Is it necessary to define pair_coeff for all atom pairs in the Soft Potential?


The following is my input file for using the Soft Potential:

units real
boundary p p p
newton on

atom_style full
pair_style soft 14.0
bond_style harmonic
angle_style harmonic
dihedral_style multi/harmonic
improper_style cvff

read_data 20per8000tol.dat

neighbor 2.0 bin
neigh_modify delay 5

fix fixv all rigid/nvt molecule temp 0.1 0.1 100.0
dump atomdump all custom 10000 20per8000tolsoft.lammpstrj id type mol x y z ix iy iz

restart 1000000 20per8000tolsoft.restart

timestep 0.05
velocity all create 295.0 807006 dist gaussian mom yes rot yes
run 1000000

I have read in the manual that the values for “A” and “cutoff” must be defined for each pair of atoms by the pair_coeff command. I have 9 atom types and so the values for these atoms are located in my data file and look like:

Pair Coeffs

1 0.07 3.55
2 0.07 3.55
3 0.07 3.55
4 0.03 2.42
5 0.07 3.55
6 0.07 3.55
7 0.066 3.50
8 0.03 2.42
9 0.03 2.50

So these are the A and r_{c} values for pairs of identical atoms, but I haven’t defined values for pairs of different atoms (eg. 1 with 2).

Even though the value of A and the cutoff are not defined for all atom pairs - the simulation still works fine. I haven’t included any mixing of the coefficients in my scripts so I’m really not sure what values the pairs of heteronuclear atoms are taking?

Thanks, Simon

The doc page for every pair style explains how
it handles mixing (or not).