Compound bucket study

1
Compound bucket study

• Recycler Meeting
• May 2, 2007
• A. Shemyakin, C. Bhat, , D. Broemmelsiek, M. Hu

2
Motivation

• We are not using the full strength of e-cooling,
  keeping the e-beam far off axis most of the time,
  because of life time degradation
• We need the stochastic cooling only to cool
  tails, and it becomes too weak at large
  intensities
• The idea to separate in space the tails and the
  core
• E-cool holds the core and creates a correlation
  between the longitudinal and transverse tails
• Tails are cooled by a high-gain transverse
  stochastic cooling outside the core
• The hope was that pbars with a high transverse
  action would be heated by IBS longitudinally
  fast enough to go to the tail (hot) area before
  being lost transversely

3
Tails correlation in e-cooled beam- interpretation

• Two reasons for the poor e-cooling of the
  transverse tails
• Tail pbars spend less time inside the electron
  beam
• Increased transverse pbar velocities inside the
  e-beam

Measured radial dependence of the drag rate (20
24 Feb 2006, L.Prost).
A simple model of a drag force vs action Force
average of the measured F(r) over phases
?e2 /(?e2 ?p2) A side note we have to try
increasing the effective e-beam size.
4
Cooling of longitudinal and transverse tails

• All with small transverse action pbars captured
  inside the bucket are cooled
• The drag force drops by 3 times at 20 MeV/c
• The drag force drops by 10 times for pbars with
  action 10 ?, which are still far from being lost
  (RR acceptance 40?)

5
Tails correlation in e-cooled beam
In the time of transverse scraping, the
longitudinal momentum spread of an e-cooled beam
decreases dramatically. The effect depends on
how long the electron cooling has been applied.
No similar data for a stochastically-cooled beam
were found.

• Longitudinal Schottky profiles in the time of
  scraping (L. Prost, 31-Mar-07).
• Green- before scraping, 11.E10
• Red- after the horizontal scrape, 7.E10
• Yellow- after vertical scrape, 3.E10.

6
The study (24-Apr-07) RF structure

• RF and RWM (1503) profiles.
• (1)-entire beam, (2)- cold area, 165 bckt (3)-
  hot area, 192 bckt.
• Main burrier length 48, mini-barriers 13
  bckt.

3
2
1
7
The study history

• Immediately after the transfer, the pbar beam was
  moved into a desired longitudinal position and
  set to the desired length
• Np 250 E10
• Before growing the mini-buckets
• dPsig 5.3 MeV/c
• FW 5.3 ?
• Transv. Schottky emittance 6.8 ?
• 13-bckt-width, full-amplitude mini-buckets were
  grown
• The width of the cold bucket (165 bckt) was
  chosen to fit inside the standard bucket squeezed
  for a transfer of 4 batches
• At the same time, e-beam was turned on (0.1 A, on
  axis)
• Transverse stochastic cooling was gated to
  outside of the cold beam
• Longitudinal cooling was off
• Turning on and initial tuning took for about an
  hour
• Was adjusted once more an hour later
• 3 hours after raising mini-buckets, the
  stochastic cooling was turned off
• In less than an hour after, the mini-buckets were
  removed, and normal operation resumed

8
Density distribution and life time
Portion of the beam in the cold area
RRWMD67
RBEAM

• Portion of the beam in the hot area dropped from
  35 to 5
• Intensity reported by RBEAM was changing up to
  3 (difficult to interpret)
• The life time estimated by RWMD47 stayed 700 hr
  for the first 2 hours after turning e-beam on,
  but dropped to 150 hr after that.

SC power

• New parameters (P. Derwent) integrals over RWM
  distribution for three areas
• (1)- entire beam, RWMD67 (2)- cold, RWMD47, (3)-
  hot, RWMD47

9
Transverse emittances

• The cold area portion behaved as it usually does
  for the case of e-cooling only
• FW emittances were steadily decreasing, while
  Schottky emittances remained nearly constant.
• The hot portion behaved as it usually does for
  the case of stochastic cooling only
• FW and Schottky emittances were close
• Cooling rate 2.3 ?/hr for (50- 20)E10
• Schottky emittance grew fast after turning
  stochastic cooling off
• Indication of a flow from the cold area?

• Indexes 1,2, and 3 corresponds to the entire
  beam, cold, and hot areas. Averages of H and V
  are shown.

10
Cooling by e-beam only

• After turning the stochastic cooling off
• The number of pbars in the hot area stopped
  decreasing
• The total transverse Schottky power stopped
  decreasing

11
Longitudinal Schottky data

• RMS Schottky momentum spread measured in cold (2)
  and hot (3) areas as well as ungated data (1).

See A. Burovs talk for the analysis of the
longitudinal dynamics.
12
Summary

• No major discrepancies with qualitative models
  were found
• 95 of the beam was cooled into the cold bucket
• Gated stochastic cooling worked
• A strong tail correlation in an e-cooled pbar
  beam fits into observations
• There was an indication of a flow of high-action
  particles from the core to the tail area
• Cooling was not faster than normal
• In 4 hours with mini-buckets, changes in
  emittances were
• From 133 to 48 eVs
• From 6.9 to 5.7 ? Schottky
• From 5.3 to 2.4 ? FW
• Optimum stochastic cooling was applied for 2 hrs
  only
• The hot area can be expanded
• There were indications of the life time
  degradation toward the end of the study
• Usual trend for e-cooling with e-beam on axis
• Agrees with a nearly constant transverse Schottky
  emittance of the cold area
• An explanation is a too slow rate of moving the
  high action pbars into the hot area