[CPMD-list] Is the CONVERGENCE GEOMETRY enough for geometry optimization ?

[Axel Kohlmeyer]

On Sat, 21 Apr 2007, liu liuxiandong wrote:

LX> Dear CPMDers,
LX> 
LX> Since the CONVERGENCE GEOMETRY criteria checks the G
LX> then, is the CONVERGENCE GEOMETRY criteria enough for geometry optimization
LX> ?  I saw that the convergence tolerances include Energy tolerance and
LX> Displacement besides the Max. force tolerance in other code such as CASTEP.

if you want a special convergence behavior, 
you can always program it yourself (after all
that is what open source software is all about).

the different geometry (and wavefunction optimizers)
have different convergence criteria and you'll also
have to factor in the use of adaptive convergence.

please note, that CPMD is primarily used as a molecular
dynamics code (hence the MD in CPMD), so that people usually
put more focus on implementing that. 

the same goes for any details on units (internally CPMD 
uses hartree atomic units) which you have to extract
from the sources, if the information is not available
in the documentation or in the mailing list archive.

axel.

LX> 
LX> Thanks.
LX> 
LX> 
LX> BEST,
LX> 
LX> XIANDONG LIU
LX> 

-- 
=======================================================================
Axel Kohlmeyer   akohlmey at cmm.chem.upenn.edu   http://www.cmm.upenn.edu
   Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.