Discussion:
Air France 447 : Black Box FDR found !
(too old to reply)
JF Mezei
2011-06-13 07:58:37 UTC
Permalink
BTW, on May 27th Air France issued a press release
http://alphasite.airfrance.com/fr/s01/communiques-de-presse/#communique3253

AF seems to protect its pilots, saying they were monitoring weather and
that the first problem was loss of speed indicators.

Bea issued some clarifications on its May 27th release.
http://www.bea.aero/fr/enquetes/vol.af.447/vol.af.447.php

I have a feeling that the report avoids mentioning "thunderstorm"
because of ambulance chasing lawyers.

PF calls the sleeping captain that they are about to hit slightly
stronger turbulence and that he'll call back once they are through it.

They turn left by 12°. Obviously to avoid "something".

at 02:10:05 auto pilot disengages, and there is a stall warning,
followed by indicated airspeed drop from 275 down to 60 knots on primary
and a few seconds later on secondary (ISIS) indicators.

QUESTION: could this be caused by the wind in a thunderstorm which would
suddently be a strong tailwind, reducing IAS ? Or is the only
explanation pointing to faulty pitots ?


They increased pitch, started to climb at up to 7000 ft/min but pilot
then made nose down commands and got climb rate down to 700 ft/min at
altituide 37,500. To me, it looks like a controlled "climb over the storm".

If the training is to always point nose down in a stall, and they
pointed nose up, it must have been because there was something below
that they really didn't want to get into.


What I don't know is whether the initial stall warnings were due to
faulty pitots, or whether they were due to wind/storm. Since the plane
was able to climb from fl 35 to fl 38, it was obviously not stalling.

Perhaps pilots had learned to disregard such stall warnings due to
faulty air speed indicators. And the first time the stall sounded, they
ignored it without "damage" but once they reached 38k feet, they ignored
a real one.

What is still not clear is why during the climb to 38kfeet, engines were
at max (TO/GA setting), but a couple minutes later, they were at iddle.

Seems to me that when the PNF declared they had dropped to 10k feet,
someone should have though about putting thrust to the max.

The report does note some rather large oscilations in banking by up to
40°.


Is it possible that the airspeed indicators do not function when the
angle of attack is too high ? This would explain why speed indicators
came back when aircraft put nose done inputs to reduce pitch/angle of
attack. So the aircraft would have been in a real stall because the
nose was too high up and engines too slow, but the computer not issued
stall warning because the airspeed indicators were considered invalid
when they could have been valid ? (at 35° angle of attack with engines
at IDLE, what sort of airspeed would perfectly working pitots record ?


Based on what my buddy Mr Google has told me, the 330-200 has a service
celiling of 41k feet. Is it correct to state that such an aircraft
would need to have engines at far much than idle when cruising at that
altitude this early in flight (when it is still heavy with fuel) ?

Does anyone have an explanation on why the engines would have been set
to IDLE after being at TO/GA ?


Is it possible that due to violent turbulence, they had to reduce
aircraft speed because turbulence was too violent ?


They obviously need to release far more information. For instance, did
they advise the crew/passengers of upcoming turbulence that could be
significant ? Did passengers fly around the cabin or were they all
strapped in ?




Would any/all thunderstorms capable of this generate lightning ?
Lighthing should have been visible by pilots more than 2 minutes before
hitting it.
--
misc.travel.air-industry is a moderated newsgroup. Please mail messages to
***@airinfo.aero, and see http://mtai.airinfo.aero for the FAQ and policies.
Uwe Klein
2011-06-13 08:58:58 UTC
Permalink
Post by JF Mezei
What I don't know is whether the initial stall warnings were due to
faulty pitots, or whether they were due to wind/storm. Since the plane
was able to climb from fl 35 to fl 38, it was obviously not stalling.
Perhaps pilots had learned to disregard such stall warnings due to
faulty air speed indicators. And the first time the stall sounded, they
ignored it without "damage" but once they reached 38k feet, they ignored
a real one.
Seems to still be a bit diffuse.

Pitots weren't really faulty but overwhelmed by icing for a short duration.
( this sounds like a bit of hair splitting, for sure, but imho is a neccessary
differentiation. Icing problems are not limited to Thales. Goodrich has about
the same issues. )

Stall warning is AoA derived and _invalid_ ( i.e. not shown/sounded ) below
60kn airspeed.

From my view it looks like the pilots thought they were on the Stall/noStall border
while in fact they lingered at the novalid AoA/valid AoA border (+-60kn forward speed).

This will be judged pilot/training error.

Airbus will nonetheless try to alleviate such missjudgements with some helpfull
software and manual/training changes.
Those then will keep rabid conspiracy theorists busy :-(

uwe
--
misc.travel.air-industry is a moderated newsgroup. Please mail messages to
***@airinfo.aero, and see http://mtai.airinfo.aero for the FAQ and policies.
matt weber
2011-06-13 18:45:43 UTC
Permalink
Post by JF Mezei
BTW, on May 27th Air France issued a press release
http://alphasite.airfrance.com/fr/s01/communiques-de-presse/#communique3253
AF seems to protect its pilots, saying they were monitoring weather and
that the first problem was loss of speed indicators.
Bea issued some clarifications on its May 27th release.
http://www.bea.aero/fr/enquetes/vol.af.447/vol.af.447.php
I have a feeling that the report avoids mentioning "thunderstorm"
because of ambulance chasing lawyers.
I suspect there are plenty of things that the report avoids
mentioning. The Accident places the BEA in a very difficult position.
The French always put commercial interests in front of everything
else. This accidents leaves them with a choice of blaming either a
French Airline/Crew, or a French Built Airplane. I think they will do
their best to avoid blaming either.
Post by JF Mezei
PF calls the sleeping captain that they are about to hit slightly
stronger turbulence and that he'll call back once they are through it.
They turn left by 12°. Obviously to avoid "something".
at 02:10:05 auto pilot disengages, and there is a stall warning,
followed by indicated airspeed drop from 275 down to 60 knots on primary
and a few seconds later on secondary (ISIS) indicators.
QUESTION: could this be caused by the wind in a thunderstorm which would
suddently be a strong tailwind, reducing IAS ? Or is the only
explanation pointing to faulty pitots ?
That involves a rather dramatic change in winds. The 'delta' is 215kt,
so even if the winds reversed completely, you still need a 107kt wind.
That isn't a thunderstorm, that is a tornado. The most likely cause of
such a dramatic airspeed change is ice in the Pitot tube, which is in
fact a known issue with the Thales tubes. It isn't the first time the
tubes have iced up. In fact Air France actually dispatched a
maintenace crew to replace AF447's Pitot tubes with Goodrich Tubes
upon arrival. Unfortunately AF447 never arrived.....

The reality is an AD for the Thales Pitot tubes should have been
issued long before the AF447 accident.
Post by JF Mezei
They increased pitch, started to climb at up to 7000 ft/min but pilot
then made nose down commands and got climb rate down to 700 ft/min at
altituide 37,500. To me, it looks like a controlled "climb over the storm".
Maximum curise altitude is generally a function of weight.
Post by JF Mezei
If the training is to always point nose down in a stall, and they
pointed nose up, it must have been because there was something below
that they really didn't want to get into.
We just don't know. The hazard in pointing the nose up is entering a
deep stall, and that is something that swept wing aircraft often have
difficulty getting out of. An ATP friend (and Airbus driver) commented
that once you enter a stall in large jet aircraft, you are going to
chew up 10-15,000 feet of altitude to recover. Realistically keeping
the nose up IS going to make things worse. There is very little actual
stall recovery training in large aircraft from what I am told, the
training revolves around making sure you don't actually stall the
aircraft in the first place.
Post by JF Mezei
What I don't know is whether the initial stall warnings were due to
faulty pitots, or whether they were due to wind/storm. Since the plane
was able to climb from fl 35 to fl 38, it was obviously not stalling.
My own belief is the crew was never able to sort out what instruments
were still working accurately, and those that were no longer providing
accurate information. They were inundated with warnings, and as a
result made bad decision based upon bad information.

However in retrospect, that obvious thing to do was use the simplest
instruments. Neither the artificial horizon or rate of climb
indicators need a computer to operate. When all else fails, what you
are supposed to do is set 85% power and nose up about 2.5 degrees.
That is almost always going to be safe while you figure out the rest.
Post by JF Mezei
Perhaps pilots had learned to disregard such stall warnings due to
faulty air speed indicators. And the first time the stall sounded, they
ignored it without "damage" but once they reached 38k feet, they ignored
a real one.
What is still not clear is why during the climb to 38kfeet, engines were
at max (TO/GA setting), but a couple minutes later, they were at iddle.
Actual power level setting on A320/A330/A340 can be confusing. Unlike
Boeing, when auto-throttle is engaged, the power levers do NOT move.
There is an auto throttle detent, and thats where the power levers
sit. So the position of the power levers isn't necessary a good
indication of actual power setting.
Post by JF Mezei
Seems to me that when the PNF declared they had dropped to 10k feet,
someone should have though about putting thrust to the max.
The report does note some rather large oscilations in banking by up to
40°.
Is it possible that the airspeed indicators do not function when the
angle of attack is too high ?
A number of indicators are known to become unreliable at very low
airpseed. 60 Kt is apparently the lower limit for many of them.Part of
the problem is the airspeed apparently became low enough that the
stall warning became inoperative. The Airbus relies upon an audio
warning (as opposed to a 'stick shaker' or 'stick pusher'). I.E. on
many aircraft, the control yoke will actually start to shake as you
approach stall, and ultimately actually push the yoke forward.

Hard to do with a sidestick controller.
Post by JF Mezei
This would explain why speed indicators
came back when aircraft put nose done inputs to reduce pitch/angle of
attack. So the aircraft would have been in a real stall because the
nose was too high up and engines too slow, but the computer not issued
stall warning because the airspeed indicators were considered invalid
when they could have been valid ? (at 35° angle of attack with engines
at IDLE, what sort of airspeed would perfectly working pitots record ?
The Pitot system is not really subject to influence by pitch angle .
What it measures is the difference between the air pressure in the
tube, and the static source at all times (when it works). That
pressure difference is used to calculate the indicated airspeed.
Post by JF Mezei
Based on what my buddy Mr Google has told me, the 330-200 has a service
celiling of 41k feet. Is it correct to state that such an aircraft
would need to have engines at far much than idle when cruising at that
altitude this early in flight (when it is still heavy with fuel) ?
Generally with the engines at flight idle, the aircraft can be
expected to descend at several thousand feet per minute, so by
definition if they were at cruise, the engines were operating well
above flight idle. Typical would probably be about 80%, however keep
in mind that 80% at 35,000 feet is vastly different than 80% at sea
level. Af 35,000 feet, you essentially have a turbo jet, there is
essentially no contribution to thrust from the bypass air. For
example a CMF56-3 engine that can produce 23,500 pounds thrust at sea
level, can only deliver about 5300 pounds thrust at cruise altitude,
that in turn means a substantially different fuel flow than at sea
level, and that's what you have fuel controllers on the engines for.



Service ceiling refers to the maximum useful altitude under some very
specific conditions, generally very low operating weight. The Service
Ceiling isn't the maximum altitude you can attain, it is just the
point at which the climb rates becomes unacceptably low. I would be
very surprised if the Service Ceiling at the weight they were at was
above FL380. Just because you can get there with a zoom climb,
doesn't mean you can stay there!
Post by JF Mezei
Does anyone have an explanation on why the engines would have been set
to IDLE after being at TO/GA ?
Is it possible that due to violent turbulence, they had to reduce
aircraft speed because turbulence was too violent ?
There generally is a maximum speed to penetrate turbulent air, however
it is rarely a lot lower than cruise speed. At typical cruise Altitude
the actual range of speeds the aircraft can sustain flight at is on
the order of 60kt. Faster than the upper limit of Indicated airspeed,
and you get shockwave formation and loss of lift, below the lower
limit, you get aerodynamic stall. So the answer is you can slow down,
but often not a lot.
Post by JF Mezei
They obviously need to release far more information. For instance, did
they advise the crew/passengers of upcoming turbulence that could be
significant ? Did passengers fly around the cabin or were they all
strapped in ?
Since they had been in bad weather for a while, it is a safe bet the
seat belt light was on.
Post by JF Mezei
Would any/all thunderstorms capable of this generate lightning ?
Lighthing should have been visible by pilots more than 2 minutes before
hitting it.
--
misc.travel.air-industry is a moderated newsgroup. Please mail messages to
***@airinfo.aero, and see http://mtai.airinfo.aero for the FAQ and policies.
Continue reading on narkive:
Loading...