1. Two-beam instabilities in low emittance rings Lotta Mether, G.Rumolo, G.Iadarola, H.Bartosik Low Emittance Rings Workshop INFN-LNF, Frascati September 17 th, 2014

2. Outline o Introduction Two-beam instabilities The CLIC Damping rings o Electron cloud in positron machines Electron cloud formation & instability Observations & modelling Low-ε example: CLIC-DR o Ion effects in electron machines Ion trapping & fast beam ion instability Observations & modelling Effect of emittance 17 September 2014 L. Mether, Two-beam instabilities2

3. Two-beam instabilities o Collective effects induced by electromagnetic field of second charge distribution, in addition to primary beam Beam-beam effects in colliders Electron cloud in positron machines Trapped ions in electron machines o Second “beam” typically produced by primary beam Synchrotron radiation on wall: electrons, ions Residual gas ionization: electrons, ions Desorption from wall due to losses: electrons, ions, neutrals o Cause instabilities, tune shift, beam blow-up and emittance growth, beam and energy losses 17 September 2014 L. Mether, Two-beam instabilities3

4. CLIC Damping Rings 17 September 2014 L. Mether, Two-beam instabilities4 DescriptionSymbolValue Beam energyE 0 [GeV]2.86 Normalized transverse equilibrium emittancesε n,x,y [nm]500, 5 Bunch populationN [10 9 ]4.1 Bunches per trainnbnb 312 / 156 Bunch spacingτbτb 0.5 / 1 Bunch length (rms)σ z [mm]1.6 / 1.8 Synchrotron tuneQsQs 6 x 10 -3 C = 427.5 m Q x =48.38 Q y =10.39

5. Electron cloud formation 1. Primary (seed) electrons are generated inside beam chamber 2. Seed electrons are accelerated by beam field, and produce secondary electrons when hitting the wall  avalanche electron production (multipacting)  exponential growth of electron density 17 September 2014 L. Mether, Two-beam instabilities5

6. Electron cloud formation 1. Primary (seed) electrons are generated inside beam chamber 2. Seed electrons are accelerated by beam field, and produces secondary electrons when hitting the wall 3. Eventually further growth is limited by space charge a stationary state, the electron cloud, is reached 17 September 2014 L. Mether, Two-beam instabilities6 CLIC-DR wiggler CLIC-DR quad

7. Electron cloud formation 1. Primary (seed) electrons are generated inside beam chamber 2. Seed electrons are accelerated by beam field, and produces secondary electrons when hitting the wall 3. Eventually further growth is limited by space charge a stationary state, the electron cloud, is reached 17 September 2014 L. Mether, Two-beam instabilities7 CLIC-DR wiggler CLIC-DR quad  Coherent beam instabilities Single bunch Coupled bunch

8. Electron cloud observations o Build-up of electron cloud can be seen as Beam degradation Fast pressure rise, outgassing Additional heat load o Observed in several machines KEK-LER, Da  ne, CesrTA (see next presentation)… 17 September 2014 L. Mether, Two-beam instabilities8 Vertical beam size blow-up at KEK-LER, seen through streak camera From K. Ohmi, K. Oide, F. Zimmermann, et al.

9. o Frequent observations and well understood mechanism  Reliable models both for build-up and instability, for future estimates o Low ε  computational challenges, resolve small beam in large chamber e.g. CLIC-DR: aperture ≈ 10000 x σ beam o Low-ε ring example: CLIC Damping Ring, wigglers Electron cloud modelling 17 September 2014 L. Mether, Two-beam instabilities9

10. Low-ε example: CLIC-DR, Wigglers Electron cloud builds up for SEY > 1.4 Threshold e - density for instability wiggler e - density ~ 1.2 x 10 13 / m 3 Emittance growth rate fast compared to damping times ~ 2 ms 17 September 2014 L. Mether, Two-beam instabilities10 with PyECLOUD rise time τ ≈ 0.7 ms τ ≈ 0.5 ms τ ≈ 0.4 ms Vertical emittance with (Py)HEADTAIL

11. Maximum central density along train Low-ε example: CLIC-DR, Wigglers o Instability simulations with initial densities given by PyECLOUD Build-up mildly dependent on ε, instability strongly dependent  Beam is unstable for all SEY values above build-up threshold! 17 September 2014 L. Mether, Two-beam instabilities11 Electron cloud suppression crucial!

12. Outline o Introduction o Electron cloud in positron machines Electron cloud formation & instability Observations & modelling Low-ε example: CLIC-DR o Ion effects in electron machines Ion trapping & fast beam ion instability Observations & modelling Effect of emittance 17 September 2014 L. Mether, Two-beam instabilities12

13. Ion trapping and instability 1. Generation of ions inside beam chamber 2. Ions are accelerated by beam field, and possibly trapped, depending on ion mass 17 September 2014 L. Mether, Two-beam instabilities13 CO, N 2 H2OH2O H2H2

14. Ion trapping and instability 1. Generation of ions inside beam chamber 2. Ions are accelerated by beam field, and possibly trapped, depending on ion mass 3. Ions accumulate along bunch train, coupling head and tail of train Offset of each bunch is recorded into generated ion distribution, and transferred to the following bunches 17 September 2014 L. Mether, Two-beam instabilities14

15. Ion trapping and instability 1. Generation of ions inside beam chamber 2. Ions are accelerated by beam field, and possibly trapped, depending on ion mass 3. Ions accumulate along bunch train, coupling head and tail of train Offset of each bunch is recorded into generated ion distribution, and transferred to the following bunches  coupled oscillations between electrons and ions  multi-bunch “fast beam ion instability” 17 September 2014 L. Mether, Two-beam instabilities15

16. Fast beam ion instability observations o Rarely a problem in current machines under regular operation Observed in several machines under (deliberate) vacuum degradation APS, PLS, SOLEIL, SSRF, BESSY II, ELETTRA, ALBA … o Measurements at CesrTA (Dec. 2013, Apr 2014) Varying ion species, pressure, bunch charge, train structure, feedback etc. 17 September 2014 L. Mether, Two-beam instabilities16

17. Fast beam ion instability observations o Measurements at CesrTA (Dec. 2013) Vertical beam offset and size as function of bunch number, at varying pressure, with vertical feedback on and off. 17 September 2014 L. Mether, Two-beam instabilities17 From A. Chatterjee et al. IPAC 2014 0 5 10 15 20 25 Bunch #

18. o Analytical model available, with estimates for e.g. Characteristic ion oscillation frequency, coupled to beam oscillation frequency Instability rise time Fast beam ion instability models 17 September 2014 L. Mether, Two-beam instabilities18 Raubenheimer et al. Phys. Rev. E 52, 5, 5487, Stupakov et al. Phys. Rev. E 52, 5, 5499 Lower ε = faster instability

19. o Analytical model often used for estimates, but has limited applicability Linear regime, offset < rms beam size Assumes bunch train as uniform line charge Assumes ion distribution trapped within bunch distribution, σ ion ≈ σ b /√2 Fast beam ion instability models 17 September 2014 L. Mether, Two-beam instabilities19

20. o Analytical model often used for estimates, but has limited applicability Approximations valid when ω ion c << L sep Fast beam ion instability models 17 September 2014 L. Mether, Two-beam instabilities20 Raubenheimer et al. Phys. Rev. E 52, 5, 5487, Stupakov et al. Phys. Rev. E 52, 5, 5499 << 1 Model not necessarily good description for low ε!

21. o Analytical model often used for estimates, but has limited applicability Only linear regime, offset < rms beam size Assumes bunch train as uniform line charge Assumes ion distribution trapped within bunch distribution, σ ion ≈ σ b /√2  Not necessarily good approximation for low ε! o Numerical model available (FASTION) Strong-strong macroparticle simulation tool Used for CLIC Main Linac, transfer lines, modified version for CesrTA Fast beam ion instability models 17 September 2014 L. Mether, Two-beam instabilities21

22. FASTION simulation: CLIC-DR with 20nT of H 2 O 17 September 2014 L. Mether, Two-beam instabilities22 σ ion > σ b /√2

23. o Analytical model often used for estimates, but has limited applicability Only linear regime, offset < rms beam size Assumes bunch train as uniform line charge Assumes ion distribution trapped within bunch distribution, σ ion ≈ σ b /√2  Not necessarily good approximation for low-ε! o Numerical model available (FASTION) Macroparticle simulation tool, including several ingredients Used for CLIC Main Linac, transfer lines, modified version for CesrTA Model improvement and optimization for low-ε rings on-going  Quantitative comparison between theoretical predictions, simulations and measurements Low-ε  computational challenge to resolve small beam in large chamber Fast beam ion instability models 17 September 2014 L. Mether, Two-beam instabilities23

24. Summary & conclusions o Two-beam effects important for performance of low-ε machines Low ε  more sensitive to both types of instability o Electron cloud Observed frequently in running machines, good models for future machines Single-bunch instability – difficult to control by feedback  Ecloud should be suppressed in future machines Ongoing research on suppression (coating, clearing electrodes, scrubbing) o Fast beam ion instability Observed occasionally, analytical formulae typically used for estimates Detailed numerical model available  improvement & optimization ongoing Multi-bunch instability - may be controlled through feedback feedback can also be involved in triggering instability 17 September 2014 L. Mether, Two-beam instabilities24 For more details, see session “Two-Stream Instabilities” at TWIICE 2014