Regensburg 2022 – scientific programme

Parts | Days | Selection | Search | Updates | Downloads | Help

MM: Fachverband Metall- und Materialphysik

MM 22: Data Driven Materials Science: Experimental Data Treatment and Machine Learning

MM 22.2: Talk

Wednesday, September 7, 2022, 10:45–11:00, H46

A materials informatics framework to discover patterns in atom probe tomography data — •Alaukik Saxena, Nikita Polin, Baptiste Gault, Christoph Freysoldt, and Jörg Neugebauer — Max-Planck-Institut für Eisenforschung GmbH, Düsseldorf 40237, Germany.

Atom probe tomography (APT) is a unique technique that provides 3D elemental distribution with near-atomic resolution for a given material. However, the large amount of data acquired during the experiment and the complexity of the 3D microstructures poses a challenge to fully quantify APT data. Here, taking APT measurements corresponding to a Fe-doped Sm-Co alloy as an example, we present an approach based on unsupervised machine learning to extract different phases in the data. On top of this method, we have built a PCA-based workflow to reduce secondary phase precipitates with complex morphology into simpler plate-like morphologies thus enabling the quantification of in-plane compositional and thickness fluctuations, and relative orientations of the precipitates. The labeled data acquired from the PCA-based approach is used to train a U-NET to perform the same task on different APT samples of the same material automatically. The composition and thickness-related insights are expected to help understand the contribution of the particles to the confinement of the magnetic domains of the dominant 2:17 bulk phase, providing further indications to guide the design of future permanent hard magnets.