Emittance growth as mesh artefact

The space charge beam spreading calculations of the laser ion source group of CERN [K. Hanke, et al., Rev. Sci. Instrum. 73 (2002) 783] has been recomputed with IGUN. Instead of only looking to the total value of beam spreading, we also investigated the behaviour of the rms-emittance and found the following unexpected results: 1)Even zero emittance beams show considerable emittance growth.2)The emittance growth of the central part depends on both the number of meshes used and the number of trajectories used.3)The most prominent emittance growth occurs at the beam boundary, which is independent of the number of trajectories used and only depends on the number of meshes inside the beam. The more meshes used, the less this emittance growth will be.Generally, emittance growth is associated with the non-linear electric field components, integrated up along the particles trajectories. By using a fractional emittance analysis, we found two different contributions to emittance growth, which might explain the observed results:(a)the deposition of space charge and the interpolation of fields inside the meshes produces errors and provides “noise” to the particle ray tracing algorithm.(b)at the beam boundary, the filling of the space charge map depends very much on the relative location of the beam boundary to the discrete mesh structure.This mesh artefact is inherent to all programs that need meshes for the evaluation of the influence of space charge on the potential distribution. When calculating a “real” beam with given emittance, it is good advice to compare this calculation with the emittance growth, obtained for a beam of zero emittance. Furthermore, a variation of the number of trajectories and the number of meshes will allow to extrapolate to the unaccessible use of an infinite number of trajectories and meshes.