gianluigi arduini cern - beams dept. - accelerator & beam
TRANSCRIPT
Gianluigi Arduini
CERN - Beams Dept. - Accelerator & Beam Physics Group
Acknowledgements: O. Brüning, S. Fartoukh, M. Giovannozzi, G. Iadarola, M.
Lamont, E. Métral, N. Mounet, G. Papotti, T. Pieloni, L. Rossi, G. Rumolo, B.
Salvant, R. Tomàs, J. Wenninger, F. Zimmermann and all the teams involved in
LHC and LHC Injector operation
S. Cittolin
LHC layout
• Total length: ~26.7 km • 8 arcs (aka sectors): ~2.8
km each
• 8 long straight sections: ~700 m each
• 2-in-1 magnet design with separate vacuum chambers • p-p, ion/ion, or p/ion
collisions
• beams cross in 4 points
RF
Interaction regions geometry In the IRs, the beams are first combined into a single common
vacuum chamber and then re-separated in the horizontal plane,
The beams move from inner to outer bore (or vice-versa),
The triplet quadrupoles are used to focus the beam at the IP.
194 mm
~ 260 m
Common vacuum chamber
D2
D1 D1
D2
Triplet Triplet
D1,D2 :
separation/recombination dipoles
Machine geometry in H plane
IP
beam1
beam1
beam2
beam2 ~ 40 m
Separation and crossing
a (mrad)
ATLAS -145 / ver.
ALICE 145 / ver.
CMS 145 / hor
LHCb 90 / ver
~ 8 mm
Not to scale !
q
4 TeV / 2012 !
4 mm (450 GeV)
1.3 mm (4 TeV)
Because of the tight bunch spacing and to prevent undesired parasitic collisions
in the region where the beams circulate in the common vacuum chamber:
o Parallel separation in one plane, collapsed to 0 to bring the beams to collision
o Crossing angle in the other plane.
Luminosity 2010-2012
• first operational experience with low bunch intensity
• learn to handle intense beams (~1 MJ stored energy)
• production and Higgs hunt 2010
2011 2012
• 2010, commissioning: 0.04 fb-1
• 2011, exploring the limits: 6.1 fb-1
• 2012, production: 23.3 fb-1
Summary: 2010 to 2012 • Impressive progress in performance
Parameter 2010 2011 2012 Nominal
Energy (TeV) 3.5 3.5 4.0 7.0
N ( 1011 p/bunch) 1.2 1.45 1.6 1.15
k (no. bunches) 368 1380 1380 2808
Bunch spacing 150 75 / 50 50 25
Stored energy (MJ) 25 112 140 362
e (mm rad) 2.4 2.4 2.5 3.75
b* (m) 3.5 1.5 1 0.6 0.55
L (cm-2s-1) 21032 3.51033 7.61033 1034
Beam-beam parameter/IP -0.0054 -0.0065 -0.0069 -0.0033
Average Pile-up @ beg. of fill 8 17 38 26
FfkN
L b
**
2
4 eb
FfkN
L b
**
2
4 eb
Limits on b* The triplet quadrupoles in the high luminosity IRs define the machine
aperture limit for squeezed beams, b* is constrained by:
b
e
*triplet
o the beam envelope,
o a margin TCT to triplet ,
o the crossing angle
~ 8 mm
Triplet Triplet TCT TCT
≥10.5
9
TCT=Tertiary Collimator TCT
Collimation system • Complex and high performance multi-stage collimation system.
• Collimation hierarchy has to be respected in order to achieve satisfactory protection and cleaning.
• Lower b* implies tighter collimator settings as well as alignment, beam sizes and orbit well within tolerance. We could do it only after having gained experience in orbit and optics controls and thanks to the small emittance delivered by the injectors.
TCP TCS7
Aperture
TCS6/
TCDQ TCT TCLA7
beam
Secondary halo
Pri
mar
y h
alo
Tertiary halo
Kicked beam
R. Bruce
Optics control
• Lower b* means enhancement
of errors challenging
• …but proven to be feasible.
R. Tomas et al.
FfkN
L b
**
2
4 eb
Wake fields and Impedances • Intense bunches generate electromagnetic (EM) fields when passing
inside a structure (e.g. Carbon collimators – opening of ~1 mm!!!)
• → results in an EM force, called wake field in time domain, beam-coupling
impedance in frequency domain.
• Avoid the abrupt transition for the beam fields at the location of the beam
passage (taper)
• Reduce the resistivity of the material
Abrupt transition With taper
B. Salvant
Wake fields and instabilities
Wake fiels can couple the head and tail of a bunch
s
v=b c
v=b c
Orbit (pipe axis of symmetry)
Pipe wall Induced (or ”image”) currents
”Test” particle
”Source” particle
Direct EM interaction → ”direct space-charge”
EM interaction through the pipe wall→ ”wall impedance”
x
Wake fields and Instabilities Many bunches (up to 1380 with 50 ns spacing)
Bunches can interact together (or even the head of each bunch can interact with the tail of the bunch) and in some cases begin to oscillate.
Example with 36 bunches in the LHC: oscillation pattern along the bunch train (simulation result):
s
v ~36 cm ~15 m
~1.6×1011 p
→ Coupled-bunch instabilities
N. Mounet
Beam instabilities
• In 2012 instabilities have become more critical due to
higher bunch intensity and tighter collimators settings.
The LHC is one of the few machines where instabilities
are more critical at high energy.
• Interplay between impedance (mostly due to collimators)
and two-beams phenomena (mostly beam-beam)
• Cures: • Transverse feedback (‘damper’) that measures the oscillations
and sends corrective deflections,
• Non-linear magnetic fields (sextupoles, octupoles, beam-beam)
that produce a frequency spread among particles – kill coherent
motion
Beam-beam effects
Strong non linear fields when
counter-rotating beams are
sharing vacuum chamber.
spread in betatronic
frequencies risk of
overlapping resonances driven
by magnetic errors
Head-on
Long-range
S. Fartoukh
T. Pieloni
Minimize magnetic errors
Paid off for the LHC
Devise correction
schemes and sorting
Paid off for the LHC
Initially expected to have
limit at DQBB ~ 0.005/IP
Electron cloud effects
• Electron cloud effects occur both in the warm and cold regions, their intensity increases rapidly for shorter bunch spacing. Observed as soon as we started to inject bunch trains (150 75 50 25 ns spacing):
• Vacuum pressure rise (interlock levels, beam losses…)
• Single-bunch and multi-bunch instabilities beam size growth
• Incoherent beam size growth
• Heat load on the cryogenics
Secondary emission yield [SEY]
SEY>SEYth avalanche effect (multipacting)
SEYth depends on bunch spacing and population
F. Ruggiero
G. Iadarola, G. Rumolo
Electron cloud effects • Fields induced by electrons act like wake
fields (more complex) that couple different
bunches along the train and head&tail of
each bunch and have similar adverse
effects on beam stability as impedances.
• As a result of that emittance blow-up and
beam losses H. Bartosik
Cures for electron cloud effects
• At the time of the construction of the LHC:
• NEG coating would require activation to >200 oC impractical
• Conditioning by beam-induced electron bombardment (“scrubbing”)
leading to a progressive reduction of the SEY as a function of the
accumulated electron dose tested in the laboratory (on Cu
surfaces) and in the SPS (Stainless Steel vacuum chambers)
Chosen strategy for the LHC operation with bunch trains – but it
takes time!! (in particular for 25 ns operation)
• More recently a-C coating has been successfully tested in the
SPS showing SEY as low as 1.1 possible implementation
for HL-LHC (interaction regions)
Pile-up density • 25 reconstructed vertices – in
a luminous region of 4.3 cm r.m.s. length
• Peaks of >40 events per bunch crossing observed with luminosities in the range of 7x1033 cm-2s-1
• This is the maximum acceptable by the experiments with the present design upgrade is vital
• Pile-up proportional to luminosity per bunch more bunches (i.e. 25 ns) is better
Z mm
Expected peak performance – 6.5 TeV
50 ns 25 ns
Beta* [m] 0.4 0.4
e*[mm] at start of fill 1.6 1.9
Max. Bunch Population [1011 p] 1.6 1.15
Max. Number of bunches/colliding pairs IP1/5 1260 2508
Bunch length (4 )[ns]/ (r.m.s.) [cm] 1.35/10.1 1.35/10.1
Max. Beam Current [A]/population[1014 p] 0.36 / 2.0 0.52 / 2.9
Max. Stored energy [MJ] 210 300
Peak luminosity [1034 cm-2s-1] in IP1/5 2.0 1.5
Half External Crossing angle IP1/5 [mrad] 120 155
Beam-beam tune shift (start fill)/IP [0.001] 7.3 4.3
Min. beam-beam separation () dsep 9.3 12
Maximum Average pile-up (=85 mb) 120 44
Pile-up density an
issue for 50 ns
Baseline scenario Provided
scrubbing
successful
Why an LHC Upgrade? • Motivations:
• Reduce statistical error by factor 3
• Radiation damage limit of IR quadrupoles (~400 fb-1)
• Radiation damage to detectors
• Target (very ambitious) of the upgrade: 200 – 300 fb-1/y (×10 today)
• 2010-2012 experience:
• Head-on beam-beam limit higher than initially expected
• Single bunch with > 3x1011 ppb with 2.5 mm emittance accelerated in the
SPS
• Low b* optics successfully tested in MD
• Limit on total beam current in LHC due to several systems (RF, dump,
vacuum, collimator robustness, machine protection, RP, …) at ultimate
value (25 ns)
• Pile-up density replaces beam-beam as HL-LHC constraint
Ideas for the Upgrade • Operation at pile-up density limit:
choose parameters that allow higher than design pile-up
Low b*
Crab Cavities as tool to maximize overlap among colliding bunches (i.e. virtual luminosity)
leveling mechanisms for controlling performance during run
dynamic b* squeeze
crossing angle and Long-range and beam-beam wire compensators
FfkN
L b
**
2
4 eb
L [1034 cm-2s-1]
t [h]
Virtual peak
luminosity (F=1)
leveling at
5x1034cm-2s-1
teff=15 h, Tta=5 h leveling at
2.5x1034cm-2s-1
teff=30 h, Tta=5 h
Ideas for the Upgrade
F. Zimmermann
FfkN
L b
**
2
4 eb
HL-LHC Performance Estimates O. Brüning
Parameter Nominal 25ns – HL-LHC 50ns – HL-LHC
Bunch population Nb [1011] 1.15 2.2 3.5
Number of bunches 2808 2808 1404
Beam current [A] 0.58 1.12 0.89
Crossing angle [mrad] 300 590 590
Beam separation [] 9.9 12.5 11.4
b* [m] 0.55 0.15 0.15
Normalized emittance en [mm] 3.75 2.5 3.0
eL [eVs] 2.51 2.51 2.51
Relative energy spread [10-4] 1.20 1.20 1.20
r.m.s. bunch length [m] 0.075 0.075 0.075
Peak Luminosity [1034 cm-2s-1] 1 7.4 8.5
Virtual Luminosity [1034 cm-2s-1] 1.2 24 26
Very low b* (10 cm)
• …of course it requires larger aperture triplets
Beam size [mm] and dispersion (IR4IR6)
at 3.5 TeV (for e=3.5 mm)
IR4 IR5 IR6
Tunes vs. dp
ATS=Achromatic Telescopic Squeeze - S. Fartoukh
b* [m]
FfkN
L b
**
2
4 eb
Hardware for the Upgrade • Upgrade of the Experiments to
cope with higher pile-up density
• New high field/larger aperture interaction region magnets
• Cryo-collimators and high field 11 T dipoles in dispersion suppressors
• New collimators (lower impedance)
• Additional cryo plants (P1, P4, P5)
• Crab Cavities to take advantage of the small b*
– SC links to allow power converters to be moved to surface
• Upgrade of the intensity in the Injector Chain (see M. Migliorati)
• The progress in the performance of the LHC has been so far breath-taking
• This has been possible thanks to the quality of the design, construction and installation and to the thorough preparation in the injectors
• Some of the (beam dynamics) challenges that had to be faced have been outlined
• Luminosity performance and choices for the upgrade are now constrained by the acceptable detector pile-up density but
• They are pushing even further the above challenges…
Summary
