Quote from vilas:
The change to AOA protection mode disengagement was done through MOI in 2005. With this software change AOA protection is no longer active when alpha is less than alpha prot and side stick has not been deflected since the last autopilot disconnection.
Quote from my response:
Have you quoted the MOI in 2005 verbatim? If so, in an identical repeat of the A340 AIRPROX of 2001, it would have prevented the engagement of AoA mode (referred to by the AAIB as AoA protection Law). Taken literally, that would represent a radical inhibition of engagement, not an additional means of disengagement.
If, on the other hand, "is no longer active" actually means "will disengage", it would be an additional means (condition, or criterion) for disengagement. But it would not apply if either sidestick had been moved during a period of hand flying. That would seem odd, but perhaps I'm overlooking something.
Reply from vilas:
...I have quoted the MOI verbatim.
Thanks, vilas. Like OwainG, I am not familiar with (or have forgotten!) the term "MOI", but I presume it's a publication for informing operators of technical mods?
In my response, I overlooked a second possible inference that might be made from the MOI. I refer to this mid-sentence phrase:
"...AOA protection is no longer active when alpha is less than alpha prot ..."
Does it mean that, by 2005 (about 4 years after the A340 AIRPROX), AI had discontinued the phase-advanced engagement of AoA mode ("AoA protection Law") cited in the AAIB Bulletin of 2001? Page 11:
"Ten seconds after the autopilot disengaged, the corrected or phase-advanced angle of attack (a computed parameter which is is not recorded but can be calculated by [AI] from the DFDR data) reached the 'alpha prot' value. This [AoA] excursion beyond alpha prot caused a change in the pitch flight control law from normal law (Nz law) to angle of attack protection law (AoA law). If both sidesticks are at neutral, the AoA protection law seeks to hold the [AoA] constant at alpha prot until a sidestick pitch command is made."
Quote from HN39 (my emphasis):
My understanding is that activation/deactivation of any protection is recorded by the DFDR by a dedicated discrete parameter changing from OFF to ON, from ZERO to ONE, or whatever, and vice versa. I think it would have been a very strange omission if overspeed protection had been activated and caused the pitch-up and the AAIB report had made no mention of that fact.
Thanks for making that point.