Radiation Damage Question


I am having an interesting error that I can normally figure out, but this time I cannot.

I am getting the error:

Neighbor list overflow, boost neigh_modify one (…/neigh_half_bin.cpp:129)

It does not happen at the beginning of my simulation, during equilibration, so I know that my initial structure is fine.

When I invoke a 5 keV PKA with a variable time step of 0.005 Angstroms (meaning the time step is such that no atom moves more than that distance in a timestep), it seems to work fine for about 20,000 steps. Once the simulation time is at ~1.5 ps, meaning the peak number of defects has been reached already, and the system is annealing, I get the error. How can this be?

Assuming that 2 atoms moved directly towards one another, that would only be 0.01 Angstroms per timestep. I thought maybe the atoms were getting to close to one another, but if this did not happen when the system had the higher energy PKA, why would it happen now?

Any ideas why this is happening? It seems that the distance I have chosen should work, but for some reason it does not.