This work describes the implementation of the particle-in-cell simulation method with Monte-Carlo Langevin and hard sphere ion-neutral collisions in our two dimensional Cartesian and cylindrical coordinate system codes. The particle-in-cell method efficiently and accurately solves complex dynamics involving many interacting particles. In our case 'many' means simulation runs with thousands to tens of thousands of Coulombically interacting particles on a workstation. In addition, the image charge induced on the device electrodes, which is required for self-consistency in order to make the electrodes equipotentials, is incorporated with this method. For an azimuthally symmetric confinement geometry and charge distribution, results using the two dimensional cylindrical (rz) particle simulator are identical to full three-dimensional simulations, by symmetry. These codes are applied to a number of examples which are relevant to trapped ion mass spectrometry in order to demonstrate their utility. Computer experiments are carried out to study the maximum number of ions which can be confined and equilibrium cloud shapes in cylindrical ICR traps, ion cloud collisional cooling in a combined trap, merger of two off-axis charge columns, Kelvin-Helmholtz instability in ring-shaped initial distributions, Vortex crystallization, and image charge detection of coherent cyclotron motion. (C) 1997 Elsevier Science B.V.