Does Gay-Berne potential has any bearing on size/visualization of particles?

LAMMPS LAMMPS General Discussion
1

Hi there! Would really appreciate your insight here:
I am currently running a system with ellipsoidal particles in 2D, with shape 3.0 1.0 1.0. When I try to visualize the result in OVITO, I am seeing a system that is less dense (ellipsoids that are much larger) than the density packing fraction that I am trying to calculate. (I already checked out all the visual representation tools in OVITO and made sure that the standard radius of the particle is set to 1, and nothing else is magnified/minimized). I was wondering if the choice and settings of the potential in LAMMPS have any bearing on this? I am using a GB potential and was wondering if this impacts the final density of the system in any way, as its energy curve values vary based on the anisotropy coefficient.
Obviously, different configuration/alignment of the particles (side-to-side, end-to-end etc) will lead to different potential cutoff, but could changing the GB potential parameters be directly impacting the size of the particles? For instance, setting N = 10000 particles for the box of 300x300 results in 0.25 number density and 0.59 (n. density * 3pi/4) in area packing fraction density, but the system often appears to be often less dense than it should be for this packing fraction density. I have not run into the same problems with spheres before where I used LJ potential.

Thank you!

2

With LJ particles the sigma parameter has no bearing on the diameter of the spheres created by a visualization propgram. Usually, they don’t have that information. Only the position. If this looks reasonable in the visualization, than this is because usually the sigma parameter is in the same neighborhood of 3-4 angstrom.

I recently added a feature to LAMMPS-GUI to render particle diameters using dump image according to their LJ sigma (or Morse r0) and that required reading the potential parameters and setting atom diameters for each atom type explicitly.