Chronology Current Month Current Thread Current Date
[Year List] [Month List (current year)] [Date Index] [Thread Index] [Thread Prev] [Thread Next] [Date Prev] [Date Next]

Re: [Phys-l] Air France 447

On 06/11/2009 07:30 AM, Sam Sampere forwarded on Behalf Of

I am a bit puzzled by the Air France crash reports.

It's good to be skeptical. You should be extreeeemely skeptical
of initial reports in this business. Shortly after the crash of
TWA 800, news crews reported seeing "survivors in life rafts".
There were also "eyewitness" reports from those who "heard the
noise and looked up in time to see a missile hitting the aircraft".
(Hint: think about the speed-of-sound issues.) More recently,
there were reports of debris "confirmed" to be from Air France
447 that turned out to be nothing of the sort, just sea trash
picked up hundreds of miles from the actual crash site.

Even more specifically, this morning there was a press release
from the CEO of Air France, pointing out that it is too early
to assume the sensors caused the crash.

It is being
suggested that frozen over pitot tubes caused erratic air speed

I don't know about "erratic". The term being bandied about is
"inconsistent" ... which would make more sense, since the A330
has three Pitot tubes, and you can imagine scenarios where one
was frozen over while another was not. In my experience, once
they get frozen over they tend to stay frozen over for a while.

And yes, I do have some experience with this. Right after
getting my instrument rating, while the ink was still wet on
my new pilot certificate, I went on a night cross-country
trip. I filed IFR just for fun, even though the weather was
forecast to be clear with unlimited visibility. I could see
6th magnitude stars when I took off. A few miles later I
suddenly noticed that I couldn't see either wingtip. At
this point I switched on Pitot heat. A half second later,
my indicated airspeed went to zero. It stayed that was for
at least five minutes, maybe ten. I forget the exact details,
but it was long enough to have a long conversation with ATC
about it.

Since then I tell everyone to turn on Pitot heat whenever the
outside air temperature is below zero ... clouds or no clouds.

leading to loss of control. Wouldn't the pilots have
noticed this

Yes and no. Sometimes a frozen Pitot indicates zero airspeed, but
sometimes it just "locks in" a previous sample of dynamic pressure.
To a first approximation, this gives an indicated airspeed that is
independent of actual airspeed ... but it's even worse than that,
because (assuming the static port is not frozen) the altimeter will
now be acting as an _altimeter_, i.e. if you increase altitude the
bogus airspeed indication will also increase.

Also, in normal cruise, pilots don't necessarily spend a lot of time
looking at the airspeed indicator, even if they are flying by hand,
and especially not if they are using the autopilot. So it is possible
for things to get significantly messed up before anybody notices.

There are reports of a "maintenance" message received via a satellite
link indicating that the autopilot had disengaged. When it disengages,
the autopilot does not necessarily give the pilots an explanation of
_why_ it disengaged. So there are scenarios where a rather bad
situation is dumped on the pilots with little or no warning.

I'm emphatically not saying this is what happened; it is just a
hypothetical scenario that ought to be considered along with many
other scenarios.

and used their GPS equipment to get their ground speed
and use it as a reference based on the difference between their last
good airspeed measurement and ground speed? Is it that ground speed
and air speed can be wildly different around thunderstorms?

Even not near thunderstorms, groundspeed is very different from
airspeed. There is always some headwind, tailwind, and/or crosswind
... enough to make GPS useless for precise airspeed control. And of
course thunderstorms make all your problems much, much worse.

Also keep in mind that airspeed is not the only variable. Angle of
attack is also important. In most general-aviation aircraft, the pilot
is expected to infer the angle of attack from airspeed and other cues,
but a fancy thing like an A330 measures angle of attack explicitly,
and places explicit limits on it. So the flight envelope is limited
by max airspeed at one end, and by max angle of attack at the other end.
especially the section on "envelope protection".

If the angle of attack sensors iced up, instead of (or in addition to)
the Pitot tubes, bad things would happen.

Also bear in mind that the A330 is strictly fly-by-wire. The computer
is very opinionated, in the sense that it simply will not permit the
pilots to exceed certain limits. That is a good thing in 99+ percent
of the cases, but you can imagine that if the limits are being
miscalculated, the pilots might wish they had some way to override
the computer.