pyPFC logo

A Python package for GPU-accelerated 3D Phase Field Crystal simulations¶

PFC atoms

Overview¶

pyPFC is a high-performance simulation framework designed for modeling crystal structures and their evolution using the Phase Field Crystal (PFC) method. The code uses PyTorch to allow execution on both CPUs and GPUs, with particularly efficiency gained on GPU platforms. pyPFC is developed with a combined focus on computational efficiency, flexibility and accessibility, making it suitable for both research and educational purposes. Areas of application include:

Materials Science: Crystal growth and microstructure evolution
Computational Physics: Pattern formation, defect evolution and elastic interactions
Metallurgy: Grain boundary and dislocation mechanics

Key Features¶

GPU Acceleration: Full GPU support using PyTorch for large-scale simulations
Adjustable Floating Point Precision: Permits optimized use of GPU memory resources and execution efficiency
Flexible I/O: VTK, Extended XYZ and pickle format support for data exchange
Advanced Data Processing: Built-in functionalities as well as interfaces to OVITO libraries (optional)
Memory Optimization: Efficient memory management for large 3D grids
Density Field Evolution: Different time integration schemes are available

Getting Started¶

Installation Guide - Setup instructions and dependencies
Quick Start Tutorial - Run your first pyPFC simulation
Examples - Comprehensive examples and use cases
API Reference - Detailed class and method documentation
Troubleshooting - Q&A

Licencing¶

This software is released under a GNU GPLv3 license.

References¶

Further details on PFC modeling and example applications can be found in: