bond/create help

Dear Lammps Users,

I tried to understand on how the fix bond/create works.
I want to create a bond between the Carbon atom (brown, atom type 3)
with Carbon atom -(lightblue, atom type 1).
What I am trying to do is after the new bonds are created between the
type 1 and type 3, the atom type 1 change to atom type 2.

I am using dreidingFF to do the MD. However, after the simulation
finish no new bonds are created. I also tried different cut-off in fix
bond/create command to 10.0, but still no new bonds are created. Where
did I do wrong ?

Regards,
Erik

test.png

link-test.data (2.67 KB)

in.link (3.11 KB)

dear erik,

Dear Lammps Users,

I tried to understand on how the fix bond/create works.
I want to create a bond between the Carbon atom (brown, atom type 3)
with Carbon atom -(lightblue, atom type 1).
What I am trying to do is after the new bonds are created between the
type 1 and type 3, the atom type 1 change to atom type 2.

I am using dreidingFF to do the MD. However, after the simulation
finish no new bonds are created. I also tried different cut-off in fix
bond/create command to 10.0, but still no new bonds are created. Where
did I do wrong ?

there are many problems with your input.

the biggest is, you cannot use fix bond/create
(or any other of the bond/change_something) fixes with
a force field that also has angles and dihedrals.
adding and removing a bond in such a system
also requires to add the corresponding angle
and dihedral terms.

but also your neighbor list settings a "unusual"
and then your special_bonds doesn't match with
what dreiding requires and even more so, your
potential parameters look _very_ unusual for
dreiding, as well as your choice of dihedral_style.

have a look and compare to the dreiding example
in the lammps distribution. where do you get those
parameters and settings from and what kind of
a model is this?

axel.

Dear Dr. Kohlmeyer,

Do you mean that if I want to use that fix bond/create, I should not
specify angle, dihedral, and improper data in the input and molecule
data file?

For the parameter, I compared it to Mayo(1990) paper directly and
chose that I think its comparable.
Some of the number I round it up, may be thats why its unfamiliar.
I use charmm as dihedral also because its looks like the one in the
paper which E = 0.5K{1-cos[n(angle - angle_eq]}.

Regards,
Erik

From the fix bond/create doc page:

When a bond is created, you might wish to induce new angle and
dihedral interactions that include that bond. However, LAMMPS does not
create these angles and dihedrals, even if your simulation defines an
angle_style or dihedral_style.

This fix requires that the pairwise weightings defined by the
special_bonds command be 0,1,1 for 1-2, 1-3, and 1-4 neighbors within
the bond topology. This effectively means that the pairwise
interaction between atoms I and J will be turned off when a bond
between them is created. It also means that the pairwise interaction
of I with J's other bond partners will be unaffected by the new bond.

Steve