Beyond Equilibrium: Modeling Structurally and Chemically Complex Materials via Energy Landscape Navigation

This seminar highlights the unique capability of energy landscape-based modeling to provide a fundamental, predictive understanding of the behavior of non-equilibrium materials with significant structural and chemical complexity. I will present two case studies illustrating how this modeling approach can reveal insights that are inaccessible to conventional methods.

First, I will discuss metallic glasses — a disordered and inherently non-equilibrium metastable material system. Due to the lack of long-range order, building a valid structure-property relationship in glasses has been a longstanding challenge. I will show that by scrutinizing atomic reconfiguration processes—in particular the competition between elementary activations (up-hill climbing) and relaxations (down-hill dropping) on the energy landscape—a self-consistent equation can be derived to describe the time evolution of these disordered materials under various conditions. This, in turn, allows for the explanation and prediction of many critical phenomena in glassy systems—such as the aging/rejuvenation crossover and thermo-mechanical hysteresis—without relying on too many empirical assumptions or fitting parameters adopted in classical models.

Second, I will show that even in structurally ordered crystalline materials, non-equilibrium processing can induce chemical complexity that modifies the energy landscape in ways that significantly deviate from the classical linear-tilting picture. Using Al-Si-Mg alloys as an example, we integrated realistic energy landscape sampling of various local chemical environments with machine learning and a kinetic Monte Carlo (kMC) framework. This combined approach uncovered new microstructural evolution pathways—specifically, the formation of non-conventional nanoscale precipitates—that are observed in advanced manufacturing experiments (e.g. selective laser melting, high-pressure die casting) but are missed by conventional kMC models.

Additional Information:

Yue Fan is currently an Associate Professor at University of Michigan, Ann Arbor. He received his Ph.D. degree from MIT in 2013, and then worked at Oak Ridge National Lab as a Eugene P. Wigner Fellow from 2013 to 2016. His primary research interest is to provide a substantive knowledge on mechanics and microstructural evolution in complex systems via predictive modeling, and thus facilitate the development of new science-based high performance materials with novel functions and unprecedented strength, durability, and resistance to traditional degradation and failure. Some honors and recognitions he has received include “TMS-JIMM International Scholar”, “TMS MPMD Young Leaders Professional Development Award”, “NSF Career Award”, “Ralph E. Powe Junior Faculty Enhancement Award” (by ORAU), and “Haythornthwaite Young Investigator Award” (by ASME-Applied Mechanics Division).