Evaluation of copper, aluminum, and nickel interatomic potentials on predicting the elastic properties
Summary¶
This Journal of Applied Physics study benchmarks embedded-atom method (EAM)-style interatomic potentials for Cu, Al, and Ni by computing single-crystal elastic constants \(C_{11}\), \(C_{12}\), \(C_{44}\) with molecular dynamics at room temperature and converting them to polycrystalline moduli via the Voigt–Reuss–Hill average. Potentials were taken from NIST, Sandia, and LAMMPS repositories; results are compared to experimental elastic data to recommend which parameterization is most reliable for each pure metal and elastic property. The work is classical non-reactive MD focused on mechanical property prediction rather than ReaxFF chemistry.
The motivation is pragmatic: many downstream simulations inherit EAM files packaged for alloys or defect studies without verifying elastic response against experiment for the pure elements.
Methods¶
1 — MD application (classical, non-reactive). The authors benchmark embedded-atom method (EAM) potentials for fcc Cu, Al, and Ni drawn from NIST, Sandia, and LAMMPS online repositories (as stated in the article). Molecular dynamics at room temperature, described as the most broadly applicable case in their framing, is applied to cubic single crystals; small-strain deformation protocols yield the three independent elastic constants \(C_{11}\), \(C_{12}\), and \(C_{44}\). Voigt–Reuss–Hill averaging converts these to isotropic polycrystalline bulk, shear, and Young’s moduli and Poisson’s ratio for comparison to experimental elastic data tabulated in J. Appl. Phys. Engine, supercell sizes, deformation protocol details, timestep, thermostat, explicit ensemble labels, and equilibration/production lengths are N/A — not restated in the short extract used here; use pdf_path. Reactive MD, electric fields, shear beyond the elastic test, and enhanced sampling are N/A — outside the scope of this elastic benchmark.
2 — Force-field training. N/A — published EAM files are evaluated, not refit.
3 — Static QM / DFT. N/A — DFT is not the production method for the elastic constants reported here.
Bulk fcc supercells use standard PBC. Molecular dynamics elastic benchmarks at room temperature on single-crystal cells use potentials distributed through NIST, Sandia, and LAMMPS repositories (as stated in the article); unless quoted in the primary text you have open, NVT/NPT staging, timestep (fs), equilibration/production run lengths (ps/ns), thermostat and barostat/pressure control, and reported hydrostatic pressure (if any) are N/A — not transcribed from the short extract used here—see pdf_path. Electric field driving and umbrella/metadynamics/replica-exchange sampling are N/A — not part of this elastic study.
Findings¶
Outcomes. Tabulated simulation versus experiment comparisons show that predicted elastic moduli depend strongly on which literature EAM is chosen; potentials developed for alloy or compound fits do not automatically reproduce pure-element elastic behavior.
Design levers. The side-by-side potential comparison is intended as a screening aid before large atomistic studies where elastic response sets stress fields.
The registered pdf_path (papers/Others/NiAl_review.pdf) should be reconciled with the J. Appl. Phys. record if manifests are audited; numerical error budgets are in the paper’s tables.
Limitations¶
- Room-temperature focus; temperature and size effects in nanowires or interfaces may differ from bulk crystal benchmarks.
- The registered PDF path in the corpus manifest should be verified against the JAP article identity if sources are reconciled.
Relevance to group¶
Provides force-field benchmarking context adjacent to reactive and metallic simulation workflows used in broader materials modeling.
Citations and evidence anchors¶
- DOI:
https://doi.org/10.1063/1.4953676(papers/Others/NiAl_review.pdfper manifest).