The ABINIT Users Mailing List ( CLOSED )

[gonze@pcpm.ucl.ac.be]

Quoting Steven Homolya <steven.homolya@spme.monash.edu.au>:

> Rebecca Janisch wrote:
> > Hi,
> > 
> > I'm not sure whether I remember the original question right (and I'm
> too lazy to
> > check the archive, I apologize for that), but, how about using the
> initial
> > charge density that ABINIT generates in the first scf loop as a
> reference?
> > 
> 
> That's more like the sort of thing that I was after.
> 
> Now two questions remain:
> 
> 1) Does abinit initialise the calculation using a superposition of
> atomic
> densities?
> 
> 2) If yes, should I use the density from before or after the first scf
> cycle?
> (Is it possible to write density before?)

Dear Steven, Rebecca,

Please, do NOT use that charge density to do anything else than starting
a calculation : it is only a gaussian of adjustable length !
See the routine initro.F90 ...

I think that at present, for bulk materials with a small primitive cell,
there is no other possibility than (1) to choose a suitable supercell, 
(2) to use cut3d to repeat periodically the bulk density in this supercell,
or to compute directly the density of the bulk in this supercell 
and then, (3) to subtract the
atomic densities computed separately atom by atom, each placed within this
supercell...
Additional coding is needed for more direct techniques.

Xavier

-------------------------------------------------
This mail sent through IMP: atlas.pcpm.ucl.ac.be