Hacker News new | past | comments | ask | show | jobs | submit login
A 50-year-old design came back to haunt Boeing with its troubled 737 Max jet (latimes.com)
128 points by eplanit 9 months ago | hide | past | web | favorite | 118 comments

Boeing want's to keep the type rating and rest of the infrastructure, but the airframe is getting old. They made a small tweak, moved engines forward, and gained significant fuel savings. The tweak caused instability. The fixed it with automation that introduced complication.

Type ratings, airport dimensions, etc. create legacy that puts limit to what kind of aircraft designs are economically viable. If you could start from the scratch and design new aircraft and airports in tandem , would probably have much wider wingspans and lifting bodies. Airports would need different design to move passengers and arrange aircraft.

The airframe isn't getting old. In fact there's nothing left of the old airframe but the general shape of the windshields.

It's the type rating approval process that got old and caused all this.

The 737 is a great example of a Ship of Theseus, years of small changes have left barely anything of the original besides the name... and the type certificate.

And, apparently, sitting very low on the tarmac.

What I haven't seen yet, and would love to see, is an explanation of why they went for this cascade of apparent band-aid solutions, instead of just adjusting the landing gear to give them more options for where to put the engines.

I'm sure everything I'm reading is an oversimplification, but, from what I have read, it sure sounds like the crux of the problem here is that they put the moment of thrust in the wrong spot.

Here's Gary Hamatani, chief project engineer for Boeing's 737 MAX program, explaining that they needed the main landing gear to fit the existing wheel wells: https://www.youtube.com/watch?v=F4IGl4OizM4 I find it very illustrative, as it contains the keywords that are at issue here: "we wanted to preserve commonality", "from a pilot's perspective there's absolutely nothing different", etc.

Lengthening the main landing gear would require a near total redesign of the wing, as the gear would have to be moved further outboard.

That was but one of many options, all with pros and cons.

It does appear that there is a bit of room to move the gear attachment point outward. Internal stuff might need to be relocated.

A very nice solution is to fold the landing gear as the B52 bomber does, asymmetrically with one in front of the other. Internal stuff almost certainly would need to be moved around, and then doors would be required. The doors would probably fit, unlike on the original, so that would be a nice upgrade.

Boeing did add some fancy stuff (on the 737MAX-10 only) to lengthen the gear right before lifting into the air and then unlengthen it as it gets retracted. The nice thing about this choice is that there need not be any changes to the height of airport equipment unless the fancy mechanism wasn't used for the nose gear. So that helps: food/maid truck, fuel truck, baggage truck, jetway or airstairs, etc.

Another option is to keep the aircraft low. This of course makes it much harder to increase the frontal area of the engines while keeping the aircraft with a pair of under-wing podded engines. We can imagine playing with the definition of "engine", installing units that have pairs of side-by-side turbines that can only be operated or maintained as a complete unit. We can imagine a much higher engine mounting, with some of the bypass air flowing over the top of the wing. We can imagine moving the engines outward to give more clearance due to wing shape, at the cost of being less-capable of crosswind landings and being more difficult to fly straight when an engine is out.

Or they could just do a clean new design.

Or Trigger's Broom (for those of us in the UK): https://www.youtube.com/watch?v=56yN2zHtofM

Many design choices still exist. Some design constraints were irrelevant already when the plane was introduced, yet they still exist and haunt the plane.

They made the ground clearance as small as possible, because 737 engineers assumed that people have to walk stairs to the ground level to enter and exit the plane. The luggage and serving of the plane was also done by hand without lifters at that time. This unfortunate choice led to engine placement that increased fuel consumption they tried to fix.

Newly designed aircraft would have better handling without these fixes.

How exactly is the approval process at fault here? Nobody forced Boeing to make a plane that has the same type rating.

The approval process allowing indefinite "type certificate stretching" may create a perverse economic incentive for companies to mask or downplay relevant changes to the aircraft.

Boeing's customers demanded a plane that they could deploy without significantly retraining their pilots. The only way to do that in the current regulatory framework was to produce an aircraft with the same type rating.

It would cost some number of $B and years to get different type rating.

If you want to call that "forcing" is a matter of taste, but it's certainly a very strong and involuntary incentive.

Boeing made the decision to maximize profits by retrofitting an older model. There’s nothing involuntary about it.

They’ve released newer model planes that have gone through the review process fine, correct?

FWIW, according to Wikipedia Boeing had intended to design a new plane but were forced into re-engineing the 737 by market forces:

> In 2006, Boeing started considering the replacement of the 737 with a "clean-sheet" design that could follow the Boeing 787 Dreamliner.[12] In June 2010, a decision on this replacement was postponed into 2011.[13]

> On December 1, 2010, Boeing's competitor, Airbus, launched the Airbus A320neo family to improve fuel burn and operating efficiency with new engines: the CFM International LEAP and Pratt & Whitney PW1000G.[14] In February 2011, Boeing's CEO Jim McNerney maintained "We're going to do a new airplane."[15] At the March 2011 ISTAT conference, BCA President James Albaugh was not sure about a 737 re-engine, like Boeing CFO James A. Bell stated at the JP Morgan Aviation, Transportation and Defense conference the same month.[16] The A320neo gathered 667 commitments at the June 2011 Paris Air Show for a backlog of 1,029 units since its launch, setting an order record for a new commercial airliner.[17]

> On July 20, 2011, American Airlines announced an order for 460 narrowbody jets including 130 A320ceos and 130 A320neos, and intended to order 100 re-engined 737s with CFM LEAPs, pending Boeing confirmation.[18] The order broke Boeing's monopoly with the airline and forced Boeing into a re-engined 737.


Lifting body designs might work for cargo but modeling exercises have shown they aren't practical for passenger service. The rows of seats would be wider, and so the passengers out at the ends would be subject to too much vertical movement during turns. This would cause motion sickness for many passengers, or at least be unpleasant. And more emergency exits would be needed for the same passenger count in order to meet evacuation time limits.

Really it's worse then that. The 737Max is a good example of technical debt. The plane more or less is the same fuselage as a Boeing 707 (with material modernized). Boeing went back to the well one time too many instead of doing a cleanroom design for it's narrow-body fleet.

Boeing has hinted several times that they want to look at a Blended Wing Body (BWB) design for the next generation of planes. That time may be now.

It's hard to think of many other legacy systems that have been around since at least the 60s like the 737 MAX has. The NYC subway system is one that immediately comes to mind.

The engines on many light single-engine aircraft (e.g. Lycoming O-360). While variants and modest improvements have been made, it is still the same 1950s engine at the core.

Safety requirements have evolved massively since then, but if you keep making the same engine with modest changes it can grandfathered. Kind of makes you wonder if grandfathering should have a limit (e.g. 50 years?).

The Soyuz definitely counts too. It was designed in the 60s and hasn't changed too much since then.

The craziest aspect of the Soyuz that makes it obvious how legacy it is is how its engines are ignited: https://www.popularmechanics.com/space/rockets/a19966/russia...

You literally put something not too dissimilar from a wooden chair directly beneath the rocket nozzles, then set it on fire right before launch. When the rocket's engines start up, the fuel is ignited by the fire from the burning wood, and combustion begins. Now that's legacy. It's also reliable enough that they've never had to redesign it.

"Wooden chair" seems like a rather drastic oversimplification.

Yeah, the bulk of the device, and its combustible material, is made of wood. But it's also got two igniters (for redundancy) and a sensor for ensuring that the device has successfully ignited. Some engineering thought went into this. Comparing it to putting a burning piece of furniture under the rocket nozzles makes the whole design sound a lot more jerry-rigged than it really is.

Notably, the ignition device, while seemingly low tech, works pretty much without flaws and fails safely.

“This setup is not as simple as a regular match, but it is surprisingly reliable and has worked for six decades on hundreds of rockets”

So why attempt to fix such a safety critical thing?

The one reason to go from external match sticks to internal ignition using hypergolic fuels is to be able to relight in flight after an engine shutdown.

Yep. Which seems like a fairly unnecessary requirement for an expendable first stage.

I'm no rocket engineer, but my guess would be that, if you're shutting down an engine at that stage in the flight, your only realistic options are going to be redirecting fuel to the remaining engines for a longer burn, or cutting all engines and activating the launch escape system. I can't imagine there's any practical reason to plan on shutting down and re-igniting engines during the first stage of a launch, and I also can't imagine trying to decide whether an engine is safe to re-light during the few seconds you'll have between when you executed the emergency shutdown and when it's time to jettison the first stage entirely.

> Kind of makes you wonder if grandfathering should have a limit (e.g. 50 years?).

That's not grandfathering. Grandfathering is "This engine was built before we made the law, so you can keep using it until it breaks."; what you're describing is just blatant selective enforcement.

It means "the aircraft design met applicable requirements at the time it was certified."

It's similar to how I can continue to drive an old car that doesn't have ABS, since ABS wasn't required or invented when the car was built.

The difference of course is with cars we update the rules and apply them to all new cars as of a certain date. We don't allow someone to keep making cars using a design from 1967, at least not without including current safety features.

The term "grandfathered" is used in FAA literature for both pre-date (literally) and the designed before a regulation came into being.

The reason it exists is otherwise the FAA could release a new requirement that would halt a pre-existing production line that didn't meet it.

> Because part 135 operators will have limited ETOPS operations, the FAA has decided to grandfather from today's rule all part 135 airplanes manufactured up to 8 years from the effective date of the rule. For purposes of airworthiness requirements, part 135 operators may use these airplanes in ETOPS without certification under § 25.1535. This is a change from the NPRM, which proposed grandfathering only those airplanes that were on an operator's operations specifications up to 8 years after the rule. Under the NPRM, they would then have had to remain on the operator's operations specifications to continue to operate ETOPS.


> b. This Advisory Circular (AC) provides guidance to affected U.S. operators that operate aircraft outside the United States with aircraft that were never required to be noise certified. If you have such an aircraft, this AC outlines the noise certification requirement dates so you can confirm that your aircraft indeed pre-dates the requirements and should be considered acceptable. We use the term “grandfathered” for these aircraft. We strongly recommend operators of such aircraft to use the FAA form in Appendix 1 that includes a grandfather clause.


> These seats should provide adequate vision to the pilot's panel and forward windows in visual system models. "Grandfather rights" prevail on previously approved simulators; however, efforts should be made to improve surveillance visibility. These observer seats need not represent the aircraft seats.


What’s it called when the thing was designed before the law, and therefore allowed to be manufactured for eternity?

"Criminally[0] negligent duty-derelictive non-enforcement of safely regulations"?

0: or not-technically-criminally if they planned ahead well enough.

Exactly. If the new standards are so extreme that it’s uneconomical to build to, it just makes old investments more valuable and impossible to enter the market.

Unless the old technology is ruled illegal.

Which is both fair and appropriate and proportional if it means that people don't die.

The IBM mainframe would be a good example in our world, though the really strong tie is probably MVS and TPF from the 70's, rather than the 60's.

That's much more 'legacy' than the 737 is, though. The 737 MAX was projected to be the best selling airplane of the modern era of aviation, and it's a legacy system. Yeah, there's still some old IBM mainframes lingering on in some places, but they're not remotely close to having the same market dominance.

Imagine if all of the cloud computing data centers ran off legacy IBM mainframes -- then it'd be a more akin situation.

These comparisons are off a bit, though, because development lifecycles in airplanes are longer than for computer hardware.

In our world the Unix operating system might be the best analogy. It’s an old design that’s been modernized in different ways, but ultimately has the same general interface (libc) that is still essentially the same.

It’s not a perfect anology, but probably better than systems that aren’t really changing much anymore.

Unix is a very apt example. We wouldn’t be piping text around if it were designed today - but it does work decently well.

It's not text, it's opaque character streams. (C 'char' type)

Practically speaking, for the flavors of UNIX like systems most will work with this is equivalent to streams of 8 bit, clean-transferred, characters.

Any other kind of transfer system that you might want to build can be built on top of that. This interlinking design is the minimum viable product, and also the lowest level one. Comparing to a more recent design, such as PowerShell, transfer of objects limits flexibility and requires more shared infrastructure.

Or Intel‘s x86 architecture.

The parent didn't have a "market dominant" qualifier...

Yes, the IBM System/360's instruction set is still a part of IBM's current mainframes, so customers can still run their 1964 programs. (Does anyone know when IBM dropped backwards compatibility for their 1950s architectures? They provided emulation for 1401 and 709 systems for a long, long time.) It will be interesting to see if the 360 or x86 instruction set lasts longer.

The B52 since 1955.

And the C-130. An extremely successful aircraft.

Lots of ammunition is really old, e.g. 1921 (but really a slight tweak of a 1906 design): https://en.wikipedia.org/wiki/.50_BMG

Lots of good examples in firearms. Forget just ammunition, there's old firearm designs that regularly see use today. Some examples would be the 1911 pistol, or the M2 Browning machine gun (1918). And there's entire classes of firearms that are essentially unchanged from that period, like bolt-action rifles, most of which use a Mauser action which was designed in 1895.

The key difference here, I think, is that there haven't been radical new improvements in firearms. You can clearly point out where the faults in, say, legacy airplanes lie, but guns (at least in civilian uses) haven't changed as much. Now, granted, for military uses, the modern assault rifle is a far cry from the aforementioned guns.

That's inaccurate at best. The shell size is mostly unchanged. The powder inside, the materials and the manufacturing have progressed.

Just looking at the diagrams in that page, I doubt that measurements were defined with a 0.001 inch accuracy in 1921.

The powder really isn't that different. The max pressure allowable is part of the cartridge specification, which hasn't changed since the original design. Otherwise you'd have newer ammunition blowing up older guns, which clearly would be unacceptable. You can take modern .30-06 ammo, which is still a very popular hunting cartridge, and fire it in an M1903 Springfield with no issues.

Well, not an expert, but have you got any information here?

The technology was pretty mature in 1921, there were a few rounds of refinement before that. Machinists definitely talked about "one thou" back then. The same M2s which fire these things have been in production for over 80 years. I'm sure there have been small tweaks, special versions, etc, but basic compatibility has remained.

Vernier calipers capable of measuring 0.001 inch were available in 1851. Agree on your first point, however.

SABRE dates from the 50s.

Yes, but if you draw the line at where what they had resembled what they offer today, it's around the 1970's.

American Delta rockets and Russian Soyuz rockets come to mind.

The T-38 first flew in 1959.

do you use banks?

Went back to the well for marketing reasons. Keeping the shape, but only the shape, so they could use the 737 name.

I see there are those who agree with the FAA's decision that the MCAS didn't represent a change worthy of mention, nor new materials, new factories... "nothing to see here folks." I think by now it's obvious that was a "puzzling" decision.

For those who might be interested: the 737 type certificate data sheet. Revised 64 times since 1967 and covering original, Classic, NG, and MAX.


I would be interested in an organizational investigation of Boeing, the FAA and the airlines similar to the investigation of NASA after the Challenger disaster.

I keep reading comments that it was just corporate greed or just regulatory capture or Boeing should’ve just built a new plane. But I suspect it’s not that simple, and that there are lessons to be learned beyond what the crash investigations uncover.

For example, from reading the Wikipedia on the 737 Max it sounds like Boeing was pressured by the airlines into producing the Max after Airbus was able to ship a more fuel efficient 320.

Usually you expect competition to lead to better products so it would be somewhat ironic if the competition from Airbus is what forced Boeing’s hand on producing the Max. Without that pressure maybe Boeing could have told the airlines to back off while it produced a new design.

I’m no airplane expert but I find the trend lines of the industry interesting and one trend has been how fuel prices and long lead times on fuel efficiency has lead to formerly continental planes now flying intercontinental. AirBus has really done this well and airlines love to be able to reuse resources. With the MAX there is a sense that airline execs are spooked by a new plane since it requires a ton of new training and maintenance outfitting. So Boeing, possibly smartly, tried to create what these airlines would pay for (a fuel efficient airplane) that would marketable be super similar to a very popular plane. But this lead to rushing the process and making too many affordances and hiding the differences from airlines and training material for fear of spooking business execs worried about cost.

The 737 is not uncomfortable because it is narrow but because it is not tall enough. The circular fuselage makes it so.

The much smaller Embraer 195 is vastly more comfortable than the 737 because it is shaped according to the passengers.

A clean sheet design for the 737 would have a big positive impact for the industry. Today it is another reason to take the bus.

"The change, however, affected the plane’s aerodynamics. Boeing discovered the new position of the engines increased the lift of the aircraft, creating a tendency for the nose to pitch up."

The problem wasn't just the ground clearance. the problem was that aerodynamically had issues (not so well balanced probably). So when they were trying to fix "this" something else was breaking and so on. That's why they introduced this MCAS software in the first place. It wasn't some new fancy tech but it was just a system to help with the bad 737's aerodynamics.

Right, but it sound like the aerodynamics have been deteriorating model-by-model, as they move away from the original design.

This creeping change must be really hard to regulate right. A terrible new design you could refuse to certify, but a terrible one which is only a few percent worse than the one your predecessor certified a decade ago... it would take a lot of confidence to say no.

<Boeing added larger, more powerful engines, but that required it to reposition them to maintain ground clearance. As a result, the 737 can pitch up under certain circumstances.>

This has to be THE MOST Stupidest decision i have ever read about.. either extend the landing gear or start from scratch.

Aeroplanes should be naturally neutral Longitudinally and also have longitudinal static stability it should never require complex control systems to achieve any of this.

It should never have passed certification.


"Under certain circumstances" is, as I understand it, close to stall. Many planes have less-than-desired stall characteristics. The tendency of T-tail planes to go into an unrecoverable deep stall comes to mind.

Or have you ever flown a Cessna 152 or 172? They have the same tendency to pitch up strongly when power is advanced at low speeds. This is probably the most common training aircraft in the world, and yet have a behavior that may cause them to stall at exactly the most inopportune time. And they're flown by students, not airline transport pilots!

The 737 MAX does not require the MCAS. The pilots could be trained to expect and handle this characteristic. (And they are, as it is apparently, to a lesser degree, present on all 737 models.)

The big problem lies in the design of the MCAS system and the fact that it was decided to rely on this and not train pilots on the new behavior.

putting and engine on the wing in an undesirable position because of the height of the landing gear is stupid..

just because "a pilot should be able to handle it" is not acceptable.

Aeroplanes traditionally are designed with as much passive self correcting features as possible for safety.

When a plane has a real emergency the less a pilot has to actively deal with the more likely people are going to come out alive.

The evidence is clear that this was a huge mistake.. many people have already died.

It's not nice that so many writers parrot that the solution is "more pilot training." Under the hypothesis that there needs to be a solution (one that I think is correct), then the answer of more training amounts to mere hope. One hopes that the pilot a) knows and b) remembers to turn off the autopilot when the emergency starts.

Pilots can fly planes with unusual characteristics when properly trained. That is the point of the type rating. I feel the 737 MAX needs its own type certificate distinct from the rest of the 737s and that should solve the problem, since the MAX has very different flight characteristics in certain situations.

Either that or ground the plane permanently for being poorly designed and unsafe with any amount of training.

If the FAA does anything else, it will harm my faith in the organization, which is already at a recent low.

NB: Jet pilot, not 737 pilot.

737 pilots have said that they are already trained in the tendency of a 737 to pitch up when full power is applied at low speeds, which apparently is present in all 737 models. (Stall recovery procedures for a 737 apparently specifically detail that you must lower the nose first, and then apply power (and not too much power) as applying full or higher power near stall speed will overwhelm the pitch authority.

What's not clear to me is how much worse the characteristics of the MAX are compared to the NG or earlier 737s? Would it really be unsafe to fly without the MCAS, or was this just the route Boeing took to avoid having to retrain pilots?

Money's on a last minute fix to the problem and engineering being told the launch couldn't be delayed.

We're told that MCAS only activates with the autopilot already off.

I agree with you that pilot training doesn't seem to be a plausible answer given that the Lion Air crash was world news, resulting in retraining, and yet that wasn't enough.

Thank you for the insight! Just for clarity, I'm using "autopilot" as a catch-all term that refers the automated systems that plausibly seem to have caused these crashes.

...which is a pretty confusing thing to do, given that there are multiple such systems, one of which is conventionally called "autopilot".

The solution isn't so much "more pilot trainimg".

It's that pilot training wasn't done that absolutely needed to have been.

A three page if even that description of the system, and it's inputs/outputs/controls should have been sufficient for a pilot to be able to build up a mental model to deal with a faulty AoA sensor, and modify their piloting technique to stay within a more conservative maneuvering envelope while operating the plane.

Whether that would have been sufficient to prevent both disasters I don't know, but it sure seems like it would be a small trade off that even in the absence of any other system hardening could have made the difference between no survivors, and an emergency landing.

Pilot training is a major reason air travel is so safe.

> One hopes that the pilot a) knows and b) remembers to turn off the autopilot when the emergency starts.

This is where training kicks in. After hitting the scenario 10 times in the simulator, the pilot doesn't remember; they _do_.

They should have a philosophy of building aerodynamically stable aircraft rather than relying on software and sensors to produce an illusion of stability.

Consider that before WW2, pilots had direct connection between the control and moving the surface. After WW2, planes got too big and heavy for that, and hydraulic boost was added. Starting with the 747, where it is not credible a pilot could control it with muscle, the switch was made to fully powered surfaces.

The trouble with that is the pilot's controls simply crack open a valve, and hydraulics do the rest, meaning the pilot gets no feedback via forces on the control column. He'll "overcontrol" the aircraft leading to a crash. The solution is the addition of a "feel computer" which pushes back on the stick making the airplane feel like direct control. Safe flying relies on the feel computer working properly.

Also, the travel of the control surfaces is automatically reduced as speed goes up, because full travel at high speeds will literally tear the airplane apart.

So right there are two flight critical computer systems on jets, and modern air travel would not be possible without them. Each produces an illusion to the pilot.

Right, this makes sense. The system you describe magnifies or reduces the signals going back and forth between the pilot and the control surfaces. It can be modelled linearly and done with an analog system.

However I'm referring to a computer literally trying to hide a stall that occurs when you pitch the nose up by producing an inverse signal out of thin air that has no input from the pilot. Why wasn't this needed before the max moniker was added to the plane?

Seriously, build the plane so that this doesn't happen. Don't create some MCAS system to hide it.

MCAS was there to prevent a stall, not hide it. All airplanes will stall under varying circumstances. Much of pilot training goes towards avoiding and recovering from stalls, but they still result in crashes.

The stall didn't exist at that configuration previously. MCAS prevents the pilot from even knowing about this stall by preventing the plane from even going into that configuration. In other words it hides the stall by preventing you from maneuvering the plane into it. It fits the definition of "Hide." But that's besides the point, we're arguing about semantics here.

I'm saying similar to the older design make the aerodynamic shape good enough so that it won't hit that stall when the plane is flying at that same configuration. I'm also saying that the control system should only magnify or reduce the signals the pilot sends to the control surfaces rather than actually sending it's own override signals itself. That's it... pure and simple.

> MCAS prevents the pilot from even knowing about this stall by preventing the plane from even going into that configuration. In other words it hides the stall ...[.] [T]he control system should only magnify or reduce the signals the pilot sends to the control surfaces rather than actually sending it's own.

What about fly-by-wire? Many aerodynamic controls are assisted by avionics, and have been for decades. The MCAS system is not fundamentally different. The intent of the system was to put the plane where the pilot wanted intended it to go, similar to how a fly-by-wire system will keep a plane flying straight if a pilot doesn't touch the yoke.

The problem seems to be that the MCAS system breaks under certain conditions, however the fundamental idea is sound. Planes have been doing similar control adjustments for decades.

No its not. Fighter planes like the F-16 are designed to be aerodynamically unstable for higher maneuverability. For this case an incredibly fast control system is required to make minor adjustments to the plane to maintain stability. This control system is in constant control of the plane and the entire jet aircraft is designed to operate in such a way.

The 737 MAX was made more aerodynamically unstable then the original 737. A partial control system was put in place to hide this instability should the pilot maneuver the plane into the configuration that will trigger a stall. It's an entirely different situation than the jet plane above.

Think of it like a bandaid to cover up a problem. I'm saying don't use bandaids. Design the plane correctly from the ground up.

Airbus also loads up its avionics with protections against dangerous pilot input. Quantas Flight 72 is an example of when that went awry with a random pitch down effect.


> A partial control system was put in place to hide this instability should the pilot maneuver ... It's an entirely different situation than the jet plane above.

I wasn't talking about military craft, I was referring to commercial aircraft that also have fly-by-wire capabilities, of which there are plenty of examples.[1] The MCAS is not fundamentally different from these other existing systems.

[1] https://en.wikipedia.org/wiki/Fly-by-wire#Airbus/Boeing

Right, except military aircraft is where FBW was developed and very very much required for performance. Such is not the case for passenger airliners. Only two Boeing planes use FBW, the 777 and 787. Every other plane does not use it including the 737.

The MCAS is fundamentally different in that it is a partial FBW system that they tacked on on top of an existing control scheme that is not FBW.

It's a random partial FBW component grafted onto something that's completely different. Sort of like a mod chip in a console. It's a hack. To drive home the exact opposite of what you said: "MCAS is abnormal and fundamentally different from many existing systems."

And one more thing - trim tabs should be separate airfoils, not an electronically set setting exactly for the purpose of reliability.

That way, an elevator should always have more authority than the trim tab, and be able to overpower it if it fails.

On other hand, independent trim tabs can work as reserve control surfaces if hydraulics fail.

The same design thinking is behind having engine reverse locks being physical parts, separately powered, and operated independently of electronic engine controls: a hard lock will always be able to overpower haywire electronics.

Same with emergency chassis deployment system on smaller planes, they are often designed to be able to physically disconnect/overpower motors that pull chassis, or chassis cowers if they jam or electronics bugs out.

Trim controlling stabilizer, yoke controlling elevator, and elevator being unable to counteract the more extreme ends of trim settings, is part of the original 737 design. They all behave that way, and 737 pilots are aware of this, and know they must neutralize trim to have ample elevator control.

However, only the 737 MAX has MCAS, and MCAS doesn't take airspeed into account. An appropriate corrective action for a low airspeed stall, will be way too aggressive for a high airspeed stall (a rare event, suggests overly aggressive dive recovery leading to a stall). And designers + regulators assumed that pilots would treat such behavior the same as an ordinary runaway trim incident, but does it seem ordinary or is it a shockingly bad overcorrection?

We've seen many times the scenario where pilots become so surprised at a particular situation that their rational ability to work through the problem can be compromised. Yes, transport pilots are substantially trained to avoid such a compromise, but problems at low altitude combined with the lack of a consistently positive rate of climb is an urgent situation. Pilots are hyper aware that recovery from either a dive or stall consumes a huge amount of altitude, which you simply don't have at low altitude. That itself could delay proper decision making: i.e. automation has betrayed us for unknown reasons, disable it now.

With MCAS disabled, what's the 737 MAX stall behavior like? Is it meaningfully easier to unwittingly get into a stall?

Trim would never be a separate airfoil because that would be horrendous for efficiency. Instead they have an all-moving tailplane.

They could have prevented this issue by slightly increasing the tailplane size to restore stability margin, but they didn't presumably for development cost and fuel efficiency reasons.

Guessing that changing the tail might also go beyond what would be allowed to stay within the 737 type rating?

Are you really sure that it will be "horrendous?"

That thinking remind me of people who try to economise on few microns of plastic we see in electronics OEM industry. Pro managers trying to shave microns from casing, ending up with multimillion dollar recalls, and that for savings not even amounting to few snickers bars.

You're adding an entire new system. That's more wiring, more programming, more hydraulics, more moving parts, etc. It's far from trivial.

Far from trivial, but it was done in the past nevertheless. All things above were going away one after another, and replaced with often ineffectual electronic safeguards.

Any design will have consequences, and is thus relevant. But changing it has as much of a consequence as not changing it. Example: Airbus design has "stick" type yokes that are not linked together, and are out of the field of view. If one pilot is using improper inputs, the other pilot won't know. Boeing planes have mechanically connected yokes, both pilots are aware of flight control inputs. Is one a better design than the other? I suggest it's the wrong question, the right question is, what are the consequences of the design on cockpit communication, management, and training and can we meet the ensuing requirements?

The calls for more and better training make me wonder why we don't just train for the relatively rare over pitch characteristic that the MCAS system is intended to prevent. If we can't actually train for that then why are we suggesting that the solution is just being quick to turn the system off? What level of unreliability is required to make it better to just never turn it on?

I wonder why Boeing didn’t they design new, longer landing gear instead of moving the engines forward?

Then the engines could have remained in the same position. I am sure most airports could handle a taller airplane.

Not a direct answer, but what you’re more or less asking is “why did they not modernize the 757 instead?” The answer is type ratings. Pilots can fly certain classes of planes based on common training. I.E. the 757 and 767 are common enough you can be trained for both at the same time. The issue with changing geometry of the 737 is you have to then retrain pilots against a new type instead of letting them take an orientation course on the new plane.

Honestly if the accidents did not happen we’d be calling them geniuses. After all, that’s why the 737 is the best selling plane ever.

Not really - they definitely could have extended the landing gear of the 737, as evidenced by the fact that they did exactly that for the 737 MAX 10. To handle the extra length they added telescoping landing gear that are about 10 inches longer than on the other variants

Lengthening the landings gear did nothing to change the engine placement. It was done to allow for type commonality.


Even with the extension it’s not as tall as the 757 or A320 series.

True, it wasn't done to solve the problem of the engine placement. It solved an entirely different problem. But it does show that the changing the landing gear whilst maintaining type commonality wasn't impossible.

Is there a point where the pilot is expected to remember so many procedures that it’s no longer possible to know all of them? You’d forget something critical here and there.

> Is there a point where the pilot is expected to remember so many procedures that it’s no longer possible to know all of them? You’d forget something critical here and there.

I don't think they have to remember all of them. Pilots rely heavily on checklists to make sure procedures are followed correctly and few mistakes are made.

True. However, there are "memory items" that have to be done from memory when the situation does not allow time for a checklist. Stall recovery is one of them, which is exactly when the MCAS is supposed to activate. I'm pretty sure runaway trim is another since it won't take long to end up in a very bad attitude if this happens, especially at high airspeeds.

There’s just not much time as often cited as the problem when a pilot makes a mistake in a stressful situation, so seems like the add more training mantra ignores a more fundamental issue.

When US Airways Flight 1549 lost both engines, Captain Chesley "Sully" Sullenberger intentionally skipped some items on that checklist. The checklist had been written with the assumption of higher altitude and he didn't have enough time to go through everything.

> Not a direct answer, but what you’re more or less asking is “why did they not modernize the 757 instead?” The answer is type ratings.

But if there's already an existing type out there that's a better fit, why not just use that?

The 757 didn't sell nearly as many planes in total as the 737. There's entire airlines (e.g. Southwest) that only use 737s and have their entire business built around them. The 737 MAX was designed for this use case; a newer 757 might not have sold to these carriers at all.

Plus, the 757 has some features on it that make it more of a longer range, transcontinental jet than the 737. That makes it more expensive. If all you're using it for is regional air service (the existing mission of the 737), then it's too expensive for the job. The 737 is the Honda Civic of the skies; the 757 is more like a (insert more expensive car model here).

Because the plane was basically designed for Southwest Airlines and American Airlines, and they demanded a new 737, not a new plane.

“If I had asked people what they wanted, they would have said faster horses.” -- Henry Ford (purportedly)

There is still a significant niche market for faster horses.

Designing longer landing gear might invalidate the common type rating, as the plane's entire fuselage and wings would be higher off the ground during take-off and landing, which might be non-trivial. Plus it would have different heights compared to previous 737s, which could cause problems with jetways, staircars, maintenance vehicles, luggage ramps, etc.

There's not many things they could've changed about the existing design without it being enough unlike the existing design to effectively count as a new aircraft (logistically if not certification-wise). They could move the engines, but they couldn't change the ground clearance.

The MAX-10 has longer landing gear by 10 inches and I'm assuming it shares the same type rating as the rest of the 737 line. So it is possible.

However the MAX-10 required a telescoping landing gear to achieve that so I'm assuming the real reason to not do so earlier was cost to design, cost to manufacture and weight.

I think the MAX 10 landing gear changes allow for a higher rotation angle (without tailstrike) on takeoff/landing, not actually increasing the engine-ground clearance.

Also the MAX 10 hasn’t been built or certified yet, I’m sure there will be a lot more scrutiny when that does happen.

I think I read that longer landing gear wouldn't fit into the fuselage without major reconfiguration.

The MAX-10 got around that by having a telescoping landing gear and was able to get 10 inches more as a result.

Does anyone know why Tunisia refused to work with the FAA?

"First introduced in West Germany as a short-hop commuter jet in the early Cold War"

The Cold War started in the second half of the 1940's, so I rather doubt that is true.

My family would have flown on a 737 Max this morning. Flight was cancelled. The FAA was literally the last regulatory org to ground the 737 Max... I'm livid that corruption and bias put my family at risk!

Yeah. It is strange that they were the last regulatory org to do so. . In America basically every governmental body is in the pocket of big business.

This is an interesting take on it: https://www.nytimes.com/2019/03/15/opinion/federal-aviation-....

Basically, the US has taken the view that any regulation should be strongly supported by facts and evidence, while many other countries are more willing to regulate on a precautionary basis, before all the results are in.

A speculative opinion piece. Here is another one: https://www.vox.com/policy-and-politics/2019/3/13/18263719/b...

Two planes of the same model crashing of is a statistical anomaly. Science is the correlation of hypothesis with statistical evidence. If I made a hypothesis saying that something is wrong with the 737 max then two 737 planes crashing is valid supporting evidence. Therefore, the decision to regulate is scientific.

The decision to not regulate and claim that there is no scientific evidence when their clearly is... is more likely to be a cover story to protect big businesses then an actual decision made to support your safety.

N.B. no American carriers have crashed a MAX, and they are flying quite a few of them.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact