R.S. Orr 2009 TRIUMF Summer Institute
Layers of Detector Systems around Collision Point
Generic Detector
Scintillatorscharged particle
photons
• Charged particle passing through matter distorts
• atom• molecule• lattice
• Relaxation of distortion – light emitted
* p A p A* A A
• Emitted photon may – excite another molecule Successive excitation & emission – different λ photons
*
*
*
* *
*
B
p p
B A
C BB
C
A
A
C
C
A
each step ~ 100 ps
whole chain ~ 1 10 ns
• Very fast – good timing resolution
R.S. Orr 2009 TRIUMF Summer Institute
• Can be extremely efficient ~ 2 /dE
MeV cmdx
2 4
collection efficiency
photomultiplier efficiency
1 /10 lost 10 /
eV cm
~ 100 photoelectrons per cm.
• Spatial resolution ~ size of scintillator ~10s cm 10
Scintillating fibres
R.S. Orr 2009 TRIUMF Summer Institute
Why are scintillators transparent to emitted light?
• Stokes shift
R.S. Orr 2009 TRIUMF Summer Institute
Gain = secondary emission factor N
dkV
number of stages
771 electron x
1110
00
1
nsG mA
50 50 mV
• Fast• Low noise• High gain
Photomultiplier
R.S. Orr 2009 TRIUMF Summer Institute
Modern Photodetectors
V
photocathode
focusing electrodes
siliconsensor
electron
• Micro-channel plates• Multi-anode pm• Hybrid pm• Visible light photon counters
R.S. Orr 2009 TRIUMF Summer Institute
Electromagnetic Showers
incident 0
0E
Converts to 2 electrons 0
2e
EE
0~ 2
Each electrons will have emitted a brems photon of energy 0
4 E
E
Now have 4 particles energy 0
4
E
Number of particles after 0t 2 tN
Average energy 0
2 t
EE t
0maxmax
2 t Ct
EE EAt the critical energy
0max
ln
ln 2 cE E
t
assume this is max depth
max 0~ lnt E
0max
c
EN
E
max 0N E
~ ~ E N E
shower grows as ln E
linear energy measurement
resolution improves with energy
R.S. Orr 2009 TRIUMF Summer Institute
Monte Carlo Simulation of an EM Shower
1
0
a btt edEE
b
dt
b
a max
11.0 ln
i
at y C
b
max 0~ lnt E
Cy E E 0.5 , 0.5 iC e
Calc max ln it y C
Put b=0.5
Get a from max
1
at
b
dE
dt
R.S. Orr 2009 TRIUMF Summer Institute
Transverse Shower Profile• Shower Broadens as it develops
• Pair
• Brems
• Compton
• Multiple Coulomb
• Moliére Radius 0S
MC
ER
E
2 421.2S eE m c MeV
• Like radiation length, Moliére radius scales for different materials
• In terms of Moliére radius, shower width is roughly independent of material - 90% of energy in
• Shower Broadens as it develops
• dense central core
• spreading with depth
2 MR
R.S. Orr 2009 TRIUMF Summer Institute
Hadronic Calorimeters
• Strong (nuclear) interaction cascade• Similar to EM shower
2 1 30 35IEM had g cm A
hadronic interaction length
• Shower length
0I
hadronic calorimeternearly always sampling calorimeters
~ 5 ~ 4 @100
~ 13 ~ 10 @1
m GeV
m TeV
• Energy Resolution
• leakage of energy
• shower fluctuations
• invisible energy loss mechanisms
R.S. Orr 2009 TRIUMF Summer Institute
Hadronic Lateral Shower Profile
Fe• Hadronic shower much broader than EM
• Mean hadronic versus MCS
Tp
0
16
1.8I cm
cm
Fe
R.S. Orr 2009 TRIUMF Summer Institute
Lateral Scaling with Material
90% containment does not scale
core scales
R.S. Orr 2009 TRIUMF Summer Institute
Calorimeter Energy Resolution
22 22
20 312 4ln
A A NAA E A
E EE E
0A
Esampling and shower fluctuations
EN
E
number of samplesenergy
energy deposited in a sampling step
E N E N E
E EN
E E
E E
E E
stochastic term 0 ~A E
R.S. Orr 2009 TRIUMF Summer Institute
Calorimeter Energy Resolution
22 22
20 312 4ln
A A NAA E A
E EE E
counting statics in sensor system
this term is usually negligible
1A
E
e.g. # of photo-electrons in PM, ion pairs in liquid argon
mean # of photo-electrons in PM per unit of incident energy
N n E
NE
N n E
n n n
1 1E
E n E
1
1A
n
2A shower leakage fluctuations – make calorimeter as deep as $$ allow
R.S. Orr 2009 TRIUMF Summer Institute
Calorimeter Energy Resolution
22 22
20 312 4ln
A A NAA E A
E EE E
noise - detector or electronics
increases with number of channels summed over
- important for low level signal- liquid argon electronics- detector capacitance
N channels with some intrinsic noise
3A
N E noise (energy)
E EN
E E
deposited energy1
Ebehaviour
3A E N
R.S. Orr 2009 TRIUMF Summer Institute
Calorimeter Energy Resolution
22 22
20 312 4ln
A A NAA E A
E EE E
inter - calibration uncertainty of channels
constant
- constant in E
• spatial in-homogeneity of detector energy response
4A
• fractional channel to channel gain uncertainty
• dead space
• temperature variation
• radiation damage
• all these influence spatial variation of effective energy response
E
k E
kE
R.S. Orr 2009 TRIUMF Summer Institute
Layers of Detector Systems around Collision Point
Generic Detector
R.S. Orr 2009 TRIUMF Summer Institute
Semi-empirical model of hadron shower development
radiation length
hadronic part
11
1E Hx
E
y
INC INCH
dEE E
dS
y eC
Cx e
electromagnetic part
0
0
0
E
H
S Sx
S Sy
interaction length
0.62 0.32ln
0.91 0.02ln
0.22
0.46
H E
H
E
E
E
C
0 0S - significant amount of material in front of calorimeter (magnet coil etc.)
R.S. Orr 2009 TRIUMF Summer Institute
More Rules of Thumb for the Hobbyist
• Shower maximum
max ~ 0.2 ln 0.7t E GeV
• 95% Longitudinal containment
95% max~ 2.5 ATTL t
0.13
ATT E GeV
• 95% Lateral containment
95% ~ 1R
• Mixtures in sampling calorimeters
active + passive material
0
0
0
1 ii
ieff
actact
act pass
crit criteff i
i iieff
crit actvis act
act critinc eff eff
f
mf
m m
f
Ef
E
R.S. Orr 2009 TRIUMF Summer Institute
Typical Calorimeter Resolutions
• Homogeneous EM (crystal, glass)
CLEO, Crystal Ball, Belle, CMS.....
0.5% 3.0%0.5%
E E
• Sampling EM (Pb/Scint, Pb/LAr)8% 15%
1%E E
CDF, ZEUS, ALEPH, ATLAS.....
• Non-compensating HAD (Fe/Scint, Fe/LAr)70% 110%
5%E E
CDF,ATLAS,H1, LEP = everyone
• Compensating HAD (DU/Scint)35%
1%E
ZEUS, HELIOS
R.S. Orr 2009 TRIUMF Summer Institute
Effect of e/h on Energy Resolution
Fe/Scint
DU/Scint
E E constE
Resolution tail
R.S. Orr 2009 TRIUMF Summer Institute
Weighting to Correct for e/h
Energy resolution does not improve
1
E1
e
h
Weight down EM part of shower
1e
h
' 1k k kE E c E
tune
R.S. Orr 2009 TRIUMF Summer Institute
Jet Energy Resolution
Fluctuations between EM and Hadronic component in jets will degrade the energy resolution for jets
Jet-jet Mass Resolution
Effects other than e/h dominate the 2-jet mass resoluton
R.S. Orr 2009 TRIUMF Summer Institute
Layers of Detector Systems around Collision Point
Generic Detector
LAr Calorimetry
SM Higgs Discovery Potential
• Five different detector technologies• Calorimetry coverage to
η 5
EM Calorimeter
WW FusionTag jets in Forward Cal
H
EM Calorimeter4H l
Hadronic & ForwardCalorimeters
&H ZZ H WW
Design Goals Technology
• EM Calorimeters ( ) and Presampler ( )
• Hadronic Endcap ( )
• Forward Calorimeter ( )
2.30 8.10
GeVmm 8
GeVmrad 40
GeV27.0
%7.0GeV%10
EEEEE r
%10
GeV%100
jets%3GeV%50
EEE
%10GeV
%100jets
EE
2.35.1
3 5
LAr Calorimeter Technology Overview
Lead/Copper-Kapton/Liquid Argon Accordion Structure
Copper/Copper-Kapton/Liquid Argon Plate Structure
Tungsten/Copper/Liquid Argon Paraxial Rod Structure
LAr and Tile CalorimetersTile barrel Tile extended barrel
LAr forward calorimeter (FCAL)
LAr hadronic end-cap (HEC)
LAr EM end-cap (EMEC)
LAr EM barrel
R.S. Orr 2009 TRIUMF Summer Institute
Electromagnetic Barrel
0 1.4
Barrel Module Schematicwith presampler
• 64 gaps /module
• 2.1 mm gap
• 2x3100 mm long
Half Barrel Assembly
• 2x16 modules
• I.R/O.R 1470/2000 mm
• 22 - 33
• 3 longitudinal samples
•
• presampler
0X
0.025 0.025 1.8
R.S. Orr 2009 TRIUMF Summer Institute
Electromagnetic Endcap1.4 3.2
• 96 gaps /module outer wheel 32 gaps/module inner wheel
• 2.8 - 0.9 mm gap outer 3.1-1.8 mm inner
• 2x8 modules
• Diam. 4000 mm
• 22 - 37
• 3 longitudinal samples
•
• Front sampling of 6
for , - strips.
0X
0.025 0.025
0X2.5
2.5 0.1 0.1
R.S. Orr 2009 TRIUMF Summer Institute
Accordion Structure
Pb Absorber
• Honeycomb spacer &• Cu/Kapton electrode
Hadronic Endcap Calorimeter (HEC)
Composed of 2 wheels per endFront wheel: 67 t 25 mm Cu platesBack wheel: 90 t 50 mm Cu plates
Rear Wheel
Wheel Assembly
Front Wheel