Designing hexagonal close packed high entropy alloys using machine learning

Abstract

High entropy alloys (HEAs) have drawn significant interest in the materials research community owing to their remarkable physical and mechanical properties. These improved physicochemical properties manifest due to the formation of simple solid solution phases with unique microstructures. Though several pathbreaking HEAs have been reported, the field of alloy design, which has the potential to guide alloy screening, is still an open topic hindering the development of new HEA compositions, particularly ones with hexagonal close packed (hcp) crystal structure. In this work, an attempt has been made to develop an intelligent extra tree (ET) classification model based on the key thermodynamic and structural properties, to predict the phase evolution in HEAs. The results of correlation analysis suggest that all the selected thermodynamic and structural features are viable candidates for the descriptor dataset. Testing accuracy of above 90% along with excellent performance matrices for the ET classifier reveal the robustness of the model. The model can be employed to design novel hcp HEAs and as a valuable tool in the alloy design of HEAs in the future.

Design and Selection of High Entropy Alloys for Hardmetal Matrix Applications Using a Coupled Machine Learning and Calculation of Phase Diagrams Methodology

“…Accuracy, precision, recall, and F 1 scores were all calculated to assess model performance for the phase classification task, the results of which are shown in Table 2 . [ 61,62 ] Performance scores >78% and >82% for classification models X and Z, respectively, indicate that the models successfully predict phase formation on the validation data from the dataset. Furthermore, it suggests that the models do not suffer from overfitting, enabling potential generalizability to predictions on unseen compositions within the virtual candidate search space.…”

“…This methodology, first applying ML and subsequently downselecting using CALPHAD, has a number of key benefits. [ 62 ] Performing CALPHAD analysis of over 2 million compositions considered by the ML in this study is computationally impractical and time intensive. Initial application of ML dramatically reduces the number of compositions that need to be considered and hence the computational time.…”

Accelerated prediction of stacking fault energy in FCC medium entropy alloys using multilayer perceptron neural networks: correlation and feature analysis

“…With the advent of machine learning (ML) in materials research, it is proven to be the most robust and rapid method to predict several material properties, including SFE [27,28]. However, the main limitation of ML incorporation for SFE prediction is the sparsity of reported data in the literature.…”