Announcement

Collapse
No announcement yet.

MAX takes to the skies again

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

  • #76
    Originally posted by Evan View Post
    Ok, hold on, this is not how aviation safety got so good. First there is design and certification. Then there is operational safety.
    Again, I disagree. I mean, I agree with the above timeline-wise (first comes dosing and certification, then operation). But NOT as a hierarchy.

    The pilots cannot be considered the third line of defense for design and certification safety
    Of course they can, they have, they are and they will be considered the last level of defense for any technical or operational issue.

    When the crew boards a revenue flight, they need to be boarding an aircraft that has already been made (and proven) safe by multiple layers of defense.
    I agree, but nothing is perfect, including design and certification.

    Look, you are getting quite black and white again here. Let's start with the ideal: Planes that are designed and certified correctly as passing all the requirements and have no known issues that would made them not-certifiable.

    Safety is not an absolute. There are limitations on how safe something can be made and these limitations have different sources, can be practical, technological, economic... Things can ALWAYS be made safer.

    It is not like something that barely meets the certification standards is suddenly safe and something that almost meets it is totally unsafe. It's not like a design load of 2.5G is a hard limit, where a plane that has a design load of 2.4G will break at the first sign of turbulence and one that has 2.6G is indestructible. Certification standards are in this sense arbitrary and have been becoming more strict as technology gets more available, small, light and affordable.

    We have known issues that are allowed by certification standards, unknown issues not known or foreseen at the time of certification, and MCAS that is totally different kind of animal.
    But not only that: A plane that is well designed and certified may encounter safety issues that pilots may need to address.
    Maintenance errors, flight planning and dispatch planning errors, ATC errors, errors of other pilots, unforeseen natural situations (ranging from forecasted weather to birds).

    Some examples:

    The A300 has a rudder control design where to make a full rudder deflection you need a relatively small force (on top of the force that is needed to break the friction and initiate any movement) and a very small displacement. This is still within certification standards but makes the plane more susceptible than most to overcontrol, PIO and hence rudder reversals.

    The 737 rudder hardover was not known at the time of certification, and it was a failure mode that would have been very hard to identify before it actually happened.
    Yet, when found, the fleet was not grounded until all rudders were retrofitted. Pilots were trained on how to deal with the situation. At that point you had a known design issue that was left to the pilots to address until the complete fleet was retrofitted, which took many years to complete.

    The A330 (and similarly in mostly any plane), upon a speed disagree it disconnects the autopilot and let the flight director give misleading commands. The pilot is expected to disconnect the AP/AT/FD, set climb thrust by moving the thrust levers out of the climb setting (and then back), and establish a pitch of 5 degrees. What of that cannot be done automatically? Couldn't the AP at least not disconnect and let the pilots disconnect it by themselves as to buy some valuable seconds of gaining situational awareness? This is perfectly within the certification standards and yet the AF 447 pilot could not cope with that. And you placed a lot of weight on the crew ability (or lack thereof) to cope with it, where it something that clearly (and somehow easily) could have been at least partially addressed with a better design.

    So let me ask you this: Why would you expect that the Ethiopian pilots would have reacted any different and not_crashed had the out-of-trim situation been caused by a garden-variety, non-MCAS related, trim runaway? I know, there were additional factors of distractions (like the stickshaker) but I don't feel any confident at all that they (or the LionAir crew) would have not been affected by similar startling, confusion, panic, wrong prioritization of tasks... pretty much like the AF 447 crew did in a "perfectly" designed, certified and flyable plane. And the Ethiopian crew (unlike the LionAIr one) had the benefit of the hindsight, from which they were not able to benefit.

    The comes the first line of operational defense: Familiarize pilots with the aircraft in its entirety and train them on how to deal with every system failure.
    Exactly, Ethiopian and the pilots of the accident flight failed at that, at least in a way that made a difference.

    --- 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
      Ok here's another. Back in the 90's there were two accidents in Boeing aircraft, I don't remember the model. One was out of the Caribbean and the other was off the west coast of South America. Both were caused by some bugs getting into the static system and the crews, who took off at night, couldn't fly with no airspeed and crashed. If I'm not mistaken I remember showing my crews in ground school a section in the 747 manual where Boeing addressed the loose or damage to the radome, the pitots are right in front of the windscreen. The section actually gave pitch attitudes at various flap settings. Then after these two accidents in my recurrent sim my instructor gave us a total loose of airspeed. We were sitting there flying level at that pitch attitude and I remember my instructor saying "don't stop thinking, what else have you got?" Then it popped into my head "INS ground speed" and we safely returned for the approach to land.

      Comment


      • #78
        Originally posted by kent olsen View Post
        Ok here's another. Back in the 90's there were two accidents in Boeing aircraft, I don't remember the model. One was out of the Caribbean and the other was off the west coast of South America. Both were caused by some bugs getting into the static system and the crews, who took off at night, couldn't fly with no airspeed and crashed. If I'm not mistaken I remember showing my crews in ground school a section in the 747 manual where Boeing addressed the loose or damage to the radome, the pitots are right in front of the windscreen. The section actually gave pitch attitudes at various flap settings. Then after these two accidents in my recurrent sim my instructor gave us a total loose of airspeed. We were sitting there flying level at that pitch attitude and I remember my instructor saying "don't stop thinking, what else have you got?" Then it popped into my head "INS ground speed" and we safely returned for the approach to land.
        They were both 75's

        Comment


        • #79
          Originally posted by BoeingBobby View Post

          They were both 75's
          One of them, I suppose, was Aeroperú Flight 603 where maintenance personnel put tape on all 3 static ports to wash the plane and then forgot to remove it.
          Which was the other one?

          In Aeroperú, the crew remained relatively calm and functioning as a well coordinated crew for most of the event that lasted almost 30 minutes (the communications with ATC and the CVR recordings are available on the internet).

          This is another example where the crew (and ATC) could have saved the day but made several mistakes.

          They identified that they had airspeed and altitude disagree, and yet...
          • They didn't use the: P+P=P concept (Pitch + Power = Performance). I've heard that the current UAS procedure with memory-item power and pitch and look-up tables did not exist at that time, but you should still know that if you have the throttles fully retarded, the speedbrakes fully extended, and a more-or-less level attitude, you are not accelerating towards an overspeed.
          • This is a serious one... How none of the 2 pilots or the controller realized that the altitude displayed in the ATC radar was conveyed by the mode C transponder in the airplane using the same unreliable air data?
          • At no point they attempt to use the radar altitude. Radar altitude is much more reliable and it is totally independent of any problem with any other system. If I have problems with the altimeter, radar altitude is the first thing I would go and look. Normally radar altitude is displayed at low altitude only (like below 2500 ft) unless you go dig in a pit of menus and setting in the computers. But they were flying below 2500 ft for quite a while and should have had the radar altitude displayed on the instruments.
          • If you know or strongly suspect issues with the air data which is indicating overspeed even when you have the throttles fully retarded, the speedbrakes fully extended, and a more-or-less level attitude, and you get a stall warning (stickshaker) and an overspeed warning, and need to bet your live on one, you better trust the stickshaker: You are not ovespeeding, you are close to stalling.
          • This is the most unforgivable one: When you receive low altitude alerts (caution, terrain, terrain, pull up), you better put full thrust and climb. Especially if you are in zero visibility so you cannot judge your separation from ground or terrain visually. And even more especially when you have been fighting with the air data system the whole flight and you don't rely on your altimeter. The GPWS (ground proximity waring system) is based on radio-altimeter information, which is a very reliable system and, more importantly, it is 100% independent of the air data system.

          --- 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


          • #80
            Originally posted by Gabriel View Post

            One of them, I suppose, was Aeroperú Flight 603 where maintenance personnel put tape on all 3 static ports to wash the plane and then forgot to remove it.
            Which was the other one?
            Birgenair 301.

            Comment


            • #81
              Originally posted by ATLcrew View Post

              Birgenair 301.
              Oh, that one. Thank you.

              Another one with poor crew performance that started when they did not abort the take-off after recognizing (at the 80 knots speed crosscheck) that there was a problem with the Captain's (PF) airspeed indication.

              This one was a problem with 1 of 3 pitot tubes / airspeed indicators. Doesn't even qualify as an unreliable speed (just a speed disagree).

              --- 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

                One of them, I suppose, was Aeroperú Flight 603 where maintenance personnel put tape on all 3 static ports to wash the plane and then forgot to remove it.
                Which was the other one?

                In Aeroperú, the crew remained relatively calm and functioning as a well coordinated crew for most of the event that lasted almost 30 minutes (the communications with ATC and the CVR recordings are available on the internet).

                This is another example where the crew (and ATC) could have saved the day but made several mistakes.

                They identified that they had airspeed and altitude disagree, and yet...
                • They didn't use the: P+P=P concept (Pitch + Power = Performance). I've heard that the current UAS procedure with memory-item power and pitch and look-up tables did not exist at that time, but you should still know that if you have the throttles fully retarded, the speedbrakes fully extended, and a more-or-less level attitude, you are not accelerating towards an overspeed.
                • This is a serious one... How none of the 2 pilots or the controller realized that the altitude displayed in the ATC radar was conveyed by the mode C transponder in the airplane using the same unreliable air data?
                • At no point they attempt to use the radar altitude. Radar altitude is much more reliable and it is totally independent of any problem with any other system. If I have problems with the altimeter, radar altitude is the first thing I would go and look. Normally radar altitude is displayed at low altitude only (like below 2500 ft) unless you go dig in a pit of menus and setting in the computers. But they were flying below 2500 ft for quite a while and should have had the radar altitude displayed on the instruments.
                • If you know or strongly suspect issues with the air data which is indicating overspeed even when you have the throttles fully retarded, the speedbrakes fully extended, and a more-or-less level attitude, and you get a stall warning (stickshaker) and an overspeed warning, and need to bet your live on one, you better trust the stickshaker: You are not ovespeeding, you are close to stalling.
                • This is the most unforgivable one: When you receive low altitude alerts (caution, terrain, terrain, pull up), you better put full thrust and climb. Especially if you are in zero visibility so you cannot judge your separation from ground or terrain visually. And even more especially when you have been fighting with the air data system the whole flight and you don't rely on your altimeter. The GPWS (ground proximity waring system) is based on radio-altimeter information, which is a very reliable system and, more importantly, it is 100% independent of the air data system.
                The radar altimeter on my Cub is inop!

                Comment


                • #83
                  Originally posted by BoeingBobby View Post

                  The radar altimeter on my Cub is inop!
                  But you have the Eyeball Mark IV bio-optical altimeter as backup, although I've heard it is unreliable in IMC at night as Aeroperú was facing.

                  --- 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 BoeingBobby View Post

                    The radar altimeter on my Cub is inop!
                    Can't you just do a sounding with a rope knotted in fathoms?

                    Comment


                    • #85
                      Originally posted by Gabriel View Post
                      Of course they can, they have, they are and they will be considered the last level of defense for any technical or operational issue.
                      No no no! Hard no! You cannot, during the design and certification process, say, eh, let the pilots deal with it. Not when you are seeing potential threat in a system that has supremacy over pilot commands. That line of defense is off the table. You either fix it or start over.

                      The pilot is a line of defense in operational issues BECAUSE the design and certification has assured that he is provided with the resources to be that, rather than being misled or thwarted by them.

                      There is a world of difference between an A330 giving misleading instrumentation after warning the pilots that the airspeeds are unreliable and a system that gives no warning and stealthily fights pilot commands. Perhaps you should review the design and certification standards required of Airbus FBW in the 1980's. That was before things got rotten.

                      Comment


                      • #86
                        Originally posted by Evan View Post
                        No no no! Hard no! You cannot, during the design and certification process, say, eh, let the pilots deal with it.
                        You cannot NOT say it. It is just impossible to avoid all failure modes and combinations of failure modes that could potentially cause a confusing situation for the pilots.
                        When we get to the fully autonomous plane, it will be "the pax are doomed". In the meantime, it is "the pilots will have to deal with it".
                        The only question is where the line is drawn. The certification standards draw it somewhere. But that's an arbitrary limit.

                        There is a world of difference between an A330 giving misleading instrumentation after warning the pilots that the airspeeds are unreliable and a system that gives no warning and stealthily fights pilot commands.

                        [...] Not when you are seeing potential threat in a system that has supremacy over pilot commands.
                        Which system gives no warning and stealthy fights the pilots commands and has supremacy over pilot commands? If you mean the MCAS, it is a gross misrepresentation.
                        The trim being actuated during 10 seconds in a 737 is impossible to miss (both visually and audibly), so you have warning.
                        You can FEEL the its effect in your hands too (control forces), so not stealthy either.
                        And the pilot can override it just using the normal electric trim input, or using the manual wheel, and can kill it using the cutout switch. So no supremacy over the pilot either.

                        And still, all that said...

                        Originally posted by Gabriel
                        We have known issues that are allowed by certification standards, unknown issues not known or foreseen at the time of certification, and MCAS that is totally different kind of animal.
                        So I recognize the difference between the MCAS and the rest (including the behavior of the A330 under speed disagree, which is not particularly any better or worse than in any other plane A or B).

                        --- 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


                        • #87
                          It’s maybe too late to do this, but it would be fascinating to give a nice sample of crews simulations of these events and see how they perform.

                          AND to give simulated stall exercises with and without MCAS operating and again evaluate what happens. (This one could still be done).
                          Les rčgles de l'aviation de base découragent de longues périodes de dur tirer vers le haut.

                          Comment


                          • #88
                            Originally posted by Evan View Post

                            The 737 was developed in the turbojet era for operation in regional airports including those without airstairs (jetways had not yet arived). Thus ground clearance above the minimum was only an issue with respect to the deck height.

                            The 747 was developed for the turbofan era using the revolutionary new JT9D, a high-bypass turbofan initially developed for the high-winged C5 Galaxy. Boeing thusly provided the 747 with generous ground clearance not only for the 92" fans of the JT9D but also for future evolutions. It would be serviced by large airports with jetways. By the 1970's, engineers had realized that the future would involve increasing bypass ratios and thus larger fans. Every subsequent Boeing airframe could accomodate them. The 767 could carry the same diameter CF-6's used by the 747 (as could the A300 and the DC-10). The 757 could carry the 78" fans of the PW20XX. The GEnx on the 748 carries a 111" fan, yet still fits safely below the 747 wingspan. Even the A200 can comfortably carry 73" fans.

                            The 737-Max struggles to handle 69" fans. They must be mounted forward of the wing in a configuration that compromises stability at the edge of the envelope. It should never have been considered as a viable 21st century aircraft.

                            Efficiency will always determine commercial airframe design. Environmental sustainability will hopefully also drive design decisions. The CFM56 might be a fine machine but it has to go the way of the steam engine. Boeing designers knew this long ago. The new era of ultra-high-bypass engines began to take form in the mid-1990's. Any production turbofan-powered airliner unable to handle them should have been retired ten years ago.

                            When it comes to fans, bigger is better, right up to that line where the drag penalty defeats them. Emerging technology keeps pushing the line further out. Airframers have to keep up.
                            Yep, Boeing didn't see the explosion of demand in that segment, and had no modern answer to the 320 platform. Airbus also scooped up the 220 from Bombardier (thanks to Boeing) leaving Boeing with a big hole in their offerings.

                            Comment


                            • #89
                              Originally posted by Schwartz View Post

                              Yep, Boeing didn't see the explosion of demand in that segment, and had no modern answer to the 320 platform. Airbus also scooped up the 220 from Bombardier (thanks to Boeing) leaving Boeing with a big hole in their offerings.
                              One other thing:

                              Originally posted by Evan
                              The 737 was developed in the turbojet era for operation in regional airports including those without airstairs (jetways had not yet arived). Thus ground clearance above the minimum was only an issue with respect to the deck height.
                              The other issue with ground clearance (and thus gear height) in the modern era is the need to assure that the plane is not geometrically limited, that is, to assure against tail strike at rotation. That is why the 757 has such tall gear struts. The 737 should have really reached obsolescence when the A321 appeared. Boeing's response, the 737-900, is prone to tail strike (as is the -800). Both have tail strike skid detectors installed as a crutch for that issue. Boeing never had a true competitor to the A321NEO. The MAX 10 is hard to watch. It relies on added gear complexity that allows the gear to 'telescope' in the deployed position. More crutches. More things to maintain. More things to fail. As if the 737 wasn't mechanically complicated enough already.

                              Last I checked, the A321 was outselling the 737MAX 10 by five to one.

                              Comment


                              • #90
                                Originally posted by Evan View Post

                                Boeing's response, the 737-900, is prone to tail strike (as is the -800). Both have tail strike skid detectors installed as a crutch for that issue.
                                The A321 is also "prone to tailstrike", and doesn't even have a skid detector.

                                Comment

                                Working...
                                X