Update on Air France flight 447


I do have a strange interest in air crash incidents, I am not sure where it has stemmed from, perhaps it is from working in both the Nuclear Industry and in Military Aircraft for many years that has got some sort of safety culture in my head and being interested in what caused such a piece of advanced machinery to crash.  Obviously now I am training to complete my ATPL I guess I have a different interest which is learning from mistakes that others have made, and for Flight 447 I have a particular interest because it is an Airbus A330 which is the aircraft I long to fly for Thomas Cook (I think this has something to do with hopping across the pond on A330/340s with SAS twice a month for 18 months).

Anyway, the investigation…   It would appear from reading the BEA articles that the aircraft pitot tubes failed as suspected, this in turn gave erroneous data to the Air Data Computer.  To protect the aircraft from doing something stupid, the Air Data Computer changed the aircraft mode to Alternate Law.  This is the equivalent of your computer going into Safe Mode, and many of the unique safety features of the aircraft are turned off.  One of those is flight-envelope protection, whereby Airbus aircraft simply will not let you do something stupid like put it into a 60degree banked turn, or pull back on the stick to an unusual attitude and stall the aircraft; it’s computers merely stop the aircraft from going beyond what it is capable of.

Unfortunately for flight 447, the flight envelope protection seems to be what sealed the aircraft’s fate.  The pilot flying the aircraft pulled back on the stick increasing the angle of attack and slowing the aircraft until it stalled, and instead of pitching down and recovering from the stall, it would seem he continued to pull back on the stick until the aircraft hit 40degrees at which point it entered a violent stall and basically dropped 38,000ft in 3 minutes, tail-first.

This is only the preliminary information released and obviously there will be a full verdict coming out later, as with any crash there are likely to be many circumstances that lead to the aircraft falling from the sky.  It is however very sad that it seems to be pilot error-induced, these pilots were well trained with a high number of hours like the pilots of the Buffalo Dash-8 air crash, where the pilot of the aircraft refused to push the nose down when the stall warner sounded and even when the stick-pusher pushed the stick forward he continued to pull back.

The sad facts are that after the autopilot disengagement:

  • the airplane climbed to 38,000 ft,
  • the stall warning was triggered and the airplane stalled,
  • the inputs made by the PF were mainly nose-up,
  • the descent lasted 3 min 30, during which the airplane remained stalled. The angle of attack increased and remained above 35 degrees,
  • the engines were operating and always responded to crew commands.

As a trainee pilot this really does interest me what would drive a pilot to do this, recovery from a stall is a drill that every pilot must practice hundreds of times in their career and yet 300 people seem to have needlessly died in 2 years because the pilot in command did the exact opposite of what they have trained and practised to do and should be able to do in their sleep.

As part of our training here in Jerez, we do something called “Upset Training” where we use a Slingsby Firefly aerobatic aircraft and spend 5 hours learning how to recover the aircraft from attitudes that would be extremely dangerous in a commercial aircraft.  It is something which is not required for the ATPL licence and so there are many pilots out there who have not had any training in how to do it.

I watched a documentary last year made by NOVA, they had a theory that the pilots of the Air France crash did not know how to recover the aircraft from a violent stall after pitot failure from super-cooled water droplets.  To prove the point they had a specialist test pilot from the US Air Force place an Airbus A330 simulator in a violent stall and it took him the best part of 20,000ft to fully recover the aircraft!  There is a Link here to the documentary/demonstration showing what the crew should have done when the pitot ports failed.

It is quite alarming how accurate the documentary made last year is similar to the facts that are being released this week, it took the Air France 447 crew over 2 minutes in a stall to select TOGA power and even then they continued to pitch the nose up.

The full update report made by BEA is here and there is a useful summary at Fear of Landing

Enhanced by Zemanta

~ by globalste on May 30, 2011.

4 Responses to “Update on Air France flight 447”

  1. Ooh, you are in Jerez. I’ve heard wonderful things about the airfield/people there.

    I haven’t seen the test pilot simulation of the accident, that’s very bad. I did hear an interview on NOVA the other day about commercial pilot safety and that we need to stop teaching pilots about traditional problems and instead focus on teaching them how to deal with computer malfunctions and how to deal with faulty instrument readings. This accident is interesting in that respect, because I can see it being used on either side of the argument.

  2. Hi – I recall an accident (an Aeroperu 757, I think) where the static vents were taped over for cleaning and the aircraft took off in darkness with the tape still in place. Almost immediately after takeoff the pilots lost both their instruments and situational awareness, although in this case the autopilot remained active long enough to effectively execute a 1-g roll (due to incorrect attitude data). Because the centrifugal force balanced the aircraft, and they had no external view, the crew were not even aware they were flying inverted at one point.

    My own limited experience of instrument training with a hood restricting my vision to the instrument panel, I can vouch for the fact it can feel like the plane’s attitude is different from reality some of the time (I never enjoyed those exercises!).

    In the case of AF447, it’s generally believed that the crew must have known they were falling rapidly, but I can’t help wondering whether it felt to them like they were in a dive, rather than pitched nose up, and they were trying to pull out of the ‘dive’ to avoid over-speed (bearing in mind they had no idea when/if they could trust any speed or attitude indications), only selecting full thrust for part of the descent when it didn’t feel like they were pulling out of the dive.

    Still can’t understand why they didn’t have a GPS display somewhere showing their true speed though.

    • Interesting points, I cannot understand thogh why they didnt use the standard procedure for a fault such as pitot icing; 4-5degrees nose up and 85% power which is what they should have done first and then sorted the other issues first. Additionally, they had the back-up Attitude Indicator in the centre of the panel to use, which is totally independent of any failure they had (as was the backup altimeter) one of the problems with large aircraft is recovery from an incipient stall often involves losing minimum height and applying TOGA to recover, Airbus/Boeing have recently worked with airlines to ensure that a pitch-down nose attitude is used (standard stall recovery).

      Hopefully we will find out what caused them to do what they did, but I doubt we ever will find out why the pilot flying pitched up at 7000fpm at pitot failure and continued to pitch up during stall. It is very sad that the pilot flying did not use his working instruments (eg altimeter and AI), pitch-down and TOGA to recover the stall and then select 5deg nose up and 85% power.

  3. […] autopilot and autothrust disengaged. This is undoubtedly because of the false data. As one commentator described it, at this point the plane is now similar to a  Windows PC in “fail […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: