Number of NEB images
Posted: Sun Sep 08, 2013 8:37 am
Hi, I'm a new user of VASP and NEB.
Because I'm not good at building code and it is just for my trial NEB calculation, I'm using the basic NEB implemented in VASP5.2.11 without building additional VTST codes.
Starting my trial calculation with 3 images, I've found that the old NEB implemented in VASP needs 2^n images (http://cms.mpi.univie.ac.at/vasp/guide/node140.html, written in 1996, "Mind, that IMAGES can only take values 2^n.") while the newer site does not mention that kind of restriction with the number of images (http://cms.mpi.univie.ac.at/vasp/vasp/E ... ethod.html, written in 2008).
So I wonder the latter site is just missing the warning message or it is truly unnecessary to use "2^n" images in our current version of VASP.
Some said that 3~5 images worked fine (viewtopic.php?f=2&t=603) but I'm not sure that the NEB methods in original VASP and VTST code work with the same algorithm.
Always thanks for your explanation.
Because I'm not good at building code and it is just for my trial NEB calculation, I'm using the basic NEB implemented in VASP5.2.11 without building additional VTST codes.
Starting my trial calculation with 3 images, I've found that the old NEB implemented in VASP needs 2^n images (http://cms.mpi.univie.ac.at/vasp/guide/node140.html, written in 1996, "Mind, that IMAGES can only take values 2^n.") while the newer site does not mention that kind of restriction with the number of images (http://cms.mpi.univie.ac.at/vasp/vasp/E ... ethod.html, written in 2008).
So I wonder the latter site is just missing the warning message or it is truly unnecessary to use "2^n" images in our current version of VASP.
Some said that 3~5 images worked fine (viewtopic.php?f=2&t=603) but I'm not sure that the NEB methods in original VASP and VTST code work with the same algorithm.
Always thanks for your explanation.