About this Abstract |
| Meeting |
2026 TMS Annual Meeting & Exhibition
|
| Symposium
|
Novel Strategies for Rapid Acquisition and Processing of Large Datasets From Advanced Characterization Techniques
|
| Presentation Title |
Affine Transformations to Correlate Experimental and Simulated EDS Spectra for Multi-Element Systems |
| Author(s) |
Malachi Nelson, James Zillinger, Luis Nunez, Boone Beausoleil |
| On-Site Speaker (Planned) |
Malachi Nelson |
| Abstract Scope |
Energy Dispersive X-ray Spectroscopy (EDS) is a popular technique for qualitative elemental analysis, but it is a challenge to produce more quantitative results. Proprietary software can provide “black box” estimates of elemental concentrations, but physics-based simulations provide higher fidelity analysis by considering the multitude of factors affecting the spectra. Quantitative fitting techniques are available, but these require intensive characterization and fitting methods which are difficult when performing high throughput analysis. This work provides an alternative method to correlate experimental and simulated EDS spectra by automating simulations to build a database of relevant features. Affine transformations are used to perform multivariable interpolation in Euclidean space to relate relevant features between experimental and simulated spectra, preserving properties such as the barycenter. Affine transformations are computationally efficient and implementing features beyond peak intensities using machine learning methods is discussed. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Characterization, Machine Learning, Nuclear Materials |