PPRuNe Forums

PPRuNe Forums (https://www.pprune.org/)
-   Tech Log (https://www.pprune.org/tech-log-15/)
-   -   AF 447 Thread No. 9 (https://www.pprune.org/tech-log/489774-af-447-thread-no-9-a.html)

john_tullamarine 5th Jul 2012 10:07

AF 447 Thread No. 9
 
AF 447 Thread No. 9

Thread part -

(a) #1 starts here and finishes here. Posts = 3890
(b) #2 starts here and finishes here. Posts = 2537
(c) #3 starts here and finishes here. Posts = 2071
(d) #4 starts here and finishes here. Posts = 1061
(e) #5 starts here and finishes here. Posts = 1978
(f) #6 starts here and finishes here. Posts = 1683
(g) #7 starts here and finishes here. Posts = 1354
(h) #8 starts here and finishes here. Posts = 1550


Links to the various BEA reports are given below. If I have missed any of the useful papers, please PM me with the URL and I can include it.

(a) BEA site - French, English
- Report link page - French, English

(b) Interim Report (No, 1) Jul 2, 2009 - English

(c) Interim Report No. 2 Dec 17, 2009 - English
- Update Dec 17, 2009 - French, English

(d) Estimating the wreckage location Jun 30, 2010

(e) Wreckage search analysis Jan 20, 2011

(f) Briefing and associated update May 27, 2011
- Briefing - update French
- Briefing - update English
- Briefing - update German
- Briefing - update Portugese

(g) Interim Report No. 3 July 2011 - French, English

Miscellaneous pertinent links -

(a) Airbus Operations Golden Rules
(b) ALPA FBW Primer
(c) C* and Civil Transports - Cranfield
(d) Longitudinal Flight Control Design - RAeS
(e) Longitudinal Stability: Effect of High Altitude and CG - Boeing
(f) pitot static system performance - USN (Pax River) FTM
(g) The Problem of Automation: Inappropriate Feedback and Interaction, Not Over-Automation. Donald A. Norman UCSD
(h) Upset Recovery - 16MB zip file
(i) Ironies of Automation. Lisanne Bainbridge UCL
(j) Cognitive Capability of Humans. Christopher Wickens Uni Illinois
(k) Trust in Automation: Designing for Appropriate Reliance John D. Lee, Katrina A. See; Human Factors, Vol. 46, 2004
(l) Training for New Technology. John Bent - Cathay Neil Krey's CRM site


Search hint: You can search PPRuNe threads with a filter in Google by using the following search string example -

ths af447 site:http://www.pprune.org/tech-log/

This will search for mentions of THS in the AF447 threads of tech log only.

Just change the THS in the string to whatever you want to look for. This allows one to search for any term or phrase of interest throughout the threads.

Adding the site:URL end part is the magic that restricts Google to only searching in Tech Log.

This filter technique is absolutely wonderful and can be used generically to find things of interest in PPRuNe - appears to work OK in the PPRuNe search function as well.

In respect of Google searching, JenCluse has added some suggestions -

a) indenting the text block with a (one) Tab, *and*

b) emphasizing the fact that it is a search text block with some manner of . . .
<SearchText>, or
"SearchText", or
?SearchText?, etc

4dogs 5th Jul 2012 10:53

Report Release
 
Does anyone have a time (UTC) for the scheduled release of the report?

john_tullamarine 5th Jul 2012 11:27

The BEA site suggests early afternoon.

However, I've had a long week so it's off to bed for me, I'm afraid.

jcjeant 5th Jul 2012 11:29

Hi,

Keep monitoring this page :)
FLIGHT AF 447

A media presentation will take place at the Musée de l'Air et de l'Espace in Le Bourget at 2.30 pm
So it's in one hour from now (local time)
Dunno if the report will be posted same time

Machinbird 5th Jul 2012 11:32

Quote:
Originally Posted by Machinbird
BEA has not yet provided spoiler and aileron position data, so analysis of the causal aspect is not yet possible.

Carried over from Thread No. 8

Originally Posted by Clandestino
Well, if it doesn't make it to the final report, you can always use page 112 of interim 3: lateral DFDR parameters. Ailerons and spoilers staunchly followed the sidesticks, their periods of fluttering like mad were occasionally interrupted by full LWD deflection with not much apparent effect, as expected in stall.

Clandestino, look at the data you suggested and blow it up until you can see the shape of the curves. You cannot using that data. Not enough pixels to see anything.
Just a bunch of blocks. I do not call that disclosure of the data.

pcoletti 5th Jul 2012 12:38

Time of media briefing is 2.30pm French time
 
Which is 12.30 GMT

Clandestino 5th Jul 2012 12:38

Due roughness, one can't use graphs to derive control rates but positions can clearly be seen.

Would you please extend your roll/roll rate graph a minute into the future? Graphs on pages 30 and 31 are smooth & high-res, you can't object to that.

Flying Wild 5th Jul 2012 12:44

briefing live with on the fly translation on the BBC: BBC News - Air France 2009 Paris-Rio crash final report released

fireblade2000uk 5th Jul 2012 12:47

The report is "expected to say" so nothing new as of yet!

frac 5th Jul 2012 12:58

Other Live Stream.

Live BFMTV

4dogs 5th Jul 2012 13:02

I had a boy's look!
 
Thanks JT,

I did have a quick look at that page but dismissed the Press Release as being from 2009...duh! :O

DozyWannabe 5th Jul 2012 13:03

So in terms of changes to aircraft systems:

FDs need to be changed in terms of their appearance in marginal systems behaviour
Stall Warning needs to be overhauled including dedicated visual channel and to operate in cases of low speed
Feedback in terms of systems failures (presumably ECAM) needs to be clarified and improved

Some certification changes will also be recommended.

So - any claims of a whitewash putting the finger solely on pilot error from this point onwards are clearly false.

[UPDATE - Claims final report will include computed FD commands, which should please some who wanted that information...]

roulishollandais 5th Jul 2012 13:09

GMT is UT-12h
 
Hello pcoletti,
... Just to be accurate :
14h30 local time in France (metropole=Paris) is
12h30 UT and also
00h30 GMT (reference for astronomy : the day starts at noon the sun passing the Greenwich meridian at 00h00 true solar time # 00h04m40s GMT 5.july2012) . (International convention)

... I have noting against UK !

roulishollandais (enjoying with sundials...):p

Heathrow Harry 5th Jul 2012 13:17

from Beeb

Technical failure and human error led to the loss of an Air France flight over the Atlantic in June 2009 and the deaths of 228 people, according to the final report into the crash.

The report by France's aviation authority, the BEA, says the crew did not react correctly after the Airbus A330 had technical failures.

It disappeared from radars while en route to Paris from Rio de Janeiro.

It took nearly two years for the flight recorders to be found.

Thursday's findings are in line with a preliminary report released last year by the BEA.

That report said the pilots did not follow the proper procedures after the aircraft's speed sensors - called Pitot tubes - failed during a storm two hours after take-off.

The plane stalled and lost altitude, the report said, and the captain did not retake control of the plane after taking a rest.

One of the mistakes of the crew, according to investigators, was to point the nose of the aircraft upwards, after it stalled, instead of down.
Manslaughter claims
Investigators have found fault with both Airbus and Air France, sparking a row between the two firms over their accountability for the crash.

Both companies are under investigation by French magistrates for alleged manslaughter.

A separate judicial report will be released next week. This is also expected to echo Thursday's report by the BEA, the French news agency AFP says.

Since the crash, Air France has replaced the speed sensors on its fleet of Airbus jets - made by the French firm Thales - with a newer model.

The wreckage of the plane was discovered after a long search of 10,000 sq km (3,860 sq miles) of sea floor.

After 23 months of searching, robot submarines finally found the flight "black box" recorders last year.

Peter H 5th Jul 2012 13:21

GMT
 
roulishollandais
... Just to be accurate :
14h30 local time in France (metropole=Paris) is
12h30 UT and also
00h30 GMT

Rather a surprise to us brits, who are told that they use GMT half the year.

More specifically, from wikipedia
Historically the term GMT has been used with two different conventions, sometimes numbering hours starting at midnight
and sometimes starting at noon. The more specific terms UT and UTC do not share this ambiguity, always referring to
midnight as zero hours. Astronomers preferred the latter GMT convention in order to simplify their observational data
so that each entire night was logged under a single calendar date.

konradeck 5th Jul 2012 13:49

Good update on Simon's site: Crash: Air France A332 over Atlantic on Jun 1st 2009, aircraft entered high altitude stall and impacted ocean

New thread compared to previous interim reports, which partially explains crew's behavior is that Flight Director was constantly suggesting the crew to raise aircraft's nose.

pcoletti 5th Jul 2012 14:28

Final report is...
 
... here:

http://www.bea.aero/docspa/2009/f-cp...p090601.en.pdf

noske 5th Jul 2012 14:35

Aviation Herald already has a summary of the presentation and the Q&A session. Most remarkable to me:
  • The flight directors provided indications to increase the nose up attitude.
  • During the final 4 minutes of flight the flight directors disappeared and reappeared multiple times.
  • The flight directors were NOT disengaged as they SHOULD have been.
  • The modus of flight director operation is being further looked into.

alph2z 5th Jul 2012 14:37

Final report (26 MB, Not finished downloading)

http://www.bea.aero/docspa/2009/f-cp...p090601.en.pdf

Web page for download mirrors and french
Rapport final

Enjoy ....
.

DozyWannabe 5th Jul 2012 14:49


Originally Posted by noske (Post 7279331)
The flight directors provided indications to increase the nose up attitude.

Not quite accurate. The presentation stated that the FDs indicated "slight nose-up" (presumably 5 degrees) at the point of Stall Warning activation. If they were indeed indicating 5 degrees then the FDs would have been ordering an initial small nose-up command followed by maintaining the 5 degree attitude. What was in fact commanded by the PF was a nose-up to almost 15 degrees within five seconds - way in excess of what the FD was displaying.

jcjeant 5th Jul 2012 14:51

Hi,

Orangefreak
It's the final report !
Date publication : July (Juillet) 2012
And a lot of annex
Going for a long reading now :)

llagonne66 5th Jul 2012 15:01

Some serious reading to do ... before commenting !
 
Including the appendixes it goes up to 43.5 megabytes ... so anybody posting a comment, let's say in the next 6 hours, won't be very serious ...

PJ2 5th Jul 2012 15:53


A procedural question from a non-pilot.

I assume that a stall warning requires a prompt response, while an event such as UAS requires a more considered response [and resort to a checklist].

If so, shouldn't the initial response to a stall warning be a memory item? With any caveats about ignoring stall warnings being part of that memory item.
From Day One of flight training, recognition of the stall in the airplane one is training in is a high priority item. The "Approach to the Stall" is a common exercise when transitioning to new aircraft types. The "approach to stall" is not the actual stall, either in the simulator or the aircraft as the case may be. No one stalls airplanes in transition or type-rating courses. It is the recognition of the impending stall that is taught.

Regulatory bodies may or may not require that the approach-to-stall exercise be part of recurrent training on one's current aircraft. In Canada, current regulations do not require the approach to the stall exercise in fly-by-wire aircraft after the initial type-rating is complete. One may go years without seeing the exercise again unless it is informally inserted into the sim exercise by a keen instructor. Usually sim time is so compressed with required exercises that such things as manual flying and so on aren't usually done.

On the memorized aspect of the stall response, you're right - the stall warning requires an immediate response and yes, the aircraft-specific response would be memorized - but it's memorized the same sense as one might "memorize", say, the skid-response in a car...it's not just "memorized mechanically" but should be "visceral" - ie, instant and automatic.

Two reasons why this may not be the current state of affairs are, 1) aircraft today are so reliable and so well-designed that getting oneself into such a corner is an extremely rare event, and 2) because of such reliability, the "approach to the stall" exercise has been de-emphasized (and not just for Airbus aircraft) so pilots today who may not have exerienced anything but a high degree of automation may not, (emphasis here) have the same appreciation of the danger of the stall as the previous group who grew up on cable-and-pulley aircraft that could and did stall. A sense that one's airplane "cannot stall" because of automated protections is completely foreign to me and to an entire generation of guys who are now retired/retiring but I wonder in what way the stall is currently perceived as a serious and ever-present threat if one loses airspeed? There have been at least a dozen accidents in the last eight or so years as a result of stalling one's transport aircraft. That's unusual and I think a trend.

The avoidance of the stall would be/should be innate as once again, avoiding a stall is absolutely fundamental to flying any airplane. Type-specific responses to the stall warning, (which indicates an approach to stall, and not the fact that the airplane is stalled) would be part of the FCOM for the aircraft. Typically, such responses used to focus on a minimum loss of altitude first, while "powering out" of the condition. At high altitude, the guidance was that the pitch attitude may have to be slightly below the horizon for recovery. TOGA (take off/go around) thrust was always applied as part of the memorized drill.

Since AF447, this response has been re-visited and, I think, far more intelligently designed due to the recognition that the stall is not an airspeed problem, it is an AoA [angle of attack] problem, and the first reaction now is to reduce the wing's AoA to unstall the wing, then recover with power. Increasing thrust with under-slung engines tends to pitch the aircraft up, countering the required response, so adding thrust is a secondary response now. These responses will be memorized and, I suspect, practised far more frequently now.

The sections of the FCOM I cited in the previous post would be those seen on the ECAM and would be accomplished after all memory items were completed by the crew.

The standard response has always been to respect the stall warning, period. The one caution could be confusing to someone who hadn't received thorough training but the two should never be confused to the extent that doubt is produced as to what to do if a stall warning occurs.

I hope this helps.

sensor_validation 5th Jul 2012 16:06


Originally Posted by llagonne66 (Post 7279386)
Including the appendixes it goes up to 43.5 megabytes ... so anybody posting a comment, let's say in the next 6 hours, won't be very serious ...

It's a very well written report, only typo I noticed in the English version was on bottom of P151 "mark range" should be "Mach range".

Sad, so sad.

syseng68k 5th Jul 2012 16:58

Have been lurking for some time, waiting for the final report,
but one thing that stood out for me on initial scan was the
paragraph on the horizontal stabiliser, p218, 5.3.1.2, where
the concept of "pitch runaway" is discussed.

I'm sure that this must have been discussed at length elsewhere
in the topic, but would have assumed that any sensible system
design would inhibit ap engagement, or any autoflight function,
unless airspeeds are known good. Otherwise, feedback loop
with no null possible...

Fox3WheresMyBanana 5th Jul 2012 17:11

Quoth page 185 of the report

Current training practices do not fill the gap left by the non-existence of manual flying at high altitude, or the lack of experience on conventional aeroplanes. Furthermore, they limit the pilots’ abilities to acquire or maintain basic airmanship skills.

In short, training's too short.

..and do Air France's crew training changes (p215) do anything about this? No

RR_NDB 5th Jul 2012 17:29

Otherwise, feedback loop with no null possible...
 
A case for GIGO, "as per design"?

Clandestino 5th Jul 2012 17:30


Originally Posted by syseng68k
but one thing that stood out for me on initial scan was the
paragraph on the horizontal stabiliser, p218, 5.3.1.2, where
the concept of "pitch runaway" is discussed.

Errr... discussed was the "pitch runaway" that could result from engaging the autopilot that would follow the false air data, not "stabilizer runaway".


Originally Posted by syseng68k
I'm sure that this must have been discussed at length elsewhere
in the topic, but would have assumed that any sensible system
design would inhibit ap engagement, or any autoflight function,
unless airspeeds are known good.

Problem is that it is easy to use computer to detect failed airspeed indicator but the system that can detect unreliable airspeed needs to be truly intelligent. No replacement for human pilot there is currently available.

ECAM_Actions 5th Jul 2012 17:40

A design flaw is the inhibition of the stall warning based on airspeed, instead of the weight-on-wheel switches. If the WoW switches are saying it is flying, then the aircraft must be flying, right??? Is this too obvious??

To invalidate the stall warning on airspeed alone is stupid.

There are three parameters:

* WoW switches (FLYING)
* Airspeed (STALL, though invalid)
* AoA (STALL)

3 out of 3 says it is stalled, even though the airspeed is invalid in this case.

If any of the other two parameters were not stalled:

* WoW switches (FLYING)
* Airspeed (STALL, though invalid)
* AoA (FLYING)

with the speed being unreliable (and the system knowing this), chances are it is flying! This of course ignores the fact there are MULTIPLE AoA vanes sensing a stall condition.

RR_NDB 5th Jul 2012 17:52

Simple engineering task. Problem is: Designers mut be ALWAYS, "truly intelligent."


No replacement for human pilot there is currently available.
And this unavailability is AN ABSURD! It is VERY EASY, REPEAT: VERY EASY to process the information in order to HELP THE CREW ALLOWING TO UNDERSTAND THE ISSUE, IMMEDIATELY! BEFORE ANY CONTAMINATION OF AVIONICS WITH GARBAGE!

To put (and maintain) in crew shoulders the responsibility to do the diagnosis (as Airbus SAS paper in UAS issue put) is IMO a GRAVE ERROR.

RR_NDB 5th Jul 2012 18:00

To invalidate the stall warning on airspeed alone is stupid.
 
K.I.S.S. or K.I.C.S.? design approach?

K.I.C.S. stands for: Keep It Complex Stupid

grity 5th Jul 2012 18:02

s.202: A crew can be faced with an unexpected situation leading to a momentary but profound loss of comprehension. If, in this case, the supposed capacity for initial mastery and then diagnosis is lost, the safety model is then in “common failure mode”.


what did this mean: "common failure mode" ???
is it accepted in the flying comunity to loose the control in this case?

aguadalte 5th Jul 2012 18:05

I was hopping to see at least one recommendation regarding PF SS inputs to be shown on PNF PFD in the sequence of this statement on the final report:

(...) It would also seem unlikely that the PNF could have determined the PF’s flight path stabilisation targets. It is worth noting that the inputs applied to a sidestick by one pilot cannot be observed easily by the other one and that the conditions of a nightflight in IMC make it more difficult to monitor aeroplane attitudes (pitch attitude in particular). (...)
I guess that would be an "evidence" of Airbus Industrie's misconception of the system...

jcjeant 5th Jul 2012 18:06

This is the reaction of Air France ..
It's sad that Air France stay on his position and have not understand the message convoyed by the BEA in the final report .... :*

Communiqués de presse

Google*Traduction


In its analysis and conclusions, the BEA said it's a chain and a combination of factors - technical and human - that led to the loss of the aircraft in just over four minutes. He confirmed that the crew was trained and qualified in accordance with the regulations and the aircraft systems have responded in accordance with their design and certification criteria.
The BEA report describes a crew that is based on information provided by the instruments and aircraft systems, and behavior of the aircraft as it was visible in the cockpit instrument readings, triggering alarms and shutdowns, noises aerodynamic vibrations of the device, etc.. The reading was made he was not permitted to apply the appropriate actions. In this environment degraded steering, crew, gathering the expertise of the captain and officers of the two pilots, remained engaged in the conduct of the flight until the last minute.
Air France pays tribute to their courage and determination in these extreme conditions.
And the reaction of Airbus:
AF447 Statement*| Airbus, a leading aircraft manufacturer


The publication of the BEA’s final report now provides the opportunity to further work on the lessons learned from this tragedy and measures to be applied to avoid the recurrence of such an accident. Without waiting for this final report, Airbus has already started working at industry level to further reinforce the robustness of pitot probes requirements and actively supports related activities.

Fox3WheresMyBanana 5th Jul 2012 18:49

I doubt that at any point in these pilots' training have they been placed in a situation where their corrective actions did not work, nor have they been trained to deal with it. I imagine this is what is meant by common failure mode.
Even after 2 minutes, if they'd started from square one, they could probably still have saved the aircraft.

This doesn't have to be as hard as my RAF Instructor who gave me the aircraft with a toppled horizon in a spin for an IF Unusual Attitude that was supposed to be full panel. But this, and other similar tricks (once is not enough), taught me to recognise when plan A wasn't working, and how to start again from scratch.
It would probably help in the civilian world if these things weren't graded either.

Clandestino 5th Jul 2012 18:51


Originally Posted by ECAM_actions
To invalidate the stall warning on airspeed alone is stupid.


Originally Posted by BEA, final report on AF447, page 44
If the CAS measurements for the three ADR are lower than 60 kt, the angle of attack values of the three ADR are invalid and the stall warning is then inoperative. This results from a logic stating that the airflow must be sufficient to ensure a valid measurement by the angle of attack sensors, especially to prevent spurious warnings.

Plugging WoW into stall warning computer would make it more complicated, adding one more point of possible failure. Is 60 kt really too low cut-out for widebody?


Originally Posted by RR NDB
And this unavailability is AN ABSURD! It is VERY EASY, (REPEAT: VERY EASY) to process the information in order to HELP THE CREW ALLOWING TO UNDERSTAND THE ISSUE, IMMEDIATELY!
BEFORE ANY CONTAMINATION OF AVIONICS WITH GARBAGE!

It is not easy in this universe where computers haven't moved from algorithms that can be broken down to simple IF...THEN; it is impossible. There are lines of code helping the FBW computers reject false air data if one computer doesn't tell the same story as the other two (which has shown its dark side at Perpignan) or there is such a rapid change in parameters that it's physically impossible but it takes intelligence to recognize if the combination of airspeed and normal acceleration is incompatible with sustained flight and yet aeroplane is flying, then there must be something wrong with the indications.


Originally Posted by RR NDB
To put (and maintain) in crew shoulders the responsibility to do the diagnosis (as Airbus SAS paper in UAS issue put) is IMO a GRAVE ERROR.

Every modern passenger transport aeroplane experiencing unreliable air data will throw a host of undue warnings. Warning computers are not intelligent and are not supposed to be, they just monitor parameters and cry out when pre-programed thresholds are exceeded. Have a look what happened to Aeroperu or Birgenair when air data become unreliable. They were 757s. Not made in Toulouse. It is all very nice for PPRuNers or investigation boards to ask for better UAS detection and resolution systems but without some creative genius coming with the idea of how they should be made, we won't get far.


Originally Posted by grity
is it accepted in the flying comunity to loose the control in this case?

Seemingly everything is possible with de-structured crew. Love that postmodernist terminology in the accident reports... not.


Originally Posted by aguadalte
I guess that would be an "evidence" of Airbus Industrie's misconception of the system...

Too bad this "evidence" did not make it into recommendations. Conspiracy? Recognition that non-interconnected sticks did indeed prevent CM1 from recognizing that all the mess was mainly CM2 induced is so low down the causality chain to be significant?

ECAM_Actions 5th Jul 2012 19:10


Is 60 kt really too low cut-out for widebody?
It would seem not slow enough. :ugh:

All the ADCs need to know is whether all the airspeeds are within a certin limit of each other. They were not, and the system alarmed/downgraded as it was supposed to.

Where the problem arose was in the fact that the crew were overwhelmed by the automation, instead of asking why the airspeed was indicating 60 kts and the stall warning intermittent.

Regardless of any computers, I hope people haven't already forgotten that the by now very low airspeed was showing as such on the PFD and ISIS?

Regardless of what people fly - SIT and THINK when an alert first appears. Do a sanity check - are basic flight parameters nominal for the phase of flight?

RR_NDB 5th Jul 2012 19:17

UAS IMMEDIATE DETECTION
 
It is POSSIBLE AND EASY! (Simple engineering solution)*

(*) Will brief on that important issue in Man machine interface Thread

I am an EE with Safety interests. A x B is not my agenda.


without some creative genius coming with the idea of how they should be made, we won't get far.
This is responsibility of R&D of A, B, E, etc.

Will brief on that important issue in Man machine interface Thread

jcjeant 5th Jul 2012 19:51

This is an amazing "pearl" in this report:

Page 206

4.2.1 Recommendations on Operations
Training for Manual Aircraft Handling
Examination of their last training records and check rides
made it clear that the copilots had not been trained for manual aeroplane handling
of approach to stall and stall recovery at high altitude.

They can't fly manually at high altitude .. :}
When you know that most of the flight is in high altitude .. this gives chilling (no pun intended)
If this is not a finger pointed to AF .. I don't know what is this ...

Bill Macgillivray 5th Jul 2012 20:27

Basics!
 
I suspect (and raise the anti-flack shields!), that a tremendous amount of basic flying skills are no longer required to be taught. Flying an aircraft safely has now moved to a stage where (generally) you are monitoring a computor/auto-pilot system and the basic skills do not appear "valid", we are into the "what is it doing now" syndrome. FATAL!! (I have only recently given up instructing and not just on basic aircraft!). Not all of the old skills should be forgotten! It could/has proved fatal!


All times are GMT. The time now is 10:28.


Copyright © 2024 MH Sub I, LLC dba Internet Brands. All rights reserved. Use of this site indicates your consent to the Terms of Use.