I would like to fill a container (walls/reflect) with squishy spherical particles and then compress the container to see how heat is generated during compression. My first thought is to use a LJ potential with the cutoff distance set to form a purely repulsive particle but that seems a little clunky. Is there a better model that I should be looking for?
I don’t think there is any model in LAMMPS that can directly serve the purpose that you are looking for.
The kind of heat that you would want to study would be generated from interparticle friction or from deformation of the particle.
Neither is captured by potentials like LJ and other point particle potentials. Perhaps some of the DPD models might come closer.
At the very least you would need to represent each granular particle by a group of explicit particles (but how many? possibly a lot) with some interaction model, e.g. harmonic springs, where the internal kinetic energy would then be modeled explicitly and could be studied. or you would have to devise a new granular model that explicitly considers these properties and updates them explicitly (that would mean to write a new atom style, pair style, and time integration fix plus a matching wall fix modification).
Thanks for the suggestions. I ended up building granular like particles with beads and springs.
Note that if you mean granular particles using the
granular pair styles in LAMMPS, then these typically
have dissipative terms, which means the temperature
of the particles will rapidly go to zero. You might be
able to set the dissipative coeffs to (nearly) zero and
get something closer to what you are seeking.