Beam-Beam Effects at CESR

1
Beam-Beam Effects at CESR

• Mark A. Palmer
• Cornell University
• July 14, 2001

2
CESR Pretzel Operation

• Single beampipe
• 9 trains of 4 or 5 bunches
• Bunches spaced by 14 ns
• Spacing between 1st bunches in trains
• 280 ns, 280 ns, 294 ns,.
• 1281 RF buckets not divisible by 9
• Collisions at a single IP (CLEO)
• Electrostatic Pretzel
• Provides horizontal separation in the arcs
• Vertical electrostatic separators
• Separates the bunches at the 2nd (unused) IP
• Crossing Angle at IP 2.5 mrad

3
Parasitic Crossings
Closed orbit
x(s) x0 bh(s) sin(fh(s)-f0)
Vertical Separation
Beam-beam tune shift
Ib x02sin2(fh(s)-f0)
Ibbh x2
DQh

• Horizontal separation of gt 8s except at 2nd IP
• Beam-beam tune shift is smallest for fh(s) -
  f0 p/2
• Primary limit on train length!

4
Parasitic Bunch-by-Bunch Effects

• Simulation of long range beam-beam interaction

• Strong-strong simulation
• Tunes
• O(1kHz) Spread in horizontal and vertical tunes
  by bunch
• Size of this spread has been verified by direct
  measurement
• Vertical displacement of bunches at parasitic
  crossings
• Distorts vertical closed orbit
• Occurs at both IPs
• Bunch at the start (end) of train experiences
  kick as it leaves (approaches) the IP
• Additional effects
• Variations in chromaticity
• Variations in angles and beta functions at IP

5
Beam-Beam Tuneshift
4-bunch
5-bunch

• Beam-beam tune shift
• Observe xv 0.07 in 4-bunch running
• Decreases for 5-bunch operation although improved
  net luminosity performance

6
Multi-bunch Performance Issues

• Bunch Current Limits
• Observations
• 11mA/bunch possible in 9x1 running
• See decreasing bunch current limit as increase
  the number of bunches/train
• c Parasitic crossings limit the maximum bunch
  current NOT the main
• beam-beam interaction
• Installation of superconducting IR (underway now)
  will significantly improve the first parasitic
  crossing adjacent to the CLEO IP
• Tune Spread
• Simulation indicates a bunch-by-bunch spread of
  O(1kHz) in both vertical and horizontal
• Observed tune spread is consistent in size with
  the simulation
• Width of working point in the tune plane
• 100 Hz Horizontal
• 1 kHz Vertical
• Currently investigating the use of an RFQ to
  correct the bunch-dependent tune

7
Bunch-to-Bunch Luminosity

• Monitor barrel calorimeter bhabha rate in CLEO
  detector
• Tracking information provides bunch
  identification
• Specific Luminosity

L(bunch) dt
I(bunch) dt

• Information integrated over run
• ( 1hr) for statistics
• Car c location of bunch in train
• Observe significant variations in all quantities
• 25 degradation in luminosity for worst bunch
  relative to best

8
Bunch-to-Bunch Differential Orbits

• BBI Luminosity Monitor
• Shake a particular bunch (or bunches) at
• a fixed frequency
• Measure the BBI induced amplitude in the opposing
  bunch
• Provides much faster response than CLEO
  luminosity measurement
• Adjust differential offset between e- and e
  bunches at IP (VCROSING 7 Knob)
• Vary betatron phase advance in the vertical
  separator bump at the 2nd IP
• Optimize collisions for each car
• Observations
• Car-to-car orbit differences at the 0.5s level
  (sv _at_ 4mm)
• Strong dependence on beam current
• Consistent with machine operators having to
  actively tune VCROSING 7 through the course of a
  run

Increasing time c decreasing current
9
Bunch Luminosity Optimization

• Verification of BBI Luminosity Monitor
  performance relative to the CLEO Luminosity
  Monitor
• Optimize BBI signal for a particular car at the
  beginning of run
• Integrate luminosity for approximately 1/2 hour
  and analyzed CLEO bunch luminosity

Results appear consistent given strong current
dependence of differential orbits
10
Bunch-to-Bunch Orbit Correction

• DC pedestal of the vertical feedback system
• Measures orbit of all bunches simultaneously
• Feedback monitor point located 1.16 wavelengths
  from IP
• Assuming no kicks between the monitor and the IP,
  obtain position at the IP by scaling measured
  positions with bip/bfm
• Complications
• Current dependence
• Bunch-to-bunch X-talk
• Feedback Kicker
• Modifications to allow
• bunch-by-bunch
• deflections
• Present system capable
• of 0.5 mm corrections

11
Bunch-to-Bunch Orbit Correction (2)

• Preliminary test of feed forward kicking during
  normal operations
• Measured relative differential displacements
  using BBI monitor technique
• Applied fixed feed forward kick using vertical
  feedback system
• Monitor CLEO bunch-by-bunch luminosity for one
  weekend of running

12
Bunch-to-Bunch Summary

• Have observed bunch-by-bunch specific luminosity
  variations at the 15-25 level
• Direct measurements of differential e-e-
  displacement suggests 0.5sv offsets
• The luminosity degradation cannot be explained by
  simple displacement alone (would require
  0.8-1.1sv offsets)

L L0 exp-(dy)2/4sy2

• This suggests that the poor specific luminosity
  of the worst bunch is probably due to a
  combination of effects such as blowup of the beam
  envelope in addition to a simple differential
  displacement of the electron and positron
  trajectories
• Work continues
• Improvement of measurement techniques and
  simulation
• Further modifications to feedback system to
  increase the available kicking strength for
  corrections at the 1sv level

13
Summary

• Long Range Beam-Beam Interaction at Parasitic
  Crossings
• Induces spread in horizontal and vertical tunes
  of O(1 kHz)
• Distorts vertical closed orbit of individual
  bunches
• Ongoing Efforts
• Improved monitoring of bunch-by-bunch effects
• Modifications to vertical feedback system to
  allow bunch-by-bunch correction of differential
  (e-e-) vertical orbit displacements
• Radiofrequency quadrupole for bunch-by-bunch tune
  correction
• Installation of a superconducting IR which will
  provide better bunch separation at the parasitic
  crossing point nearest the CLEO IR
• Beam-Beam Tuneshift
• Have observed a beam-beam tuneshift of nearly
  0.07 while running with 9 trains of 4 bunches
• Poorer performance in 9x5 running is consistent
  with the poor performance of the worst bunches