Seminar by Prof. Tomohiro Takaki from Kyoto Institute of Technology, Japan, entitled “Phase-Field Modeling and Simulation of Materials: Data-Driven Approaches, Multiphysics, and High-Performance Computing” – on 26/10/2026.

Abstract:

The phase-field (PF) method is a powerful tool for simulating microstructure evolution in materials.
However, its predictive applications remain limited by uncertainties in material parameters, challenges
in multi-physics modeling, and high computational costs. This talk introduces four topics from our
group’s recent research, spanning data assimilation, machine learning, multi-physics simulation, and
GPU-based large-scale computing.


First, for alloy solidification, we are developing a data assimilation approach that combines PF
simulations with in situ X-ray observations of dendrite growth. This approach aims to estimate
unknown material parameters and reconstruct four-dimensional microstructure evolution from
sequential X-ray transmission images [Acta Mater. 281 (2024) 120356].


Second, we are developing machine-learning models trained on data generated by PF simulations to
estimate material parameters and reconstruct three-dimensional microstructures, both from twodimensional cross-sectional images [IOP Conf. Ser.: Mater. Sci. Eng. 1335 (2025) 012014].


Third, for metal additive manufacturing, we have developed a coupled PF–lattice Boltzmann model
for simulating melt-pool dynamics in laser powder bed fusion, as well as a moving-frame algorithm
for efficient PF simulations of grain growth under various laser scan strategies [IOP Conf. Ser.: Mater.
Sci. Eng. 1335 (2025) 012016; Comput. Mater. Sci. 261 (2026) 114265].


Finally, for solid-state sintering, we have developed accurate models for particle rigid-body motion,
efficient computational methods for large-scale simulations, and a pseudo-periodic boundary
condition that accommodates macroscopic shrinkage while suppressing artificial surface effects [J.
Mater. Res. Technol. 34 (2025) 1803–1816; Mater. Des. 265 (2026) 115858].


The aim of this talk is to provide a broad overview of our group’s research activities and to encourage
further in-depth discussions on topics of mutual interest.

Bio:

Tomohiro Takaki is a professor in the Faculty of Mechanical Engineering and director of the High-Performance Simulation Research Center at Kyoto Institute of Technology. He received his Ph.D. from Kobe University of Mercantile Marine in 2003. After holding research positions at Kobe University of Mercantile Marine and Kobe University,he joined Kyoto Institute of Technology as an associate professor in 2007 and became a professor in 2019. His research focuses on phasefield modeling and simulation of microstructure evolution in materials, GPU-based high-performance computing, and data-driven approaches that integrate simulations with experiments.