Dear Daniel,
Your questions seem to be related to the questions about AeroDyn’s blade-azimuth angle outputs that we discussed in a separate forum topic: Azimuth Angle of the Blade.
Again, AeroDyn’s blade-azimuth angle is defined relative to the inflow-skew angle. In Case 1, there is no inflow-skew angle, so the azimuth angle is undefined (and simply output as zero by AeroDyn). In Case 2, z_disk points to the right when looking downwind and so an azimuth of zero means that blade is pointing to the right. In Case 3, z_disk points to the left when looking downwind and so an azimuth angle of zero means that blade is pointing to the left. Your results seem consistent with this azimuth convention.
Best regards,