[lammps-users] WCA potential

Dear Lammps Users,

I am have quick question regarding implementing WCA potential in lammps. For those who are not aware of it, It is essentially LJ 12-6 potential with additional epsilon added to it. It is generally used as a repulsive interactions with a cutoff at sigma*2^(1/6) (the distance at which interaction is negative epsilon). So basically the total interaction becomes zero at this distance.

I tried to look for various pair styles but I could find something like it. If I use lj/cut or lj/expand (for particles of different size) with the cutoff at sigma2^(1/6), I get the interaction energy as “-epsilon”. One option is to use pair_modify shift yes to move the interaction to zero. However, the problem with it is that it doesn’t modify the forces. So, for the distances between (sigma and sigma2^(1/6), the particles would feel attractive interaction).

This potential is often used to repulsive interactions between unlike atoms. It may be possible that I might be overlooking a particular pair_style.

I will highly appreciate if anyone could help me out here.

Regards and thanks in advance.


Why do you think that particles will feel an attractive interaction? The
minimum of the Lennard-Jones potential is at 2^(1/6), so if you cut (and
shift) the potential you will keep only the repulsive part. Even though
the Lennard-Jones potential changes its sign at sigma, this is not the
case for the derived force.


Hi Vikas,

I suppose that you just need to use pair_style lj/cut with the shift flag set to be yes. For the WCA potential between two types i and j, you then set the cutoff parameter to be 2^(1/6) = 1.122.

pair_style lj/cut 2.5
pair_modify shift yes

pair_coeff type_i type_j 1.0 1.0 1.122
# other pairs....

Hope this help,


Quoting Vikas Varshney <[email protected]...>:

Dear Lutz, Joanne,
I think I was too quick to make conclusions from the equation. Thanks to both of you for clarification. You are both right. I should have thought a bit more before posting. :)…


LJ, cutoff at 2^(1/6), is exactly WCA. The shift by
epsilon affects bookkeeping for energy, but does not
affect forces or dynamics.