Machine learning in designing amorphous alloys

HU Jingyi; XU Xiang; JI Xiaomei; XU Mingxian; JIANG Daifeng; WANG Jin

doi:10.13374/j.issn2095-9389.2022.11.11.002

Volume 45 Issue 9

Sep. 2023

Turn off MathJax

Article Contents

Article Navigation > Chinese Journal of Engineering > 2023 > 45(9): 1517-1527

HU Jingyi, XU Xiang, JI Xiaomei, XU Mingxian, JIANG Daifeng, WANG Jin. Machine learning in designing amorphous alloys[J]. Chinese Journal of Engineering, 2023, 45(9): 1517-1527. doi: 10.13374/j.issn2095-9389.2022.11.11.002

Citation:

HU Jingyi, XU Xiang, JI Xiaomei, XU Mingxian, JIANG Daifeng, WANG Jin. Machine learning in designing amorphous alloys[J]. Chinese Journal of Engineering, 2023, 45(9): 1517-1527. doi: 10.13374/j.issn2095-9389.2022.11.11.002

Citation:

PDF( 1179 KB)

Machine learning in designing amorphous alloys

doi: 10.13374/j.issn2095-9389.2022.11.11.002

HU Jingyi^{1, 2
,},
XU Xiang^{3, 4
,
,},
JI Xiaomei¹,
XU Mingxian¹,
JIANG Daifeng¹,
WANG Jin¹

1.
Science and Technology Service Platform (Overseas Chinese Pioneer Park), Qilu University of Technology (Shandong Academy of Sciences), Jinan 250000, China
2.
Shandong SME Productivity Promotion Center, Jinan 250000, China
3.
Shandong Key Laboratory of Blockchain Finance, Shandong University of Finance and Economics, Jinan 250000, China
4.
School of Computer Science and Technology, Shandong University of Finance and Economics, Jinan 250000, China

More Information

Corresponding author: E-mail: xuxiang8420@outlook.com
Received Date: 2022-11-11
Available Online: 2023-04-06
Publish Date: 2023-09-25

Abstract

Abstract

Metallic glasses have received a lot of interest because of their excellent mechanical, physical, and chemical qualities. For example, they have a stronger resistivity than crystalline metals composed of the same elements and a lower viscosity coefficient. However, the difficulty in creating alloy compositions has been a concern for researchers. Traditional amorphous alloy systems design approaches, such as empirical trial-and-error methods and methods based on density functional theory (DFT), have assisted researchers in exploring numerous amorphous alloy systems during the growth of materials science over the last few decades. However, with the continuous development of materials science, these methods have been difficult to meet the needs of researchers due to their long development cycles and low efficiency. Additionally, the complex and long-range disordered structure of metallic glasses makes it difficult to understand their structure and nature in a comprehensive and clear way by conventional methods. Amorphous alloy composition design and property analysis are now often conducted using machine learning techniques because of their low experimental cost, short development cycle, strong data processing capability, and high predictive performance, among other advantages. They present new approaches and chances to address significant key bottlenecks in the field of metallic glass. In this study, the main processes of machine learning model building were introduced. Subsequently, the related studies on data pre-processing, model construction, and model validation were presented. For data pre-processing, data selection, feature engineering, and advanced data balancing methods were primarily described. In the feature engineering part, the model performance with various input features was examined, and it was shown that either employing physical properties or directly using the alloy compositions as the model input might result in high performance. Four machine learning algorithms were used to generate the machine learning model: artificial neural networks (ANN), support vector machines (SVM), random forest (RF), and extreme gradient boosting (XGBoost). A comparison indicates that SVM models work best with small data sets, whereas the performance of all other models tends to get better as the amount of training data increases. Generally, the XGBoost method outperforms several other methods and is, therefore, often used in machine learning competitions. Model validation approaches: K-fold cross-validation and leave-one-out cross-validation methods were presented. A good metallic glass performance prediction method needs to perform well in both validation methods. Finally, this study provides several possible future research directions on feature engineering, dataset construction, validation, and machine learning models.
- metallic glasses,
- machine learning,
- alloy design,
- property prediction,
- data pre-processing,
- model selection,
- model validation

FullText(HTML)

References(61)

References

[1]	Klement W, Willens R H, Duwez P. Non-crystalline structure in solidified gold–silicon alloys. Nature, 1960, 187(4740): 869
[2]	Koledin T D, Saini J S, Xu D H, et al. Characterization of a new rare-earth-free Cu-based bulk metallic glass. Microsc Microanal, 2022, 28(Suppl 1): 2782
[3]	Ohashi Y, Wada T, Kato H. High-entropy design and its influence on glass-forming ability in Zr-Cu-based metallic glass. J Alloys Compd, 2022, 915: 165366 doi: 10.1016/j.jallcom.2022.165366
[4]	Zhang X P, Lai L M, Xiao S M, et al. Effect of W on the thermal stability, mechanical properties and corrosion resistance of Fe-based bulk metallic glass. Intermetallics, 2022, 143: 107485 doi: 10.1016/j.intermet.2022.107485
[5]	Shan F L, Sun T T, Song W D, et al. A bridge from metallic glasses to medium-entropy alloys in Ti–Cu–Zr–Pd–Co system: Design, microstructure, and deformation-induced-martensitic transformation. J Non Cryst Solids, 2022, 587: 121608 doi: 10.1016/j.jnoncrysol.2022.121608
[6]	Samuel A L. Some studies in machine learning using the game of checkers. IBM J Res Dev, 1959, 3(3): 210 doi: 10.1147/rd.33.0210
[7]	Ramprasad R, Batra R, Pilania G, et al. Machine learning in materials informatics: Recent applications and prospects. Npj Comput Mater, 2017, 3: 54 doi: 10.1038/s41524-017-0056-5
[8]	Correa-Baena J P, Hippalgaonkar K, van Duren J, et al. Accelerating materials development via automation, machine learning, and high-performance computing. Joule, 2018, 2(8): 1410 doi: 10.1016/j.joule.2018.05.009
[9]	Ren F, Ward L, Williams T, et al. Accelerated discovery of metallic glasses through iteration of machine learning and high-throughput experiments. Sci Adv, 2018, 4(4): eaaq1566 doi: 10.1126/sciadv.aaq1566
[10]	Wen C, Zhang Y, Wang C X, et al. Machine learning assisted design of high entropy alloys with desired property. Acta Mater, 2019, 170: 109 doi: 10.1016/j.actamat.2019.03.010
[11]	Liu X D, Li X, He Q F, et al. Machine learning-based glass formation prediction in multicomponent alloys. Acta Mater, 2020, 201: 182 doi: 10.1016/j.actamat.2020.09.081
[12]	Hachmann J, Olivares-Amaya R, Atahan-Evrenk S, et al. The Harvard clean energy project: Large-scale computational screening and design of organic photovoltaics on the world community grid. J Phys Chem Lett, 2011, 2(17): 2241 doi: 10.1021/jz200866s
[13]	Hachmann J, Olivares-Amaya R, Jinich A, et al. Lead candidates for high-performance organic photovoltaics from high-throughput quantum chemistry - the Harvard Clean Energy Project. Energy Environ Sci, 2014, 7(2): 698 doi: 10.1039/C3EE42756K
[14]	Jain A, Ong S P, Hautier G, et al. Commentary: The materials project: A materials genome approach to accelerating materials innovation. APL Mater, 2013, 1(1): 011002 doi: 10.1063/1.4812323
[15]	Saal J E, Kirklin S, Aykol M, et al. Materials design and discovery with high-throughput density functional theory: The open quantum materials database (OQMD). JOM, 2013, 65(11): 1501 doi: 10.1007/s11837-013-0755-4
[16]	Kirklin S, Saal J E, Meredig B, et al. The open quantum materials database (OQMD): Assessing the accuracy of DFT formation energies. Npj Comput Mater, 2015, 1: 15010 doi: 10.1038/npjcompumats.2015.10
[17]	Curtarolo S, Setyawan W, Wang S D, et al. AFLOWLIB. ORG: A distributed materials properties repository from high-throughput ab initio calculations. Comput Mater Sci, 2012, 58: 227
[18]	Allen F H. The Cambridge Structural Database: A quarter of a million crystal structures and rising. Acta Cryst, 2002, B58: 380
[19]	Kalidindi S R, De Graef M. Materials data science: Current status and future outlook. Annu Rev Mater Res, 2015, 45: 171 doi: 10.1146/annurev-matsci-070214-020844
[20]	Kawazoe Y, Carow-Watamura U, Louzguine D V. Phase Diagrams and Physical Properties of Nonequilibrium Alloys: Subvolume C: Physical Properties of Multi-Component Amorphous Alloys. Berlin: Springer, 2019
[21]	Turnbull D. Under what conditions can a glass be formed? Contemp Phys, 1969, 10(5): 473
[22]	Inoue A, Zhang T, Masumoto T. Glass-forming ability of alloys. J Non Cryst Solids, 1993, 156-158: 473 doi: 10.1016/0022-3093(93)90003-G
[23]	Lu Z P, Liu C T. A new glass-forming ability criterion for bulk metallic glasses. Acta Mater, 2002, 50(13): 3501 doi: 10.1016/S1359-6454(02)00166-0
[24]	Du X H, Huang J C, Liu C T, et al. New criterion of glass forming ability for bulk metallic glasses. J Appl Phys, 2007, 101(8): 086108 doi: 10.1063/1.2718286
[25]	Inoue A. Stabilization of metallic supercooled liquid and bulk amorphous alloys. Acta Mater, 2000, 48(1): 279 doi: 10.1016/S1359-6454(99)00300-6
[26]	Miedema A R, de Boer F R, Boom R. Model predictions for the enthalpy of formation of transition metal alloys. Calphad, 1977, 1(4): 341 doi: 10.1016/0364-5916(77)90011-6
[27]	Ramakrishna Rao B, Gandhi A S, Vincent S, et al. Prediction of glass forming ability using thermodynamic parameters. Trans Indian Inst Met, 2012, 65(6): 559 doi: 10.1007/s12666-012-0215-9
[28]	Mansoori G A, Carnahan N F, Starling K E, et al. Equilibrium thermodynamic properties of the mixture of hard spheres. J Chem Phys, 1971, 54(4): 1523 doi: 10.1063/1.1675048
[29]	Fang S S, Xiao X S, Xia L, et al. Relationship between the widths of supercooled liquid regions and bond parameters of Mg-based bulk metallic glasses. J Non Cryst Solids, 2003, 321(1-2): 120 doi: 10.1016/S0022-3093(03)00155-8
[30]	Guo S, Liu C T. Phase stability in high entropy alloys: Formation of solid-solution phase or amorphous phase. Prog Nat Sci Mater Int, 2011, 21(6): 433 doi: 10.1016/S1002-0071(12)60080-X
[31]	Angell C A. Formation of glasses from liquids and biopolymers. Science, 1995, 267(5206): 1924 doi: 10.1126/science.267.5206.1924
[32]	Wang L M, Tian Y J, Liu R P. Dependence of glass forming ability on liquid fragility: Thermodynamics versus kinetics. Appl Phys Lett, 2010, 97(18): 181901 doi: 10.1063/1.3506900
[33]	Mastropietro D G, Moya J A. Design of Fe-based bulk metallic glasses for maximum amorphous diameter (D_max) using machine learning models. Comput Mater Sci, 2021, 188: 110230 doi: 10.1016/j.commatsci.2020.110230
[34]	Yao Y, Sullivan T IV, Yan F, et al. Balancing data for generalizable machine learning to predict glass-forming ability of ternary alloys. Scr Mater, 2022, 209: 114366 doi: 10.1016/j.scriptamat.2021.114366
[35]	Butler K T, Davies D W, Cartwright H, et al. Machine learning for molecular and materials science. Nature, 2018, 559(7715): 547 doi: 10.1038/s41586-018-0337-2
[36]	Wu W, Sun Q. Applying machine learning to accelerate new materials development. Sci Sin-Phys Mech Astron, 2018, 48(10): 107001
[37]	Park S, Fonseca J H, Marimuthu K P, et al. Determination of material properties of bulk metallic glass using nanoindentation and artificial neural network. Intermetallics, 2022, 144: 107492 doi: 10.1016/j.intermet.2022.107492
[38]	Han G, Marimuthu K P, Lee H. Evaluation of thin film material properties using a deep nanoindentation and ANN. Mater Des, 2022, 221: 111000 doi: 10.1016/j.matdes.2022.111000
[39]	Reddy G J, Kandavalli M, Saboo T, et al. Prediction of glass forming ability of bulk metallic glasses using machine learning. Integr Mater Manuf Innov, 2021, 10(4): 610 doi: 10.1007/s40192-021-00239-y
[40]	Sun Y T, Bai H Y, Li M Z, et al. Machine learning approach for prediction and understanding of glass-forming ability. J Phys Chem Lett, 2017, 8(14): 3434 doi: 10.1021/acs.jpclett.7b01046
[41]	Li Z, Long Z L, Lei S, et al. Predicting the glass formation of metallic glasses using machine learning approaches. Comput Mater Sci, 2021, 197: 110656 doi: 10.1016/j.commatsci.2021.110656
[42]	Zhang Y X, Xing G C, Sha Z D, et al. A two-step fused machine learning approach for the prediction of glass-forming ability of metallic glasses. J Alloys Compd, 2021, 875: 160040 doi: 10.1016/j.jallcom.2021.160040
[43]	Fan Z, Ma E, Falk M L. Predicting the location of shear band initiation in a metallic glass. Phys Rev Mater, 2022, 6(6): 065602 doi: 10.1103/PhysRevMaterials.6.065602
[44]	Ward L, Agrawal A, Choudhary A, et al. A general-purpose machine learning framework for predicting properties of inorganic materials. Npj Comput Mater, 2016, 2: 16028 doi: 10.1038/npjcompumats.2016.28
[45]	Ward L, O'Keeffe S C, Stevick J, et al. A machine learning approach for engineering bulk metallic glass alloys. Acta Mater, 2018, 159: 102 doi: 10.1016/j.actamat.2018.08.002
[46]	Schultz L E, Afflerbach B, Francis C, et al. Exploration of characteristic temperature contributions to metallic glass forming ability. Comput Mater Sci, 2021, 196: 110494 doi: 10.1016/j.commatsci.2021.110494
[47]	Deng B H, Zhang Y L. Critical feature space for predicting the glass forming ability of metallic alloys revealed by machine learning. Chem Phys, 2020, 538: 110898 doi: 10.1016/j.chemphys.2020.110898
[48]	Li X, Shan G C, Shek C H. Machine learning prediction of magnetic properties of Fe-based metallic glasses considering glass forming ability. J Mater Sci Technol, 2022, 103: 113 doi: 10.1016/j.jmst.2021.05.076
[49]	Xiong J, Shi S Q, Zhang T Y. Machine learning prediction of glass-forming ability in bulk metallic glasses. Comput Mater Sci, 2021, 192: 110362 doi: 10.1016/j.commatsci.2021.110362
[50]	Lu Z C, Chen X, Liu X J, et al. Interpretable machine-learning strategy for soft-magnetic property and thermal stability in Fe-based metallic glasses. Npj Comput Mater, 2020, 6: 187 doi: 10.1038/s41524-020-00460-x
[51]	McCulloch W S, Pitts W. A logical calculus of the ideas immanent in nervous activity. Bull Math Biophys, 1943, 5: 115 doi: 10.1007/BF02478259
[52]	Lu F, Liang Y C, Wang X Y, et al. Prediction of amorphous forming ability based on artificial neural network and convolutional neural network. Comput Mater Sci, 2022, 210: 111464 doi: 10.1016/j.commatsci.2022.111464
[53]	Yu L P, Guo X X, Wang G, et al. Extracting governing system for the plastic deformation of metallic glasses using machine learning. Sci China Phys Mech Astron, 2022, 65(6): 264611 doi: 10.1007/s11433-021-1840-9
[54]	Zhang T, Long Z L, Peng L, et al. Prediction of glass forming ability of bulk metallic glasses based on convolutional neural network. J Non Cryst Solids, 2022, 595: 121846 doi: 10.1016/j.jnoncrysol.2022.121846
[55]	Zeng S M, Zhao Y C, Li G, et al. Atom table convolutional neural networks for an accurate prediction of compounds properties. Npj Comput Mater, 2019, 5: 84 doi: 10.1038/s41524-019-0223-y
[56]	Dan Y B, Zhao Y, Li X, et al. Generative adversarial networks (GAN) based efficient sampling of chemical composition space for inverse design of inorganic materials. Npj Comput Mater, 2020, 6: 84 doi: 10.1038/s41524-020-00352-0
[57]	Cortes C, Vapnik V. Support-vector networks. Mach Learn, 1995, 20(3): 273
[58]	Gunn S R. Support vector machines for classification and regression. ISIS Tech Rep, 1998, 14(1): 5
[59]	Wu Y C, Wang W H, Guan P F, et al. Identifying packing features of atoms with distinct dynamic behaviors in metallic glass by machine-learning method. Sci China Mater, 2021, 64(7): 1820 doi: 10.1007/s40843-020-1626-3
[60]	Cubuk E D, Schoenholz S S, Kaxiras E, et al. Structural properties of defects in glassy liquids. J Phys Chem B, 2016, 120: 6139 doi: 10.1021/acs.jpcb.6b02144
[61]	Sussman D M, Schoenholz S S, Cubuk E D, et al. Disconnecting structure and dynamics in glassy thin films. Proc Natl Acad Sci USA, 2017, 114(40): 10601 doi: 10.1073/pnas.1703927114