[lammps-users] meam/c pair_coeff

Dear developers and users,
I am using windows version LAMMPS 29/10/2020 (last stable MPI ).

I wish to use meam/c pair style with AlSiCuMgFe.meam potential in conjunction with library.meam

you need to very carefully read the pair style meam/c documentation. the handling of the potential parameter files is tricky and needs to be done in the exactly correct way.
it is easy to get wrong parameters. the situation is so complicated because the file formats are meant to match legacy codes, not LAMMPS.

the basic issue is that the second parameter file augments the settings of the first, the library file. in that second file the atom types are indexed by numbers starting from 1 (fortran conventions). thus you first “extract” generic parameters from the library file and those have to be exactly as many elements and in the exact same order as expected by the second file. you are using a second file with settings for 5 elements, but are extracting only 2 from the library. that is incorrect and therefore the error. there can be no checks on the order because of the numerical indexing, so you have to be ultra careful.


Hi Axel,
Thanks a lot. I will check the documentation and the parameter file more carefully.
I only try with the LAMMPS SiC example where the situation is more simple, with only 2 element parameter file.
Best regards