Northwestern University · Mechanical Engineering
prefer to be called Alex
I build machine-learning methods and high-rate experimentation for material and mechanics characterization, and further use that knowledge to achieve structural design in extreme environments.
Integrating scientific machine learning, high-strain-rate experiments, and physics-based computational mechanics — from material behavior to large-scale, multi-physics design under uncertainty.
About
I am a Research Associate in Mechanical Engineering at Northwestern University, working with Prof. Wei Chen on autonomous material discovery with AI. Previously, I was a postdoctoral researcher at UC Irvine, a research scientist at Corning Science & Technology, and a research associate at UW–Madison. I earned my Ph.D. in Mechanical Engineering (Mechanics of Materials) from Johns Hopkins University.
Research
My research asks a single guiding question: how can advanced experimental techniques and a broad spectrum of computational models be systematically integrated into engineering design? To answer it, I combine scientific machine learning with high-strain-rate experimentation and computational mechanics across four connected thrusts.
Material mechanics under extreme dynamic environments via high-rate, multi-axial experiments.
Physics-based surrogate models of nonlinear constitutive behavior with quantified uncertainty.
Efficient computational implementation of advanced models for large-scale simulation.
Integrated multi-material, multi-physics design across high-dimensional design spaces.
Four connected thrusts
A scalable, mesh-free neural framework for stress-constrained topology optimization — using additive gradients to push physics-informed structural design to large-scale problems that classical solvers struggle to reach.
Physics-informed Gaussian-process priors as a mesh-free surrogate for FEM, with continuous adjoint sensitivity analysis — enabling compliance minimization and multi-material, multi-physics design under thermo-mechanical loads.
Pressure-shear plate-impact experiments driving ceramic powders to ultra-high strain rates (10⁶–10⁷ s⁻¹) under combined normal and shear loading — revealing deformation and fragmentation mechanisms and calibrating advanced constitutive models.
Rate- and temperature-dependent characterization of ultrafine-grained titanium and ECAE-processed Mg-Al alloys across a broad range of strain rates, linking microstructure to hardening behavior and dynamic mechanical response.
Publications
Author names with Sun, X. in bold. Selected journal articles below; conference proceedings and works in progress follow.
Under review / in preparation
Conference proceedings
Recognition
Dissemination
Path so far
Appointments
Education
Get in touch
Interested in scientific ML for mechanics, high-rate experiments, or physics-informed design? I'm always happy to connect.