Originally Posted by
john_tullamarine
real A/C will have a tendency to pitch continuously in cruise with A/P off,
you are probably thinking of the phugoid. Bit of a roller coaster motion which is only a minor nuisance to the pilot. With the A/P height lock in the phugoid degenerates to a similar period speed oscillation unless the A/T takes care of it. Some aircraft have a pronounced phugoid, others are quite benign.
Actually, no. The description of what I reference is hidden somewhere deep in the AF447 thread.
Basically, AB FBW uses a "1g" reference for pitch control and any forward/back stick movement is then translated as "load factor" demand. Problem is that the normal acceleration pertaining to level flight is not of a constant value. When stationary on ground, one gets a constant gravity force opposed by constant centrifugal force (Earth's rotation) and normal acceleration is determined. In altitude, the gravity force is different (smaller) compared to surface value due to increased radius. Any forward tangential velocity in the direction of Earth’s rotation will increase the perceived orbiting speed and centrifugal force. Extreme case is high altitude flight over equator on 090 true track (yes, very similar to AF447). It all boils down to one thing - aircraft probably has s hard-wired "level flight 1g" constant but at level flight under described conditions the real g value is less. As the aircraft measures that g is less than expected for level flight, the FBW will attempt to compensate and pull up to achieve the hard wired value.
My small trial (of course properly approved by ATC in a NON RVSM airspace) was in a A320 cca 30°N on an easterly track. At 300 ft above assigned level -500 fpm trajectory was manually commanded and stick released. Expected behaviour of the FBW would be to keep the trajectory, however
probably due to logic described, the FBW performed a constant positive load manoeuvre with trajectory low at 100 ft below assigned FL and within 30 sec from stick release we hit 300 ft above assigned again, this time with +700 fpm. My observation is, that in high altitude and easterly track a small constant forward input on stick is required to maintain level flight.
I wonder if this behaviour is available from some newish simulators or not, ours certainly cannot replicate it.
FD (the un-real)