About this Abstract |
Meeting |
MS&T23: Materials Science & Technology
|
Symposium
|
Materials Processing and Fundamental Understanding Based on Machine Learning and Data Informatics
|
Presentation Title |
Computing Grain Boundary "Phase" Diagrams: From Thermodynamic Models and Atomistic Simulations to Machine Learning |
Author(s) |
Jian Luo |
On-Site Speaker (Planned) |
Jian Luo |
Abstract Scope |
In this talk, I will first review a series of our studies to compute the grain boundary (GB) "phase" diagrams [see a Perspective: Interdisciplinary Materials 2:137-160 (2023)]. First, GB lambda diagrams were constructed to forecast high-temperature GB disordering and related trends in sintering and other properties. In parallel, a lattice model was utilized to construct GB adsorption diagrams. Subsequently, atomistic simulations were used to compute more rigorous and accurate GB diagrams of not only thermodynamic and structural characters but also mechanical properties. To further extend prediction power, machine learning was combined with atomistic simulations to predict GB properties as functions of five GB macroscopic (crystallographic) degrees of freedom (DOFs) plus temperature and composition for a binary alloy in a 7-D space [Materials Today 38:49 (2020)] or as functions of four independent compositional variables and temperature in a 5-D space for a given GB in high-entropy alloys [Materials Horizons 9:1023 (2022)]. |