LAMMPS Supported Keywords/Laundry list

[ctk3b]

I'm just going to keep track of what needs to be done in LAMMPS here so if anyone who feels the desire to chip in on something knows where to look.

• The hyperlinks are to the LAMMPS documentation for the respective sections.
• Where appropriate, in parentheses after each entry is what we call this in InterMol. This is incomplete - I'm just kind of adding as I go.
• Please edit this list if you have anything to add
• [ ] Bonds
  • [ ] Reading
  • [ ] class2
  • [ ] fene
  • [ ] fene/expand
  • [x] harmonic (Bond)
  • [x] morse (Morse)
  • [ ] nonlinear
  • [ ] quartic
  • [ ] table
  • [ ] Writing
  • [ ] class2
  • [ ] fene
  • [ ] fene/expand
  • [x] harmonic (Bond)
  • [x] morse (Morse)
  • [ ] nonlinear
  • [ ] quartic
  • [ ] table
• [ ] Angles
  • [ ] Reading
  • [ ] charmm (Urey-Bradley)
  • [ ] class2
  • [ ] cosine
  • [ ] cosine/delta
  • [ ] cosine/periodic
  • [ ] cosine/squared (G96Angle)
  • [x] harmonic (Angle)
  • [ ] table
  • [ ] Writing
  • [x] charmm (Urey-Bradley)
  • [ ] class2
  • [ ] cosine
  • [ ] cosine/delta
  • [ ] cosine/periodic
  • [x] cosine/squared (G96Angle)
  • [x] harmonic (Angle)
  • [ ] table
• [ ] Dihedrals/Impropers
  • [ ] Reading
  • [ ] Proper
    • [ ] charmm
    • [ ] class2
    • [ ] harmonic
    • [ ] helix
    • [ ] multi/harmonic (Ryckart-Bellemans, up to 5 coefficients)
    • [x] opls (Fourier)
  • [ ] Improper
    • [ ] class2
    • [ ] cvff
    • [x] harmonic (ImproperHarmonic)
    • [ ] umbrella
  • [ ] Writing (parentheses show style of LAMMPS dihedral used to resolve case)
  • [ ] DihedralTrig
    • [x] improper (dihedral_style, charmm)
    • [x] proper where 6th and 7th coeff are zero (dihedral_style, multi/harmonic)
    • [ ] proper where 6th and 7th coeff are non-zero
  • [x] ImproperHarmonic (improper_style, harmonic)
• [ ] Handling of all input commands - see list in #73
  • [ ] resolving custom user variables (this could get really messy and I'm still not sure if we really want to support this kind of parser)
  • [x] support for more unit sets
  • [x] lj
  • [x] real
  • [x] metal
  • [x] si
  • [x] cgs
  • [x] electron
  • [x] micro
  • [x] nano
• [ ] Support for non-orthogonal boxes
• [ ] Translation of constraints
• [ ] Mixing rules
  • [x] geometric (combination rule 3)
  • [x] arithmetic (combination rule 2)
  • [ ] sixthpower

[mrshirts]

Fantastic detail. This helps us really understand what is going on and the range of time it will take. Most of these will be easy, if somewhat time intensive. Others will be harder . . .

[ctk3b]

I'm trying to figure out the best way to write out DihedralTrig in LAMMPS.

LAMMPS appears to support RB style dihedrals but only out to 5 coefficients: http://lammps.sandia.gov/doc/dihedral_multi_harmonic.html

Alternatively we have the CHARMM convention which should function just like proper 1 (or 9 with multiple entries): http://lammps.sandia.gov/doc/dihedral_charmm.html

Does using multi-harmonic for anything with c6=0 and otherwise decomposing to CHARMM style sound like a reasonable approach?

At some point, we may also need to support a distinction between CHARMM and other forcefield so that LAMMPS ends up using the correct pairstyle (I need to do more reading here).

[mrshirts]

Check out the gromacs writing logic- i had to go through pretty much the same process for the conversion. Including c6=0.

Out of contact for 24 hours or so- boarding plane'

Sent from my iPhone

On Jun 6, 2014, at 16:55, Christoph Klein notifications@github.com wrote:

I'm trying to figure out the best way to write out DihedralTrig in LAMMPS.

LAMMPS appears to support RB style dihedrals but only out to 5 coefficients: http://lammps.sandia.gov/doc/dihedral_multi_harmonic.html

Alternatively we have the CHARMM convention which should function just like proper 1 (or 9 with multiple entries): http://lammps.sandia.gov/doc/dihedral_charmm.html

Does using multi-harmonic for anything with c6=0 and otherwise decomposing to CHARMM style sound like a reasonable approach?

At some point, we may also need to support a distinction between CHARMM and other forcefield so that LAMMPS ends up using the correct pairstyle (I need to do more reading here).

— Reply to this email directly or view it on GitHub.

[b-v]

Note that LAMMPS also supports higher order terms for RB style dihedrals, via dihedral/nharmonic (http://lammps.sandia.gov/doc/dihedral_nharmonic.html), which is part of the USER_MISC package.

[JoshuaSBrown]

Where in your table would the input keywords:

bond_coeff
angle_coeff
dihedral_coeff
pair_coeff

be placed?

[ctk3b]

I'd say * Handling of all input commands is an appropriate home