Transcript Ikegami-san

Consideration on the low-
section of muon linac
5th meeting on muon g-2/EDM experiment
Mar. 25, 2009
Masanori Ikegami
KEK
At the previous meeting, …
• Two options were proposed;
– With longitudinal focusing (E0 or s ramped)
• E0 or s are ramped to keep the initial momentum spread.
– Without longitudinal focusing (On-crest)
• “On-crest” option were selected, but it was based on
an incorrect simulation result.
• Other conclusions includes…
– Space-charge effects are negligible.
– Choice of 324 MHz might have advantages.
– Higher E0 than J-PARC DTL should be pursued because of
its shorter pulse length.
At this meeting, …
• Another option is proposed;
– With longitudinal focusing but no RF-ramping
• E0 and/or s are NOT ramped to be matched with
the initial distribution.
• Injected beam undergoes a mismatch oscillation
during the acceleration.
• Can we adjust the phase advance for the entire
low- section so that the beam has the minimum
momentum spread at the exit?
• Filamentation due to the nonlinearity of RF forces
should be carefully evaluated.
Assumptions
• Muon generation with
–
–
–
–
40,000 muons per bunch
Isotropic momentum spread of 3 keV/c
RMS size of 2.5 mm
RMS pulse length of 3 psec
• Ideal initial acceleration to =0.08
– Simply adding the longitudinal momentum keeping
the RMS beam size, RMS pulse length, and the
momentum spread.
=0.08 corresponds to 3 MeV for proton.
Comparison with J-PARC MEBT
x’ [mrad]
MUON
E [keV]
MUON
E0  0.34MeV
0.36
-2.5
-0.36
0.24
2.5
-0.35
x [mm]

2.6
x’ [mrad]
 0.025 mm•mrad
E [keV]
16
J-PARC


-1.1
-2.6
1.1
 
x [mm]
2.9 mm•mrad
 [deg]
 z  0.084 keV•deg
  0.89 mm•mrad
J-PARC
0.35
-0.24
-5.8
E0  3MeV
5.8

-16
z 
 [deg]
93 keV•deg
 3.15 mm•mrad
Main parameters for a tentative design
•
•
•
•
•
•
•
•
•
•
Cavity type: DTL
RF frequency: 324 MHz
Input energy: 0.34 MeV
Output energy: 44.78 MeV
Average field E0: 5 MV/m (2 x J-PARC DTL)
Synchronous phase s: -30 deg
Number of cells: 32
Total length: 13.3 m
Focusing lattice: FODO
Quadrupole setting: Equipartitioning
Gradient of 1st DTQ = 1/4 x J-PARC DTL
Matched envelope
Momentum spread is
increased.
Horizontal
Vertical
Longitudinal
Beam size = rms x sqrt(5)
Beam envelope with assumed initial
distribution
Horizontal
Vertical
Longitudinal
Beam size = rms x sqrt(5)
Possible improvements
• Transverse focusing should be much weaker to
reduce mismatch.
• Triplet lattice might be effective in reducing the
mismatch.
• Longitudinal focusing should be weaker to
reduce mismatch.E0 or s ramped design
• Longitudinal momentum spread is increased
without RF ramping. One way to restore small
momentum spread is to introduce a debuncher
system.
Matched envelope with debuncher
Momentum spread is
restored.
Horizontal
Longitudinal
Vertical
20m drift
Beam size = rms x sqrt(5)
With assumed initial distribution
Momentum spread is
increased.
Horizontal
Longitudinal
20m drift
Vertical
Beam size = rms x sqrt(5)
With assumed initial distribution
(debuncher voltage adjusted)
Momentum spread is
restored by adjusting
debuncher voltage
(5%).
Horizontal
Longitudinal
20m drift
Vertical
Beam size = rms x sqrt(5)
Summary
• Options for longitudinal design
– E0 or s ramped design
• Potential technical difficulty in introducing significant ramping
– No-ramping design with debuncher
• Longer total length
• Necessity of debuncher tuning
• Options for transverse design
– Matching with stronger focusing
• Beam-based matching?
• Profile monitor
– Weaker focusing to reduce mismatch
• Larger aperture  shunt impedance, field uniformity
• Triplet lattice?