I have observed issues with the blade azimuth angle reported by OpenFAST V3.0.0. for the case when shaft tilt = 0, pre-cone angle = 0. The plots are attached herewith.
If my observation is valid, could the issue be due to Eulerian singularity (only a quick guess as of now)?
A similar question was discussed in the following forum topic: Azimuth Angle of the Blade. Basically, in AeroDyn v15, the local blade azimuth angle is defined based on how the rotor disk-averaged relative velocity projects onto the disk (based on the skew angle between the rotor axis and inflow) rather than some rigid-body rotation of the rotor. If you want the later, I would use the Azimuth output from ElastoDyn instead.