[lammps-users] problem with l-j potential in lammps

Dear all,
I want to use the l-j potential of Ar but I’m not sure how to set the parameters. For example, the lattice constant of Ar is 5.256A, sigma is 3.4A, epsilon is 0.0104eV, cutoff is 2.5sigma. Then how to set the lattice, pair_style and pair_coeff?

thanks,
Qingjie

It depends on what units you want to run in - see the units command.
Normally, you would run this in reduced LJ units, in which case,
you set the reduced density and sigma = epsilon = 1. You'll need
to read a textbood about LJ reduced units. However, the numbers
you give are in what LAMMPS calls "metal" units (eV, Angs, etc). So
you could run that way and use the lattice command to specify a lattice
constant of 5.256.

Steve

If I am using the “metal” unit. I set the pair_style and pair_coeff in the following two ways for Argon simulation:

pair_style lj/cut 8.5 # Here 8.5 is equal to 2.5*sigma
pair_coeff 1 1 0.0104 3.40

or

pair_style lj/cut 2.5
pair_coeff 1 1 0.0104 3.40

Which one is correct?

Thanks!

Dongshan

In metal units the cutoff is in Angstroms. Do you want a 2.5 Angs
or 8.5 Angs cutoff? The answer to that question is your answer.

Steve

I want to use 8.5 angstroms for the cutoff, which is the 2.5 times of the sigma (3.4 angstroms).

So there is an error on choosing the cutoff of the lj/cut potential for Argon in the previous mail:

http://sourceforge.net/mailarchive/message.php?msg_id=76be2dbf1002112152o52d86c0ei6a44b6420671ef18%40mail.gmail.com

Dongshan