[DFTB-Plus-User] Bug in modes or hsdparser.F90 when calculaiting vibrations of larger systems (i.e. 111 atoms).

Igor Popov rrobinovic at yahoo.com
Tue Nov 14 19:16:50 CET 2017


Hi,
I am trying to calculate vibrational modes of a graphene nanoribbonconsisting of 111 atoms. A DFTB+ run successfully generated a largehessian.out file. The dftb+ input is:
------------------------------- start of dftb_in.hsd -----------------------------------
Geometry = GenFormat {
<<< 'opt.gen'
}

Driver =SecondDerivatives {
  Atoms = 1:-1
  Delta = 1e-4
}

#Driver = ConjugateGradient {
#  MovedAtoms = 1:-1
#  MaxForceComponent = 0.00076  # 0.00076 H/Bohr = 0.04 eV/A which is used in Siesta by default
#  MaxSteps = 200
#  OutputPrefix = "opt"
#  LatticeOpt = No
#  #Isotropic = Yes  # dont change angle and relative lengths between lat. vecs.
#  AppendGeometries = Yes   # append geoms in opt
#  ConvergentForcesOnly = Yes
#  Constraints = {}
#}

Hamiltonian = DFTB {
  SCC = Yes
  SCCTolerance = 1e-7
  MaxAngularMomentum = {
    H = "s"
    C = "p"
    O = "p"
  }

  Filling = Fermi {
    Temperature [Kelvin] = 100.0
  }  PolynomialRepulsive = {
     H-H = No
     H-C = No
     H-O = No
     C-H = No
     C-C = No
     C-O = No
     O-H = No
     O-C = No
     O-O = No
  }

  SlaterKosterFiles = {
     H-H = "/home/popsi/slako/pbc-0-2/H-H.skf"
     H-C = "/home/popsi/slako/pbc-0-2/H-C.skf"
     H-O = "/home/popsi/slako/pbc-0-2/H-O.skf"
     C-H = "/home/popsi/slako/pbc-0-2/C-H.skf"
     C-C = "/home/popsi/slako/pbc-0-2/C-C.skf"
     C-O = "/home/popsi/slako/pbc-0-2/C-O.skf"
     O-H = "/home/popsi/slako/pbc-0-2/O-H.skf"
     O-C = "/home/popsi/slako/pbc-0-2/O-C.skf"
     O-O = "/home/popsi/slako/pbc-0-2/O-O.skf"
  }

  KpointsAndWeights = SupercellFolding{
    1  0  0
    0  1  0
    0  0  1
    0.0 0.0 0.0
  }
} #end of Hamiltonian section
Options = {
  WriteAutotestTag = No
  WriteResultsTag = Yes
}

ParserOptions = {
  ParserVersion = 4
}

------------------------------- the end of dftb_in.hsd -----------------------------------

However modes calculations report the ERROR:
At line 836 of file /home/aradi/sync/laptop/dftb_devel/dftbplus/prog/dftb+/lib_io/hsdparser.F90 (unit = 24, file = 'modes_pin.hsd')
Fortran runtime error: End of record
Input file for modes:------------------------------- start of modes_in.hsd -----------------------------------Geometry = GenFormat {
<<< 'opt.gen'
}

SlaterKosterFiles = {
  H-H = "/home/popsi/slako/pbc-0-2/H-H.skf"
#  H-C = "/home/popsi/slako/pbc-0-2/H-C.skf"   modes program uses only monoatomic slako
#  H-O = "/home/popsi/slako/pbc-0-2/H-O.skf"
#  C-H = "/home/popsi/slako/pbc-0-2/C-H.skf"
  C-C = "/home/popsi/slako/pbc-0-2/C-C.skf"
#  C-O = "/home/popsi/slako/pbc-0-2/C-O.skf"
#  O-H = "/home/popsi/slako/pbc-0-2/O-H.skf"
#  O-C = "/home/popsi/slako/pbc-0-2/O-C.skf"
  O-O = "/home/popsi/slako/pbc-0-2/O-O.skf"
}

Hessian = {
  <<< 'hessian.out'
}
DisplayModes = {
  Animate = Yes
}------------------------------- the end of modes_in.hsd -----------------------------------
I tested the modes program with a simple O2 molecule and it works flawlessly.
Any idea what could be problem? 

Igor
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.zfn.uni-bremen.de/pipermail/dftb-plus-user/attachments/20171114/35e8fb8e/attachment.html>


More information about the DFTB-Plus-User mailing list