Lion Air 737-Max missing, presumed down in the sea near CGK (Jakarta)

[Evan]

Click click, click click, hand full of yoke, hand full of thrust levers, fly the f*****g airplane!

That seems to have been the problem. The system that was fighting them is probably only engaged in manual flight. I have to believe autopilot was not available with an AoA disagree.

Apparently, this wasn't a complacency thing. It was a lack of training on abnormal operations and a lack of awareness about a crucial new system.

That and the distraction, stress and disorientation these situations breed. I think these guys just finally succumbed to battle fatigue.

You, of course, would have handled it just fine, but how did you learn how to do that? Were you trained on this or did it just come to you in a dream?

We cant expect pilots to fly the f******g airplane if we haven't taught them how the f******g airplane works,

[BoeingBobby]

Click click, click click, hand full of yoke, hand full of thrust levers, fly the f*****g airplane!

[3WE]

Perhaps they convinced themselves that the computer was somehow right and they were wrong?

Concur that this and or confusion may be part of it.

[elaw]

Perhaps they convinced themselves that the computer was somehow right and they were wrong?

[Gabriel]

I think you're right about that. Perhaps the A is for 'augmentation'.

I thought "automatic", but ok

Yes, but one thing stands out to me here: the AoA sensors are both moving in unison. This leads me to believe that the 'bad' sensor was mechanically functional but installed incorrectly, or there was a logic or voltage problem in traducing it's movements.

Exactly. It can be as simple as a zero setting not or wrongly done somewhere. Apparently the AoA sensor had been changed a couple of flights earlier.
I want to think that installing the AoA vane assembly in a wrong angle would not be a problem because the design would include some keying to ensure that installation is possible in the correct position only.

All true, but we can't brush aside the design failing here. The FDR is the smoking gun on this issue. You have potentially deadly flight control inputs being made by a system that favors a single AoA sensor, with no cross check and no 'voting' process. That is INSANE.

I agree and I didn't intend to brush anything aside. They go in parallel. That's why I said "irrespective of whatever modifications need to be done".

Now, for some reason I am always very curious about, whatever else went wrong with the technical aspects (and even with initial pilot mistakes that triggered the problem), given than that part did go wrong, what could the pilots have done to save the day and why they did not do it. That's the only reason why I am asking more questions about the pilot performance, which in particular in this case what was required to at least save the day (even without following the procedures) was so obvious and they even did exactly that... until they stopped doing it and crashed. Why?

[Evan]

TRIMDOWNA_PFDR with 2 values: 1 and 0.
This one is more confusing because the name of the variables are not so explicit and because the values of the variables are different, but I conclude (for reasons that will be obvious in a minute) that the TRIMDOWNA_PFDR is the MCAS adding nose-down trim in response to the erroneously high AoA.

I think you're right about that. Perhaps the A is for 'augmentation'.

There are aslo left and right AoA and left and right stickshaker plots, basically showing one of the AoA being about 20 units higher than the other one (units don't seem to be direct degrees of AoA) and one stickshaker basically on all the time since liftoff and the other one off.

Yes, but one thing stands out to me here: the AoA sensors are both moving in unison. This leads me to believe that the 'bad' sensor was mechanically functional but installed incorrectly, or there was a logic or voltage problem in traducing it's movements.

The design could be improved, but existing procedures and even basic flight instincts should have prevailed. Once again we have a serious issue with the flight crew (two crews actually), maintenance, and surely management and oversight.

All true, but we can't brush aside the design failing here. The FDR is the smoking gun on this issue. You have potentially deadly flight control inputs being made by a system that favors a single AoA sensor, with no cross check and no 'voting' process. That is INSANE.

[Gabriel]

Is the FDR plot showing the actual switch position or just the pitch trim command?

Both. If you check the AvHerald link in my previous post (and here below again), you will see three main graphics related to the trim:

One shows the stabilizer position (PITCHTRIMPOSITIONFDR) in units that go from 0 (apparently for fully nose down) to 10 (apparently for fully nose up) but never goes beyond 6.5.

Another one shows 2 plots:
THUMPUPMANUALFDR with 2 values: UP and NOTRIM.
THUMBDOWNMANUALFDR with 2 values: DOWN and NOTRIM.
From the name of the variable, it is quite clear that this represents where the thumb trim switch was wither active in the up position, or active in the down position, or inactive.

The last one with also 2 plots:
TIMUPA PFDR with 2 values: UP and NOTRIM.
TRIMDOWNA_PFDR with 2 values: 1 and 0.
This one is more confusing because the name of the variables are not so explicit and because the values of the variables are different, but I conclude (for reasons that will be obvious in a minute) that the TRIMDOWNA_PFDR is the MCAS adding nose-down trim in response to the erroneously high AoA.

There are aslo left and right AoA and left and right stickshaker plots, basically showing one of the AoA being about 20 units higher than the other one (units don't seem to be direct degrees of AoA) and one stickshaker basically on all the time since liftoff and the other one off.

I'm wondering if the new system might blend the manual command with the automation and result in a diminished command, to assure pilots cannot get themselves into a stall. We have to keep in mind that, if the AoA sensors had been giving accurate readings of high AoA, nose-up pitch trim IS something you want to prevent. If it turns out to be the case....

That doesn't seem to be the case and I say so for 2 reasons:

1: Boeing's bulletin describes the MCAS (without naming it) as "the pitch trim system can trim the stabilizer nose down in increments lasting up to 10 seconds. The nose down stabilizer trim movement can be stopped and reversed with the use of the electric stabilizer trim switches but may restart 5 seconds after the electric stabilizer trim switches are released." (Also in the AvHerald link)

2: If you look at the FDR plots you will see the the stabilizer position moving all the times in patterns that look like a square root symbol: It moves nose-down for a brief period of time, then it immediately moves nose-up for another brief period of time (there is no pause between the down motion and the up motion, that's the V-shape part of the square root), and then it stays stopped for a brief period of time, and then it repeats all the cycle again and again. The nose-down motion parts match parts where the TRIMDOWNA_PFDR is in "1" (MCAS active), that would be the MCAS. The nose-down motion changes to nose-up motion when the THUMPUPMANUALFDR goes to UP (that would be when the nose-up trim switch is presses in the yoke). The instant that HUMPUPMANUALFDR goes to UP also TRIMDOWNA_PFDR goes to "0" (MCAS stops trimming). When the THUMPUPMANUALFDR goes back to NOTRIM (nose-up trim switch is released) the motion of the stabilizer trim stops and remains stips for some seconds, until the MCAS kicks in again and the cycle is repeated.

This is 100% consistent with Boeing's bulletin.
Doing that, the pilots managed to compensate each MCAS activation with manual nose-up trim and keep the stabilizer going up and down but remaining more or less between 4 and 6.5 units. The fight between the MCAS and the manual trim was basically a tie.
That, until one of the MCAS actions was just "stopped" by a brief "click" in the trim switch but not compensated with application of nose up trim, and then the same with subsequent activations of MCAS that were reacted only with late, sparse and very brief applications of nose-up trim, which made the MCAS win and the stabilizer reach the 0 units position (full nose down). However, the sparse and brief activations of the nose-up trim switch show that the switch was working properly, and the MCAS and stabilizer reacting to it normally, until the end of the flight.

It looks like:
- Following the basic trim runaway procedure (MCAS or no MCAS, NG or MAX) would have saved the day.
- Not following the trim runway procedure and just following the basic flying practice and survival instinct of trimming up if you have to pull up too much to keep the nose form going down, would have saved the day (as doing it did save the first 12 minutes of the flight or so, but for some reason they surrendered).

Not to mention that:
- The previous flight had the same issue, the crew DID SOMETHING to stop the MCAS from continuously doing nose-down trim inputs, but did not report that in the log other than as "speed disagree" and "fell system light".
- Maintenance did something very wrong (a simple check of the QAR would have clearly shown the AoA disagree).

And the above is irrespective of whatever modifications need to be done in the manual (communication to pilots and airlines about the MCAS), training (for trim runaway), and MCAS or AoA sensors design. The design could be improved, but existing procedures and even basic flight instincts should have prevailed. Once again we have a serious issue with the flight crew (two crews actually), maintenance, and surely management and oversight.

So why or why?

https://avherald.com/h?article=4bf90724&opt=0

[3WE]

Blah blah but blah blah what if blah blah shouldn't blah blah

1. Yes, there are things that do not make sense.

2. This makes me THINK something bigger (and non-human factors) went wrong.

a. That and $5.00 will get you a latte.

3. Probably time to start waiting for the final report.

a. Speculation may continue
b. BUT, new facts may be limited.

[Evan]

Trim runaway is a very basic emergency procedure. I am not familiar with the training requirements in general, but I suspect that any crew would have been trained in trim runaway, especially considering that is it basically the same procedure in the MAX than the NG. Wether the training was good, effective, etc is another question. That said, my question was eve more basic than why didn't they use the cutout switch.

I've heard tell that runaway pitch trim was so common on the early Boeings that some pilots wore a glove on their pitch wheel hand. But that instinct alone isn't going to stop the problem here. AFAIK you are right about the procedure being the same on the NG as on the MAX.

Forget about the MCAS for a second. The FDR shows how in the first portions of the flight the pilots were consistently using the nose-up trim switch to stop the automatic nose-down trim and trim back to the previous stabilizer position. But then the trim switches inputs were more sparse and shorter, LETTING the MCAS win. Why? Forget about training, this is anti-intuitive and anti-survival-instinct.

Is the FDR plot showing the actual switch position or just the pitch trim command? I'm wondering if the new system might blend the manual command with the automation and result in a diminished command, to assure pilots cannot get themselves into a stall. We have to keep in mind that, if the AoA sensors had been giving accurate readings of high AoA, nose-up pitch trim IS something you want to prevent. If it turns out to be the case, I think the fix must include providing a third source of AoA data and a requirement that two sources are in agreement for the system to do this. And, of course, there's always the cutout switches...

[Gabriel]

Why on Earth would they not use the cutout switches after repeatedly fighting runaway trim? Why would they just continue to fight it? I'm going to make the wild assumption that [someone] didn't train them on this rather essential procedure. Perhaps because [someone] didn't require such training.

Trim runaway is a very basic emergency procedure. I am not familiar with the training requirements in general, but I suspect that any crew would have been trained in trim runaway, especially considering that is it basically the same procedure in the MAX than the NG. Wether the training was good, effective, etc is another question. That said, my question was eve more basic than why didn't they use the cutout switch.

However, perhaps there is something in the MCAS logic that lets it win in the end. Still waiting for a bit of transparency from Boeing on this system...

Forget about the MCAS for a second. The FDR shows how in the first portions of the flight the pilots were consistently using the nose-up trim switch to stop the automatic nose-down trim and trim back to the previous stabilizer position. But then the trim switches inputs were more sparse and shorter, LETTING the MCAS win. Why? Forget about training, this is anti-intuitive and anti-survival-instinct.

Or perhaps the crew became too distracted in trying to find a solution in the FCOM for a system that was left out of the FCOM...

Ah, the landing gear light bulb effect.... That could be... up to a point. If they were visual and experiencing low Gs, that would be hard to miss.

[Evan]

Why on Earth would the pilots do that? Why would they stop adding enough nose-up trim? Especially when they lost control in a dive that they could not recover with elevator alone?

Why on Earth would they not use the cutout switches after repeatedly fighting runaway trim? Why would they just continue to fight it? I'm going to make the wild assumption that [someone] didn't train them on this rather essential procedure. Perhaps because [someone] didn't require such training.

You might as well burn your entire fleet at that point. Until airplanes fly themselves, the rest is immaterial.

However, perhaps there is something in the MCAS logic that lets it win in the end. Still waiting for a bit of transparency from Boeing on this system...

Or perhaps the crew became too distracted in trying to find a solution in the FCOM for a system that was left out of the FCOM...

[3WE]

1. I am struggling to understand what the pilots were trying to do or thinking.

2. The CVR becomes critical to try to understand it.

#2 may contain comments similar to "What's it doing now", which would address #1: Potentially situationally overwhelmed with warnings and blinking lights and strange behaviors and multiple Evan modes and procedures.

[Gabriel]

Guys, sorry for distracting you from your philosophical war.

Have you seen the quote from KNKT retrieved from AvHerald that I copied in my latest post? And did you see the associated FDR plots also published in Averald?

I am struggling to understand what the pilots were trying to do or thinking. The trim automatically trimmed down a lot of times, the pilots compensated by trimming up with the yoke switch (that action both trimmed back up and stopped the automatic trimming for some seconds). And they were managing to keep the plane in control with this reaction against automatic trim inputs by keeping the trim more or less in the same position on average (combining the nose-down auto trimming and the nose-up manual trimming). Of course we could talk about cutting the trim for good with the cutout switches, but there is something much more basic. At some point the manual nose-up trim via the yoke switch diminished. They didn't disappear (which would have been perhaps less strange), but they became more sparse and shorter, allowing the autotrim to win the battle with the net effect of a progressively increasing nose-down trim.

Why on Earth would the pilots do that? Why would they stop adding enough nose-up trim? Especially when they lost control in a dive that they could not recover with elevator alone?

The CVR becomes critical to try to understand it.

http://avherald.com/h?article=4bf90724&opt=0

[ATLcrew]

Ok. so, what's the assfrontwards way to go about this?

Beats me. I don't claim to know what the answers are, I just have a pretty good idea of what they aren't.

[3WE]

***calculated risks with people's lives***

How many children have died?

https://en.m.wikipedia.org/wiki/Big_wheel_(tricycle)