Thermodynamic Descriptors for Rapid Search of Compositional Complex Spinodal Alloys with Artificial Neural Network

被引：0

作者：

H. Gong ^{[1
]}

Z. Q. Chen ^{[1
]}

Y. H. Shang ^{[1
]}

H. Wang ^{[2
]}

S. J. Hu ^{[1
]}

Y. Yang ^{[1
]}

机构：

[1] City University of Hong Kong,Department of Mechanical Engineering

[2] City University of Hong Kong (Dongguan),Department of Materials Science and Engineering

[3] City University of Hong Kong,undefined

来源：

High Entropy Alloys & Materials | 2024年 / 2卷 / 2期

关键词：

Compositional complex alloys; Machine learning; Spinodal decomposition; Nano-sized interconnected microstructure;

D O I：

10.1007/s44210-024-00047-x

中图分类号：

学科分类号：

摘要：

Designing compositional complex alloys with interconnected nanostructures through spinodal decomposition is promising for achieving advanced alloys with exceptional mechanical and functional properties. However, the lack of equilibrium phase diagrams for such compositional complex alloys has posed a significant challenge. In this study, we address this challenge by first introducing data descriptors derived from instability of solid solutions with ideal mixing and later serving as the basis for developing an artificial neural network (ANN) with other commonly used data descriptors. Using this ANN model, we have successfully designed a series of compositional complex spinodal alloys (CCSAs) within the Al–Co–Cr–Fe–Ni and Al–Cu–Fe–Mn–Ni system. Furthermore, we extended the ideal mixing model of solid solution instability by considering chemical short-range ordering and elemental de-mixing, which better explains the elemental separation of CCSAs.

引用

页码：374 / 386

页数：12