Proton Beam Structure at Fermilab. Mike Syphers Fermilab

Transcription

1 Proton Beam Structure at Fermilab Mike Syphers Fermilab 1

2 Proton Availability Daily Operation - Once a p-pbar store is arranged in the Tevatron,... - produce more antiprotons, and drive the neutrino program! time line is governed by 15 Hz Booster operation - 7 Boo batches to MI every 2.2 s! 5 to NuMI! 2 to pbar production - spare pulses to mboone - 1 pulse to SY120 once in a while... 2

3 Proton Availability Daily Operation - Once a p-pbar store is arranged in the Tevatron,... - produce more antiprotons, and drive the neutrino program! time line is governed by 15 Hz Booster operation - 7 Boo batches to MI every 2.2 s! 5 to NuMI! 2 to pbar production - spare pulses to mboone - 1 pulse to SY120 once in a while... 2

4 NOvA, after Run II Following the completion of Run II of the Tevatron, antiproton production will presumably cease, and the neutrino program -- NOvA -- will ultimately utilize 12 Booster cycles to fill the MI every s sec = 20 * 1/15 s (20 Booster cycles) - leaves 8 Booster cycles for other program(s) 3

5 Booster In order to optimally utilize unused Booster cycles, all Booster components need to run at 15 Hz - Recent upgrades to Booster bring the usable repetition rate to 9-10 Hz - Further upgrades to RF systems are necessary to reach 15 Hz (other components, such as kickers, etc., can run at this level already) - 15 Hz issues have been identified and documented; part of overall facility improvement plans over next few years 4

6 Beam Throughput and Radiation Safety New particle rates for the pbar rings: - presently, Debuncher/Accumulator receive approximately 25x10 10 particles per hour; for Mu2e, expecting ~2x10 13 /sec! Total beam intensity ~4 times pbar record! Peak current ~5-6 times present record - 1% loss per cycle (scaling)! ~290 W of beam loss power! BOO: ~500 W total, ~1 W/m (300 W, 0.6 W/m in uncontrolled regions) Will require improved safety mitigation for pbar rings - true for any other 8 GeV program envisioned, not just Mu2e 5

7 Meeting the Mu2e Requirements (Baseline Proposal) Proton Beam - Using Booster cycles not identified for neutrino program, accumulate charge in Accumulator, transfer to Debuncher, and form single bunch for slow spill - In principle, w/ 4x1012 (4 Tp) per Booster batch, Mu2e receives 18 Tp/s on target, 1.8x1020 in 107 s. - Assumes 15 Hz Booster is in operation - Will not affect NOvA proton economics 6

8 Mu2e Baseline Beam Proposal Experiment cycle time of 667 ms - Transfer 3 consecutive Booster batches into Accumulator ring; form 1 bunch x transfer bunch into Debuncher ring; phase rotate to form 40 ns (rms) bunch; slow extract repeat twice during single s NOvA cycle - NOvA uses 12/20*15 Hz cycles = 9 Hz - Mu2e would use 6/20*15 Hz cycles = 4.5 Hz! 18x10 12 p/sec (18 Tp/s) on average ( GeV) 7

9 Beam Transport from Booster Figure 2: Overview of Booster-Accumulator-Debuncher system. and Recycler Requires: - inj/extr to/from RR ring; µ-e Experiment Slow extraction Transfer to Debuncher Protons from Booster via Recycler ( into is part of NOvA) 8

10 Bunch Formation Proton Beam Energy Figure 5: Accumulator + Debuncher bunching and phase-energy rotation. The beam is first adiabatically bunched in the Accumulator using an h=1 rf system (0 to 6 kv), then transferred into the Debuncher where it is phase-energy rotated (40 kv)and then bunched at h=4 (250 kv). Figure 1.4: The left figure shows a conceptual schematic of momentum stacking. On the right is shown a simulation of the capture and momentum stacking of four Booster batches. momentum stack in Accumulator 1.4 Resonant form singleextraction bunch; x-fer to Debuncher A: initial debunched beam. Resonant extraction is a well established technique to extract beam slowly from a synchrotron. It involves moving the tune of a circulating beam close to harmonic resonance, such that beam becomes unstable and migrates to high amplitude. Generally, the high amplitude particles are intercepted by an electrostatic septum, in which the field is produced by a very thin wire plane, followed by a Lambertson magnet approximately 90 later in betatron phase. In practice, two types of resonant extraction have widely used: MJSbeen / Fermilab B: After adiabatic bunching in Accumulator. phase rotate, re-capture 40 nsec bunch,!p/p ~ 0.8% (rms) C: After φ-e rotation in Debuncher D: After h=4 bunching in Debuncher. 9

11 RF Requirements Accumulator: - 53 MHz (h=84), 80 kv (~30-50 kv presently avail.) khz (h=1), 4 kv (~2 kv presently available) Debuncher: khz (h=1), 40 kv (~0.5 kv at present) MHz (h=4), 250 kv (~0.8-2 kv at present) 10

12 Resonant Extraction Once beam is in the Debuncher, slow spill over next 9 Booster cycles (600 ms) Resonant Extraction process - adjust betatron tune to be near rational value septum - use feedback to control rate of slope particle extraction generic example... position 11

13 Resonant Extraction Half- vs. Third-integer extraction - traditionally, Fermilab has used half-integer resonances for extraction; at final stage, can put ALL beam on resonance and extract every particle - with third-integer, always some beam left over, which must be aborted prior to next injection First-pass simulation was performed using 3 rd -integer, with Debuncher ring parameters 12

14 Extra Slow Spill from Debuncher (prelim.) Mu2e Proposal Momentum stack 6 Booster batches directly in Accumulator (i.e. Table 1.1: The approximate parameters of the third order resonant extraction, with the septum located between the Q101 and Q602 quadrupoles.reverse direction) Kinetic Energy (GeV) Working tune (νx /νy ) Resonance (νx ) Normalized acceptance (x/y π-mm-mr) Normalized beam emittance (π-mm-mr) βx at electrostatic septum (m) βx at Lambertson (m) βx at harmonic sextupoles (m) Septum Position (mm/σ) Septum gap/step size (mm) Sextupole Drive Strength (T-m/m2) Initial Tuneshift (δν) Septum field (MV/m) Septum length (m) / /3 285/ π 15 π 22 π 14 11/ Capture in 4 kv h=1 RF System. Transfer to Debuncher Debuncher lattice and realistic magnetic elements used in simulation Phase Rotate with 40 kv h=1 RF in Debuncher Recapture with 200 kv h=4 RF system - (no space charge)!t~40 ns s Phase Space Results beam particle arriving at the stopping target. The first category of background be reduced by improving the electron energy resolution. The second category is ced by delivering the proton beam in short bunches separated by τµ, ensuring perfect proton extinction between bunches, and restricting the search for the ersion of stopped muons to the inter-bunch period. - inefficiency: wire thickness over step size : Table 1.2: The most significant beam backgrounds, for protons on the primary target and an extinction factor of There would be 5 signal events if Rµe ~ mm/10 mm ~ 1% Radiative π Capture

15 Extraction Issues For MR, Tevatron, MI operation, momentum spread was always relatively small (<10-4 ) Here, as seen previously, expect: MeV/8.9 GeV ~ 2% (full width) - space charge also an issue; needs careful analysis - Note: AGS uses(d) 3rd-integer w/ large!p/p, but beam was debunched! will require fine control of chromaticity, or perhaps some new scheme 14

16 Strong Space Charge Space Charge will be an issue in the baseline scenario... - For N particles uniformly distributed about the ring, 3r0 N 3 ( )( ) νs.c. = 2"γ 2 (v/c) = - Include bunching factor : 2 (20π 10 6 )(9.52 ) nsec B nsec 2π - Thus, expect at design parameters : νs.c

17 tion. Both straight sections have zero dispersion function due to special supressors on the arcs. The adjustment of the quadrupoles gives the tune near the integer+l/j. The influence of the metallic walls on the beam is investigated by changing of the vacuum chamber size. Figures 2 a,b show the phase portraits of the beam in resonance under different significance of the current: a)avt=o.ol, b)avl=0.03, c)av~=o.06 and d)avl=o.l. Every time the quadrupoles are set for tune near 10+1/3.From these figures one can see, how the space charge acts on the ori- Extraction w/ space charge The Space Charge Effect in Slow Extraction Resonance by Third Integer EPAC94 Yu.Senichev,V.Balandin Institute for Nuclear Research of RAS, 60-th October Anniversary prosp., 7a, Moscow, , Russia 1 INTRODUCTION With the development and construction of high intensive beam accelerators and storage rings more and more attention is being focussed on the problem of the self-field effect of accelerated particles on the stability of their motion. This problem endures second birth, which connected with, on the one hand, requirement to know more exactly the parameters of beam and on other hand, with more powerful computers for an investigation. At first the analytical methods were used in mainly, among which the equations of Kapchinsky and Sacherer take significant place. They gave necessary information about the envelope of high intensive beam with the elliptical distribution. As far as the improvement of computer technology, the new numerical methods have being developed for the solution of the motion and Poisson equations. However, up to right now one run of a modern full dimensions code for ring with some hundreds of elements takes hundreds hours, what makes impossible to optimize the lattice, Many authors synthesize the real tracking with fixed elliptical distribution, when the second moments of density are changed only[l]. Apparently this approach is correcting for the remote resonance system. III the case of being in resonance and even passing through resonance, the elliptical symmetry of distribution can be violated. In the paper [2] we studied the passing through half-integer resonance under different initial distribution with elliptical symmetry, and we have find the different behavior of hallo for each case. Moreover, the change in time of the second moments depends on, what kind of distribution is used. In this paper we study the excitation of the third integer resonance at high space charge for the slow extraction. We are interested, how to realize the isolated resonance under significant nonlinearity due to space charge. 2 &A&, y) - &A.&t Y), (1) of the where A,,, 9,, - the vector and the scalar potential of the exspace charge field and A,, -the vector potential ternal field. It is assumed here, that the transverse currents are absent: A,, = ;Q,,(l + ha), where u is the longitudinal velocity The vector potential of the external up to the octupole inclusive: (4 equal for all particles. field has components 2.&A (x y) = hx+ (K + h$ :(x3-32~7 + ;(z4 The Poisson equation lic boundary: v2*,c(x, is solved - K$ + - 6x2y2 + y4) on the grid with Y) = ;P(x, (3) the metal- (4) Y) Phase space distortions in the presence of space charge, 7I(A) near third-integer resonance, K = 4* lo- P3r3 can be very significant THE CODE DESCRIPTION It is obviously the third integer resonance changes the distribution significantly, in which it is difficult to recognize the elliptical distribution. We developed and use the four dimensional code for the solution of the motion and the space charge equation together. The motion equation are solved in the Hamilton s form. The Hamiltonian of system is represented in the common form for the curvilinear Friday, January 23, 2009 coordinates: For the rectangular boundary we use the FFT method for Poisson equation solution. However, there is the option for the arbitrary shape of the boundary, which uses the direct method of the matrix transformation of the equations system written in finite elements. The input coefficient for the code, which defines the space charge parameters and the beam energy, is K: In this representation equals: AvL the Laslet = K- linear FOR 2%lu shift of frequency (6) The output TWISS file of MAD, which describes the lattice parameters, is used as input file of the code. The figure 1 shows the typical screen picture, which gives the visual information in interactive regime about distribution in all planes, losses and the tunes in the horizontal and vertical planes. Figure 2: The phase portraits of the beam at a)avl=o.ol, b)avl=0.03, c)a~~=0.06 and d)ahvl=o.l 16

18 Debuncher Aperture gap height: in. = ± 30 mm good field width: 4 in. = ± 50 mm Extraction: - in simulations so far, step size = 10 mm; 20" emitt. - if double the beam dimensions, can quadruple emittance, mitigating space charge somewhat 17

19 Extraction Issues Will look further at Debuncher extraction with large momentum spread and space charge - perhaps examine other/new schemes Note: necessary equipment for the baseline design is understood - - electrostatic, magnetic septa - correctors, feedback circuitry 18

20 Alternate Proposal for Mu2e Perform all operations within single Booster cycle (66.7 ms) Perform bunch preparation in the Recycler - re-bunch Booster beam into 4 bunches using broadband RF system and h=4, h=8 RF systems (considered for Muon g-2 experiment at Fermilab) Transfer to Accumulator and Debuncher rings, performing slow extraction as before, but in 16 ms rather than 667 ms 20

21 Simulations* 8 MeV *C. Bhat and J. MacLachlan 21

22 Rotate into 4 Bunches Extraction of the 1st Bunch 22

23 rms!p/p Momentum Spread vs. Time broadband system only turn off broadband system, turn on h=4 system extract at this time p Time in cycle (s) RMS Booster in the Recycler vs. time [s] p The four maxima are at the times the bunch widths are 23

24 Possible Alternate Scenario 4 x 1 Tp REC Bunch formation scheme in Recycler developed for possible muon g-2 experiment; takes ~30 ms to perform New RF and kicker requirements in REC, ACC, DEB Would slow spill over ~16 ms, rather than over ~600 ms; rate would be ~x3 higher per burst to experiment If no NOvA, would be ~90%+ duty factor; but with NOvA, only run 6-8 out of 20 cycles (30-40% d.f.) ACC 4 Tp BOO Recycler 66.7 ms ~30 ms 4 Tp DEB 1 Tp Lowers space charge tune shift by 1/12 Reduces momentum spread by 1/4 Allows for experiment to receive beam on any unused Booster cycle Accumulator Debuncher 4 Tp 1 Tp ~16 ms 24

25 When NOvA is off If have all Booster Cycles available for use in Mu2e: Baseline scenario; spill over 4 cycles rather than 9, then, - 4/5 = 80% duty cycle - 36 Tp/sec (ave.) - 45 Tp/sec (spill) - 7.7x107 per burst Easier to time share among other users... Alternate scenario -- all within 1 Booster cycle-- ~90% duty cycle or better - 60 Tp/sec (ave.) - ~10x107 per burst to the experiment 25

26 Accelerator R&D for Fermilab Beams Mike Syphers Fermilab 26

27 Near-term R&D Kickers: - Extinction AC dipole R&D -- on-going - Transfer Kicker R&D! For Accumulator/Debuncher transfers Bunch formation studies - use MI -- systems exist to form 4 bunches from a single Booster batch; takes longer than required for Mu2e, but can test the technique Extinction studies in Debuncher - transfer above bunch to DEB ( reverse proton injection), study/measure extinction Other beam studies?? (target studies, e.g.?) 27

28 Near-term R&D (cont d) Mu2e Extraction Line Design Slow Extraction Process Design/ Modeling Extinction Process Design / Modeling Radiation Safety Mitigation Analysis / Design Mu2e Target Station Design / Modeling 28

29 Critical Issues Intensity Limitations: - space charge effects -- how much charge can we put into a 40 ns bunch? - impedance, beam loading issues - determine optimal operational scenario RF requirements: - Continue with momentum stacking in ACC, or form bunches in Recycler? 29

30 Critical Issues Resonant Extraction: - studies are just beginning - 1/2 or 1/3 integer? - effects of space charge on extraction process - extraction inefficiency needs better estimates - requirements and expectations for the slow spill feedback circuit need development 30

31 Critical Issues Extraction Line: - design, including extinction insert, beam dump, etc., is just underway; fully determine physical constraints between the ring and experimental hall. Extinction Channel: - Currently only conceptual; finalize and take to an engineering design; need appropriate specifications for the required instrumentation for measuring and monitoring the level of achieved extinction. 31

32 Critical Issues Radiation Safety: - Need careful analysis of necessary safeguards for running high intensity beams in the antiproton enclosures. Better estimate expected rates. Design, cost passive and active safety measures Instrumentation: - Perform analysis of the present instrumentation and possible modifications or upgrades necessary to monitor bunched beam. 32

33 References Fermilab Proton Plan, see: D. McGinnis, Beam Emittances and RF structure at Injection into the Main Injector for the Multi- Stage Proton Accumulator, Beams-doc J. Reid, R. Ducar, Booster RF Repetition Rate Limit, Beams-doc Mu2e Collaboration, Expression of Interest to Fermilab, MU2E-doc-15. D. Neuffer, More Rebunching Options for the mu2e Conversion Experiment, Beams-doc Mu2e Collaboration, Proposal to Search for µ#n! e#n with a Single Event Sensitivity Below 10 #16, MU2E-doc-62. E. Prebys, Optimizing AC Dipole Specifications for Beam Extinction, Beams-doc V.S. Kashikhin, D. Harding, V.V. Kashikhin, A. Makarov, D. Wolff, Conceptual Design of AC Dipole Magnet for µ to e# Experiment, MU2E-doc-263. M. Syphers, Possible Scheme to Ameliorate Space Charge and Momentum Spread Issues, MU2Edoc