Probers: Flight 447's Autopilot Was Off

(CBS/ AP)  Signals sent by Air France Flight 447 before it disappeared show its autopilot wasn't on, the head of the French agency leading the probe into the crash of 447 said Saturday.

Agency head Paul-Louis Arslanian said it was not clear if the autopilot had been switched off by the pilots or had stopped working because it received conflicting airspeed readings.

Plane manufacturer Airbus says the probe found the flight received inconsistent readings from different instruments as it struggled in a massive thunderstorm.

Alain Bouillard, head of the investigation into the crash, told reporters that, "we also saw messages that show the automatic pilot wasn't working."

An Air France memo to its pilots Friday about the crash of Flight 447 said the airline is replacing instruments that help measure airspeed on all its medium- and long-haul Airbus jets.

Investigators have focused on incorrect speed readings as one potential factor in the crash.

Arslanian said investigators are analyzing 24 messages sent automatically by the plane during the last minutes of the flight.

He said investigators are searching a zone of several hundred square miles for the debris.

It is vital to locate a beacon called a "pinger" that should be attached to the cockpit voice and data recorders, now presumed to be deep in the Atlantic, he said.

"We have no guarantee that the pinger is attached to the recorders," Arslanian said.

Holding up a pinger in the palm of his hand, he said: "This is what we are looking for in the middle of the Atlantic Ocean."

Investigators are trying to determine the location of the debris in the ocean based on the height and speed of the plane at the time the last message was received. Currents could also have scattered debris far along the ocean floor, he said.

"You see the complexity of the problem," he said.

Laurent Kerleguer, an engineer specialized in the ocean floor working with the investigation team, said the zone seen as the most likely site of the debris was 15,112 feet at its deepest point and 2,835 feet at its shallowest.

Water salinity and temperature can affect the distance that the beacon's signal can travel, Kerleguer said.

The Airbus A330 from Rio de Janeiro to Paris disappeared nearly four hours after takeoff on Sunday night, killing all 228 aboard. It was Air France's deadliest plane crash and the world's worst commercial air accident since 2001.

[gspencer4]

Old pilots told me that "There is a 6G differential somewhere in those big storms that will break the wings off of any airplane", and that I should not fly into any large storm when I was learning to fly.

The wings on Air Transport Aircraft were and are now designed then and now to withstand forces equal to about 4.5G (gravity) positive and about half of that in the negative before the wings fall off when the airplane is loaded to the maximum allowable gross load (fully loaded). Being less than fully loaded does lessen the wing loading and allows the aircraft to exceed these design limits before the wings are not structurally overloaded. Passing through a large severe thunderstorm with high velocity vertical winds in both up and down directions is dangerous.

Fly by wire is not a new technology. The F-16 has had "fly by wire" since it was developed in the 1970's, or was it the late 1960's, but the passengers had ejection seats that were used when "fly by wire" or other flight systems failed. The F-16 had several redundant electrical systems.

I remember an Airbus A330 piloted by an Airbus factory pilot that flew into the trees and crashed at some European air show years ago, and the final crash report blamed the "fly by wire" system. Maybe some cable system to control flight surfaces when there was an electrical failure would at least give the pilot and passengers a chance.

Aircraft have a maximum structural maneuvering speed that that you are supposed to slow down to before you start maneuvering like turning or entering a storm which will load the wings with additional G-forces. This is not the whole picture, but this might be a non-technical explanation for non-flying layman to understand.

The first Lear Jets had unauthorized "go-fast" switches to turn off the Autopilot and manually fly the airplane at faster than the redline indicated speed at high flight levels. This supposedly caused a few accidents.

[mecury69]

"Something inside the plane exploded, perhaps a bomb. Planes do not break apart due to turbulence, no matter how severe.
Posted by ontheleft at 1:12 PM : Jun 6, 2009"

No claim = no terrorist = no bomb.

Man continues to stick his middle finger at mother nature. Severe enough storms can tear apart a plane if the plane is not flown properly for whatever reason.

One of your ancesters must have also declared the Titanic as 'Unsinkable".

[cattlekate1]

Evidence is building that Air France has a faulty design in their air speed sensor. the autopilot is another clue. But nothing has been difinitive yet.
Posted by McHineguy at 11:15 AM : Jun 6, 2009

And what would cause the pitot tubes (all of them) to loose their ice-melting heat? What powers that heat? The fly-by-wire? If it was knocked out by lightning, all pitot's lose heat? No battery/little windmill generators kicking in fast enough?

Other questions - another crash of a 330 was due to speed indicators saying the speed was too fast - so the pilots lowered flaps, which cause the wing to shear off, and thus a dive and deaths of 75 plus people. Forgot the flight - will look up. Also can't remember the altitude. But McH - do you think that could be the cause? If a wing or part of a wing breaks off, then that would cause a dive which causes the depressurization?

[skydrifter1]

It's a very real possibility for turbulence to stall an aircraft; still, that is recoverable from 35,000 feet.

BUT - there is a deafening silence in the lack of pilot reports from the aircraft ahead, or a request for PIREPS from those flights. How bad is the line of thunderstorms likely to have been.

Hand-flying an airliner at 35,000 feet is a major task. There would have been at least two autopilot systems - why would a pilot choose to kick off the autopilot? It's far more likely that some excess parameter tripped off the autopilot, with the parameter so great that the autopilot couldn't be re-engaged. If the autopilot wasn't controlling the aircraft, hand-flying would have been the remaining choice. BUT, what happened to get to that degree, in an obviously rapid and catastrophic set of events.

With the background of the NYC Airbus crash, wherein the vertical fin & rudder came off - add both engines - I'm curious as to the probability of a design problem, leading to structural failure. While the aircraft are different design series, is there an inherent problem with the general Airbus engineering?

It will take a long time to investigate this problem. The odds of an honest investigation are questionable, for sure.

[McHineguy]

Something inside the plane exploded, perhaps a bomb. Planes do not break apart due to turbulence, no matter how severe.
Posted by ontheleft at 1:12 PM : Jun 6, 2009

It hasnt yet been established that the plane broke apart before falling out of the sky. It may have stalled due to the weather/air sensor, then during the fall weather and pilot actions tore a wing surface. Or, as you seem to be rushing to, a smallish bomb. But lets wait until we know something.

[ontheleft]

Something inside the plane exploded, perhaps a bomb. Planes do not break apart due to turbulence, no matter how severe.

[ajjaxtheleast]

If the one where the plane was half destroyed by the weather
then totlly destroyed by a bomb hasn't been taken yet,,,
I'll go with that one.

[McHineguy]

Here's the problem that I have.....The planes radar system, all indications show this was working prior to this, so if that's the case, why didn't the pilot divert and choose another fligt path? The radar shoud have revealed these storms that reached 60,000 feet and also revealed 100mph winds. Why would any pilot try and fly into that?
Posted by steelcity7989 at 8:22 AM : Jun 6, 2009

Storms are not one large mass. Often there are clear paths through the storm by flying between separate portions. Pilots often use their radar to plot these paths when it is impossible to go around the entire mass. BUT, a storm has been known to "close in behind you and in front at nearly the same time and leave you trapped between.

To compound this, radar will usually identify the intensity of different portions so the pilot can plot a course through the least troubling.

So, the pilot is "winding his way" through the storm using radar and his own eyes to show the least intense portions.

But, nothing is perfect. A huge, white cloud ahead may look very peaceful so you fly towards it as the least turbulent area. But, behind it, hidden by it, is a huge ugly thunderhead that is undetectable until you are too close to turn away.

[McHineguy]

It would have been nice if the article would have told us the significance of the autopilot being turned off or on. I suspect that it would have been normal to have autopilot off when approaching a storm.

So why is this clue important ?
Posted by hoseobama at 7:59 AM : Jun 6, 2009

Evidence is building that Air France has a faulty design in their air speed sensor. the autopilot is another clue. But nothing has been difinitive yet.

[McHineguy]

This is all baseless speculation. Really.

Autopilot off? Airspeed discrepancies? Um, Duh. The plane was being hammered and it was in the process of coming apart. Were there any airspeed discrepancies before they hit the weather? Did the autopilot disconnect before they hit the weather? Answers? No.

The media is reporting total BS because they have nothing else.
Posted by jroach31 at 8:55 AM : Jun 6, 2009

Actually, the speculation is pretty good if you watch carefully.

1. The air speed sensors are already being replaced in other Air France aircraft. No manufacturing time required? Obviously, they have known for some time the sensors were not serving well.
2. Digital data transmissions covering about 4 to 10 minutes (they havent been forthright about how much time) show the sensor discrepancies began several minutes before the crash.
3, As the engineers read the data that was transmitted they create logic diagrams that are reviewed by experts. After all experts have collaborated, they are able to give pretty accurate specualtion as to why a particular system has taken a particular action.
4. The autopilot may have been turned off by the pilots so they could handle the rough weather (like turning off cruise control when you drive on ice). The autopilot may have shut down because other readings (air speed sensors) are misleading. But there are at least two separate autopilots each with extensive self checking, each on different power supplies. If both shut down at the same time it means something caused it.
5. If the data transmission is still operating, the autopilot is also able to operate. The fact that you have transmitted data means the aircraft has not yet broken into pieces.