That initial climb (again), and some other questions
The 3rd Rapport d'etape adds a lot of detail, and clarifies a lot of what was previously just speculation, and it still tells me that the most interesting part of the event was that extraordinary climb from FL350. Not only did it put the aircraft in a hopeless energy state from which a stall was inevitable, but it also tells us a lot about the instrument flying skills of the PF. In fact it is so bizarre that I am still looking for an 'AHAH' moment that reveals just what it was he was trying to do and how he was doing it. I don't buy the argument that because he had previously thought about climbing out of the turbulence, this was now his way of doing it; no one in their right mind would have done it at that rate, nor deliberately pitched to 10 deg nose up at that level. Nor do I believe that these pitch inputs were inadvertent side effects of lateral inputs - at one point he was achieving 1.6g, and even if some of that was due to turbulence (and there isn't much sign that it was really particularly rough in the 20 sec or so before disconnect), there is still a lot to explain away. The stick force needed to do that isn't zero, some one posted the calibration some time back and I think you need around 5kg pull to do that, hardly inadvertent.
So what was he reacting to? I suggested a month ago that when the speed and Mach became invalid the altitude corrections would have been wrong causing a sudden change in indicated altitude, as happened in the Air Caraibes incident. In that case the change was -300ft, not a huge amount but just possibly enough to prompt the PF to want to pull up. the current BEA report's graphs don't show any such jump, though, intriguingly, the table on pages 92/93 shows that at the start at 2:10:05 the altitude (on the left PFD?) was 35 024ft, while 4 seconds later at 2:10:09 it is shown as 34 664ft, so I guess there was a jump. Nothing other than a complete disregard of attitude explains PF's later insistence 'Okay, okay okay je redescends', and similar phrases, when he plainly achieved no such thing. It seems the BEA are wondering whether reappearances of the FD bars might have had an influence, but if they did, this merely shows how the poor fellow seemed unable to 'look through the bars' to see and assess the underlying attitude. There may be good explanations for this behaviour, but I am struggling to find them. On the face of it, the question for Air France is whether this skill level is typical (in which case they have a humongous training problem to resolve) or unique to this co-pilot (in which case checking and evaluation needs attention).
One other puzzle, why (as had been noted) does the AP disconnect occur a few secs before the indicated speed loss?
I note Owen Glyndwr's Cm - alpha curves, which suggest that return to normal flight should have been possible. but is it enough just to correct the pitching moment? Don't you have to pitch down to within a few degrees of the actual flight path angle, which in this case, once the full stall (I also don't like calling it a deep stall because that's different) is developed, would mean pushing the nose down about 45 deg? I can't see them having the insight to do that