Design of an Electrostatic Energy Separator for the Isis Rfq Test Stand

It is planned to replace the 665 keV Cockroft Walton set of the ISIS spallation neutron source in the UK with a four rod RFQ. Before this replacement is made, the performance of the RFQ will be extensively assessed on a test stand. One of the quantities we hope to measure is the proportion of accelerated ( trapped ) beam at the exit of the RFQ for various H ion source currents and rod voltages. In order to do this it is necessary to separate the untrapped beam which is largely at the ion source extraction energy of 35 keV from the trapped beam which has an energy centred on 665 keV. This paper describes the design process for an instrument to perform this separation using parallel electrostatic plates to bend away the low energy beam while allowing all the higher energy beam to pass through, therefore measuring directly the fraction of trapped beam, and by implication the fraction of untrapped beam. 1 THE DESIGN PROCESS The aim of the investigation was to find a method of removing the low energy component from the beam while retaining the high energy component for long enough that its current could be measured. Initially, extensive studies were performed to discover if it was possible to use the chromaticity of a series of quadrupoles to over focus the low energy beam and cause it to hit the beam pipe while retaining the high energy beam. ( For the purposes of this paper, ‘high energy’ is defined as 0.665 ± 0.1 MeV and low energy is anything less than this ). Solutions were found, but due to the highly divergent beam from the RFQ the beam size grows large very quickly, making it necessary to have very large bore quadrupoles in order not to lose any high energy beam. The size, weight and cost of these quadrupoles caused us to look for another way of separating the two energy groups. The fact that the untrapped beam is at such low energy and that there is a wide gap between the two energy groups caused us to think of using an electrostatic method of separating the groups. Studies were performed to investigate the feasibility of using two parallel high voltage plates of opposite polarity to deflect the low energy beam more than the high energy beam. These were done by tracking a ‘large’ distribution of quasi-particles which had been generated by an RFQ simulation program written by A. Letchford. The final optimisation was performed using an initial distribution of 95351 quasi-particles, of which 89349 were ‘high energy’ and 6002 of lower energy, which corresponds to fractions of 93.7% and 6.3% respectively. Particle tracking was performed with Parmila [1] throughout, and various Fortran routines were written to perform tasks such as counting the proportions in the two energy groups at various stages of the device, calculating the trajectories of all the particles as they pass through the electric field and calculating the distribution of particles hits on the plate which the particles are bent towards. The parallel plate device was to be situated in the second half of a diagnostics box of outside length 430 mm, the first half of which will contain an emittance scanner. The entrance port of the box ends 163 mm after the end of the RFQ rods. At this point the low energy beam is already being lost on the 70 mm diameter beam pipe, and all components of the beam are diverging. At the entrance port to the box, the high energy component is quite round with a radius in both transverse directions of about 8 mm. By the middle of the box, e.g. after 253 mm from the port, the beam looks transversely like Figure 1. Fig. 1 Transverse distribution of all remaining particles at position of beginning of parallel plates, if there were no aperture plate The radius in the x direction of the high energy beam only at the same point in the box is about 1.5 cm. This means that if nothing else were done the low energy beam would have to be bent ~( 1.5 + 8.0 ) cm more than the high energy beam to be spatially separated. For this reason an aperture plate was added to the middle of the box, 217 mm after the entrance port, to crop the low energy component as tightly around the high energy 0-7803-5573-3/99/$10.00@1999 IEEE. 2208 Proceedings of the 1999 Particle Accelerator Conference, New York, 1999