Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 1
SuperKEKB/Belle II experiment Status of GEANT4 simulation
Takanori Hara (KEK) Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 2
Contents Introduction of experiment Geant4 simulation Standalone Geant4 Geant4 within software framework Summary Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 3 B-Factory aimed for
KEKB+Belle Kobayashi Ma er excess of Universe +Maskawa Dark ma er Origin of mass ...
Ma er Antima er Big bang Now (a 13.7 billion years later) Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 4 KEKB Accelerator + Belle Detector Large-solid-angle magnetic spectrometer Asymmetric energy e+ e- Collider tracking/vertexing 1.5T Solenoid Drift Chamber e+ (LER) : 3.5 GeV e/g detection Silicon Strip - e (HER) : 8.0 GeV CsI calorimeter Detector e+ e- Ecm = 10.58 GeV = @Y(4S) B B pair p/K/p identification Cherenkov detector Instantaneous Luminosity 34 -2 -1 2.108 10 cm s m/KL identification G (twice of the design value) Resistive Plate Chamber Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 5 Confirmation of CP violation
- 1 ~800M BB pairs
[fb ] 0 Y(1S) : ~6 fb-1 B 0 Y(2S) : ~7 fb-1 B Y(3S) : ~3 fb-1 Y(4S) : ~720 fb-1 Y(5S) : ~100 fb-1 off-reso. : remnant -1 ~950 fb
Kobayashi-Maskawa theory is confirmed. But this is not enough to explain the Universe Other source of CP violation is implied ... Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 6 New B-Factory aims for SuperKEKB + Belle-II KEKB+Belle New Physics ! Kobayashi Ma er excess of Universe +Maskawa Dark ma er (in TeV scale) Origin of mass ...
Ma er Antima er Big bang Now (a 13.7 billion years later) In order to see the CP violation in the earlier stage of Universe Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 7 Luminosity Prospect
50 ab-1 by 2020 = 50 present G 10 Physics with O(10 ) B, t , D, ...
We are here ! 35 2 8 x 10 /cm /s
~0.95 ab-1 34 2 2.1 x 10 /cm /s
3-year shutdown for detector upgrade 35 2 2 x 10 /cm /s Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 8 Detector Upgrade Beam BG tolerant Belle low material budget Super Belle be er performance Barrel : TOF+ACC TOP EndCAP : EACC A-RICH
Barrel PID
Endcap CDC: small cell, PID long lever arm
4-lyr SVD 6-lyr (2-lyr PXD + 4-lyr SVD) ECL: waveform sampling, pure CsI for end-caps KLM: RPC Scintillator +SiPM (end-caps) Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 9 Geant4
KEKB/Belle will continue taking data until 2010 ?? SuperKEKB/Belle II will start taking data from 2013
. IR Design (e.g. beam background simulation) : standalone G4 . Detector Optimization (e.g. PID, tracking) : w/ soft. frameworks
But now we have two software frameworks at this moment, . roobasf (upgraded version of Belle Analysis Frame) . ILC framework (Mokka/Marlin) Geant4 is implemented in each framework, but there is no campatibility combining good features from both sides ... Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 10 roobasf Framework
Simulation + Reconstruction
Geometry info. (ascii, DB, hard coded)
Condition DB
Tracking EvtGen g4superb Generator (GEANT4) Clustering ROOT File digitization vertexing,...
ROOT I/O ROOT I/O Generated events Detector simulation Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 11 ILC Software Framework
Simulation Reconstruction
LCIO (persistency data model)
Condition DB Marlin Pythia Mokka (Material DB LCIO Generator (GEANT4) Digitizaion Tracking...) ROOT
HEPEvt (ASCII) GEAR (XML) Generated events Geometry info. Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 12
Standalone Geant4 simulation or Geant4 with roobasf Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 13 Detector Background Component
Background component w.r.t current Belle
Synchrotron Radiation Lower ? Higher ? (upstream) Smaller beam size at Q but large bending magnet Synchrotron Radiation Much lower 1/800 (back-sca er) No QCS bending Radiative Bhabha Much lower 1/40 Larger crossing angle but no QCS bending Touschek (intra-beam) Much higher x 20-30 Very small beam size Beam-gas interaction Higher Vacuum around IP will be worse Higher current (x 10-100) Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 14 Beam-gas simulation Beam-gas sca ering : Bremsstrahkung eN geN and Coulomb eN eN 1) Upstream Beam-gas sca ering Upstream beam-gas sca ering causes beam energy loss. Then such beams may hit the beam-line component near IP and produce particle showers Affects to many (~all) detectors This BG depends on the upstream vacuum level, beam-current, beamline optics, and beampipe apertures 2) Beam-gas sca ering near IP Beam-gas sca ering near IP produces particles, and it may hit the detectors Affects to PXD, SVD, CDC, ... This BG depends on the vacuum level around IP (x10~100 worse) Important to fix the IR design Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 15 Beam-gas simulation 1/4 of the whole beamline is constructed in G4 simulation
electron beam (physics process off) (physics process on)
By changing the vacuum around IP, we’ll estimate the beam-gas BG
E (GeV) Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 16 Particle Identification (PID) . Upgrade of Belle PID system Current PID system of Belle Belle K/p - separation power : 3s Barrel : TOF + ACC End cap : ACC . Target Performance ACC : Aerogel Cherenkov Counter 3s 4s (0.6 < p < 4 GeV/c) (Threshold type) Install here Upgrade
Belle II 1.5T 1.2m Barrel : TOP e- 2.6 m e+ End cap : Aerogel RICH 8.0GeV 3.5GeV
Backward Forward Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 17 TOP counter . TOP (Time of Propagation) . A kind of RICH counter . Cherenkov radiator + time sensitive screen . Position (x, y) Position + time (x, y) . very compact & simple 500 25 ps) p K tdc (/ tdc 450 TOP
400 TOF y 1.18m
350
z
300 80 100 120 140 160 x ch Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 18 TOP counter Multiple sca ering d-ray Completely Geant4 based Electromagnetic/Hadronic shower Generation/Propagation of Cherenkov ph.
Good agreement in: Number of detected photons Detected time distribution Disagreement in: Spread of generated photon positions Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 19 g4superb A-RICH
. A be er p/K separation for a wider mom. range (0.7-4.0 GeV/c) threshold-type Cherenkov detector (Belle) does not provide sufficient separation for high-P particle. esp. high-P(~4GeV/c) region for B pp, Kp . limited available space (Forward region) Proximity focusingMokka ring-imaging Cherenkov counter (RICH) (GEANT4) Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 20 g4superb A-RICH . Geometry description . Particle tracking through A-RICH . Cherenkov ph. generation in the aerogel and the quarz window of the ph. detector . Rayleigh sca ering of the photons in the aerogel . Photons tracking through the quartz windown of the photon detector . Photon detectiion in the active area
Aerogel plane Detector plane Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 21 g4superb A-RICH
. Geometry at the moment hardcoded with the main parameters read from the text file . Currently optimizing the geometry and materials Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 22 ECL (Electromagnetic Calorimeter)
5
BSO(=Bi4Si3O12) 5
PbWO4
Incident particle : g (0.1 ~4GeV) random position & insident angle Pure CsI (<30 ) Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 23 ECL (Electromagnetic Calorimeter)
Energy resolution
Pure CsI (2cm x 2cm x 30cm)
BSO(=Bi4Si3O12) (2cm x 2cm x 20cm)
PbWO4 (2cm x 2cm x 20cm) Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 24
Geant4 with ILC framework Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 25 Mokka - BelleExp Geometry
+, Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 26 Mokka - Beam Pipe Geometry
“
R R 14.99 16.45 9.99 11.45 Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 27 Mokka - Belle II PXD Geometry
13.00 8 22.00 12 Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 28 Mokka - Belle II PXD Geometry Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 29 Mokka - Belle II SVD Geometry
38 8 2 80 10 2 66 14 3 115 17 4 95.5 10 1 140 14 1 114 17 1 Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 30 Mokka - Belle II CDC Geometry
H 2 6
4H Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 31 Mokka - Tracker Implementation Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 32 Mokka - Belle SVD Geometry
20 6 2 43.5 12 3 70 18 5 88 18 6 Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 33 Marlin - PXD Digitizer
7 7
7 Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 34 Marlin - SVD Digitizer Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 35 Tracking : Reconstructed Decay CP side: Tag side (+c.c.): B J/y Ks B D* p m m p p D p p K p p p
GEANT4 GEANT3 GEANT4 GEANT3 Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 36 Tracking : Impact Parameter Resolution
r Definition of impact parameter d 0 & z 0
PCA = point of closest approach defined in x-y plane
PCA z m+ from J/y decay z0
= GEANT3 = GEANT4 Perfect agreement between GEANT3 and GEANT4 Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 37
Merged Framework Gearbox ? ILC Framework Unified Geometry Handling Marlin Unified Condition DB Pythia Mokka (Material DB Generator (GEANT4) Digitizaion Tracking...)
roobasf I/F Common Event Data Model ROOT File
Tracking EvtGen g4superb Generator (GEANT4) Clustering vertexing,...
roobasf Framework Asia-Pacific Geant4 Workshop @ Tsukuba, Japan (8, Dec., 2009) 38 Summary
Geant4 is the basic and practical tool for Beam background simulation for IR design Performance study for Particle ID detectors Detector optimization in Belle-II experiment Geant4 works well with roobasf/ILC framework unified framework is proposed and Geant4 module has to be fit to this framework Common event data model, Mokka I/F, Geometry handling (Gearbox ?), ... have to be prepared Geometry handling is a good start point for Geant4 We have already start this.