Airbus - A310 Crash

[flyingcheesehead]

In a "normal" airplane you have cables and other mechanisms between you and the control surfaces too and those things can break or jam.

True, but you can at least feel them and maybe revert to trim-based control (in reverse, in many cases) rather than sit there and wonder "What the heck is it doing now?"

I guess I'm just a neanderthal pilot.

 

[gismo]

Surely - But Air France, Airbus, and the French government had a lot more to lose than this pilot did and they were surely in full CYA mode as well.

I think that goes without saying. In any major crash most of the players are looking to deflect as much blame as they can.

I still don't like the design philosophy. Some things should be the same on every single airplane: When I push the throttles forward the plane should respond, and the stick should function the same in any airplane. IMHO, I as a pilot should be able to jump into the cockpit of any airplane and at least have a fighting chance of landing the sucker safely without having to be trained on how the plane will decide to respond to my inputs.

I think that's very unrealistic. And particularly WRT jet engines lack of ability to produce power rapidly from idle.

Greg, I have no problem with FBW - I know Boeing's using it too. But, I'm curious - On the 777, does it have the various normal law/alternate law modes, or does it simply respond the way a regular airplane would? I think that people tend to confuse "fly by wire" with the operating logic in the Airbus system, when in reality they're two separate things... Kinda like someone saying "Intel sucks" when their USB cable doesn't work.

It seems extremely likely that Boeing and anyone else using FBW technology would include both revisionary modes ("laws") and some level of airframe protection. It appears they do lean more to the notion that the pilots in the cockpit may be better judges of what's needed than the engineers and their autopilot's programs when something unusual occurs, but there's definitely two sides to that story.

 

[ScottM]

It seems extremely likely that Boeing and anyone else using FBW technology would include both revisionary modes ("laws") and some level of airframe protection. It appears they do lean more to the notion that the pilots in the cockpit may be better judges of what's needed than the engineers and their autopilot's programs when something unusual occurs, but there's definitely two sides to that story.

You mean like the 737 drivers in the UK that when faced with an impending engine failure decided to shut down the GOOD engine by mistake? When they changed the speed on the engine that was failing, stopping the vibrational mode and they thought they were in the clear until a short time later when the engine finally self destructed.

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

This all came down to NOT paying attention to instruments and instead too much on seat of the pants experience.

 

[gismo]

Well, I haven't seen any of that. It brings to mind a saying that I heard from my dad. "Don't believe anything you hear and only half of what you see."

I might be missing something here but it seems to me that once anyone begins to examine the contents of the "black boxes" the protective covers would be removed to gain access. And since the covers themselves contain little if any useful information, it wouldn't surprise me if whatever agency was examining the devices substituted newer covers when repackaging them after downloading the information. Finally with a digital recorder I'd expect that it would be rather trivial to modify the stored data without substituting a different physical unit.

It's certainly possible that someone tampered with the data but I don't see much credibility in the concept that the change in physical appearance has anything to do with such tampering.

 

[Everskyward]

True, but you can at least feel them

You can feel they are jammed but that doesn't help you much. Like Matt I have done this in the sim. I don't have any clue what the Airbus solution is but in the CE-680 you pull a lever which disconnects the two yokes so that the left side controls the ailerons and the left elevator and the right side controls the roll spoilers and the right elevator. You figure out what part is jammed and reconnect the other part. Then the person without the jammed control flies. That's how is supposed to work, theoretically anyway.

and maybe revert to trim-based control (in reverse, in many cases) rather than sit there and wonder "What the heck is it doing now?"

Would it bother you to know that many airplanes don't have a manual trim wheel? It's all electric. Even as old a design as the Lear 35 has no manual trim. There two different trim systems, primary and secondary, but they are both electric. The CE-680 is the same way.

All that logic in the Airbus and other modern airplanes is there to help protect the airplane and the people in it and to make it more useful. It's just an evolution from the pushers, pullers and autopilots of the past.

 

[jesse]

The thing to remember is that there is a critical difference between a "fly-by-wire" system that just obeys the pilots control inputs and attempts to apply them to the physical controls -- and a "fly-by-wire" system that will over-ride the pilot at a specific pitch, bank, or AoA.

 

[Everskyward]

The thing to remember is that there is a critical difference between a "fly-by-wire" system that just obeys the pilots control inputs and attempts to apply them to the physical controls -- and a "fly-by-wire" system that will over-ride the pilot at a specific pitch, bank, or AoA.

But remember that even much older airplanes which you couldn't remotely call "fly-by-wire" had pushers to help prevent stalls and pullers to help prevent overspeeds. I just see the more modern systems as an extension of these protections.

 

[flyingcheesehead]

I think that goes without saying. In any major crash most of the players are looking to deflect as much blame as they can.

True, but this one had a lot more at stake - Had the crash been blamed on the FBW control logic, it's quite possible that the whole A320 project would have failed, and probably Airbus along with it, leading to billions in losses (not to mention national pride, jobs, etc.) for the French and other EU countries invested in Airbus.

I think that's very unrealistic. And particularly WRT jet engines lack of ability to produce power rapidly from idle.

Like I said - All I want is a fighting chance. Didn't say I'd have to make it 100%.

It seems extremely likely that Boeing and anyone else using FBW technology would include both revisionary modes ("laws") and some level of airframe protection. It appears they do lean more to the notion that the pilots in the cockpit may be better judges of what's needed than the engineers and their autopilot's programs when something unusual occurs, but there's definitely two sides to that story.

Yeah, it seems the Sullys of the world will know their aircraft well enough to make it do what they want it to do in spite of itself, while the Renslows of the world make Airbus look really smart (and pilots, not so much).

 

[flyingcheesehead]

This all came down to NOT paying attention to instruments and instead too much on seat of the pants experience.

Not really. "Dead foot, dead engine" is as seat-of-the-pants as it gets, and works every time.

I clicked a few links in the Greenspun site someone linked to about the Avidyne vs. Garmin glass, one of which was to an article about the Citation Mustang and how you don't really need to "identify" on it, you simply keep moving both power levers as if both engines were working. Cool. Funny how I don't like a computer messing with my flight controls, but I think FADEC rocks.

 

[ScottM]

Not really. "Dead foot, dead engine" is as seat-of-the-pants as it gets, and works every time.

Not really? In this case, are you serious? I don't agree at all. There feet had nothing to do with this. I am not even sure where you are going with that comment as it pertain to the accident we are discussing. [And please use that whole sentence before you go off on a half knicker diatribe on what to do in a twin when an engine goes out I know what you were saying. It just is meaningless in the discussion we are having]

They shut down the engine based on a smell. The idea being that smoke would come from the right engine because that was what they new on the older model was were the ac bleed air came from. The model they were in AC came form both engines and the smell ending had nothing to do with their actions.

This was a screw up based on them not analyzing the instruments which include a vibrational detector. The automation was working, they chose not to use it.

I think we get it though Kent. You have an issue with AirBuses and facts be damned, you are going to keep not liking them. C'est La Vie.

 

[flyingcheesehead]

You can feel they are jammed but that doesn't help you much.

Disagree - Knowledge is power! It allows you to assess what your next step will be instead of just fighting the jam. Were I to have the elevator jam in the 182, I know that I can still control it somewhat via trim (turning the wheel in the "wrong" direction - Trim down to pitch up). Had the captain of American 191 known what was going on, they might have had a chance to recover (slim, yes, but a chance). From Wikipedia: "In subsequent flight simulation testing, with all known collateral failures included in the simulation, only pilots who were aware of Flight 191's specific problems were able to recover successfully."

Like Matt I have done this in the sim. I don't have any clue what the Airbus solution is but in the CE-680 you pull a lever which disconnects the two yokes so that the left side controls the ailerons and the left elevator and the right side controls the roll spoilers and the right elevator. You figure out what part is jammed and reconnect the other part. Then the person without the jammed control flies. That's how is supposed to work, theoretically anyway.

Sounds like a very smart design to me!

Would it bother you to know that many airplanes don't have a manual trim wheel? It's all electric.

No - As long as the trim obeys what the pilot tells it to do. There are plenty of GA airplanes with no trim wheel these days. DA20, Cirrus, Sport Cub, etc.

 

[ScottM]

Knowledge is power!

.

Only if you use it. The hardest part of getting information to the user is getting useful information to the user in a format that they can quickly make and informed decision. In the case of flight 191 just having some sort of warning of what was happening with the control surfaces would not have been enough. They would need to filter through all the other alarms they would be getting and then process an action. They did not have enough time EVEN if they had gotten that info. Don't think I have not thought a lot of that accident. I knew people who died in it.

The question then is how do you get the right info, at the right time, in the right format to a decision maker? Many times you have to automate based on a set of rules as people become the unreliable part of the equation. Not always, but many times that holds true.

With about 95% of crashes being pilot error one has to look for ways to get the pilots to stop making those errors. There is not one solution but several. Some of which involve training, some are automation of tasks, and some are human factor integration.

 

[Everskyward]

Not really. "Dead foot, dead engine" is as seat-of-the-pants as it gets, and works every time.

Not necessarily. In bigger airplanes you have more systems which try to help you which can also confuse you if you depend strictly on feel. Many airplanes have rudder boost or rudder bias or something similar which kicks in when one engine fails. Depending on how and when it kicks in it can fool you if you are only going by the feel of the rudder pedals. In turbine airplanes I was taught to look at the engine instruments and other indications to determine which engine failed, which is contrary to what I learned in piston airplanes. Remember also that bigger prop planes have autofeather and jets don't have anything to feather so you can take a decent amount of time deciding which is the correct one to shut down. We are supposed to climb to a safe altitude (at least 1,500' AGL) before touching anything except for silencing the warnings.

Funny how I don't like a computer messing with my flight controls, but I think FADEC rocks.

Yeah that is funny. You don't care that the only thing connecting the throttles to the engines is wire and that a computer program is telling the engines what to do.

 

[Everskyward]

Had the captain of American 191 known what was going on, they might have had a chance to recover (slim, yes, but a chance). From Wikipedia: "In subsequent flight simulation testing, with all known collateral failures included in the simulation, only pilots who were aware of Flight 191's specific problems were able to recover successfully."

Maybe the sophisticated airplanes of the future will have a complex computer program to figure that all out and tell you what you should do in this situation. Or maybe even do it for you.