Suppose I have particles tagged type-1 and type-2, and for pairwise computations, I want the code to only execute 1–1, 1–2 computations, but skip 2–2, and retain all type-2’s locations unchanged. I am not sure if LAMMPS supports the way that I described here.
The reason is that, I want to avoid using the strategy as to first have all pairs computed, and then fix the force as zero for type-2 particles. This is because the type-2 particles may account for 80% ~ 90% of all (tens of millions of) particles, and the post-fix strategy might be still very expensive.
Thanks in advance for sharing any experience here.