Qantas computer glitch?

[Moonwalker]

Limit protection and fly-by-wire are not the same. As far as I remember, there is a clear override switch on the sidestick, which is not only for the autopilot, but also for limit override.

Fly-by-wire is essentially just the control of the hydraulic actuators by electrical systems and state of the art in aircraft design today. It is true that you still need a minimal computer in the loop, but you don't need more than proportional deflection in the worst case.

Limit protection applies, when you have rate command or AOA command control loops.

Well, fly-by-wire was just the control of the hydraulic actuators by electrical systems on the old Concorde for example, which was the first civil fly-by-wire and thrust-by-wire airplane by the way. But it was simple compared to modern Airbus aircraft. Airbus made its flight automation and flight envelope protection (which didn't exist on Concorde) an almost inseparable part of their fly-by-wire system (except on the old A300 and A310 series which didn't use fly-by-wire) as long as the flight control computers work in normal law. The only exception is the alternate law / direct law or in other words when multiple failures of redunant systems occur.

Regarding the sidestick:

there are two switches available on it. A red distinctive thumb button on top and a trigger on the backside used with the forefinger. The trigger simply is a push-to-talk switch used for the radio communication. The red distinctive thumb button on top is used to disable the autopilot.

But, indeed, the red thumb button also has an override function:

the two sidesticks are not mechanically interconnected. I bet that most people never thought about what would happen if the pilot pushes his sidestick to the left while the first officer pushes his sidestick to the right at the same time. Well, in this case the effect is zero but the distinctive thumb button can be used to take control of the sidestick on which the button is pressed. In other words: the fastest thumb wins, either the pilot or the first officer. Pressing the thumb button for 30 seconds deactivates the other sidestick. A single push on the other thumb button activates its sidestick again. That's the override function. It is not related to the flight automation and the flight envelop protection which can be considered as an essential part of the Airbus fly-by-wire system.

Those fly-by-wire dual-sidesticks, not interconnected, are critized by some pilots because actually one pilot should feel on his sidestick what the other does. This was still the case on Concorde contrary to Airbus aircraft for example. The control columns were still mechanically interconnected. But on modern Airbus aircraft the pilots influence to the control surfaces have changed significantly. While in normal law the flight automation and envelop protection is active all the time, in alternate law or direct law, which means in abnormal conditions, the fly-by-wire is "reduced" again to a minimum which corresponds to conventional functions of control surfaces, providing no flight automation and flight envelop protection anymore. The question then is if the pilots are able to fly the airplane in such a reduced law while they used to fly the airplane on its full fly-by-wire capabilities for years. Because in such a case the pilots have to use the rudder pedals to coordinate a turn fully manually again while the bank and pitch angels are not maintained automatically anymore as soon as the sidestick is released for example. And also the pilots have no "feel" on their sidesticks. However, this issue is addressed during flight training anyway.

All in all Airbus are rather reliable aircraft. It's just a question of certain circumstances like the last Qantas incident showed. But anyway such events are more than minor and still unlikely. But it's interesting why one of the ADIRU's failed that strange way recently.

 

[Urwumpe]

Moonwalker: I remember that there is more to the override button than this, after the accident of one of the first Airbus A320, as one theory of the crash contributed it to the limit protection.

 

[Moonwalker]

Moonwalker: I remember that there is more to the override button than this, after the accident of one of the first Airbus A320, as one theory of the crash contributed it to the limit protection.

Yes I remember the accident and the resulting discussions. But as far as I know the investigation concluded that it was caused by pilot errors.

The override pushbutton still has the two functions I descriped, or lets say as the A330 flight crew operating manual descripes (but also all the others, from A318 to A321 and A340 which I have). If there is a possibility to just override the flight envelope protection, the Qantas crew would have used that I think. But the flight control computers didn't even change over to abnormal conditions because of the faulty ADIRU. They still reacted to the false data. What happened was an auto pilot disconnect follwed by a pitch command of the flight envelope protection and another one later.

But of course you can switch off the flight control computers, which is not a good idea (the reversers become inoperative, the speedbrakes automatically retract in case they are deployed, the rudder trim also becomes inoperative and more things happen). But if you just have a "protection lost" situation, alternate or direct law, the only thing which doesn't work is the flight envelope protection and/or the flight automation. However, you can't select this manually. It's an abnormal condition.

Found a nice video of the side stick...

[ame="http://de.youtube.com/watch?v=LRilx-a3Hro"]http://de.youtube.com/watch?v=LRilx-a3Hro[/ame]

 

[to be]

We appear to have moved on to the 'blame the military' stage: http://www.dailymail.co.uk/news/wor...antas-passenger-jet-plunge-650ft-seconds.html

 

[Moonwalker]

We appear to have moved on to the 'blame the military' stage: http://www.dailymail.co.uk/news/wor...antas-passenger-jet-plunge-650ft-seconds.html

It is rather important to look for any potential causes. But I still tend to "believe" that the cause was somewhere within the corresponding systems aboard rather than on the ground or caused by a small passenger laptop, which seems even ridiculous to me. A lot of Airbus pilots use laptops theirselves in the cockpit (on the small tables which replaced the control columns) as well as even mobile phones.

But I remember a failure of my fathers Mercedes car a few years ago, caused by transmitter masts used for mobile phones each time he passed them. The motor stopped, there was a warning triangle instead of the usual digital readings and a warning sound. He had to pull out the key and wait for a few minutes on the shoulder until everything worked normally again. The failure was fixed by Mercedes of course.

 

[tblaxland]

The override pushbutton still has the two functions I descriped, or lets say as the A330 flight crew operating manual descripes (but also all the others, from A318 to A321 and A340 which I have). If there is a possibility to just override the flight envelope protection, the Qantas crew would have used that I think. But the flight control computers didn't even change over to abnormal conditions because of the faulty ADIRU. They still reacted to the false data. What happened was an auto pilot disconnect follwed by a pitch command of the flight envelope protection and another one later.

Thanks for all the great info, Moonwalker and Urwumpe.

Just to clarify, if the flight envelope protection commands a pitch down to avoid a stall (whether it is due to faulty data or not), the pilot can't override this by pulling back on the stick and/or pressing a button/switch? If not, how did the pilots pull out of the dive?


-----Posted Added-----

Finally a quick fact. According to the Australian broadcasting code of practice, current affairs shows (ACA, Today Tonight), etc) do not have to present a balanced argument, and they can legally make up the entire story if they so desire.
Sucks don't it?

The journalism of "A Current Affair" and "Today Tonight" may not be high quality, but what you stated above is not a fact:

Commercial Television Industry Code of Practice - 2 July 2004 (amended 6 September 2007)

News and Current Affairs Programs
4.3. In broadcasting news and current affairs programs, licensees:
4.3.1. must present factual material accurately and represent viewpoints fairly, having regard to the circumstances at the time of preparing and broadcasting the program;

 

[Moonwalker]

Just to clarify, if the flight envelope protection commands a pitch down to avoid a stall (whether it is due to faulty data or not), the pilot can't override this by pulling back on the stick and/or pressing a button/switch? If not, how did the pilots pull out of the dive?

Usually the flight envelope protection assures that the aircraft does not enter a critical situation, it reacts before already which prevents the situation within a very few seconds. The flight envelope protection does a rather good and safe job on that. Of course I was not aboard the Qantas A330 but I guess the pitch command was rather short and the flying pilot was able to end that pitch pretty soon again (a few seconds). Such a pitch doesn't take too long as the aircraft usually gains enough speed pretty soon. Once the commanded target pitch has been reached (and the speed is out of a risky condition again) you can pull the nose back again.

Beside the pitch command during a potential stall condition, the thrust is also commanded for take off / go-around thrust (TOGA) in reaction of the speed loss. To stop that, the pilot has to manually move the levers into the TOGA position (to avoid a thrust difference when the thrust is in manual mode again), disengage the auto throttle system, moving the lever back to the CL position (which is the automatic mode for climb and cruise thrust) and re-engage the auto throttle system again. Other than on Boeing aircraft, the Airbus throttles don't move on their own in automatic mode. You have to select a mode which then controls the thrust digitally without any throttle movement.

On Boeing aircraft you don't have such a flight envelope protection but you have a stick shaker which starts to shake the control columns shortly before the aircraft is going to enter a stall condition. Such a condition then is stopped manually by pushing the stick forward (in case there is a stall already) and set maximum thrust. That takes just a very few seconds too. During test flights before each single aircraft delivery this is even done intentionally (usually by the fleet chief of the corresponding customer airline who does the test flight before the aircraft is bought).

[ame="http://de.youtube.com/watch?v=3PfBDYuXK-g&feature=related"]YouTube - Boeing 777 Stick Shaker[/ame]

 

[tblaxland]

Thanks!