What Really Happened Aboard Air France 447

[P. Aaron]

Two years after the Airbus 330 plunged into the Atlantic Ocean, Air France 447's flight-data recorders finally turned up. The revelations from the pilot transcript paint a surprising picture of chaos in the cockpit, and confusion between the pilots that led to the crash.

Air France 447 Flight-Data Recorder Transcript - What Really Happened Aboard Air France 447 - Popular Mechanics

[Jinro]

Most aircraft (at least the one's I've flown) have something called "Pitot Heat" which prevents the airspeed sensors from icing over. I've never been in an Airbus, but surely it should have one as well. Engaging that switch probably would have prevented this all. Some basic flight planning would have probably helped too.

[()smoke()]

heavy article

[Big_Daddy]

It's strange reading something like that. Going over the last moments of their life. And then they just cease to exist... Heavy article. Thanks for sharing.

[The Crusader]

nothing short of terrifying

[i_got_a_mohawk]

Scary stuff

[Bob Lee (QSC)]

Fascinating article, though chilling. I saw a documentary on Netflix a few months ago on this topic.

I hope this leads to some improved training regimens, particularly on dealing with flight situations that depart from normal law. Increasingly automated flight control and management technology is great, but it also means that situations where the system gives up and goes manual are even further outside the normal envelope. It would be a crime to fail to learn important lessons from this disaster.

[Bob Lee (QSC)]

Originally Posted by Jinro

Most aircraft (at least the one's I've flown) have something called "Pitot Heat" which prevents the airspeed sensors from icing over. I've never been in an Airbus, but surely it should have one as well. Engaging that switch probably would have prevented this all. Some basic flight planning would have probably helped too.

I think the pitot heat is on constantly.

[Tituscrow]

What's equally terrifying is that I live about 5 miles from where Airbus make their wings, and I reckon 1 in 5 of the local community are employed there. And most of them I wouldn't trust to wire a plug.

Unsettling stuff.

[The Crusader]

Originally Posted by Tituscrow

What's equally terrifying is that I live about 5 miles from where Airbus make their wings, and I reckon 1 in 5 of the local community are employed there. And most of them I wouldn't trust to wire a plug.

Unsettling stuff.

yes but, in Airbus' defense, it was all the pilots fault.

[Tituscrow]

Originally Posted by The Crusader

yes but, in Airbus' defense, it was all the pilots fault.

Oh I know that. It was just a bit of a frivolous side-related comment. Though I do say a little Hail Mary every time I get on an Airbus.

[Jinro]

I think the pitot heat is on constantly.

But if the pitot tubes froze over, that would imply mechanical/technical failure of the heating system, while the article states that there were no mechanical/technical failures, just pilot error.

What about the de-ice system? Would Pitot Heat be included in that? I was under the assumption that de-ice was for wings and control surfaces while Pitot Heat was a separate switch, and the article did mention that they didn't activate the de-ice system until the pitot tubes were frozen.

[Bob Lee (QSC)]

Originally Posted by Jinro

But if the pitot tubes froze over, that would imply mechanical/technical failure of the heating system, while the article states that there were no mechanical/technical failures, just pilot error.

What about the de-ice system? Would Pitot Heat be included in that? I was under the assumption that de-ice was for wings and control surfaces while Pitot Heat was a separate switch.

At the time of the crash, Airbus was in the midst of a service campaign to replace the pitot probes. IIRC, there was a problem with the geometry of the older tube that tended to induce icing.

I believe de-icing systems are usually for wings, slats, ailerons, etc.

[aborgman]

Originally Posted by P. Aaron

Air France 447 Flight-Data Recorder Transcript - What Really Happened Aboard Air France 447 - Popular Mechanics

I've spoken with aerospace engineering professor who had early access to the flight recorder data for analysis. There were definitely some issues with the control electronics - specifically integral windup.

Pilots were pushing down on the control in an attempt to increase speed - but the plane still had a positive angle of attack (unknown to the pilots) because their stick interaction was only unwinding the integrator not actually nosing the plane down.

[Bredian]

The problem was not pitot heat, but 2 pilots not knowing what each other was doing with the stick - there is a significant difference in design between this Airbus and Boeing.... one pilot was pulling back on the stick, the other could not "feel" it. The most junior pilot was making a mistake and noone could see or feel it (because the controls weren't designed to transmit that).

The computer "averages" their inputs - FUBAR.

This is a pretty good case to show how a programmer/designer who is not a pilot, is not competent to design the system. Prove me wrong. This is huge problem with IT.

While they'll blame this on pilot error, this is a system design error, a training error.

[aborgman]

Originally Posted by Bredian

While they'll blame this on pilot error, this is a system design error, a training error.

Definitely a system design error combined with lack of training in dealing with the "manual" mode which wasn't actually completely manual.

Also, the control system seems to have had something to do with his choices in pulling back or pushing down - mostly due to a stall indicator which was DISABLED below a certain air speed(to avoid alarm on takeoff and landing).

So what was happening is they were in a stall condition, but pulling UP on the stick made the stall warning go away - because pulling up on the stick caused the speed to decrease BELOW the speed at which the stall warning disabled... and pushing forward on the stick increased speed (whenever they managed to unwind the integrator in the control loop) causing the stall warning to reappear.

Combine that with the fact that "manual" wasn't actually manual but was still dealing with integrator windup from the control system...

[The Crusader]

Originally Posted by Tituscrow

Though I do say a little Hail Mary every time I get on an Airbus.

me too.