The ABINIT Users Mailing List ( CLOSED )

["Anglade Pierre-Matthieu" <anglade@gmail.com>]

PS: I forgot the following:
By default Abinit is preconditioning the SCF cycle for an homogeneous
perfect metal. Since this is not the case in your computation, you may
prefer either to remove it preconditioning, diemac=1, or to use one
that is well suited for your case, iprcel 49.

Moreover, you didn't set "ntime" so Abinit will perform exactly zero
relaxation steps which is not enough right ? So set ntime to 10 or 20
or so...



On 8/18/07, Anglade Pierre-Matthieu <anglade@gmail.com> wrote:
> Hi,
>
> I think (but I may mistake) that the occupancy you want for dataset 3 is
> 22200
> instead of
> 22110
>
> Also I'm a bit surprised at the low value for ecut you have and for
> the size of your cell which seems a little small.
>
> Regards,
>
> PMA
>
> On 8/18/07, farah marsusi <fmarsusi@yahoo.com> wrote:
> > Dear users:
> > In order to compute affinity of CH, I made an in put file for computing
> > total energy of neutral and negative charged CH and then subtract them :
> > E(CH)-E(CH)-  ,but the result is not correct( about 3ev) and differs from
> > experiment(1.1 ev) .Is there any one tell me what is my mistake?
> > Here is my in put file for CH -  and CH  after convergence test :
> >
> > ndtset  2
> > acell 10 10 10
> > ecut  12
> > ionmov  2
> > ntypat  2
> > natom  2
> > xangst
> >           0.000000    0.000000    0.158311
> >           0.000000    0.000000   -0.949865
> > ixc  11
> >
> > charge1   0           # neutral particle
> > nband1  5
> > occ1   2 2 1 0 0
> >
> > charge2    -1          # charged particle
> > nband2  5
> > occ2   2 2 1 1 0
> >
> > typat  1 2
> > znucl  6 1
> > nkpt  1
> > toldfe  5.0d-8
> >
> >
> >  ________________________________
> > Need a vacation? Get great deals to amazing places on Yahoo! Travel.
> >
> >
>
>
> --
> Pierre-Matthieu Anglade
>


-- 
Pierre-Matthieu Anglade