TRIUMF: Canada's national laboratory for particle and nuclear physics

Proceedings of the Workshop on the Production and Transport of Secondary Beams at a KAON Factory, Vancouver,… Gill, D. R. (David R.), 1939- 1986

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


51833-TRI-1986-02.pdf [ 93.94MB ]
JSON: 51833-1.0107831.json
JSON-LD: 51833-1.0107831-ld.json
RDF/XML (Pretty): 51833-1.0107831-rdf.xml
RDF/JSON: 51833-1.0107831-rdf.json
Turtle: 51833-1.0107831-turtle.txt
N-Triples: 51833-1.0107831-rdf-ntriples.txt
Original Record: 51833-1.0107831-source.json
Full Text

Full Text

TRIUMFPROCEEDINGSOF THEWORKSHOP ON THE PRODUCTION AND TRANSPORT OF SECONDARY BEAMS AT A KAON FACTORYVANCOUVER JUNE 2-4, 1986Editor: D.R. GillMESON FACILITY OF:UNIVERSITY OF ALBERTA SIMON FRASER UNIVERSITY UNIVERSITY OF VICTORIA UNIVERSITY OF BRITISH COLUMBIAOPERATED UNDER A CONTRIBUTION FROM THE NATIONAL RESEARCH COUNCIL OF CANADA TRI-86-2TRI-86-2PROCEEDINGSOF THEWORKSHOP ON THE PRODUCTION AND TRANSPORT OF SECONDARY BEAMS AT A KAON FACTORYVANCOUVER JUNE 2-4, 1986Editor: D.R. GillPostal address:TRIUMF4004 Wesbrook Mall Vancouver, B.C.Canada V6T 2A3 December 1986PREFACEThis volume contains the texts of the talks presented at the workshop on 'The Production and Transport of Secondary Beams at a KAON Factory' held at TRIUMF June 2-4, 1986. Included are both the invited and contributed papers whose manuscripts were available at the close-out date. These proceedings have been produced via off­set printing directly from the camera-ready papers provided by the authors whom we thank very much for their co-operation.The topic of the workshop is developing very quickly, being spurred on by the competition of proposals to build kaon factories in different parts of the world and the belief that probably only one such facility will be realized. The rate at which these devel­opments are advancing was made very evident at the workshop by the lively discussions that took place during and following each presentation, and we regret that we could not reproduce those here. We thank all the presenters and participants for their contribu­tions toward the very full and frank exchange of ideas that took place.I wish to thank Pat Stewart and Krish Thiruchittampalam for their tremendous contributions to the organization of the workshop. I also thank Ada Strathdee for her tireless efforts in the production of these proceedings.D.R. GillCONTENTSPageThe experimental facilities of the European Hadron FacilityP. Blum .............................................................  1Experimental area plans for an advanced hadron facility (LAMPF II)E.W. Hoffman, R.J. Macek and C. Tschalar ..........................  10TRIUMF's experimental area plansJ.L. Beveridge ...................................................... 20BNL kaon beams, present and plannedD.M. Lazarus ........................................................  28Experimental facilities for a Canadian KAON factoryC. Tschalar ......................................................... 42The TRIUMF KAON factory acceleratorsM.K. Craddock .......................................................  60The initial experimental program at a KAON factoryP. Kitching .........................................................  68Hyperon scattering at the TRIUMF KAON factoryN.E. Davison ........................................................  79More muonsJ.H. Brewer ........................................................  88Experience with target area shielding around the CERN 26 GeV proton synchrotronA.H. Sullivan ......................................................  97TRIUMF KAON factory shielding and activation considerations:Extrapolating to 30 GeV, 100 yA proton beamsI.M. Thorson .......................................................  107A remote component servicing system for the TRIUMF KAON factory experimental hallC.R. Mark and W.M. Cameron ........................................  112Targets for high intensity beam at CERN: Design, operational experience and developmentsR. Bellone, A. Ijspeert and P. Sievers .............................. 119Targets for TRIUMFT.A. Hodges .........................................................  128Particle production and targetting experience at the Brookhaven AGSD.M. Lazarus ...........................................................  133Antiprotons at the KEK-PS K4 beam lineK.H. Tanaka, T. Tanimori, Y. Fujii, Y. Sugimoto, M. Sudo, S. Kohno,K. Kasai and Y. Morita .............................................  139Production of secondary particles at the TRIUMF KAON factoryA. Yamamoto and M. Takasaki ........................................  143Influence of source size on acceptance and separation of kaon beamsJ. Doornbos .........................................................  148Low momentum kaon beam lines: Contamination and design criteriaD.E. Lobb ...........................................................  153Development of electrostatic separatorsA. Yamamoto .........................................................  162List of participants .................................................. 176vTHE EXPERIMENTAL FACILITIES OF THE EUROPEAN HADRON FACILITYPeter BlumKernforschungszentrum Karlsruhe, Institut fur Kernphysik Institut fur Exp. Kernphysik, Universitat KarlsruheABSTRACTThe European Hadron Facility is designed as a high intensity proton synchrotron of 30 GeV primary energy. The large projected current of 100 yA is almost two orders of magnitude higher than that of existing accelerators in this energy range. This requires a careful design, which minimizes beam losses, but on the other hand offers a many- -faceted facility for a broad experimental programme. The EHF will provide both fast and slowly extracted beams; thus pulsed and con- tinous secondary beams of high quality can be designed and optimized to experimenters requirements. Particular effort is made to provide polarized protons up to 30 GeV.INTRODUCTIONThe present situation of experimental nuclear and particle physics at intermediate energies in Europe is marked particularly by the pro­blem, that the CERN-PS has been converted into a multipurpose injector for different accelerators. Thus in the near future there is almost no possibility to use the PS for experiments in subatomic physics.Even if the other existing accelerator facilities such as SIN and LEAR at CERN still enable scientifically interesting research, they only defer the end of this successful branch of physics in Europe.Competitive research in the future can only be realized with a dedicated proton acceler­ator of high intensity, which provides the tools necessary to study the manifold questions still open in the understanding of fundamental interactions. Therefore several physicists started seriously thinking about a KAON-factory either as an upgrade of SIN or as a new acceler­ator scenario - European Hadron Facility. The selected name indicates that this project is considered to be beyond the scope of a single country.As our colleages from TRIUMF we understand KAON as an acronym for many particles offering a possibility of many specific experiments. Here I think it is hardly necessary to spell out the tremendous number of fundamental questions which may be attacked and hopefully settled by such a.facility. Especially kaon beams of high intensity and excellent quality are assumed to be an appropriate tool to study such basic physical problems as the generation problem of the quark-lepton fam­ilies, the still puzzling CP-violation and many more. Precise measure­ments of rare p- or K-decays may serve as "window" for physics beyond the standard model.As the EHF-project is certainly not yet known to everybody, I would like to introduce it with some remarks and turn then to the preliminary layout of the experimental facilities, which was mainly worked out for2the SIN upgrade scenario by R. Abela, K. Gabathuler, C. Petitjean,D.Renker, C.Tscalar of SIN, C.Wiedner from MPI Heidelberg, and P.Blum from University of Karlsruhe and is based on a new targeting technique proposed by C. Tschalar 1.EHF-CONCEPTUAL DESIGNUsing the experience of accelerator engineers and physicists all over Europe, USA, Canada the EHF-design group succeeded in settling a con­ceptual design this spring 2. The design objectives of the accelerator are dictated by the research programme in Medium Energy and Modern Nuclear Physics:1) Maximum energy of the primary proton beam of 30 GeV with a possible upgrade to 40 GeV2) Average extracted current of 100 pA (6.25*10lu protons per sec)-43) Extracted beam with low (10 ) and high (1.0) duty factor4) Capability to accelerate polarized protons up to top energyThe large projected current leads to the important design criterion to minimize the amount of beam losses to a level of smaller than 1% of the beam intensity per cycle in order to achieve a high reliable perform­ance and to make cumbersome remote handling dispensable.From different scenarios discussed we favour the one sketched on figure 1. It is made of one Linac and four rings, which are: Booster, Accumulator, Main Ring, and Stretcher.European Hadron FacilityFigure 1: Proposed scenario for EHF3The circumference of the Main Ring and the Stretcher, which is needed for slow-spill ejection of a duty factor of 1.0, is about the same and has been taken to be 960 m,which would fit into the ISR-tunnel at CERN. As the Booster is halfsize the Main Ring, but cycles at twice the rate, a flat bottom magnet cycle would be required. In order to release this burden from the magnet power supplies an Accumulator is foreseen as an intermediate storage ring.The most striking point of this design is the choice of the transfer energies. The intermediate value of 9 GeV has been taken in order to preserve polarization during acceleration by the inclusion of one or more Siberian Snakes in the Main Ring. Thus polarization will be main­tained without requiring large periodicity of the lattice. A Siberian Snake is a special arrangement of transverse magnetic fields turning the spin by 180° leaving the orbit before and after the snake un­changed. Thus two successive traversals of the snake replaces the energy dependent spin tune f = G*2f by a fixed one, f = 1/2. ( G is the anomalous magnetic moment of the proton). The aperture of the magnets however have to accomodate the lateral displacement of the tra­jectories of particles of different momentum (injection to ejection). As the deflection varies with 1/Jf one tries to choose the injection as high as possible. With an injection energy into the Main Ring of 9 GeV the apertures are of the order of 11 cm for conventual 1.8 T magnets. The value of the injection energy into the Booster is determined by the problem of the incoherent space charge tune shift. We require a normalized emittance of 25 it nuiumrad in both planes, which represents a beam with a considerably high brilliance. In order to reduce the in­coherent Laslett tune shift to a value smaller than .2, so that one still finds locations in the tune diagram far from any systematic re­sonance, one has to choose an injection energy of 1.2 GeV.The disposal of the large projected current requires a compromise of the number of particles per pulse, which more or less limits the re­liability of the accelerator, and the repetition rate of the cycle, which has to be paid for. In a reasonable compromise one operates the Booster near space charge limit, which is given by 2.5*1013 protons per pulse and 25 Hz repetition rate.A linear accelerator of 12 mA and 360 ysec pulse length requires 200 turns injection, which is most efficiently achieved by a tf* - injection. In order to meet our design criterion we will use the high brightness Linac beam of 3 it mm»mrad normalized transverse emittance and .6*10 * eV sec longitudinal bunch area to fill the much larger acceptances of the Booster (25 tt mm mrad, and 0.05 eV sec). Placing the Linac bunches of successive turns at different positions in the Booster rf-bucket one fills the phase space in all three planes as uniform as possible. Thus the transverse and the longitudinal space charge forces are reduced. Beam losses during injection are minimized, if the Linac beam is modulated with the Booster rf-frequency. This is achieved by filling only two out of eight Linac bunches. In order to prevent particles from overflowing the Booster rf-bucket during the transition from coasting to acceler­ated beam, the rf-bucket will be filled only half.4In order to keep beam losses in the rest of the accelerator small we take the vacuum betatron acceptance four times the beam emittance at injection and included an allowance for closed orbit distortions. The transfer between two rings will occur in box-car fashion and bunch to bucket. As we provide a proper void in the beam for exciting the fast kicker elements, we do not anticipate beam losses during transfer and fast extraction. However the septum for slow extraction at maximum energy is expected to be a second location of major losses, which has to be studied carefully.The energy ranges of the accelerators chosen allow to avoid to cross transition energy by choosing the transition energy outside the acceleration range of both rings. Thus lattice design becomes simpler and more straightforward. We choose separated function lat­tices, which provide a larger flexibility in controlling the working point. The low periodicity enables the design of long dispersion-free straight sections to house rf-cavities, Siberian Snakes and fast extraction elements. For the slow-spill ejection out of the Stretcher, which will last 80 msec, a high beta insertion will be designed. The major parameters of the accelerators are listed in table 1.EXPERIMENTAL FACILITIES AT EHFThe most important aspect for the user of such a facility is, besides the reliability of the accelerator, what experimental facilities are offered to carry out the manifold experimental programme. As shown in figure 1 the facility provides two experimental areas: one that takes beam from the fast extraction mode, and the other from the slow extraction with high duty cycle.At this stage of the design of EHF the layout and particular the cho­sen secondary beams should be considered as very preliminary and mainly for getting a feeling for overall requirements like area size, number and momentum range of beams, technical realization possibilities, etc.PULSED BEAM AREA:The fast extracted beam of the Booster may be shared on a pulse-by-pulse basis with one pulse sent to the fast extraction area and the next two pulses are injected to the Main Ring for acceleration up to 30 GeV. The amount of sharing may be chosen at any ratio desired. Though a detailed design of this area has not yet been performed, we included in figure 2 for completeness a layout, which is assumed to be similar to that pro­posed for LAMPF II 3. The pulse of 2.5*1013 protons is hitting a one interaction length high-Z target, where copious it production occurs.A special focussing device collects a parallel beam to the long decay channel. Here the pions and kaons decay in flight:± ±,t->it -*■ y  + v *+  +  C-)K“  -*• u +y„At the end of the channel a thick steel shielding absorbs all particles except the neutrinos.5TABLE 1: General parameters of the EHF acceleratorsLinac Boaster Main Ringenergy GeV 0.002-1.2 1.2-9.0 9.0-30.0rep. rate Hz 25 25 12protons/cycle 1013 2.5 2.5 5.0peak current A 0.15 2.3 2.5beam length m 340 480 960norm, emittancetransversal mm mrad 3 it 25 TT 25 TTlongitudinal eV sec 60*lO-6 TT 0.05 TTlattice type DFO,separated functionFODO,magnetssuperperiods 6 4cells 54 52tune hor. 13.4 8.65vert. 10.2 8.83 *3 xmax ymax m 12.5*14.8 30.2*32.8Dxmax m 1.9 4.1yTR 12.6 8.2depolarizing resonaces 4 20preservation of pol. spin matching, SiberianQ-jump Snakeradio frequency Mhz RFQ:50/400 DTL: 400 SCL:120050.6-56 56-56.2Chopper:0.55peak rf-voltage MV .9 2.05cavities 18 15harmonic 90 180total rf power MW 1.2 2.3 5.5620 mFigure 2: Possible scenario for pulsed beam areaBesides the fascinating v-facility the fast extracted proton beam maybe used for a further exciting experimental extension - a pulsed muon facility. As the production cross section of low energy muons is almost isotropic, one may extract from the upstream part of the v-target looking under 120 degrees a low momentum y-beam, which may be used in a variety of experiments. However the pulse length of 1.6 ysec is only optimal for selected experiments with pulsed muons.By installing the transfer line between the Main Ring and Stretcher atthe proper location, the fast extracted proton beam of 30 GeV can alsobe transported to the v-facility, which means an essential increase of flux. But here special care has to be given to the design of the target to prevent deterioration due to the thermal load.CONTINOUS BEAM AREA:The slowly extracted beam from the Stretcher will be transported into the slow extraction hall. A dipole magnet in the transfer line will allow to switch polarized protons into the dedicated area, while the unpolarized beam will traverse a sequence of targets of typically one interaction length. As shielding requirements and target station become very delicate and expensive for high intensity beams, one has to look for the most economic layout, which implies the extraction of as many beams from a target as possible. But this is in conflict with the fact that the production cross section for high momentum particles is at maximum at small angles. Thus zero degree take-off angles are re­quired to maximize flux. In addition, high resolution beams need a small spot size target, which again means small take-off angles. And finally a flexible experimental programme requires independently tune­able secondary beams within their momentum and angular range.7In order to accomodate these conflicting requirements,a magnetic system for multiple achromatic extraction of independent momenta (MAXIM) has been proposed by C. Tschalar. This system in principle disentangles the production angle of the secondaries from the take-off angle of the channel, and allows to extract up to three beams with small production angle from a single target. An optimal layout consists of a central beam extracted at zero degrees and two symmetrically arranged lateral beams, which are extracted at about 200 mrad. The central beam of relative high momentum allows a free choice of the polarity and of the momentum. The two lateral beams of lower momentum are independent in momentum but require opposite polarity to optimize the flux.The major design criterion for the secondary beams of high beam purity has to be met by different means depending on the momentum range of the channel and on the type of secondaries tobe selected. Low momentum beams have problems with the short decay length of kaons, so that no long electrostatic separators can be used. As the contamination origin­ates from the decay of short-lived hyperons near the target and scattered particles from magnetic apertures, one counteracts best by introducing an intermediate focus, which serves as new target spot and will be imaged to the experimental target area. It is obvious, that beam purity has to be compensated by intensity due to the increased length. In a channel of medium energy a two-stage electrostatic sep­arator solution may be used, where rejection factors of up to 10s have been achieved. For momentum higher than 6 GeV/c electrostatic separat­ors become technically unfeasible and only superconducting rf-separat- ors “ can be used for a high duty cycle beam.Figure 3 shows a layout of the continous beam area using the principle of MAXIM. The primary beam first hits a 5 cm long tungsten target (half an interaction length) and is refocused to a second target of 10 cm tungsten. From these targets seven charged secondary beams are extracted, which cover the entire momentum range. Table 2 summarizes the general properties of these channels and gives flux estimations for a 100 yA primary beam current.TABLE 2: General properties of the continous beams (particle fluxes are quoted for underlined momenta)3eam Momentum Production Take-off Ap/p Afl Length particle fluxGeV/c angle angle X msr m ir"(10 1 0 )K'(10*P) (10')S2 0.02-0.2 120° 120° 15 50 12K1 0■4-0.8 0° -50 11.5° 5 5 18 1.5 0.6 0.02K3 0.7-1.5 0° -4° 5.7" 3 2 25 1.3 1.0 0.25K4 1.4-3.0 0° -3° 5.7° 5 1 35 3.0 7.0 4.0K5 2.0-6 .0 0° -3° 0° 5 .2 60 1..6 4.0 2.0K6 5.0-20.08.00° 0° 5 . 1 1000.7 0.9 0.9K0 0.6-20.010.00" 0° wide­band.1 209* 10*8Figure 3: Preliminary layout of the continous beam area9From the first thin target four charged beams are extracted. The high momentum beam K5 is produced at 0° and needs a 20 m long electrostatic separation stage. The two lateral beams are assigned for low energy kaon experiments. K1 will be a high resolution beam useful for hyper- nuclear physics experiments. K2 will be designed for high intensity. As the production cross section for low momentum pions is almost isotropic a large acceptance beam with a take-off angle of 120° and a momentum range of .02 to .2 GeV is foreseen.The two lateral extract­ed beams from target 2 are of medium energy. While K3 is again a high resolution beam with the emphasis for S = -2 hypernuclei, is the momentum range of K4 optimal to accept the maximum of the antiproton production. The central beam of target 2 has a wide momentum range and is devoted to manifold experimental studies like rare K-decay, hyperon physics (A,I,S,J2) or, if tuned to p, charmed baryon spectro- scopy, search for exotic matter, etc. This channel has to use a two cavity superconducting rf separation. In addition this channel can be switched to a 750 m long quadrupole channel to produce a high momen­tum p-beam. Before dumping the primary beam a further target of one interaction length may be traversed to provide a neutral wideband kaon beam K0. In order to reduce the neutron background to a ratio of about one two methods are foreseen:a) The remaining proton beam is bent by about 3° and the beam dumpis installed far from the K0 channel.b) The different absorption cross section of K° and n is used for afurther rejection of the neutrons. A 3 m long 12C absorber gainsabout a factor 10 in rejection.Here again beam intensity has to be paid for beam purity. The lateral dimensions of this beam have to be defined by a set of collimators.CONCLUSIONThe previous discussion has to be seen as a first draft of experimen­tal facilities on EHF. These as multi-purpose designed facilities can be optimal used in experiments of many kinds. The use of the magnetic extraction system MAXIM allows a considerable concentration of the external primary beamline. This means not only a cost optimized sol­ution, but provides also an increase in luminosity by using the whole beam intensity on the target. Thus the superiority due to the larger beam current compared to existing facilities is even enlarged. So generally speaking such a facility would enable the European Medium Energy physics community to contribute also in future in the wonted manner to the manifold open questions in fundamental interactions.REFERENCES1. C. Tschalar to be published in Nuc. Inst, and Meth. and this pro­ceedings2. EHF Feasibility Study Report, June 19863. LAMPF II Proposal 19864. A. Citron et al. Nucl. Inst, and Meth. 155 (1978) p9310EXPERIM ENTAL AREA PLANS FOR A N  ADVANCED H ADRON FACILITY*(LAMPF II)E. W. Hoffman, R. J. Macek Los Alamos National Laboratory, Los Alamos, NM 87544C. Tschalar SIN, CH-5234 Villigen, SwitzerlandABSTRACTA brief overview is presented of the current plans for an experimental area for a new advanced hadron facility for the exploration of nuclear and particle physics. Reference 1 contains a more complete discussion of this facility and the justification and plans as of this date.LAMPF II OVERVIEWThe Los Alamos National Laboratory is located on the Pajarito Plateau in the Jemez Mountains in north-central New Mexico. The Clinton P. Anderson Meson Facility (LAMPF) is situated on one of the finger mesas that form the plateau. An advanced hadron facility (LAMPF II) is presently visualized as consisting of the LAMPF linac sending 800 MeV protons to a 6 GeV booster ring followed by a 45 GeV main ring. The current available at 6 GeV is 144 /ta at 60 Hz and 0.0067% duty factor and at 45 GeV is 32 /ia at 3.3 Hz and 50% duty factor.Figure 1 shows the two experimental areas in a plan-view for LAMPF II. Space is also left in each accelerator tunnel for an internal gas-jet target and associated experimental equipment. Area N is intended to provide neutrinos via a pair of pulsed focusing horns. Area A is designed to accommodate secondary beams that span the range of useful energies up to 35 GeV/c. The planned charged-particle beamlines are listed in Table I with their intended characteristics. The charged-particle beams originate from two consecutive thick target stations and a third station will be used as a source for one or more neutral beams. This serial use of targets in a single primary proton beam provides maximum secondary particles per extracted proton.* Work performed under the auspices of the U.S. Department of Energy.11Table I. Beam specifications for Area A.Beam  lineE xtraction  A ngle (deg)P roduction  A ngle (deg)P(G e V /c )AHa(m sr)ap/p R esolution  A cceptance(%) (%)Length(m ) Separation/i0 .2 120 120 0.02-0 .2 50 0.5 15 ~ 2 0 l-s ta g e  dcK0.8 11.5 0-5 0.35-0.8 5 (8) 0.1 5 18 2-stage dcK6 0 0-3 2-6 0.07 (3) 1 5 75 1-stage dcTest beam 0 0-6 ~ l - 3 3 ~ 1 ~ 5 ~ 3 5K 1.5 5.7 0-4 0.6-1.5 2 (5) 0.05 3 25 2-stage dcK 2.5 5.7 0-3 1.0-2.5 1 (5) 0.50 5 35 2-stage dcK35 0 0 5-35 0.50 1 5 90 NoneK 35 sep.^ 0 0 5-30 0.06 1 1 ~ 1 3 0 rfTest beam 0 0 ~ 5 -3 0 0.50 ~ 1 ~ 3 ~ 7 5a The angular acceptances in this colum n result from  all restrictions in the channel. T he ones in parentheses are the largest ones allowed by the extraction  system s only. ^ These are the param eters o f Kzb if it is constructed  as a separated beam .BEAM  SPECIFICATION GOALSThe physics opportunities require that the LAMPF II experimental areas should provide a number of high-intensity, high-quality beams simultaneously to several experiments. The major requirements are for:• High-intensity beam• High beam quality (brightness, purity, resolution)• High beam availability• Multiple beam ports for high throughput of experiments• Access to beams of all energies and species that can be produced at LAMPF II, including i/, K, it, p, p, and polarized protons• Flexibility to meet the requirements of future physics opportunities.Source-Brightness OptimizationThe high-brightness requirement is met by targeting full-intensity, full-energy proton beam on the highest density target that will survive and by choosing production angles in very forward directions. We believe that a successful water-cooled, fixed copper target can be developed. Higher brightness by a factor of ~ 1.6 can be obtained with a tungsten target, provided certain technical uncertainties with target cooling and operational reliability can be mastered. Production angles for the K  and p beams are chosen as close to 0° as possible, consistent with reconstruction of the primary beam and the geometrical constraints of other secondary beams from the same target.Beam-Purity ConsiderationsHigh beam purity for K  and p beams is a crucial requirement that has major impact on our ability to use high intensity effectively. In general, some intensity will be traded for higher purity. Our goal is K /it or p/ir ratios better than 1:1. For some energies this requires it- and /^-rejection factors of 104. Below about 10 GeV/ c, dc separators are effective; above this region we must consider rf separators.A single-stage electrostatically separated beam is, in practice, limited to it- and /^-rejection factors of 100- 200, principally because of the halo of it's and p ’s generated by decays of short-lived particles in the vicinity of the production target and scattering from pole tips and collimators. In principle, these can be eliminated by careful collimation and trimming of the source halo at an intermediate focus upstream of the separator. A second stage of separation can accomplish the same goal more effectively in a comparable length of extra beam transport. In fact, two-stage separated beams for12bubble-chamber work have achieved 7r-rejection factors as high as 105.2 We choose two-stage separation wherever rejection factors greater than 200 are required.Beam-Line Acceptance and LengthIn separated beams the acceptance in the plane of separation is usually limited by the separator deflection. In the dispersion plane the acceptance is more complicated, but is often limited by the momentum resolution required by the physics and the volume of high-quality magnetic field one is willing to provide. For rf separation both planes are likely to be limited by the separator geometrical acceptance, with the deflection plane having about one-third the acceptance of the nondeflection plane.3 These principles were used to set the solid-angle acceptance of the beams shown in Table I. They are consistent with the parameters for existing beams at BNL4 and KEK5.The length of all but the separator section typically scales with p1/ 2 for fixed momentum resolution and fixed aperture magnets but the length of the separation section scales differently. We have chosen a factor of approximately 2 in the useful momentum range of dc-separated channels to avoid a large mismatch between separator sections and the rest of the channel and to minimize decay losses.EXPERIM ENTAL AREA LAYOUTFast-extracted Beam Area (Area N)Fast-extracted beam from the booster (initially) and from the main ring (eventually) is transported below grade to an underground target station at the beginning of Area N. This area is visualized as a sole-use area containing a thick (~40 cm graphite) target, a double pion-focusing horn system, a decay region for the pions, a shield, and an appropriate neutrino detector.Charged Secondary Beam Area (Area A)The experimental area layout (Fig. 2) is designed to make maximum use of the existing physical plant. The slow-extracted beam from the main-ring accelerator is transported to a reconfigured Area A via a beam tunnel that joins the present linac switchyard. Two targets for eight charged-particle beams and a third one for neutral beams are all arranged in series in a single proton channel. All three target stations can fit within the confines of existing structures.We have chosen a set of beams that will provide pions, kaons, and antiprotons of every momentum up to 35 GeV/c. Electrostatically separated beams are provided for kaons up to 6 GeV/c. With the successful development of rf separators one could have separated beams up to 35 GeV/c. With a limited number of beam lines it is not possible to provide a beam optimized for and dedicated to every special purpose; some compromises were made so all the important requirements would be included. Flexibility has been retained to adapt to new developments and opportunities that the future is certain to bring.Half-quadrupoles are proposed as the first focusing elements of most beams to provide maximal acceptance and forward extraction angles. Magnets of this type recently purchased at SIN show excellent optical properties6.One of our primary goals has been to gain the maximum utility from each target station. This has led to the present emphasis on target-cell configurations with more than one secondary beam per target. A new magnetic system that has been developed7 for multiple achromatic extraction of independent momenta (MAXIM) will allow us to maximize the number of channels per target with minimal compromises. MAXIM consists of two concentric annular bending magnets of opposite polarity centered on the production target (see Fig. 3). For this rotational symmetry and zero total magnetic flux, it can be shown7 that all charged particles emerge from MAXIM on an axis whose projection into the (horizontal) bending plane intersects the target. This is true exactly for all momenta. The momentum determines the deflection angle between the direction in which the particle is produced (production angle) and the direction in which it emerges from MAXIM (extraction angle). For small angles, the deflection angle is inversely proportional to the particle momentum.This extraction system allows particles produced in the forward direction to be extracted into beams at non zero angles for a given momentum. Particles of neighboring momenta are extracted at angles near enough to 0° not to significantly impair production rates or source geometry. An optimal layout consists of a central secondary beam, extracted at 0°, and two symmetrically arranged lateralEXPERIMENTAL AREA 'A' BLD'G13Fig. 2. LAMPF II slow-spill experimental area layout.Horizontal Cross Section Vertical Cross SectionFig. S. Schematic diagram of a MAXIM system for extraction of secondary beams from a production target.beams at 5° to 10° extraction angles. The central beam accepts particles of relatively high momentum and either charge sign, where the only condition is that the average momentum be typically at least 2-3 times higher than that of the lateral beams. The lateral beams are coupled only insofar as they carry particles of opposite charge for optimal flux. Equally charged particles may be extracted at the cost of relatively large production angles and corresponding rate reductions of 3 to 5 in one of the lateral beams.Thin-Target FacilityThe present thin-target facility at LAMPF is served by a thin target located in the switchyard upstream of Area A. This facility will be preserved at LAMPF II for use with the 45 GeV slow-extracted beam.Stopped ir-fj. Beams (/x0 .2)The two beams at ±120° for stopped x ’s and /i’s could be conventional achromatic channels or decay-in-flight or surface and cloud-muon beams. Large angle production is chosen for these beams because the forward-direction real estate is required for the higher-priority kaon beams and because large production angles have minimal affect on the yield of low-energy particles.Hypernuclear and Stopped K ± Beams (K0.8)These two very low momentum K  beams are extracted at ±11.5° through a MAXIM system. For these beams the MAXIM system provides a central momentum (0° extraction) of 0.5 GeV/c. Production angles vary from 0° to 5° for momenta between 0.35 and 0.8 GeV/c. K0.8 is envisioned as a three-bend channel with an achromatic output focus. An electrostatic separator and a mass slit are located after each of the first two bends. Overall length is about 18 m to the output focus.The momentum range has been chosen to meet the needs of high-resolution hypernuclear physics spectroscopy as well as to provide high rates of stopping kaons for a wide variety of studies, including K-mesic atoms and rare decays. Two stages of separation are needed to bring the 7r / i f  ratio below 1. A pion-rejection factor of 104 should be possible. A momentum resolution of about 0.1% can be achieved by momentum analysis accomplished with the aid of data from multiwire chambers before and after the last bending magnet. The chambers in these locations are downstream of the separators and will be exposed to beam rates that are manageable.General-Purpose K, p Beam (K6)This intermediate-energy beam of about 1% momentum resolution is extracted at 0° through a MAXIM system with production angles between 0° and 3° for momenta between 2 and 6 GeV/c and particles of either sign. The proton beam is separated from the secondary beam by a slight upward bend of about 12 mr before the target, producing a vertical separation of about 12 cm at the first15bending magnet after the extraction triplet and allowing the proton beam to pass undeflected through a v-shaped groove in the upper yoke of the bending magnet (Lambertson septum).The optics of this beam could also be based on a three-bend design with an achromatic output focus. An intermediate dispersed focus after the first bend would be used to set the momentum bite and trim the halo on the vertical image of the source. The separator section would be located after the second bend and the mass slit located before the final bend, so that background particles from the slit would be less likely to reach the final focus. With careful collimation and trimming of halo we hope to obtain a pion-rejection factor of 500. Another desirable feature would be two output legs, each with an achromatic focus.EPICS II Beam (K1.5)The current plan is that this channel should operate as an energy-loss spectrometer with a verti­cally dispersed beam on the scattering target. This design has two advantages:(1) Because momemtum resolution is not set by the channel momentum acceptance, a sizable acceptance can be used, resulting in higher beam rates.(2) The dispersion plane and scattering plane are orthogonal, thereby de­coupling momentum analysis from measurement of scattering angle.A momentum resolution of 0.05% and an acceptance of 5 msr is the goal for K1.5 and its spec­trometer (EPICS II). This would provide a kaon energy resolution of 285 keV at 0.7 GeV/c and 555 keV at 1.2 GeV/c.K1.5 is extracted at 5.7° through a MAXIM system. The corresponding production angles are 3.8° for 0.6 GeV/c, 0° for 1 GeV/c, and 1.9° for 1.5 GeV/c. Two stages of separation are needed for adequate beam purity. Channel length from production target to scattering target is about 25 m.General Purpose K, p Beam (K2.5)In the range of momenta from 1 to 2.5 GeV/c we anticipate the need for a general-purpose channel that can meet the needs of programs studying hadron-nucleon interactions and hadron-nucleus interactions where resolution of individual nuclear bound-state levels is not needed. A pion-rejection factor of 104 is expected for two stages of electrostatic separation. When this channel is used at its best resolution of 0.5%, beam rates will be down by a factor of 10 because of the required reduction in momentum bite. The extraction geometry is essentially the same as that of Kl.5. The production angles are 0° for 1 GeV/c and 3.4° for 2.5 GeV/c. For this beam, also, we imagine the optics to be based on a three-bend channel with an achromatic output focus and separator sections after each of the first two bends.High-Momentum Beam (K35)This beam provides K ± , 7r± , and p of the highest momenta available at LAMPF II. It is extracted at 0° through a double MAXIM system where the second pair of MAXIM magnets acts only on the proton beam and K35 but not on the lateral beams. The production angle is, therefore, also 0°. Separation of the secondary beam from the proton beam is achieved by injecting the proton beam onto the target at downward angles between 6 and 18 mr, depending on the charge sign and momentum of the K35 beam. Separation is enhanced by a vertical bending magnet between the elements in the second MAXIM magnet pair. The resultant K35 vertical displacement at the first bending magnet after the extraction triplet is sufficient for separation from a proton beam with an rms divergence of ±3 mr (multiple scattering in the target) by a Lambertson septum machined into the yoke of the bending magnet. This holds for K35 momenta up to 35 GeV/c for negatively charged particles and up to 30 GeV/c for positively charged particles. Separation of higher momenta would require longer separation distances.Separation by electrostatic separators is not practical at the momenta covered by K35. We assume that the necessary development of rf separators will occur and will be successful. Two output legs (ports) can be provided. In one we could have a large-acceptance unseparated beam of the highest possible intensity. The other output leg would be longer and configured to contain the rf separator section which is about 100 m in length.16Fig. 4. Neutral kaon (if£) flux densities at the target and rates at the exit of a 16-m- long K°  beam at 0° with 0.1-msr acceptance including absorption in the target and decay in flight. The top curve is the flux density at the target in units of particles per 34/xA of protons extracted from the accelerator (8/xA incident on the production target) per msr of solid angle, per GeV/c of momentum interval and for a production target of one interaction length of tungsten. The lower curve is the rate at the exit of the beam channel in particles per second per GeV/c of momentum interval.Test BeamsSimple test beams may be extracted to the right of the first downstream bending magnet after each target. Each would have momentum coupled to the K6 or K35 momentum by a fixed ratio.Neutral BeamsAll neutral beams are extracted from a third target, which is immediately followed by a beam dump. A magnetized shield is proposed to remove muons in about one-half the distance (i.e., in 14 m) required by ranging techniques. Figure 4 shows the flux density and rates.PRODUCTION CROSS SECTIONS AND RATESTo estimate particle-production cross sections at 45 GeV, we have fitted1 a collection of data from 10 to 200 GeV with a slightly modified version of a model put forward by Hojvat and Van Ginneken8 and used by them for estimating p production. In this model the invariant cross section for inclusive production from a nucleus is given by a product of three factors:E  v  F  F, o — a ’ r apa ‘ * nuc >°aba dp*where F, is the cross section, valid in the scaling region, for inclusive production by p-p collisions; F<,pj is an empirical “approach to scaling factor” that accounts for deviation from scaling at lower energies; and Fnttc is an empirical factor that accounts for modifications caused by nuclear matter.Reasonable fits were obtained to a large collection of data that, unfortunately, includes no data between 26 and 67 GeV for production from nuclei. We use the formula as a convenient means of summarizing available data and as a means of interpolating and extrapolating where there are no data. We judge that the formula is good to a factor of 2 for most of the region of interest at 45 GeV. It is less reliable for values of X r  >0.8 (radial scaling variable) and for very low energy secondary particles (below 500 MeV/c in the lab system).Figure 5 shows particle fluxes, corrected for absorption in the target and decay in flight, at the end of the secondary beams. Curves tire presented for 7T~’s, Ff± ,s, and p’s. Positive pion rates are typically less than a factor of two larger than the negative pion rates for secondary momenta less than 25 GeV/c.For completeness, a maximum rate is also shown which occurs when the full solid angle of accep­tance provided by the target extraction section (MAXIM plus first focusing elements) is transported by the secondary beam line.17Fig. 5. Particle rates at the end of the secondary beams for 34 fj,A of extracted proton beam, including absorption of both primaries and secondaries in the targets and decay in the secondary beam transport. Targets 1 and 2 are assumed to be 5 and 10 cm of tungsten, respectively. The dashed curves are for the available solid-angle of the channel when separators are used and the solid curves are for the maximum solid angle without separators.18PROBLEM AREASDrell-Yan FacilitiesA growing awareness of the value of studies of the quark structure of nuclei using the Drell-Yan process is suggesting a need for higher energy beams. We can imagine several possibilities for the site of an appropriate experiment with proton probes. The experiment might be placed in one of the long straight sections of the accelerator and used with an internal gas-jet target. It might be located in the transfer tunnel between the accelerator and the experimental area and used with a thin target. Or, a separate experimental area might be arranged in an extension of the staging building north of the planned experimental area. For probes other than protons, option 3 is the most reasonable if none of the presently planned beams are adequate. Further definition of the requirements for such an experiment are needed before the definitive site can be determined.Thermal Energy DepositionA target of one interaction length will cause a sizeable fraction of the incident beam power to be scattered into nearby components as various reaction products. These particles, in turn, will deposit their energy as heat and radiation dose. To illustrate the heating or dose rate, we have used the Fermilab code CASIM9 to calculate the power deposition in an idealized but relevant model consisting of a 150-mm (one-interaction-length) copper target surrounded by a 2-m-thick spherical iron shield beginning 1 m from the target. A conical hole 50 mm in diameter at the start of the shielding was provided to pass the uninteracted primary beam. In the model calculation about 40% of the incident beam power, or 600 kW, is deposited in the shield. For comparison, about 150 kW are scattered into the target-cell components by the current A-2 target at LAMPF. Active measures such as water cooling of components and close-in shielding are required to remove this heat from the target cell; free convection will not be sufficient. Furthermore, the high radiation dose requires radiation-hardened materials and designs for these components.TargetsThe optimal secondary-beam production target would be approximately one interaction length of the highest density material that would have a useful lifetime in the full-intensity primary beam. One interaction length is a good balance between the conflicting goals of having most of the primary protons interact to produce secondary particles and having none of the secondaries lost to interactions in the target. High density means shorter targets and reduced depth-of-field effects in the secondary beam-line optics. More fundamentally, high target density implies high source brightness, which is a most desirable starting point for producing high-quality beams.High matter density also brings the problem of high density of energy deposition and the attendant cooling problems. Two basic approaches to target cooling are being considered and both appear feasible with proper selection of materials and careful engineering. A stationary, water-cooled finned target or a radiatively-cooled rotating target are the leading candidates. A third, more conservative alternative would be a rotating water-cooled target.Targets for the fast-extracted beam areas face an additional complication of thermal shock from the ultra-high-peak power loads of a pulsed beam whose pulse length (~ l/is) is about five orders of magnitude shorter than the slow-extracted beam pulse length. In this regime the temperature change, hence thermal expansion, develops in a short time compared to the time it takes sound to traverse the beam spot. This leads to coherent stress waves that can constructively interfere to produce high stresses in the target. The phenomenon is now encountered in neutrino and antiproton production targets at CERN and BNL. The peak intensity from the LAMPF II booster is comparable, but the time-averaged beam is two orders of magnitude higher. Thus we will have comparable dynamic thermal stresses, but they will be superimposed on a much more severe steady-state load. Accurate modeling of dynamic thermal-stress phenomena for the LAMPF II neutrino target is a major effort that has not yet been attempted.Particle SeparatorsElectrostatic separators rapidly lose effectiveness as the secondary beam momentum p is raised. The maximum separable phase space in the separation phase plane falls as p 3 for p >>M . As the19momentum is raised, the maximum separable phase space becomes much smaller than could otherwise be transported. This effect sets a practical limit of around 6 GeV/c for K  and 10 GeV/c for p beams, although in principle very small phase-space (therefore low-intensity) beams of higher energy can be separated.Conventional rf separation techniques do not suffer from the same rapidly falling momentum dependence because the crucial transverse deflections of the two species are in opposite directions rather than in the same direction, as is the case with electrostatic separators. Thus they are capable of providing ir — K  separation at all the higher energies available at LAMPF II. They do, however, suffer from two limitations:(1) limited acceptance, because of the small beam apertures allowed in high- frequency cavities; and(2) short duty factor, because of the very high power (10- to 20-MW) rf source required to achieve high fields in room-temperature cavities.The geometrical limit on acceptance is tolerable, even though it is considerably less than can otherwise be transported by a beam channel optimized for transmission.More serious for LAMPF II is the duty-factor limitation, which falls short of the needed cw operation by two or three orders of magnitude (for room-temperature cavities). The development of reliable, high-field, superconducting microwave cavities would change the picture dramatically. The rf power-source requirements would become quite manageable since beam loading is negligible.CEBAF/Cornell development has produced reliable accelerating superconducting cavities with average gradients of 8 M V/m and maximums of 15 MV/m. This apparent nearly order-of-magnitude improvement over previous attempts, suggests that the needed rf cavities will be feasible for transverse mode separators.FUTURE DIRECTION OF EXPERIM ENTAL AREA DEVELOPM ENTTechnical uncertainties in target design, high-duty rf separators, and a number of issues related to innovative secondary beam design require additional in-depth research and development. We anticipate that the evaluation and development of high-field, high-duty-factor, superconducting rf separators will take considerable time and effort to come to fruition. Innovative beam design is especially needed for high-resolution, low-energy kaon beams required for nuclear physics with kaons. Although this problem has attracted the attention of a number of workers, a stronger effort is clearly warranted. A MAXIM system should be prototyped, mapped and tested to ensure its performance since it is the crucial element that allows 3 forward-angle, high-energy beams from a single target.REFERENCES1. The Physics and a Plan for a 45 GeV Facility that Extends the High-Intensity Capability in Nu­clear and Particle Physics,” Los Alamos National Laboratory document LA-10720-MS (May 1986).2. P. Eberhard et al., Rev. Sci. Instrum. 31, 1054-1063 (1960).3. M. Bell et al., “rf Particle Separators,” Proceedings of the International Conference on High Energy Acceleration, Dubna, USSR (Atomizdat, Moscow, 1964), p. 798.4. G. Bunce, “AGS Beams,” Brookhaven National Laboratory report 50874 UC-28 (May 1978).5. A. Yamamoto, “Study of Low Energy Intense Kaon Beam,” National Laboratory for High Energy Physics in Japan (KEK) report 81-30 (1981).6. D. George, R. Abela, and D. Renker, “Half Quadrupoles for Use in a High Radiation Environ­ment,” to be published in the Proceedings of the Ninth Magnet Technology Conference (1985).7. C. Tschalar, “Multiple Achromatic Extraction System” (submitted to Nucl. Instrum. Methods, September 1985).8. C. Hojvat and A. Van Ginneken, Nucl. Instrum. Methods 206 , 67-83 (1983).9. A. Van Ginneken, “CASIM Program to Simulate Transport of Hadronic Cascade in Bulk Matter,” Fermi National Accelerator Laboratory report FN-272 (1975).20TRIUMF's EXPERIMENTAL AREA PLANSJ.L. Beveridge TRIUMF, 4004 Wesbrook Mall, Vancouver, B.C. V6T 2A3INTRODUCTIONExcept for the performance of the accelerator and the scientists us­ing it, there is no other aspect of a laboratory more important than ade­quate space and facilities to carry out the experimental programme. Unfortunately, although representative experimental programmes can be put forward, it is not possible to predict the exact nature or emphasis ofthe physics which will actually be pursued through the lifetime of theKAON factory. This implies that a high degree of flexibility will be an important criterion for the design of the experimental facilities and areas so that changing requirements of the physics programme can be ac­commodated in a reasonable fashion. In this spirit TRIUMF's present experimental area plans represent only one choice out of a number of pos­sible options (none of which have been examined in sufficient detail to establish a clear indication of relative merit) and are therefore subject to change without notice.There are, of course, a number of properties of the secondary beamsto be produced at a KAON factory which will require optimization inde­pendent of the details of the physics programme and beyond the increasein raw particle intensity anticipated with the increased primary protonenergy. These include:• beam purity• beam brightness® energy resolution• stopping densityc polarization• appropriate beam availability i.e. intermittent and low intensity for set-up and testing - reliable high intensity for data-takingThese properties, associated with quality of the beams, are often crucial to the success of any given experiment and will have to be balanced against beam intensity. Again, flexibility is required in the approach to these design parameters so that final decisions are driven by current physics necessity rather than impressive (and often expensive) technological capabilities.The design of the production target areas of a KAON factory will ultimately be one of the most important factors in the success of the entire facility. These areas, once constructed and irradiated, will be extremely difficult to modify and will therefore determine a fundamental aspect of the facilities' capabilities. It is in these areas that the flexibility of response to physics demands will first be removed in the experimental areas. These areas will therefore require our utmost atten­tion and innovation in their design. However, the areas outside the immediate target shield will remain amenable to change and modification, and a serious effort should be made to avoid unnecessary or highly limit­ing constraints in these areas. In our present design of the experimen­tal areas we have tried to emphasize this aspect of the overall design21and have placed less emphasis on the optimization of the areas immediate­ly downstream of the production targets. We have done this in recogni­tion of the fact that eventually our only real flexibility in responding to current physics priorities will be in areas outside the immediate target shield.It is clear from the representative experimental programme and from our desire to respond flexibly to future physics requirements that the complete range of facilities possible at a KAON factory should be con­sidered at this time. These include:® a complete set of secondary charged particle channels of overlap­ping momentum capabilities up to the maximum momentum available® a neutral kaon beam• a neutrino facility• high-energy polarized proton beams• low energy ir, y channelsAt present we have considered the first three of these and have not ex­plored in any detail the possible exploitation of polarized proton beams or low energy it , y beams.REFERENCE DESIGNSIn order to examine a possible layout of a comprehensive set of secondary channels we have done reference designs for most of the chan­nels anticipated. The parameters of these channels are given in Table I and layouts of the beams designed for 0.7, 1.5 and 6.0 GeV/c are shown in Figs. 1, 2 and 3, respectively. All the reference designs have assumed a zero degree take-off angle (although the lower energy channels would be compatible with take-off angles up to 10°) and beam purity has been a primary design criterion. The sources of pion contamination in kaon channels have been studied at TRIUMF using the Monte Carlo program REVMOC. At low momentum (~1 GeV/c) pions come predominantly from parti­cle decay near the production target and from scattering from magnet poles. At higher momenta contamination is mainly a result of insuffi­cient separation at the mass slit. The reference channel designs are aimed at a contamination level an order of magnitude better than present day channels.Table I. Secondary beam line parameters.Beam Line Momentum Range Solid Angie T.keoH Momentum Length Length of(G eV /c) (msr) Angle (Deg.) Acceptance (%) (m) SeparatorKlK2K40.4 - 0.7 0.7- 1.5 2.0 - (6 GeV/c)100053.8317.929.71151 x 3.0 m 2x3.0 m 1 x30 mK0.30 (3 GeV/c)0.5 -10.0 0.03 6 Wide 20K3 1.25-2.5 .5 (2.5 GeV/c) 0 4 54 2x7.5 m dcK6Muon2.0 (1.5 GeV/c)up to 20 GeV/c 30-200 MeV/c0.1635013581046.318unseparated 1 x 1.5 m rfor 3 m dc22Fig. 1. Proposed arrangement of the low-energy/stopping beam channel Kl.Fig. 2. Proposed arrangement of the low-energy separated beam beam channel K2.6 G eV /c KAO N B E A M LIN E  K 4  LE N G T H  115 m0 6Q 7  0 8 0 9  QIOQII“0,B  B4QIQ2 »o_A - - 1  ---------------------------------------------1 MASS B3O H  8 2  SEP ARATO R 8Fig. 3. Proposed arrangement of the medium- energy separated beam channel K4.The 1.5 and 2.5 GeV/c channels are of similar design and will have two-stage separation. A mass slit after the first stage removes pions coming directly from the production target and defines the beam spot of pions from secondary sources seen by the second stage. These pions are removed after the second separation stage at the second mass slit. Fol­lowing this mass slit three combined function magnets form a first-order achromatic image of the slit on an experimental target. For the 6 GeV/c channel only one separation stage is contemplated. The two-stage separa­tion approach does not seem practical since each stage requires about 30 m of separators. Above 6 GeV/c, for kaons, the utilization of dc separa tors is not expected to be a practical alternative and at present only a simple unseparated beam has been investigated. It is possible that superconducting rf separators may be developed in the future which may make particle separation at these energies feasible. Space has been left at the end of the unseparated channel to allow such possibilities.The 0.7 MeV/c beam is a special design in which the emittance of the secondary particles accepted from the production target is defined by a 6.9 m long front-end section. The separator section images this beam onto the mass slit. As for the 1.5 and 2.5 GeV/c channels a first order23Fig. 4. Arrangement of the pion horns and decay channel for the proposed broad band neutrino beam.achromatic image of the mass slit on the experimental target is formed by three combined function magnets.In addition to the above charged particle secondary channels, a reference design has also been examined for a neutral kaon beam. This design assumes a 6° take-off angle and a total length of about 20 m. The beam is defined by a collimator which passes initially through a 10 m long clearing magnet to remove charged particles and then steel shielding to provide a total of 17 m of steel shielding equivalent. An option of using 15 GeV/c ir- to produce a tertiary beam of K°'s was also considered; however, although this method produces very clean beams the flux expected appears to be approximately 300 times lower than the secondary beam.No specific designs were done for low-energy muon beams although it is fully expected that such beams will be installed at the KAON factory as intensities of negative muons 20-50 times those available at present meson factories are anticipated. The production of low-energy particles required for such channels is essentially isotropic and high acceptance channels could be installed at one or more of the production targets with take-off angles of 90-135°. There are a number of possibilities to be investigated in this area including surface muon beams, superconducting solenoid channels and pulsed muon beams from the production target on the fast extracted beam. The channel designs are likely to be conventional and should be easily incorporated into the facility design.A neutrino facility capable of taking full advantage of the full in­tensity fast extracted beam will clearly be a required contribution to the proposed physics programme. The proposed pion horn system (Fig. 4) is very similar to that used in the broad band neutrino beam at Brook- haven, and its performance has been calculated using the program NUBEAM. Flux gains due to such a pion horn are expected to be about a factor 10. It should be possible to operate the horn elements at the required fre­quency of 10Hz; however, development work will be required to establish this. Alternative designs of pion-focusing devices have been studied such as superconducting quadrupoles and solenoids, butterfly dipoles and lithium lenses. Lithium lenses or current-carrying targets could provide some promise but it is not expected that superconducting devices will be operable in the immediate environment of a production target. Further developments at Fermilab and CERN will certainly influence the final design of this important facility.24EXPERIMENTAL AREA LAYOUTThe layout of the slow extracted beam area was determined by distributing the above reference channel designs in a variety of ways with the following conservative assumptions:1. Each production target station would have at most two forward take off channels2. Each production target will be provided with a proton beam with intensities up to 100 pamps.It was realized that these assumptions could be overly generous in terms of shielding and targeting costs and also beam splitting difficulties and associated costs. However, we decided to produce a design based on these assumptions and then change and/or compromise as was appropriate in the future.The experimental area proposed for the KAON factory is shown in Fig. 5. This particular layout of the representative channel designs which anticipates four proton beams was chosen primarily for its flexibility at the experimental ends of the channels. Experience at TRIUMF and other facilities has shown that the provision of a second experimental area at the end of a secondary channel can be a significant advantage in terms of scheduling and efficient utilization of the beam. This layout provides sufficient separation of the experimental areas and sufficient surrounding space to allow each of the secondary channels to be "Yed" at the ends without interference with other experimental areas. This feature of large open areas at the end of the channels provides a flexibility which may prove crucial to a successful response to changingKl 0 .7  GeV/c20 m7 5  mS E R V ICE  A N N E XFig. 5. Proposed layout of the experimental area.25physics requirements. Another feature of the layout is that it is easily coalesced into two proton lines each servicing two serial targets or, with sufficient advance planning, could be amenable to staged construction.The present remote handling concepts assume that a reasonably com­prehensive set of hot and warm cells will be provided within the experi­mental hall and that transport to these cells will be achieved with the overhead crane. The hall has therefore been located so that the neutrino target/horn area is contained within the building to provide access to these facilities. The location of the hot cells has not yet been deter­mined but would likely be in the area between the 0.7 and 2.5 GeV/c channels or in the upper right corner of the layout of Fig. 4.The design and position of the low energy muon beams has yet to be determined for the present experimental area layout. The choice is extremely flexible at this juncture as backward angle beam lines could be reasonably attached to the neutrino target and to three of the four proposed production targets. The present thinking is to create a pulsed and continuous beam muon facility between the "A" beam target and the neutrino target however the details of this proposal have not been investigated.C O N C R E T EFig. 6. Combination of a low momentum channel K1 with a high momentum channel K4 using the same production targetAs mentioned in the introduction, it is recognized that the front ends of the secondary channels will be of extreme importance to the performance of the overall facility. We have therefore done a more detailed design of the area downstream of the production target on proton line ”C." This area is shown in Fig. 6. The 0.7 and 6.0 GeV/c secondary channels are the reference designs given above and are separated by two bending magnets which provide a doubly achromatic non focusing transport system equivalent to a drift space for the higher momentum secondary particles and the proton beam. This solution to the design of the front26end of the channels is a straightforward application of conventional techniques and is dependent on the coupling of two channels of substan­tially different momenta. Such couplings are possible within the content of our present overall experimental area design and although changes to this design may force substantial changes to the details of the arrange­ment downstream of the target the general features will remain very much the same. The area depicted in Fig. 5 has therefore been used as a model of a typical target cell for the investigation of problems associated with shielding, beam dumps and power deposition in elements downstream of production targets.EXTRACTED BEAMSThe proposed layout of the extracted beams to the experimental areas is shown in Fig. 7. The fast extracted beam from the Driver ring would be shared on a pulse by pulse basis with part of the beam sent to the Extender ring to produce the slow extracted beam and the remainder going directly to the neutrino area. It is anticipated that the amount of beam in either area would be variable from 0 to 100% of the circulating beam. The splitting of the slow extracted beam into the four primary proton lines anticipated in the experimental areas is presently proposed to be done by first separating the beam in two by a combination of rf and dc separators followed by a Lambertson septum magnet. These two lines would be further split using kicker magnets operating on the 10 Hz rate of the main ring. This proposed splitting scenario has the disadvantage that the beam must be shared equally by the initial two lines (or one line takes all the beam) due to the rf separator. However, alternatives considered to date would potentially produce unacceptable levels of residual activity in the beam switchyard area and have been rejected on this basis.Fig. 7. Layout of the extraction area and external beam lines.27SUMMARYThe plans for the experimental areas for the KAON factory to be built at TRIUMF represent only one of a multitude of solutions to a com­plex and many faceted problem which has no predictable optimal solution. The proposed set of experimental facilities maintains a large degree of flexibility to respond to new ideas and changing demands while allowing problems associated with the nature of the facility to be addressed in detail appropriate to the present development of the project. The final solution (i.e. the facility that is finally constructed) is likely to be substantially different from the one proposed here as it will be constrained by budgetary considerations and hopefully driven by new directions of physics priorities and new ideas in facility design and technology. The present proposal however represents an excellent starting point which is capable of incorporating if necessary constraint but more importantly new ideas and direction.28BNL KAON BEAMS, PRESENT AND PLANNED*D. M. Lazarus AGS Department, Brookhaven National Laboratory Upton, New York 11973ABSTRACTAt the present time, there are three electrostatically purified (separated) kaon beams in operation at the Brookhaven AGS. They are optimized for .8 GeV/c, 1.1 GeV/c and 6 GeV/c. A fourth beam has been designed for a maximum momentum of 2.0 GeV/c. It features two-stage purification and higher order optical corrections. Anticipated per­formance figures for the beam, which is capable of delivering more than 106 kaons per spill, are given, as well as measured performance figures for the existing beams. Possible purification schemes for momenta above 6 GeV/c are discussed. Performance data from neutral and unpurified beams are presented.INTRODUCTIONThe experimental areas at the AGS are illustrated in Fig. 1. Elec­trostatically purified beams for kaon and antiproton research are the LESB I (C2, C4), LESB II, (C6, C8) and the MESB (B2, B4). An unsepara­ted 6 GeV/c beam, D6, originating at the D target station and two neu­tral kaon beams A3 and B5 are also active in searches for flavor chang­ing neutral current prgcesses manifested in the yet to be observed de­cays K+ ->• TT+y+e- and -»■ pe. The remaining beams are high energy un­separated beams typically used for pion reactions above 10 GeV/c i.e., gluonium and exotic spectroscopy, polarized protons and in the future for the transport of heavy ions. A stopping muon beam is also avail­able. The proton extraction to the North Area takes place over one turn and the 12 beam bunches are targetted for neutrino (pion) production in 2.75 psec with a 1.4 second repetition rate. In this mode the beam retains its 30 ns bunch structure whereas in the slowly extracted beam mode a one second debunched spill is available with a 2.4 second repetition rate.PRESENT PURIFIED BEAMSSpecifications for the three purified (separated) beams are given in Table I. Typical fluxes per 2 x 1012 protons on target along with measured purities are given for kaons and antiprotons. Each of the low energy beams has two branches one momentum recombined and one dispersed. This has proven to be very efficient allowing for experimental construc­tion in one branch while the experiment in the other branch is running and permitting rapid turnaround from one branch to another in one or two shifts. All three beams use electrostatic beam separators with heated glass cathodes and stainless steel anodes. Conditioning takes place under hard vacuum and operation with 10-1+ Torr N2. Power supplies areWork performed under the auspices of the U.S. Dept, of Energy.29Figure 1 AGS East Experimental Area30TABLE I.C2,C4 Low Energy Separated Beam (LESB I)RangeAp/pProduction angleaLw“/K“ - 104 ir/p - 60 @< 1.1 GeV/c± 2210.5° ± 1.65°2.6 mar 15 m .75 GeV/cP.rticl««/2 x 1012 Proton*PGeV/c TT- K" P TT+ K+ P0.400.751.106xl072xl08lxlO88008.0 40K 1.4x10“154xl031.4x10“8xl072xl081.6xl082xl032xl05lxlO61.6xl07 4x10 7 8xl07Notes:1. The momentum dispersed beam C2 spot size is 1/2“ ray trace program.2. A degrader can be used at the C target to increa momentum.3. Separated K”/ir” ■ 0.09, p/ir” ■ 0.017 at 0.75 GeVV)xl-1/4"(H) from a 3e the p flux at lowC6,C8 Low Energy Separated Beam (LESB II)RangeAp/4»0Lir/K“ - 16,it7p - 100 @< .8 GeV/c ± 32 15 msr 15 m 7 GeV/cParticlei/2 x 1012 proton*PGeV/c 7T~ K“ P x+ K“ P0.4 > 108 4xl03 800 > 108 1x10“ > 1080.6 > 108 1x10s 4xl03 > 108 2xl05 > 1080.8 > 10s 4xl05 1x10“ > 108 lxlO6 > 108B2,B4 Medium Energy Separated Beam (MESB)RangeAp/pProduction Angle 0 L1.5-6 GeV/c (K)1.5-9 GeV/c (p) t 323° ± 0.4°0.3 msr, Mode I; 0 81 m1 msr, Mode IIIPartldes/2 x 1012 proton*PGeV/c TT” K“ P TT+ K~ P1.5 lxlO7 600 2x10“ 2x107 1200 2xl073 3xl07 1x10s 1.6xl05 6xl07 2xl05 3xl076 6xl07 2x10s 1.6xl05 6xl07 5xl05 9xl07Measured K"*/tt” - .25 at 5 GeV/c.Measured p/tt~ - 7% at 2.5 GeV/c, 50% at 6 GeV/c.are oil insulated and are generally situated some ten meters away on top of the shielding. A significant improvement in performance would be ex­pected from using the KEK type separators described by Dr. Yamamoto at this meeting. The basic optics vary only in detail between the beams.LESB I1 (Fig. 2) uses a septum magnet to reduce the production angle from 18° to 10.5° allowing two 0°, high energy beams to share theFig.  2a. Beam Layout.APPROXIMATE FOCAL PLANE (HORIZ.IDl 01 02 02 SEP 03 04APPROXIMATE FOCAL PLANE (VERT.) MASS SLIT 05 03—  X o 'O .X i *0 .0 6 , A  p /p  * 0—  X *0 ,X '„ * 0 ,A  p /p  *0 .0 20 oV ° ' x 0 *0 -0 l,A p /p *0 .0 2—  X -0 .X ' *0 ,A p /p * 0 .0 20630 60  90 120 150 180 210 240 270 300 330 360 390 420 450 480 510 540 570 600mmy/AMASS SLITA  p /p *0 ■0,V' «O.OIO,Ap/p *0 .020  0—  Yo *0 ,Y ' *0 .010,A p /p  *-0 .020v e r t ic a lFig . 2b. Typical  ray trace for Beam F2.32production target. A quadrupole doublet is tuned to give nearly par­allel vertical optics through a dipole magnet and the two meter long separator which typically operates with 550 kV across the 10 cm gap. Another doublet then produces a dispersed double focus at the mass slit which is rotated about a vertical axis giving first order compensation for vertical chromatic aberration. A final quadrupole doublet with the last dipole in between yields the final dispersed (C2) or recombined (C4) focus. The dispersed branch has formed the kaon spectrometer sec­tion of the hypernuclear spectrometer for the past several years. The recombined branch will provide a stopping K+ beam for a BNL/Princeton/ TRIUMF collaboration searching for K+ r+vv or ir+ x°. Table II lists some of the physics topics addressed in the LESB I, which is now in its fifteenth year of operation.TABLE II. LESB I Experimental ProgramLESB I Experimental Program —  1970 - PresentI exotic atoms, magnetic-moment measurementa (total) for (K p), K+p), (pp)} S-meson reported(K-d), K+d), (pd)E- beta decayPolarization and da/dfi for (K+p), (K“p), (K+n)K p K°n a (total) and da/dftK~p backward elastic do/dfi (180°)K_p -► ttE“ da/dft (180°)pp backward elastic da/d£2 (180°)pp nn S-meson revisiteda (total) ppa (annih) pp —K+ -»■ ir+ + vvHypernuclei - 12C, ^C, xJn , ^0, ?Li , 9Be, 16Q j 16Cj 6RK” nuclear scattering 12C, 1+0CaHypernuclear lifetimesStrange Dibaryon searchesLESB II2 (Fig. 3) is a larger aperture, lower momentum beam using a pair of one meter separators and sextupole correction of chromatic and spherical aberrations. A quadrupole doublet subtends a central produc­tion angle of 5° and produces a beam that is nearly parallel in the vertical plane through the first separator but horizontally focussed at the sextupole magnet between the two separators. The 45° sector magnet between the first two quadrupoles disperses the beam at this point al­lowing for correction of chromatic aberration. The two quadrupoles between the beam separators produce a vertical focus at the mass slit330,D’ BS1 „- ,_r -> i -< ,03 u<t\ , . ji; - —  i! ii ji IV'-v. ■ - :■ U hs S  V '/  .C TA^GT 1Fig. 3. The LESB I II  ^and contain the beam horizontally so that the final doublet and 45° sector magnet bring the beam to a dispersed (C6) or recombined focus (C8). Currently an experiment studying hyperon radiative decays involv­ing a Boston/Birmingham/UBC/Budapest/Case Western Reserve/TRIDMF group is setting up in the recombined branch. Table III lists the experiment­al topics investigated in LESB II. The performance of this beam has been somewhat disappointing. The flux increase over the LESB I was initially a factor close to seven but because of a number of factors having to do with beam purity it was necessary to reduce the acceptance to 55 msr %, a factor of five greater than LESB I. The ir-/K“ ratio is presently 16/1 at 680 MeV/c compared with 10/1 in LESB I. Some improve­ment was gained by introducing a collimator in the aperture of Q2 and reducing the gap of the first beam separator from 15.25 cm to 12.7 cm allowing a higher electric field gradient. Other improvements have involved use of the short sextupole just downstream of the first sector magnet to correct spherical aberration and the introduction of the quad- rupole doublet between the beam separators which gave a thirty percent reduction in the vertical image size at the mass slit. The performance is presently compromised by two unforeseen difficulties. The beam sepa­rators which had operated with gradients of 50 kV/cm and 60 kV/cm re­spectively in the laboratory failed to exceed 43 kV/cm and 55 kV/cm respectively in the beam. Reducing the gap of the first separator to 12.7 cm brought the performance back to 47 kV/cm but the performance of the second separator has deteriorated to 40 kV/cm. Compromises and problems in the proton transport to the target station have lead to the use of a 5.0 mm high target, twice the object size assumed in the LESB II design. Recent improvements in the proton transport lead us to be­lieve most of the beam can be targeted on a 2.5 mm target and the first operation with this target is going on right now. It is hoped that the v~/ k~ ratio will drop from 16/1 to nearly 10/1. Further collimation is also planned but at present no funding is anticipated for a beam sepa­rator upgrade.The Medium Energy Separated Beam (MESB)3 which is shown in Fig. 4 uses a septum magnet to reduce the production angle from 6° to 3° sub­stantially improving the high energy kaon and antiproton fluxes. A novel feature is the quadrupole triplet front end which allows a vari­able vertical focal length so that the vertical acceptance and image size at the mass slit can be decreased with increasing momentum so that a constant image separation factor, can be maintained. Chromatic aber­rations are corrected by sextupole magnets as in LESB II. Once again, however, the large horizontal magnification leads to poor resolution and34L HjhV  .> ■ iu  MACM m//’•-Fig- 4. The MESBTABLE III» LESB II Experimental Program —  1978 - Presenta (annih) pp, np pp -*■ Y's pd -*■ (pn) + p♦ (pp) + nE+ ♦ pTr°; Z“ nir- asymmetry parameters in polarized I decayX n X _____________________________hence imperfect correction. Both branches (B2 and B4) can be recombined as a second horizontal focus is employed in the B4 branch. The B2 branch provides beam to the Multiparticle Spectrometer (MPS), a large solid angle spectrometer user facility operated by an in-house group both for their own research program and for users such as the BNL/Fla. St/S.E. Mass/Indiana group that has recently claimed to have unravelled the E-i riddle. The beam is capable of separating kaons with a separa­tion factor of two up to 6 GeV/c. The antiproton capability is limited by the magnetic field available in dipole magnets to about 9 GeV/c. Physics topics addressed are listed in Table IV.TABLE IV. MESB Program 1974 - PresentK* p++ ^ ( 1 4 2 0 ) ^  oT (K"p) oT(K"d) oT(pp) aT (pd)p p  -*■ TT- TT+} comparisonir+p -*■ Tr+ppp Charm searchK"p -*• 5 resonancesK“p+ in Multiparticle Spectrometer35FUTURE PLANS FOR PURIFIED BEAMSAt present there are three candidates for future purified beams. The first is a 2 GeV/c two stage electrostatically purified beam1* shown in Fig. 5 which would cover the range between the LESBs and the MESB (see Fig. 6) and would provide a unique facility capable of providing in excess of 106 K~ per machine cycle with a high degree of purification. Searches for bound, doubly strange, six quark states predicted in the bag model motivate the construction of such a beam but although a com­plete conceptual design exists its funding is at this time uncertain.Two features have been incorporated into the design which hold promise of greatly improved performance. The two stage separation should dis­criminate against pions resulting from K°and hyperon decays near the target. Pions of this type which are within the acceptance of the beam have, in general, trajectories which do not optically originate within the volume of the production target but rather from a diffuse cloudFig. 5. The proposed 2 GeV/c purified beam.surrounding the target. The additional constraints on velocity and trajectory imposed by two separator stages and two mass slits should greatly reduce this pathological component of background as illustrated in Fig. 7. The additional length for decay introduced by the second separator stage results in a 2-5 fold decrease in kaon flux over the range of the beam. Since the beam is still capable of delivering in excess of 106 negative kaons per pulse, the investment in enhanced purity is thought to be worthwhile.The correction of vertical aberrations through 3rd order has also been calculated to give a major improvement in beam purity. This has been carried out in the design by small pole displacements in seven of the eight quadrupoles of the separation stages of the beam. With these36improvements a contamination of pions plus muons of around fifty percent is anticipated. First order beam optics for the 2 GeV/c beam are dis­played in Fig. 8. The simi­larities to the existing beams is apparent. The ef­fect of the higher order corrections on the quality of the vertical images at the mass slits as shown in Figs.9 and 10. Calculated per­formances are given in Table V.Fig. 6. K- Flux at AGS.TABLE V.2 GeV/c Purified Beams (Proposed)Range 1-2 GeV/c Target6 cm PtProduction Angle 5°Length 31mft 1.3 msrAp/p 6ZEstimated particle fluxes/2 x 1012 incident protonp GeV/c K+ FT p1.0 4.2 x 10** 2 X 10^ .5 X 1061.5 4.3 x 105 2 X 105 1.0 X 1062.0 2.2 x 106 1 X 106 1.1 X 106Looking farther into the future to the operation of the AGS II, it would be desirable to have a high intensity, highly purified beam oper­ating above 6 GeV/c for spectroscopy in the 1—3 GeV mass range with projectiles carrying strange quarks. Such a beam coupled to a large solid angle spectrometer such as the MPS would give a new life to37searches for hitherto unobserved states. In the MESB four electrostatic separators, each 4.5 meters long with 400 kV across a 5 cm gap, provide an image separation for pions and kaons of less than 2 mm at 6 GeV/c.The 1/p3 dependence of separation by electrostatic means precludes its use at higher momenta. Radio frequency separation has in the past been employed in bubble chamber beams with high power rf pulses applied to deflecting structures for microseconds when a rapidly ejected beam was passing through them. In addition to the poor duty factor available for counter experiments, simple two deflector systems operating at fixed frequency only provide separation at several isolated momenta.Fig. 7. Discrimination against cloud pions by two mass slit system.KAONPROD.TGT.01 01 02 El' 3 0 °1+ '  1 . j03 0 4  HS□i0 5  Q6 E2 Q7 0 8  D2CM I CM2a —MSIrJ: l+l OJ o oCM 3 CM4_  TGT 0 -  (31.1 mlMS2M E T E R SFig. 8. First order optics for proposed 2 GeV/c purified beam.A method employing an alternating gradient sequence of quadrupoles interspersed with rf deflecting cavities was studied at BNL in 1970.5 6 In their rest frame particles see a travelling deflecting mode wave whose phase slips at a frequency that resonates with the betatron oscil­lation frequency. The oscillation amplitude of one particle species38grows while unwanted particles which are not resonant because of their different velocities receive smaller deflection. A stopper can be used to remove the unwanted particles. By tuning the betatron frequency, separation can be continuously obtained over a wide momentum range.This method of separation is also attractive because of the large phase space that can be separated. Unfortunately it is also expensive. Superconducting rf cavities provide another possibility for c.w. par­ticle separation.- 3 - 2 - 1  0 I 2 38 P ( % )Fig. 9. Effect of higher order corrections on vertical image at mass slit 1 of the proposed 2 GeV/c purified beam.Almost 25 years ago, A1 Maschke proposed modulating a beam by de­flecting the protons that generate it on and off the production target with an rf deflecting cavity so that separation might be accomplished with a second downstream cavity. Last year, Gerry Bunce suggested an interesting variation on this idea in which the deflection is across an electrostatic septum in the beam switchyard.7 Although 28.5 GeV/c pro­tons are rather hard to deflect, the septa presently used to split the beam are made up of .08 mm wires. This produces two modulated beams, each of which can then be purified with one additional deflector. In­deed, more than one septum can be employed, and more than N beams can be purified by N+l rather than 2N deflectors. Operation at 2.86 GHz would produce a time structure at .3 ns which would not be a problem for most experiments. Superconducting deflectors would be necessary for c.w. operation in this mode and would give enhanced performance in the beta­tron resonance scheme.Before leaving the subject of purified beams it is worth noting that because of their stability, antiproton beams can readily be puri­fied by making them sufficiently long to allow most of the pions to39decay.8 An antiproton ring such as LEAR is the logical extension of this idea. Indeed, if time structure is also present the pions and antiprotons can be separated in time as well.9 Such a beam is under consideration at Brookhaven for the study of threshold hyperon produc­tion and polarization as well as for charmonium spectroscopy.Table VI gives the specifications for Beam D6, an unpurified, 6 GeV/c beam transporting 2 x 107 positive kaons per pulse (along with some 5 x 108 protons and positive pions) to a Yale/BNL/Washington/SIN search for the flavor changing neutral current process K+ ir+y+e“ at the 10 ^  level. With AGS II, an upgraded apparatus and a more intense but highly purified beam, this experiment might push to the next order of magnitude in sensitivity.S P (% )Fig. 10. Effect of higher order corrections on vertical image at mass slit 2 of the proposed 2 GeV/c purified beam.TABLE IV. MESB Program 1974 — PresentK1 p++ A U J O ) 1?ot(K p) Oj,(K d) aT (pp) aT (pd) pp + Tr-ir+ir+p -*■ TT+p  ^ comparisonpp Charm search K~p -*■ E resonances K“p+ in Multiparticle Spectrometer S+ PY H° I°y E° + Xy H“ -*■ E~y * S”Y40Finally Table VI also gives performance figures for the two neutral beams at the AGS, A3 and B5 which are both being used to search for the decay, Kg + pe.TABLE VI.D6 Charged Kaon BeamProton Intensity of Target 1012TargetRangeA p /pProduction angle12LPPP1 cm x 1 cm x 7.5 cm Pt < 6 GeV/c .15 2 °.5 msr 15 mp(GeV/c)363.lxlO8 2.5x108K+2xl08 2.5xl081.4xl072.107 1.6xl08 108K"1 0 '9xl06 1.2xl0610*A3 Kg Neutral BeamProton Intensity on Target 1012 pppA3 Target 5 mm diameter 23 cm CuProduction Angle 0°12 45 psrL 12 mParticles/1012 incident protons M Kg2 x 109 5 x 107B5 Kg Neutral BeamProton Intensity on B1 Target 5 x 1012 pppB' Target 2.5 mm x 2 ■  x 7 cmPtProduction Angle 1° - 4°12 100 psrL 18 mParticles/5 x 1012 incident protonsM Kg> 1010 5 x 10841References1. M. Zeller, L. Rosenson and R.E. Lanou, Jr. Low Energy Separated Counter Beam, p. 193, BNL Summer Study on AGS Utilization. BNL 16000 (1970).2. D.M. Lazarus. Proceedings of the Summer Study Meeting on Kaon Physics and Facilities, p. 119. BNL 50579 (1976).3. C.T. Murray and J.D. Fox. Characteristics of MESB: Descriptionfor Users. Accelerator Department Informal Report BNL 18627 (1974).4. P.H. Pile. Beams for Kaon Research. Nuclear Physics A450, 517c (1986).5. H.N. Brown et al. Room temperature, long pulse rf beam separators. BNL Summer Study on AGS Utilization, p. 79. BNL 16000 (1970).6. J. Sandweiss et al. Medium energy, room temperature, rf separated beam for counter spark chamber experiments. Ibid. p. 55.7. G. Bunce. Separated K- beams (10 GeV and above). International Conference on Hadron Spectroscopy, University of Maryland (1985).EXPERIMENTAL FACILITIES FOR A CANADIAN K.A.O.N. FACTORYC. Tschalar, S.I.N.1 Criteria for CKF PerfomanceThe CKF is designed to maximize the quality of secondary beams. Their per­formance is generally described by three quantities:- phase space density- integrated flux- beam purityFurthermore, the wide range of proposed experiments requires a gamut of 6 to 8 independently tuneable beams for charged particles covering the entire available momentum range, as well as wide band beams for neutral kaons. While a high primary phase space density provided by the accelerators is an essential prerequisite, final performance of the CKF is determined by the particle density in the secondary beam phase space volume Q c;:(Is =  A P sA s6 sas a sbsPswhere A P$ is the momentum spread, A 5 and 6$ the macro and micro duty cyles, as and bg the vertical and horizontal r.m.s. beam widths, and as  and /3s the corresponding r.m.s. divergencies.Since the secondary particle production spectra are generally much broader in angle and momentum than the primary beam, only 4 of the 7 components of Os, namely A 5, 6s, as, and 65 are determined by their counterparts of the primary beam. Given a minimal primary phase space volume (Ip provided by the accelerators, a first optimization involves reducing the relevant components A p, 6p, ap, and bp at the expense of A Pp, ap and /3p by bunching and focusing the primary beam.Further optimization of secondary phase space including trade-offs between components of (Is are the domain of the experimental facilities.432 Production of secondary beams2.1 Pulsed beamsSlowly pulsed beams of neutrinos, muons and other particles may be produced by a fast extracted beam from the driver or the extender. The last mode may be used to empty the extender after a slow extraction cycle and before injecting a new driver pulse, thus providing a time-shared operation of the fast and slow extraction facilities.The pulsed neutrino facility consists of a production target of one interaction length followed by a pulsed neutrino horn designed to form a parallel beam of the pions emerging from the target. The pions then decay in a drift space of about 50 m length, terminated by a steel shielding to protect the neutrino experimental area from hadron and charged lepton background.The same production target may also provide pulsed low energy (surface and cloud) muons to be extracted laterally for appropriate experiments.2.2 Continuous beamsContinous beams of secondary particles may be produced by a slowly extracted beam from the extender impinging on targets of typically one interaction length or by multiple traversals of the extender beam through small internal targets. In both cases, the inherent macro duty cycle Ap  of the extender beam is converted into “continuous” secondary beams (A.s- — 1).While the slow extraction method is the more conventional one, it also has the decided disadvantage that the increase in the duty cycle (A s / Ap)  of about 3.104 is not converted into a corresponding useable reduction of the remaining components of Cls, thus effectively increasing Us by four orders of magnitude. The internal target mode avoids this drawback as will be shown in sect. Optimization of secondary beam extractionConventional secondary beams generally have angular and momentum accep­tances which are much smaller than the corresponding widths of the production spectra. Therefore, several secondary beams can, in principle, be extracted from each production target, thus minimizing the number of targets and increasing their luminosity.Since productions cross sections for secondaries of higher momenta are in­creasingly forwardly peaked, extraction angles near zero degree are required for most beams to maximize flux. Finally a flexible experimental programme re­quires all secondary beams to be essentially independently tuneable within their designed momentum range and angular acceptance.In order to accommodate these conflicting requirements, a magnetic system for multiple achromatic extraction of independent momentum beams (MAXIM) has been proposed^1'. It allows extraction of three beams with near forward44production angles from a single target. The central beam of relatively high momentum, extracted at zero degrees, is completely independent in momentum and particle charge. By a slight vertical tilt, it may be separated from the primary beam by a bending magnet with a grooved yoke through which the primary beam passes undeflected (Lambertson septum). The two lateral beams of lower momentum range, are extracted at small angles of 5° — 10° by half- quadrupoles. They are independent in momentum but require opposite particle charge for optimal flux. For equal particle charge, one of the beam’s intensity is reduced by a factor of 2 - 5.The proposed layout for the experimental facilities of CKF using a slowly extracted primary beam is shown in Fig. 1. Two tungsten targets in series accommodate 8 charged particle beams covering the entire momentum range from 0.2 - 20 GeV/c. The first target is 5 cm (1/2 interaction length) and the second 10 cm long. The extraction geometries of the two targets are shown in Figs. 2 and 3.A third target of one interaction length immediately in front of the beam dump produces neutral kaons. A steel shield reduces the hadron and muon background. It may be magnetized to minimize its tickness and maximize kaon flux as described in ref. 2.The principle characteristics of the secondary beams are shown in Table 1. They reflect essentially the properties of the extraction sections. The values for the angular acceptances are limited by the requirements of the proposed particle separators, values in parenthesis represent the maximum allowed by the extraction geometry. Detailed design of the secondary beams is still pending.Simultaneously available fluxes of n~ , K ~ , p, and K £ at the exits of the pro­posed fully separated secondary beam lines, produced by an extracted primary beam current of 100 pA, including absorption of primaries and secondaries in the targets and decay in the secondary beams are shown in Figs 3 to 7 (dashed curves). The corresponding peak luminosities are shown in Figs 8 - 11  assuming a micro pulse width of 0.5 ns.2.2.2 Minimization of Target VolumeAs mentioned in sect. 2.2., the increase of the secondary beam duty cycle A s  to unity may be used to reduce the volume Vt of the production targets by placing them into the extender and exposing them to multiple traversals of the extender beam.Assuming that the protons in the extender beam are redistributed through­out phase space during each revolution in the extender, which is a condition for beam stability (no resonances), then an internal target smaller than the beam cross section gradually reduces the primary beam’s intensity without changing its shape. Constant beam intensity at the target may be maintained as shown in Appendix A. At the end of the macro cycle, the remaining beam reduced to e times the initial current must be kicked out of the extender to make it ready45for the injection of a new driver pulse. This can be done with the fast abort kicker or with a special kicker to serve the fast extraction target.Appendix A shows that the source volume Vt for the secondary beams is reduced by ratio of secondary to primary macro duty cycle times the ratio of e.ln(l / e ) / ( l  —e), compared to the source volume of a target in a slowly extracted beam.Internal targets present a number of technicial problems which are dealt with in more detail in ref. 3. The first one is the extremely high power density of 10 to 100 MW/cm3 deposited in the target. A technically feasible solution is a ribbon target made of beryllium or a graphite compound moving horizontally through the primary beam at high velocity (see Fig. 12). The optimal shape of a ribbon target, resulting in a minimal apparent source size is derived in Appendix B for the case of two internal Be-targets each intercepting half the primary beam.Assuming a primary beam emittance of n mm mr in both directions and a maximal divergence of 2 mr limited by quadrupole apertures, we chose the following beam and target parameters as a typical example:ap — 0.5 mm, bp =  0.5 mm a p — 2 mr, Pp =  2 mrA P=  3.3 K T5 6a =  25 mr e=  2 %A= 500 mmwhich results in the target dimensionsI =  4 mm as =  0.05mmA corresponding target ribbon of 0.1 mm thickness moving at a velocity of 24 m /s through an beam of 100/uA/ 3.3 10~5 would show a temperature rise of about 500° C.The second technical problem is multiple scattering of the primary beam by the internal targets. The total projected r.m.s. scattering angle after a macro cycle is about 0.6 mr which is smaller than the assumed beam divergence of 2 mr at the targets and therefore does not increase the beam emittance significantly. Large-angle elastic and inelastic single scattering tails have to be removed by appropriate collimators. Total momentum loss in the targets during a macro cycle is about 0.7 % with a straggling r.m.s. width of 0.1 %, which can be handled by the extender.Finally, despite decreasing beam current, the beam intensity at the targets can be kept constant during a macro cycle by gradually steering the beam into the target. The beam remaining in the extender after a macro cycle must be kicked out before injecting a new driver pulse.A possible layout of an experimental facility using two internal targets is shown in Figs 13 and 14. The driver is assumed to have a straight section of 60 m length devoid of accelerator cavities or beam switching installations. The extender beam is distorted into a trapezoidal shape by two superconducting bending magnets of 5 Tesla field strength providing an 70 m long central straight section with an achromatic beam for the two internal targets. Eight secondary beams of the same type and as described in sect. 2.2.1 are extracted by two MAXIM systems. Their basic properties are listed in Table 2.The secondary particle rates and their phase space densities (peak luminosity per momentum bite) which are available simultaneously at the exits of the fully separated beams are shown in Figs 4 - 1 1  (solid curves).2.2.3 Minimization of the micro duty cycleFrom the primary beam characteristics listed in the previous chapter, it is ap­parent that the cross section of the ribbon target could be reduced by another order of magnitude if the amount of the remaining beam e were raised accord­ingly. However, even apart from problems of target heating, such a small source size would be difficult to convert into a correspondingly small secondary beam emittance because of optical aberrations.On the other hand, for applications such as RF separation of higher energy secondary beams, a reduction of the micro pulse length t$ of the secondary beam would be useful. So far, Tg was equal to the micro pulse length Tp of the extender beam which is limited to about 0.5 ns by bunching limitations.For a given primary pulse length Tp, the secondary beam pulse length Tg may be reduced at the expense of the primary beam width ap by vertical RF deflection of the primary beam as illustrated in Fig. 15.The effective beam height a*p after RF deflection iscip =  apTp/rg —  ap8p / SgThe amplitude A of the sinusoidal deflection of frequency v is A =  (l/iT)a*P / (i/Tp) = (1/tt )aP/(vTs)Assuming again a primary beam emittance of n mm mr (r.m.s), a maximum divergence of 2 mr, and a primary r.m.s. beam height ap of 0.5 mm, an increase of the beam reduction factor e to 20 %, would allow an effective beam height a*P of 5 mm resulting in a micro pulse length reduction ofTp/rg =  10For a phase length vrp of 25 % or 90°, the corresponding deflection amplitude needed is A — 6 mm at the target.Technically an RF deflection is produced by a vertical electric RF field in the shape of a standing wave excited in a resonent RF cavitity. The cavity may be47placed at the entrance leg of the trapezoidal extender orbit. A second identical cavity is needed at the exit leg after the second target to quench the deflection and reconstruct the original beam emittance to within less than 0.5 %.If we require a vertical voltage gap of the cavity of eight times the r.m.s. beam height and a phase length urp of 25 % or 90°, then, according to Appendix C, the RF deflector strength (peak voltage V times length L) isV( kV )L (m ) =  20P(G eV/c )e p(n  m m  mr)Tp/rsTherefore, a pulse length reduction of 10 and an emittance ep of n mm mr requires 6 MV m deflector strength or 200 kV in a cavity of 30 m length requiring RF power of about 500 kW.Constant beam intensity on the targets during the macro cycle may be ob­tained by varying the phase of the RF deflection relative to the beam RF thereby shifting the centre of the pulse gradually towards the target plane such as to increase to relative beam intensity at the target enough to offset the decrease in total beam current.The resulting time structure of secondary particles at the exit of the S 3, S 6, and S 20 beams are shown in Fig. 16 assuming a beam pulse length Tp of 0.5 ns (f.w.h.m.), a deflector frequency of 500 MHz, and a secondary pulse length ts of 50 ps. This secondary beam pulse structure allows complete separation of pions, kaons and antiprotons by a single RF deflector at the end of optimally short secondary beams for momenta from 1 to 20 GeV/c without pulse overlap and without need for long additional separation sections.The corresponding increase in phase space density is illustrated in Fig. 8 - 11 (dotted curves).3 ConclusionThe previous discussions show that major improvements of secondary beam quality are achieved by the extraction scheme MAXIM, by internal targets and if needed, by RF pulse “sharpening” . Apart from the physical gains, there are also a number of technical advantages inherent in these schemes: MAXIM al­lows a concentration of the experimental facilities and a shorter primary beam. Internal targets eliminate the need for a slow extraction system and a proton feeder line to external targets. Futherinore, much smaller magnet apertures are required for the small emittances of the secondary beams. Finally, RF pulse “sharpening” allows shorter high momentum beams and simpler RF separators. These technical simplifications result in considerable cost reductions. On the other hand, the technology of internal targets and the corresponding beam con­ditioning systems add complexity and development costs. The proposed RF deflectors represent substantial investments of their own and are probably only justifiable by experimental requirements beyond RF beam separation.48As an illustration, the gains in beam quality of the CKF compared to the improved version of the existing AGS facility in Brookhaven are shown in Table3.While the gain of 30 provided by the acclerator system is considerable, an­other factor of 40 in average vertical luminosity and thus in separated secondary flux is gained by MAXIM and internal targets, providing an impressive total average flux gain of 1200.In terms of four-dimensional luminosity, the total gains are 5000 and a fur­ther gain of 10 in pulse length to a total average gain in phase space density of 5 104 may be achieved by RF deflection.These improvements in secondary beam quality open up truly new fields of experimental physics and put the CKF into an entirely different class from existing hadron facilities.REFERENCES1. C. Tschalar, “Multiple Achromatic Extraction System” , to be published in Nucl. Instr. and Methods.2. LAMPF II Proposal 19863. C. Tschalar, “LAMPF II Experimental Area, A New Concept Using In­ternal Targets” , LAMPF II Technical Note 85-010.49APPENDICES  A  Stabilization of beam intensity at an inter­nal target during a macro cycleThe fraction of the primary beam Ip striking the internal target of cross section 4 apbp and length I is proportional to the beam loss A Ip in each traversal by interactions with the target:A Ip — (dip/dA) dlp/dA is the current density at the target.The average beam loss per unit time is thusdlp/dt =  A Ip/T =  (dIP/dA).as .bs .(l/X)/Twhere T is the time of flight for one extender orbit.If dlp/dt is to be constant during a macro cycle, dlp/dA must be kept constant. This can be done either by keeping the beam centered on the target and reducing the beam cross section linearly with time, or by keeping the beam cross section constant and shifting the beam’s centre gradually towards the target (see Fig. 17).In either case, the (constant) rate of beam loss of the target is determined by the aplitudes ap and bp and the current e.Ip of the beam at the end of the macro cycle:dlp/dt = (C/T) bs.(l/X)/(aPbp) where C is constant “form factor” of the beam profile:C =  (dip/dt)peakAapbp / Ip Furthermore, since, after N  traversalszip  =  Ip ~ (dlp/dt).NT =  IP -  CeIp(asbp)/(aPbp)N  we find (for c =  1)asbsl =  (1/7V).(1 — e)/e.apbpX The source volume Vint of the internal target (Vlnt =  4asbpl) is therefore Vint =  (1/IV).(1 — e)/(eln(l/e)).VKXtwhereVext — i.apbpXln(l/e)is the source volume of an external target producing the same number of inter­actions as the internal targets of source volume Vint.50B Optimal shape of a ribbon targetFor a secondary beam of given vertical acceptance angle 6a, the optimal ratio of thickness to width of a ribbon target is that for which the apparent target size is minimal. From Fig. 18 we find the apparent half width w of the target isw = as + 1/2 ltdaFor a given target cross sectional area A =  2ash, we find that w reaches a minimum ofWmin = 2-as = h&afora -  1/ 2It.6aTherefore (see Appendix A), the optimal length for a pair of equal ribbon targets(/t =  1/ 21; bs =  bp) is/0 =  \/(l/./V ).(l -  e)/e.ap\/9aC RF deflectionThe standard method to deflect a particle beam transversely is to place an RF resonant cavity of length L and gap D excited with a standing wave of frequency v, field amplitude E and voltage amplitude V =  E.D  into a “parallel” section of the beam thus inducing a transverse angular deflection. At the exit of the cavity, the beam is focussed onto the ribbon target thereby converting the angular deflection into a spacial deflection as shown in Fig. 15.The amplitude A required for a beam of vertical width ap is given in sect.2.2.3 asA =  (l/7r)(ap/ivrp).rP/r s .This spatial deflection is produced by an angular deflection 7 in the RF deflector, which is related on the one hand to the beam divergence ap in the deflector by7 =  (1/ n)ap/wTp)rp/Ts,and 011 the other hand to the RF field amplitude E  by7(mr) =  1/2 ,E.L(MV)/P(GeV/c)(The factor 1/2 is appropriate for standing wave deflection.)51If the r.m.s. beam height in the cavity is 1/8 of the gap D, and ep is the vertical r.m.s. emittance of the beam, thenD(m)  =  8.10-3ep(7r E L ( M V ) / P ( G e V / c )  =  V . L ( M V . m ) / ( D ( m ) . P  {GeV/c) )andV.L(kV.m) — (l6/ir)P(GeV/c).ep( factor utp is the fraction of the RF period filled by the micro pulse length Tpoi beam. For a conservative value of vrp — 0.25 or 900 of the phase angle, we obtainV.L(kV.m) =  20.P(GeV/c) .ep(n .Tp / ts52Secondary beam characteristics for external W-targets (slow extraction)T A B L E  1Beam Lengthmltgmminter.perprotonabsorptionfactora8mrBsmrQsmmb8mmQmsrAp/p extrac. angleprod.angleS 0.2 12 50 .4 .9(».P) 80 200 1.1 2.1 50 .15 120° 120°3 0.8 18 50 * .4 * 76(K,p) 22 71 0.9 1.8 5(8) .05 11.5° 0 - 5 °S 1.5 25 100 .4 .64(K,p) 11 56 .74 2.8 2(5) .03 5.7° 0 - 4°S 3 35 100 .4 .64(K,p) 5.5 56 .57 2.8 1(5) .05 5.7° 0 - 3 °S 6 60 50 .4 .75(K,p) 1.8 36 .50 1.0 0.2(2) .05 0° 0 - 3°S 20 100 100 .4 .55(K.p) 1.25 25 .50 1.35 0.1(1) .05 0° 0°K°kl 15 100 .14 .7 (K°) 10 2.5 .70 .52 0.1 wideband0° 0°T A B L E  2Secondary beam characteristics for internal Be-targetsBeam Lengthm1. Inter.• tg per m m  protonabsorptionfactor °smrBsmrasmmbsmmQmsrAp/p extrac. angleprod. angleS 0.2 15 4 0.5 1 80 200 .10 1.7 50 .15 120° 120°S 0.8 18 4 0.5 1 36 71 .09 .52 8 .05 11.5° 0 - 5°S 1.5 25 4 0.5 1 28 56 .075 .51 5 .03 5.7° 0 - 4 °S 3 35 4 0.5 1 28 56 .075 .51 5 .05 5.7° 0 - 3°S 6 60 4 0.5 1 18 36 .062 .51 2 .05 0° 0 - 3 °S 20 100 4 0.5 1 13 25 .056 .50 1 .05 0° 0°K°*L 30 4 0.5 1 5 1.25 .05 .50 .025 wideband0° 0°T A B L E  3Average gains In performance of CKF compared to Improved AGS (3pA)Flux Phase Spacesec. beam source size vertical horizontalmicro pulse widthIntensity 30 .Primary Emittance - 1 1 -beam Pulse width - “ 1Exper.Beam concen­tration (MAXIM)2 * -facilities Target size (Int. tar­gets)2 10 4Pulsing (RF def­lector)10Total gains 120 10 4 1053Fig. 1. External target facility (slow extraction).54q.Quadrupol e;  HQ = Ha1f -q ua drupol  e ; Q8 = " F i g u r e - o f - 8 " quadrupoleBH*H-type bending magnet; BWF«“W1ndow f r a m e “- t y p e  bending magnetFig. 2. MAXIM extraction geometry for target 1.Q = Quadrupole;  HQ = Ha1f - q u a d r u p o l e ; Q 8 = "F i g u r e - o f - 8 “ quadrupole  BH*H-type bending magnet; BWF*''W1ndow f ra me ' ' - ty pe  bending magnetFig. 3. MAXIM extraction geometry for target 2.55bo be • ^U*Dashed curves for external target, solid curves for internal targets. Dashed curves for external target, solid curves for internal targets.561 1  & |  S --O „ coE?«3AOo4>-4-J <V 4 )  CObo -gfij &S »s-*3 -S 2fe •= «2 ? c3o>u-133^ >fll >H-S ?*55 Qu4>T3 °co '"O P CO*> 0) O  bo uI <3d*PCX£odb  03p~ *-« C 5 =*. .5oo  £tH Octbo£ ;l .5p aU VMco nds sg<* s^ co MO,&COE?rj H3 C o£ «5ADPSx  S^ 4)<3&P"d ^  w a;COpOPT3n30)O'■OiOCu oco4>bo«3■H ^Ohs <  .2a © od  " bD D 4)CO K 4) boE: .£00bo• HCz-i Fig. 10. Antiproton phase space densities at the exits of the secondary beams (IP = Fig. 11. Neutral kaon phase space densities at the exits of the secondary beams 100/i A, Tp = 0.5 ns). Dashed curves for external targets, solid curves for (Ip = 100/iA, Tp = 0.5 ns). Dashed curves for external targets, solidinternal targets, dotted curves for addtional RF “pulse sharpening”. curves for internal targets, dotted curves for additional RF “pulse sharp­ening” .57Fig. 13. Internal target facility1:20058Eo _ ,6£><3ssA•  HAcJ-AboAOos-.uobf>• H‘S3OlT3V i2Qi0u01(VQ01S'• + -sfe<W3C LE3Si3C LEn>isId-jz•£Fig. 15. Geometry for RF pulse “sharpening”.59<0= oi—Q.<U-O•6T3OEaoJO(0c<uErj4)CQ<*>£nN« 0sLH n TO >O 8 1T- =KDNW0smICLii it iu jj uro> E > >CU ’> CU Eo o i_ O oL n 3uo m mII i i 1 i i i iQ . __i a . _ i0s'OCMIokC3aEa>- a5“TJCooVC/ibO• HClH»a .a«<60THE TRIUMF KAON FACTORY ACCELERATORSM.K. Craddock Physics Department, University of British Columbiaand TRIUMF, 4004 Wesbrook Mall, Vancouver, B.C., Canada V6T 2A3ABSTRACTTo accelerate the 100 pA proton beam from the TRIUMF cyclotron to 30 GeV, a chain of 5 fast-cycling synchrotrons and dc storage rings is proposed. 450 MeV H~ ions from TRIUMF are injected by stripping into the Accumulator ring. A 50 Hz Booster synchrotron then accelerates the pro­ton pulse to 3 GeV, where the frequency swing is almost complete. In the main tunnel (170 m radius) are the Collector ring, which collects 5 Booster pulse trains, the 10 Hz Driver synchrotron and the dc Extender ring, where beam is stored for slow resonant extraction. The accelerator designs have various features, such as H- stripping injection, high transition energy, and bucket-to-bucket beam transfers which will avoid or reduce beam loss. Dual frequency magnet power supplies provide a 3:1 rise:fall ratio, reducing the peak rf voltage requirements to 600 kV for the Booster (46-61 MHz) and 2400 kV in the Driver (61-63 MHz).MAIN ACCELERATOR AND INJECTORThe specifications for the KAON Factory1 call for the accelerator to provide 100 pA proton beams at 30 GeV. This choice of energy satisfies the desire for intense fluxes of high-energy kaons as well as stopping kaons, antiprotons and neutrinos. The 100 pA current (6 x 1011+ protons/s) is chosen to provide a significant (80-fold) improvement over beams which have been available in this energy region (<8 x 1012 p/s - see Table I, and to make possible experiments which have hitherto been impractical.In light of these specifications the KAON Factory accelerator system has been based on a rapid-cycling (10 Hz) 30 GeV proton synchrotron. At lower energies other types of accelerator could be considered, but above about 15 GeV a synchrotron is the only practical choice. The fast cycling rate keeps the charge per pulse down to N = 10 pC (6 x 1013 protons) and restricts the time available for instabilities to develop. The circulat­ing current, a measure of the likelihood of beam instability, is 2.8 A, not quite double the 1.5 A at which the CERN PS operates and only 20% higher than that delivered in the Argonne IPNS. Intensity-dependent effects, such as tune shift, instabilities and beam loading, should therefore lie in a well-understood region.The existing 30 GeV synchrotrons are limited in beam intensity both by their low cycling rates (~1 Hz) and by their low injection energies (<200 MeV into their first synchrotron stages). The injection energy is crucial because space charge reduces the transverse focusing strength by an amount strongly dependent on energy, and largest near injection; it also varies across every beam bunch. There results a spread in tune (number of betatron oscillations per orbit) which, at low energies, for a machine of fixed radius and magnet aperture, is given by Av « -N/62y3; here 6 and y are the usual relativistic speed and total energy param­eters. Assuming Av = 0.2 (which is the upper limit if serious resonances are to be avoided) the accelerable charge per pulse N increases as 82y3,61Table I. High-intensity proton synchrotrons.Average Rep. Protons/ CirculatingEnergy current rate pulse N current I(GeV) (u A) (Hz) (x 1013) (A)Slow Cycling5KEK PS 12 0.32 0.6 0.4 0.6CERN PS 26 1.2 0.38 2 1.5Brookhaven AGS - with Booster28.5 0.9(3)0.38 1.6(5)0.9(3)Fast CyclingArgonne IPNS 0.5 8 30 0.17 2.3Rutherford ISIS 0.55(0.8) 40(200) 50 (2.5) (6.1)Fermilab Booster 8 7 15 0.3 0.3AGS Booster (1) (20) (10) (1.25) (3)Proposed BoostersTRIUMF 3 100 50 1.2 2.7European HF 9 100 25 2.5 2.5LAMPF AHF 6 144 60 1.5 2.2KEK Booster 1-3 100 15 4 8Kaon FactoriesaTRIUMF 30 100 10 6 2.8European HF 30 100 12.5 5 2.5LAMPF AHF 45 32 3.33 6 2.2Japan - Kyoto 25 50 30 1 0.5- KEK 30 30 1 20 7aSlow extraction ^Fast extractionevaluated at injection. It was to take advantage of this effect that the injection energies for the Brookhaven AGS and CERN PS were raised from the original 50 MeV to 200 MeV and 800 MeV, respectively, providing improvement by a factor of 5 or more. To achieve the 50- to 100-fold higher intensity specified for the KAON Factory, without demanding a very high (and expensive) cycling rate, requires an injection energy of at least 400 MeV (a factor of 2.6 better than 200 MeV).The TRIUMF cyclotron provides the basic performance required for the KAON Factory injector: H“ beams at energies up to 520 MeV, with currents of up to 140 pA scheduled routinely. The only question is how to match this cw machine, producing a continuous stream of beam bunches at 23 MHz, with a 10 Hz synchrotron accelerating 3-ys-long pulses every 100 ms. Fortunately, the key to an answer is already available in the use of H- ions in the TRIUMF cyclotron. If these are injected into the first syn­chrotron stage by stripping them to protons as they pass through a thin62foil, Liouville's theorem on phase space conservation may be circumvented and beam injected over the many thousands of turns required. At present, the stripping process is used for extracting protons from the cyclotron; a new extraction system will be required to extract the H ions intact. Studies indicate that this should be straightforward, especially if beam is extracted at 440 MeV, taking advantage of radial precession induced by the vr = 3/2 resonance nearby. Extracting below 500 MeV will also avoid most of the beam spill caused by Lorentz stripping of the H ions in the cyclotron's magnetic field; this amounts to 8% between 440 and 500 MeV but is only 1% below 440 MeV. The g2y3 factor is 3 times larger at440 MeV than at 200 MeV.BOOSTER SYNCHROTRON AND STORAGE RINGSThe choice of a high cycling rate for the synchrotron immediatelyimplies a need for high energy gain per turn and therefore high aggregaterf voltage (several thousand kV). To ease this requirement, an asymmet­ric magnet cycle will be used with a rise time 3 times longer than thefall; this reduces the voltage required by one-third, and the number of cavities in proportion.There are two additional challenges presented by the rf system. One is associated with the high beam current, implying an rf power capability well above the 3 MW in the beam itself. The other is the frequencyswing, which amounts to a factor 1.37 as the energy rises from 440 MeV to 30 GeV. In order to separate the problems of providing high rf voltage and power from that of providing frequency swing, an intermediate Booster synchrotron is proposed to accelerate the proton beam from 440 MeV to 3 GeV. This would cover almost the entire frequency range (a factor1.33), but involve only 300 kW beam power and require only 580 kV rfvoltage (again with a 3:1 asymmetric magnet cycle). The Booster would be one-fifth the diameter of the main "Driver" synchrotron (radius 34 m rather than 170 m), so that 5 pulses from it would completely fill the Driver circumference; and it would cycle 5 times faster at 50 Hz. TheDriver itself is left with a 2400 kV voltage requirement but a frequencyswing of only 3%.The use of a Booster also allows the aperture and cost of the main ring magnets to be significantly reduced—the diameter of the beam being reduced by adiabatic damping during passage through the Booster.The large number of rf cavities required, together with the lowermagnetic fields used, combine to make the radii of these machines some­what larger than slow-cycling (low-intensity) machines of the same energy. The chief parameters of the two synchrotrons are listed in Table II. Figure 1 shows the proposed layout, with the Driver encircling the present cyclotron buildings.To allow time for injection or for slow beam spill for counter ex­periments, it is conventional to "flat-bottom” or "flat-top" the magnet cycle of a synchrotron. In the present case, however, starting with 100 yA beams from the TRIUMF cyclotron, such a procedure would result in average beam currents at 30 GeV of only 50 yA for neutrino production (fast extraction) or 33 yA for counter experiments (slow extraction). Instead, it is proposed to follow each of the three accelerators by a relatively inexpensive dc storage ring, so that the TRIUMF cyclotron would be followed by a chain of 5 rings, as follows:63Table II. Synchrotron design parametersBooster DriverEnergyRadiusCurrentRepetition rate Charge/pulseNumber superperiods Lattice ) (focusing structure/ (bending Number focusing cells Maximum 8x*3y Dispersion qmax _  Transition Yt = l//h Tunes vx*Vy Space charge AVy Emittances W s xx£y at injection /(eiongHarmonic Radio frequency Energy gain/turn Maximum rf voltage rf cavities3 GeV4.5 Rx = 34.11 m 100 pA = 6xl0llt/s 50 Hz2 pC = 1.2xlOl3ppp 6F0D00B0BB0B02415.8 m x 15.2 m4.0 m9.25.23 x 7.22 -0.15139irx62Tr (pm)0.064 eV-s4546.1 61.1 MHz210 keV 576 kV 12 x 50 kV30 GeV22.5 RX = 170.55 m 100 pA = 6xl0ll+/s 10 Hz10 pC = 6x1013ppp 12F0D0BBBBBOBO4838.1 m x  37.5 m 9.09 mOO11.22 x 13.18 -0.0937irxl6ir (pm)0.192 eV-s22561.1 ->■ 62.9 MHz 2000 keV2400 kV 18 x 135 kVTUNNELS— 1 ! DnLEO iBn' BOOSTER ^ 5mHMAIN RING3 GeV BOOSTERCYCLOTRONNEUTRINO  FACILITYE X P E R IM E N TA LH A LL100 m i iFig. 1. Proposed layout of the accelerators and cross sections through the tunnels.64A Accumulator: accumulates cw 440 MeV beam from the cyclotron over20 ms periodsB Booster : 50 Hz synchrotron; accelerates beam to 3 GeVC Collector : collects 5 Booster pulses and manipulates beamlongitudinal emittanceD Driver : main 10 Hz synchrotron; accelerates beam to 30 GeVE Extender : 30 GeV storage ring for slow extractionAs can be seen from the en­ergy-time plot (Fig. 2) this arrangement allows the cyc­lotron output to be accepted without a break, and the B and D rings to run continu­ous acceleration cycles; as a result the full 100 yA from the cyclotron can be accelerated to 30 GeV for either fast or slow extrac­tion.|......., , t Fig. 2. Energy-time plot show-o 100 ing the progress of the beamTime (ms) through the five rings.The Accumulator is mounted directly above the Booster in the small tunnel, and the Collector and Extender rings above and below the Driver in the main tunnel (Fig. 1). Figure 3 shows schematically the arrange­ments for beam transfer between rings and the location of rf stations.J 30 GeV>3 GeV} 440 MeVFASTXTRACTEDBEAMFig. 3. Arrangement of rf cavities and beam transfer lines (schematic).-CD - Dipole Magnet —0- Quodrupole Magnet ♦ RF Station65Identical lattices and tunes are used for the rings in each tunnel. This is a natural choice providing structural simplicity, similar magnet aper­tures and straightforward matching for beam transfer. The practicality of multi-ring designs has been thoroughly demonstrated at the high-energy accelerator laboratories, and new projects such as HERA, LEP and SSC use ever-larger numbers of stages.The need for the Accumulator ring would of course disappear if, in­stead of the TRIUMF cyclotron, a high-intensity pulsed H- linac were used as injector. The cost of such a machine, rivalling LAMPF in performance, would, however, be formidable-over $50 million even for 440 MeV, based on recent SSC estimates. By comparison, the cost of the Accumulator isestimated below to be about $5 million.The Collector ring could also be dispensed with, as in the LAMPF II proposal, although this option is not tied to the choice of a linac as injector. Whatever the injector, the lack of a C ring necessitates flat-bottoming the main synchrotron (D) magnet cycle for collection of theBooster pulses. Maintaining the same final average current (100 pA) then requires increasing either the repetition rate or the number of protons/ pulse (and hence the magnet apertures) for both B and D rings. The costs involved in such changes would considerably exceed any savings achieved by eliminating the C ring, the cost of which is $13 million.TIME STRUCTURE OF THE BEAMTo minimize beam loss, bucket-to-bucket transfer is planned between the rings. To achieve this, at injection into the Accumulator ring the rf frequency must be some simple multiple of that of the TRIUMF cyclo­tron, 23.05 MHz. Double this frequency, 46.1 MHz, has been chosen, making the frequency at top energy 62.9 MHz. For this frequency range the cavities are of reasonable size and tubes and power supplies are readily available. Moreover, operational experience is available on the fast- cycling 30-53 MHz system at the Fermilab Booster.In order to populate all the rf buckets, which are spaced only half as far apart as those at the extraction radius RT in the cyclotron, the radius of the A and B rings is chosen to be an odd half-integer multipleof TRIUMF*s, 4.5 Rx = 34.1 m. The circumference will then contain 45 rf buckets compared to the 5 bunches circulating in the cyclotron at double the spacing. 4.5 turns from the cyclotron will encircle the A ring once, populating every other bucket; the next 4.5 turns will popu­late the remainder, the bunches being automatically interleaved (Fig. 4).Fig. 4. Populating the 45 Accumulator buckets (radial sectors) over two turns of injection. Every fifth beam bunch is emphasized to indicate each cyclotron turn. Five buckets are left empty by pulse programming the ion source.66The C, D and E rings, being 5 times larger, will contain 225 rf buckets. Because of the higher radio frequency, the time separation of beam bunches will be 15.9 ns at 30 GeV, compared to 21.7 ns at 440 MeV. If larger separations should be required, they can be achieved, with some loss of intensity, by pulse programming the ion source. For instance, the 1:5 pulser already in operation would increase the bunch separation to 80 ns. Pulse programming (a 110 ns gap at 1.024 MHz) is also required to keep a group of 5 adjacent buckets empty in the A ring; this provides =110 ns for the extraction kicker field to rise, allowing the beam in the 40 filled buckets to be extracted cleanly. This 40 full + 5 empty bucket pattern is retained through the B, C, D and E rings, serving the same purpose for injection and extraction kickers, although the gap shrinks to 80 ns at 30 GeV. Each bunch in the fast extracted beam from the Driver is 2.6 ns long (FWHM); each train of 40 bunches is0.62 ys long, while the five trains extend over 3.48 us. In the Extender the bunch length can be adjusted to users' requirements by changing the transition energy and rf voltage; by reducing Yt to *ts "natural" value11.2 the bunches are stretched to 6.3 ns FWHM. A completely debunched beam is also possible.CONTROL OF HIGH-INTENSITY BEAMSSuccessful operation of a high-intensity accelerator depends cru­cially on minimizing beam losses and the activity they produce. Sources of loss must be controlled or eliminated, the beam and any spill must be carefully monitored, losses must be localized through the use of collim­ators and beam dumps, suitable materials must be used for absorption and shielding of spilled beam and, where activation cannot be avoided, equip­ment must be capable of being handled remotely. Careful attention to these features in the case of the TRIUMF cyclotron has enabled twice the originally specified beam current to be accelerated while exposing personnel to less radiation. The same principles would be followed for the five rings of the KAON Factory accelerator.Several processes which give rise to losses in existing machines have been avoided entirely in this design. The use of H ions for injec­tion into the Accumulator ring will almost entirely eliminate injection spill. The use of bucket-to-bucket transfer between the rings will avoid the losses inherent in recapturing coasting beams. The buckets will not be filled to more than 80% of capacity; this should avoid beam losses while providing a high bunching factor to minimize the space-charge tune spread at injection, and sufficient spread in synchrotron tune to give effective Landau damping. The rf voltage is programmed to maintain a constant bucket area over the early part of the cycle. The magnet lat­tices are designed to place transition, where the phase focusing passes through zero, above top energy in all the rings, thus avoiding the insta­bilities and losses associated with that passage. Moreover, with the beam always below transition, it is no longer advantageous to correct the natural chromaticity, so that sextupole magnets are needed only for error correction, and geometric aberrations of the beam are essentially reduced to zero.Beam instabilities will be suppressed or carefully controlled. Although all 5 rings have large circulating currents, the rapid cycling times give the instabilities little time to grow to dangerous levels.67Coupled-bunch modes, driven by parasitic resonances in the rf cavities and by the resistive wall effect, will be damped using the standard tech­niques (Landau damping by octopoles, bunch-to-bunch population spread, and active damping by electronic feedback). The longitudinal microwave instability is a separate case because of its rapid growth rate. It will be avoided by making the longitudinal emittance sufficiently large at every point of the cycle and by minimizing the high-frequency impedance of the vacuum chamber as seen by the beam.At this stage of the design it is not possible to make accurate estimates of beam blow-up due to instabilities or non-linear resonances, but to be safe, the magnet apertures have been designed to accommodate a 50% growth in the horizontal, and 100% growth in the vertical beam emittance.In case of the beam becoming unstable at any time through component failure or power excursions, each of the five rings is equipped with a fast abort system which will dump the entire beam safely within one turn.STATUS OF THE PROPOSALThe proposal was submitted to TRIUMF*s funding agencies, the National Research Council (which funds the facility) and the Natural Sciences and Engineering Research Council (which funds the experiments through individual grants), in September 1985. The total cost, including salaries and E,D&I, but not contingency, is estimated to be M$427 (1985 Canadian dollars). Two review committees have been set up jointly by the councils. The first, an international "Technical Panel" of particle, nuclear and accelerator physicists, met in February and has produced a very favourable report. The second "Review Committee" consisting of Canadian industrialists and scientists from various disciplines met in April and is expected to submit its report to the councils at their fall meetings.From there, it is a government decision, whose timing is intrinsic­ally uncertain—but we can't believe that it will take more than a year for them to realize what an irresistible opportunity the KAON Factory is for Canada!REFERENCE1. KAON Factory Proposal, TRIUMF (1985).68THE INITIAL EXPERIMENTAL PROGRAM AT A KAON FACTORYPeter Kitching TRIUMF, 4004 Wesbrook Mall, Vancouver, B.C. V6T 2A3During the preparation of the TRIUMF KAON factory proposal we asked a number of physicists at TRIUMF or connected with the lab to write pro­posals for a representative program of experiments. Those proposals1 will form the basis for my talk, although it is clear that by the time a KAON factory is built some of them may have been done and other, more interesting but as yet undreamt of, experiments may have emerged. To set the stage, I will list some questions which go beyond the standard model, to which a KAON factory might be expected to make some contribution.1. How many generations of quarks and leptons are there?2. Are the gauge bosons only left-handed?3. What is the origin of CP violation?4. Are neutrinos massless?5. Why do we seem to see quarks only in the combinations qcf or qqq?6. Where are the glueballs and other exotica which seem to be predicted by the standard model?7. How does the long range Yukawa force arise from QCD?8. What nuclear phenomena require quark degrees of freedom? Do dibaryons exist? Does partial deconfinement occur deep inside a nucleus?9. Are there new phases of nuclear matter such as strange matter or quark-gluon plasma?The attack on these and other important questions will be mounted on two frontiers: the very high energy frontier at the colliders and theprecision frontier at the KAON factory and underground labs. These two approaches complement each other nicely. The important topics which will be studied at a KAON factory include:1. the rare decays of the kaon2. CP violation3. neutrino-electron scattering4. quark spectroscopy5. quarks in nucleiI will now discuss some representative experiments in each of these areas.As Nishijima said at the MESON50 symposium last year, "Next to the hydrogen atom the kaon is nature's most fertile laboratory". The experi­mental limits on branching ratios for rare kaon decays are summarized in Fig. 1, together with the improvement which might result at a KAON fac­tory. The fact that rare decay processes complement high energy experi­ment is well illustrated in Table I, which shows mass bounds set by different processes. Some of these experiments can probe for masses of undiscovered objects such as horizontal gauge bosons, neutral Higgs, supersymmetric particles, leptoquarks, etc. up to 100 TeV.69<£EOzXoz<XCD10'5  -K ° - M e i / +  +K — 7r fie K.~ir*vv k ° - m V K —  l v K ^ ttV 0U -K pPRESENT1|I0-6 - " T "PROPOSED0 1 *>1 11CO1O" I "}JL10 * 9 - K-FACTORY" 1 “T01O• • « | • • •1 0 '" -10i PO 1" 1 ”i o “ 13-" 1 " " I "F ig .  1 . Branching r a t i o s  f o r  ra re  kaon d eca y s .Table I .  Mass bounds from d iffe re n t processes.Higgs Pseudoscalar VectorProcess Experiment sca la rs leptoquarks leptoquarks(GeV/c2) (TeV/c2) (TeV/c2)rCw+ey <1.7xlO“ 10 0.2r (y + a ll)r(y+ee¥) <2.4xl0-12 2r (y + a ll)T(yA->eA) <1.6xl0-11 16 16 87T(yA+vA')T(K£*yy)r(K £ + a ll) 9xl0“ 9 5 4 62r(K °+ e i)r(K£-»all) <2xl0-7 2 2 30r(K °+ ye)<10-8r(K £ + a ll) 5 3 62r(K++ir+ye)rCK+^all) <7xl0~9 0.7 0.3 4Am(K£-Kg) 3.5xl0-15 GeV/c2 150 ~ -70Number of k+ an example I will describein Experiment an experiment to search for the decay K+ ir+vv in some detail. Figure 2 shows how the limits on this process have changed with time and also shows the limit which we hope to reach in a new experiment at BNL where we are collaborating with Princeton and Brookhaven. The present limit is 1.4xl0-7 but the standard model predicts2 that the process will occur at about the 10-10 level, the exact value depending on the mass of the top quark and the lifetime of the B meson. Thus one has about three orders of mag­nitude to search for new physics (such as axions, fermions, light Higgs, supersymmetric particles, etc., and by reaching the 10-10 level one can test the standard model for higher-order weak corrections,extra generations, KM mixing angles, etc. The BNL experiment shouldreach a sensitivity of ~2xlO-10 , enough to look for new physics butprobably not enough to test the standard model.A diagramatic outline illus­trating the principles of the de­tector is shown in Fig. 3. The IT*" is stopped in a live target and decays. The decay products pass through a drift chamber where their curvature in a longi­tudinal magnetic field of 1 T can be measured. Pions then stop in a scintillator range stack where their total energy and range are measured, and where one looks for the 4 MeV muon arising from decay of the pion. Any accompanying photons are detected by the pho­ton veto. The crucial aspects are to positively identify the decay pion and to veto any decays where a photon is emitted. A view of the actual detector is shown in Fig. 4. It will be assembled at BNL before the end of the year and the experiment will run for approximately 2-3 years.Improvements which could be made in the next generation ex- Fig. 3. Apparatus for K -*■ xvv periment at a KAON factory are: search (schematic).IRON RETURN YOKEid4id6io‘io10io12m*in4 Camerinl et al 10 '5°7xl0-5 Ljun, e. alI 4 X I 0 6  K l e m s  e t  a l•j 9.4 xio7 Cable et <"LnWWWVvV^  5 6XIO7| 4X|o7 Asano et alTHEORY- 3.0 x 10IKaon Factory expt.2x103.7xl051.8x10I.IxlO910Proposed BNL expt.2 xl010kNS\S\VS -IQ1.2x 101.5x107.5x102.7x10131970 1975 1980 1985 1990Fig. 2. Limits for K'r +• ir+v^V£ branching ratio.71Fig. 4. Apparatus for K -*• irvv search (reality).1. The use of a high purity stopping K+ beam (106-107/s; ir/K < 1)2. A superconducting magnet giving a field of 3 T. This would improve the momentum resolution as well as making the apparatus more compact.3. A BaF2 photon veto instead of lead/scintillator4. Greater segmentationWith these improvements it should be possible to approach a branching ratio limit of 10-12.Another example of a rare decay search3 is shown in Fig. 5 where one is looking for the lepton flavour violating process K° + ye. The present limit is 2xl0-9, the main background arising from K? irev . With an ap-“-HJ vparatus designed to take advantage of the intense pure beams which will be available at a KAON factory one might hope to reach 10-12. Here the important point is mass resolution, since the main background is shifted in energy by 8.4 MeV. This means using extremely thin windows and wire chambers.M2( A P t * . I I 2)MIAP.11,1 /  TRIGGER SCINT.I j  T - /  - y HODOSCOPESPARTICLE IDENTIFICATION(CERENKOV +n FILTER)/% c,/ $ 3 1 SFig. 5. Spectrometer for k£ ye search.72CP violation is only known to occur in neutral kaon decay.^ Theimportant quantity to measure more accurately is the ratioKL * 7T°ir° /n+-\2 ——  which gives  I » 1 + 6e 1 /e .k °  +  it i r— \ n 0 0 /The theory of Kobayashi and Maskawa, in which CP violation arises from quark mixing, predicts that e ' /e ~ 0.01, whilst the soft CP violation theory of Weinberg, involving an additional Higgs, predicts e'/e ~ -0.02. The superweak theory, of course, predicts e'/e = 0. The best present experiments measure e'/e to ~0.005, not quite enough to distinguish between the competing theories. Presently proposed measurements areexpected to reach ±0.002. A KAON factory should push these limits down to ±0.001, with apparatus which may resemble Fig. 6, proposed by Strovink et al.1 Twin beams of 4-12 GeV/c K° are utilized with a regenerator al­ways present in one of the beams to normalize the + ttx rate in the other beam. The basic idea is to convert one or more of the y's from -r decay to provide additional kinematic constraints and improve separation between events from different beams. BaF2 is used to measure the photon energies.Another experiment which will look for evidence of CP violation is a search for a non-zero normal component of polarisation in Ky3 decay.5 Un­fortunately I have no time to describe this experiment.Fig. 6. Apparatus for measurement of e'/e. A possible K -*■ ir°ir0 event is shown, where two photons convert.73The next topic I would like to discuss is neutrino scattering. Here the most important experiment is to do a 1% measurement of neutrino—elec­tron scattering. Such an experiment would test the standard model in a number of ways. It would yield precise values of sin20w and p(=MW/MZcos0w) giving the relative strength of the charged and neutral current interac­tions. It would provide an important test of the standard model and the radiative corrections to it. The experimental situation is that thus far "■'100 events each have been collected on v^e and vee scattering, and about 500 events on vee. There is an ongoing experiment at LAMPF on vee which collects about 1 event/day. A KAON factory would increase neutrino fluxes by factors of ten to one hundred. With a 900 ton liquid argon TPC shown in Fig, 7, D. Bryman and G. Azuelos1 estimate that about 4000 events/day from vee scattering and about 20 events/day from v^e scat­tering would be detected. The ve beam would come from k £ ■+■ lie ir~e~Vg—  ^and would have an energy around 1 GeV.10mIOmV iM M 1PHOTOTUBES FOR TRIGGER ON SCINT. LIGHT1100 TONS.5x106 COLLECTIONPADS(HIGHLY MULTIPLEXED)Fig. 7. Liquid argon TPC.As an example of an experiment on light quark spectroscopy which might form part of the initial program at a KAON factory I will now describe the proposal of Hemingway and Dixit.1 The basic aim is to fill in gaps in our knowledge of spectroscopic states and to look for glue-balls, hybrids and other exotica. Table II shows the event rates byreaction for a recent bubble chamber experiment compared with the rates expected at a KAON factory. A 105/s beam of 4 GeV/c K_ would be used with the "electronic bubble chamber” shown in Fig. 8. In addition to magnetic analysis and electromagnetic colorimetry the detector utilizes time of flight for particle identification. An improvement of 2-3 orders of magnitude in statistical accuracy over previous measurements would result from ~1 month run.Turning to nuclear physics, the study of hypernuclei will obviouslyplay a central role. A chart of the known lambda hypernuclides is shownin Fig. 9, and surprisingly narrow levels in E hypernuclei have also been seen (Fig. 10). The fact that the hyperon contains a strange quark means that, while the nucleon orbitals are all occupied up to the Fermi surface, the hyperon orbitals are empty. Hence the hyperon can probe deep inside the nucleus. Most of the work so far, done at BNL and CERN, has been confined to the study of light nuclei because of considerations of count­ing rate and energy resolution. With the advent of beams of higher intensity, however, it should be possible to detect de-excitation photons74Table II. Event rates by reaction.X42133 events/pbTRIUMF 100 events/nbK-p -► K+ICA° 4852 3.65xl06K°K°A° 1451 1.09x10sK°K°A° 298 0.22xl06TT”*"tT— TT° A° 59475 45.0xl06n"’'ir+ir-ir-ir0A° 27050 20.5xl06K°K-ir+A° 620 0.46xl06k Jk NTAJ 659 0.50xl067T+ir_E? 10161 7.61xlO6— — + V +it it tr Ej 15300 11.5xl06K+it+iTE" 2123 1.59x10sK° -r”*”ir—“Y 759 0.57x10sK-K+tr_E| 2188 1.64x10sK-K°tr°El" 356 0.27x10s+  - + v ~  tr^tt irE^ 14360 10.8x10sK+K0^" 40 0.03x10sFig. 8. Electronic bubble chamber.Counts / 2 MeVNeutronsFig* 9. Chart of experimentally observed lambda hypernuclei.m h y - m a  h h v )175 225 275 325 375M h y ‘ Ma (MeV)175 225 275 325 375B^o(MeV)Fig. 10. E hypernuclei.76in the residual nucleus. Thus, for instance, one could look at (K~ ir-Y) and (K- ,tt°y ) reactors with and (K_ ,iTy ) reactors with a precision better than 1 MeV even with a poor resolution ir° detector. Olin and Measday1 have written a proposal, ini­tially to study light nuclei between Li and Ne With a beam of 106 K_/s at 550 MeV/c, then a typ­ical hypernuclear cross section of 100 pb/sr would give a count rate of three events/hour in the apparatus shown in Fig. 11. Here Ge detectors with BGO Compton shields detect the photons while theoutgoing ir- enters a spectrometer utilizing a 6 Tsuperconducting dipole and G^Agstrip detectors.Another important topic which will undoubted- ~c ly be studied at a KAON factory is that of doublehypernuclei, formed in the (K~,K+ ) reaction. So ^4 far only a handful of emulsion events have beenseen, because the cross section is so low. With a Fig. 11. Apparatus beam of 107 K-/s, however, one would expect to de-for (K- ,tt-y ) measure- tect approximately 1 event/hour to discrete finalment. states6 in light nuclei, enabling real spectros­copy to be undertaken. Such experiments could give information on the AA force, for example, in the reactionsK“ + 160 -► K+ + i6C AAK+ + I6Nwhich is not obtainable in any other way. It would also be very inter­esting to continue the search for the H dibaryon already begun at BNL.This object has a high degree of symmetry (2u, 2d and 2s quarks in arelative L=0 state) and has been predicted7 to be stable with a mass of ~2.15 GeV. One could search for it via the reactionK- + 3He -v K+ + n + Hwhere the K-1" and neutron are detected in coincidence and the missing mass searched for a peak.Finally, it seems probably that K+ inelastic scattering will form part of the initial program. As can be seen in Fig. 12, KT*- mesons have by far the longest mean free path in nuclear matter of any hadronic probe behaving in some respects like a "heavy electron". With such a probe one can also study the nuclear interior, with particular emphasis on iso­scalar, nonspin-flip single particle transitions. To do such experiments one needs an intense high resolution K+ beam, say 108 K+/s with 200 keV resolution at a low momentum around 350 MeV/c. A spectrometer suitable for this and other high resolution kaon-nuclear experiments is shown in Fig. 13.I will conclude by showing a summary of the proposals received for the representative program (Table III) with the beam parameters for which they call. This list has formed the basis for most of our planning for the experimental hall and beam lines for our KAON factory proposal.5<LUCD_  _  ♦ *7 | - / 2 0 " i ;  aX(f m)77(SewexiM.ROOF SHIELDING/  /  / s s s *SFctnOoHerafL 7~7^  2 ScUki/TiMiDET I ,/v 4 * ( 0 "/////*/////=30° /////*//s30°P, . (M eV /c ) lab2 3 4  5RADIUS (m ater*)Eit/eAiMcKUTjT>od&L£ 5-v/LM Jbrvevs. £*CH *w oc,e l-HAaujk»CLcitit. SPce.Tto3<u>#y H 1 ^ ert.^ k fct£ > V I STAi*JriJrt.eFig. 12. Mean free paths of hadrons, Fig. 13. Kaon spectrometer for nuc­lear physics.REFERENCES1. TRIUMF KAON FACTORY: Representative Experimental Proposals, G. Azuelos and D. Bryman, editors.2. J. Ellis and J. Hagelin, Nucl. Phys. B217, 189 (1983).3. V. Chaloupka, Ref. 1.4. For a recent review, see L. Wolfenstein and D. Chang, Summary of Work­shop on CP Violation.5. K.M. Crowe and C. A. Meyer, Ref. It,6. Based on calculations of Dover and Gal, Proc. Int. Conf. on Hypernuc­lear and Low Energy Kaon Physics, Jablonna, Poland, September 1979, Nuckleonika 25_, 447, 521 (1980).7. R. L. Jaffe, Phys. Rev. Lett. 38_, 195 and 617 (1977).78Table III. Facilities applicable to representative program.Proposal Requirements Beam line/Momentum range (GeV) ParticleFlux per secondfacilityHadron spectroscopy Y* resonances Light quark spectroscopy Quark structure of hyperons Charmonium0.4-2.5 4 0.5 3-7K~K"K~P105105105~107Kl K2 K3 K4 Kl K4K decayMeasurement of n+_/hQ0 CP violation in iL,k l * yeRHC in Kjj2 decay K nvV4-124.51-100.50.6K*K°K+K+1071071071077.5xl06K0K4K0KlKlHypernucleiDouble hypernuclei Gamma-ray spectroscopy Charged particle spectrometer20.5-0.7 0.3-1.1K“K~K~ ,TT +107106107K3 Kl Kl K2v physicsNeutrino elastic scattering Neutrino oscillations1-2 ve>vp 1013 vM day-1/cmv facility2p physicsPolarized proton experiments 30 P 101279HYPERON SCATTERING AT THE TRIUMF KAON FACTORY N.E. DavisonUniversity of Manitoba, Winnipeg, Manitoba, Canada R3T 2N2I. INTRODUCTIONThis review is concerned with the study of hyperons and their inter­actions at the TRIUMF KAON factory. The implicit assumption is made that hyperon physics is worth doing, and specific experiments will be dis­cussed only in so far as they constrain the characteristics of beams that should be available. The word 'beams' is also interpreted somewhat lib­erally. In most of the existing hyperon facilities a real beam of hyperons is produced, transported for some distance, and the products of the decay or other interactions studied far from the production region. As will be seen, at the energies available at the KAON factory, it may often be advisable to produce the hyperons and study their interactions in the same target volume. Since there is a great deal of important physics that falls into this category, such 'beams' are included in the following.Two comments should be made by way of introduction. Firstly, the central design problem is illustrated by the fact that a 10 GeV/c A beam is attenuated by decay to 1/e of its initial intensity in only 69 cm. Clearly, there are advantages in using higher energy and making use of time dilation to extend the distance over which the beam can be trans­ported. This conclusion is strengthened by the observations that (i) the fractional yield of hyperons produced by a high-energy proton beam inci­dent on a heavy target remains approximately constant at 10% as the pro­ton energy increases, that (ii) the distance a beam can be transported increases linearly with energy, and that (iii) the shielding necessary to protect the detectors against hadronic showers increases only logarithmi­cally. On the other hand, tagging of hyperons is possible only for low momentum (Kl GeV/c) hyperon beams. There are thus distinct advantages to both low- and high-energy beams, and it is important to identify those experiments which are uniquely adapted to the KAON factory before embark­ing on design of costly experiments that really ought to be done elsewhere.It may also be noted that the main difficulty facing hyperon physics is not lack of flux in the incident beam used to produce the hyperons. Rather the difficulties are: (1) detector capacity in the face of horren­dous background fluxes, (2) hyperon yield per incident particle, and (3) detailed knowledge on an event-by-event basis of the kinematics of the hyperons. The latter two points are precisely those on which KAON factory hyperon beams are expected to excel. Accepting a limitation to low-energy hyperon beams, it is possible to tag the hyperons so that the beam parameters are well determined. Precision experiments then become possible, and it may even be possible to push precision far enough to institute a new generation of attacks on the fundamental symmetry principles.This contribution is organized as follows. In the second section the static properties of hyperons are discussed with emphasis on those that are pertinent to the production and decay of hyperon beams. Section III is devoted to a discussion of existing hyperon beams, and finally,80Sec. IV discusses some detector characteristics in the context of a specific experiment.II. STATIC PROPERTIES OF HYPERONSThere are many kinds of hyperons, some of which are stable, that is stable against decay via the strong interaction, and some which are not. Those that can be used to form beams have lifetimes on the order of IO-10 s, and can be transported for centimetres or metres depending on their energies. Figure 1 shows the low mass spin 1/2 baryons that make up the 8 representation of SU(3) and the somewhat higher mass baryons that make up the spin 3/2 particles of the 10 representation. TheFig. 1. The particles of the 8 (top) and 10 (bottom) represen­tation of SU(3). The horizontal and vertical axes give the third component of isospin and the hypercharge, respectively.diagram is organized in the usual way with the horizontal axis being the third component of isospin I3 and the vertical axis being the hypercharge Y with the charge Q being given byQ = I3 + Y/2 .Table I. Static properties of hyperonsParticle Mass (MeV/c2 )Lifetime (10_1° s)CT(cm)Decay mode (%)aA 1115.6 2.63 7.07 pir 64.2 +0.64mr° 35.8 +0.65Z + 1189.4 0.800 2.02 pir0 51.6 -0.98nir+ 48.4 +0.071° 1192.5 0.58xl0-9 1.4xl0-9 ay 100l~ 1197.3 1.48 3.71 nir 100 -0.068E° 1314.9 2.90 6.61 Air0 100 -0.4102 1321 3 1.64 3.37 Air- 100 -0.4691672.5 0.820 1.47 AK_ 67.5 -0.017E°ir~ 23.6 -0.013E-ir° 8.9The properties of the hyperons are given in Table I. For reasons ofspace the uncertainties on the quantities shown have been suppressed; details may be found in Refs. 1 and 2. With the exception of the last column the uncertainties are generally on the order of 1-3%. The decay length ct  may be used to calculate the distance d a hyperon beam of a given energy may be transported before it decays to 1/e of its initial intensity:a = ctSy •The parameter a is important for describing the decay of hyperons as it enables one to deduce the degree of polarization of the hyperon through the angular distribution of the decay products. Hyperons are produced via strong interactions and decay via weak interactions. Since strong interactions conserve parity, the hyperon polarization, if non­zero, must be oriented perpendicularly to the scattering plane. On the other hand, the weak interaction does not conserve parity, and this per­mits the polarization of the hyperon to be determined experimentally. To fix our ideas let us consider a spin 1/2 A hyperon in its rest system that decays into a proton and a pion. In this case the relative angular momentum must be either S or P and the total spin must be coupled to 1/2. We let the complex amplitudes for the decay be fs and fp and define a asWe then find that the normalized angular distribution is82where is the polarization of the hyperon before decay and kp is thedirection of the decay proton in the rest frame of the A.The parameter a is a property of the hyperon decay, independent of the production mechanism, and once it has been measured the polarization may be determined from the decay angular distribution. From a practical point of view the degree of polarization may be determined most easily when a is large as in the cases of A, E+ , 5° and E“ . These, then, arethe particles with which we may hope to do hyperon experiments in whichspin observables are determined.III. HYPERON BEAMSThere are several ways in which hyperon beams can be produced. We shall look at 1) inclusive production with incident protons, 2) inclusive production with incident kaons, and 3) exclusive production with kaons.A. ProtonsWhen hyperons are produced in inclusive processes many reaction channels are open, and we do not know which reaction was responsible for a given hyperon. The hyperon energy distribution is usually broad, and it is generally not possible to tag the hyperon in a simple fashion. On the other hand, the energy of the hyperons may be high, and it may be possible to transport the beam for appreciable distances before the flux decays to unusable levels. We should note that the distance a beam can be transported scales with the energy while the thickness of shielding (for hadronic showers at least) scales as the logarithm of the energy.It is important to note again that lack of incident flux is not a problem at most laboratories. Typically, formation of hyperons consti­tutes about 10% of the total cross section at energies of a few hundred GeV, so that the initial hyperon fluxes are quite adequate. However, the large detectors presently in use can tolerate raw event rates up to only about 106 Hz, including all sources of background and unwanted events. A limitation on incident intensity is thus imposed by the number of unwanted particles that pass down the hyperon beam line or that leak through and around shielding.Table II shows some of the characteristics of the Brookhaven nega­tive hyperon beam. At the target, considerably more %'s are produced than it's , so there is no lack of raw flux. At the end of the channel though, the E flux has dropped to below 1% of the pion flux. The drop is even worse for the E particles. The first problem, then, is to choose a production method that maximizes the number of hyperons produced per incident particle. The new feature introduced by the KAON factories is that secondary beams of mesons (K's and it's) are so intense that it is reasonable to consider production of tertiary hyperon beams.B. Kaons (inclusive)Data are available to estimate the fluxes of hyperons using an inci­dent kaon beam, and the results are encouraging. Table III (Ref. 5) shows a comparison between results that could be obtained from LAMPF II or the TRIUMF KAON factory compared with useful fluxes available at Fermilab. The table requires some interpretation. Since the crucial83Table II. BNL negative hyperon beam parametersSolid angle 22 psrLength 4.4 mProduction angle 0°Momentum acceptance (Ap/v) 10%Incident flux 5xl0n interacting protons/pulseIncident proton 28.4 GeV/cMomentum of hyperons (typical) 20.8 GeV/cFluxes/pulse At target ObservedTT- 30,000 30,00047,100 2001,100 2Table III. Relative yields of hyperons per 'background' particleYLAMPF II Y per 106 K~FNALY per 106tt” or pZ~ 5 x 103Z+ 5 x 101* 7 x 1036 x 103 9 x 1032 x 101 1 x 101limitation to hyperon beams is the count rate in the detectors, it is useful for comparison purposes to normalize fluxes so that the total number of particles entering the target/detection system is the same inboth cases. It is also assumed that the energies of hyperons producedwith kaons of ~4 GeV will not allow the hyperons to be transported very far, and that many of the the incident kaons will be incident on the detection system. Column 2 thus shows the total hyperon flux that would be obtained per 106 kaons used to generate the hyperon beam. Column 3, on the other hand, shows the number of hyperons per 106 unwanted pions or protons in the hyperon beam. In both cases, the table gives the maximum number of hyperons that are actually available for experiments after the beams have been reduced so that the total count rate is only 106 Hz during the time the beam is on. We see that there is an advantage to using kaons except in the case of the E- .This is not the full story, however. The Fermilab accelerator ispulsed, with a duty factor of about 10%, while the KAON factory will have a duty factor of essentially 100% with slow extraction. The numbers in column 3 are therefore only one-tenth as large when interpreted as 10684Table IV. Production of hyperonsReaction Lab threshold(MeV)ir+p + K+E+ 890.4ir+n + K+E° 893.2K°E + 894.8K+A 758.4ir~p -*■ K+E~ 904.7K°E° 903.1K°A 767.8iT-n E“K° 909.2K"p + K+H“ 662.8tt°E° *ir°A **tt+E- *K+K°ft- 2688.6K"n -> K°H“ 667.6tt-A *tt~E° *—     — —   'Exothermic reactionsparticles per real second rather than 10® per second while the beam ison. The advantage of the KAON factory then becomes a factor of 7 to 70depending on the hyperon of interest.C. Kaons (exclusive)Let us now look at an option which is not available at the high energy laboratories. So far, we have discussed only beams that have been produced by inclusive reactions with incident beam energies well above the threshold for hyperon production where the unique capabilities ofKAON factory hyperon beams would not be exploited. Let us look then atlow energy beams and concentrate on production with tagging. We restrict ourselves to cases in which there are only two particles in the^final state, a hyperon and the tagging particle, except in the case of ft pro­duction, where the masses are such that even at 4.2 GeV/c 98% of the ft particles are produced by the reaction K-p ->• K+KOft- which is sufficiently simple that tagging is still possible.Table IV shows the main exclusive reactions that can be used for production of hyperons using an incident meson beam. The table also shows the reaction thresholds in the laboratory. Generally, it is advan­tageous to use an incident kaon beam as opposed to a pion beam. Firstly, the cross sections for hyperon production tend to be larger,® and second­ly, there is less pion background to interfere with identification of85hyperon decay products. In addition, since the cross section drops by almost two orders of magnitude per unit of strangeness in the resulting hyperon, it is clearly advantageous to start with a strange particle when producing E's or Si's. In the case of E's, the cleanest reaction is to start with a K- , which already contains a strange quark, and tag with a K+ , which contains a strange antiquark, leaving two strange quarks in the hyperon.IV. DETECTION APPARATUSAt this point, we should try to determine detector characteristics for a hyperon experiment. The choice of an 'easy' experiment is somewhat subjective. It is assumed here that static and decay properties are best studied at laboratories such as Fermilab where the beam can be trans­ported to a decay region well away from the target. On the other hand, the study of elastic scattering and simple reactions should be done with both high and low energy hyperons, and may benefit from the good knowledge of the kinematics available with tagged beams. For the present purposes then, the elastic scattering of hyperons from protons is chosen as the canonical 'simple reaction'. We do not expect any great surprises from a study of elastic scattering, but the data will be entirely new, and should be obtained at least at a few energies for a few of the hyper­on species.Elastic scattering appears to be feasible at least for the A, E1 and E- particles. Since the distance that hyperons can propagate at the energies available at the KAON factory is fairly small, the apparatus must be designed for the production and elastic scattering reactions to occur in the same target volume. Consideration is also limited to an incident K- beam. With these restrictions, and desiring also to be able to extract polarization data from the experiment, the simplest elastic scattering reaction is E-p -*■ S'p. The feasibility of studying elastic scattering of E- from protons has already been examined by the Heidelberg group,7 and in what follows much of their analysis has been adapted for the case of the TRIUMF KAON factory.To begin with, we assume that nearly the full K~ flux of the TRIUMF KAON factory beam line K4 (107 s-1) [Ref. 6] can be used and that a LH2 target 50 cm long by 10 cm in diameter can be used. These are reasonable parameters. Detectors capable of determining the trajectories of inci­dent charged particles at rates of 108 Hz are expected to be available in the near future,8 and the fraction of incident kaons scattered through angles large enough to trigger the detector will be well under 10%. The diameter of the target is chosen so that with a narrow kaon beam thedistance a E“ can travel in the LHj before decaying is less than the dis­tance from the production point to the boundary of the LH2 cell. The cross section for E- production is approximately 150 pb at an energy two pion masses above threshold in the laboratory. This gives a production of 3000 S per second. Most of the S~'s decay before scattering or interacting, but approximately one in a thousand scatters. This gives 3 elastic scatterings per second for 100% detection efficiency.The event geometry is shown in Fig. 2. The incident kaon entersfrom the left and produces a S- plus a K+. The elastic scattering pro­duces a recoil proton. The decay of the E“ produces a ir“ and a A. Sub­sequently the A decays producing a proton and a second ir“ . For good86Fig. 2. Event topology for the production of a 5" hyperon, its elastic scattering from a proton, and subsequent decay to a proton and two i t ' s .event definition and background rejection, one would like to detect all five particles in the final state. However, a highly overconstrained fit to the overall chain of reactions is still possible if the recoil from the S-p elastic scattering is not detected but provided the incident kaon momentum is known and the vertex detection is sufficiently good that energy loss of the particles in the liquid hydrogen of the target may be calculated. Taking into account the branching ratios of the A's into pairs of charged particles, and adopting a reasonable estimate for the fraction of events giving the A decay proton an energy too low to ensure reliable detection, the efficiency of the analysis is about 0.4. The analyzable event rate is thus about 1.2 events per second. An experiment with 5 S- energy bins, 30 angular bins and 10,000 counts per bin on the average would require a total of 1.5xl06 events. At a rate of 1.2 events per second, this would require slightly more than two weeks of beam time.The detector for this experiment is fairly large. Since the hyperons are produced with low energy, the decay products are emitted into 4tt sr and since the detection scheme relies on detecting the majority of the particles most of the solid angle must be covered by the detection sys­tem. A central detector and magnetic field must be available to determ­ine the momenta of the particles and their initial directions. With a central detector of the JADE type,9 and a magnetic field on the order of0.5 T, the particle momentum may be determined to about ±4% for particles of momentum less than 2 GeV/c and using data from a single track only.9 >l9 The magnitude of the momentum of the E after the scattering may then be determined with an uncertainty of about ±1% to ±3/£ and the direction with an uncertainty of about ±10 msr.7 The trigger would involve detection of the K+. The central detector would yield a good trajectory for the K+, but it would also be essential to distinguish it from a tt or a proton. This can be done by reaction kinematics, but there would still be some confusion with the K~p tt"1"! reaction. For low energy K+'s the delayed detection of the K+ decay provides a unique signal. Consequently, it would be valuable to include a lead glass wall of sufficient thickness to stop the majority of the K+'s. The range of a 500 MeV/c K+ is about 30 cm in lead glass. A thicker wall would certainly be possible, but at a correspondingly higher cost.87CONCLUSIONThe conclusion is that hyperon scattering experiments even of fairly high statistics appear to be quite feasible at the KAON factory. The detection systems are larger than what one normally uses in present-day intermediate energy experiments, but the technology to carry out the experiments is essentially all available.REFERENCES1. Review of Particle Properties, Rev. Mod. Phys. 56^ , #2 (1984).2. L.G. Pondrom, Phys. Rep. 122, 58 (1985).3. A. Fumio et al., Nucl. Instrum. Methods 220, 293 (1984).4. V. Hungerbuehler et al., Nucl. Instrum. Methods 115, 221 (1974).5. T.W.L. Sanford, Los Alamos Report LA-9386-MS, LAMPF II - Also a Hyperon Factory, 1982 (unpublished).6. KAON Factory Proposal, TRIUMF, 1985 (unpublished).7. W. Brueckner et al., Max Planck Institut Internal Report MPI H-1982- V9, 1982 (unpublished).8. M. Salomon, private communication.9. G. Gidal, B. Armstrong and A. Ritten, Particle Data Group, LBL-91 Supplement, Major detectors in elementary particle physics, 1985.10. H. Drumm et al., Nucl. Instrum. Methods 176, 333 (1980).88M O R E  M U O N SJess H. BrewerDepartment o f Physics, University o f British Columbia Vancouver, B.C., Canada V6T 2A3AbstractThe reported performance of the former SREL muon channel on the AGS at BNL indicates that raising the primary proton beam energy from 0.5 to 28 GeV, all other things being equal, increases stopped p + and p~  rates by factors of about 30 and 100, respectively. Combined with target and beamline adaptations at a 30 GeV KAON Factory, this can mean stopping p~  beams of nearly 500 times the intensities presently available — an increase comparable to that experienced when the Meson Factories began operation a decade ago. Such a large quantitative change can again make a qualitative difference to muon science, from rare-decay GUTs tests to trail-blazing on the Precision Frontier to muon-catalyzed fusion to the chemistry and solid state applications of pSK.1. The Intensity Frontier for Stopped M uons:Nearly two decades ago Canada, Switzerland and the USA decided alm ost simul­taneously to build Meson Factories whose main purpose was to increase the available fluxes of slow pions and muons by two to three orders of magnitude, so that (a) known ir / p interactions could be measured more precisely and thus provide better tests of theory (the so-called Precision Frontier); (b) hitherto undetected n/p be­haviour could be discovered, or at least confined to within smaller upper limits (rare decays); and (c) formerly difficult and exotic n/p applications to fundamental sci­ence could be made routine and familiar enough to qualify as applied particle physics technology (e.g., pion cancer therapy or /liSR).The Meson Factories have succeeded admirably in all these endeavours and in others as well, until today they constitute the main stage upon which intermediate energy physics is performed. The character of the science pursued at the Meson Factories has of course evolved [as intended — see (c) above] and the elementary particle physics applications have begun to refocus upon the lightest strange particles (as witnessed by the proposed KAON Factories of the coming decade) while the pion and muon programs take on a more and more technological character. One may therefore be tempted to look upon it/p science as the province of the Meson Factories and exclude it from plans for the KAON Factories; this would be a grave error.With a few extremely simple arguments it can be shown that the KAON Factories will also (perhaps unwittingly) provide an increase in the flux of slow muons by a89factor comparable to that achieved by the Meson factories over their predecessors, thus ushering in yet a new “golden age of muons” . It is easy to speculate about the possible consequences of this windfall for muon science, and most of this paper will be devoted to such musings.2. Energy Scaling of Slow M uon Production:It has been known for some tim e1 that the effective cross section for production of 200 M eV /c positive pions at a lab angle of 60° was approximately 44 times larger at an incident proton energy of Tp =  28 GeV than at Tp -  730 MeV. This suggests a “rule of thumb” that the pion production efficiency of a proton beam scales with its kinetic energy — or equivalently that the number of pions produced scales with the net power in the proton beam. This rule of thumb would imply that the same muon channel and the same production target would generate 60 times as many slow 7r+ (and therefore 60 times as many stopping p + ) if nothing were changed but to increase the primary beam energy from 500 MeV to 30 GeV.Production targets for 30 GeV may of course be several times thicker than those for 500 MeV; but the concomitant increase in the number of pions produced is par­tially offset by an increased probability of reabsorption of the pions in subsequent nuclear interactions — which in turn depends upon the geometry of the target and beamline. An overall rate improvement factor of 2-4 may be expected from typi­cal thick targets.2 We may therefore anticipate a net increase of 100-250 for p + production from a thick target at 30 GeV.For negative muon production, another factor of 4-5 is realized from the equal­ization of 7r+ and 7r~ production cross sections at high energies where the (3,3) resonance no longer dominates. This brings the theoretical increase of p~  flux (in the same beamline looking at a longer target irradiated by the same proton current) to a factor of 400-1250 — truly an impressive improvement.Additional improvements relative to present TRIUMF muon beamlines could be realized by designing new channels “from scratch” to take advantage of the last decade’s advances in muon production technology. Indeed, TRIUMF is presently undertaking an upgrade of the M9 channel with Japan’s assistance in the form of a superconducting solenoid. However, the point of this argument is to compare the performance of a KAON Factory muon channel with the best muon facilities available today, which are already well optimized — not with the present TRIUMF backward muon channel (M20), which has never been billed as a high-flux muon channel. The world standard at present is set by the pE l muon beam at SIN.1 D .F . Cochran et al., Phys. Rev. D 11, 3085 (1972); D. Berley, et al., IEEE Transactions on Nuclear Science, 1973 Particle Physics Conference; J. Fox, R.C. Co­hen, A.M. Sachs and E. Zavattini, BNL Internal Report; W.J. Kossler, Hyperfine Interactions 8, 797 (1981).2 W. J .  Kossler, private communication.903. E m p ir ica l E v id e n c e  from  th e  A G S :Such predictions “on paper” do not carry as much weight as the performance figures of an actual muon channel operating at 30 GeV proton energies; real engi­neering problems often lead to regrettable compromises. Fortunately, a group at BNL (Kossler et al.) has followed through on these encouraging projections and built a stopped muon channel from the old SREL and Nevis magnets3 on the AGS, where it has been commissioned with 28 GeV protons.4 The production target used in the tests was a 7.5 cm long, 1 cm 2 Pt target, which is presumably the sort of target one would use at the KAON Factory. The luminosity of 90 M eV/c positive muons in the final beam spot was the equivalent of 1.75 x 105 /z+ s 1cm 2jiA  1, or about 40-50 tim es the equivalent luminosity on the present TRIUMF “backward-decay muon channel, M20. The corresponding factor for n~ would be about 200. These numbers m ust be adjusted for the relative channel efficiency of M20 vs the BNL channel, which is impossible to know exactly, but is probably of the order of unity.The empirical /i*  rate improvement factors for 28 GeV vs 500 MeV thus fall short of expectations by at least a factor of two; evidently there are indeed some “engineering demons” at work. Of course, even a factor of 200 in n flux would facil­itate qualitatively new experiments. I will straddle the fence between conservatism  and optim ism  and assume that the actual rates achievable in a judiciously designed “backward” n~ channel at the KAON Factory will be roughly 500 times the presently available intensities (for the optimized SIN superconducting muon channel operating at 125 M eV /c backward muon momentum, roughly 2 x 107 n /s at 100 fiA  5) which works out to nearly 1010 n~ /s at 100 fiA.This rather astronomical number (nearly a nanoampere of muons!) belies the fact that one would not choose the high momentum (125 M eV /c) that gives the largest gross flux of muons, but the lowest momentum  possible, for the bulk of applications of the new channel (see below). The flux is a rapidly decreasing function of decreasing momentum below 100 M eV /c, and in the most interesting range of 20-30 M eV/c  present facilities offer less than 104 n~/s at 100 *tA; to increase this number to 106 is extremely enticing.W hat new experiments could be done with such muon beams, and what criteria do they define for the design of the optimum muon channel to take advantage of the new fluxes?3 H.O. Funsten, Nucl. Instr. and Methods 94 , 443 (1971).4 BNL Report.5 SIN Report 5 /78 /2000 , update August 1979.914. Experiments with M ore Muons:As with other parts of the proposed KAON Factory experimental program, it is impossible to know today which of the anticipated experiments will turn out most exciting a decade from now when they are well underway; in fact, as presaged by the “pSR  explosion” at the Meson Factories, many of the most lively future programs at the KAON Factory are probably not anticipated at all today. Nevertheless we may extrapolate present activities to higher intensities, using the lessons of the transition to Meson Factories, and predict at least part of the impact on muon science.4.1 . Rare Decays Revisited:The main obstacle to reviving old rare muon decay programs (such as the p —► search) is that the previous round of experiments were typically limited by back­grounds as much as by statistics; it will be difficult to reduce backgrounds still lower in the higher energy environment of a KAON Factory, but perhaps not impossible. As experience at SIN has shown,6 it is vital to remove the final “muon extraction” section of the channel from regions of high background and shield it carefully against scattered secondary particles coming down the vacuum pipe; this argues for a rather long muon channel in the backward direction.Assuming that backgrounds can indeed be conquered, upper limits on reactions such as p —> e7 , p + e~ —► p ~ e + and p —► e conversion could be lowered by two or three orders of magnitude, to the level of 10“ 15 or 10“ 16. This is still well above the current predictions of GUTs, but such a measurement would surely provoke some theoretical interest. At the moment the most promising of these studies would seem  to be the muonium-antimuonium conversion experiment, but this and the p —> experiments use positive muons, for which the KAON Factory provides “only” a factor of about 100 improvement.This illustrates the need to retain a positive muon beam (probably surface muon) capability at the new laboratory, even though the biggest gain will be for negative muons.4.2. New Outposts on the Precision Frontier:Another notable achievement of the Meson Factories has been the use of higher statistics and cleaner measurements to test standard-model predictions (such as in normal muon decay) to higher precision, thus eliminating uncertainties more thor­oughly (e.g., pushing the lower limit on the WR boson mass to above 400 Gev). The power of this approach is now fully appreciated. However, since this method requires the recording and analysis of more data in order to obtain more precision, and since many of the high-precision experinments at the Meson Factories are already pushing the limits of the performance of electronics and computers, one must be cautious about predicting enormous advances through higher statistics alone.6 C. Petitjean, private communication92Fortunately, higher intensity can provide more than just brute statistics; it also allows one the option of higher quality muons beams. The importance of this aspect to high precision is well illustrated by TR IU M F’s pSR experiences: when the beam can provide orders of magnitude more muons than can be utilized, one can throw away all but the very best part of the beam, selecting very small beam spots, very narrow momentum spreads, etc., and thus perform a much cleaner experiment at the same rate. It is therefore vital to include this principle in plans for the KAON Factory: the new muon channel should be designed as an instrument of unprecedented delicacy, not as the world’s biggest bludgeon.4 .3 . M u o n ic -A to m ic  P h y sics:One of the main activities in the early years of the Meson Factories was measure­ment of muonic X-ray spectra, with applications in QED, chemistry, nuclear physics and even medicine. Many of these measurements were plagued by Stark broadening of the X-ray lines due to the high-density media needed to stop the high-momentum  muon beams. Preliminary tests at TRIUM F7 indicate that a low-momentum p~ beam stopped in a gas target can alleviate these problems dramatically.The most interesting muonic atom is p~p,  muonic hydrogen. Its atomic physics properties (X-rays, fine and hyperfine structure, Lamb shift, etc.) are of fundamental importance, since muonic hydrogen is identical to normal hydrogen in every respect except the “electron” mass and its consequences. It is also a unique laboratory for the elementary process p~p  —* — the prototype semileptonic weak interaction— and the rate for this weak capture process from the triplet (F = l)  state would, if measured directly, provide an ideal test of the induced pseudoscalar weak coupling constant. Unfortunately, muonic hydrogen is so tiny that it acts like a “fat neutron” in ordinary matter, and the triplet spin state is rapidly quenched by collisions with other protons in all but the most rarefied hydrogen gas targets.8 Moreover, the weak capture process is only about 10“ 4 of the muon decay rate in the singlet (F =0) state, and only a few percent of that in the triplet state, so any experiment on the (F = l)  capture rate will require very large numbers of muons. The future of such measurements seems therefore to depend upon extremely low-momentum, high flux negative muon beams which can be stopped in low-density gas targets of m odest size.4 .4 . M u o n -C a ta ly z e d  F usion :The bright side of the small size of muonic hydrogen is that it readily approaches a second hydrogenic nucleus and forms a molecular ion such as dpt in which the deuteron and triton are drawn together closely enough to fuse, liberating energy in a fast neutron and (usually) releasing the muon to catalyze another fusion. The current record number of fusions per muon9 is more than 100, which is not enough to generate power, but is certainly enough to generate intense interest. The unpleasant aspects7 G.M. Marshall, C.J. Oram and J.B. Warren, private communication.8 A. Alberigi Quaranta et al., Phys. Rev. 177 , 2118 (1969).9 S.E. Jones et al., Phys. Rev. Lett. 56 , 588 (1986).93of large quantities of tritium can be avoided by studying dpd fusion, a less vigorous but equally interesting process. It is not inconceivable that the KAON Factory might some day incorporate a muon-catalyzed fusion target as an intense neutron source-, but such plans are now highly speculative.4.5 . M ore jtSR:As mentioned earlier, most pSR  experiments are of the “time-differential” variety, which is limited by the necessity to wait for one muon to decay before another is allowed to enter the sam ple.10 This restricts the p + stop rate to between 104 and 105 s - 1 , which is readily available already. In the last year, however, there have been new developments in the “time-integral” pSR m ethod,11 in which any number of muons may enter the target at once. This erases the limit on how much intensity can be profitably used, and opens the way for a “fast spectroscopy” with muons. It is therefore incorrect to assert that pSR has been completely satisfied by the muon beams of the Meson Factories. In particular, time-integral “resonance” /rSR techniques may offer a convenient solution to the intrinsic problems of /*~SR — low polarization and loss of decay-electron rate due to nuclear muon capture. In any case, the field of / / “ SR, which has been largely ignored to date for these reasons, should “take off” with increased intensity much as p + SR did with the advent of the Meson Factories.4.6. M uon-Labelled Chemistry:The most dramatic application of the new time-integral pSR  techniques has been to level-crossing resonance (LCR) measurements of muonic radicals — i.e., organic molecules to which a muonium atom has been added to form a large paramagnetic molecule incorporating the m uon.12 It is now easy to identify such chemical species in situ, make detailed comparisons of their structure with that of the analogous hydro- genic radicals, and study their chemical kinetics in any medium. Since these radicals differ only imperceptibly from their hydrogenic cousins, this method provides a new way to investigate the kinetics of exotic chemical reactions that may be important in biology, etc. It is impossible to predict the scope of applications of this new tool at such an early stage.4.7. Ultra-Slow  M uon and M uonium  Beams:In the last year several experiments have been performed demonstrating that thermal beams of /*."*" and M u  13 can be produced and accelerated into beams which should then have exceedingly low emittance and be nearly perfectly monochromatic. Although the efficiency for production of such beams from secondary targets is veryA. Schenck, Muon Spin Rotation Spectroscopy: Principles and Applications in Solid State Physics (Adam Hilger Ltd, Bristol and Boston, 1985).11 TRIUMF Annual Report, 1985.12 R .F. Kiefl et al., Phys. Rev. A  34, in press (1986).13 D.R. Harshman et al., Phys. Rev. Lett., in press (1986).94low, the development of such a muon source as part of the primary production target, along with increased production efficiency and advances in this still-young technology, could provide exciting possibilities such as a “muon microscope” or a variable-energy muon synchrotron.Similar advances have been made very recently in the production of thermal muonium atoms in vacuum from hot foils14 or from fine silica powders,15 but these lead to direct experimental applications (e.g., in muonium-antimuonium conversion) rather than “muon technology” developments.If these fast-breaking developments are to be exploited in the future, the KAON Factory experimental hall must include a production target which can be used for engineering research on secondary beam production technology. This is not incom­patible with other uses of the same production target, but it requires that one piece of extremely valuable real estate be effectively left undeveloped for speculative purposes.5. Summary of Desirable Design Features:The most obvious feature of “the” KAON Factory stopped muon channel that can be forseen today is that it should be optimized for “backward” negative muons, since there is a 4 -5  times larger flux gain available for than for p + . This does not mean that the channel cannot be used for p + science, nor even that it can only be used for backward muons; the p E 4 superconducting solenoid channel at SIN has been successfully tuned for surface muons and pions, and is in fact one of the best pion channels at SIN .16 However, its main strength is still negative muons, and this will be even more true at the KAON Factory.However, there should not be only one KAON Factory muon channel. The huge variety of positive muon applications that have sprung up in the past decade are already well served by the fine p + beamlines operating on the 500 MeV “future in­jector” (which will continue to provide beam to existing facilities), but at negligible expense the NMR-like techniques of pSR  can be made available to a much larger group of chemists and solid state physicists, much as has been done for synchrotron radiation users at “photon factories”. An array of extremely inexpensive low solid angle surface muon channels (essentially just holes in the shielding with a few perma­nent magnets) could provide 4 MeV p + beams of a few million m uons/sec, adequate for all known p + SR applications, if only the target shield is properly designed for beam access in the backward direction and adequate experimental space is reserved for this purpose. It should be noted that surface muon beams can be transported over long distances without significant losses; such channels are extremely compact in their transverse dimensions (witness TRIUM F’s M15) and could even be laid in trenches in the floor of the KAON experimental hall, leading to satellite areas away from the most coveted floor space, much like the present M15 building.14 A. Mills, K. Nagamine et al., Phys. Rev. Lett., in press (1986).15 A. Olin et al., Phys. Rev. Lett., in press (1986).16 C. Petitjean, private communication.95Such a “muon farm” would usher in a new era of “applied” muon science made feasible by the last decade’s development of /zSR techniques, which awaits only a wider availability of muons and ^SR facilities for “outside” users. Having made this point, I will focus again on the figures of merit for “the” stopped p~  channel.5.1 . Beating the Backgrounds:There are three main backgrounds to contend with in a backward channel: electrons, pions and neutrons. The electrons come down the first half of the channel along with the injected pions, but they should not make it through the momentum  selection in the muon extraction section, since the backward muons have a drastically reduced momentum. The key to efficient removal of electron contamination is in delivering all the electrons without scattering to a well separated secondary beam dump buried in a shielding wall. An additional cleanliness factor can be obtained by making the muon extraction section longer, with two bends — the second close to the final focus — to eliminate higher energy electrons produced by muon decay in flight.Like the electrons, pions should be completely eliminated by the final momentum  selection, but some escape down the channel by scattering off the walls. In the case of pions it should be possible to effectively discourage such scattering by a suitable arrangement of “pion baffles” in the injector so that pions cannot hit the solenoid walls. The pions are much easier to stop in the trap at the end of the solenoid, but where they stop they produce neutrons, which are more difficult to shield against.Neutrons are probably the worst kind of background for most p~  experiments, since they are very difficult to shield against and in many cases they mimic the foreground events very convincingly. There are two kinds of neutron backgrounds to worry about: the plentiful secondary neutrons from the production target and the pion-induced tertiary neutrons originating in the pion trap of the beamline and (unless precautions are taken) all down it. The first type are fast only in the for­ward direction, so a backward takeoff minimizes the problem; then it is a matter of shielding, so that a longer channel is highly advantageous. The second type must be eliminated by the geometry of the muon extraction, which must take the muons “around the corner” from where the pions are dumped.5.2. Slower is Better:For many reasons, some of which have been outlined above, the most desirable feature of a new high-intensity muon channel is low momentum, so that muons can be stopped in rarefied gases and/or thin, small condensed matter targets. The lesson of the surface muon beam 17 is that stopping luminosity is the figure of merit for muon facilities; the great opportunity of the KAON Factory for muons is the chance to duplicate the immense advantages realized by the surface muon beam for negative muons. This must be a primary concern guiding the design of the such a facility.17 T. Bowen, Physics Today (July 1985 cover story).966. Conclusions:With any luck I have made two points: ( l)  the increased efficiency of 30 GeV protons for producing slow negative pions makes possible an immense gain in intensity for negative muon beams — as much as a factor of 500 over what can now be achieved; and (2) some care should be taken in designing muon channels for the KAON Factory to exploit the most desirable capabilities that this unprecedented flux makes possible.I have not done more here than to point the way to what I believe to be the principles that should guide such a design study. These are:(a) One muon channel should be built to deliver the highest quality, lowest mo­mentum, highest polarization, highest flux negative backward muon beam possible (highest priorities in italics). This channel should be thought of as the most delicate instrument ever built for high precision, low background muon experiments.(b) This channel should be designed around a long superconducting solenoid fed by a high solid angle, good optics pion injector at backward angles to the production target. The length is needed to allow the elaborate shielding necessary to fulfill the quality requirement in (a). Special attention must be devoted to safeguards against scattering of contaminating particles into the muon beam.(c) There should also be several (as many as possible within geometrical con­straints at the production target) low solid angle surface muon beams extracted into long beamlines buried in trenches. These can be inexpensive and small if built from modular units incorporating permanent m agnets,18 and will thus be out of the way of high energy beamlines. The resultant “muon farm” will make the proven appli­cations of muons to condensed matter physics and chemistry available to a broader community.(d) One production target at the KAON Factory must be designed to be compat­ible with an experimental program of secondary beam engineering — research into such exotic possibilities as thermal or M u~  beams or the acceleration of these to higher energy monochromatic muon beams with correspondingly infinitesimal emit­tance. These dreams could only be realized using the increased production efficiency of the KAON Factory, and we cannot afford to ignore their promise for the future.18 K. Nagamine, Annual Report of Univ. of Tokyo Meson Science Lab., 1985.97EXPERIENCE WITH TARGET AREA SHIELDING AROUND THE CERN 26 GeV PROTON SYNCHROTRONA.H. Sullivan CERN, 1211 Geneva 23, SwitzerlandABSTRACTThe CERN 26 GeV proton synchrotron (PS) has been in operation for more than 26 years, mainly supplying beams of high-energy protons to target areas for secondary particle production. Many radiation measure­ments have been made around the shields of these beam areas to assure that they are being safely operated and to understand the radiation fields that are found. It is shown that radiation levels outside the lateral shield of a target area can be reasonably predicted using relatively simple relations derived from shielding experiments. The level of the highly penetrating muon component in the forward direction from targets and beam dumps has also been studied and it is shown that it can be reasonably predicted in a comparison between calculated and measured fluxes. Radiation resulting from particle cascades in the forward direction from targets are best determined from Monte Carlo calculations as is demonstrated in a comparison of high-energy particle fluxes from a beam dump.Current ideas on allowable levels of radiation and the criteria used to assess the effectiveness and acceptability of accelerator shields is presented.Shielding problems related to the CERN antiproton production faci­lity are discussed where it was found that the radiation being trans­mitted through the lateral shield of the target area is insignificant compared that escaping along holes and ducts through the shield. In par­ticular, problems encountered with negative pions coming through the beam channel in the target area shield are discussed and the proposals for reducing the radiation levels in an updated version of the machine are presented.INTRODUCTIONThe proton accelerators at CERN range in energy from the 600 MeV synchrocyclotron (SC) to the 450 GeV super-proton-synchrotron (SPS). In the very near future there will be electron machines in operation for the Large Electron Positron Collider (LEP). Relevant experience with target area shielding of interest for the KAON project is that related to beams from the 26 GeV proton synchrotron (PS). This machine is a 200 m diameter ring that has been in operation for over 26 years mainly to provide protons for secondary particle production, for experiments in halls around the machine. Intensities have continuously increased and the machine now works routinely at 1.6 x 10 protons per pulse every 2.4 sec (1.1 pA) at 26 GeV giving an average beam power of some 30 kW.At the time when radiation shielding problems were new, many measurements were made to understand the radiation spectrum produced in high-energy particle cascades in shields, in order to assure that expe­rimental areas could be properly shielded.98Since a number of years the PS has been superseded by the SPS as the source of particles for physics experiments and its rdle has been limited to that of an injector to the higher energy machine. The reali­zation that antiprotons could be collected and stored, and that the SPS could be used as a proton-antiproton collider has resulted in the PS being fully exploited in supplying the proton beam for the antiproton production facility. In the recent past the PS has been used to produce high-intensity neutrino beams in a specially designed target area, as well as being a useful source of test beams for detectors to be used in experiments in the higher-energy machines. Over the years a considerable experience has been accumulated in measuring radiation levels and, where necessary, adjusting the shielding of target areas supplied by beams from the PS machine.SHIELDING MODELS AND PARAMETERSThe earliest opportunity was taken after the PS started operating, to check nuclear cascade models and to obtain the essential data on which to base reasonable shield designs. The first shielding experiments used beams of 10 and 19.2 GeV protons at intensities of about 10 /sec, scattered out by internal targets in the PS . The beam attenuation curves that were measured in a steel absorber are shown in Fig. 1 and the shielding parameters deduced from these measurements are summarized in Table 12.Particles/protonFig. 1. Beam intensity measured over 46 mm diameter by C-11 activation as a function of depth in a steel absorber.Table I. The build-up and attenuation of particle flux in steelApparentPrimary Flux Attenuationproton build-up mean-freeenergy factor path (g/cm )19.2 GeV 3.5 17010 GeV 3 1451 2The particle flux determined is that as measured using the C -* C11 reaction in an activation detector. This reaction, induced in carbon primarily by protons, neutrons and pions, has a threshold of 20 MeV and a reasonably energy-independent cross-section at high-particle energies.99This flux can be measured with a very high sensitivity when the carbon activation of plastic scintillators is measured and hence the method can be used down to the low flux levels (< 1 part./cm2/sec) normally found outside a shield.The lateral distribution of hi^h-energy secondary particles around a beryllium target was also measured which indicates an empirical rela­tion, where flux is proportional to the inverse square of the emission angle. The measured dependence of secondary particle flux on angle around internal beryllium and copper targets in the PS machine* is shown in Fig. 2, where the data can be seen to also fit an inverse square re­lation. A similar angular dependence has been found in measurements at other accelerator centres . These measurement results lead to a relation for the high energy particle flux ♦ at a distance R(m) and at an angle of 8 degrees from a target of:♦ = particles/cm2/sec (1)R 8where I is the incident beam intensity, n the interaction probability and k a constant related to the target material with the experimentally determined values as given in Table 2.Table II. Secondary particle production constant in targetsTarget material k(deg'2 m'2)Beryllium 0.29Aluminium 0.37Copper 0.50Uranium 0.73Fig. 2. Angular distribution of high- energy particle flux around copper and beryllium targets. Lines drawn are inverse square relation.The above relation was determined for 19.2 GeV protons. Measurements outside the shield perpendicular to the beam direction at other energies show a practically linear dependence on primary proton energy. Combining the flux dependence on angle and energy with the attenuation by a shield, gives for the particle flux outside a target area shield of:Re 1 at ive fluxl.1010100_ 0-05 K ,E .Ll e-t^  particles/cm2/sec (2)R2 02where t is the thickness of shield traversed (in the direction 8) and X the mean-free path of high-energy hadrons in the shield material. For practical shielding calculations, where often the density of the shield­ing material is not precisely known, the values of X, as given in Table 3, are assumed.Table III. Attenuation mean-free path of high-energy hadronsMaterial Densityg/cmm|Pg/cm cmsIron 7.4 147 20Concrete 2.35 117 50Heavy concrete 3.2 123 37Earth 1.9 115 60Lead 11.3 213 19It has been found that the flux of particles of energy greater than 20 MeV can be converted to dose equivalent outside a shield on the basis of 1 Sv = 1013 particles m (3 particles/cm /sec equivalent to 1 mrem/h). Such a conversion is a best average value based on comparison of parti­cle flux measured using carbon-11 activatiop detectors and dose equiva­lent determined from an array of instruments . Furthermore, a conversi^ of flux to absorbed dose rate on the basis of 1 Gy = 4 x 10 particles m-2 (10/cm /sec equivalent to 1 mrad/h) has been found to agree with dosimeter measurements near targets to within about a factorof 2. .The above relation for dose equivalent can be compared with othjr shielding models, notably the Moyer model for radiation shielding , where good agreement is found over emission angles from 20 to 110 de­grees. The relation has also been tested against radiation measurements outside lateral shields around target areas and appears to predict levels to better than a factor of 3 at all angles greater than 15*. This has also been found for the radiation originating from beam dumps if the dump is considered as a target situated one mean-fjee path into the dump and there are two interactions per incoming proton .Taking into account the complicated shielding geometry that occurs in a real target area shielding layout and uncertainties in attenuation lengths and shielding thickness, this degree of agreement can be consi­dered good.Hadron shielding in the forward direction from the target cannot be so easily described, due to enhanced particle ^implication and recourse is made to detailed Monte Carlo computations . A comparison of radiation levels predicted along the side of a 40 cm diameter iron beam dump, using the above relation with that of the Monte Carlo progray FLUKA11 is shown in Fig. 3. The computation gives a result in stars/cm'101in iron which2has been empirically converted to particle flux on a basis of 30 part/cm giving 1 star/cm in iron. As can be seen the relation for the flux from a beam dump underestimates radiation levels at large distances into the dump where the secondary radiation is mainly forward directed, when compared to the detailed calculations. However, for proton energies greater than about 8 GeV, target area shielding in the forward direction is usually determined by the necessity to reduce muons to acceptable levels rather than the hadrons.Fig. 3. Estimated high-energy par­ticle flux escaping through the side of a 40 cm diameter iron beam dump as a function of distance along dump for 10 protons of19.2 GeV incident. Curve (a) is analytical expression and (b) is Monte Carlo Code FLUKAMUON SHIELDINGHigh-energy muons originate from the decay of secondary pions pro­duced in a proton interaction. They are highly penetrating and many metres of iron may be required to bring the fluxes down to acceptable levels. Although a muon of a given energy has a determined range with a negligable chance of interaction, the energy distribution that results from the spectral distribution of pions produced in the forward direc­tion from a proton interaction makes the muon flux appear to be exponen­tially attenuated in a shield. The muon flux at a distance x metres down be^ jp from a target in which protons of energy E GeV interact is given by := 8.5 x 10 2 —  e at^E p/m2/interaction (3)where A is the flight path (in metres) of the pions before they can interactt is the thickness of shield that has been traversed and a is an effective muon energy loss rate.When t is in metres, o has a value of 22 GeV/m for iron and 7.6 GeV/m for concrete. When the proton beam goes directly into a beam dump, the pion flight path is considered to be 1.8 times the hadron mean-free path in the materialThe relation is reasonably independent of target material and is valid up to values of at/E = 15 and for proton energies up to about 50 GeV.The muon flux emerging from a shield is usually in the form of a well-defined beam in the forward direction. The diameter of the beam,P a r t i  c l e s / c m 2102which is assumed to have a gaussian profile can be estimated from the relation:4.6x ..><* ■ 7 ®  * <4)where d is the full width at half maximum of the beam.The radiation levels from muons have been measured on many occa­sions behind beam dumps in experimental areas. The results of measure­ments made under reasonably controlled conditions are compared with pre­dictions based on the above formula in Table IV. Where absorbed dose has been measured this has been converted to muon flux on the basis of 3.6 x 10 p/m2 per Gy or 1 mrad/h = 10 p/m /sec.Table IV. Comparison of calculated and measured muon fluxes behind dumps in target areasRef.Protonenergy(GeV)Target Flightpath(m)Beam dumpMug>ns/m' per 10 incident protonsFe + (m)concr. (m) meas . calc.12 26 none 0.36 7.2 5.6 3 1.714 24 none 0.36 4.8 4.8 9 1914 24 2 cm W 9.7 4.8 4.8 19 3714 24 3.5 cm W 9.7 4.8 4.8 23 4815 19 40 cm Be 11 5.6 4.0 13 1015 15 40 cm Be 11 5.6 4.0 1.3 0.9M u o n s / m  2Measurement of a muon beam profile behind a dump consisting of 7.2 m of iron followed by 5.6 m of concrete with 26 GeV protons incident showed a diameter of 1.1m compared to 1.5 calculated using Eq. 4. Measurementsof muons from other target areas have shown that the beam diameter agrees with predicions to within 20% at both19.2 GeV and 10 GeV .The relation given by Eq. 3 also appears to agree with muon fluxes com­puted using the computer code MUST0P at least up to 30 GeV as is showen in Fig. 4.Fig. 4. Muon fluxes per incident pro­ton for iron beam dump as a function of depth as calculated using analytical expression (solid lines) and from pro­gram MUST0P (dashed lines).103Hence, the muon attenuation equations appear to reasonably predict radiation levels from real target areas and could be used to calculate near optimum shields for a given set of conditions.RADIATION SAFETY CRITERIA FOR DETERMINING SHIELDING EFFICIENCYThe limits to which an area outside a shield may be irradiated at CERN is determined by the CERN Radiation Safety Policy which takes into account the International Commission on Radiation Protection (ICRP) principle of keeping doses as low as reasonably achievable, while at the same time maintaining a certain flexibility for CERN operations. The annual basic limits that are set are:mSv rem1. Dose to a radiation worker 15 1.52. Dose to others on site 5 0.53. Dose to fence of CERN 1.5 0.154. Dose to nearest inhabitant 0.5 0.05To implement these limits the CERN radiation safety code requires, in addition, that areas where dose rates can exceed 2 mSv/h are inter­locked with the beam and that the dose rate in freely accessible areason the site averages less than 2.5 pSv/h. The maximum levels to be con­sidered for shield design would be half the above values and could be very much lower, where this can be achieved at reasonable cost.Radiation levels are measured using a system of four detectors to make as good an estimate of dose equivalent as possible . The levels found are mainly neutrons that decrease in energy with increasing dis­tance from the shield. The variation in the radiation levels are follow­ed with a system of installed monitors both near the machines and at the site border. A grid of some 120 passive monitors, consisting of Li and natural Li thermoluminescent crystals in a 15 cm polythene moderator in­tegrate both neutron and gamma dose over long periods and with whichannual site dose contours can be established1 . These c o n to u r s  give anoverall indication of the efficiency of the shields around the various CERN installations.RECENT TARGET AREA SHIELDING EXPERIENCEFor the past few years the radiation levels on the CERN site have been largely dominated by neutrons coming from the antiproton production facility. The facility consists of a heavily shielded target area where the antiprotons are produced, followed by a 50 m diameter accumulator ring (AA) where the particles are collected and cooled before being re­turned to the PS machine for onward transmission to experiments. The AA ring was built in an open pit and owing to the complexity of the machine104it was only possible to install relatively light shielding. The level of neutron radiation escaping through the AA shield is shown in Fig. 5 as afunction of distance round the ring. These neutrons result from inter­actions by the 3.5 GeV/c pions that are coming down the injection line from the target area with the antiprotons and account for more than 95% of the total radiation from the machine.Radiation transmitted through the shield of th| target area - whichhas an overall attenuation of nearly a factor of 10 - is low comparedto the radiation that escapes by way of the shafts through the shield for the emergency exit and ventilation ducts. During normal operation the dose rates near the surface building over these shafts go up toabout 0.1 mSv/h, but may increase by a factor of 50 if the beam is acci­dentally lost in critical regions. A similar situation was encountered around the neutrino target area where the lateral radiation attenuation was of the order of 107 and radiation could only be detected outside the shield when losses occurred in the injection line near the personnel access shaft.The antiproton facility is to be improved in the near future to increase the collection rate by afactor of up to 16 and carefulconsideration has had to go into ensuring that operation will not be limited due to radiation prob­lems The shielding required around the improved facility AC£>J. (Antiproton Collector On Line) has been estimated on the basis of experience with AA. Estimates show that even if iron is used instead of concrete, there will not be enough place for suffi­cient shielding around the ring. The solution that has been adopted has been to incorporate a spectrometer system (dog-leg) in the injection line inside the heavily shielded target area, such that more than 80% of the unwanted pions will be separated from the antiprotons and interact in a collimator Furthermore it is expected that the pions that can still get down the injection line will have a momentum bite that is matched to the acceptance of the antiproton collector ring so that they will have an increased chance of inflight decay before they can interact in the machine structure. The shield required around the AC0L ring has been reduced by nearly a factor of 2 in critical regions with the incorporation of the dog leg in the injec­tion line. The resulting layout of the target area is shown in Fig. 6. As can be seen the target will be opposite the emergency exit and venti­lation shafts that lead to the surface requiring that the surface build­ing is shielded if radiation levels from the target area are to be kept to a small fraction of that from the main ring.2Neutrons/cm /secFig. 5. Neutron flux on 80 cm con­crete shield of AA ring as a func­tion of distance around the machine with 101 protons per 2.4 sec. on target.105Fig. 6. Proposed cross-section and layout of antiproton productiontarget^rea shield to ensure less than 5 x 10 n/sec escape with2 x 10 protons per 2.4 sec of 26 GeV on target.The improved shielding and precautions in the updated antiprotonfacility should result in substantially lower radiation levels than at present in spite of an increase in particle accumulation rate of afactor of 16.CONCLUSIONSExperience gained with target area shielding at CERN shows that near optimum shields can be designed for target areas using proton beams in the 10 to 30 GeV range. Derived shielding equations, both for the lateral hadron shield and the forward direction muon shield, give a good estimation of radiation levels that are actually measured. Recent expe­rience with secondary particle production facilities shows that the sources of external radiation originating from a target area are, in de­scending order of importance:1061. Unwanted radiation accompanying the secondary particle beam.2. Neutrons leaking along ducts and shafts from the target area.3. Radiation transmitted through the shield.The lateral shielding required around a target can be readily esti­mated and radiation leaking along ducts can be relatively easily con­tained. The primary source of radiation originating from the target area, the pions accompanying the antiprotons through the channel in the shielding wall, required sophisticated particle separation techniques to reduce the radiation to proportions that can be handled within con­straints imposed on conventional shielding by the layout of the anti­proton ring. Experience with target areas shows that the bulk shieldingrequirements are generally well understood and that it is the unexpected beam loss or modification to the shield that can cause significantlevels of unwanted radiation outside a target area shield.REFERENCES1. J. Geibel and K.H. Reich, Nucl. Instr. & Meth., 32, 45 (1965).2. J. Baarli et al., Nucl. Instr. & Meth., 32, 57 (1965).3. L. Hoffmann and A.H. Sullivan, Nucl. Instr. & Meth., 32, 61 (1965).4. S. Charalambus et al., CERN DI/HP/97 (1967).5. W.H. Moore, Brookhaven N.L. Report AGSCD-62 (1966).6. K. Goebel and J. Ranft, CERN Yellow Report 70-16 (1970).7. A.H. Sullivan, Proc. Conf. Accelerator Dosimetry & Experience,USAEC Conference 691101, p. 625 (1969).8. J.B. McCaslin et al., LBL report LBL 14699 (1985).9. A.H. Sullivan, CERN HS-RP/IR/80-10 (1980).10. G.R. Stevenson, CERN TIS-RP/IR/86-04 (1986).11. J. Ranft et al., CERN TIS-RP/156/CF (1985).12. A.H. Sullivan, Nucl. Instr. & Meth. in Phys. Res., A239, 197 (1985).13. D. Keefe and C.M. Noble, UCRL 18117 (1968).14. M. Hofert fit al-, CERN HS-RP/044 (1979).15. M. Hofert and Ch. Yamaguchi, CERN HS-RP/TM/80-42.16. M. Hofert, CERN HS-RP/IR/78-36 (1978).17. G.R. Stevenson, CERN HS-RP/TM/79-37 (1979).18. J.W.N. Tuyn, CERN HS-RP/012/CF (1977).19. K. Goebel, ed., CERN TIS-RP/171 (1986).20. A.H. Sullivan, CERN PS/AA/ACOL Note 84-23 (1984).21. E.J.N. Wilson, CERN Yellow Report 83-10 (1983).22. C.D. Johnson fit al-, IEEE Trans. Nucl. Sci. NS-32 No. 5, 3000 (1985).107TRIUMF KAON FACTORY SHIELDING & ACTIVATION CONSIDERATIONS: EXTRAPOLATING TO 30 GeV, 100 yA PROTON BEAMSI. M. ThorsonTRIUMF, 4004 Wesbrook Mall, Vancouver, B.C., Canada V6T 2A31. INTRODUCTIONThis is intended to be a "broad-brush" sketch of our view of the shielding design and radioactivity production estimation problems for a KAON factory and the approach we are using to solve them.To estimate the radioactivity production and shielding requirement reliably it is necessary to rely primarily on experimental measurements and experience with the minimum extrapolation in either energy or current from existing facilities. Monte Carlo calculations, although they are useful in such extrapolations, are not in themselves a reliable enough guide, at least for estimation of low probability phenomena such as radi­ation dose transmission through thick shields. They are more useful in estimating the dominant operating radiation fields, particularly hadron fluxes in the immediate vicinity of targets and other beam spill points for purposes of estimating residual radioactivity and heat production. One must never lose sight, however, of the fact that the accuracy of the results from such calculations is directly dependent on the accuracy of the fundamental reaction data that goes into them. Less sophisticated calculational methods that allow a better understanding of the dominant phenomena, and therefore the critical data, are usually preferable to more sophisticated methods that bury such detail in a sea of untrackable data.2. DESIGN EXTRAPOLATIONThere are in principle two directions from which to extrapolate the shielding requirements and radioactivity production estimates; in current from the existing 30 GeV proton accelerators, e.g. the AGS at Brookhaven National Laboratory and the PS at CERN, and in energy from the 500-800 MeV meson factories at TRIUMF, SIN and LAMPF. Qualitatively, at least, the deep transport of radiation in shields is more easily and more confidently estimated by extrapolation by ~2 decades in current from the existing 30 GeV facilities; the final two decades of attenuation that the KAON factory will require beyond that needed for existing facilities can be fairly confidently extrapolated using the previously determined, empirical dose relaxation lengths, without excessive contingency require­ments. To obtain this definition, however, will require a much more detailed understanding of the shield geometry and source and radiation field distributions in the existing facilities than the TRIUMF KAON fac­tory designers now have. We hope we can obtain this detailed information at an acceptable cost and burden to the operators of the present 30 GeV facilities!The extrapolation of heat and radioactivity production estimates for the KAON factory design can rely on extrapolations in both proton current and energy from existing facilities. The extrapolation to higher energies from the existing meson factory facilities probably gives a clearer108GENERATION NUMBERFig. 1. Cascade collision multiplicities for hadrons with energies above 100 MeV from 0.5 and 30 GeV protons on thick iron shields as a function of hadron generation number. The 30 GeV estimate was made with CASIM code (Ref. 5) and the 0.5 GeV estimate was based on a 5-energy-group calculation using the Bertini-Alsmiller intra-nuc- lear cascade results (Ref. 6).qualitative understanding of the problems that must be solved, at least in the experimental target areas. The experience at LAMPF is seen as being particularly relevant where their more general approach to primary target design, as compared to TRIUMF, is more typical of the unavoidable residual radioactivity and heat production problems expected at 30 GeV facilities. At TRIUMF the conceptual design approach for high current primary beam target stations has been to bury the bulk of the induced radioactivity inside a large monolithic shield block into which were in­serted the required target and beam monitor and control elements. It has been by and large successful in the sense that the cost in constraints onsecondary channel designs has been more than compensated by the reducedremote handling requirements and costs. Vestiges of this approach will still be applicable in the design of KAON factory target areas in that there will always be a residual radiation field advantage in using fairly massive components that are not opened to the core except where absolute­ly necessary, and then preferably after transporting them to special facilities. The relief that this approach will afford for remote handling is not, however, expected to be as significant for the KAON factory as for the original TRIUMF. The reason for this is indicated in Fig. 1 which shows the number of hadrons of various types with energy above ~100 MeV in each generation that make collisions in an infinite iron medium, initiated by protons of 0.5 GeV and 30 GeV.^ For the 0.5 GeV case the numbers, as well as the energy, decrease monotonically from the firstgeneration. For the 30 GeV case the number of hadrons does not fall109below the original value for about 10 generations. Because of the heavy forward bias that the original primary proton momentum imposes on the progeny, each generation of reactions is displaced approximately one mean free path deeper into the shield. The reaction rate spatial distribution is thus spread over an extended volume, usually exacerbated by unavoid­able voids in the vicinity of primary beam targets. Thus containment within a monolithic target shield is not even approximately feasible, as it is in the 0.5 GeV case.3. MUON PROBLEMSBecause the muon ranges are so much shorter than the shield thick­nesses required for stopping the hadrons, principally neutrons, produced by 500 MeV protons no, even qualitatively, similar problem exists at 500 MeV facilities corresponding to the muon problem around 30 GeV facil­ities. The increase in intensity does, however, have some impact on the problem. At existing facilities with average operating beam currents of the order of 1 yA the muon radiation fields from production by decaying particles traversing solid components and shields are often below the intensity level that is biologically constraining. With the two orders of magnitude increase in average current for the KAON factory the general problem is now comparable in field strength to the particular problems found at existing facilities where the muon sources have been deliberate­ly enhanced, also by about two orders of magnitude by explicitly insert­ing a pion (and kaon) decay path of order 20 m, instead of the 0.2 m mean free for these particles in iron shielding.The estimates of the muon intensity downstream of the primary beam targets and shields, published previously,2 were based on the simplifying and conservative assumption that all of the pion momentum went to the muon, disregarding the component going to the neutrino. When this effect is included the original pion spectrum, based, as before, on the Sanford- Wang3 empirical model for Be, is degraded to that shown for muons in the forward directions in Fig. 2. Figure 3 shows the lateral distributions at various depths in a continuous iron beam-stop shield (of p=7.5 g cm-3) under three sets of model assumptions. The first two are the cases indicated above of ignoring and including the modification of the muon spectrum by the in-flight decay of the relativistic pions. Both presume that the pion spectrum from proton collisions with iron is the same as the Sanford-Wang prescription for Be and that only secondary pions con­tribute significantly. This is perceived to be an adequate description only at the higher muon momenta, and concomitantly greater depth in the shield. The estimate does not include any allowance for Coulomb scatter­ing of the muons or pions or the spreading of muon directions by the iso­tropic angular distribution of the muons in the rest frame of the pions, believed to be an adequate to good assumption at all but the lowest pion momenta. The normalization is approximately a factor 3 higher than the estimate described in the original proposal2 because of the assumption of complete reaction of primary proton beam and the 30% greater (~23 cm) pion mean free path used for the present estimate. The third model used to make the estimates shown on Fig. 3 is the empirical formulation by Sullivan.1* The agreement over the lateral and depth ranges shown is regarded as somewhere between gratifying and remarkable. The only per­ceived difficulty between the models is the implied discrepancy between11010-2101010"10'1010"-610-8MUON S P E C T R A  FROM 7T DECAYS ;  FROM 3 0 GeV PROTONS ON B e @ 0 ">x  X  tt-SP EC TR U M° 6 x x  O  /J. -S P EC TR U M° o x x°0**x° f x° 0 X Xx xo xO X O  x O  x o Xo .°  *; O :10 15 20 25 30Pp (GeV/c)Fig. 2. The it  spectrum at 0° from 30 GeV protons on Be (Ref. 3) is compared to the resulting y spec­trum after correction for the kin­ematic transformation of the un­correlated two-body decay tt y+v from the rest frame of the tt meson to the laboratory frame.ooco.crUJax=3OZ3Fig. 3. The estimated lateral dis­tributions of muons at various depths in an iron beam-stop shield (p = 7.5 g cm-3) is shown based on (a) the Sanford-Wang (Ref. 3) it  spectrum with no kinematic correc­tion for the y decay spectrum, (b) the kinematically corrected y spec­trum, and (c) Sullivan's (Ref. 4) empirical formula for muon produc­tion and penetration in shields.the Gaussian and the nearly linear exponential lateral distribution at large angles for the Sullivan model and the Sanford-Wang "data", respec­tively. This ambiguity could be of some practical significance in estimating the lateral requirements for muon-stopping shields for protec­tion of personnel and, more particularly, experimental detectors.4. HADRON SHIELDINGThe definitive experimental measurements of the hadron cascade de­velopment have yet to be done, even at the lower energies of the present- day meson factories. The essential problem is the lack of firm experi­mental data on the closely coupled secondary hadron spectra and angular distributions for all energies up to the primary proton energies. Lacking this support the Monte Carlo calculations do not provide the necessary confidence in answering the many detailed questions on operating radia­tion fields that will be crucial for many of the low branching ratio experiments likely to be mounted at any KAON factory.MUON FLUX IN Fe AT VARIO US I  DEPTHS d FOR 3 0 GeV PROTONS— # —  IM T  1986 ( S .-W .) =-S P EC TR U M— O —  IM T  1986 ( S .-W .)  tt-SP EC TR U M AHS N IM  A 2 3 9 . I9 7  :(1985)d = 8 md= 12m= 16 mIllIt may be time for a collaborative effort on a series of experimentsaimed at extrapolation to higher currents similar to those mounted in thepast to provide support for the design of shields at higher energies.REFERENCES1. C. Yamaguchi, TRI-DN-85-11, Some energy deposition and absorbed dose calculations for KAON factory, April 1985 and private communication.2. TRIUMF KAON Factory Proposal, September 1985.3. J.R. Sanford and C.L. Wang, AGS Internal Report BNL 11299, Empirical formula for particle production in P-Be collisions between 10 and 35 GeV/c, March 1967.4. A.H. Sullivan, Nucl. Instrum. Methods A239, 197 (1985).5. C. Yamaguchi, TRI-TN-85-3, May 1985. A manual for hadron cascade Monte Carlo code CASIM and its muon version CASIMU at TRIUMF.6. R.G. Alsmiller, Jr., M. Leimdorfer and J. Barish, 0RNL-4046 (1967), Analytical representation of non-elastic cross-section and particle emission spectra from nucleon-nucleus collisions in the energy range 25 to 400 MeV, and H.W. Bertini (1968), private communication of 500 MeV nucleon-nucleus collision results.112A REMOTE COMPONENT SERVICING SYSTEM FOR THE TRIUMF KAON FACTORY EXPERIMENTAL HALLC.R. Mark and W.M. Cameron TRIUMF, 4004 Wesbrook Mall, Vancouver, B.C., Canada V6T 2A3ABSTRACTWith the extent of high level residual activity anticipated for the 30 GeV experimental channels, comprehensive remote handling procedures will be essential to operation of the facility. Any KAON factory design philosophy must incorporate all systems Integrating with the remote hand­ling requirements. The proposed servicing system provides for handling of beam line components via specialized remote equipment manipulated by the overhead crane. Routine repair of components is accomplished, after their removal from the beam line, in a conventional hot cell facility. Positional alignment of beam line elements is performed in this same facility. Massive concrete piers forming the walls of beam line canyons offer the most cost-efficient means of shielding construction. Removable shielding blocks within the trench provide direct access to the compon­ents. Necessary services would be routed along the top level of theseshielding piers, with vertical supply ducts located at each element pro­viding a shielded transition of services to the radiation-hard environ­ment. Precise manipulation of the crane load is necessary for alignment of equipment to the oblique axis of experimental channels. Therefore, a hook rotational control is required in addition to the orthogonal axis of motion intrinsic to travelling cranes. Crane specifications must reflect the need for shielding beam line component transport flasks.INTRODUCTIONA significant portion of high energy beam will be deposited down­stream of all particle production targets. This beam loss will result inthe generation of substantial levels of residual radioactivity in the secondary channels. The specific level of activation expected is of the same order as existing TRIUMF beam lines. The extent of induced activity along the channels, from targets to beam dump, will encompass numerous beam line components. Routine servicing of these areas will require com­prehensive remote handling procedures.For a remote handling system to accommodate the diverse nature of equipment present in this project, a unified handling philosophy must be adopted. The criterion for this philosophy must suit all necessary sub­systems, conditions of operation, and foreseeable demands on the facili­ty. The most elegant solution must consider reliability, serviceability and cost, with the essential requirements for maintenance, repair, and provision for future modification.Detail design standards for all relevant components can then be pre­pared from this conceptual outline. The operation of any integrated remote handling system requires that these standards be strictly adhered to during all phases of construction.113Experimental hall design factors integrating directly with the remote servicing system include:• Stationary and removable access shielding• Supply of beam line component services• Beam line component alignment• Handling philosophy for:Beam line transport elements Beam configuration devices Diagnostics monitors Production targets Beam dumps• Overhead cranes• Building layoutGENERAL PHILOSOPHYThe proposed remote handling system will provide the most reliable working conditions with the shortest beam downtime. All component re­pairs, other than routine maintenance procedures, are to be performed outside of the beam line in an integral hot cell facility. Remote repair of components in situ has proven extremely inefficient. The unpredictable nature of equipment failure, and restricted beam line access, contribute to lengthy and complex remote operations that often result in additional equipment damage and high personnel exposures. Only the most specific tasks for which complete remote procedures have been previously commis­sioned should be conducted in the beam line. For the majority of repair operations the facility design should allow easy removal of equipment from service, with the remote handling system transporting these compon­ents directly to an adjacent hot cell facility.In this philosophy the major servicing of components depends upon proven conventional hot cell techniques. A comprehensive mechanical hot cell will provide the ability to diagnose faults, perform repairs, and commission equipment to operational standards, all under ideal conditions.BEAM LINE CANYONSThe layout of beam lines and details of the shielding are crucial factors in the proposed remote system. The single large building hall, beam line layout and shielding concept as described in the TRIUMF KAON Factory Proposal^ are most suitable for this conceptualized system.Massive poured-in-place concrete piers are most cost efficient for the predominant volume of shielding. Pier design should specify a semi­permanent construction method, with separation joints between sections. Considering the strict function of the KAON factory, any major beam line revisions should be limited, but might require reconstruction of the shielding piers. Provision for remote disassembly of active sections would allow for beam line revision subsequent to facility start-up.Construction of the largest discrete pier sections, consistent with the building crane capacity, would assure the most stable design. Due to the permanent nature of this construction, the large surface area on top of the pier shielding is made available for location of beam line services.114Fig. 1. Canyon cross section. Fig. 2. Access shielding arrangement.SHIELDINGBeam lines are constructed on the valley floor of the vertical wall canyons formed by adjacent pier sections, as shown in Fig. 1. Removable shielding blocks stacked within the canyon trench provide beam line access. For handling purposes there would be numerous individual blocks spanning the 3 m wide trench.Total shielding required over the beam line is estimated at 6 m: 2 m  of iron plus 4 m of concrete. For the largest feasible block size to reduce handling repetition, a three-layer stack appears most expedient: one 2 m layer of Fe and two 2 m layers of concrete block. Overlapped stacking of blocks is necessary to eliminate radiation streaming from the beam line elevation.To maintain a 50-ton crane capacity the first layer (3 m x 2 m) iron shielding blocks are limited to a maximum of 1.07 m in length. Equivalent concrete shielding blocks (3 m x 2 m) can be 3.6 m long. For uniform stacking of blocks, all length modules should be in even multiples of the bottom layer block length (1.07, 2.14 or 3.21 m) . Therefore, finalaccess block module length is determined solely by crane capacity.Short block modules (1.07, 2.14 m) only increase handling procedures without significantly improving the beam line access/shine angle. Given the inherent physical size of high momentum channel elements, most access situations will require removal of more than one iron layer block. In these instances the largest block length module (3.21 m) is advantageous as it reduces the total number of blocks removed. This configuration of beam line access shielding is shown in Fig. 2.Although residual levels on top of the iron shielding are antici­pated on the order of 20 mR/h during shutdown periods, hands-on mainte­nance cannot be considered for any elevation as radiation shine rises prohibitively with the removal of even one iron shielding block. All removable access shielding must be equipped with suitable remote liftattachment fixtures.Portions of the stationary and removable iron shielding downstream of production targets will require water cooling. Provision must be made for removal or repair of defective cooling circuits within the stationary sections. Handling of cooled access blocks must allow for supply of cooling services.115SERVICES SUPPLYOperational services for cooling, electrical power and controls are required for components in the beam line trench. The method selected for supply of these services will have significant impact on the construction, servicing and operation of the facility.It is proposed that services be routed directly down the canyon wall from pier surface to beam line elevation. Confining the shortest length of radiation-hard services in the shielding is preferable formaintenance, as well as for low initial cost of materials. Locating high current power supplies adjacent to the beam line might significantly reduce the required capacity of these supplies, as well as the initial and operating costs.Individual supply ducts service each beam line component as shown in Fig. 3. These 6 m high removable units are situated within vertical recesses in the canyon wall at discrete locations for every component. Each unit contains a full complement of service conduits, adequatelyshielded to maintain the primary shielding integ­rity. Ducts are easily removed by crane. Trans­port shielding and bag­ging of the lower end during removal reduces personnel exposure and limits the spread of re­movable contamination.Conduits transform through the duct from conventional materials at the surface to the high radiation environment at the beam line. Radiation-hard remote connec­tions suitable for use on the lower end of the ducts are readily available for all elec­trical services. The design of equivalent joints for fluid and vacuum service is an existing technology. Mating of service connections on the duct, with those on the beam line compon­ent, is through an indexed manifold. All operations in­volved with connection of ser­vices would be performed manually at the upper level.A flush canyon wall is maintained by recessing the ducts into the concrete pier, permitting standardized access shielding with the ability to remove shielding while compon­ent services remain connectedFig. 3. Service supply duct.116for 'on-line' inspection purposes. Spare service voids should be planned in the pier shielding for contingency use. Similar duct voids in the re­movable shielding blocks would create a significant sacrifice in service and handling ability.REMOTE HANDLING SYSTEMAll installation, removal and transportation of active equipment within the experimental hall would be performed by the remote handling crane system. High level residual activity in both equipment and areas serviced requires remote manipulation for reduced personnel exposures, and due to the weight of components involved a substantial crane system is necessary. The proposed remote system integrates these functions.Beam line access shielding, particularly the first layer iron blocks, will require remote handling for crane rigging and management of block stacking. A dedicated purpose framework suspended from the over­head building crane will be used, this having the necessary mechanism for remote connection and manipulation of appropriately designed blocks. A single framework would handle all three levels of shielding. Block removal procedures should allow for interim storage of highly activated Fe shielding blocks within a portion of the previously unstacked trench to provide shielding and control of removal contamination.Beam line component handling would be performed in a similar manner, with a dedicated framework supported and maneuvered by the overhead crane. Manipulative steering of the framework is performed by a speci­alized four-degree-of-freedom trolley/block assembly on the travelling crane. Remote viewing and positional feedbacks from the framework assist with location adjustments, attachment latching and guided removal of the component. Precise positional control could be provided by bracing the framework against the canyon walls for rigidity, with independently driven mechanisms for 'X,Y' location.Specialized Pb shield segments may be attached to the framework for local radiation shielding during removal of active equipment. Totally enclosed containment shielding flasks would be used for handling of pro­duction targets or other highly active components.BUILDING CRANEThe proposed experimental hall remote handling system establishes specific requirements for the design of the building crane(s).In the most suitable building layout beam line channels are arranged within a single 75 x 140 m structure. A travelling overhead crane, for general equipment handling in this area, would span the 75 m width. The massive size of high momentum channel elements will be reflected in any specified crane capacities.The remote system crane must operate in the beam line canyon with a minimum of trench shielding removed. Limited vertical crane access favours a single trolley crane design over the conventional ganged multiple hook. This single hoist must have sufficient payload capacity for the heaviest beam line component, plus the additional weight of any Pb shielding required.A single trolley overhead crane would furnish the necessary three- axis (X,Y,Z) motion to spatially locate the remote framework. Additional117Fig. 4. Remote system crane. Section through typical beam line canyon.control of the load rotational axis is necessary for alignment of the framework to the principal axis of the beam line channels.All the remote crane system requirements are satisfied by a 50-ton capacity double-girder top-running single trolley crane, with the inde­pendent rotational motion provided by a motorized turntable bearing located between a lower crane sheave block assembly of unique design and the specialized remote framework, as shown in Fig. 4. A second trolley unit with lower capacity and higher block lift speed would accommodate the majority of normal crane activities.Reduced exposures to operating personnel are attained by distance from the exposed canyon work area. An independent travelling crane cab would be advantageous in this respect. The remote manipulation of crane handling equipment will be performed entirely from the crane cab. Posi­tional feedbacks from the crane end trucks, trolley, hoist and turntable displayed in the cab would allow rapid, accurate positioning of the load. Documentation of positioning coordinates for desired tasks would be routine. A programmable controller could assist with operations.HOT CELLSThe hot cells facility for maintenance, repair and alignment of active components must be located in the experimental hall beneath cover­age of the remote crane. Separate hot cell styles will be utilized to perform the numerous varied operations required.A large mechanical service cell will accommodate repair of beam line components. It must have overhead crane access for introduction of transported equipment and would be equipped with shielded viewing windows and master-slave manipulators. Powered gantry manipulators, or an in-cell crane, may be needed for handling of heavy tools and component sections. A positioning turntable for service items would allow optimum repair access. CCTV viewing could assist with some inspection operations.The cell design should include the ability to perform diagnostics and inspection of components at full power and under normal operating conditions. It would be most valuable if these services were supplied through the actual service supply duct employed in the beam line by that component. Component alignment would also be performed in this cell.118A second hot cell of more specific design would be used for servic­ing of devices with high level residual activity. A fully enclosed, negative pressure hot cell is required to control the spread of removable contamination. All servicing, repair and routine exchange of production targets would be performed in this cell. This must have entry for fully contained, flask transported devices. Conventional hot cell technology would be used in the design of this cell.COMPONENT ALIGNMENTCritical alignment of components is necessary for beam line opera­tion. However, high radiation levels in the trench preclude any form of manual survey, and automated line-of-sight systems are unsuited to the required beam line layout. As canyon access should be limited to a mini­mum trench exposure, the accuracy of measurement performed from directly overhead might well be unsuitable. Instead it is proposed that alignment of all components be performed outside of the beam line trench. These procedures will be performed in the hot cell facility, through the use of a master alignment jig.Although additional means of confirming installed component align­ment might be advantageous, personnel access must be restricted from the exposed trench opening. Commercial systems are available to perform alignment with respect to a given benchmark, but any form of alignment will be complicated by the shielding constraints. The proposed procedure requires an alignment reference framework be built into the floor of each beam line trench during initial construction of the hall. This frame need only be installed to an accuracy well within conventional building construction practices. The purpose of the framework is to provide a series of index locations along the full length of the trench. A compre­hensive survey of the framework is then conducted to measure and record each index position with respect to a desired datum.For alignment operations the adjustable master jig, located in the hot cell, would substitute for the trench framework. The analogous index locations on the master jig could be positioned to duplicate locations for the recorded survey of any section of trench framework. The base frame of each component provides the mechanism for all necessary adjust­ments. This base locates to the index positions of both master and trench frameworks. Conventional survey techniques would be used to per­form or confirm any alignment of a component mounted on the master jig.For alignment of components with high residual activity the master jig must be located in a hot cell with provision for remote adjustment of the component and viewing of alignment instruments.REFERENCES1. KAON Factory Proposal, TRIUMF, September 1985, p. 5-1.119TARGETS FOR HIGH INTENSITY BEAMS AT CERN DESIGN,OPERATIONAL EXPERIENCE AND DEVELOPMENTSR.Bellone, A.Ijspeert, P.Sievers European Organization for Nuclear Research (CERN)Geneva, SwitzerlandABSTRACTWith the increasing energies and intensities, available today in high energy proton accelerators, the targeting of strongly focused beams is becoming especially problematic. The resulting temperature rises and, in particular for fast extracted beams, thermal shock effects and vibrations have to be considered. Solutions for the reliable targeting of 450 GeV/c-protons have been found for fast extracted beams up to 2x1013 p+/pulse and for slow extracted beams up to 3x1013p+/pulse. The design of the various target stations used for the CERN-SPS, as well as for the antiproton source at the PS are described and the operational experience and limits are given.INTRODUCTIONThe SPS-pr imary proton target stations were, with only one exception designed for 400 GeV/c and 1013p+/pulse (ppp). During the design phase in 1974-1977 only little experience existed for the targeting of multi-hundred GeV-beams1. Therefore, most of the design of the SPS-target stations was based on Monte Carlo nuclear cascade codes2 from which parameters such as temperature rises, thermal stresses, deposited dose and remanent radio activity were extracted. However, also the above computer codes contained considerable extrapolations from experimental results achieved at relatively low energy ranges of several 10 GeV, available at that time. Thus, an overall precision of only about 50% could be expected from these computations. A considerable amount of anticipation and "guess work” had therefore to be applied for the design of the SPS-target stations.After the commissioning phase of the SPS in 1976 the targeted proton intensity rose steadily from several 10l2ppp and currently up to 1013ppp are used for targets in slow extracted beams while up to 2x1013ppp have been applied for neutrino experiments with fast extracted beams.Initially three target stations were installed in each of the two slow extracted SPS-beams (west extraction and north extraction) while two stations were constructed for neutrino physics and one for an R.F.-separated beam. Later, two target stations for very high intens­ities of up to 3 x 1013ppp were built for an extension of the north area (NAHIF) while the three west stations were removed in 1982 and replaced by a single station.Moreover, in 1980 a target with high performance was built for the antiproton production of the AA project at the CERN-PS and recently, in view of the improvement of this antiproton source for ACOL, targets with even higher performance are under consideration.In the following we present an overall review of the various target designs and operational experience gained in the past at CERN.120BASIC CONSIDERATIONSA target of high efficiency consists ideally of a free rod with a small diameter which is irradiated with a beam which has at most the same diameter as the target. This ensures that at least those second­aries emitted with a non-zero angle with respect to the incident proton will quickly escape laterally from the target material and will thus have a high escape probability. Moreover, for many experiments, a small geometrically well-defined and stable particle source e.g. for momentum resolution, is important which again can be achieved with free, thin targets even when the lateral beam position and the focus are unstable. The need for thin targets with light weighted supports and irradiated with strongly focused beams leads to some technical problems which arediscussed hereafter.From extensive studies3 based on Monte—Carlo cascade computations it resulted that the most adequate material for targets used at SPS energies are beryllium and possibly graphite. Beryllium is particularly suitable due to its long radiation length, which governs the electro­magnetic part of the cascade, but also due to its exceptionally high specific heat which reduces the temperature rise caused by the energy deposition density from the cascade.The peak temperatures reached inside a target rod depend on the time structure of the proton burst. For fast extracted beams with a duration of t=23|us the radial temperature profile follows directly the energy deposition density which is strongly peaked along the target axis. Due to this non-uniform temperature field, thermal stresses are created which however, vanish after some milliseconds when the lateral thermal diffusion sets in. These stresses may lead to target failures well before the melting temperature of the target is reached. For slow extracted bursts with a duration of t = 1-2s where ample time forthermal lateral diffusion is available the temperature rises more or less uniformly over the cross-section which leads to lower peaktemperatures and negligible thermal stresses. A Be-rod with a diameterof 3mm and irradiated with a fast extracted 400 GeV/c-beam at 1013pppand a diameter of 3mm (4 sigma of a Gaussian proton densitydistribution) would be heated instantaneously along its axis by about 200°C, which creates thermal stresses close to its yield strength, while in the case of slow extraction the uniform temperature rise would be only about 100°C.In addition to the above phenomena, thermal shocks and vibrations may be produced by fast extracted beams^. An infinitely short proton burst would create an instantaneous temperature rise and a compressive axial stress in the target rod since any axial thermal expansion is delayed due to the mass inertia of the target material. When, however, the burst has a finite duration t, relaxation during the burst can already occur over a characteristic length l=t*c at both free ends of the target (c:velocity of axial stress resp. relaxation waves, i.e. the velocity of sound). With t=23 /js of the fast extracted SPS pulses, 1 is about 27 cm in beryllium. Clearly, the length of the target rods must be kept well below this value in order to minimise the shock effects which otherwise might lead to additional stresses, vibrations or even buckling. If long targets are required they must be made of a string of several short rods. In this context it should be stressed that buckling121may also occur when the fast extracted beam hits the target slightly off axis. In this case, one side of the target becomes hotter than the opposite side. This creates thermal bending moments in the target, initiating lateral vibrations and buckling.The energy deposited in the SPS targets may rise up to 1% of the incident beam energy, i.e. 6 kJ/pulse with 101^ppp at 400 GeV/c which represents a time average power of about 500W (SPS repetition time T= 12 s). Clearly, this requires adequate cooling which for "thick" targets is conveniently done by conducting the heat laterally through the target material proper into a heat sink, as e.g. air-cooled fins. For "thin" targets, however, forced gas convection cooling or, in the limit, radiation cooling has to be used.TARGET DESIGNS AND PERFORMANCEStandard target stations for slow extracted beamsThe Be-targets for the SPS slow extracted beams are placed in an air gap between the vacuum pipes of the primary proton and the secondary beams, sealed off by thin metallic windows. The targets proper consist of horizontal Be-plates with a height of 2-3 mm, a width of 160mm and lengths of up to 500mm. Thus targets are provided which are "thin" in the vertical direction while the heat can be discharged horizontally through the Be-plate into cooling fins of adequate surface which are clamped along either side of the plate. Natural air convection would, at the limit, suffice even at lO^ppp, but for further safety an air stream with a velocity of about 3m/s is directed onto the fins. To prevent vertical buckling of the plates, one side must be allowed to expand freely in horizontal direction.In general, several target plates are stacked vertically in a common frame, the target box. This is suspended via shafts which tra­verse the top steel shield through narrow slots, from a support frame mounted above the top shield. This principle is shown schematically in Fig. 1 and allows the remote selection of the desired target through a vertical displacement of the unit. The motorised mechanisms are effi­ciently protected from radiation by the top shield such that commer­cially available, standard components can be used and that rapid inter­ventions and repairs can readily be done. With the thickness of 0.5-0.8m steel on the top, average doses of up to 107 Gy have been accumulated since the start up of the SPS and the remanent activity is of the order of 1mSv/h, depending on the details of the operation mode and the cool-down time.As illustrated in Fig.1, the upstream (TBIU) and downstream (TBID) secondary emission beam monitors are installed in a similar fashion to the target box. The centre of the TBIU is aligned with the target axis and serves to steer and focus the incident beam. The TBID measures the secondary particle flux from the target. The ratio of the signals TBID/TBIU, called the multiplicity, is proper for each target material and geometry but it provides an easy means for supervising and optimising the targeting.To remove the target box or the monitors the two halves of the top shield are remotely wheeled aside such that the element can freely be removed vertically by the overhead crane. The latter, provided with122Fig.1. Schematic side view of an SPS target station for slow ex­tracted beams. Instead of the standard plate targets, in the detail a special target box is shown for a high resolution experiment.sufficient hook height wasessential for the adoptedprinciple. However, this infrastructure was foreseen anyway for the handling of the beamtransport elements and theshieldings.No failures have occurred until now with either the targets, whichhave supported up to 1018-1019 protons, or with any of the associatedcomponents, which could be attributed to beam induced damage (except for the loss of vacuum in the monitor tanks when the beams accidentally hit the flanges). Smear tests, which have regularly been made around the target stations to detect possible Be—contamination proved negative.Target stations for high intensity s l o w  extracted beamsFor the NAHIF target stations, designed for slow extracted beamswith up to 3x1013ppp, "thick" targets could be accepted. Therefore,several Be-blocks 60mm wide, 30mm high and 100mm long, were mounted into a frame one behind the other to achieve the desired target length of up to 500mm. Three targets of different lengths were clamped at their sidesinto a water cooled frame, which is shown in Fig.2. Due to the bulkytarget material considerable lateral heat diffusion occurs during the pulse such that the maximum temperature at the end of the pulse is only about 1% of that which would have been reached with the same beam,Fig. 2. Target unit with three different vertically stacked water- cooled Be-targets for high intensity slow extracted beams.123hitting a thin target with a diameter of 3mm5. Obviously, for thick targets the "multiplicity" monitoring is no longer meaningful, therefore the TBID has been suppressed for the NAHIF stations.To keep the muon background in the experimental area to minimum levels, even at the highest possible intensities, a heavy copper collimator limiting the acceptance of the secondary beam was installed immediately downstream of the target, in order to absorb at least those muon parents, falling outside of the acceptance. Thus, three conical channels of different opening angles were machined axially through the copper block at different vertical positions. The block itself is water cooled and has a length of 950mm, a height of 250mm and a width of 150mm. Again, through a vertical displacement of this unit, the desired collimator aperture can be selected. The principal elements of the NAHIF station are shown in Fig.3 during their alignment and Fig.4 shows the station ready for operation.Up to 1013ppp were used on these stations and no particular problems have occurred. However, no experience exists yet at the maximum design intensity of 3x1013ppp.Fig. 3. The princip­al elements for a high intensitytarget station for slow extraction: up­stream beam monitor, target unit anddownstream collimat­or on the alignment bench.Wide band neutrino target station for fast extracted beamsThe layout of the wide band neutrino station, described in more detail in Ref.6 is shown in Fig.5. The target proper is made of 11 thin Be-rods, each with a diameter of 3mm and a length of 100mm. The 11 rods are spread equidistantly over a total length of 2000mm, which creates effectively a target with only half the density of beryllium and which improves the escape probability for secondaries. Either end of each rod is supported by vertical Be-plates with a thickness of 2mm, which creates some small losses of secondaries. Cooling is provided by helium which is driven in a closed circuit by a ROOTS pump through a heat exchanger and thereafter laterally onto the target rods with a velocity of about 20m/s.No problems have occurred with fast extracted bursts at 1013ppp. However, once the intensity rose to 1.5x10^3ppp over prolonged periods124it was found that some of the Be-rods were bent or even broken. On the other hand, no problems have occurred with fast-slow extracted bursts at1.5x10^3ppp where the pulse duration is 1-2ms. This may well indicate thesignificant differencebetween ^us-bursts, which can initiate shocks andvibrations and ms-bursts where these effects should have vanished.Fig.4. Upstream view of a high intensity target station for slow extracted beams.BEAM,Fig.5a. Schematic side view of the wide-band neutrino target station with a thin Be-target with 2m length.High intensity target for fast extracted beamsIn order to allow fast extracted beams up to 2x1013ppp in the narrow band neutrino line a new target was built which consisted essentially of a thick graphite rod with a diameter of 30mm and a length of 500mm. This rod is thermally shrink-fitted into a thick walled aluminium pipe and sealed at either end with Ti-windows, which prevents the oxidation of graphite at elevated temperatures when in contact with air. The heat is conducted radially into the Al-pipe and is removed from125Fig.5b. besidethere through cooling fins by a high velocity (lOm/s) axial air stream. At most 2-3kW average power can be dissipated in the target. Above that the Al-container might be overheated such that the thermal contact at the Al-graphite interface might be lost. A side view of this target, described in some detail in Ref.7 is shown in Fig.6. The beam enters from the left through a thin Ti-window which is mounted in the air output nozzle. Two targets, one operational and one spare are mounted beside each other in a common frame which can be displaced horizontally by a pneumatic circuit, equipped with stainless steel bellows.This target survived without obvious deterioration a total number of 7 x 1018 protons at 450 GeV/c and peak intensities of 2xl013ppp with fast extraction. It is foreseen to inspect in the near future the gra­phite in more detail, which might reveal important information for the future use of graphite for targets as well as for beam dumps.Fig.6. Side view of the graphite target for high intensity neutrino experi­ments. The beam is incident from the left and passes through a Ti-window in the nozzle of the air-cooling system.A detail of the unit equipped with three targets mounted each other and supported at either end by Be-plates.126Antiproton production targetsThe antiproton production target was designed for the CERN-PS beam at 26GeV/c and intensities up to 10l3ppp with a repetition time of 2.4s. Initially, the target consisted of 10 tungsten or rhenium pellets, each with a length of 10mm and a diameter of 3mm. These pellets were embedded in the core of a graphite cylinder with an outside diameter of 30mm. Thus one could expect the graphite to confine the target material sufficiently well even when it was thermally cracked or otherwise damaged by radiation. Again, to insulate the graphite from the air, an aluminium container equipped with fins to improve the air-cooling was shrink fitted around the graphite. The design of the target, described in more detail in Ref. 8 is shown in Fig. 7. In the course of the commissioning of this target it was found that copper pellets gave a somewhat higher yield than tungsten or rhenium. Therefore, Cu-targets have been used up to now with intensities of 1013ppp and more recently with 2x1013ppp. The total number of protons used on one target unit was in general in the range of 2-4x1019 protons. No significant deteriora­tion of the target performance has been observed up to now.In order to increase the antiproton yield by an order of magni­tude, as anticipated by the ACOL project9 the use of pulsed targets in combination with strong focusing lithium lenses has been suggested.Those targets consist essentially of a metallic rod through which a high axial current is pulsed simultaneously with the beam. The thus created circumferential magnetic field inside (and outside) of the target helps to focus the antiprotons "at the source" which results finally in an increased phase space density. To produce the necessary magnetic gradients which must be at least 12x103T/m, peak currents of about 200kA are required. This creates additional thermal loads and magnetic forces in the target. Some prototype tests have been reported in Ref. 10 which clearly revealed the problems to overcome. A preliminary design forfurther prototype tests with liquid pulsed targets, kept in a metal con­tainer, as discussed in Ref.11, is shown in Fig.8. In this design it is assumed that by applying a sufficiently high pressure on to the liquid metal, voids and electrical arcing can be avoided. The target may beconsidered as a"self-healing" device. If, however, the above "conven­tional" designs donot prove to besufficiently reli­able, more exotic solutions such asliquid metal jet targets might have to be considered.Fig. 7. Axial cutthrough the anti­proton production target.127Fig. 8. Axial cut through a pulsed liquid metaltarget.KEY:1:Tie Bolts,2:Flange,3:0uter Container,4iCentral Insulating Disk, 5:Final Central Weld,6:Current Path,7:Ti-Window, 8: Liquid Metal Target, 9:Target Container, 10:CeramicInsulation, 11:Mica Insulation, 12:Air-Cooling Channels.CONCLUSIONAll SPS targets were initially made of beryllium and were designed for a maximum intensity of 1013ppp at 400 GeV/c. No problems were encountered so far at these intensities, neither with slow extracted nor with fast extracted beams, where in particular thin Be-rods are highly strained. With the increase of the SPS intensity, however, fast extract­ed beams with up to 2x1013ppp were used. Therefore, new targets were made of massive graphite cylinders which proved to be very reliable.The copper target for the antiproton production at the PS, used at 26GeV/c and about 10l3ppp performed perfectly reliably over long running periods. However, this target may operate close to its limits at 2x1013ppp, foreseen in the near future such that a more frequent exchange of the target might become necessary. No reliable design solu­tion for pulsed targets with higher performance has yet been found but developments are underway.REFERENCES1. M.Awschalom et al., Nucl.Inst.Meth., 131,235(1975)P.Sievers, CERN/SPS/ABT 77-1 (1977)2. J.Ranft,J.T.Routti,CERN-LabII/RA/71-4 (1971)3. W.Kalbreier et al., CERN-LabII/BT/74-1 (1974)4. P.Sievers, CERN-LabII/BT/74-2 (1974)5. P.Sievers, CERN/SPS-ABT-PS/Techn.Note/77-10 (1977)6. W.Kalbreier et al., 1977 Part.Acc.Conf.-Chicago (1977)7. R.Bellone et al., 1983 Part.Acc.Conf. Santa Fe (1983)8. R.Bellone et al., High Intensity Targeting Workshop, FERMILAB,D.Cline,University of Wise., Madison (1980)9. E.J.Wilson (Editor), CERN 83-10 (1983)10.T.W.Eaton et al., 1985 Part.Acc.Conf.-Vancouver (1985)11.A.Ijspeert et al., CERN/SPS-ABT/Tech.Note/86-3 (1986)128TARGETS FOR TRIUMFT. A. HodgesUniversity of Victoria, Victoria, B.C., V8N 2Y2ABSTRACTThe problems of providing secondary particle production targets which can operate in the intense proton beams produced by a KAON factory are discussed. Practical solutions are described for targets to be used in the slow extracted beams and a compromise solution for those in fast extracted beams. The possibility of using change-of-state targets to solve the latter problem is introduced.INTRODUCTIONProviding adequate production targets for the TRIUMF KAON factory presents a number of interesting technical problems. Extrapolation of techniques used at the present meson factories and other accelerator facilities will produce solutions to some of these problems whilst others may require a new approach.The requirement that the production target provide an intense, highly luminous source of secondary particles generates the concomitant problem of high power deposition in the target by the proton beam. For the slow extracted beam, several possible solutions exist: for the fast extracted beam, a compromise solution utilizing larger beam sizes and less dense targets may be necessary if a novel technique for dealing with the power density and thermal shock problems cannot be found.TARGETS FOR SLOW EXTRACTED BEAMSThe quality of particle beams available from the secondary channels is strongly dependant on the secondary particle source size and luminosity. The 30 GeV primary proton beam will be about 3 mm diameter at the target, and since the take-off angle proposed for secondary channels is zero or five degrees, the transverse dimensions of the source will be small. In order to maximize the luminosity and keep the depth- of-field effects in the channels small, the target length should be kept to a minimum. Further, the target length must be chosen to optimize particle yield by balancing particle production against loss to interactions before they escape the target.A typical production target will be of the order of one interaction length thick. It will be made of a material that is dense and posesses good thermo-mechanical properties.Stationary targetsAs an example of such a target, consider a tungsten cylinder, 6 mm diameter and 150 mm long placed in a 30 GeV, 100 pA proton beam such that the beam traverses the cylinder along its axis. Fig. 1 shows thedeposited energy density as a function of distance along the target axisfor several beam spot sizes.Fig. 2 shows the radially integrated energy density and totaldeposited energy in such a target. The latter calculations were made1129Fig. 1 Deposited energy density in radial interval 0 S R < 0.5 ox by 30 GeV protons in tungsten target 5 mm dia. by 150 mm long.Fig. 2 Radially integrated deposited energy density and total energy deposition by 30 GeV protons in tungsten target 6 mm dia. by 150 mm long.using the CASIM2 code. A one interaction length target (- 95 mm long) would have 93 kW deposited in it and would require suface cooling of 6 kW/cm2. Clearly the surface power dissipation would be too high for a practical target, as would be the internal temperature (-3300 K).Substituting copper for tungsten as the target material would reduce the surface dissipation at the expense of increased target length; further reduction could be made by increasing the target diameter and/or adding fins to the surface but this would raise internal temperatures.Temperature differences between the interior and cooled surface of the target produce thermal stresses in the material. At currents above - 25 pA, beam interruptions 50 ms or more in duration will produce sufficient strain cycling in a copper target to make it of doubtful utility.Rotating targetsThe energy deposited in the target may be spread over a larger volume of material by moving the target relative to the beam. A large reduction in both deposited energy density and the surface heat flux required to remove the deposited heat is easily achieved. A similar result might be achieved by "painting" the beam over a larger area of target, however this would be the same as increasing the beam spot size, unless techniques were available to track and correct for the beam spot movement.Fig. 3 shows schematically a radiation cooled rotating target.Using a dense refractory material such as tungsten, operating in a 100 pA beam, the surface temperature would be -2100 K. A cooled target enclosure would serve to prevent heating of the target vacuum containment vessel and drive bearings, and also limit the spread of material evaporated from the hot surfaces. There is some difficulty with this type of target of obtaining reliable operation of bearings in vacuum and with migration of active material evaporated from the target.130Fig. 3. Schematic representation of radiation cooled rotating target.Fig. 4 shows a proposed water cooled rotating target. Use of vacuum tight water seals is avoided by using the flow of cooling water to power the rotation of the target through a set of turbine blades. Rotation at a few tens of Hertz would keep temperature cycling to a modest value without incurring undue bearing wear. Such rotational frequencies have been found in protoype models to be easily achieved at very modest water flow rates.A 100 mm O.D. 88 mm I.D. by 60 mm long cylindrical target of tungsten in a 30 GeV, 100 yA beam, cooled in this fashion, would develop internal temperatures of -200° C and produce a surface heat flux of -250 W/cm2. The bearings of the targets are made of conventional metal loaded graphite3. Irradiation1* of graphite at 1021 nvt produces a small incease in its mechanical properties and has no effect on its frictional properties, so one could expect these bearings to operate well even in the hostile radiation environment encountered in the immediate proximity of the target. Targets of this design are currently being developed at TRIUMF for use in meson production targets with high beam current densities (5~10 mA/cm2).131TARGETS FOR FAST EXTRACTED BEAMSThe p r o t o n  beam i m p i n g i n g  o n  t h e  t a r g e t  i n  t h i s  mode  o f  o p e r a t i o n  i s  a p u l s e d  beam w i t h  6 x i o 13 p r o t o n s  i n  e a c h  p u l s e ,  a r e p e t i t i o n  r a t e  o f  10 H z ,  a nd  a  p u l s e  w i d t h  o f  3 p s .  The  e f f e c t s  o n  t h e  t a r g e t  o f  t h i s  mode  o f  o p e r a t i o n  a r e  p r o f o u n d .  The  i n s t a n t a n e o u s  p o w e r  d e p o s i t i o n  r a t e  i n  a t u n g s t e n  t a r g e t  w o u l d  b e  -  1 / 3  GW/cm3 , p r o d u c i n g  v e r y  l a r g e  t e m p e r a t u r e  t r a n s i e n t s  c a u s i n g  d e s t r u c t i o n  o f  t h e  t a r g e t  b y  m e l t i n g  and  t h e r m a l  s h o c k .The  t h e r m a l  s t r e s s e s  g e n e r a t e d  i n  a t a r g e t  by  a p u l s e d  beam a r e  a c o m b i n a t i o n  o f  q u a s i  s t e a d y  s t a t e  s t r e s s e s ,  p r o d u c e d  b y  t h e  t e m p e r a t u r e  d i f f e r e n c e  b e t w e e n  t h e  h e a t e d  c o r e  a nd  c o o l e d  e x t e r i o r  o f  t h e  t a r g e t ,  and  d y n a m i c  e l a s t i c  s t r e s s  w a v e s 5 , c a u s e d  b y  t h e  i n t e r a c t i o n  o f  t h e  t h e r m a l  e x p a n s i o n  o f  t h e  beam h e a t e d  r e g i o n  and  i t s  i n e r t i a l  m a s s .  F o r  t h e  p u l s e d  beam d e s c r i b e d  a b o v e ,  t h e  d y n a m i c  s t r e s s e s  a r e  4 0 - 6 0 $  o f  t h e  q u a s i  s t e a d y  s t a t e  s t r e s s e s .To  r e d u c e  t h e  s t r e s s e s  t o  a  l e v e l  w h e r e  a t a r g e t  c o u l d  o p e r a t e  w i t h  r e a s o n a b l e  l o n g e v i t y  w i l l  r e q u i r e  a  c o m b i n a t i o n  o f  c o m p r o m i s e s  r e g a r d i n g  beam s p o t  s i z e , t a r g e t  d e n s i t y ,  p l u s  r o t a t i o n  o f  t h e  t a r g e t  t o  t a k e  a d v a n t a g e  o f  t h e  10 Hz r e p e t i t i o n  r a t e .  F o r  n e u t r i n o  p r o d u c t i o n ,  t h e  u s e  o f  a l o n g e r ,  l o w e r  d e n s i t y  t a r g e t  s u c h  a s  g r a p h i t e  m i g h t  n o t  b e  a  s e v e r e  p e n a l t y ,  b u t  u l t i m a t e l y ,  d e v e l o p m e n t  o f  s h o r t  t a r g e t s  w i l l  b e  r e q u i r e d  t o  o b t a i n  t h e  b e s t  p o s s i b l e  beam q u a l i t y  i n  t h e  s e c o n d a r y  c h a n n e l s .CHANGE-OF-STATE TARGETSP r o v i d i n g  t a r g e t s  t h a t  c a n  o p e r a t e  i n  t h e  p u l s e d  beam r e q u i r e s  t h e  s o l u t i o n  o f  tw o  p r o b l e m s ;  f i r s t l y ,  t h e  e n o r m o u s  d e p o s i t e d  p o w e r  d e n s i t y  p r o d u c e s  e x c e s s i v e  t e m p e r a t u r e  c h a n g e s  i n  t h e  t a r g e t ,  a n d  s e c o n d l y ,  t h e  t e m p e r a t u r e  c h a n g e s  i n  t h e  t a r g e t  p r o d u c e ,  v i a  t h e r m a l  e x p a n s i o n ,  d e s t r u c t i v e  s t r e s s e s .By m a k i n g  u s e  o f  t h e  l a t e n t  h e a t  o f  f u s i o n  o f  a t a r g e t  m a t e r i a l  i t  s h o u l d  b e  p o s s i b l e  t o  d r a m a t i c a l l y  r e d u c e  t h e  t e m p e r a t u r e  r i s e  p r o d u c e d  b y  beam h e a t i n g .  F u r t h e r ,  b y  c h o o s i n g  a m a t e r i a l  w h i c h  h a s  a p p r o x i m a t e l y  z e r o  v o l u m e  c h a n g e  when l i q u i f y i n g  o r  s o l i d i f y i n g ,  t h e  s t r e s s e s  g e n e r a t e d  b y  t h e  c h a n g e  o f  s t a t e  w o u l d  b e  s m a l l .  M a t e r i a l s  w i t h  t h e  l a t t e r  c h a r a c t e r i s t i c s  a r e  r a r e ,  b u t  t h e r e  a r e  a  f e w  w i t h  n e g a t i v e  v o l u m e  c h a n g e  on  l i q u i f y i n g ;  t h e s e  m a t e r i a l s  c o u l d  o n l y  p r o d u c e  s m a l l  s t r e s s  t r a n s i e n t s  i n t e r n a l l y .T a b l e  I  g i v e s  t h e  p r o p e r t i e s  o f  t h e  p o s s i b l e  e l e m e n t a l  c a n d i d a t e s .T a b l e  I . E l e m e n t s  w i t h n e g a t i v e v o l u m e  c h a n g e  o n l i q u i f y i n gE l e m e n t M e l t i n g  p o i n t , CD e n s i t y 3g / c m 3L a t e n t  h e a t ,  J / gV o l u m e 13 c h a n g e ,fA n t i m o n y 6 3 0 . 7 6 . 4 9 1 6 3 . 4 - 0 . 9 4B i s m u t h 2 7 1 . 3 1 0 . 0 3 5 0 . 2 - 3 . 3 2G a l l i u m 2 9 . 8 6 . 0 9 7 9 . 8 - 3 . 1Germanium 9 3 7 . 4 5 . 2 7 4 6 6 . 1 - 5 . 1a )  D e n s i t y  a t  t h e  m e l t i n g  p o i n t .b )  V o lu m e c h a n g e  ( n e g a t i v e  m eans  c o n t r a c t i o n )  o n  l i q u i f y i n g .132The v e r y  l a r g e  l a t e n t  h e a t  o f  f u s i o n  o f  g e r m a n iu m  i s  w o r t h y  o f  m e n t i o n .  I f  a t a r g e t  o f  t h i s  m a t e r i a l  w e r e  p r a c t i c a l l y  f e a s i b l e ,  i t  c o u l d  a b s o r b  t h e  p o w e r  d e p o s i t e d  b y  a beam p u l s e  i n  a v o l u m e  w i t h  c r o s s  s e c t i o n  o f  0 . 2 5  cm 2 .A n o t h e r  p o s s i b i l i t y  w o u l d  b e  t o  u s e  an  a l l o y  c o n t a i n i n g  a m a t e r i a l  f r o m  T a b l e  I .  The  w e l l  known 56  A t ?  B i - P b  a l l o y  h a s  t h e  r e q u i r e d  n e a r  z e r o  v o l u m e  c h a n g e  o n  l i q u i f y i n g ,  and  m e l t s  a t  1 2 5 °  C b u t  h a s  a l o w  l a t e n t  h e a t .  A number  o f  a l l o y s  e x i s t ,  i n c l u d i n g  e u t e c t i c  a l l o y s  o f  g e r m a n i u m  w i t h  a l u m i n u m ,  c o p p e r ,  s i l v e r  and  g o l d ,  w i t h  m e l t i n g  p o i n t s  f r o m  - 3 6 0 °  C t o  t h e  m e l t i n g  p o i n t  o f  g e r m a n iu m  i t s e l f .S e v e r a l  t e c h n i c a l  p r o b l e m s  w o u l d  h a v e  t o  b e  s o l v e d  b e f o r e  a t a r g e t  o f  t h i s  t y p e  c o u l d  b e c o m e  a  p r a c t i c a l  p r o p o s i t i o n ,  n o t  l e a s t  among t h e s e  b e i n g  t h e  c o n t r o l l e d  r e m o v a l  o f  h e a t  w h i l s t  k e e p i n g  t h e  t a r g e t  m a t e r i a lc l o s e  t o  i t s  m e l t i n g  p o i n t .TARGET SERVICE AND REPAIRThe t e c h n i q u e s  e m p l o y e d  f o r  s u p p l y i n g  c o o l i n g ,  e l e c t r i c a l  s e r v i c e s ,  e t c .  t o  t h e  t a r g e t  a s s e m b l i e s  a nd  t h e  m e t h o d  o f  r e m o v i n g  o r  r e p l a c i n g  t a r g e t s  i n  t a r g e t  a r e a s  w i l l  b e  e x t e n s i o n s  o f  t h o s e  c u r r e n t l y  u s e d  a tTRIUMF. S h i e l d i n g  t h i c k n e s s  w i l l  b e  s o m e w h a t  g r e a t e r  i n  t h e  t r a n s v e r s ed i r e c t i o n  n e a r  a  t a r g e t  b u t  w i l l  e x t e n d  f o r  a much g r e a t e r  d i s t a n c e  d o w n s t r e a m .S h i e l d i n g  i n t e g r a l  w i t h  an d  a b o v e  t h e  t a r g e t  a s s e m b l y  w i l l  b e  b r o u g h t  t o  a  p o i n t  w h e r e  s e r v i c e  c o n n e c t i o n s  may b e  s a f e l y  made b y  h a n d .  A l l  w o r k  d o n e  o n  t a r g e t s  b e l o w  t h i s  l e v e l  w i l l  b e  c a r r i e d  o u t  i n  a h o t  c e l l ,  t r a n f e r  o f  t h e  t a r g e t  t o  t h e  h o t  c e l l  b e i n g  made a f t e r  v e r t i c a l  e x t r a c t i o n  o f  t h e  t a r g e t  i n t o  a s h i e l d e d  f l a s k  b y  c r a n e .CONCLUSIONSP r o d u c t i o n  t a r g e t s ,  s u i t a b l e  f o r  u s e  i n  t h e  s l o w  e x t r a c t e d  beam and  s u f f i c i e n t l y  d e n s e  t o  p r o v i d e  t h e  i n t e n s e  a nd  l u m i n o u s  s o u c e  o f  s e c o n d a r y  p a r t i c l e s  r e q u i r e d ,  c a n  b e  d e v e l o p e d  on  t h e  b a s i s  o f  known t e c h n o l o g y .Low d e n s i t y  t a r g e t s  c a p a b l e  o f  o p e r a t i n g  i n  t h e  f a s t  e x t r a c t e d  bea m,  p r o v i d e d  t h e  beam s p o t  s i z e  i s  e n l a r g e d ,  c a n  b e  d e v e l o p e d  f o r  n e u t r i n o  p r o d u c t i o n .P r o v i s i o n  o f  medium o r  h i g h  d e n s i t y  t a r g e t s  f o r  t h e  f a s t  e x t r a c t e d  beam w i l l  r e q u i r e  new a p p r o a c h e s ,  s u c h  a s  o u t l i n e d  f o r  t h e  c h a n g e - o f -  s t a t e  t a r g e t  a b o v e .REFERENCES1 .  C.  Y a m a g u c h i ,  p r i v a t e  c o m m u n i c a t i o n ,  M a r c h  1 9 8 5 .2 .  A.  Van G i n n e k e n ,  CASIM, P r o g r a m  To  S i m u l a t e  T r a n s p o r t  o f  H a d r o n i cC a s c a d e s  i n  B u l k  M a t t e r ,  F e r m i  l a b  R e p o r t  FN 2 7 2  ( 1 9 7 5 ) .3 .  M e t a l l i z e d  C a r b o n  C o r p o r a t i o n ,  O s s i n i n g ,  New Y o r k  1 0 5 6 24 .  R . A . W u l l a e r t  e t  a l .  i n  " E f f e c t s  o f  R a d i a t i o n  o n  M a t e r i a l s  an dC o m p o n e n t s " ,  e d s .  J . F .  K i r c h e r  and  R . E .  Bowman,( R e i n h o l d  P u b l i s h i n g  C o r p o r a t i o n ,  New Y o r k ,  1 9 6 4 ) .5 .  P .  S i e v e r s ,  " E l a s t i c  S t r e s s  Waves  i n  M a t t e r  Due t o  R a p i d  H e a t i n g  byan Intense High-Energy Particle Beam".CERN R e p o r t ,  L a b . I I / B T / 7 4 - 2  ( 1 9 7 4 ) .133PARTICLE PRODUCTION AND TARGETTING EXPERIENCE AT THE BROOKHAVEN AGS*D.M. Lazarus AGS Department Brookhaven National Laboratory Upton, New York 11973ABSTRACTExperience in production of secondary pions (neutrinos), kaons and antiprotons by 28.5 GeV/c protons incident on various target materials is given. The problems associated with various target materials with respect to target heating, physical degradation and in some cases, dis­integration, are discussed. The effect of target length and production angle on secondary beam flux and optical quality will be illustrated by some incomplete but nonetheless informative data.INTRODUCTIONAlthough the primitive (manual) target mechanisms and target design used at the AGS may not be applicable to kaon factory facilities, the experienced gained may still be useful to those designing targets for such machines. Data on particle production at various angles and with several target materials suggests that 0° production may be optimized with low A target materials of less than one interaction length and that separated beams are enhanced by finite angle production despite the larger cross section at 0°.TARGETSThe typical target mounting arrangement for the slowly extracted beam program at the AGS uses target heads that can be easily attached or removed from a i m  long handle by means of locking device operated from the opposite end of the handle. The target with a prealigned A102 flag mounted on the head, is located by means of a track and a stop on a sled installed on a concrete shield pier. The target station instrumen­tation package also is located on the sled following prealignment on a jig*A well in the top of the concrete pier allows temporary storage of the target by simply pulling the target back off the track and rotating the handle to park the target in the well. A number of nearby vaults provide long term storage by inserting the target head, releasing it from the handle and closing the vault door. The method has proven fast and reliable. Targets with activation in excess of 100 rem/hour at a few cm are changed in tens of seconds with personnel exposures of 5-10 mr in areas whose ambient levels are up to several rem/hr. Were the levels to increase by a factor of 50-100, as they would at the proposed kaon factories, this method would clearly not be viable. TargetWork performed under the auspices of the U.S. Department of Energy.134temperature is monitored by means of a thermocouple mounted in the aluminum holder adjacent to the downstream end of the target.Hevimet and iridium targets have decomposed and contaminated target stations even in the slowly extracted beam areas. Hevimet is, of course, a machinable form of tungsten, which is in sintered form with copper. Growth of whiskers or snow and swelling of the target was observed at 2 x 1012 protons/second. For the past decade, platinum targets held on an aluminum holder have been successfully employed in the SEB with a minimum of problems. Low A, low density targets such as berylium and copper, have also given relatively trouble-free performance in the SEB.During fast extraction to the neutrino area, the rate of energy deposition is increased by more than 107 in going from a one second spill to 12 bunches of 30 ns in 2.75 ps. In the single bunch extraction mode where one of the twelve bunches is extracted to an experiment, the first failures of platinum targets were recently observed at intensities five times below normal SEB operation with a one second spill.Fast extraction has destroyed sapphire (A120 3) and tungsten-rhenium targets: the latter a copy of a similar graphite encapsulated targetdestroyed at CERN. In the past year, a titanium target was successfully used for a nine week wide band neutrino run. Ti has Z = 22, A = 48, and p =4.54 gm/cm3. Its melting point is 1660°C or about 100° less than platinum. Although titanium has great strength, it has rather poor thermal conductivity. Nevertheless, the low density reduces the density of energy disposition enabling it to survive in situations where target length is unimportant. Copper has been successfully used in narrow band neutrino operation where a relatively short target is required to allowmomentum selection of the pions prior to their decay. As will be seenin the next section, copper is not a bad choice for pions produced at 0°.The conclusions are:1. Pulsed operation with short intense bunches is far more de­structive of targets than a slow spill.2. Platinum can be used at intensities below 1013/sec for a beamspot greater than 4mm2.3. The target materials and techniques used at the AGS are not applicable to kaon factories where intensity increases in ex­cess of one order of magnitude are anticipated.4. Beware of materials such as hevimet which are likely to con­taminate a target station. It is not obvious that pure tung­sten can be machined to make targets at reasonable cost or whether or not it can be used without problems of contamination.SOME COMMENTS ON PARTICLE PRODUCTIONIn order to resolve a targetting conflict between experiments and to make intelligent choices for neutrino (pion) production targets fol­lowing the failures described above, we undertook a simultaneous mea­surement of 1.4 GeV/c pion and antiproton fluxes at 0° production angle and .8 GeV/c negative kaon fluxes at 10.5° in the Low Energy Separated Beam I as a function of target materials and length. The measurements were normalized to the secondary emission chamber (SEC) immediately135X!00AT 0 MIC200W E IG H Tupstream of the target, but no attempt was made to correct the measurements for beam acceptance to obtain an absolute normaliza­tion. The counter telescope which monitors targettlng is, of course, sensitive to target ma­terial and length and is there­fore not useful for normalization.The targets used and the data obtained are given in Table I. The interaction length for Pt was obtained by interpolation using a plot (Fig. 1) of inter­action lengths vs atomic number for a variety of materials taken from the "Blue Book.”2Fig. 1. The dependence of interaction length weight.SLq on atomicTABLE I.JI 1.88" Pt 3.5" Pt .74" Cu 2.96" Cu 5.25" Cu 5.25" Al*/*o .522 .972 .125 .499 .886 .3380° Datait-/SEC 67.3 60.7 55.4 103.5 156.4 124.8p/SEC .120 .120 .0532 .159 .156 .09410.5° DataK-/SEC 14.7 21.0/18.7 5.65 14.5 15.9 8.0tT/K- 5.2 4.9/5.9 5.3 6.2 11.1 12.2The 0° data obtained indicate both the pion and antiproton pro­duction saturates at less than half an interaction length whereas the 10.5° kaon data shows no clear saturation at one interaction length. This is easily understood a the result of absorption of secondary par­ticles produced at small angles with a relatively large average amount136of material between the point of production and the downstream surface of the target. Secondaries produced at 10.5° pass through comparatively little material before exiting from the side of the target in the direc­tion of the LESB I.Figures 2 and 3 indicate that at 0° the Cu targets yield a greater number of both pions and antiprotons than the equivalent length of Pt. The 10.5° negative kaon data is shown in Fig. 4. Figure 5 indicates a marked degradation of the ir“/IT ratio for target lengths greater than 4 inches.CONCLUSIONSA somewhat higher yield of K" is obtained from Pt than from Cu, but the primary benefit for kaon users arises from the shorter interaction length of Pt which allows good kaon production from a target of less than 3 inches reducing the depth of focus to provide a better image at the mass slit and therefore a better tt- / K -  ratio.Small angle negative pion production is enhanced by the use of a Cu target. The acceptance of the 0° beam is ± 20 mr vertically and ± 40 mr horizontally. The narrow band neutrino horn acceptance includes the angular range of 10 mr to 140 mr and therefore a target length between .5 < H I  < 2 is called for. Cu appears to be the preferred target material to simultaneously optimize flux and depth of focus for the narrow band v beam.REFERENCES1. J.W. Glenn, D.M. Lazarus, P.H. Pile, J. Sculli and J.C. Walker,EP&S Division Technical Note No. 116, Brookhaven National Laboratory (1985).3. Particle Properties Data Booklet - Particle Data Group, Lawrence Berkeley Laboratory (1984).1371.4 G e V / c  7t ~  0 °x P t• Cu 4  A l30CUj>Fig. 2. The yield or number of it- per SEC count at 0° and 1.4 GeV/c vs. target length/inter­action length.2001001.4 G e V / c  p 0 °_i____ I____ !____ !____2 .4 6 .8 1.0T A R G E T  L E N G T H  ( / / / 0 )Fig. 3. The yield or number of p per SEC count at 0° and 1.4 GeV/c vs. target length/inter- Q action length. qJ>2 -x  P t  •  C u  A  A l—I—.---- 1---- 1_____I___ I_____|_2 A 6 .8 1.0 1.2TARG ET  LEN G TH  <f//0)YIELD13887  GeV/c 1 0 .5 'X20x Pt •  C u  A  Al10I 2Fig. 5. The ir-/K" ratio vs. target length in the LESB I at 10.5° and .87 GeV/c.Fig. 4. The yield or number of K~ per SEC count at 10.5° and .87 GeV/c per SEC count vs. target length/interaction length.<crtA>t=10-6 7  G e V / c 10 5 ‘x pt•  Cu A  A lAI 2 3 4 5 6T A R G E T  L E N G T H  ( i n c h e s )139ANTIPROTONS AT THE KEK-PS K4 BEAM LINEK. H. Tanaka, T. Tanimori Physics Department, KEK National Laboratory for High Energy Physics Oho-machi, Tsukuba-gun, Ibaraki-ken, 305Y. Fujii Laboratory of Nuclear Science Tohoku University Mikamine, Sendai, 982Y. Sugimoto Department of Physics, Faculty of Science Kyoto University Kita-shirakawa, Sakyo-ku, Kyoto, 606M. Sudo Institute for Nuclear Study University of Tokyo Midori-cho, Tanashi-shi, Tokyo, 188S. Kohno, K. Kasai, Y. Morita Department of Physics, Faculty of Science Hiroshima University Naka-ku, Hiroshima, 730ABSTRACTPerformance of the antiproton beam-line K4 with a double-stage electrostatic separator at KEK-PS is described. Due to the fine tuning of both the primary 12-GeV proton beam-line and the K4 beam-line, the 7T /p ratio less than unity has been attained with an applicable p yield.K4 BEAM-LINEThe K4 beam-line is one of secondary beam-lines at the KEK proton synchrotron (KEK-PS), which was designed for low energy antiprotons with good separation from background pions. For this purpose, the K4 consists of two electrostatic separators and two mass slits as shown in Fig. 1. The first operation of K4 was in Oct., 1983 and followed a long scheduled shut-down of KEK-PS due to the TRISTAN construction. The design parameters and the results of the first beam commissioning of K4 are reported in ref. 1. When the PS recovered in June, 1985, the K4 beam-line ^as also become active for the physics experiment . Then the tuning of K4 was made finely and a better performance of antiproton yield and of pion rejection was achieved than the previous K4 operation.*Present Address: Physics Department, KEK.140The maximum energy of our KEK-PS is 12 GeV, which is not so sufficient for intense production of antiprotons.This constraint leads us to design secondary beam-lines with large acceptance and with high performance of background rejection. The good background rejection was achieved by high field-gradient electrostatic separators with the direct built-in high voltage power supply . For instance, the field-gradient of K4 DC separators is 40 kv/cm in normal operation for both separators. The large acceptance of the secondary beam-line requires very good quality of primary beam since pions which may be produced at the upstream of the primary beam-line can easily be accepted by the secondary line. During the long shut-down, the primary beam-line of 12-GeV protons for K4 was reconsidered and reconstructed. Therefore the background pions have been drastically reduced by the innovation of the primary line.PERFORMANCE OF K4The typical intensity of.protons incident on the K4 production target is 1.8 x 10 ppp ( protons per pulse, one pulse width is 500 ms per 2.5 s ). The size of the primary beam on the target is approximately 6 mm in diameter, which is monitored all the time by a segmented wire ion chamber (SWIC). The production target normally used is a Pt rod141with a size of 6 mm (vertical) x 10 mm (horizontal) x 60 mm (long). The antiproton yield and the n /p ratio with this target are plotted in Fig. 2 against the secondary beam momenta by open and closed circles, respectively. Gaps of the mass slits are ±6 mm for the first one and ±20 mm for the second one.P Momenta (MeV/c)Fig. 2 The p yield (open circles) and the 7T~/p ratio (closed circles) at K4.Especially for low energy antiprotons, the absorption in the production target_may be a serious problem. From this point of view, the p yields with various target length were studied. Targets used were cylinders of Pt with 6 mm0 x 20mm, 40 mm and 60mm long. The results are summarized in Fig. 3. Antiproton yield do not depend on the target length at momenta lower than 500 MeV/c. This result indicates that the low momentum antiprotons, which are accepted in K4, are produced at the downstream end of the production target in the part and are enough for experiments by a short production target.Images of antiprotons and pions at the second mass slit were measured to improve the pion-rejection efficiency. Firstly beam-line elements including both DC separators were142tuned to transport 570MeV/c antiprotons. Both mass slits were centered at a medium plane of the beam-line and the gap of the first mass slit was set to 12mm (±6 mm from the medium plane). Then we moved the second mass slit with a step of 4 mm and the p and rr yields were measured as a function of the second-mass-slit position with respect to the beam level. The measurement without production target was also made. Results are plotted in Fig. 4, which indicate that we can obtain antiprotons more than pions if the gap of the second mass slit would be 30 % of the present gap. This fact encourages us that we can handle secondary rare particles even from such high intensity primary beam as the proposed kaon factory by using the double stage DC separation technique.We would like to express our thanks to staff and crew of KEK accelerator department for their excellent operation of PS.Fig. 3 (Left) Dependence of p production on target length. Fig. 4 (Right) Images of p and 7r at the second mass slit.References1. M. Takasaki et al., Nucl. Instr. Meth., A242, 201 (1986).2. T. Fujii et al., KEK-PS Research Proposal E-131.3. A. Yamamoto et al., Nucl. Instr. Meth., 148, 203 (1978).4. S. Kurokawa et al., Nucl. Instr. Meth., 212, 91 (1983).PRODUCTION OF SECONDARY PARTICLES AT THE TRIUMF KAON FACTORYA. Yamamoto and M. TakasakiNational Laboratory for High Energy Physics (KEK),Oho-machi, Tsukuba-gun, Ibaraki, 305, JapanABSTRACTWe describe production cross sections of secondary particles at the KAON factory estimated by using an energy scaling formula with a scaling variable x( =£*/£($) . The nuclear mass number dependence on the production cross section for each secondary particle in p - N reaction is discussed in the calculation.1. SCALING OF THE PARTICLE PRODUCTION AT pt = 0In recent years, it has been known that invariant cross sections ofhigh energy hadron productions in p - p or p - N reactions could beexpressed as functions of a scaling variable x=E¥/E^^(2E¥/a/S) where E* is the energy of the produced particle, Ej is the maximum energykinematically availabe and s is the square of the total energy of the interaction.Especially, low energy hadron production at pt = 0  has been studied at KEK and it was found that the sacling with the variable x was applicablein the region of the low energy hadron productions at pt =0.^Experimental studies were carried out at the K2 beam line of 12 GeV protonsynchrotron. The measured data cover the momentum range of 0.74 - 2.1 GeV/c, in the laboratory system, at pt = 0. The differential production cross sections of pions, kaons, protons and antiprotons in p - Be reaction are given in references 4 and 6. The production cross section converted to the invariant cross sections, I = E ■ d3o/dp3 , were compared with other two sets of measurements at pt = 0 in p - Be reaction by Dekkers et al.7) andby Atherton et al.8) Figure 1 shows three sets of data from Dekkers etal., Atherton et al. and our data as a function of x. It is apparent that x dependences of the invariant cross sections are very similar and that the absolute normalizations are consistent with each other. Reasonable fits were obtained with a simple form of1 = E dp3 *  (1_X)i> (1)where a is the normalization parameter and 6 indicates the falling-off behaviour. This form is also used by Taylor et al. in the analysis of hadron productions at pt > 0.2 GeV/c.3) The solid curves in Fig. 1 represent the best fits. The results of the fits at p - Be interactions are given in Table 1.The above expression can be converted into the differential cross section per interacting proton in the laboaratory system, which is given by144Fig. 1. Invariant cross sections of hadron productions at pt = 0 in p - Be reactions145where oa is the absorption cross section of the incident proton in the Be target. The terms of p and E are the secondary particle momentum and energy, respectively, in the laboratory system. The terms of mp , pp and Ep are the rest mass, momentum and total energy of the incident proton, respectively. This expression is reasonably consistent with the measured results and also with other calculation by the use of the Sanford-Wang’s formula. ^2. SECONDARY PARTICLE PRODUCTION AT THE TRIUMF KAON FACTORYParticle production cross sections at pt = 0 in p - N reaction with the incident proton momentum of 30.9 GeV/c at the KAON Factory were calculated with the above equation. Figures 2 (a) and (b) show thecalculated production cross section of secondary particles in p -Be and p - Pt reactions, respectively.It should be notified that the individual nuclear mass number (A) dependence exists on each secondary particle production cross section at Pt = 0 in p - N reaction, and the power a of the A dependence varies as a function of the secondary particle momentum. This feature is smilar to the one at high pt region.’*^ Figure 2 shows the power a of A dependence on the secondary particle production cross section as a function of the laboratory momentum of the secondary particle, at pinc = 18.8 and 23.1 GeV/c. They were calculated from the cross sections, in p - N reations, measured by Dekkers et al.7) It is clearly shown that the power a rises upto about unity with decreasing the secondary paticle momentum. The production cross sections in p - Pt reation were calculated by the use of the modified expression including the A dependence as follows,- & L  = ^  £ L ri_ TxbcEldp ABe oa cKldp ' {ABe oa E }• i a ^ i (E- i ^ k y pr>kwhere A is the atomic number of the nucleus. The parameter a was obtained from averaged experimental data indicated by dotted curves in Fig. 2. We assumed the same A dependences in this calculation as those at pinc = 18.8 and 23.1 GeV/c. The negative kaon production cross section per nucleus shown in Fig. 3-(b) was more consistent with the measured results at CERN and BNL(given in the KAON factory proposal), n) which were measured results in p - W and p - Pt reactions, respectively. In the KAON factory proposal, it is remarked that our prediction is too low for momenta below 2 GeV/c in comparison with the measured values at CEF^ and BNL.U) We remind, however, that the A dependence does not only exist in the absorption cross section but also exists in each secondary particle production cross section as mentioned above and, therefore, the measured data should be compared to our calculation shown in Fig. 2 (b), in which our prediction is more reasonably consistent with the measured data at CERN and BNL.As a conclusion, we consider that our scaling formula is still valuable to predict the secondary particle production cross section even in the low momentum region.146REFERENCES1) K. Kinoshita and H. Noda, Prog. Teor. Phys. 49 (1973)1639.2) E. Yen, Phys Rev. DIO (1974) 836.3) E. F. Taylor et al., Phys. Rev. D14 (1976) 1217.4) A. Yamamoto, KEK Report, KEK 81-13.5) A. Yamamoto et al., Nucl. Instr. and Methods, 203 (1982) 35.6) D. Axen, Proceedings of the Second Kaon Factory Physics Workshop, TRIUMF, Vancouver (1981) 177., TRI 81-.7) D. Dekkers et al., Phys, Rev. 137 (1965) B962.8) H. W. Atherton et al., CERN Report, CERN 80-17.9) C. L. Wang, Phys. Rev. Lett., 25 (1970) 1068.10) J. W. Cronin et al., Phys. Rev. Dll (1975) 3105.11) KAON Factory Proposal, TRIUMF (1985) 5-10.Table 1. Parameters a and 5 in the fitting form of E.d?o/dp2=a.. (l-x)b .Particle a bTl* 269 4.13Tt 256 5.67r 20.8 3.57K~ 18.1 6.87p 24.3 -1.71p 10.0 8.70* in p - Be reaction.A a ( p f N  —  h -*• X )____________DEKKERS ct ol.1.0a0.80.6------ 1-------1-------1------ 1-------1-------1------(a l  tr*•  P p =  l8 .8G eV /c ' \  *  P p - 23.1 GeV/c. ' 9xo(b l 7r *\*No- - o -----Q------( c )  K + ( d )  K “1.0 ,\ sOa O0.80.6 - ■te l p ( f )  P1.0 .  s XaNXNNNV0.8 - o.40.6 1 ! 1 1 1 1 ....................................2 4 6 8 10 12 2 4 6 8 10 12Ph [G e V /c l Ph [ G eV/clFig. 2. Plots of power a as a function of PhTable 2. Calculated hadron production cross sections* at pp =30.9 GeV/c in p - Pt reaction______Momentum %+/%- K+/K-____________ P/P0.5 GeV/c 1.2 / 1.1 3.4E-2 / 1.3E-2 0.39 / 0.7E-51.0 2.2 / 2.0 1.IE—1 / 5 .8E-2 0.57 / 6.4E-42 .0 3.3 / 2.8 2.5E-1 / 1.4E-1 0.83 / 9.1E-34 .0 3.5 / 3.1 3.7E-1 / 1.9E-1 1.27 / 2.5E-26 .0 3.3 / 2.6 3.8E-1 / 1.6E-1 1 .65 / 2.3E-28,.0 2.9 / 2.0 3.4E-1 / 1. IE—1 2.05 / 1.6E-2* d2N/dfidp = ( particles / sr. GeV/c / interacting proton )Production angle = 0-degreedftdp  ^ParticIes/sr/GeV/c /Interacting proton]147Fig. 3. Calculated production cross sections of secondarypartides at zero degree i n p + N  - h + X  reactions at pp =30.9 GeV/c.148INFLUENCE OF SOURCE SIZE ON ACCEPTANCE AND SEPARATION OF KAON BEAMSJ. DoornbosTRIUMF, 4004 Wesbrook Mall, Vancouver, B.C., Canada V6T 2A31. INTRODUCTIONIn this note a few remarks are made about the importance of the size of the vertical proton beam spot on a kaon production target for the ac­ceptance and separation of kaon beams. In Sec. 2 some simple relations between first-order matrix elements and beam sizes using the traditional TRANSPORT matrix formalism are described. In Sec. 3 these relations are applied to the relative and absolute separation between pions and kaons. Section 4 gives examples for a 6 GeV/c and a 2 GeV/c beam. The conclu­sion in Sec. 5 is that it is important to make the vertical beam spot smaller than 3 mm but that smaller than 0 .5-1.0 mm is useless.2. SOME TRANSPORT RELATIONSIn first order the vertical half-size Y and half-divergence f at a point in the beam line can be expressed in terms of size and divergence at another point in the beam line by the relationC l ■ C L  C lwithdet(R) = 1 . (2)The 2x2 matrix R contains all the information about the optics. In par­ticular we can express the co-ordinates in the separator as a function of the source parametersYsep = Rsep x yO + Rsep x  ^ (3a)()> seP = R ^ P  x yO + Rsep x 4,0 , (3b)We have point-to-parallel focusing between source and separator. This meansR ^ P  = 0 (4)(4) + (1) + R^fp = -1/Rf®p (5a)(3b) + (4) + (5a) -► <(>sep = -Y°/Rf®p . (5b)We can also express the co-ordinates at the mass slit as a function of the co-ordinates at the sourceYms = R™f x Y° + R®® x <t>° , (6 a)4,ms = gms x yO + Rms x 4,0 # (6b)149There is a vertical focus at the mass slit. This meansC  = 0. (7)Alternatively, we can express the co-ordinates at the mass slit as a function of the co-ordinates in the separatorYms = R“3s_sep x Ysep + R™f-sep x ^ sep  ^ (ga)^ms gms-sep x Ysep + Rms-sep x ^ sep . (gb)We have parallel-to-point focusing between separator and mass slit. This meansgms-sep = 0 # (9)Since (R)2-q = (R)2 —i x (R)i~o we haveRms = Rms-sep x Rsep + Rms-sep x Rsep # (1Q)(5b) + (9) + (10) ■> R“s_seP = -Rjs x r JJP . (u)3. RELATIVE AND ABSOLUTE SEPARATIONWe want to separate pions out of the kaon beam. A crossed field separator gives the pions a kick4+(mr) , £(fcV/cm) x L ( c )  A  _ 1 \  1P(MeV/cm) x 1000 Sk/ P3Example: E = 75 kV/cm L = 5 mA<j> = 4.00 for 1 GeV/c= 0.50 for 2 GeV/c= 0.16 for 3 GeV/c= 0.02 for 6 GeV/cThe relative separation is given bySr _  A<j> (from separator field)______2x<|> (divergence of beam in separator)(Note: The factor 2 is necessary because <j> is the half-divergence)(12) + (5b) + SR = x R ? , e P  .( 12 )R = — ^  x R?uF • (13)R 2x Y° 3kThe relative separation depends only on the source size (Y°) and on the optics before the separator (Rf®p). The separation Atp in angle is con­verted into a separation in position at the mass slit in accordance with (8a) and (9).We obtain for the absolute separation150Sabs 4<|> x R;ms-sep■34 (14a)A<() x R™® x Rf^P> (14b)The absolute separation at the mass slit depends only on the optics after the separator. The product of R™| and R?^p is always equal to R™® seP# We add one more equation. Consider (3a). The source size is always very small and contributes only marginally to the beam size in the sepa­rator. To a very good approximation we can writeConsider (13). For a given separation SR the smaller the source size, the smaller Rf®p, the bigger the acceptance (j>°. In other words, for a given relative separation SR the source size determines the accep­tance of the separator.We see from (13), (14) and (15) the pivotal role of the matrix ele­ment R|®p» It is in general completely determined by the layout of the beam line which forces unique settings of the quadrupoles. The same is true for the absolute separation. If we want to increase that we must increase the drift before the mass slit. Therefore, we have two very important questions:1) What is a reasonable assumption for the source size?2) What is the minimum practical aperture of the mass slit?MESB at BNL has a 1.0 mm source size and a 1.5 mm separation. Is it really possible to obtain good results with such a small separation? At a KAON factory it seems possible to have a 1 mm vertical proton beam spot on the kaon production target, even for a 100 pA beam.Figure 1 gives the layout of a possible 120 m long 6 GeV/c beam line. It consists of three sections. A doubly achromatic section with two opposite bends prepares the beam for a 30 m long separator where the beam is parallel. A final doubly achromatic section transports the beam to an experiment. The initial section has a special advantage because it is possible to vary Rf®p (and thus the vertical acceptance) in the separator while at the same time keeping the beam doubly achromatic and parellel in both planes in the separator. The tuning and layout of theYseP = r3s^ P x 4,° . (15)For a given size of gap in the separator the smaller Rg^P, the bigger theacceptance <J>0 .4a. EXAMPLE FOR 6 GeV/c BEAM0 6 0 7 0 9 0 9  0I0QII]no noc3—013 0 ' ^0102 *—BlSEPARATOR MASS B3 S L I TFig. 1. Proposed arrangement of the 6 GeV/c kaon beam line K4.151beam line after the separator has been chosen so that the vertical it - K  separation for a 6 GeV/c beam is 3 mm at the mass slit, independent of the tuning before the separator. We have S r  = 2, E = 75 kV/cm; verti­cal gap is 10 cm.Table I gives the results for different source sizes. The vertical angular acceptance of the separator and the full beam spot sizes are given. A momentum acceptance of ±2.5% Ap/p is assumed.Table I. Influence of vertical proton beam spot.Source AcceptanceBeam spot at mass slitfirst order + second-order aberrations geometric chromatic3.0 mm ±1 mr 1.5 mm 1.6 mm 12 mm0.6 5 1.5 1.6 850.15 20 1.5 10.0 1050We see that in principle indeed the acceptance in the separator is bigger for a smaller source size. However, for the rest of the beam linethe acceptance is smaller than ±5 mr. Maybe it can be increased to±7 mr. Unfortunately, the chromatic aberrations are ferocious. Require­ments of corrections could force a major change in the design. Since several tunes are possible, sextupoles and octupoles should be tunable independently of quadrupoles and bending magnets.The present design allows an acceptance of ±5 mr for a 3 GeV/c beam, even for a 3 mm proton spot. Reducing this spot to 1 mm does not there­fore help the acceptance at the lower momentum end of the beam line.4b. EXAMPLE FOR A 2 GeV/c BEAMCalculations were also performed for a 2 GeV/c beam line. The front end consists of a bending magnet followed by a doublet. A 6 m long sep­arator with a gradient of 75 kV/cm and a gap of 10 cm gives a pionseparation A<|> = 0.62 mr. If we require Sr = 2 we obtain:proton spot psep3*+ acceptance3.0 mm 1.00 ±5 mr1.5 0.50 100.75 0.25 20However, in practice it was impossible to make Rf^P smaller than 0.40 without blowing up the beam in the horizontal plane. Reversing the polarity of the doublet quadrupoles did not help. A triplet did not give better results. In other words the vertical acceptance is smaller than ±12.5 mr.1525. CONCLUSIONS Two conclusions follow from these explorations:1) It is very important to make the proton beam spot on the pro­duction target less than 3 mm high to improve acceptance and separation.2) It is useless to make the proton spot height smaller than0.5-1.0 mm.153LOW MOMENTUM KAON BEAM LINES:CONTAMINATION AND DESIGN CRITERIAD.E LobbTRIUMF, University of Victoria, Victoria, B.C. V8W 2Y2 CanadaABSTRACTThe decay near the production target of K° into charged pions is a significant source of pion contamination in low momentum kaon beam lines. Results from a Monte Carlo calculation are presented which indicate that this process may account for about half the experimentally observed contamination. Other processes that may produce significant numbers of contaminant pions are glancing angle scatter and transmission of particles around or through the shielding.A design for a .8 GeV/c kaon channel is presented. This design features an initial stage to remove contaminant particles followed by a second stage containing the separator.CLOUD PIONSPions produced by the decay of K° into tt+ and iT are referred to as cloud pions. When these cloud pion trajectories are projected back to the plane through the production target normal to the proton beam direction (the source plane), these pions form a halo surrounding the proton beam spot. The Monte Carlo calculations and the results presented below are discussed in greater detail in another paper1.Due to the absence of sufficient data on the production of K°, we calculate the production cross section for K° by using the parameters appropriate to K+ in the Sanford-Wang formula2 ’3 ’14’5’ 6. Kinematic reflection7’8 is used for very low momentum particles. Yamamoto2 has compared measured low momentum production cross sections and found the Sanford-Wang results in agreement within a factor of better than 1.5 over the range 0.5 to 2.0 GeV/c. The results presented below are in the form of ratios; we may expect these ratios to be accurate within a factor of two.All results have been obtained for a point source and an angular acceptance of 50 mrad (7.9 msr) symmetric about the primary proton beam direction.In the Monte Carlo calculations, a momentum and a direction are randomly chosen for a K° ray originating from the point source, this ray is allowed to decay in a random direction in the center of mass of the K° at a distance from the origin proportional to the natural logarithm of a random variable uniformly distributed over the range 0 to 1.The Sanford-Wang value is carried along as a weighting factor to be accumulated in the appropriate histogram bins. Daughter pion rays are rejected if they do not pass through a specified downstream circular154aperture corresponding to the specified angular acceptance. An accepted daughter ray is traced back to the source plane. The Sanford-Wang value for the parent K° ray is accumulated in the appropriate histogram bin for radial position of the daughter ray at the source plane and for the daughter ray momentum. Upon completion of all trials, the contents of each bin are multiplied by 0.1715 and divided by the total number of accepted rays.A l s o  momentum d i s t r i b u t i o n s  f o r  K* and  it1 a r e  a c c u m u l a t e d  f o r  r a n d o m l y  c h o s e n  momenta  a nd  d i r e c t i o n s ;  u p o n  t h e  c o m p l e t i o n  o f  a l l  t r i a l s  t h e  h i s t o g r a m  b i n  c o n t e n t s  a r e  d i v i d e d  b y  t h e  t o t a l  num ber  o f  t r i a l s .Finally, the ratio of appropriate bin contents are calculated to yield, as a function of momentum, the ratios: (cloud pion)/iT~, (cloud pion)/ir+, (cloud pion)/K~, (cloud pion)/K~, tT/K~ and tt+/K+.The r e s u l t s  a r e  p r e s e n t e d  i n  F i g s .  1 ,  2 and  3 b e l o w ,  r e p r o d u c e d  f r o m  r e f e r e n c e  1 .  The  e f f e c t i v e  s o u r c e  s i z e  o f  t h e  c l o u d  p i o n s  i s  q u i t e  l a r g e :  a c i r c l e  o f  -  2 cm r a d i u s  c o n t a i n s  50% o f  t h e  p a r t i c l e s .  F o r  ap r a c t i c a l  beam l i n e ,  t h e  q u a n t i t y  o f  i n t e r e s t  i s  t h e  f r a c t i o n  o f  t h e s er a y s  t h a t  l i e  w i t h i n  t h e  b a n d  o n  t h e  s o u r c e  p l a n e  p r o d u c e d  b y  i m a g i n g  b a c k  t h r o u g h  t h e  s y s t e m  t h e  m ass  s l i t  a s  s e e n  b y  p i o n s :  i t  i s  f o u n d  t h a ta b o u t  14% o f  t h e  c l o u d  p i o n  r a y s  l i e  w i t h i n  t h i s  b a n d .F o r  . 5  G e V / c  p a r t i c l e s ,  f r o m  F i g .  3 we s e e  t h a t  t h e  r a t i o ,  e x p r e s s e d  a t  t h e  s o u r c e  p l a n e ,  o f  c l o u d  p i o n s  t o  d i r e c t l y  p r o d u c e d  k a o n s  i s  £ 2 ;  f o r  a 16 m c h a n n e l  t h i s  r a t i o  w i l l  b e  e n h a n c e d  b y  a f a c t o r  o f  16x  b y  t h e  f a s t e r  d e c a y  o f  k a o n s  r e l a t i v e  t o  p i o n s  and  we e x p e c t  o f  t h e  o r d e r  o f  14% o f  t h e  c l o u d  p i o n s  i n c i d e n t  o n  t h e  m as s  s l i t  p l a n e  t o  p a s s  t h r o u g h  t h e  m ass  s l i t :  an  e x p e c t e d  c l o u d  p i o n  t o  k a o n  r a t i o  a f t e r  t h e  m ass  s l i t  o f  £ 4 . 5 : 1 .  T h u s ,  c l o u d  p i o n s  may a c c o u n t  f o r  a b o u t  h a l f  t h e  c o n t a m i n a t i o n  ( ~  1 0 : 1 )  o b s e r v e d  i n  l o w  momentum k a o n  c h a n n e l s .GLANCING ANGLE SCATTERINGREVMOC9 c a l c u l a t i o n s  h a v e  b e e n  p e r f o r m e d  f o r  . 5  G e V / c  p i o n s  i n c i d e n t  a t  g l a n c i n g  a n g l e s  o n t o  c o p p e r  and  i r o n  s u r f a c e s 1 0 . ( T h e  r e a d e r  i s  c a u t i o n e d  t h a t  t h e r e  a r e  t y p o g r a p h i c a l  e r r o r s  i n  t a b l e  4 o f  t h e  l a t t e r  p a p e r :  i n t e r c h a n g e  t h e  l a b e l s  IRON and COPPER; i n  t h e  f i r s t  l i n e  o f  t h e  t a b l e  f o r  - 7 7 . 5  r e a d  - 4 1 . 7 ,  f o r  8 5 . 6  r e a d  5 5 . 4 . )  A n g l e s  o f  i n c i d e n c e  o f  10 t o  20  mrad r e s u l t  i n  50% t o  35% o f  t h e  r a y s  b e i n g  " r e f l e c t e d "  b a c k  w i t h  rms momentum l o s s e s  o f  o f  t h e  o r d e r  o f  5% t o  7% and  rms c h a n g e s  i n  r a y  d i r e c t i o n  o f  o r d e r  o f  70  t o  100  m r a d .  I t  d e p e n d s  v e r y  much o n  t h e  d e s i g n  o f  a  p a r t i c u l a r  c h a n n e l  w h e t h e r  o r  n o t  r a y s  w i t h  s u c h  p a r a m e t e r s  w i l l  b e  t r a n s m i t t e d  b y  t h e  d o w n s t r e a m  o p t i c a l  s y s t e m .PRACTICAL CONSIDERATIONS FOR DESIGNS FOR A KAON FACTORYTo  d a t e  l o w  momentum k a o n  beam l i n e s  h a v e  b e e n  a s  s h o r t  a s  p o s s i b l e  t o  m a x i m i z e  t h e  k a o n  f l u x  a t  t h e  e x p e r i m e n t a l  l o c a t i o n .  T h e s e  e x p e r i m e n t a l  l o c a t i o n s  h a v e  b e e n  f a i r l y  c l o s e  t o  t h e  p r o d u c t i o n  t a r g e t  and  t h e  beam t r a n s p o r t  s y s t e m s  t o  t h e  m a ss  s l i t  h a v e  h a d  o n l y  tw o  b e n d i n g  m a g n e t s .155For such designs it is not clear how many of the undesired particles travel by line of sign from the target region to the experimental apparatus and it is not clear how many travelled along paths outside the beam pipe. Berien11 reported for the CERN K26 channel that reducing the field in the first bending magnet to zero reduced the contamination at the experiment by a factor of only . 5x.Secondary particles incident on any material near the production target can produce tertiary particles that when traced back to the source plane would appear to originate from near the primary proton beam spot; another source of a halo. Therefore, it seems desirable to have as little material as possible close to the downstream proton beam. For this reason, septum magnets and half quadrupoles as first elements in the extraction system are less desirable compared to large aperture dipole magnets.Electrostatic separators are not maintenance-free devices: they will have to be well shielded from the production target so that residual radioactivity levels are low enough to allow maintance. This probably means that they cannot appear in the early stage of a channel at a kaon factory.A 0.8 GeV/c KAON CHANNELBerien11 suggested that an initial stage of the channel be used to form a stigmatic achromatic image of the source spot at a slit location to remove the halo of "cloud" particles and that this be followed by a stage containing the separator. A design for a .8 GeV/c channel12 is presented in Fig. 4. The distance from the source to the mass slit is 20.62 m (3.24$ and 63.1$ survival for .8 GeV/c kaons and pions respectively). Dipoles B1 and B2 form a MAXIM system13. Magnets B5 and downstream are superferric magnets1 1 5  with dipole field strengths < 3 T and quadrupole gradients  ^20 T/m. The channel acceptance is -56.0 to 64.0 mrad in the x (bend) plane, -12.0 to 15.0 mrad in the y (non-bend) plane, -1.0 to1.5$ momentum band; all values are full width at half maximum.The channel from location Intf (the first stigmatic achromatic image) to location MS (the mass slit) has reflection symmetry about plane B6A and the separator has been split into two sections to conform to this symmetry. This does not represent "two-stage separation" since there is no y image at B6A: the rays from a point source on the axis at the production target are parallel to the y = 0 plane at this location.The separator parameters are 1 MV across a 14 cm gap (71.4 kV/cm) with a total separator length of 4.76 m.A sextupole correcting element is located at magnet B6 to produce a smaller y beam spot at MS (the mass slit). Even with this sextupole correction, the tail of the y beam spot overlaps the kaon beam spot. However, a TURTLE16 (see Fig. 5) run shows that the rays in the tail of the pion beam spot at MS (y  ^0.9 cm) are spatially well correlated in the plane B6A, and can be removed by a triangular slit at B6A. With this triangular slit the resulting TURTLE calculated y beam spots at MS are as given in Fig. 6; there is a clear separation of the two distributions.156It will be noted from Fig. 4 that the channel configuration is a very favorable one for shielding: there is ample room to locate shielding to stop line-of-sight rays from the production target travelling to the experimental location and the total bend produced by the channel is about 180° so that the section of the channel downstream of B7 looks in a direction almost parallel to that of the proton beam line.REFERENCES1. D.E. Lobb, Nucl. Instr. and Meth., in press (1986)2. A. Yamamoto, KEK81-13 (1981)3. J.R. Sanford and C.L. Wang, BNL 11279 (1967)4. J.R.Sanford and C.L. Wang, BNL 1 1 479 (1967)5. C.L. Wang, Phys. Rev. Lett. 25, 1068 (1970)6. C.L. Wang, BNL 17218 (1972)7. M. Zeller et al, BNL 16000,193 (1970)8. D. Berley, BNL 50579, 257 (1976)9. C.J. Kost and P.A. Reeve, Proc. Conf. on Computers in Accelerator Design and Operation, Lecture Notes in Physics 215 (1983) 158.10. D.E. Lobb, IEEE Trans. Nucl Sci. NS-30, 2827 (1983)11. P. Berien, Proc. Int. Conf. on Hypernuclear and Kaon Physics,Heidelburg, 371 (1982)12. D.E. Lobb, A .8 GeV/c Kaon Channel for LAMPF II, LAMPF Report LA-10599-MS (1985)13. C. Tschalar, this Conference (1986)14. D.E Lobb and P.A. Reeve, Proc. Tenth Int. Conf. on Cyclotrons and their Applications, (1984) 85.15. D.E. Lobb, Proc. Ninth Int. Conf. on Magnet Technology, (1985) 722.16. D.C. Carey, K.L. Brown and Ch. Iselin, SLAC Report 246,UC Report 28 (1982)157Fig. 1. The spectra of if±, K1 and cloud pions as a function of momentum obtained by numerical integration of the Sanford-Wang formula for a cylindrically symmetrical angular acceptance of 50 mrad half angle (7.9 msr) for the four incident proton momenta of 13. 19, 25 and 31 GeV/c. Each curve has been normalized separately to have a maximum value of 10 arbitrary units.158p (GfV/c) p (G«V/c)F i g .  2 .  The  r a t i o  a t  t h e  p r o d u c t i o n  t a r g e t  a s  a f u n c t i o n  o f  momentum o f  d i r e c t l y  p r o d u c e d  tt+ t o  K+ and  tt~ t o  K~ f o r  1 3 ,  1 9 ,  25 and  31 G e V / c  i n c i d e n t  p r o t o n s . 1-------- ra o u o  p io n  t o  t t *  r a t io.•*f‘o°0000° a » o♦a °° o°°o0°°0 ° o 0Xo o°o  v  ♦ o  o° ***$ Oo %  v aQ O O PROTON MOMENTUMo aO°° o (G#V/C)°  A  31. . . »  * »O O 19o  13Og*«* o •V A 7„A_L_ _L_0.5 1.0 p (G«V/t)1.5 2.0402.0050.20.1T T" "T "CLOUO PK)N TO K *  RATIOPROTON MOMENTUM (G*V/d A 31 7  23 O 19 o  13a0 * 7* 7 ***o  a  * 7  AO o _ _   ^7 7 An  a „  „ O a ° ° ° „o a o u o _ p000 °°0A , o ° o .o O o0. I -------------------------- rCLOUO PION TO TT ~  RATO-xxx*1ftooOo°o O7  AA♦ o7 a p * nO0°a □ oO. o o o °a -0oPROTON MOMENTUM (G # V *)A  31 v  2 3  O 19 O 13 I___________________ L_O Q0.5 10p (GeV/c)CLOU) PION TO K '  RATIOPROTON MOMENTUM(G «V*)A 31 7  23 a  19 o  13° 7 aA A ^ A a a .0  4 4 t n 5 7 “ ’ : f t Ao o ’MS*.*o 0 ooaao ’pp “a»o D° o 0°o„0.3 10 P (GeVtt)IS 2 0 0.5 1.0 p (G«V/c)15ea _2.0F i g .  3 -  The  r a t i o s  o f  c l o u d  p i o n s  t o  ir+ , ir' o f  momentum.K+ and K a s  a f u n c t i o n159+ C2Bl B2Fig. 4. A .8 GeV/c kaon beam line. The centers of curvature of the central trajectory in the bending magnets are labelled C1 to C7. The four deflecting magnets at the entrances and exits of Sep 1 and Sep 2 are not illustrated. The x axis is normal to the central trajectory in the plane of the diagram; the y axis is normal to the central trajectory and normal to the plane of the diagram.160- 5  0 0 0  T O  - 4  7 3 0  T O  - 4  5 0 0  T O  - 4  2 3 0  T O  - 4  0 0 0  T O  - 3  7 5 0  T O  - 3  5 0 0  T O  - 3  2 5 0  T O  - 3  0 0 0  T O  - 2  7 5 0  T O  - 2  5 0 0  T O  - 2 .  2 5 0  T O  - 2  0 0 0  T O  - 1  7 5 0  T O  - 1  5 0 0  T O  - 1  2 3 0  T O  - 1  0 0 0  T O  - 0  7 3 0  T O  - 0  5 0 0  T O  - 0  2 5 0  T O  0  0 0 0  T O  0 . 2 3 0  T O0  5 0 0  T O  0 . 7 3 0  T O1 0 0 0  T O  1 2 5 0  T O  1 5 0 0  T O1 7 3 0  T O2  0 0 0  T O  2  2 3 0  T O  2  5 0 0  T O  2  7 3 0  T O0 0 0  T O  2 3 0  T O  5 0 0  T O  7 3 0  T O  0 0 0  T O  2 5 0  T O  5 0 0  T O  7 5 0  T O000 -15 000 -10 000 * *  * * —  - * *     — -*12 4  1 3  1 2 2 2 4 7 3 4 3 2 1  1 1  1 2 7 5 8 2 21 1 1 2 3 1 6 4 5 3 4 3 2  11 1 1 1 3 4 7 3 6 6 4 2 2 1 1  2  2 1 5 4 1 6 3 4 3 2 1 1I  2  2 4 3 6  4 4 7 3  1 3 1 8 3 3 4 1 4 2 3 1  1 2 1 1 5 4 3 3 1I I  1 7 3 1 3 5 4 1  \ j  1 1 2 1 3  1 2U  222-5  000 0 000 5 . 0 0 0 10 000 15 000• 1 « ------- » •* ••--Fig. 5. TURTLE16 results for the location at B6A of the pion rays that have y £ 0.9 cm at MS. The vertical axis is y (cm), the horizontal axis is x (cm); both are oriented in the local coordinate system of magnet B6: to obtain values in the global coordinates multiply both x and y values by -1.161-3 coo t o  - i  *00 t o- I  0OO TO - »  7 0 0  TO - 1  4 0 0  TO - I  5 0 0  TO - 1  4 0 0  TO - I  3 0 0  TO - 1  3 0 0  TO - I  IOO TO - 1  COO TO - O  4 0 0  TO - O  8 0 0  TO - O  7 0 0  TO - o  4 0 0  TO- O  4 0 0  TO - O  3 0 0  TO - O  3 0 0  TQ - O  1 0 0  TO  0  0 0 *  TO0 .  IO O TO  O 3 0 0  TO O 3 0 0  TOJLSflflL-Ifl--O 5 0 0  TO 0  4 0 0  TO O 7 0 0  TO 0  6 0 0  TO 0  4 0 0  T O1. 0 0 0  TO I .  1 0 0  TO- 3  0 0 0- 1  4 0 0  - J  8 0 0  - »  7 0 0  - I  4 0 0  - 1  3 0 0  - I  4 0 0  - 1  3 0 0  - 1  3 0 0  -I 100 - I  000 - 0  4 0 0  - 0  BOO - O  7 0 0  - O  4 0 0  - 0  3 0 0• 3 0 0  - o  3 0 0  -O IOO0 000 0. 100 O 3000  3 0 0  0  4 0 09. aofl..0  4 0 0  0  7 0 0  0  BOO0  4 0 01 OOO I 100 I  3 0 0K beam spot at MS_u4 03 0 77 3 81 0 4 811184 1 44 4 33 0 0_383 3XXXXXXx x x x x xx x x x x xx x x x x xx x x x x xx x x x x xx x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x xx x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x  X X X x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x I x x x x x x x x x x x x x x x x x x x x x xXX X X X X X X X X X X X X X X X x x x x x x xI N T E R V A LL £ B 8  TH AN - I  0 0 0 0TT- 1  COO TO - 0  4 0 0 0- O  4 0 0 TO - 0  8 0 0 0- O  6 0 0 TO - O  7 0 0 0- O  7 0 0 T O - 0  4 0 0 0- O  4 0 0 T O - 0  3 0 0 0- O  5 0 0 TO - 0  4 0 0T o " - 0  3 0 0 0- 0  3 0 0 TO -0 3 0 0 0- O  3 0 0 TO - 0  1 0 0 0- O  IOO TO 0  0 0 0 00  0 0 0 TO 0  1 0 0 00  1 0 0 TO 0  3 0 0 00  3 0 0 TO 0  3 0 0 00  3 0 0 TO 0  4 0 0 00  4 0 0 TQ 0  3 0 0 00  3 0 0 TO 0  6 0 0 60 600 TO0  7 0 0  TQ0  7 0 00  8 0 0 -I beam spot at MSSlits to transmit the kaon beam0  eoo TO 0  4 0 0 10  4 0 0 TO 1 0 0 0 71 0 0 0 TO 1 t o o 1 8 XX1 10 0 TO 1 3 0 0 6 8 XXX XXX1 3 0 0 TO 1 3 0 0 2 0 1 XXX XX X X X X X ! x x x x x x x x x1 3 0 0 TO 1 4 0 0 4 1 7 XXX (XX X XX X X X 1 XX XX X XX1 4 0 0 TO 1 3 0 0 7  l O XXX x x x x x x x x x x x x x x x x x1 5 0 0 TO I  6 0 0 7 3 6 XXX x x x x x x x x x x x x x x x x x1 4 0 0 TO 1 7 0 0 7 8 0 XXX x x x x x x x x x x x x x x x x x1. 7 0 0 TO 1 BOO 5 4 4 XXX x x x x x x x x x x x x x x x x x1 8 0 0 TO 1 4 0 0 4 8 1 XXX x x x x x x x x x x x x x x x x x1 4 0 0 TO 3  OOO 3 1 2 XXX x x x x x x x x x x x x x x x x x3  0 0 0 TO 2  1 0 0 2 0 0 XXX x x x x x x x x x x x x x x x x x3  IOO TO 3  3 0 0 1 3 33  3 0 0 TO 3  3 0 0 6 6 XXX XXX3  3 0 0 TO 3  4 0 0 3 0 X X3  4 0 0 TO 2  3 0 0 4 13  5 0 0 TO 2  6 0 0 33  tOO TO 2  7 0 0 13  700 TO 3 BOO 63  8 00 TO 3 4 0 0 0x x x x x x x x x x x x x x xx x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x xX x x x x x X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X  x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x X X X X X X X X X X X X X X X X X X X X X I  x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x xx x x x x x x x x x x x x x x x x x x x x x xXXFig. 6. TURTLE16 results for the kaon and pion y beam spot distributions at MS with sextupole correction at B6 and a triangular slit at plane B6A. Each distribution is normalized separately. The vertical axis is y (cm); the horizontal axis is the number of rays per bin.162DEVELOPMENT OF ELECTROSTATIC SEPARATORSAkira Yamamoto National Laboratory for High Energy Physics (KEK),Oho-machi, Tsukuba-gun, Ibaraki, 305, JapanABSTRACTDevelopment and recent status of electrostatic separators used in secondary beams and in other applications are reported. The future developments to realize higher fields beyond 100 kV/cm are also discussed.1. INTRODUCTIONIn the field of high energy and nuclear physics, electrostaticseparators have been widely used to purify secondary particle beams. Especially for the separation of low energy kaons and antiprotons from other particles , the electrostatic separators are essential elements in the separated beams. The developments of the electrostatic separators were started at LBL11 and CERN2-^  before 1960’s and many kinds of basic studies were carried out at the both laboratories. An idea of heated-glasselectrodes was proposed by Murray and it was established at LBL, BNL andR A L 1 5,6) Qn Qthg,- hancj, anodized aluminum was found to be suitable ascathode material and it was investigated at CERN and at KEK.3,71 A new concept of electrostatic separators with built-in high voltage generators were proposed by Kusumegi and it was established at KEK.8^In recent years, the electrostatic separators have been also used in colliding accelerators to separate the orbits of electrons and positrons (protons and antiportons) each other in the colliding sections. 3,1 ^  An unique new application of electrostatic separator has been found in a mass spectrometer system for nuclear physics experiments.1512. MASS SEPARATION BY ELECTROSTATIC SEPARATORSThe electrostatic separator is basically composed of a pair of parallel plane electrodes assembled in a vacuum tank. The separation of particle trajectories of different masses is performed by the use of electric and magnetic fields, at right angles each other and to the track of the particles. When charged particles transverse the electric field perpendicularly, they are deflected with the angle (0) expressed as9)where e, E, I and p are the electric charge, the electric field, thelength along the beam axis and the particle momentum. The term c is thelight velocity and /3 is the normalized velocity of the traversingparticle, which is related to the particle rest mass m by /3 = p/mc2j. In a practical beam line, the angular deflection 0 is cancelled by small magnetic field B (= E/c/3) perpendicular to the vertical electric field to keep the trajectry of wanted particles straight along the beam axis.Typically, an electric field of £ = 100 kV/cm corresponds to B = 333 gauss at /3 = 1. The separation of the deflection angle (A0) between a wanted particle (w) and an unwanted particle (u) can be expressed asVOLTAGE C kV163(2)For highly relativistic region, A0 is approximated byp^ c'(3)This expression shows that the angular deviation is inversely proportional to the third power of the particle momentum and the way in which electric field is related to the separation and, more importantly, how very high electic fields are needed for high energy particle beams.3.1. High voltage break-down in vacuumThe electrostatic separator can be used in the pressure region of IO"4 Torr in which the discharge in vacuum shows an interesting feature of pressure dependence on the break down voltage, as shown in Fig. 1. The relation of the pressure to the breakdown voltage has been empiricallywhere Vo is the breakdown voltage in high vacuum ,k is the parameter and p is the pressure. The term Vo is mainly determined by the species of the material used in the electrodes and their surface condition in ultra high vacuum. Gap dependence rarely exists in large gap ( > a few cm) system. It has been generally accepted that the dominant mechanism causing breakdown is both the field emission and the vaporization of the micro particles3'n) . In the region of 10~4 - 10-5 Torr, phenomena are still vacuum break down, but the effect of the residual gas exists. The term k is determined by both the properties of residual gas and electrode material. The possible reasons of the pressure effect are charge transfer and elastic collision between microparticles (or ions) and residual gas molecules. They can suppress the kinetic energy of the particles striking the electrode, and consequenty reduce the probability of the growth of the breakdown.3. HIGH VOLTAGE TECHNOLOGY IN VACUUMexpressed7^ asV*ox ~  Vo +  k  ■ p  ( @ P < P c ) (4)electrodes: electrodes /V= V0 .k PV0 = t ( t c , - - )k=g(lg,g,m)I I I I I1x10-5 1x10-4 PRESSURE C Torr ]Fig. 1. Pressure dependence of the breakdown voltage in log and linear scales.164In the region of 10“3 Torr (p>pc ), the breakdown behaviour changes completely. The breadown voltage strongly depends upon the pressure and follows basically Paschen’s law in gaseous discharge. The critical pressure is determined by the maximum gap distance between electrodes and the ground surface (of vaccum tank). It corresponds to te mean free path of the electron along electric field flux line in residual gas.3.2 ElectrodesFrom the view point of vacuum discharge, cathode should be carefully investigated because the cathode is the origin of electron emission which takes an important role in the growth of discharge in vacuum.Murray had proposed the heated glass as the cathode material because of its reasonably adjustable resistivity depending on the temperature and its thermal conductivity to reduce the current density and to decouple the energy during the discharge. The heated glass cathode was successfully established and adopted in the separators at LBL, BNL and other laboratories. At CERN, excellent performances were obtained with anodized aluminum electrodes, which were stable and well reproducible, without heating.3) Similar results to those at CERN were also obtained at KEK. The optimum thickness of anodized surface seemed to be 5 - 10 pm. The possible reasons for the excellent characteristics are its reasonably high resistivity, high melting temperature, high mechanical strength and the smooth surface due to chemical erosion. They help to prevent the electron emission and the growth of wiskers. Furthermore, aluminum as the base metal has good thermal conductivity, which can diffuse quickly the local heating and, consequently, can suppress the further thermal electron emission leading the discharge. The anode is considered to be more tolerable than the cathode, then stainless steel is usually used as the anode material because of its chemical stability, high melting point and other convenient characteristics.3.3. Gas and pressure effectThe conditioning process is inevitable before the cntinuous operation in the beam line. It is started in high vacuum (lowest pressure) under the condition of a constant current of a few tens micro ampere until the voltage reachs up to the saturation point, then the gas is injected gradually with increasing the voltage. When the votage reaches up to the ultmate value in the pressure region of 10”4 Torr, the conditioning is finished. On the other hand, deconditioning behaviour excists due to surface contamination. The characteristic (V - P) curve looks to shift towards higher pressure during the continuous operation, as shown in Fig.2. The pressure width between the critical pressure pc and the vacuum breakdown boundary at the operation voltage is called "pressure plateau”.The effects of the gas injected into the vacuum tank were studied by Danloy et al.10-1 and by Yamamoto et al.7) . Typical results are shown in Fig. 3. The neon/helium mixture or pure neon gas seems to be superior in terms of both the voltage maximum and the pressure plateau. It was remarked by Danloy that impurity of organic vapour and water-vapour reduced the maximum voltage and accelerated the deconditioning speed.I0) The wide pressure plateau and slow deconditioning rate enables the stable and long continuous operation.3.4. Size effect on the critical pressureThe maximum distance between the electrodes and the inner wall of the165IO O C P - V  C U R V E  J O C  S E P A R A T O R  M A R K  I  /9 0 0F O R  j —  1 ~  2  S p a rk s /ftB U B B L E  C H A M B E R  L IN E  f \  ELECTROCE j j j8 0 0 P O S IT IV E ---------S U S  3 0 4  .mfjN E G A T IV E ---------------A n . A l  I j !G A P ------------------------1 0 0  m m  /  • fG 4 S ------------------------ N . - H .  ? & / / !H  V  mo *  ■ IO O  5  M V  / . ' /____ _ O ct. 2 97 0 0 m mm* NOV. 6-------- *  Nov. I I6 0 0 OECCNOITIONING RATE / / /  ■ E 5 . I 0 - *  T w r / d o y  / / /5 0 0r * J//// ' / /4 0 C ■—* — *— •o  3 0 0►-s :r\I t i 0 * «  I « I0 * «  I * l 0 * 4p«essu«c ( Torr)Fig. 2. Pressure dependence of breakdown voltage after theconditioning at the KEK-Mark II separator. Behaviour of the deconditioning is also shown.pressure C Tor r ]Fig. 3- Gas effect on the breakdown voltage.166vacuum tank determines the critical pressure in relation to the mean free path of the electron in the residual gas. The small radius (distance) of the vacuum tank brings the shift of the critical pressure towards higher pressure, and it enables the wider pressure plateau and the higher breakdown voltage, as shown in Fig. 4.3.5. Crossed magnetic fieldThe crossed magnetic field in the electrostatic separator causes complex influence on the pressure dependence of the breakdown. Figure 5 shows the effect of the magnetic field on the breakdown voltage reported by Sanford.5) Especially, the critical pressure at the low electric field shits towards lower pressure due to the effect of the Penning discharge induced by existing crossed magnetic field. When the breakdown occurs, it is required to swith off the magnetic field in order to recover the electric field. It is preferable to install the magnets locally at both ends of the separator, if it was tolerable in the beam design. It makes the system simple and stable in the practical use.3.6. High voltage systemIn a conventional way, high voltage cables are used to feed high voltage from high voltage generators to electrodes through feedthrough insulators. The high-voltage cable brings many technical problems into the system, such as insulation, connection and others. Furtheremore, the shielded coaxial cable has unnecessary capacitance and, therfore, the electrostatic stored energy in the cable causes additional damage of the electorodes, insulators and other elements in the breakdown. Even if a protection resistor is located on the way of the cable, the unnecessary capacitance remains in the cable.These problems were completely solved with a new techique to mount the high voltage generators directly onto the vacuum tank. It was realized as a result of the development of the compact high voltage generator as a key element. Figure 6 shows cross sections of the electrostatic separator with built-in high voltage generators (Mark II at KEK). The main parameters of the high voltage generators are summarized in Table 1. The built-in high voltage generator has brought high and stable performance of the separator at KEK.The built-in high voltage generator has been adopted in the separator systems at TRIUMF and LAMPF, and has been considered to be adopted in a new separator system at BNL.4. RECENT STATUS OF THE ELECTROSTATIC SEPARATORS4.1. Status at CERNAlthough operations of the electrostatic separators at CERN were terminated in 1981, due to the close of separated beam lines, their excellent performances should be mentioned. In 1960’s - 1970’s, the technology of the electrostatic separator at CERN was established. To realize them, improvements were made with respects to 1) the size(reduction) of the vacuum tank, 2) new and optimized design of theinsulator made of fine ceramic and 3) optimization of the species of gas mixture. A short separator, shown in Fig. 7, achieved the highestperformance of 100 KV/cm (800 kV / 8 cm) as a working field.10'1 The mainparameters of the short separator are given in Table 2.167Fig. it. Size effect of the vacuum tank.T o n h  P r t s j u r o  ( T o r r )Fig. 5- The effect of the crossed magnetic field.mean free path168Fig. 6. Cross sections of the electrostatic separator KEK-Mark II with built-in high voltage generator.Fig. 7. Cross sections of the CERN short separator.169Table 1. Main parameters of a Built-in High Voltage generator for the KEK electrostatic separator Mark I and Mark II. _____Model Mark I Mark IIType Cockcroft-WaltonDimensionsMain circuit Box 1.Z7 x 0.7 x 0.3 m3 0.7 x 0.7 x 0.2 m3Feedthrough Bushing 0. l<p x 0.9 m 0.1?) x 0.6 mMaximum Voltage 600 kVMaximum Current 1 mAStability at 600 kV IE-4Ripple at 600 kV IE-4Input frequency of C. V. 16- 20 kHzNumber of Steps of C. W. 30Feedback resisiter 6000 M0Shunt resister 6000 MQSerial protection resister 100 MQFilter time constant 0.1sTable 2. Electrostatic separators used in separated beam linesItems CERN BNLName MARK III BS1 BS2Beam Line LESB IIVacuum TankMaterial SUS304 SUS304 SUS304Inner Diameter 64 cmElectrodeMaterials (Cathode) Anodised Al Heated Glass Heated Glass(Anode) SUS304 Heated Glass ST. SteelWidth 40 cm 61 cm 61 cmLength 2 m 2.1 m 2.1 mGap 8 cm 15 cm 10 cmV-max. for Conditioning 1100 kV 800 kV 800 kVWorking Voltage 800 kV 680 kV 680 kVMaximum field strength 135 kV/cm 53 kv/cm 75 kV/cmWorking field 100 kV/cm 45 kV/cm 68 kV/cmBreak down rate 0.1 0.1 sparks/h 0.1 sparks/hWorking pressure 10E-4 Torr 10E-4 Torr 10E-4 TorrGas Ne-He N2 N2Deconditioning rate 7E-5 Torr/lOOhPressure plateau 1.5E-4H.V. Feed Cable Cable CableMagnetic Field Crossed field Crossed field Crossed fieldOperation - 1981 1979 - 1979 -1704.2. Status at BNLTwo new short separators were developed in 1970’s for the low energy separated beam LESB II.115 The heated glass cathode was used, and a new configuration of the electrodes and shield covers was adopted. A cross section of the new electrodes is shown in Fig. 8. The main parameters of the new separator are given in Table 2. Another electrostatic separator of 10 cm in gap and 6 m in length has been operated in the medium energy separated beam MESB.4.3 Status at KEKThree types of electrostatic separators with built-in high voltage generators have been developed. These parameters are summarized in Table3. The first one, Mark-I, had a large vacuum tank of 1.4 m in diameter and relatively large high voltage generators. Figure 9 shows an end view of Mark I and the high voltage generator. One of two Mark-II’s has achieved the highest performance of 1005 kV / 10cm at the maximum field and 950 kV / 10cm at the working field in the continuous operation.12^ . It was achieved with an effort for the improvement about 1) optimization of the configuration of the electrode and the shield-cover, 2) reduction of the size in the vacuum tank and in the high voltage generators, 3) new design of the insulator made of high purity alumina fine ceramic, 4) the use of minimum organic materials in fabrication process especially in polishing process of the electrodes and the vacuum tank and 5) an automatic control system for the conditioning and the continuous operation in order to minimize the damages due to the breakdown. Figure 10 shows an end view of Mark-II.Mark-III is a 2 m short separator with crossed magnetic field. It was specially designed for the low energy separated beam K3 with the momentum range of 0.5 - 1.0 GeV/c. Its basic design is similar to the one ofMark-II, except for the configuration of the high voltage genrators attached to the top and the bottom of the vacuum tank. A view of Mark-III in the K3 beam line is shown in Fig. 11. Since 1979, Mark-III has been operated in a high raidation area (HO3 rad/h) near the K3 production target location and the primary beam dump (10^  rad/h)13) for 1012 protons per pulse. No indications of the radiation damage has been observed in the high voltage system since the start of the operation.4.3 Status at TRIUMF and LAMPFThe electrostatic separators has been used in muon beam lines at TRIUMF and at LAMPF, to separate muons from electrons. Their operational fields are 400 - 500 kV per 10 - 20 cm by the use of the heated glass electrodes.I4,l5) The TRIUMF separator has equipped built-in high voltage generators recently developed.4. OTHER APPLICATIONS OF THE ELECTROSTATIC SEPARATORSRecently, electrostatic separators has been used in colliding accelerators to separate beam vertically between electrons and positrons (protons and antiprotons) each other in the colliding sections. I7) The main parameters of the electrostatic separators at CERN-SPS and KEK-TRISTAN are given in Table 4. Figures 12 (a) and (b) show an outlook and an end view, respectively, in the construction stage of the KEK-TRISTAN separator. This separator has two differrent operation conditions from the one of the secondary beam separator. It has been used171Fig. 8.Cross section of the new electrode configuration in the BNL separator.Fig. 9- End vlev of the KEK-Mark I separator and a bullt-lnn high voltage generator.Fig. 10. End viev of KEK-Mark II. Fig... 11. A Viev of KEK-Mark IIITable 3. Electrostatic separators used In separated beam linesItems KEKName MARK I MARK II MARKIIIBeam Line Kl Kl(Kh,K2) K3Vacuum TankMaterial SUS30U SUS30k SUS301*Inner Diameter lhO cm 80 cm 80 cmLength 9 m 6(3) m 2 mElectrodeMaterials (Cathode) Anodised Al Anodised Al Anodised ALs(Anode) SUS301* SUS301* SUS30UWidth UO cm 1*0 cm 1*5 cmLength 9 m 6(3) m 1.9 mGap 10 cm 10 cm 15 cmV-max. for Conditioning 900 kV 1005 KV 870 kVWorking Voltage <800 kV <900 kV <750 kVMaximum field strength 90 KV/CM 100.5 kV/cm 75 kV/cmWorking field 70-80 70-90 U0-50 KV/cmBreak down rate < 1 sp/h < lsp/h < 1 sp/hWorking pressure 2E-1* Torr 2E-U Torr IE-1* TorrGas Ne-He Ne-He Ne-HeDeconditioning rate IE-5 Torr/d 3E-6 Torr/d 2E-5 Torr/dPressure plateau § W.F. 6E-h Torr IE-3 Torr 7E-1*H.V. Feed Built-in Built-in Built-inMagnetic Field Separated Separated Crossed fieldOperation 1977-1981 1979 - 1979 -Table h. Electrostatic separators used In colliding acceleratorsItems CERN KEKName L SAccelerator SPS TRISTANVacuum TankMaterial A1(A6063-T5)Inner Diameter 58 cm 30 cm -Length 3.2 m 5.»» 3-8ElectrodeMaterials (Cathode) Ti(-AL) A1(A6063-T1)(Anode) Ti(-Al) A1(A6o63-T1)Width 16 cm lU.5 cmLength 3 m U.7 m 3-3 mGap 2 - l6 cm 8 cmV-max. for Conditioning 300 kVWorking Voltage 200 kV 100-200 130-225 kVMaximum field strength 75 kV/cm 2k kV/cm 28 kV/cmWorking field 50 kV/cm 12-2U 16-28 kV/cmBreak down rate <0.1 sp/hWorking pressure IE-10 Torr IE-10 IE-10 TorrWorking current 100 uA 1*0 uA 1*0 uA173Fig. 12. (a) layout of the separator in TRISTAN e+e- collider and (b) an end viev of the separator in assemble stage.Table 5 Main parameters of the CARP systemMode FI / Mode F2Mean orbit radius in D 100 cmB-max. In D 10 kGMax. magnetic rigidity 100 kG.cmMean orbit radius in E 300 cmMax. electric field 25 kV/cmMax. electric rigidity 7-5 MeVMaximum energy 3-8 MeV/chargeTotal path length 7.3 m / 8.6 mFig. 13. (a) layout of the CARP system and (b) a viev of the separator vith built-in high voltage generator installed in the system.174in an ultra high vacuum region of 10“10 Torr and enormous heat induced by RF structure of the electron beam has to be removed from the electrodes to the outside. A very quick swith-off (<1 ms) of the field is also required in its operation.An unique application of the electrostatic separator was made by Morinobu et al. A reaction product mass separator named CARP was designed and built as a new experimental facility in the AVF cyclotron at RCNP of Osaka University.I8) It was intended to separate unslowed recoiling products in nuclear reactions from the primary beam, and to analyze them according to their charge-to-mass ratio. An electrostatic mass separator having a pair of curved electrodes has been developed as the heart of this system. A layout and a view of the CARP system are shown in Figs 13 (a)and (b), respectively. The system has been succesfully operated since last year.5. FUTURE DEVELOPMENTSOn the basis of the recent experience of the electrostatic separator, we consider the feasibility of field strengths beyond 100 kV/cm in 10 cm gap in near futrue.In principle, it shoud be feasible from a view point of theoretical limits of both field electron emmision and emission of microparticles. Their limits are considered to be E > 107 V/cm and a difference of the order of magnitude exists between the theoretical limits and the present practical overall field strength between electrodes (E =10^ V/cm). It means that enormous field enhancement excists at local spots, edges and generated whiskers on the surface of the electrodes and of the inner wall of the chamber.It is sure that a simple and smooth configuration and very smooth and clean surface of the elements help to suppress the local field enhancement. Especially, the smooth surface of the electrode made of the optimized material having high melting temperature, good thermal conductivity and high mechanical strength must be important to get andkeep the high voltage and field in vacuum. On the other hand, the pressureeffect is also important to get the higher voltage (field). Thus, followings are pointed out as items to be investigated.1) Material for the elecrode and the vacuum tank.(for example, anodized aluminum vaccum tank.)2) Finishing of the electrode and the vaccum tank.3) Configuration of each element inorder to eliminate the field enhancements.4) The way to reduce the stored energy or the way to minimize the energy dump in the electrodes.5) Reduction of the size of the vacuum tank (about 60 cm).6) Compact and reliable voltage feedthrough (insulator).7) Species of the gas mixture (for examples, Xe , other innert gas andtheir mixture) and exclusion of impurity.8) Compact and reliable high voltage generator (1000 kV).As a conclusion, a field strength of 150 kV/cm seems to be feasible in near future with hard efforts for the items mentioned above. The field strength beyond 200 kV/cm seems neither being feasible nor being effectivebecause that other serious effects such as primary beam halo and other175background particles at the mass slit may become dominant in the practical mass separation, even though it would be realized.ACKNOWLEDGEMENTSThe author would like to thank Professors H. Hirabayashi, Y. Yoshimura, H. Ishimaru and T. Momose for their valuable discussions. Thanks are also due to Mr. H. Ishii for his continuous efforts for the developments and the operation of the electrostatic separators. He wish to appreciate Prof. S. Morinobu of Osaka University for his kind informations and discussions.REFERENCES1) J. J. Murray, Proc, Int. Conf. on High-Energy Phys. Instr. (1960) 25.2) C. Germain and R. Tinguely, Nucl. Instr and Meth. 20 (1963) 21.3) F. Rohrbach, CERN report CERN 64-50.4) F. Rohrbach; High Voltage Technology, ed L. L. Alston(0xford Univ. Press, London, 1968) 350.5) J. R. Sanford, BNL-AGS, Internal Report JRS-2 (1964).Proc. Int. Symp. on Insulation of High Voltage in Vacuum,6) K. H. Davies, RAL Report RHEL/R 222 (1971)7) A. Yamamoto et al., Jpn. J. Appl. Phys. 16, 2 (1977) 343.8) A. Yamamoto et al., Nucl. Instr. and Meth. 148 (1978) 203.9) A. P. Banford, The transport of charged particle beams, (E & F.N.Spon Limited, London, 1966) 121.A. Yamamoto, KEK Report KEK-81-13.10) L. Danloy and P. Simon, Proc. Vth Int. Symp. on Discharges and and Electrical Insulation in Vacuum, (Poznan-Poland, 1972) 367.11) V. Kovarik et al., Nucl. Instr. and Meth. 158 (1979) 371.12) H. Ishii et al., Proceedings of the 3rd Symp.on Accelerator Science and Technology, RCNP of Osaka Univ., (1980) 259.13) J. McElroy, private communications.14) H. A. Thiessen, private communications.15) R. Brown et al., Proc. Xth Int. Symp. on Discharges and Electrical Insulation in Vacuum, Columbia (1982) 451, IEEE Cat. No. 82CH1826-7.16) T. Momose and H. Ishimaru, J. Vac. Sci. Technol., A4(2) (1986) 244.17) S. Morinobu et al., Proc. 4th Int Conf. on Nuclei For From Stability, Helsingor (1981) 717. CERN Report 81-09.176LIST OF PARTICIPANTSR.E. ABEGG, TRIUMF, Vancouver, B.C., Canada T. ANDERL, Universitat Bonn, Bonn, West GermanyE.J. ANSALDO, University of Saskatchewan, Saskatoon, Sask., Canada E.G. AULD, University of British Columbia, Vancouver, B.C., Canada J.L. BEVERIDGE, TRIUMF, Vancouver, B.C., CanadaE.W. BLACKMORE, TRIUMF, Vancouver, B.C., CanadaP. BLUM, Inst. f. Kernphysik, Kernforzentrum Karlsruhe, Karlsruhe, West Germany J. BREWER, University of British Columbia, Vancouver, B.C., Canada G.S. CLARK, TRIUMF, Vancouver, B.C., CanadaM.K. CRADDOCK, TRIUMF and Univ. of British Columbia, Vancouver, B.C., Canada N.E. DAVISON, University of Manitoba, Winnipeg, Manitoba, Canada P. DEPOMMIER, University de Montreal, Montreal, Qudbec, Canada J. DOORNBOS, TRIUMF, Vancouver, B.C., Canada G. DUTTO, TRIUMF, Vancouver, B.C., Canada J. ERNST, Universitat Bonn, Bonn, West GermanyD. FITZGERALD, Los Alamos National Laboratory, Los Alamos, NM, U.S.A.D.G. FLEMING, University of British Columbia, Vancouver, B.C., CanadaD.R. GILL, TRIUMF, Vancouver, B.C., CanadaD.P. GURD, TRIUMF, Vancouver, B.C., CanadaE.W. HOFFMAN, Los Alamos National Laboratory, Los Alamos, NM, U.S.A.C. HOJVAT, Fermilab, Batavia, IL, U.S.A.R.R. JOHNSON, TRIUMF and Univ. of British Columbia, Vancouver, B.C., Canada K.R. KENDALL, TRIUMF, Vancouver, B.C., Canada P. KITCHING, TRIUMF, Vancouver, B.C., CanadaD.M. LAZARUS, Brookhaven National Laboratory, Upton, NY, U.S.A.D.M. LEE, Los Alamos National Laboratory, Los Alamos, NM, U.S.A.D.E. LOBB, University of Victoria, Victoria, B.C., CanadaJ.A. MACDONALD, TRIUMF, Vancouver, B.C., Canada G.H. MACKENZIE, TRIUMF, Vancouver, B.C., Canada C.R. MARK, TRIUMF, Vancouver, B.C., CanadaS. MARTIN, Inst. f. Kernphysik, Kernforschungsanlage, Jiilich, West Germany T. NUMAO, TRIUMF, Vancouver, B.C., CanadaC. ORAM, TRIUMF, Vancouver, B.C., Canada A.J. OTTER, TRIUMF, Vancouver, B.C., CanadaD.J. OTTEWELL, TRIUMF, Vancouver, B.C., CanadaM.A. PACIOTTI, Los Alamos National Laboratory, Los Alamos, NM, U.S.A.D. RENKER, SIN, Villigen, SwitzerlandJ.R. RICHARDSON, UCLA, Los Angeles, CA, U.S.A.R. RUI, Univ. di Trieste, Trieste, ItalyM. SALOMON, TRIUMF, Vancouver, B.C., CanadaJ.T. SAMPLE, Science Council of B.C., Burnaby, B.C., CanadaM. SEVIOR, Univ. of British Columbia, Vancouver, B.C., CanadaP. SIEVERS, CERN, Geneva, SwitzerlandG.M. STINSON, University of Alberta, Edmonton, Alberta, Canada A. SULLIVAN, CERN, Geneva, SwitzerlandH.A. THEISSEN, Los Alamos National Laboratory, Los Alamos, NM, U.S.A.I. THORSON, TRIUMF, Vancouver, B.C., Canada R. TRELLE, TRIUMF, Vancouver, B.C., Canada C. TSCHALAR, SIN, Villigen, SwitzerlandE.W. VOGT, TRIUMF, Vancouver, B.C., CanadaR.D. WERBECK, Los Alamos National Laboratory, Los Alamos, NM, U.S.A.U. WIENANDS, TRIUMF, Vancouver, B.C., Canada Z. WU, IHEP, Beijing, People's Republic of ChinaA. YAMAMOTO, KEK, National Laboratory for High Energy Physics, Ibaraki, JapanS. YEN, TRIUMF, Vancouver, B.C., Canadam N £;*,^  .,*/ ;• ' HR4Cm m I4 %  It .•*>•% f: jmm1. r s';X  y" : ' >,v ’r .,.; ■* - r - i .»-v ,fv ■ v ■ ,4V■ « s .*>■ ’i-fl-f:{* : .Vjj* ■/ . ;- -!S  | '-■ ..V J, r ' #' ..CLJV/-w  ■ < i £ ^ N•. X-* • -hM a » ;:-,M\  W ; *  i  •i t o ,1- f * P  - -s •*r •« ;v.P i  i i  " ,-sm W . „ - : wW w ^ m  ,.11 ®S * ;/.;*5s r ! S f ! ’H*(iw s a  ,i # W 'I ® :-- WSJte* V-.■-; »f* 4s4'',  * » &-; v-' «i/f  V ‘i J a r  t'■, ‘ ->■' ..c:,.■ v . f•-■* -  -  ' 2  V - V .  ,»■ * ■?■?• , : , i ;-r- i ■ * 1 v - v. . !■’ ::.i. /  -  ,  :  I ,* • 'A •, :-■H  • ^d  -:; / W  V  ' ■ W r  4'iV % 9 S *-■ ...“ ‘Mi S ®  '


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items