Problem modeling LJ potential with two unique sigmas (sigma12 and sigma6)

Using version 22Feb13.

I am working on thermodynamic integration of a system where some of
the particles in my system are changing from type 1 to type 2 which
uses a different set of LJ parameters: (epsilon and sigma). The
thermodynamic integration progresses across a series of simulations
according to a third parameter (lambda). When I introduce this lambda
parameter into my LJ potentials, my sigma12 and sigma6 become unique
functions of lambda.

Therefore, my question is this. In LAMMPS, is there a way to model an
LJ potential with three parameters: (epsilon, sigma12 and sigma6)? I
have looked and haven't had any luck short of coding it in myself, and
I thought I would ask here first before going with that option.

Thank you,

well, you can always create a table of potentials for the
potential you want and use the pair_style table command.

i don't know if what you need exists specifically, but of course
you can try coding it in yourself. i don't think it
would be extremely difficult considering you have already
the lj code overthere. but maybe the table is good enough


2013/7/1 Travis Russell <[email protected]>:

You could use pair table, but it won’t help with
thermodynamic integration if you want to
change the values of the 2 sigmas over time.

What you describe really is no longer a LJ potential.
You many need to code it up as a 3-param pair
style. Then it is pretty easy to get it to
work with fix adapt to vary the params on the fly.