Open maaren opened 7 years ago
I don't think it is very clearly documented, but I think that the LJ14 interactions are added in the dihedral potential. See: lammps-17Nov16/src/dihedral_charmm.cpp
if (implicit) forcecoul = qqrd2e * q[i1]*q[i4]*r2inv;
else forcecoul = qqrd2e * q[i1]*q[i4]*sqrt(r2inv);
forcelj = r6inv * (lj14_1[itype][jtype]*r6inv - lj14_2[itype][jtype]);
fpair = weight[type] * (forcelj+forcecoul)*r2inv;
if (eflag) {
ecoul = weight[type] * forcecoul;
evdwl = r6inv * (lj14_3[itype][jtype]*r6inv - lj14_4[itype][jtype]);
evdwl *= weight[type];
}
I have managed to create a lammps charmm dihedral topology which reproduces the gmx dihedral energy for 1-butanol, but I still have not been able to get the lammps E_vdwl equal to gmx LJ 14+SR
When I run intermol with https://github.com/dspoel/liquid-simulations/blob/master/FF/GROMACS/CGenFF/TOP/GAS/1-butanol.top as input I get:
when I remove the [ pairtypes ] section in 1-butanol.top and rerun I get:
So something goes wrong with 1-4 input of lammps. After reading: http://lammps.sandia.gov/doc/pair_charmm.html
I tried to use: pair_style lj/charmm/coul/charmm 19.999 20.0
and experimented with the pair_coeff parameters to get epsilon_14 and sigma_14 in there, but until now I have not got something that reproduces the energies with lammps.