besiii simulation status

34
BESIII Simulation Status Z.Y. Deng M. He H.M. Liu Y.J. Mao Z.Y. You Y.Yuan BES Annual Meeting June 1, 2005

Upload: nathaniel-ewing

Post on 30-Dec-2015

77 views

Category:

Documents


1 download

DESCRIPTION

BESIII Simulation Status. Z.Y. Deng M. He H.M. Liu Y.J. Mao Z.Y. You Y.Yuan BES Annual Meeting June 1, 2005. Outline. Introduction Current status Performance Future work Summary. BOOST Project ( B ESIII O bject O riented S imulation T ool). Event GENBES Generator. - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: BESIII Simulation Status

BESIII Simulation Status

Z.Y. Deng M. He H.M. Liu Y.J. Mao Z.Y. You Y.Yuan

BES Annual MeetingJune 1, 2005

Page 2: BESIII Simulation Status

2/34

Outline

Introduction

Current status

Performance

Future work

Summary

Page 3: BESIII Simulation Status

3/34

GeometryGeant4

Tracking

DetectorDigitization

Response

Hitobjects

EventGENBESGenerator

HepEvtformat

Raw dataMC truth

BOOST Project (BESIII Object Oriented Simulation Tool)

Page 4: BESIII Simulation Status

4/34

BOOST Project (BESIII Object Oriented Simulation Tool)

Components Generator Particles and physics processes Magnetic field Material MC truth Data I/O User interface Geometry Hits recording Digitization

sub-detector parts

common parts

Page 5: BESIII Simulation Status

5/34

boost/ bes/ common part src/ source c++ codes include/ header files

gen/ … event generators phy/ … physics processes tru/ … MC truth gmk/ … common makefiles mac/ … common card files dat/ … geometry data files exe/ … main program for execution doc/ … documentations mdc/ … main drift chamber tof/ … time of fight emc/ … electromagnetic calorimeter muc/ … muon chamber

BOOST code structure

Page 6: BESIII Simulation Status

6/34

BOOST Project (BESIII Object Oriented Simulation Tool)

Developers Liu Huaimin: common parts Yuan Ye: MDC Deng Ziyan: TOF & common parts Fu Chengdong/He Miao: EMC Peking University: MUC & xml geometry

Page 7: BESIII Simulation Status

7/34

BOOST Project (BESIII Object Oriented Simulation Tool)

Phase 1: Framework prototype, simple geometry with hit Finished, BOOST prototype realized in May,2003

Phase 2: Detailed detector with simple digitization Finished, now digits and MC truth output can be used for

reconstruction

Phase 3: Detailed digitization, data/MC comparisons In progress, need information from beam test and data.

A long-term work!

Page 8: BESIII Simulation Status

8/34

Geant4 and BOOST releasesGeant4

Release 6.1 March, 2004 Release 6.2 June, 2004 Release 7.0 December, 2004 Patch of 7.0 February, 2005

BOOST (mantained by cvs) ‘boost-1-0’ (March, 2004) -----Geant4.6.1 ‘boost-1-1’ (June, 2004) -----Geant4.6.2 …… ‘boost-2-0’ (February, 2005) ----Geant4.7.0 ‘boost-2-1’ (May, 2005) ----- Geant4.7.0+p01

Page 9: BESIII Simulation Status

9/34

BOOST Current Statusgeometry

Detailed BESIII geometry is constructed in BOOST

Geometry added since last annual meeting Accurate stereo cells in MDC New End-cap EMC SCM (Super-Conducting Magnet )

BESIII Geometry in XML format Geometry in BOOST now can also be constructed with XML Simulation, reconstruction, event display will share the same BES

III geometry See detailed information in Mao Yajun’s report

Page 10: BESIII Simulation Status

10/34

BOOST Current Status MDC stereo cell

Accurate description New class (J4TwistedTubs) introduced all stereo wire cells are constructed

Display for checking Twisted tube can’t be displayed directly hit positions in stereo cell used 3D scatter plot with ROOT package

Page 11: BESIII Simulation Status

11/34

BOOST Current Status end-cap EMC geometry

Quadrangular crystal Irregular boxoid Can’t be described by existing classe

s in Geant4 releases New class (G4IrregBox) introduced

Pentangular crystal Divided into 2 boxoids 2 boxoids share the same copy numb

er

quadrangular crystal

pentangular crystal

Page 12: BESIII Simulation Status

12/34

BOOST Current Status end-cap EMC geometry

construct all crystals in one sector

Place other sectors through rotation and reflection

Rotate east end-cap to place west end-cap

end-cap EMC

sector

rotation

reflection

Page 13: BESIII Simulation Status

13/34

BOOST Current Statusgeometry

From boost-2-1

barrel: MDC, TOF, EMC, SCM, MUC end-cap: TOF, EMC, MUC

Page 14: BESIII Simulation Status

14/34

BOOST Current Statusgenerator

genbes: (BESII generator system) Generate events in an ASCII file which supports HEPEVT inter

face with Geant4

The same user interface as BESII• genbes.cards , genbes.user

All BESII generators can be used in BOOST

tester: (single-particle generator for debugging) Shoot particles of given type to a given direction with given kinetic energy Users can randomize above quantities

Page 15: BESIII Simulation Status

15/34

BOOST Current Statusgenerator: howl

J/ψKs Ks π+π-π+π-

Hits in MDC

Hits in barrel TOF(no threshold cut)

Hits in end-cap TOF

Hits in end-cap MUC

Page 16: BESIII Simulation Status

16/34

BOOST Current Statusgenerator: rhopi

J/ψρπ π+π-

Hits in barrel MUC(π- punch through)

Shower in EMC

gamma

Page 17: BESIII Simulation Status

17/34

BOOST Current Statusprocesses & magnetic field & material

Physics processes BesPhysicsList constructed with Geant4 classes Many hadronic models installed, LHEP, LHEP-GN, Q

GSP-GN, …… More research on them needed to choose suitable

ones for BESIII physicsMagnetic field

1 Tesla uniform magnetic field defined inside SCMMaterial

All materials needed in BOOST defined in XML

Page 18: BESIII Simulation Status

18/34

BOOST Current Statushits and digitization

Hits recording Finished last year

digitization MDC

• Drift distance-> drift time• dE/dx• Wire resolution added• Simple background added

TOF• Detailed digitization• Light emission & light propagation• PMT response & threshold discrimin

ation

EMC • waveform added• Simulate main amplifier output• Provide risetime to eliminate no

ise• No noise, so not used now

MUC• Digit information added

Page 19: BESIII Simulation Status

19/34

BOOST Current StatusMC truth

Purpose To provide truth information for

reconstruction debugging Similar to MCMADE in BESII

What to save True state of the particles and their

association with detector response• Particle

• particles from generator • uninterrupted decayed daughters

• Detector response• hits in sub-detectors associated with particles

Page 20: BESIII Simulation Status

20/34

BOOST Current StatusMC truth

Track and vertex (associated with particle)

Track • PDGcode• charge• original vertex• terminal vertex• four-momentum• decayed daughters

Vertex• parent track• position • time

truth hits (associated with track)

Considering demand of each reconstruction developer

MDC• Hit history

TOF• First-hit

EMC• Total energy loss

MUC• Hit history

Page 21: BESIII Simulation Status

21/34

BOOST Current StatusData I/O

Ascii data I/O It’s temporary, but useful in software (SIM + REC) development

Ascii data format fixed for the moment

MC truth (TRUTH) & Raw data (DIGI) included in one file

• EVHEAD• TRACKTRUTH, VERTEXTRUTH• MDCTRUTH, MDCDIGI• TOFTRUTH, TOFDIGI• EMCTRUTH, EMCDIGI• MUCTRUTH, MUCDIGI

Users can turn on/off TRUTH and DIGI part of each sub-detector in run cards

Page 22: BESIII Simulation Status

22/34

BOOST Current StatusUser interface

Friendly to both users and developers source ~dengzy/.boostenv cvs co boost edit boost.cards & boost.user more controls in run cards

Run background/terminal/interactive/PBS job freely boost -b/-t/-i/-q use ‘boost –h’ for detail

Reconstruction developers can generate needed data files themselves

Page 23: BESIII Simulation Status

23/34

BOOST PerformanceSpeed

From LHC experience G4 direct comparison with G3 G4 is slower (1.5~2) than G3 G4 tracking particle to 0 energy

Compared with SIMBES SIMBES (G3/BESII), BOOST (G4/BESIII)

How to compare Similar cuts

• CUTE CUTM (10KeV), range cuts (1mm) Same platform (Besfarm7) Physics events used (dimuon, rhopi, bhabha)

Page 24: BESIII Simulation Status

24/34

BOOST PerformanceSpeed

CPU time (s) for one eventBesfarm7 (PIII/933)

Event SIMBES BOOST

ee -> uu 0.122 0.189

J/ψ->ρπ 0.471 0.763

ee -> ee 0.736 1.224

BOOST 1.5 times slower than SIMBES

More to do about speed improvement Threshold cut in digitization Speed of hadronic models to be compared

Stable (no crash for 100,000 events)

Page 25: BESIII Simulation Status

25/34

BOOST PerformanceMDC

P distribution (1GeV/c e-)

σp =6.1MeV

P distribution (1GeV/c mu-)

σ p=5.8MeV

Reconstruction result using BOOST data sample

Page 26: BESIII Simulation Status

26/34

1GeV/c e- vertical incidence

total energy loss in one scintillator (5cm width)

BOOST PerformanceTOF

Page 27: BESIII Simulation Status

27/34

BOOST PerformanceTOF

e- 1GeV/c Z=0 : forward TDC resolution = 121ps backward TDC resolution = 122ps

T : 1/sqrt(Q)

time-walkforward

backward

Page 28: BESIII Simulation Status

28/34

BOOST PerformanceTOF

e

μπ

kp

Randomly generate particles: e-, mu-, pi-, kaon-, protonMomentum: uniform between 0-1.6 GeV/cOnly intrinsic time resolution

Page 29: BESIII Simulation Status

29/34

BOOST PerformanceEMC

total energy loss of pi- e-

Page 30: BESIII Simulation Status

30/34

BOOST PerformanceMUC

Probability of single particle(μπ) entering MUC

Page 31: BESIII Simulation Status

31/34

BOOST Integration to BOSSframework

Different developing environment BOOST

• On besfarm• Configured with gmak

e• Based on GEANT4

BOSS• On koala• Configured with CMT• Based on GAUDI

Integration Status BOOSTAlg algorithm/packages c

reated in BOSS

Key functions of BOOST realized in BOSS

Interface of saving data to TDS is ok

Codes imported into cvs

Page 32: BESIII Simulation Status

32/34

Future Work

Move developing environment from besfarm to lxplus

Magnetic field mappingPhysics validation, optimizationData I/O

Objectivity? Raw data in online format, MC-Truth format?

Background, noise?Detailed detector response

Page 33: BESIII Simulation Status

33/34

Summary

Detailed BESIII detector is built in BOOSTAll BESII generators can be used in BOOSTRaw data/MC-truth data are successfully

used in reconstructionFriendly user interfaceBOOST integrated into BOSSPrototype of TOF detailed digitization is

realizedMore to do about detector response !

Page 34: BESIII Simulation Status

34/34

Thank you !