About this Abstract |
Meeting |
2024 TMS Annual Meeting & Exhibition
|
Symposium
|
Additive Manufacturing: Process-induced Microstructures and Defects
|
Presentation Title |
Unveiling Metal Additive Manufacturing Microstructure through Data-Driven Unsupervised Clustering of Crystallographic Texture |
Author(s) |
Aashique Alam Rezwan, David Montes de Oca Zapiain, Daniel Moser, Michael J Heiden, Theron M Rodgers |
On-Site Speaker (Planned) |
Aashique Alam Rezwan |
Abstract Scope |
Metal additive manufacturing (AM) parts often have microstructural features that are not well described by traditional quantification metric and have variations within an AM part that are challenging to quantify. This work presents a data-driven approach to quantifying microstructures that incorporates grain morphology, crystallographic orientation and phase information. A combination of generalized spherical harmonics (GSH), 2-point spatial correlation, and a dimensionality reduction technique (i.e., Principal Component/Auto-encoder) is used to discern the complex microstructural data into human-readable results for comparison. The method shows sensitivity to minor microstructural differences within specimens created with nominally identical processing parameters, when applied to an experimental dataset of AM 316L-SS. The method can identify different build batches, orientation and outliers present in the process and can potentially serve as a tool in detecting AM process changes for quality control applications.
SNL is managed and operated by NTESS under DOE NNSA contract DE-NA0003525. |
Proceedings Inclusion? |
Planned: |