High-throughput and data-driven machine learning techniques for discovering high-entropy alloys

High-throughput and data-driven machine learning techniques for discovering high-entropy alloys

2024 | Lu Zhichao, Ma Dong, Liu Xiongjun & Zhaoping Lu
This review discusses the application of high-throughput techniques and data-driven machine learning (ML) in the discovery and optimization of high-entropy alloys (HEAs). HEAs, characterized by their chemically disordered but topologically ordered structures, have attracted significant interest due to their unique properties such as high strength, hardness, and resistance to high-temperature softening. The review highlights the challenges in designing HEAs with specific properties, particularly the vast composition space and the complex interplay between constituents and properties. High-throughput techniques, including laser additive manufacturing (LAM), combinatorial thin film deposition, diffusion multiples, and welding, are used to efficiently explore this composition space. These techniques enable the rapid synthesis and characterization of HEAs, providing valuable materials information for ML modeling. ML methods, such as machine learning algorithms and data-driven workflows, are employed to predict and optimize the properties of HEAs based on large datasets. The review also discusses the integration of experimental, theoretical, and computational methods to uncover the underlying structure-property relationships in HEAs. Finally, the review outlines future research directions, emphasizing the need for more robust datasets, improved classification systems, and the development of reliable thermodynamic databases for HEAs.This review discusses the application of high-throughput techniques and data-driven machine learning (ML) in the discovery and optimization of high-entropy alloys (HEAs). HEAs, characterized by their chemically disordered but topologically ordered structures, have attracted significant interest due to their unique properties such as high strength, hardness, and resistance to high-temperature softening. The review highlights the challenges in designing HEAs with specific properties, particularly the vast composition space and the complex interplay between constituents and properties. High-throughput techniques, including laser additive manufacturing (LAM), combinatorial thin film deposition, diffusion multiples, and welding, are used to efficiently explore this composition space. These techniques enable the rapid synthesis and characterization of HEAs, providing valuable materials information for ML modeling. ML methods, such as machine learning algorithms and data-driven workflows, are employed to predict and optimize the properties of HEAs based on large datasets. The review also discusses the integration of experimental, theoretical, and computational methods to uncover the underlying structure-property relationships in HEAs. Finally, the review outlines future research directions, emphasizing the need for more robust datasets, improved classification systems, and the development of reliable thermodynamic databases for HEAs.
Reach us at info@study.space
[slides] High-throughput and data-driven machine learning techniques for discovering high-entropy alloys | StudySpace