Originally Posted by
pptp
When we say non-fly-by-wire, or mechanically controlled, we are generally talking about the control surfaces being connected in some way to the control inputs. On older planes, and many newer planes, there is a connection via cables or other mechanical means and the input is augmented with hydraulics, airflow or other methods to make flying the plane less fatiguing.
With some caveats, Boeing uses the above method. Airbus uses fly-by-wire which means the the control inputs are read by a computer and translated into appropriate moment of the control surfaces with zero physical connection. Although time has shown that both methods are incredibly safe, some feel that there can be something "lost in translation" between what the pilot wants the plane to do and what the computer thinks he wants the plane to do.
If there is catastrophic system failure, the Airbus HAS to have some sort of backup power/computer system to operate the plane. With a non fly by wire, theoretically you can have a completely dead plane and still fly it, albeit at times with some great effort.
Anyone feel free to correct me if I got anything wrong.
FBW has absolutely proven itself to be safe, no question. One of the principles behind FBW is that the computer will generally not allow human error to create a catastrophic situation (i.e. force the plane into a stall).
Aviation writer William Langewiesche wrote a book called "Fly by Wire" in which he makes the argument that the FBW technology on the Airbus Captain Sullenberger ditched in the Hudson played a central role in the crash landing being so successful.
On the other hand, FBW with no mechanical backup requires some power to control the plane.
There is also the issue the the computer could actually
cause a crash. Although pure speculation, some have argued that the crash of the AF Airbus in June 2009, might have been caused, at least in part, by FBW.
Of the information that is known, it appears clear that the aircraft was receiving inconsistent air speed measurements. There was an identified problem with the pitot tube that measures airspeed. In this case, however, the pitot tubes might have simply iced up, thus rendering them ineffective as well.
But on the Airbus, inaccurate air speed measurements, espeicially if combined with strong winds, could lead the computer to mistakenly send the plane into a stall, if the computer believed the plane was reaching its maximum allowable speed. In other words, the computer mistakenly believes the plane is flying too fast (when it is not) because of faulty sensors, and slows it down so much that it goes into a stall (meaning it drops out of the sky).
Once the Airbus entered the stall, it might have been impossible for the pilots to take any kind of effective action to get it out of the stall, and the computer might have forced it to remain in the stall if it still believed the plane was exceeding its maximum speed.
No one knows this for certain, but it is not an inconceivable scenario in a FBW aircraft.