I did this flight today, to run-in a repaired cylinder and generally remain within glide range of the airport for the first hour (FR24)
A close-up from the GPS log:
I have been told so many times that if you do this, gyros will eventually lose the roll angle accuracy and will erect to show zero roll. But
all just kept working for the whole hour and were all spot on.
The whole flight was done at 4300ft on autopilot, with the turns in ROL and ALT modes. As you can see there was NE wind; about 15kt.
It is defiantly a thing – I know certain aircraft in certain specialist roles that maintain continuous orbits have experienced the issue, but I believe it is after many hours in orbit and I suspect you weren’t long enough to see the effect.
Peter wrote:
I have been told so many times that if you do this, gyros will eventually lose the roll angle accuracy and will erect to show zero roll.
I was tempted to try the same one day ;-) I think the electronic AHRS solutions are different as they can use the magnetic field, detect the turn rate etc. to stabilize. Regarding the mechanic solutions if you fly in a circle will the gravity stabilization not cancel out the error? Probably flying for a long time with a side slip will disturb the gyro a lot more.
Because of the way a mechanical gyro works, its erection mechanism can indeed be fooled if a constant false vertical can be maintained long enough. However when flying in circles, the false vertical is not in a constant direction as the aircraft constantly changes heading and so does the direction of the false vertical. There is a marginal error during turns but it cancels itself every 360°.
Also sideslip won’t work either because although the direction of the vertical doesn’t change during a sideslip, the vertical is not a false vertical but the actual vertical so the gyro will keep showing the aircraft banked because it actually is.
Impressive TB20 thermaling 
I believe those Rate 1 360 turns on AP will produce few gyros issues? compared to say 60 degrees turns on hand flying that may topple the gyro and force its re-alignment?
I think my brain might have toppled after that many orbits……
Nooo; I was taking sunset pics, while constantly watching the GPS to make sure I didn’t end up in the dungeon at Gatwick. I have zero lives left now
Thanks Alex for your explanation. I can’t say I understand it, but it sounds like you do and that’s good enough for me When at univ, I went to a fascinating presentation by Eric Laihwaite on gyros and the whole subject was a mystery to me.
Spinning objects are more weird than most people think. Even NASA found this out the hard way
It makes you wonder what would happen if you were using a rate based a/p to do this.
