A Project to convert TLS Booster to hadron accelerator 1. Basic design TLS is made of a 50 MeV electron linac, a booster from 50 MeV to 1.5 GeV, and a storage ring. The TLS storage ring is currently operating at TBA mode. The TLS booster is made of 12 missing-dipole-fodo cells. The dipole length is 2.4 m with bending radius 4.58 m. The circumference is 72 m. The horizontal betatron tunes for electron beam acceleration was 4.4 in order to minimize electron beam emittance. The magnetic field strength of the TLS-B can provide proton beams from 7 MeV to 850 MeV, and carbon ion beams from 6 MeV/u to about 400 MeV/u. The energy is most suitable for ion-beam cancer therapy. Proton C 6+ ion electron KE/u (MeV/u) 7 300 6 400 50 1500 Bρ (Tm) 0.383 2.695 0.706 6.347 0.167 5.00 β 0.121 0.653 0.113 0.715 1 1 γ 1.007 1.320 1.006 1.429 98 2935 B (T) 0.084 0.588 0.154 1.385 0.036 1.09 T 0 (μs) 1.977 0.368 2.126 0.336 0.240 0.240 f rev (MHz) 0.506 2.717 0.470 2.975 4.16 4.16 h 120 120 Since the vacuum chamber of the TLS-B is about 35 mm, we expect that the vertical admittance of the ion beam is about 34 π-mm-mrad or an rms vertical emittance of 5.8 π-mm-mrad. The horizontal admittance is larger than 50 π-mmmrad, and thus the corresponding space charge limit should be more than 2e10 particles per bunch, well beyond the required dosage of 2-3 Gy/minute for radiation therapy applications. 2. The injection systems: a) The ion beam injection requires a 7 MeV H source, i.e. an ion source with 3 MeV RFQ and 4 MeV DTL. The RFQ/DTL can be replaced by a commercially available 13 MeV cyclotron with negative charge extraction. b) For the carbon ion source, one can use similar source that has been used in the Heidelberg Heavy Ion therapy center. The injection line can be achieved by the injection chicane and strip-injection system. For proton and carbon ion injection, one needs two straight sections. c) Because the space charge limit for the vacuum chamber is about for rapid cycling, the injection can be accomplished by single turn kick injection. This is also a very clean injection scenario. The requirement is an ion linac with current of at least 1 ma.
3. Acceleration system: The low frequency broadband rf system can be designed using the MA or ferrite. These quarter-wave structures are well understood, and can be designconstructed easily. The power supply of main dipole is 10 Hz resonance circuit. With the resonance circuit, the rf voltage requirement is high to maintain 10 Hz ramp rate. Cavity design is challenging. Depending the capability of the cavity, the ramp rate may be limited. High gradient MA cavity has been shown to provide 15 kv per cavity, and the project needs two of such cavity. Other options are ferrite loaded cavities. Requirement of rf voltage in rapid accelerating accelerators Proton acceleration in IUCF cooler ring from 45 MeV to 500 MeV in 1 second where ρ 2.4 m. Using R 14 m, we obtain V sin φ s 240 Volts. Acceleration of protons from 9 GeV to 120 GeV in 1 s at the Fermilab Main Injector would require db/dt 1.6 Tesla/s. The circumference is 3319.4 m with ρ=235 m. The voltage requirement becomes Vsin s =1.2MV. Acceleration of proton and heavy ion at 10 Hz needs Vsinφ s 10 and 25 kv for proton and Carbon ion respectively. This means that one may need 2-3 rf cavities for the acceleration and bunching of the beam. Proton C 6+ ion electron KE/u (MeV/u) 7 300 6 400 50 1500 T 0 (μs) 1.977 0.368 2.126 0.336 0.240 0.240 f rev (MHz) 0.506 2.717 0.470 2.975 4.16 4.16 h 2 2 2 2 120 120 Beam delivery and transport system: 4. Extraction and beam transport systems: The logistic of the radiation physics room is important from the lab integration. The lab should allocate a most convenient location for the future expansion. The slow extraction system requires a wire-septum and a Lambertson extraction septum. Since the 10 Hz operation is favorable, I recommend 10 Hz rapid cycling fast extraction for the TLS-B proton project. The average current is about 3 na. Fast extraction can avoid the complication of slow extraction.
5. Beam Injection schemes and intensity limitation a) For the proton option, we propose to use the standard option of 3 MeV RFQ and 4 MeV DTL H- ion source. The beam is striped injected into the ring and adiabatic captured, or chopped into the bucket. The adiabatic capture efficiency is typically 90% or higher with a much smaller final phase space area. b) For carbon ion, we use the EBIS source with RFQ and DTL that can deliver 6 MeV/u at about Summary: Extending the capability of TLS-B to encompass the hadron beams is important to NSRRC. At 10 Hz ramp rate with about 2e10 proton per pulse, the average current is about 32 na. Such a system has the advantage of cyclotron for low peak current rad-scan, and advantage of easy energy variation by employing super cycle control system. This system has also the advantage of avoiding the complicated slow extraction design and manipulations. The extraction will be very clean. The conceptual design can be accomplished before the TPS commissioning. The construction of can be accomplished after the TPS operation. The cost of the proton source and 7 MeV RFQ/DTL or a 13 MeV cyclotron is about $3M. The cost of beam delivery system depends on the length of the beam line and where the experimental station. The project cost will be about $10M + civil construction. The cost of light ion needs additional evaluation on cost of ion source, rf requirements! Injection system and space charge limit Gillespie s formula
Possible Injectors LINAC FFAG E (MeV/u) C 4+ 7 6 N/pulse (10 9 ) 60 5.9 Emittance (πμm) 6.4 8.8 N_turn (10 9 ) 0.40 5.9 Pulse length (μs) 300 Tolerable foil hits 12 10 Accumulation turns 150 19 N_total (10 11 ) 0.60 1.0 Emittance (πμm) final 17 17 Space charge potential for Gaussian distribution Space Charge potential: Expansion: 4 th order Septum separation Injection beam clears from main dipoles Enough space for stripping foil assemply Systematic resonances are located at ν=3 and 2. 6 th order Using the 7 MeV/u linac or the 6 MeV/u FFAG C 4+ sources, we can easily accumulate a beam of 10 11 C 6+ with an emittance of 17 π-mm-mrad in the synchrotron. Linear:, 2
If one chooses the bare tunes at (6.23, 6.20), and 1000 injection-turns, with a total tune-shift of 1.1. Note that the tune shift of SNS is only 0.15! The tune for small amplitude particles continue to decrease as the particle is injected! Phase space map at the end of injection. Incoherent Laslett tune spread parameter for paraxial Gaussian beam 2 2 1 4 2 2 8 = Now we compare the space charge tune shift parameters of the Gaussian and the uniform KV-like distribution? 2 16 4 The space charge tune shift parameter of KV-like distribution function is reduced by a factor of 2. Space charge limit: p C 12 A 1 12 Z 1 6 Circum(m) 72 72 Einj/u (MeV) 7 7 beta_inj 0.12 0.12 gamma_inj 1.0075 1.0075 Eext/u (MeV) 250 400 p (MeV/c)/u 729 951 Brho (T-m) 2.43 6.35 dnu_sc 0.1 0.1 epsn (μm) 6 6 N_sc 6.0E+10 4.0E+10 To reach this number of particles in the accelerator, we need injector current less than 1 ma for 20 turn-accumulation! The resulting beam intensity in the ring is far beyond requirements of dose intensity in cancer therapy. Space charge effects: Linac delivers about 30 ma beam current to the Fermilab Booster, i.e. about 4.2 10 11 particles in one injection turn. a 0 b t b 1 t 2 0 K sc dt
Adiabatic capture Maximum Current Beam Current Ramping 1.5GeV Extraction Point PS dc Output Current 50MeV Injection Point 0 58% 99% Chen-Yao Liu, et al. Horizontal IPM measurements of the Fermilab Booster
Work on magnetic field ramping 1. Need multi-particle simulation on injection adiabatic capture, 2. RF cavity design and LLRF control 3. Need evaluate power supply requirement of fast ramping 4. Need to evaluate nonlinear field induced by fast ramping, i.e. eddy current on vacuum chamber 5. Need to calculate dynamic aperture based on nonlinearity of eddy currents on vacuum chamber Adiabaticity condition In a rapid-cycling synchrotron, magnetic field produced by the eddy current on vacuum chamber can be analytically obtained. S.Y. Lee, NIM A300, 151 (1991).