Particle-in-cell plasma discretization

In one dimension a simple linear shape distributes a particle's charge between neighboring nodes. The weights decrease with distance and sum to one within each cell. Multiplying those weights by the represented charge produces nodal charges that add to the original charge. Converting them to a charge density additionally requires the declared nodal control volumes and, in a reduced-dimensional model, the cross-sectional convention. A list of charges in coulombs should never be silently relabeled as a three-dimensional density.

The field seen by a particle is gathered from grid values using an interpolation rule. Matching deposition and gathering consistently can improve symmetry and reduce spurious self-interaction, but energy conservation also depends on the field solver and temporal coupling. Higher-order shapes distribute information over more nodes and filter short spatial scales. They can reduce some noise or aliasing while broadening the represented charge, so a smoother picture is not proof of greater physical resolution at every wavelength.

A moving particle changes the charge assigned to neighboring nodes. To preserve a discrete Gauss constraint under an electromagnetic field update, the current deposition must match that charge change through the solver's discrete divergence. One can conserve the total charge while putting current through the wrong interfaces. A local continuity identity is therefore stronger than checking one global sum. Charge-conserving deposition schemes construct the current from the particle's path and the chosen shape and grid operators.

The electrostatic variant solves a Poisson problem rather than advancing all electromagnetic wave degrees of freedom. Periodic electrostatics requires a compatible net charge: summing the discrete Laplacian gives zero, so a nonneutral source cannot be accommodated merely by choosing a potential gauge. A uniform neutralizing background may be a justified physical approximation in a particular model, but adding it changes the source. Particle loss at an absorbing boundary must likewise be accompanied by the intended boundary charge or current accounting.

Increasing particle number can reduce sampling fluctuations for a specified loading procedure without reducing spatial truncation error. Refining the grid can resolve shorter wavelengths while increasing noise per cell if the particle count is left fixed. Time stepping must resolve the particle motion and the relevant plasma frequencies; an explicit electromagnetic field solver also has its own Courant condition. Conditions involving Debye length or plasma period depend on the algorithm and regime, rather than forming a universal guarantee for every PIC scheme.

Useful studies vary particle count, shape, mesh spacing and time step separately, keeping the represented total charge and physical parameters controlled. Ensembles or loading comparisons address statistical variability; mesh convergence addresses a deterministic approximation. The probability lab shows exact binomial count moments for independent trials. Converting those count moments to normalized fluctuations illustrates a sampling distinction, but the instrument has no particle shapes, self-consistent forces, charge deposition or plasma stability calculation.