Production and Characterization of High Entropy Alloy for Validation of Artificial Intelligence Method

• Grant number: 24/01404-2
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: March 01, 2024
• End date: February 28, 2025
• Field of knowledge: Engineering - Materials and Metallurgical Engineering - Physical Metallurgy
• Principal Investigator: Francisco Gil Coury
• Grantee: Roberta Benedita Costa
• Host Institution: Centro de Ciências Exatas e de Tecnologia (CCET). Universidade Federal de São Carlos (UFSCAR). São Carlos , SP, Brazil
• Associated research grant: 22/02770-7 - Integrated Computational Materials Engineering ICME: applied to modeling, production, characterization and testing high entropy alloys, AP.PNGP.PI

Abstract

Artificial intelligence algorithms play a crucial role in the learning process by analyzing and training predictive models based on data. The quality and diversity of data used in training are determining factors in the predictive capacity of the model. The more comprehensive and varied the data, the greater the accuracy and flexibility of the algorithm. In this specific project, we will use high-performance CALPHAD-generated data to train an artificial intelligence model. This model will later be employed in alloy selection, in conjunction with a genetic algorithm. The main objective of the project is the selection and manufacturing of three alloys using this combination of algorithms. We will test different training configurations for the genetic algorithm and artificial intelligence. This will allow us to challenge the algorithms in various ways, thus identifying their limitations and potential new promising alloy compositions. Three alloys will be selected and manufactured through electric arc furnace casting using high-purity elements. We will then conduct characterizations in two conditions: "as-cast" and "homogenized." The characterizations will include techniques such as optical microscopy, electron microscopy, and X-ray diffraction. This way, we will have a comprehensive overview of the alloy properties and the employed algorithms, enabling us to evaluate their potential for future applications.The project in question will use data generated by high-performance CALPHAD, guaranteeing a wide range of information. It is also worth highlighting that data quality may vary between systems and some databases may not be completely described for all binaries and ternaries, so that, to improve the reliability of the artificial intelligence code, the project counts on with the choice of three alloys with incomplete descriptions in the database, so that after their subsequent production and characterization, it is possible to obtain the necessary data to expand the reliability of the model.In this sense, the project aims to select and manufacture three alloys, which will be manufactured by casting in an electric arc melting furnace, using high purity elements. The resulting samples will be characterized under "raw fusion" and "homogenized" conditions, in order to identify the phases present in each of them. Prototype phases will be based on similar systems described in the literature and additional trials will be listed as necessary.