Simulation/Reconstruction for CPP Experiment

1. Simulation/Reconstruction for CPP Experiment David Lawrence JLab Sept. 27, 2013 simulation status - D. Lawrence - JLab

2. The GlueX Detector in Hall-D • CPP Experiment will use GlueX detector in Hall-D: • Linearly polarized photon source (~9GeV for GlueX) • 2T solenoidal magnetic field (dp/p = few %) • Drift chambers • High resolution Time-of-flight detector • FCAL segmented lead-glass calorimeter • Modifications to standard GlueX setup: • Replace LH2 target with thin Sn target • Move target upstream to improve low-angle acceptance • Remove start-counter • Add muon/pion detector • Move coherent peak to 6GeV simulation status - D. Lawrence - JLab

3. Experimental Setup Signal reaction • All occur via the Primakoff effect (interaction with the Coulomb field of nucleus) • All result in very forward going particles • Low t (-t < 0.005 GeV2) Normalization Beam polarization Solenoid FDC Muon Detector TOF FCAL Target simulation status - D. Lawrence - JLab

4. Motivation for geometery changes in target region Without start counter With start counter • Smaller z values = target is further upstream • GlueX LH2 target extends from z=50cm to z=80cm simulation status - D. Lawrence - JLab

5. Start Counter is in the way Location: (X, Y, Z) = (2.3, 0, 96.6) =============================================== Volume: STRC material: Scintillator density: 1.032 g/cm^3 rad. length: 42.1442 cm A: 11.0618 Z: 5.56867 ancestory: STRC -> STRT -> LASS -> HALL -> SITE Nose angle=17.5o Trajectory=1.4o Material=3mm/sin(18.9o) = 9.3mm or 0.022 rad. lengths simulation status - D. Lawrence - JLab

6. GlueX Simulation Tool Chain Generators (pythia, genr8, …) • Multiple generators exist including: • bggen (pythia + low energy reactions) • genr8 (isobar t-channel configurable) • coherent bremstrahlung photons hdgeant (hddsGeant3.F) • GEANT3-based simulation engine: • detailed geometry defined • partial hit digitization mcsmear generated.hddm hdgeant.hddm • Final stage smearing: • cumulative-level smearing done • dark hits/noise hits added • future: dead-channel removal hdgeant_smeared.hddm reconstruction Simulation geometry defined in XML HDDS XML CentralDC_HDDS.xml ForwardTOF_HDDS.xml … simulation status - D. Lawrence - JLab

7. simulation status - D. Lawrence - JLab

8. Without Vertex Constraint simulation status - D. Lawrence - JLab

9. With Vertex Constraint simulation status - D. Lawrence - JLab

10. Kinematics of Experiment fppis angle between pp system and incident photon polarization vector in CM frame yppis angle between pp scattering plane and polarization vector in helicity frame simulation status - D. Lawrence - JLab

11. Linear Polarization of incident photon beam helps distinguish Primakoff from coherent ro production Primakoff + ro Primakoff only (invariant mass of p+p- system) (invariant mass of p+p- system) Wpp(GeV/c2) fpp ypp simulation status - D. Lawrence - JLab

12. Extracting the Cross Section(sort of) • Generated (not simulated) events representing 10 days of running with 70% polarization and r background only • Fits done for different W bins to extract Primakoff and coherent r contributions Black: MARK-II data Red: Statistical errors (from fits) Curves: Theory simulation status - D. Lawrence - JLab

13. Beam Polarization via poPrimakoff Use asymmetry between vertical and horizontal polarization data to extract polarization q f Exercise done with 100% polarization and no backgrounds Plots by Alexander Mushkarenkov simulation status - D. Lawrence - JLab

14. Software Level-1 Event filter • Event is kept if either L1a_fired or L1b_fired is true • Roughly 50% of hadronic events should be accepted by L1 trigger • Some tuning of FCAL cut will be needed • Trigger implemented in TRIGGER library (DMCTrigger objects) • hd_eventfilter A. Somov GlueX-doc-1043 Nominal goal for L3 is to discard 90% of L1-accepted events simulation status - D. Lawrence - JLab

15. Energy deposition in FCAL GEANT4 GEANT3 • Trigger depends strongly on FCAL response • Particle ID depends on BOTH FCAL and m/p detector response • Currently have limited ability to simulate FCAL response to hadronic showers 1 GeV FCAL m/p detector m+ p+ 2 GeV GEANT3 simulation by LubomirPentchev simulation status - D. Lawrence - JLab

16. Some ideas on what is needed next … • Implement muondetector in HDDS and test with hdgeant4 • Devise code management system to maintain CPP specific geometry while keeping aligned with geometry shared with GlueX (e.g. FDC) • Simulations of m/p separation with refined muon/pion detector design • Tracking in m/p detector ? • Complete implementation of relevant reactions in GlueX ANALYSIS package • gp -> pp+p- • gp -> ppo • Full-scale simulation and extraction of signal reaction with all relevant backgrounds • s meson • Incoherent p+p- ? • Simulate Polarization measurement using Primakoff asymmetry with: • 70% polarization • Realistic trigger • Backgrounds simulation status - D. Lawrence - JLab

17. Backup Slides simulation status - D. Lawrence - JLab

18. Backgrounds • Experiment will measure reaction: gPb -> Pbp+p- • Primary background will be coherent ro production followed by r->ppdecay • Will use angular distributions to separate Primakoff from coherent ro production (see later slides) • Currently gathering list of other potentially relevant backgrounds including: • s meson production (angular distributions same as Primakoff) • incoherent p+p- production • … simulation status - D. Lawrence - JLab

19. Relating cross-section to ap-bp Figure 5. from Pasquini et al. Phys. Rev. C 77, 065211 (2008) gg->p+p- Cross-section for gg->p+p- calculated based on two values of ap-bp: ap-bp = 13.0 x 10-4 fm3 (top, dotted line) ap-bp = 5.7 x 10-4 fm3 (solid and dashed lines) Cross-section varies by ~10% for factor of 2 variation in ap-bp Need measurement of s(gg->p+p-) at few percent level ap-bp = 13.0 ap-bp = 5.7 (Data points from MARK-II) Invariant mass of p+p- dotted: subtracted DR calculation with ap-bp= 13.0 dashed: subtracted DR calculation with ap-bp= 5.7 solid: unsubtracted DR calculation with ap-bp= 5.7 simulation status - D. Lawrence - JLab

20. Detector Rates/Acceptance • 107 tagged photons/second on 5% radiation length Pb target • 500 hours of running • Wpp acceptance down to ~320 MeV/c2(working to improve acceptance to even lower Wpp) • Estimated ~36k* Primakoff events(contrast this with the ~400 events in the acceptance of the MARK-II measurement) * before detector acceptance simulation status - D. Lawrence - JLab

21. Charged Particle Tracking • Tracking code development began in 2004 • Tracking is done in multiple stages: • Track Finding • Wire-based fitting (wire positions only) • Time-based fitting (drift times used) • Fitting done using a Kalman Filter (replaced original least-squares fitter) Source code is checked out and built via nightly cron job on 3 platforms Twice a week cron jobs automatically simulate and reconstruct single track events and multi-track b1p events Plots from 5/4/12 semi-weekly single track tests simulation status - D. Lawrence - JLab

22. Charged Particle Tracking Results of recent semi-weekly test doing full reconstruction of b1p events Final state: p p+ p+ p- p- gg (5 charged tracks) Some mis-identification of p+ and proton exists >60% reconstruction efficiency of “X” meson Plots from 5/7/12 semi-weekly b1p tests simulation status - D. Lawrence - JLab

23. Calorimetry • FCAL • Code developed based on experience with Rad-f experiment in Hall-B • Full reconstruction (GlueX-doc-823) • Depth corrections (GlueX-doc-1093) • Calibration procedure established FCAL reconstruction efficiency Red: minblocks= 1 Black: minblocks= 2 • BCAL • 1st generation Code developed copied from KLOE and adapted to GlueX • 2nd generation currently under development by GlueX • Improved angular resolution • Better error estimation • Increased po reconstruction efficiency (61%->73% 11/28/11 report) BCAL reconstructed energy resolution simulation status - D. Lawrence - JLab

24. pp invariant mass w/ vertex constraint w/o vertex constraint w/ vertex constraint w/o vertex constraint simulation status - D. Lawrence - JLab