Dear all,

I have a system with 4 atom types. I would like to compute the *epair = evdwl + ecoul + elong + etail* for each pair of the atom types. While using *thermo_style* command i can output the *epair* for the entire system, but I am unsure of how to obtain it for each pair of atoms. An alternative to this I considered was to define multiple *compute pe* with *pair* keyword, but from the doc pages it seems that the group ID needs to be ‘all’ for this compute. Again, using *compute group/group* seems to give the total energy of the between the defined groups. Any suggestions will be helpful.

Thanks

Ganesh

Well… compute group/group does not include kinetic energies of the group. It will calculate the first two terms of your equation. etail and elong is tricky (probably much more complex as well) so it does not calcuate that. The manual says that the interaction energy is the pairwise energy between two groups.

As Vikas said, you can get the evdwl+ecoul interactions between

2 groups (group A = type1, group B = type 2) with compute group/group.

You could also compute etail yourself' it's just a formula using an integral

from the cutoff to infinity, and the relative densities (assumed constant)

of the 2 types. There is no way to compute elong for just the

two types interacting. LAMMPS does all the charges together for PPPM.

You would have to do a PPPM calc with only the atoms of the 2 groups, and

take out the self-interactions.

Steve

Thanks Vikas and Steve. Then from what I understand, the sum of the outputs for *evwl+ecoul* from *thermo* command should be same as that from *compute group/group* when both groups are ‘all’

Please suggest me if I am correct.

Ganesh