Designing Fe-based high entropy alloy–a machine learning approach. Issue 5 (August 2020)

• Record Type: Journal Article
• Authors: Debnath, B; Vinoth, A; Mukherjee, M; Datta, S
• Is Part Of: IOP conference series. Volume 912:Issue 5(2020)
• Publication Date: 2020-08
• Subjects: Materials science -- Periodicals; 620.1105
• DOI: 10.1088/1757-899X/912/5/052021
• Languages: English
• ISSNs: 1757-8981

Abstract

Abstract: High Entropy Alloys (HEAs) are constituted by at least five elements and can even increase to seven or eight different elements. Due to high entropy of mixing, the solid solutions of so many elements become stable and the tendency to form intermetallic compounds decrease. As a result, it is possible to develop alloys with high strength and high hardness using this approach. Changing the composition of such alloys, the mechanical properties of the alloy can be varied widely. In this work HEAs with high toughness are designed computationally using machine learning and artificial intelligence approaches. With so much potential in this new breed of alloys, iron-based alloys (having high iron content) is designed in the present work to reduce the cost of the alloy. Here we have used supervised machine learning technique to map the relation between composition and properties of HEA. Multi-objective optimization is employed to search suitable composition for Fe-based HEA having increased strength and ductility, which will lead to improved toughness of the alloy.