How to upload custom kinetic Monte Carlo simulation data to NOMAD (no built-in parser)?

NOMAD and FAIRmat
1

I’m trying to archive kinetic Monte Carlo (kMC) simulation results for thin-film growth of fcc metals (Ag, Au, Cu, Ni, Pd, Pt) on (111) surfaces. The simulations were performed with a custom Python kMC code (not a supported package like VASP or LAMMPS).

I’ve prepared structured JSON files per simulation:

`{`
`“simulation_id”: “Ag_deposition_431K”,`
`“conditions”: {`
`“process_type”: “deposition”,`
`“temperature_K”: 431,`
`“partial_pressure_Pa”: 113,`
`“sticking_coefficient”: 1`
`},`
`“system”: {`
`“chemical_formula”: “Ag”,`
`“structure_type”: “thin_film”,`
`“crystal_system”: “cubic”,`
`“space_group”: “Fm-3m”,`
`“lattice_type”: “fcc”,`
`,`
`“simulation_domain_angstrom”: [`
`60.32147284415487,`
`58.04436430747917,`
`56.871629999016115`
`],`
`“materials_project_id”: “mp-124”`
`},`
`“workflow”: {`
`“type”: “kinetic_monte_carlo”,`
`“code”: “Custom Python kinetic Monte Carlo simulator”`
`}`
`}`

and final_structure.json (atomic positions):

{
  "lattice_vectors": [[60.32,0,0], [0,58.04,0], [0,0,56.87]],
  "atomic_positions": [[0.0,0.0,0.0], [2.89,2.89,0.0], ...],
  "species": ["Ag","Ag",...],
  "pbc": [true,true,true\]
}

Problem: When I upload a ZIP containing these files to NOMAD, the files appear under “Files” but processing stalls at “Step 1: Upload”, no metadata parsing, no 3D viewer, and no error messages.

What is the minimal valid file structure for NOMAD to recognize a custom kMC simulation and activate the 3D viewer?

This data accompanies a recently published paper on substrate-controlled morphology of fcc metals. I’ve already archived the full dataset on Zenodo and would like to complement it with NOMAD entries for discoverability and interoperability.

2

Hello @aldanads

Upload of custom simulations is possible via the H5MD-NOMAD schema and file format, described in details HERE, with examples on the central NOMAD deployment HERE

Note that within this framework you are free to store any custom values / parameters you wish, however, the parsed values (i.e., those extracted from your raw upload files and stored within the unified database structure) are limited to quantities that are already included in NOMAD’s simulation schema. In particular relevant for you: We do not yet have explicit schema support for MC simulations (i.e., the detailed MC runtime parameters will not be parsed, although they will still be stored within the entry in the updated hdf5 file).

If you require further assistance with this approach, I suggest joining the NOMAD Discord server and posting in the #forum-comp (an issue forum within the Computational category). There you can get assistance must more quickly and connect directly with our entire development team.

Best,

Joseph Rudzinski (Coordinator of Theory & Computation, FAIRmat)

3

Thank you so much for your fast response!