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.
more »
« less
Deep Learning-Based Hardness Prediction of Novel Refractory High-Entropy Alloys with Experimental Validation
Hardness is an essential property in the design of refractory high entropy alloys (RHEAs). This study shows how a neural network (NN) model can be used to predict the hardness of a RHEA, for the first time. We predicted the hardness of several alloys, including the novel C0.1Cr3Mo11.9Nb20Re15Ta30W20 using the NN model. The hardness predicted from the NN model was consistent with the available experimental results. The NN model prediction of C0.1Cr3Mo11.9Nb20Re15Ta30W20 was verified by experimentally synthesizing and investigating its microstructure properties and hardness. This model provides an alternative route to determine the Vickers hardness of RHEAs.
more »
« less
- Award ID(s):
- 1946231
- NSF-PAR ID:
- 10218830
- Date Published:
- Journal Name:
- Crystals
- Volume:
- 11
- Issue:
- 1
- ISSN:
- 2073-4352
- Page Range / eLocation ID:
- 46
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
more » « less
Abstract Severe lattice distortion is a prominent feature of high-entropy alloys (HEAs) considered a reason for many of those alloys’ properties. Nevertheless, accurate characterizations of lattice distortion are still scarce to only cover a tiny fraction of HEA’s giant composition space due to the expensive experimental or computational costs. Here we present a physics-informed statistical model to efficiently produce high-throughput lattice distortion predictions for refractory non-dilute/high-entropy alloys (RHEAs) in a 10-element composition space. The model offers improved accuracy over conventional methods for fast estimates of lattice distortion by making predictions based on physical properties of interatomic bonding rather than atomic size mismatch of pure elements. The modeling of lattice distortion also implements a predictive model for yield strengths of RHEAs validated by various sets of experimental data. Combining our previous model on intrinsic ductility, a data mining design framework is demonstrated for efficient exploration of strong and ductile single-phase RHEAs.
-
Refractory high entropy alloys (RHEAs) have gained significant attention in recent years as potential replacements for Ni-based superalloys in gas turbine applications. Improving their properties, such as their high-temperature yield strength, is crucial to their success. Unfortunately, exploring this vast chemical space using exclusively experimental approaches is impractical due to the considerable cost of the synthesis, processing, and testing of candidate alloys, particularly at operation-relevant temperatures. On the other hand, the lack of reasonably accurate predictive property models, especially for high-temperature properties, makes traditional Integrated Computational Materials Engineering (ICME) methods inadequate. In this paper, we address this challenge by combining machine-learning models, easy-to-implement physics-based models, and inexpensive proxy experimental data to develop robust and fast-acting models using the concept of Bayesian updating. The framework combines data from one of the most comprehensive databases on RHEAs (Borg et al., 2020) with one of the most widely used physics-based strength models for BCC-based RHEAs (Maresca and Curtin, 2020) into a compact predictive model that is significantly more accurate than the state-of-the-art. This model is cross-validated, tested for physics-informed extrapolation, and rigorously benchmarked against standard Gaussian process regressors (GPRs) in a toy Bayesian optimization problem. Such a model can be used as a tool within ICME frameworks to screen for RHEAs with superior high-temperature properties. The code associated with this work is available at: https://codeocean.com/capsule/7849853/tree/v2.more » « less
-
Additive Manufacturing (AM) has opened new frontiers for the design of refractory high-entropy alloys (HEAs) for high-temperature applications. The thermal conductivity of the AM feedstock is among the most important thermo-physical properties that control the melting and solidification process. Despite its significance, there remains a notable gap in both computational and experimental research concerning the thermal conductivity of HEAs. Here, we use density functional theory (DFT) to systematically investigate the alloying effects on the transport properties of Ti-Cr-Mo-W-V-Nb-Ta RHEAs, including electrical and thermal conductivities and the Seebeck coefficient. The relaxation time of charge carriers is a key underlying parameter determining thermal conductivity that is exceedingly challenging to predict from first principles alone, and we thus follow the approach by Mukherjee, Satsangi, and Singh [Chem Mater 32, 6507 (2022)] to optimize the relaxation time for RHEAs. We validated thermal conductivity predictions on elemental solids, binary and ternary alloys, and RHEAs and compared them against thermodynamic (CALPHAD) predictions and our experiments with good correlations. To understand observed trends in thermal conductivity, we assessed the phase stability, electronic structure, phonon, and intrinsic- and tensile strength of down-selected RHEAs. Our electronic structure and phonon results connect well with the observed compositional trends for thermal transport in RHEAs. Our DFT assessment and CALPHAD predictions provide a unique design guide for RHEAs with tailored thermal conductivity, a critical consideration for AM and thermal-management applications.more » « less
-
A simple Gaussian process regressor (GPR) model is employed to predict steel hardness and toughness response for tempered martensitic steels. A dataset of over 2000 hardness values from over 250 distinct alloys was compiled, with the aim of incorporating a diverse set of quenched and tempered martensitic steels. The Izod impact toughness was included for over 450 of these alloy/temper conditions. The GPR exhibited an increase in accuracy for both the predicted hardness and Izod impact toughness over linear regression trained on the same dataset. Shapley additive explanations (SHAP) were used to assess the importance of the input features of tempering temperature, tempering time, and 15 elements. Tempering temperature and carbon content were the most important input features in all models. The relative importance of the other 14 alloying elements varied depending on the target property. The SHAP analysis highlighted the complex relationships between composition and mechanical properties that are able to be captured by machine learning approaches.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.3390/cryst11010046
