[lammps-users] questions to make data file for posistional data and force field

LAMMPS LAMMPS Mailing List Mirror
1

Thanks a lot for your answer really, Steve.
I’ve made data file (data.graphene) for graphene using topotools of VMD from pdb file following advice of Axel (many thanks for it, Axel).
I think I should edit it afterward to meet my system.
I’d like to use airebo potential and I’ve read “read data” parts of manual carefully.
However, I have six questions on it as follows and anaswers on them would help me to understand LAMMPS greatly.

  1. I suppose molecule ID is not required additionally for my system and the format of “atom_ID atom_type coord_x coord_y coord_z” will be sufficient for “Atoms section”.
    In addition, here, atom_type should be 1 if I write “pair_coeff * * CH.airebo C” in my input file.
    That is, the following may be O.K.

1 1 17.028999 34.431000 5.132000 # C1 1
2 1 17.028999 7.371000 5.132000 # C2 1
…

What do you think on it?

  1. The format for Bonds section is “bond_ID bond_type ID_1st_atom ID_2nd_atom”.
    Here, where is bond_type defined generally?
    If I use airebo potential, is it described in CH.airebo? (It is hard to find bond_type notation in that file.)
    What should I write bond_type for graphene using airebo potential?
    Should bond_type be “1”?

  2. There are only atoms and bonds section in data.graphene made from topotools.
    If I use airebo potential for graphene (which is composed of carbon only and, maybe, there will be merely one atom type carbon, I guess.), will other sections for angle, dihedral,improper etc. be needless?

  3. In questions 3, if sections for angle, dihedral, improper are also required in the use of airebo potential, where can I find description for types (1~ Ntypes for angle_type, dihedral_type, improper type) of angle, dihedral, improper?

  4. I use periodic boundary condition for my system.
    In spite of the use of PBC, may I omit optional arguements of nx,ny,nz (for image)?
    That is, will LAMMPS consider the most closed atom for a certain atom_ID automatically (considering various image ones)?

  5. Generally speaking (not only for CH.airebo), where can I find detailed “type description” for each sections (Bonds, Angles, etc.)?

Cheers,

Luke

2

2010/5/31 JhonY. I. <[email protected]...>:

Thanks a lot for your answer really, Steve.
I've made data file (data.graphene) for graphene using topotools of VMD from
pdb file following advice of Axel (many thanks for it, Axel).
I think I should edit it afterward to meet my system.
I'd like to use airebo potential and I've read "read data" parts of manual
carefully.
However, I have six questions on it as follows and anaswers on them would
help me to understand LAMMPS greatly.

1. I suppose molecule ID is not required additionally for my system and the
format of "atom_ID atom_type coord_x coord_y coord_z" will be sufficient for
"Atoms section".
In addition, here, atom_type should be 1 if I write "pair_coeff * *
CH.airebo C" in my input file.
That is, the following may be O.K.

1 1 17.028999 34.431000 5.132000 # C1 1
2 1 17.028999 7.371000 5.132000 # C2 1
...

What do you think on it?

first of all, this entry has to match what you use as
atom style in your input.. if you use topotools, then
it will write the format accordingly with the atom style
given after the filename. it defaults to full, but can
write several others, including atomic (like above).

2. The format for Bonds section is "bond_ID bond_type ID_1st_atom
ID_2nd_atom".
Here, where is bond_type defined generally?
If I use airebo potential, is it described in CH.airebo? (It is hard to find
bond_type notation in that file.)

you should first read the paper describing AIREBO.
this potential has no fixed bonds and thus there
must be no bonds given in the data file.

What should I write bond_type for graphene using airebo potential?
Should bond_type be "1"?

no.

3. There are only atoms and bonds section in data.graphene made from
topotools.

topotools writes what information is available to VMD. if desired it
can reconstruct angle, dihedral, and improper definitions from bond
topology and angles. however, for AIREBO, you have to use
topo clearbonds before writing the data file.

If I use airebo potential for graphene (which is composed of carbon only
and, maybe, there will be merely one atom type carbon, I guess.), will other
sections for angle, dihedral,improper etc. be needless?

see above. since you have only one atom type, you should do the
following after reading your .pdb file into VMD.

set sel [atomselect top all]
$sel set type C
$sel delete

that will reset all atom types to the same type. with that and
clearing all bonds, you should be ready for using AIREBO.

4. In questions 3, if sections for angle, dihedral, improper are also
required in the use of airebo potential, where can I find description for
types (1~ Ntypes for angle_type, dihedral_type, improper type) of angle,
dihedral, improper?

see above. to understand you should really reed the paper(s)
on the potential that you are using. similarly, for other potentials,
where bonded interactions are explicitly defined. it is the force
field that determines. what different types are required, typically
the depend on the selection of atom types. again, topotools
script can help here, as they allow to "retype" bond, angles
and dihedrals and then the respective types have to be
matched against a database. for other force fields, these are
determined based on residue templates.

5. I use periodic boundary condition for my system.
In spite of the use of PBC, may I omit optional arguements of nx,ny,nz (for
image)?

read the lammps manual more carefully.

That is, will LAMMPS consider the most closed atom for a certain atom_ID
automatically (considering various image ones)?

with periodic boundary conditions, yes.

6. Generally speaking (not only for CH.airebo), where can I find detailed
"type description" for each sections (Bonds, Angles, etc.)?

see above. and the bond_style/angle_style/dihedral_style/improper_style
of the lammps documentation.

axel.