Announcement

Collapse
No announcement yet.

Air France 447 - On topic only!

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Originally posted by Evan View Post
    ...I don't buy it...
    In cases of pilots doing crazy unthinkable illogical and wrong things you have to buy in a little bit...because with any of these theories, it is easy to say why every theory is wrong...

    That theory that the pilot was saying, "oh crap, we lost airspeed and I don't want to overspeed, so let me simply pull back and let HAL's stall protection give me a good slow speed"...

    And I guess I'd also say that people temporarily being overwhelmed and getting stupid happens...and the events AFTER they stalled but were seemingly unaware that they had stalled (the attitude was halfway normal!).

    I guess when a computer control system is designed to make an airplane act NOT like an airplane goes haywire you can't be sure what you are seeing or what to do to help the situation.

    For a computer system that helps an airplane act like an airplane- and it goes haywire- the human beings still know to treat it like an airplane.
    Les règles de l'aviation de base découragent de longues périodes de dur tirer vers le haut.

    Comment


    • Originally posted by 3WE View Post
      I guess when a computer control system is designed to make an airplane act NOT like an airplane goes haywire...
      Right. I mean look at this thing: The pilot moves the stick aft and the crazy haywire computer makes the nose go up. HAYWIRE! And then he moves the stick forward and, get this, the nose goes down! Side-to-side, it results in roll! They firewall the thrust, they get full thrust of all things. In fact, every command they made resulted in a control surface movement corresponding to the command itself. How is any pilot supposed to fly this death trap! This is not how an airplane is supposed to act!

      Remind me again how an airplane is supposed to act...

      Comment


      • You just did a good job of illustrating my point!

        You correctly point out that during most of the accident sequence, the airplane responded to the controls the way a non-Airbus-FBW airplane would be expected to respond to the controls.

        However the plane in question is an Airbus with FBW controls. And the pilots had been extensively trained that that airplane did NOT fly like an aircraft with mechanical controls. Instead, they had it drummed into them that the stick and other controls were inputs to a computer, and the computer was tasked with interpreting the pilots' inputs and manipulating flight controls to achieve the desired flightpath. And I'll speculate that 100%, or extremely close to 100% of their experience flying that aircraft model was with the flight controls operating that way (ie not in direct law).

        So while I think it's quite fair to criticize the pilot(s) for not recognizing that the flight control mode had changed and the aircraft was stalled, IMHO I think the explanation goes deeper than that they were just dummies. The pilots' behavior represented not just their individual characteristics, but also their training and experience.
        Be alert! America needs more lerts.

        Eric Law

        Comment


        • Originally posted by Evan View Post
          Right. I mean look at this thing: The pilot moves the stick aft and the crazy haywire computer makes the nose go up.

          [more BS deleted]

          Remind me again how an airplane is supposed to act...

          Indeed
          you do need a reminder.

          ...as the nose goes up and speed slows and the nose goes back down again, as opposed to staying there frozen nose up without further stick input.

          Ask Gabriel to further describe phugoid behavior and the tendency for a trimmed traditional aircraft to maintain speed all by itself- whether it's a heavily computerized 777 or a Piper Cub.

          Sure, you're right, this is all trivial, closed-mided, traditional seat of the pants stuff until, I dunno...maybe you loose your speed and some other important indicators due to ice over the ocean on a dark night in an Airbus?

          ...and Eric's comments on how you have been trained and have flown for the past 10/whatever years versus how the controls are working on this one strange night are also very relevant.
          Les règles de l'aviation de base découragent de longues périodes de dur tirer vers le haut.

          Comment


          • Originally posted by elaw View Post
            However the plane in question is an Airbus with FBW controls. And the pilots had been extensively trained that that airplane did NOT fly like an aircraft with mechanical controls. Instead, they had it drummed into them that the stick and other controls were inputs to a computer, and the computer was tasked with interpreting the pilots' inputs and manipulating flight controls to achieve the desired flightpath. And I'll speculate that 100%, or extremely close to 100% of their experience flying that aircraft model was with the flight controls operating that way (ie not in direct law).
            The aircraft was in Alternate Law, not Direct Law, throughout the descent. Alternate Law still uses the C* method of interpreting pilot commands, as in Normal Law. The main difference is that the envelope protections are lost. Sidestick coordination is unaffected.

            Not that is would have mattered. The inputs were not subtle.

            Comment


            • Originally posted by 3WE View Post

              Indeed
              you do need a reminder.

              ...as the nose goes up and speed slows and the nose goes back down again, as opposed to staying there frozen nose up without further stick input.
              Wrong. If you pull a traditional yoke back and hold it there, the nose does not go back down. The out-of-trim condition just makes it increasingly harder to keep it up there. Only if the yoke is released does it have this longitudinal stability effect. Pilot forces (intent) are the issue here, not trim forces.

              This pilot pulled back and held it there. And so the nose went up and stayed up, until he pushed it forward, whereupon it went down again. Like a real airplane.

              This unstable Airbus argument gets really tiring. You have to think of it this way:

              Physical Flying................Sidestick Manual Flying................Automated Flying

              Ok, Yoke flying is physical. You feel the out-of-trim physically and use your physical senses intuitively. This is the traditional way of flying and I do think Boeing is wise to simulate this in their FBW. But pilots should also be able to evolve to the current technology of...

              ...sidestick manual flight, which, in the case of Airbus, provides no physical feedback for pitch or roll (relying instead on envelope protections). If a pilot is actually flying the plane, he doesn't need trim-force feedback, and relieving him of this outdated need to retrim was the idea behind the newer method. True, in a scenario like this, with those protections degraded, it is a more dangerous regime, which is why pilots have to be well-trained on how to fly an Airbus jet manually and paying attention to the instruments.

              Automated flight is, of course, the same regardless of the means of control.

              Now, this plane stalled because the pilot (for whatever reason) maintained a climb attitude despite stall warning. If he was flying a 767 I imagine he would have done the same despite trim forces (maybe even retrimming them). And for the last time the FBW did not go haywire or fail in any way.

              So... what's your point? Airbus FBW is unsafe? Rubbish!

              AF447 crashed because the pilots were not well-trained on how to fly an Airbus jet manually and paying attention to the instruments. Or each other. It was and will always be 100% Air France's fault for that deficiency.

              Comment


              • Originally posted by elaw View Post
                You just did a good job of illustrating my point!

                You correctly point out that during most of the accident sequence, the airplane responded to the controls the way a non-Airbus-FBW airplane would be expected to respond to the controls.

                However the plane in question is an Airbus with FBW controls. And the pilots had been extensively trained that that airplane did NOT fly like an aircraft with mechanical controls. Instead, they had it drummed into them that the stick and other controls were inputs to a computer, and the computer was tasked with interpreting the pilots' inputs and manipulating flight controls to achieve the desired flightpath. And I'll speculate that 100%, or extremely close to 100% of their experience flying that aircraft model was with the flight controls operating that way (ie not in direct law).
                In fact, you are wrong (almost).

                The plane was not in direct law, but in alternate law.
                The control algorithm for pitch is the same in normal and alternate law: Stick deflection is proportional to load factor (G) at high speeds, and to pitch rate (deg/sec) at low speed.

                The only difference is that most protections are lost, including all the low speed / stall protections (Alpha floor, alpha prot, alpha max). Instead, a stall warning adivises the pilot when the AoA approaches the critical value.

                So he didn't find himself flying a plane that would respond to his stick inputs in a different way than it ususally did, except that it would not forbid the pilot to pull up past the stall AoA.

                --- Judge what is said by the merits of what is said, not by the credentials of who said it. ---
                --- Defend what you say with arguments, not by imposing your credentials ---

                Comment


                • 3we, Evan, I think both of you are right and wrong.

                  In a traditional plane (or one FBW that simulates the traditional plane), if you pull up the nose will go up, will overshoot the equilibrium (trim condition), will start to lose speed, and then the nose will come down again. Of course, if the pull-up is excessive, the nose-down will come with the stall.

                  In an Airbus, even if you pull up just a bit, the nose will start to go up and keep going up. The computer will automatically add more and more elevator as speed goes down to keep the plane going up (even with the stick just a bit back) and, once consumes all the elevator travel, it will add more and more nose-up trim. This airbus was basically with the elevator and the stabilizer (trim) fully nose-up to the stops for most of the fall.

                  Of course, in this particular case this would not make a difference.
                  When the first interim report came out after the download of the CVR and FDR, I said that what this pilots did would have equally swiftly stalled an Airbus, a 777, a DC-8 or a Cessna 152.

                  So, ok, Airbus compensates the lack of speed and AoA stability with the envelope protections. That's not so fun (no more phugoids) but I am ok with that.

                  What I find unacceptable, even if in this particular accident it was not relevant and no accident so far happened because of this, is that they have an "intermediate" law, alternate law, that lacks BOTH speed/AoA stability and envelope protections. If I was an Airbius pilot maybe I would thing differently, but if you ask me now today, if I lose normal law I will push the buttons and turn the knobs as necessary to go to direct law (or press the "gimme my frigging plane back" button, if there was one).

                  I really don't even understand how they got away with that through certification. The standards clearly require a given minimum "pound per knot" gradient in the elevator. I understand that Airbus, that offers ZERO pond per knot, might have offered the envelope protections as an acceptable alternate means of compliance (AMOC). But the alternate law?

                  --- Judge what is said by the merits of what is said, not by the credentials of who said it. ---
                  --- Defend what you say with arguments, not by imposing your credentials ---

                  Comment


                  • Originally posted by Gabriel View Post
                    3we, Evan, I think both of you are right and wrong.

                    In a traditional plane (or one FBW that simulates the traditional plane), if you pull up the nose will go up, will overshoot the equilibrium (trim condition), will start to lose speed, and then the nose will come down again. Of course, if the pull-up is excessive, the nose-down will come with the stall.
                    By 'keeping it up there' I meant maintaining the target pitch on the artificial horizon (the intent). In traditional control, it's going to stay where you want it by force if you are flying the plane regardless of trim condition. My point is that longitudinal speed stability is seen as a safeguard against not flying the plane.

                    Originally posted by Gabriel
                    I really don't even understand how they got away with that through certification. The standards clearly require a given minimum "pound per knot" gradient in the elevator. I understand that Airbus, that offers ZERO pond per knot, might have offered the envelope protections as an acceptable alternate means of compliance (AMOC). But the alternate law?
                    Alternate law still retains artificial high and low speed stability. The difference is that it can now be overridden by the pilot with increased input (with stubborn intent, just like traditional flying). AIrbus provided Alternate Law only as a control law reversion for situations where significant redundant system failures have made envelope protection logic unreliable. As a provisional control law it requires close, disciplined attention of the flight crew. I think it is reasonable for Airbus to expect flight crews experiencing such failure conditions to be focused on flying within the safe envelope (including automated flight) as this is an EMERGENCY condition following a MASTER CAUTION and a ALTN LAW:PROT LOST 'wake up' call. The only dangers I see in having neither protections nor traditional speed stability 'feel' reside in either pilot complacency or pilot ineptitude. Neither of these things should ever be expected to exist in the cockpit following an Alternate Law reversion. Alternate Law is not a routine flight mode; once Normal Law is lost, Airbus expects a crew to get the airplane down as carefully and quickly as possible and to fly with extreme vigilance until that is possible.

                    DIrect law is probably more dangerous because a sidestick is not a good replacement for a large yoke in proportional pitch control. Small inputs are far more sensitive. It simply wasn't designed to be used this way. Direct law in pitch exists only out of necessity in a more worse-case scenario. However, under Alternate law, once you are established on final and have dropped the gear, you ARE in Direct Law (because the Normal Law transition to Direct Law in the flare is no longer available). So in the most critical phase of flight, you have DIRECT proportionate control and you BETTER have the skills training to fly this way.

                    One other thing: if an inept pilot gets into an unsual attitude in Alternate law (as in this case), the high and low speed stability are suspended (control law becomes Abnormal Alternate Law) to facilitate recovery. Of course this assumes a pilot will know and be applying the correct recovery procedure. This is the ONLY scenario that matches your description of Alternate law and requires a very poorly trained pilot to ever get there and a very well-trained pilot to get out of there. It is a scenario where traditional speed stability is of no use anyway.

                    Air France didn't provide this training. Their pilots were only skilled in routine flight and marginally trained for upsets. They lacked real practiced skills for flying the A330 in provisional flight modes and for recovering from unusual attitudes. They had no discipline for CRM or procedures. Add these skills back into the equation and then tell me how dangerous Alternate law is.

                    What's dangerous here is the lack of training, regardless of what you're flying.

                    Comment


                    • Originally posted by Evan View Post
                      Wrong. If you pull a traditional yoke back and hold it there, the nose does not go back down.
                      Very cool.

                      Clearly you have never flown a traditional plane.

                      The bold makes it so much better.
                      Les règles de l'aviation de base découragent de longues périodes de dur tirer vers le haut.

                      Comment


                      • Originally posted by Evan View Post
                        Wrong. If you pull a traditional yoke back and hold [your intended target pitch] there, the nose does not go back down [before the stall].
                        I forgot how exhaustively specific I have to be here.

                        Comment


                        • Originally posted by Gabriel View Post
                          In fact, you are wrong (almost).

                          The plane was not in direct law, but in alternate law.
                          Doh! Clearly I got a wrong impression somewhere along the line. Foot now lodged firmly in mouth...
                          Be alert! America needs more lerts.

                          Eric Law

                          Comment


                          • Originally posted by Evan View Post
                            Alternate law still retains artificial low speed stability.
                            Could you provide a source please?
                            Not that I don't believe you, but I want to investigate more.

                            --- Judge what is said by the merits of what is said, not by the credentials of who said it. ---
                            --- Defend what you say with arguments, not by imposing your credentials ---

                            Comment


                            • Originally posted by Gabriel View Post
                              Could you provide a source please?
                              Not that I don't believe you, but I want to investigate more.
                              Thjis link shows everything at a glance:

                              Comment


                              • Originally posted by Evan View Post
                                Thjis link shows everything at a glance:
                                I lurk quite a bit on here-- and rarely post. This has been an interesting f/up discussion... anyway.

                                According to your link. When in "Alt Law"

                                A low speed stability function replaces the normal angle-of-attack protection
                                ◦This command CAN be overridden by sidestick input. ---> which he was
                                ◦The airplane CAN be stalled in Alternate Law. ---> which it was
                                ◦An audio stall warning consisting of "crickets" and a "STALL" aural message is activated. ---> which occured
                                ◦The Alpha Floor function is inoperative.

                                My "speculation" is that the PF didn't realize he was in Alt Law and figured the protections where still in place (among other things). Hence the near constant pulling back on the stick.

                                Pitch looked like it was around 10 deg for quite a bit and higher at times.

                                I noticed that throttles were at near idle (or at idle) for a little bit there, then pushed all the way up (with the nose still up in the air).

                                Were they so confused that not even the alt dropping like crazy didn't give them a clue that they were indeed stalled?

                                Ok-- dumb question now. Don't they have GPS on those planes and given what was going on couldn't it have helped a bit w.r.t. how fast they were going (yeah, I know it would have been ground speed-- but still better than nothing)

                                EDIT:
                                Net-net-- I would have thought that one of the 3 in the cockpit at that point would have said follow the procedure (bunch of thrust and a little nose up, after the stall recovery of course).

                                Comment

                                Working...
                                X