I was trying to learn NEB calculation in LAMMPS, by studying the simple problem of vacancy migration process in BCC Iron. But I am not able to get the correct energy barrier for the process, even after several attempts.
I have a few questions/doubts.
- When using the neb command, should I have a non zero value for energy tolerance? Or should I set zero for energy tolerance and a nonzero value for force tolerance?
- In some of my simulations I see that the Reaction coordinates near 1, collapse onto a single RD value. For eg; RD’s 0.7, 0.8 and 0.9 all collpase onto 0.6, thus leaving a gap, between RD=1 and RD=0.6. Why is this happening?
Comments below.
Steve
I was trying to learn NEB calculation in LAMMPS, by studying the simple
problem of vacancy migration process in BCC Iron. But I am not able to get
the correct energy barrier for the process, even after several attempts.
I have a few questions/doubts.
1. When using the neb command, should I have a non zero value for energy
tolerance? Or should I set zero for energy tolerance and a nonzero value for
force tolerance?
It's up to you. If you set the tolerance to 0.0, you will likely run until
your timestep limit is exceeded. The more important question is
whether you have converged to a stable answer.
2. In some of my simulations I see that the Reaction coordinates near 1,
collapse onto a single RD value. For eg; RD's 0.7, 0.8 and 0.9 all collpase
onto 0.6, thus leaving a gap, between RD=1 and RD=0.6. Why is this
happening?
Could be lots of reasons. Maybe your spring constant is too small.
Maybe you are trying to do NEB on a bad starting/ending configuration.