Comments

DFWCapt, I am nominally not as qualified to comment as you. Remove a couple of zeros from your TT and that's still more than I have. The most engines installed in an airplane I flew is one, and one with four little pistons. And the most automation that the airplane had was an auto-squelch. That said, I am more qualified to comment on AF447 than you simply because I have read BEA's third interim report that includes the full CVR transcripts and FDR plots, something that you obviously haven't. You'll find the link at the end of this post, but let me summarize: 1) Due to an unreliable airspeed event, the AP and AT self-disconnected and the flight control law reverted to alternate law. 2) The PF took control of the plane, COMMANDING a pull-up so "intense" that resulted in two brief activations of the stall warning, more than 1.4Gs, more than 10° of ANU attitude, and more than 7000fpm. That, in an airplane that was very heavy and near it's ceiling for the conditions (heavy weight and temp above ISA). 3) Additionally, he did so without adding an inch of thrust, which had been previously reduced a bit from the cruise setting by the autothrottle in response to the pilots' command on the speed setting to slow down from 0.82 M to 0.80M, in preparation for turbulence penetration. 4) Of course, this climb is not sustainable even at low altitude, with less weight, and with TOGA, let alone a heavy plane at 35,000ft and reduced thrust. 5) The steep climb started to smooth down due to, in part, the huge loss of speed and in part due to the pilot application of nose-down stick input. But it was too late. The plane had climbed too much (to 37,500ft) and lost too much speed. 6) It is interesting to note that, during this initial sequence, the pilots had correctly recognized the disconnection of the AP and AT, the unreliable airspeed, and the reversion to alternate law. 7) At this point, the stall warning started again. The pilot responded to that with application TOGA (finally) and... PULLING UP HARD AGAIN!!! The airplane had still enough margin over stall that the nose rose again and the plane reached 380,000ft before starting to descend. 8) The stall warning continued to sound uninterruptedly for almost one full minute, during which the pilot made mostly nose-up stick inputs. The description above covers more than the first one and a half minute of the event. It is interesting to note that during that time: 1) The speed was unreliable but everything else worked Ok, including three attitude indicators, three altimeters, two vertical speed indicators, and all the engine instruments. 2) By the then current procedures, the pilots were required to respond to the correctly identified UAS event simply with AP/AT/FD off (two of them had already happened by themselves), thrust to CLB and pitch 5° ANU. Of the three things they had to do, they did none. Also, the pilots were required to respond to the stall warning with TOGA and a reduction in pitch. They did the first but not the second (which is the most crucial, since the stall is a matter of AoA, you can recover from a stall with no thrust but you can't recover without reducing the AoA). 3) There was no overspeed warning, true or false. 4) There was no false stall warning and, at least during this first more than one minute and a half, no false stall warning silence. There was, however, nearly one full minute of correct stall warning. 5) The plane was in alternate law which is the same than direct law except for the loss of envelope protection (not a minor issue, I agree, especially since low speed and stall protections are among them): It's a blend between load factor demand and pitch rate demand. Except for said envelope protections, the reaction of the plane to a given stick input is the same regardless of whether they are at low altitude or up high, in normal law or alternate law. The FCCs will take care of applying the needed elevator deflection to get the commanded load factor or pitch rate. And regarding the loss of protections, they knew they wee in alternate law so they knew (or should have known) that the stall protection was lost, that the plane could stall, and that pulling up and keeping pulling up is a damn good way to stall a plane (A330 or J3) 6) Everything the plane did was COMMANDED BY THE PILOTS, not the result of atmospheric disturbances or of the computer responding to spooky data. So we really know what they did and what they should have done. The only question is why they did what they did instead of what they should have done. And thee is one word for it: Training. Starting from PPL.