Measurements and simulations of electron-cloudinduced tune shifts and emittance growth at CESRTA

Size: px

Start display at page:

Download "Measurements and simulations of electron-cloudinduced tune shifts and emittance growth at CESRTA"

Loreen Wilcox
5 years ago
Views:

1 Measurements and simulations of electron-cloudinduced tune shifts and emittance growth at CESRTA Stephen Poprocki, J.A. Crittenden, D.L. Rubin, S.T. Wang Cornell University ECLOUD 8 June -7, 8 Elba Island, Italy

2 Electron Cloud (EC) can cause instabilities and emittance growth, and can be a limiting factor in accelerator performance An increase in vertical beam size due to electron cloud has been seen in many e+ rings: PEPII, KEKB, DAPHNE, CESR EC has been studied at CESRTA (Cornell Electron-Positron Storage Ring Test Accelerator) since 8 Local and ring-wide EC measurements EC mitigation techniques Inform ILC damping ring design Review Emittance growth has been measured along trains of positron bunches, and compared to simulations ECLOUD 8: Measurements and simulations of electron-cloud-induced tune shifts and emittance growth at CesrTA Stephen Poprocki!

3 This talk will present our measurements, describe the simulations, and compare results Focus on recent developments: Improved tune shift measurements Fitting e-cloud model to tune shift measurements at various bunch currents & beam energies Improved modeling of photons from synchrotron radiation & generation of primary electrons (Jim Crittenden s talk Wednesday morning) Effect on emittance growth simulations These improvements greatly enhance the predictive power of the model Overview Can be applied to any storage ring given a lattice and vacuum chamber information ECLOUD 8: Measurements and simulations of electron-cloud-induced tune shifts and emittance growth at CesrTA Stephen Poprocki!

4 Electron Cloud positron bunch photons photo-electron secondary electron Buildup of electrons hitting the vacuum chamber wall and generating secondary electrons Main source: photoelectrons from synchrotron radiation Also beam-gas ionization or stray protons hitting the wall Bunches accelerate the electrons as they pass Positron bunches pull the cloud towards it ( pinch effect ) EC builds up along a train of bunches ECLOUD 8: Measurements and simulations of electron-cloud-induced tune shifts and emittance growth at CesrTA Stephen Poprocki!

CESRTA CESR (Cornell Electron-Positron Storage Ring) 768 m in circumference Starting in 8, CESR was reconfigured into a low emittance damping ring as a Test Accelerator (CESRTA) for the ILC Damping

5 CESRTA CESR (Cornell Electron-Positron Storage Ring) 768 m in circumference Starting in 8, CESR was reconfigured into a low emittance damping ring as a Test Accelerator (CESRTA) for the ILC Damping Ring specifically, and future high intensity, ultra low emittance storage rings in general The goal was to: Characterize the build-up of EC in each of the key magnetic field regions Study the most effective methods of suppressing EC in each region Electron and positron beams.8 6 GeV Flexible bunch patterns Superconducting wigglers at low energy ( GeV) Generate 9% of the synchrotron radiation ECLOUD 8: Measurements and simulations of electron-cloud-induced tune shifts and emittance growth at CesrTA Stephen Poprocki!5

6 Beam:. GeV positrons or electrons (5. GeV for additional tune shifts) Horizontal emittance:. nm, fractional energy spread: 8x -, bunch length: 9 mm bunch train,. ma/b and.7 ma/b, ns spacing (.6x and.x bunch populations) witness bunch,.5 to. ma, bunch positions to 6 Measure: Witness bunch position probes cloud as it decays Witness bunch current controls strength of pinch effect (cloud pulled in to e+ bunch) Betatron tunes: using digital tune tracker Drive an individual bunch via a gated kicker that is phase locked to the betatron tune Vertical bunch size: from X-ray beam size monitor Bunch-by-bunch, turn-by-turn Horizontal bunch size: from visible light gated camera Bunch-by-bunch, single-shot witness Bunch-by-bunch feedback on to minimize centroid motion Disabled for a single bunch when measuring its tunes Measurements ECLOUD 8: Measurements and simulations of electron-cloud-induced tune shifts and emittance growth at CesrTA Stephen Poprocki!6

7 Beam size Vertical emittance growth along a train of positron bunches above a threshold current of.5 ma/b Vertical Bunch Size [µm] Avg. Bunch Current [ma].7 ma.7 ma.7 ma.6 ma.58 ma.5 ma.5 ma. ma. ma.6 ma Trains Bunch Number ECLOUD 8: Measurements and simulations of electron-cloud-induced tune shifts and emittance growth at CesrTA Stephen Poprocki!7

8 Beam size Trains of e- bunches do not blow-up Indicates e+ emittance growth is due to EC, not another effect Trains Vertical Bunch Size [µm] ma, e. ma, e.7 ma, e +. ma, e +.7 ma/b e+.7 ma/b e Bunch Number. ma/b e+. ma/b e- ECLOUD 8: Measurements and simulations of electron-cloud-induced tune shifts and emittance growth at CesrTA Stephen Poprocki!8

9 Horizontal beam size also blows-up in.7 ma/b e+ train Horizontal beam size Trains Horizontal Bunch Size [µm] ma, e. ma, e.7 ma, e +. ma, e +.7 ma/b e+.7 ma/b e Bunch Number ECLOUD 8: Measurements and simulations of electron-cloud-induced tune shifts and emittance growth at CesrTA Stephen Poprocki!9

10 One witness bunch to a bunch.7 ma/b e+ train Start with witness at bunch #6, vary current, eject bunch, move to #55 For a given witness bunch #, the cloud it sees is the same Emittance growth strongly depends on current (pinch effect) Witness bunch beam size ma/bunch train Witness Bunch Current [ma] Vertical Bunch Size [µm] Bunch Number ECLOUD 8: Measurements and simulations of electron-cloud-induced tune shifts and emittance growth at CesrTA Stephen Poprocki!

11 Tune shift measurements Tune shifts can be measured various ways:. Pinging : Coherently kicking entire train once, measuring bunch-by-bunch, turn-byturn positions, and peak-fitting the FFTs Fast measurement (whole train at once) Multiple peaks from coupled-bunch motion contaminate signal Vertical Tune Shift [khz] ) Pinged data 6. ma/b 5. ma/b. ma/b. ma/b. ma/b 5. GeV (Revolution frequency: 9 khz) Unable to measure horizontal tune shifts from dipoles (vertical stripe of cloud moves with train). Single bunch : Feedback on all bunches except one. FFT its turn-by-turn position data Cleaner signal if kicking the single bunch with gated kicker Measures horizontal tune shift. Digital tune tracker : Enhancement on above technique, driving the bunch transversely in a phase lock loop with a beam position monitor Best method; used here Vertical Tune Shift [khz] ma/b 5 ma/b ma/b ma/b ma/b Bunch ) Digital tune tracker 5. GeV 5 5 Bunch ECLOUD 8: Measurements and simulations of electron-cloud-induced tune shifts and emittance growth at CesrTA Stephen Poprocki!

12 Tune shift measurements Tune shifts measured at. and 5. GeV at various currents:. GeV Horizontal Tune Shift [khz] ma/b. ma/b Vertical Tune Shift [khz] ma/b. ma/b Bunch Bunch 5. GeV Horizontal Tune Shift [khz] ma/b ma/b ma/b Vertical Tune Shift [khz] ma/b 5 ma/b ma/b ma/b ma/b Bunch Bunch ECLOUD 8: Measurements and simulations of electron-cloud-induced tune shifts and emittance growth at CesrTA Stephen Poprocki!

13 Simulations involve four codes which feed into each other. Tracking photons from synchrotron radiation (SynradD) Information on photons absorbed in vacuum chamber. Photo-electron production (Geant) Quantum efficiencies Photo-electron energies. Electron cloud buildup (ECLOUD) Space-charge electric field maps. Tracking of beam through the lattice with EC elements (Bmad) Betatron tunes Equilibrium beam size Simulations The separation of steps and makes this a weak strong simulation More on this later ECLOUD 8: Measurements and simulations of electron-cloud-induced tune shifts and emittance growth at CesrTA Stephen Poprocki!

14 ) Photon tracking SynradD Simulates photons from synchrotron radiation Tracks photons through vacuum chamber including specular & diffuse reflections Input: lattice, D vacuum chamber profile, material Output: information on absorbed photons: Azimuthal angle Energy Grazing angle with vacuum chamber wall Fraction per bin (%) SYNRADD: CesrTA 5. GeV e+ beam: Entire ring, phantoms Average absorbed photon rate:.8 photons/m/e Entries 59 Photon Rate ( /m/positron) Entries 59 Azimuthal angle (degs) 6 8 Absorption S Position (m) x [cm] Wigglers 5 5 Entries Entries S [m] Number of Reflections (Cutoff at ) 6 8 Origin S Position (m) ECLOUD 8: Measurements and simulations of electron-cloud-induced tune shifts and emittance growth at CesrTA Stephen Poprocki!

) Photo-electron production Geant Input: Absorbed photons Azimuthal angle Energy Vacuum CO layer Aluminum Grazing angle with vacuum chamber wall Simulates electron production from photo-electric and

15 ) Photo-electron production Geant Input: Absorbed photons Azimuthal angle Energy Vacuum CO layer Aluminum Grazing angle with vacuum chamber wall Simulates electron production from photo-electric and Auger effects Vacuum chamber material (Aluminum) and surface layer (5 nm carbon-monoxide) Output: Quantum efficiency vs azimuthal angle Photo-electron energy distributions QE depends on photon energy & grazing angle which vary azimuthally Improvement on ECLOUD model Big improvement to predictive ability See Jim Crittenden s talk Wednesday morning Incident photons ev 5 deg. grazing angle ECLOUD 8: Measurements and simulations of electron-cloud-induced tune shifts and emittance growth at CesrTA Stephen Poprocki!5

16 ) EC buildup simulation Start with EC buildup simulations with ECLOUD in both dipole and field-free regions Use element-type ring-averaged beam sizes Dipole: 7 x um Drift: 8 x um The large horizontal size is dominated by dispersion Obtain space-charge electric field maps from the EC for time slices during a single bunch passage, in ±5σ of the transverse beam size t = ps Transverse EC charge distributions in an 8 G dipole for bunch of a.7 ma/b positron train y [σ] y [σ] y [σ] y [σ] t=-.5σ z - - x [σ] x [σ] t=+.5σ z - - x [σ] t= x e E x [V/m] E y [V/m] Only ~.% of electrons are within this beam region Necessary to average over many ECLOUD simulations y [mm] x [mm] y [mm] x [mm] ECLOUD 8: Measurements and simulations of electron-cloud-induced tune shifts and emittance growth at CesrTA Stephen Poprocki!6

17 Calibration of model to tune shift measurements ECLOUD simulations depend strongly on vacuum chamber secondary yield (SEY) parameters Direct SEY measurements provide a good starting point, but it s hard to accurately determine all the parameters Still, the condition in the machine may be different To improve agreement between the ECLOUD model and our various measurements: Use a multi-objective optimizer to fit the SEY parameters to tune shift data At each iteration, run ECLOUD simulations in parallel varying each parameter by an adaptive increment Calculate Jacobian & provide to optimizer TABLE I. Main parameters of the model. Copper Stainless steel Emitted angular spectrum (Sec. IIC) Backscattered electrons (Sec. IIIB) P ;e..7 ^P ;e.96.5 ^E e (ev) W (ev) 6.86 p.9 e (ev).9 e.6.6 e Rediffused electrons (Sec. IIIC) P ;r..7 E r (ev). r. q.5. r.6.6 r True-secondary electrons (Sec. IIID) ^ ts.888. ^E ts (ev) 76.8 s.5.8 t t.8.8 t.7.7 t Total SEY a ^E t (ev) 7 9 ^ t..5 a Note that ^E ^E and ^ ^ P P provided M. Furman & M. Pivi, Probabilistic Model for the Simulation of Secondary Electron Emission, Phys. Rev. ST Accel. Beams 5, (Dec. ) ECLOUD 8: Measurements and simulations of electron-cloud-induced tune shifts and emittance growth at CesrTA Stephen Poprocki!7

Electric field gradients from cloud space-charge fields 9 Tune shifts calculated from the cloud space-charge Dipole electric field 8 Drift gradients 6 Gradient just before a bunch passage coherent

18 Electric field gradients from cloud space-charge fields 9 Tune shifts calculated from the cloud space-charge Dipole electric field 8 Drift gradients 6 Gradient just before a bunch passage coherent tune shift Demonstrated in witness bunch tune measurements (left) Gradient during pinch incoherent tune spread, emittance growth Demonstrated in witness bunch size measurements (right) Vertical Tune Shift [khz] Horizontal Tune Shift [khz] Bunch Current [ma] Train Bunch E x gradient [kv/m ] Current [ma] Train 7 5 GeV.7mA/b, bunch ECLOUD 8: Measurements and simulations of electron-cloud-induced tune shifts and emittance growth at CesrTA Stephen Poprocki Vertical Bunch Size [µm] Time slice E y gradient [kv/m ] ma/bunch train Dipole Drift GeV.7mA/b, bunch Time slice Figure : Top: vertical electron cloud space-ch tric field gradients for the time slices within e bunches, for dipoles and drifts. Bottom: electric fi ents for the time slices in bunch, showing of the bunch at time slice 6. Witness Bunch Current [ma] SUMMARY We have obtained improved measurements o tune shifts along trains of positron bunches in the and vertical planes for a range of bunch popul abling advances in the predictive power of elec buildup modeling. We employed the SynradD a simulation codes Bunch Number to eliminate ad hoc assumption electron production rates and kinematics charac prior buildup simulations (see [6] for details). Elec model parameters for secondary electron yield were determined through tune shift modeling ca optimized to the measurements. This work achie!8

19 Simulated tune shifts at. GeV Results from fitting ECLOUD model to data Simultaneously fit to:.7 ma/b at. GeV,, and 6 ma/b at 5. GeV.8 (next slide) Parameters varied include: Peak energy of true secondary. yield True secondary yield 's' parameter True secondary yield Rediffused secondary yield Elastic yield at energy Same SEY parameters used in simulations at all currents & energies TS77_GeV_SD: Coherent Tune Shifts for Jobs 69/69. ma/b Dipole Drift Dipole + Drift Apr/7 Positron,. ma ΔQ X ΔQ Y.8.8 Dipole Drift Dipole + Drift Apr/7 Positron,. ma TS77_GeV_SD:.6 Dipole Coherent Drift Tune Dipole Shifts.6 + Driftfor Jobs 69/69 Apr/7 Positron,. ma Dipole Drift Dipole + Drift Apr/7 Positron,. ma ΔQ X ΔQ. ΔQ X Y.8. ΔQ Y ΔQ.8.8 X ΔQ Y Time (ns). Time (ns)...7 ma/b Time (ns) Time (ns).5 6 ΔQ X Time (ns) ΔQ Y Time (ns) 5 Time (ns) Time (ns).5.5 TS77_GeV_SD: Coherent Tune Shifts for Jobs 69/69 TS77_GeV_SD: Coherent Tune Shifts for Jobs 69/6.6. Time (ns) Time (ns) (Revolution frequency: 9 khz) ECLOUD 8: Measurements and simulations of electron-cloud-induced tune shifts and emittance growth at CesrTA Stephen Poprocki!9

20 SD: Coherent ΔQ X Tune Shifts for Jobs 69/69 ΔQ Y rift Dipole + Drift Apr/7 Positron,. 7_GeV_SD:.5 Coherent Tune Shifts for ma/b Jobs 69/69. Time (ns).... Time (ns) Simulated tune shifts at 5. GeV CH75_SD: Coherent Tune Shifts for Jobs 69/695CH75_SD: Coherent Tune Shifts for Jobs 7/7.5 Dipole Drift Dipole + Drift CHESS May//7. ma Dipole.5 Dipole Drift Dipole + Drift Apr/7 Positron,. ma ΔQ Y ΔQ X ΔQ Y CH75_SD: Coherent Tune Shifts for Jobs 696/697 CH75_SD: Time (ns) Coherent Tune Shifts for Time Jobs (ns) 7/7 Dipole Drift Dipole + Drift CHESS May//7.6 ma.5 ΔQ (ns) X.5 ΔQ Time (ns) Y.5 Time (ns).5 Time ma/b (ns) Time (ns) Time (ns) CH75_SD: Coherent Tune Shifts for Jobs 698/699 Dipole Drift Dipole + Drift CHESS May//7.6 ma Dipole Drift Dipole + Drift CHESS May//7 5.6 m ΔQ X ΔQ Y Dipole Drift Dipole + Drift CHESS May// m ΔQ X Time (ns) ΔQ Y 5 ma/b 6 ma/b Time (ns) ΔQ X Time (ns) ΔQ Y ma/b Time (ns) ECLOUD 8: Measurements and simulations of electron-cloud-induced tune shifts and emittance growth at CesrTA Stephen Poprocki!

21 ) Tracking simulations Use the time-sliced electric field maps in EC elements at the dipole and drifts Track particles in bunch through the full lattice (using Bmad) for multiple damping times, with radiation excitation and damping weak-strong model: does not take into account effects on the cloud due to changes in the beam Tracking: Weak: beam; Strong: EC EC buildup simulations: Weak: EC; Strong: beam Justification: EC buildup simulations are rather insensitive to vertical beam size Strong-strong simulations are too computationally intensive to track for enough turns Damping times at CesrTA are ~, turns We want equilibrium beam sizes ECLOUD 8: Measurements and simulations of electron-cloud-induced tune shifts and emittance growth at CesrTA Stephen Poprocki!

22 Bunch size from simulation is the average over last k turns (of 6k) See vertical emittance growth in.7 ma/b simulations Vertical bunch size growth - train Vertical Bunch Size [µm] ma/b positron train Bunch ma, e +. ma, e + Sim:.7 ma, e + Sim:. ma, e Turns Bunch Number ECLOUD 8: Measurements and simulations of electron-cloud-induced tune shifts and emittance growth at CesrTA Stephen Poprocki!

23 Horizontal bunch size growth - train See some horizontal emittance growth in.7 ma/b simulations compared to. ma/b but needs investigation Horizontal Bunch Size [µm] ma/b positron train Bunch ma, e +. ma, e + Sim:.7 ma, e + Sim:. ma, e Turns Bunch Number ECLOUD 8: Measurements and simulations of electron-cloud-induced tune shifts and emittance growth at CesrTA Stephen Poprocki!

24 Vertical Bunch Size [µm] Vertical bunch size growth - witness bunch.7 ma/bunch train Witness Bunch Current [ma] Sim:. Sim: Bunch Number More emittance growth with: shorter distances from train (more cloud) higher witness bunch current (more pinch) Simulations show similar behavior ECLOUD 8: Measurements and simulations of electron-cloud-induced tune shifts and emittance growth at CesrTA Stephen Poprocki!

25 We have obtained various measurements of tune shifts and emittance growth from electron clouds Our e-cloud model has been improved with precise modeling of synchrotron radiation photons & generation of primary electrons The model has been validated with improved tune shift measurements for a range of bunch currents at. and 5. GeV A witness bunch at a range of currents gives a direct measurement of the pinch effect Vertical emittance growth scales with pinch Coherent tune shift does not Our weak-strong incoherent model is consistent with this data The simulations can uncover the largest contributions to tune shifts and emittance growth EC mitigation methods can be targeted to these regions and tested in simulation Future work: Further investigate horizontal emittance growth in simulation Use model to predict EC effects at future accelerators Use model to understand underlying factors driving emittance growth New approaches to mitigating emittance growth from EC Summary ECLOUD 8: Measurements and simulations of electron-cloud-induced tune shifts and emittance growth at CesrTA Stephen Poprocki!5

26 Thank you for your attention ECLOUD 8: Measurements and simulations of electron-cloud-induced tune shifts and emittance growth at CesrTA Stephen Poprocki!6

27 Witness bunch to a. ma/b train (below threshold) 8 6. ma/bunch train Witness Bunch 5 6 e+ e ma/bunch train Witness Bunch 5 6 Vertical Bunch Size [µm] 8 6 Vertical Bunch Size [µm] Witness Bunch Current [ma] Witness Bunch Current [ma]. ma/bunch train. ma/bunch train 8 6 Witness Bunch 5 6 Witness Bunch 5 6 Horizontal Bunch Size [µm] 8 6 Horizontal Bunch Size [µm] Witness Bunch Current [ma] Witness Bunch Current [ma] ECLOUD 8: Measurements and simulations of electron-cloud-induced tune shifts and emittance growth at CesrTA Stephen Poprocki!7

Simulations of synchrotron-radiation-induced electron production in the CESR vacuum chamber wall

45th ICFA Beam Dynamic Workshop June 8 12, 2009, Cornell University, Ithaca New York Simulations of synchrotron-radiation-induced electron production in the CESR vacuum chamber wall Jim Crittenden Stephen