AF 447 Thread No. 5

[GarageYears]

My take on what happened

After reflecting on what I've read from the report (and thanks to the various posters who translated and then refined the CVR transcript over the course of the last few hours), here's what I think happened taking a big picture look:

- AP disconnect took the crew by surprise with immediate confusion related to speed
- At no point was UAS identified or the appropriate procedure/SOP
- The PF input was significant in terms of SS deflection resulting in the over-correction of the initial mild roll, coupled with a 3/4-travel deflection NU input leading to a climb
- Despite the climb, the focus appears to have been on the lost speed indications
- Since speed was the focus (and was unavailable), the true loss of speed due to the zoom-climb was masked
- When the speeds did return the aircraft was truly slow and correspondingly trimmed nose high with negative AoA (the stall warnings and attitude of the aircraft were not acknowledged)
- The aircraft was descending NU in the stall when the Capt returned to the cockpit
- Since he was not in the cockpit at the beginning of the event, he arrives to find the aircraft altitude unwinding, the attitude indicators showing pitch up, with the power in TO/GA
- At no point was the initial climb discussed to allow the Capt to understand where all the airspeed went
- Speed continues to be a focus, but distractions due to roll seem to predominate
- Altitude was not commented until very late in the sequence and seems to be identified with surprise (as if the loss of altitude was not noticed prior)
- Once the altitude loss was identified, the confusion at this point led to continued NU and TO/GA, seemingly in an attempt to power out of the condition, but with no diagnosis of the AoA or trim
- Since NU (climb) did not seem to be working, ND was tried, but this occurred way late and was not sustained
- At FL100 control was passed to the PNF, but NU/climb commands continue basically until impact

I have purposefully not added details in the sequencing above, since it is too easy to get lost in the fine print and ignore the bigger picture.

Above all, the initial loss of the speeds and the transition from "loss of speed indications" to "we've no speed due to stall" seems to have been the root here.