Empowering Alloy Design with AI and CALPHAD: Applications in Recycled Aluminum

Guest Speaker：Song-Mao Liang Song-Mao Liang，CompuTherm LLC, USA

Inviter: Prof. Da Shu

Date&Time: Monday, 14 July, 9:30-11:00

Venue: Meeting Room 308, Xu Zuayo Building

 Biography:

Dr. Song-Mao Liang is currently a Senior Materials Scientist at CompuTherm LLC. He earned his B.Eng. from Beijing University of Chemical Technology and a Ph.D. through a joint program between the Institute of Metal Research, Chinese Academy of Sciences, and Brunel University London. He worked at Clausthal University of Technology (Germany) and the University of Wisconsin-Madison (USA), with over a decade of research experience in phase diagrams, alloy thermodynamics, and kinetic simulations. He currently serves as Chair of the Alloy Phases Committee at The Minerals, Metals & Materials Society (TMS, USA). To date, he has published more than 40 academic papers with 1200+ citations. He serves as a reviewer for multiple international journals and has received the "Outstanding Contribution in Reviewing Award" multiple times from journals including Materials & Design and Calphad.

 Abstract:

Integrated Computational Materials Engineering (ICME) has emerged as a powerful approach to accelerate materials development by linking composition, processing, microstructure, and performance quantitatively. The CALPHAD method has evolved from its original focus on thermodynamics and phase diagrams into a comprehensive simulation framework that includes diffusion, solidification, precipitation kinetics, and more.  This seminar will introduce the Pandat software, a CALPHAD-based platform for computational materials design. Two case studies will demonstrate an AI-assisted, high-throughput CALPHAD framework for analyzing how alloying elements affect phase formation in the Al–Si–Fe–Mg–Mn system—a critical area for impurity control and solidification behavior in secondary aluminum alloys. By integrating AI with CALPHAD, this approach enables materials researchers—even those without extensive coding experience—to fully utilize CALPHAD in alloy design and process optimization.