Skip to comments.Air France Disaster – Focus On Airbus Air Data Inertial Reference Unit (ADIRU) System
Posted on 06/03/2009 9:12:38 AM PDT by luckybogey
The Air France disaster should be watched very closely as there are numerous political implications in the aircraft industry. The Australian is reporting investigators may be looking at the Airbus Air Data Inertial Reference Unit (ADIRU) System that sent a Qantas A330 on a wild ride over Western Australia last year.
The Qantas incident last October, and another in December last year also involving an Airbus 330 near Western Australia, involved a problem with a unit called an air data inertial reference unit, which prompted flight control computers to twice pitch down the nose of one of the jets.
Fast action by the crew limited the extent of the plane's fall but 14 people were seriously injured.
The incidents raised questions about a potential wider problem with ADIRUs, which collect raw data on parameters such as air speed, altitude and angle of attack, and process the information before sending it to flight computers. They led to European authorities issuing a global alert to A330 operators.
After the Air France disaster, The Seattle Times reported yesterday that experts were already examining these malfunctions...
(Excerpt) Read more at theaustralian.news.com.au ...
I wonder how much computer data the A330 can burst to Air France as the disaster was happening. I think it can send data via satellite.
Some A330 pilots hashing this around would be very interesting.
Nobody was killed during the first two incidents...now some were. Maybe they’ll do something about it now...
At 35,000 feet the pilots would have time to recover from a nose pitch down, wouldn’t they? They can override?
Aviation Ping List
Air France Flight 447:
A detailed meteorological analysis
Gee, I think I’ll stick with Boeing.
You're assuming that the problem really is with the ADIRU. However, there's an easier explanation based on the storms in the area at the time:
A former A330 pilot said yesterday that strong turbulence mentioned by the Air France captain in his last message could have been responsible for the aircraft breaking up if load limits on the plane had been exceeded.
"When you're going through the edge of the (storm) cell, you can sometimes go from a sudden downdraft to a sudden updraft," he said.
It seems more likely than a one-off failure that's never been seen before. (Note that the Qantas failure was on a different unit made by a different manufacturer.)
Air France Bomb Threat Before Flight 447 Crash
My understanding is the plane’s data was being provided back to Air France data center including the malfunctioning systems up to the time of the event.
I’ve been sailing on the ocean (done a few major trans ocean passages) and I have vast respect for “mere” thunderstorms. I’ve been hit by line squalls with 60+mph wind at the deck that went from zero to sixty plus in bare seconds, driving white water in front. I assume this is the result of a microburst hitting the ocean surface, or something similar. Hail, twisting shifting wind blasts, etc.
The winds at 35,000’ must be even worse. What can a pilot do if confronted with a line of storms hundreds of miles across, and over 50K’ high? Look for a “sucker hole?” Go hundreds of miles around?
If flying straight into one of those walls of storms, I would not rule out ANYTHING, including flying directly into up and down drafts of over 100 knots, accompanied by golfball sized hail and lightning.
I think that sometimes, modern transoceanic jets are just going to fly into something that can destroy anything with wings.
Just an ocean sailors opinion. At those times that I’ve experienced this type of “out of nowhere” storm fury from “mere” thunderstorms, I’ve been mighty glad I was in a 48’ steel-hulled vessel, and not up in the sky in an aluminum or carbon fiber tube.
God bless the pilots who must wrestle with such monsters. They can kill you, don’t doubt it.
An Airbus mechanic/engineer called into Dennis Prager yesterday, and this “override” question came up. It seems that fly-by-wire Airbuses only have a limited ability to override. The computers really do fly the plane. THey depend on great robustness and redundancy, but in the end, the computer will do what it thinks it must, regardless of pilot input.
At least, that’s what I heard the guy tell Dennis Prager.
On the aircraft I’m most familiar with (primarily business-jets), it’s a regulatory requirement that the pilot be able to physically overpower the autopilot and its associated actuators in the case it runs away and tries to go hard-to-the-stops in any particular direction.
The flight test to demonstrate this capability involves:
1. Warn the pilots (who are expecting it; it’s a planned test flight) ... Three ... Two ... One ... Now!
2. Pilots wait two seconds to simulate how long it would take to recognize the hardover were it was unanticipated.
3. Pilots override controls to recover.
The problem the AF447 guys may have run into is that they had a hardover, but due to the turbulence didn’t recognize it until they were significantly diverted from straight-and-level. If at the same time they lost instruments, they wouldn’t have known what to do to recover even if they did recognize it.
For the best analysis of the possible cause of the incident, please go to http://www.weathergraphics.com/tim/af447/.
Superb analysis with detailed information.
I highly recommend.
At least, thats what I heard the guy tell Dennis Prager.
The override needs teleprompters.
Give me Boeing or I ain’t going.
"the planes involved in the Australian incidents were the bigger A330-300s, rather than the 200 series involved in the Air France disaster, and it is understood their ADIRUs are made by a different manufacturer"
I heard the same thing but I wondered how much data the system can send. I guess a human looking at the gauges and the weather would tell them even more. Sad.