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

AF447 Thread No. 3

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

AF447 Thread No. 3

Old 29th May 2011, 06:14
  #521 (permalink)  
Join Date: Aug 2005
Location: Dublin
Posts: 3
Desitter :- 228 people went to the bottom because a pilot hearing a stall warning pulled the stick back. Given the pilot's actions. a Sopwith Camel in this event would have have had the same outcome, except the Sopwith wouldn't have been screaming "stall" in his ear.
alanp is offline  
Old 29th May 2011, 06:34
  #522 (permalink)  
Join Date: Oct 2004
Location: Floating around the planet
Posts: 383
Everybody talks now about the pilot`s performance , but it`s not economic interesting to talk about the 2 F/Os.

If instead of 2 F/Os they had 2 Capts and 1 F/o how it was years ago, probably they had more chances.

Aviation is a big more or less.
A-3TWENTY is offline  
Old 29th May 2011, 06:36
  #523 (permalink)  

the lunatic fringe
Join Date: May 2001
Location: Everywhere
Age: 63
Posts: 617
Given the pilot's actions. a Sopwith Camel in this event would have have had the same outcome, except the Sopwith wouldn't have been screaming "stall" in his ear.
Absolutely agree.

This accident looks more and more like a flying skills issue, rather than design or software issue. And by implication an Air France training problem.

It falls under the heading.. "Flight with unreliable airspeed."

In general terms if you suspect that you have "unreliable airspeed" as a pilot, you stick to "datums." Put the power at a sensible figure, and set the attitude at a sensible attitude for that phase of flight. That will keep you safe. Having settled that down, diagnose the problem.

Plane, Path, People. .... or similar.

What you do not do is plant the sidestick in the back stop and expect a good outcome.
L337 is offline  
Old 29th May 2011, 07:24
  #524 (permalink)  
Join Date: Jul 2009
Location: Not far from a big Lake
Age: 78
Posts: 1,458
Thoughts on the initial climb and later, the implications of deep stall.
  • Alt 2 is a roll direct law. This means that the pilot flying will have to control bank angle. Only the aircraft's natural stability will tend to level the wings. Any turbulence will impart a bank angle until corrected.
  1. If PF had to fly the wings to keep them level, could he have inadvertently pulled the stick back while making corrections, either through nervous tension or bad personal geometry?
  2. And could his instrument scan have been so rusty that he didn't realize he was inducing a climb?
  3. Was there anything unusual about the copilots' stature, or technique in gripping the stick? (Realizing we ourselves don't really know who was PF but some guesses can be made).
  • Although ACARS does not specifically state that the aircraft later switched into Direct Law, and BEA has not told us (yet), It appears that after a significant period of deep stall, the AOA and A/S on all 3 ADRs had to have been considered unreliable by the system. How can even one Prim stay operational in this situation? If all 3 Prims are not available, the aircraft must be in Direct Law. That would explain the pitch trim staying at 13 degrees after the aircraft actually achieved a deep stall.
Machinbird is offline  
Old 29th May 2011, 07:28
  #525 (permalink)  
Join Date: Nov 2007
Location: Munich
Posts: 3
Originally Posted by gums View Post

Good math, Cog. Maybe go further and calculate the AoA required for those parameters.

Then re-read my re-post about the Viper deep stall AoA and descent rate and indicated airspeed.
Good math, poor aeronautics :-( Groundspeed is the horizonal component of the (inertial) velocity vector, if the airplane is going straight up or down, groundspeed is zero. Vertical speed can be either airdata (baro pressure) or inertial, the differerence between the two would be vertical wind components (up/downdrafts).
fruitflyer is offline  
Old 29th May 2011, 08:00
  #526 (permalink)  
Join Date: Nov 2007
Location: Scotland
Posts: 21
Sound asleep? Smooth ride?

If the pilot resting had to be called to the flight deck by the PNF, does this infer that the plane wasn't doing anything abnormal, at that point, that would have alerted the sleeping pilot?
Bill G Kerr is offline  
Old 29th May 2011, 08:10
  #527 (permalink)  
Join Date: Mar 2006
Location: USA
Posts: 1,616
Many have asked why the PF continued to pull back as the plane descended since it clearly indicated a stall. Maybe the FO disregarded the Altitude/VS data as well. Consider this:

He almost certainly would have disregarded any airspeed indications, so he's already thinking there's a pitot problem. Next thing he knows, the VSI varies between +7000 and -10,000. He's never seen those values before. Could he have assumed that the entire pitot-static system had gone bad?

The poor fellow now resorts to his last instrument- the attitude indicator. When he pulls back, the stall warning goes off, release the back pressure, and it returns. Catch 22 in his mind. Even if he knew that the stall warning would disappear below 60kt, if he had any idea that his airspeed was indeed below 60, he would probably have reacted more appropriately.

They probably put it all together when the RA started functioning and gave a "TOO LOW TERRAIN" alert, but at 10,000fpm, 15 seconds wasn't nearly enough time.

It must truly have been horrific in the cockpit. Flying in IMC, and none of your instruments or control inputs making any sense. RIP.
Check Airman is offline  
Old 29th May 2011, 08:26
  #528 (permalink)  
Join Date: Sep 2010
Location: by the seaside
Age: 70
Posts: 885
Bearfoils earlier post

What would explain the need for constant roll left, to maintain SL?

damage? AIR ? cg? Likewise NU? cg? ICE? All we see is the response, not the stimulus.

Whilst I can visualize the stall I cannot understand the aileron inputs.

The possible causes are updraft/turbulence - very unlikely at altitude in my experience.

Dutch roll - wrong duration of control input.

Asymmetric ice accretion ??? doubt it but..

flight control fault - side stick - computer or hydraulics or roll damper.

If the published data is correct and the aircraft wasn't in a stalled condition then 30 secs of full sidestick in a non protected aircaft should see it upside down.

My money is there is something missing which confused the pilots even more.....

Forgive my ignorance if alternate law will stop this happening - I didn't fly the bus although I had the chance.
blind pew is offline  
Old 29th May 2011, 08:34
  #529 (permalink)  
Join Date: Oct 2009
Location: UK
Posts: 1,271
Abnormal Attitude Law

Hi Check Airman,

I tend to agree. In just 46 seconds (between 2h 10 min 05 and 2 h 10 min 51) the aircraft climbed with ROC up to 7,000 ft/min from 35,000ft to “37,500 ft and the recorded angle of attack was around 4 degrees.” (which is a sensible angle of attack).

They must have ridden in the updraught of a CB to climb that rapidly and still have flying speed. Subsequently, they encounter the downdraught, their instruments don't make sense, stall warnings come and go, and I think they have mistakenly gone for Wind Shear QRH attitude 17 degs with TOGA power.

Edit. Once they were in Abnormal Attitude Law, the Stab Trim remained where it was last. (In this case full nose up)

"The system applies an abnormal-attitude law in pitch and roll if the aircraft exceeds any of these limits in flight.......

Angle of attack > 30° or < - 10° (- 15° for A319 and A321 (??? for A330))
Speed > 440 knots or < 60 knots

The law in pitch is the alternate law with no protection except load-factor protection and without auto trim. In roll it is a full-authority direct law with a yaw mechanical.

When the aircraft has recovered from its abnormal attitude, the flight control laws in effect are :
in pitch : alternate law without protection with autotrim.
in roll : full authority direct law with yaw alternate law."

Last edited by rudderrudderrat; 29th May 2011 at 09:45. Reason: syntax
rudderrudderrat is offline  
Old 29th May 2011, 08:38
  #530 (permalink)  
Join Date: Jan 2008
Location: uk
Posts: 806
Originally Posted by bratschewurst View Post
What's not clear from the report is why the PF made any stick inputs at all. My understanding is that, even in alternate law, the computers maintain the pitch and roll attitudes previously commanded until the stick is moved, so presumably the aircraft would have continued straight and level, assuming that was the case when the a/p disengaged.
But it does appear, if my understanding of the Airbus FBW system (gleaned mostly from these threads) is correct, that the aircraft would have continued in straight and level (and unstalled) flight absent the left nose-up input by the PF.
I think roll drops to direct in alternate law, but it is still a good question - if the aircraft was level it should stay that way. BEA don't indicate the cause of the first left roll - maybe turbulence, maybe the wing stalled, or maybe some input from FBW. They made a left turn shortly before - were they still in the turn when normal law dropped out, and what would that do ?
infrequentflyer789 is offline  
Old 29th May 2011, 08:45
  #531 (permalink)  
Join Date: Jun 2009
Location: Bangkok,Thailand
Posts: 106
This new generation of low time pilots might not cost much pay wise but what do you think Air France paid for this crash?
IMHO Air France seems set up to take the brunt of the blame along with it's pilots and procedures. We now know (it seems) that the least experienced pilot was at the controls when the hit the fan. This will be a hard pill for the public to swallow, and AF's reputation may already be in a downward spiral. Could AF be going the way of Swiss Air? It may just be heading that way.......
Razoray is offline  
Old 29th May 2011, 08:49
  #532 (permalink)  
Join Date: Jun 2005
Location: Australia
Posts: 6
CVR crucial

When can we expect to see a transcript of the CVR - I would have thought that this is now crucial to a proper understanding of what the pilots thought they were dealing with.
terry walls is offline  
Old 29th May 2011, 09:27
  #533 (permalink)  
Join Date: Sep 2010
Location: by the seaside
Age: 70
Posts: 885
No AF will not go the same way as Swissair.

Everyone in SR was worried about flying with crossair but management was using cheaper pilots to undermine SR pay scales.

Unlike the French the swiss had no qualms about buying foreign expertise.

Air France will not fail as the government will always protect it financially and AF have a protected market as the average french citizen believes if it's not french then it's not worth eating, using or visiting.
blind pew is offline  
Old 29th May 2011, 09:31
  #534 (permalink)  
Join Date: Oct 2007
Location: netherlands
Age: 52
Posts: 730
Some rude articles come up in not so marginal media outlets..

Wait for notorious pre occupied Fox news to have a go on the french..

'Baby' pilot at controls of doomed Air France Airbus | The Australian
keesje is offline  
Old 29th May 2011, 09:49
  #535 (permalink)  
Join Date: May 2011
Location: PARIS
Age: 58
Posts: 37
I'm afraid when I read all these comments about speed

Stall has nothing to do with speed : it has to do with angle of attack.

It's incredible to read at all those posts wondering about the speed.

A real pilot earing a stall alarm should not even care about the speed.

Stall alarm = nose down instead you are dead.
JJFFC is offline  
Old 29th May 2011, 10:13
  #536 (permalink)  
Join Date: May 2011
Location: Paris
Posts: 4
Airspeed and software


I read yesterday a part of this thread about what to do when the airspeed is not available.

Originally Posted by David Horn
No, in this situation pitch attitude and power is the primary reference.
Originally Posted by Graybeard
The aerodynamics of the plane is a perfectly good airspeed indicator.

For a given weight and cg, the pitch angle (nose up/down), as shown on the Attitude Indicator, will indicate airspeed. Pitch down, and the plane speeds up; pitch up and the plane slows down. It's there staring at the pilot throughout every flight.

Oh, that happens to be the basic of the Unreliable Airspeed Indicator training.
If, as said earlier in this discussion, the pitch and thrust behavior should give enough clues to the pilot to find a solution to the problem, why can't these clues be submitted to a computer, and have the computer calculate the airspeed from these clues ?

In other words, shouldn't the autopilot software be improved to take into account this calculation and enable the autopilot to go on instead of tripping off (1)?

In other words, imagine we are not talking about an airliner but about a costly military drone. Wouldn't you want the drone to come back safely when this kind of problem occurs and to develop the necessary software ?

(1) " the computers controlling the flight switched off the autopilot after becoming confused by conflicting speed readings, caused by the icing up of pitot tubes monitoring the plane's velocity" : Air France crash inquiry details pilots' battle for survival | World news | The Guardian ; "From 2 h 10 min 05, the autopilot then auto-thrust disengaged" :http://www.bea.aero/fr/enquetes/vol....mai2011.en.pdf;
andreg is offline  
Old 29th May 2011, 10:24
  #537 (permalink)  
Join Date: May 2010
Location: PLanet Earth
Posts: 818
Regarding the initial stall warning:

From 2 h 10 min 05, the autopilot then auto-thrust disengaged and the PF said "I have the
controls". The airplane began to roll to the right and the PF made a left nose-up input. The stall
warning sounded twice in a row.
The airplane’s pitch attitude increased progressively beyond 10 degrees and the plane started
to climb. The PF made nose-down control inputs and alternately left and right roll inputs.
These two paragraphs make me wonder if the stall alarms were really spurious.
At that altitude the buffet margin at M0.82 for G - onset should be around 1,5 - 1,6g, maybe even lower given the fact that additional Roll input was commanded..
Also speed might already have been reduced to M0.8 at that point further reducing the margin.
Edit: Stall warning activates even before that, thanks @HN39 for the comment. /Edit
A firm Nose-Up input should be able to achieve this.
Please note: Protections were already lost at that point and the Stall warning appears to have been subsequent to the Nose-Up stick command.
Please also note the subsequent Roll inputs by the PF.
either they were in significant turbulence or it was a sign that the aircraft was already in the incipient stage of the stall. These big Irons with full wing tanks and immense roll inertia normally don't have much tendency to 'wobble'.

Last edited by henra; 29th May 2011 at 11:02.
henra is offline  
Old 29th May 2011, 10:30
  #538 (permalink)  
Join Date: Jul 2009
Location: France - mostly
Age: 80
Posts: 1,689
Originally Posted by rudderrat
I tend to agree. In just 46 seconds (between 2h 10 min 05 and 2 h 10 min 51) the aircraft climbed with ROC up to 7,000 ft/min from 35,000ft to “37,500 ft and the recorded angle of attack was around 4 degrees.” (which is a sensible angle of attack).
They must have ridden in the updraught of a CB to climb that rapidly and still have flying speed.
Not quite. quote: "In level flight at FL350 and M.82 the pitch attitude (=AoA in level flight - HN) was 2.5 degrees". AoA=4 degrees is approx. the stall warning threshold at M.8 and results in a normal load factor of 1.39. If that LF is maintained during 9 seconds V/S=7000 ft/min.

P.S.: OTOH, at FL375 and M.68, AoA=4° corresponds to LF=0.78, i.e. a reducing V/S. At 215 kCAS and 700 fpm the flight path angle is about 1° up, i.e. deck angle is 5° up.

Last edited by Jetdriver; 29th May 2011 at 12:27.
HazelNuts39 is offline  
Old 29th May 2011, 10:32
  #539 (permalink)  
The Reverend
Join Date: Oct 1999
Location: Sydney,NSW,Australia
Posts: 2,020
Not the first one to question pilot response but how would you handle a stall warning with the HST at 13dg nose up ??
HotDog is offline  
Old 29th May 2011, 11:11
  #540 (permalink)  
Join Date: May 2011
Location: Paris
Posts: 4
Originally Posted by jcjeant

Unreliable speed:
Airbus presentation dated 26-28 September 2006.
Unreliable Speed – Latest Improvements. Presented by Sourya Kahoul and Marie-Helene
The use of rectangles and diamond boxes on page 8 and henceforth differs from conventional flowcharts where "yes" and "no" are usually written along the lines rather than inside boxes. In flowcharts, diamond boxes are usually used for questions rather than answers. The use of the green color in combination with the word "no" is confusing.
andreg is offline  

Thread Tools
Search this Thread

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

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