The CW and Pulsed SRF Linac Development Activities at KEK 2017/8/9 Topical workshop on high repetition-rate XFEL physics and technology @Shanghai, China Kensei Umemori (KEK) on behalf of KEK-SRF group
Outline • Introduction • SRF activities at KEK • Development of pulse SRF accelerator ➢ ILC ➢ STF • Development of CW SRF accelerator ➢ Compact ERL (cERL) ➢ ERL based EUV-FEL • SRF R&D ➢ Nitrogen doping (for High-Q operation) ➢ Nitrogen infusion (for High-Q and high-gradient operation) ➢ CW SRF gun • Summary
Introduction Merit of Superconducting cavity • High accelerating voltage can be obtained due to very small surface loss. 𝑸 = 𝑾 𝟑 ➢ For example, typical surface losses for 1.3GHz cavity are 𝑺 10W for 10MV and 100W for 33MV (CW) Surface resistance is 6 order smaller than normal conducting cavity (1.3GHz) Demerit of superconducting cavity • He refrigerator is essential to cool SRF cavities and to keep temperature. ➢ Several hundreds times of electric power is needed for operation of He refrigerator to keep 2.0K. Lower heat load (high-Q operation) is important • High-Q (low heat load) ⇒ Reduce surface loss ⇒ Reduce cryogenic load ⇒ Reduce construction and operation cost
Important parameters for SRF One example of measurement for SRF cavity Rs [Ω] Qo 1/T [1/K] Qo =2πf x U / P Eacc[MV/m] U: Stored energy in cavity [J] P: Loss at cavity surface [W] Qo ∝ 1/Rs Rs: Surface resistance[Ω] Eacc : Accelerating gradient [MV/m]
SRF Activity at KEK
SRF R&D Nano-beam R&D cERL CW SRF R&D Cavity fabrication 6th IHEP-KEK SCRF Collaboration Meeting (July 15, 6 IHEP)
ILC / STF
ILC Acc. Design Overview (in TDR) Damping Ring e- Source e+ Main Liinac Physics Detectors Item Parameters e+ Source C.M. Energy 500 GeV e- Main Linac Key Technologies Length 31 km 1.8 x10 34 cm -2 s -1 pre-accelerator Luminosity few GeV source Repetition 5 Hz Nano-beam Technology KeV Beam Pulse Period 0.73 ms damping extraction SRF Accelerating Technology Beam Current 5.8 mA (in pulse) ring & dump few GeV 250-500 GeV final focus few GeV Beam size (y) at FF 5.9 nm SRF Cavity G. 31.5 MV/m IP bunch main linac Q 0 Q 0 = 1x10 10 compressor collimation 6th IHEP-KEK SCRF Collaboration 8 Meeting (July 15, IHEP)
ILC Site Candidate Location in Japan: Kitakami • Preferred site selected by JHEP community, • Endorsed by LCC, in 2013 High-way Oshu Express- Ofunato Rail Kesen-numa Sendai Ichinoseki IP Region 6th IHEP-KEK SCRF Collaboration Meeting (July 15, 9 IHEP)
STF Cryomodules STF-1 (4 cavities) in 2008 S1-Global (4+4 cavities) in 2010 4 cavities per batch Proc. in SRF11 Proc. in SRF09 STF-2 (12 cavities) in 2014~ STF Cavity Package Quantum Beam (2 cavities) in 2012 Capture CM in STF-2 accelerator Proc. in IPAC12 Proc. in IPAC16 15/Jul/2017 6th KEK-IHEP Collaboration Meeting 10
STF-2 Accelerator Layout To be constructed Cold box CM1 + CM2a (Twelve cavities) View from downstream Waveguide system To be constructed Cold box Capture CM (Two cavities) Chicane RF Gun View from upstream 15/Jul/2017 6th KEK-IHEP Collaboration Meeting 11
Summary of Cavity Performance Not measured in 3 rd C.T. Significantly degraded in 3 rd C.T. K t for CAV#1 and #8 was never changed! 15/Jul/2017 12 6th KEK-IHEP Collaboration Meeting
Vector-sum Operation with 8 Cavities Trend graph of Ave. E acc and Δ f Each pulse for 8 cavities during vector-sum operation 35.9 MV/m 35.7 MV/m 31.8 MV/m 28.3 MV/m 30 min 30.6 MV/m 27.3 MV/m 30.1 MV/m 27.8 MV/m 30.9 MV/m ± 0.02 MV/m They satisfy the ILC specification! 46Hz ± 1.8Hz 15/Jul/2017 6th KEK-IHEP Collaboration Meeting 13
Compact ERL
Main linac module Compact ERL HOM damped (for 100mA circulation to suppress HOM-BBU in design) Prototype for 3-GeV class ERL light source 9-cell cavity (ERL-model2)× 2 or high-current CW linac RF frequency: 1.3 GHz 20MeV Recirculation Input power : 20kW CW (SW) (return) loop 15 MV/m （ design ） E acc : Q 0 > 1 10 10 Unloaded-Q: Dump Cryostat Two 9-cell SC cavities HOM Main-linac absorber Merger Tuner Injector-linac Photocathode Target Energy : 35MeV 3MeV DC gun Change 20MeV due Input couplers e- (Not SRF Gun) to field emission Requirement was satisfied at V.T. Heavy F.E Tuner 2-cell Cavity was met @9-10MV/m after string assembly. Injector module 2-cell cavity × 3 Design parameters of the cERL Cryostat Double coupler 35 MeV 20MeV Nominal beam energy RF frequency: 1.3 GHz 5 MeV 2.9MeV Nominal Injector energy Input power : Beam current 10 mA (initial goal) 10kW/coupler (10mA, 5MeV) 180kW/coupler (100mA, 10MeV) 100mA (final) Input Coupler 7.6MV/m （ 5MeV ） E acc : 0.1 ‒ 1 mm mrad Normalized emittance 15MV/m (10MeV) HOM Coupler & e- Q 0 > 1 10 10 RF Feedthrough Unloaded-Q: Bunch length 1-3ps (usual) Requirement was satisfied at V.T and for initial 10mA requirement . (bunch compressed) 100fs (short bunch) 15
Beam current upgrade 2016 Japan Accelerator Society Meeting, Shogo Sakanaka • Beam current increased step by step. • Now 1mA(CW) electron beam is operated with energy recovery mode. Experiment ① Laser Compton scattering Experiment ② THz generation Bright X-ray LCS beam can be generated by Electron bunch was compressed to ~250fs using using 0.9 mA with low emittance beam. sextupole magnet. THz component generated by Imaging was successfully taken. a coherent transition radiation (CTR) monitor is analyzed by a Michelson interferometer. HyPix-3000 from RIGAKU From a rough estimation by CTF spectrum, 250 fs bunch length was achieved by bunch compression SDD THz radiation was successfully generated Apr/2015 An X-ray image of a hornet which was taken using LCS-produced X-ray (2015.Apr)
ML2 ML1 Energy recovery at main linac E=2.9MeV E=2.9 MeV E=19.9 MeV Pin – Pref ～ Ploss + Pbeam Δ(Pin – Pref) ～ Pbeam ← Beam loading ML ML1:Acc ML2: Decc ML2 Energy recovery ML1+ML2 Energy loss measured from the graph = 4 W. (+-4W) Required power without recovery is : ML1 17.14 MV x 900 uA = 15.4 kW Energy Recovery is almost 100.0% (error +-0.03%) Cavity voltage ： ※ different slop of ML1/ML2 come from energy 8.56 MV (ML1), 8.57 MV (ML2) difference of (acceleration – deceleration) beam Current: 0 ～ 900uA
Typical one day operation of cERL SRF Synchronized with 80K line temperature Injector Piezo tuner 55 m 3 /h; ML cavities ON INJ Control voltages for piezo Vacuum pressure in cavities (~ 10 -7 Pa) tuners (for 3 cavities) Piezo FB: Field & Vac ON 34 m 3 /h; ML cavities OFF GV open & loss GV close E acc in cavities (~3.2 MV/m) D f = 2 kHz No data 3.5 hours 12:00 12:00 18:00 18:00 Drift due to temperature change of tuner system Main linac (Stop 2K operation during night time) Vc (ML1&2) 8.57MV Field & Vac 8 hours ML Piezo tuner 8 hours Piezo voltages D f = Vac.(ML1) (~10^-7Pa) 1kHz QL~1*10^7 Vac. (ML2) (~10^-8Pa) Piezo feedback works well 12:00 24:00 12:00 24:00
Parameter Specification Example of high current ERL-FEL Wavelength 13.5 nm Proposal of 10kW class Output power 10 kW ERL based EUV-FEL Bunch chare 60 pC Beam energy 800 MeV Presented by Norio NAKAMURA ERL2015 （ https://www.bnl.gov/erl2015/ ） Accelerating gradient 12.5 MV/m (main linac) 1 st Arc 9-cell cavity × 64 Number of SRF cavity Bunch compressed Beam repetition 162.5 MHz Beam current 9.75 mA Beam Dump Design strategy (main linac) 10MeV, 10mA Epeak/Eacc is 1.5 times reduced from cERL cavity to overcome field emission. 8.6 MV/m 12.5MV/m 10kW FEL output Energy recovery 2 nd Arc Merger Is needed. Injector Linac Gun 10MeV, 10mA decompressed
SRF R&D
High-Q & high-G for pulse linac (ex. ILC) High-Q for CW linac (ex. CW XFEL) 25mTorr N2 25mTorr N2 N-dope N-infusion @120 deg C, 48hours @800 deg C, 2min High-Q & High-Q High-G A. Grassellino 「 High grad/high Q via N TTC meeting （ 2014/Dec ） A. Melnychuk infusion 」 (LCWS2016) 「 Update on N doping at Fermilab 」 N goes to N goes to around several ten nm tens um S. Aderhold / A. Grassellino (TTC@Saclay) A. Romanenco (Linac14)
Results of N-doping at KEK large Without N-dope ↓ furnace (2.7 .7Pa N-dope, , 20min in ) All cases, Q-value became worse after N-doping. ↑ • We tried N-doping at KEK furnaces, but Without N-dope results were not successful. • KEK furnace pumped by a diffusion pump.
N-dope/N-infusion trial using J-PARC furnace • J-PARC has oil-free furnace with cryo-pump(10,000 litter/sec) and three TMPs(3,000 litter/sec x 3). • Vacuum level reached to ~1e-6 Pa. • Normally used for degassing of beam-duct and components.
VT results of N-doping VT4(reference measurement) N-doping 15um EP VT5 Additional 10um EP → VT6 • Magnetic field canceled. (< 1mG) • Cooled down with thermal gradient Quench VT4(reference) Rs =3.3nΩ VT5(N-dope) Rs =1.8nΩ VT6(add EP) Rs =1.2nΩ Very high Q up to high field Q = 2.4e11@11MV/m, 1.4K Q = 3.3e10@14MV/m, 2.0K Quench at 19MV/m No field emission First success in Japan
Recommend
More recommend
