precession error for artificial horizon?

[wolle]

I recently noticed that the artificial horizon can get out-of-whack after long periods of flying, i.e. the ground on the artificial horizon is at an angle with the real ground.

 

Is this due to precession errors (e.g. during sustained turns the g-vector does not point down any more, and after a while the artificial horizon adapts to this new "down" direction)?

 

Does anyone know: is the artificial horizon in the Huey susceptible to such precession errors, and is it modeled in DCS:UH-1?

[Cake]

For those who want to know more about precession in Attitude Indicators, see below from FAA Instrument Flying Handbook:

 

For an AI to function properly, the gyro must remain

vertically upright while the aircraft rolls and pitches around

it. The bearings in these instruments have a minimum of

friction; however, even this small amount places a restraint

on the gyro producing precession and causing the gyro to tilt.

To minimize this tilting, an erection mechanism inside the

instrument case applies a force any time the gyro tilts from

its vertical position. This force acts in such a way to return

the spinning wheel to its upright position.

 

The older artificial horizons were limited in the amount of

pitch or roll they could tolerate, normally about 60° in pitch

and 100° in roll. After either of these limits was exceeded,

the gyro housing contacted the gimbals, applying such a

precessing force that the gyro tumbled. Because of this

limitation, these instruments had a caging mechanism that

locked the gyro in its vertical position during any maneuvers

that exceeded the instrument limits. Newer instruments do

not have these restrictive tumble limits; therefore, they do

not have a caging mechanism.

 

When an aircraft engine is first started and pneumatic or electric

power is supplied to the instruments, the gyro is not erect. A

self-erecting mechanism inside the instrument actuated by the

force of gravity applies a precessing force, causing the gyro to

rise to its vertical position. This erection can take as long as 5

minutes, but is normally done within 2 to 3 minutes.

 

Attitude indicators are free from most errors, but depending

upon the speed with which the erection system functions,

there may be a slight nose-up indication during a rapid

acceleration and a nose-down indication during a rapid

deceleration. There is also a possibility of a small bank angle

and pitch error after a 180° turn. These inherent errors are

small and correct themselves within a minute or so after

returning to straight-and-level flight.

 

Note sure about how it's modeled in DCS. Does it correct itself after leveling out for a minute or two?