Workshop facts and figures

1. Summary of TWIICE(Topical Workshop on Instabilities, Impedances and Collective effects)16-17 January 2014, SOLEIL, SaclayFrance H. Bartosik, E. Koukovini Platia, K. Li, Y. Papaphilippou, S. Persichelli, G. Rumolo and B. Salvant for all CERN participants CLIC seminar 31 January 2014

2. Workshop facts and figures • 74 participants: • 16 from CERN • 10 from SOLEIL • Several representatives from US (Argonne, Brookhaven, Fermilab, Berkeley, SLAC, Cornell) and light sources around the world (Brazil, Switzerland, UK, Sweden, Germany, China, Italy, France) • Large majority of attendants were from the light source community and not so much from damping rings • 35 talks and 5 summaries/discussion sessions in 2 days  very dense program!

3. Disclaimer: This subjective summary only reflects our point of view!

4. Main messages • Beam dynamics teams are very small in light sources  difficult to find a critical mass for deep beam dynamics studies. • G. Rehm: The question is not “Do we need a feedback in our machine?” but “For what reason would you not install a feedback?”. It is off the shelf, not that expensive and could be tremendously helpful in case of issues. • It is important to have feedback systems in our codes, as it changes significantly the beam dynamics: strong request from the light source community for a code that could predict Headtail instability and TMCI thresholds with damper. • Experimental observation of damping of single bunch instability by bunch by bunch feedback (ESRF, Spring8, SOLEIL) • Intricate interplay between impedance heating, outgassing, fast Ion instability and feedback to yield instabilities in SOLEIL • Still difficulties in IBS modeling. Theories can explain what happens in the core, but different physics in the tails. Need for a strong effort on the codes. Can play on the impact factor to fit measurements, but not satisfactory. • Many benchmarks between impedance codes. Very dangerous to blindly trust 3D codes in a large range of frequencies, material parameters and geometries! • Most of our colleagues in low emittance rings use GdfidL and ECHO as their main code for 3D impedance simulations • Methods to identify optimal bunch length below which there is no additional physics information in the simulated wake potential • Several labs are building impedance models of their machine or projects and performing comprehensive collective effects assessment (BAPS in China, MAX IV in Sweden, NSLS II in the US) • CSR instabilitieswere added into tracking codes. Maybe important effect for TLEP and CLIC. • Always interesting to test negative momentum compaction factor.

5. Main open questions • Impact of space charge on TMCI threshold? Alexey Burov predicts that TMCI threshold is relaxed in the presence of strong space charge • Interest in EM properties of NEG at high frequencies and impact on impedance (CLIC DR and ANL). • IBS: Scattering in tails is less evident (Touschek-like effect dominant?) • How to disentangle IBS from other collective effects? • Amor Nadji: “In future ring-based light sources (DLSR), the combination of ultra low emittance, high current and narrow chambers will mean that collective effects will be important!”

6. Proposed actions for us • Need to gain expertise on CSR instability and include it in tracking codes (Headtail). Assess if issue for our future lepton colliders • Light sources look like an ideal testbed for our beam dynamics codes (TMCI with damper in particular, NHT and DELPHI). • See the applicability of the method of B. Podobedov to find the optimum bunch length for 3D simulations of our devices. • New possible collaborations: NEG coating RF measurements with ANL, 3D simulations with ESRF. • Need to gain more expertise in GdfidL and ACE3P and to discuss more with other labs who have similar issues.

7. Parameter range for various machines

8. Impedance and instabilities

9. Alexej Burov (FNAL) • Very important to add transverse damper to assess beam stability • Stability diagrams are very asymmetric for low emittance rings (due to x >> y ) and can be shifted due to longitudinal to transverse Landau damping

10. Na Wang (IHEP Beijing) • Impedance model and associated collective effects for the new Beijing Advanced Photon Light source project. • Impedance model contains resistive wall, RF cavities, undulatortapers and wiggler tapers • Comprehensive evaluation of limits from collective effects (e.g. microwave instability, bunch lengthening, TMCI, transverse coupled bunch instability, fast beam ion instability)

11. Thomas Perron (ESRF) • Working on ESRF upgrade to reduce horizontal emittance • Lower momentum compaction factor  lower Qs  more sensitive to instabilities • Asked for help to simulate equipment and build impedance model • Designed HOM-free cavity • Factor 3 to 4 increase of TMCI threshold thanks to using bunch by bunch damper

12. Eirini (CERN/EPFL) • Measurement of NEG conductivity with waveguide method • Very important input for resistive wall impedance and instability thresholds, as skin depth is very small in the frequency range excited by the CLIC bunch. • More at the CLIC workshop next week!

13. B. Podobedov (BNL) • The hunt for the wake function, as cannot simulate point charge. • Problems of applicability found with scaling law of Stupakov, Bane, and Zagorodnov (PRSTAB 2011) • Exposed method to find the largest exciting bunch length gfor which the wake potential would contain all the physics needed to reconstruct the short range wake function (PRSTAB 2013). “This g is easy to find for arbitrary geometries.” • Beware of reducing the bunch length until the wake converges!

14. Henrique de Oliveira Caiafa Duarte (LNLS, Brazil) • Benchmark of CST, GdfidL, ACE3P and ECHO for shallow cavities and collimators. • Comparison between simulations and CST for longitudinal kicker • GdfidL and ACE3P are less prone to dispersion issues than CST

15. Slide of Henrique from yesterday

16. Question of Giovanni Rumolo yesterday to Henrique: So there is no way to improve the convergence of CST for this collimator structure?

17. New CST simulationperformed this night  We can do better with CST, but I agree one has to be very careful !

18. T. Gunzel • Heat load of the CLIC-striplineusing GdfidL • Computations performed mode by mode and no coherent heating assumed (Ploss M*Nb2 and not (M*Nb)2 ) • If feedthrough ideal, most of the power leaves the stripline through the ports • “In terms of heat load the design of the CLIC stripline is perfect.” G. Rehm: “ beware of non ideal feedthroughs.”  similar studies at Diamond (heating workshop in Oxford in 2013)

19. O. Frasciello (LNF/INFN) • Impact of geometric impedance of LHC collimators on LHC transverse impedance model  additional ~20% on tune shift

20. G. Skripka (MAX IV, Sweden) • Short range wake with method of B. Podobedov (computing optimum exciting bunch length)  successful also for periodic cavity • Impedance model of MAX IV (cavities, BPMs, flanges, tapers)

21. A. Blednykh (BNL) • Impedance of insertion devices • Available Resources for Impedance Simulations: 42 nodes with 8 Cores per node and 1334GB Total Memory (RAM) • Benchmark between GdfidL, CST and ECHO, CST performing quite bad and showing unphysical drifts. • Benchmark between GdfidL and Stupakov’s dipolar and quadrupolar impedance equations for rectangular smooth tapers

22. C. Belver-Aguilar (IFIC) • Study of kicker for CLIC damping ring: stripline design proposed and prototype built and will be installed in ALBA for tests • Good agreement between analytical calculations and simulations • Bench measurements with wire under way at CERN

23. J. Byrd (LBNL) • Playing with higher harmonic cavities to control bunch shape and landau damping • Current study of ALS-II, welcoming collaborations for higher harmonic cavity design and beam dynamics requirements • Outstanding questions: • Effect on single bunch instabilities? • Effect of Landau damping on HT instabilities? • Effect of overstretching (multiple longitudinal fixed points)?

24. Marit Klein (MAX IV) • transverse beam instabilities in MAX IV using the multibunch code mbtrack • Use of wakes from GdfidL and harmonic cavities

25. Yong-ChulChae (ANL) • Off-line request for NEG coated pipe impedance measurements • Optimizing the 3D taper to reduce the vertical impedance (only, and in fact only Ky is it sufficient to check only the dependence with Ky?) • Simulations with GdfidL • New chamber’s kick factor should be < 60% of old one, despite the smaller aperture.

26. N. Biancacci (CERN/Sapienza) • Localization of impedance sources using intensity dependent phase advance measured by the BPM system around the machine • Learnt important measurement constraints: High performance BPM system, need for high ratio between local machine impedance and measurement noise, High quality optic model. • Successful application to the PS machine

27. P. Brunelle (SOLEIL) • Impact of incoherent tune shift on orbit. • Incoherent tune shift due to low vertical aperture (vacuum chamber and insertions) • Since undulator position is changing, impact on tunes  feedback implemented to keep the tunes fixed

28. Particle scattering • See also summary of Yannis at the workshop!

29. Theo Demma (LAL)

30. Fanouria Hannes Antoniou Bartosik(CERN)

31. S. Wang (CESR-TA) • Model and data agree well for “weak IBS” regime

32. Two stream instabilities

33. Summary of the two-stream instability session G. Rumolo, R. Cimino Based on input from the presentations of G. Iadarola, H. Bartosik, R. Nagaoka, N. Wang, T. Perron

34. Positron machines Electron machines • Primary electrons (mainly photoemission) • Acceleration and secondary electron production • Ions generation (mainly gas ionization) • Acceleration and trapping • Multi-bunch electron cloud build up • Detrimental effects • Mitigation/suppression needed • Multi-bunch accumulation • Beam instability • Very good vacuum and vacuum composition needed

35. CLIC e+ damping ring C = 427.5 m E-cloud aspects have been investigated in three families of devices Wiggler a=40mm, b=6mm Ltot = 104 m Quadrupole a=9mm, b=9mm Ltot = 86 m Dipole a=40mm, b=9mm Ltot = 58 m

36. ELECTRON CLOUD BUILD UP s Positron bunch train Primary/secondary electron production(PEY, SEY) PyECLOUD x y ** This process is only slightly dependent on the beam transverse emittance

37. Challenging simulation scenario • Short bunches  Short time step • Small emittance  Beam size 104 smaller than chamber size • In the cases of wigglers and dipoles e- in a narrow stripe close to the beam  Fine grid needed for Poisson solver

38. Thresholds and saturation values lower for 0.5 ns • Large e- densities (>1e13) at the beam location • E- in narrow stripe in wigglers/dipoles, around the quadrupole field lines in quads. • Local low SEY coating or clearing electrode for full e-cloud suppression in all cases possible

39. ELECTRON CLOUD DRIVEN SINGLE BUNCH INSTABILITY s Positron bunch train Equations of motion of the beam particles PyECLOUD HEADTAIL x y ** This process is strongly dependent on the beam transverse emittance

40. Beam becomes unstable (few turns rise time) as soon as electron build-up reaches saturation in wigglers • Chromaticity does not help • Consistent with threshold density found with uniform electron distributions (1.3 x 1013 m-3)

41. Tolerate e-cloud but damp the instability: feedback system MITIGATION/SUPPRESSION TECHNIQUES Clearing electrodes installed along the vacuum chambers (only local, impedance) Solenoids (only applicable in field-free regions) • Machine scrubbing during operation • Limited by reachable SEY • Depends on e- energy • Relies on surface graphitization Possible Solutions • Applying on the wall thin films with intrinsically low SEY • NEG coating (helps vacuum) • C coating (no activation) • Surface roughness to stop secondary electrons • Grooves • Rough material coating • Sponges

42. MITIGATION/SUPPRESSION TECHNIQUES  SPONGES Impedance impact, vacuum behaviour, desorption properties are still under study  seems very promising

43. Positron machines Electron machines • Primary electrons (mainly photoemission) • Acceleration and secondary electron production • Ions generation (mainly gas ionization) • Acceleration and trapping • Multi-bunch electron cloud build up • Detrimental effects • Mitigation/suppression needed • Multi-bunch accumulation • Beam instability • Very good vacuum and vacuum composition needed

44. Mainly estimations based on analytical formulae for trapping condition and instability rise time Applied to Beijing Advanced Photon Source (BAPS) and ESRF upgrade Detailed simulations foreseen, possibly including a transverse damper

45. The CLIC Main Linac H2O peak Vacuum specifications for CLIC long transfer line, Main Linacand BDS made with strong-strong multi-species FASTION code Different vacuum compositions investigated, NEG & baked vacuum most favorable

46. Observations Ion instabilities observed in APS (with additional He injection), PLS (with additional H2 injection), SOLEIL, BESSY II, ELETTRA, ALBA Fast beam ion instability observed in electron rings During commissioning/start up (chamber not yet conditioned, bad vacuum, feedback system not yet operational) Because of some local pressure rise (e.g., directly connected to impedance induced heating) Artificially induced by injecting gas into the vacuum chamber and raising the pressure by more than one order of magnitude (for studies) Usually less severe than predictions, stabilizing effects not included in existing models ? Quantitative comparison between theoretical predictions, simulations and measurements yet to be made Experiment planned at Cesr-TA (April 2014)

47. Observations (II) • Ions enhanced by local heating (outgassing) seem to trigger some recently observed high current instabilities @ SSRF and SOLEIL

48. Observations (III) • mbtrack simulations suggest that SOLEIL instability results from an intriguing interplay between resistive wall, ion effect and transverse feedback

49. Wrapup Two-stream effects often affect the performance of running accelerators and can be a serious limitation for future low emittance rings Electron cloud formation and instabilities in CLIC DRs Studies carried out with detailed modeling Electron cloud in wigglers not acceptable for beam stability Promising ongoing research on mitigation or suppression techniques (C and sponge coating, scrubbing mechanism) Ion accumulation and instabilities Mainly analytical formulae used for future machine design, detailed simulations needed Observations in running machines usually made in presence of vacuum degradation and with high intensity  important interplay between several effects (RW, FII, damper) observed Beam-induced outgassing enhanced for machines with low-gap chambers and high intensity short bunches,  FI effects possibly more serious for future low emittance light sources

50. Instrumentation and feedback