The current LJ implementation for a 5-atom cluster gives a minimum energy of -9.10, which fits experimental values. When using LJ with the ase calculator, however, a minimum value of 9.049 is obtained. Mathematically speaking, this should not occur unless ASE is doing extra manipulation or our ase calculator is set up wrong.
The ase calculator has a cutoff set to 15 units. I don't know far these units extend in our simulation, so maybe increasing them will fix the issue. To solve this issue, you'd want to both find the root cause of the discrepancy and suggest a modification that would fix it.
The current LJ implementation for a 5-atom cluster gives a minimum energy of -9.10, which fits experimental values. When using LJ with the ase calculator, however, a minimum value of 9.049 is obtained. Mathematically speaking, this should not occur unless ASE is doing extra manipulation or our ase calculator is set up wrong.
The ase calculator has a cutoff set to 15 units. I don't know far these units extend in our simulation, so maybe increasing them will fix the issue. To solve this issue, you'd want to both find the root cause of the discrepancy and suggest a modification that would fix it.