1 / 29

The Santa Cruz Linear Collider Group Simulation Effort

The Santa Cruz Linear Collider Group Simulation Effort. SiLC Workshop Prague, Czech Republic April 26, 2007 Bruce Schumm. What are we working on?. Tracker/Reconstruction/Fitter Validation Package Working on making it platform-independent Pulse Development Simulation

Download Presentation

The Santa Cruz Linear Collider Group Simulation Effort

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. The Santa Cruz Linear Collider Group Simulation Effort SiLC Workshop Prague, Czech Republic April 26, 2007 Bruce Schumm

  2. What are we working on? • Tracker/Reconstruction/Fitter Validation Package • Working on making it platform-independent • Pulse Development Simulation • Resolution, Efficiency vs.  • Strip pitch / Readout pitch • Tracking with an All-Silicon Tracker • Non-prompt tracks • Performance vs. z segmentation

  3. People Myself, plus a number of very dedicated undergraduate physics majors: Tyler Rice Lori Stevens (summer: SLAC Pope Fellow) Chris Meyer Luke Kelley We’re working on results for LCWS07; not too much news for this meeting

  4. Tracking Efficiency and Momentum Resolution Analysis Chris Meyer UCSC ILC Simulation Reconstruction Meeting March 13, 2007

  5. SODTrack Analyzer • SODTrack reconstruction written by Fred Blanc; extends VXD stubs into central tracker. For now uses cheating for VXD stubs. Fitter: simple helix fit. • Geared towards SiD detector concept • Original tracking efficiency code written by Eric Wallace (UCSC Undergraduate, now at U of W) was implemented to create root output files • The SID-01 files were used for ZPole bbbar events, and SID-AUG-05 for 500 GeV uds events.

  6. Definition of findable track (filters) • Cos() < 0.5 • rOrg < 400 mm • Pathlength < 500mm • Final or Intermediate State Particle • Purity of hits on track > 0.79 from same MC Particle Also: Events only accepted if cos(T) < 0.5 and thrust value > 0.94 Criteria developed by Bruce Schumm, Lori Stevens, and Tyler Rice after some study

  7. ZPole bbbar All tracks originating outside 1.4 cm aren’t found…

  8. Total Efficiency for Prompt Tracks 500 GeV uds Making a pT > 5 GeV cut increases efficiency about 0.3% 98.46% efficiency for pT >.75 GeV/c 98.77% efficiency for pT > 5 GeV/c

  9. Efficiency vs pT-1 500 GeV uds At higher pT-1 (lower pT) there is some inefficiency

  10. Fake Rates vs. Momentum Calculated from all SODTrack hits; “fake” is < 80% of hits from same MC Particle.

  11. Efficiency vs.  500 GeV uds Let  be the angle between a track and the thrust axis No obvious effect as you go into the jet core (small ) pT > 0.75 GeV pT > 5.0 GeV

  12. “Tri-plots” Residuals Billior Calculation Expectation Error Matrix Expectation

  13. Tracking Validation Package • Chris Meyer working on making package platform-independent – will run on ROOT output from any code framework. User would need to • Apply selection algorithm to filter “findable” tracks (framework dependent, but our criteria are good guide) • Do hit-by-hit MC Truth association • Output track-by-track information in specified ROOT format • Provide look-up table (in p and cos) ofexpected momentum resolution (can be gotten by running LCDTRK, or your favorite routine)

  14. Pulse Development Studies Have been exploring angle effects and strip/readout pitch for long (60 cm) ladders Starting to get results now; expecting to have organized body of results for discussion in Hamburg Preliminary (very) sense: Efficiency vs. entrance angle not a problem for L=60cm. Was a problem for L=167cm; exploring transition now

  15. Resolution vs. Incidence Angle  for 25 m Pitch Sensors Read Out with a 50 m Pitch (60 cm Ladder) Track through un-read-out channel Average Track through read-out channel

  16. Resolution vs. Incidence Angle  for 37.5 m Pitch Sensors Read Out with a 75 m Pitch (60 cm Ladder) By way of comparison, for sensors w/ equal sense and readout pitch d d = 50 m  = 4.6 m d = 75 m  = 7.1 m

  17. RECONSTRUCTING NON-PROMPT TRACKS • Snowmass ‘05: Tim Nelson wrote axial-only algorithm to reconstruct tracks in absence of Vertex Detector • UCSC idea: use this to “clean up” after vertex-stub based reconstruction (VXDBasedReco) • About 5% of tracks originate beyond the VXD inner layers • For now: study Z-pole qq events • Work done with Tyler Rice, Lori Stevens

  18. Cheater • VXDBasedReco had not yet been ported to org.lcsim framework, so… • Wrote “cheater” to emulate perfectly efficient VXDBasedReco; assume anything that can be found by VXDBasedReco is found and the hits flagged as used • Loops over TkrBarrHits and MCParticles, finds particles with rOrigin < 20mm and hits from those particles, removes them from collections • rOrigin defined as sqrt(particle.getOriginX()^2 + particle.getOriginY()^2)

  19. AxialBarrelTrackFinder (Tim Nelson, SLAC) • Loops over all hits in each layer, from the outside in, and finds 3 “seed” hits, one per layer • Performs CircleFit (alogrithm provided by Norman Graf) to seed hits • If successful, looks for hits on the remaining layers that can be added to seed fit, refitting after each hit added. • If at least 4 hits on track, and Chi^2 of fit reasonable, creates track object and adds to collection • Only two (half-barrel) segments in z for now

  20. AxialBarrelTrackFinder Performance Define “findable” particle as • Pt > 0.75 • Radius of origin < 400 mm (require four layers) • Path Length > 500 mm • |cos| < 0.8 Calculate efficiency for finding such non-prompt tracks

  21. Particle is “found” if it is associated with a track with four or more hits, with at most one hits coming from a different track. All non-associated tracks with pt>0.75 and DCA < 100mm are labeled “fake”. Particles Fakes Found 5 Hits 131 (43%) 1 Found 4 Hits 100 (33%) 305 Not Found 73 (26%) -------- Total 304 (100%)

  22. Further Exploration of AxialBarrelTracker (new) • Four-hit tracks very impure… • Restrict to particles that have exactly one hit on each of the five layers  five-hit track efficiency increases to 68% • Additionally require that all hits on three-hit “seeds” come from same particle  five-hit track efficiency increases to 87%

  23. AxialBarrelTracker Modifications Under Development • Require that  of hits on track be at least within /2 of one another  may reduce 4-hit fakes by up to 50% • Project underway (Lori Stevens) to introduce z segmentation into pattern recognition (have only been requiring all hits have same sign in z up until now)  may clean up three-hit seeds if segmentation is fine enough (?) • Multiple passes through AxialBarrelTracker with initially pass having very stringent requirement on seed quality • The challenge now is to try to reconstruct particles with fewer than five hits!!

  24. Some Other Studies (if Time Permits)? • Chris Meyer is beginning to look into jet energy resolution assuming perfect measurement of neutrals • How low do we need to go in pt? • What tracking efficiency is needed? Wrap-up UCSC undergrads working on basic but important studies – mostly, but not all, SiD-related. Tracker validation and pulse simulation more generic, but five-layer tracker studies clearly geared towards SiD. More results at Hamburg, and over summer.

  25. RANDOM BACK-UP SLIDES

  26. What’s Left after “Cheating”? (258 events, no backgrounds) Total hits:30510 100% Good hits: 1754 5.7% Looper hits: 13546 44.4% Knock-on hits: 10821 35.5% Other hits: 4389 14.4% Total tracks: 6712 100% Good tracks: 445 6.6% Looper tracks: 459 6.8% Knock-on tracks: 3303 49.2% Other tracks: 2505 37.3% “Good” Track Momentum “Other” “Knock-on” (less than 10 MeV) “Looper” Number of hits on track

  27. And where do these tracks originate? “Looper” “Good” “Knock-on” “Other” Radius of Origin (mm) Of Tracks

  28. Particles can be found more than once… (but there’s only one entry per particle in the previous table) Number of times each found particle is found

  29. Motivation To explore the effects of limited detector resolution on our ability to measure SUSY parameters in the forward (|cos()| > .8) region.

More Related