Super B Positron Production and Capture Freddy Poirier

Super. B Positron Production and Capture Freddy Poirier – LAL R. Chehab, O. Dadoun, P. Lepercq, A. Variola, 1 poirier@lal. in 2 p 3. fr

Super. B Positron Production Study 10 n. C Primary beam Linac for e 100 s Me. V e+/e. Target Pre-injector Linac for e+ ~280 Me. V ~300 Me. V 2. 856 GHz AMD e- gun Geant 4 1 Ge. V Damping Ring Parmela/Astra/G 4 Accelerating Capture Section 2. 856 GHz e(0. 6 to 1 Ge. V) Present study: 600 Me. V Tungsten Target W: 1. 04 cm thick ACS 2

Target Yields Studies Target Geant 4 simulation (O. Dadoun – LAL): 1. 7 For a 600 Me. V e- beam, the optimum yield is 1. 7 e+/e- with a W-target thickness of 1. 04 cm Previous studies* were using a yield of 1. 58 e+/e- with a target thickness of 1. 40 cm 3 * As shown at Super. B 2009 Meeting in Frascati.

The AMD Bl(z) (Gauss) • The Adiabatic Matching Device is based on a slowly decreasing magnetic field system which collect the positrons after the target. • AMD has a wide momentum range acceptance (with respect to systems such as Quarter Wave Transformers) • The AMD for the present Super. B Studies is 50 cm long with a longitudinal field Bl starting at Bl(0)=6 T decreasing down to Bl(50 cm)=0. 5 T Transverse emittance in AMD is transformed: 95% Ellipse of actual positrons distribution in AMD Where a = 22. m-1 4 Z (cm)

The AMD • Input: – 300 mm bunch out of the target – Yield = 1. 7 e+/e- for a 600 Me. V e- bunch Large Energy Spread <E>=~20 Me. V Erms=~40 Me. V Zrms=~2. 2 cm (tail!) Geant 4 Px (Me. V) Energy (Me. V) • Output from the 6 T 50 cm long AMD: Astra Good agreement (px/pz) 5 Z (m) X (m)

The ACS • Accelerating Capture Section (ACS) Goal: Collect and Accelerate positrons up to at least 300 Me. V. • The ACS is encapsulated in a 0. 5 T solenoid and includes several tanks: – Example: • • 6 tanks for Full Acceleration at 2. 856 GHz – Travelling Wave 1 tank = 84 cells (+2 couplers), ~3. 054 m RF: 2. 856 GHz, 2π/3 0. 9466 cm of aperture (constant radius) • The ACS is here simulated using ASTRA (gain in flexibility) 3. 054 m Solenoid Tanks ~300 Me. V Cells • The positrons are then accelerated up to DR energy requirements (~1 Ge. V). 6

RF in tanks • P. Lepercq (LAL) has calculated the Travelling Wave Fields in Super. Fish and adapted them for ASTRA’s simulation: Field line in a 6 cells TW cavity 2π/3 mode Longitudinal field in a single tank (2. 8 GHz): 25 MV/m Seen by ref. particle Adaptation and normalisation Adjustment of irises, RF to the required 2π/3 mode, group velocity, . . 7 1 tank = ~3. 054 m

The ACS • Capture and Acceleration depends on: – Strategy employed for collecting positrons • A) Full Acceleration – Straight out of the AMD, the particles are accelerated (25 MV/m) • B) Deceleration (first cavity) + Acceleration – The positrons are decelerated to form a small bunch (free peak gradient = ~10 MV/m) – This gives different scenarios depending on the type of RF cavities within the ACS • • 1 st scenario = 2. 846 GHz full acceleration 2 nd scenario = 2. 846 GHz deceleration + acceleration 3 rd scenario = 1. 428 GHz deceleration + acceleration 4 rd scenario = (A. Variola’s idea) combination of RF types (using 3 GHz TM 020 mode for deceleration and 1. 4 GHz TM 010 acceleration). This 4 scenarios are under investigation 8

What are the simulated ACS? ACS scenarios: 1 2 S-Band (Acc) (Dece) 4 3 S-Band/L-Band (Dece) Scenario Strategy (1 st tank) Frequency (GHz) Total Nb of tank to reach ~300 Me. V Aperture (cm) Length of the ACS (m) 1 Acceleration S-Band (2. 856) 6 0. 95 18. 94 2 Deceleration S-Band (2. 856) 7 0. 95 22. 01 3 Deceleration L-Band (1. 428) 4 2 25. 01 4 Deceleration S-Band / L-Band (3 (2 harmonic) / 1. 428) 4 2 21. 84 nd 9

2. 856 GHz: Acceleration scenario New results using ASTRA including the z. RMS at exit of the AMD of ~2. 2 cm Energy (Me. V) End of tank number 6: Population Energy (Me. V) • End of 1 st tank results: Z (m) Energy (Me. V) <E>=40 Me. V Erms=20 Me. V We used here a very stringent cavity phase which limitates the capture Zrms=5 mm Z (m) Total yield is 2. 8% with an Erms/E of 7% at 300 Me. V There is still room for further optimisation in the ACS tanks, we could increase 10 the yield and keep a relatively low Erms. and short bunch. Scenario 1

2. 856 GHz: Deceleration scenario • Find the RF phase which gathers a maximum of particles within a bunch • Find the peak gradient which helps this At the exit of the 1 st tank: Population -Maximise particles in this bunch -Minimise its length 280 o Population Goal: 200 o 280 o 10 MV/m -Minimise the other bunches Z (m) E (Me. V) 11 Scenario 2

At end of ACS – 2. 856 GHz • Full acceleration in downstream tanks (7 in total) is used after deceleration: Zrms=6. 4 mm Gaussian fit: ~289 ± 12 Me. V Energy (Me. V) Z (m) Total Yield here: 7. 5% Calculated yield for particles within 287± 10 Me. V: 3. 9% 12 Scenario 2

End of 1 st Tank – 1428 MHz • • 1 Tank = ~6. 10 m Deceleration mode Scenario 3 Energy (Me. V) – 250 o – 6 MV/m Again game is to catch maximum of particles in a small bunch Energy (Me. V) Z (m) Gaussian Fit sz = 4. 6 mm Rather short bunch and low 13 energy distribution

At end of 4 th tank – 1428 MHz ~25 m long beam line Energy (Me. V) • Acceleration on crest up to the 4 th tank leading to an average energy of roughly 300 Me. V Energy (Me. V) sz=8. 89 10 -03 m se=16. 9 (9. 09) Me. V ( but energy Tails!) Total Yield = ~32. 3% sx’=1. 69 10 -3 rad, sy’=1. 74 10 -3 rad sx=8. 0 10 -3 m, sy=8. 2 10 -3 m Transverse Emittance = 1. 35 10 -5 rad. m (=sx*sx’), Longitudinal Emittance=0. 08 Me. V. m (=sz*se) 14

At end of 4 th tank – 1428 MHz At 1 Ge. V, we want ± 1% i. e. ± 10 Me. V of energy dispersion. Energy (Me. V) Having an idea of the yield for ± 10 Me. V at 300 Me. V gives us an idea of how well our scenario work. • Yield for particles between 300 ± 10 Me. V: – – 19. 6% (994 positrons) sz=6. 4 mm sx’=1. 84 10 -3 rad, sy’=1. 76 10 -3 rad sx=7. 7 10 -3 m, sy=8. 3 10 -3 m 15

A 4 th Scenario • 3000 MHz TM 020 for Deceleration and 1428 MHz for acceleration – 1 st tank is a 3 GHz tank = 2. 93 m • Iris – Aperture larger (Here we constrained the radius opening to 20 mm) • compactified bunch length when deceleration • Shorter beam line for the 3 GHz TM 020 case (wrt 1428 MHz only) – 2 nd up to 4 th are 1. 428 GHz tank = 6. 10 m each • Tank gradient = 25 MV/m • Tank phase optimised for maximum acceleration on crest for the considered bunch • Because of the RF (1. 428 GHz), the wavelength is rather large and the energy dispersion due to acceleration on crest is minimised • 21. 84 m from the beginning of the AMD needed here to reach at least 300 Me. V 16

End of 1 st Tank – 3000 MHz • Length of 1 st Scenario 4 tank = ~2. 93 m – Cell length= 3. 331 cm Gaussian Fit sz = 3. 66 10 -3 m Energy (Me. V) • Tank Phase f 1= 280 o • Tank Gradient G 1=10 MV/m Energy (Me. V) 17 Z (m)

At end of 4 th tank – 3000 MHz ~21. 9 m long beam line Energy (Me. V) • Average Energy = ~333 Me. V Energy (Me. V) Z (m) Total Yield = ~31. 9% sz=3. 5 10 -03 m Z (m) se=5. 2 (3. 2) Me. V sx’=1. 4 10 -3 rad, sy’=1. 46 10 -3 rad Scenario 4 sx=8. 1 10 -3 m, sy=8. 1 10 -3 m 18

Recap • 4 Scenarios under investigation Scenario 1 2 3 4 RF (MHz) – strategy 2856 - acc 2856 – dec 1428 – dec (S-band) (L-band) 3000 dec + 1428 - acc Mean Energy (Me. V) 302 287 295 333 Erms (Me. V) 21. 4 32. 3 (12) 16. 83 (9. 09) 5. 2 (3. 2) Zrms (mm) 2. 7 6. 4 8. 89 3. 5 Xrms (mm) 3. 8 4. 4 8. 0 8. 1 X’rms (mrad) 1. 02 1. 11 1. 69 1. 4 Ex =X’X (mm. mrad) 3. 8 4. 6 13. 0 11. 4 Total Yield (%) 2. 8 7. 53 32. 3 31. 9 Yield ± 10 Me. V (%) 1. 3 3. 9 19. 6 29. 3 With a positron injection of 10 n. C and a yield of 3. 9%, we will have 2. 43 109 positrons at 300 Me. V ± 10 Me. V (scenario 2 – 2. 8 GHz) These values are a good indication of how well the scenarios work but it’s not enough 19

Extension to 1 Ge. V • What are the implications at 1 Ge. V (DR entrance)? – To get answers: • A simplified lattice was built up with a 0. 5 T solenoid extension from 300 Me. V up to 1 Ge. V (total length=~65 m) – Similar model as the lattice up to 300 Me. V. • A more complex lattice is under investigation including quadrupoles. 20

Extension to 1 Ge. V • No optimisation of the simplified lattice • Assumption on the DR requirements: ex geometric (10 -6 m. rad) 6 ey geometric (10 -6 m. rad) 6 Injection current (n. C) 10 Energy Acceptance at DR entrance (1 Ge. V) 1% (± 10 Me. V) Bunch current at end of linac (25 Hz) 1. 0 109 CDR 2 • Implication on the bunch current: Scenario 3 4 RF (MHz) 1428 3000 + 1428 Nb of particles 1. 5 109 2. 9 109 21

Conclusion • A first lattice from target up to ~300 Me. V and a simplified one to the DR was built up • Several Energy strategies have been studied as well as several RF scenarios • Some of the Scenarios lead to the required yield for the DR – Still a lot of room for optimisation of the lattice • We have not taken into account any “Safety knobs” which would increase the nb of e+ or help to relax requirements such as: – Higher drive e- beam energy – 10 bunches in the DR – Higher DR energy (will reduce the emittance by adiabatic damping) or larger transverse acceptance AMD, lattice optimisation might give some leverage Low energy primary beam can offer a good candidate to provide a sufficient and good quality positron beam. 22

More Slides 23

Pre-Conclusion • Comparisons (preliminary): – Acceleration / Deceleration • Deceleration allows to gain in total Yield • Bunch length shorter, Energy dispersion smaller – 2856 / 1428 MHz • Maximum total Yield for 1428 MHz • Energy dispersion (smaller for 1428 MHz) – 1428 / 3000 (TM 020) MHz • 1 st tank: – Total Yield higher for 1428 MHz – Same particle density (to be shown properly) – Shorter beamline for 3000 MHz 24

ACS previous studies (XIth Super. B Meeting) • Full acceleration case simulated with G 4+Parmela • But no realistic bunch length out of the AMD • This corresponded to the best we can get for the full acceleration case (without optimisation). AMD type Yield e+/e(production) AMD Yield (e+/e- in %) ACS type Yield at end of 1 st cav. (e+/e- in %) 6 T – 50 cm (600 Me. V) 1. 58 59. 7 L-Band (1. 3 GHz, r=1. 8 cm) 6 T – 50 cm (600 Me. V) 1. 58 59. 7 5. 9 T - 15 cm (200 Me. V) 0. 52 4. 6 Total Yield at end of ACS (e+/e- in %) 38. 9 34. 2** S-Band (2. 856 GHz, r=0. 95 cm) 18 (9. 7*) 5. 9^ S-Band 3. 3 (2. 0*) 1. 6** - Large impact of the accelerating technology used (mainly due to aperture) - Combined impact of the primary beam + AMD design Latest investigations are using ASTRA& &We gained flexibility and more realistic bunch length using ASTRA 25

ACS’s RF cavities • The Accelerating Capture Section (ACS) is encapsulated in a 0. 5 T solenoid and includes 7 tanks (6 for full acceleration scenario) – 1 tank = 85 cells, ~3. 054 m – RF: 2. 856 GHz, 2π/3 – 0. 9466 cm of aperture (constant radius) – Pierre Lepercq (LAL) calculated the field for the Travelling waves using Super. Fish Exemple including the field lines with the 6 cells travelling wave for Parmela Simulation: 26

ACS energy strategy – 2 (extreme) possible energy strategy scenario: • Acceleration mode – Straight out of the AMD the particles are accelerated – we use 25 m. V/m • Deceleration mode – The particles are decelerated to form straight a small bunch – choice of peak gradient for the cavities (free ~10 MV/m) Acceleration phase Deceleration phase Exemple with a 1. 4 GHz Cavity, 4 MV/m 27

Comparison – end of 1 st tank RF (MHz) Tank length (m) 1428 6. 10 3000 2. 93 Gradient (MV/m) Phase (best so far) sz (mm) 6 250 4. 6 10 280 3. 7 Yield 0<z<1 cm 26. 4% 26. 2% (1332) (1322) 37. 2% 32. 6% (1877) (1643) Yield -0. 5<z<1. 5 cm s. E (Me. V) -0. 5<z<1. 5 cm 3. 9 4. 9 • short tank length (3 GHz) • shorter bunch length (3 GHz) • same density of particles for a given short z size • More particles (1. 4 GHz) • smaller energy dispersion (1. 4 GHz not surprising) 28

At end of 4 th tank – 3000 MHz • As an indication: – 333 Me. V ± 10 Me. V Energy (Me. V) 323 343 29. 4% (1481 positrons) sz=3. 5 mm Note yield for e+ within 331 and 339 Me. V = 15. 9% 29

At end of 4 th tank – 3000 MHz • 1 st tank is a 3 GHz tank = 2. 93 m • 2 nd up to 4 th are 1. 428 GHz tank = 6. 10 m each – Tank gradient = 25 MV/m – Tank phase optimised for maximum acceleration on crest for the considered bunch – Because of the RF (1. 428 GHz), the wavelength is rather large and the energy dispersion due to acceleration on crest is minimised • 21. 84 m from the beginning of the AMD needed here to reach at least 300 Me. V 30

Population AMD (6 T – 50 cm) exit Energy (Me. V) Energy distribution at the exit from the AMD, using Geant 4 simulation 31

Deceleration Approach • Idea from SLAC in the late 70’s (SLAC-pub-2393) • Arranging the phase and amplitude of the fields so that the distribution in longitudinal phase space of the incoming positrons lay along one of the orbits in longitudinal phase space: 32

Simulation Specifics • Tools in use for simulations of the Adiabatic Matching Device (AMD) and the Accelerating Capture Section (ACS): – Parmela (LAL version) • AMD + ACS were simulated initially with Parmela – Though the AMD field inputs for Parmela was rather difficult to modify and to implement (as based on coils) – Some problems, due to lost particles with large angle at entrance of AMD, not resolved. – New Cavity field implementation for Parmela is time consuming. – Geant 4 (LAL version) • AMD field simulation done (analytical longitudinal and radial field) • No bunch length so far (work in progress) – Astra • AMD field simulation done (analytical) • ACS field with inputs from Super. Fish relatively fast to implement • Each code has its drawbacks – Though benchmarks have been done and show relatively good agreement: This work is in progress • • Geant 4 (AMD) + Parmela (ACS) have been used for the first batch of simulation (continued work) ASTRA is presently being used to simulate both ACS and AMD. – We gained in flexibility 33