UT theoretical chemistry code forum

Dear all:

One of my NEB jobs was killed by the computer supervisor. It had been running for several days, finishing 27 ionic steps. My question is: Is that possible to restart this job using the configuration from 27th ionic step, instead of restarting it from stracth? If it's possible, how do I do that?

Thanks!

Sure, you can copy the CONTCAR files to the POSCAR files and continue. We also have a little script for doing this, which works for regular, neb, and dimer calculations called vfin.pl. If you are in your main run directory and you execute 'vfin.pl run1', all of the previous calculation information will be stored in the directory run1. The main directory will then be ready for continuing the calculation.

Hi, Prof. Graeme. Thank you so much for all the tips.

Then, after that the vfin.pl has been run, how about the CHGCAR in the 0X/？ The coverged CHGCAR and WAVECAR are usefull for a restart run? And we need to set ISTART = 1and ICHARG = 1 in the INCAR?

You an restart an NEB using the WAV/CHGCAR as you would a regular calculation.

Then, why the nebclean.sh code only copy and tar the CHGCAR and WAVECAR to the run1, but without leave a copy in the original directory?

It is safe to change the code to make it?

Ah, good point. We generally do not restart using the CHGCAR and WAVCAR. Yes, it would be reasonable to modify the scripts to keep a copy in the original directories.

To strop ca NEB run, to set LSTOP = TRUE in STOPCAR in run directory?