Air France 447 "What really happened"

[cpxxx]

Had Dubois done so, he almost certainly would have understood, as a pilot with many hours flying light airplanes, the insanity of pulling back on the controls while stalled.

This is telling, I didn't actually realise until now that the Captain wasn't actually at the controls. This issue has been debated on several pilot forums. It's been suggested that a lack of basic flying skills is becoming more evident in younger pilots.

Remember in Europe generally pilots go to flight school and finish up with about 300 hours sitting in the right seat of something like an A320 or B737. They will have covered stalls during training but that's it. Now they're flying an electric jet that 'doesn't stall'. So holding full back stick with TOGA will fly you out of trouble. Except that in this case alternate law and the laws of physics apply. Which is what happened. The biggest failing though was crew coordination.

I met the Father of one of the passengers recently. Ironically he's a very experienced private pilot and knows all too well what happened to his daughter. He has difficulty understanding how the pilots reacted as they did. Sad, very sad.

[Tordon22]

It's my understand that in a situation where you have no speed indicators that the pilots are taught to apply a predetermined pitch and power (or some such). They aren't taught to "yank back" as you say.

It's true and one thing that a lot of people have gone to.

I'm not sure how they couldn't have known the plane was in alternate law. Knowing when it is or isn't and the differences in the controls would be a huge part of emergency training, I'd think.

It is a massive failure and I personally have no idea what these guys were looking at. The length of time that this happened over is the worst part to read about.

[deSelys]

As a low-time PPL, I'm not qualified to judge on how the pilots reacted, but for the sake of the discussion, here is what I learned from the interim reports (3rd here) and various discussions on pilot forums (mostly pprune.org):


Facts:

- The events happened during a sleep phase of the pilots circadian rythm
- When the Captain (Dubois) leaves the flight deck to take his rest, there is no briefing about the responsibilities of each FO (Robert PNF and Bonin PF at that time)
- When auto-pilot disconnects and the facts that speeds are unreliable is acknowledged, pitch & power procedure is not mentioned
- Robert (PNF) did acknowledge the degradation to Alternate Law at 02h 10' 22" (CVR transcript, page 90 of 3rd interim report)
- as mentioned in the article, lack of CRM
- When Dubois returns, there is no report of what happened (unreliable speeds, subsequent climb, alternate law)
- CRM doesn't improve after captain's return


Speculations:

- It is possible at 1st that Robert and Bonin were thinking that the initial climb and the beginning of the descent were caused by up- and downdrafts, as the plane was flying in a turbulent area
- When CAS falls below 60 kt, the stall warning is inhibited (3rd interim report page 20). The logic behind is the following: at such low speeds, the AoA vanes are unreliable. This may have contributed to the crew's confusion: while deeply stalled, no warning when the pilots are pulling on the sticks, but if they begin to push the nose down, the speed increases and AoA readings become valid again, triggering the stall warning
- While the AoA readings are sent to the flight computers, there is no AoA dial or tape for the pilots. Maybe they would have realized that they were in a deep stall if an AoA indication had been available for them
- While it didn't contribute to the crash as no stall recovery was attempted, the THS (Tail Horizontal Stabiliser) auto-trimmed itself up to the stops although speeds and, a little later, AoA readings were unreliable. As the auto-trim is actuated by relatively slow electric motors, it could have delayed for too long a recovery started with enough altitude
- Like someone said, training and culture seemed to fail the crew
 


Blue skies to the crew and their passengers.


On a side note, the 'bus aerodynamics proved a pretty solid stability because the plane didn't enter a spin while so deeply stalled (40+° AoA) in turbulent air and engines power going from TOGA to IDLE and back to TOGA.

[Golfer]

It's my understand that in a situation where you have no speed indicators that the pilots are taught to apply a predetermined pitch and power (or some such). They aren't taught to "yank back" as you say.

I never said they could would or should apply full back pressure due to a loss of or an unreliable airspeed indication. That's exactly the wrong thing to do in that situation however it is a technique an Airbus will let you get away with when you're in a maximum performance situation. Examples of which can be described as a stall recovery, windshear recovery or ground prox recovery (not hitting the mountain in front of you) because the airplane will let you. The idea being command every bit of performance you can get and let the airplanes computers not actually go to 11. It works when everything is working.

Now having said that applying full back pressure is the very much wrong thing to do in an unreliable speed condition, here we are with a pilot who did exactly that and would have done so all the way to the ground. The million dollar question is "why?"

Its a technique I was taught in a FBW airplane for windshear and EGPWS warnings. Not for stall recoveries however I'll credit the company that wrote the manuals, training culture, FAA special emphasis items and the very seasoned instructors of that particular program for that more than the airplane.

Regarding dual inputs, like FTJR said there is an override button and associated audible warnings to go along with it. Even so you'd be amazed at what you DON'T hear when you tunnel in on something with all that fixation. I've seen it happen with simple things such as warning chimes and audible warnings that can be muted.  They drive me nuts but if you're focused and saturated your won't process any more inputs until you snap out of it so to speak.

[Angus]

Stalling an aircraft through more than 30K with a reasonably good nose up attitude and all the means to detect exactly that, while having all control surfaces and engines in good order is something that encourages me to....being seaborne.
Just saying.

[B4Buster]

Seems as though there was some Spatial Disorientation thrown in there too.

[Reschke]

All I can say is WOW! What a tremendous cluster%^&$ that must have been.

[SFRT - Frenchy]

Those glass cockpit and fancy autopilots mush the brains of pilots, the more glass time you have the less you react to basic airmanship. The worst F/Os we have on the Metros are the 2000h CFIs & 5000h+ jet/glass pilots 

[F22RaptorDude]

I think I recall hearing about this on seconds from disaster, or something similar

[Babalonian]

Makes me sick, honestly.  The day I found out AB's fly-by-wire dual controlls were unsynchronised I knew this was only a matter of time before one idiot would get the other driver and the rest of their passengers killed.   

[Golfer]

Makes me sick, honestly.  The day I found out AB's fly-by-wire dual controlls were unsynchronised I knew this was only a matter of time before one idiot would get the other driver and the rest of their passengers killed.   

Oh please...

The airplane didn't kill anyone.

[guncrasher]

Oh please...

The airplane didn't kill anyone.

think george carlin said once:  the day the make airplanes of the same stuff they make black boxes, nobody would die in airplane accidents.


semp

[Krupinski]

think george carlin said once:  the day the make airplanes of the same stuff they make black boxes, nobody would die in airplane accidents.


semp

That'd be a pretty frikkin heavy airplane.

[Tupac]

That'd be a pretty frikkin heavy airplane.

With enough thrust a refrigerator will fly. Not well nor far, but it will fly.

[RufusLeaking]

Is there no stall buffet in an Airbus? No stick shaker?

I was in the last KC-135 instructor pilot class to practice stall recognition. The plane was slowed to just above stall speed. Airflow separation on wings would be felt on the elevator, starting with a tickle and progressing to 'elephants.'

The practice was stopped due to an unrelated incident. For the maneuver, the forward and aft body tanks had to be empty. While I was in training, a KC crew left an aft body pump on after the tank was dry. As they descended, the air got dense enough to provide the oxygen to the fuel fumes and overheated pump for an explosion that took the tail off. After that, forward and aft tanks were no longer allowed to be emptied, so stall recognition training was ceased.

Back to the Airbus, can't simulators simulate the stall vibrations? Certainly, loss of airspeed indications can and is practiced in simulators.