Development of a new, fully automated system for electron backscatter diffraction (EBSD)-based large volume three-dimensional microstructure mapping using serial sectioning by mechanical polishing, and its application to the analysis of special boundaries in 316L stainless steel
Abstract
We report the development of a fully automatic large-volume 3D electron backscatter diffraction (EBSD) system (ELAVO 3D), consisting of a scanning electron microscope (ZEISS crossbeam XB 1540) with a dedicated sample holder, an adapted polishing automaton (Saphir X-change, QATM), a collaborative robotic arm (Universal Robots UR5), and several in-house built devices. The whole system is orchestrated by an in-house designed software, which is also able to track the process and report errors. Except for the case of error, the system runs without any user interference. For the measurement of removal thickness, the samples are featured with markers put on the perpendicular lateral surface, cut by plasma focused ion beam (PFIB) milling. The individual effects of both 1 μm diamond suspension and oxide polishing suspension polishing were studied in detail. Coherent twin grain boundaries (GBs) were used as an internal standard to check the removal rates measured by the side markers. The two methods for Z-spacing measurements disagreed by about 10%, and the inaccurate calibration of the PFIB system was found to be the most probable reason for this discrepancy. The angular accuracy of the system was determined to be ~2.5°, which can be significantly improved with more accurate Z-spacing measurements. When reconstructed grain boundary meshes are sufficiently smoothed, an angular resolution of ±4° is achieved. In a 3D EBSD dataset of a size of 587 × 476 × 72 μm3, we focused on the investigation of coincidence site lattice ∑9 GBs. While bearing predominantly a pure tilt character, ∑9 GBs can be categorized into three groups based on correlative 3D morphologies and crystallography.
- Publication:
-
Review of Scientific Instruments
- Pub Date:
- September 2022
- DOI:
- 10.1063/5.0087945
- Bibcode:
- 2022RScI...93i3707T