Bond/create issue in Images of the system

LAMMPS LAMMPS General Discussion
1

Hello,

I’m studying a bulk containing polymer chains (A), waters, and ions, in All-Atom MD using charmmff and periodic in 3 dimensions.
I already ran this system for a few ns successfully and I have values for “Image Flags” in my datafile.
In the second step, I added the second polymer (B) to the previous system using “read_data add append…”, and also it runs perfectly without any problem.

Then I used the “fix bond/create” command to create bonds between polymers (A and B). Everything goes well, except I get the famous warning “Inconsistent image flags”.

I checked the dump and it seems there are a few bonds that are created between polymer B and Images of polymer A, near the edge of the box.

Is there any solution to prevent such bond creations?

P.S. I’m working with the stable version of lammps (29Sep2021), and I work with write_data and read_data to continue my dynamics in each step.

2

The bond creation is correct. The problem is that when creating the bond(s) also the image flags should be made consistent.

While you are getting the warning, it should not have any bad consequences since the actual interactions will be computed between the respective closest periodic images of atoms not the original ones. Only when you visualize the trajectory with unwrapped coordinates you will have bonded interactions between atoms that are far apart. With wrapped coordinates, those should be close.