| About this Abstract |
| Meeting |
TMS Specialty Congress 2024
|
| Symposium
|
Accelerating Discovery for Mechanical Behavior of Materials 2024
|
| Presentation Title |
Mechanical Properties Prediction of Functionally Graded Metallic Materials Through High-throughput Characterization and Machine Learning |
| Author(s) |
Christopher Bean, Dhruv Anjaria, Rephayah Black, Jackson Nie, Marie Charpagne, Jean-Charles Stinville |
| On-Site Speaker (Planned) |
Christopher Bean |
| Abstract Scope |
Advances in high-throughput material characterization, complemented by machine-learning techniques, have brought the longstanding goal of accelerating material discovery within grasp. With these advancements, materials of interest with favorable composition can be identified using computed approaches or combinatorial synthesis. However, mechanical characterization considerably hampers this material development cycle. High-fidelity macroscopic testing is time-consuming and remains the sole method for obtaining advanced mechanical properties. The present study presents an accelerated route for evaluating advanced mechanical properties by leveraging inverse analysis of large datasets of nanometer-scale plastic localization events collected through high-throughput, high-resolution digital image correlation correlated with electron backscattered diffraction measurement and analyzed by computer vision and graph neural network. We identified correlations between the characteristics of the plastic localization events (twinning and slip) and macroscopic properties. These correlations enabled a swift evaluation of advanced mechanical properties, including fatigue life, of stainless functionally graded materials. |
| Proceedings Inclusion? |
Definite: Other |