UT theoretical chemistry code forum

The adaptive kinetic Monte Carlo script is more complicated than our other scripts. A brief description is available at:
http://theory.cm.utexas.edu/vtsttools/akmc/

Lijun Xu, who wrote the script, has been working on a manual (which may be available). He can also answer specific question.

Our script is extremely simple; it just integrates the DOS times the energy, in the simplest possible way. You should be able to check it yourself - with a spreadsheet, for example. We have noticed that the center of the d-band is somewhat sensitive to the number of unoccupied band that you allow. I...

Yes, both PAW and the LAECHG=.TRUE. flag.

It looks like you may not have included core charges.

Follow the instructions at http://theory.cm.utexas.edu/bader/vasp.php

and see if the Bader surface are further from the atomic centers.

It's also possible that there is something else going wrong with your NEB calculation, but we can't tell from the force and energy on the one image.

We have not implemented this yet.

The min-mode following methods (dimer and Lanczos) can both be run in serial. This is a little bit of a trick, but you can also run a single climbing image NEB. To do this, you can generate a MODECAR vector between your initial and final states, and then do a dimer calculation starting from the midp...

It's not necessarily incorrect - this is a periodic crystal. It is, however, a crystal of 3 atoms in which two are moving a lot. This corresponds to infinite array of translating atoms, or some kind of phase transition. From that point of view, I don't see how the energy barrier could have much mean...

You need to contact the person who built vasp on the machine, or someone who understands how to build it from the source. Our posted instructions should make sense to that person.

You could, but these rotations also take time, so as long as the force is steadily dropping, it might be faster to keep going as it is. If it were not for the last two iterations (78 and 79) things would be looking great. And it may be converging just fine, as long as these last two iterations are a...

There are two parts to the vtsttools. First, there is a collection of scripts which you can download and put in your path: http://theory.cm.utexas.edu/vtsttools/scripts/ Second, there is code that links into vasp. This does not alter the normal operation of vasp, but instead just provides additional...

It looks to me as if it is converging, but a bit slowly. It also looks like there might be some noise in the force because it keeps rotating a bit, and the torque is not dropping to zero. This may be due to LREAL, but that's a bit of a guess. I expect that the force will eventually drop below 0.01, ...

The script does not properly deal with changes in the cell geometry. There needs to be additional code to interpolate the cell geometry between your initial and final POSCAR files. Using the NEB with a changing cell geometry is an interesting, but new area of research (to us, that is: John Wilkin's ...

It seems like there is some problem with the reading of the CHGCAR. But first, from your error message relating to Charge.f90, it seems that you might have a very old version of the code. Is that right? If your run 'bader -v', what version do you have. The Charge.f90 file was replaced by charge_mod....

This is a good question, and one that we don't know much about. You can set the external pressure, and ISIF=3 to have each image relax the cell volume. Images with different cell sizes, however, will appear to be separated by the distance moved by all atoms. This doesn't mean that the NEB will not c...