Dear all,

I have several dimer configurations as a function of distance obtained from quantum calculations (a relaxed PE scan but freezing the COM-COM distance between the two molecules). I am trying to do the same optimization in lammps as well and for this I tried using 1) fix spring/self 2) fix spring/chunk ( I am aware they both work differently). The distance though remain fixed, but both of them give different energy values, relative as well as absoulte.

Moreover when the dimers are far apart by 12 angstrom distance, ideally the energy value of such configuration should have been double of a monomer energy, but that’s not the case.

Can anyone guide me if I am missing something important here?

Part of input file of 1) fix spring/self

###################################

group com1 id 1 2 4 5 7 8

# atoms part of calculation of com of resid 1

group com2 id 34 35 37 38 40 41

# atoms part of calculation of com of resid 2

fix 45 com1 spring/self 10000

fix 46 com2 spring/self 10000

###################################

Part of input file of 2) fix spring/chunk

###################################

group com1 id 1 2 4 5 7 8

# atoms part of calculation of com of resid 1

group com2 id 34 35 37 38 40 41

# atoms part of calculation of com of resid 2

compute 8 com1 chunk/atom type

compute 9 com2 chunk/atom type

compute 10 com1 com/chunk 8

compute 11 com2 com/chunk 9

fix restrain1 com1 spring/chunk 1000 8 10

fix restrain2 com2 spring/chunk 1000 9 11

#################################################

Thanks,

Megha