Generic Question about hybrid potential

LAMMPS LAMMPS Mailing List Mirror
1

Hi there,
I am having a general question regarding using hybrid potential by combining LJ and Tersoff scheme. In accordance with the literature, parameters are used as per the following scheme:

There are 4 type of atoms in the simulation: H O C Si as 1,2,3 and 4.

pair_style hybrid lj/cut 9 tersoff

pair_coeff * * lj/cut 0.000 0.000
pair_coeff * * tersoff albe.tersoff NULL NULL C Si
pair_coeff 2 2 lj/cut 0.1533 3.166 10 # O-O
pair_coeff 2 3 lj/cut 0.55 3.6 7 #CO
pair_coeff 2 4 lj/cut 6.2 1.3807 6 # SiO2

Energy minimization however reveals following values which are evidently erroneous. Any alert on the mistake in the numerical parameter or the way in which above arguments are written will be sincerely appreciated.

Memory usage per processor = 4.59562 Mbytes
Step Temp E_pair E_mol TotEng Press
0 268.35702 6.4486463e+16 0 6.4486463e+16 1.1154065e+17
100 268.35702 26956612 0 26960871 41128441
Loop time of 5.76321 on 512 procs for 100 steps with 122770 atoms

2

Hi there,
I am having a general question regarding using hybrid potential by combining
LJ and Tersoff scheme. In accordance with the literature, parameters are
used as per the following scheme:

There are 4 type of atoms in the simulation: H O C Si as 1,2,3 and 4.

pair_style hybrid lj/cut 9 tersoff
pair_coeff * * lj/cut 0.000 0.000
pair_coeff * * tersoff albe.tersoff NULL NULL C Si
pair_coeff 2 2 lj/cut 0.1533 3.166 10 # O-O
pair_coeff 2 3 lj/cut 0.55 3.6 7 #CO
pair_coeff 2 4 lj/cut 6.2 1.3807 6 # SiO2

Energy minimization however reveals following values which are evidently
erroneous. Any alert on the mistake in the numerical parameter or the way in
which above arguments are written will be sincerely appreciated.

can be anything. bad geometry, bad settings, bad choice of parameters,
wrong units.
how should anybody be able to tell with practically no useful information?