How to Get People Flying on the 737 Max

[RayDunzl]

* And, as I like to tell my students: we used to have a cat who was grey on the average.

If they give you a hard time about that, you could follow up with the fact that the average student has one tit and one testicle.

 

[jomark911]

Their biggest mistake was corporate rot. I do not believe that the problem can ever be fully remedied since the design sucks. I read the backstory on the MAXX development. It was driven entirely by the threat of the superiority of the Airbus competition particularly with regards to fuel efficiency. In order to save money (and not require complete retraining of pilots on a new airplane so as to make the cost to the buyer lower) they did not do a complete redesign. They just retrofitted the previous 737 with huge engines that created a lot of aerodynamic problems (and ground clearance problems) which they knew about but figured that they could correct in the software. See https://www.latimes.com/business/story/2019-12-23/boeing-737-max-dennis-muilenburg and https://www.theguardian.com/busines...andal-the-internal-boeing-messages-and-emails Clearly cost savings (read ”greed”) trumped safety. These guys are killers like the Sacklers at Purdue Pharma and like the Sacklers they will not do jail time. Read the two links and if you are not disgusted, I don’t know what will disgust you.

Being an aviation pro , since 1990 , I can assure you , there are thousands of things that disgust me in aviation.
This is not one of em.

 

[orangejello]

Not in aviation. So not tainted. Those articles disgusted me. Boeing seem like just another example of a company not doing the required investment which then leads to mediocrity and irrelevance. They can continue for a while longer because Airbus cannot satisfy the demand. But that culture seems rotten.

 

[jomark911]

Well it is so , if you believe so.
Airbus has been accused for utilizing the worst pilot-aircraft interface in aviation. Not allowing pilots to take full control of the aircraft , without having to pull some 50 cb's to take two of the three computers out .
I call this way of design rotten too.
Depends on the perspective you're looking at it.

 

[ahofer]

I believe I heard that both Southwest and Alaska Air are going to one-plane fleets - the 737.

 

[MrPeabody]

Phenomenally stupid design. Reacting to the input from a faulty angle-of-attack sensor, the stall-avoidance system forced the aircraft into a dive. The pilots hadn't been educated as to how the system worked, and reportedly it was not even mentioned in the pilot manuals for the plane. It isn't likely that BA could have made certain that every pilot that flew the plane would have undergone the requisite training and would have understood how to determine whether it is working correctly and how to override it. Given that BA could never have had that kind of certainty, no automated system of this sort should ever have been installed in the plane. I haven't seen BA admit that it was a dumb idea, and I haven't seen them say that they have removed the system to make absolutely certain that in the future the automated system can't do this under any circumstances. I'm concerned that their "solution" may have been to add more angle-of-attack sensors along with software that will decide which of them are reporting the angle-of-attack correctly and which aren't. Somewhere there is probably some writeup of what they did exactly, but I haven't seen it. I just think it is stupid as hell to allow any possibility for software to take control of the plane and potentially cause it to crash, without or without the ability for a shrewd pilot to override the thing.

 

[blueone]

Their biggest mistake was corporate rot. I do not believe that the problem can ever be fully remedied since the design sucks. I read the backstory on the MAXX development. It was driven entirely by the threat of the superiority of the Airbus competition particularly with regards to fuel efficiency. In order to save money (and not require complete retraining of pilots on a new airplane so as to make the cost to the buyer lower) they did not do a complete redesign. They just retrofitted the previous 737 with huge engines that created a lot of aerodynamic problems (and ground clearance problems) which they knew about but figured that they could correct in the software. See https://www.latimes.com/business/story/2019-12-23/boeing-737-max-dennis-muilenburg and https://www.theguardian.com/busines...andal-the-internal-boeing-messages-and-emails Clearly cost savings (read ”greed”) trumped safety. These guys are killers like the Sacklers at Purdue Pharma and like the Sacklers they will not do jail time. Read the two links and if you are not disgusted, I don’t know what will disgust you.

I don't work for Boeing or Airbus, but having worked for my share of companies that made dumb strategic errors, I think the 737 Max root cause is best summed up by this photo. Note that this photo mentions seat width, which makes me prefer the A320 to actually fly in, but the real issue I'm trying to depict is landing gear height:

The 737 is on the right.

The latest geared turbofan engines, largely responsible for the Neo and Max fuel efficiency increases, have a larger outside diameter than the previous non-geared engines. The 737s have such a low landing gear height that even the conventional turbofans needed bottom-flattened nacelles to fit under the 737 wings:

And that tweak wasn't going to be sufficient for the geared turbofans. Nor was redesigning the 737 landing gear, which would seem to have required an airframe redesign. A redesign would have certainly slipped the Max program, already years behind the Airbus Neo, and would have huge development costs. (Boeing's original strategy on a redesign of the 737 was apparently that it wasn't worth the cost for a short haul plane, and neither were geared turbofan engines. Perhaps not their best strategic accomplishment, but then Airbus has the A380 strategy to make Boeing strategists feel better.) So Boeing's decision was to stay with the 1960s-era short landing gear design, push the engines forward under the wing for more clearance, and tilt them upward, allowing the geared turbofans to fit. Well, sort of. The entire operation reminds me of a Chevy big-block V8 retrofit into a Chevette I saw decades ago. I'm being facetious, but not that much. Just like the big block affected the handling, such as it was, of the Chevette, so apparently did the larger geared turbofans on the 737. And, of course, a big block Chevette wasn't a production car, while the 737 Max was to be Boeing's volume sales leader. And, apparently, the Max's handling "issues" were ultimately limited to corner cases, relatively speaking, and this is where I think your charge of corporate greed is accurate.

Boeing very much wanted the Max and the 737NG (the 737 600-900 series) to have nearly identical pilot training, and that is now known to be the root of their MCAS fiasco, which is so well documented in the press that recounting it here would be redundant. Perhaps greed is the wrong term. Perhaps it's what happens when you choose a risky and questionable path that ultimately doesn't work out well, and by the time reality becomes obvious it's so disastrous to change paths that you are are forced into really dumb alternatives to stave off perceived catastrophe.

So, would I fly on a Max in 2021? Yes, I would, but I'd still rather fly the A320, for the better seat width, especially on all-coach class Southwest flights.

 

[MrPeabody]

This is a great set of comparative data; thanks for posting it.

I also just want to chime in with a word about relative risk. This is one of my hot buttons pet peeves nonlinearities (as I like to say) -- so many folks make choices about things they're willing to do, or not willing to do, with no understanding of the actual relative risk (or safety) of one activity vs. another. Drives me nuts, OK?!

Dangerousness relative risks of activities by Mark Hardy, on Flickr

Ca. (edit) two orders of magnitude more risky (in terms of risk of death per unit time invested in the activity) to travel by car than by plane.

I apologize that I no longer have the original citation for this particular data set. I first saw a presentation along these lines when I was in grad school by Bruce Ames (the "Ames Test" guy), who was disheartened by the way that data from "his" test were misused/misrepresented. In his own way, he was trying to recast the context of Ames Test data, but he was largely "preaching to the choir". I've never found a copy of Ames' slide (mind you, I was in grad school a long time ago now). This graphic is taken from a Boston Globe article, but that's all I remember at this point.

Schizophrenic, arthritic lumberjacks who use motorcycles for rock climbing in the Himalayas are an especially high-risk group.

 

[jomark911]

Phenomenally stupid design. Reacting to the input from a faulty angle-of-attack sensor, the stall-avoidance system forced the aircraft into a dive. The pilots hadn't been educated as to how the system worked, and reportedly it was not even mentioned in the pilot manuals for the plane. It isn't likely that BA could have made certain that every pilot that flew the plane would have undergone the requisite training and would have understood how to determine whether it is working correctly and how to override it. Given that BA could never have had that kind of certainty, no automated system of this sort should ever have been installed in the plane. I haven't seen BA admit that it was a dumb idea, and I haven't seen them say that they have removed the system to make absolutely certain that in the future the automated system can't do this under any circumstances. I'm concerned that their "solution" may have been to add more angle-of-attack sensors along with software that will decide which of them are reporting the angle-of-attack correctly and which aren't. Somewhere there is probably some writeup of what they did exactly, but I haven't seen it. I just think it is stupid as hell to allow any possibility for software to take control of the plane and potentially cause it to crash, without or without the ability for a shrewd pilot to override the thing.

Well this is actually what AIRBUS is doing.

 

[MrPeabody]

I don't work for Boeing or Airbus, but having worked for my share of companies that made dumb strategic errors, I think the 737 Max root cause is best summed up by this photo. Note that this photo mentions seat width, which makes me prefer the A320 to actually fly in, but the real issue I'm trying to depict is landing gear height:

View attachment 101340

The 737 is on the right.

The latest geared turbofan engines, largely responsible for the Neo and Max fuel efficiency increases, have a larger outside diameter than the previous non-geared engines. The 737s have such a low landing gear height that even the conventional turbofans needed bottom-flattened nacelles to fit under the 737 wings:

View attachment 101342

And that tweak wasn't going to be sufficient for the geared turbofans. Nor was redesigning the 737 landing gear, which would seem to have required an airframe redesign. A redesign would have certainly slipped the Max program, already years behind the Airbus Neo, and would have huge development costs. (Boeing's original strategy on a redesign of the 737 was apparently that it wasn't worth the cost for a short haul plane, and neither were geared turbofan engines. Perhaps not their best strategic accomplishment, but then Airbus has the A380 strategy to make Boeing strategists feel better.) So Boeing's decision was to stay with the 1960s-era short landing gear design, push the engines forward under the wing for more clearance, and tilt them upward, allowed the geared turbofans to fit. Well, sort of. The entire operation reminds me of a Chevy big-block V8 retrofit into a Chevette I saw decades ago. I'm being facetious, but not that much. Just like the big block affected the handling, such as it was, of the Chevette, so apparently did the larger geared turbofans on the 737. And, of course, a big block Chevette wasn't a production car, while the 737 Max was to be Boeing's volume sales leader. And, apparently, the Max's handling "issues" were ultimately limited to corner cases, relatively speaking, and this is where I think your charge of corporate greed is accurate.

Boeing very much wanted the Max and the 737NG (the 737 600-900 series) to have nearly identical pilot training, and that is now known to be the root of their MCAS fiasco, which is so well documented in the press that recounting it here would be redundant. Perhaps greed is the wrong term. Perhaps it's what happens when you choose a risky and questionable path that ultimately doesn't work out well, and by the time reality becomes obvious it's so disastrous to change paths that you are are forced into really dumb alternatives to stave off perceived catastrophe.

So, would I fly on a Max in 2021? Yes, I would, but I'd still rather fly the A320, for the better seat width, especially on all-coach class Southwest flights.

The handling issues are well documented? How severe? Of course it is generally true that handling issues are corner cases; this is the nature of handling issues.

 

[blueone]

The handling issues are well documented? How severe? Of course it is generally true that handling issues are corner cases; this is the nature of handling issues.

Here's one site that does a pretty good job describing MCAS operations:

http://www.b737.org.uk/mcas.htm

 

[MrPeabody]

Well this is actually what AIRBUS is doing.

Yes, and both Airbus and BA are moving toward elimination of the pilot. I suspect that the main motivation for this is pressure from the airlines, who simply want to eliminate the cost of paying the pilot. It is inevitable. Probably not more than a decade or two down the road. It will start with freight carriers, including FedEx, UPS and future AMAZON-AIR, and once that hurdle is crossed, it will be only the blink of an eye before the passenger airlines follow suit. The only thing that would stop it would be if the pilot's association had the same kind of clout that the railroad workers union had back in the middle decades of the previous century. No union has that same kind of power anymore, and the pilot's union certainly does not.

 

[tvrgeek]

Actually, as the plane has been reviewed with far more detail than anything else flying, it is probably a good bet. On the other hand, I like old steam gauge overpowered tanks like the original 737. Pre E model. Big fan of the early 727's. I am not a fan of digital controls due to my background in understanding how analog and digital systems fail. Analog tend to be predictable. Digital not.

This year? Not getting anywhere near a COVID cesspool. Next year? I spent too many years doing field service. I expect to never get on another plane as long as I live.

I do wish we had better than a third world train system. Embarrassing actually. Any trip less than 300 miles would be better served by trains. Leave the planes where there is water or mountains to cross, or when the trip would be more than 3 or 4 hours and that 350 knot speed is important.

 

[MrPeabody]

Here's one site that does a pretty good job describing MCAS operations:

http://www.b737.org.uk/mcas.htm

Thanks for that. I will read through it more later. But even with a quick glance, I noticed this:

"The original design of MCAS was that it would only activate 'at extreme high speed pitch-up conditions that are outside the normal operating envelope' (see extract from the Mainenance Training Manual below). However during flight testing it became apparent that the engine nacelles were also creating a pitch-up effect under certain conditions at very low speeds. So the scope of MCAS was broadened to include low speed activation as well as high speed activation."

Translation: MCAS as it was implemented was a "solution" for a serious handling issue that was directly due to the installation of engines that were too large to fit properly under the wing.

Which is the gist of what you had said.

 

[tvrgeek]

Yes, and both Airbus and BA are moving toward elimination of the pilot. I suspect that the main motivation for this is pressure from the airlines, who simply want to eliminate the cost of paying the pilot. It is inevitable. Probably not more than a decade or two down the road. It will start with freight carriers, including FedEx, UPS and future AMAZON-AIR, and once that hurdle is crossed, it will be only the blink of an eye before the passenger airlines follow suit. The only thing that would stop it would be if the pilot's association had the same kind of clout that the railroad workers union had back in the middle decades of the previous century. No union has that same kind of power anymore, and the pilot's union certainly does not.

I suspect you will be shown incorrect over the time line you suggest. I can see moving to a single pilot as automation systems manage more of the workload, but never completely autonomous. I can see a secondary backup pilot being ground based where one "operator" could be backup for multiple planes. It would require a complete next generation of avionics, so the lead time is longer than the technology. Do not bet the pilot union is as weak as you think.

The major cost is not the pilot, but fuel. Another useless cost is our antiquated air-route system. Attempts to build a reliable point to point control system have so far been un-successful, but maybe in 20 years. Knowing a bit about "life threat" level software, make that 30 to 40.

 

[tvrgeek]

Thanks for that. I will read through it more later. But even with a quick glance, I noticed this:

"The original design of MCAS was that it would only activate 'at extreme high speed pitch-up conditions that are outside the normal operating envelope' (see extract from the Mainenance Training Manual below). However during flight testing it became apparent that the engine nacelles were also creating a pitch-up effect under certain conditions at very low speeds. So the scope of MCAS was broadened to include low speed activation as well as high speed activation."

Translation: MCAS as it was implemented was a "solution" for a serious handling issue that was directly due to the installation of engines that were too large to fit properly under the wing.

Which is the gist of what you had said.

I believe it was due to the shift in the center of lift to the center of gravity. Maybe that was caused by the larger diameter engines.

 

[MrPeabody]

Actually, as the plane has been reviewed with far more detail than anything else flying, it is probably a good bet. On the other hand, I like old steam gauge overpowered tanks like the original 737. Pre E model. Big fan of the early 727's. I am not a fan of digital controls due to my background in understanding how analog and digital systems fail. Analog tend to be predictable. Digital not.

This year? Not getting anywhere near a COVID cesspool. Next year? I spent too many years doing field service. I expect to never get on another plane as long as I live.

I do wish we had better than a third world train system. Embarrassing actually. Any trip less than 300 miles would be better served by trains. Leave the planes where there is water or mountains to cross, or when the trip would be more than 3 or 4 hours and that 350 knot speed is important.

So much there that I agree with very strongly. My first ever flight in any airplane was in an early 727 flown by Eastern airlines. Back then the fares were regulated and airplanes often flew half empty. That had to change of course, no question about that, but the deregulation that occurred in the '80s changed the airline industry in a dramatic way. Digital control systems would be mostly okay if they didn't involve complicated software that can't ever be adequately tested. Having spent a good part of my life as a software developer in mostly embedded systems, I know well that most any software has "corner cases" that can't even be identified much less properly tested. I saw so much stuff get past testing. For the most part the stuff that gets caught during testing is just the low-hanging fruit. This is why I was always a strong advocate for more thorough code reviews (and code reviews that actually meant something, as opposed to the ones that were mostly about whether everyone likes the object model). Analog control systems are much preferable for anything that can be done with analog, but I doubt whether the complexity needed for full pilot elimination could be done in analog. Like you, I absolutely, positively will not go near an airport, much less an airplane, until the Covid-19 thing is over. And I doubt I will even after that. Since deregulation, flying on a commercial airline has been one of my least favorite things to do.

And it is indeed an embarrassment that North America is so far behind Europe and Asia in high-speed trains. Why aren't we working of this instead of working on a manned mission to Mars? It makes no sense to me. Our electrical power grid is also an embarrassment, consisting of a hodgepodge of fragments. The one thing that we do particularly well as a nation is fight wars. Not to suggest that this is a bad skill to have, but only that we ought to be better at doing other things in addition.

 

[MrPeabody]

I believe it was due to the shift in the center of lift to the center of gravity. Maybe that was caused by the larger diameter engines.

The web site indicated, in the excerpt I quoted, that it was an aerodynamic effect attributable to the shape of the engine nacelles. Moving the engines forward no doubt shifted the center of mass forward, but surely they would have compensated for that, somehow, because it is not good practice to rely on aerodynamic effects to correct a fore/aft misalignment of the center of lift and the center of mass. Anyone who has ever played much with balsa wood gliders will likely have figured out this much. If the wing is too far back and you try to fix this using the angle of the horizontal stabilators, you quickly discover that the angle of the stabilator needs to be more pronounced at low air speed and less pronounced at high air speed. Aerodynamic forces generally vary as the square of airspeed, naturally. You easily find yourself with a toy airplane that loops when you throw it hard and then dives after it slows down. Hopefully BA was not relying on software control of the stabilators to fix a problem with center of mass being too far forward of the center of lift. I would hope that they didn't do this. But who knows.

 

[tvrgeek]

We gave up on code reviews in favor of faster try it process. Why? Well, so much of the code was plug in modules you had no idea what was inside, how much good does doubling the development time for formal review to inspect 10% of the code? We spent more time doing integration testing and kept each deployment very small so roll-back could be quick. We did use an extensive set of automated test tools. Once integrated, they took no time from the developers. I found the developers would do wonderous things if you let them develop, not sit in review meetings, planning meetings,. status meetings, management feel-good meetings and even scheduled stand-ups. I let them call stand ups when they thought it was needed ( not often) or when the noise over the cubicles was too quiet or too noisy. Hint someone was stuck.

Now, if you have people with the right personality, Extreme programing can work. This was DoD, but not life threat. If L-T or Nuclear, we would not be using Java and not have any external supplied routines. If your architecture is proper in each module having clear gaza-inna and gazz-outa, then white box testing can be highly effective. Put more than a day's worth if coding in a module, and then you get into trouble.

Well, it seems Boeing did try to fix inherent imbalance with software. The technology of fly by wire has made the old concept of "if all else fails, let go" and planes whose the natural position was to fly. Makes sense in fighter jets which have been unstable for years, but airlines? Not too thrilled. If I were the architect, there would be a button on the yoke that says " I have the airplane" but I guess they decided it was too hard to actually fly so they made the switch where you have to hunt for it. I am sure there is far more to it than that. Single sensor, not three, was just plain stupid.

That is not what is stopping me from getting on one. It's having to pay to be abused. I sure would jump on one before an L1011. Hated those things. Flew at about a 20 degree angle. I hope they are all gone.

 

[terasankka]

What do you mean , you will not fly max. As a pax or as a crew member?
Indeed they have made mistakes , but they've remedied the whole thing.
Their biggest mistake, trying to copy AIRBUS.

As a passenger.

They just retrofitted the previous 737 with huge engines that created a lot of aerodynamic problems (and ground clearance problems) which they knew about but figured that they could correct in the software.

Yes, exactly this. I will not step into that abomination as a pilot nor a passenger.