If you have a given lattice (say fcc) in which you specify the simulation cell size as exactly the upper and lower limits in all three directions. If you use s s s as boundary condition, you dont have any problem because if any atom crosses the specified limit, the boundary condition will take care of it. But if you use f f f as bound cond’n then atom will be lost. In that case you can specify higher bounds than your initial structure to prevent losing atoms. My question is if you do similar analysis in those two cases, do they give same result?