Preliminary Results of the

1
Preliminary Results of the LEDA Beam-Halo
Experiment
R. Garnett, T. Wangler, J. Qiang, and P.
Colestock Los Alamos National Laboratory SNS
Mini-Workshop, ORNL June 25-26, 2001
2
Halo Experiment Scientific Team
P. Colestock J. D. Gilpatrick M. E. Schulze H. V.
Smith T. P. Wangler C. K. Allen K. C. D. Chan K.
R. Crandall R. W. Garnett W. Lysenko J. Qiang J.
D. Schneider R. Sheffield
3
Experiment Motivation

• Computer simulations and linac operating
  experience show that a small fraction of
  particles can acquire a large transverse energy
  to form beam halo. We want to validate our codes.
• Search for evidence that halo formation is
  described by the particle-core model. This
  implies that parametric resonances exist to drive
  particles to large amplitudes. Our codes should
  contain the correct space-charge physics.
• Demonstrate that the simulation codes predict the
  mechanisms that produce beam halo and predict the
  maximum amplitude correctly. We hope to be able
  to say we know how to select design parameters
  including apertures that will minimize beam loss
  and radioactivation.

4
Envelope Modes of Mismatched Bunched Beams
y
y
Symmetric (Breathing) Mode
x
z
Antisymmetric Mode (not measured)
Quadrupole Mode
5
Beam-Halo Experiment

• 6.7-MeV, 75-mA pulsed beam from LEDA RFQ
• 30-msec pulse, 1-Hz rep-rate (slow data
  acquisition)
• FODO transport line with 52 quadrupoles beam
  diagnostics.
• First 4 quadrupoles used for matching /
  mismatching.
• Wire scanners and scrapers used to make profile
  measurements. Large dynamic range 1051.
• Vary mismatch and current. Measure and compare
  with codes 1) Rms emittances, 2) Maximum
  detectable amplitudes,
• 3) Kurtosis (beam shape parameter).
• Data taken for Breathing and Quadrupole Modes
• I16 mA, 50 mA, 75 mA
• 0.65 ? ? ? 2.00
• 0.71 ? ?/?o ? 0.93
• Also search for additional halo from other
  sources.

6
Fully-Instrumented LEDA Beam-Halo Lattice
11 meters
52 quadrupole FODO lattice
RFQ
HEBT
BPMs
PMTs
RWCs
T
Steerers
WS 4 /HS
WS 45,47,49,51 /HS
WS 22,24,26,28 /HS
52 Quadrupoles 4 in the HEBT 9 Wire
Scanners/Halo Scrapers (Projections) 1 in the
HEBT 3 Toroid (Pulsed Current) 2 in the HEBT 5
PMT Loss Monitors (Loss) 2 in the HEBT 10
Steering Magnets 2 in the HEBT 10 Beam Position
Monitors (Position) 5 in the HEBT 2 Resistive
Wall Current Monitors (Central Energy)
First 4 quadrupoles independently powered for
generating mismatch modes.
7
LEDA Facility Halo Lattice
8
Wire-Scanner / Scraper is our main halo
diagnostic tool (J.D.Gilpatrick, et al.).

• Wire-scanner uses 33-mm carbon fiber to measure
  core.
• Scraper is graphite plate brazed onto copper.
  Scraper measures halo-Graphite is 1.5 mm thick
  so protons stop in graphite.-Scraper bias
  voltage about 10V to suppress secondary electron
  emission.-Copper is water cooled.
• Dynamic range - 1031 for wire alone (4 rms),
  1051 for wire scraper (5 rms)
• Wire and scraper data acquired separately and
  combined later.

Scrapers
9
Wire-Scanner / Scraper DataMatched Beam, 75 mA,
WS22x
10
Beam is quickly de-bunching.
RMS Beam Sizes, 75 mA
11
Measurement Cycle

• RF blanking pulse de-energizes RFQ.
• 75-keV beam injected into unpowered RFQ as
  injector beam approaches steady state.
• RF blanking pulse is removed and RFQ is excited.
  (T5 ms rise time)
• Beam profile monitors are in fixed position so
  only one wire or scraper is in beam at a time.
  All other wires or scrapers are outside beam pipe
  aperture.-Wire or scraper collects beam-induced
  charge over about 30 ms-30-ms limit is set by
  onset of thermionic emission of the scanner
  wire.-Integrated charge is digitized. -Only
  last 10 ms of integrated charge is selected for
  data.
• After 30-ms interval, dc injector turned off.
• Before next pulse (1 Hz), scanner wire and
  scraper are moved to next position.

12
Procedures for Matching and Mismatching

• Initial matched-beam quad settings determined
  from TRACE3D.
• Quad scans used to determine partial derivatives
  of the beam size as functions of the matching
  quad strengths.
• Least-squares fitting procedure is used to
  determine quad settings that produce equal rms
  sizes in x and y at WS 22-28 (middle of the
  channel) The beam is rms matched.
• TRACE3D is used to determine quad values that
  give the appropriately scaled ellipse parameters
  at WS 22-28 for the pure-mode mismatches.
• Mismatch parameter, ? ? ratio of initial
  mismatched rms size to matched beam rms size.

13
Matching procedure appears to work!
14
Other Beam Measurements / Calculations

• Rms Emittance Calculated from quad scan data
  using a least- squares fitting procedure and rms
  beam sizes measured at the profile monitors
• Maximum Detectable Amplitude Determined from
  intersection of transverse profile curve with
  background noise level.
• Shape of Distribution Characterized using a
  kurtosis parameter defined in terms of ratio of
  4th moment to 2nd moment of the measured
  distributions .

15
Combined Data - 75 mA, ?1.0 and ?1.5
?1.0
?1.5
Scanner 22
Scanner 22
Our simulations do not predict distributions like
this.
Scanner 51
Scanner 51
16
Max. amplitude vs. z shows good agreement with
simulations.
Measurement Results
Simulation Results
17
Maximum Amplitude Particle-Core Model and
Simulation Predictions, 100 mA
18
Maximum AmplitudeComparison of Data and
Simulation
WS/HS 51 Data

• Results independent of beam current (tune
  depression).
• Code correctly predicts trend of the data
  including minimum for the matched-beam case.
• Supports particle-core model.

19
Comparison of Transverse Emittance vs. z
IMPACT Simulation Results
Measurement Results
Simulations using the nominal RFQ output beam
from PARMTEQM appear to underpredict the measured
emittance growth.
20
Kurtosis Definition
-Similar definition applies for y and z
coordinates. -Dimensionless shape parameter.
Independent of beam intensity. -Easily calculated
from moments of measured or simulated beam
profiles. -Zero for uniform-density 2-D
elliptical (KV) or 3-D ellipsoidal beam. -Equal
to or near unity for Gaussian profile. -Matched
beams without halo have values between 0 and 1.
Increases as tails develop, but can decrease and
go negative if beam profile becomes square.
21
X-Kurtosis Data - 75 mA

• General trends -
• -Kurtosis decreasing as a function of ?.
• Plausible since distribution is seen to
  become square.
• -No particular z-dependence.
• Kurtosis may not be a good figure-of-merit for
  halo.

22
Low-energy tail on RFQ output beam?

• Assumed PARMTEQM output distribution for core
  (100 keV energy spread).
• Superimpose 10 low-energy tail
• Uniform distribution
• 6.0 MeV ? W ? 6.7 MeV
• Assume low-energy beam is mismatched
• ? 1.5 to get agreement with measured widths.
• Good agreement with data.
• Seems physically plausible.
• Channel transports beam as low as ?2 MeV.
• Preliminary measurements do not support
  hypothesis.

?1.5, 75-mA Simulation Results
23
Matched beam distribution shape also not
predicted by nominal PARMTEQM distribution.
Measured Profile Matched Beam, WS 22
Simulation Results Matched Beam, All WS
24
Transverse tails in RFQ output beam explain our
observations?
Simulation Results
Artificially superimposed tails
Shoulders produced at Wire-Scanner 49
X-profile
Y-profile
25
Summary

• Measurements have been made at 15, 50, and 75 mA.
  Initial analysis for 75 mA data.
• Rms-emittance grows along the channel Growth
  rate increases as mismatch increases.
• Kurtosis decreases with increasing mismatch
  strength as shoulders develop.
• Profile measurements are not in good agreement
  with simple multiparticle simulations using the
  nominal output beam from the RFQ.
• Unexpected halo structure shoulders and
  asymmetries
• General trends of the data are predicted by the
  codes and the Particle-Core Model.
• Better understanding of profile data will require
  better characterization of RFQ output beam
  spectrometer, slit collector emittance
  measurement?