fix bond/break changes

Hello all,
I am attempting to run a simulation with several defined types of monomers. I’ve used “fix bond/create” to create a reversible bond between two monomers of “type 3” in my simulation. I use the keywords iparam and jparam to change the type of the monomers to “type 4” once a bond is formed. This works well, and the type change is reflected in the dump file I dump my atom information out to. However, because the bonds I’m simulating are reversible, I want the bonds I’ve formed to break over time. When this happens, I want the monomers that were bonded to change back to “type 3”. This is where I’ve encountered trouble.
I know that there is no provision for such an action in the “fix bond/break” command, so I have had to attempt to modify the code of the command to create such a provision. I cannot figure out why I cannot change types back when editing the file. I see the code where the bond is “broken”, and have tried to insert statements into that function of the code, but to no avail. For instance, if I put into the “void FixBondBreak::post_integrate()” function a statement such as “atom->type[i] = 3;” just as an example to force what I believe to be the location of the atom I want to change types of to be “type 3”, when I remake LAMMPS, copy it and run my simulation, I notice that the dump file still contains the incorrect type.
What am I missing here? Do I have some sort of communication issue that should be rectified by a comm statement? If so, what am I lacking? What comm statement or packing statement do I need? Am I completely misguided in my understanding of the atom->type statement in changing the “type” part of the atom file? Is there something I need to edit in the dump command? Where is the issue? Please help me figure this out. Any advice or suggestions are greatly appreciated.
Best regards,
Joey Emfinger

Don't know. So long as you are changing the
type of an owned atom, and not a ghost atom,
which means i < nlocal is required, then that
should change the atom type permanently.