I have used Reaxff forcefiled in LAMMPS to perform MD simulation and I want to observe bond breaking and formation during the simulation. I heated the system to 1800 K with the fixed heating rate to see whether there is bond breaking. I got the trajectory from the lammps, and then I exported the trajectory to VMD to observe the bond breaking. However, I got strange results. The first picture (attachment) is the initial structure; the second picture (attachment) is the structure from specific temperature. It seems that there is bond breaking in second one, but it looks so strange. My questions are
(1) Whether VMD can be used to analyze results from Reaxff forcefiled in LAMMPS
(2)If not, is there other tools can be used to analyze results from Reaxff forcefiled in LAMMPS
I have used Reaxff forcefiled in LAMMPS to perform MD simulation and I
want to observe bond breaking and formation during the simulation. I heated
the system to 1800 K with the fixed heating rate to see whether there is
bond breaking. I got the trajectory from the lammps, and then I exported
the trajectory to VMD to observe the bond breaking. However, I got strange
results. The first picture (attachment) is the initial structure; the
second picture (attachment) is the structure from specific temperature. It
seems that there is bond breaking in second one, but it looks so strange.
My questions are
(1) Whether VMD can be used to analyze results from Reaxff forcefiled in
LAMMPS
(2)If not, is there other tools can be used to analyze results from Reaxff
forcefiled in LAMMPS
there are two issues with your inquiry that make it difficult to answer.
1) you seem to be mixing up "visualization" and "analysis". what you are
asking about specifically are *visualization* issues and particularly about
VMD. this kind of question is best directed to the VMD mailing list. had
you done some (more) web searching, you should have found that inquiries
about visualizing systems with bond forming or bond breaking is a recurring
topic on the VMD mailing list. VMD in its default settings assumes a
conventional classical MD simulation, where bonds cannot be broken and it
is optimized for that. it has representations, though, that can recompute
bonding patterns based on simple heuristics and with Tcl scripting, more
complex approaches or external topology information can be applied. but
this doesn't happen automatically; it requires some experience, programming
skills and some understanding of the underlying architecture.
2) you seem to have an incomplete understanding of what constitutes a bond
and how it is determined. this is not an on/off thing and most
visualization tools use extremely simple heuristics. as far as the
underlying physics is concerned, bonds are an empirical concept based on
rationalizing observed behavior that finds a strong correlation between
distance and interaction strength. reaxff tried to model this by using a
"bond order" term (a floating point number, not an integer) that represents
bond strength. to monitor, you have to use the analysis tools integrated in
LAMMPS in the USER-REAXC package.