Predicting Elastic Constants of Refractory Complex Concentrated Alloys Using Machine Learning Approach

Abstract

Refractory complex concentrated alloys (RCCAs) have drawn increasing attention recently owing to their balanced mechanical properties, including excellent creep resistance, ductility, and oxidation resistance. The mechanical and thermal properties of RCCAs are directly linked with the elastic constants. However, it is time consuming and expensive to obtain the elastic constants of RCCAs with conventional trial-and-error experiments. The elastic constants of RCCAs are predicted using a combination of density functional theory simulation data and machine learning (ML) algorithms in this study. The elastic constants of several RCCAs are predicted using the random forest regressor, gradient boosting regressor (GBR), and XGBoost regression models. Based on performance metrics R-squared, mean average error and root mean square error, the GBR model was found to be most promising in predicting the elastic constant of RCCAs among the three ML models. Additionally, GBR model accuracy was verified using the other four RHEAs dataset which was never seen by the GBR model, and reasonable agreements between ML prediction and available results were found. The present findings show that the GBR model can be used to predict the elastic constant of new RHEAs more accurately without performing any expensive computational and experimental work.

Recent advances in computational design of structural multi-principal element alloys

“… 74 Elastic constants: C 11 , C 22 , C 33 , C 12 , C 13 , C 23 , C 44 , C 55 , C 66 1229 binary alloys from Materials Project (DFT) , , VEC, , Regression: random forest (RF) RMSE: 34.61 GPa, MAE: 23.46 GPa, R 2 : 0.779 Bhandari et al. 75 Elastic constants: C 11 , C 12 , C 44 370 MPEAs (DFT) , , , , , VEC, , , name of alloys Regression: gradient boosting regressor (GBR) RMSE: 12.68 GPa, MAE: 9.62 GPa, R 2 : 0.853 Vazquez et al. 76 Elastic constants: C 11 , C 12 , C 44 110 binary and 60 ternary systems (DFT) , , , , , …”

“… 73 The model was reported to have comparable performance to neural network models, having high interpretability at the same time. Other usages include the prediction of various mechanical properties, such as elastic constants, 74 , 75 , 76 hardness, 77 , 78 yield strength, 79 , 80 and Young’s modulus. 80 , 81 , 82 A recent study combined DFT and ML to explore the optimal compositional space of the non-equiatomic MoNbTaTiV system using the element composition as the only feature and identified Ti as the critical factor that could potentially compromise Young’s modulus of the MPEA.…”

Machine learning prediction of the mechanical properties of refractory multicomponent alloys based on a dataset of phase and first principles simulation

“…The dataset was further used to train various ML models, including random forest regressor, gradient boosting regressor, and XGBoost regression models, to predict the mechanical properties of CCAs. For example, the elastic constants in the dataset were used to train those ML models (Bhandari et al, 2022) which were evaluated by the root-mean-squared error, the average coefficient of determination, and mean absolute error. It is found that gradient boosting regressor has higher prediction accuracy on elastic constants.…”

High-throughput materials screening algorithm based on first-principles density functional theory and artificial neural network for high-entropy alloys

“…Other methods also exhibit high accuracy prediction, e.g., a combined ML and CALPHAD technique, an artificial neural network technique (ANN) coupled with experimental data 9 , 10 , 12 , 23 – 28 . In addition to phase formation of HEAs, machine learning was recently employed to predict the mechanical properties of HEA bulk materials, including microhardness 10 , 27 , yield strength 12 , 23 , dislocation density 12 , elastic modulus 29 , Young’s modulus 30 , hardness 11 , 31 , and elastic constant 32 . These shed light on the ML-based high-throughput screening of HEA materials.…”