Hacker News Reader

Comment by hurricanetc | Hacker News Reader

I can't believe Boeing has been able to get away with the 737-MAX. It is a unique plane with unique flight characteristics running a complex software simulation so that it "feels" like an existing plane. It's an abomination designed to sell a new plane without having to retrain pilots.

I think the fact that this has been allowed to happen is evidence of your third point. I'm not sure if the MAX plane or MAX brand is salvageable but I don't think Boeing as a company will be allowed to fail.

View on HN

>It is a unique plane with unique flight characteristics running a complex software simulation so that it "feels" like an existing plane. It's an abomination designed to sell a new plane without having to retrain pilots.

Well...not to excuse the poor decisions that lead to the problems with the MAX, but this arguably isn't one of them. Military jets have been unflyable without computer assistance for decades, and it was only a matter of time before such tech made it into commercial aerospace. Though the advantages of maneuverability conferred to fighters from dynamic instability aren't useful in passenger flight, I'd argue that computerized flight controllers are sufficiently mature and battle tested that building a commercial jet which is dependent upon computer control isn't really an oversight, provided the systems are well designed and redundant.

Remember, the issue here wasn't with the flight characteristics per se, it was penny pinching administrators who removed redundancy from a critical system to save (and attempt to mark up) a couple hundred dollars per plane.

>Remember, the issue here wasn't with the flight characteristics per se,

Well, it was and it wasn't. In order to meet the stated demands of their customers (large airlines who didn't want to retrain pilots on a new type), they had to stick to the same 737 type certificate. The engines on the MAX change the centre of gravity such that the handling characteristics were too different. The regulators would not have allowed the MAX to share the type certificate with the 737NG as it was, so MCAS was born.

I'm convinced that the MAX 8 itself could have been certified as a distinct type (without MCAS, using some other system that wasn't half-baked) and we would not be having this conversation because those two crashes would not have happened.

At least as I see it, the flight characteristics led to MCAS because of penny pinchers.

The 737-Max could not pass 25.173 certification standard. This was the result of larger engines needing to be mounted farther forward due to the constraint of the 737 landing gear.

It was a desire to keep a common type rating that led to the need for MCAS to pass 25.173 because of the landing gear causing an insufficiently (though still positively) longitudinally stable aircraft, not a desire to keep the handling characteristics "similar" to the other 737s, but to keep them legal at all.

1 - https://www.law.cornell.edu/cfr/text/14/25.173

>The 737-Max could not pass 25.173 certification standard. This was the result of larger engines needing to be mounted farther forward due to the constraint of the 737 landing gear.

What I'm suggesting is that if they were to just call it say, the 797 (or whatever), then they could make the handling characteristics different from the 737 and solve the pitching moment problem in a different way that would have been safer.

Yes. They could put the larger engines under a different wing with taller landing gear and call it a 757. ;)

No need. The plane is perfectly flyable without MCAS. It just handles differently to other 737s in parts of the flight envelope.

There is a legal need (to pass 25.173 requirements). The Max doesn't have MCAS because it feels different from another 737; it has MCAS because it feels different from the legal requirement to certify any civilian transport aircraft.

> has MCAS because it feels different from the legal requirement to certify any civilian transport aircraft

Source for that claim?

I’ve read extensively on the topic and have internal Boeing sources and everything I’ve seen says MCAS was needed to maintain the type rating, not that without MCAS the plane wasn’t airworthy.

25.173.a requires the aircraft to meet this section at speeds as low as the minimum speed for steady unstalled flight.

25.173.c requires the stick force curve to have a stable and minimum slope of 1 lbf/6 knots.

At high angles of attack (low speeds), the stick force reduces slightly as compared to lower angles of attack (higher speeds). This requires a negative slope at a point on the stick force curve, violating both the minimum slope and the stable slope requirement of 25.173. (indirect source: https://www.wsj.com/articles/how-boeings-737-max-failed-1155... ). They couch it as "pushes the nose up" (though the image cited is dual sourced to Boeing and a WSJ graphics author, so I'll wager that "pushes the nose up" is WSJ content rather than Boeing content).

I've never seen the claim regarding 25.173 properly sourced.

Don't know if any of these will satisfy or not, in advance of a final NTSB/FAA/Boeing finalization/post-mortem (I doubt all the internal docs are ever going to come out, of course):

http://newsinflight.com/2019/03/29/what-causes-boeing-737max...

https://abnormaldistribution.org/index.php/2019/04/30/ieee-s...

https://www.quora.com/Does-the-fact-that-the-pilots-of-the-7...

https://seekingalpha.com/article/4286602-boeing-737-max-misc...

They could have if they'd been willing to accept a new type rating, but airlines aren't keen on retraining their 737 pilots unless they have to. Boeing's retention of the same rating was a selling point to airlines otherwise if they're having to train people anyways maybe Airbus would be worth a look?

There's nothing wrong with fly-by-wire per se but it has specific engineering requirements -- including requirements for the skills of the engineers who design it -- that absolutely must be obeyed. Boeing's leadership treated it as "a little bit of software" as though it was merely an Excel macro. Safety critical software is just not in the same category as a mobile app but Boeing execs ignored the distinction.

Airbus controls have functioned this way since the 80s. A given stick deflection will produce the exact same amount of rotation no matter the aircraft weight or speed.

Well...not to excuse the poor decisions that lead to the problems with the MAX, but this arguably isn't one of them.

The situation of Boeing is that they've boxed themselves into a corner from an engineering and regulatory viewpoint. The plane could be made flyable with enough modifications - of course. But those modifications are supposed to meet regulatory constraints AND the FAA is facing heat for letting them get away with what they got away with.

But with enough pressure, sure, the FAA may approve WTF Boeing does to make the plane work. But then, there's a big spotlight on this stuff and there are a lot of people saying right they'll never fly an 737 MAX ever again. Sometimes bad publicity can distracted or waited out but I don't see this as possible now - the story too photogenic, Boeing can't wait years. Just as much, the world's regulatory agencies aren't necessarily deferring to the FAA any more after this.

So this looks really bad for Boeing.

>Military jets have been unflyable without computer assistance for decades

I think you'll find those have ejector seats fitted. I'm not sure that kind of dynamic instability will ever be worth the it for passenger jets. It's not as if they are getting into dog fights much.

Military jets come with a different risk tolerance from civilian airliners. I think there's a defensible position that says this kind of pitch-up behaviour is an unnecessary risk on a civilian plane, however reliable you make the computer system that compensates for it.

Military jets aren't really comparable. The pilots have signed up for a job with an elevated risk of death, and they are often asked to do far riskier things like "go dogfight this MIG near the North Korean border" or whatever. The "acceptable" level of fatalities in military aviation exists, whereas in passenger flights it's basically treated as zero.

> Military jets have been unflyable without computer assistance for decades, and it was only a matter of time before such tech made it into commercial aerospace.

Er...no? The difference is that for the military jets there is an advantage, a requirement, that is helped by the plane being unstable and thus needing assistance (manoeuvrability).

For commercial airliners, there is no such requirement or advantage in normal circumstances (that is, unless you are trying to salvage an old airframe and press it into a role it is unsuited for because you are to cheap to develop the appropriate new airframe). You want a regular plane to be stable.

> Military jets have been unflyable without computer assistance for decades

Yes, but the maneuverability demands on military jets are what caused that to be a requirement. Passenger aircraft are not generally supposed to be unstable on any axis, and don't pull a couple of Gs in any direction for dogfights (as if dogfights are still a thing).

>> Remember, the issue here wasn't with the flight characteristics per se, it was penny pinching administrators who removed redundancy from a critical system to save (and attempt to mark up) a couple hundred dollars per plane.

This sounds good (blame it on cheapo mgmt) but really is not true. For one, the system never had redundant sensors, so they couldn't have been "removed". Secondly, the systems were designed by engineers -- with the mandate of no retraining necessary, true -- but I don't see where anybody stood up and said this is a failed/doomed system (except outsiders, after the fact). Third, "hundreds of dollars" is a total joke. The point of compatibility with previous 737s was a major selling point, worth billions of dollars in orders from airlines who liked the 737, instead of requiring them to choose between an entirely new Boeing design vs a semi-familiar (for many airlines) Airbus. It also had side effects of getting to market faster and reducing recertification and training expenses, likely totalling billions in additional savings.

Sources: https://www.theverge.com/2019/5/2/18518176/boeing-737-max-cr... https://www.seattletimes.com/business/boeing-aerospace/a-lac...

Reply on news.ycombinator.com