Page 1 of 1

Short questions regarding dimer method (it stops at one ionic run)

Posted: Thu Oct 13, 2016 12:04 am
by tp.joe
Hello,

I generated a modecar file using modemake.pl on initial structure and the likely transition state from CNEB. However, when I started it only run once and finished, the energy given is apparently the initial structure energy (POSCAR). In DIMCAR, this is what written

Step Force Torque Energy Curvature Angle
1 0.06844 --- -275.47229 --- ---

Not sure why there is no Torque, Curvature, and Angle. Am I doing something wrong here?

Re: Short questions regarding dimer method (it stops at one ionic run)

Posted: Thu Oct 13, 2016 1:25 am
by graeme
If you post a tar.gz file of the calculation, I'll take a look.

Check to see if it converged on the first iteration, i.e, if the initial force is lower than your ediffg criterion.

Re: Short questions regarding dimer method (it stops at one ionic run)

Posted: Thu Oct 13, 2016 1:58 am
by tp.joe
Hello sir,

I attached the file (Dimer.tar.gz) that shows DIMCAR only 1 step. The POSCAR is the initial structure file, and the MODECAR file is generated from modemake.pl of the presumably transition structure from NEB and the initial state.



I think this could have been much easier if I use neb2dim.pl, but I am not able to get it to run from the NEB file. I always get the message

" FORMING DIMER BETWEEN IMAGES 3 and 4

ZERO DISTANCE BETWEEN IMAGES AND"

and no new folders are created. I also attached the NEB.tar.gz for the NEB run that I intended to use neb2dim.pl on just in case you're interested. I do realize that I suppose to run "vfin.pl" prior to neb2dim.pl. But when I ran vfin.pl it creates an empty file instead.


Thank you very much Prof. Graeme,
Joe