ATLAS/CMS Upgrades

Yasuyuki Horii

Nagoya University

on Behalf of the ATLAS and CMS Collaborations

/26Outline

LHC/HL-LHC plan

ATLAS/CMS upgrades

Physics prospects

2

LHC/HL-LHC Plan

/26Overview 4

SM precision studies and BSM searches with 13-14 TeV and 3000 fb-1.

Peak instantaneous luminosity: 5-7x1034 cm-2s-1 — a lot of challenges.

Two upgrade phases: Phase 1 (2019-2020) and Phase 2 (2024-2026).

http://hilumilhc.web.cern.ch/about/hl-lhc-project

/26Luminosity levelling 5

Lower pileup in the experimental detectors

Lower energy deposition by the collisions in the interaction region magnets

The average luminosity is almost the same.

HL-LHC is designed to operate with levelling.

CERN-ACC-2015-0140

ATLAS/CMS Upgrades

/26Challenges 7

Increased luminosity provides a significant challenge for the experiments.

Upgrades are essential to exploit the full potential of LHC and HL-LHC.

Higher radiation dose

Higher pileup

Higher particle rate

Higher event rate

Replacement of some of the detectors

Replacement of the electronics

Overall modifications

on the trigger and readout scheme

→

/26Inner tracker 8

Inner trackers will be in an extreme environment at HL-LHC. 1 MeV neutron equivalent fluence up to 2 x 1016 /cm2.

Ionisation dose up to 10 MGy.

Particle rates up to 2 GHz/cm2 — high occupancy, high bandwidth.

CERN-LHCC-2015-010; LHCC-P-008

Pileup 140 expected at L = 5 x 1034 cm-2s-1

CMS ATLAS

CMS

/26Inner tracker 9

CERN-LHCC-2012-022; LHCC-I-023. CERN-LHCC-2015-020; LHCC-G-166.

Phase 2 ATLAS

Channel occupancy [%] for 200 pileups

Ratio of reconstructed to generated tracks

Entire tracker replacement (all-silicon tracker) at the Phase 2 upgrade.

Radiation tolerance, increased granularity, reduced material, extension to forward, …

No pileup dependence with ≧ 11 hits

Pixel thickness possibly 150 µm, pixel size possibly 50 x 50 µm2

/26Inner tracker 10

Pixel detector replacement in the end of 2016 (as a Phase 1 project).

Entire tracker replacement at the Phase 2 upgrade.

Radiation tolerance, increased granularity,reduced material, extension to forward, …

CERN-LHCC-2015-010; LHCC-P-008

Pixel

Pixel +Strip

Strip

Pixel size considered: 25x100 µm2 and 50x50 µm2

CMS Phase 1/2

/26Calorimeter 11 CMS Phase 2

Endcap calorimeter will be replaced — longevity and performance issues.

CERN-LHCC-2015-010; LHCC-P-008

Frac

tion

of th

e re

spon

se

Hadron fluence 2 x 1014 /cm2 at |η| = 2.6. Defects in lead tungstate scintillating crystal of the electromagnetic calorimeter.

Response degradation also expectedfor the hadron calorimeter.

Light transmission loss

A high-granularity sampling calorimeter with a tungsten/silicon electromagneticpart (EE) followed by brass/silicon (FH) and brass/scintillator (BH) hadronic parts.

High performance at high pileup

/26Muon spectrometer 12

CERN-LHCC-2013-006; ATLAS-TDR-020

Micro-mesh gaseous detector (MM)

ATLAS Phase 1

New Small Wheel will be installed to cope with a relatively high hit rate(~15 kHz/cm2 at L = 7 x 1034 cm-2s-1) and also to improve muon trigger.

Both MM and sTGC for precision tracking and trigger.

Position resolution per layer: ~100 µm.

Segment angle resolution at first-level trigger: ~1 mrad.

Coverage: 1.3 < |η| < 2.7.

/26Trigger 13

More luminosity — more interesting events but also more background.

Without changes, trigger rates exceed the limits of trigger/readout system.

Simply increasing the threshold would kill the signal.

CERN-LHCC-2012-022; LHCC-I-023. CERN-LHCC-2015-019; LHCC-G-165. CERN-LHCC-2015-020; LHCC-G-166.

Choice of ATLAS and CMS at Phase 2 upgrades

Increase trigger rates. First level: ~100 kHz → 750-1000 kHz Storage level: ~1 kHz → 5-10 kHz

Increase latency — improve algorithm. First level: ~3 µs → 6-12.5 µs

Electronics replacements for all sub-systems.

CMS ATLAS

/26Trigger 14

Track trigger implementation in the first-level trigger.

Benefits: improved pT determination, better identification of charged leptons, …

Technologies: studies ongoing for Associative Memories, FPGA, …

Electron trigger

CERN-LHCC-2015-010; LHCC-P-008.

Muon trigger

CMS Phase 2

/26Trigger 15

Calorimeter trigger upgrade Muon trigger upgrade

Higher granularity information provided at first-level trigger. Less sensitive to pileup.

Extend muon trigger acceptancein the barrel by additional chambers.

CERN-LHCC-2013-017; ATLAS-TDR-022-2013. O. Kortner, VCI 2016.

Current

Phase 1

Additional RPCs

Phase 2

Muon A x ε in barrel could be improvedfrom ~70% to ~95%.

Trigger rate reduction for e, γ, …

ATLAS

Physics Prospects — Examples

/26ttH 17

ATL-PHYS-PUB-2014-012

Direct probe of Higgs-top coupling.

Observation expected for ttH, H→γγ.

ATLAS expected: 8.2σ (3000 fb-1).

gg→H and H→γγ indirect (loops). [GeV]γγm

100 120 140 1600

200 Background subtracted eventsSignal Fit

100 120 140 160

Even

ts /

( 2 G

eV )

0

100

200

300

=14 TeVs, -1 L dt = 3000 fb∫ SimulationBackground Fit

ATLAS Simulation Preliminary

/26H→bb 18

ATL-PHYS-PUB-2014-011

[GeV]bbm50 100 150 200 250

Even

ts /

20 G

eV

0

10000

20000

30000

40000

50000

60000

70000

80000

90000 ATLAS Simulation Preliminary

> = 140µ <-1L dt = 3000 fb∫ = 14 TeV s > 200 GeVV

T1 lep., 2 jets, p

VH(bb)x10VZWWMultijettt

t, s+t-chanWtW+bbW+blW+ccW+clW+lZ+bbUnc.

Observation expected for VH, H→bb (V = Z or W).

ATLAS expected significance at 3000 (300) fb-1: 8.8σ (3.9σ).

[GeV]bbm50 100 150 200 250

Even

ts /

20 G

eV

0

500

1000

1500

2000

2500ATLAS SimulationPreliminary

> = 140 µ, <-1 = 14 TeV, 3000 fbs

2 lep, 2 jets, 2 tags, > 200 GeVTZp

ZH x 10DibosonttZ+bbZ+blZ+ccZ+clZ+lUnc.

Access to Higgs-bottom coupling.

/26H→µµ 19

CERN-LHCC-2015-010; LHCC-P-008. ATL-PHYS-PUB-2013-014.

Reduction of the material and betterspacial resolution for tracking at Phase 2. Mass resolution expected:40% better with respect to ‘Phase 1 aged’(radiation damage for 1000 fb-1 assumed).

Observation expected for H→µµ.

ATLAS expected: 7.0σ (3000 fb-1).

Access to Higgs-muon coupling.

[GeV]µµm80 100 120 140 160 180 200

Even

ts /

0.5

GeV

210

310

410

510

610

710

810

910

1010 ATLAS Simulation Preliminary

-1dt = 3000 fb L ∫

= 14 TeVs=125 GeV

H, mµµ →H

µµ →Z

ttνµνµ →WW

[GeV]µµm100 110 120 130 140 150 160

(Dat

a - B

ackg

roun

d) /

0.5

GeV

-5000-4000-3000-2000-1000

010002000300040005000

ATLAS Simulation Preliminary = 14 TeVs

-1dt = 3000 fb L ∫

S+B toy Monte Carlo

S+B model

B-only model

/26Higgs couplings 20

)YκXκ(∆=XYλ∆

0 0.05 0.1 0.15 0.2 0.25

)Zγ(Zλ

Zγλ

gZλ

Zµλ

Zτλ

bZλ

tgλ

WZλ

gZκ

ATLAS Simulation Preliminary = 14 TeV:s -1Ldt=300 fb∫ ; -1Ldt=3000 fb∫

ATL-PHYS-PUB-2014-016. arXiv:1307.1347 [hep-ph].

For various coupling scale factor ratios,the precision of % level expected at 3000 fb-1.

Similar precision expected for ATLAS and CMS.

Fit with a fully generic parametrisation

No assumption on the total width

κgZ (= κgκZ/κH) overall scale parametercommon to all signal channels

No assumption on new particle contributionthrough loops

Hashed areas: current theory systematic uncertainties

/26Higgs couplings 21

CERN-LHCC-2015-019; LHCC-G-165

Significant improvement expected with 14 TeV, 3000 fb-1. Precision test of Yukawa terms for various ‘flavors’: t, b, τ, and µ.

mass (GeV)0.1 1 10 100

1/2

or (

g/2v

)λ

-410

-310

-210

-110

1WZ

t

bτ

µ

68% CL

CMSProjection

(14 TeV)-13000 fb

/26Bs,d→µµ 22

Some of new physics scenariosmay boost the Bs,d→µµ decay rates.

Bs/Bd ratio provides a stringent testof various models beyond SM.

B (Bs→µµ) = (3.65 ± 0.23) x 10-9

B (Bd→µµ) = (1.06 ± 0.09) x 10-10

C. Bobeth, et al., PRL 112, 101801 (2014)

Bs,d→µµ decays are only proceedthrough FCNC processesand are highly suppressed in SM.

B (Bs→µµ) = (2.8+0.7-0.6) x 10-9

B (Bd→µµ) = (3.9+1.6-1.4) x 10-10

D. M. Straub, arXiv:1012.3893

B (Bs→µµ) = (0.9+1.1-0.8) x 10-9

B (Bd→µµ) < 4.2 x 10-10 (95% CL)

CMS and LHCb, Nature 522, 68 (2015)

ATLAS, arXiv:1604.04263 [hep-ex]

/26Bs,d→µµ 23

(GeV)µµm4.9 5 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9

S/(S

+B) W

eigh

ted

Even

ts /

( 0.0

2 G

eV)

0

20

40

60

80

100

120CMS Simulation

)|<1.4µ(η|datafull PDF

-µ+µ→sB-µ+µ→dB

combinatorial bkgsemileptonic bkgpeaking bkg

-1Scaled to L = 300 fb

(GeV)µµm4.9 5 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9

S/(S

+B) W

eigh

ted

Even

ts /

( 0.0

1 G

eV)

0

100

200

300

400

500 CMS Simulation

)|<1.4µ(η|datafull PDF

-µ+µ→sB-µ+µ→dB

combinatorial bkgsemileptonic bkgpeaking bkg

-1Scaled to L = 3000 fb

300 fb-1 3000 fb-1

σ x B predicted by SM assumed.

B (Bs→µµ) precision: 13% B (Bd→µµ) precision: 48% (2.2σ)

B (Bs→µµ) precision: 11% B (Bd→µµ) precision: 18% (6.8σ)

CERN-LHCC-2015-010; LHCC-P-008. K. F. Chen, EPS-HEP 2015.

/26Bs→J/ψφ 24

ATL-PHYS-PUB-2013-010. PRD 91, 073007 (2015).

Luminosity 250 fb-1 3000 fb-1

σ(φs) (Stat.) 0.064 rad 0.022 rad

) [GeV]0s(B

Tp

0 10 20 30 40 50 60 70 80

) [ps

]0 s

(B τσ

00.020.040.060.08

0.10.120.140.160.180.2

> = 20µATLAS 2012 <> = 60 µIBL Layout, <> = 200 µITK Layout, <

ATLAS simulationPreliminary

Run 1

Run 2, …

CP violation due to interference betweendirect decay and decay with Bs-Bs mixing.

New physics can show up in the mixing.

Phase difference between interfering

amplitudes φs extracted from decay time

defined on the transverse plane: .

Improve decay time resolution στ by 30% with respect to Run 1 at ATLAS.

0 0_

φs = -0.0365 rad+0.0013-0.0012

SM global fit by CKMfitter

Bs

Bs

J/ψφ0_

0

Method improvement in arXiv:1601.03297 [hep-ex].

/26t→qγ, qZ, and qH 25

ATL-PHYS-PUB-2013-007. ATLAS-PHYS-PUB-2013-012. CMS PAS FTR-13-016.

FCNC top quark decays are highly suppressed in SM: B < 10-13.

New physics scenarios may enhance the rate up to B ~ 10-4.

HL-LHC expected limits at 95% CL are B = 10-4–10-5.

)γ q→BR(t

-510 -410 -310 -210 -110 1

qZ)

→BR

(t

-510

-410

-310

-210

-110

1

LEP

(q=u only) ZEUS

(q=u only) H1

D0

CDF

)-1ATLAS (2 fb

)-1CMS (4.6 fbATLAS

preliminary (simulation)extrapolated to 14 TeV:

-1300 fb(sequential)

-13 ab(sequential)

-13 ab(discriminant)

95% C.L.EXCLUDEDREGIONS

)4 cH) (x10→Br(t 1 1.5 2 2.5 3

SC

L

-310

-210

-110

1ATLAS Preliminarys = 14 TeV√, -1L dt = 3 ab∫

cutsT

Expected, tight jet p cuts, conservative bkg

TExpected, tight jet p

cutsT

Expected, loose jet p cuts, conservative bkg

TExpected, loose jet p

95%

/26Conclusion 26

Aim for SM precision studies and BSM searcheswith 300 fb-1 (LHC) and 3000 fb-1 (HL-LHC) at ATLAS and CMS.

Potential observation of the processes related with ‘flavors’:ttH, H→bb, H→µµ, Bd→µµ, …

Potential CP-violation measurement of Bs→J/ψφ, …

Increased luminosity (5-7 x 1034 cm-2s-1) provides a significantchallenge for the experiments.

High radiation dose, pileup, particle rate, and event rate.

Overcome the difficulties by the upgrades in various aspects.

Backup Slides

/26Calorimeter 28 ATLAS Phase 2

Maintain required performance under HL-LHC conditionsand therefore do not need replacement with possible exception for FCal.

FCal replacement with high-granularity one (100 µm gap) under discussion.

Addition of timing detector (intrinsic resolution O(10) ps) under discussion.

LAr: radiation hardness

CERN-LHCC-2015-020; LHCC-G-166

/26Calorimeter 29 CMS Phase 2

Radiation dose at 3000 fb-1 for the scintillating tiles of the endcap hadron calorimeter will reach up to 300 kGy— response degradation expected.

CERN-LHCC-2015-010; LHCC-P-008

For the new endcap calorimeter, exploit advances in silicon detectors in terms of cost per unit areaand radiation tolerance. The silicon sensors to be usedwill be simple, large area, andsingle-sided.

/26Muon spectrometer 30

CERN-LHCC-2013-006; ATLAS-TDR-020

• Current drift tube chambers: inefficiency and resolution degradationwith hit rate above 300 kHz/tube.

• Impact on the endcap inner layer with L > 1034 cm-2s-1.

ATLAS

/26Muon spectrometer 31 CMS Phase 2

(i) new irradiation tests must be performed to confirm that all types of existing muon detectors will survive the harsher conditions.(ii) additional muon detectors in the forward region 1.6 < |η| < 2.4 to increase redundancy and enhance the trigger and reconstruction capabilities.(iii) extension of muon coverage up to |η| = 3 or more behind the new endcap calorimeter to take advantage of the pixel tracking coverage extension.

Possible additional chambers GEM — micro-pattern gas amplification detector RPC — time resolution of ~100 ps for pileup mitigation

CERN-LHCC-2015-010; LHCC-P-008

/26Higgs couplings 32

Scenario 1: all systematic uncertainties unchanged.

Scenario 2: improved theoretical/systematic uncertainties.

CERN-LHCC-2015-010; LHCC-P-008. ATL-PHYS-PUB-2014-016.

iy

-310

-210

-110

1 Z

W

t

b

τ

µ

ATLAS Simulation Preliminary

= 14 TeVs

νlνl→WW*→4l, h→ZZ*→, hγγ→hγZ→, hµµ→bb, h→, hττ→h

]µκ, τκ, bκ, tκ, Wκ, Zκ[

=0i,uBR

-1dt = 300 fbL∫-1dt = 3000 fbL∫

[GeV]im

-110 1 10 210R

atio

to S

M

0.80.9

11.11.2

/26Bs,d→µµ 33

• Without trigger upgrade,

unsustainable event rate at HL-LHC.

• Track trigger with upgraded CMS

detector plays an essential role.

• Invariant mass mµµ resolution at

Level-1 trigger expected: ~70 MeV.

• Level-1 trigger rate expected:a few hundred Hz (<< 1 MHz).

(GeV)µµm4 5 6

Even

ts /

(0.0

2 G

eV)

1

10

210

310

410

510

610

710

810 CMS Simulation

-µ+µ→sB-µ+µ→dB

Background

Total signal

-1Scaled to L = 3000 fb

L1TrkMu (PhaseII) Trigger) > 3 GeVµ(

Tp

)| < 2µ(η|) > 4 GeVµµ(

Tp

)| < 2µµ(η|)| < 1 cmµµ(z d∆|

) < 6.9 GeVµµ3.9 < m(

at Level-1 trigger

CERN-LHCC-2015-010; LHCC-P-008. K. F. Chen, EPS-HEP 2015.

/26Bs→J/ψφ 34

Opposite-side tagging studied and calibratedby B±→J/ψK± (flavor provided by kaon charge).

Di-muon trigger with pT > 11 GeV (both muons)assumed at ATLAS at HL-LHC.

Systematic error of Run 1 analysis:

uncertainties in flavor charge tagging,

likelihood fit modelling,

trigger efficiency determination,

contribution of B→J/ψK* decays,

inner tracker alignment

— will benefit from the larger data samples.

ATL-PHYS-PUB-2013-010

Number of reconstructed PV

0 10 20 30 40 50 60 70 80 90 100

) [ps

]0 s

(B τσ

00.020.040.060.08

0.10.120.140.160.180.2

> = 20µATLAS 2012 <

> = 60 µIBL Layout 11,11 <

> = 200 µITK Layout 11,11 <

ATLAS simulationPreliminary

Run 1Run 2, …

Slight στ increase (14%) in Run 2 with number of primary vertices — but stable at > 40.

/26t→qH 35

arXiv:1509.06047v2 [hep-ex]

Current 95% CL upper limit on the branching ratio at the order of 10-3.