
I am a Mechanical Engineering PhD Candidate in Micro and Nano Mechanics Group at Stanford University. My current work focuses on understanding how material defects and microstructures govern macroscopic mechanical properties using computational approaches, such as metal strain hardening and homogenization of digital rocks. I was a Research Scientist Intern at Tokyo Electron working on computational modeling of semiconductor engineering (3DI). Previously, I got my MS in Mechanical Engineering from Cornell University and BS in Theoretical and Applied Mechanics from Shanghai University. I'm from Beijing.
I am interested in solving problems in mechanics of materials from atomistic to continuum scales. My research interests are combining computational mechanics, computational materials science, scientific machine learning, and inverse optimization for structural and materials modeling and design, with potential impact on energy, biotechnology, and advanced manufacturing.
[Ad for Stanford students]: Please consider enrolling in ME335A Finite Element Analysis, for which I am to be the TA.
![]() |
Zhai.
Stress predictions in polycrystal plasticity using graph neural networks with subgraph training. Computational Mechanics (2025).
TL;DR: GNNs are used to efficiently predict stress in polycrystal plasticity, achieving over 150x speedup compared to FEM with high accuracy. |
![]() |
Zhai & Yeo.
Computational design of antimicrobial active surfaces via automated Bayesian optimization. ACS Biomat. Sci. & Eng. (2023).
TL;DR: A new workflow to digitally design antimicrobial surfaces using Bayesian optimization. |
![]() |
Zhai et al.
Benchmarking inverse optimization algorithms for materials design. APL Materials (2024).
TL;DR: We benchmarked optimization algorithms for designing elemental crystals for targeted properties. |
![]() |
Zhai & Yeo.
Controlling biofilm transport with porous metamaterials designed with Bayesian learning. J. Mech. Behav. Biomed. Mat. (2023).
TL;DR: We use Bayesian optimization to design porous materials with enhanced biomass transport and insights into growth mechanisms. |
2025: We present our recent work on a new theory of dislocation link statistics in strain hardening at CompFest 2025 at Berkeley. | ![]() |
![]() |
![]() |
![]() |
![]() |