1. ADSR meeting@Cockcroft Inst.July 2009 From PAMELA to ADSR, T.Yokoi From PAMELA to ADSR Takeichiro Yokoi (JAI)

2. ADSR meeting@Cockcroft Inst.July 2009 From PAMELA to ADSR, T.Yokoi ADSR scenarios

3. ADSR meeting@Cockcroft Inst.July 2009 From PAMELA to ADSR, T.Yokoi Particle therapy to ADSR Particle therapyADSR Max Energy250MeV(proton) 400MeV/u(HI) 1GeV Current10~100nA (proton) >1mA Extraction energyVariableFixed Energy spread<1MeVO(MeV) Beam size 4  4~10  10mm - Repetition rate1KHz- Beam structurePulse/CWCW is preferable ReliabilityHigh(>99%) Operation cycle10h/day (>150day/cnt.) 24h/day (>150day/cnt.) Two big challenges……. 1.High intensity(>1mA)  space charge, beam loading, loss control 2. High reliability

4. ADSR meeting@Cockcroft Inst.July 2009 From PAMELA to ADSR, T.Yokoi Space charge tune shift Incoherent space charge tune shift Assuming an accelerator similar to PAMELA proton ring, ∆ H ~-0.27, ∆ V ~-0.40 @ 30MeV N bunch 6.24  10 12 (*1)  (H,V)200  mm mrad(unrm) Bunch factor0.1(*2) R6.5m (H/V)/ring 9.1/3.1  avr (H/V) 1/4 m *1: =1mA in 1kHz *2: =assuming h=1 To achieve ∆ <0.25, factor of 1.6 reduction is required

5. ADSR meeting@Cockcroft Inst.July 2009 From PAMELA to ADSR, T.Yokoi Space charge tune shift(cnt’d) To remedy the situation….. (1) Reduce N tot (Increase the rep. Rate) (2) Increase the beam size:  (ex painting, mismatched injection)  Large DA is essential (Need improvement from PAMELA) (3) Increase injection energy (difficult: cost of injector) Repetition rate of more than 1kHz is a requirement Horizontal Vertical

6. ADSR meeting@Cockcroft Inst.July 2009 From PAMELA to ADSR, T.Yokoi Reliability To obtain high reliability in an accelerator …… 1.Redundancy (injection and extraction are key element) 2.Degraded operation 3.Less pulsed elements(ex 10 12 msec =12days)

7. ADSR meeting@Cockcroft Inst.July 2009 From PAMELA to ADSR, T.Yokoi Slow Extraction in NS-FFAG ∆ v <0.5 ~2% of F/D ratio can change the vertical tune more than 0.5  In a lattice with vertical tune drift, by changing the D/F ratio, resonance energy can be varied  Half integer resonance can be used for the extraction : “ Energy variable slow extraction in fixed field accelerator” Resonance point H v

8. ADSR meeting@Cockcroft Inst.July 2009 From PAMELA to ADSR, T.Yokoi Slow extraction (Vertical) ESS ∆B 1 /B 1 (perturb):3  10 -3 @ V :3.5, ∆V/turn:90kV It can drastically ease the requirements for RF of PAMELA by reducing the repetition rate or multibunch acceleration Problems … 1. Dynamics of vertical motion (sensitive to fringing field distribution  3D field tracking is now under preparation) 2. Energy resolution is expected to be worse than that obtained in synchrotron. (*For the application of proton driver, it is not a problem) @ESS @septum ESS Tracking after ESS FDF ESS Septum

9. ADSR meeting@Cockcroft Inst.July 2009 From PAMELA to ADSR, T.Yokoi Charge exchange injection in PAMELA lattice FFAG in Ibaraki IBT project made use of main field in charge exchange injection  Need to place a foil in the magnet  impractical in PAMELA magnet Introducing an orbit chicane with DC bump is an another approach ( h /cell~0.75) Stripping foil DC bump Injection beam Combining with the slow extraction, it might be able to eliminate pulsed magnets from the ring Tracking study has just started…

10. ADSR meeting@Cockcroft Inst.July 2009 From PAMELA to ADSR, T.Yokoi Multi-bunch acceleration time Energy 1ms Option 1 time Energy 1ms Option 2 Option 1: P  N rep 2 Option 2: P  N rep Multi-bunch acceleration is preferable from the viewpoint of efficiency and upgradeability Low Q cavity (ex MA) can mix wide range of frequencies

11. ADSR meeting@Cockcroft Inst.July 2009 From PAMELA to ADSR, T.Yokoi Multi-bunch acceleration 2-bunch acceleration using POP-FFAG (PAC 01 proceedings p.588) ∆f  4 f sy Multi-bunch acceleration has already been demonstrated In the lattice considered, typical synchrotron tune <0.01  more than 20 bunches can be accelerated simultaneously (6D Tracking study is required)

12. ADSR meeting@Cockcroft Inst.July 2009 From PAMELA to ADSR, T.Yokoi N f :1 N f :2 N f :3 N f :4 N f :10 N f :15 N f :20 N f :25 N f :1 N f :2 N f :3 N f :4 N f :10 N f :15 N f :20 N f :25 Amplitude Distribution in Multi-Frequency Superposition N f :2 N f :4 N f :1 N f :3 Frequencies of constant frequency separation are superposed

13. ADSR meeting@Cockcroft Inst.July 2009 From PAMELA to ADSR, T.Yokoi Amplitude Distribution in Multi-Frequency Superposition Ex. In the case of 20 bunch case, ~97% of the time portion is <0.5V max.  Is the higher part crucially important from the viewpoint of beam acceleration? If not, requirement of maximum power of RF driver can be considerably reduced.  It can be experimentally examined using real FFAG beam (ex KURRI’s 3-FFAG) N f :1 N f :2 N f :3 N f :10 N f :20 N f :30 Multi-bunch acceleration is degraded operation in its nature

14. ADSR meeting@Cockcroft Inst.July 2009 From PAMELA to ADSR, T.Yokoi Remaining issues… Injector --- How to accumulate ~10 13 proton/bunch ? (in 1kHz operation, assuming 10  s accumulation time, the duty factor of injector =0.01) 1GeV system would be a cascaded system  new ideas other than those for 250MeV ring are needed.

15. ADSR meeting@Cockcroft Inst.July 2009 From PAMELA to ADSR, T.Yokoi Summary The biggest challenge of ADSR is handle huge beam current >1KHz rep. rate will be required to overcome space charge. Kicker-less FFAG might be possible using slow extraction and charge exchange injection Multi-bunch acceleration with low Q cavity might reduce the power requirements considerably 1 GeV ring needs further consideration