in general, please understand, that this mailing list is not a debugging service. everybody has to debug their own inputs.
at best, people can provide you with some general advice:
- can you run your system without adding the coating and the indenter? or only substrate and coating?
- does LAMMPS print any warnings? (and what version of LAMMPS are you using, anyway?)
- have you visualized your system? where does the massively accelerated escaping atom originate?
two things stand out when looking at your input:
- your hybrid potential input doesn't make sense. the Al and Ni only EAM pair_coeff entries are meaningless and will be overwritten by the following mixed potential file. that one is the only correct way to handle this anyway, since EAM is a manybody potential and not pairwise additive, which your input seems to assume
- it looks as if you have atoms thermostatted multiple times. that is guaranteed to produce bogus trajectories.