LHC & ATLAS UPGRADES

R. S. Orr

Toronto Group Meeting

• Physics Case for LHC Upgrade

• LHC Issues

• General ATLAS Issues

• ATLAS Canada R&D

Alors, c’est fini!Et maintenant?

Diameter 25 mBarrel toroid length 26 mEndcap end-wall chamber span 46 mOverall weight 7000 Tons

ATLAS Detector Systems

LHC Prospects• Date for 7 TeV beam commissioning: July 2008

• Initial physics run starts “late” 2008 collect ~10 fb-1 /exp (2.1033cm-2 s-1) by “end of 2009”

• Depending on the evolution of the machine… collect 200-300 fb-1 /exp (3.4-10.1033cm-2 s-1 ) in 5-6 years time

Already time to think of upgrading the machine

Two options initially discussed/studied• Higher luminosity ~1035cm-2 s-1 (SLHC)

–Needs changes in machine and particularly in the detectors Start change to SLHC mode some time 2012-2014 Collect ~3000 fb-1/experiment in 3-4 years data taking.

• Higher energy? – LHC can reach s = 15 TeV with present magnets (9T field) s of 28 (25) TeV needs ~17 (15) T magnets R&D + MCHf needed

–Don’t discuss today

LHC context in 2011 - 2015

SLHC make be only game in town for a LONG time.

Physics Case for the SLHC

The use/need for for the SLHC will obviously depend on how EWSB

and/or the new physics will manifest itself

This will only be answered by LHC itself

What will the HEP landscape look like in 2012??

Rough expectation for the SLHC versus LHC Improvement of SM/Higgs parameter determination Improvement of New Physics parameter determinations, if discovered Extension of the discovery reach in the high mass region Extension of the sensitivity of rare processes

Indicative Physics Reach

Approximate mass reach machines: s = 14 TeV, L=1034 (LHC) : up to 6.5 TeV s = 14 TeV, L=1035 (SLHC) : up to 8 TeV s = 28 TeV, L=1034 : up to 10 TeV

Units are TeV (except WLWL reach) Ldt correspond to 1 year of running at nominal luminosity for 1 experiment

† indirect reach (from precision measurements)

Ellis, Gianotti, ADRhep-ex/0112004+ updates

PROCESS

LHC

14TeV

100 fb-1

SLHC

14TeV

1000 fb-1

SLHC

28TeV

100 fb-1

LinCol

0.8 TeV

500 fb-1

LinCol

5 TeV

100 fb-1

Squarks 2.5 3 4 0.4 2.5

WLWL 2σ 4σ 4.5σ

Z’ 5 6 8 8† 8†

Extra Dim (δ=2) 9 12 15 5 - 8.5† 30 - 55†

q* 6.5 7.5 9.5 0.8 5

Λcomp 30 40 40 100 400

TGC (λγ) 0.0014 0.0006 0.0008 0.0004 0.00008

The Higgs at the LHC

• First step–Discover a new Higgs-like particle at

the LHC, or exclude its existence• Second step

–Measure properties of the new particle

to prove it is the Higgs • Measure the Higgs mass

• Measure the Higgs width

• Measure (cross sections x branching ratio)s

• Ratios of couplings to particles (~mparticle)

• Measure decays with low Branching ratios (e.g H)

• Measure CP and spin quantum numbers (scalar particle?)

• Measure the Higgs self-coupling (HHH), in order to reconstruct the Higgs potential

Make sure it really is Higgs

SLHCStatisticsNeeded

LHC~1 good year of data

Higgs Self Coupling Measurements

~ v mH

2 = 2 v2

Once the Higgs particle is found, try to reconstruct the Higgs potential

Djouadiet al.

SM/2 << 3SM/2

Not possible at the LHC

Too much backgr.

Higgs Self Couplings

LHC : (pp HH) < 40 fb mH > 110 GeV + small BR for clean final states no sensitivity

SLHC : HH W+ W- W+ W- jj jj studied

S S/B S/B

mH = 170 GeV 350 8% 5.4mH = 200 GeV 220 7% 3.8

6000 fb-1

-- HH production may be observed first at SLHC: ~150 <MH<200 GeV-- may be measured with statistical error ~ 20-25%

LC : precision up to 20-25% but for MH < 150 GeV (s 500-800 GeV, 1000 fb-1)

Beyond the Standard Model

Supersymmetry Extra dimensions

•New physics expected around the TeV scale •Stabilize Higgs mass, Hierarchy problem, •Unification of gauge couplings, Cold Dark Matter,…

+…

+ a lot of other ideas…Split SUSY, Little Higgs models, new gauge bosons, technicolor, compositness,..

Bulk

SM Brane

G

G

Discovery reach for squarks/gluinos

Time mass reach

1 month at 1033 ~ 1.3 TeV1 year at 1033 ~ 1.8 TeV1 year at 1034 ~ 2.5 TeV

ATLAS 5 discovery curves

SUSY : Discovery Reach

LHC 2.5 TeVSLHC 3 TeV

5 discovery reach )g~( m ),q~( m

SUSY Higgses h,H,A,H

• Green region only SM-like h observable with 300 fb-1/exp

• Red line: extension with 3000 fb-1/exp• Blue line: 95% excl. with 3000 fb-1/exp

Heavy Higgs observable region increased by ~100 GeV at the SLHC.

Time Scale of an LHC upgrade

L at end of year

time to halve error

integrated L

radiationdamage limit~700 fb-1

•Life expectancy of LHC IR quadrupole magnets is estimated to be <10 years due to high radiation doses

• Statistical error halving time exceeds 5 years by 2011-2012 → it is reasonable to plan a machine luminosity upgrade based on new low- IR magnets around ~2014-2015

design luminosity

ultimate luminosity

Jim Strait, 2003

Machine Upgrade in Stages

• Push LHC performance without new hardware– luminosity →2.3x1034 cm-2s-1, Eb=7→7.54 TeV

• LHC IR upgrade – replace low- quadrupoles after ~7 years

peak luminosity →4.6x1034 cm-2s-1

– low- quadrupoles plus dipoles, plus crab cavities…. peak luminosity →15.5 x 1034 cm-2s-1

• LHC injector upgrade– peak luminosity →9.2x1034 cm-2s-1

• LHC energy upgrade– Eb→13 – 21 TeV (15 → 24 T dipole magnets)

Beam-Beam Limit Luminosity Equation

1

2 2 22 *

1revprofileb bb

p

fL n Q F

r

IR upgrade

injector upgrade

LHC +injectorchanges

LHC+injectorchanges

2

1;

21c z

x

F

Piwinski angle

luminosity reduction factor

nominal LHC

Nominal Crossing Angle “at the edge”

Summary of Luminosity Upgrade

Scenarios for with acceptable heat load and events/crossing 35 2 110L cm s

25-ns: push to limit*

• Slim magnets inside detector?• Crab Cavities• High Gradient, Large Aperture Quads 3Nb Sn

50-ns: Fewer bunches, higher charge

• Realizable with • Beam-Beam tune shift due to large Piwinski angle?• Luminosity leveling via bunch length and tuning

NbTi

*

Early Separation (ES)

• ultimate LHC beam (1.7x1011 protons/bunch, 25 spacing)

• squeeze * to ~10 cm in ATLAS & CMS

• add early-separation dipoles in detectors starting at ~ 3 m from IP

• possibly also add quadrupole-doublet inside detector at ~13 m from IP

• and add crab cavities (Piwinski~ 0)

→ new hardware inside ATLAS & CMS detectors, first hadron crab cavities

ultimate bunches + near head-on collision

stronger triplet magnetsD0 dipole

small-angle

crab cavity

optionalQ0 quad’s

challenges: • D0 dipole deep inside detector (~3 m from IP),• optional Q0 doublet inside detector (~13 m from IP),

• strong large-aperture quadrupoles (Nb3Sn)

• crab cavity for hadron beams (emittance growth), or shorter bunches (requires much more RF)

• 4 parasitic collisions at 4-5 separation,• low beam and luminosity lifetime ~*

merits:• most long-range collisions negligible,• no geometric luminosity loss,• no increase in beam current beyond ultimate, • could be adapted to crab waist collisions (LNF/FP7)

ES Scenario

Large Piwinski Angle (LPA) • double bunch spacing to 50 ns, longer & more intense bunches with Piwinski~ 2

• *~25 cm, do not add any elements inside detectors• long-range beam-beam wire compensation

→ novel operating regime for hadron colliders

fewer, long & intense bunches + nonzero crossing angle + wire compensation

wire

compensator

larger-aperture triplet magnets

merits:• no elements in detector, no crab cavities,• lower chromaticity,• less demand on IR quadrupoles

(NbTi expected to be possible),• could be adapted to crab waist collisions (LNF/FP7)

challenges: • operation with large Piwinski parameter unproven for

• hadron beams (except for CERN ISR),

• high bunch charge,

• beam production and acceleration through SPS,

• larger beam current,

• wire compensation (almost established),

LPA Scenario

Principle of Early Separation

D0 D0

D0 D0

Full Early Separation (50 ns only if D0 not in

inner detector)

Partial Early Separation

(25 or 50 ns)We need a residual crossing angle

First encounter

First encounter

Stronger focusing with cancellation of the geometrical luminosity loss

25ns preferred by ATLAS

D0 is just in front of FCal

Possible locations in ATLAS

A

B

C

D

• Stay out of A

• B,C possible location of D0, but need more calculations to avoid to damage the muon system

• D possible location of Q0 or D0, probably the least problematic one

Detectors: General Considerations

Normalised to LHC values.

104 Gy/year R=25 cm

LHC SLHC

s 14 TeV 14 TeVL 1034 1035

Bunch spacing t 25 ns 25/50 ns

pp (inelastic) ~ 80 mb ~ 80 mbN. interactions/x-ing ~ 20 ~ 300/400(N=L pp t) dNch/d per x-ing ~ 150 ~ 2000/2500<ET> charg. particles ~ 450 MeV ~ 450 MeV

Tracker occupancy 1 10/20Pile-up noise in calo 1 ~9Dose central region 1 10

In a cone of radius = 0.5 there is ET ~ 200GeV.

This will make low Et jet triggering and reconstruction difficult.

Detector Upgrade• ATLAS has begun studying what needs to be upgraded for 1035cm-2s-1

instantaneous luminosity– ~10 harsher pileup, radiation environment– Also constrained by existing detector: what can be moved/stored where/when

• Major ID overhaul foreseen– TRT replaced by Si Strips– Pixels move to larger radius– New technology for innermost layers

• Calorimeters– New FE electronics for HEC– New cold or warm FCAL– Opening endcap cryostat implies a long installation schedule (~2-3 years)

• Schedule to fit 2016 timescale– Aim for upgrade TDR in 2010 to allow adequate procurement/construction

• Also Trigger, FE in general, etc.. etc.. etc…

LAr Calorimeters at sLHC - Overview

• Critical issues– ion build up and heat load

• The HiLum ATLAS Endcap Project

• Radiation hardness: –R&D for HEC cold electronics;

FCal - Heatload

V Strickland

• Simulation of LAr FCAL beam heating

– Maximum temperature 93.8K

– recently conclude unlikely that will LAr boil

• Improve FCAL cooling (open endcap cryostat)?

– ~2-3 year round-trip – big timing challenge

• New “warm” FCAL plug?

• Main FCAL issue is Voltage drop on protection resistors

Mini-FCalNeutron Shielding

EMEC HEC

FCal

Pu

mp

300mm

+ve Ion Buildup – Distorts Electric Field

•EMEC and HEC OK•FCAL: may go above 1 close to inner wall•Turn off inner part of FCAL•Instal mini warm FCAL in front – reduce energy deposited in FCAL1

limit r=1 @LHC

sLHC

limit r=0.1 @sLHC

HiLum ATLAS Endcap Project

• Goal: establish limitations on the operation of the endcap calorimeters (FCAL, EMEC, HEC) at highest LHC luminosities.

• R&D: 'mini modules‘ of FCAL, EMEC and HEC type, each in one separate cryostat;

• IHEP Protvino: beam line # 23: from 107 up to 1012 p/spill; E= 60/70 GeV;

• Arizona, Dresden, JINR Dubna, Kosice, Mainz, LPI Moscow, MPI Munich, BINP Novosibirsk, IHEP Protvino, TRIUMF, Wuppertal.

FCAL module

• 4 'standard' HEC gaps (HEC1)• 4 read-out channels• 4 HV lines (one per subgap)

HiLum Test Modules

• Order of magnitude increase in Data rates, Occupancy, Irradiation• No TRT – Si strips• Pixels moved to larger radius• New technology for inner layers• R&D required on sensors, readout, and mechanical engineering

Inner Detector Replacement

Tracks

• Preserve (improve?) tracking performance in SLHC environment• Need to replace TRT: all-silicon ID• Minimize material

TRT

silicon

P Nevski

Strawman Layout of Tracker

b layer + 3 pixel layers

3 short-strip layers(2.5cm) + stereo

2 long-strip layers(9cm) + stereo

moderator

|η|<2.5

Semi-projective gaps

Pixel-layer Technologies

Si pixel sensorBiCMOS analogue

CMOS digital

Cluster3

Cathode (drift) plane

Integrated Grid (InGrid)

Cluster2 Cluster1

Slimmed Silicon Readout chipInput

pixel

1mm,100V

50um, 400V

50um

• Diamond pixels

Rad hard, low noise, low current

– cost, signal, uniformity?• May test in pre-SLHC b-layer replacement

(~2012)

Harshest radiation environment (R~4cm)

– investigate new technologies• 3D Si

Highest signal - fabrication?

• Thin silicon + 3D interconnectsConservative – high voltage

• Gas over thin pixel (GOSSIP)

Low material – sparks?

Schedule

Strawman & options fixed Dec 2006

ID R&D, conceptual design 2007-2009

TDR Feb 2010

ID cooling PRR April 2010

Silicon sensor PRR July 2010

ID FE electronics PRR Oct 2010

b-layer replacement Ready 2012

Procure parts, component assembly 2010-2012

Start surface assembly March 2012

Stop data taking Sep 2014

Remove old detectors, install new 2014-2015

Data April 2016

Tracker Upgrade work in Canada

Diamond Sensors – Toronto, Carleton, Montréal, Victoria +…..– Prove radiation tolerance of pCVD diamond pixel prototypes– Industrialize bump-bonding– FE electronics– Mechanical structure– Test beam program 2008-2009

Electronics – Carleton, UBC, York, TRIUMF +…..

– FE ASICS – Si FE module controller– Initially FPGA, Move to ASIC

– Contribute to system design – develop expertise– Backend (eg RODs) later in upgrade path– TRIUMF Technical manpower

• Superconductor HTS-BSCCO or

• Well above demonstrated

• 20++ years for R&D, ? years production

• ?G$

LHC Energy Doubler 14*14 TeV

Dipoles: Bnom=16.8T, Bdesign=19T

• Superconductor

• 16T demonstrated at 4K

• 10 years for R&D, 10 years production

• 3G$

3Nb Sn

LHC Energy Tripler 21*21 TeV

Dipoles: Bnom=25T, Bdesign=29T

3Nb Sn

3Nb Sn