Review of the European XFEL Bunch Compression System Summary Torsten Limberg

Topics and Speakers Introduction and ConceptT. Limberg Optic and Tolerances W. Decking Simulation CalculationsM. Dohlus Tuning T. Limberg Bunch Compression Options M. Dohlus Diagnostic Overview & FB H. Schlarb Diagnostic Sections Lay Out C. Gerth Diagnostic Tools and Optical Replica B. Schmidt & M. Yurkov Vacuum N. Mildner, T. Wohlenberg, K. Zapfe

Design Goals and Considerations Electron bunches out of the gun: 50 A peak current, small energy spread BC system has to convert that to: –5 kA peak current –< 25  m Bunch Length (shorter pulses?) –< 1.4 mm-mrad slice emittance –< 1 MeV slice energy spread (stay about a factor of two below that from synchrotron radiation in undulator) –Compensate rf structure wake field induced correlated energy spread as good as possible with rf induced energy chirp for compression (mimimize laser bandwidth) –avoid high gain for micro-bunch instability –avoid big projected emittance (> 2.5 mm-mrad) –< 10% peak current jitter (SASE jitter <10 %) –arrival time jitter has mainly to be measured and taken care of by the experiments

Bunch Compression Scheme (TADR)

Bunch Compressor Beam Line Optics Diagnostic Section Drift through shielding Dogleg (R 56 ≈ m) 18 deg deflection to commissioning dump

W. Decking: To Do List Optics and Tolerances Include BC Diagnostic Sections in Master Deck Increase BC chicane middle dipoles distance to include diagnostics Calculate transverse wakefield effects of 3 rd harmonic cavities Adjust phase advance between BC1 and BC2 to n*pi Magnet tolerance studies (field quality and alignment of dipoles)

‘laser heater’ (LCLS layout) rms = 2 keV (Gaussian) rms = 10 keV (Gaussian) rms = 10 keV (from laser heater) slice energy distribution P(  E) M. Dohlus: Simulation Calculations

gain curves “real” heater: rms = 10keV after BC1 after BC2 after dogleg dogleg, r56 = 0.84mm  0 TDR: gaussian distribution rms = 10 keV shot noise TDR: “real” heater:

energy to current modulation: “real” heater: rms = 10keV TDR: gaussian distribution rms = 10 keV after BC1 after BC2 after dogleg ASTRA simulation: 5% modulation at cathode, = 0.2 mm  injector dogleg (~45m after cathode): ~2keV energy / eV 5% current / A s / mm 0.5% cathode after 45m (130 MeV)

Setup Using ‘Multiknobs’ Make knobs to change independently the first, second and third derivative of the combined accelerating voltage of Injector Linac and 3 rd harmonic RF, using linac and 3 rd harmonic phase and 3 rd harmonic amplitude. –V(s)= V 1 cos(k 1 s+  1 ) + V 3 cos(k 3 s +  3 ) =  V + g ∙ s + x 1 ∙ ∙s 2 + x 2 ∙ ∙s 3 + o(s 4 ) Use gradient knob for peak current, 2 nd derivative to balance beam distribution in the center region and 3 rd derivative knob for adjusting the tails. Linac Amplitude is still used to keep beam energy constant.

Things to Do Practical design of multi-knobs for FLASH Prepare detailed tuning scheme for FLASH Test it and learn…

BC System – Review Options ● BC2 working point (energy-charge-compr.) ● 2BC (rf-rf-bc-rf-bc-rf) ● table: 2BC (rf-rf-bc-rf-bc-rf) dogleg + 2BC (rf-dog-rf-rf-bc-rf-bc-rf) n3BC (rf-bc-rf-rf-bc-rf-bc-rf) 3BC (rf-rf-bc-rf-rf-bc-rf-bc-rf) rollover compression ● laser heater ● cases in detail peak current projected emittance slice emittance uncorrelated energy spread remaining chirp µ-bunch stability parameter-sensitivity arrival time stability

M. Dohlus bc system optimization sheet rf knobs r56 knobs compression factors absolute tolerances (amplitude & phase_deg) µ-bunching gain chirp minimal relative tolerances shot noise due to µ-bunching gain

inverse tolerance noise: I rms /A Balancing the micro-bunch instability strength vs. the rf jitter sensitivity inverse tolerance noise I rms /A

…continued 20,500 10,500 10,400 20,400 5,400 C1, E1/MeV inverse tolerance nois e I rms /A E1 = 400 MeV r56BC1 = 90mm, C1=5 r56BC2 = 75mm, C2=20  L2 = 10 deg min(ampl_tol) = 0.1% min(phas_tol) = deg noise: I rms = 147 A min(phas_tol) = deg noise: I rms = 260 A

2BC rf (1+3) -bc-rf-bc-rf-c rollover compr. rf (1+3) -bc-rf-bc-rf-c dogleg+2BC rf-d- rf (1+3) -bc-rf-bc-rf-c n3BC rf-bc- rf (1+3) -bc-rf-bc-rf-c 3BC rf (1+3) -bc- rf (1+3) -bc-rf-bc-rf-c E=400MeV 2GeV 17.5GeV C= r56=-90mm -75mm 0.84mm ampl_tol=0.1% ph_tol=0.023deg noise= 147 A  L2 = 10 deg E=500MeV 2GeV 17.5GeV C= r56=-100mm -200mm 0.84mm ampl_tol=0.2% ph_tol=0.055deg small  L2 = 40.5 deg E=130MeV 400MeV 2GeV 17.5GeV C= r56=40mm -90mm -87.2mm 0.84mm ampl_tol=0.11% ph_tol=0.040deg noise= 270 A 130MeV  L2 = 10 deg t566_dog=1m E=130MeV 500MeV 2GeV 17.5GeV C= r56=-30mm -90mm -45.0mm 0.84mm ampl_tol=0.09% ph_tol=0.048deg noise= 95 A 130MeV  L2 = 10 deg E=130MeV 400MeV 2GeV 17.5GeV C= r56=30mm -80mm -83.7mm 0.84mm ampl_tol=0.11% ph_tol=0.045deg noise= 93 A 130MeV  L2 = 10 deg t566_dog=1m

Diagnostics overview BC1 proposed beam line design: SRF 1.3GHz SRF 1.3 GHz Bunch compressor TDS X&Y Diagnostic section SRF 3.9 GHz Spectrometer Dump Standard diagnostics: TORtoroid system for transmission measurements (1,3&4 for interlock) DCdark current monitors (upstream BC1, downstream BC1) BPMbeam position monitor ~ 20 (not yet determined … every quad?) purpose: orbit correction, transfer measurements, dispersion correction OTRoptical transition screen (with wire scanners WS?)

Diagnostics overview BC1 proposed beam line design: SRF 1.3GHz SRF 1.3 GHz Bunch compressor TDS X&Y Diagnostic section SRF 3.9 GHz Spectrometer Dump Special diagnostics: TDStransverse deflecting structure X & Y EOelectro-optic longitudinal beam profile monitor BCMbunch compression monitors (CSR at D4 and CDR/CTR) SRsynchrotron radiation monitor (energy and energy spread) BAMbeam arrival time monitor -> B Schmidt

Diagnostics overview BC1 proposed beam line design: SRF 1.3GHz SRF 1.3 GHz Bunch compressor TDS X&Y Diagnostic section SRF 3.9 GHz Spectrometer Dump Additional devices: COLcollimators (1 st & 2 nd to remove dark current, 3 nd & 4 th for kicked e - ) KICfast kicker to off-axis screens (2 x and 2 y) Alignlaser for optics alignment BLMbeam loss monitors (about 8-10 sufficient)

Horizontal kicker Vertical kicker FODO lattice 6 off-axis OTR screens (y and x) 3 cells = 11.4 m OTR1OTR3OTR5OTR2OTR4OTR6 VK2 HK2 VK1 Horizontal slice emittance / vertical streak Vertical slice emittance / horizontal streak HK1 Screen / Kicker arrangement (2) 45deg 76deg HK1OTR1 OTR1 HK1OTR2 OTR3 HK2OTR4 OTR4 HK2OTR6 OTR6 45deg 76deg VK1OTR1 OTR2 VK1OTR2 OTR3 VK2OTR4 OTR4 VK2OTR6 OTR5 Bend plane of BCs defines the OTR arrangement

VK1 HK1 VK2 HK2 CSR T1 T2 EOSD BAM ABCMTDS-xTDS-y ABCM 2.5m Booster Linac FODO lattice SR Alignment laser RES 10 modules 3.8 m 5 modules 7.6 m Lattice can be divided into modules: Diagnostic Section Engineering layout (3)

Conclusions (1): Conclusions For which bunch rep rate, 5MHz or 1MHz, shall the on-line slice emittance diagnostics be designed in BC1: Desired resolution can easily be reached at 1 MHz but is just at the theoretical limit for 5 MHz. Kickers with the required kick strength for 1MHz are in operation in several machines at DESY (‘off-the-shelf’). 5 MHz would requires new design and prototype development. If standard FEL operation will be 5 MHz slice emittance diagnostics cannot be operated parasitically if designed for 1 MHz (or might not be used if resolution is not sufficient). If standard FEL operation will be 1 MHz one would lose at least a factor of 1.6 in resolution if designed for 5 MHz

Conclusions (2): Conclusions Dump defines the horizontal streak direction in BC2. If the BCs are installed vertically slice emittance could be measured in the bend plane of BCs. Number of quads in current layout BC1 was 22 now 22 BC2 wsa 13 now 19 Layout of the dignostics sections can be arranged in modules. Components can be prealigned and tested. This saves time during installation and commissioning. Layout of BC1 diagnostic section almost finalized. After beam dynamic and sensitivity studies (2 months) the vacuum and engineering layout could be started New lattice layout requires slightly more space BC1: 1.5 m in BC m in diag section = 2.4 m BC2: 1.0 m in BC m in diag section* = 2.5 m * Additional FODO cell for 45 deg lattice requires 7.6 m more space

Coherent radiation Status : - spectrally resolving single shot instrument developed (multi stage grating spectrograph with parallel read out) -Advanced prototype running at FLASH (THz beam-line) - Existence of spectroscopic fingerprints shown down to µm scale To be done : - develop compact monlithic version - explore and establish feedback capabilities - detailed planning of station lay-out existing detector unit Potential layout for 4-stage spectrograph

Electro-optical monitors Status : - different methods under study at FLASH - integrity and validity of data largely explored - spectral decoding method proven to be sufficiently simple - dedicated fiber-laser version under construction To be done : - step from ‘experiment’ to ‘on-line tool’ - more robust and reliable laser system (fiber-laser) - fast (parallel) read-out system (line camera) - direct (optical) coupling to optical timing system

Requirements / implications : - EO crystals inside beam pipe (r ~ 2-5 mm), retractable - optical ports for laser in/out space underneath beam pipe : ~ 2 m 2 optical table (laser +spectrometer + camera). beam laser ~0.6 m

T. Wohlenberg: Bunch compressor section BC1 and BC2 General remarks Lengths of the vacuum system BC1 and BC2: BC1: total length ~ 69m → chicane length ~ 27m → deflection of the chicane ~ 0.68m BC2: total length ~ 90m → chicane length ~ 25m → deflection of the chicane ~ 0.33m Vacuum requirements:  Pressure needs to be in the range of mbar (next to cold sections)  Pump system: sputter ion pumps and titan sublimations pumps Both sections are particle free :  The design of all vacuum components needs to be according to the particle free conditions. Early discussion of concept of all components including beam diagnostic is necessary!  All vacuum components have to be cleaned under particle free condition (clean room).  Installations needs to be done under local clean room conditions.

Bunch compressor section BC1 and BC2 General remarks  From the point of view of vacuum technology both BC sections should be  treated similar. This should be valid for the aspect of material choice,  joining technology, support for the chambers etc..  The design concept for the flat chamber in the chicane is similar to FLASH!

Bunch compressor section BC1 and BC2 Schedule Draft:  Components layout + girder and frames concept including electronics/diagnostics units concept ~ 1 year  Design of BC1 and BC2 ~ 1 year  Fabrication of all components ~ 1.5 years  2007, A rough concept should be settled for the girders/frames concept including electronics and diagnostics as well as part of the layout of the components.→ layout for the arrangement of the components should be available!  2008, The detailed concept for the layout of the components, electronic concept and the girder and frames concept should be finished.

Bunch compressor section BC1 and BC2 open issues  Do we have the BC‘s chicane to be installed vertically or horizontally? → we prefer vertical installation!  Do all components need to be copper coated in both BC’s?  Can the RF-shielding remain the same as for FLASH or do we have to design a new concept for the flange connections, bellows, valves and pump connections?  Is a massive lead shielding necessary ? → need to be included into the girder and frame design!  How does the dump section for BC1 and BC2 look like?  What diagnostic installations will be needed next to the beam line?