The ABINIT Users Mailing List ( CLOSED )

[verstraete@pcpm.ucl.ac.be]

Hello,

your acell looks very large, and the increase +0.5 bohr too. If you start 
too far from equilibrium (we didn't get forces or energies in the log...) 
it is very hard to force convergence: is the experimental value in the 
middle of the range of acell you cover? Do you actually need all these 
values?

If this is correct, you probably have too many kpoints, but that can't 
hurt. You can always add the smearing (not too large, 0.01 Ha), and if the 
gap is small, it can improve convergence.

You can also use an improved (and more costly) preconditioning scheme with 
iprcel 45. Not guaranteed, but it often speeds up convergences. Are you 
converging at all? I also prefer iscf 3 instead of the default 5; 3 is 
more robust.

You should provide a condensed version of your output with energy 
convergence etc... You should also (always!) check the archive of the 
forum for mails regarding convergence, there have already been plenty of 
problems like yours, and some answers...

Matthieu

--
===================================================================
Matthieu Verstraete               mailto:verstraete@pcpm.ucl.ac.be 
PCPM, Boltzmann, pl. Croix du Sud, 1            tel: 010/ 47 33 59
B-1348 Louvain-la-Neuve  Belgium                fax: 010/ 47 34 52