UT theoretical chemistry code forum

It looks to me as if the forces are not accurate enough to properly evaluate the curvature and rotational force. When things are working, the Torque should drop during the 4 rotation iterations that you are allowing per step, and the Curvature should also drop systematically. In steps 3-5, this is c...

Can you post the DIMCAR - this is a short file that will give a good idea of what is going on. There is some overhead to get the dimer oriented at the start, but I'm a little surprised that it moved so little after 45 force calls. Are you using conjugate gradients (IBRION=3, POTIM=0, IOPT=2)? To con...

V2.02 (05/03/07) released. Changes include:

- A global lbfgs optimzer for the NEB (use LGLOBAL=.TRUE. with IOPT=1)
- More robust writing of the CONTCAR and CENTCAR from the dimer method

If you're willing to invest a little time to try a new system, I think our akmc script is the best way to go. We could also use some feedback - if there are parts of it that don't make sense, or are confusing, we'll fix, change, or add documentation for them. The way this script works is that you fi...

Thanks for the comments. I've changed the 32 bit linux binary to one built using ifort instead of pgf. For a 32 bit machine, you can build the program with the portland group compiler using Makefile.pgf_32. The Makefile.pgf is for 64 bit machines (as that error indicates). It does seem that we have ...

How does this relate to a Bader analysis or transition state calculations? If this is a general vasp question, the appropriate forum is at: http://cms.mpi.univie.ac.at/vasp-forum/forum.php

The forces and energies are very large. I agree with Andri that it would be best to start by looking at the path. If your initial path is somewhat unphysical (a long way from the minimum energy path) it is best to start with a conservative optimizer (IBRION=3 or IOPT=3 (quickmin) or IOPT=7 (fire)) u...

I think you would have to test it to know for sure. My guess is npar=1 would be best (a higher values doesn't really make sense), and the other values the same.

This is possible, but it needs to be set by your queuing system or mpirun initiation of the job (not npar). My guess is that your job has been given 4 mpi slots to run on, since you have 4 processors. If you are running directly, you can use mpirun -np 8, and duplicate the nodes in your mpi machine-...

The energy that vasp reports in the 1F line is the sum of the energy of all images. In this case, it looks like you have 2 images so that the final energy is twice that in the loop.

PGI (Portland Group) is just a fortran compiler. It is generally good to build both serial and parallel versions of vasp. To use the NEB, you must have the parallel version. The dimer and lanczos methods can both be run in serial or parallel.

I can't imagine how this happened. We add a few modules which might slightly increase the binary size. Our binaries, built with PGI, are 8MB. The decrease in binary size is very strange.

These questions are a little basic for this forum. A short description of harmonic transition state theory can be found in the following class notes: http://theory.cm.utexas.edu/statmech/notes/notes12.pdf The equation that Dan gave you is 12.13 in these notes, and the text (along with any statistica...

You are a brave person running this on windows; but in principle it should work. If you do come across anything that breaks, and see a way to fix it, we'll update the scripts. I think there are so few people running on windows, however, that we won't do any cygwin testing ourselves. As a side issue,...

If you want to find a path that is qualitatively different from the linear path, you need to manually construct a different initial path. You're pretty much on your own here because we do not have automatic ways of sampling different paths. One possibility is to use different initial and final state...