| About this Abstract |
| Meeting |
MS&T23: Materials Science & Technology
|
| Symposium
|
Additive Manufacturing of Metals: Microstructure, Properties and Alloy Development
|
| Presentation Title |
Prediction of Thermal Conductivity of Al-Alloys: Finite Element Simulations Combined with Statistical Analysis and Machine Learning |
| Author(s) |
Shuvodeep De, Sunyong Kwon, Dongwon Shin, Yousub Lee |
| On-Site Speaker (Planned) |
Shuvodeep De |
| Abstract Scope |
Most cast Al-alloys are strengthened by precipitates, and their microstructure features significantly influence their thermal/mechanical properties. A nearly infinite number of precipitate features challenge understanding and predicting optimal composition, processing, and microstructural conditions to achieve desired properties. We developed a simulation framework to predict thermal conductivity of Al-alloys (e.g., Al-3Ni, Al-3Ni-0.45Zr) using Finite Element Method (FEM), synthetic microstructure, statistical analysis, and Machine Learning (ML) approach. The influence of precipitate’s morphology and orientation on thermal conductivity has been investigated with microstructure generated by Cellular Automata (CA) and statistically analyzed using two-point statistics. Principal Component Analysis (PCA) finds correlation between the microstructure and thermal conductivity. The workflow can be extended to other Al-alloys by computation on a High-Performance Computing (HPC) platform. This virtual exploration of precipitate features will not only fill a gap resulting from limited or missing data points but also provides selection guidance for rapid optimization of experimental conditions. |