1 / 25

Status of GlueX Particle Identification

Status of GlueX Particle Identification. Ryan Mitchell September 10, 2004. Outline. Overview of Particle ID Components. Physics Examples. Look at each subdetector: Central Drift Chamber (CDC) Barrel Calorimeter (BCAL) Cerenkov (CKOV) Forward Time of Flight (TOF)

job
Download Presentation

Status of GlueX Particle Identification

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Status of GlueXParticle Identification Ryan Mitchell September 10, 2004

  2. Outline • Overview of Particle ID Components. • Physics Examples. • Look at each subdetector: • Central Drift Chamber (CDC) • Barrel Calorimeter (BCAL) • Cerenkov (CKOV) • Forward Time of Flight (TOF) • Likelihoods for Particle Hypotheses. • Angular Efficiencies.

  3. Particle ID with the • GlueX Detector: • dE/dx from the CDC • 2. Time from the BCAL • 3. Photo-electrons from • Cerenkov • 4. Time from the TOF

  4. Starting Momentum Spectra • Starting momenta • for four reactions: • p  +-p • 2. p  K+K-p • 3. p  K*K*p • K+-K-+p • p  K1K-p • K+K-p • K++-K-p Black = proton, Blue = Kaons, Red = Pions

  5. PID Cases (p  K*K*p) • 0. No PID Info • CDC • CDC, BCAL • CKOV • CDC, CKOV • CDC,BCAL,CKOV • CKOV,TOF • CDC,CKOV,TOF • CDC,BCAL,CKOV,TOF Central Forward Black = proton, Blue = Kaons, Red = Pions

  6. An Example From GEANT (pK*K*p)

  7. dE/dx from the CDC Calculated with Argon Gas Good K/p separation below ~2 GeV/c. Black = Proton Blue = Kaon Red = Pion Green = Electron

  8. Measured dE/dx in the CDC Measurements for pK*K*p. Calculating Likelihoods: Given a track with momentum p hitting the CDC, CDCLi = 1/i(2)1/2exp(-(x-xi)2/ 2i) i = particle hypothesis xi = predicted measurement i = predicted error (10%) x = actual measurement A dE/dx resolution of 10% is assumed. (Case 2 tracks) Black = proton, Blue = Kaons, Red = Pions

  9. Time of Flight from BCAL With 200 ps resolution, resolve: -- /K up to ~1 GeV/c -- K/p up to ~2 GeV/c Flight distance of 1.0 meters. Black = Proton Blue = Kaon Red = Pion Green = Electron

  10. Measured Time from BCAL Calculating Likelihoods: Given a track with momentum p and pathlength L hitting the BCAL, BCALLi = 1/i(2)1/2exp(-(t-ti)2/ 2i) i = particle hypothesis ti = predicted measurement i = predicted error (200ps time resolution, 1% momentum res., 1% length res.) t = actual measurement Measurements for pK*K*p. Time resolution of 200ps (Case 2 tracks) Black = proton, Blue = Kaons, Red = Pions

  11. CKOV Photo-Electrons The CDR design: Index of Refraction = 1.0015 Length = 1.0 meters Efficiency = 90 cm-1. Black = Proton Blue = Kaon Red = Pion Green = Electron

  12. Measured CKOV NPE Calculating Likelihoods: Measurements for pK*K*p. Given a particle of momentum p, calculate the expected number of photoelectrons under different particle hypotheses:  = N0sin2c. (Case 7 tracks) If cerenkov “fires”: CKOVLi = (1-e-) + a – a(1-e-) If cerenkov is quiet: CKOVLi = e-(1-a) N0 = cerenkov efficiency  = length of cerenkov  = cone of radiation angle a = accidental rate Black = proton, Blue = Kaons, Red = Pions

  13. Time from the TOF Wall With 100 ps resolution, resolve: -- /K up to ~2.5 GeV/c -- K/p up to ~4.0 GeV/c Black = Proton Blue = Kaon Red = Pion Green = Electron

  14. Measured TOF from the TOF Wall Calculating Likelihoods: Given a track with momentum p and pathlength L hitting the TOF, TOFLi = 1/i(2)1/2exp(-(t-ti)2/ 2i) i = particle hypothesis ti = predicted measurement i = predicted error (100ps time resolution, 1% momentum resolution, 1% length resolution) t = actual measurement Measurements for pK*K*p. Time resolution of 100ps (Case 7 tracks) Black = proton, Blue = Kaons, Red = Pions

  15. Likelihoods • Combine the likelihoods from all the detectors into a single likelihood, e.g., LK = LKCDCLKBCALLKCKOVLKTOF • Make decisions based on likelihood ratios. For now, use: /K separation = 2ln(L/LK) K/ separation = 2ln(LK/L) p/K separation = 2ln(Lp/LK)

  16. /K Separation /K separation = 2ln(L/LK) for pK*K*p Look at the ratio for each detector individually. (case 2) (case 2) (case 7) (case 7) Blue = Kaons, Red = Pions

  17. /K Separation Combine likelihoods into an overall likelihood. Now look at the /K separation. Blue = Kaons, Red = Pions

  18. K/p Separation K/p separation = 2ln(LK/Lp) for pK*K*p Look at the ratio for each detector individually. (case 2) (case 2) (case 7) (case 7) Black = proton, Blue = Kaons

  19. K/p Separation Combine likelihoods into an overall likelihood. Now look at the K/p separation. Black = proton, Blue = Kaons

  20. Particle ID Cuts • To get an idea of efficiencies, make the following simplifications… • A Pion is considered identified if: 2ln(L/LK) > 2.0 • A Kaon is considered identified if: 2ln(LK/L) > 2.0 • A Proton is considered identified if: 2ln(Lp/LK) > 2.0

  21. Momentum Efficiencies For pK*K*p, there is definite structure in momentum efficiencies… Inefficiencies due to… -- high momentum central tracks. -- forward tracks from 2 to 3 GeV/c Black = proton, Blue = Kaons, Red = Pions

  22. Angular Efficiencies for K*K* Starting After PID Efficiencies in Gottfried-Jackson angles. Preference for forward- backward decays.

  23. Angular Efficiencies for K+K-

  24. Angular Efficiencies for K1K

  25. Conclusions • There are no overwhelming problems in GlueX particle ID. • In the present setup, inefficiencies occur due to: • High momentum central tracks • Forward tracks between 2 and 3 GeV/c • These inefficiencies lead to sculpted momentum spectra, but only slight variations in angular efficiencies.

More Related