Boeing 737 MAX

[Revoltingest]

This would have been my expectation as well, but it seems not to be the case. The software is confused by one sensor and doesn't do checks you and I would expect it to.

This is how things appear based upon what we see in the news.
But I'm just speculating, & lack enuf info to state it as fact.

That is what I'd expect from a capable system. This one, it seems, had a fatal flaw. As I read about it, the FAA and Boeing were aware of the issue, but unable to go ahead with the software update due to government shutdown in the US. Whether that's true, I'm uncertain.

Sadly an event like this will probably cause pushback for it is good in general and the only solution in the long run.

Usually, an accident is the result of multiple flaws in simultaneous occurrence.
And some flaws aren't even flaws, except in the presence of other flaws arising.
(Sounds strange, eh.) Combine the probabilities of each occurring, & this is why
airliner accidents have a low probability per mile flown.

I heard a talking head on NPR the other day...an expert in something...say
that designers should make everything fundamentally safe, & not rely upon '
statistics. Ugh....He prolly doesn't even know that even the strongest
aircraft aluminum alloys have only a statistically limited fatigue life.
Alas, the public will get an impression that designers are recklessly
depending upon chance. They remind me of creationists' criticism
of evolution...."It's just random chance!".

The accident is being investigated. Lessons will be learned. And every
aspect of the system, from airplane to human, will see improvements.

 

[Jumi]

This is how things appear based upon what we see in the news.
But I'm just speculating, & lack enuf info to state it as fact.

News generally don't go much into things like sensors. I'd never consider "newsmedia" primary sources. Most journalists don't have any technical training.

This guy(avionics engineer and a former Boeing avionics supervisor), goes more into it:

Satcom Guru: Comparing Ethiopian ET302 to Lion Air JT043, JT610

A Boeing revision to the MCAS FCC software is not available, even at this time, to require dual AoA sources, to limit the authority, and to remove the ability to retrigger.

The FAA was fully cognizant of the hazards presented by an erroneous AoA vane. The AD was their response, believed to be sufficient and adequate. This AD was carried forth globally. The AD addressed the biggest concern raised by the pilot community which was to bring awareness of the function. The whole point of an AD is to ensure that no future accident would result from these circumstances.

Usually, an accident is the result of multiple flaws in simultaneous occurrence.
And some flaws aren't even flaws, except in the presence of other flaws arising.
(Sounds strange, eh.) Combine the probabilities of each occurring, & this is why
airliner accidents have a low probability per mile flown.

I think most people are cognizant of this. There's the saying "it never rains, but it pours" among others...

I heard a talking head on NPR the other day...an expert in something...say
that designers should make everything fundamentally safe, & not rely upon '
statistics. Ugh....He prolly doesn't even know that even the strongest
aircraft aluminum alloys have only a statistically limited fatigue life.
Alas, the public will get an impression that designers are recklessly
depending upon chance. They remind me of creationists' criticism
of evolution...."It's just random chance!".

I'm the exact opposite of the talking head you mention. Everything is statistics and probabilities. There's no way to avoid them. You can just do the best to prepare for possibilities and eventualities like fatigue.

 

[Revoltingest]

News generally don't go much into things like sensors. I'd never consider "newsmedia" primary sources. Most journalists don't have any technical training.

This guy(avionics engineer and a former Boeing avionics supervisor), goes more into it:

Satcom Guru: Comparing Ethiopian ET302 to Lion Air JT043, JT610




I think most people are cognizant of this. There's the saying "it never rains, but it pours" among others...


I'm the exact opposite of the talking head you mention. Everything is statistics and probabilities. There's no way to avoid them. You can just do the best to prepare for possibilities and eventualities like fatigue.

The AD providing different procedures in the case of AOA disagreement
will be effective only if all affected pilots are thoroughly retrained for it.
This is necessary for proper reactions in times of urgency & stress.
The airline then has responsibility to ensure this.

 

[Jumi]

Yes, the interim AD for MAX planes' design flaw of relying on only one sensor seemed sufficient for FAA.

The results from the black boxes in Ethiopia will show if this is the same flaw still. There seemed to be some difference as analyzed by satcom guru.

 

[Bob the Unbeliever]

Was going to comment, but Revolt & Jumi covered what I was going to say.

I will say this, though: modern airplanes that carry more than about 10 people, are far past purely mechanical controls.

For starters, with supercritical wing design (which increases efficiencies multiple times over, needed for reasonable fuel range) a mere human cannot possibly hope to keep the plane's flight stable. Humans simply cannot process information fast enough, nor make the constant micro-corrections often enough to get the job done. Robots are an absolute must.

So even if some idiot built a modern large jet, with pure mechanical controls? There would need to be some system that assisted these, and took care of the micro-corrections, just to stay in the air and not go out of control and crash. Speed is also a factor.

Note that the only airplanes that still retain these older systems? Are on the small side, and slow, too. Typically prop jobs as well. Or are pure stunt/sport type airplanes, and thus flown by literal experts. But-- take note of the crash-per-hours-flown rate of these exotic craft. Not dissimilar to that of race cars versus ordinary passenger cars, come to think of it...

But hey! Let's return to the Good Old Days, of Lighter Than Air Flight, where Men were Men, and Hot Air Balloons were Cutting Edge, and God Only Knows Where You End Up while Flying! Each Flight Is A Literal Adventure! One you may not even survive!

... right?

 

[Revoltingest]

Was going to comment, but Revolt & Jumi covered what I was going to say.

I will say this, though: modern airplanes that carry more than about 10 people, are far past purely mechanical controls.

For starters, with supercritical wing design (which increases efficiencies multiple times over, needed for reasonable fuel range) a mere human cannot possibly hope to keep the plane's flight stable. Humans simply cannot process information fast enough, nor make the constant micro-corrections often enough to get the job done. Robots are an absolute must.

So even if some idiot built a modern large jet, with pure mechanical controls? There would need to be some system that assisted these, and took care of the micro-corrections, just to stay in the air and not go out of control and crash. Speed is also a factor.

Note that the only airplanes that still retain these older systems? Are on the small side, and slow, too. Typically prop jobs as well. Or are pure stunt/sport type airplanes, and thus flown by literal experts. But-- take note of the crash-per-hours-flown rate of these exotic craft. Not dissimilar to that of race cars versus ordinary passenger cars, come to think of it...

But hey! Let's return to the Good Old Days, of Lighter Than Air Flight, where Men were Men, and Hot Air Balloons were Cutting Edge, and God Only Knows Where You End Up while Flying! Each Flight Is A Literal Adventure! One you may not even survive!

... right?

For everyone's enjoyment....
Supercritical airfoil - Wikipedia

 

[Jumi]

Two interesting articles...

How a 50-year-old design came back to haunt Boeing with its troubled 737 Max jet

The 737Max and Why Software Engineers Might Want to Pay Attention

 

[Jumi]

A couple of quotes from the latter article, of the approaches I found problematic myself.

MCAS; this is an automation system intended to keep the plane from stalling. One of the things it measures is the angle-of-attack: if the plane exceeds a particular flight angle, it will fall out of the sky. The MCAS, in the default configuration, uses a single sensor as input to measure this critical metric.

And

...they added additional monitoring and control systems. One of these systems was added “for safety,” but it was also capable of overriding operator input during critical operational phases, where activity is high tempo and high consequence.

This new safety automation is capable of overriding operator input in silence and in ways that were poorly documented by designers, unclear to operators, and promised by developers that nobody had to get new training on — a selling point — and this safety automation proved to cause the system to become critically unrecoverable in, at least, one case.

And a quote from the former:
Skow criticized Boeing’s MCAS system, saying it acted only on the basis of angle of attack. The Lion Air jet was traveling so fast that when MCAS ordered the stabilizer to pitch the nose down it was a violent reaction. The software should have factored in air speed, he said, which would have better calibrated the pilots’ reaction.

 

[Revoltingest]

Two interesting articles...

How a 50-year-old design came back to haunt Boeing with its troubled 737 Max jet

The 737Max and Why Software Engineers Might Want to Pay Attention

I've reached my monthly limit of free articles on the first link.

Regarding the 2nd one....

is system detects a dangerous flight condition, it trims the aircraft, attempting to prevent a stall by pushing the nose down. Trim is not a fancy, new fangled technology: the Cessnas I fly have trim wheels, and autopilots manipulate trim to fly aircraft in an automated way. What is different here is: the MCAS commands the trim in this condition without notifying the pilots AND to override the input, the pilots must deactivate the system via a switch on a console, NOT by retrimming the aircraft via the yoke, which is a more common way to manage the airplane’s trim.

The cockpit has a continual problem of balancing information overload with providing
needed info. (Fighter planes are the worst in terms of info volume.) To not notify the
pilot of something, can be a very good thing. Listening to CVRs (cockpit voice recorders),
one finds that pilots in stressful situations sometimes won't pay attention to even audible
warnings of imminent danger. Too many bells, buzzers, lights & commands....& pilots
will be overwhelmed.
Systems are designed for optimum balance, but what will work well for some pilots
in some situations, will utterly fail in others. Thus, this balance must be continually
re-evaluated in light of experience. Automation will make things better, but it will
be a long & bumpy road to improve safety.

The point of the above is that we'll be presented with criticisms of designers & pilots,
& it will seem obvious to us that simple mistakes were made. But don't be tempted
to see it thus. Things are too often more complex than they appear.

Disclaimer:
I don't claim to understand these systems.
I'm only aware that there is a vast process going on which I know little of.

 

[Bob the Unbeliever]

For everyone's enjoyment....
Supercritical airfoil - Wikipedia

*whew* that's a relief! My memory did not betray me...

Thank you for the link-- I wrote my other piece from memory, and I had *thought* that supercritical wings were more efficient, and your Wiki agreed with my memory.

From your link: " ... it [supercritical design] has since been mainly applied to increase the fuel efficiency of many high-subsonic aircraft."

 

[Bob the Unbeliever]

Two interesting articles...

How a 50-year-old design came back to haunt Boeing with its troubled 737 Max jet

The 737Max and Why Software Engineers Might Want to Pay Attention

Wow. From the first link, I had no idea the darn plane is nearly as old as I am!

It's truly a legacy of a time long past-- walking **outside** (in the possible rain) to get onto an air plane, by going up a set of steps (that were often bolted to a vehicle)...

I remember when airline engines were loud, and not very fuel efficient. But back in those days? Anyone could just go out to the airport, and walk up into the terminal, and go wandering around, looking out of any of the windows as you pleased. And I often did, back in the day... I've always found air planes fascinating.

These days, if I tried that? Well... I'd be sitting for a *long* time in a grey metal room, wondering "what just happened?" ...

But Orwell would understand why, I think...

 

[Bob the Unbeliever]

I've reached my monthly limit of free articles on the first link.

Far be it from me, to advocate bypassing an obvious need to pry money from intereseted parties?

But I use something called "scriptsafe" and did not get the nag pop-up at all, and therefore, had no problems reading the article.

Very interesting history, the venerable 737. Decades past it's "best if used by" date.

Reminds me of the venerable B-52, only not in a good way.

A key point that seemed to shout out at me, worth noting. It's easier to FAA certify an old airplane, that has upgrades, than it is a new one from scratch. That seems exactly backwards, to me...

.

Over the years, the FAA has implemented new and tougher design requirements, but a derivative gets many of the designs grandfathered in, Moss said.

“It is cheaper and easier to do a derivative than a new aircraft,” said Robert Ditchey, an engineer, aviation safety consultant and founder of America West Airlines, which purchased some of the early 737 models. “It is easier to certificate it.”

.

 

[Revoltingest]

Very interesting history, the venerable 737. Decades past it's "best if used by" date.

Reminds me of the venerable B-52, only not in a good way.

Big difference....
B52 production ended in 1961. Any still in service would have an airframe over half
a century old. Modifications can extend life, but aluminum has a limited fatigue life.
All those flying hours, takeoffs & landings....they take their toll. They can reinforce
areas of high stress, but they cannot go on forever because....you know....aluminum.

The 737 Max series is based on an older airframe design, but is all modern production.
With the modifications, it's competitive with Airbus 320 series planes.

A key point that seemed to shout out at me, worth noting. It's easier to FAA certify an old airplane, that has upgrades, than it is a new one from scratch. That seems exactly backwards, to me...
.

Completely new designs have more aspects to test.
Reworking an old one is more of a known quantity,
so easier certification makes sense, but I'm not
familiar with it.

 

[Jumi]

The cockpit has a continual problem of balancing information overload with providing
needed info. (Fighter planes are the worst in terms of info volume.) To not notify the
pilot of something, can be a very good thing. Listening to CVRs (cockpit voice recorders),
one finds that pilots in stressful situations sometimes won't pay attention to even audible
warnings of imminent danger. Too many bells, buzzers, lights & commands....& pilots
will be overwhelmed.

Agreed. In the article you couldn't access the cockpit display design was shortly criticized, but because the expert criticizing it is part of a company prioritizing it "better" I didn't take him 100% when he said his company's solution could have prevented it.

Systems are designed for optimum balance, but what will work well for some pilots
in some situations, will utterly fail in others. Thus, this balance must be continually
re-evaluated in light of experience. Automation will make things better, but it will
be a long & bumpy road to improve safety.

Let's also not forget that if design is bad it can be made in a way that no matter what you do, you will fail. Of course designers learn, but a new generation of designers sometimes have to learn through making disasters. Safety standards are improved all the time, but even there's a catch. Not everyone understands them fully who is expected to know them or understands all the relevant ones.

The point of the above is that we'll be presented with criticisms of designers & pilots,
& it will seem obvious to us that simple mistakes were made. But don't be tempted
to see it thus. Things are too often more complex than they appear.

Many have pushed aside any design issues and criticized the pilots that they should have had more training or shifted the blame to the airliner if you look at this thread. You can see a new system sold as no new training required and then when it causes problems blame the operator's training for not "getting" the new one as intended.

It remains hard to think of reliance on one sensor without checks for airspeed and other factors as not a problem.

 

[Jumi]

Wow. From the first link, I had no idea the darn plane is nearly as old as I am!

It's truly a legacy of a time long past-- walking **outside** (in the possible rain) to get onto an air plane, by going up a set of steps (that were often bolted to a vehicle)...

Yeah the 737 is old, but it still works. The revisions sometimes keep problematic parts of the design. The 737 MAX inherited much of the design of the older 737s because it was easier to grandfather designs than get new ones approved according to the article. Designing one from scratch... it's quite expensive.

I remember when airline engines were loud, and not very fuel efficient. But back in those days? Anyone could just go out to the airport, and walk up into the terminal, and go wandering around, looking out of any of the windows as you pleased. And I often did, back in the day... I've always found air planes fascinating.

These days, if I tried that? Well... I'd be sitting for a *long* time in a grey metal room, wondering "what just happened?" ...

But Orwell would understand why, I think...

There's good and bad about that. Poking around was easier in the past, but yeah airport security is no joke these days at least not on any airport I've been to. There's positives to it too, hijackers and such will need more creativity to achieve something.

I think it's part of why 2017 was the safest year with no passenger airplanes crashing. I think 2018 and 2019 weren't expected to be so much worse, mostly due to the Lion air and Ethiopian air crashes.

 

[Revoltingest]

Agreed. In the article you couldn't access the cockpit display design was shortly criticized, but because the expert criticizing it is part of a company prioritizing it "better" I didn't take him 100% when he said his company's solution could have prevented it.


Let's also not forget that if design is bad it can be made in a way that no matter what you do, you will fail. Of course designers learn, but a new generation of designers sometimes have to learn through making disasters. Safety standards are improved all the time, but even there's a catch. Not everyone understands them fully who is expected to know them or understands all the relevant ones.


Many have pushed aside any design issues and criticized the pilots that they should have had more training or shifted the blame to the airliner if you look at this thread. You can see a new system sold as no new training required and then when it causes problems blame the operator's training for not "getting" the new one as intended.

It remains hard to think of reliance on one sensor without checks for airspeed and other factors as not a problem.

Aye, they'll all learn from this one too.
We'll all become safer than we'd otherwise be.

 

[esmith]

It is said that the Navy's NATOPS manual is written in blood.
I would have to say any safety instructions whether military or civilian is written in blood.

 

[Jumi]

It is said that the Navy's NATOPS manual is written in blood.
I would have to say any safety instructions whether military or civilian is written in blood.

Lots of times sacrifices have been made, but often still no understanding arises. Sometimes common sense could avoid Titanic-like accidents. Sometimes politics is used to stop advancements from being made. I've seen some blame for US government shutdown for delaying the software update.

Often the security improves when scientific theories mature and leads to improvements in standards. Of course simulation and experimentation feed into that. There are many types of situations we never have to go through thanks to them. If it's true that they were overconfident in their sensors being reliable enough to never fail, then they made a mistake. One that's already been payed with blood way before these planes...

 

[Bob the Unbeliever]

Big difference....
B52 production ended in 1961. Any still in service would have an airframe over half
a century old. Modifications can extend life, but aluminum has a limited fatigue life.
All those flying hours, takeoffs & landings....they take their toll. They can reinforce
areas of high stress, but they cannot go on forever because....you know....aluminum..

Oh, I agree, re: B52-- but lest you forget, the wings are all new on those still in the air. The Air Force replaced those with new, stronger alloys quite a while back. This let them upgrade the fuel tanks, engine mounts, etc, etc. I also seem to remember a re-skinning project as well. These are basically all new airplanes. But not. You know-- like my great-great-grandfather's axe. Grandad had to replace the handle. My dad had to put on a new head, as the old one had worn down to a nub. But it's me Great-great-grandfather's Axe all right!

The 737 Max series is based on an older airframe design, but is all modern production.
With the modifications, it's competitive with Airbus 320 series planes.

Completely new designs have more aspects to test.
Reworking an old one is more of a known quantity,
so easier certification makes sense, but I'm not
familiar with it.

So, basically, it's an All New Airplane, and ought to be treated as such-- but a cheap loophole lets them use untested cray-cray stuff, which results in ... fatal crashes?

That was my lament. If an old design is updated? There should never be permitted a "grand father" clause here. Treat it as new.

It's not as if you can coast into the breakdown lane, and calmly step out onto the road, is it?

 

[Revoltingest]

Oh, I agree, re: B52-- but lest you forget, the wings are all new on those still in the air. The Air Force replaced those with new, stronger alloys quite a while back. This let them upgrade the fuel tanks, engine mounts, etc, etc. I also seem to remember a re-skinning project as well. These are basically all new airplanes. But not. You know-- like my great-great-grandfather's axe. Grandad had to replace the handle. My dad had to put on a new head, as the old one had worn down to a nub. But it's me Great-great-grandfather's Axe all right!


So, basically, it's an All New Airplane, and ought to be treated as such-- but a cheap loophole lets them use untested cray-cray stuff, which results in ... fatal crashes?

That was my lament. If an old design is updated? There should never be permitted a "grand father" clause here. Treat it as new.

It's not as if you can coast into the breakdown lane, and calmly step out onto the road, is it?

I disagree, & will leave certification debates to manufacturers & the FAA.