Bonds affected by neighbor lists?

LAMMPS LAMMPS General Discussion
5

Sorry, I don’t think I understood.

Yes, I see that pair_style list needs an explicit list of pairs and their interactions like

15 259 lj126     1.0 1.0      50.0
15 603 morse    10.0 1.2 2.0  10.0 # and another comment
18 470 harmonic 50.0 1.2       5.0
19 332 quartic  10.0 5.0 -1.2 1.2

but it does not have a FENE bond that I am using for physical reasons (it behaves ~like an entropic polymer of finite length). pair_style list has harmonic and I could try to choose harmonic parameters to reproduce fene behavior as much as possible but it will not be great.

So was I suggesting that if I had a long bond between say atoms 1 and 2 then I could still have a fene bond between 1 and 2 but also add something like 1 2 harmonic 1e-10 0 100 in the pair_style list so that 1 and 2 are never lost (within r_c = 100). But how I understood your last answer is that this will not work and the interactions from pair_style list will not make other pair_style-s to keep track of selected atoms - is that right?

And another question about a more complicated molecule. I found a related thread here and you said there that “Perhaps, with setting special_bonds lj/coul 0.0 1.0 1.0 you can avoid it, but I am not certain.” and the guy answered that you were “spot on” but did not give much detail.

So, my questions:

  1. Do you now think your suggestion about effectively enabling 1-3 and 1-4 relieves the requirement for them to be inside comm_modify cutoff Rc? I see you were “not certain”, so - did anything change?

I tried (using special_bonds fene is the same as you suggested, right?) and it seems to be working but I am not sure if I can rely on this enough to launch many tasks and not check each one for crushes manually.

  1. If doing special_bonds fene does not really relieve the worries about 1-3 and 1-4, then can I not use special_bonds explicitly at all - will it help? It again seems to be working on a sample run but I am not sure if that’s just luck.