French Jet Vanishes on Flight From Brazil to Paris

Recent reports suggest Airbus may suspect the pitot tube airspeed measuring systems as a possibly related factor in the crash. It is likely that their recent acceleration of pitot system replacements is a precaution, based on this possibility.

Certainly abrupt changes in pressure altitude or indicated airspeed as measured by these analogue devices could cause excursions in the outputs of numerous computer systems - including that controlling the aircraft autopilot - that use this data as a vital input.

Though I don't know if it was a factor, one of the important principles of flight in my experience was not to use autopilots when encountering turbulent conditions, particularly in an "alktitude hold" mode which commands the autopilot to maintain a constant altitude. Better to fly the aircraft at nearly constant angle of attack and ride out the excursions.

Modern fly-by-wire aircraft (electrically powered flight control systems) are easily linked to increasingly complex autopilots that incorporate stability augmentation features that make the aircraft much more aerodynamically efficient under a broad range of flight conditions. The down side of this is that they can be more vulnerable to catastrophic failure modes in some conditions.

I don't know that these considerations were factors in the crash, but it appears they might be.

When the airliner first disappeared from radar, the tv heads were "analyzing it" . One head was talking about how it was impossible for a pilot to get the AP to relinquish control to a human.
In the Airbus in question, do you know whether its difficult to take over from AP in such weather conditions as what it apparently experiencing

All the aircraft I flew and all I know of had/have redundant means of quickly disengaging the autopilot (or AFCS = automatic flight control system in jargon). We routinely disengaged it in any mode but clear air level flight.

Modern transport aircraft, necessarily constrained by the economics of fuel consumption, increasingly invoilve the far more extensive use of ever-more capable and programmable autopilots. For example it is relatively simple to use a computer input to program the optimum fuel consumption profile for whatever air traffic control constraints might be applied (i.e. best angle of attack, meaning you slow down as the aircraft burns fuel and gets lighter.)

The best and safest technique for penetrating an area of heavy turbulence is to slow down a bit from the best efficiency speed and fly at a constant angle of attack (i.e. pitch attitude), riding the currents up and down as you go. That generally involves manual operation of the flight controls.

Much speculation is afoot regarding the possibility of ice accumulation in the Pitot tube, which could have reduced the measured ram air pressure (=Vsquared/2g) relative to measured static pressure, yielding an erroneously reduced indicated airspeed. This is a primary flight instrument both for the pilot and the AFCS computers, however large transport aircraft have multiple pitit tubes & backups. There are other backup sources including angle of attack and GPS-derived groundspeed instrtuments, but, since indicated airspeed is much less than groundspeed at high altitude , and many civil pilots are unused to closely monitoring angle of attack, a number of subtle and insidious scenarios can easily be envisioned.

Thanks for your takes, George.. they're helpful.

Indeed they are.

That seems like a lot of stuff to just randomly fail in a short period of time. And if cabin pressure was lost, does that mean that the pressurization system stopped working, or that they actually lost cabin pressure suddenly?

Maybe the cause of the problem was catastrophic damage to some portion of the plane, which then resulted in a series of error messages from the onboard systems?

The failures probably weren't random at all, and, instead probably reflected the rapid sequence of events leading to the catastrophy that occurred while the automatic data link was still working. Very likely the cabin pressure loss occurred when the fuselage fractured and the aircraft broke apart. The time interval (if it is known) between the autopilot disengagement and the rest of the reported events will tell the investigators a lot about what happened and what might have been the likely ubderlying cause, if not the initiating event. If the g force ecperienced by the aircraft is part of the data stream they will know even more.

You're implying that it broke apart in flight, giving the electronics enough time to report their cascade of failures?

Perhaps the issue here is that I don't know what you are implying.

My understanding from the news that has been reported is that the data link reported an autopilot disengagement, followed later by the reports of loss of cabin pressure and of electrical power. It is not clear from the news reports how much time elapsed between those reports, but the context so far suggests that the last two (loss of cabin pressure and loss of electrical power) came close together or nearly simultaneously. That implies that the aircraft broke up in flight due to some cause - very likely due to the turbulence it was encountering, but even that is uncertain.

In my experience all bets are off in a large, growing thunderstorm. I've gone through them and encountered nothing but dark clouds & rain: I've also seen icing and violent turbulence sometimes involving huge up and downdrafts that moved me up or down several thousand feet. Once I came out of one over the Florida panhandle with all the paint on the leading edges of the wings & several feet back peened off by ice abrasion - after less than a minute in the storm.

Such thunderstorms grow rapidly and generate huge vertical circular (or elliptical) currents of air. Much depends on exactly where the aircraft is relative to these rather unpredictable currents. The appearance of the clouds is not a particularly reliable indicator of the violence within.

I saw that they found several chunks of metal {floating?} and several bodies today. Maybe the data recorders will be found after all.

I'm not trying to imply anything, just trying to make sense of the data. And I agree that the info we're getting is missing a lot of detail, particularly the timeline.

I was just surprised to hear that they had such a cascade of system errors which were reported. Originally I thought that just one malfunction (in the flight speed indicators) might have led to an out-of-control plane which crashed into the ocean. But that wouldn't explain a sequence of systems failures.

Then the loss of cabin pressure was a surprise to me because that might indicate significant structural damage at a high altitude, something which is not explained by simply losing control during a storm. Storms may cause a plane to crash, but they don't "break" planes in flight no matter how strong they are. Commercial airliners are extremely durable.

I guess the data I'm reading so far just doesn't paint a very clear picture of what likely happened.

Ailerons, Flaps & tail control surfaces are generally hollow or foam-filled & will float for days. Some fuselage & wing sections are also built of honeycombed aluminum or carbon filament and will float for a while.

Autopilots have several modes of operation and in modern aircraft can be connected to engine throttle control systems. The systems can be operated to hold altitude, airspeed or both. However, these modes of operation are positively dangerous in turbulent air or thunderstorms where it is more important to limit g force and airspeed excursions by maintaining a constant wing angle of attack. I doubt that the pilot would intentionally enter a thunderstorm in either mode of operation, but the evidence remains to be found.

If the pilot became disoriented by a combination of turbulence and an incorrect (low) airspeed indication and added power & pushed the nose over to recover his "lost" airspeed, or if the autopilot did the same owing to an erroneous data input, the aircraft could end up in a steep dive which, with the clouds & turbulence could well have ended up in an overstressed condition on the wings or fuselage as the pilot struggled to recover. Alternatively if, in a strong downdraft, they raised the nose excessively to hold altitude the aircraft could have stalled violently. Either situation in a violent thunderstorm could easily overstress the aircraft structure.

Alternatively a cabin window could have failed causing a rapid decompression and in the ensuing confusion (one has about 25 seconds of consciousness in a high altitude decompression) they could have lost control. There are lots of possibilities. We don't know which: but we do know the aircraft crashed.

Tropical oceanic thunderstorms can be very violent at altitude: when you are in one it is quite easy to imagine a catastrophic structural failure.

It may seem so, but I believe the specs on these planes rate them as being extremely difficult to damage without hitting something (assuming they are build correctly of course). Even lightening strikes don't usually damage them (unless they ignite the fuel or something).

I look forward to seeing the timeline of events reported by the equipment. Maybe that will clarify things.

you believe so?

I believe that it is very unlikely that the storm alone, or any degree of out-of-control flying would crack the fuselage sufficiently to cause depressurization at high altitude.

But if the depressurization reading was just a false message created by some other damage, then it won't matter.

I really can't believe why anyone is refusing to take seriously my contention that we are facing just another incident of Bermuda Triangleism phenomenon.

I don't think the triangle has an effect on french folks.

mebbe it's all that cheese they eat, a density thing...

Just for the hell of it, it is worth noting that the mortality on this tragic accident was much greater than occurred at (or after as a result of) the nuclear reactor accident at Chernobyl. Do you think we will as a result stop flying airplanes? What explains the difference in our popular reactions?