Gradient Smoothing Methods
Matlab and Fortran, in-house code, 2024
The GSMs source code is released with the GSMs book.
Software
GPU accelerated mixed-precision SPH framework
CUDA and Fortran, Open source, 2023
Herein, we introduce a GPU-accelerated mixed-precision SPH framework by using low precision FP16 in NNPS while maintaining high-precision FP64 in other components.
Adaptive remeshing for GSM
MATLAB, in-house code, 2022
This software enables adaptive remeshing based on moving material features for Gradient Smoothing Method.
1D Phase Field software
Matlab GUI, Tutorial freeware, 2021
This tutorial freeware (1D Phase Field software) aims at providing students/new beginners a direct computational tool for understanding material microstructure evolution in engineering and material science.
1D diffusion freeware
Matlab GUI, Tutorial freeware, 2021
This tutorial freeware (1D diffusion software) aims at providing students/new beginners a direct computational tool for understanding diffusion type of problems in engineering and material science.
Optimal Controller of microstructure’s evolution
FORTRAN & MATLAB & Python, in-house code, 2021
This software enables optimal control of microstructure’s evolution through coupling phase-field models with reinforcement learning algorithms. The microstructure can be guided from any initial, unstructured state to any target state with the least cost. Please refer to the examples.
Phase-Field platform
FORTRAN & MATLAB, in-house code, 2021
The Phase-Field platform solves the general Allan-Cahn and Cahn-Hilliard equations using Finite Difference Method (with uniform grid) or Gradient Smoothing Method (with adaptive grid). Two numerical solvers are provided: a simple explicit solver and an efficient implicit solver. The in-house code will be released as planned associated with upcoming paper publication.
L-GSM platform
FORTRAN & MATLAB, in-house code, 2021
The L-GSM platform provides a meshfree method for handling large deformation problems with a much better performance in stability and efficiency. Applications can be found here. The in-house code will be released as planned.
Physic-based models for material processing
MATLAB, in-house code, 2019
Three highly efficient physics-based models are developed in Matlab for predicting the mechanical behaviors of material during processing: large deformation, bound strength, and failure. The software is archived by P&G and here are some samples.