The ABINIT Users Mailing List ( CLOSED )

[verstraete@pcpm.ucl.ac.be]

Strange: optcell 1 should have found the lowest energy, and I doubt there 
is a local minimum. Is there a reason to use chkprim 0? It eliminates some 
uses of symmetry, and could be messing up your angles or atomic positions 
during the relaxation (checked this?), although the acell will remain 
identical. Send your full output for the relaxation (or at least excerpts 
+ header and footer).

Further, did you set dilatmx and ecutsm in the same way in all cases (ie 
relaxation and static calculations)? These change the PW basis and can 
influence your total energy if you're not very tightly converged... 1.4 is 
much too large, for instance, and makes your effective pw cutoff huge, but 
distorted by the kinetic energy smoothing function.

That the result could be different for Murnaghan vs optcell 1 is 
imaginable, but that the lowest energy structure was not found is 
bizarre.

Matthieu

--
===================================================================
Matthieu Verstraete               mailto:verstraete@pcpm.ucl.ac.be 
PCPM, Boltzmann, pl. Croix du Sud, 1        tel: +32 10 47 33 59
B-1348 Louvain-la-Neuve  Belgium            fax: +32 10 47 34 52