Olivier Callot On behalf of the LHCb collaboration

Olivier Callot On behalf of the LHCb collaboration 12 November 2012 LHCb detector status and upgrade n General overview n Status of the various components n Global performance n Upgrade guidelines HCP 2012, Kyoto , Japan 12 -16 November 2012

Detector layout u. LHCb is studying beauty (and charm) n At LHC, the production is peaked forward/backward n The detector is a single arm spectrometer l Both b go together forward (or backward) l Acceptance 2 < η < 5 n The b meson / baryon is boosted l It flies a several millimetres before decaying l This is the signature for selecting events u. General detector layout n The vertex detector is a key component n Dipole magnet, and tracking stations after, to measure accurately the momentum n Particle identification by 2 RICH detectors, electromagnetic and hadronic calorimeters, and a muon system Olivier Callot HCP 2012 : LHCb detector status and upgrade 12 November 2012 2

Collision point Y X Z Olivier Callot HCP 2012 : LHCb detector status and upgrade 12 November 2012 3

Velo u 84 silicon micro strip sensors n 44 mm radius n R or Phi geometry u. Open and closes for each fill n Centred l It around the current beam position doesn’t move during a fill n Mechanically reproducible to ~ 5 µm Average distance between right-side PV and left-side PV ± 5 µm Olivier Callot HCP 2012 : LHCb detector status and upgrade 12 November 2012 4

u. Impact n Very parameter resolution is essential good resolution l ~20 µm at high Pt èThis means over 2 Ge. V… n Slightly better in MC l Work in progress to understand it l Material description suspected n Primary l ~16 vertex resolution excellent µm in x, y, ~76 µm in z (20 tracks) u. Radiation damage n Starts to be visible in dark current n It behaves as expected n Doesn’t affect data quality. Olivier Callot HCP 2012 : LHCb detector status and upgrade 12 November 2012 5

Tracking system u. One n Si station (TT) before the magnet strips, 4 layers u. Dipole n ∫Bdl u 3 magnet = ~ 4 Tm ; polarity switched regularly stations after the magnet n Si strips in centre (IT), straw tubes outside (OT) n 4 layers per station x-u-v-x, (5 degree stereo angle) u. Tracking n For Olivier Callot efficiency over 96% tracks traversing the whole detector, over 5 Ge. V momentum HCP 2012 : LHCb detector status and upgrade 12 November 2012 6

RICH u. Particle identification from 2 to ~100 Ge. V n 2 RICH, 3 radiators n Readout by HPD l ~500, 1024 channels l High efficiency l Very low noise l 10 -20 replaced each year u. Particle n ~95% ID performance efficiency for 5% contamination l Averaged n Well Olivier Callot over B daughter tracks described by simulation HCP 2012 : LHCb detector status and upgrade 12 November 2012 7

Calorimeters u. ECAL made of shashlik blocks n Lead - scintillator stack n ~6000 channels, readout by PMT n ~10% / √E + 1% u. HCAL: scintillating tiles in iron n ~1500 channels, same readout and electronics as ECAL n ~70% / √E + 9% n Mainly used for trigger u. Pre. Shower and SPD n same geometry as ECAL n Scintillator tiles readout by MAPMT n Identify electron/photon, used in L 0 trigger Olivier Callot HCP 2012 : LHCb detector status and upgrade 12 November 2012 8

Muon system u 5 stations interleaved with iron walls n First station before the calorimeters n Projective geometry l Allows n Muon l Olivier Callot it to be used in the L 0 trigger identification performance ~97% efficiency for 3% miss-ID. HCP 2012 : LHCb detector status and upgrade 12 November 2012 9

Trigger u. L 0 hardware trigger, custom electronics n High Pt local cluster in HCAL (3. 6 Ge. V) or ECAL (2. 6 Ge. V) n High Pt muon (1. 4 Ge. V) or di-muon n Accept rate limited to 1 MHz, latency < 4 µs u. HLT software trigger, 30, 000 copies on 1500 nodes n HLT 1 mainly a topological trigger: l At least one track with Pt >1. 6 Ge. V and impact parameter >100 µm l Accepts around 50 k. Hz n HLT 2 selects by physics channel, inclusive or exclusive l Full track reconstruction but no particle-id. n 25% of the input events are deferred, i. e. stored on disk and processed during inter-fills ! n Total accept rate around 5 k. Hz. Olivier Callot HCP 2012 : LHCb detector status and upgrade 12 November 2012 10

Luminosity u. LHCb was designed for single collision crossings n Worries l We even intended to reject multiple interaction at level 0 trigger… n Design u. We of ambiguities to assign the B to the proper primary vertex luminosity 2 10 32 cm-2 s-1 for ~2700 colliding bunches found that higher collision rate was OK n We now run at 4 1032 cm-2 s-1 with only 1262 colliding bunches l 50 ns separation between bunches, 25 ns nominal and from 2015. l This means 4 times more collisions per crossing than in the design l The average number of visible collisions per crossing is ~1. 8 èHigher occupancy in the detector èMore challenging for the trigger, see later n The luminosity is kept constant thanks to the ‘luminosity levelling’ l The beam separation is adjusted a few times per hour to maintain the luminosity constant. l Routine operation since 2011 Olivier Callot HCP 2012 : LHCb detector status and upgrade 12 November 2012 11

Olivier Callot HCP 2012 : LHCb detector status and upgrade 12 November 2012 12

Luminosity is frequently adjusted (± 3% around target value) Olivier Callot HCP 2012 : LHCb detector status and upgrade 12 November 2012 13

Global performance u. Data taking is very efficient n Minor losses due to HV not OK, or Velo not in working position n 2. 4% dead time, due to some hardware devices n “DAQ” is in fact a lot of short stops to give a “kick” to some systems Olivier Callot HCP 2012 : LHCb detector status and upgrade 12 November 2012 14

Reconstruction performance u. Momentum resolution excellent n From 0. 4% at 5 Ge. V to 0. 6 % at 100 Ge. V/c n Important ingredient to study narrow states n Momentum scale OK, checked by comparing the mass of B mesons to the PDG values. J/Ψ → μμ : σ ≈ 15 Me. V/c² Olivier Callot Υ(1 S) → μμ : σ ≈ 54 Me. V/c² HCP 2012 : LHCb detector status and upgrade BS→J/ΨΦ : σ ≈ 8 Me. V/c² 12 November 2012 15

u. PID is working nicely n Important to disentangle various decays B 0 → Kπ B 0 → ππ Λb → pπ Olivier Callot Λb → p. K HCP 2012 : LHCb detector status and upgrade BS → KK 12 November 2012 16

Integrated luminosity u. Well Olivier Callot on track for over 2 pb-1 this year! HCP 2012 : LHCb detector status and upgrade 12 November 2012 17

The LHCb upgrade u. The (installation in 2018) current main limitation is the L 0 trigger n Rate limited to 1 MHz (in practice ~950 k. Hz) due to the readout time of the front-end l 36 n The clock cycles -> 900 ns main hadronic trigger is based on HCAL clusters l Threshold is high, to decrease the rate to ~700 k. Hz l About half the interesting B decays are lost. u. Basic upgrade strategy: Remove the L 0 trigger n Readout the events at 40 MHz n Implement a fast software trigger to select events based on their topology l Gain a factor 2 in efficiency on hadronic channels l Improve the detectors to provide a fast reconstruction Olivier Callot HCP 2012 : LHCb detector status and upgrade 12 November 2012 18

Consequence u. All front-end electronics has to be redone n Change in output rate requirement l No more pipe-line or buffers locally l A lot more optical links, of a newer generation u. Rebuild chips the detectors with embedded front-end n Silicon strip detectors, i. e. Velo, TT, IT n RICH photo-detectors (front-end chip inside the HPD) u. Target luminosity n Similar value of pile up (µ=2) but twice the number of bunches n Initial target : 10 1032 cm-2 s-1 , i. e. 2. 5 times the current value l Factor ~10 in the yield of hadronic channels (14 / 7 Te. V helps) 32 -2 s-1 and prepare the detectors for it. Foresee to run at 20 10 HCP 2012 : LHCb detectorcm status and upgrade 12 November 2012 n. Callot Olivier 19

New detectors u. Two options for the Velo n Upgraded micro strip detector l Similar overall geometry, 20% more channels per sensor, smaller pitch and thickness n Pixel detector l 53 x 53 µm 2 cells l ~780 k channels per module l 50 modules è 40 millions pixels… Olivier Callot HCP 2012 : LHCb detector status and upgrade 12 November 2012 20

Proposed layout of the Pixel detector Olivier Callot HCP 2012 : LHCb detector status and upgrade 12 November 2012 21

Tracking stations u. TT similar but improved n Better coverage by overlapping sensors and getting closer to the beam pipe n Less material with thinner sensors Olivier Callot HCP 2012 : LHCb detector status and upgrade 12 November 2012 22

Main tracking stations (after magnet) u. Two options for the T stations n Similar to the current detector, with an increased Inner Tracker l Outer Tracker straw tubes are sensitive to spill-over from the previous and next crossings, as the drift time is around 40 ns. èThis will be the case with 25 ns crossing period. l Increasing IT will decrease the occupancy in OT by a factor~2, to reach similar levels as the current ones. l IT made thinner and lighter è 4 times more channels Olivier Callot HCP 2012 : LHCb detector status and upgrade 12 November 2012 23

u. Scintillating n 2. 5 fibres m long fibres, 5 layers of 0. 25 mm diameter n Readout by Si. PM with readout cells 0. 25 x 1. 32 mm n Similar geometry as now (reuse of supports) l 3 stations, 4 layers each l But less material in acceptance n R&D ongoing to validate this option l Radiation hardness of Si. PM èNeed dedicated shielding, low temperature… èNoise reduction by cluster shape analysis in front end l Accuracy of the fibre mechanics èNeed to keep the fibre straight to ~50 µm and flat to ~200 µm over 2. 5 m! Olivier Callot HCP 2012 : LHCb detector status and upgrade 12 November 2012 24

RICH u. Replace the HPD by MAPMT u. Remove the aerogel radiator n Limited use in a very dense environment n Decrease dead material. u. Studies on-going to improve the optics n Reduce the number of MAPMT (cost…) n Better resolution n Less magnetic field to be shielded where the MAPMT will be u. More fancy proposal also investigated n TORCH detector (time of flight with quartz bars to ~15 ps, for low momentum particles) for ~2020 Olivier Callot HCP 2012 : LHCb detector status and upgrade 12 November 2012 25

Calorimeter and muon u. No major change needed n The electronics of the calorimeter should be remade, but will stay very similar, only the output stage will change n M 1 will be removed as we don’t need a L 0 Muon any longer n The muon electronics will stay as is, as for the L 0 Muon the readout works already at 40 MHz. u. Possible n The improvements central part of M 2 may become GEM instead of MWPC l Handles n Electronics l Dead Olivier Callot more rate may be re-implemented to be safe at 20 1032 cm-2 s-1 time due to long signals HCP 2012 : LHCb detector status and upgrade 12 November 2012 26

Trigger u. Suppress the L 0 n In fact one can keep it at the beginning as a “Low Level Trigger” to minimize the load on the CPU farm l But no longer limited to 1 MHz u. Implement a very fast software trigger n 30 MHz input implies that every millisecond requires 30, 000 copies of the code running in parallel n Tune the detector (and readout) to be fast l Pixel seems better than strips for the Velo l Fibre Tracker seems better than IT + OT n Investigate l GPU, n We new trends in computing large number of core, … need a factor 30 compared to the current system! l We expect ~10 from the computing industry… l We need to find a factor 3 in our code ! Olivier Callot HCP 2012 : LHCb detector status and upgrade 12 November 2012 27

Summary u. LHCb is performing very well in difficult conditions n Four times more luminosity per crossing than in the design n All detectors over 99% operational n Resolutions in position and momentum conform to expectations u. High statistics available, a lot of results expected n Many already published, see Burkhard’s talk just after this. l And a special talk by Johannes after l See also parallel sessions : Tom (SUSY), Andrea (QCD), Kurt (EW) , Johannes, Mika, Stefania and Angelo (Heavy Flavour) n Expect l And u. The than 2012 will give more than twice 2011 statistics… we will continue to take data at the same rate from 2015 upgrade program is progressing n Get up to 20 times more hadronic events per second ! n Challenging, but realistic Olivier Callot HCP 2012 : LHCb detector status and upgrade 12 November 2012 28

Thanks you for your attention … LHCb detector And collaboration Olivier Callot HCP 2012 : LHCb detector status and upgrade 12 November 2012 29

… and for the occasion to visit Japan Saturday morning at Yamanakako lake Olivier Callot HCP 2012 : LHCb detector status and upgrade 12 November 2012 30