lhc beam performance and luminosity upgrade scenarios

Name Event DateName Event Date11

F. Ruggiero Univ. “La Sapienza”, Rome, 20–24 March 2006CERN

LHC beam performance and LHC beam performance and luminosity upgrade luminosity upgrade

scenariosscenariosperformance limitations, possible performance limitations, possible scenarios and milestones for the LHC scenarios and milestones for the LHC

upgradeupgrade

http://care-hhh.web.cern.ch/care-hhh/

See also slides on Measurements, ideas, curiosities

triplet magnets

BBLR



F. Ruggiero LHC beam performance and luminosity upgrade scenarios

CERN

OutlineOutline• Time scale and potential of an LHC upgradeTime scale and potential of an LHC upgrade• LHC commissioning strategy and beam LHC commissioning strategy and beam

parametersparameters• Machine performance limitationsMachine performance limitations• Luminosity optimization and upgrade pathsLuminosity optimization and upgrade paths• Luminosity lifetime: peak vs integrated Luminosity lifetime: peak vs integrated

luminosityluminosity• LHC luminosity upgrade scenarios: LHC luminosity upgrade scenarios:

• ultimate performance without hardware changesultimate performance without hardware changes• upgrade of the Interaction Regionsupgrade of the Interaction Regions• new bunch-shortening RF system and cryogenic loadsnew bunch-shortening RF system and cryogenic loads• collimation, beam-beam compensation and crab collimation, beam-beam compensation and crab

cavitiescavities• milestones and baseline designmilestones and baseline design



CERN

Time scale of an LHC upgradeTime scale of an LHC upgrade

L at end of year

time to halve error

integrated L

radiationdamage limit~700 fb-1

• the life expectancy of LHC IR quadrupole magnets is estimated to be <10 years owing to high radiation doses

• the statistical error halving time will exceed 5 years by 2011-2012

• therefore, it is reasonable to plan a machine luminosity upgrade based on new low-ß IR magnets before ~2015

design luminosity

ultimate luminosity

courtesy J. Strait

03 November 2005 - LHC seminar F. Ruggiero & W.Scandale, LHC luminosity upgrade - report from LHC-LUMI-05 4

CARE-HHH

luminosity versus energy upgradeCourtesy of Michelangelo Mangano



CERN

Chronology of LHC Upgrade Chronology of LHC Upgrade studiesstudies• Summer 2001: two CERN task forces investigate physics potential

(CERN-TH-2002-078) and accelerator aspects (LHC Project Report 626) of an LHC upgrade by a factor 10 in luminosity and 2-3 in energy

• March 2002: LHC IR Upgrade collaboration meeting http://cern.ch/lhc-proj-IR-upgrade• October 2002: ICFA Seminar at CERN on “Future Perspectives in High Energy Physics”• 2003: US LHC Accelerator Research Program (LARP) • 2004: CARE-HHH European Network on High Energy

High Intensity Hadron Beams

• November 2004: first CARE-HHH-APD Workshop (HHH-04) on “Beam Dynamics in Future Hadron Colliders and Rapidly Cycling High-Intensity Synchrotrons”, CERN-2005-006 • September 2005: CARE-HHH Workshop (LHC-LUMI-05) on “Scenarios for the LHC Luminosity Upgrade” http://care-hhh.web.cern.ch/CARE-HHH/LUMI-05/



3 November 2005 - LHC seminar F. Ruggiero & W.Scandale, LHC luminosity upgrade - report from LHC-LUMI-05 6

CARE-HHH

The CARE-HHH Network

• Roadmap for the upgrade of the European accelerator infrastructure (LHC and GSI accelerator complex)

o luminosity and energy upgrade for the LHCo pulsed SC high intensity synchrotrons for the GSI and LHC complexo R&D and experimental studies at existing hadron acceleratorso select and develop technologies providing viable design options

• Coordinate activities and foster future collaborations• Disseminate information

Coordinate and integrate the activities of the accelerator and particle physics communities, in a worldwide context, towards achieving superior High-Energy High-Intensity Hadron-Beam facilities for Europe

Mandate

• HHH coordination: F. Ruggiero (CERN) & W. Scandale (CERN)1. Advancement in Acc. Magnet Technology (AMT): L. Rossi (CERN) & L. Bottura (CERN)

2. Novel Meth. for Acc. Beam Instrumentation (ABI): H. Schmickler (CERN) & K. Wittenburg (DESY)

3. Accelerator Physics and Synchrotron Design (APD): F. Ruggiero (CERN) & F. Zimmermann (CERN)



CERN

Nominal LHC parametersNominal LHC parameterscollision energydipole peak fieldinjection energy

Ecm

BEinj

2x7 8.3 450

TeVT

GeV

protons per bunchbunch spacingaverage beam current

Nb

∆tI

1.15 25 0.58

1011

nsA

stored energy per beamradiated power per beam

362 3.7

MJkW

normalized emittancerms bunch length

nz

3.75 7.55

mcm

beam size at IP1&IP5beta function at IP1&IP5full crossing angle

**c

16.6 0.55 285

mmrad

luminosity lifetimepeak luminosityevents per bunch crossing

LL

15.5 1034

19.2

hcm-2s-1

integrated luminosity ∫ L dt 66.2 fb-1/year



CERN

• Peak luminosity at the beam-beam limit L~ I/*

• Total beam intensity I limited by electron cloud, collimation, injectors

• Minimum crossing angle depends on beam intensity: limited by triplet aperture

• Longer bunches allow higher bb-limit for Nb/n: limited by the injectors

• Less ecloud and RF heating for longer bunches: ~50% luminosity gain for flat bunches longer than *

• Event pile-up in the physics detectors increases with Nb

• Luminosity lifetime at the bb limit depends only on *

⇒ reduce Tturnaround to increase integrated lumi

0 1 2 3 4 5 6number of bunches nb1000

1

2

3

4

5

6

hcnubnoitalupop

Nb0111

LHC upgrade LHC upgrade paths/limitationspaths/limitations

I=0.86 A

I=1.72 A

nom

in al

ultimate

I=0.58 A

large

r cro

ssing

angle

long

er

bunc

hes

bb limit more bunches



CERN

The peak LHC luminosity can be multiplied by: factor 2.3 from nominal to ultimate beam intensity (0.58 0.86 A) factor 2 (or more?) from new low-beta insertions with ß* = 0.25 m

Tturnaround~10 h ∫Ldt ~ 3 x nominal ~ 200/(fb*year)

Expected factors for the LHC Expected factors for the LHC luminosity upgrade luminosity upgrade

Major hardware upgrades (LHC main ring and injectors) are needed to exceed ultimate beam intensity. The peak luminosity can be increased by:

factor 2 if we can double the number of bunches (maybe impossible due to electron cloud effects) or increase bunch intensity and bunch lengthTturnaround~10 h ∫Ldt ~ 6 x nominal ~ 400/(fb*year)

Increasing the LHC injection energy to 1 TeV would potentially yield: factor ~2 in peak luminosity (2 x bunch intensity and 2 x emittance) factor 1.4 in integrated luminosity from shorter Tturnaround~5 h thus ensuring L~1035 cm-2 s-1 and ∫Ldt ~ 9 x nominal ~ 600/(fb*year)



CERN

Challenge of a Cold MachineChallenge of a Cold Machine



CERN

LHC Cleaning System LHC Cleaning System (R. Assmann)(R. Assmann)

43

Pilot

No collimationNo collimation

Single-stage cleaningSingle-stage cleaning

Two-stage cleaning (phase 1)Two-stage cleaning (phase 1)

Two-stage cleaning (phase 2)Two-stage cleaning (phase 2)



CERN

Collimation & Machine Collimation & Machine ProtectionProtection



CERN

Constraints for LHC Constraints for LHC commissioningcommissioning

Only 8/20 LHC dump dilution kickers available during the first two years of operation total beam intensity in each LHC ring limited to 1/2 of its nominal value

According to SPS experience and to electron cloud simulations, the initial LHC bunch intensity Nb can reach and possibly exceed its nominal value for 75 ns bunch spacing, while it may be limited to about 1/3 of its nominal value for 25 ns spacing

Machine protection and collimation favours initial operation with lower beam power and lower transverse beam density. Simple graphite collimators may limit maximum transverse energy density to about 1/2 of its nominal value

Emittance preservation from injection to physics conditions will require a learning curve do not assume transverse emittance smaller than nominal, even for reduced bunch intensity

Initial operation with relaxed parameters is strongly favoured higher ß*, reduced crossing angle, and fewer parasitic collisions



CERN

LHC beam commissioningLHC beam commissioningMike Lamont, Chamonix 2005 workshopMike Lamont, Chamonix 2005 workshop

• No parasitic encounters• No crossing angle• No long range beam-beam• Larger aperture

• Instrumentation• Good beam for RF, Vacuum…• Lower energy densities

• Reduced demands on beam dump system• Collimation• Machine protection

• Luminosity• 1030 cm-2s-1 at 18 m• 2 x 1031 cm-2s-1 at 1 m

43 on 43 with 3 to 4 x 1010 ppb to 7 TeV



CERN

Phase R1/2 Time [days] Total1 Injection 2 1 22 First turn 2 3 63 Circulating beam 2 3 64 450 GeV: initial commissioning 2 4 85 450 GeV: detailed measurements 2 4 86 450 GeV: 2 beams 1 2 27 Nominal cycle 1 5 58 Snapback – single beam 2 3 69 Ramp – single beam 2 4 8

10 Single beam to physics energy 2 2 411 Two beams to physics energy 1 3 312 Establish Physics 1 2 213 Commission squeeze 2 4 414 Physics partially squeezed

TOTAL 60

LHC beam commissioningLHC beam commissioningMike Lamont, Chamonix 2005 workshopMike Lamont, Chamonix 2005 workshop



CERN

parameter units 75 ns spacing

25 ns spacing

nominal

number of bunchesprotons per bunch

nb

Nb [1011]9360.9

28080.4

28081.15

normalized emittance

rms bunch lengthrms energy spread

n [µm]z [cm]E [10-4]

3.75 7.55 1.13

3.75 7.55 1.13

3.757.551.13

IBS growth timebeta at IP

full crossing angle

xIBS [h]ß* [m]c [rad]

1351.0250

304 0.55285

106 0.55285

luminosity lifetimepeak luminosity

events per crossing

L [h]L [1034cm-2s-

1]

22 0.127.1

26 0.122.3

151.0

19.2

∫ over 200 runs L dt Lint [fb-1] 9.3 9.5 66.2

Steps to reach nominal LHC luminositySteps to reach nominal LHC luminosity

Possible scenarios with 75 ns and 25 ns bunch spacing for early LHC runs with integrated luminosity of about 10 fb-1 in 200 fills, assuming an average physics run time Trun = 14 h and Tturnaround=10 h.



CERN

Luminosity optimizationLuminosity optimization

Collisions with full crossing angle c

reduce luminosity by a geometric factor Fmaximum luminosity below beam-beam limit ⇒ short bunches and minimum crossing angle (baseline scheme)H-V crossings in two IP’s ⇒ no linear tune shift due to long rangetotal linear bb tune shift also reduced by F

INNfnLn

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



CERN

If bunch intensity and brightness are not limited by the injectors or by other effects in the LHC (e.g. electron cloud) ⇒ luminosity can be increased without exceeding beam-beam limit Qbb~0.01by increasing the crossing angle and/or the bunch lengthExpress beam-beam limited brilliance Nb/εn in terms of maximumtotal beam-beam tune shift Qbb, then

2

*zc

*nb

2bb

2p

rev*bb

p 21

2

nQ

rfIQ

rL

At high beam intensities or for large emittances, the performancewill be limited by the angular triplet aperture

2

θc

*

*bbp /20

/1,1min2 triplA

IQr

L



CERN

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

amplitudes, where particles come close to opposing beam

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

• increasing problem for SPS, Tevatron, LHC, i.e., for operation with 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 m75.3A5.0

36m75.31032

3

INnd

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

*θ angular beam

divergence at IP



CERN

Schematic of a super-bunch collision, consisting of ‘head-on’ and ‘long-range’ components. The luminosity for long bunches having flat longitudinal distribution is ~1.4 times higher than for conventional Gaussian bunches with the same beam-beam tune shift and identical bunch population (see LHC Project Report 627)



CERN

Schematic of reduced electron cloud build up for a longbunch. Most electrons do not gain any energy when traversing the chamber in the quasi-static beam potential [after V. Danilov]negligible heat load



CERN

Scenarios for the luminosity Scenarios for the luminosity upgradeupgrade ultimate performance without hardware changes (phase 0)

maximum performance with only IR changes (phase 1) maximum performance with “major” hardware changes

(phase 2)

Phase 0: steps to reach ultimate performance without hardware changes:1) collide beams only in IP1 and IP5 with alternating H-V

crossing2) increase Nb up to the beam-beam limit L = 2.3 x 1034 cm-2

s-1

3) increase the dipole field to 9T (ultimate field) Emax = 7.54 TeV

The ultimate dipole field of 9 T corresponds to a beam current limited by

cryogenics and/or by beam dump/machine protection considerations.

• beam-beam tune spread of

0.01• L = 1034 cm-2s-1 in ATLAS and

CMS• Halo collisions in ALICE• Low-luminosity in LHCb

Nominal LHC performance



CERN

Phase 1: steps to reach maximum performance with only IR changes1) Modify the insertion quadrupoles and/or layout ß* = 0.25 m2) Increase crossing angle c by √2 c = 445 µrad3) Increase Nb up to ultimate intensity L = 3.3 x 1034 cm-2s-1

4) Halve z with high harmonic RF system L = 4.6 x 1034 cm-2s-1

5) Double the no. of bunches nb (and increase c ) L = 9.2 x 1034 cm-2s-1

excluded by electron cloud? Step 5 belongs to Phase 2 Step 4) requires a new RF system providing an accelerating voltage of 43 MV at 1.2 GHz a power of about 11 MW/beam longitudinal beam emittance reduced to 1.8 eVs horizontal Intra-Beam Scattering (IBS) growth time decreases

by ~ √2 Operational consequences of step 5) exceeding ultimate beam

intensity upgrade LHC cryogenics, collimation, RF and beam dump

systems the electronics of all LHC beam position monitors should be

upgraded possibly upgrade SPS RF system and other equipment in the

injectors

Scenarios for the luminosity Scenarios for the luminosity upgradeupgrade

24

luminosity upgrade: baseline scheme

increase Nb

bblimit?

restore F2/12

*zc

21

F

no

yes

c>mindue to LR-bb

crab cavities

BBLRcompen-sation

reduce z

by factor ~2using higherfrf & lower ||

(largerc ?)

2.3

reduce c

(squeeze *)

use large c

& pass each beamthrough separatemagnetic channel

reduce * byfactor ~2

new IRmagnets

increase nb byfactor ~2

if e-cloud, dump &impedance ok

9.2

1.0

4.6

simplified IR design with large c

peak luminosity gain1.72 A

0.86 A

0.58 A

0.86 A

beam current

or decouple L and F

25

luminosity upgrade: Piwinski scheme

reduce * byfactor ~2

new IRmagnets

decrease F2/12

*zc

21

F

increase zc

increase Nb

nobb

p

nb

2 QFr

N

?

yes

reduce #bunchesto limit total current?

flatten profile?

7.7 15.5

1.0

0.86 A 1.72 A

luminosity gain

beam current

superbunches?0.58 A



CERN

Various LHC upgrade options Various LHC upgrade options parameter symbol nominal ultimate shorter

bunchlonger bunch

no of bunches nb 2808 2808 5616 936proton per bunch Nb [1011] 1.15 1.7 1.7 6.0bunch spacing ∆tsep [ns] 25 25 12.5 75average current I [A] 0.58 0.86 1.72 1.0normalized emittance

n [µm] 3.75 3.75 3.75 3.75

longit. profile Gaussian Gaussian

Gaussian

flat

rms bunch length z [cm] 7.55 7.55 3.78 14.4ß* at IP1&IP5 ß* [m] 0.55 0.50 0.25 0.25full crossing angle c [µrad] 285 315 445 430Piwinski parameter c z/(2*) 0.64 0.75 0.75 2.8peak luminosity L [1034 cm-2 s-

1]1.0 2.3 9.2 8.9

events per crossing 19 44 88 510luminous region length

lum [mm] 44.9 42.8 21.8 36.2



CERN

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

parameter symbol nominal ultimate shorter bunch

longer bunch

protons per bunch Nb [1011] 1.15 1.7 1.7 6.0bunch spacing ∆tsep[ns] 25 25 12.5 75average current I [A] 0.58 0.86 1.72 1.0

longitudinal profile Gaussian Gaussian Gaussian flat

rms bunch length z [cm] 7.55 7.55 3.78 14.4Average electron-cloud heat load at 4.6–20 K in the arc for R=50% and δmax=1.4 (in parentheses for δmax=1.3)

Pecloud [W /m] 1.07(0.44)

1.04(0.59)

13.34(7.85)

0.26(0.26)

Synchrotron radiation heat load at 4.6–20 K P [W /m] 0.17 0.25 0.50 0.29

Image currents power at 4.6–20 K P [W /m] 0.15 0.33 1.87 0.96

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

Pgas [W /m] 0.038(0.38)

0.056(0.56)

0.113(1.13)

0.066(0.66)



CERN

Events per bunch crossing and Events per bunch crossing and beam lifetime due to nuclear p-p beam lifetime due to nuclear p-p

collisionscollisionsrev

bb

bing-Xevents

fnL

bb=60 mb total inelastic cross section

TOT

bbN 2

/

LNn

beam intensity halving time due to nuclear p-p collisions at two IP’s with total cross section TOT=110 mb

NgasxIBS

L 54.122

11

luminosity lifetime: assumes radiation damping compensates diffusion

54.1)1( N

N e exponential luminosity

lifetime due to nuclear p-p interactions

*bb

p

rev

bb 2 Q

rf

NnL nuclear scattering

lifetime at the beam-beam limit depends only on * !



CERN

Optimum run time and effective Optimum run time and effective luminosityluminosityL

run

L

turnaroundrunL

T

eTT

][-1,-ProductLog1 L

turnaround-1-

L

turnaround

L

run

T

eTT

The optimum run time and the effective luminosity are universal functions of Tturnaround/L

wwezzw ][ProductLog where

][-1,-ProductLog

1

L

turnaround-1-turnaroundrunL

L

Teff

eTTL

L

When the beam lifetime is dominated by nuclear proton-proton collisions, then L~N/1.54 and the effective luminosity is a universal functions of Tturnaround/

0.5 1 1.5 2TturnaroundL

0.20.40.60.8

11.21.4TrunL

0.5 1 1.5 2TturnaroundL

0.20.40.60.8

1LeffL



CERN

Effective luminosity for various upgrade Effective luminosity for various upgrade options options

parameter symbol nominal ultimate shorter bunch

longer bunch

protons per bunch Nb [1011] 1.15 1.7 1.7 6.0bunch spacing ∆tsep [ns] 25 25 12.5 75average current I [A] 0.58 0.86 1.72 1.0longitudinal profile Gaussian Gaussian Gaussia

nflat

rms bunch length z [cm] 7.55 7.55 3.78 14.4ß* at IP1&IP5 ß* [m] 0.55 0.50 0.25 0.25full crossing angle c [µrad] 285 315 445 430Piwinski parameter c z/(2*) 0.64 0.75 0.75 2.8peak luminosity L [1034 cm-2 s-1] 1.0 2.3 9.2 8.9events per crossing 19 44 88 510IBS growth time xIBS [h] 106 72 42 75nuclear scatt. lumi lifetime N/1.54 [h] 26.5 17 8.5 5.2luminosity lifetime (gas =85 h)

L [h] 15.5 11.2 6.5 4.5

effective luminosity Leff [1034 cm-2 s-1] 0.4 0.8 2.4 1.9(Tturnaround=10 h) Trun [h] optimum 14.6 12.3 8.9 7.0

effective luminosity Leff [1034 cm-2 s-1] 0.5 1.0 3.3 2.7 (Tturnaround= 5 h) Trun [h] optimum 10.8 9.1 6.7 5.4



CERN

Interaction Region upgradeInteraction Region upgrade

factors driving IR design: • minimize * • minimize effect of LR collisions• large radiation power directed towards the IRs• accommodate crab cavities and/or beam-beam compensators. Local Q’ compensation scheme?• compatibility with upgrade path

goal: reduce * by at least a factor 2

maximize magnet aperture,minimize distance to IR

options: NbTi ‘cheap’ upgrade, NbTi(Ta), Nb3Sn new quadrupoles new separation dipoles



CERN

IR ‘baseline’ schemesIR ‘baseline’ schemes

short bunches & minimum crossing angle &BBLR

crab cavities & large crossing angle

triplet magnets triplet magnets crab cavity

BBLR



CERN

alternative IR schemesalternative IR schemes

dipole first & small crossing angle

triplet magnets

dipole magnets

dipole first & large crossing angle &long bunches or crab cavities

triplet magnets

dipole

reduced # LR collisionscollision debris hit D1



CERN

3535

Dipoles first and doublet focusingDipoles first and doublet focusing

IP D1

D2

D2

Q1

Q2

Features

• Requires beams to be in separate focusing channels

• Fewer magnets

• Beams are not round at the IP

• Polarity of Q1 determined by crossing plane – larger beam size in the crossing plane to increase overlap

• Opposite polarity focusing at other IR to equalize beam-beam tune shifts

• Significant changes to outer triplet magnets in matching section.

Focusing symmetric about IP

Tanaji Sen, Doublet optics


F. Ruggiero LHC upgrade scenariosCERN

Flat beamsFlat beams• Interesting approach, flat beams could increase

luminosity by ~20-30% with reduced crossing angle• Symmetric doublets studied by J. Johnstone (FNAL)

require separate magnetic channels, i.e. dipole-first, Crab cavities or special quads

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

• Probably requires BBLR compensation• Recently S. Fartoukh has found a more interesting flat

beam solution with anti-symmetric LHC baseline triplets

37

Beam aspect ratio vs triplet aperture (1/5)• Beam screen orientation for H/V scheme

S. Fartoukh, ABP-RLC meeting, 28-10-2005

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)

In both cases, H-separation of about 9.5*max(x,b1 ,x,b2)

In both cases, V-separation of about 9.5*max(y,b1 ,y,b2)

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

38

Pushing the luminosity by 10-20%

Case x* [cm] y

* [cm] *[rad]

n1 in the triplet

Geometric loss factor [%]

L/Lnom

Nominalr=1.055c

m

55.00 55.00 285 ~7 83.9 1.00

Flat r=2.055c

m

110.00 27.50 201 ~7 95.1 1.13

Flat r=1.655c

m

88.00 34.37 225 ~7.5 92.7 1.10

Flat r~1.751c

m

88.00 30.00 225 ~7 92.7 1.18All these cases being allowed by the nominal LHC hardware:layout, power supply, optics antisymmetry, b.s. orientation in the triplets (only changing the present H/V scheme into V/H scheme)!

S. Fartoukh, ABP-RLC meeting, 28-10-2005



CERN

‘‘cheap’ IR upgradecheap’ IR upgrade

short bunches & minimum crossing angle &BBLR

triplet magnets

each quadrupole individually optimized (length & aperture) reduced IP-quad distance from 23 to 22 mconventional NbTi technology: *=0.25 m is possible

BBLR

in case we need to double LHC luminosity earlier than foreseen



CERN

• quadrupole-first and dipole-first solutions based on conventional NbTi technology and on high-field Ni3Sn magnets, possibly with structured SC cable

• energy deposition, absorbers, and quench limits • schemes with Crab cavities as an alternative to the

baseline bunch shortening RF system at 1.2 GHz to avoid luminosity loss with large crossing angles

• early beam separation by a “D0” dipole located a few metres away from the IP (or by tilted experimental solenoids?) may allow operation with a reduced crossing angle. Open issues: compatibility with detector layout, reduced separation at first parasitic encounters, energy deposition by the collision debris

• local chromaticity correction schemes• flat beams, i.e. a final doublet instead of a triplet. Open

issues: compensation of long range beam-beam effects with alternating crossing planes

Several LHC IR upgrade options are being explored :



CERN

Crab cavities vs bunch Crab cavities vs bunch shorteningshortening

Crab cavities combine advantages of head-on collisions and large crossing anglesrequire lower voltages compared to bunch shortening RF systemsbut tight tolerance on phase jitter to avoid emittance growth

KEKB Super-KEKB

ILC Super-LHC

x* 100 m

70 m 0.24 m

11 m

c +/- 11 mrad

+/-15 mrad

+/-5 mrad

+/- 0.5 mrad

t 6 ps 3 ps 0.03 ps

0.08 ps

Comparison of timing tolerances

RF Deflector( Crab Cavity )

Head-onCollision

Crossing Angle (11 x 2 m rad.)

Electrons PositronsLERHER

1.41 MV

1.41 MV

1.44 MV

1.44 MV



CERN

Crab Cavities



CERN

Motivation of Beam-Beam Motivation of Beam-Beam compensation studiescompensation studies

• Provide more LHC luminosity earlier Space is reserved in the LHC for the wires. An early

test in RHIC will determine the effectiveness of the compensation and possibly address the challenges to the compensation. If effective, will allow a smaller crossing angle and more beam intensity.

• Provide direction to an IR upgrade path If compensation is proven to be effective, then the

quadrupole-first option, possibly with flat beams, seems to be a more natural path for the IR upgrade. Otherwise, the dipole-first option will be more attractive.



CERN

Lessons from the SPS Lessons from the SPS experimentsexperiments

• Compensating 1 wire with another wire at nearly the same phase “works”

• Compensation is tune dependent

• Current sensitivity • Alignment sensitivity• Equivalent crossings in the

same plane led to better lifetimes than alternating planes

• Beam lifetime ~ d3 d is the beam-wire distance Higher power law expected

given the proximity of high order resonances

Both wires on 1 wire on

Nearly perfect compensation

No wires activated



CERN

2nd prototype BBLR in the CERN SPS has demonstrated benefit of compensation

G. Burtin, J. Camas, J.-P. Koutchouk, F. Zimmermann et al.



CERN

Lessons from RHIC Lessons from RHIC experimentexperiment• Study at injection

energy with 1 bunch and 1 parasitic interaction per beam

• There is an effect to compensate, even with 1 parasitic

• Drop in lifetime seen for beam separations < 7 σ

• Effect is very tune dependent

• How important are machine nonlinearities and other time dependent effects?

• Did they change with the beam-beam separation?



CERN

Wire compensation at Wire compensation at RHICRHIC• Compensation of 1

wire by another wire worked well in the SPS under LHC conditions.

• Real test of the compensation principle requires 2 beams

• Beam studies in RHIC show that parasitic interactions have strong influence on beam loss

• Favorable location for wire has been found in IR6, phase advance to parasitic ~6 degrees at top energy

Proposed wire location

Location of parasitic



CERN

Milestones for future LHC Milestones for future LHC Upgrade machine studiesUpgrade machine studies

• 2006: installation and test of a beam-beam long range compensation system at RHIC to be validated with colliding beams

• 2006/2007: new SPS experiment for crystal collimation, complementary to Tevatron results

• 2006: installation and test of Crab cavities at KEKB to validate higher beam-beam limit and luminosity with large crossing angles

• 2007: if KEKB test successful, test of Crab cavities in a hadron machine (RHIC?) to validate low RF noise and emittance preservation

• 2007-2009: LHC running-in and first machine studies on collimation and beam-beam



CERN

Tentative conclusions Tentative conclusions for the LHC IR Upgradefor the LHC IR Upgrade

• We do need a back-up or intermediate IR upgrade option based on NbTi magnet technology. What is the maximum luminosity?

• A vigorous R&D programme on Nb3Sn magnets should start at CERN asap, in parallel to the US-LARP programme, to be ready for 1035 luminosity in ~2015

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



CERN

Towards a baseline designTowards a baseline design

• Define a Baseline, i.e. a forward looking configuration which we are reasonably confident can achieve the required LHC luminosity performance and can be used to give an accurate cost estimate by mid-end 2006 in a “Reference Design Report”

• Identify Alternative Configurations and rate them in terms of technological and operational risks/advantages

• Identify R&D (at CERN and elsewhere)• To support the baseline• To develop the alternatives

Following the approach proposed by Barry Barish for the ILC, we propose to:



CERN

Reference LHC Luminosity Reference LHC Luminosity Upgrade: Upgrade: workpackages and workpackages and

tentative milestonestentative milestonesaccelerator WorkPackage 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 after 2015LHC Main Ring Accelerator Physics

High Field SuperconductorsHigh Field MagnetsMagnetic MeasurementsCryostatsCryogenics: IR magnets & RFRF and feedbackCollimation&Machine ProtectionBeam InstrumentationPower converters

SPS SPS kickers

Tentative MilestonesBeam-beam

compensation test at RHIC

SPS crystal collimation test

LHC collimation tests

LHC collimation tests

Install phase 2 collimation

LHC tests: collimation & beam-beam

Install new SPS kickers

new IR magnets and RF system

Other Tentative Milestones Crab cavity test at KEKB

Low-noise crab cavity test at

RHIC

LHC Upgrade Conceptual

Design Report

LHC Upgrade Technical Design

Report

Nominal LHC luminosity

10^34

Ultimate LHC luminosity 2.3x10^34

beam-beam compensation

Double ultimate LHC luminosity

4.6x10^34

LHC Upgrade Reference

Design ReportReference LHC Upgrade scenario: peak luminosity 4.6x10^34/(cm^2 sec)

R&D - scenarios & models Integrated luminosity 3 x nominal ~ 200/(fb*year) assuming 10 h turnaround timespecifications & prototypes new superconducting IR magnets for beta*=0.25 mconstruction & testing phase 2 collimation and new SPS kickers needed to attain ultimate LHC beam intensity of 0.86 Ainstallation & commissioning beam-beam compensation may be necessary to attain or exceed ultimate performance

new superconducting RF system: for bunch shortening or Crab cavitieshardware for nominal LHC performance (cryogenics, dilution kickers, etc) not considered as LHC upgradeR&D for further luminosity upgrade (intensity beyond ultimate) is recommended: see Injectors Upgrade



CERN

LHC upgrade scenariosLHC upgrade scenariosSummarySummary

• The upgrade scenario currently assumed as The upgrade scenario currently assumed as baseline includes a reduction of baseline includes a reduction of * to 0.25 m, an * to 0.25 m, an increased crossing angle and a new bunch-increased crossing angle and a new bunch-shortening RF system.shortening RF system.

• The corresponding peak luminosity with ultimate The corresponding peak luminosity with ultimate beam intensity is 4.6x10beam intensity is 4.6x103434 cm cm-2-2 s s-1-1 at two IP’s. at two IP’s. Electron cloud effects and/or cryogenic heat Electron cloud effects and/or cryogenic heat loads may exclude the possibility to double the loads may exclude the possibility to double the number of bunches.number of bunches.

• Milestones for future LHC Upgrade machine Milestones for future LHC Upgrade machine studies include RHIC tests on Long Range Beam-studies include RHIC tests on Long Range Beam-Beam compensation and possibly on crab cavity Beam compensation and possibly on crab cavity operation, after KEKB tests, SPS and Tevatron operation, after KEKB tests, SPS and Tevatron studies on crystal assisted collimation, as well studies on crystal assisted collimation, as well as collimation, electron cloud, and beam-beam as collimation, electron cloud, and beam-beam studies at the LHC itself.studies at the LHC itself.

• Several LHC IR upgrade options are currently Several LHC IR upgrade options are currently being explored: we need to converge to a being explored: we need to converge to a baseline configuration and identify a few baseline configuration and identify a few alternative options.alternative options.



CERN

Additional SlidesAdditional Slides



Alternative ways to avoid Alternative ways to avoid luminosity lossluminosity loss

1) Reduce crossing angle and apply “wire” compensation of long range beam-beam effects

2) Crab cavities ⇒ large crossing angles to avoid long range bb effects w/o luminosity loss. Potential of boosting the beam-beam tune shift (factor 2-3 predicted for KEKB, what about LHC?)

3) Early beam separation by a “D0” dipole located a few metres away from the IP, as recently suggested by JPK at the LHC-LUMI-05 workshop. The same effect could be obtained by tilted experimental solenoids, but the experiments don’t seem to like the idea.

A potential drawback of 2) and 3) is that Qbb is no longerreduced by the geometric factor F ⇒ lower beam-beam limit?

55

Back to the Xing angle issue

Q1

Q2

Q3

Orbit corrector

An “easy” way to reduce or cancel the Xing angle at the IP and gain 20% to 50% in luminosity.

Is it possible for the detectors?

J.-P. Koutchouk, LHC-LUMI-05



New idea: D0 magnet a few New idea: D0 magnet a few meters away from the IPmeters away from the IP

Advantages• Cheap and elegant solution to increase luminosity• No need of a new bunch shortening RF system• Cleans collision debris from Q1?

Possible drawbacks• Reduced separation at first few parasitic encounters?• Collision debris and background in the experiments?• Compatibility with detector layout and integration

into the experiments



LR beam-beam LR beam-beam compensation: remarks and compensation: remarks and

open issuesopen issues• Simulations of LR compensation with 2 wires indicate that lifetime is recovered over a wide tune range but not for all tunes.

• The measured SPS lifetime is 5 ms x (d/)5. Extrapolation to LHC beam-beam distance (9.5 ) would predict 6 minutes beam lifetime! Tevatron observations with electron lens show cubic dependence. Further SPS tests at different energy are needed.

• Lifetimes predicted by simulation codes are much larger than those observed, even though sensitivity to parameters seems correct. Needs further understanding and beam tests, e.g. at RHIC.

• For extreme PACMAN bunches there is overcompensation which causes the footprint to flip over or to increase instead of shrinking. To avoid degraded lifetime for PACMAN bunches, the wire should be pulsed train by train. It is rather challenging to make a pulsed wire for BB compensation: the required average pulse rate is 439 kHz and the turn-by-turn amplitude stability 10-4.

• Experiments at RHIC (Fischer) with a single LR encounter show that the BB effect is visible starting from a 5 separation, consistent with Tevatron and Daphne observations, but contrary to LHC simulations and possibly earlier observations at the SPS collider.

58

RHIC experimentRHIC experiment

•Studied at injection energy with 1 bunch and 1 parasitic interaction per beam•There is an effect to compensate, even with 1 parasitic•Drop in lifetime seen for beam separations < 7 σ•Effect is very tune dependent

SPS : (d/)5

Tevatron: ~ d3

RHIC : ~ d4 or d2

[measured 04/28/05, scan 4]



LHC-LUMI-05 workshop: LHC-LUMI-05 workshop: some conclusions on the IR some conclusions on the IR

UpgradeUpgrade• Local correction à la Raimondi, via dispersion inside triplet magnets and two pairs of sextupoles, can correct chromaticity and geometric aberrations look for a solution that can be implemented and removed anytime by varying quads and sexupole strengths

• Three IR layout options were identified that should be studied in more detail:

1) dipole-first based on Nb3Sn technology with ℓ* = 19 m2) quad-first layout based on Nb3Sn technology ℓ * = 19 m3) low gradient quad-first layout based on NbTi technology• Still need to fix ℓ* and required length for TAS upgrade.

Agreement to assume ℓ* = 19 m as a reasonable estimate

• CARE-HHH web repository with optics solutions is very desirable we should all use the same input (MADX)

• Update the 3 proposals by the end of 2005


F. Ruggiero LHC IR Upgrades, LARP Workshop, WG1 summaryCERN

Energy Deposition Issues in Energy Deposition Issues in LHC IR Upgrades, LHC IR Upgrades, N. Mokhov (FNAL)N. Mokhov (FNAL)

• All three aspects, i.e. i) quench limit, ii) radiation damage (magnet lifetime), and iii) dynamic heat load on the cryo system should be simultaneously addressed in the IR magnet design. i) and ii) are linked

• Peak power deposition at non-IP end of IR magnets ~proportional to ∫Bdℓ FDFD “quadruplet” focusing?

• Estimated dipole field with TAS in quad-first option to reduce peak energy deposition “well below” quench limits

15-20 Tm for magnetic TAS• Estimated thickness of internal absorbers a 5 mm thick

SS absorber reduces peak power by a factor ~2• Impact of orbit corrector D0 inside the experiment on energy

deposition in downstream magnets, including detector solenoid field

more work needed, modest impact of solenoid field on energy deposition (more from fringe fields)



Action items/comments on energy Action items/comments on energy deposition, deposition, Nikolai MokhovNikolai Mokhov

• Refine and test scaling law for energy deposition in IR magnets with MARS simulations (including dependence on ℓ*)

• Introduce quench limits to JPK’s spreadsheet for NbTi and Nb3Sn

• Address radiation damage/lifetime issues in all IR magnet design analyses: 7 years at 1034 become 8 months at 1035 with currently used materials new (ceramic type) materials for 1035?

• Launch R&D program on beam tests for SC and insulating materials asap: BNL, FNAL, MSU

• Arrive at a clear picture on Dynamic Heat Load limits. How serious is the current 10 W/m limit or 120 W on each side of IR? This becomes 100 W/m and 1.2 kW for 1035. Cooling scheme? Cryoplant capability?



Potential impact of novel magnet Potential impact of novel magnet technology for IR elements, technology for IR elements, Peter Peter

McIntyreMcIntyre• Designs have been suggested for novel magnet technology to mitigate limitations from heat deposition and radiation damage from deposition of secondary particles in the quadrupole triplet and separation dipole. One example is an ironless quadrupole using structured-cable Nb3Sn conductor, which could provide 390 T/m gradient at a location as close as 12 m from the IP, and compatibility with supercritical helium flowing throughout the coils. A second example is a 9 T levitated-pole dipole for D1, which would open the transverse geometry so that secondaries are swept into a room-temperature flux return.

• In order to evaluate the potential benefit of these concepts it is necessary to model the heat deposition and radiation damage in the more compact geometries, and to examine potential interference with the performance of the detectors.

• Of particular importance is to undertake a consistent examination of the impact of reducing ℓ* on the ensemble of issues that impact achievable * the interface of the IR with the machine lattice (chromaticity and dispersion, multipole errors, orbit errors, etc.), and the strategy for accommodating long-range beam-beam effects.

• Also of interest is to evaluate the pros and cons of the alternatives for operating temperature (superfluid, two-phase, or supercritical cooling) for the IR elements that must operate with substantial heat loads.

63

Latest design: 9 Tesla @ 4.5 K

Support each pole piece using tension struts (low heat load). 56 mm clear aperture

All windings are racetracks.

Only pole tip winding is Nb3Sn.

All others are NbTi.



Dipole-First optics (R. De Maria)• matched optics solution for dipole-first layout for

Beam1 and Beam2 with squeeze and tunability study:• 18 km -max requires additional Q’ correction • dispersion of 15 cm from D1/D2 arrangement for free• could be increased for D’ ≠ 0 at the IP• dispersion changes sign left and right from IP• S. Fartoukh proposed a `kissing scheme’ could allow equal

signs of D but vertical D is quite small• optics study relies on Nb3Sn technology:

• 10 m long dipole magnets with B = 15 T• quadrupole magnets with 260 T/m and 80 mm aperture 11 T coil field• IR layout provides magnetic TAS for “free”



Alternative Dipole-First optics (O.Brüning)

proposal of a low-gradient solution that could be realized with NbTi technology

• 18 km -max requires additional Q’ correction • maximum gradient of 70 T/m allows more than 200 mm diameter with a peak coil field of 5.5 T• Dispersion inside the triplet could be increased for D’ ≠ 0 at the IP• Layout still requires an improved TAS absorber



CERN

CERN: the World’s Most Complete CERN: the World’s Most Complete Accelerator Complex (not to scale)Accelerator Complex (not to scale)



CERN

Injector chain for 1 TeV proton beamsinjecting at 1 TeV into the LHC reduces dynamic effects of persistent

currents, i.e.: persistent current decay during the injection flat bottom snap-back at the beginning of the acceleration easier beam

control decreases turn-around time and hence increases integrated luminosity

LL00 [cm[cm-2-2ss-1-1]]

LL

[h][h]TTturnaroundturnaround

[h][h]TTrunrun

[h][h]∫∫200 days200 days L dtL dt [fb [fb--11] ] gaingain

10103434 1515 1010 14.614.6 66 66 x1.0x1.010103434 1515 55 10.810.8 85 85 x1.3x1.310103535 6.16.1 1010 8.58.5 434 434 x6.6x6.610103535 6.16.1 55 6.56.5 608 608 x9.2x9.2

run

L

run

0 Lturnaroundrun

turnaroundrun

L

0

L

turnaroundrun

run

1 (optimum) T

T

TTTTLLdt

eTT

T

with gas = 85 h andx

IBS= 106 h (nom) 40 h (high-L)



Injector chain for 1 TeV proton beams injecting in LHC more intense proton beams with constant

brightness, within the same physical aperture will increase the peak luminosity proportionally to the proton intensity

• at the beam-beam limit, peak luminosity L is proportional to normalized emittance n = , unless limited by the triplet aperture

• an increased injection energy (Super-SPS) allows a larger normalized emittance n in the same physical aperture, thus more intensity and more luminosity at the beam-beam limit.

• the transverse beam size at 7 TeV would be larger and the relative beam-beam separation correspondingly lower: long range beam-beam effects have to be compensated.

2

*zc

*2p

repn2bb 2

1

r

fQL

n

*

csep

d



LHC injector complex upgradeLHC injector complex upgrade• CERN is preparing a road map for an upgrade of its

accelerator complex to optimize the overall proton availability in view of the LHC luminosity upgrade and of all other physics users

• Scenarios under consideration include a new proton linac (Linac 4, 160 MeV) to overcome space charge limitations at injection in the PS Booster and a new Superconducting PS reaching an energy of 50-60 GeV

• This would open the possibility of a more reliable production of higher-brightness beams for the LHC, with lower transmission losses in the SPS thanks to the increased injection energy

• It would also offer the opportunity to develop new fast pulsing SC magnets in view of a Super-SPS, injecting at 1 TeV into the LHC

lhc beam performance and luminosity upgrade scenarios

Documents

beambeam limit

mb luminosity lifetime

h beam lifetime

bunch crossing

f luminosity upgrade

nuclear pp collisionssbb

cerneffective luminosity

beam aperture