Metallurgy-Informed Bayesian Optimization: Rethinking Alloy Discovery and Design

Abstract: The discovery of advanced alloys for extreme environments remains one of the grand challenges of materials science. Conventional approaches to alloy development are often slow, costly, and limited by the vast complexity of the chemistry–processing–structure–property design space. In this talk, I will discuss recent progress in accelerating alloy discovery through metallurgy-informed Bayesian optimization. By integrating physical models, high-throughput CALPHAD simulations, microstructural knowledge, and experimental feedback, we transform materials discovery from a “black box” search into a “gray box” process that leverages prior scientific understanding. I will present results from multi-objective, multi-constraint optimization campaigns in high-entropy alloys and refractory systems, highlighting strategies to incorporate feasibility awareness, property correlations, microstructural sensitivity, and chemistry/process intuition into the optimization loop. The talk will also address the challenges and opportunities in bridging autonomous experiment design, human expertise, and metallurgical insight, ultimately charting a path toward self-driving metallurgical laboratories for accelerated, physics-informed alloy innovation.

This event is sponsored by FAMU-FSU College of Engineering and Department of Materials Science and Engineering.

Raymundo Arróyave, Ph.D.

Chevron Professor II

Department of Materials Science and Engineering 

Texas A&M University

Speaker Bio: Raymundo Arróyave is the Chevron Professor II, Segers Family Dean’s Excellence Professor, Presidential Impact Fellow, and Chancellor EDGES Fellow in the Department of Materials Science and Engineering at Texas A&M University, where he also holds joint appointments in Mechanical Engineering and Industrial and Systems Engineering. He earned his Ph.D. in Materials Science from MIT in 2004, followed by a postdoctoral appointment at Penn State, and joined Texas A&M in 2006. His research focuses on computational and AI-assisted materials design, spanning thermodynamics, kinetics, and machine learning approaches to accelerate alloy discovery and unravel process–structure–property relationships. He has published over 300 peer-reviewed papers and leads several large-scale multidisciplinary projects on alloy discovery for extreme conditions. His contributions have been recognized with the Acta Materialia Silver Medal (2023), the TMS Brimacombe Medal (2019), and election as Fellow of ASM International (2020). He also serves as Editor-in-Chief of Materials Letters and Associate Editor for multiple leading journals, and has delivered more than 150 invited talks worldwide.