Proton Driver Scenarios at CERN and Rutherford Appleton Laboratory
Total Page:16
File Type:pdf, Size:1020Kb
PHYSICAL REVIEW SPECIAL TOPICS - ACCELERATORS AND BEAMS 16, 054801 (2013) Proton driver scenarios at CERN and Rutherford Appleton Laboratory J. W. G. Thomason STFC/ISIS, Didcot, OX11 0QX, United Kingdom R. Garoby and S. Gilardoni CERN, Geneva, Switzerland L. J. Jenner Imperial College London, SW7 2BW, United Kingdom, and Fermilab, Batavia, Illinois 60510-5011, USA J. Pasternak Imperial College London, SW7 2BW, United Kingdom, and STFC/ISIS, Didcot, OX11 0QX, United Kingdom (Received 19 October 2012; published 6 May 2013) The concept of sharing a high-power proton accelerator (proton driver) between neutrino production and other facilities such as a high-energy collider or a spallation neutron source is an attractive, cost- effective solution which is being studied in site-specific cases as part of accelerator upgrade plans at CERN and at the ISIS facility at the Rutherford Appleton Laboratory. Descriptions of these proton-driver scenarios will be given primarily in the context of the Neutrino Factory, but other neutrino production plans at CERN will also be noted. DOI: 10.1103/PhysRevSTAB.16.054801 PACS numbers: 29.20.DÀ I. INTRODUCTION pole, a chopper, and a linear accelerator. In the solution adopted at CERN, the final energy of the proton driver The proton driver at the Neutrino Factory (NF) [1]is is delivered by the linac and the beam time structure required to deliver a proton beam of 4 MW at a repe- must be obtained with the help of charge-exchange tition rate of 50 Hz to the pion-production target. The injection into an accumulator ring followed by fast proton-beam energy must be in the multi-GeV range in phase rotation in a dedicated compressor ring. An alter- order to maximize the pion yield. In addition, the NF native approach, as adopted in the Rutherford Appleton specifies a particular time structure consisting of three Laboratory (RAL) solution, is to use a chain of circular very short bunches separated by about 120 s. To allow accelerators, typically rapid cycling synchrotrons the muon beam to be captured efficiently, short, 1–3 ns (RCSs), where bunch compression is accomplished adia- rms, bunches are required. Each bunch from the proton batically in the final RCS. driver will become a separate muon bunch train. The bunch separation is constrained by beam loading in the downstream muon accelerator systems and by the time TABLE I. Proton-driver requirements. A proton kinetic energy scale for disruption of the mercury-jet target. The in the range 5 to 15 GeV has been shown to provide adequate à proton-beam parameters necessary to produce the de- performance. The number of protons, beam radius, , and sired number of muons in the storage rings of the NF geometric emittance correspond to the values for an 8 GeV proton beam. are listed in Table I. In order to achieve such short bunches, a dedicated bunch compression system must Parameter Value be designed to deal with the very strong space-charge Kinetic energy 5–15 GeV forces. Average beam power 4 MW Several proton-driver schemes fulfilling these require- (3:125  1015 protons=s) ments have been proposed. Typically they consist of an Repetition rate 50 Hz HÀ -ion source followed by a radio-frequency quadru- Bunches per train 3 Total time for bunches 240 s Bunch length (rms) 1–3 ns Published by the American Physical Society under the terms of Beam radius 1.2 mm (rms) the Creative Commons Attribution 3.0 License. Further distri- Rms geometric emittance <5 m bution of this work must maintain attribution to the author(s) and à at target 30 cm the published article’s title, journal citation, and DOI. 1098-4402=13=16(5)=054801(7) 054801-1 Published by the American Physical Society THOMASON et al. Phys. Rev. ST Accel. Beams 16, 054801 (2013) II. PLANS AT CERN be kept below 2000 K. The beam parameters after accu- mulation are obtained as a compromise between the com- A. The neutrino factory peting requirements of minimizing the heating of the The CERN NF scenario would be based on the proposed injection foil, maximizing the aperture, and adequate com- 5 GeV, high-power version of the Superconducting Proton pensation of the space-charge forces and are set to allow Linac (SPL) [2], which can deliver 1014 protons at the for rf phase rotation in the downstream compressor ring. repetition rate of 50 Hz [3]. In the recent past, the SPL The size of the two rings is determined by the requirement study evolved into an international collaboration whose that successive bunches must arrive at the correct location aim is the optimization of the architecture of a pulsed in the compressor ring. The compressor ring has a large superconducting high-power proton linac. The most recent phase slip factor, which is needed for the fast phase rota- design of the SPL and the description of the goals of the tion. Tracking simulations in the compressor ring have collaboration can be found in [4]. been performed using the ORBIT code [6]. The good per- In the CERN scenario, the chopped beam from the SPL formance of the compressor ring is demonstrated in Fig. 2 would be injected into an isochronous accumulator ring in (left panel). The simulations have also been used to inves- which 120 ns long bunches are formed without the need for tigate the transverse phase space. Figure 2 (right panel) an rf system. The absence of synchrotron motion in the shows that the transverse space charge can be tolerated due accumulator ring makes it important to study the stability to the limited number of turns of the beam in the compres- of the beam in the presence of space charge. As presented sor ring and the relatively large dispersion, which effec- in [5], transverse stability can be obtained with a suitable tively lowers the tune shift by enlarging the beam size. The choice of chromaticity as shown in Fig. 1 (left panel) and parameters of the accumulator and compressor rings are longitudinal stability can be achieved by limiting the lon- listed in Table II. More details of the CERN proton-driver gitudinal broadband impedance to a few ohms as shown in scenario can be found in [7]. The low energy normal- Fig. 1 (right panel) [5]. Two-dimensional phase-space conducting part of the SPL is currently under construction painting is used in the stripping injection into the accumu- and should become operational in the following few years lator ring, allowing the temperature of the stripping foil to as part of the LHC injector chain. The existing proton 1 1 Q’x=0 Zl=30 Ω Q’x=-7 Zl=20 Ω Q’x=7 0.8 Zl=10 Ω Q’x=10 Zl=8 Ω Ω 0.1 Q’x=-10 Zl=6 0.6 Zl=4 Ω Zl=3 Ω (m) x σ 0.4 0.01 0.2 longitudinal emittance (eV.s) 0.001 0 0 100 200 300 400 500 600 0 200 400 600 800 1000 no. of turns no. of turns FIG. 1. Horizontal beam size evolution for various values of chromaticity assuming a transverse impedance of 1M =m, QR ¼ 1, fR ¼ 1GHz (left). Longitudinal emittance evolution for different values of shunt impedance (right). The lines for 3 and 4lie on top of each other in the figure. FIG. 2. Phase-space plots before and after bunch rotation. 054801-2 PROTON DRIVER SCENARIOS AT CERN AND ... Phys. Rev. ST Accel. Beams 16, 054801 (2013) TABLE II. Parameters of the accumulator and compressor geometry is constrained by the location of Linac4 and rings for the CERN proton-driver scenario. the space needed for the muon front-end and muon accel- Parameter Value eration chain. Accumulator ring Circumference 185 m B. Super beams and beta beams No. of turns for accumulation 640 As well as provision for the NF, other new proposals are Working point (H/V) 7:37=5:77 being made for experiments at CERN requiring higher Total bunch length 120 ns beam power to produce neutrinos by either exploiting À3 rms momentum spread 0:863  10 existing machines or assuming the availability of beam Compressor ring from the 5 GeV, high-power version of the SPL. Circumference 200 m No. of turns for compression 86 1. Neutrino experiments with existing accelerators rf voltage 1.7 MV Gamma transition 2.83 The existing CERN accelerators deal with multiple Working point 4:21=2:74 types of particles and supply beams to numerous experi- ments (see layout in Fig. 4). The CNGS experiment [9] is currently operating using a 500 kW proton beam from linac, Linac2, will be replaced soon by the more modern the SPS and sending neutrinos to the Gran Sasso under- Linac4 [8] that will accelerate HÀ up to 160 MeV, before ground laboratory 730 km away. The SPS performance injection into the Proton Synchrotron Booster. The linac is expected to improve by the end of the decade, as a performances will match the requirements of the program result of the on-going upgrade program of the LHC of increasing the LHC luminosity. injectors [10]. The recently started LAGUNA-LBNO Figure 3 shows a preliminary layout of the NF on the Design Study [11] is aimed at making use of the in- CERN site, using the SPL followed by a transfer channel creased SPS beam power, tentatively set at 750 kW, for towards the accumulator and compressor rings. This generating a conventional beam and sending it to a SPS TUNNEL NEAR DETECTOR CAVITES RLA II 3.6 - 12.6 GeV RLA I 0.9 - 3.6 GeV MUON DECAY RING PRE LINAC LHC TUNNEL ROTATOR TARGET STATION STACKED ACCUMULATOR & COMPRESSOR RINGS TRANSFER TUNNEL SPL MEYRIN SITE PS LINAC 4 FIG.