Disconnecting your AT is always a good starting point in a recovery.
Why is that relevant ? In both crashes the AT disconnected itself due to the airspeed disagree caused by the left AoA failure. The stupidity of the MCAS is that it put the aircraft toward the ground precisely because of the same left AoA failure _AND_ because the AT disconnected (because of the same left AoA failure) !!!
Strangely, at that point of the flight it was safer to stay with AT enabled to prevent MCAS action...
You are confusing Autothrottle (AT) with Autopilot (AP).
AP disconnected but AT in ET302 stayed engaged for basically the entire flight at Full Thrust - even when it was pointed down. Not a good thing.
Oups, you are right. My post is coherent about the Auto Pilot. The ET302 preliminary report did not have any reference to Auto Throttle, so my confusion.
But the ET302 preliminary report did contain this:
"At 05:39:42, Level Change mode was engaged. The selected altitude was 32000 ft. Shortly after the mode change, the selected airspeed was set to 238 kt"
So the pilots did correctly set the speed at a good value. This is precisely because of the left AoA failure that the throttles was still at the setting it was when the Auto Pilot disconnected because of the Speed Disagree, because of the left AoA failure.
I know it's a bit complex to understand, but the way the 737-8/9 MAX was designed produced a lots of safety consequences just from the single left AoA failure alone, and those consequences promptly puts the pilots in a situation where the established trained procedures are not enough to safely save the flight. Not that it was impossible to save the flight, but that it was not possible at the expected safety level with the established trained procedures
1) Left AoA failure with high value
-> Speed Disagree indication -> Speed Disagree procedure.
2) Left AoA failure with high value
-> Speed Disagree indication -> Auto Pilot disconnect -> Throttles stay at the setting to keep 238 kt while climbing. Aircraft still climbing at that time so nothing to do.
3) Left AoA failure with high value
-> Speed Disagree indication -> Auto Pilot disconnect -> Enable MCAS operation.
4) Left AoA failure with high value
-> MCAS generate repetitive horizontal stabilizer fast rate trim nose down commands -> Ground Proximity Warning System “DON’T SINK” alerts. -> Stab runaway procedure.
5) Left AoA failure
with high value[/b] -> MCAS generate repetitive horizontal stabilizer fast rate trim nose down commands -> attitude change from climb to descent -> speed increase with Throttles stay at the setting to keep 238 kt while climbing. -> Manual stab trim wheels unusable for the above still ongoing procedure -> crash.
My concern is that the officially published EAD did only present the analysis that contains the point 3 and part of the point 4 in the list (point 1 and 2 are not explicitly listed as a concern), while a proper analysis at that time must have see all the points in the list. Especially the point 5 that combined with the previous point create a very unusual situation far away from any trained procedure that all the contributors in this forum was able to show. Remember that it take days after the second crash before some experts did raise the possibility that the speed did play a major role in the crashes. This understanding was not available to the ET302 pilots.
The risk assessment of a left AoA failure with high
value was incredibly negligent, not only at the design time, but at the certification time, and even more incredibly at the analysis that motivated the EAD.