B-physics trigger for the ATLAS detector at LHC: recent developments ⋆ B. Epp, V.M. Ghete and D. Kuhn Institute for Experimental Physics, Innsbruck ur Bildung, Wissenschaft und Kultur, ¨ ⋆ Arbeit unterst¨ utzt vom Bundesministerium f¨ Osterreich.
Introduction ATLAS classical B-physics menu B 0 d → J/ψK 0 with J/ψ → e + e − CP-Violation studies: S J/ψ → µ + µ − B 0 d → ππ B 0 s → J/ψφ B 0 B 0 s → D s π , B 0 s oscillation studies: s → D s a 1 with D s → φπ B d,s → µ + µ − X ′ , K ∗ 0 , ρ 0 Rare decays: with X = ‘ . . . but B-physics is and will remain for the next five years a very dynamic domain, both theoretically and experimentally (BaBar, Belle, CDF, D0, . . . ), therefore new channels may become interesting ( B c , B 0 d → K ∗ γ , . . . ). B-physics general programme: to be pursued in the ‘low luminosity’ run of the LHC (latest luminosity target: 2 × 10 33 cm 2 s − 1 ). Rare decays: both low and high luminosity ( 10 34 cm 2 s − 1 ) runs. ¨ ATLAS B-Physics Trigger OPG FAKT03, Oct. 4, 2003 V . M . Ghete
Experiment overview Inner detector: • discrete semiconductor pixel and strip detectors • continuous straw-tube tracking detectors with transition radiation • inside the solenoid: 2 T magnetic field. Calorimetry: • highly granular LAr EM calorimeter: | η | < 3 . 2 • hadron calorimeter (scintil- lator-tile): | η | < 4 . 9 Muon spectrometer: • air-core toroid system on average ∼ 0 . 5 T. ¨ ATLAS B-Physics Trigger OPG FAKT03, Oct. 4, 2003 V . M . Ghete
ATLAS trigger architecture ATLAS-TDR-016; CERN-LHCC-2003-022 (May 2003) ¨ ATLAS B-Physics Trigger OPG FAKT03, Oct. 4, 2003 V . M . Ghete
B-physics trigger in the DAQ/HLT Technical Proposal Start-up luminosity target: 10 33 cm − 2 s − 1 Level-1 Trigger: • single muon, with p T > 6 GeV. Rate (Hz) Level-2 Trigger: HLT selection Selected Event B - channels (in addition to muon p T > 6GeV) LVL2 Filter • µ confirmation within LVL1 RoI in B s → D s π 3 hadrons p T > 1.5 GeV, 270 41 D s → φ (K + K - ) π invariant mass cuts for φ and D s EF: invariant mass, vertex fit quality, trans- the Inner Detector and precision Hadron verse decay length and angle cuts. channels B d → ππ 2 hadrons p T > 4 GeV, 80 5 muon chambers B d → Κπ angle, ∑ p T and loose invariant mass cuts EF: invariant mass, vertex fit quality, trans- verse decay length and angle cuts. • track reconstruction in the whole bb → µ B d (J / ψ ( ee ) K 0 ) 2 tracks p T > 0.5 GeV, TRT ee identification, 660 32 ∆η , ∑ p T , ∆ Z and invariant mass cuts Electron ID (‘full scan’) channel EF: invariant mass, vertex fit quality, trans- verse decay length and angle cuts. bb → e 90 single e , p T > 5 GeV, identification in • track kinematic cuts and channel- B d (J/ ψ ( µµ )K 0 ) TRT+ECAL of electron reconstructed with Not p T > 4 GeV in the Inner Detector covered To be in this studied specific mass cuts B d → J/ ψ ( µµ )(K/K * ), second µ 170 note B s → J/ ψ ( µµ ) φ , (p T > 5 GeV, |η| < 2.5) B → µ µ , identification in muon chambers + matching Event Filter: B → K 0* µµ , etc., with the Inner Detector Λ b → Λ 0 J/ ψ ( µµ ), B c → J/ ψ ( µµ ) π • channel-specific selections done Total B - physics trigger rate 1270 ~100 using set of loose offline cuts. ATL-DAQ-2000-031 (Jun 2000) ¨ ATLAS B-Physics Trigger OPG FAKT03, Oct. 4, 2003 V . M . Ghete
From beauty to reality: after HLT Technical Proposal Changes in detector geometry: • increased beam-pipe diameter (41.5 mm → 50.5 mm) • increased pixel length in B-layer (300 µ m → 400 µ m) Financial constraints: B-physics trigger resources have to be minimized. Financial uncertainties: some items to be deferred → reduced detector at start-up • only 2 of the 3 pixel layers (inner B-layer maintained), TRT only at η < 2 • reduced HLT system → reduced computing resources at LVL2 and EF. Priority given to high- p T physics, B-physics hadronic triggers first to be affected. LHC parameters: luminosity target for start-up doubled to 2 × 10 33 cm − 2 s − 1 Alternatives to reduce the resource requirements: • require at LVL1, in addition to single µ trigger, a second muon, a JET or an EM RoI, then reconstruct at LVL2 and EF within RoI • re-analyze thresholds and use a flexible trigger strategy, depending on luminosity. ¨ ATLAS B-Physics Trigger OPG FAKT03, Oct. 4, 2003 V . M . Ghete
B-physics trigger strategy in HLT TDR Strategy adapted to limited bandwidth: • Start with a di-muon trigger for higher luminosities LHC fills. • Add further triggers (hadronic final states, final states with electrons and muons): in the beam coast for the low luminosity fills. = ⇒ always fill the available bandwidth in the HLT system. Trigger types: • di-muon trigger: two muons at LVL1 • hadronic final states triggers: single muon at LVL1, followed by – RoI reconstruction in ID at LVL2, from Jet RoI trigger at LVL1 – full-scan in ID at LVL2 • triggers for final states with electrons and muons: single muon at LVL1 – RoI reconstruction in TRT at LVL2, from EM RoI trigger at LVL1 – full-scan in TRT at LVL2 ¨ ATLAS B-Physics Trigger OPG FAKT03, Oct. 4, 2003 V . M . Ghete
Di-muon triggers Examples of final states with two muons: • J/ψ → µµ : B 0 d → J/ψK S , B 0 s → J/ψφ , B 0 s → J/ψη , Λ b → Λ 0 J/ψ • B 0 s, d → µµ , B 0 d → ( K ∗ 0 , ρ 0 , φ ) µµ LVL1: at least two muons. Minimum thresholds: p T � 5 GeV Muon Barrel p T � 3 GeV Muon End-Cap • actual thresholds: ⇐ = LVL1 rate. • mainly due to muons from heavy flavour decays, plus single muons double counted in end-cap trigger chambers. LVL2 and EF: confirmation of muons us- ing ID and precision muon chambers. Spe- cific selection at EF: mass and decay length Luminosity: 1 × 1033 . cuts, after vertex reconstruction. Di-muon rates: single muon all + second muon. ¨ ATLAS B-Physics Trigger OPG FAKT03, Oct. 4, 2003 V . M . Ghete
Hadronic final states Channels studied so far: B 0 d → ππ and B 0 s → D s π , B 0 s → D s a 1 . LVL1 trigger: single muon, threshold determined by trigger rate and luminosity. Threshold values considered: p T > 6 ÷ 8 GeV. LVL2 trigger: confirms muon using ID and precision muon chambers; reconstruct tracks in ID, select D s → φπ or B 0 s → ππ based on mass cuts. Options for ID track reconstruction: • require a low E T LVL1 Jet RoI, in addition to a single muon trigger, and reconstruct the tracks at LVL2 in the RoI only. Advantages: modest resources required. Disadvantages: lower efficiency expected. Problem: low E T LVL1 Jet difficult to trigger. • track reconstruction within entire SCT, Pixel and (optionally) TRT detectors; ‘full-scan’: better efficiency, but greater resources. EF: refit ID tracks in LVL2 RoI, select D s → φπ or B 0 d → ππ based on mass cuts and vertex cuts. ¨ ATLAS B-Physics Trigger OPG FAKT03, Oct. 4, 2003 V . M . Ghete
Hadronic final states: simulation studies Two sets of studies: • Fast simulation of calorimeter trigger based on ATLFAST + parameterised calorimeter simulation. Rather complex: – B-field, longitudinal and transverse shower profiles – pulse history, digitization and Bunch Cross Identification system (BCID) – complete LVL1 trigger algorithms • A full detector simulation with an incomplete simulation of the LVL1 calorimeter – e / γ / τ algorithm: tested, debugged, validated. – Jet algorithm: tested, debugged, but not yet fully validated. – E miss , E sum T : coded, not yet debugged/validated. T – noise added for both calorimeters and for trigger towers – calibration (ECal) slightly updated. – towers still built from CaloCells, version using LAr/Tile tower simulation in work = ⇒ no Bunch Cross Identification system yet. ¨ ATLAS B-Physics Trigger OPG FAKT03, Oct. 4, 2003 V . M . Ghete
Hadronic final states: simulation studies Procedure: • use a sliding window of size | ∆ η | × | ∆ φ | = 0 . 8 × 0 . 8 in the LVL1 calorimeter trigger to find a jet associated to the second B-hadron. • determine the Jet RoI multiplicity • match the RoI found with the B-hadrons and determine the Jet RoI efficiency. Samples: • b ¯ b → µ ( p T > 6 GeV) X events • signal events: B 0 s → D s π and B 0 d → ππ Sample production: all events simulated with a LVL1 µ 6 trigger. • Generator: Pythia 6.2. • Simulation: initial-detector geometry. • Reconstruction: Athena reconstruction (trigger tower noise on, Tile noise on, LAr complete RDO and cell production with noise on/off). ¨ ATLAS B-Physics Trigger OPG FAKT03, Oct. 4, 2003 V . M . Ghete
Hadronic FS, full simulation: number of LVL1 Jet ROIs Sample: b ¯ b → µ 6 X events, with tower noise on, Tile noise on, and LAr noise on (left plot) off (right plot). 100 200 E T > 6 GeV Mean: 5.91 E T > 6 GeV Mean: 1.93 RMS: 2.30 RMS: 1.54 80 150 60 100 40 50 20 0 0 0 5 10 15 20 0 5 10 15 20 Jet ROI multiplicity Jet ROI multiplicity • increased Jet RoI multiplicity: mainly due to LAr noise. ¨ ATLAS B-Physics Trigger OPG FAKT03, Oct. 4, 2003 V . M . Ghete
Recommend
More recommend
