ELFCAR ERROR

[sungeng]

I am trying calculate the electron localization function of simple water molecular.
The result is strange, the shape of final electron density cloud does not agree with
the position of the water molecular.

Acturally , this question had been asked previously by others , But still no answers ?

http://cms.mpi.univie.ac.at/vasp-forum/ ... ELF#p13408

I am hopping any admin could notice this problem, and give any suggestions.

Thank you !

[admin]

This is not observed in our test calculations.
Please display POSCAR and INCAR.

[sungeng]

This is not observed in our test calculations.
Please display POSCAR and INCAR.

Sorry for so late.

This is my INCAR
SYSTEM = Structure_Optimization

ISTART = 0 # job : 0-new 1-cont 2-samecut
ISPIN = 1 # IMPORTANT

ENCUT = 400 # IMPORTANT
# ENAUG = 650
PREC = Normal # IMPORTANT
NELM = 100 # give the maxmium number of electronic SC steps
NELMIN = 4 # gives the minmium number of electronic SC steps
EDIFF = 1E-4 # IMPORTANT specifies the global breakdown conditions for the electronic SC-loop
EDIFFG = -0.05 # IMPORTANT


NSW = 0 # IMPORTANT sets the maximum number of ionic steps
IBRION = -1 # determines how the ions are updated and moved
ISIF = 0 # determines what to be relaxed
# ADDGRID = .TRUE



ISMEAR = 0 # IMPORTANT how the partial occupancies fnk are set for each orbital
SIGMA = 0.005 # IMPORTANT electron temperature with of smearing


ALGO = Fast # convenient option to specify the electronic minimisation algorithm
LREAL = .FALSE.
GGA = PE

LWAVE = .FAlSE
LCHARG = .FALSE

LELF = .TRUE.


And this is my POSCAR for a simple water

O H
1.00000000000000
10.0000000000000000 0.0000000000000000 0.0000000000000000
0.0000000000000006 10.0000000000000000 0.0000000000000000
0.0000000000000006 0.0000000000000006 10.0000000000000000
O H
1 2
Selective dynamics
Direct
0.2236001803083601 0.1369449502137714 0.5873257195763708 T T T
0.2418170022345913 0.2241321460549819 0.6264542334024383 T T T
0.1918768174570487 0.1572789037312509 0.4976680470211881 T T T

0.00000000E+00 0.00000000E+00 0.00000000E+00
0.00000000E+00 0.00000000E+00 0.00000000E+00
0.00000000E+00 0.00000000E+00 0.00000000E+00

ELFCAR is visulized by VESTA, and the position of water molecular and the electron density is not even at the same position in the cell.

Thank You !

~

[admin]

This is really a bug. When not setting NCORE a default value NCORE=1 is taken
and produces corrupted ELFCAR (CHG and CHGCAR are correct).
Setting NCORE to some reasonable value (e.g. NCORE = 4) gives you
the correct ELFCAR.

[gniding]

Dear VASP community

I am having a similar problem with other molecules (including ammonia). To Use NPAR = 4 doesn't solve the problem.
I am using vasp.5.4.1, I wonder if the problem has already been solved in a more recent version

Thanks.

[henrique_miranda]

The suggestion above is to use NCORE=4 and not NPAR=4.
The meanings are different:
wiki/wiki/index.php/NPAR

Do you still have an issue when using NCORE=4?

Kind regards,
Henrique Miranda

[gniding]

Yes, NCAR=4 or NPAR=4 with 16 core, I have the same problem

[henrique_miranda]

You are supposed to use

Code: Select all

NCORE=4

and not

Code: Select all

NCAR=4

[gniding]

Sorry, I miss spelled my last post, but my INCAR is correct. Using NCORE = 4 do not solve the bug in my case

Here there is my INCAR

----

SYSTEM = amonia ELF
Starting parameters for this run:
NWRITE = 2 write-flag
ISTART = 0 job : 0-new, 1-Cont.Diff.Gset, 2-Cont.Same.Gset
ICHARG = 2 charge: 0-wave, 1-file, 2-atom, >10-const


Electronic Relaxation:
PREC = Normal
ENCUT = 400


EDIFF = 1E-5 energy stopping-criterion for ELM
LREAL = Auto real-space projection (.FALSE., .TRUE., On, Auto)
IALGO = 38 algorithm (8=CG for small, 48=RMM for big systems)
NSIM = 1 NSIM bands are optimized simultaneously (IALGO=48 only)

LCHARG = .TRUE.
LAECHG = .TRUE.
ADDGRID = .TRUE.
LELF = .TRUE.
LVHAR = .FALSE.
LWAVE = .FALSE.

Ionic Relaxation:
NSW = 1000 max number of geometry steps
IBRION = 2 ionic relax: 0-MD, 1-quasi-Newton, 2-CG, 3-Damped MD
EDIFFG = -0.01 force (eV/A) stopping-criterion for geometry steps
ISIF = 2 (1:force=y stress=trace only ions=y shape=n volume=n)
IVDW = 11

DOS related values:
ISMEAR = 0 (-1-Fermi, 1-Methfessel/Paxton)
SIGMA = 0.05 broadening in eV

Parallelization and R-Wigs flags:
NCORE = 4

----

Here there is my POSCAR
block_1
1.00000000000000
30.0000000000000000 0.0000000000000000 0.0000000000000000
0.0000000000000000 30.0000000000000000 0.0000000000000000
0.0000000000000000 0.0000000000000000 30.0000000000000000
N H
1 3
Direct
0.5000003248253690 0.4981245955136073 0.4968313048129741
0.5000032302059226 0.5284436731196527 0.4812566945464334
0.5280979297971611 0.4817167496568520 0.4859554348343810
0.4718985151715472 0.4817149817098972 0.4859565658062142

0.00000000E+00 0.00000000E+00 0.00000000E+00
0.00000000E+00 0.00000000E+00 0.00000000E+00
0.00000000E+00 0.00000000E+00 0.00000000E+00
0.00000000E+00 0.00000000E+00 0.00000000E+00

---

Thanks

[henrique_miranda]

Sorry for the delay in answering this post.

1. Could you please post a figure with the ELFCAR you obtain?
In the issue reported previously in this post the densities are not localized on the positons of the ions, is this the case for you as well?

2. Have you tried to run the calculation in serial as a test?
Do the results differ from when NCORE=4 and you run with mpirun -np 4