vef.pl and nebresults.pl provide different forces for the same run - expert advice needed!

[ravindra]

Hi folks,

I am in need of some expert help. This question may be trivial to some of you but your input will be quite helpful for my manuscript.
NEB calculation has converged but if I go into individual image directory and run vef.pl or vef.py, I get a certain Force vs Ionic steps graph....

BUT, if I run nebresults.pl script in the parent directory, the forces are different and much smaller.
I don't know which graph to trust. I have attached both graphs here and your advice would be very much appreciated.

[graeme]

If you attach a tar.gz file of the simulation I can help with this question. Most likely there is a simple explanation - the NEB forces are composed of only the real forces perpendicular to the path plus a spring force to keep the images equally spaced. For convergence, you should look at the NEB forces.

[ravindra]

Many thanks for your prompt response!
I have attached the entire simulation with OUTCARs in the respective image folders. [File name = neb.zip, around 396 MB]
Your explanation will help me understand the conceptual framework better. Very grateful for your help.

[ravindra]

I have attached the results in the post above.
Many thanks for your Prof. Henkelman.

graeme wrote:
> If you attach a tar.gz file of the simulation I can help with this
> question. Most likely there is a simple explanation - the NEB forces are
> composed of only the real forces perpendicular to the path plus a spring
> force to keep the images equally spaced. For convergence, you should look
> at the NEB forces.