The accident of the flight Air France 447 was deadliest in the history of the airline and, since it happened almost 3 years ago, has appeared every now and then in the media. Today, while having lunch with two colleagues I learnt about a documentary about it that was shown in French TV channel France 5 last Wednesday.
The documentary, produced by Bernard Vaillot, is titled “Vol Rio-Paris, les raisons d’un crash“, lasts about 50 minutes, and the programme in which it was emitted included an interview afterwards (complete programme duration 1h08′).
The documentary includes as main attraction a reconstruction of the last 4 minutes of the flight. This reconstruction is built from the cockpit voice recorder (CVR) and the flight data recorder (FDR), often referred as black box. A complete transcription, with comments, interpretation and an exhaustive description of the events can be found in the different interim reports, briefings and releases made by the French BEA (Bureau d’Enquêtes et d’Analyses), the body in charge of the accident investigation.
In the BEA reports [PDF, 10.3MB, New Findings, pgs. 77-78] some facts were established: such as the loss of correct speed indication due to the icing of the pitot probes, the disconnection of the autopilot and the subsequent reaction from the pilot, pulling the stick towards him nosing up the aircraft, increasing the angle of attack until the aircraft enters in stall.
Other interesting features of the documentary are the testimonies of different pilots, including a real stall exercise performed by a flight instructor (from minute 25). He first shows how to recover from a stall, then he simulates the wrong reaction from the AF447 pilot, pulling the stick towards him, pitching up, while the stall alarm was still sounding.
You may see the following trailer below before venturing to see the whole programme.
EDIT: The trailer video has been deleted from Youtube, however, now the full video is available:
Unfortunately, I haven’t yet seen a subtitled or doubled version of it.
The Blog by Javier 
Pingback: Stall exercise | The Blog by Javier
An english transcript of the last minutes in the cockpit, plus some reasoning on what may have gone through the pilot’s minds during those terrible mintues are in this fine article:
http://www.popularmechanics.com/technology/aviation/crashes/what-really-happened-aboard-air-france-447-6611877
There is one glaring fact being missed. The stabilizer was in the nose up (leading edge down) position throughout the descent. The auto trim function was inoperative. This was shown by the lack of any stabilizer movement when the pilot move the sidestick forward during the descent. This moved the elevator but NOT the stabilzer. This means that the system was in basic law, not normal or alternate law. Thus they were locked into the stall. They could have used a manual input to get the stabilizer into a correct position BUT THEY DID NOT KNOW THAT IT WAS NEEDED. Why? No indication of Stabilizer position? Too many distracting warnings? They were dead unless they could get the nose down, but this was not possible with the stabilzer locked.
The auto trim function in normal and alternate law is nice and convenient, but as it moves the stabilzer to relieve the pilot from retrimming the pitch
(Unlike previous aircraft or Boeings) a failure is dangerous unless recognized.
I personally experienced a similar encounter in the 757. We experienced unreliable airspeed indications (actual speed: 290, unreliable indications: 140 knots) due to the pitot-static system icing up while at -15 degrees SAT at FL280 in the weather. However, we survived because I hand flew the airplane using pitch and power as I was taught 25 years ago in the USAF as a fighter pilot. I did not panic. I did not stall the airplane. Perhaps the safety system should focus more on the issue that the Air France pilots panicked and stalled the airplane because they didn’t uphold their number one priority…maintain aircraft control. Their improper response to the unreliable indications put them into a stall and was their demise. If a pilot can’t manually hand fly an airplane using pitch and power, then how would we expect them to be able to recognize and recover from a stall. Maybe the FAA should have made airline pilots practice hand flying in a variety of different unexpected scenarios in the simulator versus have them demonstrate a stall recover from a scripted training scenario. More information on my encounter in the 757 at the link below.
http://acesaero.com/air-france-447-a-non-bureaucratic-assessment
Pingback: On why blog posts are read | The Blog by Javier