Hello everyone,
I am using Lammps 14May-2016 version. I will describe my system first before asking my question. Basically it is a bead spring model. I have 4 beads to make up one chain. And there are many chains arranged together in a bundle. All the beads are of the same atom type and have the same mass.
Besides the harmonic bond style, angle style and dihedral styles, there are lj 6-12 potential fit between each beads in a chain and a lj 9-6 potential between beads in the adjacent chains.
Now my question is by specifying a command like “pair_coeff 1 1 100.0 3.5 9.0” , I can specify the interactions between beads in each of the chains. However what do I do for capturing the lj 9/6 potential which is basically between chains?
I thought of one solution. I can change the atom types in each chain and modify the data file. For example, if there are 8 chains I can have atom types 1-8. The lj9/6 can be captured by using something like the “pair_coeff 1 2 100.0 3.5 9.0” and so on. However, the problem is if I have many chains (around 50), it is very tedious to capture each cross term pair_Coeff command. Is there a better way for this? Any feedback will be much appreciated.
Thank you so much for your help.
Regards,
Hello everyone,
I am using Lammps 14May-2016 version. I will describe my system first
before asking my question. Basically it is a bead spring model. I have 4
beads to make up one chain. And there are many chains arranged together in
a bundle. All the beads are of the same atom type and have the same mass.
Besides the harmonic bond style, angle style and dihedral styles, there
are lj 6-12 potential fit between each beads in a chain and a lj 9-6
potential between beads in the adjacent chains.
Now my question is by specifying a command like "pair_coeff 1 1 100.0 3.5
9.0" , I can specify the interactions between beads in each of the chains.
However what do I do for capturing the lj 9/6 potential which is basically
between chains?
I thought of one solution. I can change the atom types in each chain and
modify the data file. For example, if there are 8 chains I can have atom
types 1-8. The lj9/6 can be captured by using something like the
"pair_coeff 1 2 100.0 3.5 9.0" and so on. However, the problem is if I have
many chains (around 50), it is very tedious to capture each cross term
pair_Coeff command. Is there a better way for this? Any feedback will be
much appreciated.
please check, if i do understand you correctly:
- a "molecule" would be a 4-bead chain
- the 12-6 LJ terms are only for *intra*molecular pairs
- the 9-6 LJ terms are only for *inter*molecular pairs
- you are currently using special_bonds lj/coul 1.0 1.0 1.0
if yes, i may have a suggestion for you.
if no, please try explaining again
axel.
You are totally correct in understanding the problem. Please do provide your feedback. Thank you so much.
You are totally correct in understanding the problem. Please do provide
your feedback. Thank you so much.
ok. then i would suggest the following.
- define the pair style only for the *inter*molecular interactions, i.e.
lj9-6 interactions.
- use "neigh_modify exclude molecule/intra all" to turn off those
interactions within each molecule
- switch your bond style to using hybrid combining your current bond style
with bond style table
- create a table for bond style table, that has the desired 12-6 LJ
intra-molecular potential.
- define *additional* bonds for all intramolecular pairs and use the bond
table on those
axel.