Accelerator Physics Issues in the ILC

1
Accelerator Physics Issues in the ILC

• James Jones
• ASTeC, Daresbury Laboratory

2
Layout of Talk

• Justification for the ILC
• Why a linear Collider?
• Luminosity Scaling in a LC
• SCRF
• Linear Collider Subsystems
• General Layout
• Sources
• Damping Rings
• Bunch Compressors
• Linac
• Beam Delivery System

3
Why the I(nternational) L(inear) C(ollider)?

• If we look at the energy frontier in particle
  physics accelerators of the last 40 years we see,
  almost without exception, discoveries have been
  made using circular machines.
• Sowhy build a linear collider?
• Lets just make a bigger ring?
• Remember This is a Lepton Machine!

4
Synchrotron Radiation
Power radiated in a dipole at energy E, and
bending field B

• The issue is one of radiationway too much
  radiation.
• Synchrotron arises from an accelerating particle
  in a magnetic field

Which leads to the Energy loss per turn (that
which needs to be replaced)
5
Synchrotron Radiation RF Power

• The synchrotron radiation is replaced by the RF
  system, which becomes the major cost factor in a
  large collider.
• We have seen that the RF costs vary as
• Whilst the linear costs vary as
• And thus the optimum cost varies with
• We can demonstrate this using the LEP machine as
  an example

6
LEP as an Example

• If we imagine a new Super-LEP, or even some sort
  of HYPER-LEP, how much is it likely to cost?

7
The Solution

• The solution is to go for a Linear Collider!
• This consists of a long linac of RF accelerating
  structures with gradients in the region of 25MV/m
• This is for ILC, can have gt100MV/m
• Most Important

e
e-
10 km
8
Luminosity Design Issues

• Energy Reach

Luminosity
9
The Luminosity Issue
10
Luminosity Scaling Law
11
Luminosity Scaling Law
12
Luminosity Scaling Law
tiny vertical emittancestrong focusing at IP
(short bunch length sz)
13
Luminosity Scaling Law
Beamstrahlung
degrades luminosity spectrumbeam-beam
backgrounds (pair production)generally
constrained to a few
14
The Luminosity Issue

• High current (nb N)
• High efficiency(PRF ?Pbeam)

• High Beam Power
• Small IP verticalbeam size

• Small emittance ey
• strong focusing(small by)

15
The Luminosity Issue

• High current (nb N)
• High efficiency(PRF ?Pbeam)

Superconducting RFTechnology

• Small emittance ey
• strong focusing(small by)

16
Why SCRF?

• Low RF losses in resonator walls(Q0 ? 1010
  compared to Cu ? 104)
• high efficiency hAC ?beam
• long beam pulses (many bunches) ? low RF peak
  power
• large bunch spacing allowing feedback correction
  within bunch train.

17
Why SCRF?

• Low-frequency accelerating structures(1.3 GHz,
  for Cu 6-30 GHz)
• very small wakefields
• relaxed alignment tolerances
• high beam stability

18

• End Of Section 1

19
The layout of a linear collider

• This is a generic layout!

20
Civil Engineering of the ILC

• Possible layouts of the ILC _at_ 500GeV 1TeV
• Linac will follow the earths curvature, whilst
  the DR and BDS will be laser straight
• Tunnel depth and design will depend on site
  considerations

21
Source
1000ms _at_ 5Hz Several nC at least on the e-
side

• The e- and e sources must meet several
  requirements
• Produce long bunch trains of high charges
• Produce small emittance beams
• Produce spin polarised beams

• The e- source can be produced from either an
  advanced RF gun or a photo-cathode gun.
• The e source is much more difficult requires
  either a multi-GeV electron source and target or
  an undulator and target

22
e- Source

• A laser-driven photo cathode is one of the likely
  designs for the ILC source.
• A polarised laser impinges on a photo-electric
  substance often semi-conductor based (e.g.
  GaAs, CsTe) due to their high quantum efficiency.
• Laser pulse can be modulated to give the required
  time structure of the pulses.

• Semi-conductor cathodes require very high vacuum
  (lt10-11 mbar)
• Beam emittance is dominated by space-charge
  effects
• mm-mrad (10x, 500y)
• Typical energy at the exit of the gun is MeV

23
e Source

• There are two main mechanisms for producing
  positrons
• A conventional source consisting of an electron
  accelerator and a thick target
• Or an undulator source consisting of a long
  undulator producing photons, and a thin target

? 10MeV
e
e-
e-
150m
24
Synchrotron Radiation - Undulators

• An Undulator creates a periodic magnetic field,
• Permanent Magnets or Pulsed
• Lorentz contracted magnetic period ? ? / ?
• Relativistic Doppler effect ? ? / 2 ?
• Observed Radiation Wavelength ? / 2 ?2 (150GeV
  ?294000)

Magnetic Field periodic in one plane
Linearly Polarised Light
Electron Motion periodic in the other plane
25
Helical Positron Undulator

• If we want circularly polarised e we need
  circularly polarised photons.
• This is done with a helical undulator.

• The S/C undulator is the chosen design
• Uses S/C ribbon wire alternately wound around a
  4mm vacuum chamber, creating a helix.
• Current flows in alternate directions in
  alternate strands

26
Helical Positron Undulator - Layout

• Layout with the helical undulator places it at
  150GeV point in the electron linac electrons
  used to produce the photons that hit the target
  and create positrons.
• Positron beam is then captured and transported
  all the way to the other end of the ILC to the e
  damping ring

27
Buncher Cavity Bunch Timing Structure

• Electrons bunches from both sources are too long
  (ns) for the linac (ps _at_ 1.3GHz)
• Long electron trains are bunched in a
  Sub-Harmonic-Buncher.
• This defines the train length of 1000ms

28
Damping rings

• Beams have emittances much larger than required.
• Damping ring used to reduce the emittance
• Use a large storage ring and synchrotron
  radiation to damp the emittance of the beam
• The damping of the beam is described by
• With the damping time given as
• That is, twice the time it takes to radiate all
  of its initial energy

29
Damping rings

• Damping ring must cope with (at least) one bunch
  train (see why next!)
• This is 950ms x c 285km!
• Therefore need to compress the bunch train into
  smaller bunch spacing
• Damping ring size depends directly on the rise
  and fall times of injection kicker(s)
• Currently have 6ns kicker rise time ? 6.6km ring
• The damping rings have many issues surrounding
  instabilities
• Major problem is the electron cloud in the e
  damping ring
• Electrons collect in vacuum chamber and are
  resonantly accelerated by the positrons fields
  then cause instabilities
• Can be mitigated by varying the bunch spacing or
  by trapping the electrons in a grooved section of
  the vacuum chamber.

30
Transverse Damping

• When an e-/e emits a photon of radiation,
  assumed to be in the forward direction, it loses
  transverse momentum.
• The RF system only replaces longitudinal
  momentum.
• Since we can see that

31
Transverse Damping

• We can calculate the damping time scaling in
  relation to the ring design
• We know and
• Which can be reformulated as
• And thus
• For, say, LEP ,

32
Quantum Excitation

• The final emittance from the damping ring also
  has an anti-damping component due to the
  quantum nature of particle emission.
• When a particle emits a photon in an area of
  dispersion, the change in particle energy leads
  to a change in the particles orbit an increased
  Betatron oscillation ? an increased emittance

If we call the excitation rate Q, the equilibrium
emittance is then achieved when the damping and
the excitation rate are equal With
33
Transverse damping times

• The damping time varies as
• However, the RF costs vary as
• And the equilibrium emittance as
• As an example
• We therefore require almost 1sec to damp the
  beam!
• Use damping wigglers to increase the damping
  rate
• e- damping time 50ms
• e damping time 25ms due to larger initial
  emittance

34
Vertical Emittance

• There is no (designed) dispersion in the vertical
  plane.
• From the previous slides, this would imply the
  vertical emittance damps to zero!
• Actual vertical emittance is theoretically
  limited by
• Space charge effects
• Intra-beam scattering processes
• Opening angle of the photon radiation
  (diffraction and electron beam size)
• In reality the vertical emittance will probably
  be dominated by magnet errors
• Cross-plane coupling from displaced quadrupoles
  and sextupoles.
• Typical alignment tolerances 30mm
• Require extensive beam based alignment techniques

35
Low Emittance Tuning

• To maintain the low vertical emittance required,
  need to correct errors such as coupling and
  spurious vertical dispersion.
• Errors come from a variety of source at different
  frequencies
• Weather (very low frequency), Ground (low),
  Civilisation (medium), Equipment (medium-high)

• The correction of these errors involves the use
  of dipolar correction magnets (to steer the
  beam), and skew quadrupole magnets on movers (to
  correct the vertical dispersion and the coupling).

Uncorrected Ground Motion
Corrected 1/day
36
Bunch Compressors

• From radiative effects in the damping ring, the
  equilibrium bunch length is of the order a few
  mm, at the Interaction Point we want a bunch
  length 300mm
• A Bunch Compressor is used to compress the bunch
  longitudinally, at the expense of a corresponding
  increase in the energy spread of the bunch.

?E
-?E
37
Bunch Compressors

• There are many designs for the bunch compressors
  not just dipole chicane!
• Can use a wiggler to give a large path length
  change.
• Important quantity is R56 term gives
  relationship between path length change with
  energy.
• If the energy spread becomes too large, the
  higher order terms (T566) become important
• This makes the beam become non-linear in
  longitudinal phase space
• Introduces unwanted correlations at the IP
• The bunch compressor must also not increase the
  transverse emittance of the beam. This can arise
  due to two main effects
• Chromatic aberrations occurring due to non-linear
  dispersion in the chicane
• ISR and CSR instabilities, which constrain the
  strength of the chicane dipoles.

38
LINAC

• Baseline design for the ILC linac is the TESLA
  9-cell 1.3GHz Nb cavity.
• Design used in the Tesla Test Facility an X-ray
  FEL demonstration and will be used in the X-FEL
  project _at_ DESY.
• Mature super-conducting design that regularly
  makes gt25MV/m, and has achieved 35MV/m

39
Cavity Modifications

• As soon as the decision for S/C RF was made, new
  ideas to improve the TESLA cavity were announced

Re-entrant cavity, may provide even higher
gradients, but will require new infrastructure to
implement
10 increase in gradient
40
Fabrication Techniques
41
Electro-polishing

• Electro-polishing remove surface impurities from
  the cavity walls
• These cause increased magnetic fields and
  electron emission which lower the Q-factor of the
  cavity ? Lower achieveable gradients

Buffered Chemical Polishing
Electro- Polishing

• Several single cell cavities at g gt 40 MV/m
• 4 nine-cell cavities at 35 MV/m, one at 40
  MV/m
• Theoretical Limit 50 MV/m

42
Cryomodules

• The S/C RF cavities must be housed in a
  cryomodule, with a suitable method of RF power
  distribution.
• The design of the cryomodules is complicated
  how many cavities per module, spacing, quadrupole
  location.

• The Input couplers are also a major design
  consideration high powers and warm-to-cold
  transitions make them a challenging design.

TTF Cryomodule Design
Saclay Input Coupler
43
LINAC tunnel housing
Single tunnel solutiona la TESLA TDR(and for
the XFEL)
44
LINAC tunnel housing
Two-tunnel optionklystrons/modulators(?)/LLRF/PS
in Service Tunnel to allow access during
operation (availability arguments).
45
Linac Cost

• The linac is the major cost driver for the ILC -
  16,000 S/C cavities!
• gt10 years of RD by the Tesla Collaboration has
  produced a mature scaleable design that has
  already been heavily cost-optimised.
• Much work needs to be done on technology transfer
  to industry mass production of high gradient
  cavities has yet to be proven
• In Europe, have a head start with the XFEL project

• Still need to decide on final accelerating
  gradient
• 27MV/m SAFE
• 31.5MV/m BASELINE
• 40MV/m AMBITIOUS
• Directly affects the length of required linac
• Could limit the upgrade potential of the final
  machine.

46
Beam Delivery System

• The beam delivery system has several major
  functions
• Focus and collide nanometre sized beam at the
  interaction point
• Remove (collimate) the beam halo, producing less
  background at the IP
• Provide diagnostics for the upstream section of
  the collider
• The beam delivery system is 2.5km long, and
  contains 8 distinct sections. In order, they are
• Machine Protection
• Skew correction section
• Emittance Measurement section
• Energy diagnostics and Polarimeter
• Energy Betatron collimation
• Final Focus System
• Extraction Line Beam Dump

47
Skew Correction Section

• Skew correction section removes cross plane
  coupling introduced in the linac.
• Measure the cross-plane terms using orthogonal
  wire scanners.

4 orthogonal skew quadrupoles correct xy, xy,
xy, xy
4 laser wires for emittance measurement
48
Collimation

• Collimation occurs to remove Beam Halo and other
  backgrounds
• Tend to degrade the luminosity spectrum
• Betatron collimation is performed by several
  collimators
• Each one covers a different phase in transverse
  phase space

49
Collimators

• The actual collimators must be able to withstand
  at least one bunch train.
• They are therefore spoiler/absorber pairs the
  spoiler disrupts the beam and causes a shower
  which is mopped up by a downstream absorber.
• Collimators are a major source of wakefields in
  the BDS

50
Final Focus

• At the IP, very strong defocusing of the incoming
  beams is required for maximal luminosity
• This requires very strong quadrupole magnets,
  which leads to strong chromatic aberrations.
• 2 designs of final focus to correct these
  chromatic effects at the IP

Local correction using a finite D at the IP
Raimondi (L 500m) Non-local correction (CCS
scheme) Brown (L 1.5km)
51
IP Fast (Orbit) Feedback
Long bunch train 3000 bunches tb 337 ns
52
Ground motion spectra
53
Long Term Stability
beam-beam feedback upstream orbit control
No Feedback
beam-beam feedback
example of slow diffusive ground motion (ATL law)
54
BDS Strawman Model

• Baseline BDS model is now 14mrad x 2
• Ideally want two different crossing angles
  small large
• Small angle allows better hermiticity and smaller
  backgrounds
• Large angle is easier, and can be upgraded to
  ?-? (?)

55
IR (BDS) Civil Engineering
T. Markiewicz (SLAC) MATLAB Tool to study
constraints from civil engineering
56
Detector Hall

• Construction and placement of detectors is still
  under discussion
• Detector concepts all very different
• Construction timescales for all concepts very
  tight
• Possibility of building detectors on surface a la
  CMS

57
Beam Delivery System

• There are several important issues in the BDS
• Crossing Angle _at_ IR
• 2mrad (head-on) and a 20mrad IR? Proposal is
  now for two 14mrad IRs.
• 1 IR or 2?, 1IR with a push-pull detector?
• Lots of work on making designs work for all
  operational options (i.e. high lumi).
• Collimation System
• Do we have renewable spoilers or not?
• Non-linear collimation, ala CLIC?
• High-powered dumps
• The beam power is 11MW/dump.
• Water or gas filled dump?
• Combining dumps from different IRs or tune-up
  sections?
• How will the beamstrahlung dump work?
• Work on the next strawman model with two 14mrad
  IRs is ongoing.
• Work has not stopped, however, on either of the
  other 2(3) options

58
ATF-2

• BDS Test Facility _at_ KEK
• Extension to ATF damping ring test facility

59
End Station - A

• End Station A is an old experimental hall at the
  end of the SLAC linac.
• Beam energies up to 28.5GeV _at_ 10Hz
• Used for BDS diagnostics tests
• Also collimator wakefield tests
• Future uses still uncertain

60
The Global Design Effort GDE

• 3 Regional Design Teams
• Central Group with Director
• Goal Produce an internal full costed ILC
  Technical Design Report by 2008

61
ILC Projected Time Line
2005
2006
2007
2008
2015
2010
2012
preparation
construction
operation
EURO XFEL
EUROTeV
UK playing a significant role(both detector and
machine)
CARE
62
Summary

• The ILC is ambitious project which pushed the
  envelope in every subsystem
• Main SCRF linac
• sources
• damping rings
• beam delivery
• Still many accelerator physics issues to deal
  with, but reliability and cost issues are
  probably the greater challenge
• Probably in excess of 3000 man-years already
  invested in design work.

cost driver
L performance bottleneck
63
Thanks!

• Nick Walker Whom I stole most of this talk
  from
• All the people in ASTeC/Cockcroft who contributed
  slides