Hacker News Reader

>Secondly, its not a software problem, the plane is structurally deficient.

Aerospace engineer here. This is incorrect. The MAX exhibits different pitch behavior at high angles of attack than the 737NG, which would normally require a different type certificate. MCAS was their attempt to change the handling characteristics of the MAX aircraft so that it was sufficiently similar to previous iterations of the 737.

This is not the same as being "structurally deficient". The fatalities arose from a combination of ethic failures at Boeing, financial pressures from large customers, bad safety design of the MCAS system, and inadequate pilot training.

Plenty of aircraft flying today exhibit nose-up pitch tendencies at high angles of attack. This is not an intractable problem; it's just not a problem that Boeing could solve while keeping to the 737NG type.

"This is not the same as being "structurally deficient". The fatalities arose from a combination of ethic failures at Boeing, financial pressures from large customers, bad safety design of the MCAS system, and inadequate pilot training."

Boeing used a 737 airframe for economic reasons (it wwas way cheaper, didnt require pilot retraining), but needed more ground clearance for bigger engines. The 737 design cant be practically modified to have taller main landing gear. The solution was to mount to engines higher on the plane. This made the plane more unstable, so they slapped the software solution on. This "fixed" the problem, but then the plane was more complex to fly, which led to the crashes. But Boeing didn't want a plane that required retraining, so they hid the complexity of the software solutions, which led to crashes because the pilots didn't know what was going on.

Sure you can say the plane was structurally fine, but this is semantics, they created a plane that was unstable because it was cheaper.

It really annoys me when people just dismiss a point with the phrase "this is semantics". Semantics is meaning! You're ignoring the meaning of what you responded to. If you don't think the point was valid, you should give a reason.

No I'm not, the plane is flawed. Boeing created a flawed plane to cut costs. Then they slapped software on it and called it fixed. Sure because it can fly you could call it "structurally sound", but its not. The plane is a pile of trash

I'm not saying you're wrong, just that it's irritating that you aren't supporting your opinion. You just state your opinion and follow it with something dismissive. Twice now.

Interpreting your comments in the most charitable way, you must have reasons for your opinions that convince you, so it would contribute more to the discussion if you explained why.

Nobody rejects "semantics".

Yeah. If anything the phrase should be more like "syntactics". Semantics is anything but semantics.

When someone says “it’s just semantics”, they mean that the argument is only over the definition of words, not over that actual meaning.

That may be what they think they mean, but I can't recall an instance where that was accurate, including here.

Apart from that, I don't have to like people saying "it's just semantics" if they mean the exact opposite.

>Sure you can say the plane was structurally fine, but this is semantics, they created a plane that was unstable because it was cheaper.

In the aerospace industry, semantics matter. "Structurally deficient" implies that the aircraft structure itself (i.e. the spars, ribs, load-bearing skin, etc.) is unsafe. This is not accurate, and neither is it accurate to say the MAX is "unstable". These are terms of art with precise engineering meanings.

Precision in language matters, and to dismiss it as "semantics" pre-empts any meaningful discussion of the issue at hand. It's like saying float vs char is "semantics". Those differences matter.

Yes it is accurate. The plane crashed twice. They put engines on top of the plane that weren't made for the body.

It did crash twice, but not because it is structurally unsound. And not because it is unstable.

edit: Also, when you say the engines were put "on top of the plane" it makes me question if you've ever actually seen a MAX 8.

You should stop posting about issues that you are completely clueless about. I'd point out where you're wrong, but in short: everything.

Well, the plane did crash twice. Everything else seems wrong though.

Aerospace engineer here. This is incorrect. The MAX exhibits different pitch behavior at high angles of attack than the 737NG, which would normally require a different type certificate. MCAS was their attempt to change the handling characteristics of the MAX aircraft so that it was sufficiently similar to previous iterations of the 737.

I'm sure the pitch up tendency is worse on the MAX, but I don't believe MCAS was added to enable the re-use of the same type certificate. The handling problems were worse than anticipated as is evidenced by the low-speed activation of MCAS.

My guess is that the 737 MAX on its own wouldn't meet the handling requirements. It's been a while since the Ducommun scandal, but more recently Boeing was caught knowingly installing unsafe parts (NG/MAX slat tracks that didn't meet spec) and forging repair documentation for Air Canada. Pardon me if I take any claims that Boeing is spewing with a lump of salt.

inadequate pilot training.

It's hard to train for something that the manufacturer won't disclose or implement in the sims. It's even harder when the remedy (moving the trim wheels) is physically impossible.

>the re-use of the same type certificate.

That was the demand from the airlines; that there be no training delta. For that you need to remain within the same type, or at least close enough that they're equivalent.

>My guess is that the 737 MAX on its own wouldn't meet the handling requirements.

Correct.

>It's hard to train for something that the manufacturer won't disclose or implement in the sims. It's even harder when the remedy (moving the trim wheels) is physically impossible.

Yes. The training the pilots received on the differences between the MAX and the NG (i.e. no training) was inadequate to prepare them to fly the aircraft safely.

That was the demand from the airlines; that there be no training delta. For that you need to remain within the same type, or at least close enough that they're equivalent.

Those are two separate issues. There are currently four generations of 737 on the TC[1] and sim time is required between the different generations. Look at the 777 vs 777X, there are radical changes and yet the 777X is (for now) grandfathered on the 777 TC. A change requiring a new TC (or STC) is an order of magnitude more dramatic.

The training mandate drove the poor design of MCAS but wasn't the reason MCAS was implemented. Even with training the MAX would never have met FAA regs without something like MCAS.

The list price of the MAX is around $100 million, and Southwest wanted $1 million per aircraft if training were reqiured. If it were simply a matter of more appropriate training Boeing would've already taken the 1% haircut and moved on. American talked a good game about not wanting the MAX at all if training were required, but let's face it they'd probably rather take the training hit than wait in line for an A320 production slot. Instead Boeing shuttered the production line because the problem is simply not that easy to solve.

If it were simply a matter of stick force gradient Boeing could've likely just modified the elevator feel computer. Instead Boeing went the route of moving the whole damn stabilizer, a particularly odd choice because the stabilizer moves very slowly -- slower than a reasonable person would want to prevent an impending stall.

1: http://rgl.faa.gov/Regulatory_and_Guidance_Library/rgMakeMod...

Thanks for the clarification; I don't work for the airlines so I wasn't aware of the sim time requirements.

Here's the detailed list of differences training requirements:

http://fsims.faa.gov/wdocs/fsb/b737_rev_15.pdf

All these arguments are about engineering.

It's no longer an engineering problem.

Now it's a trust problem.

Boeing, the FAA, and the airlines have demonstrated that they can't be trusted to make the right decision when passenger safety conflicts with quarterly earnings per share.

They'll roll the dice and hope they can fix the problem, as cheaply as possible, before the next catastrophe.

So it doesn't matter if they find an engineering fix.

The public will never trust that it's really fixed, and re-branding this dud won't fool anybody.

The sooner Boeing figures this out, the less money they'll lose.

This is why Boeing never should have been allowed to self certify. Decisions about Matters of safety should have been given to people who didn’t have to weigh it against earnings per share.

>> The handling problems were worse than anticipated as is evidenced by the low-speed activation of MCAS.

I'd say the handling is bad enough to require something like MCAS in order to be safe. If not, they would have just disabled it and ate the cost of training pilots for the new variant. Instead they think it's better (cheaper) to try to fix MCAS while letting planes pile up that will likely require extra training on the new fix anyway.

In other words I agree. It probably doesnt fly well without help, as opposed to just flying differently.

> The handling problems were worse than anticipated as is evidenced by the low-speed activation of MCAS

Where the heck did you get that from? The MCAS failures were caused by a failure of the AOA sensors, not any kind of handling characteristics.

Where the heck did you get that from? The MCAS failures were caused by a failure of the AOA sensors, not any kind of handling characteristics.

MCAS was originally intended to handle high speed, high altitude situations. Flight testing found that there were handling problems at lower speeds and altitudes that MCAS could "fix". So MCAS was given roughly twice the authority it was originally designed with and it was then programmed to activate in lower speed situations as well. The FAA, of course, was kept in the dark.

If you're curious go through the Seattle Times articles.

I tend to agree with assessment. Now, the original article to which these comments apply would have strength if it had shown where the putive losses are hidden. Still there appears to broad parallels between ethical lapses and claims that Boeing cares more about financial engineering than customer safety at the top.

Oh so sweet this history revisionism.

“ The 737 engines used too much fuel so to make a more attractive plane for poorer countries ...“

It was AMERICAN Airlines and Southwest (of yes the USA) who pushed Boeing into creating the Max, not “poorer countries” (unless you count the US among those)

Yes this is more accurate—Southwest in particular is a huge 737 customer, and they wanted to be add longer legs with minimal changes (same ground crew, berthing and piloting). The versatility of crew is a key part of their business model.

> These planes cannot fly without a computer adjusting them to keep the balance, something that is not true of any other major aircraft.

I think you mean any other major civilian aircraft. No recent major combat aircraft can fly at all without significant computational assistance.

This does not excuse the 737-Max shitshow in the slightest...if anything it makes it worse.

> any other major civilian aircraft

Aside from Airbus? Isn't basically every Airbus in the sky at this point flown by a computer with the pilots making requests for what they'd like it to do? This has long been a fundamental ideological difference between Boeing and Airbus, with Boeing just now starting to get into the same game Airbus has been doing for a while.

I think only the A300 and the A310 are not fly-by-wire.

Don't forget Boeing's own 777 and 787, both are fly-by-wire also.

Stretching my knowledge about those planes a bit, so it'll probably show, but I think the distinction I'm going for is not how the controls are activated (hydraulic vs electrical), but what role computers play in day-to-day flying. Boeing has long held that pilots are in charge, and as a result Boeing planes are pretty manual without a great deal of automation. Airbus has a differing philosophy, where the computer is fundamentally in control of the plane, and the pilots manipulating the controls are asking the computer to do something for them -- it decides if that request is valid for the flight mode and then acts to make it happen. One of the most well-known relatively recent failures of the Airbus system was the Air France flight, and IIRC a main cause was the computer switching to a mostly manual flight mode (due to a sensor failure) and the resulting change in how the flight controls work confused the copilot to the point that he stalled the plane.

In any case, that's what I was trying to get at. The game Boeing is playing with MCAS is a step in the direction that Airbus has a lot more experience with.

> One of the most well-known relatively recent failures of the Airbus system was the Air France flight, and IIRC a main cause was the computer switching to a mostly manual flight mode (due to a sensor failure) and the resulting change in how the flight controls work confused the copilot to the point that he stalled the plane.

That's a rather misleading take, given by people pushing a certain agenda. The first officer in question was hauling the stick back as far as it would go. This resulted in the plane stalling. That's not a matter of confusing flight modes or computer behaviour - pretty much any aeroplane will stall if you pull the stick back all the way (and what else would you expect to happen?)

I think the confusion came from the fact that the captain didn’t get haptic feedback from the joystick that the first officer was pulling it in the opposite direction, plus a lack of communication between the two, or was that a different crash?

> That's not a matter of confusing flight modes or computer behaviour - pretty much any aeroplane will stall if you pull the stick back all the way (and what else would you expect to happen?)

As far as I know, on the normal flight mode (normal law) the computer behavior of an Airbus is that it will not stall if you pull the stick back all the way - it will go as far as it can before stalling, and no more. But that depends on the sensors which had failed, so the computer had to switch to a mode which didn't have that protection.

Airbus is definitely a civilian manufacturer and all their aircraft are strictly fly-by-wire. But as far as I know they are at least dynamically stable, unlike many combat aircraft and the 737 Max.

Update after reading na85's comment: The 737 Max is dynamically stable.

Also apparently a few Airbus models are not FbW.

> Also apparently a few Airbus models are not FbW

Only the legacy ones before the A320, namely the A300 and A310. Not really relevant nowadays.

> These planes cannot fly without a computer adjusting them to keep the balance, something that is not true of any other major aircraft. The problem can be solved with software, but its more like a bandaid.

Why is it wrong to rely on software?

I'm not sure I understand the issues you're talking about but many military aicraft are deliberately built to be dynamically unstable and need constant software control to fly.

Read my update, this was a criminal action taken to cut costs

Because people died and will continue to die unless the airframe/engine configuration , which they knew was faulty, is changed to not be garbage.

It is crazy to accept a flawed design when people's lives are at stake. A plane isnt a movie or a song. You cant just fix stuff in post. Do it right the first time.

>I'm not sure I understand the issues you're talking about but many military aicraft are deliberately built to be dynamically unstable and need constant software control to fly.

Some specific fighter variants are built this way, yes. But the risk tolerance for a single-seat fighter aircraft that has to fly combat sorties is very different than for a civilian airliner with hundreds of people onboard.

> many military aicraft are deliberately built to be dynamically unstable and need constant software control to fly.

Yes. It's also well understood that combat pilots accept the possibility of death when they sign up for military service. That's not the risk tolerance that is accepted for passenger travel.

> These planes cannot fly without a computer adjusting them to keep the balance...

First I have heard of this issue, though I haven't been following it closely. Citation, please?

https://theaircurrent.com/aviation-safety/what-is-the-boeing...

http://www.boeing.com/commercial/737max/737-max-software-upd...

I dont have time to find all the sources for this. But basically Boeing tried to make a flawed plane because it was cheaper and they could fly more people around on less fuel. The plane was also designed for runways in third world countries. Then they hired 8$ an hour software engineers to patch up the issues that they knew about (which ended up taking down the plane anyways).

Not all the sources. Just one, reasonably well documented source.

From what I understand, MCAS functions to emulate the similar flight experience of prior 737 aircraft to minimize pilot retraining..l

I think that there are two issues. MCAS was needed because the airplane wouldn't meet FAA rules without it. Fixes for things like that are not necessarily bad. Problem comes from two issues. Not wanting to require retraining meant they tried to bury the MCAS's existence to the point of not having a way to disable it and only it. And second the flight envelope problems were worse than originally expected. So they gave MCAS more control authority.

They ended up with a problem where bad inputs to the MCAS system could only be countered by disabling the electric trim motors. Once you did that the only way to fix the trim was via a manual wheel which didn't have enough mechanical advantage when the elevators were loaded. That itself was a festering problem carried forward from the original design.

Look up MCAS.

Yeah, I know about MCAS, and nothing I have found suggests the airplane is inherently unstable.

In fact, a cursory search shows the opposite.

https://news.ycombinator.com/item?id=19556707

https://www.youtube.com/watch?v=H2tuKiiznsY

If you haven't been following closely, then do a modicum of your own research first before you ask for citations.

The MCAS is exactly the cause of the issues and was the software that caused the crashes because it behaved unexpectedly to the pilots. The whole reason why Boeing leaned on software was because their sales pitch was that this new plane was essentially the same as a 737 so it didn't need any extra training. But this was false. They needed this MCAS because the mechanics of flying the plane were vastly different, I think because of the location of the engines, etc. That's the whole issue, where Boeing lied to the FAA and sold the 737MAX as a drop-in replacement for the 737 when it really wasn't. So if you don't even know this then you need to start from scratch and do your own research.

Everything you wrote it true, but doesn't support the statement:

> These planes cannot fly without a computer adjusting them to keep the balance.

I haven't seen anything suggesting that. If MCAS was completely ripped out of the computers, these planes still can be flown, right? The pilots would need to be trained on the differing characteristics, but it isn't inherently unstable like some military jets are.

The planes would still physically fly without MCAS. (They are not unstable.)

The planes could not legally fly without MCAS, because they fail a required stick force linearity certification requirement. (FAR 25.173)

> the mechanics of flying the plane were vastly different

I suggest you need to continue doing your own research on when MCAS comes into play and why. Especially if you're going to condescend to others.