1. Review and Perspective of Muon Studies Marco Destefanis Università degli Studi di Torino PANDA Collaboration Meeting GSI (Germany) September 5-9, 2011 for the PANDA Collaboration

2. Overview Motivation Review of Drell-Yan process and background  Generators  Layout (MISS)  Cut studies and Results  Trigger Future perspective Summary

3. Drell-Yan Process and Background Background studies: needed rejection factor of 10 7 Drell-Yan: pp ->  +  - X cross section   1 nb @ s = 30 GeV 2 Background: pp ->  +  - X, 2  + 2  - X,…… cross section   20-30  b m  = 105 MeV/c 2 ; m  145 MeV/c 2 average primary pion pairs:  1.5

4. Phase space for Drell-Yan processes x 1,2 = mom fraction of parton 1,2  = x 1 x 2 x F = x 1 - x 2  = const: hyperbolae x F = const: diagonal PAX @ HESR symmetric HESR collider 1 1.5 GeV/c 2 ≤ M  ≤ 2.5 GeV/c 2

5. A. Bianconi Drell-Yan Generator for pp Antiproton beam Polarized/Unpolarized beam and target Drell-Yan cross section from experimental data Selects event depending on the variables: x 1, x 2, P T,, ,  S from a flat distribution Cross section: A. Bianconi, Monte Carlo Event Generator DY_AB4 for Drell-Yan Events with Dimuon Production in Antiproton and Negative Pion Collisions with Molecular Targets, internal note (PANDA collaboration) A. Bianconi, M. Radici, Phys. Rev. D71, 074014 (2005) & D72, 074013 (2005)

6. Simulation Layout DPM generator 10000 simulated background events (elastic + inelastic events) Pythia8 generator 10 5 simulated background events P8gen->SetParameters("PhaseSpace:pTHatMin = 0.001"); P8gen->SetParameters("PhaseSpace:mHatMin = 0.001"); P8gen->SetParameters("PhaseSpace:pTHatMinDiverge = 1.0000e-05"); P8gen->SetParameters("PhaseSpace:minWidthBreitWigners = 0.00001"); P8gen->SetParameters("HardQCD:all = on"); P8gen->SetParameters("SoftQCD:all = off"); P8gen->SetMom(15.); Comparison: 10000 simulated events

7. PANDA Detector Setup Design Signal: A. Bianconi Drell-Yan generator Background: PYTHIA8 generator Framework: Muon Independent Simulation Software (MISS) Next step: complete the work in PANDAROOT

8. MISS Muon Independent Simulation Software Pure GEANT4 simulation ROOT used for storage and plots Geometry: magnet design: nov-dec 2007 release magnetic field: Genova design Muon Detector: 1 layer each 2cm of iron Sensible volumes filled with Ar+CO 2 Results presented in the Physics Book (credits G.C. Serbanut)

9. ABDYG Signal Distribution Most of the signal composed of two muons in the ENDCAP Smaller contribution from B+E couples Contributions from the FS region are not negligible

10. Background and Cuts Sources of background Primary background: Primary  & Secondary  from Primary  Secondary background: Secondary  & Secondary  from Secondary  Cuts and their effect on signal Iron At least 1 hit in the first 2 layers q T > 0.75 GeV/c Signal Linear rejection depending on Fe thickness Reject 30% of the signal ~ 35% of the signal is reconstructed Primary Background Linear rejection depending on Fe thickness Rejection ≈ 10 3 Rejection ≈ 10 4 Secondary Background Almost no effectRejection ≈ 10 4 Rejection >510 6 Next Step: Kinematic refit Rejection factor of 10 7

11. DY Asymmetries @ Vertex UNPOLARISED SINGLE-POLARISED 500KEv included in asymmetries Acceptance corrections crucial! 1 < q T < 2 GeV/c 2 < q T < 3 GeV/c xPxP xPxP xPxP xPxP xPxP xPxP

12. 15 March 2011MPB - PANDA week@GSI12 Barrel & EndCapBarrel & Muon Filter background hits from LoLo L 123 L5L5 L6L6 L7L7 ≥  0.360.160.050.040.03 ≥  0.080.0210 -3 ≥ K 0.027 10 -3 2 10 -3 10 -3 DY signal ≥  0.890.860.670.660.65 ≥  0.450.350.160.150.14 Triggering on hits – MDT Layers Acceptance for background and signal events

13. 15 March 2011MPB - PANDA week@GSI13 Event rates trigger LoLo L 123 L7L7 ≥ 1  1.5 10 5 8 10 4 2 10 4 ≥ 1  2 10 -1 2 10 --1 1.5 10 -1 ≥  4 10 4 6 10 3 5 10 2 ≥  10 -1 3 10 -2  BKG and DY signal by Pythia 8  Cross sections by Pythia 8  DY signal for  >1 GeV  Luminosity 10 31  BKG cross section 50 mb  DY cross section 25 nb  DY cross section for  >1.5 GeV 5 nb

14. Future Perspectives Complete the simulations in the TS Perform simulations with the full FS PID studies with the muon system PID with the combination of muon system and central tracker informations Muon system reconstruction algorithm Integration of mu-det TS and FS informations Muon system stand alone tracking software Integration between muon and inner tracking Possible PandaRoot Tools

15. Summary Generators for Drell-Yan studies MISS and PandaRoot layout Cuts for background rejection Kinematically constrained refit still to be investigated Extensive simulations needed on the GRID (~10 8 Ev) with PANDAROOT in order to: check rejection factor set kinematic cuts Occupancy studies Trigger studies Retracking routines: stand alone + combined Complete detector design NEEDED in both TS and FS

16. Computing Environments – INFN Grid Motivation CPU time consumption collect statistics  Operational Nutshell with precompiled FairSoft, PandaRoot, gcc and system libraries Jobs: Jobs are started with scripts (BASH+AWK) Nutshell is moved to the local machine (  40s) Environment variables set Run process (40 min) Output send to the database and saved (2min, 910MB) Nutshell and output removed from the local machine Full automatization (credits G.C. Serbanut) (few s) 10000 signal events

17. Computing Environments – Farm  Operational Nutshell with precompiled FairSoft, PandaRoot, gcc and system libraries Jobs:  Jobs are started with scripts (BASH+AWK) Nutshell is created and then moved to the local machine (few s) Jobs are running locally (40 min) Output moved to storage elements (2min, 910MB) Clean the computing element (few s)  Check free space on disks Balance of the amount of data on storage elements (credits G.C. Serbanut) 10000 signal events

18. Computing Environments – Cluster 4 clients with 2 CPUs dual core -> 16 threads at once Jobs are started with scripts (BASH+AWK)  Passwordless ssh connection among all the machines  Clean input/output (few s) Copy locally input data (few s) Start and monitor the simulations (35 min) Retrieving the data (2min, 910MB) PC Clients Server (SSH v.2) (credits G.C. Serbanut) 10000 signal events

19. Pythia 8 Interface to use Pythia8 for simulations PndP8Generator* P8gen = new PndP8Generator(); P8gen->Init(); Set momentum → P8gen->SetMom(double); Set parameters → P8gen->SetParameters(char*); “PhaseSpace:pTHatMin = 0.001” pythia.readstring(PhaseSpace:pTHatMin = 0.001); Call to TRandom1 or TRandom3 classes pytr1rng and pytr3rng

20. Possible PandaRoot Tools Muon system reconstruction algorithm Integration of mu-det TS and FS informations Muon system stand alone tracking software Integration between muon and inner tracking

21. Summary B 20cm Fe x E 34cm Fe rejection factors with q T,μμ > 0.75 GeV/c reaction vertex in the target region 1.5 < M μμ < 2.5 GeV/c 2 kinematically constrained refit still to be investigated Rejection factor achieved for secondary background: > 5 10 6 Few months of data taking are enough to:  evaluate unpolarised and single-spin asymmetries with good accuracy  investigate their dependence on q T, μμ Extensive simulations needed on the GRID (~10 8 Ev) with PANDAROOT in order to: check rejection factor set kinematic cuts