p - I. Physikalisches Institut B RWTH Aachen

First steps
towards a
Balloon­borne electron­positron­spectrometer
Henning Gast
I. Physikalisches Institut B
Bad Honnef, 29 August 2006
Overview
●
Motivation
●
High-altitude balloons
●
Design and projected performance
–
Tracker
–
Electromagnetic calorimeter (Ecal)
–
Time of Flight system (TOF)
–
Transition Radiation Detector (TRD)
●
Scintillating fibres
●
Introduction to software
Henning Gast •
Bad Honnef 29 August 2006
•
p 2/31
Motivation
Previous measurements of cosmic-ray positron
fraction: deviations from expected shape, but
large errors.
No primary source for positrons known: probe
for new physics.
Dark Matter Candidate: SUSY-Neutralino χ
Annihilations can occur, e.g. in the galactic
halo:
χχ → bb, W+W-,... → ... → e+e- ... (stable)
Primary source of positrons in cosmic rays!
Secondary background arises from
hadronic interactions of cosmic ray
protons and subsequent decays.
W. de Boer et al.
Atmosphere inhibits ground-based measurement: 20X0, 8I , therefore use high
altitude balloon as carrier. Advantages: Reproducibility, post-flight calibration
possible. Background by atmospheric secondaries must be considered.
Henning Gast •
Bad Honnef 29 August 2006
•
p 3/31
Cosmic ray fluxes
Proton suppression of
O(10000) needed for
measurement of the
positron fraction.
Spectra steeply
declining with energy:
Φ~E-(2.7-3.0)
Composition:
90% protons
8% helium
1% electrons
positrons, antiprotons, heavy nuclei
Henning Gast •
Bad Honnef 29 August 2006
Systematic effects
widely cancel out when
measuring positron
fraction.
•
p 4/31
Having fun with balloons
Artist's impression of a ULDB in flight
(NASA)
CREAM trajectory (NASA)
Henning Gast Ultra-long duration balloons (ULDB) under
development by NASA:
Payload
3,630 kg
Scientific Payload
1,500 kg
Altitude
138,000 ft (42 km)
Duration
up to 100 days
Record-breaking CREAM flight lasted 41d 22h.
Launch pads in Palestine, TX, Fort Sumner, NM
(test flights), McMurdo Station, Antarctica, ...
After the CREAM flight: “Payload recovery
operations are in progress.” (NASA website)
CREAM at launch (NASA)
•
Bad Honnef 29 August 2006
•
p 5/31
Detector layout
Time of Flight (TOF)
trigger system
velocity measurement
Tracker
momentum measurement
disks
Transition Radiation
Detector (TRD)
proton rejection 101-102
Electromagnetic
calorimeter (ECAL)
proton rejection 103-104
0.2 T
Rasnik system
stabilty control
100 x 6 mm
1mm Pb, 6x1 mm Scintillator
Henning Gast •
Bad Honnef 29 August 2006
•
p 6/31
Momentum measurement
Trajectory of a charged particle in a uniform
magnetic field is a helix:
p cos=0.3
GeV /c
z BR
Tm
curvature (κ=1/R) error arises from two effects:
● finite measurement resolution
 k res=


720
N 4
L'2
● multiple scattering
 k ms ≈
improve position measurement
increase detector length (bad for acceptance)
0.016GeV /c z
L p  cos 2 

L
X0
minimize material budget
addition in quadrature leads to
p
2
=  a 2ms b res
p2
p
maximize magnetic field
equidistant spacing:
 k res=

L'
2

720
N 4
concentrate
layers at centre
and sides
“sagitta” geometry:
k res=

L'
2

256
N
s=
x 1 −x 3
−x2
2
Henning Gast •
Bad Honnef 29 August 2006
•
p 7/31
Designing the tracker
purpose: momentum measurement inside magnetic field
principal choices:
silicon semiconductor detectors
delicate and expensive
time projection chamber
gas-filled detector
problematic in near-space
environment
scintillating fibres
cheap and effective
uniform tracker/ECAL R+D
low material budget
Henning Gast •
Bad Honnef 29 August 2006
•
p 8/31
Scintillating fibres: principle
Saint-Gobain Crystals
standard size 0.25 mm to 5 mm square or round cross section
emission peak
430 ~ 530 nm
decay times
~3 ns
1/e length
2~4 m
photons per MeV
~8000
radiation length
42 cm
operating temperature -20°C to +50°C
Henning Gast •
Bad Honnef 29 August 2006
•
p 9/31
Scintillating fibres: Simulation
physics of optical photons implememented
in Geant4
scintillation
● reflection, refraction, transmission at
optical boundaries
● absorption and scattering in matter
●
free parameters of the simulation
fibre geometry
● surface quality
●
number of photons:
dE
⋅G
⋅
⋅d
Mirror
scint
dx core
n  =0.2⋅0.073⋅1.6⋅8000 MeV −1⋅40 keV =7.5
n  =SiPM⋅trap⋅f
fibre with core and cladding
see talk by G. Roper
Henning Gast •
Bad Honnef 29 August 2006
•
p 10/31
Scintillating fibres: Picture
Henning Gast •
Bad Honnef 29 August 2006
•
p 11/31
Tracker layout
tracker:
62 cm
2x8 layers of staggered modules, mounted on 7cm wide,
250µm thick CF skin rings, 1cm Rohacell
12 layers measuring bending-plane
4 layers rotated to measure slope
~10% X0 (Tracker) + ~2.5% X0 (TRD)
tracker module:
2x4x128 250µm scintillating fibres
2x
100µm CF skin
1x
5 mm Rohacell
32 mm
single tracker module
tracker +
TRD
(upside down)
Henning Gast •
Bad Honnef 29 August 2006
•
p 12/31
Tracker readout scheme
A
x =
∑ x i Ai
∑ Ai
x
4x1 readout scheme
(column-wise) with
weighted cluster mean
p resolution depends
on width, not height
lower number of
channels
Henning Gast 16x1 silicon photomultiplier, strip width 380 µm
need 32x1, 250µm strip width
•
Bad Honnef 29 August 2006
•
p 13/31
Readout electronics
top view
fibres to be read out by silicon
photomultipliers
~8x2x17x256=69632 channels
~1.5 mW per channel
→ 100W power consumption
side view
front view
Henning Gast •
Bad Honnef 29 August 2006
•
p 14/31
Tracker performance
momentum resolution for 4x1 readout
from full detector simulation
protons
dominated by
multiple scattering
for low energies
curvature resolution
for high energies
Henning Gast p
=  0.1162 0.004⋅p/GeV 2
p
•
Bad Honnef 29 August 2006
•
p 15/31
Designing the Ecal
6
Limitations:
Weight
Readout channels
~250 kg
<10000
Choice of absorber material:
lead
X0/λI
ρX0
X0 in 250kg
0.0328 6.37 g/cm2 15.3
3D sampling calorimeter:
80 layers
1mm absorber
6x1mm scintillating fibres
(light guides to be included in simulation soon)
Henning Gast •
Bad Honnef 29 August 2006
•
p 16/31
Ecal shower
e+
p
dE
b 1  −bt
=E 0
t e
dt
 1
t=x/X0
1 GeV e+ ECAL shower simulated by Geant4 simulation
Henning Gast •
Bad Honnef 29 August 2006
•
p 17/31
Ecal digitization
dE/dx
3 mm
before
digitization
after
digitization
amplitude
layer number
3x3 mm array: 8100 pixel SiPM
digitization:
photons at
fibre end
Henning Gast 8100
pixels
photons
pixel array
•
Bad Honnef 29 August 2006
•
p 18/31
Ecal energy resolution
preliminary
E 

=
E
Henning Gast •
Bad Honnef 29 August 2006


2
a
b 2
E
•
p 19/31
Shower shape analysis
simulation of Ecal showers
positrons
Henning Gast protons
•
Bad Honnef 29 August 2006
•
p 20/31
Cuts on shower parameters
e+
p
bremsstrahlung
fraction
shower shape fits for parameters E0, tmax=(a-1)/b from
full simulation
ratio of Ecal
energy within
one Moliere
radius from
shower axis
angle between
reconstructed
track and shower
axis
pMC/pmeas
Henning Gast •
Bad Honnef 29 August 2006
•
p 21/31
Ecal performance: Proton rejection
Ecal proton rejection of O(10000) at ~55% electron efficiency
Henning Gast •
Bad Honnef 29 August 2006
•
p 22/31
Misreconstructed showers
“good”
positron
missed
positron
“good”
proton
misidentified
proton
TRD design
based on AMS02 TRD
2x4 layers of TRD modules
each module consists of
2cm radiator fleece (PE/PP mixture, ρ = 0.06 g/cm3)
72 µm kapton-based straw tube wall (16x)
6mm straw tube diameter, filled with Xe/CO2 (80:20)
Carbon fiber stiffeners
TRD module
physics:
Geant4 offers several TR models,
here: irregular fibre radiator with
gamma-distributed fibre diameter
Photo-absorption ionization model for
energy loss in thin material (gas tube)
Bug found in Geant4
Henning Gast •
Bad Honnef 29 August 2006
•
p 24/31
TRD performance: electrons vs protons
from reconstructed TRD track
proton MPV at 1.35 keV
AMS02 testbeam:
1.3 keV
electrons
protons
30-40 GeV
AMS02 testbeam results:
higher electron
energy depositions
from TR photons
apparent
rejection analysis
to be done
Henning Gast •
Bad Honnef 29 August 2006
•
p 25/31
TRD performance: charge separation
p/Q > 5 GeV
within arrows:
80% of 11B
3% of 10Be
10% of 12C
Be
10
B
11
C
12
Henning Gast •
Bad Honnef 29 August 2006
•
p 26/31
TOF design and principle
2x2 layers in x/y direction
each made of 8 modules:
10cm wide, 5mm high
polystyrene scintillator
geometrical optimization due
digitization:
at the moment, time of
passage of first particle used
upper TOF
upper TOF
t1
=
L
lower TOF
t2
Henning Gast •
L
=
t 2 −t 1

1
m2
1 2
p
Bad Honnef 29 August 2006
p from rigidity
measurement and
charge determination
→ m → particle ID
•
p 27/31
TOF performance
with ideal electronics resolution
electrons
protons
C-12
Henning Gast •
Bad Honnef 29 August 2006
•
p 28/31
Introduction to software
●
●
A software package for simulation, reconstruction, event display
and analysis has been developed.
Written in C++, uses GEANT4, ROOT, Qt, and some external
libraries.
parallelized by condor
simulation
reconstruction
event display
analysis I
GEANT4 core:
geometry,
tracking,
physics,
interaction,
digitization
track finding,
cleaning, fitting
event mode
ECAL mode
shower mode
fill histograms
calculate analysis
variables
documentation
analysis II
automatically
generated with
doxygen
fast access
draw plots
calculate efficiencies
ROOT files for
persistency
shower axis
and fit
TRD tracks
extended ROOT files
CVS
Henning Gast •
Bad Honnef 29 August 2006
•
p 29/31
Software: Event display
Fast event display
exists for browsing
simulation and
reconstruction result
files.
event mode:
subdetector digis
reconstructed tracks
reconstructed
shower
digis used in tracks
ECAL mode:
ECAL shower
histogram
Henning Gast •
Bad Honnef 29 August 2006
•
p 30/31
Summary
●
●
●
●
Indirect dark matter search relies on precise measurement of the
cosmic-ray positron fraction, not yet done.
Dedicated balloon-borne detector proposed, based on scintillating
fibre tracker, sampling e.m. calorimeter and transition radiation
detector for proton suppression, scintillator trigger and time-offlight system.
Software framework for simulation,
reconstruction, display and analysis
exists.
Atmospheric background under study.
(see talk by Philip von Doetinchem)
●
●
First measurements of fibre properties
have begun.
A lot of work ahead...
Henning Gast •
Bad Honnef 29 August 2006
•
p 31/31