I am looking at the interpolation part of pair_eam. In pair_table I think it is a normal cubic spline function as in Numerical Recipe.
However the spline used in pair_eam looks different. What kind of spline function does LAMMPS use? Thanks.
It is a cubic spline, but there are different flavors of them. The
EAM splines are using the same machinery that the original EAM
code, called DYNAMO, used 25 years ago. This was intentional
so that we could match results and get the same level of accuracy.