HMA

[ajschult]

Description

This branch implements harmonically mapped averaging (HMA) for thermodynamic properties of crystals (potential energy and pressure). HMA is enabled by listing energy_HMA and/or pressure_HMA as desired properties in the input file. The initial configuration must be the lattice so that the code knows how to compute displacements from lattice sites.

Related Issue

Issue #94

How Has This Been Tested?

I've run Cassandra LJ FCC crystals with various temperatures and cutoffs and found results that match other codes. I've add a 500-atom simulation input file to examples/NVT/HMA.

Backward Compatibility

(none)

Post Submission Checklist

Please check the fields below as they are completed

• Suitable new documentation files and/or updates to the existing docs are included.
• One or more example input decks are included.

Further Information, Files, and Links

HMA paper: https://doi.org/10.1103/PhysRevE.92.043303
HMA in LAMMPS paper: https://doi.org/10.1063/1.5129942
HMA for VASP paper: https://doi.org/10.1016/j.cpc.2020.107554
LAMMPS pull request: lammps/lammps#1503

[rsdefever]

Thanks for the PR! I'll try to get through it in the next week or so.

When you get a chance, can you add the required input file sections to our ReadTheDocs pages here?

Also, I think the AZP issues are not from your PR. I have a PR open to fix those (#102); perhaps rebase your PR on that once I merge it into master.

[ajschult]

The CI runs merged into current master, so it's seeing your AZP fixes.

I added the properties to the list in the docs. It seems like there's still not much in terms of explaining HMA. Is there some section of the docs where that ought to go?

[rsdefever]

The CI runs merged into current master, so it's seeing your AZP fixes.

Right, good catch.

I added the properties to the list in the docs. It seems like there's still not much in terms of explaining HMA. Is there some section of the docs where that ought to go?

Yeah, I was struggling to decide this too. The docs need some reorganization. Can you create a section under "Guides" for now that covers HMA. At this point my primary goal is just to get the content in place. I don't think you need an in-depth description of the theory of the method, but perhaps a quick overview and point to the papers. Then describe how to use the HMA features you have implemented, any limitations of the implementation, and point to the example that you have created, etc.