Announcement

Collapse
No announcement yet.

Volaris A321 loses both ELACS

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

  • #76
    Originally posted by Evan View Post
    Well, as I said, yes, if the crew needed to take over it would then be in Alternate law, but the difference in control law is irrelevant to autoflight. If the autoflight can get along in an UAS mode until the speeds return there is no reason to degrade from Normal law. In other words, the mode reversion should only occur if the autoflight is disconnected before the speeds return to agreement.
    Evan,

    Even in an Airbus, with it's advanced FBW and very high systems integration, there is a clear distinction between the flight controls behavior and the autoflight system.

    The control law is independent of who is flying the plane, whether it's the PIC, the FO or R2D2.

    And the envelope protections, if the conditions are met, will kick in regardless of who is flying the plane. While it is harder to imagine how the envelope protection may be needed when in autoflight, it can happen due to system failures, weather and even other reasons like the pilots telling the autoflight to do something that the envelope protections would rather avoid (maybe this would not happen in the middle of a UAS event, but I am talking of the general concept here).

    You cannot tie the reversion of control laws to whether the AP and AT are both engaged or not.

    So if the flight control transfer function being enforced is the same than in normal law but you have degraded envelope protections, then you are in effect in alternate law. And I didn't mention the status of the autoflight in this sentence.

    --- 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


    • #77
      As I read about all these computer failures, and the serious implications of them, I'm wondering if the A319/20/21 needs to be on my no-fly list?
      Les règles de l'aviation de base découragent de longues périodes de dur tirer vers le haut.

      Comment


      • #78
        Originally posted by Gabriel View Post
        The control law is independent of who is flying the plane, whether it's the PIC, the FO or R2D2.

        And the envelope protections, if the conditions are met, will kick in regardless of who is flying the plane. While it is harder to imagine how the envelope protection may be needed when in autoflight, it can happen due to system failures, weather and even other reasons like the pilots telling the autoflight to do something that the envelope protections would rather avoid (maybe this would not happen in the middle of a UAS event, but I am talking of the general concept here).

        You cannot tie the reversion of control laws to whether the AP and AT are both engaged or not.
        Now you've lost me again. Gabriel, the SEC's and ELAC's limit what the autopilot can command. It cannot command anything that would trigger the various protections of the control laws, even if the pilots ask it to. Aside from that, control laws are a relationship between stick deflection and control surface movements. They are irrelevant to autoflight.

        Comment


        • #79
          Originally posted by 3WE View Post
          As I read about all these computer failures, and the serious implications of them, I'm wondering if the A319/20/21 needs to be on my no-fly list?
          You should add the 330/340/350/380 to that list, too. Not to mention the 318/ACJ, should you run into one of them by some chance.

          Comment


          • #80
            Originally posted by ATLcrew View Post
            ...(319/320/321/)330/340/350/380...Not to mention the 318/ACJ...


            Originally posted by Another ATP sort of
            "I think, based on the types of aircraft listed, you're pretty much guaranteed not a fiery death, but a long, slow, out-of-control, terrifying, tumbling fall from flight level something something oh to a smoking hole in the ground."
            Les règles de l'aviation de base découragent de longues périodes de dur tirer vers le haut.

            Comment


            • #81
              My understanding (which can be wrong because it is mostly based on industry practices, not Airbus specific)

              Originally posted by Evan View Post
              Now you've lost me again. Gabriel, the SEC's and ELAC's limit what the autopilot can command.
              Yes, in the same way that they will limit what the human pilot can command, or mre accurately, what they (the SECs and ELACs) will do with what the pilot (human or otto) commands.

              It cannot command anything that would trigger the various protections of the control laws, even if the pilots ask it to.
              Are you sure?

              Aside from that, control laws are a relationship between stick deflection and control surface movements. They are irrelevant to autoflight.
              That's not correct. Pitch-wise, the only difference between normal law and alternate law (standard or abnormal) is the protections, because the transfer function remains stick-on-G / stick-on-pitch rate.

              If Airbus follows the typical design philosohy for autopilot, we can think in three systems:

              One system (the MCP or the FMS) has the desried performance (example: track the glide slope).
              Another system, the flight director, tells what is the required maneuver to achieve the desired performance (example, pitch up 2 deg).

              The third part is the pilot who provides the required control inputs to perform the maneuver indicated by the flight director.
              These first two parts are independent of the status of the autopilot (on or off) and a human pilot can provide the control inputs to follow the commands of the flight director displayed in the PFD. Or it can be the autopilot doing exactly the same.

              In a classic (non FBW) plane, the autopilot has servos that do the mechanical control inputs exactly as a human pilot would (ok, not exactly: more accurately!).

              You mentioned before the "virtual TLA". In the Airbus (or Boeing 777) I would expect the AP to have a virtual flightstick (or virtual yoke), providing to the flight control computers (SECs and ELACs) the same kind of electronic control input than a pilot would.

              If you are in an Airbus, select TOGA with the thrust levers and a VSI descent of 5000 fpm in the MCP, I would expect that the flight director will indicate what pitch is necessary to establish -5000 fpm, the autopilot would move its virtual sidestick providing electronic inputs to the SECs and ELACs, and the SECs and ELACs will comply with this input which will result in the plane pitching down to the angle indicated by the flight director and descending at 5000fpm... until the overspeed protections kick in and the SECs and ELACs overrides the pilot inputs (human or, in this case, automatic) and pitch up.

              --- 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


              • #82
                Originally posted by Gabriel View Post
                That's not correct. Pitch-wise, the only difference between normal law and alternate law (standard or abnormal) is the protections, because the transfer function remains stick-on-G / stick-on-pitch rate.
                No, that is correct. The difference between Normal and Alternate law is still the relationship between stick deflection and control surface movement. For example, In Normal law, exceeding alpha prot, the stick becomes proportional to angle-of-attack until the AoA drops below alpha prot again. Another example is that in Alternate law, control law changes to Direct when the gear is extended, but only in manual flight. If the A/P is in use, the control law remains Alternate after gear deployment and only reverts to Direct when the AP is disconnected.

                But in any case control law is only relevant to manual flight control. In autoflight, the FMGC (essentially the autoflight brain) is getting speed envelope data directly from the FAC's (flight augmentation computers) as well as air and IR data from the ADIRU's, the SEC's and ELAC's enforce limitations, so it is in itself a form of envelope protection. We're not talking about some primitive autopilot here.

                Control law concerns manual flight. Once you remove the stick from the equation, it is irrelevant. Once you regain autoflight, the danger is abated until you disengage it again.

                Comment


                • #83
                  -Three questions:

                  1- What happens in the scenario I presented above (normal law)?
                  2- You are flying an apporach on AP, get hit by a wind shear / microbrust and AoA exceeds alpha prot. Do the AoA protection kick in and give the relevant ECAM messages (normal law)?
                  3- Same as 2 but in alternate law.

                  These are honest questions, I don't know what happens, I've already said what I believe would happen.

                  --- 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


                  • #84
                    Originally posted by Evan View Post
                    No, that is correct.
                    Evan, you are saying the same than me here: That, except for the envelope protections (alpha prot is an nevelope protection, hence the "prot" in there), the relationship between stick deflection and elevator movement is the same in Normal and 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


                    • #85
                      Originally posted by Gabriel View Post
                      1- What happens in the scenario I presented above (normal law)?
                      2- You are flying an apporach on AP, get hit by a wind shear / microbrust and AoA exceeds alpha prot. Do the AoA protection kick in and give the relevant ECAM messages (normal law)?
                      3- Same as 2 but in alternate law.

                      These are honest questions, I don't know what happens, I've already said what I believe would happen.
                      Autoflight has a mission to stay in the envelope (to keep the aircraft in stabilized flight). It knows the AoA. It's going to do everything possible to keep it healthy and nothing to exceed it. If it can't it gives up and hands control to the pilots (it will disengage when AoA exceeds alpha prot). And, as you know, there are certain extreme circumstances where alpha max can be exceeded momentarily even under Normal law and perhaps under automation. Normal and Alternate law in FBW are a blend of automation and manual flight. The blend is in the way the aircraft protects itself from the pilot and the difference between Normal and Alternate law is the degree to which the automation does that. Hence, control laws are specifically for manual flight.

                      Originally posted by Gabriel View Post
                      Evan, you are saying the same than me here: That, except for the envelope protections (alpha prot is an nevelope protection, hence the "prot" in there), the relationship between stick deflection and elevator movement is the same in Normal and Alternate law.
                      Yes, the difference is only in the level of blended automation. But this has nothing to do with full autoflight.

                      Comment


                      • #86
                        I can sum it up this this way:

                        I feel very safe flying with any reputable operator an the A320/30/40/80/50 as long as either a) Normal law is in force or; b) flying in Alternate law when autoflight is engaged. Either way, I am protected from hand-flying error. I feel far less safe in an aircraft without cockpit trim forces or envelope protections in manual flight. Therefore I always want the industry to minimize the conditions that prevent autoflight engagement in Alternate law. They have mostly done this by adding redundancy, so it mainly occurs when multiple redundant systems fail (multiple computer failures, multiple hydraulic system failures, etc.) and the likelihood of such failures is very remote UNLESS a common point of failure exists. UAS is a good example (the environment: the only current defense is to avoid threatening weather systems or to ride it out). Structural damage from uncontained engine failure is another (proximity: some systems need to be in close proximity to one another and a rotor blast radius can be quite large). But common design vulnerabilities within a redundant system should not be tolerated and removed wherever possible.

                        If it turns out that a common point-of-failure exists between the two ELAC's, I hope that a service bulletin or AD will follow. I also dislike the idea of dispatching an Airbus in a condition where a single system failure can result in the elevated danger of Alternate law without autoflight. This should also be the trade-off for the lack of trim-force feedback in manual flight: a stricter MEL requirement for the automation.

                        Comment


                        • #87
                          Originally posted by Evan View Post
                          A pilot can be a back-up or a fuck-up
                          Best flight school t-shirt slogan ever!
                          Be alert! America needs more lerts.

                          Eric Law

                          Comment


                          • #88
                            Originally posted by Gabriel View Post
                            - You could keep the degraded slow speed protection of the standard alternate law using the AoA vanes (if they agree). Yes, stall AoA depends on Match, and Match depends on speed, but you can select a reasonable fixed AoA trigger value that will pitch down slightly faster than required at low Match, and still prevent a stall at high Match.
                            I don't know if you remember this but Airbus offered an optional Back-Up Speed Scale function at the time AF447 dissappeared (I think it is now a standard feature). It uses the AoA vanes but cannot accurately infer airspeed. Instead it displays a green range to stay within. Apparently it works pretty well. Not accurate enough for envelope protections I guess....

                            Still, I think the autopilot could use this as well during a temporary UAS mode.

                            Comment


                            • #89
                              I just hope that the warning system that tells pilots when the ELACs fail doesn't fail.
                              Les règles de l'aviation de base découragent de longues périodes de dur tirer vers le haut.

                              Comment


                              • #90
                                Originally posted by Evan View Post
                                I don't know if you remember this but Airbus offered an optional Back-Up Speed Scale function at the time AF447 dissappeared (I think it is now a standard feature). It uses the AoA vanes but cannot accurately infer airspeed. Instead it displays a green range to stay within. Apparently it works pretty well. Not accurate enough for envelope protections I guess....
                                On the contrary, it is not accurate enough to hold speed, because at high speed significant speed variations require small changes in AoA to keep 1G.

                                But it is perfect for AoA protection. That's what the stall warning is based on even in direct law or in no law (mechanical backup). Just put an alpha thrshold short of the stall warning where the high AoA protection kicks in a similar style than the slow speed protection in standard 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

                                Working...
                                X