Results are presented for electron beam transport experiments in a 1-m-long straight section consisting of a solenoid and five short printed-circuit quadrupoles. A linear computer code for rms envelope matching, SPOT, is used for channel design, while final simulations with more realistic elements are obtained with a 2 1 2 D version of WARP, a particle-in-cell code. Reasonable agreement is found between calculations and the effective beam envelope obtained from pictures of the beam on a movable phosphor screen. The results validate, within experimental errors, the use of short magnetic quadrupoles for the transport of space-charge dominated beams. The straight section constitutes the prototype matching section for an electron recirculator to be built at the University of Maryland. [S1098-4402(98)

Two Panofsky quadrupoles with rectangular aperture and fast rise time are proposed for the injection area in the University of Maryland electron ring project (UMER) [1,2,3]. The theoretical current distribution needed on the surface of a rectangular aperture to produce a quadrupole field is derived [4]. The conductor location is determined mostly by the theoretical current distribution, with some free factors to optimise the field. The design is based on the linearity of longitudinal integrated field. Each quadrant of the Panofsky quadrupole consists of 10 loops of conductors to minimize inductance while retaining the field quality. A 2:1 scaled model has been made and the magnetic field was measured. Two 1:1 models have been made to measure inductance, mutual inductance and the rise time. Simulations of beam propagation with Panofsky quadrupoles demonstrated the linearity of the design.

No abstract

Abstract.Numerical simulation using particle-in-cell codes is a powerful tool in understanding the nonlinear dynamics of space-charge-dominated beams.The University of Maryland Electron Ring (UMER) will explore the transport of beams with intensity previously inaccessible to circular machines. The ring will also function as a testbed for accelerator codes. Applications such as heavy ion fusion and colliders require the preservation of beam quality during transport over large distances. This need for low beam emittance and small particle losses constrains the design and fabrication of the lattice and the injector. Furthermore, the non-zero energy spread leads to dispersion in the circular lattice. Simulations using the WARP code address these issues: the magnets, including the fringe field nonlinearities, are modeled realistically; dispersion matching is attempted; and effects of lattice and beam errors are examined. The simulations aid in understanding experimental results, such as the transverse density waves observed in the injector.

The end effect in a bunched beam is caused by the space-charge forces, which accelerate particles at the bunch head and decelerate particles at the bunch tail. It occurs in high-current linear accelerators and rings [1]. In the University of Maryland Electron Ring (UMER) project [2,3,4], the energy of particles at the very ends of a rectangular bunch is 15% different from that of the main part of the bunch. Assuming a negligible transverse space charge effect at the edge of the bunch, one dimensional single particle calculation is performed by a matrix formalism [6] to estimate the ratio of particle loss due to the end effect. Two possible cases are examined: with no induction gap, and with three induction gaps. If three induction modules are used, for a small number of turns, the particle loss is less than 0.15%, which is small enough to be acceptable. Simulation by code CIRCE [8] is in progress.