for the LHC
R. Assmann (CERN)
V. Previtali (CERN, EPFL)
for the LHC Collimation Project
Crystal Collimation Workshop, Fermilab
RWA&VP, FNAL 12/07
LHC Collimation Staged Concept
• The LHC collimation system has been proposed and is being
implemented as a staged system:
– Phase 1: Collimation and machine protection functionality for beam
intensities up to 40% of nominal.
– Accept predicted limitations in cleaning efficiency and impedance
during the first few years of LHC operation.
– Phase 2: Upgrade of the system for requirements of nominal &
ultimate parameters and beyond.
• Gain factor 10 in efficiency.
• Reduce impedance by factor 4 or more.
– Several routes will be considered in parallel to achieve these goals.
– Most likely: Several smaller improvements will bring required gain.
RWA&VP, FNAL 12/07 2
The Staged LHC Path
Energy density Stored energy Number of
at collimators in beams collimators
(nominal 7 TeV)
State-of-the-art in SC colliders
(TEVATRON, HERA, …)
1 MJ/mm2 2 MJ
Phase 1 LHC Collimation 400 MJ/mm2 150 MJ * 88
Nominal LHC 1 GJ/mm2 360 MJ 122
Ultimate & upgrade scenarios ~4 GJ/mm2 ~1.5 GJ ≤ 138
Limit (avoid metal damage or
~50 kJ/mm2 ~10-30 mJ/cm3
* Limited by cleaning efficiency (primary) and impedance (secondary)
RWA&VP, FNAL 12/07 3
Specified Allowed Proton
Losses with Collimation
• Collimators are the LHC defense against unavoidable losses:
– Irregular fast losses and failures: Passive protection.
– Slow losses: Cleaning and absorption of losses in super-conducting
– Radiation: Managed by collimators.
– Particle physics background: Minimized.
• Specified 7 TeV peak beam losses (maximum allowed loss):
– Slow: 0.1% of beam per s for 10 s 2.1 MW
– Transient: 5 × 10-5 of beam in ~10 turns (~1 ms) 83.3 MW
– Accidental: up toMJMJ in 200 ns into 0.2 mm2
4.2 1 20.8 TW
RWA&VP, FNAL 12/07
CERN White Paper
• There is no “magic bullet” for gaining a factor 10 in performance without
major side effects.
• All proposed improvements require careful R&D on principle and
• Two phases in implementation:
– Up to 2010/11: R&D on phase 2 secondary collimators and advanced
– 2010 – 2012: Production and installation.
– 2012: Operational usage of first collimator upgrades.
• Important milestones:
– 2010: Tests of phase 2 secondary collimators in the LHC.
– 2011: Tests of advanced collimation in the LHC (this includes crystals).
RWA&VP, FNAL 12/07 5
Collimation R&D Scope for LHC
• Phase 2 secondary collimators for warm regions (WP, LARP, FP7):
– Prototype from SLAC/LARP.
– Prototypes from CERN.
• Cryogenic collimators for dedicated places or regions (FP7 – collaboration
with GSI in Germany).
– Developed and built for the FAIR project (collimation in SC ring).
– Adapt to LHC parameters if cryogenic collimators are required.
– Requires shorter and higher field SC magnets.
• Other options pursued (no LHC prototyping or construction agreed yet):
– Non-linear collimation (FP7).
– E-lens scraper for the LHC (tbd).
– Crystal-enhanced collimation (WP, LARP, FP7):
– Magnetic collimators (tbd).
RWA&VP, FNAL 12/07 6
Tevatron Input Crucial
• LHC should start from the present state-of-the-art and extrapolate into
unknown territory: Tevatron lessons are essential!
– Maximum loss rates.
– Collimation set-up and tuning.
– Electron lens effect on losses.
• Concerning crystal-enhanced collimation:
– Does crystal-enhanced collimation reliably work with stored beam and
– What is the gain in efficiency?
– Lessons of using crystals in the environment of a high energy physics collider.
• CERN Accelerator&Beam department and EPFL participate through
existing LHC collimation project – LARP connection. Important studies…
RWA&VP, FNAL 12/07 7
Beam propagation Impact Collimator
Unavoidable losses 1. Phase 2 materials for system improvement.
2. Crystals AP under study (surface effects,
halo (p) Secondary dilution, absorption of extracted halo).
Crystal p Shower
≤ 1 mm
Phase 1 Colli- 1 Colli-
Shower Absorber SC magnets
Super- and particle
conducting physics exp.
CFC & CFC Phase 2 W/Cu W/Cu
Crystal material Low electrical resistivity, good absorption, flatness, cooling, radiation, …8
RWA&VP, FNAL 12/07
Layout is 100%
RWA&VP, FNAL 12/07 9
Implementation into LHC
• Reference (further work ongoing):
R. Assmann et al, “Optics Study for a Possible Crystal-Based Collimation
System for the LHC”. EPAC06.
• LHC requires efficient collimation during all stages of operation (injection
– ramp – physics – dump).
• Crystals can only be placed close to primary collimators in IR3 and IR7.
• In this case, we have two scenarios:
– Crystal not tuned up: “Conventional” collimators will establish multi-stage
– Crystal tuned up: Halo will be extracted to dedicated downstream dump (up to
2 MW over 10 s).
• Beam divergence at crystal varies with beam energy (present optics):
– Adjustment required during energy ramp to maintain crystal enhancement.
RWA&VP, FNAL 12/07 10
Beam Divergence at Crystal
RWA&VP, FNAL 12/07 11
Downstream Trajectory for
normalized amplitude (sigma)
Normalized offset [s]
-10 Phase shift
0 100 200 300 400 500
RWA&VP, FNAL 12/07 12
Crystal Effect: Increase of
Amplitude and Phase Shift
crit ~ 2-2.5 urad
v.r. ~ 3-4 urad
chan ~ 50 urad
RWA&VP, FNAL 12/07 13
Phase Advance Versus Distance
Required dumps: ~ 70 m downstream of crystal for channeling
~ 230 m downstream of crystal for reflection
RWA&VP, FNAL 12/07 14
– Reflection angle: - 4 mrad (~1.5 times critical angle)
– Amplitude increase: 6 6.3 s
– Opt. phase advance: 160 degrees
– Distance to dump: ~ 230 m
– Channeling angle: 50 mrad
– Amplitude increase: 6 27 s
– Opt. phase advance: 77 degrees
– Distance to dump: ~ 70 m
RWA&VP, FNAL 12/07 15
Energy Deposition (FLUKA)
Primary Dump channeled Dump reflected Dump channeled Primary
Collimators particles particles particles Collimators
RWA&VP, FNAL 12/07 16
Consequences of Diffusive Impact
RWA&VP, FNAL 12/07 17
Always Channeling and Reflection?
• Diffusion generates exponential distribution of impact parameter from
zero to typically 100-200 nm.
• A small fraction of halo will always see the channeling, even if crystal is
oriented for reflection. Same is true vice versa.
• Need to install dumps (absorbers) for both processes.
• Further understanding required to quantify effects.
RWA&VP, FNAL 12/07 18
Assume ideal crystal:
2.5 mm length,
15 mrad@450 GeV
8 crystals with
1 mm interdistance.
RWA&VP, FNAL 12/07 19
Issues for LHC Implementation
• Several issues have been identified for crystal implementation into the
LHC (some of them illustrated in previous slides):
– Halo beam dynamics with crystals (effect of small impact parameter).
– Dump of extracted energy (up to 2 MW over 10 s).
– Surface effects (interested in 1-100 nm regime).
– Implementation into high radiation zone (radiation damage).
– Operational stability and efficiency (change in divergence).
– Multi-turn turn behavior and associated issues (e.g. heating of crystal).
• Simulations ongoing for multi-turn collimation with crystals (crystal routine
implemented into CERN tracking code). However, basic uncertainties.
• Tevatron is the best place to experimentally assess basic feasibility of
• LHC efforts on crystal will depend on outcome at Tevatron.
RWA&VP, FNAL 12/07 20
• Results that we are interested in:
– Accurate geometry of bent crystal measured.
– Measured change in multi-turn beam halo distribution with crystal channeling and
reflection (resolution 10-4 level).
– Measured beam loss maps around the ring in different regimes.
– Measured temperature in the crystal and/or its holder.
• Tests proposed:
– Change diffusion speed and therefore impact parameter on crystal (allows determining
impact of amorphous layer, ratio channeling/reflection).
– Change multi-turn dynamics by changing collimator settings.
– Calibrate absolute loss rates with known intensity loss.
– Check on “spiky” loss behavior when crystal close to beam?
– Evaluate crystals with different bent angle, length, material, ...
RWA&VP, FNAL 12/07 21
Contributions from LHC Collimation
• Multi-turn simulations of crystal-enhanced collimation with different
diffusion models and parameters (will be done for LHC – numerical
models can also be used for Tevatron).
• Comparison with simulations results for Tevatron from other codes.
• Participation in experiments and in analysis of experimental data,
including comparison with simulation results.
• Support for measurements of beam loss around the Tevatron (might
include participation from CERN beam instrumentation group, including
LHC hardware – if needed and if agreed).
• Information exchange concerning precision movement of material inside
the vacuum (as done for LHC phase 1 collimators) – if needed.
• Evaluation of results for possible LHC implementation of crystals.
• People: R. Assmann, V. Previtali, S. Redaelli
(maybe + BI if requested and agreed)
RWA&VP, FNAL 12/07 22
• Crystals could possibly enhance the efficiency of the LHC collimation system.
• Basic feasibility in a storage ring with diffusive beam loss must still be proven.
• There are a number of issues to be addressed in order to arrive at a system
design for crystal-enhanced collimation in LHC.
• We are working on these issues but many input parameters are not fully defined
and require experimental input. Several questions mentioned in this talk.
• Tevatron can perform the needed experiments and clarify key issues and input
• Possible contributions from the LHC collimation team have been defined and
• Looking forward towards productive years on studying crystal collimation. Results
will determine LHC effort on this direction (goal are beam tests in 2011).
• Thanks to N. Mokhov and Fermilab for providing this unique opportunity.
RWA&VP, FNAL 12/07 23
RWA&VP, FNAL 12/07 24
Diffusion Process & Impact Parameter
Slow loss: Beam lifetime: 0.2 h Loss rate: 4.1e11 p/s
Loss in 10 s: 4.1e12 p (1.4 %)
(~ 40 bunches)
Assume drift: 0.3 sig/s
5.3 nm/turn (sigma = 200 micron)
Almost all particles
y ≤ 0.2 mm
RWA&VP, FNAL 12/07 25
pp, ep, and ppbar collider history Higgs +
SUSY + ???
~ 80 kg TNT
The “new Livingston plot“ of proton colliders: Advancing in unknown territory!
A lot of beam comes with a lot of garbage (up to 1 MW halo loss, tails, backgrd, ...)
Collimation. Machine Protection.
Comparison: Phase 1 Collimators
RWA&VP, FNAL 12/07 27
LHC Need for Collimation
• Ideally, stored proton beams would have infinite lifetime and no protons
would be lost.
• However, a multitude of physical processes will limit the lifetime of the
beams and unavoidable proton losses must be taken into account.
• Conditions for quenching a SC magnet:
– Transient loss of 10-9 fraction of beam (within 10 turns)
– Slow loss of 3×10-8 fraction of beam per s and per m (< 10000 h lifetime)
• Proton losses must be intercepted and absorbed by specifically
designed devices, namely collimators. These constrain the aperture.
• Multi-turn process: protons diffuse to limiting aperture bottleneck. Process
also called beam cleaning.
• 2 out of 8 straight sections in the LHC are dedicated to collimation!
RWA&VP, FNAL 12/07 28
LHC Collimator Gaps
5 - 6 s (primary)
6 - 9 s (secondary)
s ~ 1 mm (injection)
s ~ 0.2 mm (top)
Small gaps lead to:
1. Surface flatness
tolerance (40 mm).
RWA&VP, FNAL 12/07 29
Allowed Quench threshold
intensity (7.6 ×106 p/m/s @ 7 TeV)
Illustration of LHC dipole in tunnel
p Rq Ldil /c Cleaning inefficiency
Number of escaping p (>10s)
Number of impacting p (6s)
Beam lifetime Dilution
(e.g. 0.2 h minimum) length
Collimation performance can limit the intensity and therefore
RWA&VP, FNAL 12/07 30
Intensity Versus Cleaning Efficiency
For a 0.2 h
99.998 % per m efficiency
RWA&VP, FNAL 12/07 31
The LHC Phase 1 Collimation
• Low Z materials closest to the beam:
– Survival of materials with direct beam impact
– Improved cleaning efficiency
– High transparency: 95% of energy leaves jaw
• Distributing losses over ~250 m long dedicated cleaning insertions:
– Average load ≤ 2.5 kW per m for a 500 kW loss.
– No risk of quenches in normal-conducting magnets.
– Hot spots protected by passive absorbers outside of vacuum.
• Capturing residual energy flux by high Z absorbers:
– Preventing losses into super-conducting region after collimator insertions.
– Protecting expensive magnets against damage.
• No shielding of collimators:
– As a result radiation spread more equally in tunnel.
– Lower peak doses.
– Fast and remote handling possible for low weight collimators.
RWA&VP, FNAL 12/07 32
Efficiency in Capturing Losses
Beam1, 7 TeV
Local inefficiency [1/m]
99.998 % per m Quench limit
(nominal I, =0.2h)
Beam2, 7 TeV
TCDQ Ideal performance
99.998 % per m
(nominal I, =0.2h)
99.998 % needed
Local inefficiency: #p lost in 1 m over total #p lost = leakage rate 99.995 %
RWA&VP, FNAL 12/07 33
5) Beyond Phase 1
• The LHC phase 1 system is the best system we could get within the
available 4-5 years.
• Phase 1 is quite advanced and powerful already and should allow to go a
factor 100 beyond HERA and TEVATRON.
• Phase 2 R&D for advanced secondary collimators starts early to address
expected collimation limitations of phase 1.
• Phase 2 collimation project was approved and funded (CERN white
paper). Starts Jan 2008. Should aim at complementary design compared
• Collaborations within Europe through FP7 and with US through LARP are
crucial components in our plans and address several possible problems.
• We also revisit other collimation solutions, like cryogenic collimators,
crystals, magnetic collimators, non-linear schemes.
RWA&VP, FNAL 12/07 34
Draft Work Packages
White Paper (WP), Europe (FP7), US (LARP)
WP1 (FP7) – Management and communication
WP2 (WP, FP7, LARP) – Collimation modeling and studies
WP3 (WP, FP7, LARP) – Material & high power target modeling and tests
WP4 (WP, FP7, LARP) – Collimator prototyping & testing for warm regions
Task 1 – Scrapers/primary collimators with crystal feature
Task 2 – Phase 2 secondary collimators
WP5 (FP7) – Collimator prototyping & testing for cryogenic regions
WP6 (FP7, LARP) – Crystal implementation & engineering
RWA&VP, FNAL 12/07 35