On Boundary Conditions in the sub-mesh interaction of the Particle-Particle-Particle-Mesh Algorithm

Abstract: The Particle-Particle-Particle-Mesh algorithm elegantly extends the standard Particle-In-Cell scheme by direct summation of interaction that happens over distances below or around mesh size. Generally, this allows for a more accurate description of Coulomb interactions and improves precision in the prediction of key observables. Nevertheless, most implementations neglect electrostatic boundary conditions for the short-ranged interaction that are directly summed. In this paper a variational description of the Particle-Particle-Particle-Mesh algorithm will be developed for the first time and subsequently used to derive temporally and spatially discrete equations of motion. We show that the error committed by neglecting boundary conditions on the short scale is directly tied to the discretization error induced by the computational grid.