1. GAUSS - GEANT4 based simulation for LHCb
GEANT4 Delta Review
9 October 2002
W. Pokorski / CERN

2. Contents Introduction to LHCb Overview of Gauss project
GiGa – Gaudi interface to GEANT4
Physics validation: RICH1 and HCAL results
Summary
09/10/2002
LHCb Simulation

3. LHCb Experiment
Precision Measurements of CP violation in the B Meson System
Large Sample of Events with Bd and Bs Mesons.
Most of the b hadrons are produced at small polar angles.
LHCb: Single Forward Arm Spectrometer with Open Geometry.
From the CP asymmetries in the final states of B-meson decays,
Measure CKM Angles.
This design is
being modified
to optimize the
performance of LHCb.
09/10/2002
LHCb Simulation

4. GAUDI – LHCb software framework
all of the LHCb event processing software is built within a framework – Gaudi framework
separation between data and algorithms
data store-centred architectural style
separation between transient and persistent data
isolation of user codes/algorithms from underlying persistency technologies
components interact through their abstract interfaces
09/10/2002
LHCb Simulation

5. GAUDI - Object Diagram Converter Interactive Service Application
Manager
Converter
Converter
Event
Selector
Transient Event Store
Data
Files
Message
Service
Persistency
Service
Event Data
Service
JobOptions
Service
Algorithm
Algorithm
Algorithm
Data
Files
Transient Detector
Store
Particle Prop.
Service
Persistency
Service
Detec. Data
Service
Other
Services
Data
Files
Transient
Histogram Store
Persistency
Service
Histogram
Service
09/10/2002
LHCb Simulation

6. Gauss application Int.face GiGa Digi Alg Geant4 (GiGa) Pythia etc
JobOpts
JobOpts
JobOpts
Int.face
GiGa
Digi
Alg
Geant4
(GiGa)
Pythia
etc
Geant4
MCParticle
MCVertex
MCHit
Digit
MCDigit
HepMC
Cnv
Cnv
Cnv
Geometry
Generator
Detector Simulation
Digitization
09/10/2002
LHCb Simulation

7. GiGa overview
GEANT4 Interface for Gaudi Applications or Gaudi Interface to GEANT4 Applications
GEANT4 callable and controllable from within GAUDI environment
common detector geometry source used by other applications (reconstruction, visualisation)
communication via Transient Stores (Event, Detector Data) as any other service or algorithm in Gaudi
use of common services (ParticlePropertySvc, RandomNumberSvc, MagneticFieldSvc, etc.)
09/10/2002
LHCb Simulation

8. GiGa structure Data Files Persistency Service Application Manager
GiGaKine
Conversion
Service
G4 Kine
Transient Event Store
Event
Service
Geant4
GiGaHits
Conversion
Service
G4 Hits
Algorithm
Converter
Algorithm
Cnv
GiGa
Service
Algorithm
Cnv
Transient Detector
Store
GiGaGeom
Conversion
Service
G4 Geom
Action
Detec.
Service
Action
we can see two parts of GiGa
GiGaCnv – converts data (geometry, events) from and to transient stores (and loads it/extracts it to/from G4)
GiGa – configures, triggers and control the actual simulation
Other
Services
Data
Files
Persistency
Service
09/10/2002
LHCb Simulation

9. GiGa features it’s a Gaudi service
provides access to internal G4 event loop via GiGaRunManager
all interactions with Geant4 only through abstract interfaces of GiGa Service
minimizes the couplings to Geant4
allows loading external physics lists
instantiates (using Abstract Factory pattern) different “actions” (makes them to be plugable components)
09/10/2002
LHCb Simulation

10. GiGa - Geometry Conversion
Xml description
Materials
Volumes
Xml Cnv
Geo Conversion Service
Converter
Converter
Gaudi transient store
Geant4
Materials
Geant4
Volumes
09/10/2002
LHCb Simulation

11. Geometry conversion ex. (RICH 1)
XmlGiGa
G4OpenGL
09/10/2002
LHCb Simulation

12. GiGa - Kinematics Conversion
HepMC Event
LHCb
Vertices
LHCb
Particles
Kine Conversion Service
Converter
Converter
Converter
Geant4
Trajectories
Geant4
Primary
Vertices
09/10/2002
LHCb Simulation

13. Hits Conversion Service
GiGa – Hits Conversion
LHCb
Ecal Hits
LHCb
Muon Hits
LHCb
Velo Hits
Converter
Hits Conversion Service
Converter
Converter
Geant4
Hits
Geant4
Hits
Geant4
Hits
09/10/2002
LHCb Simulation

14. Detector Simulation – “physics lists”
crucial part of the whole simulation program – will certainly require several tuning iterations
most of the stuff already implemented in Geant4
some specific processes needed implementation
for RICH: photoelectric process (creation of photoelectrons in HPDs), energy loss: in the silicon of HPDs
new feature recently added to GiGa : modular physics lists
allows dynamic loading (via jobOptions) of particular physics “sublists”
expected to increase flexibility and to make validation easier
09/10/2002
LHCb Simulation

15. Sensitive Detectors & Hits
GiGaSensDetTracker
G4TrackerHits
(Geant4)
creates
GiGaTrackerHitsCnv
ProcessHit()
Geant4 world
converts to:
invoked when particle passed through the sensitive volume
MCHits
(/Event/MC/OT/Hits)
lvVolume
(XmlDDDB)
Gaudi world
<logvol name="lvU_ActiveLayer" … sensdet="GiGaSensDetTracker/myDet">
09/10/2002
LHCb Simulation

16. Physics Validation
- RICH
- HCAL

17. RICH detectors in LHCb
To identify charged particles in the momentum range GeV/c.
Two detectors: RICH1, RICH2.
Momentum range
RICH1: Aerogel 10 GeV/c
C4F < 70 GeV/c
RICH2: CF <150 GeV/c.
Photo Detectors: Baseline solution- HybridPhotodiodes (HPD).
RICH test beam presented: To test the performance of the
Aerogel radiator.
S. Easo
09/10/2002
LHCb Simulation

18. Test beam Set-up at CERN
S. Easo
Beam from CERN-PS: πˉ and p/π in the range 6 – 10 GeV/c (Δp/p = 1%)
09/10/2002
LHCb Simulation

19. Simulation of the Testbeam Setup.
Mirror Rad. of Curvature=1185 mm.
Four Pad Hpds are used.
Hpd
Mirror
Vessel
Filter
S. Easo
Aerogel
09/10/2002
LHCb Simulation

20. A Typical event in the Testbeam
Red lines: Charged particle
Green lines : Photons.
Transmission in Aerogel, Cherenkov radiation, Rayleigh scattering, etc, simulated properly
S. Easo
09/10/2002
LHCb Simulation

21. Photoelectric Effect at the HPD
Standard Geant4 processes not applicable in this case
(tabularized quantum efficiency data available from the manufacturer)
A Special class created to generate the photoelectrons,
which is derived from a GEANT4 base class.
This process uses the quantum efficiency data
and the results of Fountain focussing tests.
Electron Energy: High Voltage applied.
Direction: From Fountain focussing.
The quantum efficieny data includes the loss of photons by
reflection at the Hpd quartz window surface.
S. Easo
09/10/2002
LHCb Simulation

22. Hit Creation in the Si Detector
Implemented using a special process class since the standard
Geant4 procedure too complex for this purpose
the only important point: photoelectrons loose all their energy in the Si
The backscattering causes a loss of efficiency in creating hits
Parameterized efficiency = B* N/S
where N = threshold cut in terms of width of the pedestal = 4
S= Signal to noise ratio=10
B= backscattering probability=0.18.
S. Easo
09/10/2002
LHCb Simulation

23. RICH1 with SingleParticleGun
RICH1 Event
Pion with 7 GeV/c.
Cherenkov Photons
In Aerogel and C4F10.
Rayleigh scattering
Switched off for
Illustration.
S. Easo
09/10/2002
LHCb Simulation

24. RICH1 Hits
09/10/2002
LHCb Simulation

25. HCAL Test beam
HCAL is a sampling device made out of steel as absorber and
scintillating tiles are active material.
The scintillating tiles run parallel to the beam axis.
It will provide data for the LHCb hadron trigger.
Testbeam studies of the response to particles incident at
various angles is studied
comparison to simulation
I. Belyaev +
A. Berdiouguine et al.
09/10/2002
LHCb Simulation

26. Energy Response in HCAL
Response to 50 GeV/c Pions
Histogram : Real Data
Dots: Simulation
Testbeam (1) Data , GEANT3
(MICAP +FLUKA).
HCAL TDR.
Testbeam (2) Data, GEANT4.
I. Belyaev +
A. Berdiouguine et al.
09/10/2002
LHCb Simulation

27. Energy Resolution of HCAL
Data and G3
Energy Resolution of HCAL
Testbeam Data, GEANT3
GEANT3 with GEISHA, FLUKA,MICAP
Testbeam Data , GEANT4
G4+GEISHA agrees with G3+GEISHA.
Need help to understand and use
G4 with QGS+CHIPS
Data
G4 (QGS+CHIPS)
G4(GEISHA)
Data and G4
I. Belyaev +
A. Berdiouguine et al.
09/10/2002
LHCb Simulation

28. Panoramix view of MCHits
09/10/2002
LHCb Simulation

29. Panoramix view of MCHits (2)
09/10/2002
LHCb Simulation

30. Summary Current Status:
we can run GEANT4 simulation with all the subdetectors included
we are testing/validating subdetector specific code and physics processes
RICH test beam data comparison – well advanced
HCAL test beam data comparison – well advanced
other subdetectors participation increasing
09/10/2002
LHCb Simulation

31. Summary (2)
Planning:
by end of the year: to have a complete GEANT4 based simulation application
by the summer: to perform some “test productions”
by … : to definitely move from the old GEANT3 based simulation program to the new one, GEANT4 based
09/10/2002
LHCb Simulation