[lammps-users] about special_bonds command

LAMMPS LAMMPS Mailing List Mirror
1

Dear ALL,

I don’t know how to use the special_bonds command? If I do Langevin simulation of polymer with coarse-gained bead spring model and FENE bond potential. How can I give the special_bonds numeric coefficients ? What is the difference of special_bonds 0.0 1.0 0.0, special_bonds 0.0 1.0 1.0, and special_bonds 1.0 1.0 1.0? And, how can I choose them? Could some experts give me any suggestion?

Thanks in advance!

2

hi steve and all,

I want to use the following potential:

e (ij) = A/r + Bexp(-Cr) - D(sigma/r)^^12 - E(sigma/r)^^6

to model the two types atom interaction.

where r (rij) is the distance between atom i and j. A,B, C,D is constant.

I am aware that the long range potential should use the Ewald summation. I was wondering that how to achieve the mixed potential? Do I need to write the code myself? or Lammps already has all these seperated potential terms. If that, where can I start? If someone had similar experience, please give me some suggestions. I’ d appreicate that.

Thanks

AC

3

If you understand what 1-2, 1-3, and 1-4 interactions are in a molecule
with bonds, then special_bonds will make sense. For a linear FENE
chain, just number the beads in the chain from 1 to N. The 1-3 partners
are beads (1,3), (2,4), (3,5), etc. Pairwise (LJ) interactions between
those pairs will be turned off if the 2nd special_bonds coeff (1-3) is 0,
otherwise they will be turned on partially or fully.

Steve

4

Not sure what you mean by two types and "mixed". But if
you just want the potential in your formula, then I suppose
you could use pair hybrid/overlay with a sub-styles coul (A term),
lj/cut (D and E terms) and buck (B and E terms, but set the 1/r6 coeff
to 0).

Or you could just take pieces of those 3 files and create your
own new pair_style spiff.

Steve

5

Thanks Steve and AEI for kind reply. In the special_bonds doc page: http://lammps.sandia.gov/doc/special_bonds.html

you can read that" For a lj units system with FENE bonds a setting of special bonds 0 1 1 should be used. ". I can’t understand why we must turn on the 1-3, and 1-4 interactions and turn off the 1-2 interaction.

Thanks.

wende

2008/9/15 Steve Plimpton <[email protected]>

6

That's the typical FENE bead-spring model when you don't have angle
or dihedral terms in your chain.

Steve