Air France plane lost over Atlantic

[bradl]

The real concern I have is that, personally I dislike the fly by wire system on the Airbus. I feel it takes too much direct control away from the pilot. For example, with the Airbus fly by wire system in normal law when you move the sidestick you are not actually commanding the flight control surfaces to deflect, you are commanding a change in load to the flight computer, and the computer figures out how to move the control surfaces. This differs from the approach of Boeing with their airplanes that utilize fly by wire where you still basically command the control surfaces to move directly through the yoke. It is possible to directly control the flight control surfaces of an Airbus in direct law, however that is the third level of redundancy built into the system and is designed to hopefully never be used (if it is, you're having a bad day).

Don't get me wrong, the Airbus a very nice airplane to fly most of the time. Very very easy. The thing has auto trim and all kinds of other helpful things. On a normal day these systems work fine, it's on a not normal day or not normal situation that I get concerned. I do not want a computer system sitting between what I want the airplane to do and what the airplane actually does especially when that computer system has the capacity to override the pilot. Boeing believes (correctly in my opinion) that the pilot should always have the final say, and as such can always override anything regarding the fly by wire. Airbus believes the opposite.

There are valid points made for and against the Airbus philosophy, you can see where I stand.

I don't like it when machines try to out think me. Same reason I will only ever own a car with a manual transmission.

Hope that answers your question.

Umm.. you do realize that the B787, B777, the entire EJets series (E170, E175, E190, E195, Lineage 1000), CSeries, the upcoming MRJs, and the next B737 replacement are all FBW, or will be FBW, right? If you are concerned about FBW, you'll either have to get over it, or start to take the train everywhere you can, because everything else is going to that.

Isn't the fly by wire system used to prevent a terrorist for example from changing the course of the plane? Sounds like a good security measure if you ask me.

No, not really. FBW isn't there to prevent any manual intervention on the pilot's part. The big difference is the means of the response between what the pilot is doing and how the aircraft responds to that use. Pre-FBW was done by hydraulics or some other means, whereas FBW is completely electrical. But even then, most operations, even on everything Boeing, is handled by FMC, especially with relation to RVSM and RNAV. There isn't much manual in flying passenger jets unless you've lost some major component.

BL.

 

[alphaone]

Umm.. you do realize that the B787, B777, the entire EJets series (E170, E175, E190, E195, Lineage 1000), CSeries, the upcoming MRJs, and the next B737 replacement are all FBW, or will be FBW, right? If you are concerned about FBW, you'll either have to get over it, or start to take the train everywhere you can, because everything else is going to that.

Perhaps I wasn't very clear but but I by no means have a problem with FBW itself, only Airbus' philosophy on how it should work. Simply put, Boeing (and now Embraer) believe that the pilot should have the final say always as to what the aircraft does, Airbus believes that the airplane should have the final say.

Here's a recent good article which briefly touches on the different FBW philosophies then goes in depth on Embraer's brand new FBW system (which on paper I like): http://www.flightglobal.com/blogs/flightblogger/2011/02/embraer-defines-hybrid-philoso.html

Also, most of the time FBW isn't completely electrical. Usually the way it works is that the inputs from the control stick/yoke (or in Airbus' case, the flight computers) get sent through wires to servos which in turn move the flight control surfaces though hydraulics. Wasn't sure if that was what you meant.

 

[Fuzzy14]

Umm.. you do realize that the B787, B777, the entire EJets series (E170, E175, E190, E195, Lineage 1000), CSeries, the upcoming MRJs, and the next B737 replacement are all FBW, or will be FBW, right? If you are concerned about FBW, you'll either have to get over it, or start to take the train everywhere you can, because everything else is going to that.

As [alphaone] explained earlier, Boeing fly by wire has a different operational philosophy to Airbus. Personally I think the Airbus philosophy is the correct one as we all know most incidents are caused by pilot error (or wrong interpretations of the situation by the pilot.) Remember it isn't Windows flying the plane!

It's astonishing how Quantas have been able to avoid any major fatalities when you take into consideration all the long haul flights travelling to the other side of the Earth and how far away it is from the plane manufacturers.

An unfortunate myth which disrespects the 63 people they've killed.

 

[aristobrat]

An unfortunate myth which disrespects the 63 people they've killed.

Not to disrespect those, but it is (IMO) astounding that they haven't had any fatalities since 1951. 60 years?

 

[Fuzzy14]

Yes, their safety record is exceptional, I just get fed up with the perpetuation of the Rainman myth. They haven't had a fatal accident in a jet since 1951, but due to Hollywood people don't know those details.

Nobody bangs on about British Airway's safety record if you don't include Zagreb (ATC error) and Manchester (technically not BA.)

 

[alphaone]

As he explained earlier, Boeing fly by wire has a different operational philosophy to Airbus. Personally I think the Airbus philosophy is the correct one as we all know most incidents are caused by pilot error (or wrong interpretations of the situation by the pilot.) Remember it isn't Windows flying the plane!

It is an interesting argument. What you said is exactly why Airbus has the FBW philosophy it does. For example, there was AA965 which crashed in Columbia in '95. Putting it very briefly, by inputting an incorrect navaid into the FMS the pilots got themselves off course in mountainous terrain. They had the spoilers fully extended to help them in the descent to the destination airport. Soon the GPWS activated alerting them of imminent impact with terrain ("too low, terrain", "whoop whoop pull up!") and the crew reacted by applying full power and pitching for best climb. However, the crew forgot that the spoilers were still extended which hampered their climb performance so that they could not clear the terrain, which they impacted. Cause is crew error, they absolutely should have recognized the configuration problem, but somewhat understandably in their panic they didn't. Maybe there should have been another warning system (bell or master caution or whatever) in place to alert them of the configuration issue.

Note: the 757 does not have FBW in the first place. In this case, Airbus would argue that their FBW would recognize that the spoilers were still extended and retract them automatically when full power was applied, thus probably averting the accident. This is a perfectly valid point in a perfect world.

The reason I'm opposed to Airbus' philosophy in giving the airplane final authority into its operation is as follows. I have a dialogue with the other pilot I am flying the airplane with. He/she understands why I am doing something. Maybe it's something non standard or unusual, but there may be a perfectly good reason for me to be doing it, it may even be crucial to the safe outcome of the flight (we're just talking hypotheticals here, I have no specific scenario to give). There is no such dialogue between the pilot flying and the flight computers in the FBW system. It does not know why, only how it is programmed to react to certain inputs (in which there also exists the potential for those inputs to be incorrect, same as for the pilot). If it starts overriding the pilot flying, to me that has a potential for danger.

 

[Fuzzy14]

It's a natural reaction for those with PPLs (I fly a Diamond Katana) to want compete control over the plane, I'd certainly not want a computer anywhere near a 2 seater.

But when I read the AAIB reports commercial planes seem to be more about managing (esp CRM) rather than flying. I guess when in trouble you revert to first principals, want the computer out of the way and fly by the seat of your pants.

 

[yg17]

So many accidents in the recent pass (both fatal and non fatal) have been caused by things like blatant pilot error, ATC error, terrorism, maintenance failure, or other forms of human error (I'm thinking ValuJet 592 here - that was a major human **** up) that I just don't think you have enough data to make the claim that Airbus's FBW system is better than Boeings or vice versa. If you could take out all non-flight computer factors that cause plane crashes, I think the accident rate would, pardon the expression, plummet. We can argue Airbus or Boeing until we're blue in the face, but at the end of the day, I feel safe getting on either one. After all, I have a much better chance of dying in a car accident on the way to the airport.

 

[rovex]

Yes, their safety record is exceptional, I just get fed up with the perpetuation of the Rainman myth. They haven't had a fatal accident in a jet since 1951, but due to Hollywood people don't know those details.

Nobody bangs on about British Airway's safety record if you don't include Zagreb (ATC error) and Manchester (technically not BA.)

Agree that BA's safety record is understated, however what about that crash from BOAC (if I'm not mistaken) years ago killing everyone on board in the ocean. I believe BOAC is BA as it is today.

 

[Huntn]

So many accidents in the recent pass (both fatal and non fatal) have been caused by things like blatant pilot error, ATC error, terrorism, maintenance failure, or other forms of human error (I'm thinking ValuJet 592 here - that was a major human **** up) that I just don't think you have enough data to make the claim that Airbus's FBW system is better than Boeings or vice versa. If you could take out all non-flight computer factors that cause plane crashes, I think the accident rate would, pardon the expression, plummet. We can argue Airbus or Boeing until we're blue in the face, but at the end of the day, I feel safe getting on either one. After all, I have a much better chance of dying in a car accident on the way to the airport.

I've not seen or read the final report about the Air France crash (if it has come out), but my impression is that a pitot static design flaw in combination with possible pilot error in dealing with this situation is the prevalent theory.

If you are familiar there has been a long history of pilots crashing planes with blocked pitot tubes due to things like indications of over speeding the airplane and stalling the airplane trying to deal with it. For this specific situation, identifying the issue is the hardest part. Once identified, it is a relatively easy problem to deal with (fly known attitudes and power settings).

 

[Kebabselector]

I believe BOAC is BA as it is today.

BOAC and BEA merged to be British Airways. (think the last BEA crash was in the 70's when a Trident crashed and killed everyone)

 

[0007776]

Looks like they released a bit more information. It sounds like the plane stalled and fell pretty fast, about 3 and a half minutes.

http://www.bea.aero/fr/enquetes/vol.af.447/point.enquete.af447.27mai2011.en.pdf

 

[iJohnHenry]

Looks like they released a bit more information. It sounds like the plane stalled and fell pretty fast, about 3 and a half minutes.

http://www.bea.aero/fr/enquetes/vol.af.447/point.enquete.af447.27mai2011.en.pdf

Ahem.

Now the question is, why??

 

[Fuzzy14]

They probably weren't aware that they had dropped below stall speed, probably had a stall warning on one system and an overspeed warning on another.

Normally the ultimate stall warning is the stick-shaker because this is a mechanical device linking the yoke to the horizontal stabiliser. Anybody know how this works on fly by wire?

 

[iJohnHenry]

Normally the ultimate stall warning is the stick-shaker because this is a mechanical device linking the yoke to the horizontal stabiliser. Anybody know how this works on fly by wire?

This is a good question.

I am guessing that nothing happens.

Perhaps they need a hydraulic system, to simulate the yoke warning.

 

[dmr727]

Normally the ultimate stall warning is the stick-shaker because this is a mechanical device linking the yoke to the horizontal stabiliser. Anybody know how this works on fly by wire?

This is a good question.

I am guessing that nothing happens.

Perhaps they need a hydraulic system, to simulate the yoke warning.

I'm not an Airbus guy, but I talked about this a bit with my wife, and she confirms there's no stick shaker - at least on the 319/320/321. I'd guess it's the same on the 330 as well. That said, there's a slew of other aural and visual warnings indicating the onset of a stall.

Understand that a typical stick shaker is nothing more than an off center motor attached to the yoke column, and is activated by angle of attack vanes attached to the fuselage. I wouldn't think such a system would be any different on a FBW aircraft. Since it's all AOA based (as it should be), problems with the pitot/static systems shouldn't affect it.

 

[iJohnHenry]

Understand that a typical stick shaker is nothing more than an off center motor attached to the yoke column, and is activated by angle of attack vanes attached to the fuselage. I wouldn't think such a system would be any different on a FBW aircraft. Since it's all AOA based (as it should be), problems with the pitot/static systems shouldn't affect it.

Thanks.

I thought it was more a matter of direct feedback from a control surface.

So it's "do-able", if the other warnings you mention are not enough?

 

[rovex]

So, was this more of a pilot error rather than a mechanical error? Should the
Pilots really have known when they dropped below stall speed even if the plane was giving all kinds of different indications?

 

[iJohnHenry]

So, was this more of a pilot error rather than a mechanical error? Should the Pilots really have known when they dropped below stall speed even if the plane was giving all kinds of different indications?

This is vastly unfair to the pilots.

They are trained to trust their instruments.

The problem lies with the design of the aircraft.

 

[Huntn]

Ahem.

Now the question is, why??

The Answer: Recognition is the key. You are flying along fat dumb and happy on the autopilot, one minute every thing is normal, the next second overspeed warnings are going off and if you look at the airpseed it shows the airplane suddenly flying in the red, over speeding the design limits of the airplane.

A quick reaction would be to pull the power back to idle, but even worse would be pulling back on the stick making the plane climb, bleeding off airpseed. The problem is that with a blocked pitot tube, the airspeed may stay in the red no matter how long you climb until you have stalled your aircraft and you are now plummeting out of the sky with full back stick, trying desperately to get an indication (over speed) to go away, and it's not going to go away as long as it's iced over.

The first response should be to reduce power, just in case somehow you actually did get fast, but if you normally cruise at 85% power and you are level and now have 50% power selected and there is no sign the airspeed is dropping off, it's time to get real suspicious real fast. Once you recognize this is a false indication, you should return to known attitudes and power settings regardless of what the airspeed says. I have not read the report but apparently they spent the last minutes of their lives with full back stick and falling fast.

I fly the Airbus 320 and although I'm not an expert on this issue I know enough. This situation has periodically been trained throughout my career flying on the 727, the DC9, but honestly I don't remember if much has been said about this in the last few years I've been on the 320. Several if not many airplanes have been lost over the years for this specific reason, pitot heat that was not turned on or had failed.\

Additionally, my understanding is that some of the European airline Airbuses had pitot static systems with known design flaws. U.S. airline Airbuses did not, or were changed early. If your pitot tube is working properly, it is heated at all times and should never ice over. New commercial aircraft don't have pitot tube heat on-off switches, they are always on. They have warnings if the pitot tube heat fails, but with a design flaw, the heat could be working, or the aircraft thinks it's working, and the tube still plugs with ice. I don't know specifically if this is what happened to Air France, but I think it is. This is why recognition of the problem is the hardest part. Once recognized, it is easy to deal with.

This is vastly unfair to the pilots.

They are trained to trust their instruments.

The problem lies with the design of the aircraft.

Normally this is not an issue. Even if an aircraft system is designed to warn the pilots there could be the rare circumstance where the situation exists, but a warning does not occur. In this case pilots must be trained to recognize the situation (obviously before it happens).

 

[bradl]

This is vastly unfair to the pilots.

They are trained to trust their instruments.

The problem lies with the design of the aircraft.

Hate to say it, but this logic is what got Buddy Holly, Richie Valens, and the Big Bopper killed. Their altimeter was off by 500ft, flying in the opposite direction of another aircraft. Trusting their instruments, they collided with that aircraft.

First and foremost, they see and avoid, especially in VFR. IFR, obviously, is another story.

BL.

 

[iJohnHenry]

Hate to say it, but this logic is what got Buddy Holly, Richie Valens, and the Big Bopper killed. Their altimeter was off by 500ft, flying in the opposite direction of another aircraft. Trusting their instruments, they collided with that aircraft.

First and foremost, they see and avoid, especially in VFR. IFR, obviously, is another story.

BL.

Great. I am aware of "the day the music died".

But how can you fault pilots drilled in the infallibility of their instruments??

My original statement stands. Look to the engineering of the aircraft.

 

[Huntn]

Hate to say it, but this logic is what got Buddy Holly, Richie Valens, and the Big Bopper killed. Their altimeter was off by 500ft, flying in the opposite direction of another aircraft. Trusting their instruments, they collided with that aircraft.

First and foremost, they see and avoid, especially in VFR. IFR, obviously, is another story.

BL.

Well, I'm not going to outright say you are wrong, but my understanding this (the Buddy Holly crash) was a one airplane accident that occurred either due to 1)flying in icing conditions 2)a pilot flying in IFR conditions (clouds/poor vis) unqualified to do so 3) an overweight airplane (which in light aircraft is very common) or 4) a combination of these factors. It does not take much ice on the wings of a small aircraft to take it down, although the CAB thinks it was an unqualified pilot in instrument conditions.

Buddy Holly Civil Aeronautics Board Report Text.

 

[Huntn]

One more thing about Airbuses, the ones I'm familiar with are the A319-A320. They do not have stick shakers, an independent system that warns of stall. What they do have are flight laws. When you fly an Airbus you, the pilot, are negotiating with the airplane's computers to control it. The computers will not allow you to take the aircraft beyond a predetermined limit. Under "normal" law, the aircraft will not allow itself to stall, not allow itself to be rolled upside down, however in the case of stalling, this is based on knowing what its real airspeed is. Now if it's getting bad info, the pilot is not going to get a stall warning. An independent stick shaker system, something Boeing used to use (I assume they still do) would be a good thing, but it is still going to take a pilot who might be faced with simultaneous over speed and stall warnings to figure out just what is happening.

 

[rovex]

One more thing about Airbuses, the ones I'm familiar with are the A319-A320. They do not have stick shakers, an independent system that warns of stall. What they do have are flight laws. When you fly an Airbus you, the pilot, are negotiating with the airplane's computers to control it. The computers will not allow you to take the aircraft beyond a predetermined limit. Under "normal" law, the aircraft will not allow itself to stall, not allow itself to be rolled upside down, however in the case of stalling, this is based on knowing what its real airspeed is. Now if it's getting bad info, the pilot is not going to get a stall warning. An independent stick shaker system, something Boeing used to use (I assume they still do) would be a good thing, but it is still going to take a pilot who might be faced with simultaneous over speed and stall warnings to figure out just what is happening.

Surely the clever thing to do was to follow the stall speed warning? I mean, what implications could that of had?

They knew their altitude so why would they follow the over speed indications?