[DFTB-Plus-User] Problem with MD calculations with Nose Hoover thermostat
Benjamin Hourahine
benjamin.hourahine at strath.ac.uk
Sun Aug 17 17:14:32 CEST 2014
Dear Susmita,
the problem comes, probably, from the thermostat coupling strength. If no units are specified then atomic units will be used. You should choose the frequency to be on the scale of the Raman frequency of Si, so around 600 cm-1.
The diagonalization failure happens for one of two reasons (usually), either a crazy structure (probably this case, due to the coupling strength) or the bug in the implementation of several implementations of the divide and conquer eigensolver present in several implementations of lapack.
There are two other things worth changing. As this is pure Si, SCC and d-orbitals are both not needed (p should be sufficient).
Regards
Ben
Dr. B. Hourahine, SUPA, Department of Physics,
University of Strathclyde, John Anderson Building,
107 Rottenrow, Glasgow G4 0NG, UK.
+44 141 548 2325, benjamin.hourahine at strath.ac.uk
2013/4 THE Awards Entrepreneurial University of the Year
2012/13 THE Awards UK University of the Year
The University of Strathclyde is a charitable body,
registered in Scotland, number SC015263
________________________________________
From: DFTB-Plus-User [dftb-plus-user-bounces at dftb-plus.info] on behalf of Susmita Basak [basak at fhi-berlin.mpg.de]
Sent: 15 August 2014 11:02
To: dftb-plus-user at dftb-plus.info
Subject: [DFTB-Plus-User] Problem with MD calculations with Nose Hoover thermostat
Dear DFTB+ community Members,
I am a new user of DFTB+. I have tried to do MD calculations for
Silicon using Nose Hoover chain thermostat. I am getting the following
error-- 'Failure in diagonalisation routine zhegvd, non-positive
definite overlap! Minor 1 responsible.' This error was mentioned
couple of years ago in the mailing list but the solution to the problem
is not available in the forum. I am copying the input files and md.out
below. I have used the latest Slater-Koster files for Silicon which
include d orbitals. Can you please help me with this problem ?
With best regards,
Susmita
Fritz Haber Institute of the Max Planck Society
=========================
##dftb_in.hsd
Geometry = GenFormat {
<<< "input_geometry.gen"
}
Driver = VelocityVerlet{
MovedAtoms = 1:-1
Steps = 100
TimeStep [Femtosecond] = 1 # 1 femtosecond = 1e-15 second
OutputPrefix= "geo_end"
MDRestartFrequency= 1
Thermostat = NoseHoover{
Temperature [Kelvin] = 300.0
CouplingStrength= 500
}
ConvergentForcesOnly = Yes
}
Hamiltonian = DFTB {
SCC = Yes
SCCTolerance = 1e-5
SlaterKosterFiles = Type2FileNames {
Prefix =
"/mnt/lxfs1/home/basak/dftb/dftb+_1.2.2_autotest/slako/matsci-0-3/"
Separator = "-"
Suffix = ".skf"
}
MaxAngularMomentum {
Si = "d"
}
Charge = 0.0
SpinPolarisation = {}
Filling = Fermi {
Temperature [Kelvin] = 300.0
}
KPointsAndWeights = SupercellFolding {
8 0 0
0 8 0
0 0 8
0.5 0.5 0.5
}
}
Options = {}
ParserOptions {
ParserVersion = 4
}
============================================
## Input Geometry
2 S #number of atoms, geometry type supercell in Cartesian
co-ordinates
Si
# Index Type Coordinates
1 1 0.00 0.00 0.000
2 1 1.3565168814 1.3565168814 1.3565168814
0.000000 0.000000 0.000000 # coordinate origin in Å
# Lattice vectors
0 2.71303376291 2.71303376291
2.71303376291 0 2.71303376291
2.71303376291 2.71303376291 0
====================================================
## md.out
MD step: 0
Nose-Hoover chain variables
x:
0.1000000000E+01 0.1000000000E+01 0.1000000000E+01
v:
0.0000000000E+00 0.0000000000E+00 0.0000000000E+00
g:
0.0000000000E+00 0.0000000000E+00 0.0000000000E+00
Pressure: -0.117899E-02 au -0.346870E+11
Pa
Potential Energy: -2.7420859013 H -74.6160
eV
MD Kinetic Energy: 0.0014250669 H 0.0388
eV
Total MD Energy: -2.7406608344 H -74.5772
eV
MD Temperature: 0.0009500446 au 300.0000 K
MD step: 1
Nose-Hoover chain variables
x:
NaN NaN NaN
v:
NaN NaN NaN
g:
NaN NaN NaN
Pressure: NaN au NaN
Pa
Potential Energy: -2.7420803400 H -74.6158
eV
MD Kinetic Energy: NaN H NaN
eV
Total MD Energy: NaN H NaN
eV
MD Temperature: NaN au NaN K
MD step: 2
Nose-Hoover chain variables
x:
NaN NaN NaN
v:
NaN NaN NaN
g:
NaN NaN NaN
_______________________________________________
DFTB-Plus-User mailing list
DFTB-Plus-User at dftb-plus.info
http://www.dftb-plus.info/mailman/listinfo/dftb-plus-user
More information about the DFTB-Plus-User
mailing list