The milling behaviour will depend on the material.

Polycrystalline materials which do not amorphize under ion bombardment tend to develop structures (called curtaining effect or waterfall effect).

https://www.globalsino.com/EM/page4499.html

https://www.globalsino.com/EM/page4502.html

For a lamellar microstructure the orientation of the lamellae in the sample might affect the development of curtaining.

Here, different orientations of the lamella could be checked (parallel, perpendicular, 45°?)

Are there publications about similar materials? PhD thesis often contain more experimental details than papers.

My experience with Ga beams:

0) Make sure that the beams are well adjusted (focussed and alignment of different currents)

1) Use a deposit layer (Pt, C) of sufficient thickness.

2) Decreasing the beam current will make sharper edges (beam diameter, opening angle), but milling time will increase.

3) Development of curtaining can be reduced by:

- Rocking Beam (manual, see: http://dx.doi.org/10.1016/j.ultramic.2015.11.001, or by stage, if available)

- beam parameters (voltage, current)

- milling with precursor, if available (CopperRx, Nanoflat, A-Maze)