[ase-users] VASP module in ASE

[Chaitanya Krishna]

Hi Jonas,

> Thank you, yes we need to think about this tag. My idea is the
> following:
>
> In a standard case where the ionic relaxation/molecular
> dynamics/vibrational study/etc... is taken care by ASE, Vasp does not
> need to know about the constraints since only several statics
> calculations will be done in Vasp and ionic positions are updated by
> ASE. This makes it possible to make use of the more flexible constraints
> which ASE offers.
But I think ASE should not restrict what is already available in VASP.
Adding to what is already available is always nice though.

> However, if the user prefer to let Vasp deal with the ionic relaxation
> or molecular dynamics (for example by setting ibrion=2 and nsw>0) the
> constraints should be written out explicitly to the POSCAR file, so
> should the initial velocities (which I totally forgot about).
You might already be knowing, but just to be clear ...

Selective dynamics is just a special case if you want to do a
constrained optimization, or a constrained molecular dynamics
simulation.

Constraints:
It's not necessary to write out the constraints on the atoms
explicitly all the time. You need to mention the constraints only if
you use the Selective Dynamics tag.

Velocities:
The velocities can be skipped for all of the IBRION tags. If you are
doing a MD (IBRION=0) then you can mention the velocities. From VASP
manual 'If no initial velocities are provided, the file [POSCAR] may
end here. For molecular dynamics the velocities are initialised
randomly according to a Maxwell-Boltzmann distribution at the initial
temperature TEBEG (see section 6.28).'

Good night!!

Chaitanya.