https://theaircurrent.com/aviation-s...b7JuyicCOXJVhU

Another shell crashes into the Bismarck.

"More than half responded with the wrong procedure."

So these were presumably head of class pilots expecting to perform some sort of AOA failure response and they fluffed it. Is my reading accurate?

Not sure they were head of class pilots, my impression was that they were going to do this testing with an "average" selection of line pilots.

But yes it seems the testing did not go as planned.

Seems like the cure is to simplify the checklist:
Are the following happening:
1) autopilot kicked out unexpectedly
2) pitch trim moving unexpectedly
3) pitch control force increasing unexpectedly

Return pitch control force to zero with control wheel switches and then immediately disable trim motors.
Set the pilot not flying the job of operating the trim wheel as directed by the PF.
If that does not work, pilot not flying will re-enable the switches as directed by the PF only long enough to use the trim switches to zero pitch control forces and immediately disable them when completed.

As the pilot is able, immediately declare emergency and land ASAP.

These are indicators the pilots can memorize.

Would you buy a car where the:
1) cruise control kicked out unexpectedly
2) Steering moves unexpectedly
3) Steering control force increased unexpectedly ?

Really, all you have to do is recycle the cruise control switch and get your passenger to hang onto the steering wheel. Anyone can remember that.

Personally, I'd prefer a little more predictability out of my vehicles. Especially the flying ones.

MechEngr, how about the stick shaker noise, IAS disagree warnings going off at the same time? Which are you going to attend to first? The stall warning, a runaway stabilizer or airspeed unreliable? ALL of these are memory items and needed to be attended to ASAP, and all are potential symptoms of an AoA failure. Not so simple is it.

Run this while trying to figure it out.

For some reason the Emergency AD said DO NOT re-enable the switches (no reason given just don't do it).

Also turning them back on did not work for the Ethiopian crew - Question. Why did they turn them back on, if they did not intend to use them to trim?

So when a Boeing representative said that the problem was a simple paperwork formatting issue that could be addressed in "a matter of days" that wasn't entirely accurate, then? Nothing about "oh, by the way half of the pilots failed to implement the correct procedure in our revised plan". That part wasn't important, I guess.

I'm frustrated here because they are an engineering company but they sound more like a bloody hedge fund that has been caught cooking the books. Engineering firms are supposed to have a culture of accuracy, not deception.

Sorry- Boeing used to be an engineering company- that changed starting around the mid 90's when MDC bought Boeing and installed way to many Jack welch interns. With over 30 years at Boeing, I can say they were doing pretty good when I left in the mid 90's- along with about 9000 others.

Name a successful program since the 777 ...

There was both a trim increment from the control wheel switch from one of the pilots and an autopilot fail warning that would only be issued if they tried to re-engage the autopilot. Both re-enabling the motors and re-enabling the autopilot are cautioned against in this situation. It's not clear why they did not continue to trim with the trim switch.

There is no "runaway stabilizer." That term I specifically avoided because it's subjective. Sometimes the trim runs. And then stops. If it reaches the mechanical stop was it running away anymore? Nope, it has stopped running. But I bet the control forces are too high. So deal with what the real problem is and not make the pilot make some guess. If STS trims too much or too little, pilots input manual corrections. Was that runaway? What if the trim runs for more than a few seconds but the control forces are acceptable? Is that runaway?

It should take less than 10 seconds to determine what to do about the stall warnings and that is independent of control forces. In both crashes there was only one-side with the error and the performance of the plane was unchanged, indicating the warning was false. Am I mistaken or is that not part of all 737 training for 20+ years? How old is the NG series? The ET302 disaster unfolded over nearly 36 times that time span.

Tweak away, but the AD instructions were terrible with only vague actions to take and weak symptoms to determine it was the correct procedure.

Take Captain Sully's testimony and now we find greater than 50% of the pilots tested (knowing they would get) an AoA failure messed up the correct response and the fact that ALL 3 MCAS real life events took longer than (much) 10 seconds.

I feel at this stage you need some evidence to support your words.

It seems very clear the 737 MAX is a different beast to the 737 NG - simple fact if both crash aircraft were NG we would not be here.

Reading between the lines and other HF information 30 seconds would be a minimum but up to 2 minutes for a "average" pilot to start to correctly identify and carry out required memory items. They expect the memory items will not be carried out correctly in most cases.

My gut feeling (which isn't worth much and needs some data from Boeing to actually decide) is that the whole reason for this situation is that the contribution of MCAS in normal operations is slight and that it was entirely there for the control linearity requirement.Based on my gut, there isn't much to protect from as it does not do stall recovery only feel, and this is the main reason that no one thought of what MCAS would do if the AoA system was lying. If it was intended to do very little how could anyone imagine it would do very much?


I expect that the new version of software will have a big chunk of changes to validate AoA inputs and that this particular situation will happen as often as "trim runaway" does on the NG, which seems to be - has never happened.

The 10 seconds is for the independent determination of which set of instruments is reliable in a one-side stick shaker. I don't know how to tell if pilots are unable to feel control forces, but the ET302 flight indicates the PIC countered the MCAS input starting at 12 seconds after the MCAS input began. I think it was even less time for the Lion Air captain on the accident flight, sometimes as little as 5 seconds, cutting off MCAS mid-operation. What neither was primed to do was neutralize the forces and immediately shut off the trim system. They did not take 2 minutes to realize a need to retrim, though the ET302 PIC never neutralized the trim.

In the event of a depressurization event at altitude, 2 minutes to react can mean everyone is dead, so I am not buying 2 minutes of dithering when things go badly. I would like to see if they changed that list, because if they are looking at the stab indicator and saying "I don't think it's running away" because MCAS or an intermittent short let off for a few seconds, while hauling with 80 pounds on the control wheel to keep the nose up, then just maybe that is not a good way to write a procedure.

It's also the case that none of those in the test planted the plane, which is more evidence that something is wrong with the procedure as written. More detail about what deviations they made would be interesting. One thing that isn't necessary is to diagnose the reason for the trim problem. It's as likely to be caused by a wiring fault as something predictable and in neither case can the crew effect repairs mid-flight. So the emphasis should be on symptom relief.

80 pounds of pull represents a strongly expressed opinion by the pilot, and it seems unhelpful that a "feel-augmentation" system can override it.

Adding another column switch that (temporarily) disables MACS when the column pull reaches these levels might be helpful.

PS It might be best implemented in the wiring, rather than keeping the computers in this time-critical loop.

[QUOTE=CurtainTwitcher;10638642]MechEngr, how about the stick shaker noise, IAS disagree warnings going off at the same time? Which are you going to attend to first? The stall warning, a runaway stabilizer or airspeed unreliable? ALL of these are memory items and needed to be attended to ASAP, and all are potential symptoms of an AoA failure. Not so simple is it.

Run this while trying to figure it out.

Flaps out: a/p and F/D off, 80% N1 10° up, 75% N1 4° up clean aircraft will not fall out of the sky.

David P. Davies' solution to a slight pitch up into a stall in the 707 used only the elevator, not the much larger horizontal stab, and the solution was safe, effective, and benign. The MAX uses the stab, and trims forward for up to 10 seconds at a shot. That's at least an order of magnitude, maybe two, larger trim input than Davies'. I refuse to believe that a correction that large is "only feel."

I would absolutely love to hear what that man would have to say about the current state of Boeing and the MAX. He was a big fan of the Americans engineering and very well regarded by Boeing if I'm remembering his interview correctly.

I really think it's past time to drop the pilot-blaming, which is exactly what this "should have done" conversation amounts to.