Status of the LHCb experiment Elie Aslanides CPPM, IN2P3-CNRS et - - PowerPoint PPT Presentation

Status of the LHCb experiment

Elie Aslanides

CPPM, IN2P3-CNRS et Université de la Méditerranée, France

• n behalf of the LHCb Collaboration

LISHEP Itacuruçá, Rio de Janeiro, April 4, 2006

Introduction LHCb detector status Expected performances Conclusion

Introduction

LHCb is the dedicated B physics experiment at the LHC devoted to the precision study of CP violation and rare decays.

The LHCb Collaboration includes 47 institutes from 15 countries and more than 600 members.

• Extend B-physics results obtained in B-factories and the Tevatron
• Search for new physics in a complementary way to ATLAS/CMS

LHCb may be the only running Beauty Physics experiment, after the B-factories (if no Super B-factories are approved) !

Baseline physics goals of LHCb…a second generation experiment

With a luminosity <L> ≈ 2×1032 cm−2s−1, 107s/year, 2 fb-1 /year, LHCb should reach in 5 years unprecedented precisions σ(>5 y) SM (expect) φs(Bs→ccss) ~0.013 ~ 0.035 γ (DsK, D0K*0) ~1° ~60° (tree only) γ (ΚΚ+ππ) ~2° ~60 (tree + penguin) Br(Bs→µ+µ−) ~0.7×10−9 ~3.5×10−9 Bd→K∗0µ+µ− AFB(s) sensitive to NP 22k events expected >5y (Super-KEKB similar… by 2020!) Belle 357 fb−1 114 events

1000 fb−1 by ~2009 AFB(s)

±0.04

Beauty production at the LHC

At √s = 14 TeV, pp collisions have large σbb ~ 500 µb but small compared to the total, σbb/σtot ~ 5·10-3 interesting B decays have low b.r. ~10-5 Bunch crossing rate at the LHC is 40 MHz LHCb average L ~ 2 ×1032 cm-2s-1 → 2 fb-1 / year (107 s) → 1012 bb produced/year! → most events due to single interactions per bunch crossing!

1 2 3 4

Luminosity [cm− 2 s− 1 ]

1031 1032 1033 0.2 0.4 0.6 0.8 1.0

Probability

pp interactions/crossing

Beauty production at the LHC

Forward peaked, correlated " b anti-b" pair production"

1 10 10 2

• 2

2 4 6

eta of B-hadron pT of B-hadron

ATLAS/CMS LHCb

p p

LHCb is a forward spectrometer (10 – 300 mrad)

pT vs η for detected B hadrons 100 µb 230 µb

Experimental requirements …

• Efficient and flexible Trigger
• High quality Event Reconstruction

particle identification hadrons…,µ’s and e’s, as well as γ’s, π0’s excellent tracking and vertexing good p, E, Mass and τ resolutions

• Powerful Readout and on line processing (HLT)

~ 1 m m

b

b

Spectrometer

p p

250 mrad 10 mrad Vertex Locator Dipole magnet Tracking system Calorimeters Muon system RICH detectors

Shielding wall Electronics + CPU farm Offset Interaction Point Detectors can be moved away from beam-line for access

LHCb at Point 8

Introduction LHCb detector status Expected performances Conclusion

November 2005

The beam pipe…

Located in the high rapidity region where particle density is high, it is a major source of secondaries! UX85/1 UX85/2 UX85/3

1mm 1-2 mm 3-4 mm

VELO window Al flanges and bellows Stainless steel flanges and bellows UX85/4 3 Beryllium sections Stainless steel section

Production is progressing well…

VeLo window prototype

UX 85/1 (Be) COMPLETED

UX 85/2 (Be) tested at IHEP, Protvino

Acceptance tests NEG coating …

UX 85/3 ( Be) under construction by Kompozit, Moscow, All components in production. Installation early summer’06, fits in general planning.

The magnet

Warm dipole magnet ∫B dl = 4 Tm Iron yoke 15 ton; Power 4.2 MW Nominal field reached on Nov. ‘04

Magnetic field measurements

60 3D Hall probes

VeLo

TT magnet

• ∫ B dl in VELO-TT region

needed for fast online pT

• B inside RICH is ok for

the HPD operation!

• Dec. ’04; June ’05 including RICH1

shield and all iron

• Dec. ’05 [final and Polarity ±]

Vertex Locator

Key element around the IP

~1m 8cm 8cm

21 stations of Silicon 300µ-strip detectors r-φ geometry variable pitch [ r (40-102µ);φ (36-97µ) ] 172 k channels

VELO detector vacuum box … 300µ AlMg3

RF box corrugations RF box corrugations

RF shield for sensors + electronics guides the beams mirror charge suppresses dynamic vacuum phenomena suppresses electron multipacting

To give precise vertex reconstruction VELO approaches to 8 mm from beam Radiation ~1.5 1014 n eq./cm2/y Expected lifetime ~3 years; Si to be replaced Detector stations in 2 retractable halves Complex mechanics to allow retraction during beam injection (~completed) VELO uses vacuum like a « roman pot » VELO operated at –5°C (CO2 cooling)

Sensors characteristics: n+n type, double metal layers 300 µ, laser cut FE electronics (Beetle chip) mounted on thin kapton, connected to the sensors via Pitch Adapters Modules production started ! should be completed ~end summer’06 RF boxes installed September ’06 Vacuum tests October ‘06

VELO modules installed in RF boxes >January ’07

Trigger Tracker Si µ-strip detector 144 k channels

TT T1 T2 T3

Outer Tracker Straw Tubes 56 k channels Inner Tracker Si µ-strip detector around the beam pipe 130 k channels

Tracking…

Trigger Tracker

Four Si µ-strip detection layers, ~ 8 m2 of silicon, covering the nominal LHCb acceptance. Arranged in two double layers (0°,+5°) and (-5°,0°) 30 cm apart. Together with VELO, the TT measures the pT of the high IP tracks for use in the trigger. Offline: decay tracks of long-lived neutral particles decaying outside the VELO fiducial volume.

500 µ silicon, CMS OB2-type sensors Strips: pitch 198 µ; length 11, 22, 33 cm Radiation: ≤ 9×1012 neq/cm²/10 year Operated at 5°C

Almost all sensors and components in hand… TT support structure will be installed in April TT installation in UX85 between June and October ’06.

3 Tracking stations

Outer Tracker Inner Tracker

T stations: Outer Tracker

3 stations, each made up of Kapton/Al straws glued together to form modules Installation in UX85 November ’06 Commissioning starting December ‘06

4 double-layers (0°,+5°) and (-5°,0°) Ar/CF4/CO2

modules 64-cells wide modules only ~0.7% of 1 X0: “light” panel (Rohacell core with carbon skins) “light” straws

5 mm

T stations: Inner Tracker

Silicon strip detectors close to beam pipe, in region of high occupancy:

• nly 2% of area, but 20% of tracks

arranged in boxes around beam pipe 410 µm thick for two-sensor ladders 320 µm thick for single sensors

1.2 m

0.4 m

Same sensors as Trigger Tracker Strip length 11, 22 cm, pitch 198 µ Four layers (0°, +5°, -5°, 0°) Cooled -5 °C Radiation ≤ 9×1012 neq/cm²/10 years

Installation in UX85 in June ‘06

Ring Imaging CHerenkov

RICH1

RICH2

Three radiators in two detectors to give π-K separation from 2-100 GeV

Θ 25 to 250 mrad p 2 to 60 GeV/c Θ 15 to 100 mrad V 120 mrad H p 17 to 100 GeV/c Aerogel + C4F10 200k channels 295k channels CF4

Novel photon detectors: Hybrid Photon Detectors

Pixels 5 10 15 20 25 30 Pixels 5 10 15 20 25 30 500 1000 1500 2000 2500 3000

~500 tubes, each with a 32x32 pixel sensor array Pixels size (500x500 µm2) Operated at 20 kV

~150 HPD’s already in hand! Production at a rate ~30 HPD /month…

Test beam image

RICH1 combines the use of aerogel n=1.03 and C4F10 n=1.0014 radiators for low momentum particles

7 x 14 HPD array Light weight spherical mirrors Glass plane mirrors

• utside the acceptance

Quartz windows to HPD

material budget ~7.5% X0

High clarity aerogel

Switched from Be to C spherical mirrors, quite recently! New design tested April ’06; production expected < end ’06. RICH-1 installation completed ~ March 2007.

High clarity aerogel was developed with a Novosibirsk group … now in production!

Magnetic shielding of RICH1 now installed

Magnetic measurements in the HPD plane show residual field of less than 25 Gauss

RICH2 uses CF4 gas radiator for high p particles

(n=1.0005)

Spherical Mirror Photon funnel Shielding Central pipe Support Structure Flat mirror

7 m Gas vessel: 100 m3

RICH-2 completed end 2006 In position 11.19.2005

Mirrors alignment ~150 µrad mirror movement ~100 µrad

• cf. RICH-2 Cherenkov angle resolution

~ 700 µrad!

Calorimeter system

HCAL ECAL SPD/PS

SPD/PS

2 planes of Scintillating Pads + 2 X0 Pb (1.5 cm); 0.1λI

ECAL

Pb – scintillator Shashlik calorimeter, 25 X0; 1.1 λI

HCAL

Fe – scintillator tile calorimeter, 5.6 λI

19k channels, R/O by WLS fibres to PM or MaPMT’s

Pre-shower SPD / PS

2 scintillator pad planes

• n either side of a Pb absorber

MAPMT PS SPD support structure

MAPMT SPD VFE + ECAL PS

particles

Clear fibers

16 Super modules x (2 x 13 modules)

2 x 5952 channels

Inner Middle Outer Modules

4x4 cm2 6x6 cm2 12x12cm2 4 super modules per half detector MAPMT+ VFE R/O cables Moving cable trays

PS/SPD installation has started…

…the Pb modules are in place! Detector installation April to ~September ‘06

Electromagnetic (shashlik) calorimeter

Two retrievable halves

Chariot Electron. platform modules Beam plug

3312 modules, 25X0 Pb σE/E = 10% /√E ⊕ 1% stacked: ~ 6 m high positioning agrees to specifications to <1mm!

In place since June ’05…

Commissioning 6000 ch.with LED monitoring going on!

Electron. platform Chariot modules Beam plug

particles PMT scintillators fibers light-guide

Hadronic Calorimeter HCAL: σE/E = 80% /√E ⊕ 10%

HCAL module

Tile calorimeter of 52 HCAL modules

8 blocks of Sc/Fe

HCAL 1500 ch. commissioning using LED monitoring going on!

±1.5 mm

HCAL in place … since September ‘05

2005 was the year of “detector assembly and installation”

HCAL ECAL

2006 is the year of R/O electronics installation and detector commissioning

The Muon system

M1 M2 M3 M4 M5

provides muon identification and contributes to the Trigger

The muon detector is composed of 5 stations

MWPC’s are used everywhere except in the highest rate (>100 kHz/cm2) inner part of M1 where triple-GEM’s are used.

~500 kHz/cm2 < 184 kHz/cm2 > Efficiency > 96% in 20ns Ar(45)/CO2(15)/CF4(40)

The muon system has a projective (x,y) geometry pointing to the IP to facilitate the search of µ candidates in the L0 trigger processor. 20 types of chambers: 1368 MWPC’s and 24 « triple-GEM’s ». 125k physical channels; 26k logical channels.

Status of the Muon detector …

MWPC’s and 3-GEM’s are under construction 1053 chambers have been produced (~tested) so far (March 31)

M1R4 M1R1

Expected end of production M2, M3. M5 summer ’06; M4 October ’06; M1 end ’06.

Muon detector installation and commissioning

Muon towers assembled with electronics and gas racks

M2, M3, M4, M5 ≤ Jan 2007; M1 Jan. – March 2007 MUON commissioning Feb –March ‘07

The LHCb trigger

Pile up system Calorimeters Muons Level 0 pT µ, e, h, γ HLT Confirms L0 Associates pT/IP Explores µ, e, h, γ Selects event types 1 MHz

storage

Full detector information 40 MHz Custom Electronics 4 µs latency Processor farm 2 kHz Event size ~35kB

Event selection

A current thought for the band width at the HLT output L0 efficiencies

ET >3.6 GeV pT >1.1 GeV ET >2.8 GeV ET >2.6 GeV

B (data mining) Trigger Inclusive b (e.g. b→m) 900 Hz Charm (mixing & CPV) PID D* candidates 300 Hz J/ψ, b→J/ψX (unbiased) Tracking High mass di-muons 600 Hz B (core program) Tagging Exclusive B candidates 200 Hz Physics Calibration Event type HLT rate

Trigger status

L0 components production started L0 commissioning: components in autumn ’06 trigger early ’07 HLT software ready June ’06 Event Filter Farm installed early ‘07

L0µ Processor card

9U × 220 mm

18 layers, Class 6 32+2 optical links at 1.6 Gb/s 5 StratixGX FPGAs 96 copper serial links at 1.6 Gb/s

Computing

A Filter Farm of ~2000 CPU nodes at pit 8 and Extensive use of LCG for offline

Fully assembled demo-rack used in the RTTC ‘05

Real Time Trigger Challenge in 2005

( test-bed with 44 CPU nodes) implemented and tested DAQ architecture run the trigger algorithms, test and control of the EB and the data tranfer by the ECS R/O of ‘complete detector’ at 1 MHz All components finalized; many ordered. Installation: ECS and DAQ hardware ’06 Event Filter Farm 2007

Introduction LHCb detector status Expected performances Conclusion

Expected performances

Studied … using fully-simulated events

• Pythia tunned on CDF, UA5
• GEANT4
• Multi-particle interactions
• Spill-over effects
• full pattern recognition

Interactive analysis display PANORAMIX

[See talk of Leandro de Paula for the Physics Expectations]

• Reconstruction of tracks passing through full spectrometer:

efficiency ~ 95%, with a few percent of ghost tracks

• Momentum resolution ∆p/p ~ 0.4%
• Impact parameter resolution σIP ~ 20 µm for high-pT tracks
• From VELO vertex detection the proper time resolution ~40 fs

Tracking

Particle Identification

• RICH system provides excellent hadron identification 2–100 GeV

→ K tagging and clean separation of two-body B decays

• Lepton ID: for e (µ) in ECAL (Muon)

efficiency ~ 90% for π misid rate of < 1%

20 40 60 80 100 20 40 60 80 100

Momentum (GeV/c) Efficiency (%)

Κ → Κ π → Κ

250 500 750 1000 1250 1500 1750 2000 2250 5 5.1 5.2 5.3 5.4 5.5

Invariant mass [ GeV/c2 ] With RICH

1000 2000 3000 4000 5000 6000 7000

Events / 20 MeV/c2 No RICH

Bd ππ Bd πK Bs πK Bs KK Λb pK Λb pπ

→ → → → → →

π–K separation

Neutrals reconstruction

Reasonable efficiency for π0 has been achieved for B0 → π+π−π0 study, using both “resolved“ (separate clusters) and “merged” cluster shapes in the calorimeter (unassociated to charged tracks).

Resolved Merged

25 50 75 100 125 150 175 200 225 0.35 0.4 0.45 0.5 0.55 0.6 0.65

π+π−π0 mass (GeV)

Recent study of η → π+π−π0 gave a mass resolution ~ 12 MeV (resolved)

KS → π+π− efficiency ~ 75% if decay in VELO, lower otherwise. Modes with multiple neutrals will be challenging …

Flavour tagging

SV QVTX

tagging B

2.1 1.0 2.4 0.4 1.0 LHCb 2.2 2.3

• 0.5

0.7 CMS 2.1 1.6

• 0.3

0.7 ATLAS Same side Jet/vertex charge Kaon Electron Muon Tag Tagging power εD2 = ε(1−2w)2 (in %) Combined tagging power for BS in LHCb is ~ 6%

Note ~2% at the Tevatron ~30% at B-Factories

Tagging power for B0 ~ 4%

(reduced same side tagging)

e, µ from semi-leptonic decays K± from the b→c →s jet/vertex charge same side π/K

signal B

Recent Neural Network based study achieved 9% for BS tagging!

Illustration of B physics sensitivity

• bservation of Bs–Bs oscillation
• Use mode Bs → Ds

−π+ ; 1 year of data (80k selected events)

for ∆ms = 20 ps-1 (SM preferred) + Dilution by flavour tagging: εD2 ~ 6% for Bs decays + Proper time resolution ~ 40 fs + Signal/Background ~ 3 (from 107 inclusive bb events) + Effect of acceptance Oscillations still clearly seen!

Proper time (ps) Events

1000 800 600 400 200 1 2 3 4 5 Perfect reconstruction

Illustration of B physics sensitivity

• bservation of Bs–Bs oscillation
• Use mode Bs → Ds

−π+ ; 1 year of data (80k selected events)

for ∆ms = 20 ps-1 (SM preferred) + Dilution by flavour tagging: εD2 ~ 6% for Bs decays + Proper time resolution ~ 40 fs + Signal/Background ~ 3 (from 107 inclusive bb events) + Effect of acceptance Oscillations still clearly seen!

Proper time (ps) Events

1000 800 600 400 200 1 2 3 4 5 Perfect reconstruction + flavour tagging

Illustration of B physics sensitivity

• bservation of Bs–Bs oscillation
• Use mode Bs → Ds

−π+ ; 1 year of data (80k selected events)

for ∆ms = 20 ps-1 (SM preferred) + Dilution by flavour tagging: εD2 ~ 6% for Bs decays + Proper time resolution ~ 40 fs + Signal/Background ~ 3 (from 107 inclusive bb events) + Effect of acceptance Oscillations still clearly seen!

Proper time (ps) Events

1000 800 600 400 200 1 2 3 4 5 Perfect reconstruction + flavour tagging + proper time resolution

Illustration of B physics sensitivity

• bservation of Bs–Bs oscillation
• Use mode Bs → Ds

−π+ ; 1 year of data (80k selected events)

for ∆ms = 20 ps-1 (SM preferred) + Dilution by flavour tagging: εD2 ~ 6% for Bs decays + Proper time resolution ~ 40 fs + Signal/Background ~ 3 (from 107 inclusive bb events) + Effect of acceptance Oscillations still clearly seen!

Proper time (ps) Events

1000 800 600 400 200 1 2 3 4 Perfect reconstruction + flavour tagging + proper time resolution + background 5

Illustration of B physics sensitivity

• bservation of the Bs–Bs oscillation
• Use mode Bs → Ds

−π+ ; 1 year of data (80k selected events)

for ∆ms = 20 ps-1 (SM preferred) + Dilution by flavour tagging: εD2 ~ 6% for Bs decays + Proper time resolution ~ 40 fs + Signal/Background ~ 3 (from 107 inclusive bb events) + Effect of acceptance Oscillations still clearly seen!

Proper time (ps) Events

1000 800 600 400 200 1 2 3 4 5 Perfect reconstruction + flavour tagging + proper time resolution + background + acceptance

Illustration of B physics sensitivity

the Bs–Bs oscillation frequency

Plot of the uncertainty σAon the fitted

• scillation amplitude vs ∆mS

LHCb could exclude the full SM range in one year! If observed, σstat(∆mS)~0.01 ps-1 could make a 5σ observation of ∆mS up to 40 ps-1 well beyond the SM, in 5 weeks!

]

− 1

[ps

s

m ∆ 40 60 80 100 120

Α

σ

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

= 1.0 ± 0.1

simulated

τ ∆ / τ ∆

≥ 5σ observation

• f Bs oscillations

for ∆ms < 68 ps–1 with 2 fb–1 LHCb

Conclusion

The LHCb detector combines all qualities required for the study

• f B-physics at the LHC: powerful trigger, precision vertexing

and tracking, excellent particle id and proper time resolution. The expected performance of LHCb, and the high b production rate at the LHC, should allow to extend the results of the B- factories, to overconstrain the unitarity triangle and to search for new physics in a complementary way to ATLAS and CMS. Construction and installation of all detector components is progressing well. LHCb is on schedule to be ready for global commissioning by end 2006 and…the first collisions in 2007!