But then, stabilizer position dropped slowly over 2.5 minutes by 0.2 units to nose down, while crew was keeping nose up by force on both control columns. Did mechanic trim not work at all?
That would explain the discussion at 5:41:46 - the mechanic trim wheels were not operating. There is also no attempt to use the electric trim switches before and during that discussion.
My understanding is that full manual trimming can be quite stiff due to aerodynamic forces and sometimes requires both crew members to apply force to the trim wheel, and AFAIK, that applies at more typical flight speeds. At or above Vmo, it would be all the worse.
There *does* seem to be an attempt to use the electric trim switches in the discussion at 5:41:46:
1)The captain asked if trim was working.
2)The first officer stated it was not.
3)The first officer offered to try using manual trim.
4)The captain gave him the goahead.
5)The first officer reported that manual operation of the trim wheel was not working.
From the fact that the first officer presented manual operation of the trim wheel as an alternative option under 3), it would seem that his statement that trim was not working at 2) was regarding something other than brute-force manual trim, i.e, operation of the electric trim switches.
Then again, he ought to have known that the electric trim switches would not work, because he had (correctly) *suggested and executed* the task of actuating the stab-trim cutout switches, which disabled all trimming except by muscle power operating directly on the trim wheel.
I will note here that distinguishing between the trim switches and direct operation of the trim wheel in this context is a bit confusing. We have three types of trim input that are relevant here:
1) Fully automatic trim inputs from MCAS, executed by the trim motor.
2) Pilot-commanded trim inputs executed by the trim motor.
3) Direct muscle-operation of trim by the pilot.
2) and 3) are both "manual" in as opposed to 1) being "automatic", in that the decision to change trim is taken by the pilot, not by an aircraft system, but 3) can also be considered manual as opposed to 2) being "powered", in that the trim change is executed by the pilot applying muscle power to the trim system. Considering that all three had a bearing on the outcome of the flight, this has the potential to create confusion as to whether "manual" trim refers to 2) or 3).
In any case, the whole exchange above took place over eight seconds, which to me indicates that the First Officer likely tried briefly to move the trim wheel and gave up, and neither asked for nor received assistance from the captain in moving it, which may well have been required.
Still, I wonder why they did not react to the overspeed warnings - the whole situation looks like an information overflow, with ice warnings also appearing along the way.
It certainly does.
At 5:43:04, the overspeed condition made controlling pitch by both control columns impossible. At that point they activated electric trim again, because mechanic trim had no effect. That allowed to trim nose up again.
I'm not sure that it actually wasn't possible for them to control pitch. They did not seem to actually be descending yet at this point, and while up-pitch authority was reduced, they were banked for a turn, and could have regained some vertical authority by rolling wings-level. Also, the graphs in the report show control column position, and when MCAS puts them into their final noseover, they put in a noticeable amount of extra aft-stick over what they had been using prior to that point, which indicates that they had at least some strength margin to pull back further on the stick.
---------- Post added at 14:57 ---------- Previous post was at 14:44 ----------
The Concorde accident likely is the closest you can ever get to a single point of failure.
And even that report had multiple events leading to the metal part on the runway...
Actually, the rudder hardovers Boeing had on the 737 back in the 90s are what come to mind for me. As the rudder was hardly used at cruise, the hydraulic servos in the rudder cooled down to ambient temperature, while the remainder of the hydraulic system remained hot. On approach, with the rudder being used more, hot hydraulic fluid was injected into the cold servos, and thermal expansion caused components to jam in such a way that the rudder either went full-travel to one side, or gave a reversed response to the rudder pedals. It was uncertain at the time, but I think that history has showed that crashes were pretty much single-cause results of that defect.
But even then, while it's a fairly clear-cut single-point-of-failure case *now*, contemporary reports were written before the problem was understood, and listed multiple potential causes.