View Single Post
Old 5th Apr 2019, 11:44
  #3294 (permalink)  
fdr
 
Join Date: Jun 2001
Posts: 584
Originally Posted by yanrair View Post
So. MCAS being sorted. Good thing. Does that leave us still with a much bigger problem? The elephant in the room! What’s going to happen next time an automated function badly misbehaves- and it will.
Y
Having done reversion testing, I was wondering what I had missed in the wording, so have dusted off the manual and had another look at the FCTM wording for the runaway stabiliser and the follow on manual trim. In the latter it mentions excessive airloads, and then discusses having both pilots applying effort to the system....in extreme cases airloads have to be relieved, and the next sentence covers returning the aircraft to the actual in-trim speed while attempting to trim using the manual trim wheels.

... the problem with the wording is that having the aircraft now trimmed to an undesired speed (....Mach 2 etc from the runaway trim case...) then the crew are to match the trim case while using both pilots if needed to manually change the trim. So, the crew bunt over to match the in trim speed for the nose down runaway case... which is all well and good except that is generally where planet earth is also located. If you are in an A-10, that is possibly second nature, or if you are current on Stukas, but for an RPT aircraft, that seems to be rather light on reasonableness. If this is reasonable, then perhaps the OEM should load up one of their aircraft with the management and DERs involved, and go demonstrate the technique at low level and high speed and determine how reasonable that is in the real world where the crew are confronted with an unexpected surprise in close proximity to the ground or water. On second thought, it would appear unfair to ask the OEM to do that... or anyone else for that matter, pilots, cabin crew and passengers.

The 737 is the worlds most prolific jet transport, by a sizeable margin, the manner that the trim issue is worded, and the limited protection that exists with the Part 25 wording for compliance doesn't seem to be impressive, particularly for a design where current activity is making "a safe aircraft safer". Perhaps runaway trims just don't happen and there is no risk from the underlying design, but I'm pretty sure that the last person who tried to pass an FAA check ride with two pilots having all hands on the controls at the same time ended up with more than just a discussion item.

25.255 provides for a 3 second trim error to be introduced... that is not what either of the recent crews experienced. Perhaps 3 seconds is perfectly fine as a value, in which case, the crew just need the chicken entrails to know when the system may exceed 3 seconds of uncommanded motion.

Humans are adept at adaptation, we can find workarounds for almost anything, but some 50 years on, if the cure for the system is as described, it may be time to look closely at grandfather rights vs public safety. Boeing builds a good plane, however perhaps it deserves some more thought into the system architecture, and at the very least some reinforcement training for crews that may need to deal with the issues related to the design.
fdr is offline