Originally Posted by soylentgreen
upon re-reading my words, they did come across snarky and like a cheap shot, so I apologize.
No need to apologize, harshness was mine.
Originally Posted by soylentgreen
I said "it would be fun" and "The outcome would be quite interesting."
Agree it could turn out to be fun & interesting, not necessary for the same reasons, though.
Originally Posted by soylentgreen
Some things we can: training and machines. I'm simply saying that
we should think of the big picture here, and improve the things that we do control.
Agree, but big picture is autopilots resigning almost daily in cruise levels (no need to have UAS) and most of the occurrences getting sorted out without anyone noticing. AF447 is just a detail, very significant and very ugly though.
Originally Posted by soylentgreen
How many miles do we need google's robot cars to have accident-free before we trust them?
Aviation has it covered in certification processes. As soon as robots show functionality and reliability of human pilots, we'll have pilotless aeroplanes. Don't hold your breath waiting for it, key issue is still functionality, with latest computers being no significantly better than ENIAC.
Originally Posted by soylentgreen
The "naturalistic" study shows that 1 of 37 crews in similar situations crashed. As I mentioned, that's such a small sample size that we can't say whether the actual percentage is closer to 0% or closer to 10%.
My point was that since a) level of safety currently achieved makes sample very small b) there are so many variables that affect the outcome, if we want to learn something in order to further advance aviation safety, we can not rely on statistics to get any meaningful result. That is something known since at least mid 90ies (that's about time I started paying attention) - at the time "swiss cheese" and "accident chain" analogies were accepted and it was recognized it wasn't enough just to analyze accidents but also a close calls, where all safety features failed except one or two and result was no one hurt, no damage when it could easily ended otherwise.
Now if we apply numerology, and boldly and deceptively call it statistics, to the sample provided by BEA in interim 3 by assuming that survival means no problem while crash means manual handling of the aeroplane deficiency, we'd come to conclusion that it's not such a big deal if just 3% of crews are unable to control the aeroplane when hit with UAS. Of course it is deeply wrong conclusion and you were on the right track when you mentioned "survived for wrong reasons". To discover what were wrong reasons, we need detailed case studies, not just statistical analysis of contributing factors. BEA has analyzed 13 out of 36 discovered cases of UAS and added one which couldn't be analyzed; case of TAM 332 on Nov 12 2003. Both pilots pulled as airspeed was lost but then pushed when stall warning went off so we need to resolve why some pilots perceived and understood the warning while some didn't.
Chapter 1.17.1.5.4 is also pretty damning.
Originally Posted by Nemrytter
It's not yet published but should be in a month or two.
I'm not holding my breath.
Originally Posted by deSitter
At what point was the "deep stall" unrecoverable?
Insufficient data to tell whether it was unrecoverable. No one felt suicidal enough to fly the AF447 profile for real.
Originally Posted by deSitter
What exactly is a deep stall?
Something of no concern on A330/340.
Originally Posted by deSitter
Was the trim issue the main obstacle to recovery?
Not trim, crew.
Originally Posted by triple bravo
every significant "stick down" input was followed by a) THS forward movement, b) pitch decrease and c) airspeed increase.
Not quite, THS remained at full nose-up, elevators moved from full nose up to half-nose up and that was enough for nose and AoA to go down. Good indication there was no deep stall but not definite proof - now this sentence should give ammo to conspiracy theorists.
Originally Posted by notfred
Is this true for other modern jets?
I guess not, especially for those that need to have mach trim e.g. 757. Issue is that if you know procedures, recognize when they need to be applied and follow them, you need not to know whether your aeroplane suffers from mach buffet or not.
Originally Posted by Sadpole
We are fooling ourselves that human beings are capable of logical reasoning
Most of the time, they are.
Originally Posted by Sadpole
true that a true ace pilot, one that committed his whole life to aviation, could NOT be affected by even most illogical configuration of plane controls.
False. Air has no respect for anybody. It is ignorant who coined the phrase "beginner's mistake". Experience is double edged sword, as the experienced "ace" often does not recognize the times he tempted the fate and is bound to repeat the feat. With a bit of luck, such an "ace" can live to retirement.
Originally Posted by SadPole
the whole system was demonstrated to be a complete failure
You condemned the whole system based on single accident. Such a feat can not be excessively serious.
Originally Posted by SadPole
imagine vegetable engineers, people who never had any desire to create new things,
Vegetable is as vegetable does. Just because you don't like some aspect of the machine, it doesn't mean it's bad or that improving it would cause massive degradation of other aspects.
Originally Posted by Triple Bravo
mishandling of the Mulhouse case
Care to elaborate?
Originally Posted by TripleBravo
Why are certification requirements not adapted to recent findings?
Accident investigators have only the power of advice. Also their scope is limited to accident at hand.
Originally Posted by TripleBravo
They weren't suicidal, they fought for their own lifes as well and sadly lost.
They fought for their lives by tying the noose around the neck and jumping off the stool, without ever recognizing there's a rope.
Originally Posted by TripleBravo
How come that flying manually is something not much appreciated in the industry?
PPRuNe is not the industry.
Originally Posted by TripleBravo
The captain wasn't in his seat, and due to his CV I'm convinced he would have had the abilities to handfly without major problems.
Have a go at BEA's report. While there were deficiencies noted, 36 crews given the manual control of the aeroplane performed well enough not to kill anyone. I doubt they had more manual training at high levels than AF crews. Some were AF crews.
Originally Posted by DozyWannabe
Then please explain why there are at least five dead astronauts and cosmonauts who were lost in training accidents.
If extend the scope to display flying, we may add Rimantas Stankevičius, Buran test pilot, to the list. Devoted, experienced, professional. One day overhead Salgareda he entered split-S too low.