About this Abstract |
Meeting |
2024 TMS Annual Meeting & Exhibition
|
Symposium
|
Additive Manufacturing: Materials Design and Alloy Development VI – Closed-Loop Alloy Design
|
Presentation Title |
Machine Learning Discovery of Optimal Processing Zones in Laser Powder Bed Fusion via High Throughput Mechanical Experiments |
Author(s) |
Salahudin M. Nimer, Mary Daffron, Steven Storck |
On-Site Speaker (Planned) |
Salahudin M. Nimer |
Abstract Scope |
The combined optimization of composition, processing, and post-fabrication treatment in materials is slow and costly. This is largely due to the use of conventional tests which significantly limit the number of variables that can be studied, and thus restricts the generation of process-structure-property relationships. In this work, we present an application of a novel method employing a high-throughput compressive shear test for rapid exploration of the laser powder bed fusion parameter space of IN625. A machine learning model was developed to define optimal processing zones of the material across a wide parameter space exploring the laser power, hatch spacing, laser spot size and scan speed. The method was validated by alignment with manufacturer-recommended parameters, and results indicate the potential for parameters that may enable faster build speed and/or enhanced performance. Considerations for future implementations of this technique leveraging automation to accelerate the materials design and development loop will be presented. |
Proceedings Inclusion? |
Planned: |
Keywords |
Additive Manufacturing, Characterization, Machine Learning |