[CPMD-list] error optimising geomerty
Axel Kohlmeyer
axel.kohlmeyer at theochem.ruhr-uni-bochum.de
Thu Jul 29 16:44:16 CEST 2004
>>> "AA" == A Alaka <A.Alaka> writes:
AA> I am trying to optimise the geometry of a structure that only repeats
AA> in the Z-direction; its a semiconducting carbon nano-tube. I used the
AA> input below and the error I got is below that. This is a spin-polarised
AA> calculation. Is there anything I am doing wrong? Any help at all would
AA> be great. thanks in advance.
hi,
a few comments:
last time i checked, PCG MINIMIZE did not work with
k-points (only PCG). but both ODIIS and PCG with k-points
are a bit tricky (AFAIK, it is not exactly a frequently used
part of CPMD). running in parallel could make it even worse.
so you may want to try using a diagonalization scheme instead,
or try a gamma-point only calculation.
the ZFLEXIBLE CELL option is _very_ experimental
(as is pointed out in the manual!).
anyway, if you are using a variable cell method,
you need to make sure, that your stress tensor is
converged with respect to the plane wave energy cutoff
(i.e. you usually have to use a very large cutoff).
for your system you may probably be better off by using
fractional coordinates and doing a series of calculations
where you only change the length of the z-axis of the supercell.
if you use CONVERGENCE ADAPT, you should also set
CONVERGENCE INITIAL to the convergence you want to
start with. otherwise the first step will be converged
to the (full) default and the adaptive convergence will
be disabled in the first step.
do you have a specific reason to explicitely specify
the LDA correlation? usually, just using
FUNCTIONAL REVPBE (which uses Pade approximation instead)
should do the trick. some of the LDA functionals do not
seem to be fully ported to NEWCODE and/or LSD yet.
regards,
axel.
AA> &CPMD
AA> OPTIMIZE GEOMETRY XYZ SAMPLE
AA> 1
AA> STRUCTURE BONDS ANGLES DIHEDRALS
AA> RHOOUT
AA> ENERGYBANDS
AA> CONVERGENCE ORBITALS
AA> 1.D-5
AA> CONVERGENCE GEOMETRY
AA> 1.D-4
AA> CONVERGENCE ENERGY
AA> 0.005
AA> CONVERGENCE ADAPT
AA> 0.02
AA> SPLINE POINTS
AA> 3000
AA> PRINT FORCES
AA> 1
AA> MIRROR
AA> LSD
AA> LBFGS
AA> PCG MINIMIZE
AA> !fREE eNERGY fUNCTIONAL
AA> !eLECTRON tEMPERATURE
AA> !0
AA> !iNITIALIZE wAVEFUNCTION rANDOM
AA> !oRTHOGONALIZATION lOWDIN
AA> &END
AA> &PROP
AA> LOCALIZE
AA> CUBEFILE DENSITY
AA> PROJECT WAVEFUNCTION
AA> POPULATION ANALYSIS MULLIKEN
AA> CONDUCTIVITY
AA> &END
AA> &DFT
AA> NEWCODE
AA> LDA CORRELATION PZ
AA> GRADIENT CORRECTION REVPBEX REVPBEC
AA> &END
AA> &SYSTEM
AA> STATES
AA> 137
AA> ZFLEXIBLE CELL
AA> ANGSTROM
AA> CELL VECTORS
AA> 11.0 0.0 0.0 0.0 12.0 0.0 0.0 0.0 4.165069
AA> ENERGY PROFILE
AA> CUTOFF
AA> 40
AA> KPOINTS MONKHORSTPACK
AA> 4 4 4
AA> &END
AA> &ATOMS
AA> *C_revPBE_NLCC KLEINMAN-BYLANDER NLCC
AA> LMAX=P LOC=P
AA> 32
AA> 4.3000000000 7.4488870000 1.2860690000
AA> 4.3000000000 7.4488870000 4.1650690000
AA> 3.0949730000 7.2091920000 3.4210000000
AA> 3.0949730000 7.2091920000 2.0000000000
AA> 2.0734010000 6.5265990000 1.2860690000
AA> 2.0734010000 6.5265990000 4.1650690000
AA> 1.3908080000 5.5050270000 3.4210000000
AA> 1.3908080000 5.5050270000 2.0000000000
AA> 1.1511130000 4.3000000000 1.2860690000
AA> 1.1511130000 4.3000000000 4.1650690000
AA> 1.3908080000 3.0949730000 3.4210000000
AA> 1.3908080000 3.0949730000 2.0000000000
AA> 2.0734010000 2.0734010000 1.2860690000
AA> 2.0734010000 2.0734010000 4.1650690000
AA> 3.0949730000 1.3908080000 3.4210000000
AA> 3.0949730000 1.3908080000 2.0000000000
AA> 4.3000000000 1.1511130000 1.2860690000
AA> 4.3000000000 1.1511130000 4.1650690000
AA> 5.5050270000 1.3908080000 3.4210000000
AA> 5.5050270000 1.3908080000 2.0000000000
AA> 6.5265990000 2.0734010000 1.2860690000
AA> 6.5265990000 2.0734010000 4.1650690000
AA> 7.2091920000 3.0949730000 3.4210000000
AA> 7.2091920000 3.0949730000 2.0000000000
AA> 7.4488870000 4.3000000000 1.2860690000
AA> 7.4488870000 4.3000000000 4.1650690000
AA> 7.2091920000 5.5050270000 3.4210000000
AA> 7.2091920000 5.5050270000 2.0000000000
AA> 6.5265990000 6.5265990000 1.2860690000
AA> 6.5265990000 6.5265990000 4.1650690000
AA> 5.5050270000 7.2091920000 3.4210000000
AA> 5.5050270000 7.2091920000 2.0000000000
AA> &END
AA> LINE SEARCH : LAMBDA=.188 PREDICTED ENERGY = -131.1599112891
AA> 369 5.142E-04 4.343E-05 -131.159883 -6.100E-05 13.30
AA> LINE SEARCH : LAMBDA=.167E-01 PREDICTED ENERGY = -131.1598543150
AA> 370 1.592E-03 1.931E-04 -131.159911 -2.777E-05 12.30
AA> LINE SEARCH : LAMBDA=.547E-01 PREDICTED ENERGY = -114.3005782022
AA> 371 4.284E+00 7.265E-01 -131.159856 5.503E-05 22.79
AA> UNIV| THE LEADING MINOR OF ORDER 4 IS NOT POSITIVE DEFINITE,
AA> UNIV| AND THE FACTORIZATION COULD NOT BE COMPLETED.
AA> PROGRAM STOPS IN SUBROUTINE UINV| ILLEGAL RESULTS DPOTRF [PROC= 0]
AA> UNIV| THE LEADING MINOR OF ORDER 4 IS NOT POSITIVE DEFINITE,
AA> UNIV| AND THE FACTORIZATION COULD NOT BE COMPLETED.
AA> PROGRAM STOPS IN SUBROUTINE UINV| ILLEGAL RESULTS DPOTRF [PROC= 2]
AA> Fatal error, aborting.
AA> Fatal error, aborting.
AA> UNIV| THE LEADING MINOR OF ORDER 4 IS NOT POSITIVE DEFINITE,
AA> UNIV| AND THE FACTORIZATION COULD NOT BE COMPLETED.
AA> PROGRAM STOPS IN SUBROUTINE UINV| ILLEGAL RESULTS DPOTRF [PROC= 1]
AA> Fatal error, aborting.
AA> UNIV| THE LEADING MINOR OF ORDER 4 IS NOT POSITIVE DEFINITE,
AA> UNIV| AND THE FACTORIZATION COULD NOT BE COMPLETED.
AA> PROGRAM STOPS IN SUBROUTINE UINV| ILLEGAL RESULTS DPOTRF [PROC= 3]
AA> Fatal error, aborting.
AA> Job sge.9239 on fimbrata: received signal ABRT.
AA> _______________________________________________
AA> CPMD-list mailing list
AA> CPMD-list at cpmd.org
AA> http://cpmd.org/mailman/listinfo/cpmd-list
--
=======================================================================
Axel Kohlmeyer e-mail: axel.kohlmeyer at theochem.ruhr-uni-bochum.de
Lehrstuhl fuer Theoretische Chemie Phone: ++49 (0)234/32-26673
Ruhr-Universitaet Bochum - NC 03/53 Fax: ++49 (0)234/32-14045
D-44780 Bochum http://www.theochem.ruhr-uni-bochum.de/~axel.kohlmeyer/
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.
More information about the CPMD-list
mailing list