Originally Posted by Lookleft
KC-135 Kyrgyzstan and see that the Air Force also has a problem with over reliance on automation. Due to flightcontrol problems the aircraft developed Dutch Roll and it broke up in flight. The report suggested that the crew might have been able to recover the aircraft if they had taken manual control but they instead had attempted to engage the autopilot.
So its not just a civilian problem. It has a lot to do with reducing training to the absolute minimum and hoping that the only non-normals experienced by a crew fit neatly into the QRH and ECAM/EICAS.
Thank you Lookleft for that worthful reference. It shows too that never described strong forces build a glued 3D net around the plane during developped dutch roll, not explained to pilots, AND not understood by classical aerodynamic rules who matters only about airflow very near of the plane.
It is how successive airfoils wristed around the plane and around oneanother with different AoAs.
Systems and autopilots ignore that modell, so "automation" (what I used to call "butterfly" from the fractal equations building lift and drag) cannot bring a good response to that degenerating system, if time delays and sensitivity of the FCS are too low (and too cheap, of course). A good handflying pilot with aerobatics training (fighter pilots) is better able to feel more quickly the limit of acceptable actions.
Developped dutch roll is the easiest PIO to overcome because degree of equations are the lowest, but that degree increases with bad reactions.
We know that PIO often develop in FBW rate limitation resulting from digitalizing vs analogic solutions like Concorde (Mc Ruer and others).
Automation is like a drug near the airports : folks buy houses near the airport because they are cheaper, and then complain about noise to get money. Airports ask greater precision in lateral path, airlines request more automation during departure and arrival.... There is no limit ! At some moment you get a safety saturation.