OK, so instead of going from 50 MPH to 200 MPH, we go to 440 MPH. So average speed during downhill run is 245 mph, while accelerating at 0.75 g’s to final speed. 18 seconds required, during which time they lose 3234 feet of altitude.
pulling out of a dive at 440MPH and 3.42 g’s (sticking with the figure I used previously), the turn radius is 3782 feet. But - and I forgot to account for this earlier - if the plane is only in a 30-degree dive, then it doesn’t use up 3782 feet of altitude to achieve level flight; it only needs (1-cos(30))*3782 = 507 feet. 45-degree dive? OK, 1108 feet.
Add in the 1819 feet of altitude they lose while trying to achieve a nose-down attitude, and you get 1819+1108+3234 = 6160 feet of altitude as the bare minimum required for recovery, assuming skilled pilots who know what 3.42 g’s feels like, and assuming the aircraft can generate enough lift at 440 MPH to achieve 3.42 g’s without stalling the wings.
The behavior where the stall warning sounded when they were less stalled but stopped when they were more stalled may have been by design. It certainly wasn’t known to the crew. So from their POV it amounted to a malfunction; systems designed and intended to guide their behavior did in fact guide their behavior: in exactly the wrong direction.
The real punch line to the whole thing is your statement “… they had no idea WTF was going on.” And they had 3-ish minutes to figure it out. And the symptoms were changing as they descended and as the pitot tubes deiced. Not an easy problem.
It has a lot more to do with modern airplane design than it does a generation gap.
Complex systems fail in complex & subtle ways.
In the Olden Dayes airplanes were inherently simple machines with inherently simple failure modes. Systems interaction was minimal. Pilots were taught block-diagram level understandings of how each system was built and operated, and were expected to interpret the signs of malfunction themselves, arrive at a diagnosis, and perform whatever mitigation seemed appropriate.
Nowadays that approach is more than deprecated; it’s prohibited and will get your butt chewed by the Feds or your boss if you approach malfunctions that way.
So now we’re supposed to run the written fault tree analysis and if there isn’t one, or the situation is ambiguous, we’re totally on our own. Try to keep the shiny side up & land ASAP.
That’s not to say basic airmanship and aircraft handling skills are deprecated. They’re just harder than ever to get and keep sharp. For pilots who’ve spent their whole careers on highly automate3d aircraft in highly automated corporate cultures those skills may have never been developed to a high level in the first place.
When the first highly automated jets came out and were being flown by guys with a whole career on primitive jets the standing joke was “I’ve forgotten how to fly but can now type 50 wpm”
The underlying problem AF447 faced was one of these: Byzantine fault - Wikipedia Which is also the problem the Airbus engineers faced. How to interpret and react when some systems are lying to you and you’re not sure which ones they are? Or you’re not even sure whether they’re lying or not?
IMO … This accident, and another almost-accident I’ve read about, demonstrate cases where the Airbus engineers designed a system which is simultaneously too smart & too dumb. It’s a brittle failure mode where it’s helping you a bunch until suddenly and with little warning it switches sides and starts trying real hard to kill you.
I don’t know where you’ve gotten the 3.42Gs from. I’m not asserting you’re wrong, I’m asking the question. The minimal info I have on Airbus flight controls would seem to indicate they’d limit the pilots to 2.5G.
Regardless of that quibble, it’s clear that for a sizeable fraction of the plunge from cruise altitude the situation was, as a matter of physics, recoverable. Whether it was recoverable as a matter of man / machine interface and crew dynamics for the given crew on the given day is a separate and much more interesting question.
It came from md2000’s post (#5). I used it in my first analysis, then just stuck with it. It’s not too different from the “150% of 2.5 Gs” that you described (although FBW won’t let you deliberately exceed that 2.5-G limit).
If 2.5 G is the limit, then that means the turn radius increases by ~50%, so add another 500 feet to the altitude requirement. FBW also means skilled aerobatic piloting isn’t required; just pull back on the stick as hard as you can and let the FBW system limit you to 2.5 G.
They were descending at roughly 10,000 ft. per minute. So if we assume, arguendo, that 10,000 ft. was the lowest altitude a perfectly flown maneuver would have cleared the ocean, then they had roughly 2 minutes, 1 minute, and 30 seconds respectively to sort the problem out starting from your three altitudes. If someone wanted to believe recovery from 5000 above the ocean was possible, then add 30 seconds to each estimate.
F-16A.
There aren’t any special regs about passenger jets vice other aircraft. So 1000’ above the ground or nearby obstacle would be a generally applicable lower limit speaking just to regulations.
As a practical matter very short hops might only climb to 2 or 3000 feet. e.g. flying between any of Oakland, San Jose, or San Francisco. Or between Newark, JFK & LGA. Which is also about the altitude we’d use at any airport if making a go-around and needing to reposition for another approach.
Sustained longer-duration cruise will normally climb at/above 10,000 above sea level to get above the 250 knot speed limit. We used to fly the 35 miles between PUB & COS at 10,000 or 11,000 feet so we could go 330 knots. Given the terrain there on the front bench is around 6000 feet above sea level we were 4000 - 5000 feet above the ground.
Any longer flights the most efficient procedure is to climb until it’s time to descend. For even a 100 mile flight we’ll be apexing at 20-some thousand. So while one could legally fly lower, it Simply Isn’t Done.
Cool. My brother flew the F-16 when he was a zoomie. Before that he was in the FB-111. Don’t know which flavors of those, though. I’ll ask him when I see him next week at our Christmas family reunion. When he was at Luke in the mid-1990s he got me into the sims and we did a low-lever buzz of the Vegas strip at night. That was a nice light show. Every time he tried to get me to land the F-16, though, I crashed and burned. No matter how much he tried to help me. I was better at artillery FDC, so I should stick to that (USMC).
He also worked at Johnson Space Center with the shuttle landing scenarios, and he got us into those sims too. That was fun. He proved that you could barrel roll and then loop the space shuttle before bringing it to a nice smooth landing at Edwards.
Anyway back to AF447 - that PM article that Xema linked to upthread was riveting reading. I’ll read it again because the flight situations don’t come too quickly to a ground-pounder like me. I’d like to discuss it with my brother, too.
But from my first reading it sounds like the crash is mostly due to pilot error - Bonin, specifically. I’ll re-read it tonight, though.
In the incident described in that article, the captain had accidentally wedged a camera in between his seat and the control stick and unintentionally commanded the plane into a dive when he adjusted his seat. When the co-pilot pulled back on his stick to correct for the dive, they got the warning.
No. They just retrained and reminded all the pilots of this “feature”. There are two philosophies with respect to Fly-by-wire designs. Airbus prefers the simplicity of their system, and they’re sticking with it. Boeing has a design that more closley mimics traditional, mechanical controls.
Speaking of Yeager, in a section of this interview he talks about two of his more famous stalls. He stalled the X1-A at 80,000 feet and needed 51 seconds to recover, during which time the X1-A fell to 25,000 feet.
The second stall was with the F-104, as seen in The Right Stuff. It stalled at 104,000 and Yeager was never able to get the nose down. He ejected at 6,000 feet.
It’s been fixed in the sense that they realize it’s a potential problem and most aircraft delivered in the last couple years have the revised software which sounds the alert as pointed out by engineer_comp_geek.
The fix has not been retrofitted to all Airbus aircraft everywhere, nor is it (yet) required to be, at least not by FAA.
I understand what you are saying, but is this really true? If the pilots had removed their hands from the controls not long after all the warnings, would the plane have had enough time and height to keep flying, even if they had to skim the waves? And isn’t this the essence of this thread?
I’ve just had a browse through the final report and this issue isn’t really addressed at all. It is not listed as a causal factor or discussed in the text of the report. Also as I understand it the issue wasn’t so much of dual inputs being received, the FO was applying the sole input for most of the time, but more that the fact of the FO’s nose up input was not available to the other crew. The audio warning “DUAL INPUT” doesn’t solve this.
The first alarm they received was simply an idicator that the plane came off auto pilot. Had Bonin touched nothing at this point, everything would have probably been fine.
“So here is the picture at that moment: the airplane was in steady-state cruise, pointing straight ahead without pitching up or down, and with the power set perfectly to deliver a tranquil .80 Mach. The turbulence was so light that one could have walked the aisles—though perhaps a bit unsteadily. Aside from a minor blip in altitude indication, the only significant failure was the indication of airspeed—but the airspeed itself was unaffected. No crisis existed. The episode should have been a non-event, and one that would not last long. The airplane was in the control of the pilots, and if they had done nothing, they would have done all they needed to do.”
From here: Should Airplanes Be Flying Themselves? | Vanity Fair