Development of high hardness glasses guided by Machine Learning

• Grant number: 22/15867-9
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: February 01, 2023
• End date: January 31, 2024
• Field of knowledge: Engineering - Materials and Metallurgical Engineering - Nonmetallic Materials
• Principal Investigator: Edgar Dutra Zanotto
• Grantee: André Tamake Yoshioka
• Host Institution: Centro de Ciências Exatas e de Tecnologia (CCET). Universidade Federal de São Carlos (UFSCAR). São Carlos , SP, Brazil
• Associated research grant: 13/07793-6 - CEPIV - Center for Teaching, Research and Innovation in Glass, AP.CEPID

Abstract

Glasses have been selected for numerous domestic and high-tech applications, including fiber optics, bioglass, ion conductors, smartphone screens and ballistic armor. These applications are due to the various properties that some glassy compositions manage to combine, such as aesthetic beauty, transparency, chemical durability and high values of hardness, elastic modulus and resistance to scratching. Artificial intelligence, and especially machine learning, has often been used to predict glass properties based on chemical composition, which notably accelerates the development of new products and greatly increases industrial competitiveness. In this work it is desired to develop high hardness glasses through simulations using machine learning algorithms. For this purpose, predictive models of artificial neural networks (ANN) and genetic algorithms will be used. It is expected that with the correlation between composition and hardness performed by predictive models, it will be possible to predict and produce vitrifiable compositions with optimized mechanical properties, especially hardness. After selecting compositions predicted by the models, with high Vickers hardness (>7.5 GPa) target value for various applications, a study will be carried out, according to the fundamentals of glass science, to certify the ability to form glass, chemical durability and low density. Based on these criteria, various glassy compositions will be simulated, produced and characterized for verification and proof of properties predicted by artificial intelligence algorithms.