Atomistic Modelling and Simulation tools for material properties
A Python library for computing material properties from atomistic simulations, developed by the AMS department at ICAMS, Ruhr University Bochum. Built on top of ASE (Atomic Simulation Environment).
- Structural optimization — stepwise, isotropic, and specialized relaxations
- Equations of State — Polynomial EOS fitting
- Elastic constants — full elastic matrix via energy methods
- Phonons — phonon dispersion and density of states via Phonopy
- Surface properties — slab-based surface energy and decohesion calculations
- Generalized stacking faults — gamma line and gamma surface calculators
- Transformation paths — Bain paths and other lattice transformations
- Point defects — vacancy and interstitial formation energies
- Convex hull — thermodynamic stability and hull distance
- High-throughput pipeline — chain calculators, serialize results to JSON, resume interrupted runs
- DFT integration — VASP and FHI-aims wrappers
- Materials Project — download and cache reference structures via
mp-api
pip install .Or for development (editable install):
pip install -e .Verify installation:
import amstools
print(amstools.__path__)conda env create -f environment.yml
conda activate ams
pip install -e .from ase.build import bulk
from ase.calculators.emt import EMT
from amstools import MurnaghanCalculator
atoms = bulk('Al', 'fcc', a=4.05)
atoms.calc = EMT()
calc = MurnaghanCalculator(atoms)
calc.calculate()
print(calc.value) from amstools import StepwiseOptimizer, MurnaghanCalculator, ElasticMatrixCalculator
pipeline = StepwiseOptimizer() + MurnaghanCalculator() + ElasticMatrixCalculator()
pipeline.run(atoms, atoms.calc)Interactive Jupyter notebooks are in the tutorial/ folder,
covering equations of state, phonons, elastic constants, pipelines, and high-throughput workflows.
python -m pytest test/ -vTests use the EMT toy potential — no DFT setup required.
See CONTRIBUTING.md for instructions on adding new property calculators, writing tests, and understanding the pipeline architecture.
If you use amstools in your research, please cite:
Lysogorskiy Y. et al., amstools: Atomistic Modelling and Simulation tools, ICAMS, Ruhr University Bochum (2026). https://github.com/ICAMS/amstools
amstools is published under the Academic Software License (ASL). Free for academic non-commercial use. Contact yury.lysogorskiy@aceworks.works for commercial use. See LICENSE for full terms.
Developed at the Interdisciplinary Centre for Advanced Materials Simulation (ICAMS),Ruhr University Bochum, Germany.