Regarding Potential Energy Calculation

Dear Sir,

I would like to calculate the change in potential energy of a given system by a small distance to displace all the atoms in one dimension(i.e. say x and other two dims are fixed) without any change in kinetic energy. For that I was trying using different commands like change_box, displace_atoms, fix_defrom etc. After that i was used zero step run (run 0) but there is no change in potential energy. then I was trying single step simulation but both potential and kinetic energy are chenges. Is there any option to calculate the change in potential energy without change in kinetic energy?
I would greatly appreciate it if you kindly give me some suggestion on this.

Thanking You.

With best regards,

Atanu K Metya

If you displace all the atoms, why should the potential energy change?

I.e. why isn’t the PE translationally invariant?

Steve

Since the potential energy as a function position. If I change the position of all atoms by display_atom command then potential energy should change without any change of the dynamics (i.e KE should constant)

Since the potential energy as a function position. If I change the position of all atoms by display_atom command then potential energy should change without any change of the dynamics (i.e KE should

Not unless you define an additional external potential.

Please provide examples, because discussing in this generality is difficult.

Axel

constant)

I would like say if change the position of all the particles then potential energy should change.

for example, if I consider LJ potential for non-bonding interaction which is function of position( r or x,y,x). Therefore, if I displace the position of all particles (i.e coordinate change x to x+delta_x and other cordinates are fixed) then the potential energy will change. Is it possible such calculation using displace_atom or change_box command in lammps (here I have not change the velocity of the atoms i.e KE is fixed)

I would like say if change the position of all the particles then potential energy should change.
for example, if I consider LJ potential for non-bonding interaction which is function of position( r or x,y,x). Therefore, if I displace the position of all particles (i.e coordinate change x to x+delta_x and other cordinates are fixed) then the potential energy will change. Is it possible such calculation using displace_atom or change_box command in lammps (here I have not change the velocity of the atoms i.e KE is fixed)

No. Pairwise additive potentials are translationally invariant as Steve said. The potential does not use absolute but relative positions. Just think of a harmonic oscillator as the simplest example.

Axel.

PE is not a function of absolute position as Steve has already told you.