Research
Physics-based machine learning
Explore the capability of physics-informed machine learning in establishing underlying rules in engineering data to reduce the lead time in the discovery and design of complex systems in a wide range of applications.
Accelerate reliable atomistic simulations with machine learning force field potentials.
Develop structure-property relationship of hybrid materials with complex microstructures
Predict history-dependent material behaviors to enable design and health monitoring
Microstructure-aware multitask learning
Data fusion for improving model efficiency and fidelity
Uncertainty quantification of machine learning based models to ensure model reliability
Multiscale Multiphysic Simulation
In this research, multi-physics simulations are developed to understand the coupled mechanical-electro-chemical material phenomena that impact the performance of materials and structures, which will generate new knowledge for the design of functional materials and structures.
Corrosion and mechanical performance of dissimilar material joints.
Failure mechanisms and capacity degradation of battery electrodes.
Multiphysics simulation for dissimilar material joints
Mech-electro-chemical coupled simulation for failure model analysis, health monitoring and co-design of battery
Smart Manufacturing
This research focuses on developing physics-informed AI for smart manufacturing
Physics-informed machine learning for automated inline inspection system in electronic assembly.
Physics
Nanomechanics
This research focuses on exploring and developing atomistic simulations towards computationally characterizing the complex nanoscale interfacial behaviors in hybrid functional materials system for establishing design principles of functional materials with unique effective properties.
Develop atomistic simulations towards computationally characterizing the complex nanoscale interfacial behaviors in hybrid functional materials system for establishing design principles of functional materials with unique effective properties.
Guided polymer self-assembly by 2D materials has the great potential to create hybrid van der Waals heterostructures using a bottom up energy-efficient approach. Hybrid van der Waals heterostructures is an excellent candidate for further flexible electronics.
Bio-polymer has the great potential to reduce dependence on petrochemicals, thereby addressing one of the biggies environmental crisis, plastic pollution. Specifically, soy protein has been receiving ever-increasing attentions to be used as a functional material in an effort to modify the properties of polymeric materials, due to its diverse structures and interactions as well as great abundance.
Interface guided polymer assembly for polymer design
Single cell mechanical response for bio-inspired antifoul surface design
Biopolymer design based on soy protein
Sponsors