The ABINIT Users Mailing List ( CLOSED )

[Gilles Zerah <gilles.zerah@cea.fr>]

Hello,

The only thing I know of, is the "parareal" coupling scheme.
The idea is to predict the trajectory by a classical potential and
correct it by ab initio -in parallel-
It is still experimental, and the implementation is not complete at this
time.
A new version is in test.
The main idea of thie new version is to use Thomas-Fermi instead of
classical potential as
predictor. This is less accurate, but more flexible:
you do not need to provide classical potential parameters for each
material.
To see what it is, look at the keyword "parareel" in Abinit.

Sincerly,

Gilles Zerah

> I received a question about the possibility
> of a "mixed ab-initio and empirical molecular dynamics" with abinit.
> Does anyone know if it is possible to do with abinit something
> similar, or if it is possible to interface abinit with a classical
> molecular dynamic
> code, in such a way to treat only some degrees of freedom from an ab
> initio
> point of view, and the rest classically?
> thank you
> Valerio Olevano