[PPT] - GEANT4 Simulation of the abBA/Nab Spectrometer: Progress Report PowerPoint Presentation

SLIDE 1

GEANT4 Simulation of the abBA/Nab Spectrometer: Progress Report

Emil Frleˇ z for the abBA/Nab Collaboration

frlez@virginia.edu

University of Virginia, Charlottesville abBA/Nab/Panda “Common Magnet” Meeting North Carolina State University, Raleigh, NC January 8, 2006

– p. 1/12

SLIDE 2

Choice of Simulation Software

– p. 2/12

SLIDE 3

Choice of Simulation Software

Mathematica, Penelope, Simion, GEANT3, GEANT4 . . . ?

– p. 2/12

SLIDE 4

Choice of Simulation Software

Mathematica, Penelope, Simion, GEANT3, GEANT4 . . . ?
State-of-the-art object-oriented toolkit written in C++ for

the simulation of the passage of particles through matter

– p. 2/12

SLIDE 5

Choice of Simulation Software

Mathematica, Penelope, Simion, GEANT3, GEANT4 . . . ?
State-of-the-art object-oriented toolkit written in C++ for

the simulation of the passage of particles through matter

A world-wide collaboration of institutes, experiments, and

national organizations contributing resources to the GEANT4 production service and providing mutual support

– p. 2/12

SLIDE 6

Choice of Simulation Software

Mathematica, Penelope, Simion, GEANT3, GEANT4 . . . ?
State-of-the-art object-oriented toolkit written in C++ for

the simulation of the passage of particles through matter

A world-wide collaboration of institutes, experiments, and

national organizations contributing resources to the GEANT4 production service and providing mutual support

Extensive documentation, user manuals, user forum, problem

reporting and user support, workshops, video presentations

– p. 2/12

SLIDE 7

Choice of Simulation Software

Mathematica, Penelope, Simion, GEANT3, GEANT4 . . . ?
State-of-the-art object-oriented toolkit written in C++ for

the simulation of the passage of particles through matter

A world-wide collaboration of institutes, experiments, and

national organizations contributing resources to the GEANT4 production service and providing mutual support

Extensive documentation, user manuals, user forum, problem

reporting and user support, workshops, video presentations

G4: work in progress, while GEANT3/PAW support is

discontinued

– p. 2/12

SLIDE 8

Choice of Simulation Software

Mathematica, Penelope, Simion, GEANT3, GEANT4 . . . ?
State-of-the-art object-oriented toolkit written in C++ for

the simulation of the passage of particles through matter

A world-wide collaboration of institutes, experiments, and

national organizations contributing resources to the GEANT4 production service and providing mutual support

Extensive documentation, user manuals, user forum, problem

reporting and user support, workshops, video presentations

G4: work in progress, while GEANT3/PAW support is

discontinued

We used GEANT4 version 6.2.p01 (free ;8-)

– p. 2/12

SLIDE 9

General Code Layout

– p. 3/12

SLIDE 10

General Code Layout

GEANT4 version 6.2.p01

– p. 3/12

SLIDE 11

General Code Layout

GEANT4 version 6.2.p01
User code written in C++

– p. 3/12

SLIDE 12

General Code Layout

GEANT4 version 6.2.p01
User code written in C++
Installation from UVa http server:

http://dirac.phys.virginia.edu/neutron/G4.tar.gz, size 1.8M

– p. 3/12

SLIDE 13

General Code Layout

GEANT4 version 6.2.p01
User code written in C++
Installation from UVa http server:

http://dirac.phys.virginia.edu/neutron/G4.tar.gz, size 1.8M

Modular user code, contains ∼125 files

– p. 3/12

SLIDE 14

General Code Layout

GEANT4 version 6.2.p01
User code written in C++
Installation from UVa http server:

http://dirac.phys.virginia.edu/neutron/G4.tar.gz, size 1.8M

Modular user code, contains ∼125 files
New modules can be easily added by users without intimate

knowledge of over-all code structure

– p. 3/12

SLIDE 15

Spectrometer Geometry I

– p. 4/12

SLIDE 16

Spectrometer Geometry I

Coordinate system: z = neutron beam axis, x = detector axis

– p. 4/12

SLIDE 17

Spectrometer Geometry I

Coordinate system: z = neutron beam axis, x = detector axis
Sensitive detectors:
two 100×100×2 mm3 Silicon detectors

– p. 4/12

SLIDE 18

Spectrometer Geometry I

Coordinate system: z = neutron beam axis, x = detector axis
Sensitive detectors:
two 100×100×2 mm3 Silicon detectors
Passive material:
two pairs of split Helmholtz coils, transport solenoid

magnet, polarized neutron beam coils, and 4 accelerating electrodes

– p. 4/12

SLIDE 19

Spectrometer Geometry I

Coordinate system: z = neutron beam axis, x = detector axis
Sensitive detectors:
two 100×100×2 mm3 Silicon detectors
Passive material:
two pairs of split Helmholtz coils, transport solenoid

magnet, polarized neutron beam coils, and 4 accelerating electrodes

Magnetic and electric fields defined on 1 mm3

three-dimensional grid

– p. 4/12

SLIDE 20

Spectrometer Geometry I

Coordinate system: z = neutron beam axis, x = detector axis
Sensitive detectors:
two 100×100×2 mm3 Silicon detectors
Passive material:
two pairs of split Helmholtz coils, transport solenoid

magnet, polarized neutron beam coils, and 4 accelerating electrodes

Magnetic and electric fields defined on 1 mm3

three-dimensional grid

Individual detector components can be positioned or switched
ff

– p. 4/12

SLIDE 21

Spectrometer Geometry II

– p. 5/12

SLIDE 22

Spectrometer Geometry II

Geometry of magnetic spectrometer with two split pairs,

transport solenoid and polarized neutron beam coils

– p. 5/12

SLIDE 23

Electric and Magnetic Fields

– p. 6/12

SLIDE 24

Electric and Magnetic Fields

Axial and radial components of Nab’s magnetic and electric

fields used in GEANT4 charged particle transport and spin tracking

– p. 6/12

SLIDE 25

Input: Cold Neutrons Energy Spectrum

– p. 7/12

SLIDE 26

Input: Cold Neutrons Energy Spectrum

Event generator with realistic neutron energy spectrum at the

input of the FNPB neutron guide

– p. 7/12

SLIDE 27

Input: Cold Neutrons Energy Spectrum

Event generator with realistic neutron energy spectrum at the

input of the FNPB neutron guide

– Long wavelength structure is artificial – bin aliasing combined with low statistics.

– p. 7/12

SLIDE 28

e and p Tracking and n Spin Transport

– p. 8/12

SLIDE 29

e and p Tracking and n Spin Transport

Integrate 12 variables: x, y, z, px, py, pz, E, t, s, sx, sy, sz

– p. 8/12

SLIDE 30

e and p Tracking and n Spin Transport

Integrate 12 variables: x, y, z, px, py, pz, E, t, s, sx, sy, sz
Use Cash-Karp Runge-Kutta-Fehlberg 4/5 method

[ref. Numerical Recipes in C, 2nd Ed.]

– p. 8/12

SLIDE 31

e and p Tracking and n Spin Transport

Integrate 12 variables: x, y, z, px, py, pz, E, t, s, sx, sy, sz
Use Cash-Karp Runge-Kutta-Fehlberg 4/5 method

[ref. Numerical Recipes in C, 2nd Ed.]

Equations of motion in a combined electric and magnetic field

– p. 8/12

SLIDE 32

e and p Tracking and n Spin Transport

Integrate 12 variables: x, y, z, px, py, pz, E, t, s, sx, sy, sz
Use Cash-Karp Runge-Kutta-Fehlberg 4/5 method

[ref. Numerical Recipes in C, 2nd Ed.]

Equations of motion in a combined electric and magnetic field
Spin components are treated utilizing Thomas-BMT equation

– p. 8/12

SLIDE 33

e and p Tracking and n Spin Transport

Integrate 12 variables: x, y, z, px, py, pz, E, t, s, sx, sy, sz
Use Cash-Karp Runge-Kutta-Fehlberg 4/5 method

[ref. Numerical Recipes in C, 2nd Ed.]

Equations of motion in a combined electric and magnetic field
Spin components are treated utilizing Thomas-BMT equation
0.1 mm step size, processing time ∼ 0.2 sec/per event

– p. 8/12

SLIDE 34

Input to the Code

– p. 9/12

SLIDE 35

Input to the Code

Detector version

– p. 9/12

SLIDE 36

Input to the Code

Detector version
Magnetic and electric field maps

– p. 9/12

SLIDE 37

Input to the Code

Detector version
Magnetic and electric field maps
Neutron beam time structure, neutron beam (y, z) profiles,

neutron energy spectrum and neutron polarization

– p. 9/12

SLIDE 38

Input to the Code

Detector version
Magnetic and electric field maps
Neutron beam time structure, neutron beam (y, z) profiles,

neutron energy spectrum and neutron polarization

Choice of integration method, maximum tracking step size,

minimum integration step, maximum time-of-flight

– p. 9/12

SLIDE 39

Input to the Code

Detector version
Magnetic and electric field maps
Neutron beam time structure, neutron beam (y, z) profiles,

neutron energy spectrum and neutron polarization

Choice of integration method, maximum tracking step size,

minimum integration step, maximum time-of-flight

Detector energy thresholds, energy/time resolutions,

pedestals, random coincidences, and noise

– p. 9/12

SLIDE 40

Output of the Code

– p. 10/12

SLIDE 41

Output of the Code

Pre-defined HBook4 or ROOT histograms

– p. 10/12

SLIDE 42

Output of the Code

Pre-defined HBook4 or ROOT histograms
PAW Ntuples or ROOT trees digitizing individual events

– p. 10/12

SLIDE 43

Output of the Code

Pre-defined HBook4 or ROOT histograms
PAW Ntuples or ROOT trees digitizing individual events
Simulated energy depositions and (energy,time) pairs in

sensitive detectors on event-per-event basis in ASCII format

– p. 10/12

SLIDE 44

Output of the Code

Pre-defined HBook4 or ROOT histograms
PAW Ntuples or ROOT trees digitizing individual events
Simulated energy depositions and (energy,time) pairs in

sensitive detectors on event-per-event basis in ASCII format

Single event display using OpenGL or Wired

– p. 10/12

SLIDE 45

Monte Carlo Statistics Sample

– p. 11/12

SLIDE 46

Monte Carlo Statistics Sample

Required experimental statistics ∼ 5 · 109 decays

– p. 11/12

SLIDE 47

Monte Carlo Statistics Sample

Required experimental statistics ∼ 5 · 109 decays
Number of Monte Carlo events should be an order of

magnitude higher

– p. 11/12

SLIDE 48

Monte Carlo Statistics Sample

Required experimental statistics ∼ 5 · 109 decays
Number of Monte Carlo events should be an order of

magnitude higher

So far we analyzed simulation runs with 107 events executed
n a single Linux node

– p. 11/12

SLIDE 49

Monte Carlo Statistics Sample

Required experimental statistics ∼ 5 · 109 decays
Number of Monte Carlo events should be an order of

magnitude higher

So far we analyzed simulation runs with 107 events executed
n a single Linux node
Physics parameters will be extracted from comparison of data

with simulation

– p. 11/12

SLIDE 50

Monte Carlo Statistics Sample

Required experimental statistics ∼ 5 · 109 decays
Number of Monte Carlo events should be an order of

magnitude higher

So far we analyzed simulation runs with 107 events executed
n a single Linux node
Physics parameters will be extracted from comparison of data

with simulation

Practical options are

(1) to use parallelized G4 code on Linux clusters or (2) to run a code on a supercomputer

– p. 11/12

SLIDE 51

To-Do-List (Preliminary)

– p. 12/12

SLIDE 52

To-Do-List (Preliminary)

Refine EM field maps

– p. 12/12

SLIDE 53

To-Do-List (Preliminary)

Refine EM field maps
Refine the (active/passive) geometry of the detector

– p. 12/12

SLIDE 54

To-Do-List (Preliminary)

Refine EM field maps
Refine the (active/passive) geometry of the detector
Code in the adiabatic tracking of charged particles

– p. 12/12

SLIDE 55

To-Do-List (Preliminary)

Refine EM field maps
Refine the (active/passive) geometry of the detector
Code in the adiabatic tracking of charged particles
Test the parallelized code

– p. 12/12