with all the barriers placed to prevent them of such actions

PJ2

28th Sep 2014, 16:22

Hi Dozy, thanks for your replies. Points made and understood. It's complex, no?!

Machinbird, interesting thoughts. I wonder about the notion that the stick was there in one's hand and so one had to 'do something' with it. It is an extremely powerful extension of the hand as you know. It takes time, a bit of experimentation, discussion and particularly paying attention in class to learn what the stick actually does and as a result, how to use the stick well.

Stirring the stick was a well-known action. I recall many years ago, (early 90's), examining some data of an Airbus aircraft on approach in gusting conditions. The autopilot was on and the aileron parameters showed small, sometimes slightly larger deflections as the bank angle was held as commanded, (level, on an ILS approach, not turning). The autoflight system has roll information more accurately and quicker than pilots are able to perceive through the PFD and their human sensory equipment; the autoflight anticipates more quickly and makes smaller adjustments. ....(Even when disconnected, the Airbus remains in what Boeing drivers today would call CWS, control-wheel-steering; within certain limits, the airplane stays in its last-commanded pitch-and-bank attitudes. This is because lateral movement of the stick is a roll-rate request and longitudinal movement is a 'g' request - many here will know this from the extensive discussions on the autoflight system's design to maintain 1g flight, but what may not be appreciated is the implications for manual flight.)

When the autopilot was disconnected just prior to touchdown, one could see the difference immediately, (different, not bad, not good, just different) - the aileron deflections using stick control were notably greater in amplitude. I did not note any PIO by the way but it seems to me that that would have been an ideal circumstance in which PIO was "in-forming" the airplane-pilot combination.

The important understanding in use of the stick here is of course, neither bank nor pitch change unless commanded, ergo, in gusting conditions, leave the airplane alone and it will return to it's last-known-and-commanded attitude.

Very difficult though for a pilot to "leave the stick alone" when in hand, and watching the rocking-and-rolling in turbulent conditions; one must by virtue of being a pilot "do something" to make the machine do as one wants. The notion that it will if one just leaves it alone has not been "exorcised" either in training or in the industry; stirring the pot is endemic because, (the subtle thought or 'the psychology' goes), "when one has one's hand on the stick, one does things, for why is the hand there if not to move the stick?"

I know the notion seems simplistic; the vast majority of pilots are keenly purpose-driven and will quite legitimately state that unless one "controlled the airplane" it wouldn't do as one wanted in manual flight. I think that is largely true, but not exclusively - I tried "leaving the stick alone", (holding, not moving) and it is a challenge.

One could leave the stick alone in a gusting approach to wait to see what the autoflight (not the autopilot) system was going to do when a wing dropped momentarily on short final..., or one can do what comes naturally to any pilot and use the available control (column or stick) to pick the wing up and settle the airplane.

FWIW, I think the stick arrangement is more condusive to PIO than the wheel/column arrangement which is why it is important to learn how to fly the stick. When one makes inputs, one must make small movements, (move the hand nearer to the base of the stick, use only thumb-and-forefinger and move from the wrist...those kinds of things).

In doing so, one must learn when to leave it alone, even while it is in-hand. It is old, old hat for Airbus pilots now, but one can rotate the airplane on takeoff to 15deg and just let go of the stick until the next requirment for attitude change arrives. The airplane maintains the attitudes, despite gusts.

In practise one doesn't do this because one is "in command" of the airplane; the subtlety is that one must move things to "be in control", but knowing when to "do nothing", (different than the AF447 "do nothing - that entreaty would apply to all airplanes, not just the Airbus), is as important an understanding with such a system.

Of course in Alternate Law, it is just an airplane, and (essentially, though not wholly) WYSIWYG, and the discussions on AltB and gain have been had.

In the vast debates regarding stick-or-column one wonders how a stick-as-column, (like a fighter) would fare in the "visibility-of-the-other-pilot's-input" discussion. I'm sure Airbus considered it. To me it is a non-sequitur because the PFD certainly says sufficient to know what needs doing to correct whatever the other pilot may be doing incorrectly, and there is always the take-over button which one can either take over momentarily using the Priority button or take over permanently by holding the button in for, (IIRC), 15 seconds.

I hope this little input is helpful. I'm sure you know most of this but there are some subtleties from an Airbus guy's p.o.v.

jcjeant

28th Sep 2014, 17:00

there is always the take-over button which one can either take over momentarily using the Priority button or take over permanently by holding the button in for, (IIRC), 15 seconds.
We have also had seen how this can be used .. in the AF447 case !

_Phoenix_

28th Sep 2014, 18:50

The control technique used seems to be relatively unimportant in Normal Law but has critical implications in Laws that involve direct motions of control surfaces such as Alt2b.

Only the roll was in direct law. Pitch was never direct.
"When the airplane's nose was pointed up, it stayed pointed up - to maintain the same g-load and pitch rate.."
Bonin's "heavy hand" fully controlled the roll, he kept the wing at horizontal during 38000ft stalled descent, this might say something about his piloting skills, apart of imaginary PIO
Bonin held the sidestick back in that initial pull up, just long enough to pitch the airplane to 12°, we still don't know why... however it happened relatively fast and similar to AF471, see post #524.

Consider this analogy, if you want: the driver sees the traffic light turning from green to yellow, initially he brakes (pulls up), but it is not good (stall alarm goes off). The driver decides to go through the intersection, he bottom the acceleration pedal but for some reason the car's computer continue to apply brake and service brake all together (THS max NU and elevator half way NU), the engine's torque simply cannot overcome the brake applied to disks, the driver is confused "I lost all control of the car I don’t understand anything, I've tried everything"... he brakes again, the car stops in the middle of intersection, independent of any other action of the driver.

jcjeant

28th Sep 2014, 19:56

Hi,

"Two test flights were conducted with crews at Airbus Airbus Air France without being informed. Anyways .. those test flights are useless
1- Not performed in the same environment of the one of the AF447 encoutered
2- The pilots knows in advance what will happend .. what alarm will sounds .. etc ...
We know already by BEA report that the aircraft performed "as per design"

Only interesting if those test flights were performed by the AF447 pilots (same as a reconstitution on a crime scene)
Unfortunately the pilots are no more there ...

DozyWannabe

28th Sep 2014, 23:26

It appears that there are two types (at least) of control inputs being made on the stick by Airbus pilots.
One group applies pressure on the stick to create the desired aircraft response.
The other group makes short pulsing movements of the stick to make quick successive adjustments to the aircraft's flight path.
I suspect that pulsing, twitching the controls,... is symptomatic of the PlayStation generation of pilots. It certainly is a popular technique among computer games players to speed up reaction times and maintain finer control.
Nope. The "pulsing" method of control is the one recommended by Airbus, and it has nothing to do with video games, it's simply that it provides a method for observing the effect of the input and thus moderating the next input (or indeed figuring out whether further input is necessary or not).

I'm pretty sure that the term "mayonnaise stirring" as applied to Bonin's input patterns related not to the "pulses", but to the abnormally large degrees of deflection and the unusual number of reversals.

The nadir of this was his extreme and sustained pitch-up input between 02:11:40 and 02:12:15:
http://i1088.photobucket.com/albums/i331/turricaned/fdr-munge.png

I deliberately picked an example of a car which is available to purchase today rather than a self driving car, which I agree would have been a more accurate comparison.
Right, but the point I was making was that traction control and Normal Law roll and pitch rate techniques are not automation.

The point being that automation covers many deficiencies in technique to the point where a less experienced pilot can be unaware of, or forgets to check trim settings
If he hadn't pulled up so aggressively, the trim setting would have been a moot point.

In the interests of fuel economy and cost saving, aircraft are being flown automatically at and beyond the ability of the best human pilots. When that automation fails, seeking to maintain extreme altitude for a given aircraft weight is not a wise course of action.
I wouldn't say that about ability, though it is true that even the most basic gyro-driven wing leveller will be able to react to and damp out turbulence-driven bumps and rolls better than human pilots just because of the inherent lag in human reaction time.

The aircraft was not flying at an extreme altitude at the start of the sequence. Bonin discussed going higher with the Captain prior to the latter taking his rest break, and the latter said it was inadvisable for safety reasons.

It is however contrary to a pilot's instincts to deliberately fly to a lower altitude at night, in bad weather, over the middle of an ocean.
He didn't need to go to a lower (or higher) altitude, just maintain the altitude he had.

I've uploaded a grab of Fig. 64, as I was describing before:
http://i1088.photobucket.com/albums/i331/turricaned/af447_fig64.png

Again, note in particular the brief and significant drop in the right crosswind at 02:10:07, which would go some way to explain the initial overcontrol to the left, and also note the pink plot in the "Roll Attitude" graph, indicating that the turbulence would not have rolled the aircraft over.

@PJ2 - Cheers. Depends what you mean by "complex"! Certainly the final paragraph of the VF article would indicate that the author did not do their research when it comes to the combined Airbus/Boeing overhaul of stall training and Airbus's move to increase manual handling training as part of the A350 programme. :)

@_Phoenix_ - Again, the "intersection" analogy is muddled. The THS movement happened as a *consequence* of the pilot's input, not in opposition. He never applied enough nose-down input to have a significant effect. The aircraft was complying with the commands given the whole time, not acting against them!

_Phoenix_

29th Sep 2014, 02:22

Yes, again... He never applied enough nose-down input to have a significant effect..

If you wish, zoom again (200%) in segment 2:12:30 to 2:12:45. The average input value sent to computer was clear ND. But the elevator doesn't rotate more than half NU and THS doesn't bother from max NU.
Doesn't matter the law, if PF demand ND rotation A/C should start rotate that direction not opposite, especially under stall warning. By the way, maybe you noticed that stall "wakes up" with each ND input 2:12:35 and 2:12:40.

See the video below at 25:08, how much time was required in this case?
https://www.youtube.com/watch?v=WAiA6ep95Sc#t=1508

Machinbird

29th Sep 2014, 03:37

The "pulsing" method of control is the one recommended by Airbus, and it has nothing to do with video games, it's simply that it provides a method for observing the effect of the input and thus moderating the next input (or indeed figuring out whether further input is necessary or not).
Dozy, where did you find that information? I haven't seen that in print before.

RetiredF4

29th Sep 2014, 07:30

_Phoenix_
Doesn't matter the law, if PF demand ND rotation A/C should start rotate that direction not opposite, especially under stall warning. By the way, maybe you noticed that stall "wakes up" with each ND input 2:12:35 and 2:12:40.

Just a reminder, with a nose down side stick input a ND rotation is intended by the PF, but a reduction of a former g-demand is executed by the fbw system. The aircraft was already falling with less than 1g, not having been able to execute the former demand of 1g (if stick neutral) or positive g (if stick aft). That overall small demand change moved the elevators a bit from the full nose up position, but was not enough to move the THS.

@ Dozy
How would those small intervals between the SS inputs give any feedback how the aircraft is responding to the input? Only the effect of the sum of multiple inputs would show in a reaction of the airframe, but not the single input itself.

If that pulsing is recommended by airbus (which i doubt until proof is shown) there should be an explanation available to this steering method we all would sure be interested in.

rudderrudderrat

29th Sep 2014, 08:53

If that pulsing is recommended by airbus (which i doubt until proof is shown) there should be an explanation available to this steering method we all would sure be interested in.

FCTM OP-020 Flight Controls
Normal Law.

"OPERATIONAL RECOMMENDATION:
From the moment the aircraft is stable and auto-trimmed, the PF needs to perform minor corrections on the sidestick, if the aircraft deviates from its intended flight path.
The PF should not force the sidestick, or overcontrol it. If the PF suspects an overcontrol, they should release the sidestick.
AT TAKEOFF AND LANDING
The above-mentioned pitch law is not the most appropriate for takeoff and flare, because the stable flight path is not what the PF naturally expects.
Therefore, the computers automatically adapt the control laws to the flight phases:
GROUND LAW: The control law is direct law
FLARE LAW: The control law is a pitch demand law.
OPERATIONAL RECOMMENDATION:
Takeoff and landing maneuvers are naturally achieved. For example, a flare requires the PF to apply permanent aft pressure on the sidestick, in order to achieve a progressive flare. Whereas, derotation consists of smoothly flying the nose gear down, by applying slight aft pressure on the sidestick."

Alternate Law.
"Alternate law characteristics (usually triggered in case of a dual failure):
In pitch: same as in normal law with FLARE in DIRECT
In roll: Roll DIRECT
Most protections are lost, except Load factor protection.
At the flight envelope limit, the aircraft is not protected, i.e.:
In high speed, natural aircraft static stability is restored with an overspeed warning
In low speed (at a speed threshold that is below VLS), the automatic pitch trim stops and natural longitudinal static stability is restored, with a stall warning at 1.03 VS1G.
In certain failure cases, such as the loss of VS1G computation or the loss of two ADRs, the longitudinal static stability cannot be restored at low speed. In the case of a loss of three ADRs, it cannot be restored at high speed.
In alternate law, VMO setting is reduced to 320 kt, and α FLOOR is inhibited."

In Normal Law, most pilots use the "pulsing" method i.e. they release the side stick between attitude changes.

In ALT Law, FCTM says roll is DIRECT but makes no recommendation on how to handle it.
Bonin seems to have found the difference in handling qualities to be significant.

PJ2

29th Sep 2014, 14:52

A "pulsing" characterization of stick movement makes sense when one does it...the notion is "small movement, neutralize-wait...small movement, neutralize-wait...", rather than ordering pitch and roll almost randomly or at least in a series of opposing orders.

When there is stick movement, there are roll-rate and 'g' orders being sent. No movement, no orders - the airplane remains at its last "set" attitude. When change of either is desired, move the stick, neutralize it, - wait.

In a sense the ideal motion with the stick is what we do with the control column & wheel. We don't see a lot of push-pull on the column, nor left-right on the wheel, we don't "stir the column", (nor would we tolerate it!); - the movements are "squeezed" on the column and so should they be on the stick.

I wouldn't have used the term "pulsing" as that could imply a jabbing motion and it's not - it's a "squeeze", just like moving the thrust levers/throttles.

Sidestick movement:
Example of roll angles & rates, autoflight engaged then disengaged at about 300ft, or where the yellow line is no longer flat.

Touchdown is where the black line is more or less flat. Crosswind was 25kts+ from the left.

http://www.smugmug.com/photos/i-ZGh2XRC/0/L/i-ZGh2XRC-L.jpg

CONF iture

30th Sep 2014, 02:09

To me it is a non-sequitur because the PFD certainly says sufficient to know what needs doing to correct whatever the other pilot may be doing incorrectly
That's how Airbus wants to justify to the world its technical choice, but AF447 again tells a different story.
Having a direct knowledge of the PF control inputs is a GOLDEN mine for a PM in his capacity to thoroughly analyze a situation, to be deprived of it is a waste of precious data.

jcjeant

30th Sep 2014, 02:56

Indeed .. when Robert ask Bonin to go down .. and bonin answer "I go down ..." but instead pull the stick .. we do not hear Robert say .. " if you go down why are you pulling on the stick ? "
And he don't ask because he see nothing of the stick

vilas

30th Sep 2014, 05:39

jcjeant
This is very interesting thread which keeps running but the question is did the pilots know that the aeroplane had stalled? If so how do you recover without looking at the PFD? conventional dual yoke has an advantage for the other pilot being able to follow on controls but that is not the cause of this accident. Lack of recent training for the failure leading to inappropriate control response and virtually zero scan by all including the captain on his return to cockpit is the cause. Even a blind folded pilot on hearing stall warning would push the stick forward here he keeps pulling back. It is more of shock and awe rather than any planned action. All things being equal this crew may have done the same thing even in B777. There is an incident in Airbus where pilot started both engines with fuel pumps off, ECAM warning was twice cancelled without noting the failure, take off and climb to FL 380 was made with pumps off. Dual engine flame out and successful restart and landing was done without noticing the pumps. So propensity to ignore indications direct or indirect is unlimited.

PJ2

30th Sep 2014, 08:13

CONF iture, jcjeant;

I have read and understand your arguments regarding the sidestick.

The sidestick solution is successful as millions of flight hours, thousands of crews quietly going about their work and the incident-accident rates all can attest. It is neither reasonable nor possible to point to one control solution over the other by citing the record in support of a personal opinion; the record is not there. Nor is it reasonable to cite one accident in support of a general theory. Even so, that door swings both ways.

What is reasonable is personal preferences, even using the same reasons; but a preference isn't an argument.

Regards.

Cool Guys

30th Sep 2014, 12:03

It has been clarified previously that statistically there is no evidence to prove whether the unconnected side sticks or connected yokes are better. Where the differences become apparent is under high workload situations where it is advantageous to know what the other pilot is doing. The sample base of these events is pretty small so we cannot make a judgement.

However there have been numerous pros here that have provided good reasoning why one is better than the other. Some have even provided a bit of scientific evidence to back up their beliefs. Granted, Airbus cannot take this information and change their cockpit layout based on a few peoples’ beliefs on an anonymous forum but on the other hand it can’t be discounted completely because there is no statistical evidence to prove otherwise.

Cool Guys

30th Sep 2014, 12:27

Regarding the above discussion about the 2 types of control inputs made on the stick by airbus pilots (1. Applying the appropriate pressure for the appropriate aircraft response 2. short pulsing movements). My interpretation of this is if a pilot has been using the “pulse” method in normal law and the system degrades to alt law he has to change his input method. This pilot may have been pulsing the side stick for thousands of hours and one dark stormy night he gets into alt law and along with handling all the issues that caused the law reversion he also has to change the way he controls the plane.

I really think the designers of this system could of been a bit kinder to the pilots.

roulishollandais

30th Sep 2014, 12:46

It i a design issue, but it is not a designer's issue,It is a regulators' issue. The regulators must do sure that training programs and airlines's operations comply with design . Designers are inventors, they must be free to design.

vilas

30th Sep 2014, 12:49

Cool Guys
In alternate law pitch control, handling remains similar to normal law except that it has no protections. Only in roll it is different where it behaves like other aircrafts with natural stability.

CONF iture

30th Sep 2014, 13:39

What is reasonable is personal preferences, even using the same reasons; but a preference isn't an argument.
Personal opinion or preference would not count much.
You can fly millions of hours and don't need to bother what kind of setup is in your cockpit, but then AF447 arrives where 3 crew members are disconnected if and how flight controls inputs are applied or not ... The Airbus setup is not the cause for the crash, but one of the contributory factor and had to be listed as such in the BEA report.

ManUtd1999

30th Sep 2014, 14:18

very much do not think the crew ever knew they had stalled.

I agree. I was reading the vanity fair article (excellent btw) and it was mentioned how the stall alarm cuts out at obscenely high AoA. This created a situation where Bonin would lower the nose and the stall alarm would kick in again, adding further confusion and loss of confidence in the instrumets I'm sure.

It's also an interesting inight into system design. It might work perfectly for all sensible flight regions, but there's always that 1 untested case where a fault in the logic appears.

jcjeant

30th Sep 2014, 15:22

vilas
This is very interesting thread which keeps running but the question is did the pilots know that the aeroplane had stalled?Apparently they don't know ... but ..
Before the stall occur ... Robert ask Bonin to descend (go down) and he answered "Ok I go down" and in the same time he was pulling on the stick
So if Bonin had follow the Robert advise ... no stall
Helas .. Robert don't see that Bonin was pulling instead of pushing ....
The first reactions of Robert were good .. asking to go down and level wings

noske

30th Sep 2014, 15:41

In some ways, this ingrained perception that pilots who made mistakes must have been in some way deficient was negated as far back as "Fate Is The Hunter". The only pilot in the book that Gann explicitly refers to as being incompetent and having obtained his position by deception is eventually found out and loses his career, but is not killed as a result of his incompetence.
In that "Capt. Dudley" episode it is also interesting to see what Capt. Gann does when he has to realize that this fellow pilot, one that he fully trusted and even kind of admired, is totally mishandling their aircraft. (Thanks for making me pick up that book again. :ok: )

vilas

30th Sep 2014, 15:46

jcjeant
They were in instrument conditions. Airbus or Boeing what does one look at? Going down or up looking at what? the side stick? Airbus pilot looks at N1 or EPR gauge for thrust not at throttle movement similarly the pitch change is to be noticed on the PFD not on side stick. These guys are talking and doing things but from time they lost auto pilot no one has looked at the PFD till they crashed. It cannot be put down to anything else.

PJ2

30th Sep 2014, 15:48

CONF iture;

Thank you for your response. As you know though the data is thin, we have more examples of control columns in the full-aft position than we do sidesticks when loss-of-control or a full-stall has occurred.

It is true prima facie that when you can't observe something you can't make a judgement of it but there are numerous clues to the condition of the aircraft still present. There are only two examples in which these clues did not help the crew - the Perpignan crew, because the AoA sensors were frozen and the crew would not have been able to diagnose the problem to respond correctly, and this one which had the descent rate indication with the nose-up and full power.

If the stick is "contributory" then the control column logically (for the argument to work), must not be and this is not the case.

Cool Guys, re, "I really think the designers of this system could of been a bit kinder to the pilots."

Well, in what way?

I can only think of one, and it occurred to me the first time I flew the A320. IIRC correctly, during the takeoff roll the "iron cross" symbol was in view and one could see what the other pilot's stick was doing. As a "beginner" on Airbus I thought it was an excellent solution to the relative lack of visibility. But then the symbol disappeared just after rotation and I honestly thought, "whaa...?", because, I thought, it was so useful to see what the other guy's stick was doing. And it never came back during any phase of flight, and I thought it should any time the autopilot was disengaged, particularly on final.

But as I flew the Airbus types, "knowing" became second-nature and the position of the sidestick not being a source of direct information, (primarily because the movements are so tiny that it is not possible to judge position relative to the stick's 'null' position), became irrelevant and the PFD and engine instruments became primary - subconsciously, as we do as pilots. The "sources" of aircraft state changed from what I knew in the Boeings/Lockheeds/Douglas's, and though subtlely different, it worked as well in all the circumstances I encountered. I suspect you'd get roughly the same notions from any Airbus pilot.

Though there are almost two dozen examples of crews stalling their transport aircraft over the past dozen years or so, I doubt if we'll see an accident like this one again. But we'll see others for reasons discussed and well understood.

jcjeant

30th Sep 2014, 18:52

vilas
These guys are talking and doing things but from time they lost auto pilot no one has looked at the PFD till they crashedRobert certainly (on the beginning of the sequence "Bonin climb episode" ) looked at the PFD and this is why he asked Bonin to go down and level wings
Bonin responded OK to the requests but performed one only (by also looking at the PFD)

jcjeant

30th Sep 2014, 18:59

Thank you for your response. As you know though the data is thin, we have more examples of control columns in the full-aft position than we do sidesticks when loss-of-control or a full-stall has occurred. Indeed when the two pilots are on the same frequency (or in the same analysis error) control colums in the full-aft position will occur
In the AF447 case .. it's clear that Bonin and Robert are not on the same frequency .. on the beginning of the event
Bonin want to climb and Robert want to descend
If it was control columns ... who whon ?

PJ2

30th Sep 2014, 20:08

jcjeant;

You know as well as I do that the sidestick takeover buttons are for that very purpose - when you need to either take control momentarily, (to avoid dual sidestick inputs), or completely locking-out the sidestick on the other side due to a more serious problem, (pilot incapacitation, for example).

In the control column case IIRC you can break the elevator control-cable "channels" to separate control columns in a B767, (Egyptair case) and in the L1011. Other than that, (DC8, DC9 case), the "winner" will be he who, for whatever reason in the moment, is the physically stronger.

I'd take the sidestick lockout solution over the other solutions.

jcjeant

30th Sep 2014, 20:17

You know as well as I do that the sidestick takeover buttons are for that very purpose - when you need to either take control momentarily, (to avoid dual sidestick inputs), or stop inputs from the other side due to a more serious problem, (pilot incapacitation, for example). On Airbus .. you don't check the stick position of the pilot at your right or left .. you check simply the PFD and you know all .. I'm right ?
OK ...
Now Robert ask to descend (as good Airbus pilot he check the PFD) and Bonin answer "OK I descend"
Robert continue "as good Airbus pilot" to check the PFD and see the climb continue ...
Or Bonin just answered "I descend" ..
So now "as good Airbus pilot" Robert can think .. maybe that is something wrong with instrument ... the doubt can take root .. the catasrophe sequence can unfold .....
What if Robert had seen the Bonin stick position ?
This is a infinite debate :)

PJ2

30th Sep 2014, 20:37

If you parse the argument thus, yes, it is atomized and therefore infinite in variations of "reading PFDs, knowing sidesticks or columns, listening to the airplane, etc., etc.

But if you "gestalt" what went on over a period of about two minutes, what teases out of the background noise is the clear potential for assessing what had happened and what was happening.

One indeed can see by referring to the PFDs alone, that that is unquestionably the case. Put another way, one can rescue the airplane by reference to the PFDs.

The infinite quest is not the what or how, it is the why.

john_tullamarine

30th Sep 2014, 21:59

It might work perfectly for all sensible flight regions, but there's always that 1 untested case where a fault in the logic appears

Which is why it is so critical that original and modification design having any direct impact on piloting must be assessed rigorously by experienced flight test folks. I am sure that this occurred throughout at Airbus .. just tragic that this outlier caught everyone out ...

But if you "gestalt" what went on over a period of about two minutes, what teases out of the background noise is the clear potential for assessing what had happened and what was happening.

One of the underlying problems - here and in general - is mental overload.

I'm sure we have all experienced the sort of problem.

I recall one such very clearly. An experienced 727 crew, but quite a few years' worth of being rusty on the Type, requalifying on the bird over at DFW. First sim session, engine failure, F/E missed this and that, with the result that I ended up with hydraulics out as well. Hadn't had any flying at all for the previous couple of years and was working well and truly above my comfort zone level. I quite clearly can remember seeing the bank continue to increase and fixating on control input to correct. The other driver, being relatively relaxed in the other seat .. put two and two together quickly and pulled back the other pod thottle, giving us time to sort out the rear panel oversight. Very sobering experience and resulted in some soul searching that night over a beer or ten.

Regardless of what we may think of AF447 overall, one really needs to maintain a degree of empathy for what must have been an extremely confusing and confronting situation for the two guys in the hot seats. Just a pity one of them wasn't sitting back relaxed and able to think a little about what might have been going on at the time .. rather than both getting too involved.

Philosophically, from a command management and SOP point of view, not all that different, for example, to the Eastern trimotor Everglades mishap of many years ago ? Trivial problem, breakdown in cockpit discipline, and the bird is lost ..

It's very easy for us greybeards to opine that we wouldn't have been caught out on the day ... but, change the circumstances a little, here and there, and the situation might have been an eye-opener for us in a similar manner .. ?

PJ2

30th Sep 2014, 22:19

Hi John;

Yes, mental overload is certainly an issue - the sim sessions usually reserved a few of those moments for everyone and the Airbus was good at it...!

In fact I was thinking of the examination of the overall, known picture to emphasize that situational awareness can be derived from the PFD even as one may attempt to make the argument about not "seeing the other stick" prevented comprehension, etc. But I agree here, that mental overload quickly became an issue and focus narrowed, (I doubt if anyone heard the stall warning.)

Regarding empathy with these guys, thank you for the reminder - it is indeed easy to type words on a keyboard...

DozyWannabe

30th Sep 2014, 22:46

Regarding empathy with these guys, thank you for the reminder - it is indeed easy to type words on a keyboard...
I should point out that when I make statements along the lines of "should have done this" or "maybe this would have helped", I'm not saying the buck stopped with the crew - far from it. There was supposed to be a whole system that included the builders, the regulators and the airline who should have had their backs - and in this case the approach to the situation could at best be described as "disjointed".

I repeat the point I made a few posts ago - namely that the concept of "blame" is not a useful one when it comes to discussing airline safety.

[PS. PJ2 - I think I'm going to save/bookmark your posts #592 and #595 whenever that subject comes up again - it says everything I've ever said in a much more concise and clear way, and I can only hope that because it comes from a respected pro, folk will be more inclined to listen to you than to me! :ok: ]

john_tullamarine

30th Sep 2014, 22:58

Don, I had no doubt that you would be on side with my comments.

However, and I am not immune to the personal deficit either, some of our PPRuNe colleagues at times in discussions of this mishap .. do appear to forget the human fragility aspects.

That more rigorous training, exposure and experience may have changed the outcome is a quite separate topic ..

For those who haven't read much on the general aspects this training paper (http://www.casa.gov.au/wcmswr/_assets/main/lib100215/hf-engineers-res-ch3.pdf), while targeting maintenance folk, is a useful plain language telling of the story and may be a useful read.

As Dozy observes, we are all around the fireplace with each having a role to play in minimising the chances that any might get burned ..

PJ2

30th Sep 2014, 23:46

Hi John;

The career always reminded me that I wasn't immune either. We're all smart until we're not. There were a number of times when after reflecting I thought it could have gone 'the other way' and I know I'm not alone.

G0ULI

30th Sep 2014, 23:53

John,
All the examples in the training paper had a common factor, the maintenance was interrupted, carried out in a hurry or stages were skipped in order to get the plane out on time. The root of nearly all commercial aviation accidents is money. Commercial pressures to keep aircraft available, pilots abbreviating pre flight checks to get in the air on time, overloading or incorrect weight and balance checks before a hurried departure. Training programmes have been cut back or compressed and maintenance directives deferred for as long as possible in order to maximise profits for the airline and its' shareholders.

Airbus arrived at their cockit design through wanting to save as much weight as possible. Less weight in the aircraft fixtures and fittings means more passenger and cargo revenue and lower fuel costs. Side stick controls meant the cockpit could be made smaller allowing a bit more room for the passenger cabin and an extra row of seats. Not a bad thing in itself, but it can have a negative effect on ergonomics, as has been theorised in this thread.

Maximising the return on investment is all that matters to airlines these days, aircraft and crew are just resources to be used to the fullest extent possible without actually losing an aircraft through fatigue, crew or mechanical. You can be sure that the company analysts at all of the big airlines have factored in the acceptability of losing an aircraft or two against profits and insurance costs.

I recall reading about the existance of a confidential memo many years ago which indicated that paying for the loss of three aircraft during their operational life would be cheaper than delaying further production for modifications and doing a recall to fix a design fault that had been discovered. This mentality is by no means confined to the aircraft industry. Every day products are released to market with known deficiencies. The manufacturers find it more profitable to just replace faulty goods rather than redesign the product.

Even when human lives are at stake, it often takes an extraordinary amount of campaigning to to get the fault fixed. How many vehicle manufacturers have maintained that their products were perfectly safe before issuing a recall notice to all owners? It all comes down to money.

roulishollandais

1st Oct 2014, 00:18

John :D:D:D

DozyWannabe

1st Oct 2014, 00:41

Airbus arrived at their cockit design through wanting to save as much weight as possible. Less weight in the aircraft fixtures and fittings means more passenger and cargo revenue and lower fuel costs. Side stick controls meant the cockpit could be made smaller allowing a bit more room for the passenger cabin and an extra row of seats.
With respect, I beg to differ.

Sure, weight and complexity savings were a factor and it would be daft to deny it. But such things were in fact very much secondary factors. I've touched on this before so I'll try to keep it brief, but the Airbus FBW flight deck is designed as it is for a primary specific purpose, and that is to have an unprecedented level of flight deck commonality across the fleet, regardless of whether the type is a short-haul narrowbody or a long-haul widebody.

In the early '80s, Airbus was still something of a newcomer and they used the fact that FBW technology was tried-and-tested by the military and space programmes of the '70s to offer something no other manufacturer could. In effect, they turned their status as a newcomer with only two proven legacy types into an advantage. The fact that the British and French aero research industry had collaborated on the Concorde "minimanche" project (from which the Airbus sidesticks were developed) also influenced the decision.

Another factor leading to the "clean slate" thinking regarding the flight deck was that the traditional linked yoke design was an outgrowth of the days when the flight surfaces were directly linked to the light surfaces by cables. That one crew member could see/feel what the other was doing was actually a side-effect. The *real* reason they were linked was because in a situation where one person's muscle force was insufficient to work against the aerodynamic pressure against the flight surface, it was possible to combine the muscle force of two to do so. In real terms, there was no significant need for "follow through" outside of a trainer.

With the advent of large widebody types came all-hydraulic control systems (i.e. there was no manual reversion) - simply because the flight surfaces (and the forces acting against them) were so large that muscle power alone would be insufficient to move them. The last western narrowbody jetliners to have manual reversion were the B737 and the DC-9, and I've read plenty of posts on here from airline veterans which made clear that even airliners of that size were very tough to handle in the manual reversion scenario. So much so that Boeing's intended replacement for the B727 (i.e. the B757), despite being smaller than the widebodies, had no manual reversion at all (i.e. it was all-hydraulic).

Once you remove the "extra muscle power" requirement that goes with direct cable connection, then you effectively remove a significant reason (arguably the most significant - as this is an airliner, not a trainer) for the controls to be connected.

Obviously, back in 1988 there was a degree of consternation from the piloting community as to how this evolution would pan out - after all on paper it looked like a significant change from what had gone before. But I'd say that a safety record that holds its own now we're at 26 years and counting of this technology being on the line suggests that the design has proven itself. Undoubtedly there are pros and cons in an ergonomical sense, but that comes with the territory.

As PJ2 has kindly pointed out, the fact is that there are currently more accidents involving aircraft with traditional controls where the PNF has not picked up on what the PF was doing than there have been with the passive sidesticks.

The scuttlebutt that has Airbus working in league with the beancounters to make pilots' lives harder is as pernicious as it is false.