Status of LHCb The pre upgrade years. Running scenario A few selected channels. The LHCb Upgrade The luminosity upgrade The LHC machine and higher luminosities. What is limiting LHCb to profit from larger L? Base-line upgrade scenario. Projected yields of Super LHCb Conclusions - 1 -
Muon MWPC RICH2 HPDs OT: straws 5 mm 4 Tm Magnet VELO Si IT DAQ HCAL & ECAL Be beam-pipe - 2 - Si Trigger Tracker RICH1: mirrors
LHCb commissioning in full swing. Expect to be ready for first collisions. - 3 -
LHCb pre upgrade years Till summer 2008 Beam-gas, first collisons? Commissioning/shake down detector 2nd half 2008 Expect (hope for) 0.1 fb 1 @14 TeV. Final shake down and Trigger commissioning. First look for NP compatible with low L 2009-2012 Stable running at L = 2 5.10 32 cm 2 s 1 Develop full physics program: L 10 fb 1. 2013-2020 Upgrade LHCb SuperLHCb Aim for L = 2.10 33 cm 2 s 1, L 100 fb 1. - 4 -
CP Violation Charge conjugation: particle anti-particle Parity: reflection in the origin of the space coordinates of a particle. CP violation distinguishes matter from anti-matter: baryogenesis. The CPV in the standard model too small to explain our existence: Leptogenesis? Since ν s have mass, they mix and they will probably have CPV. New physics new source(s) of CPV Studying CPV can reveal new physics, beyond the reach of direct searches. The observation of CPV in the K-system, led Kobayashi and Maskawa to the prediction of the third generation. Hence... But: no smoking gun... - 5 -
The Cabibbo-Kobayashi-Maskawa Matrix CKM-matrix describes the charged current interactions of quarks: couplings of W + -boson to up-down quark pairs: V = V ud V us V ub V cd V cs V cb. V td V ts V tb The matrix can be parametrized with four independent parameters, including one phase, which introduces CP violation. Wolfenstein parametrization: λ(= sinθ Cabibbo ), A, ρ and η parameters. 1 λ 2 /2 λ Aλ 3 (ρ i η) V λ 1 λ 2 /2 Aλ 2 + δv, Aλ 3 (1 ρ i η) Aλ 2 1 There are nine unitarity relations in the matrix: triangles. δv contains λ 4 terms - 6 -
Compare CPV with CP conserving Non-CPV CKM values from: V ud and V cs from nuclear β-decay and K πlν resp: fixes λ. V ub and V cb : from tree diagram B-decays at LEP/CLEO: fixes A. m d : B-factories. m s : measured by CDF. CPV CKM values from: ɛ K : CPV in K ππ. α, β, γ from B-decays. - 7 -
Triangle Status KM nature of CP-violation well established. Next step: LHCb flavour physics program - 8 -
LHCb Physics Program Indirect searches for NP in loop induced decays. BR(B s µµ) B mixing parameters. CPV in exclusive b sss hadronic penquin decays. CPV in B decay amplitudes Measurements of exclusive b sll and b sγ Determination of γ using B DK tree decays. Search for LFV in leptonic B-decays NP search in charm sector (mixing, CPV, rare decays) b-hadron spectroscopy, heavy quarkonia etc.. - 9 -
Expectations based on: Examples of LHCb measurements: full GEANT, spill-over, pile-up, LHCb PYTHIA tuning at 14 TeV (25 % larger multiplicity than CDF tuning @ 14 TeV), mean-l = 2 10 32 cm 2 s 1. Following typical channels: γ measurements, B s µµ, B s J/ψφ, B s φφ, B K µµ - 10 -
Present world average: δ(γ) 25 o Measuring γ - 11 -
Unitary Triangle and γ (a) 2006: Tree only, ///// NP (b) 2006: Loop, NP? (c) 2012: 10 fb 1, γ =? Expect σ(γ) 3 for 10 fb 1. σ(γ) from sides (UTFIT) 6 Lattice improvements could bring this down to σ(γ) sides 1. While increasing statistics, have to resort to theoretically cleanest decays. Upgrade aim: increase statistics with factor 20-12 -
BR(B s µµ) Rare loop decay, sensitive to NP. SM: BR=(3.55±0.33)10 9. Example CMSSM (hep-ph/0411216), can be strongly enhanced at large tan(β) and interesting gaugino masses. CDF: 47 10 9 @90% CL. D0: 75 10 9 @90% CL. BR(Bs µµ) - 13 -
BR(B s µµ) Challenge: background rejection: B s mass-resolution: 18 MeV Excellent vertex resolution isolation. Background dominated by double B µx decays. Trigger: p µ T 1 GeV 0.05 fb 1 to overtake Fermilab, 6 fb 1 to have 5σ signal of SM BR. - 14 -
B s J/ψ(µµ)φ Extract φ s from golden mode: J/ψ(µµ)φ. φ s strange counterpart of φ d = 2β φ SM s = 0.0368 ± 0.0018 NP (?) in box could increase φ s. 65k J/ψ(µµ)φ/fb 1 σ(φ s ) = 0.044 with 0.5 fb 1. D0 (1.1 fb 1 ): φ s = 0.79 ± 0.56 + 0.14 0.01 NP amplitude parametrized with h s and σ s LHCb with 10 fb 1 : obtain 3σ evidence of CPV if SM value. - 15 -
B s φφ Golden hadronic penguin counterpart: φφ V ts cancels in mixing-decay, hence: CPV-φ 0 NP! 2k φφ/fb 1 (with BR=1.4 10 5 ) σ(φ NP ) = 0.042 with 10 fb 1 Compare: 0.4k B 0 φk S /fb 1 σ(sin(2β)) = 0.14 with 10 fb 1 Expect ±0.12 with 2 ab 1 from B-factories. - 16 -
B K µµ Loop decay with BR 1.2 10 6, i.e. sensitive to NP. 3.8k events/fb 1. Estimated B/S=0.4±0.1 σ(a FB (s) = 0) : 0.28 GeV 2 for 10 fb 1. Measure ratio of Wilson coefficients C 7 /C 9 with 7% statistical error. - 17 -
B K µµ Other asymmetries: Describe decay as a function of 4 parameters: M 2 µµ, θ l, θ K, φ Transversity amplitude A (2) T : (d) 2 fb 1, 1 σ s=(m µµ ) 2 [GeV 2 ](e) 10 fb 1, 90 % CL Sensitive to MSSM with tan(β)=5 (hep-ph/0612166), hence complementary to B s µµ. - 18 -
LHC and Luminosity LHC: we (will..) have the machine! L peak LHC 1034 cm 2 s 1, 201n Assume σ visible = 63 mb. @2.10 32 : 10 MHz xings with 1 int. @10 33 : 26 MHz xings with 1 int. nr-int/xings: only factor 2 increase up to @10 33, but spill-over goes linear with L!. SLHC: LHCb does not need it but... L peak SLHC 8.1034 cm 2 s 1, 201n + 4 Baseline scheme 25 ns bunches, but: 50 ns I high I high bunches Interleave with 50 ns I low I low bunches Atlas/CMS xings: I h I h, I l I l, I h I h LHCb xing: I h I l, I l I h, I h I l etc.. - 19 -
B-rates and Int/xing Assume σ visible = 63 mb. B-rate: @10 33 take all xings with 5 int! @2.10 33 take all xings with 10 int! Note: σ z (interactions) 50 mm. Typical B-decay length 10 mm. - 20 -
Ramping up Peak Luminosity Simulated luminosities (including pile-up, spill-over) up to 2.10 33 cm 2 s 1, i.e. 10 design peak luminosity. Trigger: first sub-system which does not scale: Brief overview of the trigger Explanation why it fails Base-line upgrade proposal Tracking Radiation damage - 21 -
Level-0: Largest E T hadron, e(γ) and µ. Bottleneck: 1 MHz max-output rate. L0 limiting yield for larger L peak : High Level Trigger: LHCb Trigger for pedestrians Access to all detector info from day 1. Limitation: CPU (brain?) power. Will improve with Moore s law automatically : plan to replace CPU boxes every 3 years anyway. M5 M4 M3 M2 Pile Up System # interactions per crossing HCAL Calorimeter Triggers PS SPD ECAL Highest E T clusters: hadron, e, γ, π 0 SPD multiplicity M1 RICH2 T3 T2 T1 Magnet RICH1 TT VELO All DATA HLT Event Filter Farm VELO Pile Up Muon Trigger Two highest p muons T L0 Decision Unit defines L0 trigger To FE 40 MHz - 22 - Readout Supervisor timing & fast control Level 0 1 MHz HLT 2 khz Storage
LHCb Trigger Performance (@L = 2.10 32 ) ɛ Trigger : % off-line reconstructed with good B/S: E hadron T 3.5 GeV: ɛ(b hadronic) 25 35% E γ T 2.5 GeV: ɛ(b γx) 30 40% E µ T 1. GeV: ɛ(b µµx) 60 70% Hence: increase lumi AND improve ɛ trigger! Typical trigger storage rates: 200 Hz exclusive B: core program. 600 Hz M µµ > 2.5 GeV: IP-unbiased B J/ψX 300 Hz D : Charm mixing&cpv. For un-triggerable channels: 900 Hz inclusive B (B µx): Data mining. 550 Hz of true B µx B µx trigger: tagging ɛd 2 0.15 1.5 10 9 fully contained, µ-tagged, /2 fb 1. - 23 -
Trigger and Luminosity Trigger: L peak 2 3.10 32 no hadron-trigger gain. hadronic-channels: yield (time 2.10 32 ) µ-channels: yield L Why? @L > 2.10 32 : L0-retention 10% @L > 10 33 : L0-retention 3% @L = 2.10 33 4 pp/xing. Result: E T threshold M B. - 24 -
Need both p T and Impact Parameter (IP). Upgrade Trigger Proposal Take larger E T seeds, i.e. µ, e-, γ- and hadron-clusters Assume they originate at (0,0,0), and extrapolate to T1-T3. Reco tracks in window around extrapolated candidate: precise p T Connect to VELO tracks: gives IP. - 25 -
Trigger Upgrade number of seeds small, hence: Perform whole trigger on CPU farm. Read-out all sub-systems at 40 MHz. Preliminary studies: Event building at 40 MHz, CPU power. Hadron trigger efficiency: L peak, and ɛ hadron 2. Replace all FE-electronics replace VELO/SiT and RICH-HPD, OT-FE, Cal-FE-boards. - 26 -
Tracking and Luminosity Tracking & Particle-ID: Note: standard algorithms. VELO tracking not a problem. Straws: L peak 10 33 spillover is a problem. Address occupancy, especially at large η: R&D... OT straws-occ(no-spillover): 6(4.5) 25(10)% Hence: faster gas, increase Si-IT coverage, scintillating fibres? Tracking environment/algorithms in high(er) occupancy environment. VELO Tracking [%] Tracking efficiency [%] Full spill-over Limited spill-over Ghost rate: full spill-over Ghost rate: limited spill-over Luminosity [ 10 32 ] - 27 -
Radiation: Integrated L: Radiation designed for L 20 fb 1 radiation damage. Safety factor? Need first running. Affects mainly large η. Address radiation L 20 100 fb 1. Need to replace VELO (anyway): rad-harder Si, pixels? Inner part of Shashlik Calorimeter: crystals? Inner part of Si-trackers. Remove µ-chamber before Calorimeter. - 28 -
Projected Yield vs Time Good B/S [k-events] LHCb L [fb 1 ] B K µµ B s φφ B φk S Year in 21st century KEKB [ab 1 ] LHCb< 2013: L 15fb 1 L = 5.10 32 2010 Yield h : constant> 2.10 32 Upgrade: Assume ɛ trigger µ unchanged Assume ɛ trigger h 30 60% L = 2.10 33 2015 Include tagging efficiency: LHCb ɛd 2 = 0.07 Belle ɛd 2 = 0.3-29 -
Conclusions LHCb is (almost) ready to take over from first generation B-factories. Accumulate 10 fb 1 between 2008-2012. 1st LHCb Collaboration Upgrade Workshop Jan/2007 in Edinburgh: well attended. LHCb Upgrade: draft circulating. EOI... Ramp up upgrade effort to be able to take decision 2010. Upgrade specs: Accumulate 100 fb 1 by 2020 Increase ɛ hadron trigger by factor 2. Maintain tracking and PID performance. Start data-taking with SuperLHCb 2014. - 30 -