FeTiC 2n MEAM potential implementation

Dear Lammps users,

I’m trying to generate the potential files for the 2nn meam for the FeTiC
system as described in Kim, H.-K., Jung, W.-S., & Lee, B.-J. (2009). Acta
Materialia, 57(11), 3140–3147. doi:10.1016/j.actamat.2009.03.019

I managed to generate a first version of the “MEAM library” and the “MEAM
parameter” files (FeTiClibrary.meam and FeTiC.meam) according to LAMMPS format
and following the instructions given in https://cmse.postech.ac.kr/home_2nnmeam

I started doing some first trials for single-element systems to verify if the
implementation is correct. In the case of pure Fe and C I am able to retrieve
the correct lattice constants and cohesive energies. (files in.C_lattice and
in.Fe_lattice).

In the case of Ti (files in.Ti_lattice), I get NAN values for the energy and
the pressure:
Step Temp E_pair E_mol TotEng Press Volume
0 0 nan 0 nan nan 73353.3794046816

By playing a bit with the parameter file, i noticed that if I turn off the 2nn
interaction and comment the definition of Cmin, the simulation runs fine (BUT
without 2nn interaction):
Step Temp E_pair E_mol TotEng Press Volume
0 0 -19441.0031366012 0 -19441.0031366012 -41119.7811533008 73353.3794046816

Do you have any idea how to correct this behaviour?

I am running LAMMPS (6 Jul 2015-ICMS)

Regards,
Sebastian ECHEVERRI RESTREPO

FeTiC.meam (2.18 KB)

FeTiClibrary.meam (536 Bytes)

in.C_lattice (499 Bytes)

in.Fe_lattice (500 Bytes)

in.Ti_lattice (496 Bytes)

I think you’ll have to look inside the MEAM library files

(Fortran, in lib/meam) and determine where the NAN is

occuring due to some numerical operation. Valgrind

can probably help with that.

Steve

Dear Steve, lammps-users,

It seems that the error is in the ICMS version of lammps

I built a new version of lammps from source and now it works fine

Best Regards,
Sebastián