Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
CLIC ACE 2-4 September 2008 R. Corsini, J.B.Jeanneret, F. Stulle
Drive Beam Generation ComplexDrive Beam Generation Complex
Will talk about:
• The concept
• Previous work
• Critical issues
• Status of present activities
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
Electron beam manipulation
Long RF PulsesP0 , n0 , 0
650 Klystrons
low frequencyhigh efficiency
Power stored inelectron beam
Short RF PulsesPA = P0 N1
A = t0 / N2 A = 0 N3
143000 Accelerating Structures
high frequency high gradient
Power extracted from beamin resonant structures
The CLIC RF power source can be described as a “black box”, combining very long RF pulses, and transforming them in many short pulses, with higher power and with higher frequency
What does the RF Power Source do ?
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
Drive Beam Acceleratorefficient acceleration in fully loaded linac
140 s total length - 24 24 sub-pulses - 4.2 A2.4 GeV - 60 cm between bunches
240 ns
Drive beam time structure - initial
24 pulses – 100 A – 2.5 cm between bunches
240 ns5.8 s
Drive beam time structure - final
Power Extraction
Drive Beam Decelerator Sector (24 in total)
Combiner ring 3
Combiner ring 4pulse compression &
frequency multiplication
pulse compression & frequency multiplication
Delay loop 2gap creation, pulse
compression & frequency multiplication
Transverse RF Deflectors
RF Power Source Layout
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
Full beam-loading acceleration in TW sections
RF in
“short” structure - low Ohmic losses
RF Power Source “building blocks”
RF to load
most of RF power (≥ 95%) to the
beam
High currentbeam
No
No beam
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
P0 , 0
P0 , 0
Beam combination/separation
by transverse RF deflectors
RF Deflector,
Deflecting
Field
Transverse
02 P
0 , 2
0
RF Power Source “building blocks”
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
Counter propagation from central complex
Instead of using a single drive beam pulse for the whole main linac, several (NS = 24) short ones are used.
Each one feed a 800 m long sector of TBA.
(DLDS-like system)
Counter-flow distribution allows to power different sectors of the main linac with different time bins of a single long electron pulse. The distance between pulses is 2 LS = 2 Lmain/NS. The initial drive beam pulse length is equal to 2 Lmain= 140 s/c.
RF Power Source “building blocks”
pulse 2 pulse 1
main linacdecelerator sector
main beampulse
From central complex
12
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
odd buckets
even buckets Delay
Loop
RF deflector
Combination scheme
Gap creation & first multiplication 2
Ldelay = n 0 = c Tsub-
pulse
Phase coding
180 phase switch
Acceleration 0
Deflection 0 / 2
Sub-Harmonic Bunching 0 / 2
How to “code” the sub-pulses
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
666 ps
1 8
Fast phase switch from SHB system (CTF3)
8.5 · 666 ps = 5.7 ns
Streak camera – 500 ps/mm
… or use a laser + photo-injector
satellite
main
3 TW Sub-harmonic bunchers,each fed by a wide-band TWT
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
RF injection in combiner ring (factor 4 for simplicity)
3rd
o/4
4rd
2nd
Cring = (n + ¼)
injection line
septum
localinner orbits
1st deflector 2nd deflector
1st turn
o RF deflector field
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
A complete zero-order conceptual design on the drive beam complex has been published in ’99 (Yellow report CERN 99-06). Since then, the parameters changed (in general in a favorable direction) and some of the concepts evolved and/or were tested in CTF3.
It is now time to fully review all components.
•Some conceptual work on overall drive beam complex layout remains to be done
- drive beam phase stability, see later
•Drive beam injector
- CTF3 a good example, RF injector preferred (no satellites) – to be tested in CTF3 as well
•Drive beam accelerator
- design existed for previous parameter set, needs to be redone (no major problems)
•Delay loop and combiner rings
- design needs to be revisited, modified and evaluated
- this is a critical item
•Bends into long transfer line
- does not exist but should be relatively easier than other bends
•Transfer lines and turn-around into decelerator and compressor/feedback
- baseline exists
•Drive beam decelerator
- design exists, some improvements, cost driver
Status (derived from Daniel`s talk)
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
1404.27.8
L. RinolfiLC ’99 Workshop
Injector
Scaling from CTF3 & new parameters reassuring, but satellites problem, stability…
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
• Smaller transverse emittance
• Shorter bunches, no energy tails
• No satellites
• Issues (to be tested in CTF3)
• High charge, long pulse laser…• Stability
Injector – RF gun option
PHIN test stand in former CTF II
PHIN RF gun
CTF3 thermionic injector
RF gun implementation in CTF3
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
Accelerator
CTF3 basically OK
In CTF3~ 1mm from
injector
Need full study (phase stability)
R. CorsiniLC ’99 Workshop No big problems expected, but design needed for realistic start-to-end
simulations (especially longitudinal gymnastics/phase stability)
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
R. CorsiniLC ’99 Workshop
favorable param. scaling
CTF3
CTF3
Delay Loop & Combiner Rings
to be checked
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
B. Jeanneretwork-in-progress
~ +/- 0.2 mm /arc at +/- 2%
~ +/- 0.3 mm /arc at +/- 2%
lattices, 2nd order isochronicity, chrom. control
R. CorsiniCERN 99-06
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
6 4 2 0 2 4 61.12
1.14
1.16
1.18
1.2
c t (mm)
U (G
eV)
1 1.5 2 2.5 3 3.51 10
6
1 105
1 104
1 103
0.01
Beam Energy (GeV)
Tota
l Rel
ativ
e E
nerg
y L
oss
Energy loss from SR and CSR
Both rings - = 10 m - 27 m
= 2 mm, Qb = 7.8 nC
CSRShielded (h=20mm)
SR
total
NB: I have kept the same in the rings as the old parameter set – now we need only a factor 2 in final compression
can accept more longitudinal phase space distortion will be less sensitive to energy variation for drive beam phase stability
6 4 2 0 2 4 61.12
1.14
1.16
1.18
1.2
c t (mm)
U (
GeV
)
‘99 parameters
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
Another ring issue – transverse stability in the RF deflectors
Optimum ring tune
D. Schulte – R. Corsini
‘99 parameters
OK in CTF3, but vertical polarity (trapped) critical – should be OK with damping or couplers
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
Old & last 30 GHz parameters
Drive beam current initial 5.7 A 4.2 A
Drive beam current final 181 A 100 A
DB bunch charge 12.1 nC 8.4 Nc
Drive beam energy 2.4 GeV ( 240 MeV) 2.4 GeV ( 240 MeV)
DB acceleration frequency 0.937 GHz 1 GHz
DB bunch frequency initial 0.46 GHz 0.5 GHz
DB bunch frequency final 15 GHz 12 GHz
DB pulse length initial 94 s 140 s
DB pulse length final 70 ns 240 ns
Combination factor 2 4 4 = 32 2 3 4 = 24
Number of sectors/linac 21 24
Sector length 670 m 868 m
Length delay loop/line 21 m 72 m
Length combiner ring 1 84 m 145 m
Length combiner ring 2 334 m 434 m
Rms bunch length final 400 m 1 mm
Power per PETS 640 MW 136 MW
2005 - CLIC Note 627 August 2008
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
1) INJECTION LINAC
a. RF Structures
b. Linear optics
c. Collective effects
d. dE/dz correlation for later compression
2) BUNCH STRUCTURE to build trains (rise time / flat section / Fall time)
3) DELAY LOOP
a. Isochronous & achromatic optics, beta-beating control
4) COMBINER RINGS
a. Isochronous & achromatic optics, beta-beating control
5) RF DEFLECTORS FOR DL, CR1, CR2
6) TRANSFER LINE DOWN TO TUNNEL
a. Optics
b. layout
7) FINALIZE THE LONG TRANSFER LINE
a. Optics & layout
b. Kick-out and matching to turnarounds
c. Beam-based alignment
8) TURN-AROUND TO DECELERATOR
a. Matching
b. Layout & compatibility with other beam lines
9) EXTRACTION OF SPENT BEAM after decelerator
a. Optics
b. Compatibility with other beam lines
10) DUMP CONCEPTUAL DESIGN
11) TIMING Drive Beam / Main Beam
12) ALL THE ABOVE
a. Transverse feed-back
b. Longitudinal feed-back
c. Magnet specification and inventory (at least for cost)
d. Specification for vacuum
e. Specification for instrumentation
f. Machine protection
g. Collimation
Task list B. Jeanneret
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
• Optics
• Parasitic dispersion + |dp|=2% => emittance dilution
• Logistics
• Beam line suspended to the ceiling
• Light & transversely thin magnets preferred
• Cost & power
• Weak magnets preferred
• Collective effects
• Trains : 100 A , 73 m long
• Strong transverse multi-bunch resistive wake-fields
• Ions => detuning & instabilities
Deserves global optimisation
Issues :
DB long transfer line : simple optics but 21 km long
B. Jeanneret
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
B. Jeanneret
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
B. Jeanneret
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
B. Jeanneret
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
B. Jeanneret
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
B. Jeanneret
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
B. Jeanneret
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
B. Jeanneret
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
Bunch Compressor Chicane 10 m
Turn Around Loop 77 m = 1x 60deg arc + matching + 3x 80deg arc
Bunch Compressor Chicane 20 m
1m 2.5m 1m
10.02 deg
10m
0.75m 8m 1m
5.53deg
20m
R56=−0.2 m
R56=−0.16 m
Phase measurements
Layout of final BCs & turn-around loop
F. Stulle
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
Bunch Compressors BC1 and BC2:
BC1 is not only a bunch compressor, but is also used to convert an incoming
energy jitter into a measurable phase jitter
for the energy jitter measurement its R56 should not be too small:
the phase error measured in front of the loop is corrected in BC2 by changing
the path length of the bunches
its R56, i.e. the bending angles, should be large enough
to allow the usage of weak correction kickers:
the influence of ISR is small due to the huge beam emittance
and the rather low electron energy
CSR (transverse) is also rather easy to control due to the huge beam emittance
Design Considerations / Constraints - BCsF. Stulle
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
Turn Around Loop:
the turn around loop has to be achromatic and should be isochronous
it has to be compact and simple since 2 x 24 loops are required
the R56(s) should stay close to zero, since the bunch has an energy chirp
it might be compressed to short lengths, i.e. it might radiate a lot CSR
this is the main complication for the lattice design,
a trade-off has to be made between R56(s), T566(s) and chromaticity
the influence of ISR is small due to the huge beam emittance
and the rather low electron energy
CSR (transverse) is also rather easy to control due to the huge beam emittance
Design Considerations / Constraints - TALsF. Stulle
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
Turn Around Loop latticeF. Stulle
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
longitudinal
phase space
transverse, hor
phase space
initial final
CSR in Loop
T566 in Loop
Simulation Results, 1D CSR
F. Stulle
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
• The last extensive and coherent study of the drive beam complex has been completed in ’99 (Yellow report CERN 99-06).
• Since then, the parameters changed (many times).
• New parameters are in most respects much more favorable than old ones.
• Still, it is now time to fully re-design and assess all components, including the ones neglected at the time. Work just started.
• The experience accumulated since then in CTF3 extremely useful. It relieved many worries we had, and raised a few new ones. Past and future studies in CTF3 are a valuable input to steer the new studies.
• Some of the critical issues (beam power and losses management, machine protection, stability...) need strong effort – scarce at present. This is especially important since they are addressed only partially in CTF3.
CONCLUSIONS
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
Reserve slides
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
Since the power per sector is fixed, the product Ibeam Ebeam is constant
For: is good to have
Transverse beam stability in decelerator high current, large aperture, long PETS
Drive beam combination in rings small current, high energy (but below about 2.5 GeV)
Drive beam current & energy trade-off
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
Several issues can put an upper limit to the beam energy in the combiner rings:
•Too high field in magnets not an issue (long rings – long RF pulse length)
•Synchrotron radiation:
•Energy loss not limiting
•Power loss in vacuum chamber potential limit
•Energy spread & emittance increase negligible
•Coherent synchrotron radiation
•Beneficial effect
•Deflectors
•Higher power for given angle
•Constant power from damping of real emittance → neutral
Drive beam energy limits
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
DRIVE BEAM LINAC
CLEXCLIC Experimental Area
DELAY LOOP
COMBINERRING
CTF3 – Layout
10 m
4 A – 1.2 s150 Mev
30 A – 140 ns150 Mev
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
CTF3 – R&D Issues - where
fully loaded acceleration
recombination x 2
phase-coding
bunch length control
recombination x 4
bunch compression
PETS on-off
two-beamacceleration
structures 12 GHz
structures 30 GHz
deceleration stability
R1.1 – structures
R1.2 – DB generation
R1.3 – PETS on-off
R 2.1 – structure materials
R 2.2 – DB decelerator
R 2.3 – CLIC sub-unit
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
CTF3 – R&D Issues - when
structures 12 GHz
fully loaded acceleration
recombination x 2
phase-coding
bunch length control
recombination x 4
bunch compression
PETS on-off
two-beamacceleration
structures 30 GHz
deceleration stability
2008
2009
2010
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
MKS03 MKS07MKS06MKS05
Spectrometer 10Spectrometer 4
Measured RF-to-beam efficiency
95.3 %
Theory 96% (~ 4 % ohmic losses)
RF pulse at structure output
RF pulse at structure input
analog signal
1.5 µs beam pulse
SiC load
damping slot
Dipole modes suppressed by slotted iris damping (first dipole’s Q factor < 20)and HOM frequency detuning
Full beam loading acceleration stability - efficiency
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
666 ps
1 8
8.5 · 666 ps = 5.7 ns3 TW Sub-harmonic bunchers,each fed by a wide-band TWT
Streak camera image
mainsatelliteFast phase switch from SHB system
phase coding
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
Satellite control, RF gun option
• Better emittance
• Shorter bunches
• No satellites
• Lower current
Drive Beam Generation ComplexCLIC ACE2-4 September 2008 R. Corsini, J.B. Jeanneret, F. StulleR. Corsini, J.B. Jeanneret, F. Stulle
longitudinal
phase space
transverse
phase spaceinitial final
CSR in Loop
T566 in Loop
Simulation Results, 1D CSRF. Stulle