| About this Abstract |
| Meeting |
TMS Specialty Congress 2025
|
| Symposium
|
The 7th International Congress on 3D Materials Science (3DMS 2025)
|
| Presentation Title |
Informed Unsupervised Machine Learning Analysis of Dislocation Microstructure From High-Resolution Differential Aperture X-Ray Structural Microscopy Data |
| Author(s) |
Khaled N. SharafEldin, Bryan D. Miller, Wenjun Liu, Jon Tischler, Benjamin Anglin, Anter El-Azab |
| On-Site Speaker (Planned) |
Khaled N. SharafEldin |
| Abstract Scope |
This study leverages high-resolution Differential-Aperture X-Ray Structural Microscopy (DAXM) to probe the local microstructure and strain at sub-micron resolution in deformed 304 steel at 2% strain. We developed a machine learning technique to understand the multimodal statistical nature of the lattice rotation lattice rotation and deviatoric elastic strain. The multi-peak nature of the rotation distribution measured over a disoriented grain not aligned with the loading axis of a polycrystalline specimen was fit to a collection of multi-variate Cauchy distributions, which were then mapped back to form a contiguous regions of the crystal rotated near the average values of each distribution, representing the coarsest grain subdivision scale. The dislocation density tensor was also extracted, and its norm was laid over the rotation field to confirm the subgrain boundaries. This study highlights the potential of integrating advanced X-ray microscopy techniques with data-driven analysis methods to uncover detailed microstructure scales in deformed crystals. |
| Proceedings Inclusion? |
Undecided |