| Abstract Scope |
Integrated Computational Materials Engineering (ICME) has emerged as a powerful tool for materials design in manufacturing. While thermodynamic models like Calphad are widely used in ICME, their accuracy in capturing complex process physics and microstructure-property relationships remains a challenge. This limitation can be addressed by incorporating machine learning (ML) models, which can compensate for the missing physics. Here we present two case studies demonstrating the efficacy of integrating ML into Calphad-based ICME for additive manufacturing design. In the first case, we developed a heat treatment design for wire-arc additive manufacturing of Haynes superalloy to enhance strength. The second case explores the design of functionally graded alloys, transitioning from steel to superalloy. Our results demonstrate that integrating ML into ICME significantly enhances the design effectiveness of additive manufacturing processes. This approach allows for accurate prediction of structure-property correlations, identifying a pathway for an efficient and customized material design in advanced manufacturing. |