OT: If it Ain't Boeing, I Ain't Going

[Bullethead 12]

I was just reading the March 2012 issue of Popular Mechanics, which has a partial transcript of the cockpit voice recording from Air France 447, which crashed in the Atlantic in 2009. The full transcript apparently has been available in France since late last year but this is the 1st I'd heard of it.

 

Anyway, the root cause of the crash was that one of the 2 copilots pulled the stick back all the way, stalled the plane, then kept the stick all the way back. Thus, the plane remained stalled and mushed all the way down from 38,000 feet into the ocean, belly-first. A contributing cause seems to have been the nature of the Airbus flight control system, in the way it handles problems, and the way pilots are trained and become habituated to using that system.

 

The interesting thing to me is that, prior to the black boxes being recovered, folks had speculated that the loss of airspeed indicators due to icing had been the main cause. This did, in fact, happen, but the avionics thawed out long before the crash, but the crew didn't know how to get themselves out of the mess anyway.

 

So if it ain't Boeing, I ain't going.

Edited February 17, 2012 by Bullethead

[FastCargo 412]

I could regale you with a few stories of Boeing software not behaving correctly either...at least one of which has caused a crash. So don't generalize a specific brand based on one crash incident.

 

Now, what you can generalize is a training fault. Up until a little while ago, stall recovery was based on the idea of 'powering' out of the stall. The idea is that modern airliners have tons of power, and that typical stall situations were close to the ground. Therefore, the idea was to lose as little altitude as possible in a stall recovery. Most airlines were training to this standard...using the aircraft's power to gain airspeed back.

 

This accident shows the fallacy of that recovery technique. The stall was so deep, that the engines would never power the aircraft out of the stall...it would simply fall like a leaf (a very heavy leaf) into the water. Part of the reason was the training in only lowering the nose enough to accelerate, not enough to fully break the stall. This accident (along with others) has finally started shifting the mentality back to what we all learned in basic flying...break the stall by lowering the AOA, get some flying airspeed back, then pull back up...especially if you have buttloads of altitude below you.

 

FC

[Capitaine Vengeur 263]

If it is Tupolev or Ilyushin, I ain't. Definetely.

[Slartibartfast 153]

Must admit the plane that I worried about is the 737... and as FC pointed out is this is a training issue... Personally I will not fly on a plane unless there is Duct tape showing somewhere on the plane otherwise I will have to use my own...

 

Never trust a plane without duct tape...

[RAF_Louvert 101]

.

 

During my time in the AF our crew once flew with a pilot who couldn't keep his RC-135 on the runway during a landing. We decided he must have pronounced it "Boing!"

 

.

[UK_Widowmaker 571]

Boeing 737's?....those crates are just waiting to plummet from the Skies!

[Olham 164]

What a terrible mistake! If he was pulling the stick fully back all the time, he must have been thinking that he was in a steep descend, nose first?

I wonder what the artificial horizon was showing? He should have seen it there?

[Flyby PC 23]

I'd like to wave a little flag for the de Havilland Comet.

 

It pioneered the way in civilian jet air transport but lost it's safety reputation when a number of catastrophic accidents were unexplained. The Comet paid the price for being the first to encounter the problem, but every other airliner in the world benefited from the lessons eventually learned about metal fatigue.

 

The Comet was otherwise a fine aircraft, and way ahead of it's time. It was heavily modified and updated, but you can trace the lineage of the Hawker Siddley Nimrod which was on active duty from 1969 until prematurely retired in 2011 all the way back to the de Havilland Comet from 1949.

 

It's a tremendous injustice that the Comet is primarily remembered for the air crashes which destroyed it's reputation but at the same time made the world a safer place for those following in their pioneering footsteps.

Edited February 17, 2012 by Flyby PC

[Bullethead 12]

This accident shows the fallacy of that recovery technique. The stall was so deep, that the engines would never power the aircraft out of the stall...it would simply fall like a leaf (a very heavy leaf) into the water. Part of the reason was the training in only lowering the nose enough to accelerate, not enough to fully break the stall. This accident (along with others) has finally started shifting the mentality back to what we all learned in basic flying...break the stall by lowering the AOA, get some flying airspeed back, then pull back up...especially if you have buttloads of altitude below you.

 

There's a lot more to it than a particular stall recovery technique. When you have a flight control system that not only does most of the flying but can override the crew, the crew become reliant on it. Eventually, you end up with aircrew who lack the basic piloting skills to perform the simple task of keeping an airplane anywhere at all above sea level given 35,000 feet to work with. I'm sure anybody reading this forum, even if they've never flown a real plane in their lives, could, if dropping into a totally unfamiliar cockpit, at least manage this feat. Maybe they'd have trouble holding a constant altitude and heading, but keeping the plane somewhere, anywhere, above the deck shouldn't be a problem at all. That's what scares me, which is why I won't fly on an Airbus.

 

This might be old news to our European members but it's new to me, so bear with me while I outline the incident:

 

First off, setting the stage. Besides the above flight system, the plane had 1 senior pilot (Dubois) and 2 copilots (Robert and Bonin). When the trouble started, the 2 copilots were at the controls and the captain wasn't in the cockpit. The guy in the right seat (Bonin) was "flying" the plane, as in being responsible for any deviations from the autopilot's programmed course. Finally, the sticks of the 2 seats move independently, so that the guy in the left seat couldn't tell from his own stick that the guy in the right seat had his stick all the way back.

 

The sequence of events was as follows:

 

02:06:50

Plane is flying on autopilot at 35,000, in clouds. All is fine. The copilots discuss but ultimately decide not to turn on pitot tube heat.

 

02:08:03

Robert tells Bonin to turn a little to the left to avoid some rough weather ahead. While this is happening, the pitot tubes ice up and an alarm sounds indicating the autopilot is switching off due to a lack of airspeed data. Bonin recognizes he has control of the airplane. The PM article says neither Bonin nor Robert were trained to fly in manual without airspeed indicators, which makes me as WTF?

 

02:10:06

Bonin pulls back on the stick in an attempt to fly over the rough weather, going into a 6700 feet/minute climb. The plane slows to 93 knots and the stall warning sounds. Despite this, Bonin keeps pulling back on the stick.

 

02:10:25

Robert turns on the wing deicers and 1 pitot tube resumes working, so the crew now have airspeed data again. Robert realizes they are too slow and tells Bonin to descend.

 

02:10:36

Bonin keeps pulling back on his stick but not as hard, and tells Robert he's descending. Robert can't tell what Bonin's doing with the stick because his own isn't following Bonin's inputs. Airspeed goes to 223 knots and the stall warning stops. Everything is now fine and dandy: the airplane is under full control and the instruments are working. The only thing really "wrong" is that the autopilot is still off. Robert calls for Captain Dubois.

 

02:10:49

For no apparent reason, Bonin again pulls full back on the stick, and keeps it there for most of his remaining lifespan. Airspeed falls and the stall warning sounds again. At the same time, the remaining pitot tubes thaw out. All instruments are now working normally.

 

02:11:03

Bonin asks Robert to confirm that he's in TOGA (take off go around) mode. He's still trying to climb with the stick full back. The engines are at full power.

 

02:11:21

The plane reaches 38,000 feet and can go no higher. Bonin is still pulling full back on the stick, so the plane enters its fatal stall, nose-up 15^, IAS 100 knots, descending at 10,000 feet/minute, stall warning going full blast. Robert says he doesn't understand what's going on. Both he and Bonin might have forgotten that with the autopilot off, the flight control system will no longer prevent pilots from stalling the airplane, and thus regarded the stall warning as spurious, despite the evidence of all the other instruments.

 

02:11:32

Bonin says he doesn't have any control of the plane. In reality, he's the one causing the stall by continuing to hold the stick all the way back. If he'd let go, the plane would have righted itself.

 

02:11:43

Captain Dubois enters the cockpit and asks WTF? He does not take the controls. He can't see that Bonin is holding back on the stick (it's a small side stick) and can't figure out from the instruments why the plane is falling. It apparently never occurs to him that Bonin is holding the stick back, any more than it had to Robert.

 

02:11:45

Bonin says they've lost control of the plane. At this point the stall warning stops because forward speed is so low that the flight control system rejects the AOA inputs as invalid. The crew then debate whether they're climbing or descending, and eventually agree on descending. Nobody mentions the word "stall".

 

02:13:39

Robert says "Climb, climb, climb". Bonin says "But I've been holding the stick back the whole time!" Dubois says not to climb but to descend. Robert demands and receives control from Bonin and pushes the stick forward. The plane is now at about 2000 feet, probably too late to recover, and low-altitude alarms start sounding. But in any case, Bonin somehow takes the controls back from Robert without telling anybody and again pulls the stick all the way back.

 

02:14:23

The crew make various "Oh sh*t" comments

 

02:14:27

Splat.

[+Julhelm 266]

"But I've been holding the stick back the whole time!"

O_o

[MikeDixonUK 5]

I can't say I much like the idea of an Airliner with a flight stick, I prefer them to have yokes - not for any practical reason of course - I just think that big planes should have yokes and fighters should have sticks - it is the way of things!

 

Like them pesky P-38s with their yokes - it's just not right! Or these new fangled Submarines that don't even have periscopes!

 

With regards to the fact that the sticks aren't joined (not sure why they wouldn't be) - I'm presuming that one stick is the primary stick and one is the secondary stick, so the controls of one would always over-rule the controls of the other ie, if the primary was pulling up and the secondary was pushing down - the plane would pull up, as opposed to trying to pull up and push down at the same time! And presumably there's a button to change between the two?

 

As for the incident, it seems from what I've read here that most of the fault is with poor training of the crew, mixed with the fact that Pilots these days just aren't as experienced at actually flying, as they spend most of the time letting a computer do it, then the issues of unjoined sticks not physically telling the co-pilot what the pilot is doing, and also it not being easy to see what they're doing due to the sticks being small and in the corner - both of which you'd hope would be overcome with simple communication between one man and the other, but these are the things that are often overlooked in a crisis I guess.

Edited February 17, 2012 by MikeDixonUK

[Hasse Wind 46]

If pilots lack basic flying skills, bad things will happen, sooner or later. It's not good to learn to rely too heavily on hi-tech systems. And this is not restricted only to aviation. For example, how many people today can do some basic maths without using a calculator?

[MikeDixonUK 5]

If pilots lack basic flying skills, bad things will happen, sooner or later. It's not good to learn to rely too heavily on hi-tech systems. And this is not restricted only to aviation. For example, how many people today can do some basic maths without using a calculator?

 

I can count the number of times I've used simple maths in the past year on the fingers of one hand!

 

7!

[Hasse Wind 46]

I can count the number of times I've used simple maths in the past year on the fingers of one hand!

 

7!

 

[JFM 18]

I'm not doubting any of this but how--how??--could any pilot be so f***ing daft as to hold the stick all the way aft and not figure out what's causing the stall? That's just an unfathomable lack of basic stick-and-rudder aviating that every Cessna private pilot has drilled into his head WAY before his/her first solo. Again, how?? Even the worst pilots I've ever known--guys I wouldn't even taxi with--knew how to recognize and recover from a stall.

 

As far as Airbus, IMO they're noisier than Boeings but I didn't mind them. Except for the A-340; that thing (at least the two I've been on) is a dutch rolling barge.

[Bullethead 12]

With regards to the fact that the sticks aren't joined (not sure why they wouldn't be) - I'm presuming that one stick is the primary stick and one is the secondary stick, so the controls of one would always over-rule the controls of the other ie, if the primary was pulling up and the secondary was pushing down - the plane would pull up, as opposed to trying to pull up and push down at the same time! And presumably there's a button to change between the two?

 

From what I can tell, there is a button to switch between which stick "has the airplane". This is presumably how Bonin took the controls back from Robert just before the crash.

 

As to why the sticks aren't connected to each other, that's not surprising because they're not connnected to anything physical. They're really no different than our gaming joysticks, being merely electro-mechanical input devices for a computer. IOW, the only moving part is what you see sticking up out of the armrest, so there's nothing behind the instrument panel or under the floor to link together physically. The only way to make the sticks move in unison would therefore be a type of "force feedback" mechanism.

 

Aparently, this wasn't thought worthwhile or it would have been there from the start. I suppose it might be suggested now to prevent future accidents. But I think that instead, the flight control system will get a few tweaks to stay running under more adverse conditions and be given even more authority to disregard crew inputs. I mean, the whole Airbus philosophy is that pilots are only there because the law requires them. The almighty flight control system is supposed to do everything. This is a strategy intended to own the market of the developing world, places where there's not even an education system capable of turning out pilot-quality graduates in sufficient quantity, let alone an adeguate flight-training infrsstructure. And because Airbus is so heavily subsidized that it's effectively a branch of government, it can get away with this.

 

As for the incident, it seems from what I've read here that most of the fault is with poor training of the crew, mixed with the fact that Pilots these days just aren't as experienced at actually flying, as they spend most of the time letting a computer do it

 

Well, when the goal of the whole program is to take a guy out of a grass hut and put him in the cockpit of an airliner, what do you expect? Even in countries that have the education system and flight schools to turn out real pilots, if the main job of the crew is to tell the air traffic controllers what the airplane has decided to do for itself, then why bother training them how to fly? IIRC, in this Airbus crash, even Bonin, an employee of Air France and presumably a Frenchman, had 5000 hours. You'd think that somewhere in all that time he'd have learned that when the stall warning is going off, pulling back on the stick is a bad idea. I learned that, without benefit of a stall horn, flying models before I was 10. When I took flying lessons, I learned that before I even got in the airplane the 1st time. And once I was off the ground, all we did the 1st few lessons was demonstrate that the airplane will stall if you hold the stick back long enough, and what to do after it stalled.

 

So from my admittedly dinosaur POV, there's absolutely zero excuse for anybody with wings on his chest to 1) get into this situation at all or 2) fail to get out of it should some inexplicable paranormal sequence of events cause such a problem. When 3 guys, each with thousands of hours of so-called flying experience, let this happen, what does that tell you?

[FastCargo 412]

Nowhere in that transcript does it show the aircraft is overriding the pilot's commands. The aircraft followed the pilot's commands the entire time...which is in fact why they were killed.

 

Let me clear something up right now. 99% of the the time, a Boeing and an Airbus are flown in the exact same manner. Yes, that's right, the same so called bad over-reliance on automation is done the exact same way by both types of aircrew. Your argument for what got this crew killed is based on a flawed premise.

 

Now, at the limits, there are some fundamental differences. In a modern Airbus (as opposed to the older A300/A310 series), there is a hard G limiter and a hard bank limiter. That's it. In full manual mode, an Airbus won't let you put it on it's back or rip the wings off...while a Boeing will. Otherwise, in the common normal modes that both types have, there are similar protections in place that, yes Virginia, will override the pilot's commands to keep the aircraft from hitting the ground or getting too slow.

 

In fact, there is a beautiful system on the Boeing 777 that if you lose an engine at V1 (the most dangerous time to lose an engine), it flies the aircraft so well, a few friends have remarked they didn't notice anything right away except for the actual warning signals. That's with a 100,000 lb thrust class engine going to zero in a few seconds, and the system compensating with flight controls so the aircraft doesn't go into a gigantic yaw (which normally takes an assload of pilot rudder to counteract on a modern airliner).

 

Lets talk about the accident a bit more shall we? First, this was the middle of the night, with weather. Bells and horns are going off, the aircraft is shaking, and SA is already low. The aircrew is relying on that training I mentioned eariler...training that most airline aircrew had been receiving concerning stall recovery. They are expecting to power out of a stall that was so deep, they have never encountered it before...yep, not even in the simulator. They don't have a big sky to look at to see just how nose high they were (and at about 15 degree pitch, they aren't that nose high...typical rotation angle is 18 degrees, windshear recovery is 17.5 degrees, GPWS alert recovery is 20 degrees) or how fast they are dropping. In other words, the actual aircraft attitude is within their experience...it's just the flying speed and AOA that isn't. And because their airspeed is unreliable due to the icing, they don't have reason to trust it's giving them accurate readings. Which leaves the AOA...which most airliners, again of both types, don't show on the primary attitude display.

 

Starting to get the picture now? This isn't some Cessna on a 'clear and a million' day...it's a multi-ton automated airliner in the middle of the night in a regime of flight the aircrew has never seen before...even in a simulator. Also, unlike the stall characteristics of a high asymmetric cambered, straight winged aircraft like a Cessna, where the stall is followed by an immediate nose drop and airspeed gain (a sort of self fixing situation), the supercritical swept wing of a modern airliner in a full stall acts more like some high performance aircraft, where there isn't a definite drop in pitch, but the sink rate continues to increase at a high rate. The T-38 has a similar characteristic, like the 'Century series' aircraft it was training you for. You could sit there in the stall, the nose above the horizon, with lots of buffet, in full afterburner, and think you were okay...until your saw your VVI pegged at 6000ft per minute...in the descent. And the T-38 killed plenty of folks this way early on (yep, even in 'clear and a million' weather) until an AOA gauge was installed and brought into the cross check, and training emphasized the 'gotchas' in the flight envelope.

 

As far as side sticks go in modern airliners, that I can't speak to. After all, these are 'side sticks', like in an F-16, and so there isn't a whole lot of movement in the first place. They probably work similar...in that there is no direct feedback in the stick other than resistance reaching the limit and a 'stick shaker'. However, in the place of opposite inputs, one stick may be a 'master' and the other a 'slave' (probably Capt and F/O in that order). However, that may not make a difference...one only has to look at Egypt Air 990 to see that even with a conventional yoke setup, the F/O can override the Captain.

 

Personally, I'd love to fly an airliner with a side stick. Having flown 'stick and throttle' jets (small and large) my whole military career, aircraft with yokes just don't feel right and are like driving a bus. And, having been lucky enough to fly an F-16D, I am convinced that if you're going to go FBW, go with the side stick...it just feels right.

 

In conclusion, inadequate training killed this crew, not the Airbus philosophy. I personally like Boeing's more than Airbus, but I don't consider the Airbus philosophy to be any less safe. And I've been qualified in Boeing, McDonnell Douglas and Airbus airliners, so I have a little bit of knowledge on the subject.

 

FC

[JFM 18]

Great stuff.

 

Even in a Cessna, as I'm sure you know, although it recovers itself from a stall as stated, it'll stay in one if you hold the yoke fully aft. But, why do that? This is what I don't understand with this particular accident. Night, confusion, weather, yadda, okay, but why hold the stick fully aft? My experience is flying recip ASEL/AMEL machines so I don't know: Is there ever a situation where full aft stick/yoke would be used for, well, any aspect of flying an airliner? (To be clear, that's an actual question, not a rhetorical one.)

[FastCargo 412]

Great stuff.

 

Even in a Cessna, as I'm sure you know, although it recovers itself from a stall as stated, it'll stay in one if you hold the yoke fully aft. But, why do that? This is what I don't understand with this particular accident. Night, confusion, weather, yadda, okay, but why hold the stick fully aft? My experience is flying recip ASEL/AMEL machines so I don't know: Is there ever a situation where full aft stick/yoke would be used for, well, any aspect of flying an airliner? (To be clear, that's an actual question, not a rhetorical one.)

 

That is a little puzzling. In stall recoveries, especially low altitude, the idea is to max perform the aircraft....get the AOA just a small margin below stall to get the most lift for your current airspeed.

 

If I were theorizing, I would suspect that they thought they were not in an actual stall due to the unreliable airspeed indicator and the fact that their pitch hasn't dropped at all. They thought they could power out of the situation and stabilize their altitude and so were pulling to the G limit.

 

This is not the first time a blocked pitot static system has killed an airliner. Here's another...a Boeing 757 in 1996.

 

http://en.wikipedia.org/wiki/Aeroperú_Flight_603

 

Note some similar circumstances...night, low or no visual references, unreliable and contradictory airspeed (and altitude) indications. Basically, the aircrew flew an otherwise perfectly flying aircraft into the water because of lack of SA. Had they simply maintained a known pitch and power setting, they probably could have been okay. But with all the bells and whistles going off, they probably got overloaded, and simply missed the RALT giving them their true AGL..

 

FC

[JFM 18]

I'm just puzzled (or not learned enough to understand) what they thought full aft control input was going to do for them within any flight parameter, unless they wanted to loop the plane or something. Although, via my aerobatic "experience" (one hour of SNJ time), even the loops didn't require full aft input.

 

Yeah, I'm familiar with the demise of Flight 603. Helluva thing a piece of tape can lead to.

[FastCargo 412]

And folks, an addendum.

 

I am not saying that one particular aircraft maker is better than another. Both large players in the airliner game do things and have faults that drive people crazy. Nor am I necessarily trying to change anyone's mind about their favorite manufacturer. What I am doing is trying to clear some misconceptions about how the companies do things, and why particular accidents have occurred.

 

FC

[FastCargo 412]

Well, we don't really know at this point. Sometimes reflex can be a hell of a thing to let go of...our natural instincts if in a descent of not one's own making is to pull back on the stick, and if that doesn't seem to be making the aircraft respond when it feels like it should, well, maybe you aren't pulling enough...

 

I'll be the first to say it sounds simplistic, but weirder things have happened under stress.

 

Here's an example of a friend of mine almost becoming a statistic because he reacted the way you normally would.

 

He's flying in a T-38, solo, when a large bird appears to be headed straight toward his forehead (by the way, if he had not reacted, he would have hit the bird dead on in the windscreen...there's HUD video). My friend, not wishing to have a turkey vulture lunch with a side salad of plexiglass, decides to push forward on the stick to avoid the bird. The problem is that he's only about 175 knots, gear and flaps down, in a 45 degree bank, about 400 feet, with a 2500 VVI in the final turn. His rapid push forward just about doubles the VVI with less than 400 feet left until ground impact...yep, slightly less than 8 seconds to kill that much VVI while fully configured. He got immediate ground rush, and so did a stall recovery...but pulled straight to 1.0 AOA verses the recommended .8-.85 AOA. The reason we normally don't teach to pull to 1.0 AOA is that any misstep means now you're on the wrong side of the lift curve...vs .8-.85 gives you some wiggle room.

 

That decision to go to 1.0 AOA saved his life.

 

There are some small telephone poles off the end of the runway about 1/2 of a mile. They aren't very tall (standard small ones like you see in residential areas) with wires strung across the top. His aircraft went under those wires, 'sabre dancing' and leaving a giant rooster tail of dirt and grass. He crossed a highway (thank god no cars were in the way...he wouldn't have seen them anyway because his nose was so high), roostered some more onto the field, took out a chain link fence with his left main gear, which then snapped off, and eventually got out of ground effect. He was able to eventually bring up what remained of the landing gear and do a belly landing (a T-38 is not considered landable with only one main gear).

 

He received a lot of flak for pushing the stick down that got him into trouble. But think about what you would instinctively do if you saw a bird the size of a large house cat about to smash into your face at 175 knots...

 

Sometimes, instinct can be real hard to overcome, even with dedicated training.

 

FC

[hood 2]

What a fascinating discussion.Thanks to all you gentlemen.

[Bullethead 12]

Nowhere in that transcript does it show the aircraft is overriding the pilot's commands. The aircraft followed the pilot's commands the entire time...which is in fact why they were killed.

 

I'm sorry, I must not have been clear enough. The autopilot, when operating, will override crew input if that input will cause the airplane to go beyond its flight envelope. For instance, had the system been working when Bonin pulled the stick all the way back, the amount of up elevator he actually got would have decreased over time as the plane slowed, until he was getting zero elevator response despite still having the stick full back.

 

It's my contention that such a system breeds hamfisting. When the system is on, you only get what the airplane wants to give you no matter how you move the stick or pedals. Thus, I think it would naturally habituate the crews to just move the stick fully in all directions, relying on the system to smooth things out and limit effects. So, when the system is suddenly gone, there'd be a tendency to keep moving the stick the same way as with it on. IOW, I can easily see Bonin thinking there was nothing at all extraordinary about having the stick fully back. I'm sure he did it every day, knowing he wasn't going to get anywhere near that much elevator. But he'd either forgotten that the autopilot was off at the time, or at least how to fly without it.

 

So yes, the crew, particularly Bonin, flew the plane into the ocean, not the system. I'm not arguing that. But other than over-reliance on the system actually flying the airplane, and being trained under that principle, how else do you explain Bobin's full back stick all the time. He obviously wasn't trying to loop the plane, he just wanted to climb a little, but he pulled the stick full back anyway. Why? Because to him, pulling the stick back wasn't flying the plane, it was just asking the plane to go up and it would decide how much.

 

And the T-38 killed plenty of folks this way early on (yep, even in 'clear and a million' weather) until an AOA gauge was installed and brought into the cross check, and training emphasized the 'gotchas' in the flight envelope

 

So the question really is, why wasn't this crew thoroughly schooled on the "gotchas" of this plane's envelope? The only reason I can think of is that the system was supposed to prevent you from finding the "gotchas", so why spend the time and money training for something that will never happen.

 

In conclusion, inadequate training killed this crew, not the Airbus philosophy. I personally like Boeing's more than Airbus, but I don't consider the Airbus philosophy to be any less safe. And I've been qualified in Boeing, McDonnell Douglas and Airbus airliners, so I have a little bit of knowledge on the subject.

 

I don't doubt your knowledge and we're not disagreeing that lack of training was the main killer. But why was there such a lack of training? Why would a guy with thousands of hours only use full stick input in situations that clearly didn't call for it, and when he obviously didn't want full control response? I'm talking before the fatal stall, when he was just trying to climb over rough weather. To me, that speaks of a much greater training failure than merely changing a method of getting out of (or preventing) a stall. It's like the guy really didn't know how to fly at all. And that, I think, comes from knowledge that most of the time, he really wasn't flying the plane, just making suggestions to its control system.

[FastCargo 412]

I'm sorry, I must not have been clear enough. The autopilot, when operating, will override crew input if that input will cause the airplane to go beyond its flight envelope. For instance, had the system been working when Bonin pulled the stick all the way back, the amount of up elevator he actually got would have decreased over time as the plane slowed, until he was getting zero elevator response despite still having the stick full back.

 

It's my contention that such a system breeds hamfisting. When the system is on, you only get what the airplane wants to give you no matter how you move the stick or pedals. Thus, I think it would naturally habituate the crews to just move the stick fully in all directions, relying on the system to smooth things out and limit effects. So, when the system is suddenly gone, there'd be a tendency to keep moving the stick the same way as with it on. IOW, I can easily see Bonin thinking there was nothing at all extraordinary about having the stick fully back. I'm sure he did it every day, knowing he wasn't going to get anywhere near that much elevator. But he'd either forgotten that the autopilot was off at the time, or at least how to fly without it.

 

So yes, the crew, particularly Bonin, flew the plane into the ocean, not the system. I'm not arguing that. But other than over-reliance on the system actually flying the airplane, and being trained under that principle, how else do you explain Bobin's full back stick all the time. He obviously wasn't trying to loop the plane, he just wanted to climb a little, but he pulled the stick full back anyway. Why? Because to him, pulling the stick back wasn't flying the plane, it was just asking the plane to go up and it would decide how much.

 

No.

 

Airbus airliners do NOT fly this way.

 

In normal operations with airliners (Boeing, Airbus, McDonnell, etc), when the autopilot is on, the aircraft is not flown with the stick, it is usually flown with the control panel. You dial in altitudes, headings, speeds, etc, either through the upfront panel or through the Flight Management System (FMS) and the aircraft goes there. You do not touch the stick or throttles.

 

When you start moving the stick past a certain range, with the autopilot engaged, the autopilot will click off. The stick now controls the aircraft. The hard limits are still in place (various G, bank, rate limits, etc.) and alpha speed and floor are still in place (throttle up if you get too slow or too low without gear). You can still handle the aircraft way too roughly. You can slam people to the ground, break bones, or hit heads into the sides with excessive roll and pitch rates, all without exceeding the bank and G limits. You click off the auto throttle control, and you get rid of the alpha speed and floor safties too. You CANNOT handle the Airbus by going full deflection on the control stick and not start injuring or killing people doing it, even with the hardcoded limits.

 

There was incident in fact where an A320 crew got the aircraft too fast in the descent and pulled right to 2.5Gs trying to prevent an overspeed. Just that one pull broke bones in the passenger compartment (admittedly, unexpected...folks were standing at the time)...all without exceeding the Normal Law limits.

 

That is NOT how the aircraft is flown.

 

So the question really is, why wasn't this crew thoroughly schooled on the "gotchas" of this plane's envelope? The only reason I can think of is that the system was supposed to prevent you from finding the "gotchas", so why spend the time and money training for something that will never happen.

 

As they say in Monty Python: 'No one expects to be less than 100 knots airborne in an airliner!'

 

I've said before that I've qualified in a Boeing 737, McDonnell Douglas DC-10, an Airbus A300/A310, and the Rockwell B-1B. In absolutely NONE of those aircraft, was the idea that I could be something less than close to stall speed EVER addressed. Training consisted of stall avoidance, recognition and recovery...NOT in trying deal with a fully stalled aircraft (my term 'deep stall' is actually a misnomer...that applies to T-tails only). We didn't deal with that part of the envelope because you weren't ever expected to GET there. This is NOT an aircraft type specific problem. We deal with wind shear and GPWS (terrain) recoveries as well...but it's still assumed you're somewhat close or over minimum flying airspeed.

 

I don't doubt your knowledge and we're not disagreeing that lack of training was the main killer. But why was there such a lack of training? Why would a guy with thousands of hours only use full stick input in situations that clearly didn't call for it, and when he obviously didn't want full control response? I'm talking before the fatal stall, when he was just trying to climb over rough weather. To me, that speaks of a much greater training failure than merely changing a method of getting out of (or preventing) a stall. It's like the guy really didn't know how to fly at all. And that, I think, comes from knowledge that most of the time, he really wasn't flying the plane, just making suggestions to its control system.(FC emphasis added)

 

I have friends on both sides of the fence on the Boeing/Airbus control command preference issue. Frankly, we are all scratching our heads on this part because training does tend to be universal...why the full back stick deflection the whole time? As I postulated before, my theory is that they were so overloaded, they didn't realize they were in a stall, and so never thought to lower the nose to gain airspeed to fly out of it.

 

I hate to say this, but 95% of the time of a modern airliner's flight envelope, that's all the pilots are doing anyway is telling the aircraft where to go, and letting the autopilot take care of the getting there part. Doesn't matter if it's Boeing, Airbus, Embrarer, Canadair, etc...they ALL fly this way. In fact, above FL290 in most countries, the aircraft is required by law to fly with the autopilot working the system in normal operations (there are exceptions of course, but not for most civil airliners).

 

What I'm trying to say is this: In my opinion, you are correct in that over reliance on automation and safety limits can result in making the wrong decisions when you lose all that protection. However, you are incorrect in assuming that this is somehow a problem unique to Airbus. It is not...there are tons of incidents across all brands of aircraft where over reliance on automation has resulted in accidents.

 

Where automation gets pilots in trouble is when you use it as a crutch to get you to 100%. That is not the way it is to be used. Its true purpose is to keep you from getting fatigued, so that you are 100% when the time comes that you need to be.

 

FC