Before they do that, I’d be curious to know if they’ve reproduced the conditions of this accident in a sim. I don’t know if it’s common practice for all accident investigations, but I have heard of it in some. Try a few flight crews and see if they can correctly diagnose the problem and get it under control without crashing. Maybe even have the co-pilot “panic” and pull back on his stick to see if the other pilot notices. Ideally, you’d want to do this with crews who haven’t read the transcript and don’t know what to expect.
Thank you for your feedback, I am glad you found it to be a good analysis.
I sincerely hope that these kind of measures are being put in place as it would seem remiss of any carrier not to train their pilots for all reasonable cofigurations of the aircraft. Surely icing of the pitot tubes in cruise cannot be considered so remote as to not train for it? But perhaps I betray my position as a non-pilot with this lack of knowledge. I will await the actual crash report with interest.
One hopes the lessons that will be learned from this terrible tragedy will be heeded.
Regards.
I got an email forward about flying and piloting today. A couple of them reminded me of this thread.
Speed is life. Altitude is life insurance.
And
Aircraft Identification:
If it’s ugly, it’s British.
If it’s weird, it’s French.
If it’s ugly and weird, it’s Russian.
What about Concorde? By that logic, it would be both ugly and weird, and yet it is not Russian. Maybe its ugly weirdness is what prompted the Russians to copy it with the Tu-144, though.
Here they are together in Germany at the Auto & Technik Museum Sinsheim. Couldn’t find a better picture, but seeing them both sided by side it is quite easy to tell which is [del]weird[/del] French and which is [del]ugly and weird[/del] Russian.
Did they tack on the Dumbo ears to make it fly better?
They’re called “canards”. According to Wikipedia:
Basically yes as Robot Arm linked. 
Something you can’t see in the picture, but is very obvious in person is the shape of the wing. My understanding is that the difference in shape is a significant factor in the need for cannards on the Tupolev. From above the wing shape is very similar on the two aircraft, however viewing them from behind there is a noticeable difference. The Tupolev wing is straight, while the Concorde wing is curved downward at the ends. You can kind of see the curvature in this image.
Thanks for the explanations and the picture, Robot Arm and Hbns. The Concorde really is a beautiful and amazing piece of engineering!
Interesting analysis. It’s scary that something as simple as “pilot pushing stick the wrong way for a few minutes” is what caused a major commercial airliner to plummet into open ocean in the middle of the night.
Push the stick the wrong way on any airplane long enough and the same result will occur. What is scary about this is that the other two pilots did not recognize the problem and that the control design did not give them the feedback needed to recognize the problem in time to correct it.
At a holiday party I had an interesting conversation with a man who is an A330 pilot. He had not had a chance to read the details of the AF447 crash.
When I told him that Bonin appears to have held the controller back through the whole event, he mentioned that there is a mode known (IIRC) as “Alpha Floor”: if you hold the controller aft, the plane will climb to the maximum altitude it can maintain, and stay there, at a low but fully controllable airspeed (or so I gathered). The catch is that this is true only under Normal Law, which was not in effect after the pitot problem. But if we assume that Bonin was not aware of this difference, or did not properly note the shift from Normal to Alternate Law, this could be an explanation for his (otherwise rather strange) action.
How does the plane know when all the pitot tube inputs are in error? There was certainly a cascade effect of bad information fed to the computers that should never happen and has been addressed with pitot heat upgrades.
As I understand it, there is a narrow window that the pilots have to correct such a situation and they need to know what is going on for that to happen.
This isn’t like the crash in Buffalo where the crew was given correct information and basically did the opposite of what needed to be done. And yes, piss poor design seems to be how I’d describe it.
So… now that the final report is out, can any experts glean any added insight? Does the final report add anything new? The message I get from the report is gross pilot error (obviously in an incredibly difficult situation, I’m not judging).
First, the copilot’s action in holding back the stick was not totally illogical. You are flying along, the stall warning is off, the engines are close to full throttle, a climb is standard procedure and will probably take you above some of this bucking and shaking which probably is the storm you are in. In fact, when the captain pushes the nose down a bit, the stall warning comes on. Nose down does not seem to be right. Climbing seems right.
Of course, if you stop and think (in the middle of what must have been wild shaking and a cockpit full of noises and contradictory instruments) nose down should NOT cause a stall. However “if I do A it stalls, stop doing it” is not an unexpected response.
IMHO they don’t need completely mechanically linked sticks, but some sort of “force feedback” system, higher resistance, maybe a bit of shaking, when pushing against a contrary stick would go some way toward avoiding the “single channel of input” issue. They did it 15 years ago for a cheap joystick, surely they can do it reliably to two sticks considering what an airbus cost.
Also, the stall warning was a problem; ***the ***problem. Even if they had a different stall warning - two-toned or whatever (thinks french police sirens) for “I think you are stalling but I’m not sure” that would have been better than “you are not stalling”. Absent the stall warning, it’s not completely surprising what the copilot did and it’s sad that this thread especially seems to be heaping the blame on his action.
I had my instructor once demonstrate the reliability of seat-of-pants flying. He took me into a cloud, no visual cues, and said “just fly what feels right”. After 30 seconds, he says “now check the instruments”. What felt 'straight and level" to my inner ear had turned into a spiral downward, a slow turn while descending. And that was soft, bump-free flying.
All the discussion of multiple modes, failsafe auto-climb options, etc. imply means the machine has become so complex that it is difficult to remember all the picky details in times of imminent panic. I suppose the best legacy these fatalities give us is that all pilots now have a slap-in-the-face refresher course on what could possibly go wrong in this situation, and all the odd trivia associated with the fly-by-wire system. That makes all aircraft that much safer until the next serious but different unusual situation comes along.
Of course, the altimeter should be your most reliable instrument. It does not need a pitot, it can take input from any unpressurized area of the plane.
What the copilot did isn’t completely surprising - but it was far outside what should be expected of anyone competent to fly this aircraft. His failure to understand that the plane was in Alternate Law, or that his action - acceptable under Normal Law - was now not appropriate was an essential part of the process that led to this crash.
IOW, the evidence strongly suggests that this action is fully deserving of blame.
My understanding is that the only instrument that wasn’t working was the airspeed, correct? The artificial horizon, the altitude, that was all working, right? Shouldn’t they have been able to see that they were pointing too far above the horizon in a situation where they were worried about stalling?
As someone suggested in a much earlier thread, once the plane stalled, odds are it was bouncing like crazy - it would not be a smooth, soft ride, more like it was being shaken and rocked at the same time. Add to that the turbulence of the top end of a thunderstorm and the problem might have been compounded by no steady indicators to read. Everyone says “nose up 40 degrees” but that’s the extreme, it was about 15-20 and with a lot of shaking, and no horizon reference, odds are it was hard to get an idea without instruments exactly what the attitude of the plane was.
Maybe the copilot remembers the “autoclimb” rule but forgot it did not apply in Alternate Law mode…
*“Shake me like a British nanny!” *- Stewie, Family Guy
When you are being shaken all over, it’s difficult to stop and do some introspection on what may be happening or what does and does not make sense.
Is there anything in the report about how turbulent their aircraft was during the problem?
Even the airspeed indicators were only displaying incorrect information for about 180 seconds in total. Every instrument in the cockpit was functioning accurately again before the plane even started to descend. The problem was that brief disconnection caused the plane to change into “alternate law” and the pilots to fail to trust the instruments from then on.
[QUOTE=md2000]
As someone suggested in a much earlier thread, once the plane stalled, odds are it was bouncing like crazy - it would not be a smooth, soft ride, more like it was being shaken and rocked at the same time. Add to that the turbulence of the top end of a thunderstorm and the problem might have been compounded by no steady indicators to read. Everyone says “nose up 40 degrees” but that’s the extreme, it was about 15-20 and with a lot of shaking, and no horizon reference, odds are it was hard to get an idea without instruments exactly what the attitude of the plane was.
Maybe the copilot remembers the “autoclimb” rule but forgot it did not apply in Alternate Law mode…
“Shake me like a British nanny!” - Stewie, Family Guy
When you are being shaken all over, it’s difficult to stop and do some introspection on what may be happening or what does and does not make sense.
Is there anything in the report about how turbulent their aircraft was during the problem?
[/QUOTE]
The captain managed to walk back to the cockpit from his sleeping quarters during the stall descent. So the plane may well have been shaking but it surely can’t have been “bouncing like crazy”. They would have been able to read the instruments; in fact the final sentence the Captain uttered before the plane hit the water was a reading from the pitch indicator (“15 degrees of pitch…”).
Anyone following this?
I don’t think manslaughter is appropriate.