Originally Posted by
bearfoil
DOZY: As a sw boffin, do you expect much of what is needed for an understanding has a particular timeframe?
Well, from a software perspective, what you'd need to do is load up a test rig made up of the exact hardware installed on that aircraft with the exact software installed (down to point releases), apply the DFDR data to that test rig's inputs to see what the software and hardware does - then repeat this test anything up to hundreds or thousands of times to see if any anomalies appear. I suspect this would not be a small or inexpensive undertaking, but it would be do-able.
How did this START?
It's looking very much like an inappropriate response to the situation as it developed, starting with the nose-up inputs shortly after FMS disconnect, then progressing through selection of TOGA and the later aggressive and maintained nose-up inputs all the way down, and it wouldn't be the first time it has happened (as I've always said, I think this was another Birgenair-type situation).
Way back when I started commenting on these threads earlier in the year I said that unreliable instrument readings in the wee hours at night, in IMC with unsettled weather hundreds of miles from land was a nightmare situation for any pilot to confront, and that any findings of mishandling on the part of the pilots *must* take this factor into account, and this is why I got very agitated when people said that by saying software failure was unlikely to be a cause I was blaming the pilots.
Lack of training appears to be a significant issue here, along with poor CRM practice when it comes to rest periods. It would appear that the ITCZ is a known problem area when it comes to aviation, and many pilots on the thread have expressed the opinion that they would not have left the two F/Os in charge until safely out the other side.
Colganair 3407 was the wake up call that the airline industry had bred complacency in two distinct areas, one of which was the effect of fatigue and the other was poor recognition of stall conditions and application of the correct response to those conditions - but that investigation was still ongoing when AF447 crashed and the final NTSB report not released until February 2010, 8 months after AF447.
In short I think AF447 was a "perfect storm" of the problems within aviation. Birgenair and Aeroperu had shown what could happen if the pitot-static system was compromised, but both of those incidents happened in the climb phase. Not much thought was given to what would happen if something similar happened at cruise altitude, with the attendant limits on possible escape procedures. Airlines had been training pilots to respond to approach to stall and the warnings generated without getting into what would happen if you were to actually stall, how to recognise it and - crucially - how to get out of it, because it requires going against the human instinct to cram on power and pull up when the correct response is actually to get the nose down and hold it there until the speed comes back and the wings are flying again. AF also deserve to come in for criticism for failing to expedite the replacement of pitot tubes which were known to have problems.
@jcjeant - if you're honestly suggesting that airlines were training their Airbus pilots on approach to stall only and the rest of their crews were getting full stall recognition and escape training, I think you need to get some perspective.