| About this Abstract |
| Meeting |
2025 TMS Annual Meeting & Exhibition
|
| Symposium
|
Novel Strategies for Rapid Acquisition and Processing of Large Datasets from Advanced Characterization Techniques
|
| Presentation Title |
Correlative Microscopy and AI for Rapid Analysis of Complex Material Structures |
| Author(s) |
Hugues Francois-Saint-Cyr, Alice Scarpellini, Bartlomeij Winiarski, Rengarajan Pelapur |
| On-Site Speaker (Planned) |
Hugues Francois-Saint-Cyr |
| Abstract Scope |
High-temperature coatings require thorough validation of their chemical and mechanical properties in harsh environments. Correlative microscopy (CM) workflows, combined with advanced data acquisition and processing techniques, provide a comprehensive understanding of these materials. Multi-dimensional analysis of samples is achieved by combining X-ray, electron-beam, and ion-beam techniques. This study employs strategies for rapid acquisition of high-resolution datasets and AI-driven data processing algorithms, to efficiently extract meaningful microstructures and their associated chemical phases.
AI-assisted image analysis (IA) accelerates the interpretation of results for both specialists and non-specialists. Deep learning models, trained with "human-in-the-loop" approaches, ensure accurate analysis. This study showcases the synergy of CM and AI-assisted IA through the example of a Thermal Barrier Coating (TBC) for a turboramjet engine. The workflow includes BIB milling, SEM-EDS, and FIB cross-sectioning, facilitating rapid characterization of TBC microstructure and composition. Time-to-results could be shrunk up to 4x-fold. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
High-Temperature Materials, Machine Learning, Characterization |