The ABINIT Users Mailing List ( CLOSED )

[verstraete@pcpm.ucl.ac.be]

Hello Thierry,

Some hints (no guarantees):
try "iprcel 45" i.e. preconditioning of the unit cell using the dielectric 
matrix.

Otherwise use more vacuum: the difference for an extra Angstroem can be
enormous. Slab calculations can often be painful, but I've always managed
to make them work.

Normally you don't need a smearing if everything stays semiconducting, but
it doesn't hurt unless you go too high in temperature (0.01 Ha is usually
the maximum I use).

Another thing: make sure you have enough bands. Put these in by hand, and 
add about 10 bands beyond your valence ones. The effects of 
degeneracies around the Fermi level throw off the calculation.

Using enough k-points is also important: for very small numbers of 
k-points (1 or 2) I have once or twice had problems (ie. not frequently)


Matthieu




On Thu, 17 Oct 2002, Thierry Deutsch wrote:

> Hi,
> 
>       I would like to calculate the Si surface with ABINIT using geometry 
> optimization. The goal is not to have an accurate result but to test for a 
> new development.
> 
> I have tried to tune some parameters  tsmear (0, 1000K and 5000K). but 
> ABINIT 
> does not converge in the SCF cycle.
> 
> diemac is 12 maybe an higher value is better.
> 
> Maybe someone has already calculated the Si(100) with ABINIT ?
> 
> Thierry Deutsch

-- 
===================================================================
Matthieu Verstraete               mailto:verstraete@pcpm.ucl.ac.be 
PCPM, Boltzmann, pl. Croix du Sud, 1            tel: 010/ 47 86 81
B-1348 Louvain-la-Neuve  Belgium                fax: 010/ 47 34 52