# perturbed hamiltonians

Hello,

I would like to post a question about thermodynamic integration.

Suppose that my system is characterised by a Hamiltonian H=K+V where K is the kinetic part and V a pairwise potential. I use LAMMPS to generate an equilibrium configuration for H at a certain temperature T, and pick a configuration r0 as a reference.

I then consider the perturbed system

H’/T=H/T+lambda N (r-r0)^2

where N is the number of particles, r0 the 3N-dimensional vector of the reference configuration and r the vector for a current configuration. Is it possible to add the quadratic spring as an interaction potential to LAMMPS? The closest tool that I have found is fix ti/spring tool, but it does not match with my approach since every change in lambda affect both the quadratic and the original term.

Solving, this would allow me to compute thermal average and, finally, perform the thermodynamic integration of whatever observable in my model.

Thank you for any suggestion!

Hello,

I would like to post a question about thermodynamic integration.

Suppose that my system is characterised by a Hamiltonian H=K+V where K is
the kinetic part and V a pairwise potential. I use LAMMPS to generate an
equilibrium configuration for H at a certain temperature T, and pick a
configuration r0 as a reference.

I then consider the perturbed system

H’/T=H/T+lambda N (r-r0)^2

where N is the number of particles, r0 the 3N-dimensional vector of the
reference configuration and r the vector for a current configuration. Is it
possible to add the quadratic spring as an interaction potential to LAMMPS?
The closest tool that I have found is fix ti/spring tool, but it does not
match with my approach since every change in lambda affect both the
quadratic and the original term.

​what you describe sounds a lot like what fix spring/self does.

axel.​