Velocity bunching experiment SPARC

1
Velocity bunching experiment _at_ SPARC

• Daniele Filippetto...

2
... On behalf of SPARC team
3
Outline

• The velocity bunching concept
• SPARC hardware overview
• VB experiment _at_ SPARC
• Emittance degradation by solenoid
  misalignment
• Conclusions

4

• The velocity bunching concept

If the beam injected in a long accelerating
structure at the 0 crossing field phase and it is
slightly slower than the phase velocity of the RF
wave , it will slip back to phases where the
field is accelerating, but at the same time it
will be chirped and compressed.
Compression and acceleration take place at the
same time within the same linac section, actually
the first section following the gun, that
typically accelerates the beam, under these
conditions, from a few MeV (gt 4) up to 20-25 MeV.
5
Phase space rotation
S. G. Anderson et al., Velocity bunching of
high-brightness electron beams, Phys. Rev.
ST-AB, 8, 014401 (2005).
6
Peak current vs RF compression phase
SPARC nominal case Initial parameters 1 nC
beam 10 ps long
7
... What happens in transverse plane?
If the transverse emittance can be preserved
during the longitudinal focusing, the beam
brightness can be increased
!!! RF Compression is made at lower energy than
magnetic compression. Can be used at SPARC to
shorten the beam and use it also for other
applications foreseen at SPARC (ICS, THZ
production, etc... )
8
SPARC overview
9

• Iron joke (blue) for field lines guiding
• 1 single and 4 triplet coils surrounding two
  LINAC section,
• indipendently powered

10
Diagnostic hardware
Spectrometer system T14 deg Lm26.7cm Ld2.13m
Pixel size30um Energy resolution 8keV/pix _at_
150MeV Overall resolution (RMS) 10-4 DE/E
10-2
11

• VB run _at_ SPARC (high charge case)

Laser parameters
Xrms 358um Yrms 350um TFWHM7.3ps Energy _at_
cathode 170uJ
Gun parameters
Gun Input Power7.5MW Gun Peak
Field105MV/m e-Energy out of the
gun4.7MeV Working inj.phase30 deg. e-beam
charge _at_30deg280pC
12
C-Factor Vs RF compressor phase
First linac section used as compressor
240fs rms
Maximum energy
13
E-beam parameters _at_ LINAC exit, C1
Max energy on crest 147.5MeV Total
DErms 0.16MeV DE/Erms
0.11 Charge 280pC Bunch Length RMS
3.01ps Slice Peak Current 30Amps Longitudinal
emittance 159.6 keVmm
Beam slice current profile

• The slice is chosen to be the FEL slippage length
  (Nupx?r210µm Sl250um)
• The sliding slice approach is used in data
  analysis

14
Effect of solenoid TW solenoids OFF Vs ON
(66Amps, 590Gauss) C1
TW sol on
TW-SOL on ex1.85um ey1.65um
Best emittance after solenoid scan with TW-SOL
off ex1.4um ey1.5um
Isol161 A
Isol158 A
15
TW solenoids Off VS ON, slice emittances

• The solenoid misalignment leads to an increase of
  the projected emittance, which is
  not found looking at the slice emittances
• the mismatch parameter is similar in the two
  cases
• The difference is due to slice centroid
  misalignement (will be treated more in detail
  further on in the presentation)
• A beam based alignment is mandatory to reach
  lower projected emittances

16
Beam after compression _at_C3

• Beam loses more than 45MeV
• Increase of correlated energy spread
• Increase of slice current by factor approx 4

17
Beam after compression _at_C3
Emittance without TW solenoids (Gun solenoid
current157Amps) Ex6.2 mm mrad Ey4.0 mm mrad
For a compression factor C3 Gain of a factor
3.7 on the maximum slice current (30 Vs
110) Loss of a factor 1.15 on the minimum slice
emittance (1.2 Vs 1.4) Gain of a factor 2.7 on
the slice Max Brigthness (0.41 Vs 1.1x1014)
B/C0.9
18

• Low charge/max Compression Case

Bunch Charge 60pC Bunch
length rms 1.95
ps Longitudinal emittance 54.2
keVmm Laser spot size rms
250um
19
Beam _at_ C-17 (TW sol 45Amps)
Energy97.6 MeVDE/Erms 1 Ipeak217.5
Amps Ex1.52 mm mrad Ey1.62 mm mrad
Proj. emittance
B 2x1014 Amps/m2
TW solenoids OFF Gun sol Current(151Amps) Ex4.1
mm mrad Ey3.4 mm mrad
! Slice length equal to the minimum time
resolution of the system (32fs)
20

• Critical point
• Proj. emittance degradation due to solenoids
  misalignment

• The solenoid force is energy dependent
• KLqB0/2m0cß?
• strong energy-time correlation in VB conditions
• different focusing forces for different time
  slices
• if the beam is propagating off axis respect to
  the magnetic field, the slice centroids will
  experience time dependent kicks

Lower Energies
higher Energies
Induced longitudinal-transverse correlation,
proj. emittance increase
21
Example 1mm solenoid misalignment (H)
Effect on transverse beam shape along the Linac
measurements
Out linac1
On crest
VB conditions
Out linac2
PARMELA runs simulating the two TW solenoids 1mm
off axis respect to the rf cavity, on crest and
in the VB conditions
VB conditions
22
Effect on emittance measurement
QS for projected emittance
QS for slice emittance
same quad currents
higher emittance value
Simulated X e Y vs phi at linac output
X-phi
Y-phi
Beam dimensions
Quad strength
23
Slice centroid spread exclusion Projected
emittance from slice
an, ßn, ?n, en twiss parameters of slice n
Dattoli et al. Slice Emittance, Projected
emittance and properties of the FEL SASE
radiation, MOPC32, this conference
24
Slice centroid contribution to the emittance
M.Ferrario, V.Fusco, M.Migliorati, L.
Palumbo,Int. Journal of Modern Physics A ,Vol 22,
No. 3 (2007) 4214-4234
uses the slice centroid
different from 0 only if slice centroids do not
lie on the same axis
correlation between slice centroid spread and
single slice dimension in ph.sp.
exenv0
excent0
excross?0
25

• From the slice emittance with the quad scan, the
  values of alpha beta and emittance for each
  slice are calculated at one precise position
• From the QS measurements also the system for
  slice centroids (both in X and X) can be written
  and solved (first order system)
• All the 3 emittance terms can be calculated

26
Conclusions

• Demonstrated transverse emittance preservation
  in the VB regime for medium compression
  factors
• At extreme compressions the emittance is not
  perfectly compensated but still good enough to
  further accelerate the beam and use it for FELs.
• Higher total energy spreads make the beam
  emittance sensible to magnetic components
  misalignment (quads, sol., etc...)
• The centroid emittance contribution can be
  isolated and measured

Next steps

• BBA on TW solenoids
• emittance studies as function of TW solenoid
  fields
• Longitudinal phase space detailed studies
• FEL radiation experiments with velocity bunched
  beams

27
RFD
RFD off
SPARC typical parameters
28
RF compression experiments in the world
29
Operational consideration on Machine tuning
sx555.6 (3.0) mm
Isol157 A
Not VB data
sx515.5 (0.93) mm
Isol158 A
sx595(1.3) mm
Isol160 A
30
Emittance degradation induced by chromatic
effects in quadrupoles
Single quadrupole
Quadrupole doublet