Spring = 5 or -5?

[drgao]

Hi graeme,

Greetings!

I haven't linked the VTST code to my vasp, so I'm trying to use the original NEB method in VASP to get MEP. It seems that the converged result could be obtained more easily for spring = 5 than spring = -5. Is it reasonable to use IBRION = 3 and spring = 5? I found spring = 5 from this link (http://theory.cm.utexas.edu/wiki/index. ... EP_methods).

In addition, I have tested spring = 0 and sping =5 for the same job. The energies obtained from spring = 0 seems much more reasonable than that from spring = 5, but the spring = 0 calculation consumed much more CPU time. And spring = 0 and 5 gave completely different MEPs. Which one is more reliable?


Thanks very much

[kambiz]

I think this is not related to this forum.

Taken from VASP manual:

For SPRING=0, each image is only allowed to move into the direction perpendicular to the current hyper-tangent, which
is calculated as the normal vector between two neighboring images.

The nudged elastic band method is applied when SPRING is set to a negative value.

[drgao]

Thanks very much, kambiz.

By the way, what is the range of negative spring? from -1 to -5, or others?

Thanks

[kambiz]

Hi,
yes, I usually use a value between -5 to -1.

Cheers

[drgao]

Thanks a lot, kambiz.

Good luck