Replies, below.
Best wishes,
Ray
Dear Ray,
Thanks for the clarification. Certainly what you have said is true and I might have been misunderstood before. I will be grateful if you may kindly help me to figure it out.
If I used two configurations (one for the reactants and one for the products), would it be true to say that we have one data file and one replica?
My input script is as follows:
dimension 3
units real
boundary p p p
atom_style charge
atom_modify map array sort 0 0.0
read_data initial
pair_style reax/c lmp_control checkqeq yes safezone 1.8 mincap 150
pair_coeff * * ffield.reax H O
neighbor 2 bin
neigh_modify every 10 delay 0 check no
fix 1 all qeq/reax 1 0.0 10.0 1.0e-6 reax/c
timestep 0.001
minimize 1e-8 1e-8 999000 999000
min_style quickmin
Replicas
fix 2 all neb 1
timestep 0.001
min_style quickmin
dump 1 all custom 1 dump.coordNVT1 id x y z
dump_modify 1 format “%5d %25.10g %25.10g %25.10g”
neb 1e-8 1e-8 999000 999000 50 each Replica1
Is the way I set it up correct? The reactants configuration is provided as initial and read by LAMMPS using read_datra. right? And the products configuration is replica1.
Once I am running this input file, the calculations always show that the climbing replica is the last replica. I do not see a way how to automatically generate the replicas given the two fixed configurations. Can you please help me on this issue?
Thanks.
I suggest you start from the NEB example that LAMMPS provides. Carefully read the LAMMPS NEB doc page and familiarize yourself with NEB simulations and their analyzes and visualizations. Then you can modify the EAM script to adapt ReaxFF.
Note that you can supply replicas along the path, but that is rather advanced and delicate. I suggest you start from a more conventional way, which is supplying the start/end images.
Hope this helps.
If the climbing replica is always the last replica, this implies that your final replica is not a local minimum, but is uphill from all of the other replicas. In order to find a TS, you first need initial and final replica geometries that are local minima i.e. forces are relaxed close to zero and energy change is uphill in all directions (for small but not zero size displacements).