COMPASS force field is supported in LAMMPS by the ‘class 2’ option. It gives ‘class 2’ option bond_style, angle_style, dihedral_style and pair_style.
However, for simple inorganic molecules (SiO2) for example, the COMPASS paper by Sun, ( http://pubs.acs.org/doi/pdfplus/10.1021/jp992913p ) says that because of the simplicity of molecules in this domain, "The total energy is written into valence contributions of bond stretching, angle deformation, cross-coupling of bond-bond and bond-angle, and nonbonded terms that include the electrostatic and van der Waals (vdW) terms ". However, I cannot find the options for cross-coupling of bond-bond and bond-angle terms. LAMMPS gives dihedral_style, as an option.
Where am I to input parameters for bond-angle and bond-bond interaction terms in LAMMPS?
Details of my system -
It is an amorphous Silicon Dioxide unit cell (57x57x57 A^3, 5808 atoms), built with VMD plugin ‘inorganic builder’. Goal is to define force field for this system, as a learning exercise. I have chosen COMPASS force field as it is an atomistic force-field. The unit cell VMD-inorganic builder generates, has Silicon and Oxygen atoms explicitly defined. If snapshots of my system will help, I shall post them.
Regards,
Brahm Prakash Mishra
Mechanical Engineering
NIT Trichy
COMPASS force field is supported in LAMMPS by the 'class 2' option. It gives
'class 2' option bond_style, angle_style, dihedral_style and pair_style.
However, for simple inorganic molecules (SiO2) for example, the COMPASS
paper by Sun, ( http://pubs.acs.org/doi/pdfplus/10.1021/jp992913p ) says
that because of the simplicity of molecules in this domain, "The total
energy is written into valence contributions of bond stretching, angle
deformation, cross-coupling of bond-bond and bond-angle, and nonbonded terms
that include the electrostatic and van der Waals (vdW) terms ". However, I
cannot find the options for cross-coupling of bond-bond and bond-angle
terms. LAMMPS gives dihedral_style, as an option.
Where am I to input parameters for bond-angle and bond-bond interaction
terms in LAMMPS?
as class2 angle parameters. for details read the documentation for
read_data and class2 angles. while those are externally visible as
separate entities, internally LAMMPS handles angles, bond-bond, and
bond-angle interactions all as angles. similar for dihedrals.
Details of my system -
It is an amorphous Silicon Dioxide unit cell (57x57x57 A^3, 5808 atoms),
built with VMD plugin 'inorganic builder'. Goal is to define force field for
this system, as a learning exercise. I have chosen COMPASS force field as it
is an atomistic force-field. The unit cell VMD-inorganic builder generates,
has Silicon and Oxygen atoms explicitly defined. If snapshots of my system
will help, I shall post them.
good luck with that. VMD generally assumes a CHARMM-like force field,
which is quite different from a class2 force field. you are most
welcome to contribute code to build suitable topologies with VMD to
the topotools plugin. i don't know COMPASS well enough and never cared
enough to invest the necessary time.
axel.
You should reflect a bit more on why the chosen FF. I mean, what are the properties you are trying to compute that justify such a choice. Your reference paper seems to deal with the molecular state but then you mention dealing with a bulk system. Many properties of bulk SiO2 have been studied over the years with much more simpler atomistic FFs such as BKS, induced dipole Tangney-Scandolo potentials, etc. Again, it all depends on what you are trying to accomplish from your “learning exercise”. Just make sure you don’t end up killing a flea with a cannon…
Carlos