The CERN Antiproton Decelerator AD

1
The CERN Antiproton Decelerator AD

•  
•  Performance, developments and future
  possibilities
•  

2
The CERN Antiproton Decelerator
3
Run statistics
4
HW breakdowns in 2004

• PS ejection septum
• Water leak May 1 week stop to replace with
  spare unit.
• Spare unit develops same failure in July 3
  weeks stop to repair spare and install.
• AD electron cooler vacuum
• 2 weeks stop in June excessive outgassing in
  collector region caused de-activation of NEGs.
  Dissasembly, inspection, replacement of all
  suspect parts, bakeout.

5
Peak beam intensity
6
Current status, ejected beam
() nominal/peak
7
Multiejection

• Simple scheme introduced at 100 MeV/c using
  existing RF-HW minimal modifications.
• Bunching on h1,3 or 6 is possible with present
  RF-HW.
• 2.4 s. rep.rate imposed by ejection magnets.
• Good efficiencies and beam lifetime obtained. (12
  s. longer coast at 100 MeV/c _at_ h6)

8
Current limits beam intensity

• Proton beam 1.5E13 on target, only marginal
  improvements can be obtained in the PS-complex,
  mainly limited by spacecharge effects in PSB
• 25 increase can be had with 5-bunch beam (now
  4), need 2 injections/cycle into PS and 3.6 (now
  2.4) s cycle. Modifications in PS required
• Stacking mode studied, a certain increase can be
  expected but tradeoff against cycle length.
  Modifications in PS required
• Target/collector set-up for reliability rather
  than max. yield
• Transverse acceptances close to optimum
• Losses during cycle balanced vs. cycle
  length/rep. rate

9
Current limits cycle length
10
Current limits cycle length

• From 140 to 84 seconds in 5 years
• Ramp length
• Field lag modifies tunes/orbits during ramps and
  arrival at plateaus, only B-main is compensated
  at plateau arrival losses if ramps are too steep
  (15 2GeV/c 300MeV/c)
• Fast eddy currents in vacuum chamber can modify
  tunes as much as 0.01 and provoke large orbit
  excursions of up to 35mm _at_ 300MeV/c with current
  slope, nothing can be done about this
• Tunes cannot (yet) be measured on ramps
• Ramp shape can be further somewhat optimised

11
Current limits cycle length

• Beam cooling
• Stochastic cooling well optimised. Some losses of
  particles with large momentum offset at 3.5GeV/c
  are caused by limited momentum acceptance of
  cooling system
• Electron cooling is slower than foreseen due to
• Alignment problems e-Pbar, better correctors
  needed
• Large initial emittances (see above) due to
  adiabaticresonant blowup during ramps
• Slow cooling at beginning of plateau due to field
  lag
• Attempts to find better parameter settings with
  Betacool simulation will be done
• Other.

12
Current limits cycle length
13
Current limits beam densityCapture and ejection
14
Current limits beam density

• Cooled beam at 100MeV/c
• Vertical planeok
• Horizontal planeDense core with halo, caused
  or aggravated by interaction between the
  RF-system and the electron cooler

15
Transverse beam profile at 100 MeV/c

• Severe halo structure develops mainly during
  ejection bunching.
• Debunching with ecooler at nominal voltage also
  contributes. (was done to gain time)
• Some overlap Vrf Vec was found at ejection
  bunching.

16
Transverse beam profile at 100 MeV/c, present
situation

• Ecooler voltage is close to zero during
  debunching at beginning of plateau.
• At ejection, ecooler voltage is ramped down just
  before Vrf reaches max.
• Bunch length around 100ns.
• Small residual halo still there.

17
Reduction of dp/p

• Longitudinal phase-space tomography with moving
  bucket
• Deceleration 300-100 MeV/c with reduced voltage
  (3kV 500V)
• dp/p reduced by 50

18
Current limits other

• Experimental area beamline switching
• Delays due to inherent AD stability problems at
  low energy
• Re-tuning of lines often necessary, slow process
  due to slow repetition rate and destructive BPMs
• Non-destructive BPMs would allow on-line
  measurements and corrections
• Tests of switching each cycle could be done if
  needed, uncertain if possible with present
  magnetic elements

19
2005 activities

• Improvement of ecooler
• Vacuum renew and add NEGs
• Ecooler BPMs improve shielding
• Improved correctors for better alignment at 300
  MeV/c
• AD consolidation programme worked out and
  included in the general AB plan.

20
2006 run

• On the draft schedule (for the moment)
• Reduced physics run June 5 September 3
• 7d/7, 24h/24 approx. 2000 h of physics.
• Approval issues remain.

21
Future possibilities

• 2nd RFQD was discussed..

22
Future possibilities ELENA

• How do we gain in intensity with an added
  decelerator/cooler ?
• Beam from AD 3 107 antiprotons per 84s cycle at
  5.3 MeV kinetic energy with transverse emittances
  1 to 2 p mm mrad.
• Use a small ring to further decelerate beam to
  100 keV, increase density by electron cooling
• Use of a much thinner degrader will significantly
  reduce adiabatic blowup and scattering gt two
  orders of magnitude gain in intensity is expected
  for ATHENA and ATRAP.
• Beam emittances after deceleration and cooling in
  ELENA will be much smaller than after the RFQD gt
  one order of magnitude gain in intensity is
  expected for ASACUSA.
• 100 keV kinetic energy is close to optimal both
  from the point of view of beam intensity,
  momentum spread and separation of transfer line
  and trap vacuum.

23
Future possibilities ELENA

• Compact machine located inside of AD Hall with
  minimum of rearrangement.
• Energy range from 5.3 MeV to 100 keV.
• Equipped with electron cooler
• Machine assembling and commissioning has to be
  done without disturbing current AD operation.
• A similar ring for decelerating antiprotons
  from LEAR was proposed by H.Herr in 1982.

24
ELENA ring configuration
25
ELENA main parameters
26
ELENA cycle

• No electron cooling is performed at injection
  beam is already cooled in AD. After injection
  beam is decelerated immediately.
• One intermediate cooling (at 40 MeV/c probably)
  is needed to avoid beam losses

27
ELENA in the AD hall
28
ELENA in the AD hall
29
Current status

• ELENA proposal supported by SPSC
• AB department is waiting for decision by CERN
  management
• Consequences for AB if we follow SPSC
• Need to do a study that defines
• Expected beam density increase
• Costs
• Beam sharing between users
• Set-up and commissioning plan

30
Conclusions

• AD routinely surpasses the design goals for pbar
  flux and beam density
• Further (limited) improvements could be
  considered
• Tuning and development is slow due to long cycle
  and low beam intensities
• ELENA would open up new possibilities and
  challenges..

31
Longitudinal phase-space tomography, 2 GeV/c
32
Longitudinal phase-space tomography, 3.5 GeV/c
33
Longitudinal phase-space tomography, 100 MeV/c
34
Future possibilities ELENA

• Beam from AD 3 107 antiprotons per 84s cycle at
  5.3 MeV kinetic energy with transverse emittances
  1 to 2 p mm mrad.
• How antiprotons are decelerated further today
• Experiments with antihydrogen program (ATHENA and
  ATRAP) use degraders to slow the beam further
  poor efficiency due to adiabatic blow up and
  scattering in the degrader.
• ASACUSA uses RFQD for deceleration down to around
  100 keV kinetic energy. Due to absence of
  cooling, beam deceleration in RFQD is accompanied
  by adiabatic blow up (factor 7 in each plane)
  which reduces trapping efficiency.

35
ELENA how we define limit on tune excursion?

• MD studies in AD for investigation
• of the beam stable area in tune
• diagram.
• Machine is stable when tunes are
• inside of polygon. Beam is lost
• when tunes approach 5.5
• (2nd order resonance) and 5.33
• (3rd order resonance).
• CERN Booster experience tune excursion of
  0.4 is possible for a short time with careful
  compensation resonance driving terms. CERN PS
  experience tune excursion of 0.2 is possible
  with similar precautions.

36
ELENA Lattice considerations

• Beam focusing is achieved by proper choice of
  edge angles of the dipoles. Economical solution
  for cost and space saving neither gradient
  magnets, nor quadrupoles needed!
• Large area in tune diagram should be available
  for tune excursion caused by space charge.
  Conservative estimate for coherent tune shift ?Q
  0.10 was accepted which is based on CERN Booster,
  PS and AD experience.
• Tunes Qx1.45, Qy1.43 (with similar fractional
  parts as in the AD) fit requirements.
• Choice of tunes together with required straight
  section length defines machine circumference
  about 22m.

37
ELENA Beam lifetime considerations

• Intrabeam scattering (IBS) is important at very
  low energies in a short bunch with small
  emittances. With reasonable choice of beam
  parameters (1.5 107 particles, emittances 5p mm
  mrad and ?p/p10-3) emittance rising times for
  coasting beams are more than 1 minute. For
  bunched beam 1.3m long they are in the order of 1
  second.
• Residual gas scattering produces beam blow up
  0.5p mm mrad/s at 100 keV and pressure 3 10-12
  Torr.
• Electron cooling at 100 keV will be strong enough
  to successfully counteract intrabeam and residual
  gas scattering.
• For fast extraction, the beam blow up is limited
  by the time of beam bunching and bunch rotation
  (if needed), which takes a few hundred msec.