About this Abstract |
Meeting |
2022 TMS Annual Meeting & Exhibition
|
Symposium
|
Seeing is Believing -- Understanding Environmental Degradation and Mechanical Response Using Advanced Characterization Techniques: An SMD Symposium in Honor of Ian M. Robertson
|
Presentation Title |
Graph-based Analysis of Deforming Polycrystals |
Author(s) |
Darren C. Pagan, Austin R. Benson, Matthew P. Kasemer |
On-Site Speaker (Planned) |
Darren C. Pagan |
Abstract Scope |
A polycrystal can naturally be thought of as a network of connected grains through which load and deformation is transferred. With the recognition of this network structure, it is apparent that polycrystals can be represented more formally by graphs. In addition, the huge range of tools that have been previously developed to analyze and predict the behavior of various structures represented by graphs, including social networks and traffic patterns, can be taken advantage of. Here we describe the application of graphs to analyze and understand large three-dimensional descriptions of polycrystal microstructure and micromechanical response, including those measured using high-energy diffraction microscopy and generated synthetically. A focus will be the use of graph-based machine learning (graph neural networks) for the prediction of mechanical response of polycrystals. |
Proceedings Inclusion? |
Planned: |
Keywords |
Characterization, Computational Materials Science & Engineering, Machine Learning |