5 ms·

I think whats different about the 737 Max versus other past troubled models of planes is the lack of empathy. Prior troubled models are reflown or rebranded an

I think whats different about the 737 Max versus other past troubled models of planes is the lack of empathy. Prior troubled models are reflown or rebranded and flown - of course after addressing the critical flaw. The 737 Max’s flaw is that it doesn't really fly well and needs software to regulate it, and a human to override the software, which is just odd. But the company’s complacency, self regulatory status, lack of empathy - probably because saying sorry has legal consequences for them, or they’re leadership really are sociopaths, or both - is what damns this line for good. Will consumers really be so gullible this time or have more pressing needs for the cheaper flight that they ignore the model being flown? I’m really curious

yutyut 7y ago
>needs software to regulate it it This is a standard pattern on most modern aircraft. Control is abstracted away by some computers, however—and as you said—the pilot can override the computers to more directly interact with the aircraft’s flight controls. Not an odd principle at all.
- salawat 7y ago
  There's a difference there though. That's Fly-By-Wire in two different contexts. One way to do it is indirect linkage to control surfaces, and using the computer to generate the appropriate signaling to actuate the control surfaces in a more or less direct mapping of pilot intent to control surface output. Depending on design philosophy, more in depth envelope protection may be implemented, (see Airbus) but there's no given that says Fly-By-Wire must be able to countermand and pilot commands. The other way Fly-by-Wire is used (and to me personally is an anti-pattern) is as a pilot/plane babysitter like MCAS. The plane should be free of bad aerodynamic behaviors so that even with total electrical system casualty, a passive glide and controlled landing is possible. Once MCAS is taken out of the picture, the MAX doesn't have that. It does not meat the full set of airworthyness regulations for Civil Transport Aircraft as laid out in FAR 25.173. I understand some in the industry may scoff at having to comply with the directive as written, but it's now been clearly demonstrated what happens when you don't.
  - jdsully 7y ago
    FAR 25.173 says nothing about electrical failure. Indeed at the sizes of aircraft we’re taking about the pilot is just not going to be strong enough without hydraulic or electric assistance.
    - salawat 7y ago
      The regulation doesn't, no, the technical details of MCAS as a piece of software necessary to maintain airworthyness however, demand the flight control computer be able at all times to work as designed to maintain the regulated condition of airworthyness. Namely, no control stick force inversions on approach to stall. It's kind of hard to do that with a computer if you've lost your electrics. Therefore either the system needs to be bulletproof, or you should have designed out the need for the crutch in the first place. Not gone and made a Frankenstein's monster of a plane.
    - bdavis__ 7y ago
      leaving out the discussion of how much force is required and how strong a pilot is, the hand crank for the stab trim has mechanical leverage. the pilot turns a crank that is geared to move a cable. the amount of leverage is determined by mechanical design. it could be geared to move a stabilizer on a plane N times as large...more turns though.
      - salawat 7y ago
        It still runs into problems where in general, the electronic manipulation is what pilots muscle memory develops around, and the more mechanical advantage is utilized, the faster the wheel turns in the cockpit during normal operation. Those wheels are already notorious for the punishment they can subject an errant kneecap to.
  - temac 7y ago
    IMO MCAS (the version properly respecified and reimplemented, of course) really fits in the "envelop protection" category (even if slightly pre-emptive), why do you classify it otherwise?
megous 7y ago
> The 737 Max’s flaw is that it doesn't really fly well and needs software to regulate it, and a human to override the software, which is just odd. It's not really odd though. One of the first A320 crashes (https://en.wikipedia.org/wiki/Air_France_Flight_296 https://en.wikipedia.org/wiki/Air_France_Flight_296) was also influenced by computer doing opposite of what the pilot tried to achieve (pilot flying at extremely low altitude into the trees tried to pull up to avoid the trees, but computer was set up to ignore that and instead lower the nose first to gain speed, before pulling up, to avoid a stall). The plane would have probably crashed even if the computer did what pilot commanded, but hey, my point is that computers override pilots these days, and it's not odd. It's probably for the good, overall.
eCa 7y ago
> The 737 Max’s flaw is that > it doesn't really fly well Actually, no. It's flaw is that it doesn't fly like previous 737's, and they need to make it do that.
- msbarnett 7y ago
  Actually, if you’re going to correct someone with “Actually,”, you should understand the issue. The MCAS exists to prevent stick forces from inverting as the nose pitches up towards the stall angle. If the stick forces invert, it becomes easier to continue into a stall angle than to move away from it. Commercial airframes are required to never exhibit this characteristic as a condition of certification. MCAS was Boeing’s idea of a solution to this; command the stabilizers down to provide counter force against the stick, making it harder for the pilot to pull into the stall than pull away from it at every point. Even if previous 737s had never existed and the 737MAX as it exists today were for some reason an all new design, it would need the MCAS counter-force to get certified as air worthy no matter how much new training pilots got. It (claims to) solve a fundamental airworthiness requirement; it is not just there to “fly like previous 737s”. The behaviour of the airframe is simply not considered commercially airworthy without it.
  - LaserToy 7y ago
    That is my understanding as well. Not sure why a lot of folks are sticking to pilot certification idea. IMO, removing G-force sensor from the system should be looked into. I just can’t see any sane engineer agrees to do so. They used it to make sure system only engages on high G but will limited authority, now the system needs to engage with low G, but higher authority - they should’ve introduce another signal (air speed?), not remove one. It is either incompetence or greed.
    - msbarnett 7y ago
      It reads like broader organizational incompetence to me. From what I’ve seen: 1. They did safety analysis based on the original design: what needs to happen for MCAS to misfire (both the g-sensor and the AoA need to be incorrect) and what happens if it does (it runs once, to maximum authority of 0.6 degrees). They concluded from this that the issue was “major failure” rather than “hazardous” or “catastrophic”, and that the design met the reqs for “major” (2 sensors involved here, and major only requires 1). 2. Much later, with the analysis already rubber stamped by the FAA, they discover the low speed issue with stick forces and decide they can modify MCAS to resolve this as well, by radically increasing its authority over the stabilizers (more deflection needed to generate the same force at lower speeds) and removing the gsensor (which wouldn’t trip at low speeds). (This also seems to be where they introduced running it in a loop? Although some things we’ve seen out of Boeing seem to indicate that they might have not understood at any point that this was the way the system was written, or at least anyone involved in any analysis never knew to consider that) 3. Again later, someone concludes that the changes meet their initial analysis classification requirements, because they’ve already done a safety analysis and concluded “major”, which allows for 1 sensor with no redundancy, which they continue to have. Nobody seems to have noticed that they’ve now fallen into a circular reasoning loop — the system must be safe to “major” level because we previously decided the system was safe to “major” level. They look at whether or not the changes still meet the requirements for “major” failure, but neglect to re-open the analysis and decide whether “major” is still the correct classification of failure in the first place. Probably this means 1, 2, and 3 were being done by multiple groups of people with poor visibility and communication between them. Someone was asked to check whether the new setup met the requirements for the failure analysis the FAA had approved, but there was nobody looking at the bigger picture question of whether the validity of that analysis still applied.

yutyut 7y ago

>needs software to regulate it it This is a standard pattern on most modern aircraft. Control is abstracted away by some computers, however—and as you said—the pilot can override the computers to more directly interact with the aircraft’s flight controls. Not an odd principle at all.

salawat 7y ago
There's a difference there though. That's Fly-By-Wire in two different contexts. One way to do it is indirect linkage to control surfaces, and using the computer to generate the appropriate signaling to actuate the control surfaces in a more or less direct mapping of pilot intent to control surface output. Depending on design philosophy, more in depth envelope protection may be implemented, (see Airbus) but there's no given that says Fly-By-Wire must be able to countermand and pilot commands. The other way Fly-by-Wire is used (and to me personally is an anti-pattern) is as a pilot/plane babysitter like MCAS. The plane should be free of bad aerodynamic behaviors so that even with total electrical system casualty, a passive glide and controlled landing is possible. Once MCAS is taken out of the picture, the MAX doesn't have that. It does not meat the full set of airworthyness regulations for Civil Transport Aircraft as laid out in FAR 25.173. I understand some in the industry may scoff at having to comply with the directive as written, but it's now been clearly demonstrated what happens when you don't.
- jdsully 7y ago
  FAR 25.173 says nothing about electrical failure. Indeed at the sizes of aircraft we’re taking about the pilot is just not going to be strong enough without hydraulic or electric assistance.
  - salawat 7y ago
    The regulation doesn't, no, the technical details of MCAS as a piece of software necessary to maintain airworthyness however, demand the flight control computer be able at all times to work as designed to maintain the regulated condition of airworthyness. Namely, no control stick force inversions on approach to stall. It's kind of hard to do that with a computer if you've lost your electrics. Therefore either the system needs to be bulletproof, or you should have designed out the need for the crutch in the first place. Not gone and made a Frankenstein's monster of a plane.
  - bdavis__ 7y ago
    leaving out the discussion of how much force is required and how strong a pilot is, the hand crank for the stab trim has mechanical leverage. the pilot turns a crank that is geared to move a cable. the amount of leverage is determined by mechanical design. it could be geared to move a stabilizer on a plane N times as large...more turns though.
    - salawat 7y ago
      It still runs into problems where in general, the electronic manipulation is what pilots muscle memory develops around, and the more mechanical advantage is utilized, the faster the wheel turns in the cockpit during normal operation. Those wheels are already notorious for the punishment they can subject an errant kneecap to.
- temac 7y ago
  IMO MCAS (the version properly respecified and reimplemented, of course) really fits in the "envelop protection" category (even if slightly pre-emptive), why do you classify it otherwise?

megous 7y ago

> The 737 Max’s flaw is that it doesn't really fly well and needs software to regulate it, and a human to override the software, which is just odd. It's not really odd though. One of the first A320 crashes (https://en.wikipedia.org/wiki/Air_France_Flight_296 https://en.wikipedia.org/wiki/Air_France_Flight_296) was also influenced by computer doing opposite of what the pilot tried to achieve (pilot flying at extremely low altitude into the trees tried to pull up to avoid the trees, but computer was set up to ignore that and instead lower the nose first to gain speed, before pulling up, to avoid a stall). The plane would have probably crashed even if the computer did what pilot commanded, but hey, my point is that computers override pilots these days, and it's not odd. It's probably for the good, overall.

eCa 7y ago

> The 737 Max’s flaw is that > it doesn't really fly well Actually, no. It's flaw is that it doesn't fly like previous 737's, and they need to make it do that.

msbarnett 7y ago
Actually, if you’re going to correct someone with “Actually,”, you should understand the issue. The MCAS exists to prevent stick forces from inverting as the nose pitches up towards the stall angle. If the stick forces invert, it becomes easier to continue into a stall angle than to move away from it. Commercial airframes are required to never exhibit this characteristic as a condition of certification. MCAS was Boeing’s idea of a solution to this; command the stabilizers down to provide counter force against the stick, making it harder for the pilot to pull into the stall than pull away from it at every point. Even if previous 737s had never existed and the 737MAX as it exists today were for some reason an all new design, it would need the MCAS counter-force to get certified as air worthy no matter how much new training pilots got. It (claims to) solve a fundamental airworthiness requirement; it is not just there to “fly like previous 737s”. The behaviour of the airframe is simply not considered commercially airworthy without it.
- LaserToy 7y ago
  That is my understanding as well. Not sure why a lot of folks are sticking to pilot certification idea. IMO, removing G-force sensor from the system should be looked into. I just can’t see any sane engineer agrees to do so. They used it to make sure system only engages on high G but will limited authority, now the system needs to engage with low G, but higher authority - they should’ve introduce another signal (air speed?), not remove one. It is either incompetence or greed.
  - msbarnett 7y ago
    It reads like broader organizational incompetence to me. From what I’ve seen: 1. They did safety analysis based on the original design: what needs to happen for MCAS to misfire (both the g-sensor and the AoA need to be incorrect) and what happens if it does (it runs once, to maximum authority of 0.6 degrees). They concluded from this that the issue was “major failure” rather than “hazardous” or “catastrophic”, and that the design met the reqs for “major” (2 sensors involved here, and major only requires 1). 2. Much later, with the analysis already rubber stamped by the FAA, they discover the low speed issue with stick forces and decide they can modify MCAS to resolve this as well, by radically increasing its authority over the stabilizers (more deflection needed to generate the same force at lower speeds) and removing the gsensor (which wouldn’t trip at low speeds). (This also seems to be where they introduced running it in a loop? Although some things we’ve seen out of Boeing seem to indicate that they might have not understood at any point that this was the way the system was written, or at least anyone involved in any analysis never knew to consider that) 3. Again later, someone concludes that the changes meet their initial analysis classification requirements, because they’ve already done a safety analysis and concluded “major”, which allows for 1 sensor with no redundancy, which they continue to have. Nobody seems to have noticed that they’ve now fallen into a circular reasoning loop — the system must be safe to “major” level because we previously decided the system was safe to “major” level. They look at whether or not the changes still meet the requirements for “major” failure, but neglect to re-open the analysis and decide whether “major” is still the correct classification of failure in the first place. Probably this means 1, 2, and 3 were being done by multiple groups of people with poor visibility and communication between them. Someone was asked to check whether the new setup met the requirements for the failure analysis the FAA had approved, but there was nobody looking at the bigger picture question of whether the validity of that analysis still applied.