Hello,
I am planning to simulate interaction between a methyl radical (CH3) and a carbon surface to "see" what happens (sticking, reflexion...).
I am using REBO potential to model interaction between C and H atoms.
I have two questions concerning the modelling of interactions between atoms of the CH3 radical (and actually also of the carbone substrate) using LAMMPS.
First I was wondering if I should use the bond_style command or not. Is the pair_style command not sufficient for modelling interaction between atoms ?
Second, what would be the difference whether the bond_style command is set or not ?
Thanks
Laurent.
Hello,
I am planning to simulate interaction between a methyl radical (CH3) and
a carbon surface to "see" what happens (sticking, reflexion...).
I am using REBO potential to model interaction between C and H atoms.
I have two questions concerning the modelling of interactions between
atoms of the CH3 radical (and actually also of the carbone substrate)
using LAMMPS.
First I was wondering if I should use the bond_style command or not. Is
the pair_style command not sufficient for modelling interaction between
atoms ?
no. in LAMMPS bond order potentials like REBO compute bond
contributions on the fly as part of the non-bonded calculations.
Second, what would be the difference whether the bond_style command is
set or not ?
if you use a bond style, you would define *additional* and
*fixed* bonded interactions. this will automatically lead to
pairs being removed from the neighbor list due to the default
special_bonds settings and thus completely mess up the
bond order potential (unless you set special_bonds to lj/cut 1.0 1.0 1.0)
any other special bonds setting will lead to inconsistent
results for the bond order potential.
let us know if you have any additional questions about this.
axel.