Summary of the ECLOUD04 Workshop

1
Summary of the ECLOUD04 Workshop

• Robert Macek, LANL and Miguel Furman, LBNL, Oct
  19, 2004

2
Outline

• Program
• Electron cloud formation and buildup
• Observations
• Simulations (see summary reports of sessions D
  and E in proceedings)
• Electron cloud driven instabilities
• Observations
• Theory and simulations (see summary report of
  session F in proceedings and reviews by
  Zimmermann and Ohmi)
• Progress on cures
• Some issues for the future
• Conclusions

3
ECLOUD04 program and some statistics

• 6 sessions plus summary session and a panel
  discussion
• 2 sessions on observations at existing
  accelerators and concerns for future machines (17
  talks)
• one session on surface properties, measurements
  and treatments (7 talks)
• 2 sessions on simulations of e-cloud buildup (17
  talks)
• an all day session on theory and simulations of
  e-cloud instabilities (12 talks)
• Panel discussion on future needs and future
  directions (10 panelists moderator)
• Winery tour (optional) and banquet Wed afternoon
  and evening
• Some statistics
• 59 participants from 20 institutions in North
  America, Europe and Asia
• 3 invitees unable to attend because of problems
  with entry into the U.S.
• 53 talks plus 6 summary talks
• Talks posted to the web and proceedings are in
  progress
• See http//icfa-ecloud04.web.cern.ch/icfa-ecloud04
  /

4
List of electron cloud effects (ECE) observed at
accelerators

• Beam induced multipacting (amplifies stray
  electrons)
• Short-bunch (APS, KEKB, PEP-II, PS/SPS/LHC)
• Trailing-edge multipactor (PSR, KEK-PS, ISIS?)
• Vacuum pressure rise (electron-stimulated gas
  desorption)
• Often the first indication of ECE and observed at
  many machines
• A limiting factor for certain modes of RHIC
  operation
• Transverse instabilities (aka electron cloud
  instabilities, two-stream instabilities)
• Coupled bunch (B factories, BEPC, PS, SPS)
• Single bunch (KEKB, SPS, PSR)
• Emittance growth (KEKB, PEP-II, SPS)
• Tune shifts (KEKB, AGS Booster)
• Heat load on cryogenic wall (anticipated for LHC,
  measured in SPS beam experiments)
• Judged a critical issue for LHC performance
• E-cloud induced spurious signals in beam
  diagnostics (PS, SPS, PSR)

5
Amplification by beam-induced multipactor
Beam-induced multipactor in short bunchring
(schematic for LHC from F. Zimmermann talk at
Santa Fe Workshop 2000 )
Trailing-edge multipactor in long bunch ring
(e.g. PSR)
6
Parts of the e-cloud effects (ECE) problem

• E-cloud formation, buildup and dissipation
• Primary or initial electrons (PE, residual gas
  ionization, losses etc)
• Buildup by beam-induced multipactor and/or
  trapping
• Codes ECLOUD, CLOUDLAND, POSINST, ORBIT, CSEC
• Interaction of a given cloud with the beam to
  produce instabilities and/or emittance growth
• Analytical models e.g. centroid model for
  coasting beam
• Simulation codes HEADTAIL, PEHTS, PEI, QUICKPIC,
  BEST, NCSEC, ORBIT, WARP
• Moving towards a combined, self-consistent
  treatment of buildup and instabilities/emittance
  growth
• Codes NCSEC, ORBIT, POSINST WARP, PARSEC
• Mitigations of ECE
• Control or reduce cloud buildup
• Surface treatments e.g. coatings, beam scrubbing,
  grooved surfaces
• Clearing electrodes
• Damping of the instabilities

7
Electron cloud formation, buildup and dissipation

• Initial (or primary) sources of electrons are
  ubiquitous at accelerators
• 3 regimes of buildup
• Long train of short bunches with beam-induced
  multipactor amplification (B factories, PS, SPS,
  LHC)
• EC density saturates after some number of
  bunches from space charge or from equilibrium
  between generation and losses
• Long bunch (many electron oscillations during
  bunch passage) with trailing edge multipactor
  amplification (PSR, SNS, ISIS?)
• Trapping of initial electrons (during a long
  bunch) e.g. coasting beams, heavy ion linac or in
  quadrupoles
• SEY characteristics of chamber surfaces are very
  important physics inputs to the simulations
• Multipactor gain strongly dependent on dmax
• Dissipation after the beam passes can be
  surprisingly slow because of high SEY for low
  energy electrons
• e.g. 200 ns time constant observed at PSR

8
SEY Characteristics
Recent measurements at CERN by R. Cimino et al
and presented in his talk at ECLOUD04
Typical curve for partially scrubbed StSt used in
simulations (in ECLOUD code from G. Bellodi talk
at ECLOUD04)
9
Direct observations of e-cloud buildup

• Vacuum pressure rise
• Gas desorption from electrons hitting the wall
  beam scrapping?
• Biased collection electrodes (BPM plates) at SPS,
  PSR
• Difficult to relate to electron flux hitting the
  wall
• Measures net current at the plate
• Bias fields change the secondary
  emission/multipacting at the surface
• Strip detectors at CERN SPS
• RFAs and similar devices at APS, KEK-PS, SPS,
  PSR, and RHIC
• Measures e-flux striking the wall and with
  repeller voltage can provide information on
  energy distribution
• Simulations give reasonable agreement with
  measurements given the uncertainties on SEY input
  parameters and seed electrons
• Some observations unexplained e.g. electron
  bursts at PSR, recovery after sweeping at PSR
  and KEK-PS, microwave transmission anomaly at SPS
  may point to missing physics or missing details

10
RHIC data on pressure rise and e-cloud
Presented by W. Fischer

• Proton fill with 108 ns spacing
• E-cloud detector measures e-flux (above 10eV)
  hitting the wall
• The ED signal tracks log(P)! Why?
• E-cloud density measured at RHIC can be
  reproduced with simulations (CSEC) if SEY (1.9)
  is adjusted

11
Main Results at 25 ns Bunch SpacingBuild-up
measured using a strip detectorfrom talk by M.
Jimenez at ECLOUD04
12
Electron signals from RFA in straight section at
PSRfrom talk by R. Macek at ECLOUD04
Signals averaged for 32 beam macropulses, Stable
beam 8 mC/pulse beam intensity, Device is
labeled ED42Y, Transimpedance 3.5 k?, opening
1 cm2
POSINST simulations by Furman and Pivi give
reasonable agreement with these data
Bk95, p6-12
13
Recovery after Clearing Gap of electrons at PSR
Bk 98, p 50-51
14
Electron detector data from KEK-PS (main ring)
Presented by T. Toyama, KEK
15
Some unexplained EC observations

• Recovery behavior after sweeping electrons from
  the pipe observed at PSR and KEK-PS
• Microwave transmission through section SPS beam
  pipe to measure index of refraction of e-cloud
  yields surprising results (T. Kroyer talk)
• Electron bursts observed at PSR
• Missing physics or missing detail in EC models?

Multiturn (110) sequence of signals from 2
electron detectors and a local loss monitor near
end of accumulation
16
ECI observations, theory and modeling

• Many observations of thresholds, mode spectra and
  growth rates
• B factories, Photon factories with e rings, SPS
  with LHC beams, PSR
• Analytical models
• e.g coasting beam centroid models for ISR, PSR
  provided insight into ECI
• Models using approximate analytic wake fields
• Simulations (see session summary by Zimmermann
  Wolski, review talks by Zimmermann and Ohmi at
  ECLOUD04)
• A number of codes in use e.g., HEADTAIL, PEHTS,
  PEI, QUICKPIC, BEST, NCSEC, ORBIT, WARP
• A number of approaches and some tailored for
  particular regimes
• Benchmarking of codes on some standard problems
  underway
• Comparisons with experimental data
• Simulations are in generally good agreement with
  thresholds, mode spectra, and growth rates for
  KEKB and reasonable agreement on thresholds and
  modes at SPS/LHC
• Analytical models and simulations agree with
  observations on mode spectra for PSR and in rough
  agreement on thresholds BEST code also gives
  rough agreement on growth rates

17
Examples of observation of thresholds for ECI
KEKB
At KEKB observe sudden increase in beam size as
beam current is varied
At PSR a given beam intensity is accumulated and
stored for 500 ms, rf voltage is lowered until
unstable motion and beam loss suddenly appear.
Threshold intensity is linear in rf voltage
PSR
18
Comparison of experimental mode spectra with PEI
simulations
Presented by K. Ohmi
19
ECI observations at SPS with LHC beams (from G.
Arduini talk)
Emittance growth alongthe batch for 1st 48
bunches
Snapshot of BPM signals for 1st 48 bunchesof
the batch
Beam Intensityevolution showing Losses in the
tail
20
Single bunch ECI at PSR
Mode spectra at threshold for 2 intensities
(agree with centroid model)
Time evolution of instability at PSR
Spectrogram of dipole motion (R. Macek)
Calculation of e-cloud wake using CSEC (M.
Blaskiewicz)
21
Progress on cures

• Weak solenoids were very effective in reducing
  e-cloud and ECI at B-factories (KEKB and PEP-II)
• Tests of NEG coatings for reducing SEY are very
  encouraging (e.g. see talk by A. Rossi, also M.
  Pivi summary of session C)
• Neg coatings planned for warm sections of LHC
• Test of grooved metal surface showed 30
  reduction in effective SEY (see talk by G.
  Stupakov)
• Beam scrubbing/conditioning to reduce SEY shown
  to be effective for LHC beams at SPS, also
  effective at PSR
• Tests at CERN SPS also suggest scrubbing maybe
  slower on a cold surface
• Damping of ECI by feedback effective at SPS for
  coupled-bunch instabilities in the horizontal
  plane (see talk by G. Arduini)
• Landau damping of e-p by increasing tune spread
  in various ways effective at PSR as is coupled
  Landau damping

22
Main Results at 25 ns Bunch SpacingPressure
decrease resulting from both vacuum scrubbing and
Beam conditioning (M. Jimenez at ECLOUD04)
(hours)
23
Effect of beam scrubbing on prompt electron
signals at PSR
Data for 8 mC/pulse beam presented by R. Macek
at ECLOUD04
Electrons surviving the gap also showed some
reduction ( factor of 2) with beam conditioning
for higher intensity beams (8 mC/pulse)
24
Panel discussion on future directions

• General consensus that surface science underlying
  secondary emission and gas desorption is very
  important and needs more work and would benefit
  from greater inter-laboratory collaboration
• Systematic benchmarking codes against one another
  and against experimental data is needed
  (international collaboration)
• Need for self-consistent combined treatment of
  e-cloud buildup and instability dynamics in the
  simulation codes
• Measurement of the e-cloud density at the beam
  locations is of fundamental importance and
  requires new diagnostic methods
• Careful evaluation of NEG coatings for long term
  effectiveness in reducing SEY and its
  effectiveness in reducing gas desorption is
  recommended
• Better understanding and characterization of gas
  desorption (also electron emission) by beam
  collisions with walls are a priority for heavy
  ion machines (RHIC, HIF linacs, future GSI
  machines)

25
Conclusions

• Good progress towards better understanding of ECE
  and means of control but much remains to be
  resolved in order to predict performance of new
  machines with high confidence
• The B factories are running quite well after
  controlling ECE largely by means of weak solenoid
  windings on a good fraction of the ring
  circumference
• May need additional mitigation of ECE for the
  proposed super B factory projects
• The systematic program at the CERN SPS in
  preparation for LHC keeps yielding valuable
  information on many aspects of ECE especially on
  beam scrubbing which is a key element of the LHC
  strategy to control ECE
• Vacuum pressure rise is a major limitation on
  RHIC performance.
• Convincing evidence for ECE at RHIC but beam
  scraping losses appear to also contribute
  significantly to vacuum pressure rise
• Simulations of e-cloud buildup are in reasonable
  agreement with observations given uncertainties
  on input parameters e.g. effective SEY and
  certain source terms
• Effect of unstable beam motion on multipacting
  generally not included
• Models and simulations of ECI give generally good
  agreement with experiments on mode spectra
  thresholds and growth rates are in good agreement
  for KEKB data and comparisons less complete for
  other rings
• Comparisons and benchmarking between codes and
  with experimental data is underway
• Official goal of US-LARP and CARE program in
  Europe
• Codes that self consistently combine buildup and
  beam dynamics are needed but challenge present
  computing capabilities

26
Many thanks to all the participants for an
excellent workshop!