How hherpa/hea_l12 works
Purpose and scope
This is a notebook- and file-driven computational-materials workflow for the L1₂ strengthening phase of the Al–Co–Cr–Fe–Ni–Ti high-entropy alloy. It progresses from disordered crystal-structure generation through first-principles calculations to MEAM-based molecular dynamics. It is not an application service; execution entry points are Jupyter notebooks and simulator input/run files.
Structure generation
SQS generation produces L1₂ CIF structures at 32, 64, and 128 atoms, providing the principal structural inputs for downstream simulation:
• CIF/SQS generator/HEA_L12_SQS_32atoms.ipynb
• CIF/SQS generator/HEA_L12_SQS_128atoms.cif
• CIF/SQS generator/README.md
The key boundary is exported CIF: it captures lattice, atomic positions, species, and sublattice disorder independently of the generating notebook. An archived alternative uses pymatgen random structures: CIF/archive_random_pymatgen_structure/HEA_L12_random_pymatgen_structure.ipynb.
DFT workflow
The primary DFT entry point is DFT/HEA_L12_QE_32atoms.ipynb, which represents a 32-atom Quantum ESPRESSO calculation path. QE consumes structure-derived inputs and local pseudopotentials from:
• DFT/pseudo/
• DFT/pseudo/Al.pbesol-n-kjpaw_psl.1.0.0.UPF
• DFT/pseudo/Ni.upf
GPU-oriented GPAW, PySCF, and QE notebooks are explicitly archived, so they are alternative/prototype backends rather than the primary path: DFT/archive_DFT_GPU/.
Reference-system DFT and analysis
Ni, Al, and Ni₃Al are maintained as interpretable reference calculations for validating formation/electronic-property workflows. QE inputs, results, and analysis are colocated:
• DFT/example_Ni3Al/QE_run_01/QE_run.ipynb
• DFT/example_Ni3Al/QE_run_01/Ni3Al/Ni3Al.pw.in
• DFT/example_Ni3Al/QE_run_01/Ni3Al_QE_analysis.ipynb
QE output boundaries include optimized structures (*_opt.cif), XML, charge/Hartree fields, and projected DOS, enabling postprocessing without rerunning the solver.
VASP calculation corpus
A separate VASP-based reference corpus organizes multiple runs by material and calculation stage. Each case has standard VASP inputs, a shell execution entry point, and persisted outputs:
• DFT/example_Ni3Al/run_01/Ni3Al/run.sh
• DFT/example_Ni3Al/run_02/Ni3Al/INCAR
• DFT/example_Ni3Al/run_06/run_06_VASP_analysis.ipynb
This is parallel to, not automatically integrated with, the QE workflow. CONTCAR, DOSCAR, OUTCAR, vasprun.xml, and—where present—charge/local-potential files form the interface to analysis.
Molecular dynamics
LAMMPS performs thermal treatment and uniaxial tensile simulation using a multicomponent MEAM potential. The executable input is:
• LAMMPS/run.hea
• LAMMPS/FeMnNiTiCuCrCoAl.meam
• LAMMPS/library.meam
Its flow is anneal/cool/load simulation → trajectory/log and tabular measurements → notebook postprocessing. Existing outputs, including per-stage CSV files and stress–strain data, are retained in LAMMPS/logs/.
Runtime and integration boundaries
Required material-science runtimes are Python/Jupyter, sqsgenerator, pymatgen, Quantum ESPRESSO, and LAMMPS; GPAW/PySCF and VASP appear as supported experimental/reference backends. Compute infrastructure is external to the repository: notebooks are suitable for local/Colab execution, while run.sh files imply shell/HPC-style VASP execution. Data handoffs are file based—CIF/POSCAR structures, DFT solver inputs/outputs, and LAMMPS potential/trajectory/CSV artifacts—with no orchestrator that automates the full pipeline.
Main entry points: HEA_L12_QE_32atoms.ipynb HEA QE calculation · HEA_L12_SQS_128atoms.cif L1₂ SQS CIFs · QE_run.ipynb Ni₃Al QE reference · run.sh VASP shell run
Components 18
Structure generation
- SQS generator · Generates disordered L1₂ structures.
CIF/SQS generator/HEA_L12_SQS_32atoms.ipynb - L1₂ SQS CIFs · 32-, 64-, and 128-atom structural handoff boundary.
CIF/SQS generator/HEA_L12_SQS_128atoms.cif - pymatgen prototype · Alternative random-structure generator.
CIF/archive_random_pymatgen_structure/HEA_L12_random_pymatgen_structure.ipynb
First-principles workflows
- HEA QE calculation · Primary 32-atom first-principles path.
DFT/HEA_L12_QE_32atoms.ipynb - QE pseudopotentials
- QE result artifacts · Optimized CIF, XML, fields, and projected DOS support postprocessing.
- Ni₃Al QE reference · Interpretable reference calculation.
DFT/example_Ni3Al/QE_run_01/QE_run.ipynb - Ni₃Al QE input
DFT/example_Ni3Al/QE_run_01/Ni3Al/Ni3Al.pw.in - Ni₃Al QE analysis
DFT/example_Ni3Al/QE_run_01/Ni3Al_QE_analysis.ipynb - GPU DFT prototypes · GPAW, PySCF, and alternative QE experiments.
VASP reference corpus
- VASP shell run · Shell/HPC-style VASP entry point.
DFT/example_Ni3Al/run_01/Ni3Al/run.sh - VASP case inputs · Material and calculation-stage specific inputs.
DFT/example_Ni3Al/run_02/Ni3Al/INCAR - VASP result artifacts · CONTCAR, DOSCAR, OUTCAR, vasprun.xml, and field files.
- VASP analysis
DFT/example_Ni3Al/run_06/run_06_VASP_analysis.ipynb
Molecular dynamics
- Thermal and tensile run · Defines anneal, cooling, and uniaxial loading.
LAMMPS/run.hea - MEAM potential set · Multicomponent alloy interaction model.
LAMMPS/FeMnNiTiCuCrCoAl.meam - MD logs and measurements · Per-stage measurements and stress–strain data.
- MD postprocessing
Connections 15
- SQS generator → L1₂ SQS CIFs: exports
- pymatgen prototype → L1₂ SQS CIFs: alternative export
- L1₂ SQS CIFs → HEA QE calculation: structure input
- QE pseudopotentials → HEA QE calculation: solver input
- HEA QE calculation → QE result artifacts: runs QE
- Ni₃Al QE input → Ni₃Al QE reference: calculation input
- Ni₃Al QE reference → QE result artifacts: produces reference results
- QE result artifacts → Ni₃Al QE analysis: postprocesses
- GPU DFT prototypes → HEA QE calculation: alternative backend
- VASP case inputs → VASP shell run: run inputs
- VASP shell run → VASP result artifacts: executes VASP
- VASP result artifacts → VASP analysis: postprocesses
- MEAM potential set → Thermal and tensile run: potential input
- Thermal and tensile run → MD logs and measurements: anneal, cool, load
- MD logs and measurements → MD postprocessing: trajectory and CSV