Normal-conducting Accelerating System for SuperKEKB

Transcription

1 Normal-conducting Accelerating System for SuperKEKB Tetsuo Abe KEKB-RF/ARES-cavity group High Energy Accelerator Research Organization (KEK) Outline 1. Overview of the KEK B-factory (KEKB) 2. Fundamentals of the ARES-cavity system 3. Upgrade for SuperKEKB 4. R&D topics from ARES 4. Summary and the future APPI For more details on (Super)KEKB, see (KEKB homepage), (SuperKEKB), (KEKB review 2005).

2 KEK B-factory (KEKB) Super-conducting 8 cavities for HER Circumference: ~3km Normal-conducting 20 cavities (ARES) for LER IP Normal-conducting 12 Cavities (ARES) for HER Asymmetric-energy e+e- collider Low Energy Ring (LER): 3.5GeV e+ High Energy Ring (HER): 8GeV e- Lorentz boost: βγ = CMS energy: 10.58GeV (Υ(4S)) Finite crossing angle: 11mrad 2 Producing huge number of B etc. Various physics researches Operation since

3 KEKB Achievement (1) ~~Peak luminosity~~ 2

4 KEKB Achievement (1) ~~Peak luminosity~~ Reached the KEKB Design in May /cm /sec 3

5 KEKB Achievement (2) ~~Total integrated luminosity~~ logged by Belle KEKB PEPII 4

6 KEKB Achievement (2) ~~Total integrated luminosity~~ logged by Belle KEKB PEPII Reached 100/fb in Oct

7 2005 KEKB Achievement (3) ~~Daily luminosity~~ Logged by Belle 6

8 2005 KEKB Achievement (3) ~~Daily luminosity~~ Pass a milestone again! Logged by Belle >1/fb/day! Continuous Injection L >~14/nb/sec 7

9 Yielding Many Physics Results Examples of the Belle results Observation of Large CP Violation in the neutral B Meson System, PRL 87, (2001) Observation of the Decay B K l+ l-, PRL 88, (2002) Observation of Large CP Violation and Evidence for Direct CP Violation in B0 p+p- Decays, PRL 93, (2004) Evidence for Direct CP Violation in B0 K+p- Decays, PRL 93, (2004) Compelling Evidence for Direct CP Violation in B->pi+pi-, hepex/ Discoveries of the new hadrons Etc 8

10 For Higher Luminosities Assuming that both beams have the same size and β*s, N N f Luminosity = R 4πσ σ γ + * * x y I L N: number of particles in a bunch of e+, e- f: collision rate σ : beam size at IP RL: geometrical reduction factor (not far from 1) Beam current: I± = N± f e ± ξ ± ± y Beam-beam tune shift parameter * * r N e β y 2er e β ξ y ± y = R * * * 2πγ ± σy ( σx + σy) (for a flat beam with short bunches) y Increase the beam-beam parameter. Make the beam size (b) at IP smaller. Increase the beam currents (I). 9

11 With higher beam currents Measures against, ex. Higher-power HOM Damages to various components (bellows, gate valves, etc.) Worse vacuum, vacuum leak More SR Serious in the arc sections Beam blowup due to electron clouds, etc. Solenoids, and any others? Coupled bunch instabilities (CBIs) Transverse Driven by the beam-ion, photoelectron effects, etc. Can be suppressed by the bunch-by-bunch feedback system Longitudinal Driven by the accelerating mode Can be suppressed by the accelerating-cavity system Vacuum-group task Feedback group task 10

12 Longitudinal Coupled Bunch Instability (CBI) driven by the accelerating mode In case of multi-bunch operation (1293 bunches in KEKB ( )) Synchrotron oscillation with interaction among bunches via wake field Coupled oscillation with # of degree of freedom equal to # of bunches Mostly caused by the longitudinal impedance of accelerating cavities Wake field Ex.M=4 11

13 For mode: µ 1 rf 2ET 0 revωs Growth Rate Ieηω τ µ R + ω µω { ZL( ω rf µωrev + ωs) } R ZL( rf rev ωs) (growth) Longitudinal impedance { } (damp) (ω s << ω rev << ω rf) Cavity resonance freq. Impedance Asymmetry w.r.t. wrf causes CBI. Resonator impedance 12

14 Higher beam current requires a larger detuning. Optimum detuning (For min. power with fixed V c) ω = ω hω R rev = = I sinφ 2V c s Pb tanφ 4πU R Q s a 0 f a Proportional to the beam current Ex. for the KEKB/LER design: Ι=2.6A ω = 2π 200 khz > ω = 2π 99kHz Inversely proportional to the stored energy If a large energy is stored, the detuning can be reduced. CBI suppression rev ARES 13

15 Accelerator Resonantly-coupled with Energy Storage 3-cavity system stabilized with the p/2-mode operation consists of HOM-damped accelerating cavity (A-cav), Energy-storage cavity with TE013 (S-cav), Coupling cavity (C-cav) with a parasitic-mode damper. Perpendicular to the beam axis Along the beam axis 14

16 Accelerator Resonantly-coupled with Energy Storage 3-cavity system stabilized with the p/2-mode operation Input Coupler Port consists of Top View HOM-damped accelerating cavity (A-cav) Energy-storage cavity with TE013 (S-cav) Coupling cavity (C-cav) with a parasitic-mode damper 15

17 Operation with the Accelerating p/2 Mode (Advantages) The field of the p/2 mode is the most stable against Beam loading, Detuning of A-cav ( = f a ) The stored-energy ratio: Us/Ua can be changed Us Ua f The parasitic 0 and modes k k 2 a = 2 s can be damped selectively out of C-cav (C-damper) f a π /2 = 1 +U / s Ua 16

18 Energy-storage Cavity (S-cav) Q0(S-cav) =~ 1.7x10^5 Can store a large electromagnetic energy in TE013 To suppress the longitudinal CBIs Optimum detuning f = ω hω S-cav s R Isinφ = 2V c s 0 R Q Pb tanφs = 4πU a 0 f a Ua f a f = f = π a Ua + Us 1 + Us Ua f a f = 200 khz in KEKB/LER 2.6A, 20 sets a f a = 710 khz in SuperKEKB/LER 9.4A, 28 sets Cf. Ua :energy in A-cav Us :energy in S-cav =9 (in KEKB) f rev = 99kHZ Movable tuner on A-cav ~2m 17

19 ARES in the KEKB Tunnel Design Parameters Vc Ra/Q0 Q0 Pin Pc Us/Ua 9 0.5MV 15 W 11x10^5 400kW 150kW (Waveguide from klystron) 18

20 ARES Operation (Jan.~Feb, 2005) LER: 20 cavities Total Vc: 8.0MV (0.4MV/cav) Beam current: ~1.6A Input RF power/cav: ~300kW HOM power/cav: ~5kW Trip rate: ~1/cav/3months HER: 12 cavities (+ 8 SCCs) Total Vc = 15MV( 13MV) = 4.09MV(ARES)+10.91MV(SCC) (0.34MV/cav) Beam current: ~1.20~1.27A Trip rate (ARES): ~1/cav/3months Stable Operation!!! LER(e+) HER(e-) ARES cavities in the FUJI RF section 19

21 Upgrade toward 20

22 Main Strategy Make the beam size smaller: y* = 6 mm y* = 3 mm x2 Final focus magnets closer to Belle Tighter clearance More SR background Increase the beam-beam parameter by: ξ y = 0.05 ξ y = 0.14 Can be achieved by the successful Crab crossing Increase the beam currents: 2.6 A (LER) / 1.2 A (HER) 9.6 A (LER) / 4.1 A (HER) x3 x4 Target: cm sec x24 21

23 Measures against Larger detuning On ARES Increase the energy ratio: Us/Ua = Higher HOM powers HOM-load upgrade f f a π /2 = 1 +Us / Ua Higher input RF powers (400kW/cav 800kW/cav) TiN coating on the coaxial line Coupler test stand upgraded for ~800kW(CW) 22

24 ARES R&D Programs [1] Design modification of A-cav for Us/Ua: 9 15 [2] Construction of a new L-band HOM-load test stand Using 1.25GHz klystron (1.2MW, CW) The 1 st stage just finished [3] Input couplers with TiN coating Against multipactoring in the coaxial line TiN(Titanium Nitride) has low secondary-electron yields and is good for vacuum. Two couplers have been completed. Being tested in the upgraded coupler test stand up to 800kW. [4] New highly-pure copper electroplating for S-cav The old facility has been retired. Reusing a facility being used for J-PARC. 23

25 [1] To Change the Energy Ratio /2 mode & eq. circuit Applying Slater s tuning curve to the /2 mode Changing the aperture size: h btwn A-cav and C-cav, Us/Ua, Qext,s/Qext,a can be changed. 24

26 Changing the Aperture Size of A-cav Simulation tool: HFSS S-parameter calculation The phase of the reflection coefficient: f P arg( S ) 2tan P = + P 1 RF P1 frf Geometry A-cav C-cav 25

27 Aperture Size v.s. Energy Ratio Modification of the aperture size: +15mm ( mm) for Us/Ua=15 Acceptable for the current mechanical structure Re-design of A-cav Aperture Size: h [mm] 26

28 Wall Loss in Energy-storage Cavity Pwall=150kW for Us/Ua=15 Record: Pwall>200kW in the teststand 27

29 Longitudinal Coupled-bunch Instability driven by the p/2 mode Larger detuning Larger instability driven by the accelerating /2 mode Cured by increasing the energy ratio: Us/Ua f f a π /2 = 1 + Us / Ua Impedance/cav of the p/2 Mode fa = 200kHz (in KEKB) fa = 710 khz (in SuperKEKB) Cf. f rev = 99kHZ We need the feedback? 28

30 Growth Rate(1) SuperKEKB LER ARES 28 sets 10 5 Ua : Us = 1 : 9, Vc = 0.5 MV (S. Yoshimoto) 10 4 Growth rate [1/s] µ = -1 µ = -2 µ =- 3 µ =- 4 Rad. damping rate Beam current [A] 29

31 Growth Rate(2) SuperKEKB LER ARES 28 sets Ua : Us = 1 : 15, Vc = 0.5 MV 10 5 (S. Yoshimoto) 10 4 Growth rate [1/s] µ = -1 µ = -2 µ =- 3 µ =- 4 Rad. damping rate Beam current [A] 30

32 Growth Rate(3) Ua : Us = 1 : 18, Vc = 0.5 MV SuperKEKB LER ARES 28 sets 10 5 (S. Yoshimoto) 10 4 Growth rate [1/s] µ = -1 µ = -2 µ =- 3 µ =- 4 Rad. damping rate Beam current [A] 31

33 Longitudinal Coupled-bunch Instability driven by the accelerating p/2 mode (SuperKEKB LER with 9.4A, ARES 28 sets, Vc=0.5MV/cav) Radiation damped Growth Time Can be suppressed by the upgraded RF feedback system using a comb filter (KEK-PREPRINT-96-79) 32

34 [2] HOM-load Upgrade ARES HOM-damped Structure HOM HOM HOM (Side view) 33

35 SiC Absorbers In the HOM Wave Guide (WG) Direct water cooling Limit: >26kW/cav (HPT) In the Grooved Beam Pipe (GBP) Indirect water cooling via the copper plate Limit: ~3.6kW/cav (HPT) Max. power which can be supplied by the old L-band klystron. 34

36 HOM Extrapolation for Super-KEKB LER HOM in WG load (LER): 26kW/cav (HPT) 80kW/cav (SuperKEKB) Need HPTs over 26kW Increase # of absorbers /WG Enhanced water cooling (KEKB) Nb: 1224 σz: 7mm (Super-KEKB) 4896 (full) 3mm HOM in GBP load (LER): 3.6kW/cav (HPT) 20kW/cav (SuperKEKB) Need direct water cooling like in 35

37 Winged chamber loaded with SiC Absorbers (used in the movable-mask sections) Y. Suetsugu et al., Development of Winged HOM Damper for Movable Mask in KEKB, Proc. PAC2003. Can be a prototype. Directly water-cooled SiC bullet 36

38 New A-cav Design with Winged Chambers Directly water-cooled SiC bullet 37

39 We cannot test HOM loads with high powers for SuperKEKB. More powerful klystron is needed to upgrade the HOM loads. 38

40 D01C/ARES HOM-load Test Stand L-band klystron Freq. = 1.25GHz 1.2 MW (CW) Water dummy load Output from the klystron We would like to thank Fukuda-san and Hirano-san for their courtesy to allow us to use their klystron, RF windows and WG components. (Taken on ) 39

41 The 1 st RF Power Comes! Output power beyond the max. obtained in the old L-band test stand (3.3kW) Tuning to deliver more RF power up to 100 kw 40

42 [2] Input Couplers with TiN Coating The problem is the multipactoring in the coaxial line. Coupling Loop Coaxial Line (WX77D) Example of the TV-camera snapshots of the multipactoring in the coaxial line 41

43 Studies on Total gas pressure Gas mixture ratio (Ar:N2) set Thickness meas. done by direct observation using SEM TiN coating Glass plate Secondary electron yields For the details, see the slides presented in the 6th Workshop on a Higher Luminosity B Factory: 42

44 Two input couplers have been TiN-coated with the final condition. (taken on ) 43

45 After Coating Before coating 44

46 Fabrication Leak test Tested in the upgrade coupler test stand 45

47 Coupler Test Stand Upgraded for Higher Power Capability: kW Input coupler used as output coupler S-cav Input coupler to be tested 1MW DL To 1- MW Water Dummy Load From 1- MW CW Klystron 46

48 Power History (Data taken in TS) To be compared with results on a coupler w/o TiN coating 47

49 [4] New Copper Electroplating for S-cav S-cav is made from Iron with copper electroplating. H. Ino, et. al, "Advanced copper lining for accelerator components", Proc. of LAC2000, Monterey, CALIFORNIA, 1015 (2000) Ex. DTL tank for J-PARC Linac 48

50 Difference between J-PARC and SuperKEKB Studies on Thickness (targets) Ground(alkalinity): ~50µm Main(acidity): ~150µm Electrolytic Polishing: about -40µm Less defect Electric performance Check Q0. 49

51 Pillbox Test Cavity Diameter: 451.2mm Height: 260.0mm Made from iron (SS400) (After copper electroplating) (Before copper electroplating) 50

52 Theoretical Calculation of Q0 (=Q0(cal)) Analytical solution of the electromagnetic field in the pillbox cavity TEmnp mode Q0(m,n,p) = Assuming ω jωµ 0 m j' mn pπ z Er = A J ( )sin sin 2 m r mθ k r b d jωµ 0 j' mn j' mn pπ z Eθ = A J' ( )cos sin 2 m r mθ k b b d E = 0 1 pπ j' mn j' mn pπz Hr = A 2 J' m( r)cosmθ cos k d b b d 1 pπ m j ' pπz H = A J ( r)sinmθ cos mn 2 k d r m b d j ' mn pπ z Hz = AJm( r)cosmθ sin b d 100%IACS electric conductivity (=1/1.72E-8Ωm) flat surface (i.e. no defect) z θ mnp U P w a ll Wall loss power Stored energy P TMmnp mode U ε 2 dv E µ = = cavity 2 dv H cavity wall ωµ = 2 2σ cavity ds H 2 1 pπ j j pπz E = A J' ( r)cosmθ sin r mn mn 2 k d b m b d 1 pπ m j pπz E = A J ( r)sinmθ sin θ mn 2 k d r m b d jmn pπ z Ez = AJm( r)cosmθ cos b d jωε 0 m jmn pπ z Hr = A J ( )sin cos 2 m r mθ k r b d jωε 0 jmn jmn pπ z Hθ = A J' ( )cos cos 2 m r mθ k b b d H = 0 z 51

53 IACS International Annealed Copper Standard 100%IACS electric conductivity: 1/1.72E-8Ωm The electric conductivity of the highest-class oxygenfree copper: 102%IACS Cf. Electroplating in an acid sulfate bath w/o brightener: 102%IACS 52

54 53

55 54

56 After Trial and Error Copper Electroplating in an acid sulfate bath w/o brightener (PR process) Barrel Fine! Endcap 55

57 After Trial and Error Copper Electroplating in an acid sulfate bath w/o brightener (PR process) Barrel Fine! Electrolytic polishing Endcap 56

58 After Trial and Error Copper Electroplating in an acid sulfate bath w/o brightener (PR process) Barrel Fine! Electrolytic polishing Gorgeous! Endcap 57

59 Setup of the Q0 Measurement 58

60 Setup (close view) Loop couples with magnetic field. Antenna couples with electric field. 59

61 Results of the Q0 Measurements Electroplating in a pyrophosphate bath with brightener (applied to the present S-cav s) (no temperature correction) 60

62 Results of the Q0 Measurements DC electric conductivity (102%IACS) Frequency dependence comes from defects on the surface. TM modes TE modes Electroplating in an acid sulfate bath without brightener before electrolytic polishing Electroplating in a pyrophosphate bath with brightener (applied to the present S-cav s) (no temperature correction) 61

63 Results of the Q0 Measurements DC electric conductivity (102%IACS) No defect! Electroplating in an acid sulfate bath without brightener after electrolytic polishing Frequency dependence comes from defects on the surface. TM modes TE modes Electroplating in an acid sulfate bath without brightener before electrolytic polishing Electroplating in a pyrophosphate bath with brightener (applied to the present S-cav s) (no temperature correction) 62

64 Summary SuperKEKB is being proposed. The main strategy is To increase the beam-beam parameter, To make the beam size at IP smaller, To increase the beam currents. Every group is working hard! ARES R&D programs ongoing against higher beam currents For larger detuning, increase the energy ratio: Us/Ua: 9 15 HOM load upgrade with a new test stand. TiN-coated input coupler for the doubled input RF power (800kW, CW) Employing a new highly-pure copper electroplating for S-cav 63

65 (from K.Oide s review 2005) Official Goal Next Milestone Installation of Crab Cavity We are here. 550 /fb will be reached before the Crab Cavity. The next milestone (=1 /ab) will be achieved before Summer

66 (from K.Oide s review 2005) SuperKEKB Installation of Crab Cavity We are here. Shutdown Shutdown for months in for upgrade. 0.6 /ab/month in

67 Fin. 66