A gridded ion thruster pulls xenon (or krypton) ions through two closely-spaced perforated grids: the positive screen grid and the negative accel grid, which sets up the accelerating field. One representative aperture pair is shown here (a real grid has thousands, arranged hexagonally).
Perveance: P = J_b / |V_a|^1.5
Sputter yield: Y(E) = Y_max · [(E − E_th) / (E_ref − E_th)]^1.5, 0 ≤ Y ≤ Y_max
Erosion rate: dh/dt ∝ (J_b/J_ref) · Y(|V_a|) · f_propellant · f_impingement(P)
Charge-exchange (CEX) erosion: a small fraction of beam ions collide with slow neutral propellant atoms downstream and swap charge. The resulting slow ion is born with almost no forward velocity but is still inside the accel grid's negative field, so it gets pulled straight back and slams into the downstream edge of the accel aperture — this is what carves the classic "pit-and-groove" erosion pattern that eventually limits grid life (this is the dominant wear mechanism on real thrusters like NSTAR and NEXT).
Perveance and direct impingement: raise the current density or drop the accel voltage and the perveance P = Jb/|Va|1.5 climbs. Past a critical perveance the beamlets can no longer be focused through the aperture and start striking the accel grid webbing directly — the beam visibly fans out and the erosion rate accelerates sharply, shown here as the "impingement risk" readout switching to high.
Sputter yield: impact energy above a threshold Eth knocks grid atoms loose; yield grows with energy above threshold, following a Bohdansky-type curve. Krypton's lower atomic mass gives it a lower sputter yield than xenon at the same energy, at the cost of somewhat worse thrust efficiency — a real propellant trade explored on missions considering krypton for cost reasons.
Grid erosion normally accumulates over thousands of operating hours, so the time acceleration slider fast-forwards the simulated clock — the numeric readouts are physically derived from the formulas above, while the particle animation is an illustrative visualisation of the same process.