injection, extraction and protection of the clic damping rings
DESCRIPTION
Injection, extraction and protection of the CLIC damping rings. R. Apsimon TE-ABT-BTP CLIC Workshop 2013 30 th January 2013. Design requirements. Primary goal: Minimise length of injection & extraction insertions Minimises beam instability due to collective effects - PowerPoint PPT PresentationTRANSCRIPT
Injection, extraction and protection of the CLIC damping rings
R. ApsimonTE-ABT-BTP
CLIC Workshop 201330th January 2013
CLIC Workshop 2013 2
Design requirements
• Primary goal:– Minimise length of injection & extraction insertions• Minimises beam instability due to collective effects
– Must respect physical limitations of elements• Maximum field in kicker and septum magnets
– Homogeneity and stability requirements• Aperture constraints
• Secondary goal:– Protect machine from injection/extraction failures
30/01/2013
CLIC Workshop 2013 3
Damping ring requirements
• Keep equilibrium emittance small– Relies on high degree of symmetry• Injection & extraction cells must be symmetrical
– Kicker and septa orders swapped between injection and extraction
– Identical design for both systems– Injection emittance larger than extraction• More kicker length required• Longer cell length• Injection system more critical for design
30/01/2013
CLIC Workshop 2013 4
Kicker and septum magnet parametersParameter Injection Extraction
Aperture (mm) 12*Repetition rate (Hz) 50
Vacuum (mbar) 10-10
Pulse voltage per stripline (kV) ±12.5Stripline pulse current [50Ω load] (A) ±250
Rise and fall times (ns) 1000Pulse flat top duration (ns) ~160
Flat top reproducibility ±1x10-4
Stability ±2x10-3 ±2x10-4
Field homogeneity (%) ±0.1 ±0.01
Parameter Thin septum Thick septumGap height (mm) 5 5
Septum thickness* (mm) 5 / 10 20 / 25Pulsed mode DC DCLength (m) ~ 0.85 ~ 2.00
Injection stability ±2x10-4 ±2x10-5
Extraction stability ±2x10-5 ±2x10-6
Deflection angle (mrad) 17 190Gap field (T) 0.19 0.91
Coil current (kA) 0.76 3.60
Kicker parameters Septa parameters
Kicker and septa designs are optimised for their geometries to provide the maximum field, without exceeding their respective stability requirements.
For the optimisation of the injection and extraction optics, their lengths are the only parameters which will varied.
* The kicker aperture for the baseline design is 20mm; 12mm is used for the injection cell optimisation. This is the aperture of the long straight sections and is near optimal.
30/01/2013
CLIC Workshop 2013 5
Parameterisation of injection systemLcell
Lthick Lthin Ldr1 Ldr2 Lkick
Optimisation parameter
System constraint Beam constraints
Cell symmetry 6σ beam < aperture
Minimise Beam to fit d-quad
Clear thin septum blade Beam to fit kicker ap
Clear thick septum blade Match beam after cell
Clear quad radius
Turns out matching cell is required after injection cell to meet aperture constraint.
Extraction system is identical but septum and kicker orders reversed.
30/01/2013
Septum magnets
Stripline kicker
CLIC Workshop 2013 6
Injection/extraction parametersSepta in vacuum Septa not in vacuum
Kicker parametersAperture 12 mm 12 mmVoltage ±12.5 kV ±12.5kV
Kicker length 2.4 m 2.6 mThin septum parameters
Gap field 0.2T 0.2T
Length 0.9 m 0.9 m
Thick septum parameters
Gap field 1T 1T
Length 2.0 m 2.0 m
Inj/ext cell length 7.9 m 9.4 mMatching cell length 2.4 m 3.1 m
Total length 10.3 m 12.5 m
30/01/2013
CLIC Workshop 2013 7
Failure modes
• Fast failures– Particles hit aperture within few turns• E.g. injection and extraction kicker failures
– Passive protection needed (collimators, absorbers)• Slow failures– Failure slow enough to abort/dump beam before
it hits aperture • E.g. magnet power supply failure
– Use extraction system to remove beam
30/01/2013
CLIC Workshop 2013 8
Injection kicker failure modes
• Inductive adder level failure– 20 levels: supply ~700V each (See J. Holma’s talk)– Consider up to 3 levels failing simultaneously
• Assumed to be caused by failure of FETs on level• ~8σ event, so realistic worst-case scenario.
• Total inductive adder failure• Likely to be due to a trigger timing error• ALL particles considered dangerous and hit aperture shortly
downstream of injection• Injection collimator designed to capture full 6σ beam (+
tolerances)
30/01/2013
CLIC Workshop 2013 9
Collimator considerations [1]
• Number of σ that can pass through aperture
Region A1/2 (mm) H-plane V-plane H-plane V-planeLSS 12 ≥13.3 ≥65.3
Arc 20 ≥33.7 ≥126.1
Injection cell Extraction cell
1st quad 20 17.1 246.5 17.1 246.5
Septum - 7.1 242.7 9.3 110.8
Kicker 12 9.9 263.0 8.7 119.9
δ = alignment tolerance = 2mmA1/2 = physical half-aperture
Acceptance calculations at injection emittance, assuming there is a pre-damping ring
30/01/2013
CLIC Workshop 2013 10
Collimation considerations [2]
• Beam aperture critical in injection/extraction regions– Use absorbers to protect septa (fixed position)– Collimators to protect rest of machine (moveable)
• Collimation scheme depends on whether septa are in vacuum or not– Dependence on injection trajectory
30/01/2013
CLIC Workshop 2013 11
Septa in vacuum: H-plane
30/01/2013
Quad
Septum
Kicker
Collimation absorbers
Matching cell
CLIC Workshop 2013 12
Septa not in vacuum: H-planesteeper angle
30/01/2013
Quad
Septum
Kicker
Collimation absorbers
Matching cell
CLIC Workshop 2013 13
Comments on collimator plots
• Beam envelope– 6σ envelope ± 2mm tolerance
• First collimator– Needed to stop particles hitting aperture before
reaching second collimator• Second collimator– Designed to completely capture beam for total
kicker failure• Scattering + secondary particles not yet considered
30/01/2013
CLIC Workshop 2013 14
Comparison of schemes
• Septa in vacuum• Smaller beams; good aperture clearance• >4 m reduction in total length of DR
– This is almost entirely drift length
• Septa not in vacuum• Efficient collimation• Simpler septum design and operation
30/01/2013
Tracking simulations• Tracking done for failure of 3 inductive adder levels– 1000 particles for 100 turns
• Uniform random number generators: 6σ ± 2mm phase space• Polar coordinates to create oval beams
– 340 “dangerous” particles• Exceed 6σ ± 2mm phase space of nominal orbit• ~3.45% loss for Gaussian beam
Turn number % absorbed
At injection 37.4%
1 turn 52.1%
2 turns 92.4%
3 turns 95.9%
4 turns 97.4%
10 turns 99.1%
All particles captured by absorbers + collimators; no losses in kickers or elsewhere.
Remaining 0.9% of particles on edge of phase space limit and survive for many turns.
CLIC Workshop 2013 16
Emittance estimates
• Taken from tracking data– After 100 turns
Horizontal Vertical
@ Injection 4.5 x 10-8 1.1 x 10-9
Nominal orbit 1.2 x 10-7 1.2 x 10-9
Poorly injected 2.9 x 10-7 3.1 x 10-9
30/01/2013
CLIC Workshop 2013 17
Phase space: no collimationPhase space plot at second injection collimator
30/01/2013
CLIC Workshop 2013 18
Phase space coverage: 1 turn
Blue: phase space of nominal orbitGreen: Phase space of poorly injected beam (3 levels failed) without collimationRed: Phase space of poorly injected beam (3 levels failed) with collimationBlack: Phase space confined by collimation
30/01/2013
CLIC Workshop 2013 19
Phase space coverage: 2 turns
Blue: phase space of nominal orbitGreen: Phase space of poorly injected beam (3 levels failed) without collimationRed: Phase space of poorly injected beam (3 levels failed) with collimationBlack: Phase space confined by collimation
30/01/2013
CLIC Workshop 2013 20
Phase space coverage: 3 turns
Blue: phase space of nominal orbitGreen: Phase space of poorly injected beam (3 levels failed) without collimationRed: Phase space of poorly injected beam (3 levels failed) with collimationBlack: Phase space confined by collimation
30/01/2013
CLIC Workshop 2013 21
Phase space coverage: 4 turns
Blue: phase space of nominal orbitGreen: Phase space of poorly injected beam (3 levels failed) without collimationRed: Phase space of poorly injected beam (3 levels failed) with collimationBlack: Phase space confined by collimation
30/01/2013
CLIC Workshop 2013 22
Dump system considerations
• Latency– How many turns before beam can be dumped?
• Location and space constraints
30/01/2013
CLIC Workshop 2013 23
Breakdown of latency
• Signal time of flight to dump kicker• ~1μs
• Latency of electronics• <1μs
• Kicker rise time• ~1μs
• Time for 1 turn of ring (circumference: 400-450m)• 1.3-1.5μs
• ~2-3 turns of ring required to dump beam
30/01/2013
CLIC Workshop 2013 24
Location + space constraints
• Avoid– Regions with synchrotron radiation– High dispersion regions• Near injection or extraction only suitable places.
• Dedicated dump cell?– Would add ~10m in each straight section• Unacceptable increase in length
– Can extraction cell be used as dump system?
30/01/2013
CLIC Workshop 2013 25
Technical challenges
• Kicker must fire in two modes– Extraction mode (±12.5kV)– Dump mode (±17.5kV)• Need to extract beam with injection emittance
• Separate dumped beam from extracted
30/01/2013
CLIC Workshop 2013 26
How to achieve 2 kicker modes
• Separate inductive adder into 2 banks of levels– “Bank 1” contains 20 levels– “Bank 2” contains 8-10 levels– Extraction trigger discharges Bank 1– Dump trigger discharges Banks 1 and 2
• Triggering system likely to be challenging– Need to test reliability of 2-mode trigger
30/01/2013
CLIC Workshop 2013 27
Kicker triggering
Bank 1 Bank 2
Trigger select
“Extract”
Bank 1 Bank 2
Trigger select
“Dump”
30/01/2013
CLIC Workshop 2013 28
Kicker failure modes
• Extraction mode– Both banks fire: beam dumped → safe– Bank 1 fires: beam extracted → safe– Bank 2 fires: beam absorbed by septum absorber and collimator → safe– Neither bank fires: beam remains in ring
• Dump mode– Both banks fire: beam dumped → safe– Bank 1 fires: beam extracted → NOT SAFE– Bank 2 fires: beam absorbed by septum absorber and collimator → safe– Neither bank fires: beam remains in ring
30/01/2013
CLIC Workshop 2013 29
Separate extracted and dumped beams
• Start of extraction line– Kicker gives larger deflection to dumped beam– Use defocussing quad to further separate beams
• Septum magnet to separate extraction and dump lines– Use same septa design as in extraction system
30/01/2013
CLIC Workshop 2013 30
Current design: h-plane
30/01/2013
Dumped beam
Extracted beam
Septum magnets
Dipole
CLIC Workshop 2013 31
Consideration of damping time [1]
• Time needed to damp beam:– Injection: 54 μm rad (x), 1.3 μm rad (y)– Extraction: 500 nm rad (x), 5 nm rad (y)– Equilibrium: 470 nm rad (x), 4.8 nm rad (y)
t
eqinjeq et
30/01/2013
CLIC Workshop 2013 32
Consideration of damping time [2]
• ~8.5 damping times to reach design emittance– 17ms (injection period 20ms)
• How long to charge inductive adder?– Currently unknown: If not sufficient then…• Add levels in Bank 2 to compensate missing charge?• Reduce storage time by ~1 damping time?
– 4% increase in extraction emittance; acceptable?• Reduce damping time?
– New wiggler design reduces damping time to 1.8ms– Increases equilibrium emittance slightly; net gain?
30/01/2013
CLIC Workshop 2013 33
Comments on design
• Septa in vacuum?– Easier if extraction septa NOT in vacuum• More lever-arm; less length needed to separate beams• Twiss parameters more controllable
• Final quad needed in dump line– Control spot size at dump block
30/01/2013
CLIC Workshop 2013 34
Radiation length
• Need minimum 5 rad. lengths for 2.86 GeV e-
– Use 10 rad. lengths for dump block– Use 5 rad. lengths for absorbers and collimators
Material Density (kg m-3)
Radiation length (m) ΔTinst (Inj)(K)
ΔTinst (Ext)(K)
Beryllium 1.84 X 103 0.353 0.002 4.9
Carbon 2.25 X 103 0.188 0.006 15.0
Titanium 4.50 X 103 0.036 0.007 18.1
Copper 8.93 X 103 0.014 0.01 24.2
Tungsten 19.3 X 103 0.0035 0.025 60.2
Higher density means more back scattering, but shorter radiation length
30/01/2013
CLIC Workshop 2013 35
Material choice
• In DR, space is limited– short radiation length and low back-scattering• Use titanium: ~20cm for collimators and absorbers
• Dump block– Space not limited• Use carbon for dump block• Surround block in higher mass material (e.g. concrete)
to contain radiation.
30/01/2013
CLIC Workshop 2013 36
Conclusions
• Injection and extraction optics– Fully optimised for both septum magnet designs• Septum outside vacuum seems better
• Machine protection– Injection• Tracking simulations show DR collimation is sufficient
– Extraction• Combined dump looks promising; needs further studies
30/01/2013