JLCNLC Baseline Design and TRC Report

1
JLC/NLC Baseline Design and TRC Report

• Tor Raubenheimer Kaoru Yokoya
• ISG9
• KEK, December 10th, 2002

2
X-band Linear Collider Goals

• Stage 1 Initial operation (length 18 km)
• 500 GeV cms (650 GeV at lower luminosity)
• L 25 20x1033 in JLC and NLC
• Stage 2 Add additional X-band rf components
  (length 32 km)
• 1 TeV cms (1.3 TeV at lower luminosity)
• L 25 30x1033 in JLC and NLC
• Higher Energy Upgrades
• 1.5 TeV with upgrade of linac rf system or length
  increase
• NLC injector and beam delivery built for gt1.5 TeV
• 3 TeV with advanced rf system and upgraded
  injector
• See CLIC parameters A 3-TeV ee- Linear
  Collider Based on CLIC Technology, CERN-2000-008
• Beam delivery sized for 3 to 5 TeV collisions

3
Beam Parameters

• Parameters also exist for operation at the
  Z, W, low-mass Higgs, and top

• Stage 2 has 5x1033 at 1.3 TeV

Trade cms energyfor beam current
4
Change to Baseline Design

• Initial NLC design presented to the TRC was too
  difficult(3.2 ms klystrons and 8x DLDS pulse
  compression)
• 3.2 ms klystrons were not on the horizon
• DLDS demonstration was years away
• TESLA has been making rapid progress both
  technically and politically
• However with progress on the structure gradient,
  the potential for X-band was clearer than it had
  been for years!
• Needed an rf configuration that could be
  conclusively demonstrated in 1 year timescale to
  compete with TESLA
• 1.6 ms klystrons were making great progress at
  KEK
• SLED-II only requires 2 klystrons rather than 8
  klystrons for DLDS
• Concept was already demonstrated with NLCTA
  operation
• At 4x pulse compression, compression efficiency
  is still quite good

5
Stolen from Nobu Toge!
6
JLC/NLC RF System Tests
SLED-II Demonstration Spring 2003
JLC/NLC SLED-II Rf System Layout
Delay line in tunnel or gallery
500 MW in 400 ns
7
Desired Energy Range

• Would like to cover the Z-pole and W-production
  and then have continuous energy coverage at
  energies above LEP (gt220 GeV)
• Upper energy reach needs are not known
• Opinions vary around the world (HEPAP and ACFA
  specify 1 1.5 TeV while ECFA calls for 400
  800 GeV)
• Lower energy operation is limited by e
  production schemes
• Current TESLA scheme does not work between cms
  200 300 GeV
• Highest energy reach can be attained by trading
  beam loading for energy
• TESLA has 50 loading ? 50 greater energy reach
  (providing cavity limits are not exceeded)
• NLC has 30 beam loading cavities are qualified
  at unloaded gradients

8
Energy Reach
Both designs can trade luminosityfor energy in
Stage I TESLA reaches 750 GeV NLC reaches
650 GeV NLC can do the same in Stage II to reach
1.3 TeV
NLC Stage 2
1300
25 Bunches
NLC trades tighter tolerances for a greater
energy reach
1200
Trade cms energyfor beam current
1100
192 Bunches
1000
0
0.5
1
1.5
2
2.5
3
Luminosity (1034)
9
Upgrade Routes and Costs

• NLC and TESLA costs are similar in value for 500
  GeV
• FNAL review of TESLA costing methods estimated
  6.1 B with US accounting and 20 contingency
• NLC (and JLC?) costs are estimated at 7 B with
  30 contingency
• The error on the costs is much greater than any
  differences!
• JLC/NLC upgrade requires adding structures,
  klystrons, etc. in the 2nd half of the linac
  tunnel
• Cost to upgrade to 1 TeV is roughly 30 of
  initial TPC
• TESLA upgrade route install 35 MV/m cavities at
  onset, double rf system, upgrade cryo-plant
• Assuming initial installation of 35 MV/m
  cavities, cost to upgrade to 800 GeV cms is 20
  of initial project cost
• Upgrade from 800 GeV to 1 TeV is another 25 for
  a total of 45 of the initial project cost

10
SLED-II Implications

• Changing the baseline rf system to SLED-II
• Faster demonstration of both power source and
  later a full rf sub-unit, i.e. 1, 2, or 4
  klystron pairs powering structures
• Depends on final modulator configuration
• Doubled number of klystrons, modulators, and PC
  systems and increased the linac length by 8
• Obvious cost impact perhaps some advantage in
  reliability
• Decreased unloaded gradient by 7
• Reduced breakdown rate
• Very small effect on tolerances and emittance
  preservation
• Tolerances scale with the square root of the
  gradient
• Makes 8-pack failure slightly more invasive to
  beam operation
• With DLDS 8-pack failure probably did not have to
  scale quadrupole lattice (LEM) with SLED-II
  almost certainly will

11
Why X-Band?

• Why X-band instead of superconducting?
• Superconducting has loose tolerances and
  stability however these are not necessary nor are
  they necessarily easier to achieve in cryostats
• Entering new regime with massive superconducting
  facilitywill limiting effects be found that
  cannot be seen in TTF?
• Many sub-systems are very different from anything
  demonstrated
• Particularly true for damping ring complex
• Are the cost of TESLA and NLC similar for
  comparable energy reach?
• Why X-band instead of lower rf frequencies?
• Support higher gradients 70 ? 100 MV/m for
  cheaper collider
• The technology exists and the tolerances are
  attainable
• Present prototypes are better than tolerances
• Directly use knowledge from SLC Its the
  diagnostics, stupid!
• X-band provides essential knowledge-base for
  continued progress in e/e- colliders at higher
  gradients and possibly higher rf frequencies

12
Why X-Band? (2)

• Why not higher rf frequencies?
• Power sources are not available for component
  testing
• X-band klystron sources took 10 years to develop
  and can now reliably deliver 500 MW in 150ns
• CLIC just recently delivered 200 MW in 40ns at 30
  GHz in CTF-II but lots of jitter and breakdown
  even with 40 MW
• This will change with time, making higher f more
  attractive however it will take years to develop
  the suite of rf components
• Pulsed heating becomes a severe limitation
• 900C pulsed rise calculated in CLIC heavily
  damped structure
• Cracking in copper observed with 80C rise
• Do not understand breakdown limits well enough
  but cannot expect a linear gradient rise with f

13
Why X-Band? (2)

• Why klystrons instead of Two-Beam Accelerator at
  X-band?
• Experience building reliable repeatable
  klystronsTBA drive beam is quite difficult
• Beam must be steered but beam halo makes
  diagnostics difficult
• Drive beam is basically unstable with BBU and
  large energy spread
• Operational issues associated with TBA are very
  difficult
• TBA is years from being tested as a system
• CTF-II is demonstration of concept but not
  feasibility
• Cost is not significantly different at 1 TeV
  higher initial value for drive beam complex but
  lower slope with increasing energy
• However, higher frequency normal conducting is
  probably the only route to higher energy
  collisions
• X-band provides essential knowledge base for
  continued progress in e/e- colliders

14
LC Technical Review Committee
Two working groups (plus reliability)
Energy Technology Daniel Boussard
(Chair) Chris Adolphsen, SLACHans Braun,
CERN Yong-Ho Chin, KEK Helen Edwards, FNALKurt
Hubner, CERNLutz Lilje, DESY(Pavel Logatchov,
BINP)Ralph Pasquinelli, FNALMarc Ross,
SLAC(Tsumoru Shintake, KEK)Nobu Toge, KEKHans
Weise, DESYPerry Wilson, SLAC
Luminosity Gerald Dugan (Chair) Ralph Assmann,
CERNWinnie Decking, DESY Jacques Gareyte,
CERNWitold Kozanecki, SaclayKiyoshi Kubo,
KEKNan Phinney, SLACJoe Rogers, CornellDaniel
Schulte, CERNAndrei Seryi, SLACRon Settles,
MPIPeter Tenenbaum, SLACNick Walker, DESY Andy
Wolski, LBNL
15
LC Technical Review Committee

• The Technical Review Committee (TRC) has led to
  extensive review of many JLC/NLC and TESLA
  subsystems
• Many JLC/NLC people are involved directly or
  indirectly
• Energy/technology Group present status is being
  studied (compared)
• Luminosity Group breaking new ground
  (collaborative effort)
• Mainly damping ring studies and DR ? IP
  luminosity studies
• Benchmarking simulation codes against each other
• Pushing to include all relevant effects
• Static alignment procedures
• Vibration and stability effects with feedback
  systems
• Beam-beam effects at IP
• Integrated luminosity performance very difficult
  to evaluate
• Reliability Group pushing everybody to arrive at
  realistic reliability goals with necessary
  changes to designs
• Leading to improvements in all designs TESLA and
  JLC/NLC

16
TRC Conclusions

• Greg Loew reported at ICFA Seminar
• Four levels of RD priority R1 feasibility R2
  required for design R3 before final
  engineering R4 optimization.
• Energy and technology issues were not unexpected
• Lots of work on luminosity issues
• Conclusions not very sharp
• Two unexpected items found in JLC/NLC design
• Importance of collimator wakefields
• Jitter amplification of parasitic crossings
• Many unexpected items found in TESLA
• 10x better vacuum and BPM resolution required in
  TESLA DR
• Wigglers severely limit dynamic aperture in TESLA
  DR
• 100 150 emittance growth in TESLA linac versus
  3 in TDR
• Large difficulties with head-on collisions
• Also learned of things that work in TESLA design
  ? apply to NLC?

17
TRC Energy Technology Conclusions
JLC/NLC
TESLA
R1

• Test of complete accelerator structure at design
  gradient with detuning and damping, including
  study of breakdown and dark current
• Demonstration of SLED II pulse compression system
  at design power level
• Test of a complete main linac RF sub-unit (as
  identified in machine description) with beam
• Full test of KEK 75 MW, 1.6 µs PPM klystron at
  150 or 120 Hz
• Full test of SLAC induction mod.

• Building and testing of a cryomodule at 35 MV/m
  and measurement of dark current
• Test of a complete main linac RF sub-unit (as
  identified in machine description) with beam
• Testing of several cryomodules at nominal field
  (23.4 MV/m) over long enough periods to verify
  breakdown and quench rates, and measure dark
  current
• Test of RF components at higher powers for 800
  GeV operation

R1
R2
R2
18
TRC Luminosity Conclusions
Common (JLC/NLC and TESLA)
TESLA Only

• Electron cloud and ion instabilities need study
• Additional simulations and experiments on e
  correction are needed for damping rings
• Demonstrate extraction kicker with better than
  0.1 stability
• Complete static DR?IP tuning simulations with
  dynamic effects
• Develop most critical beam instrumentation,
  including intra-train diagnostics
• Develop sufficiently detailed prototype of linac
  girder/cryostat to provide information on
  vibration

• Further optimization of damping ring dynamic
  aperture
• Study tighter alignment and electron cloud and
  ion instability requirements for 800 GeV upgrade
• Development of TESLA DR kicker
• Review trade-offs between head-on and
  crossing-angle collisions
• Detailed analysis of the tradeoffs between one
  and two-tunnel layouts
• Detailed evaluation of critical sub-system
  reliability

R2
R2
19
Luminosity Issues Only few x 10,000 larger than
SLC!

• Increased beam power from long bunch trains
• SLC 120 Hz x 1 bunch _at_ 3.5x1010
• NLC 120 Hz x 192 bunches _at_ 0.75x1010 ? 200x
• Generation of uniform multi-bunch trains sets
  source requirements
• Control of long-range wakefields is essential to
  prevent BBU
• Larger beam cross-sectional densities N / (sx
  sy)
• SLC 3.5x1010 x 1.6 mm x 0.7 mm (FFTB 0.6x1010
  x 1.7 mm x 0.06 mm)
• NLC 0.75x1010 x 250 nm x 3.0 nm ? 330x SLC
• Factor of 5 from energy (adiabatic damping) and
  factor of 4 from stronger focusing (similar to
  Final Focus Test Beam)
• Factor of 15 30 from decrease in beam
  normalized emittances at IP

20
Luminosity Issues for FY03

• Concentrate on improving source designs
• Evaluate undulator-based e source
• Improve energy spectrum at damping ring injection
  and review multi-bunch limitations
• Improve damping ring designs
• Evaluate e-cloud and ion instabilities and
  improve wiggler models
• Consider alternate ring designs for better
  stability
• Improve linac (LET) emittance preservation
  simulations
• Aim towards evaluating integrated luminosity
• Must understand reliability issues
• Experiments?
• Continue BDS optimization
• More inclusive background/collimation studies
• Stabilization studies

21
Polarized Electron Sources

• JLC/NLC e- source based on successful SLC
  injector design
• Emittance estimates reasonable from SLC operation
• Charge Limit limited current from SLC polarized
  photocathode
• Demonstrated charge and polarization during E158
  run

20 mm spotin 100 ns

• NLC requires 100 nC per 100 ns
• Greatly exceeded in E-158 tests
• Polarization was 85

14 mm spot in 100ns
NLC Req.
SLC photocathodein 300 ns
22
Positron Source Targets

• SLC e target failed after 5 years at stress
  levels 2x lower than previously measured and
  predicted for WRe targets
• Radiation damage problem may be worse in Ti
  targets ? starting RD program to investigate

SLC e target
Beam direction

• SLC target studied at LANL and modeled at
  LLNL
• Problem due to embrittlement from
  radiation damage

23
Positron Source Concepts

• Baseline design uses abrute force solution toe
  target limitations
• Need to investigateother conventional concepts
• Studying a variation of TESLA undulator source
  for polarized positron production
• Studies of e yield as well as helical undulator
  and targets
• 150 GeV location eases operational limitations
• Possible test in FFTB line at SLAC

24
Damping Ring Issues

• Beam is stored for a relatively long time in the
  rings (ms)
• More accelerator physics in the rings than
  elsewhere in LC
• Stability is essential for collider performance
• JLC/NLC rings are similar to the ATF and the 3rd
  generation SRS
• Damping rings have beam currents and bunch trains
  similar to the high operating luminosity
  factories
• However they have much smaller beam sizes (higher
  densities) and are much more sensitive to weak
  instabilities
• Tougher dynamic aperture and stability
  requirements
• They also require much better alignment to get
  flat beams 40 um
• High beam density pushes frontier in electron
  machines
• Space charge tune depression and microwave
  instabilities
• Ion trapping and electron cloud effects
• Intrabeam scattering and short Touschek lifetime

25
Operating Ring Comparisons

• Compare random alignment and jitter tolerances
• Uncorrelated misalignments or jitter that would
  lead to equilibrium emittance, jitter equal to
  the beam size, or Dn 0.001
• These are not specs. on alignment but they are
  measures of the sensitivity
• Looking for significantly better alignment and
  stability than has been previously attained

26
Dynamic Aperture Studies

• Dynamic aperture in rings is difficult for three
  reasons
• Strong focusing for small emittance
• Long wigglers for damping time
• Large injected beam size from high-power injector
  (60 kW)
• Current calculations look good but have narrow
  energy acceptance
• Want to improve model of incoming beam and
  reduce sDE
• Need to improve wiggler model and add field
  errors
• Recent work by Marco Venturini looks like a good
  start

27
Impedance Issues

• Calculated chamber impedance model using MAFIA
  and Omega3P in 1995 and again in 1998
• Ring layout and components have not changed much
• Used Oide-san Vlasov code to calculate thresholds
• Close to weak threshold and factor of 3 or below
  strong
• Very similar results in 1996 ZDR
• Recently started calculating CSR impedance from
  bends and wigglers (Juhao Wu)
• Thresholds look similar to that of chamber
  impedance
• Need to combine all sources of impedance and
  update threshold calculations
• Also need to consider methods of increasing ap
• Increases bunch length and increases threshold

28
Electron Cloud and Ion Instabilities

• Electron cloud and ion instabilities are
  important concerns
• Designed ring for 1 nTorr vacuum pressure to
  reduce ion instability growth times
• Analytic estimates still give 100 us
• Will need to reduce SEY to handle electron cloud
• Experiments planned at LBNL to study methods of
  reducing SEY
• Need to understand e- lifetime
• Need full simulations ions are easy electron
  cloud more difficult.

Calculation by Mauro Pivi
29
Linac Emittance Preservation

• Improved optics to avoid known SLC problems
• BBU from short- and long-range wakefields must be
  controlled
• Measurements at ASSET verify wakefield reduction
  procedures
• Tight alignment tolerances
• NLC alignment and jitter tolerances are 10x and
  100x tighter than SLC TESLA alignment and jitter
  tolerances are 1x and 10x tighter than SLC
• Sets tight requirements on diagnostics and
  controls
• NLC design includes more controls and diagnostics
  than FFTBTESLA design include less control and
  diagnostics than SLC
• Demonstrated NLC-level alignment at FFTB
• Demonstrated emittance tuning and BNS damping at
  SLC
• Will require excellent stability (both vibration
  and drift)
• Need to understand limitations
• Requires improved simulation tools ? integrated
  luminosity

30
Coolant Vibration Studies

• Vibration of accelerating structure caused by
  turbulent water flow
• Measured 300 nm on structure ltlt tolerance
• Coupling to quadrupole was measured at about 3
  nm tolerance is 10 nm!
• Direct coolant induced quad vibration 2 nm
• Direct effect from rf pulse is negligible!
• Next step model complete girder with quad

300 nm at nominalflow
31
Stabilization Studies
Inertial Stabilization Test Block

• Final doublet tol. 1nm
• Inertial system with 6 d.o.f. tested at SLAC
• Limitation due to sensors
• Reduced vibration gt10x
• Other inertial studies atCERN
• Optical anchor systemstudied at UBC
• Developing new non-magnetic sensorwith better
  noise characteristics
• Working on details before considering a real
  implementation

32
FONT Feedback On ns Timescales
Magnet assembly and X-band BPM installed onto
NLCTA latency of 40 ns cable delay Measured
10x reduction of beam motion
Beam direction
Feedback loop
33
Conclusions

• SLED-II demonstration and Gradient RD are 1st
  priority
• Luminosity RD has been very broad and addressed
  most issues
• This is a real strength of the JLC/NLC design
• Many test facilities, RD, and simulation studies
• Most sub-systems and diagnostics are modest
  extrapolations from operating accelerators or
  demonstrations at test facilities
• Collider is designed with an energy reach from 92
  GeV to 1.3 TeV
• Continue to develop models to evaluate
  reliability and integrated luminosity
• Think about next-generation simulation tools
• Work on multi-bunch issues in source studies
• Improve damping ring designs and beam injection
  systems
• Continue to work on stability issues and BDS
  optimization

34
WG1 (Luminosity Issues) Schedule