In the midst of his discussion with Owain Glyndwr, Concours77 has quoted a paragraph I wrote in
post #1359:
"The function of any auto-trim system is to remove the necessity for the PF or AP to continue to command the deflection of a primary flight control. In this case the auto-trim is concerned with pitch control, and adjusts the incidence (angle) of the THS (trim-able horizontal stabiliser) to enable the elevators to return to neutral in a few seconds after they have been deflected to comply with a command from the pilot or AP for a pitch-change (actually, in the case of C*, a change in normal G)."
That was a half-decent explanation of auto-trim, mainly for the benefit of pilots who haven't used it. But perhaps it can be improved in relation to Airbus FBW aeroplanes like the A330. It may also be helpful to reflect that auto-trim is merely an automated version of the various trimming systems with which virtually all aeroplanes have been fitted since the early days of aviation.
So here goes:
The function of a flight-control trimming system is to remove the necessity of a PF or AP to apply a continuing command to deflect a primary flight control in one direction. Pitch trim has been achieved traditionally by a hinged trim-tab which forms part of the trailing edge of an elevator. The pilot can adjust the position of the tab, which then exerts an aerodynamic force on the elevator as a substitute for the force previously exerted by the pilot's elevator control, which will then maintain the desired position "hands-off." Most large jet transports use a different system, in which changes of the angle of incidence of the "trimmable" horizontal stabiliser (THS) allow the elevators to return to neutral. On aircraft with conventional flight-control systems the elevators are controlled by the PF or the AP. In manual flight the PF uses a trim wheel or piccolo switches to move the THS, but with the AP engaged an auto-trim system moves it. In Airbus FBW with the AP disengaged the EFCS interprets the PF's sidestick position according to C* logic, which at cruise speed effectively means normal acceleration, and any resulting displacement of the elevators is reduced in due course by the auto-trim slowly moving the THS in the appropriate direction.
Please don't hesitate to inform me of any errors or possible improvements.