Performance limitations of the present LHC

Name Event DateName Event Date11

F. RuggieroCARE-HHH-AMT WAMDO Workshop, CERN, 3–6 April

2006CERN

Performance limitations Performance limitations of the present LHCof the present LHC

and LHC luminosity upgrade pathsand LHC luminosity upgrade paths

triplet magnetsBBLR


F. Ruggiero Performance limitations of the present LHCCERN

OutlineOutline

• Beam-beam limit and nominal LHC Beam-beam limit and nominal LHC performanceperformance

• Luminosity optimization and operational Luminosity optimization and operational marginsmargins

• LHC upgrade paths and beam induced heat LHC upgrade paths and beam induced heat loadsloads

• Catalog of beam performance limitationsCatalog of beam performance limitations• IR aperture: flat beams and quad re-IR aperture: flat beams and quad re-

alignmentalignment• Magnet quench limitsMagnet quench limits• Collimation, impedance, and beam Collimation, impedance, and beam

intensityintensity• Electron cloud effects Electron cloud effects • Feedback systems and emittance Feedback systems and emittance

preservationpreservation



4pb

HO

rN 2)/(

2

sep

HOparLR d

n

/ IP / IP no. of IP’sno. of IP’s QQbbbbtotaltotal

SPSSPS 0.0050.005 33 0.0150.015

Tevatron (pbar)Tevatron (pbar) 0.01-0.020.01-0.02 22 0.02-0.040.02-0.04

RHICRHIC 0.0020.002 44 ~0.008~0.008

LHC (nominal)LHC (nominal) 0.00340.0034 3 3 ~0.01~0.01

tune shift from head-on collisions (primary IP’s)

tune shift from long-range collisionsnpar parasitic collisions around each IP

conservative value for total tune spread based on SPScollider experience

Beam-Beam tune spread for round beams

increases for closer bunches or reduced crossing angle

limit on limits Nb/(

*

ccsep

)(

s

sd relative beam-beam separation for full crossing angle c

high-lumi in IP1 and IP5 (ATLAS and CMS), halo collisions in IP2 (ALICE) and low-lumi in IP8 (LHC-b)



Beam-Beam tune footprintsBeam-Beam tune footprints

Tune footprints corresponding to betatron amplitudes Tune footprints corresponding to betatron amplitudes extending from 0 to 6extending from 0 to 6 for LHC nominal ( for LHC nominal (red-dottedred-dotted), ), ultimate (ultimate (green-dashedgreen-dashed), and “large Piwinski parameter” ), and “large Piwinski parameter” configuration (configuration (blue-solidblue-solid) with alternating H-V crossing only ) with alternating H-V crossing only in IP1 and IP5. in IP1 and IP5.



LHC working points in collisionLHC working points in collision

The beam-beam tune footprint has to be accommodated The beam-beam tune footprint has to be accommodated in between low-order betatron resonances to avoid in between low-order betatron resonances to avoid diffusion and bad lifetime. More resonance-free space near diffusion and bad lifetime. More resonance-free space near the coupling resonance the coupling resonance good coupling compensation good coupling compensation may allow may allow QQbbbb~0.015~0.015



Luminosity optimizationLuminosity optimization

Collisions with full crossing angle Collisions with full crossing angle cc

reduce luminosity by a geometric factor reduce luminosity by a geometric factor FF

maximum luminosity below beam-beam limit maximum luminosity below beam-beam limit ⇒ ⇒ short bunches and minimum crossing angle (baseline short bunches and minimum crossing angle (baseline

scheme)scheme)

H-V crossings in two IP’s ⇒ no linear tune shift due to long H-V crossings in two IP’s ⇒ no linear tune shift due to long rangerange

total linear bb tune shift also reduced by total linear bb tune shift also reduced by FF

INNfn

Ln

b*2

2brevb

44

peak luminosity for head-on collisions

round beams, short Gaussian bunches

transverse beam size at IP

I = nbfrevNb total beam current

• long range beam-beam• collective instabilities• synchrotron radiation• stored beam energy

2

n normalized emittance

Nb/n beam brightness• head-on beam-beam• space-charge in the

injectors• transfer dilution

2

*211/

zcF

FrN

Qn

pbyxbb 2



Minimum crossing angleMinimum crossing angleBeam-Beam Long-Range collisions:Beam-Beam Long-Range collisions:• perturb motion at large betatron perturb motion at large betatron

amplitudes, where particles come amplitudes, where particles come close to opposing beamclose to opposing beam

• cause ‘diffusive’ (or dynamic) cause ‘diffusive’ (or dynamic) aperture, high background, poor aperture, high background, poor beam lifetimebeam lifetime

• increasing problem for SPS, increasing problem for SPS, Tevatron, LHC, i.e., for operation Tevatron, LHC, i.e., for operation with larger # of buncheswith larger # of bunches

higher beam intensities or smaller * require larger crossing angles to preserve dynamic aperture and shorter bunches to avoid geometric luminosity loss

baseline scaling: c~1/√* , z~*

nθ

c

n11bpar

θ

cda m75.3

A5.036

m75.3

10323

INnd

dynamic aperture caused by npar parasitic collisions around two IP’s

*θ angular beam

divergence at IP



Various LHC upgrade options Various LHC upgrade options

parameterparameter symbol [unit]symbol [unit] nominalnominal ultimateultimate shorter shorter bunchbunch

longer longer bunchbunch

number of bunchesnumber of bunches nnbb28082808 28082808 56165616 936936

protons per bunchprotons per bunch NNbb [10 [101111]] 1.151.15 1.71.7 1.71.7 6.06.0

bunch spacingbunch spacing ttsep sep [ns][ns] 2525 2525 12.512.5 7575

average beam average beam currentcurrent

II [A] [A] 0.580.58 0.860.86 1.721.72 1.01.0

normalized normalized emittanceemittance

nn [µm] [µm] 3.753.75 3.753.75 3.753.75 3.753.75

longitudinal profilelongitudinal profile GaussiaGaussiann

GaussiaGaussiann

GaussiaGaussiann

flatflat

rms bunch lengthrms bunch length zz [cm] [cm] 7.557.55 7.557.55 3.783.78 14.414.4

ß* at IP1&IP5ß* at IP1&IP5 ** [m] [m] 0.550.55 0.500.50 0.250.25 0.250.25

full crossing anglefull crossing angle cc [µrad] [µrad] 285285 315315 445445 430430

Piwinski parameterPiwinski parameter c c zz/(2/(2**)) 0.640.64 0.750.75 0.750.75 2.82.8

peak luminositypeak luminosity LL [10 [1034 34 cmcm-2-2 s s--

11]]1.01.0 2.32.3 9.29.2 8.98.9

luminosity lifetimeluminosity lifetime L L [h][h] 15.515.5 11.211.2 6.56.5 4.54.5

events per crossingevents per crossing 1919 4444 8888 510510

luminous region luminous region lengthlength

lumlum [mm] [mm] 44.944.9 42.842.8 21.821.8 36.236.2



Heat loads per beam aperture Heat loads per beam aperture for various LHC upgrade options for various LHC upgrade options

parameterparameter symbol [unit]symbol [unit] nominalnominal ultimateultimate shorter shorter bunchbunch

longer longer bunchbunch

protons per bunchprotons per bunch NNbb [10 [101111]] 1.151.15 1.71.7 1.71.7 6.06.0

bunch spacingbunch spacing ttsep sep [ns][ns] 2525 2525 12.512.5 7575

average beam currentaverage beam current II [A] [A] 0.580.58 0.860.86 1.721.72 1.01.0

longitudinal profilelongitudinal profile GaussianGaussian GaussiaGaussiann

GaussiaGaussiann flatflat

rms bunch lengthrms bunch length zz [cm] [cm] 7.557.55 7.557.55 3.783.78 14.414.4

Average Average electron-cloudelectron-cloud heat load at 4.6heat load at 4.6––20 K in 20 K in the arc for the arc for RR =50% and =50% and maxmax=1.4 (in parentheses =1.4 (in parentheses for for maxmax=1.3)=1.3)

PPecloudecloud [W [W /m]/m]1.071.07

(0.44)(0.44)1.041.04

(0.59)(0.59)13.3413.34

(7.85)(7.85)0.260.26

(0.26)(0.26)

Synchrotron radiationSynchrotron radiation heat load at 4.6heat load at 4.6––20 K20 K PP [W [W /m]/m] 0.170.17 0.250.25 0.500.50 0.290.29

Image currentsImage currents power at power at 4.64.6––2020 KK PP [W [W /m]/m] 0.150.15 0.330.33 1.871.87 0.960.96

Beam-gas scatteringBeam-gas scattering heat load at 1.9 K for 100-heat load at 1.9 K for 100-h beam lifetime (in h beam lifetime (in parentheses for a 10-h parentheses for a 10-h lifetime). It is assumed lifetime). It is assumed that elastic scattering that elastic scattering (~40% of the total cross (~40% of the total cross section) leads to local section) leads to local losses.losses.

PPgasgas [W [W /m]/m]0.0380.038

(0.38)(0.38)0.0560.056

(0.56)(0.56)0.1130.113

(1.13)(1.13)0.0660.066

(0.66)(0.66)



LHC performance LHC performance limitations from IR optics limitations from IR optics

constraintsconstraints• The triplet aperture is The triplet aperture is

completely filled for nominal completely filled for nominal LHC conditionsLHC conditions

• However there are two ways to However there are two ways to better use the available better use the available aperture with “minimal” aperture with “minimal” modifications:modifications:• Flat beamsFlat beams• IR quadrupole re-alignmentIR quadrupole re-alignment



Luminosity with Flat BeamsLuminosity with Flat BeamsFlat beams means aspect ratio Flat beams means aspect ratio r r ≠≠11 at the IP: at the IP:

*

*

*y

*x**** ,/ , ,

y

xyxyx rrr

The X-ing plane is always the plane where the The X-ing plane is always the plane where the beam size is largest at the IP (beam size is largest at the IP (i.e.i.e. smallest at smallest at the triplet):the triplet):• To gain aperture in the triplet (smaller X-ing angle To gain aperture in the triplet (smaller X-ing angle

and better matching of beam aspect ratio to beam-and better matching of beam aspect ratio to beam-screen shape)screen shape)

• To gain luminosity (geometric loss factor closer to To gain luminosity (geometric loss factor closer to unity)unity) ,

21

2

*zc

on-head roundflat

r

LL **

y*x *



Flat beamsFlat beams• Interesting approach, flat beams may increase Interesting approach, flat beams may increase

luminosity by ~20-30% with reduced crossing luminosity by ~20-30% with reduced crossing angleangle

• Symmetric doublets studied by J. Johnstone Symmetric doublets studied by J. Johnstone (FNAL) require separate magnetic channels, (FNAL) require separate magnetic channels, i.e.i.e. dipole-first, Crab cavities or special quads dipole-first, Crab cavities or special quads

• Tune footprints are broader than for round Tune footprints are broader than for round beams,beams, since there is only partial since there is only partial compensation of parasitic beam-beam compensation of parasitic beam-beam encounters by the H/V crossing scheme. More encounters by the H/V crossing scheme. More work needed to evaluate nonlinear resonance work needed to evaluate nonlinear resonance excitation.excitation.

• Probably requires BB Long Range Probably requires BB Long Range compensationcompensation

• Recently S. Fartoukh has found an interesting Recently S. Fartoukh has found an interesting flat beam solution with anti-symmetric flat beam solution with anti-symmetric LHC LHC baselinebaseline triplets triplets



Beam aspect ratio vs triplet apertureBeam aspect ratio vs triplet aperture beam screen orientation for H/V schemebeam screen orientation for H/V scheme

⇒ Find the optimum matching between beam-screen and beam aspect ratio

Effect of increasing thebeam aspect ratio at the IP(and decreasing the vert. X-angle)

Effect of decreasing thebeam aspect ratio at the IP(and increasing the vert. X-angle)



Pushing the LHC luminosity by 10-20%Pushing the LHC luminosity by 10-20%

CaseCase xx**

[cm][cm]yy

** [cm][cm]

cc[[rarad]d]

nn11 at at triplettriplet

geometric geometric lumi loss lumi loss

[%][%]L/LL/Lnomnom

NominalNominal

r=1r=15555 cmcm

5555 5555 285285 ~7~7 83.983.9 1.001.00

Flat Flat

r=2r=25555 cmcm

110110 27.527.5 201201 ~7~7 95.195.1 1.131.13

Flat Flat

r=1.6r=1.65555 ccmm

8888 34.434.4 225225 ~7.5~7.5 92.792.7 1.101.10

Flat Flat

r~1.7r~1.75151 ccmm

8888 3030 225225 ~7~7 92.792.7 1.181.18 All these cases are allowed by the nominal LHC hardware: layout, power supply, optics anti-symmetry, beam screen orientation in the triplets (only changing the present H/V scheme into V/H scheme)



IR quadrupole re-alignment IR quadrupole re-alignment ((R. R. TomTomààss))

• Aperture gain of up to 6 mm by Q2 re-Aperture gain of up to 6 mm by Q2 re-alignmentalignment

• Find optimum for aperture and/or energy Find optimum for aperture and/or energy depositiondeposition

• Present orbit correctors may not be strong Present orbit correctors may not be strong enoughenough



Magnet quench levelsMagnet quench levelsfollow-up of CARE-HHH-AMT workshop (P. Pugnat)follow-up of CARE-HHH-AMT workshop (P. Pugnat)

• Review of past estimates for LHC dipoles Review of past estimates for LHC dipoles (D. Leroy)(D. Leroy)• Continuous losses: 10 mWContinuous losses: 10 mW//cmcm3 3 or 0.4 W/m of cable produces produces TT

< 0.2< 0.2 K K with the insulation selected for MBswith the insulation selected for MBs ~10 ~1077 p/s at 7 p/s at 7 TeVTeV• Transient losses: Transient losses: enthalpy marginenthalpy margin 1 mJ/cm1 mJ/cm3 3 from insulated from insulated

conductorconductor andand 35 mJ/cm35 mJ/cm3 3 from LHefrom LHe (if (if ttlossloss > 8 ms)> 8 ms)• LHC & Magnet Operation LHC & Magnet Operation (R. Schmidt & S. Fartoukh)(R. Schmidt & S. Fartoukh)

• During the ramp, quench margins of MB’s & MQ’ decrease During the ramp, quench margins of MB’s & MQ’ decrease significantlysignificantly

• During the squeeze the margin of some quadrupoles in During the squeeze the margin of some quadrupoles in experimental insertions could decrease.experimental insertions could decrease.

• Quench Levels and Transient Beam Losses at HERA Quench Levels and Transient Beam Losses at HERA (K. (K. Wittenburg)Wittenburg)• Empirical approach: Empirical approach:

• adiabatic approximation for quench level: 2.1 mJ/cmadiabatic approximation for quench level: 2.1 mJ/cm33 for for TTcscs = 0.8 K = 0.8 K

• cooling & MPZ concept taken as safety margins,cooling & MPZ concept taken as safety margins,• x16 the threshold in p/s for continuous loss rate (from x16 the threshold in p/s for continuous loss rate (from

Tevatron)Tevatron)

• Experiences & Lessons: Experiences & Lessons: • Quenches occurred at about a factor 5 below expectationQuenches occurred at about a factor 5 below expectation• BLM’s cannot protect against instantaneous lossesBLM’s cannot protect against instantaneous losses

Insertion Magnets and Beam Heat Loads

R. Ostojic, AT/MEL 17

Conclusions for LHC IR magnets

• Heat loads associated to pp collisions are considerable in the experimental insertions, in particular in the low-beta triplets.

• Thermal properties of the coils of both types of low-beta quadrupoles were experimentally studied, and confirm a safety factor of 3 with respect to expected heat load for nominal luminosity.

• MQM and MQY quadrupoles have insulation schemes analogous to the MB. Similar thermal properties could be expected, but have not been experimentally verified.

• Magnets operating at 4.5 K are expected to have higher quench limits for transient losses, but lower for continuous losses than at 1.9K.

Magnet Coil insulationOperating

temperatureConditions/Reference

Temperature margin

Heat reserve (transient losses)

Peak power density

Temperature margin

Heat reserve (transient losses)

Peak power density

MB 2x50mu (50% overlap) + 73 mu (2 mm gap) 1.9 K 7 K 38 mJ/cm3 10 mW/cm3 1 K 0.8 mJ/cm3 5 mW/cm3 LPR 44; Meuris et al. (1999)MQXA 2x25mu (50% overlap) + 60 mu (2 mm gap) 1.9 K 8.2 K 55 mJ/cm3 1.3 K 1.3 mJ/cm3 4 mW/cm3 Kimura et al, IEEE Tran SC., 9(1999)1097MQXB 2x25mu (55% overlap) + 50 mu (2 mm gap) 1.9 K 8 K 50 mJ/cm3 1.2 K 1.2 mJ/cm3 0.4 mW/g Mohkov et al., LPR 633MQM 2x25mu (50% overlap) + 55 mu (2 mm gap) 1.9 K 7.5 K 50 mJ/cm3 10 mW/cm3 1 K 1.0 mJ/cm3 5 mW/cm3MQM 2x25mu (50% overlap) + 55 mu (2 mm gap) 4.5 K 6.5 K 75 mJ/cm3 1.2 K 5 mJ/cm3 2 mW/cm3MQY 2x25mu (50% overlap) +55 mu (2 mm gap) 4.5 K 6.5 K 75 mJ/cm3 1.4 K 5 mJ/cm3 2 mW/cm3MQTL B-stage epoxy impregnated 4.5 K 6.5 K 75 mJ/cm3 2 K 5 mJ/cm3 1.0 mW/cm3 R.Wolf, Pr comm., 28 July 2004

Injection Collision



Estimate of Quench LimitsEstimate of Quench Limits Example of Results for transient losses Example of Results for transient losses

(Available for all LHC magnet types)(Available for all LHC magnet types)

Magnet typeMagnet type Cable typeCable type Op-T (K)Op-T (K)

Enthalpy (mJoule/cmEnthalpy (mJoule/cm33))

Fast perturbationFast perturbation Slow perturbation Slow perturbation (no insulation)(no insulation)

< 0.1 ms < 0.1 ms > 100 ms> 100 ms

MBMB Type-1Type-1 1.91.9 1.541.54 56.5556.55

MBMB Type-2Type-2 1.91.9 1.451.45 56.4156.41

MQMQ Type-3Type-3 1.91.9 4.244.24 70.5370.53

MQMCMQMC Type-4Type-4 1.91.9 1.511.51 49.9749.97

MQMLMQML Type-4Type-4 1.91.9 1.511.51 49.9749.97

MQMMQM Type-7Type-7 1.91.9 1.511.51 49.9749.97

MQMMQM Type-7Type-7 4.54.5 2.412.41 9.879.87

MQMLMQML Type-4Type-4 4.54.5 2.412.41 9.879.87

MQYMQY Type-5Type-5 4.54.5 2.892.89 12.1512.15

MQYMQY Type-6Type-6 4.54.5 3.803.80 15.3115.31

from A. Siemkofrom A. Siemko et al., et al., CERN LTC 19 October 2005CERN LTC 19 October 2005

G. Robert-Demolaize

Efficiency of the Cleaning System

The LHC Cleaning System should allow to run the machine close to the quench limit of the super-conducting magnets for the specified lifetime:

cdilqp LRN /max

Allowed

intensity

Quench threshold

(7.6 ×106 p/m/s @ 7 TeV)

Dilution

Length

(50 m)

Beam lifetime

(e.g. 0.2 h

minimum)

Cleaning inefficiency

=Number of escaping p (>10)

Number of impacting p (6)

=> SIMPLIFIED DEFINITION OF QUENCH LIMIT !

=> Major role of the quench limit on maximum intensity of the machine !

G. Robert-Demolaize

Maximum allowed intensity

To achieve LHC design intensity, we require the following local cleaning inefficiencies:

=> used as input for quench limits in loss maps!

(assuming simplified quench limits).

G. Robert-Demolaize

Phase 1 – Injection & Early Physics

G. Robert-Demolaize

Phase 2 – Collision Optics

LHC Collimation Team 23

Overall System Status 7 TeV

• Status Chamonix 2005:

S. Redaelli et al, Chamonix 2005

Up to 5 times above quench limit at various locations in experimental insertions…

0.2 h lifetime. Perfect cleaning & beam set-up.

LHC Collimation Team 24

Latest 7 TeV Results with Collimation Full LHC System

• Understand LHC collimation system better and better…

Black thin lines: CollimatorsBlue lines: SC apertureRed lines: Warm aperture

Fixed successfully all quench problems around the ring (tertiary collimators), except basic system limitation downstream of IR7! (Convert blue spikes into black spikes)

Compatible with expected limitation from impedance (~50%).

Improve with phase 2!7 TeV, 0.2 h lifetime, perfect cleaning&beam

Frank Zimmermann, GSI Meeting 31.03.2006

25

prototype LHC collimator installed in the SPS (R. Assmann)

R. AssmannR. Assmann

2626

F. Zimmermann et al

Collimator-Induced Tune Change Collimator-Induced Tune Change (Changing Collimator Gap)(Changing Collimator Gap)

Gap: 2.1 51 mm

SPS tune depends on SPS tune depends on collimator gap!collimator gap!

Expected tune change observed within factor 2!

Impedance estimates are strongly confirmed by experiment!

M. Gasior, R. Jones et al

Frank Zimmermann, GSI Meeting 31.03.2006

27

new generalized formula

measurement

b

yb yx

XY

y

co

y

co

cflatyBL

pbgeneraly

co

x

y

z

dYdXYXGyb

erf

yberf

bdeZ

LrNQ

2

42

4

4

2

2/1

2

2/

,

4,

2

3

2

3

24

2

222

222

generalized formula: combine correct frequency dependence of Burov-Lebedev with nonlinear dependence on transverse coordinates from Piwinski, assuming that the two dependencies remain factorized

Frank Zimmermann:nearly perfect agreement!



LHC graphite collimatorsLHC graphite collimators

d

a

beam

current

Induced

Usual regime: New regime:

d

a

beam

ad ,

current

Induced

,ad d

aaeff

• One may think that the classical “thick-wall” formula applies also for 2 cm thick graphite collimators about 2 mm away from the beam

• In fact it is not The resistive impedance is ~ 2 orders of magnitude lower at ~ 8 kHz!

dδaaeff when



LHC stability diagram (maximum octupole LHC stability diagram (maximum octupole strength) and collective tune shift for the most strength) and collective tune shift for the most

unstable coupled-bunch mode at 7 TeV (E. Metral, unstable coupled-bunch mode at 7 TeV (E. Metral, 2004) 2004)

0.0008 0.0006 0.0004 0.0002

0.000025

0.00005

0.000075

0.0001

0.000125

0.00015

All the machine

All the machinewith Cu coated (5 μm) collimators

Without collimators

(TCDQ+RW+BB) QRe

Q Im

Elias Métral, RLC meeting, 03/02/06

30/20

Mode 0

Mode 1

Mode 2

Mode 3

10 8 6 4 2ReQ10 4

0.250.5

0.751

1.25

1.5

ImQ10 4

10 8 6 4 2ReQ10 4

0.25

0.50.75

1

1.251.5

ImQ10 4

10 8 6 4 2ReQ10 4

0.25

0.50.75

11.25

1.5

ImQ10 4

10 8 6 4 2ReQ10 4

0.25

0.50.75

1

1.251.5

ImQ10 4

10 8 6 4 2ReQ10 4

0.25

0.50.75

1

1.251.5

ImQ10 4

10 8 6 4 2ReQ10 4

0.250.5

0.751

1.25

1.5

ImQ10 4

0yQ 1yQ

2yQ 3yQ

4yQ 5yQ

Stability diagrams (vertical plane)

LHC at 7 TeV

Phase 2 collimators:

~70% of the nominal LHC

intensity can be stabilized using

Landau octupoles at zero

chromaticity



Machine Protection and Machine Protection and Collimation challengesCollimation challenges

• Magnet quench limits need to be experimentally validated

Fresca test facility and LHC sector test• Beam Loss Monitors need proper calibration for efficient

machine protection LHC sector test• Learn how to set-up routinely a complicated three-stage

collimation system control beta-beating at ~10% level• Phase-2 collimation system is not compatible with

nominal LHC intensity at 7 TeV, if we want to stabilize the beams using Landau octupoles at zero chromaticity: • use low-noise transverse feedback and chromaticity to stabilize

the beams?• octupoles are “passive” and more reliable ideal to push

machine performance and reduce experimental background levels

• active feedback may increase emittance and reduce luminosity• investigate crystal assisted collimation and/or develop new low-

impedance collimators (e.g., longitudinally segmented or incorporating Cu stripes to carry low-frequency image currents?)

Frank Zimmermann, LHC Electron Cloud, GSI Meeting 30.03.2006

32

LHC strategy against Electron Cloud

1) warm sections (20% of circumference) coated by TiZrVgetter developed at CERN; low secondary emission; if cloud occurs, ionization by electrons (high cross section ~400 Mbarn) aids in pumping & pressure will even improve

2) outer wall of beam screen (at 4-20 K, inside 1.9-K cold bore) will have a sawtooth surface (30 m over 500 m) to reduce photon reflectivity to ~2% so that photoelectrons are only emitted from outer wall & confined by dipole field

3) pumping slots in beam screen are shielded to preventelectron impact on cold magnet bore

4) rely on surface conditioning (‘scrubbing’); commissioning strategy; as a last resort doubling or triplingbunch spacing suppresses e-cloud heat load

uniquevacuum system!


33

arc heat load vs. intensity, 25 ns spacing, ‘best’ model

calculation for 1 train

R=0.5

computational challenge! higher heat load for quadrupolesin 2nd train under study

max=1.7

max=1.5

max=1.3max=1.1

max=1.3-1.4 suffices

BS cooling capacity

injectionlow luminosity

high luminosity

ECLOUDsimulation


34

is “scrubbing” needed in LHC? still lacking experimental data, e.g., on max

uncertainty in heat load prediction of factor ~2 also incomplete understanding of scrubbing

(COLDEX data vs. prediction, RHIC, DAFNE) if max~1.3 reached in commissioning, no scrubbing

is needed for heat load and fast instabilities pressure should be ok too according to N. Hilleret one concern: long-term emittance growth and poor

lifetime (observed in SPS after scrubbing) we still believe we need to prepare a scrubbing

strategy in case it turns out to be necessaryto go to max~1.3 (e.g., tailor train spacings & train lengths at nominal bunch intensity)



Instabilities & emittance Instabilities & emittance growthgrowth

caused by the electron cloud caused by the electron cloud 1) Multi-bunch instability – not expected to be a problem

can be cured by the feedback system2) single-bunch instability – threshold electron cloud

density 0~4x1011 m-3 at injection in the LHC3) incoherent emittance growth new understanding! (CERN-GSI collaboration)

2 mechanisms: periodic crossing of resonance due to e- tune shift

and synchrotron motion (similar to halo generationfrom space charge)

periodic crossing of linearly unstable regiondue to synchrotron motion and strong focusingfrom electron cloud in certain regions, e.g., in

dipoles



Resonance Trapping Resonance Trapping (G. Franchetti, GSI)(G. Franchetti, GSI)

The same resonance trapping mechanism can explain slow emittance growth and beam losses observed with space charge in the PS (left) and with electron cloud in the SPS (below)

• Particles with large synchrotron amplitudes reach larger and larger betatron amplitudes and are lost bunch shortening

• Particle losses are enhanced by chromaticity


37

single-bunch “TMC” instabilityfast growth above e- densitythreshold; slower growth below

= 1 x 1011 m-3= 2 x 1011 m-3

= 3 x 1011 m-3

“Transverse Mode Coupling Instability (TMCI)” for e- cloud ( > thresh)

Long term emittance growth ( < thresh)

E. Benedetto LHC, Q’=0,at injection



Electron density vs LHC beam intensity

Challenge: how to go from max~1.7 to 1.3?Scrubbing should be done at nominal Nb

(stripes)

typical“TMCI”instabilitythreshold

calculation for 1 bunch train

max=1.7

max=1.5

max=1.3 max=1.1

R=0.5

ECLOUDsimulation



LHC bunch train at injection in the LHC bunch train at injection in the SPSSPS

Evolution of bunch length and bunch population for the first and the last bunch in an LHC bunch train of 72 bunches. SPS measurements with electron cloud in Aug 2004. Courtesy G. Rumolo, G. Arduini, and F. Roncarolo.

rms

bunch

leng

th (

ns)

bunch

inte

nsi

ty (

au)

time (min) time (min)

Qx=26.135

Qy=26.185

ξx=0.15

ξy=0.1

VRF ~ 3 MV

dampers on coupling:

0.008



Tentative ConclusionsTentative Conclusions• Some “safety nets” in the original LHC conceptual

design (low-impedance, stabilization by octupoles, full triplet aperture without beam screens) have been sacrificed to guarantee a more robust collimation system and a safer IR vacuum behaviour

• Machine downtime caused by magnet quenches may be initially frequent, until collimation and machine protection are fully mastered

• A shorter machine turnaround time implies reliable tables of quench levels, BLM calibrations, and a dynamic optics control (reference magnets)

• Emittance control will be challenging and may require crystal assisted collimation and/or new low-noise feedback systems.

• A longitudinal feedback may enable shorter bunches and reduce geometric luminosity loss for lower *.



Tentative Conclusions (continued)Tentative Conclusions (continued)• Reaching nominal LHC performance is challenging• Some uncertainties remain in connection with

electron cloud effects and vacuum behaviour of the cold arcs: exceeding nominal beam current may be impossible or take several years operation with 75 ns bunch spacing would reduce e-cloud & long range beam-beam effects and maximize luminosity

• Operation with flat beams can help relaxing IR aperture constraints and/or increasing luminosity

• A re-alignment of the IR quads would further relax aperture constraints, increase luminosity, and minimize energy deposition in the magnet coils. This option should be considered also for the IR upgrade.



Tentative Conclusions for Tentative Conclusions for the LHC IR Upgradethe LHC IR Upgrade

• We do need triplet spares and thus a back-up or intermediate IR upgrade option based on NbTi magnet technology. What is its luminosity reach?

• A vigorous R&D programme on Nb3Sn magnets should start at CERN asap, complementary to the US-LARP programme, to reach an LHC luminosity of ~1035 after 2015

• Alternative IR layouts (quadrupole-first, dipole-first, D0, flat beams, Crab cavities) will be rated in terms of technological and operational risks/advantages by the end of 2006



Additional SlidesAdditional Slides



HERA operational experienceHERA operational experience

From K. Wittenburg

HERA:Ring of 6.3 km- 422 sc main dipoles- 224 sc main quads- 400 sc correction quads- 200 sc correction dipoles



Heat load in the Low-Heat load in the Low- TripletTriplet

N. Mokhov et al, LHC Project Report 633Peak power density:

0.45 mW/g



Trapped modes for tertiary LHC Trapped modes for tertiary LHC collimator chambers collimator chambers (A. Grudiev, (A. Grudiev,

2006)2006)


8th ICFA Seminar, Daegu, Korea 29/09/2005CERNF. Ruggiero

Vertical growth rate of head-tail modes in the LHC as a function of chromaticity at injection energy, for

~3000 bunches of nominal intensity

10 5 5 10Qy'

5

10

15

20Instability growth ratessec1

Mode 3

Mode 2

Mode 1

Mode 0

At injection head-tail modes with growth rates up to about 4 sec-1 are stabilized by lattice nonlinearities (assuming an amplitude detuning of 0.002 at 6 sigma).

The rigid mode m=0 has to be stabilized by the transverse feedback.



Stability diagrams (vertical plane)

LHC at 7 TeV

Phase 1 collimators:

~50% of the nominal LHC

intensity can be stabilized using

Landau octupoles at zero

chromaticity

10 8 6 4 2ReQ10 4

0.250.5

0.751

1.251.5

ImQ10 4

10 8 6 4 2ReQ10 4

0.250.5

0.751

1.251.5

ImQ10 4

10 8 6 4 2ReQ10 4

0.250.5

0.751

1.251.5

ImQ10 4

10 8 6 4 2ReQ10 4

0.250.5

0.751

1.251.5

ImQ10 4

10 8 6 4 2ReQ10 4

0.250.5

0.751

1.251.5

ImQ10 4

10 8 6 4 2ReQ10 4

0.25

0.50.75

1

1.251.5

ImQ10 4

Mode 0

Mode 1

Mode 2

Mode 3

0yQ 1yQ

2yQ 3yQ

4yQ 5yQ

Performance limitations of the present LHC

Documents

Transcript of Performance limitations of the present LHC