Page 73 - Read Online
P. 73
Hu et al. J Mater Inf 2023;3:1 I http://dx.doi.org/10.20517/jmi.2022.28 Page 13 of 15
such a model persists or not. Extending the physical mechanism and model prediction of glass formation in
single-component and binary systems to multi-component systems is an intrinsically important and intrigu-
ing direction for future work. Another interesting related topic would be machine learning study of the phase
selection of high-entropy alloys. With mainly five elements of similar size, which is close to one set of our
current simulations, high-entropy alloys usually form a finite number of simple crystals. This is an ideal case
as a classification problem. The driver of the phase selection and local chemical ordering is the energetic com-
petition. By using a similar simulation protocol or with an advanced patchy particle model, these issues can
be well tackled by combining computer simulations and machine learning methods.
DECLARATIONS
Acknowledgments
Hu YC has been focusing on the computational study of the glass-forming ability of metallic glasses since
he started his postdoc research with Prof. Corey O’Hern at Yale University in 2018. At Yale, he carried out
most of the simulations to quantify the critical cooling rates of thousands of systems and studied the statistical
physics of the glass-forming ability. Supported by a JSPS fellowship, Y.C.H. has also worked with Prof. Hajime
Tanaka at the University of Tokyo to study the atomic-scale structural mechanism of glass formation and
crystallization of binary alloys. Without these experiences and thoughts, this work would have never come to
fruition. Y.C.H. is very grateful for all the support from all the members of the O’Hern lab and the Tanaka lab.
Hu YC acknowledges the technical support from Yale Center for Research Computing. Hu YC thanks Y.C. Wu
for carefully reviewing the manuscript before submission.
Authors’ contributions
Proposed and supervised theproject, conducted the simulations, built themachinelearning model, performed
the analysis, and wrote the manuscript: Hu YC
Contributed to generalizing the machine learning codes and performing analyses to respond to the referees:
Tian J
Availability of data and materials
The dataset and the machine learning codes are available at the high-quality dataset and the machine learning
package at https://github.com/yuanchaohu/ML_GFA_JMI or at https://github.com/jzt5132/ml_code.
Financial support and sponsorship
None.
Conflicts of interest
All authors declared that there are no conflicts of interest.
Ethical approval and consent to participate
Not applicable.
Consent for publication
Not applicable.
Copyright
© The Author(s) 2023.
REFERENCES
1. Klement W, Willens RH, Duwez P. Non-crystalline structure in solidified gold–silicon alloys. Nature 1960;187:869–70. DOI
2. Lu ZP, Liu CT. A new glass-forming ability criterion for bulk metallic glasses. Acta Mater 2002;50:3501–12. DOI