symmetry and geometry optimization

[panda]

I am trying to perform a geometry optimization and upon each ionic step (each NSW) the symmetry drifts further and further from what it should be. The initial coordinates have been verified from the crystallographic information file but I do not understand why the symmetry is not being maintained and gets sequentially further and further away from the starting structure with each ionic relaxation. I am using auto k-points and my INCAR file looks like:

ISMEAR = -5 # method for calculating density of states -5 (tetrahedron) 0 (Gaussian) 1 Methfessel-Paxton
RWIGS= 1.50 0.8 1.0
EMIN=-15
EMAX= 10
NEDOS=2501

Start Parameter for This Run:
NWRITE = 2

Electronic Relaxation 1:
PREC = Normal

Ionic Relaxation:
NSW = 100
NBLOCK = 1
KBLOCK = 1
IBRION = 2
POTIM = 0.5
ISIF = 3

DOS Related Values:
SIGMA = 0.2
LELF = T
LWAVE = .FALSE.
LCHARG = .TRUE.


At the end of the run the symmetry is completely lost. Any help would be greatly appreciated!

Thanks,

Amanda

[crajesh]

Hi Use ISYM=2 tag

[panda]

I was using paw potentials so ISYM = 2 was set to default. The problem I was having involved "guessing" the space group and running the calculation for polymer/molecular type calculations. I obtained a table of bond distances, so I basically drew the structures out by hand using the bond distance table and then got the space group right. So the fault of the user, not the software