Air France Jet Missing Off Brazil Coast

[humble]

Uhmm, humble, without fly-by-wire, the F-16 would have never gotten off the ground.  FBW can do a lot of things for the pilot to ease the load of flight.  The F-16 is a shining example of FBW done right, in my opinion.  In the 30 years of F-16 production, not one aircraft has been lost due to anything related to FBW.  Make no mistake about it, FBW in the F-16 is keeping the pilot alive.

The throttle on most cars today is FBW.  I have a few hundred thousand miles on those systems myself and have never had a failure of any kind.  There has never been a problem at all, as a matter of fact.

Stating all FBW is bad is like saying all cars are bad because of the Yugo.

Thats exactly what I was trying to say here Skuzzy...
FBW was originally designed to enable military pilots to fly airplanes that might not be flyable with "normal" controls. The computer "translates" normal control inputs into more complex ones required to fly the aircraft. For many cutting edge planes FBW is the only way to fly.

[Skuzzy]

But you also said the following.

My "issues" are simple. I'm a firm believer in the KISS principle and complexity for marketing value is a bad idea. The mere fact that you need 3-4 backup systems is a sure indication your heading down a bad rode IMO. Imagine if all the cars on the road had FBW instead of steering linkages.

This seems to imply you think FBW is just a bad idea all around as it violates the KISS principle and you do not think it can handle simple steering chores in a car, yet they have no problem handling the throttle, which is a more complex system.

Look at my car.  It has no air meter.  The throttle pedal activates some type of variable resistor, sending its values to the computer.  The computer takes those values and adjusts the air flow by altering the amount of valve lift and duration, then also adjusts the duration of the direct injector firing time.  It has to do all these things thousands of times a second and not miss a beat or problems could arise.

Steering would be done as follows.  The computer reads the resistor value inputs from the steering wheel, then outputs the pulses to a servo control motor to activate the steering mechanism.  Now, do you trust a servo control motor anymore or less than the power steering pump?

I am just playing devils advocate to your analogy.  I realize it could have been a different one.

I may be wrong on this, but seems to me a good FBW system, in an airplane, is probably the simplest system you can have in terms of number of components and weight.  I see no violation of the KISS principal.

[Golfer]

I have 1000 hours in FBW airplanes and the redundancy built into the systems had me more than satisfied to feel completely safe during those operations.  Including traversing the Amazon, over water and in the best/worst of weather.

To lose all control capability requires such a huge number of things to go wrong that you're screwed long before you get to the fact your can't control the airplane.

Before we get all hussy over the fly-by-wire aspect that enough here clearly do not understand don't forget that we don't know anything about this accident beyond the airplane is in the ocean.  FWIW the only flight control failure I've exerpienced came in an aircraft with cables, pulleys and pushrods.  A rudder cable pulley decided to throw in the towel which threw the rudder and aileron controls out of whack due to an interconnect system with the airplane.  This resulted in an uncommanded aileron input and a significant roll rate with a very high force required to counter it.  After verifying the ailerons were in fact still attached and moving freely using full trim into the roll lessened the force required to maintain wings level.


No such thing has/could/would happen with a fly by wire airplane.  In my case there was triple redundancy of the hydraulic systems, the redundancy of the computers as to which electrical and hydraulic system power which actuators including the RAT (Ram Air Turbine) should your emergency degrade to that point you still were afforded with workable flight controls.  The airplane was very much overengineered (by Boeing engineers who in effect built the airplane like an Airbus no less) and I found it to be reliable while maintaining a very comfortable level of redundancy for the unfortunate eventualities.

[humble]

Humble, just because I am interested and you said both planeset have their issues and can be complex and unforgiving. What do you Boeing's issues are or what could they improve? It has been an interesting discussion on FBW, composites and resins so far .. so I am just curious as to where you can see Boeing has issues.

The reality is that most of what Airbus is pioneering is customer (airline) driven, both in regard to safety and operational affordability. I have no ability to discuss the economic metrics but the flight safety issues are readily observable. Lets look at the two recent crashes (yes the -8 is not a Boeing plane). Both the Buffalo crash and the turkish 737 are a combination of pilot workload, inexperience and potential issues (mechanical, icing etc). The end result is that the pilots action and inaction directly contributed to the crash. As mentioned above somewhere "pilot error" is an ever increasing factor in a majority of air frame losses. The 737's lack of an "alpha floor" and auto throttle clearly contributed to the crash. The colgan FO's pitch setting contributed heavily to the buffalo crash. My understanding is that in normal mode the avionics in the airbus series would have prevented both incidents. So we have the very well intentioned efforts to protect airlines from both the deterioration in experience and training in the potential new hire pilot pool and the reality that ever increasing congestion creates a tremendous pilot workload and creates the potential for a pilot to get behind the plane for numerous reasons.

My "issue" here is the combination of technology in design and concept that attempts to deal with real issues via ever more complex solutions. Pilots rely on avionics in both plane sets way to much. Boeing (correctly IMO) draws the final line of responsibility clearly with the pilot. However we can see that this choice does contribute indirectly to crashes.

[Golfer]

The colgan FO's pitch setting contributed heavily to the buffalo crash.

The what?  You really don't have any informed clue about that accident.  I could go on for pages about it but suffice to say the blame for the Colgan accident lies squarely on the shoulders of a blabby captain who through the course of the flight caused the crew to be behind in their housekeeping.

[Die Hard]

Humble, unlike the F-16 the Airbus' are not inherently unstable designs. In the event of a total electrics failure an Airbus will still fly along quite happily... Of course, it will fly where it wants to go rather than where the pilots would like, but that's no different than in a conventionally controlled airliner. Large airliners need servo power to move the control surfaces, and the electrics power the hydraulics. Wether fly by wire or not you need power to operate the control surfaces. In a conventional aircraft you push the stick and it pushes a hydraulic cylinder which moves the surface; in a FBW aircraft you move a small computer control that tells the computer to move the hydraulics. In either case without power to the hydraulics the surface isn't going to move. To move the surfaces on a 747 with old fashioned wires and the strength of the pilot, as in a Cesna, would require a tug-of-war team on the flight deck.

The chance of a total loss of electric or computers is infinitesimal compared to the chance of pilot error. The NTSB reckons that half the accidents in civil aviation are caused by human error... Half! The computers not only adds a layer of protection against common pilot errors, but also aid the pilots in controlling the aircraft under unusual circumstances such as engine failures and partially or even complete hydraulic failures where the computers will help the pilots control the aircraft with engines and whatever controls remain.

It is my opinion that, even if the computers add complexity and probably will lead to some equipment related accidents, they are worth it for the lives they save by preventing human errors and helping the pilots overcome (even previously hopeless) problems.

http://www.youtube.com/watch?v=xPt0rMVbTmk

http://www.youtube.com/watch?v=s-zqo2QpIC0

[humble]

"Fly by Wire" is not driven by the replacement of mechanical linkages with electronic ones (although that is a potential issue) but by the interrelationship between the control surface, control input and the software itself.

[68Wooley]

...Boeing (correctly IMO) draws the final line of responsibility clearly with the pilot...

Out of curiosity, is this the case with Boeing's FBY models? Or does the system in the 777 have the authority to override the pilot in a similar maner to an Airbus?

[thrila]

Some details leading up to the crash have been revealed:

http://www.telegraph.co.uk/news/worldnews/europe/france/5444168/Air-France-pilots-battled-for-15-minutes-to-save-doomed-flight-AF-447.html

[Die Hard]

Out of curiosity, is this the case with Boeing's FBY models? Or does the system in the 777 have the authority to override the pilot in a similar maner to an Airbus?

The computer in the Airbus do not have authority over the pilots unless the pilots so chose. The pilots can set the computers to give them full control; if a pilots wants to he can still over stress an Airbus.

[Die Hard]

Some details leading up to the crash have been revealed:

http://www.telegraph.co.uk/news/worldnews/europe/france/5444168/Air-France-pilots-battled-for-15-minutes-to-save-doomed-flight-AF-447.html

I'm not so sure I believe the "source". The pilots would have reported problems if they were having cascading systems failures over a period of 15 minutes.

[humble]

The computer in the Airbus do not have authority over the pilots unless the pilots so chose. The pilots can set the computers to give them full control; if a pilots wants to he can still over stress an Airbus.

Not true, the only time a pilot has full authority would be in direct law, this is far beyond "pilot choice"....

There is no pilot option to select direct law in any airbus that I am aware of...

[humble]

The what?  You really don't have any informed clue about that accident.  I could go on for pages about it but suffice to say the blame for the Colgan accident lies squarely on the shoulders of a blabby captain who through the course of the flight caused the crew to be behind in their housekeeping.

A lot of factors contributed and in the end the decision to override the stick pusher is baffling. Neither pilot was really qualified to fly a commerical airliner of any flavor IMO. However the decision to set the control lever to max and "flat plane" the props is highly questionable given the need to keep the plane fast. As it relates to my earlier comments the entire point of the airbus software is to preclude pilot induced upset or to override or ignore pilot input and if needed respond to a lack of input....

[humble]

It is my opinion that, even if the computers add complexity and probably will lead to some equipment related accidents, they are worth it for the lives they save by preventing human errors and helping the pilots overcome (even previously hopeless) problems.

The problem here is that your incorrect, in an emergency the avionics in the airbus provide no help and might make an upset more likely or recovery from an upset impossible. We certainly do not know what happened but I think its very probable that "normal flight law" might have contributed to creating the upset and that alternate flight law may have precluded recovery...thats not coming from me its coming from guys who fly the 330 for a living...

[Golfer]

Reading PPRuNe doesn't make you a defacto expert.  Being able to copy and paste someones opinion doesn't qualify it as your experience.

That said regarding the Buffalo crash I still don't follow what you're trying to say with "the FOs Pitch Setting"

Outside the recline of her seat there's nothing that makes sense.  The condition, propeller and power levers all belong to the flying pilot.  It is standard to increase the propellers to full rpm for landing however those engines require a power increase because without the extra torque you just put two efficient drag discs onto your airplane with them in full rpm.

Humble your talking about the various control laws of the Airbus (which you appear not to have a thorough understanding of nor any experience with) is akin to discussion the turn radius of the Zero over a Spitfire during the aftermath of the Pearl Harbor attack.