About POTIM in different IBRION

Problems running VASP: crashes, internal errors, "wrong" results.


Moderators: Global Moderator, Moderator

Post Reply
Message
Author
cwandtj
Newbie
Newbie
Posts: 8
Joined: Tue Feb 21, 2006 2:45 pm
Location: Gainesville, Florida, US

About POTIM in different IBRION

#1 Post by cwandtj » Wed Jun 18, 2008 1:25 am

Hi, VASP master and VASP users.

1) I have questions in the meaning of POTIM in different IBRION. Can we think that POTIM value in certain IBRION (say IBRION=2) is directly applied to different IBRION?

2) In IBRION=2(CG), I got mainly trialstep >1.0 with my POTIM=0.2. However, for some few steps, I got trialstep<1.0.
This means I need to use small POTIM?

3) Even though it's mentioned that POTIM has no physical meaning, I am not sure whether I can have "different" local minima by different POTIM value. Is it possible?

For me, I used POTIM=0.3 (IBRION=3) for some complicated oxide surface. I am testing with POTIM=0.1 to see my previous POTIM is OK. Do you think that this is meaningful?

Thank you for your help in advance.

CW
Last edited by cwandtj on Wed Jun 18, 2008 1:25 am, edited 1 time in total.

admin
Administrator
Administrator
Posts: 2921
Joined: Tue Aug 03, 2004 8:18 am
License Nr.: 458

About POTIM in different IBRION

#2 Post by admin » Wed Jun 18, 2008 10:26 am

you can find a nice summary of POTIM in the VASP-workshop tutorial optionic.pdf (pp. 19 ff).
in principle, different choices of POTIM can lead to different local minima, depending on your potential hypersurface. The ionic optimization is based on the minimization of the forces (with additionat trial steps in order to avoid getting trapped in local minima). Therefore if POTIM is large and takes you to an adjacent local minimum, or if the guess of the starting geometry is too far from the global minimum (but close to a local minimum with a high transition barrier to the global minimum) the final geometry may vary for different choices of POTIM.
Last edited by admin on Wed Jun 18, 2008 10:26 am, edited 1 time in total.

Post Reply