Ethiopian Airlines Crash; Another 737 Max

ok...say it....racist comments. :lol:
I actually typed my response without seeing your reply (you submitted it while I was typing) but it is a somewhat clear example of what I’m talking about...

Also don’t you think you should consider the source of the info? Boeing would love to offset any responsibility it can. It (and to some extent it’s employees) of course is going to look for explanations that blames a 3rd party.
 
Then is was a solution to a problem that doesn't exist.
What? Just because a problem hasn’t occurred and isn’t likely to occur doesn’t mean you don’t provide a solution. Furthermore it did provide a “solution” to the problem of maintaining the 737 type commonality. Whether or not that is a desirable thing to solve and whether or not MCAS is the best solution is another discussion.

Also what was mryan’s source on that statement? Do we even know it is true? Furthermore, is he referring to normal MCAS activations where it activates as designed or just erroneous ones due to bad input information?
 
Last edited:
No MCAS activations at all or just none that required emergency actions?
I doubt I can find it, but in one of the articles I read, it was either one of the union guys or an airline executive who said they had studied it and found exactly zero MCAS activations in US fleets. Which makes sense - why should you be slow with no flaps and a high angle of attack in a 737? I've thought from the beginning that in the end there will be an environmental aspect to both of these accidents. Extremely high DA in both places, plus throw in dust and sand in Addis Ababa.
 
Last edited:
I cannot find this info again but I understand the US airlines had purchased the option of using both AOA indicators and having that as a safety feature (where they both must agree before activating the MCAS system) while the other two airlines in questions only used the single AOA, thereby creating a single-source activation for the MCAS. This could be a big factor in why no US airlines have had an activation.
 
Take a look at the Jepp SID for HAAB. Not a great place to have an uncommanded nose down trim problem right after takeoff.
 
Last edited:
I've thought from the beginning that in the end there will be an environmental aspect to both of these accidents. Extremely high DA in both places...
That's a new insight, and plausible.
 
I cannot find this info again but I understand the US airlines had purchased the option of using both AOA indicators and having that as a safety feature (where they both must agree before activating the MCAS system) while the other two airlines in questions only used the single AOA, thereby creating a single-source activation for the MCAS. This could be a big factor in why no US airlines have had an activation.

No, there was an option for an AoA disagree warning but not for MCAS to use more than one AoA sensor. That is coming in a future software update.
 
No, there was an option for an AoA disagree warning but not for MCAS to use more than one AoA sensor. That is coming in a future software update.
And even with that light, if pilots aren’t even informed that MCAS exists and how it works and it’s limited fault tolerance, they would not be able to deduce the possible consequences of an AOA disagree.
 
Last edited:
there were zero incidents in the US. What's up with that?

The birds in the US know better than to hit our AoA vanes. :rolleyes:

There have also been zero incidents in China, Canada, Turkey, Germany, Ireland, India, Japan, the Czech Republic, Singapore, Kazakhstan, Morocco, Qatar, Norway, Mauritania, Iceland, Brazil, Dubai, Fiji, Poland, Panama, South Africa, Mexico, and Argentina... All countries that also have airlines that have taken deliveries of the 737 Max.

But we're Americans, our pilots are better than all those. USA! USA! :rolleyes:
 
Complacency.

It happened to NASA, and it may have happened to Boeing. It can be insidious. Look at the safety records of both - excellent, until they weren't.

In NASA's case: Challenger was launched in conditions the SRBs hadn't been tested. 30F at launch, the SRB assembly was qualified down to 40F. O-rings were too stiff from the cold to maintain a seal when the SRB sections changed shape during launch. Until then? Great safety record. Why should this flight be any different, all the others worked just fine?

Boeing: The 737 may have the best safety record of any airliner ever made. Sure, it's had a lot of variations over the years, but they just work. Designers (engineers, managers, and certainly accountants) make decisions on what to add, when, why, what it will cost, and what it will gain in sales, safety, and profitability. MCAS was a good idea to solve a problem - nudge down the nose to counteract upward pitch forces when inside certain parameters. Why did that system only used a single AOA? AOA never fails, it's rock solid...until it isn't. MCAS was probably meant to be so "behind the scenes" that the crew wouldn't even notice it activated unless they saw the trim wheels turn occasionally and that might be why it wasn't publicized. I *think* automatic trim movements are normal (someone with experience can correct me), so the blip of the wheels once in a while might not even raise an eyebrow. That's OK, until the crew doesn't realize that what's normal has become abnormal and then Boeing expected the crew to jump into the runaway trim procedure to stop it. That's where Boeing might have missed. We expect flight crews to be expert at what they are doing, and the systems they are flying, and rightly so. But not all experts are as good as the others. I think Boeing forgot that pilots are human - AOA never fails; then it fails at a low altitude, the crew gets overwhelmed or task saturated, human factors and physics begin to clash, and physics wins. Lion Air was an interesting example - the captain corrected almost 2 dozen times and it didn't seem to spark anything in him to turn off the stab trim. Why not? We can look back and say training, experience, or whatever, but it may just be human factors (possibly a cultural difference) in how problems are diagnosed and handled. This is the kind of thing that needs to be put into design decisions, too. We've automated some things so much that if one thing fails, the cascading problems can put the whole system into a bad place. MCAS gets bad input? Turn it off, but turning it off means you lose the manual electric trim and now you may be in a situation where the trim wheels themselves can't be turned.

I'm not clearing the pilots and I'm not putting all the blame on Boeing. But we do need to design systems to be operated by the lowest common denominator. If that denominator gets lower (inexperienced pilots at some airlines, poor training, poor mx), there's one problem. But if the system requires a higher denominator (additional training, depending on certain crew reactions within a tight time limit), that's a problem too. I don't believe in coincidence, so I'm really curious about the whole story on these two accidents.

Sometimes it takes an accident to remind you just how close you are to the edge of the envelope. Trying to maintain a safety margin when you've never had an accident can lead to complacency. I expect Boeing will make a couple of fixes to the system, test them under more extreme conditions, and MAX will be another in a long line of 737s with an excellent safety record.
 
I'm not clearing the pilots and I'm not putting all the blame on Boeing. But we do need to design systems to be operated by the lowest common denominator.

Designing aircraft around the lowest common denominator is not generally how it's done. A certain level of skill is usually expected when flying an airliner or even an Ercoupe. Gotta give the designer a little leeway.
 
Designing aircraft around the lowest common denominator is not generally how it's done. A certain level of skill is usually expected when flying an airliner or even an Ercoupe. Gotta give the designer a little leeway.
My intent was to say we expect an ATP to be able to meet minimum qualifications and we need to design to that minimum. What's the old saying? What do you call the guy who graduates last in his medical school? "Doctor". Similar with an ATP. What happens if that minimum starts to drop? Not that the PTS, or whatever, requires lower standards, but the candidates are edging closer to the bottom edge.
 
word around the water cooler and my friends who worked the field and Boeing say.....poor ground handling causes those AoA vanes to fail. Could be why no US carriers have reported this failure mode.

But...even if true...this ignores the issue of a critical safety system being totally unaware of crosscheck redundancy when one of two available sensors fails. Unless, apparently, you pay for the "extra safety" package. It's easy to blame the operators, but so why put them in this position in the first place when it is unnecessary and readily avoidable at little cost? Boeing is about to find out that safety enhancement is a lot cheaper in the long run than wrongful death lawsuits.
 
The birds in the US know better than to hit our AoA vanes. :rolleyes:

There have also been zero incidents in China, Canada, Turkey, Germany, Ireland, India, Japan, the Czech Republic, Singapore, Kazakhstan, Morocco, Qatar, Norway, Mauritania, Iceland, Brazil, Dubai, Fiji, Poland, Panama, South Africa, Mexico, and Argentina... All countries that also have airlines that have taken deliveries of the 737 Max.

But we're Americans, our pilots are better than all those. USA! USA! :rolleyes:
I'm not talking zero "incidents", I'm talking zero activations of MCAS in US fleets.
 
But...even if true...this ignores the issue of a critical safety system being totally unaware of crosscheck redundancy when one of two available sensors fails. Unless, apparently, you pay for the "extra safety" package. It's easy to blame the operators, but so why put them in this position in the first place when it is unnecessary and readily avoidable at little cost? Boeing is about to find out that safety enhancement is a lot cheaper in the long run than wrongful death lawsuits.
The extra cost option was for an AOA disagree light as has been stated many times in this thread. AOA cross checking and redundancy for MCAS purposes was not available, although is slated to be implemented in a future Boeing software update.
 
Designing aircraft around the lowest common denominator is not generally how it's done. A certain level of skill is usually expected when flying an airliner or even an Ercoupe. Gotta give the designer a little leeway.

So are you saying a critical safety system that doesn't use all readily available information to prevent inappropriate and potentially dangerous deployment acceptable design leeway or responsible safety engineering? I would not.
 
I cannot find this info again but I understand the US airlines had purchased the option of using both AOA indicators and having that as a safety feature (where they both must agree before activating the MCAS system) while the other two airlines in questions only used the single AOA, thereby creating a single-source activation for the MCAS. This could be a big factor in why no US airlines have had an activation.

Where do you get this stuff? Ya know, exactly where are you pulling it out of?
 
The extra cost option was for an AOA disagree light as has been stated many times in this thread. AOA cross checking and redundancy for MCAS purposes was not available, although is slated to be implemented in a future Boeing software update.

Even the AOA disagree light seems like a bandaid. Should MCAS even be allowed to initiate maneuvers if it can't verify accurate input data? Seems like from a safety engineering standpoint this is a critical point of consideration.
 
In the end I believe Boeing will get this right, and the 737 Max may well become one of the most popular commercial aircraft ever flown. It is just unfortunate it will come at the cost of a little too much safety design complacency and after the fact heel-dragging that led to the loss of several hundred lives in the process. Hopefully lessons will be learned.
 
In general, I think we'll see more accidents. Not in the US, but globally. More and more startup airlines are popping up around the world, and more and more foreign airlines are expanding. That's going to deplete the availability of experienced pilots in those countries and lead to more of what we saw in Ethiopia, 300 hr FOs, essentially a single pilot operation. Can you get away with that? Sure, but eventually it will bite you. As the population on the left side the curve increases, I do think we'll see more accidents. And how will public perception towards the mfg, like Boeing, swing? "Their airplanes are so complex they aren't safe!", might be a typical reaction. Followed by, "Then why don't they crash in the US?" I've seen those arguments in this set of accidents already.
 
That makes sense theoretically, but it’s described as “up to 9.3 seconds”, not every activation will be 9.3 seconds.
In the case of an unschedule MCAS activation due to faulty AoA data then each activation will run for the full 9.3 seconds unless interrupted by either the use of primary electric trim or the stab trim switches being selected to cutoff.

I’m having a hard time imagining that big ol’ trim wheel spinning away during a critical phase of flight and knowing when to trust it and when not to.
It's not the trim wheel that you're trusting, or not. It is the repeated need to retrim the airplane nose-up. Every time you re-trim the airplane it is going out of trim again five to ten seconds later. After a few cycles, you should know that you have a trim runaway. The problem, on the accident flights, was that the crews (F/O on Lion Air, Capt on Ethiopian) were not retrimming each time so they didn't detect the repeating pattern and arrest the trim problem before it had run the trim to nearly full nose-down.

It sounds like using manual trim is impossible at certain speeds.
Only if the stab trim is already at the extreme end of its range and, even then, not impossible; difficult.

The trim wheels have fold-out handles that are positioned 90deg (of rotation) apart. That ensures that one pilot, or the other, always has good leverage to turn the wheel when they are working together. That is only required when the airplane is significantly out of trim. When it approaches full scale then you also have to unload the stabilizer while trimming. This is the same trim system with manual backup that has flown for decades in the B707, KC-135, B727, and all B737s.

That’s not the end world, but it sucks, so you’ve got to deal with it. So it’s flaps up and hit the AP, which seems to be drunk, then kicks itself off. MCAS has no idea what’s going on, but it kicks in and you get half your nose up trim taken away.
Why would you retract the flaps when you have a stick shaker (stall warning) going off?

Why would you engage the autopilot when you have having unreliable airspeed and altitude indications? The reason the A/P "seemed to be drunk" and quickly disengaged was because it was trying to follow the unreliable airspeed and altitude data.

Why not fly the airplane yourself so you will be aware of its condition and have a chance of noticing the repeated need to trim nose-up? De-automate. Fly the airplane first.

I don't believe the procedure says to wait to execute the runaway trim procedure until the aircraft is in relatively neutral trim.
It does not because is presumes that the pilots have been flying the airplane instead of allowing a malfunctioning system to run the trim to the nearly full nose-down position.

I don't believe the other incidents commenced seconds after the airplane rotated off the runway
They did. Stick-shaker at liftoff, unschedule MCAS activation at flap retraction. All three flights.
 
In the case of an unschedule MCAS activation due to faulty AoA data then each activation will run for the full 9.3 seconds unless interrupted by either the use of primary electric trim or the stab trim switches being selected to cutoff.


It's not the trim wheel that you're trusting, or not. It is the repeated need to retrim the airplane nose-up. Every time you re-trim the airplane it is going out of trim again five to ten seconds later. After a few cycles, you should know that you have a trim runaway. The problem, on the accident flights, was that the crews (F/O on Lion Air, Capt on Ethiopian) were not retrimming each time so they didn't detect the repeating pattern and arrest the trim problem before it had run the trim to nearly full nose-down.


Only if the stab trim is already at the extreme end of its range and, even then, not impossible; difficult.

The trim wheels have fold-out handles that are positioned 90deg (of rotation) apart. That ensures that one pilot, or the other, always has good leverage to turn the wheel when they are working together. That is only required when the airplane is significantly out of trim. When it approaches full scale then you also have to unload the stabilizer while trimming. This is the same trim system with manual backup that has flown for decades in the B707, KC-135, B727, and all B737s.


Why would you retract the flaps when you have a stick shaker (stall warning) going off?

Why would you engage the autopilot when you have having unreliable airspeed and altitude indications? The reason the A/P "seemed to be drunk" and quickly disengaged was because it was trying to follow the unreliable airspeed and altitude data.

Why not fly the airplane yourself so you will be aware of its condition and have a chance of noticing the repeated need to trim nose-up? De-automate. Fly the airplane first.


It does not because is presumes that the pilots have been flying the airplane instead of allowing a malfunctioning system to run the trim to the nearly full nose-down position.


They did. Stick-shaker at liftoff, unschedule MCAS activation at flap retraction. All three flights.
So based on all that, it sounds like the first crew error was retracting the flaps with the stick shaker going off. Once that error was made, the situation could still have been salvaged by making sure that the airplanes were properly re-trimmed each time the primary electric trim was used, and once it was realized that there was a trim runaway, making sure that the airplanes were at least approximately trimmed when the stab trim switches were selected to cutoff. Do I have that right?
 
So based on all that, it sounds like the first crew error was retracting the flaps with the stick shaker going off. Once that error was made, the situation could still have been salvaged by making sure that the airplanes were properly re-trimmed each time the primary electric trim was used, and once it was realized that there was a trim runaway, making sure that the airplanes were at least approximately trimmed when the stab trim switches were selected to cutoff. Do I have that right?
I wouldn't say that retracting the flaps was an error, especially not for the first crew. It's just something that I wouldn't do. Just seems a little strange that they would retract them with a stall warning going off.

It's clear that Boeing was wrong to not include a comparison check between the two AoA readings but what I'm trying to say is that it doesn't seem like they were flying the airplane. If they were, they would have been keeping it close to in-trim and they would have recognized the stabilizer runaway sooner. Also, why did the power stay at 94% for the entire flight (Ethiopian) while their airspeed was reaching excessive speeds affecting their ability to trim manually? There should always be one pilot who's primary job is flying the airplane. See EAL401 as an example of what happens when there isn't.
 
How about we stop striving to lay blame and just let them figure out what happened? Boeing didn't build a system knowing that yea, it'll work but if the AOA goes tango uniform they're gonna crash. The crews apparently did a pretty decent job in a major WTF? moment to try save it. There is a news story right now about ten babies who have died in a Fisher-Price Rock n' Play sleeper. The world's not perfect, yet...
 
Fly the airplane first.
And as being discussed elsewhere, part of the "flying the aircraft" would have been to reduce power. It is reported that the engine power settings where kept at takeoff levels through the entire sequence. Not only did this add air flow pressure to the out of trim condition but brought them to the maximum maneuvering speed rather quickly, instead of giving them option of proceeding at the minimum air speed to work the nose down problem.
 
How about we stop striving to lay blame and just let them figure out what happened? Boeing didn't build a system knowing that yea, it'll work but if the AOA goes tango uniform they're gonna crash. The crews apparently did a pretty decent job in a major WTF? moment to try save it. There is a news story right now about ten babies who have died in a Fisher-Price Rock n' Play sleeper. The world's not perfect, yet...
The purpose of trying to figure out whether doing something differently could have saved the flight is to learn, not to lay blame.
 
And as being discussed elsewhere, part of the "flying the aircraft" would have been to reduce power. It is reported that the engine power settings where kept at takeoff levels through the entire sequence. Not only did this add air flow pressure to the out of trim condition but brought them to the maximum maneuvering speed rather quickly, instead of giving them option of proceeding at the minimum air speed to work the nose down problem.
Engine thrust is below the cg, so I'd suspect more thrust may have made things better.
 
The purpose of trying to figure out whether doing something differently could have saved the flight is to learn, not to lay blame.
Most of this thread has been a pretty good analysis. We've all seen "Well, if they would have only...", a lot. But they didn't, and the "why" is where the learning comes in. "They" could be the crew's decisions and the Boeing design decisions.
 
The purpose of trying to figure out whether doing something differently could have saved the flight is to learn, not to lay blame.

The blame comes out of the investigation whether we like it or not. We’re not so PC that every captain gets a trophy whether it’s his fault or not. We’re talking skilled captains here, sometimes with 20,000 hours plus.
 
Engine thrust is below the cg, so I'd suspect more thrust may have made things better.
I don't think so. The pitch-up from full thrust isn't all that much. They were flying roughly 150kts faster than they needed to, though, which adds a lot of authority to the mis-trimmed stabilizer and increases the air loads on it making manual trimming much more difficult. I think the excessive airspeed was a lot bigger problem than the nose-up moment from thrust was a help.
 
And as being discussed elsewhere, part of the "flying the aircraft" would have been to reduce power. It is reported that the engine power settings where kept at takeoff levels through the entire sequence. Not only did this add air flow pressure to the out of trim condition but brought them to the maximum maneuvering speed rather quickly, instead of giving them option of proceeding at the minimum air speed to work the nose down problem.

HABB is in a bowl, if you can call it that, at 7k msl. There is high terrain all around when you look at a topographical map, though less so in the direction of depsrture. . I'm not sure I can blame these guys for not reducing power heading toward higher terrain at high DA, clawing for altitude, terrain warning going "Don't Sink, Don't Sink", especially if they weren't sure of their own airspeed, altitude and other flight data due to the bad sensor ( it's been reported they asked the tower for a speed check) and not acutely aware of the effect the speed would have on the use of manual trim. I haven't compared their speed to the timeline of events, but I believe the extreme speeds were weIl beyond the point of trim cut-out when manual trim would have been required. The first waypoint on the SID is 12 or 14k. Flighttracker shows they never got above 9.
 
Last edited:
these guys for not reducing power heading toward higher terrain at high DA
Perhaps. But as stated elsewhere by flying 737 pilots the excessive engine power on a severly out of trim 737 will get you down quicker than up over an obstacle. Their examples point to if the 302 crew had throttled back to 80% or 75% with flaps 1, they could have retrimmed and dealt with the miscompare air data. Fly the aircraft is was their main point. Same as the Lion pilot did before handing off controll to the SIC. They also state that selecting engine power at 75% and flaps 1 would have removed the mcas and allowed the option of manual trim that wasn't readily possible at 94% engine power.
 
  • They also state that selecting engine power at 75% and flaps 1 would have removed the mcas and allowed the option of manual trim that wasn't readily possible at 94% engine power.

    I'd have to check but I don't think the Lion Air crew was headed toward higher terrain.

  • And how we're they to know that? 5 seconds after flap retract the a/p cut off due to the erroneous air data. Ten seconds after that was the first MCAS attack. With all that going on simultaneously, and the stick shaker, and the terrain proximity alert, they probably didn't even notice they had a trim problem as yet. They trimmed up using the electric trim, and then came the second attack, and then they started to realize they had a problem, but by then it was too late.
  • How fast and how out of trim does the manual say you can be before the manual trim won't work?
 
Last edited:
The blame comes out of the investigation whether we like it or not. We’re not so PC that every captain gets a trophy whether it’s his fault or not. We’re talking skilled captains here, sometimes with 20,000 hours plus.

I was speaking about my own motivations. I'm interested in what the pilots could have done to save their behinds, because in spite of the fact that I will never fly an airliner, there is at least some possibility that it could help me someday if I get in a situation where I'm getting a confusing array of fault indications going on at the same time. For me, an important step in understanding these accidents is finding out what actions would have been most effective at dealing with the situation. Another step is to understand what factors worked against the accident crews discovering what needed to be done soon enough to avoid the accident. For starters, with a stick shaker activated immediately after takeoff, imagining myself in that situation I find myself wondering whether I would realize quickly enough that the plane was not in fact stalled, and that the nose needed to be trimmed upwards.

I really can't see a point in worrying about whether my musings could cause some third party to blame the pilots. For one thing, Pilots of America is not an official investigatory body, so I doubt that there is much chance of my kibitzing having much real-world effect.
 
So we have, going back to the 60s:
  1. a decision by Boeing engineers to design an airplane low to the ground
  2. an aircraft design that requires a so-called "roller-coaster" technique to relieve enough aerodynamic force on the stabilizer to use the manual trim wheels
  3. a decision to remove the "roller-coaster" technique from flight manuals starting with the 737-300 despite it still being relevant
  4. airlines that want a more fuel-efficient successor to the 737 ASAP rather than a clean-sheet design, thus the invention of the "Max" series
  5. airlines that don't want the new "Max" to have a different type rating
  6. more efficient engines which won't fit under the wing (see point #1) so they are mounted farther forwoard
  7. differing pitch characteristics due to point #6 that, while seemingly not dangerous, do not conform to a decades-old, arguably obsolete certification standard
  8. invention of MCAS to correct point #7
  9. decision by Boeing engineers to feed MCAS off of only one AoA sensor
  10. decision by Boeing engineers to make an AoA disagree warning light optional
  11. decision by Ethiopian Airlines and others to not purchase the AoA disagree warning light
  12. a failure of an AoA sensor, that just so happens to be the one that feeds the MCAS system
  13. pilots not keeping the airplane in-trim with the manual electric trim following an unwarranted MCAS activation due to point #12
  14. pilots not following the MCAS stab-trim-cutout procedure quickly enough, possibly due to point #3
  15. pilots unable to manually retrim the airplane due to points #2, #3, and #14
  16. pilots then re-activate a malfunctioning electric trim system due to point #15
  17. pilots are timid re-trimming the airplane due to violent reactions from small trim movements at the aircraft's current speed, now over 300 knots
  18. another reactivation of MCAS due to point #17 which results in a violent pitch-down due to the high speed
  19. aircraft colliding with terrain due to high terrain and low altitude
While most accidents occur as a result of a chain of events, some are more aptly described by what's known as the "swiss cheese" model, where a collection of random, seemily-unrelated events and decisions conspire to cause an accident.

Reason.gif
 
Back
Top