1-5 neighbors

Hey all: I am thinking of using a polymer model different from the
canonical FENE bonds in my simulations. There is a pairwise potential
for the 1-5 neighbors along the chain,which is not the default
nobonded pair potential (LJ potential in FENE bonds). And the
potential is in a tabulated form.

So I will need to turn off the default nonbonded interaction between
1-5 neighbors. Special_bond command in LAMMPS can be used to turn off
default nonbonded interaction between the 1-2 (bond), 1-3(angle), and
1-4(dihedral) neighbors. Can it also be used to turn off the 1-5
interactions?

Besides, LAMMPS supports 1-2, 1-3, and 1-4 or bond, angle and dihedral
potentials in tabulated forms . Is there a similar command in LAMMPS
for 1-5 interaction in the tabulated form?

I would expect myself to modify the codes or add new commands. Just
want to make sure the features I mentioned have not yet been
incorporated into current LAMMPS.

Thank you.

Best,
Ting

Hey all: I am thinking of using a polymer model different from the
canonical FENE bonds in my simulations. There is a pairwise potential
for the 1-5 neighbors along the chain,which is not the default
nobonded pair potential (LJ potential in FENE bonds). And the
potential is in a tabulated form.

So I will need to turn off the default nonbonded interaction between
1-5 neighbors. Special_bond command in LAMMPS can be used to turn off
default nonbonded interaction between the 1-2 (bond), 1-3(angle), and
1-4(dihedral) neighbors. Can it also be used to turn off the 1-5
interactions?

no.

Besides, LAMMPS supports 1-2, 1-3, and 1-4 or bond, angle and dihedral
potentials in tabulated forms . Is there a similar command in LAMMPS
for 1-5 interaction in the tabulated form?

no.

I would expect myself to modify the codes or add new commands. Just
want to make sure the features I mentioned have not yet been
incorporated into current LAMMPS.

please have a look at previous discussions about
non-bonded potentials that can distinguish between
atoms in the same or different "molecule".
that is more likely a reasonable approach.
you just have to make sure that your set of
ghost atoms is large enough to include all
1-5 neighbors.

axel.

Hi

If you need to specify custom forces between atoms separated by more
than 3 bonds (for example 1-5 interactions), then you may have to give
all the atoms in the polymer a unique type, and manually specify
custom pair forces between atoms of type i and i+4 (for example 1 5,
2 6, 3 7, 4 8, ...) This was discussed here:
http://lammps.sandia.gov/threads/msg04534.html

The only problem with this approach if you have multiple polymers.
These pair forces will effect the interactions between pairs of atoms
even if they belong to different molecules. This is what Axel brought
up. One way to fix this is to create your own custom pair_style table
which checks to see whether two atoms have the same moleculeID. There
are some instructions for how to do this contained in this thread:
http://sourceforge.net/mailarchive/message.php?msg_id=29471770

... and there a working pair-style which already does this in this thread:
http://sourceforge.net/mailarchive/message.php?msg_id=29508433
(This code calculates Lennard-Jones forces, not tabular forces, but it
is useful to look at if you get stuck.)

Recently, a couple people requesting special 1-5 and 1-6 interactions
in LAMMPS. I don't think this will happen. If my understanding is
correct, then the point of having custom 1-4 interactions (and the
special bonds command) is to weaken the 1-2, 1-3, and 1-4 interactions
that would otherwise interfere with the desired bond-distances,
3-body-angles, or 4-body dihedral angles specified by the user. (I
guess that's why it stops at 4. It would be fun to post a request for
a 4-D version of LAMMPS with 5-body angle interactions. I'm for it.)
Cheers!
Andrew

Hi Andrew and Axel: Thank you for your reply. I will try using the
method you suggested.

Since I am using a table for the calculation of 1-5 neighbor
interaction, I will subtract the default nonbonded potential from the
table. This way I won't have to modify the special_bond command to
turn off default nonbonded potential between 1-5 neighbors.

Best,
Ting

2012/7/9 Andrew Jewett <[email protected]...>:

It's fine to subtract the 1-5 interactions, and replace them with
something else. But lammps does not keep track of them for you.
Again, if you give each atom in a polymer a unique type, then you can
determine if two atoms are 1-5 interactions by by checking the atom
types (atom->type[i], atom->type[i]). (You could also try looking at
the atom-ID using atom->tag[i]. But this seems more difficult.)

For comparison, lammps keeps a list of all of the 1-4 interactions
(and the 1-2, 1-3 interactions). (These are the lists of bonds,
angles, and dihedrals in the data file.) For charmm style
force-fields, the 1-4 interactions are recalculated by the code which
calculates the dihedral angle forces (dihedral_charm.cpp). But there
is no list of 1-5 interactions.

Good luck.
Andrew