Unit of loading Potentialfield

[TeetotalingTom]

Summary

The datar from a PotentialField was erroneously being rescaled upon loading. the fields provided by abinit are intensive, but unlike the density, they are not defined "per unit volume". Thus, no rescaling from the default length format (Bohr radii) to the default Pymatgen format (Angstrom) is required.

Beforehand, this would be the plane-averaged total potential of an 11-layer aluminium slab cut along (100) and referenced to the Fermi level:

here one would mistakenly infer that the work function of this slab is almost 50 eV.

This is the result with the corrected code:

Reaching a work function of approximately 4.2 eV, consistent with the literature for 11-layer slabs with the PBE functional.

As a minor detail, the code now loads the only two type conversions bohr_to_angstrom and Ha_to_eV that are being used.

[TeetotalingTom]

After ensuring that the potentials are correctly loaded in core/tests/test_fields.py, the relevant test is passed accordingly. The test flowtk/tests/test_vzsisa.py still fails, but this is due to the QHA code and unrelated to the current MR.

[gmatteo]

Hello Tom, thank you for fixing the bug. Regarding the error in test_vzsisa.py, I’ll take care of it.