Page 1 of 1

Neb calculation fails

Posted: Thu Jun 15, 2006 8:05 am
by Warda
hello all,
Itried to do neb calculation as suggested to me in the forum but Ihave got this message error :
running on 16 nodes
each image running on 2 nodes
distr: one band on 1 nodes, 2 groups
vasp.4.6.3 06Feb03 complex
01/POSCAR found : 2 types and 33 ions
LDA part: xc-table for Ceperly-Alder, standard interpolation
No initial positions read in

The POSCAR located in the 01 subdirectory is as :


fcc:
3.52
2.0000000000000000 0.0000000000000000 0.0000000000000000
0.0000000000000000 2.0000000000000000 0.0000000000000000
0.0000000000000000 0.0000000000000000 2.0000000000000000
32 1

Direct
0.0000000000000000 0.0000000000000000 0.0000000000000000
0.0000000000000000 0.5000000000000000 0.0000000000000000
0.5000000000000000 0.0000000000000000 0.0000000000000000
0.2500000000000000 0.7500000000000000 0.0000000000000000
0.2500000000000000 0.2500000000000000 0.0000000000000000
0.5000000000000000 0.5000000000000000 0.0000000000000000
0.7500000000000000 0.2500000000000000 0.0000000000000000
0.7500000000000000 0.7500000000000000 0.0000000000000000
0.0000000000000000 0.0000000000000000 0.5000000000000000
0.0000000000000000 0.5000000000000000 0.5000000000000000
0.2500000000000000 0.2500000000000000 0.5000000000000000
0.2500000000000000 0.7500000000000000 0.5000000000000000
0.5000000000000000 0.0000000000000000 0.5000000000000000
0.5000000000000000 0.5000000000000000 0.5000000000000000
0.7500000000000000 0.2500000000000000 0.5000000000000000
0.7500000000000000 0.7500000000000000 0.5000000000000000
0.2500000000000000 0.0000000000000000 0.2500000000000000
0.0000000000000000 0.2500000000000000 0.2500000000000000
0.0000000000000000 0.7500000000000000 0.2500000000000000
0.2500000000000000 0.5000000000000000 0.2500000000000000
0.5000000000000000 0.2500000000000000 0.2500000000000000
0.5000000000000000 0.7500000000000000 0.2500000000000000
0.7500000000000000 0.5000000000000000 0.2500000000000000
0.7500000000000000 0.0000000000000000 0.2500000000000000
0.2500000000000000 0.0000000000000000 0.7500000000000000
0.0000000000000000 0.2500000000000000 0.7500000000000000
0.0000000000000000 0.7500000000000000 0.7500000000000000
0.2500000000000000 0.5000000000000000 0.7500000000000000
0.5000000000000000 0.2500000000000000 0.7500000000000000
0.5000000000000000 0.7500000000000000 0.7500000000000000
0.7500000000000000 0.5000000000000000 0.7500000000000000
0.7500000000000000 0.0000000000000000 0.7500000000000000
0.2361111111111140 0.2361111111111140 0.2361111111111140

and the INCAR file is as :

ISTART = 0 ; ICHARG=2
ISMEAR = 1 ; SIGMA = 0.2
LORBIT=11
LREAL=.TRUE.
PREC = High
IALGO = 48
ENCUT = 400
ISPIN=2
SPRING=-5
IMAGES=8
IBRION=1

Can any one tell me what the problem?

Thanks for your suggestion.

Posted: Thu Jun 15, 2006 5:11 pm
by dwyuan
Did you edit your poscar file on windows machine?
If you did so, please check your poscar file format.

Posted: Thu Jun 15, 2006 10:41 pm
by andri
I don't really see any problem with the POSCAR file but we recommend setting the element symbols in the top line for our scripts to work properly, like:

Ru N H
10.7876184200000
1.0000000000000000 0.0000000000000000 0.0000000000000000
0.0000000000000000 0.8660436140083640 0.0000000000000000
0.0000000000000000 0.0000000000000000 1.6685795973862401
64 1 2
Selective dynamics
Direct
0.0625047100503906 0.2155906743045861 0.1823673465958322 F F F
0.3125045347787747 0.2155910720305769 0.1823674902077457 F F F
0.5625038698466653 0.2155909667141103 0.1823676050354095 F F F
0.8125040829204693 0.2155906614048746 0.1823676717670395 F F F
0.0625047007585451 0.7155908312694663 0.1823674341060055 F F F
.
.
.

As dwyuan points out then files that are altered on a window$ machine can have trailing characters at the end of each line when ported to a UNIX/Linux system, e.g. the EOL or NUL character. These need to be stripped off. In a tcsh shell, this is how I strip the NUL character (000 in Oct) from files (ascii.tcsh):

#!/bin/tcsh

foreach a ($argv)
echo $a
tr -d "\000" < $a > ascii_tmp ; mv -f ascii_tmp $a
end

then I run ascii.tcsh <input_files> .

Also, if you are using our implementation of NEB then the initial and final configurations should be in directories 00 and 09 (assuming 8 images as you seem to be using), respectively. If you are using the VASP implementation of the elastic band then the question is more appropriately directed to the offical VASP community forum,
http://cms.mpi.univie.ac.at/vasp-forum/news.php