A group of Scholtech scientists used machine learning (ML) methods to predict superhard materials based on their crystal structure.
Posted in Journal of Applied Physics.
Superhard materials have recently attracted increasing research interest due to their potential implications for industries ranging widely from oil production to high technology manufacturing. A superhard material has two important characteristics, stiffness and fracture toughness, which represent resistance to deformation and crack propagation, respectively.
Materials with properties suited to specific industry requirements can be found computationally using advanced methods of computational materials science which is supported by a good theoretical model to calculate the desired properties for superhard materials.
Ephim Mazhanik, a Ph.D. Artem R., a professor at Scholtech and MIPT, a student at the Scholtech Center for Energy Science and Technology (Computational Materials Discovery Laboratory). Guided by Oganov, was successful in building such a model using a methodological neural network (CNN) on the graph, an ML method that enables to predict the properties of a material from its crystal structure. By using a set of materials with known properties, you can first teach CNN to calculate those properties for unfamiliar structures.
“Facing a lack of experimental data on stiffness and fracture stiffness to properly train the model, we turned to more abundant data on elastic moduli and properties sought using the physical model we had previously built To achieve their predicted values, “says. Efim Mazhanik.
“In this study, we applied ML methods to calculate hardness and fracture toughness for more than 120,000 crystal structures, both known and hypothetical, most of which have never been detected in terms of these properties. While our model confirms that diamond is the hardest known material. It suggests the existence of several dozen other potentially very hard or superhard materials, “says Artem Oganov.
Scientists predict new, harder and superhard ternary compounds
Ifim Mazhanik et al. Application of machine learning methods for the prediction of new superhard materials, Journal of Applied Physics (2020). DOI: 10.1063 / 5.0012055
Skolkovo Institute of Science and Technology is provided by
Quotes: Scientists Predict New Superhard Material (2020, 10 September) Retrieved 11 September 2020 from https://phys.org/news/2020-09-scientists-superhard-materials.html
This document is subject to copyright. No part may be reproduced without written permission, except for any impartial behavior intended for personal study or research. The content is provided for information purposes only.