[lammps-users] Constant potential energy using fix atom/swap


I am trying to model a high entropy alloy using LAMMPS. In particular I am just characterizing the alloy potential energy as a function of the main elements concentration. I am also using the hybrid Monte Carlo – Molecular Dynamics method (fix atom/swap) in order to get a more stable mixed alloy. However I noticed that the atom/swap command does not improve the system pe with all the interatomic potentials. More specifically some potentials show a decrease in the system pe with the increase of number of swaps, while other interatomic potentials show a constant pe behavior. As the initial configuration was yet the most stable possible.

(I tried using the vey same model with also same alloy elements and even equal random seeds, changing just the potential library)

Also, these potentials show a very linear behavior of pe with alloy concentration while others show a pe minimum in the areas of nearly equimolar concentrations.

Is this normal? I was expecting to see a change in the pe with every potential while swapping atoms.

Thank you.



I’m CCing Aidan in case he wants to comment.
My guess is that this is simply a function of the pair styles you are using
and their parameterization. I.e. not a LAMMPS issue per se.


Hi Steve,

Thank you for your answer. Do you think I should change the pair style? Or another type of interatomic potential would be necessary?