Post on 18-Mar-2020
ILC Project
Kaoru Yokoya
KEK
KEK Theory MeetingMarch 3, 2005
1
ICFA Decision
ICFA chose Superconducting Technology at ICHEP04 Beijingfollowing the recommendation of ITRP
ITRP Report lists advantages of SC
• Simpler operation
◦ Large cavity aperture ⇒ less sensitive to ground motion
◦ Large bunch distance ⇒ inter-bunch feedback
• Lower risk of main linac
• XFEL provides prototype
• Industrialization underway
• Less power consumption
2
The ITRP Report also states
• We are recommending a technology, not a design
• We expect the final design to be developed by a team drawn
from the combined warm and cold linear collider communities
What has to be done
• Form international collaboration structure
• Decide design choices and details
• Industrial design
• Cost reduction
3
Official Time Schedule
2004.8 Adopted ‘Cold’ at ICHEP, Beijing
2004.11 1st ILC Workshop at KEK
2005.2 Decide the director and location of Central GDI
2005. Establish Regional GDIs
2005.8 2nd ILC Workshop at Snowmass. Decide design outline
(acc.gradient, 1/2 tunnel, dogbone/small DR,
e+generation etc)
2005 end Complete CDR
2007 end Complete TDR, role of regions, start site selection
2008 Decide the site, budget approval
2009 Ground breaking
2014 Commissioning starts
4
Can TESLA be the baseline?
Yes, but still many alternatives remain after the SC/NC decision
• Accelerating gradient: 35MV/m or higher ?
• Tunnel: Single or double (or triple) ?
• Damping ring: dogbone or small ?
• Positron production: undulator or conventional ?
• Crossing angle: zero or small or large ?
5
Gradient
Result of questionnaire in WG5 in 1st ILC Workshop
25MV/m Technology in hand
35MV/m Needs essential work
45MV/m Future upgrade
6
7
Tentative 500 GeV Linac Parameters versus GradientTESLA USSC 30 MV/m 35 MV/m 40 MV/m
Ecms (GeV) 500 500 500 500 500N 1010 2.00 2.00 2.00 2.00 2.00Nb 2820 2820 2820 2820 2820Tsep (ns) 336.9 336.9 307.7 295.4 270.8Buckets 1.3GHz 438 438 400 384 352Iave (A) 0.0095 0.0095 0.0104 0.0108 0.0118Gradient MV/m 23.4 28.0 30.0 35.0 40.0Cavities/10MW klys 36 30 24 20 16Q0 1010 1.00 1.00 1.50 1.00 0.70Qext 106 2.50 2.99 2.93 3.28 3.44Tfill (us) 420.0 502.7 491.9 550.9 577.1Trf (ms) 1.37 1.45 1.36 1.38 1.34frep (Hz) 5 5 5 5 5Linac overhead 0 5% 5% 5% 5%Total no of cavities 20592 18096 16656 14240 12480Total no of klystrons 572 603 694 712 780Active 2 linac length (km) 21.6 18.7 17.3 14.8 12.9Total 2 linac length (km) 30.0 27.0 23.5 20.1 17.9Pb (MW) 11.3 11.3 11.3 11.3 11.3Pac (linacs) (MW) 94 105 104 121 142Luminosity 1034/cm2/s 3.0 2.6 2.0 2.0 2.0
8
Higher Gradient
• Must reach 1TeV
• Gradient gives an
impact on the site
length
• Presumably not an
issue for US and
Europe, but
• Is an issue for Japan.
• A few candidates>∼45km in Japan,
but prefer higher
gradient for flexibility
of site selectionSite length vs. Gradient for 1TeV
9
10
1/2 Tunnels?
• TESLA design adopts single
tunnel, accomodating
◦ Klystron
◦ Linac cryomodule
◦ 2 Damping Ring lines
• TESLA says
◦ Save cost ∼300MEuro
◦ Double tunnel has ground
motion problem
• But many problems of
operation
• Not a big impact on overall
design
11
Positron Production
-e
e+
AdiabaticMatchingDevice
IPto the
target
to
RingDamping
solenoids
Ti-alloy0.4 X0
250 GeV
undulator ~100 m
electronbeam
beamγ −
acceleratingstructure
Pros
• Polarized positron
• Single target
Cf: X-band e+ source requires 3
targets
• Better emittance of created
e+ beam
Cons
• Complex commiss./operation
(e+/e− operation coupled)
• Low energy operation
• IP energy spread of e− beam
0.05% ⇒ 0.15%
12
Conventional Positron Production
• Hit electrom beam on target rather than photon.
• Can decouple e+/e− beams
• Thick target (several rad length) ⇒ more energy deposit
• TESLA rejected the conventional method since the design start,
because of the large pulse charge (40 times X-band)
• But, later turned out 1ms pulse is long enough to prevent stress
accumulation.
• US estimation says comparable to X-band
• Target damage test being proposed at KEKB
◦ Ring total charge ∼ 10µC, close to TESLA pulse charge◦ Extraction in 10µs by existing abort system
(Extraction in 1ms requires advanced kickers)
• No polarized positron
13
Damping Ring
Problems of the Dogbone design
(a) Commissioning/operation
◦ DR commissioning only after linac completion◦ No DR tuning during linac repair
(b) Stray field from linac can disturb beam extraction from DR.
O(µTesla) matters.
(c) The long straight sections
◦ Space-charge force (Coulomb force within a bunch)◦ Long wiggler section needed ⇒beam stability range
(d) Fast kicker (<∼20ns) needed for injection/extraction
14
• (a)+(b) (coming from sharing tunnel) can be partially solved by
in the 500GeV stage (But not in 1TeV stage)
• Space-charge problem is solved by ‘coupling bump’ in theoreti-
cal/computer level
• Kicker under development at DESY (data as of Apr.2004)
spec. measuredRise time (10%-90%) 8ns 4.9nsMicro pulse rep rate 3MHz 2MHzMacro pulse rep rate 5Hz 5HzAmplitude stability 0.05% 1.2% (0.2% with 30 kickers)
Residual kick 0.5% 2.75%
15
Alternative Design of DR Is a smaller ring possible?
• Compress more the bunch interval
• Main motivation is to avoid interference with linac
• but not the cost issue
Numerous ideas on injection/extraction
• Stripline kicker
New kicker to be tested at
ATF by summer 2005
• Fourier kicker
N−1∑
k=0
ei(ω0+k∆ω)t
16
Difficulties of Small DR
• Even more ambitious kickers required
• Collective instabilities harder (higher beam current)
Also, note:
• We are going to adopt 35MV/m as baseline
• TESLA 35MV/m (800GeV) requires 4886 bunches (11.5ns in
DR), not 2820 bunches (20ns)
• Even more bunches preferred at higher gradient
(in order not to loose power efficiency)
17
Crossing Angle
• Basically no big difference from warm design
• except for items related to the bunch distance
◦ No bunch-to-bunch interference
◦ IP fast feedback easier
• Three different designs
Original TESLA zero crossing angle
GLC small angle (7mrad)
→ bunch-bunch interference (warm collider)
NLC large angle (20mrad)
• 2nd IP: e+e− and γ-γ compatible?
• Linac orientation is another issue
18
Energy Upgrade Senario
Assume the gradient G is established by construction start
Senario 1500GeV
1TeV>=
G
G G'
• 1 stage cheaper
• possible gradient improve-
ment in 2nd stage
• Can accomodate dogbone DR
in the empty half in 1st stage
Senario 2500GeV
1TeV
G/2
G
• Do not need to close the pro-
duction line in industries for
2nd stage
19
Parameter Subcommittee
• Formed under ILCSC in Apr.2003
• Members
S.Komamiya, D.Son (Asia)
R.Heuer (chair), F.Richard (Europe)
P.Grannis, M.Orglia (N.America)
• Report Sep.2003
20
• Energy
◦ 500GeV (at any energy 200GeV ≤ ECM ≤ 500GeV)
◦ Upgradable to ∼1TeV
◦ Energy stability ≤ 0.1%
◦ Calibration run needed at 90GeV
(L can be lower, long shutdown acceptable)
• Luminosity
◦ ∫ Leqdt ≥ 500fb−1 in first four years of running (year zero for
commissioning not included)
(Leq: 500GeV-equivalent value using L ∝ ECM)
◦ Design luminosity by year 4
◦ After first 4 years,∫ Leqdt ≥ 1ab−1 in 2 years (option)
◦ For ECM upgrade,∫ Leqdt >∼ 1ab−1 in 3-4 years
21
• P(e+)≥ 80%
• IPs
◦ 2IPs of similar energy reach and luminosity should be planned
◦ One IP must be compatible with γ-γ
◦ 2IP switch within a few % time loss
• Options
◦ e−e−
◦ P(e+)>∼50% for ECM ≥50GeV
◦ GigaZ: L >∼ several ×1033/cm2/s
◦ WW threshold: L >∼ several ×1033/cm2/s
Energy calibration <∼ a few 10−5
◦ γ-γ: L 30-50% of e+e−
22
History of SLClate 1970’ proposal
Oct 1983 construction starts
mid-1987 construction ends
Apr.1989 first Z0 event
1992 SLD data taking start (pol beam av.22%)
1983 flat beam, strained lattice (pol ∼62%)
1994-5 damping ring chamber and FF optics improved
1995 thin strained lattice (pol ∼80%)
mid-1998 end of run
damping ring
damping ring
e-
e+
SLAC
arc
collision
point
e+target
23
Design and Final Parameters of SLC
Design Achieved Units
Beam charge 7.2×1010 4.2×1010 e±/bunch
Rep. rate 180 120 Hz
DR εx 3.0×10−5 3.0×10−5 m·radDR εy 3.0×10−5 0.3×10−5 m·radFF εx 4.2×10−5 5.5×10−5 m·radFF εy 4.2×10−5 1.0×10−5 m·radIP σx 1.65 1.4 µm
IP σy 1.65 0.7 µm
Pinch factor HD 2.2 2.2
Luminosity 6×1030 3×1030 /cm2/sec
EPAC2000, Vienna ’SLC- The End Game’, R.Assmann et.al.
24
History of SLC Luminosity
LINAC1998, Monterey, CA, ’SLC Final Performance and Lessons’, N.Phinney
25
Can ILC Do Better than SLC?
• Most problems in SLC were related to alignment/control/diagnosticsrather than acceleration issues
• Remarkable progress on these issues through SLC and after SLC
◦ Diagnostics devices, many of them non-destructivewire scanner, laser wire, bunchlength, cavity BPM
◦ Progress in computer simulation
◦ Feedback systempulse-to-pulse fdbk, collision fdbk, algorithm
◦ Progress in tuning algorithm
• Wakefield effects much weaker than in SLC (true even withwarm collider)
◦ Wakefield affects the accuracy of simulation
◦ When SLC was designed, the linac beam was known to beunstable under signle-bunch beam breakup.
26
• There are still some beam dynamics problems pending. . .
◦ electron cloud and fast ion instabilities in DR
(Note: electron cloud was known before KEKB construction)
◦ stray field effects on DR
◦ banana control
but these are less serious than the predicted beam breakup at
SLC
• We will have more experience if ATF2 is constructed
27
Asian Organization for ILC Research
Working groups (same structure as in the ILCWS)
WG1 Overall design K. Kubo
WG2 RF H. Hayano
WG3 Injectors M. Kuriki
WG4 Beam Delivery System T. Sanuki
WG5 Cavities K. Saito
? Facility A. Enomoto
? Site R. Sugahara
Japan LC Office
F. Takasaki (chair), K. Yokoya, H. Hayano, N. Toge, S. Yamashita
28
Research in Asia
• Active participation from China and Korea
• Existing facilities and technology base(workshops in the labs, SC technology for nuclear fusion, etc.)
• Working Groups formed
• Many from China/Korea are visiting/going to visit KEK
• Possible production of components
IHEP contact person PAL Task Force Team
(Head) Won NAMKUNG
WG1 GAO Jie Jin-hyuk CHOI, Moo-yun CHO
WG2 CHI Yun-long Jong-seok OH, Won-ha HWANG
WG3 PEI Guo-xi Sung-ju PARK, Eun-san KIM
WG4 Wang Jiu-qing Jung-yun HUANG, Hee-seock LEE
WG5 ZHAO Sheng-chu Sang-ho KIM, Young-uk SOHN
29
Critical Reaserch Areas
• Establishment of the technology for acceleratinggradient 35MV/m
◦ Technologies for cost reduction
◦ Technologies for mass production
• Pursuit of possible higher accelerating gradient
◦ Larger operational margin of gradient
◦ Wider possibility of site selection
• Beam technology development using KEK-ATF
◦ Unique storage ring to reach ILC emittance
◦ Make full use of ATF capabilities
30
Remaining Problems on SC Technology (H. Hayano)
• Reliability of cavity gradient >35MV/m
• Complexity and cost of input coupler
• Rigidity of cavity-jacket relating to Lorentz detuning
• Reliability of tuner mechanism, Reliability of Piezo in cold
• Cavity alignment after cooling down
• Cost optimization of RF Waveguide System
• Cost optimization of cryomodule
31
Development of Higher Gradient
• Single-cell test soon
• 9-cell cavity design done. Fabrication started.
• Individual vertical test of four 9-cell cavities by Sep.2005
◦ Just in time for CDR completion
◦ In existing facilities (AR east)
◦ If expected performance not obtained,
⇒ change to slower plan for ILC 2nd stage
• Cryomodule test by end of 2006 ⇒ STF Phase 1
• Industrial design by TDR
32
STFSuperconducting RF Test Facility
• Linac R&D building for
JPARC
• Emptied by summer 2005
• 93m tunnel underground
• Reuse existing facilities
◦ Refregerator from AR east
◦ Power supply and
modulator
33
34
STF Phase 1 (2005-2006)
• Crymodule for 4 45MV/m cavities
• Crymodule for 4 35MV/m cavities
• RF source and cryogenic system (mostly recycled)
• Electron beam and its diagnostics system
• Synthetic test of 35MV/m & 45MV/m cryomodules with beam
and in parallel
• Construct cavity treatment facility in KEK
(can fabricate cavities also for overseas use and non-ILC use)
• Nb-Cu clad and seamless cavity fabrication for cost reduction
35
Targets of STF Phase 1
• Stable and reliable gradient 35MV/m with reasonable yield rate.
• Establish feasibility of 45MV/m
• KEK SC engineering experience
• First step to the industrialization
36
STF Phase 2 (2007-2009) still uncertain
• 3 Cryomodules each containing 12 cavities (35 or 45MV/m)
• Reinforcement of RF and cryogenic systems
• Synthetic test with a beam
Targets
• Industrial design
• Establish Asisn/Japanese technology for ILC component pro-
duction
37
Extension of ATF Extraction Line : ATF2• Stable collision of 5nm beams is essential in ILC
• SLAC-FFTB realized 60nm beam by ∼1997
• Extension of ATF extraction line can produce ∼35nm beam
This beam size itself is not a big progress but, with nanoBPM
technology being developed at ATF, a stabilization test down to
nanometer is possible
Time Scale
• Very useful for ILC design even after TDR
◦ Final Focus Design work will continue after ILC ground break-
ing (under constraints such as total length and crossing an-
gle)
• Can experience tuning work (Minimize ILC commissioning time)
38
39
Purposes of ATF2
(A) Small beam size
(A1) Obtain σy ∼35nm(Note: The principles of chromaticity correction are differentbetween FFTB and ILC)
(A2) Maintain for long time
(B) Stabilization of beam center
(B1) Down to <∼2nm by nanoBPM (cavity BPM)
(B2) Bunch-to-bunch feedback of ILC-type beam (∼300nm inter-val) when fast kicker extraction is ready
But technological ones are not all. . . ATF2 should be
• World center of beam-handling technology
• Recruit/education center for new comers
as the present ATF is.
40
International Collaboration on ATF2
• Design study going on by international collaboration
◦ Mini-workshops: Dec.11 at KEK, Jan.5 at SLAChttp://www-conf.slac.stanford.edu/mdi/ATF2.htm
◦ Submit proposal to PAC in May
◦ Complete the design by BDIR workshop in June
• Budget
◦ Total ≈ 3.0 Oku Yen (floor, beamline, diagnostics)
◦ Desirable to share the expense among Asia, North America,Europe
◦ Basic agreements with SLAC
• Present plan
◦ Floor construction in summer 2006
◦ Start operation in Jan.2007
41
42