Old news but good article by Wm Langewiesche on Air France 447

[MAKG1]

I'm sorry, but that's just ludicrous at best.

Every engineering decision is a cost/benefit analysis at one level or another.

Even a very simple software system will have many shortcuts made in the interest of feasibility. An aircraft software system is exceedingly complex.

This is astoundingly poorly understood outside the field. A ludicrously complex mechanical system might have a few hundred moving parts. An FMS has hundreds of thousands, for example. Such a system will never be completely spec'd or tested. Ever. It's not possible, even in principle and even with infinite budget (it also requires infinite schedule). Which means a cost/benefit for everything.

 

[RotorAndWing]

Yeah, engineers never make mistakes. Which is why we(and the French) don't need an AD system for aircraft.

Oh - wait....

That recent column about the P210 is an engineering nightmare. I really liked the part where faulty design spec helped kill the four doctors. Sure - the average pilot should have been able to get the damaged plane down partial panel no prob. Sure.

Since I've been involved with Airbus I have noted they are very proactive on maintaining the active fleet and often issue various bulletins and system upgrades whenever they discover a potential problem. They keep a constant tab on the data from each aircraft and watch trends to stay ahead of potential problems.

People here keep comparing Airbus with GA aircraft, which is just silly. Does Cessna plot all of the data from their airplanes? Does Beech have a system to track all of the flight parameters of their airplanes and report back to Wichita?

Do the GA manufacturers take the data when they see a trend, redesign the system and issue an upgrade to the operator, or even put out a safety warning to the operator?

Somehow the idea here is once the airplane (Airbus) is built and sold it becomes like a GA aircraft and never hears or sees the factory again.

 

[RotorAndWing]

No not really. Cost/benefit analysis happens all the time.

Seriously, keep this in the context of the discussion (AF447).

You don't think after this happened the engineers ran simulation after simulation, analyzed thousands of hours of data and tried to find a problem in the system and a correction?

Again, the problem with AF447 was pilot error, the pilot flew the airplane into the ocean. It's not some secret hidden system flaw that Airbus doesn't feel needs correction due to a cost/benefit analysis.

 

[RotorAndWing]

This is astoundingly poorly understood outside the field. A ludicrously complex mechanical system might have a few hundred moving parts. An FMS has hundreds of thousands, for example. Such a system will never be completely spec'd or tested. Ever. It's not possible, even in principle and even with infinite budget (it also requires infinite schedule)..

Do you even know what an FMS is in relation to Airbus?

 

[docmirror]

This is astoundingly poorly understood outside the field. A ludicrously complex mechanical system might have a few hundred moving parts. An FMS has hundreds of thousands, for example. Such a system will never be completely spec'd or tested. Ever. It's not possible, even in principle and even with infinite budget (it also requires infinite schedule). Which means a cost/benefit for everything.

We had a hospital that was interested in one of our network systems. Someone in procurement got a spec back from the engineering dept and one of the spec requirements was a stochastic test of all combinations of connections and speeds. So, they gave it to me to get started on, and I just laughed and told mgmt the testing would take about 3-4 years to complete. They said no way, they wanted to see the math.

So, I got out my fancy calculator, and said; 'let us begin. We have a system with 544 input/outputs, each of those has to be tested as input and output. Each of them can also be operating in five different speeds. So if they want stochastic tests that's n!/(n-r)!(r!) which gives about 3.6ee11, then we have to test them all as output types so double that, then we have to start testing each one as input as we traverse the network, then the same thing as we set all to input and traverse the network setting one at a time as output. I came up with a best guess at 2.455ee41 stochastic examples. Breaking it down to 5 seconds of test time per combination, the time worked out to 4 and a half years. If they had the PO ready, I was ready to code it. We didn't hear back from them.

 

[MAKG1]

Seriously, keep this in the context of the discussion (AF447).

You don't think after this happened the engineers ran simulation after simulation, analyzed thousands of hours of data and tried to find a problem in the system and a correction?

Again, the problem with AF447 was pilot error, the pilot flew the airplane into the ocean. It's not some secret hidden system flaw that Airbus doesn't feel needs correction due to a cost/benefit analysis.

Not secret.

Every engineering change has side effects. They may be concerned that they may introduce more problems than they solve by changing control laws. And they would probably be right.

Engineering is very seldom so simple as "works vs. doesn't work." Virtually everything works after a fashion.

 

[RotorAndWing]

Not secret.

Every engineering change has side effects. They may be concerned that they may introduce more problems than they solve by changing control laws. And they would probably be right.

Engineering is very seldom so simple as "works vs. doesn't work." Virtually everything works after a fashion.

Uh, got news for you sparky, they change parameters on control laws on occasion when they deem a change needs to be made. On some airplanes the parameters may be different to correct a problem discovered through data analysis. The Airbus is a constant evolving program as more data is gathered.

Again, most here believe once the airplane leaves the factory it remains in it's same condition much the same as a GA light airplane.

 

[dmspilot]

Seriously, keep this in the context of the discussion (AF447).

You don't think after this happened the engineers ran simulation after simulation, analyzed thousands of hours of data and tried to find a problem in the system and a correction?

Again, the problem with AF447 was pilot error, the pilot flew the airplane into the ocean. It's not some secret hidden system flaw that Airbus doesn't feel needs correction due to a cost/benefit analysis.

Your failure to see the world in anything other than black and white makes it seem almost as if you have never heard of the concept of "human factors". Yes, the pilot screwed up. You're a real Sherlock Holmes. The question is why did he screw up, and probably more importantly, why did two other more experienced pilots let him?

 

[docmirror]

Since I've been involved with Airbus I have noted they are very proactive on maintaining the active fleet and often issue various bulletins and system upgrades whenever they discover a potential problem. They keep a constant tab on the data from each aircraft and watch trends to stay ahead of potential problems.

People here keep comparing Airbus with GA aircraft, which is just silly. Does Cessna plot all of the data from their airplanes? Does Beech have a system to track all of the flight parameters of their airplanes and report back to Wichita?

Do the GA manufacturers take the data when they see a trend, redesign the system and issue an upgrade to the operator, or even put out a safety warning to the operator?

Somehow the idea here is once the airplane (Airbus) is built and sold it becomes like a GA aircraft and never hears or sees the factory again.

You know eff-all about engineering. Maybe you're the hottest stick since Wolfgang, but systems design, testing, human factors, closed/open feedback systems, inertia, heuristics, fault tolerance, loads, stresses, strains(not the same), energy, conservation, coupling, and materials are beyond anything you can imagine. Heck - I'm just a lowly EE, with a modest understanding of the aero side of things and even I'm finding serious faults. What does that say about Airbus vaunted engineers? The guy who came up with the idea of independent stick controls was wrong. Wrong, wrong, wrong. Yes it works, yes it is trainable, yes it was contributory in this accident. To quote one of the best engineers ever: 'to change something is easy; to improve is very, very difficult.' Was the change in stick operation a change? Yep. Was it an improvement? Nope, not even close. Same with the stall warning. Was the plane stalled? Why yes it was. Was the stall warning blaring? Why no it wasn't. The designer/engineer decided that a plane in a stalled condition lower than 60 knots was not stalled, therefore don't report that. Oops.

 

[docmirror]

And while I'm on a roll, what about that Airbus that was on some kind of victory tour right after it was unveiled and crashed on a low pass I think? I'm not sure on the details, but as I recall, a perfectly operating and perfectly designed plane did exactly what the pilot told it to do - right up to the point where it didn't. Something about the engines not responding to throttle inputs? No - I'm sure the pilots were wrong in that too, cuz who would think that pushing the throttles all the way forward means 'more power Mr Scott!'

Wonder what happened to that little design perfection?

 

[RotorAndWing]

Your failure to see the world in anything other than black and white makes it seem almost as if you have never heard of the concept of "human factors". Yes, the pilot screwed up. You're a real Sherlock Holmes. The question is why did he screw up, and probably more importantly, why did two other more experienced pilots let him?

Actually I've had considerable training on Human Factors, Human Factors in Aviation Accident Investigation from the Transportation Safety Institute at the MMAC in OKC.

It's quit evident why the pilot screwed up and has been discussed and evaluated in detail in the airline industry. Low experience pilot in a high tech airplane, cultural issues as well as a lack of CRM, to name a few.

Speaking of Sherlock Holmes, you are the one in that role as you keep trying to blame Airbus for a "needless layer of obfuscation" yet you don't even understand how the systems actually work.

 

[Everskyward]

Seriously, keep this in the context of the discussion (AF447).

You don't think after this happened the engineers ran simulation after simulation, analyzed thousands of hours of data and tried to find a problem in the system and a correction?

How is this not in the context of AF447? Everything is a cost/benefit decision. How do you know what tests they ran during the initial design of the system? Way up in the thread someone posted that the 777 has yokes that work together, not because they are connected mechanically but because of computer design. Someone made this decision, not because it was cheaper, but because pilots wanted it that way, according to the poster.

Again, the problem with AF447 was pilot error, the pilot flew the airplane into the ocean. It's not some secret hidden system flaw that Airbus doesn't feel needs correction due to a cost/benefit analysis.

Sure it was pilot error but I think that there may have been some other things which contributed.

 

[RotorAndWing]

You know eff-all about engineering. Maybe you're the hottest stick since Wolfgang, but systems design, testing, human factors, closed/open feedback systems, inertia, heuristics, fault tolerance, loads, stresses, strains(not the same), energy, conservation, coupling, and materials are beyond anything you can imagine. Heck - I'm just a lowly EE, with a modest understanding of the aero side of things and even I'm finding serious faults. What does that say about Airbus vaunted engineers? The guy who came up with the idea of independent stick controls was wrong. Wrong, wrong, wrong. Yes it works, yes it is trainable, yes it was contributory in this accident. To quote one of the best engineers ever: 'to change something is easy; to improve is very, very difficult.' Was the change in stick operation a change? Yep. Was it an improvement? Nope, not even close. Same with the stall warning. Was the plane stalled? Why yes it was. Was the stall warning blaring? Why no it wasn't. The designer/engineer decided that a plane in a stalled condition lower than 60 knots was not stalled, therefore don't report that. Oops.

I didn't realize you had so much experience with Airbus, and such an in-depth knowledge of its systems, as well as having actually sat in the cockpit and having actual operating experience.

 

[jesse]

Do you honestly think that Airbus hasn't been studying this Rotor? You write as if they just waited for the NTSB report and were like Meh, pilot error, let's go play golf.

I can guarantee you there is a team of people at Airbus that have been analyzing exactly what we've been analyzing in this thread and they've been looking at their systems and looking at the cost/benefit of making a change. Could a change be made that would have prevented this? Most certainly. Will they make the change? Depends on what they determine the cost/benefit to be.

I can promise you there was an engineer out there at one point that thought something should have been different and that change would have prevented this crash and was shot down by the numbers folks. That's how engineering works.

 

[docmirror]

I didn't realize you had so much experience with Airbus, and such an in-depth knowledge of its systems, as well as having actually sat in the cockpit and having actual operating experience.

Your lack of realization is epic.

Most catastrophic failures like this one are cascaded. I can't believe you have had any training in human factors and still pin this on the pilots to the exclusion of everything else: "the plane did exactly what it was commanded". That's one of those bon mots like; 'well, the operation was a complete success madam, sadly the patient died.'

Oye vey.

 

[Everskyward]

Do you honestly think that Airbus hasn't been studying this Rotor? You write as if they just waited for the NTSB report and were like Meh, pilot error, let's go play golf.

I can guarantee you there is a team of people at Airbus that have been analyzing exactly what we've been analyzing in this thread and they've been looking at their systems and looking at the cost/benefit of making a change. Could a change be made that would have prevented this? Most certainly. Will they make the change? Depends on what they determine the cost/benefit to be.

I can promise you there was an engineer out there at one point that thought something should have been different and that change would have prevented this crash and was shot down by the numbers folks. That's how engineering works.

Even if there is no system change I'll bet there will, or already is, a training change, even if it's only awareness. As someone else pointed out, it took a while, but the stall demos were revised.

 

[RotorAndWing]

Do you honestly think that Airbus hasn't been studying this Rotor? You write as if they just waited for the NTSB report and were like Meh, pilot error, let's go play golf.

I can guarantee you there is a team of people at Airbus that have been analyzing exactly what we've been analyzing in this thread and they've been looking at their systems and looking at the cost/benefit of making a change. Could a change be made that would have prevented this? Most certainly. Will they make the change? Depends on what they determine the cost/benefit to be.

I can promise you there was an engineer out there at one point that thought something should have been different and that change would have prevented this crash and was shot down by the numbers folks. That's how engineering works.

I have been trying to say that yes, Airbus has analyzed this every which way to Sunday, gathered thousands upon thousands of hours of data to try to find a problem.

The "change" was to replace the pitots, yes they were defective. As far as procedures? There may have been some minor changes, but system wise they couldn't find fault with the airplane. I do know Airbus has ran hundreds (thousands?) of hours of simulation to replicate the failures but the outcome wasn't fatal as in AF447. Competent, trained crews worked their way through the problem, the system redundancy worked as it should and the plane flew.

Should Airbus have placed some incompetent pilots at the controls and simulated the failure? Should the system be designed for the lowest common denominator and be designed to be operated by an imbecile?

Go read this thread on Air France and how they took a perfectly operating airplane and almost lost it.

Because of the crew confusion on this should Airbus redesign the airplane?

 

[nauga]

I can promise you there was an engineer out there at one point that thought something should have been different and that change would have prevented this crash and was shot down by the numbers folks. That's how engineering works.

That's only one way of any that it can go. Have you ever been involved in the design or test of a full-authority FBW flight control system? Are you familiar with DO-178B/C? FMEA? Design and testing of flight critical software? This thread has degenerated into "Something should have been done! Think of the children!"

Nauga,
who can name a handful of accidents that could have been prevented by a singing porcupine

 

[RotorAndWing]

Your lack of realization is epic.

Most catastrophic failures like this one are cascaded. I can't believe you have had any training in human factors and still pin this on the pilots to the exclusion of everything else: "the plane did exactly what it was commanded". That's one of those bon mots like; 'well, the operation was a complete success madam, sadly the patient died.'

Oye vey.

 

[jesse]

That's only one way of any that it can go. Have you ever been involved in the design or test of a full-authority FBW flight control system? Are you familiar with DO-178B/C? FMEA? Design and testing of flight critical software? This thread has degenerated into "Something should have been done! Think of the children!"

Nauga,
who can name a handful of accidents that could have been prevented by a singing porcupine

I don't think something SHOULD have been done. You can't win them all. The idea that people aren't and shouldn't analyze if something should be done is what is pretty funny in this thread because there is no doubt Airbus is/has done just that.

 

[MooneyDriver78]

And it took 3 accidents to realize this... Pinnacle 3701, Colgan 3407, AF 447

4, add Birgenair 301
Just like AF447, except is was insects instead of ice, and it was the Captain who stalled the plane into the ocean, 2 other pilots were telling the Captain he was stalling the plane, and it was a Boeing instead of an Airbus.
The stick-shaker going off, and the Captain still yanking back for all that he was worth.

 

[RotorAndWing]

Even if there is no system change I'll bet there will, or already is, a training change, even if it's only awareness. As someone else pointed out, it took a while, but the stall demos were revised.

I read the Airbus OEB's as they come out, plus our training department gets regular revisions from Toulouse as they continuously update training. That and we have a FOMS department that analyzes our flight data continuously looking for red events.

AAF447 is discussed in detail, systems review is emphasized and everyone has experienced similar failures in training to understand why following published procedures are important as well as to emphasis CRM and procedures.

 

[nauga]

The idea that people aren't and shouldn't analyze if something should be done is what is pretty funny in this thread because there is no doubt Airbus is/has done just that.

I'm not getting that from any poster in this thread. It has clearly been analyzed, by both Airbus and independent parties. The big divide seems to be whether or not there is a significant problem with the control laws.

For the record (for other posters, not necessarily you specifically), Airbus is *not* the only manufacturer who uses non-interconnected sidesticks nor are they the only ones with sticks that are not backdriven to indicate autopilot function. There are military examples of both (e.g. two-seat F-16's for the former, F-16, F/A-18 for the latter), but they tend to have somewhat more rigorous standards for pilot selection and training.

Nauga,
and his closed-book NATOPS test

 

[RotorAndWing]

I'm not getting that from any poster in this thread. It has clearly been analyzed, by both Airbus and independent parties. The big divide seems to be whether or not there is a significant problem with the control laws.

For the record (for other posters, not necessarily you specifically), Airbus is *not* the only manufacturer who uses non-interconnected sidesticks nor are they the only ones with sticks that are not backdriven to indicate autopilot function. There are military examples of both (e.g. two-seat F-16's for the former, F-16, F/A-18 for the latter), but they tend to have somewhat more rigorous standards for pilot selection and training.

Nauga,
and his closed-book NATOPS test

Exactly!

 

[weirdjim]

Yeah, engineers never make mistakes.

Corvair, Titanic, Pinto, GM ignition switches, Hindenburg, ... all come immediately to mind.

Jim
An engineer who has made his share.

 

[MAKG1]

Uh, got news for you sparky, they change parameters on control laws on occasion when they deem a change needs to be made. On some airplanes the parameters may be different to correct a problem discovered through data analysis. The Airbus is a constant evolving program as more data is gathered.

Again, most here believe once the airplane leaves the factory it remains in it's same condition much the same as a GA light airplane.

Changing parameters is not the same as changing the laws. "Sparky."

And they will NOT be changed arbitrarily.

Some of us really do engineering.

 

[RotorAndWing]

Changing parameters is not the same as changing the laws. "Sparky."

And they will NOT be changed arbitrarily.

Some of us really do engineering.

That's hard to tell from your postings. Real engineers use correct terminology, something you seem to have trouble with.

 

[SCCutler]

Aren't Dash-8 yokes interconnected? How well did that work for Colgan 3407? I think it's fair to say that the lack of interconnect and/or backdrive compounded the confusion, and I believe the accident report states as much, but I don't think it's fair to say with absolute or even strong certainty that had the sticks been interconnected that 447 would not have hit the water.


Nauga,
and his SOP

Irrelevant, because...

You can never say anything like that with certainty but the difference between the two accidents is that AF447 took a number of minutes to develop while Colgan3407 was over in seconds.

...of that.

But the common thread in those two accidents were lack of basic airmanship skills.

Superficially, yes. But, in he case of the Air France crew, the poor airmanship led (in turn) to utter confusion and, ultimately, disbelief, disbelief which continued for a Very Long Time, in the face of some reasonably and readily available evidence.

I fly an Airbus, and I don't understand all of this fascination with the side sticks. The way it works in the airplane is actually a good system, it has warnings ( visual and aural) if both sticks are moved at the same time, and it has a lock out function.

Remember the Egypt Air plane that crashed in the Atlantic? The PF flew it into the ocean while the PNF fought him on the control yoke. A conventional system yet because the PF held it and the PNF couldn't overcome him.

Had the side sticks been interconnected do we think the outcome would have been different? The PF (AF447) was determined to keep the control stick full aft.

Interesting observations, but if you actually read what I wrote above, you'd get somewhere. The side sticks are not, in and of themselves, bad, and (like I noted) the entire Airbus FBW fleet could have lived out its existence without this issue ever coming up (and with a little luck, and especially with the knowledge of this occurrence and its contributing elements, nothing exactly like this will ever again occur).

But, seriously, you're going to analogize a suicidal pilot, intent on crashing the plane? In EgyptAir, both pilots knew very well what was going on, and it was a battle to the death.

Even if there is no system change I'll bet there will, or already is, a training change, even if it's only awareness. As someone else pointed out, it took a while, but the stall demos were revised.

Exactly.

My point, again, is not that the Airbus system is *bad* (and I made a point of that, didn't I?), but rather that, from a human factors point of view, a big, honkin' yoke in your gut is a much more obvious indicator to the other guy that the airplane is being commanded to pitch up to greatest extent possible. Would it have made a difference? Dunno. Hope we never have to find out again.

 

[Matthew]

My point, again, is not that the Airbus system is *bad* (and I made a point of that, didn't I?), but rather that, from a human factors point of view, a big, honkin' yoke in your gut is a much more obvious indicator to the other guy that the airplane is being commanded to pitch up to greatest extent possible. Would it have made a difference? Dunno. Hope we never have to find out again.

Not focusing on Airbus here, this is more generic:

It's not always a question of is it "bad", many times it's a question of "can it be better". Eventually you will get to the point where it's good enough, or the it's reached the point of not being economically feasible - but even then, somebody or somebodies has to define the limits.

It really is a lessons-learned topic that shouldn't just stop at pilot error.

 

[Silvaire]

...Way up in the thread someone posted that the 777 has yokes that work together, not because they are connected mechanically but because of computer design. Someone made this decision, not because it was cheaper, but because pilots wanted it that way, according to the poster...

What I said was that they moved in accordance with the flight controls because they are driven by the backdrive actuators but you can push the Captian's yoke full forward and pull the F/O's yoke full back and I believe the result is the same as with the Airbus. It's still a Fly By Wire system. There is no need for a Dual Control annunciation or warning message because it would be visually obvious and would require probably 20 lbs or more force by each pilot to accomplish.

I think the only difference between the Boeing and Airbus is that the Airbus stick (and rudder pedals) have no force feedback, they are just electronic controls. The Boeing yoke is an electronic control as well but it has an artificial force feedback system.

 

[Everskyward]

What I said was that they moved in accordance with the flight controls because they are driven by the backdrive actuators but you can push the Captian's yoke full forward and pull the F/O's yoke full back and I believe the result is the same as with the Airbus. It's still a Fly By Wire system. There is no need for a Dual Control annunciation or warning message because it would be visually obvious and would require probably 20 lbs or more force by each pilot to accomplish.

I think the only difference between the Boeing and Airbus is that the Airbus stick (and rudder pedals) have no force feedback, they are just electronic controls. The Boeing yoke is an electronic control as well but it has an artificial force feedback system.

Thanks for the correction. But what is a "backdrive actuator"?

 

[RotorAndWing]

But, seriously, you're going to analogize a suicidal pilot, intent on crashing the plane? In EgyptAir, both pilots knew very well what was going on, and it was a battle to the death.

Yes! seriously.

I flight instructed in airplanes and helicopters. A few times I had students lock up on the controls and had to fight the controls away from them, even though we thoroughly briefed transfer of control.

The worst was a female student in an Apache. We shut down an engine and she began putting back pressure on the yoke. I was trying to take control and keep her from pulling back further. She was 98 pounds wet, but yet it was a struggle keeping the nose down. Why she felt she had to pull so hard, and why she locked up, I don't know.

 

[docmirror]

Piper Apache <> Airbus A3XX.

At least, that's what I have been reliably informed previously.

 

[RotorAndWing]

Piper Apache <> Airbus A3XX.

At least, that's what I have been reliably informed previously.

Bottom line, no matter what I write, you are going to find fault with it.

But I guess you have to entertain yourself somehow.

Please continue to tell us how you are an expert on an airplane you know nothing about.

 

[Silvaire]

Thanks for the correction. But what is a "backdrive actuator"?

They are actuators under the cockpit floor that move the rudder pedals and yoke to correspond with the positions of the flight control surfaces. If you turn the captains control wheel full left the F/O's wheel will also turn left but you can hold the captains wheel full left and forceably turn the F/O's wheel full right with some effort.

I think the design of these controls (Airbus included) were never undertaken with a consideration that "hey one guy might be pulling and the other guy pushing" outside of the notion that a simple audible and visual annunciation would solve the problem. The 777 backdrive system certainly wasn't designed to prevent such a thing, it was simply meant to have the FBW system mimic the traditional systems that preceded it.

 

[RotorAndWing]

If you turn the captains control wheel full left the F/O's wheel will also turn left but you can hold the captains wheel full left and forceably turn the F/O's wheel full right with some effort.

The B727/737/747/757/767 all do this in roll. You can hold one yoke while over riding the other yoke in the opposite direction.

 

[ColoPilot]

I think the design of these controls (Airbus included) were never undertaken with a consideration that "hey one guy might be pulling and the other guy pushing" outside of the notion that a simple audible and visual annunciation would solve the problem. The 777 backdrive system certainly wasn't designed to prevent such a thing, it was simply meant to have the FBW system mimic the traditional systems that preceded it.

Even a non-FBW aircraft would not prevent the push/pull scenario -- on the Egypt Air one pilot was pushing, one pilot was pulling and, as designed, the columns split. The right elevator follows the right column and left elevator follows the left column. I'm not sure that is any better than summing the inputs from the 2 controls as is done on some FBW aircraft.

 

[Norman]

This discussion has become about as productive as the discussion of MH 370's disappearance on PPRUNE. 11,500 useless posts and counting.

 

[nauga]

I think the design of these controls (Airbus included) were never undertaken with a consideration that "hey one guy might be pulling and the other guy pushing" outside of the notion that a simple audible and visual annunciation would solve the problem.

They all kind of are, in that the systems have to be capable of independent operation if one side jams. Different manufacturers have different ways of managing it, but whoever said that opposing inputs are a possibility even in non-FBW airplanes was absolutely correct.

As for the cue of a yoke in your lap, we are in agreement that it's clearer than decoupled sticks. Where me might disagree is whether that was necessary and sufficient to break through the fog in the cockpit. We seem to agree that it's clearly NOT sufficient in some cases (Colgan), so apparently it's a matter of degree.

One thing I'm fascinated by is the focus on this issue, both in PoA and elsewhere including professional circles, with an apparent lack of concern for some of the other glaring issues put forth in the final report. It's true that breaking even the smallest link in the accident chain can avert an accident, but to focus on that small link and leave other huge and easily broken links untouched, or to berate people for touching them, seems short sighted.

Nauga,
who reads and heeds

 

[Hello]

And while I'm on a roll, what about that Airbus that was on some kind of victory tour right after it was unveiled and crashed on a low pass I think? I'm not sure on the details, but as I recall, a perfectly operating and perfectly designed plane did exactly what the pilot told it to do - right up to the point where it didn't. Something about the engines not responding to throttle inputs? No - I'm sure the pilots were wrong in that too, cuz who would think that pushing the throttles all the way forward means 'more power Mr Scott!'

Wonder what happened to that little design perfection?

I've just registered on this forum to reply to this user.

First, if you are not sure about the details (you're wrong about everything, but anyway), use the wonders of the internet to enlighten you.

It's amazing how a person who is somehow related to aviation can display such extreme ignorance about this accident

The Air France Flight 296 was 100% pilots' fault, period.

Instead of making a pass at 100 feet like it was planned, they botched and did it at 30 feet instead, and when they reached the end of the runway and realized their mistake, applied TOGA and tried to climb, but the elevators didn't respond, because alpha prot had been automatically engaged to avoid a stall. So the a320 continued flying straight, ingested leaves and crashed in a fairly controlled manner (only 3 people died)

Had it been an airplane without flight envelope protection, it would have stalled, crashed and possibly killed everyone on board.

Do you get it now?

Best regards.