FaVAD: A software workflow for characterization and visualizing of defects in crystalline structures

The analysis of defects and defect dynamics in crystalline materials is important for fundamental science a wide range applied engineering. With increasing system size the molecular-dynamics simulation data becomes non-trivial. Here, we present workflow semi-automatic identification classification structures, combining new approach description with several already existing open-source software packages. Our addresses key challenges posed by often relatively tiny volume fraction modified parts sample, thermal motion presence potentially unforeseen atomic configurations (defect types) after irradiation. local environment any atom converted into rotation-invariant descriptive vector (‘fingerprint’), which can be compared to known types also yields distance metric suited classification. Vectors cannot associated structures indicate defects. As proof-of-concept apply our method on an iron sample analyze caused collision cascade induced 10 keV primary-knock-on-atom. obtained results are good agreement reported literature values. Program Title: Fingerprinting Visualization Analyzer Defects (FaVAD). CPC Library link program files: https://doi.org/10.17632/bmv9kxkzg3.1 Developer’s repository link: http://gitlab.mpcdf.mpg.de/NMPP/favad.git Licensing provisions: GPLv3. Programming language: Python 3, Fortran 90, C++. Nature problem: damage evolution level commonly analyzed Wigner–Seitz or topology Voronoi tessellation based methods. However, these approaches exhibit specific shortcoming, especially at elevated temperatures. In order improve upon that, more robust quantifiable as well unpredicted desirable. Solution method: A fingerprint-like proposed detail material augmented probabilistic interpretation. It calculation descriptor each sample. These vectors represent compact form individual environments atoms. For standard (i.e. interstitial atoms) corresponding precomputed used rapid Unexpected less common identified applying principal component vectors. Vacancies computing radii largest empty spheres embedded followed thresholding process. This easy use requires only modest computational resources. Finally, classified visualized using open source VisIt. Additional comments including restrictions unusual features: computed command line interface QUIP Gaussian Approximation potential (GAP) package. done scripts make extensive NumPy KDTree2 code employed calculate location single vacancies voids. VisIt visualization point We provide Dockerfile automatically creating portable Docker container installs all programs together script example damaged molecular shell install locally Linux-based server desktop included also.