Dear Experts,
I am facing the convergence issue in one of the NEB calculations for one 2d system. After 132 iterations, I didn't get the convergence, and the run time was over.
Please suggest about it.
Thanks and regards,
Poonam S.
NEB convergence problem
Moderator: moderators
-
- Posts: 4
- Joined: Wed Oct 06, 2021 1:46 pm
NEB convergence problem
- Attachments
-
- pd_image.zip
- (62 MiB) Downloaded 749 times
Re: NEB convergence problem
The biggest problem is that you don't have our vtstcode linked into your vasp binary. You have appropriate settings for the VTST code, but as it is, you are telling vasp to do nothing (IBRION=3, POTIM=0), so there has been no relaxation from your initial path.
-
- Posts: 4
- Joined: Wed Oct 06, 2021 1:46 pm
Re: NEB convergence problem
Dear Expert,
I linked vtstcode with vasp binary and I rerum my calculations. Now I am getting the meaningful barrier but now the problem is after completed 90 iterations it is not converging and the run time is over. I am using the same input files. Please suggest about it.
I linked vtstcode with vasp binary and I rerum my calculations. Now I am getting the meaningful barrier but now the problem is after completed 90 iterations it is not converging and the run time is over. I am using the same input files. Please suggest about it.
- Attachments
-
- convergence_problem.zip
- (51.7 MiB) Downloaded 679 times
Re: NEB convergence problem
It looks fine to me. The max force is down to 0.02 eV/Ang and the barrier is dropping slowly but steadily. Just continue the run. There will be faster optimizers, but start by getting it to converge with your current settings and then you can try second order optimizers if you want to increase efficiency.