High power test results of X-band deflecting cavity Jianhao Tan, - PowerPoint PPT Presentation

High power test results of X-band deflecting cavity Jianhao Tan, Wencheng Fang, Qiang Gu, Zhentang Zhao, SINAP Toshiyasu Higo, KEK June. 15, 2017, Valencia

Outline Background Brief introduction of deflector Conditioning history Breakdown position analysis results Summary

SXFEL Test facility linac layout X-band

X-Band TDS layout for SXFEL User Facility Feedback, Feedforward Set point …… D/A A/D LLRF 11424MHz SSA 50MW, 100ns Specification of SXFEL user facility ： Oil tank Energy ： 1500MeV Bunch length ： 76um （ RMS ） load 3 dB splitter Beam size ： 36um Load Install space ： <1.3 m Load 23MW ， 0.5us Resolution ： 20fs Deflecting voltage ： 30MV 0.6m 0.6m Input power: 20MW

Demand of X-Band TDS 1. For now, three RF units of x-band deflecting structures have been confirmed, including 6 deflecting structures, each is 60 cm long. 2. As diagnostic tool and RF kicker, at least 5 units normal conducting x-band deflecting structures will used on HXFEL (hard x-ray free electron lasers), 1 meter long for each structure.

Preliminary design of cell shape Working mode Two caved-ins Non-working mode on cell ID surfaces Working mode Non-working mode Two holes (LOLA Structures)  Scheme 1 ： Cave type  Scheme 2 ： Hole type

Optimization of Regular Cells a (a/b) Group Transvers Quality Attenuati (mm) velocity e shut factor on factor (%c) impedanc (1/m) e (MΩ/m) 5.5(0.37) -2.46 38.35 6662 0.730 5.2(0.35) -2.98 42.24 6622 0.606 5.0(0.34) -3.17 46.04 6778 0.564 4.5(0.30) -3.18 54.09 6924 0.543 4.0(0.26) -2.62 61.53 7072 0.646 Structure type Constant impedance 3.5(0.23) -2.02 68.81 7821 0.757 Operating frequency 11.424GHz t Transverse shut Group Q Attenuatio Operating mode Disk-loaded waveguide impedance(MΩ/m value (m velocity( n 2pi/3 m) ) %c) factor(1/m Total length L 0.3 m ) Resolution 20fs 2.2 Deflecting voltage 10MV 45.62 3.17 6345 0.595 Input power 30 MW 2.0 49.07 3.17 6778 0.557 Group velocity Vg -2.62%c 1.8 50.31 3.16 6965 0.544 Filling time t F 23 ns

Fabrication of cells and couplers Cave With caves on outside Box with Cover Box wall of the cups to fix regular cup the direction. The machining model of couplers, with box and cover, and an regular cup attached to the coupler box.

Field measurement and tuning results The cage, with several metal wires surrounded on the carrier which could be scrip or Teflon, the performance affected by Diameter(D) Length(L) metal diameter(b) and Number(N)

Installation on SHIELD-A Nextef RF source for A SINAP_DEF Layout of Deflector Shield A A

Conditioning history

Operation Plan 4 stages operation Power 50 MW 40 MW 45 MW 35 MW 2 nd 4 th 3 rd 1 st Pulse width 51 ns 91 ns 132 ns 173 ns 4, April, 2016 30, June, 2016 Totally conditioning time: nearly three months, but take the holidays and maintain days out, left 8 weeks to test deflector. (Test of TD24R#05 begin July 2016) The first x-band structure design, fabrication, brazing and test, so we want to test more pulse width, and get more breakdowns information at different width, feed back to the design and technology of fabrication and brazing. Start from 51ns, final target power is 50MW @173ns. Due to the limit of conditioning time, and worried about lots of time will used on higher power and longer pulse width, so the conditioning process operated as we planed. Power up to nearly 45 MW, lots of breakdowns prevent power increase. Lots of problems, not bad things.

Conditioning history For accelerating structures  5 30 E t  a p const BDR For deflecting structure 30 . 𝑢 𝑞5 𝐹 𝑒𝑓𝑔 𝐶𝐸𝑆 = 𝑑𝑝𝑜𝑡𝑢 𝑊 𝑒𝑓𝑔 𝐹 𝑒𝑓𝑔 = 𝑀

RF Conditioning Results @ 51ns Operation start from 14 April @51ns Two weeks running, reach to power 30MW. On the condition of this power level, meet the requirement of bunch length measurement. BDR evaluation continue two days running.

BDR Evaluation Results @ 51ns 33MW 30MW Over 35 hours running, 8 breakdowns record at 30 MW. 1.2*e-6 breakdowns per pulse. 11 breakdowns record during 13 hours at 33MW. 4.7*e-6 breakdowns per pulse.

RF Conditioning Results @ 91ns Operation start from 9 May @ 91ns Power target set 40MW 6~7 days increase power up to 40MW, no breakdown rate evaluation

BDR Evaluation @ 132ns 12 hours 25 breakdowns Threshold program error Run#16 analysis results Power keep at 38MW 78 Acc-BD record over 62 hours, 1.25 breakdowns happen per hour,7.99*e-6 breakdowns per pulse About several hours conditioning at 40MW

BDR Evaluation @ 173ns Run#24 analysis results Power keep at 43MW @173ns 20 breakdowns record during 1215 minutes, about 20 hours BDR= 5.49 *e -6

BDR Evaluation @ 173ns Run#25 analysis results Power keep at 41MW @173ns 25 breakdowns record during 2678 minutes, about 45 hours BDR= 3.11 *e -6

Breakdown Position Analysis Results

BD position Analysis program-Deflector KEK program Instead of deflector parameters 𝐺 𝑨 = 2𝑨 𝑊𝑕 − 𝑀 𝑊𝑕 ---> z= 𝐺 𝑨 𝑊𝑕+𝑀 2 𝐺 𝑨 = ∆𝑆𝑡 − ∆𝑈𝑠 − 𝑈 𝑔𝑗𝑚𝑚 (∆𝑆𝑡−∆𝑈𝑠−𝑈 𝑔𝑗𝑚𝑚 )𝑊𝑕+𝑀 ---> z= 2 (∆𝑆𝑡−∆𝑈𝑠−𝑈 𝑔𝑗𝑚𝑚 )𝑊𝑕+𝑀 ---> z= 2 𝑨 𝑒𝑨 𝑀 𝑒𝑨 𝑔𝑗𝑚𝑚 = 𝑀 𝐺 𝑨 = න − න 𝑈 𝑤 𝑤 𝑕 𝑕 0 𝑨 𝑊𝑕 (∆𝑆𝑡−∆𝑈𝑠)𝑊𝑕 ---> z= 2

Breakdown Position Analysis Results Different pulse width operation have the same trend, all have two “hot spot” in the up and downstream; Divide the structure into three parts, middle parts, upstream parts and downstream parts. In the middle parts, the number of breakdown almost the same each cell. A few of breakdowns out of range of structure, accuracy of the breakdown position from program is not enough, need more analysis. Lots of breakdowns happen in the upstream and downstream of the structure, not reasonable .

Breakdown Position Analysis Results 132ns Running Run#10 Take 132ns running as an example Run#17 Population of breakdown cell as processing ▽  △  ◇  □  ○ Lots of breakdowns in upstream cells Further processing, appeared breakdowns in the downstream Suspected that the breakdowns resulted by leaking of alloy into the structure

Cutoff analysis-cut view Disk Coupler aperture Location of leaking @ first brazing Cutting view of deflector, and made several samples to observe in the optical microscope and SEM; The dirty things in the cavities, they are not caused by breakdown. Liquid with acid to clean the cavity after wire-cutting

Image under optical microscope Image under optical microscope, it is Results under optical microscope obvious that the positions in the coupler upstream and downstream where breakdown frequently are concentrated in the coupler aperture. Damaged heavily More analysis will carry out in scanning electron microscope. Input side Disks from upstream and downstream Output side

Test results to design 1mm The test results show that the frequently breakdown in the coupler prevent the higher power running. Optimization of coupler will used on new deflector design.

Summary The first prototype of x-band deflecting structure has been finished, including design, fabrication, brazing, test, tuning and high power test …… Have found some problems of fabrication and brazing, analysis about BD position, such as cut structure to halves have done and the analysis results basically reflected the problem of design. Optimization of coupler will used on new deflector design.

Thank you!!!

Acknowledgement Many thanks to … T. Higo, S. Matsumoto, T. Abe, A. Karube, A. Asai, A. Kawaba, Y. Arakida, T. Takatomi, P. Matsui, X. Wu… and many other members of KEK contribute to SINAP Deflector test.