1. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas Beam Tests at FLASH John Carwardine (Argonne) April 29, 2009 Slide 1

2. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas Outline Brief overview of the FLASH facilityBrief overview of the FLASH facility 9mA experiment overview9mA experiment overview Experiment results, data analysisExperiment results, data analysis Planning for August/September studiesPlanning for August/September studies Looking ahead to FY10...Looking ahead to FY10... US Resources for the 9mA programUS Resources for the 9mA program Wrap-upWrap-up Slide 2

3. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas Cryomodule String Test Goals The highest priority goal is to demonstrate beam phase and energy stability at nominal currentThe highest priority goal is to demonstrate beam phase and energy stability at nominal current Important goals because of their potential cost impact:Important goals because of their potential cost impact: –demonstrate operation of a nominal section or RF-unit –determine the required power overhead –to measure dark current and x-ray emission –and to check for heating from higher order modes Goals needed to understand linac subsystem performance:Goals needed to understand linac subsystem performance: –develop RF fault recognition and recovery procedures –evaluate cavity quench rates and coupler breakdowns –test component reliability Slide 3 System Integration Tests

4. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas Long-pulse high beam-loading (9mA) demonstrationLong-pulse high beam-loading (9mA) demonstration –800  s pulse with 2400 bunches (3MHz) –3nC per bunch –Beam energy 700 MeV ≤ E beam ≤ 1 GeV Primary goalsPrimary goals –Demonstration of beam energy stability Over extended periodOver extended period –Characterisation of energy stability limitations Operations close to gradient limitsOperations close to gradient limits –Quantification of control overhead Minimum required klystron overhead for LLRF controlMinimum required klystron overhead for LLRF control –HOM absorber studies (cryo-load) –…–…–…–… Major operational challenge for FLASH !Major operational challenge for FLASH ! –Pushes many current operational limits Primary objectives of 9mA program Primarily a LLRF experiment 4

5. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas TTF/FLASH facility overview

6. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas FLASH Accelerator Layout Slide 6 Comparison of machine parameters XFELILCFLASH design 9mA studies NMLProj-XHINS Bunch chargenC13.213 # bunches32502625720024002850 Pulse length ss 650970800 95012501000 Average currentmA599992030 Cavity Loaded-Q3e6 1.25e6 Cavity FrequencyMHz1300 325 GradientMV/m2531.520-3225 35 Cavities / klystron32268-24 241624

7. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas FLASH operations TTF/FLASH is used for different purposesTTF/FLASH is used for different purposes –VUV and soft X-ray FEL photon source –Test bench for accelerator R&D Typical accelerator parameters for photon users:Typical accelerator parameters for photon users: –2-30 bunches (up to 200), ≤ 1nC per bunch, 1MHz bunch rate –400MeV-1GeV 2008 machine time allocations2008 machine time allocations –FEL user operation:161 days –FEL studies:119 days –Accelerator studies: 49 days The 9mA program has received 16 8hr shifts during the last three accelerator studies periodsThe 9mA program has received 16 8hr shifts during the last three accelerator studies periods

8. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas The 9mA Experiment

9. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas The (International) Team FLASH Experts (DESY)FLASH Experts (DESY) –Siggi Schreiber- laser/gun injector set-up –Bart Faartz- general set-up –Lars Froehlich- TPS installation / commissioning, BLM calibration –Florian Loehl- optics matching & emittance –Holger Schlarb- optics & steering –Nina Golubeva- optics calculations –Vladimir Balandin- optics calculations –Valeri Ayvazyan- - LLRF set-up and tuning –Mariusz Grecki- LLRF set-up and tuning –Waldemar Koprek- LLRF set-up and tuning (mostly gun) –Jacek Sekutowicz- HOM absorber measurements –Stefan Simrock- LLRF (general) –Kay Rehlich- controls (DAQ) –Kay Wittenburg- diagnostics –Dirk Noelle- diagnostics (BPM) –Nick Walker- overall coordination –Katya Honkavaara- planning –Mikhail Krasilnikov- RF gun modelling ANLANL –John Carwardine- LLRF / overall coordination –Xiaowei Dong- data analysis, optics modeling FNALFNAL –Brian Chase- LLRF (experiment & data analysis) –Gustavo Cancelo- LLRF (experiment & data analysis) –Michael Davidsaver- DAQ applications programming –Jinhao Ruan- laser setup KEKKEK –Shinichiro Michizono- LLRF (experiment & data analysis) –Toshihiro Matsumoto- LLRF (experiment & data analysis) SLACSLAC –Chris Adolphsen- LLRF (experiment & data analysis) –Tom Himel- Planning & scope –Shilun Pei- LLRF (experiment & data analysis) SACLAYSACLAY –Abdallah Hamdi- TPS installation / commissioning ~40 subscribers to ttf9mA mailing list (not all shown here) 9 RF/LLRF collaborators: DESY, KEK, FNAL, SLAC, ANL

10. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas The 9mA experiment in context: Addressing R&D issues of several projects ILC: GDE stated milestone for Technical Design PhaseILC: GDE stated milestone for Technical Design Phase –Primary motivation and driver for the experiment –Important for evaluating RF overhead and gradient control XFEL: Close collaboration with world-wide LLRF groupsXFEL: Close collaboration with world-wide LLRF groups –Focus (potentially accelerate) development and planning –Critical design/dev’t input from “operation at limits” experience FLASH: Addresses many operational issuesFLASH: Addresses many operational issues –Focus / accelerate upgrades and enhancements –Better characterization of machine –Towards routine high-power long-pulse operation for users NML, Project-X, HINS, STF-IINML, Project-X, HINS, STF-II –LLRF for heavily beam-loaded cavities is a key R&D topic –Important design/dev’t input from “operation at limits” experience –Opportunity to develop and test hardware, software, and algorithms Slide 10

11. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas 9mA experiment chronology First run (May 08)First run (May 08) –Hardware failures (power-out) effectively made shifts unusable –Poor machine setup made bypass optics/steering virtually impossible Second run (September 08)Second run (September 08) –Significant progress on all fronts –Careful set-up of injector (3nC, 1MHz) resulted in ‘loss-free’ transmission to dump (via by-pass) –Vacuum incident resulted in aborted programme Third run (January 09)Third run (January 09) –Beam loss studies –LLRF regulation, beam loading compensation algorithms –Run cavities at higher gradients Upcoming fourth run (August/September 2009)Upcoming fourth run (August/September 2009) –2 weeks of 24/7 dedicated 9mA studies 11

12. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas Results to date compared with the 9mA goals Achieved in Sept 08Goal for Sept 09 Bunch charge to dump 2.5nC @ 1MHz3nC @ 3MHz Bunches/pulse550 @ 1MHz2400 @ 3MHz Beam pulse length550uS800uS Beam power6kW (550x3nC/200mS @ 890MeV) 36kW (2400x3nC/200mS @ 1GeV) Gradient in ACC4-6Ensemble avg: ~19MV/mEnsemble avg: to ~27MV/m Single cavities: to ~32MV/m Also achieved... All RF systems operating routinely with 800us flat tops. Improved characterization of the bypass and dump line optics Characterization of LLRF feed-forward and feedback performance. Collected and analyzed cavity data for RF power overhead study

13. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas Examples of studies by US collaborators Analysis of RF overhead, gradient, stability/jitterAnalysis of RF overhead, gradient, stability/jitter –Shilun Pei, Chris Adolphsen (SLAC) LLRF performance and improvementsLLRF performance and improvements –Gustavo Cancelo, Brian Chase (FNAL) Analysis of beam loss, optics modelingAnalysis of beam loss, optics modeling –Xiaowei Dong, John Carwardine (Argonne) ‘Beta testing’ of upgraded DAQ data archiver and tools‘Beta testing’ of upgraded DAQ data archiver and tools –Michael Davidsaver (FNAL) 13 US collaborators have participated in all the studies shifts

14. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas Energy profile over long bunch trains (Sept 08 studies) Slide 14 Artifacts in energy profile 1.Beam turn-on transient (regulator) 2.Negative slope over flat top (Note: vector sum has a positive slope) 3.Droop at end of the flat top (beam loading compensation not set correctly) 4.Pulse-to-pulse jitter 5.Energy jumps if beam trips off early (1) (2) (3) Planned regulation improvements... Increase feedback gain (only 20 now) Upgrade to latest generation LLRF system (SimconDSP) Better beam loading compensation – ideally set through DOOCS Better feed-forward table generation Beam energy over 550 bunches (many pulses) Beam loss signal at end of bypass G. Cancelo (4) (5)

15. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas Gun dark current losses at 9DUMP vs energy Slide 15 02/18/09 Vector Sum vs time Dark current loss at 9DUMP vs time Dark current loss vs Vector Sum (overlay of many pulses) Bunch-train terminated by MPS Gives an indication of energy acceptance for energies above nominal Why two peaks? Is the BLM in a shadow?

16. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas FLASH cavity gradient limits: Assessing maximum achievable gradients Slide 16 J. Branlard Vector Sum: 27MV/m Avg (C1-4): 32MV/m Max cavity power: 330KW Qexts are set equal: gradients are flat with no beam Alternative: set Qexts to give flat gradients at 9mA (tilts occur at lower currents)

17. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas Preparing for August/September 9mA studies

18. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas Now to Aug 17:Now to Aug 17: –Deliver beam to users + fel studies –No machine studies scheduled Aug 17-Sep 21: 5 weeks of dedicated machine timeAug 17-Sep 21: 5 weeks of dedicated machine time –Tunnel access to repair dump vacuum line (3 weeks) –Two weeks of 24/7 dedicated 9mA beam studies Sept 21: FLASH shutdown beginsSept 21: FLASH shutdown begins Slide 18

19. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas Important operational experience... So far, the main challenges have been operational issuesSo far, the main challenges have been operational issues –Controlling peak + integrated beam loss –LLRF system performance and tuning We need a better model of the bypass line opticsWe need a better model of the bypass line optics A well-tuned and well-matched gun & injector is essentialA well-tuned and well-matched gun & injector is essential LLRF systems require expert attention when increasing beam current or pulse lengthLLRF systems require expert attention when increasing beam current or pulse length –Tuning beam loading compensation –Reduce energy spread over long bunch trains: transients at start and end of pulse; slope over flat-top Slide 19

20. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas Preparatory work prior to August LLRF system upgrades at ACC456 (ACC23)LLRF system upgrades at ACC456 (ACC23) –Upgrade hardware to latest generation (SimconDSP) –Upgrade rf signal down-converters for higher IF –Algorithm improvements: beam loading compensation, feed-forward waveform generation, … –LLRF system modeling, study planning Optics workOptics work –Improve alignment between model and measured lattice –Improve understanding of loss points and apertures –Refine the bypass lattice –Develop a methodical tuning process, online physics model Prepare gun laser and timing system for operation with a 3MHz bunch rate: install pockel cells, timing systemPrepare gun laser and timing system for operation with a 3MHz bunch rate: install pockel cells, timing system DESY DESY FNAL KEK DESY ANL

21. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas Aug/Sept 2009 studies planning Slide 21

22. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas US role FY09/Q3-Q4 Main task is to prepare for 2 weeks of studies in SeptemberMain task is to prepare for 2 weeks of studies in September Participate in LLRF system characterization & upgradesParticipate in LLRF system characterization & upgrades Participate in planning / preparation of the studiesParticipate in planning / preparation of the studies Continue data analysis and optics simulations, co-develop studies plans for tuning for low beam lossContinue data analysis and optics simulations, co-develop studies plans for tuning for low beam loss Participate and lead LLRF and accelerator studies shiftsParticipate and lead LLRF and accelerator studies shifts Studies goals for September are very challenging!Studies goals for September are very challenging! Slide 22

23. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas Looking ahead...

24. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas FLASH long-range schedule Shutdown for FLASH upgrade: Sept 21 – March 09Shutdown for FLASH upgrade: Sept 21 – March 09 Re-commission + machine & FEL studies: ~ 3monthsRe-commission + machine & FEL studies: ~ 3months Restart operation for photon users: Summer 09Restart operation for photon users: Summer 09 User operation continues until end 2011User operation continues until end 2011 Shutdown for FLASH-II upgrade: early 2012Shutdown for FLASH-II upgrade: early 2012

25. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas Possible future studies at FLASH If the machine can be operated reliably with high beam power, then the 9mA program could continue.If the machine can be operated reliably with high beam power, then the 9mA program could continue. –Continue with the major 9mA program topic area –Add: priority 2 an 3 items not covered earlier –Add: new ILC-related studies, eg RTML studies –Add: studies of mutual interest to ILC and XFEL Add studies of mutual interest with NML, Project-X, HINS,...Add studies of mutual interest with NML, Project-X, HINS,... –Opportunity to develop and test LLRF hardware, firmware, algorithms, operational issues,...

26. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas Potential RF System Integration studies Slide 26

27. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas Wrapping up...

28. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas US Resources for 9mA program FY09 resources for LLRF (9mA study is one activity):FY09 resources for LLRF (9mA study is one activity): –LLRF modeling & analysis at APS: 1 FTE –LLRF development at FNAL: 0.92 FTE Additional 9mA program support has come fromAdditional 9mA program support has come from –GDE project management office (Carwardine) –HLRF Technical Area Group (Adolphsen, Pei) The budget allows us to participate in machine studies, perform some data analysis, co-lead the program with N. Walker (DESY)The budget allows us to participate in machine studies, perform some data analysis, co-lead the program with N. Walker (DESY) Slide 28 LLRF resources are insufficient to make substantial technical contributions or to support GDE design/integration activitiesLLRF resources are insufficient to make substantial technical contributions or to support GDE design/integration activities ART Guidance for FY10: expect a flat budgetART Guidance for FY10: expect a flat budget

29. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas Summary TTF/FLASH is the only facility that can address the GDE goal of running beam through a cryomodule string by 2012TTF/FLASH is the only facility that can address the GDE goal of running beam through a cryomodule string by 2012 A relatively small US team is making important contributions on HLRF/LLRF, machine physics studies/modeling, analysis of key operational dataA relatively small US team is making important contributions on HLRF/LLRF, machine physics studies/modeling, analysis of key operational data The team is gaining important experience in SCRF linac ‘operation at the limits’The team is gaining important experience in SCRF linac ‘operation at the limits’ Continued US participation with the 9mA program is important to the GDE global collaboration and it will benefit new and ongoing SCRF projects in the USContinued US participation with the 9mA program is important to the GDE global collaboration and it will benefit new and ongoing SCRF projects in the US Slide 29

30. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas Backups Slide 30

31. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas High Beam-Loading Long Pulse Operation (Sept 08) 450 bunches achieved with stable operation450 bunches achieved with stable operation –Few hours of archived data –Currently under analysis –(vacuum OK) Long bunch trains with ~2.5 nC per bunch:Long bunch trains with ~2.5 nC per bunch: –550 bunches at 1MHz –300 bunches at 500KHz –890 MeV linac energy All modules (RF) running with 800us flat-top and 1GeV total gradientAll modules (RF) running with 800us flat-top and 1GeV total gradient Increase from 450 to 550 bunches eventually caused vacuum incidentIncrease from 450 to 550 bunches eventually caused vacuum incident –The “straw that broke the camels back!” Slide 31 10 MeV over 550 bunches (~1%) (~4 MeV over 1 st 500) 10 MeV over 550 bunches (~1%) (~4 MeV over 1 st 500)

32. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas ACC4-6 RF Power Distribution > 5MW Waveguide component limits DesignDesign –High power part: 5MW max –Low power part: 230kW max LimitsLimits –High power part: ca. 5-7MW –Low power part: ca. 350kW S. Choroba

33. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas FLASH Gradient limits

34. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas

35. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas Beam energy at the dump This is a similar picture but includes energy plot for a total of 2713 pulses.This is a similar picture but includes energy plot for a total of 2713 pulses. –Not all pulses reach 550 bunches, but most reach 480. –For every pulse we see the same two areas where the energy deviates the most from the –For every pulse we see the same two areas where the energy deviates the most from the –We also see an energy increase in the middle of some pulses. MATLAB artwork by John C. Pulse number Bunch number 35

36. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas Correlation of beam loss with cavity vector sum and beam energy With cavity Vector Sum With beam energy Beam not centered in the energy aperture? 36

37. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas BPM9 Stripline BPM15 Button WINDOW with Ti-chamber SWEEPER RD13 (HSK) QUAD Q10 (QC) QUAD Q11 (QC) Chamber with pump port & bellow DUMP Dump Vacuum Section 2879mm STEERER, H7  to bend plane (CV) DIPOLE + Trim D6 (TDC) FLASH beam dump line Leak is believed to be here (caused by thermal cycling of the flange)

38. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas FLASH dump line Slide 38

39. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas Beam losses signatures in dump region during tuning Slide 39 (30 bunches @ 50KHz)

40. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas FLASH cavity gradient limits: Assessing maximum achievable gradients Gradient tilts over 1ms flat-top (9mA beam, all Qext = 3e6) Lorenz-force detuning vs gradient (1mS flat-top) Qext tuning options

41. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas Reflected Ratio for Piezo On/Off Slide 41 Piezo Off with Nominal Initial Detuning Piezo On with Nominal Initial Detuning January Measurement Red → Nominal Initial Detuning with Piezo Off. Black → Nominal Initial Detuning with Piezo On. Piezo works well to reduce the reflection ratio (hopefully will minimize the rf power overhead) but adds some jitter as expected. t A S. Pei

42. ART DOE/NSF Review April 2009: Beam Tests at FLASH (Carwardine) Americas FLASH Upgrade 2009/2010 Slide 42