EMMA—The world’s first non-scaling FFAG

https://doi.org/10.1016/j.nima.2010.08.109Get rights and content

Abstract

Due to the combination of fixed magnetic field operation with strong focusing, non-scaling FFAGs have a significant potential for future particle accelerator applications. However, this technology has a number of unique features, which must be fully studied before this potential can be realised. To do this, a proof-of-principle non-scaling FFAG, called EMMA – Electron Model for Many Applications – has been constructed at the STFC Daresbury Laboratory in the UK. It has been designed by an international collaboration of accelerator scientists and engineers. It will demonstrate the principle of non-scaling FFAGs and be used to study the features of this type of accelerator in detail.

Introduction

Over the last 10 years, considerable worldwide interest has developed around Fixed Field Alternating Gradient (FFAG) particle accelerators. As the name implies, these are azimuthally varying field (AVF) accelerators employing both fixed magnetic field strengths during the acceleration cycle and alternating gradient, or strong, focusing [1]. Although originally invented in the 1950s [2], little further development of these machines occurred until the late 1990s. At that time, the benefits of their large beam acceptance and high repetition rate were noted, initially for the acceleration of muons in accelerators based on muon storage rings, in particular the Neutrino Factory [3], but later for a number of other applications [1].

Until recently, the focus has been on scaling FFAGs, in which the azimuthal magnetic field profile is kept the same at all radii, so that the orbit shape and optics are similar and betatron tunes remain constant during the acceleration cycle. A number of these machines have been constructed and operated in Japan, in particular the so-called Proof-of-Principle FFAG [4], a 150 MeV proton FFAG [5], a 10 MeV proton FFAG to study Boron Neutron Capture Therapy [6], a prototype for ADSR [7] and the PRISM storage ring for the phase space rotation of muons [8].

Non-scaling FFAGs were invented at the end of the 1990s [9], again principally for the acceleration of muons in a Neutrino Factory. In these machines, orbits at different momenta have different shapes, i.e. do not scale, and in a single magnetic element the curvature may reverse with increasing energy, resulting in large betatron tune changes during acceleration. While this brings the potential problem of multiple resonance crossings, it also brings a number of advantages, as discussed in Section 1.1.

The scaling nature of scaling FFAGs is achieved by employing a non-linear variation of the average magnetic field strength with radius, Bavrk, where k is the average magnetic field index, defined byk(r)rBavdBavdr

In non-scaling FFAGs, as originally proposed, a linear variation of magnetic field is employed. This brings a number of advantages over both scaling FFAGs and other AVF accelerators such as cyclotrons. The linear field variation results in a large dynamic aperture, allowing the acceleration of large emittance beams. Furthermore, with the appropriate selection of parameters, it creates a parabolic variation in orbit length with energy. At the minimum of the parabola, the momentum compaction [10] approaches zero and remains very small for acceleration over a factor of 3 or more in momentum. This means that the orbit excursion in non-scaling FFAGs can be much smaller than in other fixed field accelerators.

For relativistic particles, the parabolic variation in orbit length brings another advantage: the time of flight of the beam around the accelerator is also approximately parabolic. For acceleration of up to approximately three in momentum and with a high enough RF voltage, this makes it possible to use a fixed frequency RF system. Although the phase of the beam varies continuously with respect to that of the RF system, the beam is accelerated over the entire energy range, so-called serpentine acceleration [11], [12], [13], [14], [15] (see Fig. 1). This makes it possible to accelerate the beam very rapidly.

However, these advantages come with a number of potential problems. In particular, the very small momentum compaction, approaching zero for a certain region of phase space, and serpentine acceleration are unique to this type of accelerator. Further, the constant magnetic field gradient means that the betatron tunes cross many integer values duration the acceleration cycle, though it is believed that with very fast acceleration, these resonance conditions will be crossed so quickly that there will be no significant deterioration in beam quality [16].

Although non-scaling FFAGs have several advantages for particle acceleration, experimental verification is required before their potential can be further exploited. In addition, the unique features of these machines need to be studied in detail and compared with tracking simulations. To do this requires the construction of a proof-of-principle machine. To minimise costs, while maintaining the ability to study all of these features, it has been decided to build a relativistic electron non-scaling FFAG. Further, since the concept was first proposed as an accelerator of a muon beam in a Neutrino Factory, some of the parameters of this machine have been determined by scaling down those of the muon FFAG [17].

The proof-of-principle machine is called EMMA, the Electron Model for Many Applications. As explained in more detail in Section 2, it is approximately 16.5 m in circumference and accelerates electrons from 10 to 20 MeV. It has been constructed at the STFC Daresbury Laboratory in Cheshire, in the UK, and utilises an existing accelerator called ALICE, Accelerators and Lasers In Combined Experiments [18], as the injector.

There are a number of stages of experimentation foreseen with EMMA. The first stage is the basic proof-of-principle stage: operating the accelerator and measuring the betatron tunes and time-of-flight as a function of energy and machine parameters. The second stage will focus more on the feasibility of a non-scaling FFAG for muon acceleration, verifying the large transverse acceptance, serpentine acceleration and fast resonance crossing, again as a function of machine parameters. Although the transverse beam emittance of the electron beam will not be as large as that of a muon beam, phase space scanning with the beam will be used to probe the full acceptance. The third stage will attempt more detailed studies of resonance crossing by utilising the ability of ALICE to deliver an electron beam at almost any energy between 10 and 20 MeV. Each experimental measurement will be compared quantitatively with models from particle tracking simulations and will be used to benchmark the codes. This is essential because the detailed design of non-scaling FFAGs for other applications relies on the models.

Although the machine parameters of EMMA are closely tuned to model a muon FFAG, there are several upgrade paths planned for the future to explore features of non-scaling FFAGs more directly related to other applications. Installation of a low frequency RF system, with which an electron beam could be accelerated with much slower acceleration rate, would enable a study of non-scaling FFAGs such as a proton or light ion driver for particle therapy [19] or for an accelerator driven subcritical reactor [20]. Further, adding non-linearity to the lattice would enable investigations of the reduction in betatron tune changes and the reduction in the dependence of the time of flight with transverse amplitude, while at the same time exploring the loss in dynamic aperture.

This paper will describe the design and construction of the EMMA accelerator. It is laid out as follows. Section 2 will outline the machine lattices used to specify EMMA, the specifications themselves and the resulting layout of the accelerator. Section 3 will describe the injection system and Section 4 the main components of the EMMA ring itself. As an experimental machine, it is very important to make precise and detailed measurements of the beam accelerated by EMMA. The diagnostic devices to do this will be described in Section 5 and the method by which the measurements are made available to the operators in Section 6. Conclusions will be drawn in Section 7.

Section snippets

Specifications

The aim of the EMMA is to explore, as fully as possible, the beam dynamics in a linear non-scaling FFAG and the specifications for the accelerator have largely been determined by this. To deliver a sufficient kinetic energy range, while minimising costs, EMMA accelerates an electron beam from 10 to 20 MeV. It uses a fixed- and high-frequency RF system, where “high” means that the variation over the energy range of the ring of the time of flight per turn is a significant fraction of the RF period

ALICE

The specifications for EMMA create challenging requirements for the accelerator used as the injector. In particular, this must be able to deliver a beam at any energy between 10 and 20 MeV. The beam must consist of a single bunch, with a normalised transverse emittance and bunch length, around 3 π mm mrad and 12.5 ps, respectively, sufficiently small for it to act as a point probe of the EMMA acceptance. The bunch charge should be as large as possible to avoid compromising the resolution of the

Injection and extraction system

The injection system must be able to inject single bunches at any energy from 10 to 20 MeV in to the full acceptance of the ring, for each of the eight EMMA lattices. Further, the space available to do this is extremely limited. In the initial design stages, it was studied whether it would be possible to inject through the magnets in the preceding cells. However, as these magnets have to move to provide the independent variation of their dipole and quadrupole strengths, this was considered

Overview

As EMMA is a purely experimental machine, it is very important that it has sufficient diagnostic devices to make detailed measurements of the beam throughout the acceleration cycle. These are located in the injection line, to measure beam properties on injection into EMMA, and the ring itself. Destructive devices are located in an external diagnostics beamline, which is described in Section 5.2. The extraction system into this is designed to allow extraction at all energies, so that

Control system

EMMA is operated from a central control room that is shared with the ALICE accelerator. All important aspects of the operation, monitoring and characterisation of the facility are managed by a distributed control system that provides a wide range of standard graphical interface tools as well as the ability to interface to more complex data processing and modelling tools. It is designed to meet a number of important operational requirements. These include:

  • Compatibility with the ALICE control

Conclusions

Non-scaling FFAGs have demonstrated a significant potential for a variety of applications. This is due to their mixture of fixed magnetic fields during acceleration and strong focusing. They are currently being studied for the acceleration of proton and light ion beams for the treatment of cancer, for accelerator driven systems and neutron and muon production. They are also part of the baseline acceleration system for muons in a Neutrino Factory.

However, this type of accelerator, in particular

References (37)

  • S. Koscielniak et al.

    Nucl. Instr. and Meth. A

    (2004)
  • J.S. Berg

    Nucl. Instr. and Meth. A

    (2008)
  • Harada

    Nucl. Instr. and Meth. Res. A

    (2009)
  • T. Ohkawa Bull. Phys. Soc. Jpn. (1953); A.A. Kolomensky, V.A. Petukhov, M.S. Rabinovich, Lebedev Ph. I. Rpt. (1953);...
  • C.R. Prior et al., Proceedings of the 11th European Particle Accelerator Conference, Genoa, Italy, 23–27 June 2008,...
  • M. Yoshimoto et al., Proceedings of the 19th IEEE Particle Accelerator Conference, Chicago, IL, USA, 18–22 June 2001,...
  • T. Adachi et al., Proceedings of the 19th IEEE Particle Accelerator Conference, Chicago, IL, USA, 18–22 June 2001, p....
  • Y. Mori et al., Proceedings of the 11th European Particle Accelerator Conference, Genoa, Italy, 23–27 June 2008, p....
  • M. Tanigaki et al., Proceedings of the 9th European Particle Accelerator Conference, Lucerne, Switzerland, 5–9 July...
  • A. Sato et al., Proceedings of the 10th European Particle Accelerator Conference, Edinburgh, UK, 26–30 June 2006, pp....
  • F. Mills, Proceedings of the 4th International Conference of Physics Potential and Development of μ+ μ‐ Colliders, San...
  • J. Scott Berg, Longitudinal reference particle motion in nearly isochronous FFAG recirculating accelerators, in:...
  • J. S. Berg, Dynamics in imperfectly-isochronous FFAG accelerators, in: Proceedings of EPAC 2002, Paris, France (EPAC,...
  • S. Koscielniak et al.

    Longitudinal dynamics in an FFAG accelerator under conditions of rapid acceleration and Fixed, high RF

  • J.S. Berg

    Phys. Rev. ST Accel. Beams

    (2006)
  • S. Machida et al.

    Phys. Rev. ST Accel. Beams

    (2007)
  • Cited by (49)

    • Physics with CEBAF at 12 GeV and future opportunities

      2022, Progress in Particle and Nuclear Physics
    • The PyZgoubi framework and the simulation of dynamic aperture in fixed-field alternating-gradient accelerators

      2015, Nuclear Instruments and Methods in Physics Research, Section A: Accelerators, Spectrometers, Detectors and Associated Equipment
      Citation Excerpt :

      The smaller and simpler magnet designs have led to several proposals for their use which include the PAMELA (Proton Accelerator for MEdicaL Applications) design study for particle therapy [16]. The first demonstration of a non-scaling FFAG was recently made using the EMMA (Electron Machine with Many Applications) experiment [17–19]. The fact that a particle׳s radial orbit changes significantly within the accelerator magnets means that their simulation differs from that required in many other accelerator design codes.

    • Experimental simulation of beam propagation over long path lengths using radio-frequency and magnetic traps

      2014, Nuclear Instruments and Methods in Physics Research, Section A: Accelerators, Spectrometers, Detectors and Associated Equipment
    • Design, tests and commissioning of the EMMA injection septum

      2013, Nuclear Instruments and Methods in Physics Research, Section A: Accelerators, Spectrometers, Detectors and Associated Equipment
      Citation Excerpt :

      Electron Model for Many Applications (EMMA) is a proof-of-principle electron non-scaling fixed-field alternating gradient accelerator recently built at STFC Daresbury Laboratory in UK [1–4].

    • Fixed-field alternating-gradient accelerators

      2023, Handbook Of Accelerator Physics And Engineering: Third Edition
    View all citing articles on Scopus

    Work supported by Research Councils UK, Grant number EP/EO 32869/1

    View full text