1. Francesco Devoto Study on Performance of First Level Muon Trigger of ATLAS experiment at LHC

2. Summary ATLAS Trigger System Conclusions First Level Muon Trigger  Selection Algorithm  Performance with Simulated Data  First Test with Cosmic Rays Data LHC and ATLAS Experimental Apparatus

3. Ring Radius Circumference 4.2 km ~27 km Center of Mass Energy14 TeV Luminosity10 34 cm -2 s -1 Bunch Crossing 40 MHz (t bx =25 ns) n° proton for bunch~10 11 At LHC the protons will collide in 4 points where the following 4 experimental apparatus were installed: ATLAS (A Toroidal LHC ApparatuS) CMS (Compact Muon Solenoid) ALICE (A Large Ion Collider Experiment) LHCb (Large Hadron Collider bPhysics) The beginning of data taking is scheduled for 2008 Proton-Proton Collider Large Hadron Collider

4. 46 m Solenoidal Magnetic Field + Vertex Detectors: PIXEL SCT TRT CALORIMETER: Toroidal Magnet MUON SPECTROMETR : Precision Chambers (MDT and CSC) Trigger Chambers (RPC in the barrel and TGC in the end-cap) Z (  Y X 22 mATLAS Experimental Apparatus Pseudorapidity Electromagnetic Hadronic

5.  B Physics  Function period with low luminosity ( L = 10 33 cm -2 s -1 ) Physics Programm  LHC is a factory all of Standard Model particles as well as of potenzial new particles with mass in the TeV range  Reserch of Higgs boson  Sperimental Limit (LEP): m H >113.5 Gev/c 2  LHC will be able to obseve a Higgs boson with mass up to 1TeV and it will measure its mass and pairing with great precision  Precision Measurements  W and t high production  Reserch of SuperSimmetric Particles (SUSY)

6. Trigger/DAQ Problem 1000 TB s -1 TRIGGER System Initial Rate R = L ×  p-p ≈ 1GHz DAQ ? Data Trasmission Storage

7. ATLAS Trigger System Level-1 Trigger The trigger system must be able to make a selection of the events that have some characteristics of interest Level 1 Level 2 Event Filter 1GHz 100kHz ~100Hz Muon Detector Muon Trigger Software Trigger High Level Triggers Hardware Trigger Calorimeter Trigger Central Trigger Processor Front-End LVL2 Trigger Region of Interest ~2.5  s ~10 ms ~1s

8. ATLAS Trigger System: Muon Spectrometer Barrel RPC Detector EndCap TGC Detector Barrel: |  | < 1.05 EndCap: 1.05 <|  | < 2.7

9. RPC  View  View Chamber Cover Barrel Barrel Muon Spectrometer 

10. The interest physical processes happen with high p T A p T selection is necessary to eliminate the processes that are not of interest LVL1-µ: Threshold Selection Eff pTpT 1 Fluctuation energy lost in the calorimeter Uncertainty on the position of the interaction vertex Scattering Magnetic field disuniformity p T threshold

11. Muon pT = pT Threshold Same algorithm both for low and high-pT 3 Low-pT + 3 High-pT threshold avalaible! Side  p-p Algorithm: Identification of candidate muon track coming from interaction vertex within p T range Infinite momentum track (p T  ) d+ e d- depends on the muon p T d+ d- PIVOT Plane RPC 2 Low-pT Plane RPC 1 High-pT Plane RPC 3 Coincidence Windows Calculation: Associate to each pivot strip a COINCIDENCE WINDOW in the Low-pT and High-pT system LVL1-µ: Trigger Algorithm Pivot Strip Low-pT Strip

12. Low-pT Coincidence Matrix LVL1-µ: Coincidence Matrix (CM) 6 GeV 8 GeV 10 GeV 3 slots for Low-pT Threshold avalaible: Default 6, 8, 10 GeV/c 3 slots for High-pT: Default 11, 20, 40 GeV/c Thr= 6 GeV Thr= 8 GeV Thr= 10 GeV 6 GeV+ 8 GeV+10 GeV

13. LVL1-µ: Hardware and Simulation RPC segnals are sent to CMA Synchronize input signals Make coincidence Give trigger results for  and  projection LVL1 ATHENA Analyse Data Simulated Single  The package that simulates the trigger is integrated in ATHENA the ATLAS official framework 4 CM create a PAD PAD generate the information of the RoI RoI Spatial Extension = ∆  × ∆  ≈ 0.1 × 0.1

14. LVL1-µ Performance: Efficiency Curves Low-pT Efficiency Plateau Region Plateau Efficiency ~82%

15. High-pT Efficiency Plateau RegionPlateau Efficiency ~78% LVL1-µ Performance: Efficiency Curves

16. LVL1-µ Performance: Geometrical Inefficiency Map Magnetic Ribs Apparatus Support Structures Lift  Crack    = 3  =  rad

17. Muon Selection Asymmetry Spectrometer layout is not exactly symmetrical respect to this plane then selection asymmetries are possible Muons are deflected in the r–  plane under the action of magnetic field Muons track are symmetrical respect to z=0 plane There are significant asymmetries under threshold while above it is small Events under threshold will be processed and rejected by HLTs P T = 10 GeV/c  

18. First Level Trigger Special Configuration: First Collision p-p LHC first collision will be with a luminosity of L= 10 31 cm -2 s -1 The initial program is for the tweak of machines and calibration of deterctors but also the first data on b-physics Low-p T threshold are MU4 (fully opened), MU5 (5GeV/c) e MU6 (6GeV/c) For High-p T standard threshold HLTs there are in DAQ but the algorithm disabled Fully Opened 5 GeV/c 6 GeV/c Fully Opened 5 GeV/c 6 GeV/c

19. Coincidence windows are completely opened! The maximum width is given by the Matrix hardware processor size (64x32 channels) Cosmic Trigger: Special Lvl1 algorithm configuration is used to maximize Trigger efficiencies with cosmic events! Cosmic Trigger: Special Lvl1 algorithm configuration is used to maximize Trigger efficiencies with cosmic events! First Level Trigger Special Configuration: Cosmic Ray

20. Muons hitting RPC System Muons selected by LVL1 #of muons selected by LVL1 Efficiency: LVL1 cosmic = 0.50.1% LVL1 standard = 0.10.1% MU4 Rate Expected MU4 Rate Expected ~600Hz LVL1 cosmic algorithm is applied on a semple of 300k events, P>1 GeV Coincidence matrix fully opened Standard Threshold First Level Trigger Special Configuration: Cosmic Ray

21. Cosmic events surface distribution triggerated by RPC Z (cm) X (cm) 18m 12 m Distribution obtened with sperimental data from sector 13 Simulated distribution First Level Trigger Special Configuration: Cosmic Ray Tracks muons reconstruction software was developed by Neaples group

22. LVL-1 Algorithm in M5 (November 2007) Same algorithm both projection (η e φ) Request pointing to vertex Sperimental value consistent with MC estimation First Level Trigger Special Configuration: Cosmic Ray

23. Using the Level-1 Efficiency curves we may estimate the rates with different threshold Level-1 Efficiency Inclusive μ Cross Section at LHC MU6 Barrel Trigger ~ 12 kHz (low lumi) Mu20 Barrel Trigger ~ 2 kHz (high lumi) MU6 Barrel Trigger ~ 12 kHz (low lumi) Mu20 Barrel Trigger ~ 2 kHz (high lumi) ThresholdLum (cm -2 s -1 ) TriggerTotal (Hz)  /K Contribution MU4 (Fully Open) 10 33 Low-p T 9230091% MU510 33 Low-p T 2780079% MU610 33 Low-p T 2240079% MU810 33 Low-p T 1385081% MU1010 33 Low-p T 1116083% MU1110 34 High-p T 2757083% MU2010 34 High-p T 1368083% MU4010 34 High-p T 800085% LVL1-µ Trigger Rates

24. Selection Efficiency in the Low-pT Region p T (GeV/c) N events 2.00.8 × 10 6 2.51 × 10 6 3.00.9 × 10 6 3.50.9 × 10 6 p T (GeV/c) % events with no RPC signal LVL1 Efficiency 2.099.1%(1.4 ± 0.1) · 10 -3 2.599.5%(3.4 ± 0.1) · 10 -3 3.099.5%(4.8 ± 0.1) · 10 -3 3.596.6%(33.4 ± 0.3) · 10 -3 Low-p T Events Position They are concentrated in the transition region between Barrel and EndCap

25. Event track with p T =2.5GeV/c that enter in the spectrometer EndCap region but it is selected by LVL1 RPC Barrel Barrel LVL1 RoI pTpT % events in the EndCap but triggerated by RPC 2.046.2% 2.537.4% 3.08.7% 3.59.6% Selection Efficiency in the Low-pT Region

26. Efficiency Curve LVL1, LVL2 and EF It was possible elaborate an eventual trigger menu for the first run at luminosity of 10 31 cm -2 s -1 ThresholdLevel 1Level 2Event Filter Rate (Hz) MU4PS=1000~1kPT1 1 MU62249848 MU6PS=200224PT1 1 MU10112137 MU10PS=100112PT1 1 MU151942 MU15PS=2019PT1 1 MU20i14PT14PT14 MU408PT8 8 2MU64PT4 4 2MU20< 1PT1 1

27. Risults and Conclusions Cosmic Rays Selection Efficiency = 0.5±0.1%  MU4 Rate Expected = ~ 600Hz Rate Expected MU6 = 22kHz L = 10 33 cm -2 s -1 Rate Expected MU20 = 13kHz L = 10 34 cm -2 s -1 Rate value obteined completely satisfy the request of limit LVL1-  pass band Elaboration of an Trigger Menu for first p-p run at luminosity of 10 31 cm -2 s -1 Study on Performance of First Level Muon Trigger Selection Efficiency in the region with RPC >99%  Low-p T Selection Geometrical Efficiency = 82%  High-p T Selection Geometrical Efficiency = 78%

28. Backup Slides

29. LVL1-µ: CM, PAD e RoI Estensione spaziale = ∆  × ∆  ≈ 0.1 × 0.1 I segnali dagli RPC sono inviati alla CM Sincronizza segnali d’ingresso Effettua le coincidenze Operazioni di majority Fornisce i risultati del trigger sia per la proiezione  che  4 CM costituiscono una PAD La PAD genera le informazioni relative alle RoI: Region of Interest

30. P T = 20 GeV/c The inversion of the asymmetry sign between Low and High-p T is due at intersection of muons trajectories Muon Selection Asymmetry  

31. Barrel Simulation: Cosmic Trigger Coincidence windows are completely opened! The maximum width is given by the Matrix hardware processor size (64x32 channels) Cosmic Trigger RoadMap: Special Lvl1 algorithm configuration is used to maximize Trigger efficiencies wrt to cosmic events! Standard Configuration Coincidence Windows fully opened Top-Down Algorithm Cosmic Trigger RoadMap: Special Lvl1 algorithm configuration is used to maximize Trigger efficiencies wrt to cosmic events! Standard Configuration Coincidence Windows fully opened Top-Down Algorithm

32. 40 GeV/c Fermi Function: Rates Calculus: Parameterizing efficiency curves The Rates Calculus Programm need the knowledge of efficiency curves with a step of 500MeV/c We don’t have such granularity 6 GeV/c Fermi Function Polynomial Line Fermi Function Polynomial:  (p t ) = D + Ex -1 + Fx -2 + Gx -3 Line:  (p t ) = H + Ix

33. Selection Efficiency in the very Low-p T region: p T =2GeV/c Selection Efficiency in the Barrel at 2 GeV/c= (0.14  0.01)% Number of RoI Events in the Barrel spectrometer |  |<1.05 Threshold events with Roi=1 MuCTPi Position of events with RoI=1  of the Barrel muons selected by RPC

34. Selection Efficienza in the very Low-p T region: p T =2GeV/c Selection Efficiency in the Endcap at 2 GeV = (0.083  0.004)% Events in the EndCap spectrometer |  |>1.05 Numero di RoI attivate MuCTPi Position of events with RoI=1  of EndCap muons selected by RPC Threshold events with Roi=1

35. Rates Calculus: Rates Value Variation The LVL1 selection efficency value contribute significantly to the uncertainly on rates Small oscillation of efficiency value in the low p T region produce considerable variation on rate value Efficiency value of 0.5% for p T ≈ 2 GeV/c produce an increase of rate value from ~13kHz to ~19kHz Nominal Value