CLIC workshop 2015 CERN 20150126 Recent Xband activities
CLIC workshop 2015, CERN 2015/01/26 Recent X-band activities in Tsinghua University Jiaru Shi*, Hao Zha, Xiaowei Wu, Huaibi Chen (THU) Toshiyasu Higo, Shuji Matsumoto(KEK)
Overview of X-band at Tsinghua • CLIC-study collabration – Start from 2009 – “Choke-mode damped structure”, design, producing, testing, … • NSFC project on “research of physics and key technologies of high-gradient X-band accelerating structures” – 5 -year program starting from 2012 – Focus on RF breakdown study, X-band structure design, X-band structure manufacturing… • ANL collaboration – X-band metallic PETS – X-band structure, C 3 • Industry X-band linacs
outline • Choke-mode structure • T 24_THU_#1 structure • High Power test of T 24_THU_#1 at KEK
New choke design CDS-C 1. 6 mm 2. 0 mm 1. 2 mm • Features include “two-section”, “impedance matching”, “detuning”
Damping result W⊥ (s=0. 15 m) V/m/mm/p. C Freq(G Hz) Q (R/Q) 15. 7 5. 4 80. 3 18. 6 6. 9 63. 2 23. 7 3. 8 14. 8 28. 4 12. 4 10. 9 38. 2 11. 6 9. 9 ⊥ s CDS WDS Fc 1 1 Frms 4 5 Fworst 25 20 WDS 5 CDS-A ~30 CDS-C 4~5
RF measurement • Check the load • Check the reflection of the Choke • Calibrate the connector
Measurement of absorption curve • CDS-A agrees very well • 2 -section choke agrees – (no peak up to 40 GHz, trend OK) • Errors from “multi-short” calibration – espcially low frequency end (<15 GHz)
bunch excitation to measure the damping @ AWA, ANL • Structure can be moved, driven by step motor – Offset to excite dipole modes, pickup RF signal – (No witness bunch at the moment)
Experiment 3 cells with Si. C 3 cells without Si. C RF probe installed
Single-cell SW test structure • Single cell standing wave – 1 C-SW-A 3. 75 -T 2. 6 -CHOKE 1. 6 [Ref: V. Dolgashev] – To test breakdown behavior in the Choke – Also test un-damped to compare
Choke for high power test • RF design and mechanical design • In pipeline
Assembly of T 24 • The assembly of T 24, with cells shipped from CERN, (machined at VDL) • Validate the production capability at Tsinghua while we are still improve the machining. • The procedure that includes etching and cleaning, bonding and brazing, tuning, baking
“infrastructure improvement” • Clean room for RF measurement • Bead-pull setup
“infrastructure improvement” • Temperature control of the furnace
Bonding test and Inspection at CERN • Ultrasound test OK • Crossing grains at bonding plain • Improvement: – Boundary condition – Pressure 0. 1 MPa – Precise Temperature
T 24_THU_#1 fab/test flow Design for CLIC (CERN) Fabrication of parts (VDL) High power test (Nextef-KEK) Cleaning Etching (Tsinghua Univ. ) Bench test (Tsinghua Univ. ) Vac bake (Tsinghua Univ. ) Bonding Brazing (Tsinghua Univ. ) Tuning(Tsinghua Univ. )
The assembly of T 24_THU_#1 (1) • Cleaning/etching • Brazing of the couplers
The assembly of T 24_THU_#1 (2) RF check before bonding 0 11. 1 11. 2 11. 3 11. 4 -5 S 11 (d. B 0 -15 -20 -25 freq (GHz) 11. 5 11. 6
The assembly of T 24_THU_#1 (3) Bonding • Reduced pressure at 0. 1 MPa. (New protocal? )
The assembly of T 24_THU_#1 (4) Brazing step 1 • Au. Cu 75
The assembly of T 24_THU_#1 (5) Brazing step 2 • Au. Cu 65
The assembly of T 24_THU_#1 (5) Tuning and baking • Bench test showed an excellent result after Tuning • Baking is done at 500 degree C, 4 days – Compare to 650 degree C, 10 days @KEK/SLAC/CERN
Installation in Nextef Vacuum problem with ss RF flange. Hand lapping solved the problem Edge before and after hand lapping Question of heat cycle? Photos from Higo @KEK
High power test in Nextef • Operation started from 17/10/2014 • Data is until 19/1/2015 • Reaching 100 MV/m conditioning goal at 252 ns pulse width RF-on Time [hr] Acc-BD 1 st-PBD Gradient [MV/m] 51 ns 751. 8 2989 1285 40 -> 98. 5 91 ns 277. 5 950 917 48 -> 99. 6 132 ns 155. 1 377 426 42 -> 100. 3 173 ns 128. 3 351 529 55. 6 -> 100. 9 213 ns 94. 0 293 497 51. 8 -> 100. 7 252 ns 83. 2 335 619 57. 6 -> 101. 3 Total 1489. 9 5295 4273 —
History plot of T 24_THU_#1 • Gradient in each pulse width operation increases smoothly • BDR almost keeps at 2 e-5 bpp
Normalized Gradient plot • Normalize to 252 ns, 2 e-5 bpp
Compare with other T 24 testing results • Both T 24_THU_#1 and T 24_#3 had done the full conditioning to 252 ns, 100 MV/m • T 24_Dogleg only made the conditioning of 50 ns pulse width • Baking at Tsinghua was of 500 degree C 4 days, compared to 650 degree C 10 days. • Small particles in the environment? Cleaning, handling… • Clean room for cleaning/etching ready Fabrication Assembly/Bonding/Bakin g High Power Testing time T 24_THU_#1 VDL Tsinghua Univ. Nextef (KEK) 2014 T 24_Dogleg VDL CERN Xbox-1 (CERN) 2014 T 24_#3 Morikawa SLAC Nextef (KEK) 2010
Compare with other T 24 testing results (2) • T 24_THU_#1 costs more time to reach 100 MV/m at 51 ns pulse width but having a smaller BDR compared with T 24_Dogleg • T 24_#3 shows an excellent performance with higher ramping speed and less breakdowns Full conditioning BDR~6. 35 e-5 bpp BDR~1. 97 e-5 bpp Full conditioning Only 50 ns conditioning BDR~0. 79 e-5 bpp
Compare with other T 24 testing results (3) • Gradient normalized to 252 ns and 2 e-5 bpp • T 24_#3’s ramping speed is much faster than T 24_THU_#1 and T 24_dogleg • Reason still unknown • Fabrication/bonding/vacuum baking/conditioning program ? . . .
Compare with other structures
BD cell location • Cell location calculation still need to be optimized
BD cell location (2) • More downstream Acc-BDs • No obvious hot spots • Many 1 st. P-BDs/BD clusters RF-on Time [hr] Acc-BD 1 st-PBD All-BD 1489. 9 5295 4273 9568
Dark current measurement • Two Faraday Cups (Up and Mid) • Increased at the beginning and decreased • Three beta measurements until now (2014. 12. 9/2014. 12. 16/2015. 1. 7) • Still preliminary results First beta Second beta measurement Third beta measurement t
91 ns dark current measurement results • Gradient range: 68~100 MV/m, Es=2*Eacc, φ=4. 6 e. V • Lower gradient region has error in dark current’ reading 12. 9 beta_up=35. 0 91 ns beta_mid=32. 9 12. 16 beta_up=38. 1 1. 7 beta_up=36. 1 91 ns beta_mid=39. 1 91 ns beta_mid=38. 1
132 ns and 213 ns measurement results 12. 16 beta_up=42. 1 1. 7 beta_up=36. 4 132 ns beta_mid=42. 1 213 ns beta_mid=41. 8 • Beta at 91 ns pulse width almost keeps the same or increases a bit (Mid Faraday Cup’s data) from the three measurements • Still preliminary results • Need more beta measurements to see how beta changes during operations 91 ns 132 ns 213 ns Elapsed time 12. 9 35. 0/32. 9 — — 1130. 8 12. 16 38. 1/39. 1 42. 1/42. 1 — 1296. 1 1. 7 36. 1/38. 1 — 36. 4/41. 8 1587
Future work • Continue the operation of T 24_THU_#1 to observe the behavior of BDR and Dark current • Analyze 1 st. P-BD/BD cluster in different points of view • More analysis of present data (BD location/missing energy/…) • Compare with former CLIC prototype structures • Try various experiments on T 24_THU_#1
Summary • Choke-mode damped structures – Single cell test structure in pipeline – Full length structure coming after • Several other X-band structures to validate machining/brazing – X-band power extractor @11. 7 GHz – X-band C 3 accelerating structure @11. 7 GHz – X-band linac for industrial application @9. 3 GHz
Summary • T 24_THU_#1 – successfully assembled – Several problems: stainless steel in heat cycle? Water pipes – High power test at Nextef@KEK, unloaded gradient reaches 100 MV/m@252 ns, after 1500 hours running, BDR 2 e-5 bpp. And still testing! – T 24#3 (Morikawa/KEK/SLAC) is better – Discussion: 1) baking 2) careful handling/cleaning 3) Morikawa? – next structures with cells machined in Tsinghua or/and Morikawa
Thank you!
Metallic X-band Power Extractor At 11. 7 GHz for RF power extraction at Argonne Wakefield Accelerator Disk-loaded strucure
A X-band (Copper) Power Extractor for ANL • disks brazing. (no HOM damping) • mode-launcher
C 3 structure at 11. 7 GHz • Design at ANL, 3 regular cells + 2 matching cells • Brazing design
X-band linac 1 Me. V 43
Machined, brazed, and tuned 44
- Slides: 44