incorrect result when I use the CI-NEB compiled with vtst

[ljzhang123]

when i set ISIF=2 I got the incorrect negative energy barrier, the calculation seems converged. but when i use ISIF=3, the results seems more reliable. i don't know how to solve this problem, need i recompiled the vasp?

[filipejunqueira]

I get the same thing... Wrong barriers with ISIF=2. I'm doing a tests to see if the other values of ISIF give me still wrong values. As far as I understand, my installation on our home cluster was done with vtst.

The fact that this hasn't been addressed or at least acknowledged by a dev for for 8 years is so frustrating.

[martin.schlipf]

Please provide more information:

• Do you use the built-in NEB method or the one from VTST?
• When you compute the individual structure along the transition state, do all of the calculations converge properly?
• Did you relax your initial states i.e. the fixed points of your elastic band sufficiently?
• How do you conclude that the energy barrier is wrong; by comparing to experiment or some ab-initio reference?

To investigate the system, we need to have access to the input files of the calculation. Please attach them to your reply.