![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
The Wild (and Scary?) Blue Yonder.
Oct. 1st, 2014 12:54 pmI've probably mentioned now and then that I'm an aviation buff in general--have been since I was a kid. I love planes and flight and pretty much everything related. Before 9/11 made it impossible (don't tell me the terrorists haven't already won), one of my favorite things to do was to go to the airport and stroll around the terminal and gates, planespotting and peoplewatching. We also had a great place to park right at the end of the runways at Orlando International, so I could park and watch the jets coming in to land.
But the darker side of it fascinates me too--aviation disasters. And especially the ones that simply shouldn't have happened; the accidents that were purely human error, no mechanical problems or outside interference. They're usually a series of bad decisions, overlooked information, and just sheer bad luck. Tenerife in 1977, when one 747 erroneously crossed a runway right in front of another 747 on its takeoff roll. San Diego in 1978 (my 3rd birthday, that one was), when a 727 pancaked down onto a Cessna in midair and went down into a residential area. Terrible, but fascinating to me.
Yesterday I read this Vanity Fair article about the crash of Air France 447 in 2009, which I'd not really read much about before. It's a long piece, but extremely interesting to me, and offers some very interesting insights into the nature of modern commercial aviation and the most likely causes of air crashes these days. The thing is, the planes themselves have become so reliable from a mechanical perspective, and so automated from an operational perspective, that pilots spend the vast majority of their flying time just watching the plane fly itself. The flight plan is pre-programmed for them on the ground, so all the pilots have to do is take off and climb out, and from there the plane can do the rest. The result is that many pilots these days have no real stick-and-rudder experience since everything's always been automated for them, and so when and if something out of the ordinary happens--something that makes the aircraft's computers throw up their hands and hand things back over to the human pilots--those human pilots are all too often at a loss to determine what has happened and what to do.
Air France 447 was a mystery for quite some time. The plane essentially disappeared over the Atlantic on a flight from Rio to Paris, and the pilots had never indicated that there was any trouble, never made any distress calls. Several days later some floating wreckage was discovered, and seemed to indicate that the aircraft was at least mostly intact when it hit the water. But the black boxes were nowhere to be found, so the mystery remained: what happened? Finally in 2011, almost 2 years after the crash, the fuselage of the plane was located at the bottom of the ocean, and the black boxes were recovered. And what was discovered in the flight and voice recorder data was astounding.
As the flight passed into a region of thunderstorms at cruising altitude, ice buildup caused the pitot tubes--externally-mounted tubes responsible for measuring airspeed--to clog, giving the onboard computers no idea of the aircraft's velocity. At that point, two things happened: the aircraft disengaged autopilot to give the pilots control, and also switched to a different set of flight parameters. Under normal circumstances, a modern, computer-assisted aircraft will simply not allow the pilots to make control inputs that will take the aircraft out of its safe flight envelope. It'll basically say "I'm sorry Dave, I can't do that," and remain in safe flight. But the mode Flight 447 switched into removed those restrictions, and it's quite possible--nay, quite likely--that the pilot at the controls (actually a copilot, the captain having excused himself for a nap due to being up too late partying with two women the night before) had no idea that was the case, and no experience flying outside of the usual Airliners-For-Dummies mode.
Now, it must be noted that nothing at all was actually wrong with the aircraft at this point. All the pilots had to do was absolutely nothing, and the plane would have continued cruising in level flight, eventually the pitot tubes would have thawed and unclogged, the autopilot would have re-engaged, and they'd have been in Paris by morning. But for whatever reason--subconscious panic and brainfreeze, possibly--the pilot's instinct was to climb (even though they were already about as high as the plane could go in the current environment), so he pulled back on his control stick, which raised the nose of the aircraft and the angle of attack of the wings. And he kept on pulling back until the plane began to enter an aerodynamic stall, which caused buffeting and turbulence--and caused it to begin losing altitude as the wings stopped generating sufficient lift.
Now here's where my fascination really comes in. In many aircraft with dual controls, when one control is moved, its counterpart moves as well, which gives the copilot a visual indication of what's happening with the controls. Not so in Airbus aircraft like AF447--the controls are basically a pair of joysticks off to each side, and they are not linked. So for the next 3 minutes or so, as AF447 was falling out of the sky, the pilot at the controls was holding his stick back, keeping the nose lifted, keeping the plane in a stall, and the other copilot (and the captain, once he came back into the cockpit) had no idea because the control sticks weren't linked and the pilot pulling back didn't think to mention it (if he even realized he was doing it) until past the point of recoverability. And there's the real rub. Aerodynamic stalls are Flying 101: something every pilot should know about since day one, as well as how to get out of one. If you have the altitude to spare (and AF447 did, at first), you simply lower the nose and dive into the stall, which will restore airflow and lift, and then you pull up out of the dive. It's as basic as aeronautics gets. But these 3 pilots, being so unused to anything out of the ordinary, and so out of touch with actually hands-on flying the aircraft, failed to recognize that they were in a stall. Even despite the aircraft itself sounding a stall warning (quite literally a voice saying "Stall!") no less than 75 times during the crisis. They may have simply blocked the warnings out, or they may have assumed that an actual stall was impossible--since usually it is--and written it off as a symptom of the lack of airspeed data. So finally the copilot yanking his stick back says, "I've been pulling back the whole time!" which makes things click for the captain--but by then they were out of altitude, and hence out of time, and a perfectly functional airliner with 228 souls on board bellyflopped into the ocean at an 11,000 feet-per-minute rate of descent.
I suppose I shouldn't read about these things with 8 (or more) commercial airliner flights in my immediate future. But these are the extreme exceptions, so I don't fret too much. :)
But the darker side of it fascinates me too--aviation disasters. And especially the ones that simply shouldn't have happened; the accidents that were purely human error, no mechanical problems or outside interference. They're usually a series of bad decisions, overlooked information, and just sheer bad luck. Tenerife in 1977, when one 747 erroneously crossed a runway right in front of another 747 on its takeoff roll. San Diego in 1978 (my 3rd birthday, that one was), when a 727 pancaked down onto a Cessna in midair and went down into a residential area. Terrible, but fascinating to me.
Yesterday I read this Vanity Fair article about the crash of Air France 447 in 2009, which I'd not really read much about before. It's a long piece, but extremely interesting to me, and offers some very interesting insights into the nature of modern commercial aviation and the most likely causes of air crashes these days. The thing is, the planes themselves have become so reliable from a mechanical perspective, and so automated from an operational perspective, that pilots spend the vast majority of their flying time just watching the plane fly itself. The flight plan is pre-programmed for them on the ground, so all the pilots have to do is take off and climb out, and from there the plane can do the rest. The result is that many pilots these days have no real stick-and-rudder experience since everything's always been automated for them, and so when and if something out of the ordinary happens--something that makes the aircraft's computers throw up their hands and hand things back over to the human pilots--those human pilots are all too often at a loss to determine what has happened and what to do.
Air France 447 was a mystery for quite some time. The plane essentially disappeared over the Atlantic on a flight from Rio to Paris, and the pilots had never indicated that there was any trouble, never made any distress calls. Several days later some floating wreckage was discovered, and seemed to indicate that the aircraft was at least mostly intact when it hit the water. But the black boxes were nowhere to be found, so the mystery remained: what happened? Finally in 2011, almost 2 years after the crash, the fuselage of the plane was located at the bottom of the ocean, and the black boxes were recovered. And what was discovered in the flight and voice recorder data was astounding.
As the flight passed into a region of thunderstorms at cruising altitude, ice buildup caused the pitot tubes--externally-mounted tubes responsible for measuring airspeed--to clog, giving the onboard computers no idea of the aircraft's velocity. At that point, two things happened: the aircraft disengaged autopilot to give the pilots control, and also switched to a different set of flight parameters. Under normal circumstances, a modern, computer-assisted aircraft will simply not allow the pilots to make control inputs that will take the aircraft out of its safe flight envelope. It'll basically say "I'm sorry Dave, I can't do that," and remain in safe flight. But the mode Flight 447 switched into removed those restrictions, and it's quite possible--nay, quite likely--that the pilot at the controls (actually a copilot, the captain having excused himself for a nap due to being up too late partying with two women the night before) had no idea that was the case, and no experience flying outside of the usual Airliners-For-Dummies mode.
Now, it must be noted that nothing at all was actually wrong with the aircraft at this point. All the pilots had to do was absolutely nothing, and the plane would have continued cruising in level flight, eventually the pitot tubes would have thawed and unclogged, the autopilot would have re-engaged, and they'd have been in Paris by morning. But for whatever reason--subconscious panic and brainfreeze, possibly--the pilot's instinct was to climb (even though they were already about as high as the plane could go in the current environment), so he pulled back on his control stick, which raised the nose of the aircraft and the angle of attack of the wings. And he kept on pulling back until the plane began to enter an aerodynamic stall, which caused buffeting and turbulence--and caused it to begin losing altitude as the wings stopped generating sufficient lift.
Now here's where my fascination really comes in. In many aircraft with dual controls, when one control is moved, its counterpart moves as well, which gives the copilot a visual indication of what's happening with the controls. Not so in Airbus aircraft like AF447--the controls are basically a pair of joysticks off to each side, and they are not linked. So for the next 3 minutes or so, as AF447 was falling out of the sky, the pilot at the controls was holding his stick back, keeping the nose lifted, keeping the plane in a stall, and the other copilot (and the captain, once he came back into the cockpit) had no idea because the control sticks weren't linked and the pilot pulling back didn't think to mention it (if he even realized he was doing it) until past the point of recoverability. And there's the real rub. Aerodynamic stalls are Flying 101: something every pilot should know about since day one, as well as how to get out of one. If you have the altitude to spare (and AF447 did, at first), you simply lower the nose and dive into the stall, which will restore airflow and lift, and then you pull up out of the dive. It's as basic as aeronautics gets. But these 3 pilots, being so unused to anything out of the ordinary, and so out of touch with actually hands-on flying the aircraft, failed to recognize that they were in a stall. Even despite the aircraft itself sounding a stall warning (quite literally a voice saying "Stall!") no less than 75 times during the crisis. They may have simply blocked the warnings out, or they may have assumed that an actual stall was impossible--since usually it is--and written it off as a symptom of the lack of airspeed data. So finally the copilot yanking his stick back says, "I've been pulling back the whole time!" which makes things click for the captain--but by then they were out of altitude, and hence out of time, and a perfectly functional airliner with 228 souls on board bellyflopped into the ocean at an 11,000 feet-per-minute rate of descent.
I suppose I shouldn't read about these things with 8 (or more) commercial airliner flights in my immediate future. But these are the extreme exceptions, so I don't fret too much. :)
