Prediction of phase and hardness of HEAs based on constituent elements using machine learning models

Bakr, Mahmoud; Syarif, Junaidi; Hashem, Mohamed

doi:10.1016/j.mtcomm.2022.103407

Cited by 12 publications

(6 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Machine learning is applied to the strength and hardness properties of HEAs with data and set samples, either derived from experimental studies or obtained with the help of computational simulations ( Bakr et al, 2022 ; Chen Z. W. et al, 2022; Hou et al, 2022 ). The research mainly focuses on machine learning to accurately predict the strength and hardness properties of new alloys and to guide the design of alloy compositions based on the property prediction results.…”

Section: Component Design Theory and Simulation Studiesmentioning

confidence: 99%

Bio-high entropy alloys: Progress, challenges, and opportunities

Feng

Tang

Liu

et al. 2022

Front. Bioeng. Biotechnol.

View full text Add to dashboard Cite

With the continuous progress and development in biomedicine, metallic biomedical materials have attracted significant attention from researchers. Due to the low compatibility of traditional metal implant materials with the human body, it is urgent to develop new biomaterials with excellent mechanical properties and appropriate biocompatibility to solve the adverse reactions caused by long-term implantation. High entropy alloys (HEAs) are nearly equimolar alloys of five or more elements, with huge compositional design space and excellent mechanical properties. In contrast, biological high-entropy alloys (Bio-HEAs) are expected to be a new bio-alloy for biomedicine due to their excellent biocompatibility and tunable mechanical properties. This review summarizes the composition system of Bio-HEAs in recent years, introduces their biocompatibility and mechanical properties of human bone adaptation, and finally puts forward the following suggestions for the development direction of Bio-HEAs: to improve the theory and simulation studies of Bio-HEAs composition design, to quantify the influence of composition, process, post-treatment on the performance of Bio-HEAs, to focus on the loss of Bio-HEAs under actual service conditions, and it is hoped that the clinical application of the new medical alloy Bio-HEAs can be realized as soon as possible.

show abstract

Section: Component Design Theory and Simulation Studiesmentioning

confidence: 99%

Bio-high entropy alloys: Progress, challenges, and opportunities

Feng

Tang

Liu

et al. 2022

Front. Bioeng. Biotechnol.

View full text Add to dashboard Cite

show abstract

“…For instance, care must be taken while extracting the data from mixed manufacturing routes and tackling an imbalanced dataset. This becomes clear from the fact that although the ANN model used in Bakr et al 28 study achieved an accuracy of 93.4% but could not correctly predicted the existence of amorphous phase. Hence, proving the fact that even after achieving 93.4% of accuracy, their model resulted in erroneous predictions while treating the imbalanced data.…”

Section: Discussionmentioning

confidence: 86%

“…In the literature concerning phase prediction of HEAs using ML algorithms, no study can be seen that targets one particular synthesis route to extract the data reliably from experimental studies which can help avoid the spurious effect of an alternative synthesis routes on the resulting phase. For example, Bakr et al 28 used neural network on 775 samples of HEAs synthesized from mixed manufacturing routes (Arc-melting, sintering, SLM, and others) and obtained 93.4% accuracy in predicting the existence of different phases (AM, BCC, FCC, and IM). Their study did not consider the effect of manufacturing method on the resulting phase of HEAs.…”

Section: Mixing Enthalpy ∆S MIXmentioning

confidence: 99%

Phase prediction and experimental realisation of a new high entropy alloy using machine learning

Singh

Katiyar

Goel

et al. 2023

Sci Rep

View full text Add to dashboard Cite

Nearly ~ 108 types of High entropy alloys (HEAs) can be developed from about 64 elements in the periodic table. A major challenge for materials scientists and metallurgists at this stage is to predict their crystal structure and, therefore, their mechanical properties to reduce experimental efforts, which are energy and time intensive. Through this paper, we show that it is possible to use machine learning (ML) in this arena for phase prediction to develop novel HEAs. We tested five robust algorithms namely, K-nearest neighbours (KNN), support vector machine (SVM), decision tree classifier (DTC), random forest classifier (RFC) and XGBoost (XGB) in their vanilla form (base models) on a large dataset screened specifically from experimental data concerning HEA fabrication using melting and casting manufacturing methods. This was necessary to avoid the discrepancy inherent with comparing HEAs obtained from different synthesis routes as it causes spurious effects while treating an imbalanced data—an erroneous practice we observed in the reported literature. We found that (i) RFC model predictions were more reliable in contrast to other models and (ii) the synthetic data augmentation is not a neat practice in materials science specially to develop HEAs, where it cannot assure phase information reliably. To substantiate our claim, we compared the vanilla RFC (V-RFC) model for original data (1200 datasets) with SMOTE-Tomek links augmented RFC (ST-RFC) model for the new datasets (1200 original + 192 generated = 1392 datasets). We found that although the ST-RFC model showed a higher average test accuracy of 92%, no significant breakthroughs were observed, when testing the number of correct and incorrect predictions using confusion matrix and ROC-AUC scores for individual phases. Based on our RFC model, we report the development of a new HEA (Ni25Cu18.75Fe25Co25Al6.25) exhibiting an FCC phase proving the robustness of our predictions.

show abstract

“… Performance Measure FCC BCC IM Model This work Accuracy - - - 92.74 Recall% 97.17 94. 03 83.59 - Precision% 97.72 97.37 84.67 - F1 score 97.44 95.62 83.82 - Bakr et al [21] Accuracy% - - - 93.42 Recall% 95.24 94.9 84.98 - Precision% 96.30 96.03 86.41 - F 1 score 95.76 95.41 85.54 - Alshibany et al [20] Accuracy% - - - 90 F 1 score 89 96 82 a - Lee et al [15] Accuracy% - - - 93 Pei et al [16] Accuracy% - - ...…”

Section: Data Generationmentioning

confidence: 99%