Repulsive interaction with adp potential

Hi all,

I’ve been going through the LAMMPS manual and previous mail threads, but I cannot seem to find a solution to my problem.

I’m simulating a side crack in aluminum in LAMMPS. To model the crack, I’ve deleted interactions across the crack plane by using a neigh_modify exclude command. This works at low temperature. However, when I increase the temperature to 300 K, it seems that atoms across the crack plane start bumping into each other, as if they do not see each other. I’ve tried to find a way of keeping repulsive interactions while excluding attractive interactions, but it seems that the pair potential I’m using, which is an Angular Dependent Potential (ADP; it’s an extension of the EAM), does not allow this. Have any of you encountered a similar problem before, and if so, how did you go about solving it?

Thank you very much,

Thom John

With neigh_modify exclude the atoms across the crack plane will not be in each other’s neighbor list, so it is correct that the atoms do not see/feel other. It seems to me you can delete an extra layer or two of atoms on the crack plane (like the attached snapshot), increase your temperature gradually, and/or tabulate your ADP potential so the atoms across the crack plane only has repulsion (this then has to be used with pair hybrid to have a regular ADP and a pure repulsive ADP through pair_style table).