Bond atom missing

LAMMPS LAMMPS Beginners
1

Dears Hi
Am running simulation of polymer and I get error [ERROR on proc 0: Bond atoms 1314 1315 missing on proc 0 at step 1048 (src/ntopo_bond_all.cpp:61)], I tried to solve it but I couldn’t. I will attach my input script with molecule file, can somebody help me please
in.input.mod.4.lmp (4.1 KB)
PDMS.mol (2.2 KB)

2

Save people time and effort so they don’t give you advice that you have already followed and explain what you did try to solve it?

3

Dear akohlmey
I want to run simulation of PDMS and prepare it to make it as a membrane. First I did a simple simulation just for PDMS and I computed the properties like pressure, and know want to add soft potential to avoid overlaps but I can not get the end of running because this error [ ERROR on proc 0: Bond atoms 1314 1315 missing on proc 0 at step 1048 (src/ntopo_bond_all.cpp:61).

4

There are several issues here:

  • I mentioned before that your force field is bogus since PDMS unlike alkanes has polar atoms
  • I see only a bunch of monomers, no polymer. I also mentioned before that your system preparation is not creating the kind of system you say that you want
  • I have run your input and I see no problem with the soft potential. That run completes, your problem starts after you change the pair style
  • If you closely observe your output, you should see very high potential energy once you switched the pair style and then quickly the bond atom missing. That is quite a common situation and thus you need to apply at least one more of the techniques that are commonly suggested when this kind of error happens and you have close contacts between atoms
  • As (also) mentioned before, if you don’t find yourself a properly knowledgeable tutor that will train you in doing correct and meaningful MD simulations, you will continue to struggle with very elementary issues like you currently are. From what I see here it does not look like you have learned anything substantial since I gave this advice the last time and thus you are not very likely to succeed. Very few people have the ability to teach themselves MD without a tutor or collaborating with people that have the expertise. Keep in mind that having an MD simulation that does not crash is only a minor challenge. It can still be fundamentally wrong in many ways and figuring out what is wrong under such circumstances is much harder and essentially impossible with no previous experience or people to talk to. Posting questions in a forum is not very effective. Just look at how little progress you have made so far. If you would work with a proper tutor, you would probably running something meaningful within a week or two.
1 Like
6

Please post new questions as a new topic, and do not reply to an unrelated topic.