http://www.bea.aero/en/enquetes/flight.af.447/flight.af.447.php

The 3rd report was published July 29th.

There is also a news conference available. (in french). (29th of july).
Question/answers is only in text mode (so I assume available in english).

Short story: the plane stalled.

First loss of indicated airspeed is probably due to icing. Ambiant noise
in cockpit changed just before in a way that was consistant with plane
flying through ice crystals.

Afterwards, loss of airspeed was NOT due to icing, but rather an angle
of attack that was too high for pitots to give reliable speeds, so the
computer ignored them.

Ironically, when pilots went with nose down attitide, the stall warning
restarted, and when then had nose up, the stall warning disapeared
(since speeds became invalid). They say this could lead to confusion.

However, the angle of attack is NOT a piece of information presented to
pilots. (and this is something BEA now recommends be added as standard
information in cockpits so pilots can know just how close/far they are
from a stall.

In essence, in phase 3 of the flight, the stall warning was disabled due
to high angle of attack which was so high it caused the stall.

At the very first instance of loss of speed, the pilots failed to
perform the documented procedure in such a case. Such a proceduce wants
a pitch of 5°. The pioots went for a pitch of 10 to 12° whic caused
plane to climb at up to 7000 fpm to 38,000 feet. (the report has graphs
which show how high a plane can fly at a certain speed and weight).

The loss of speed during the climb to 38k feet made the plane stall
becuse it was too high, too slow and too heavy for that altitude.

The initial loss of speed indicators  is not explained, but they say
that ambiant noise in the cabin changed in a way consistent to flying
through ice crystals. So the theory of pitots freezing up has not been
ruled out, In the press conference the work "likely" is used to almost
confirm this is what happened.

The BEA has formed a "human factors" group to study the ergonomics of
the systems and why/how the crew appeared confused and unaware of what
was really happening. (personnal question: wouldn't people have felt the
vertical acceleration when plane went from positive climb rate to a very
high negative fall rate ?

Also, the BEA guy confirmed that one of the 2 pilots moved the throttle
to idle from TOGA. (initially, they put the pedal to the metal and later
reduced speed to idle (in the middle of a stall nonetheless).

Angle of attack indication is not something which is presented to
pilots. BEA recommends this be added to give them an idea of how close
to a stall they are.

BEA also wants a camera to record the cockpit instrumentation board.
Much information such as the radar echos are not recorded.

Right now, the 330 transmits its position every 10 minutes. BEA
recommends that transmission of location not only every 10 minutes but
also when certain conditions are triggered (alarms, stall etc). This
would greatly help reduce the area to be searched after a crash.

None of the 2 pilots in the cockpit had had training on high altitude
manual piot in alternate law which is what happened when pitots
(allegedly) froze up and they reaction was to climb. The procedure
called for 5° pitch. HOWEVER, thet had had training for such situationa
at low altitudes where procedures calleed for 10-15° pitch, and this is
what they apparanety did in high altitude.

Another aspect: he FDR records data from the left side of the cockpit,
not the right side.  So it is not know what the PF (who was on the right
seat) actaually saw in terms of airspeeds.
--
misc.travel.air-industry is a moderated newsgroup. Please mail messages to
m...@airinfo.aero, and see http://mtai.airinfo.aero for the FAQ and policies.