The ABINIT Users Mailing List ( CLOSED )

[Xavier Gonze <gonze@pcpm.ucl.ac.be>]

e120231@metu.edu.tr wrote:
> Hello,
>
> I am a new user of abinit.While I was working on bcc iron, I experience some 
> difficulties. The
> results did not converge to an acceptable value ( on the basis of acell).
>
> My input file is below
>
>
> getwfk -1 nsppol 2 nband 6 nline 6
>
> occopt 4 tsmear  0.04
>
> kptopt 1 ngkpt  14 14 14
>
> acell 3*6.0 rprim   -0.5  0.5  0.5 0.5  -0.5  0.5 0.5  0.5  -0.5
>
> optcell 2 ionmov 3 dilatmx 1.5 ecutsm 0.5
>
> ntypat 1 znucl 26
>
> natom 1 typat 1 xred 0.0  0.0  0.0
>
> ixc 1 iscf 3
>
> ecut 20.0 ntime  40 nstep 40 toldfe 1.0d-6
>
> I have tried to change "iscf" (3 & 5) "nband" (4 & 6) "ionmov" (2 & 3) but no 
> improvement was seen.
>  What am I missing ? I will be very glad if anyone informs me. Thank you.

You might try to decrease also tsmear : for d-electron metals, the adequate
smearing is usually much smaller than for sp-metals like Al . you might have
to increase further ngkpt for small values of tsmear. Especially,
monitor the appearance and convergence of a spin-polarized charge (look 
inside the output
file), at fixed unit cell size (e.g. experimental), before trying to converge 
acell.

Xavier