[Info-vax] OT: Aircraft pitot tubes and clustering.

[JF Mezei]

FrankS wrote:

> Surface winds can vary tremendously from winds even a few thousand
> feet above the surface.  

So yes, winds make a huge difference. BUT, during flight, the systems
should know about both airspeed and ground speed because they also
calculate ETA at airport, and required fuel to destination. If you have
a tremendous headwind the pilots didn't know about, they may have to may
a pit stop to refuel because they wouldn't have enough fuel to make it
to destination.

So a computer should have sufficient information to question a changing
airspeed measurement that it still considers valid because 2 pitots are
exhibiting the same error.

Consider a case where a plane is flying through idle air. It's airspeed
would be about the same as ground speed. But all of a suddenm it gets
into a nice big tailwind. At that point, the aircraft's airspeed drops.
BUT, the aircrasft also starts to accelerate and the airspeed will begin
to rise again, as will groundspeed. Airspeed will return to "normal"

Similarly, when the aircraft leaves that area of tailwind, its airspeed
will increase significantly, and then decrease as the aircraft
physically decelerates without the favourable wind in its back.


So, if airspeed decreases to stall speed, but groundspeed does not
change, the computers could signal that there is an error, because a
change in wind should only cause temporary change in airspeed intil
aircraft accelerate/decelerate until airspeed becomes normal again.
(aka: with a certain amount of thrust, the aircraft will stay around the
same airspeed for its weight).

This is something computers could deal with. Whether aircraft computers
are programmed to look at trends over the last 5 minutes to do sanity
checks is the question. Perhaps they are programmed only to look at real
time data and react instantly to changes.