Finite Rotations with a Step (Abaqus)

[cmvoss]

Has anyone here worked with applying time-varying (amplitude driven) finite rotations within a step? I have an application where I rotate an analytical rigid surface through a motion cycle from its reference point. This motion cycle should bring it back to its starting point, but the final point is off slightly. Translation is fully fixed and there is a different sinusoidal amplitude used to drive UR1 and UR3. UR2 is fixed in the BCs, but does not come out fixed when you look at the output.

The discrepancy between the start and endpoint is not huge, but it will add up as I need to repeat this step 40 times. Any ideas what I'm misunderstanding?

 

[cmvoss]

Oops, I forgot there is a dedicated Abaqus forum. I'll move (well, repost) this there instead.