[CPMD-list] =?GB2312?B?SG93IHRvIHBpY2sgYSBwcm9wZXIgQ09OVkVSR0VOQ0UgT1JCSVRBTCCjvw==?=

[liu liuxiandong]

Dear CPMDers,

When I run a OPTIMIZE GEOMETRY, I find that the CONVERGENCE GEOMETRY has
been reached and the programme terminates normally before the ORBITAL
converges to 1.0D-6.
My system is a solid-water interface system containing 76 atoms in total.
Is the convergence of orbital 5D-6 is "tight" enough for such systems ?
And how to pick this tolerance properly ?

Thanks for your suggestions.

Xiandong


My input is as follows,


 RESTART WAVEFUNCTIONS COORDINATES HESSIAN
  OPTIMIZE GEOMETRY XYZ
  HESSIAN UNIT

  CONVERGENCE ADAPT
  0.02
  CONVERGENCE INITIAL
  5.0D-6
  CONVERGENCE ORBITALS
  1.0D-6
  CONVERGENCE GEOMETRY
  2.0D-4

    LBFGS NREM
      20


&END

&DFT
  FUNCTIONAL BLYP

&END

&SYSTEM

  SYMMETRY
   14
  ANGSTROM
  CELL ABSOLUTE DEGREE
   21.7780 10.3365 11.5168 90.0000 90.0000 90.0000
  CUTOFF
   100.0
&END

&ATOMS
*O_MT_BLYP.psp KLEINMAN-BYLANDER
 LMAX=P
 38

*H_MT_BLYP.psp KLEINMAN-BYLANDER
 LMAX=S
 20

*Si_MT_BLYP.psp KLEINMAN-BYLANDER
 LMAX=D
 8

*Al_MT_GIA_BLYP.psp KLEINMAN-BYLANDER
 LMAX=D LOC=0
 8

&END
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://cpmd.org/pipermail/cpmd-list/attachments/20070513/c1076be7/attachment.html