Are OT vs. non-OT simulations comparable

Joseph Han jhh... at gmail.com
Fri Dec 23 00:10:09 CET 2011


I'm a new user to cp2k, and was curious about the OT versus non-OT
simulation results.  Specifically, assuming that the number and types
of atoms in two simulations are the same, are the energies directly
comparable in an OT versus non-OT simulation.

More specifically, suppose that I am looking at different crystal
structures of the same stoichiometry.  In one cell configuration, I
need to use

      &OT ON
        MINIMIZER CG
        LINESEARCH 3PNT
        PRECONDITIONER FULL_ALL
      &END OT

to achieve convergence.  For the other configuration, I used

      ADDED_MOS 30
      &SMEAR  ON
         METHOD FERMI_DIRAC
         ELECTRONIC_TEMPERATURE [K] 600
      &END SMEAR
      &DIAGONALIZATION
         ALGORITHM STANDARD
      &END DIAGONALIZATION
      &MIXING
         METHOD BROYDEN_MIXING
         ALPHA   0.1
         BETA    1.5
         NBROYDEN  8
      &END

Can the total energies be compared when using different SCF
convergence techniques?  Does changing ADDED_MOS affect the ability to
compare results (i.e. does it represent a change in basis set)?

Thanks

Joseph



More information about the CP2K-user mailing list