particles and universe: particle detectorshep.fuw.edu.pl/u/zarnecki/pau16/wyklad05.pdf · particles...
TRANSCRIPT
Particles and Universe:Particle detectors
Maria Krawczyk, Aleksander Filip Zarnecki
April 12, 2016
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 1 / 49
Lecture 5
1 Introduction
2 Ionization
3 Particle detectorsGas detectorsSilicon detectorsScintillation detectorsCherenkov detectorsCalorimeters
4 Collider experiments
5 Reconstructed events
6 Event selection
7 Simulation
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 2 / 49
Introduction
Hypothesis of Luis de Broglie (1923): wave-particle duality
Diffraction on hexagonal structures:
Light Electrons
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 3 / 49
Introduction
Classical MechanicsIf we know initial positions and velocities off all constituents of the system(eg. Solar system objects), we can foresee the future state of the system(as well as determine its history).
Quantum mechanicsMotion of a particle described as a propagation of probability wave(according to quantum wave equations)
Probability of detecting particle at given point depends on the amplitudeof its wave function.
We can not be sure where the particle is until we make a measurement.Only the dedicated measurement can “probe” the wave function and giveits actual position. Without the measurement, we can only guess...
Also, we are not able to measure particle state with infinite precision -uncertainty principle
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 4 / 49
Introduction
Classical MechanicsIf we know initial positions and velocities off all constituents of the system(eg. Solar system objects), we can foresee the future state of the system(as well as determine its history).
This reasoning fails on the subatomic scales!
Wave-particle duality forces us to look for new methods to describeparticle behavior...
Quantum mechanicsMotion of a particle described as a propagation of probability wave(according to quantum wave equations)
Probability of detecting particle at given point depends on the amplitudeof its wave function.
We can not be sure where the particle is until we make a measurement.Only the dedicated measurement can “probe” the wave function and giveits actual position. Without the measurement, we can only guess...
Also, we are not able to measure particle state with infinite precision -uncertainty principle
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 4 / 49
Introduction
Classical MechanicsIf we know initial positions and velocities off all constituents of the system(eg. Solar system objects), we can foresee the future state of the system(as well as determine its history).
Quantum mechanicsMotion of a particle described as a propagation of probability wave(according to quantum wave equations)
Probability of detecting particle at given point depends on the amplitudeof its wave function.
We can not be sure where the particle is until we make a measurement.Only the dedicated measurement can “probe” the wave function and giveits actual position. Without the measurement, we can only guess...
Also, we are not able to measure particle state with infinite precision -uncertainty principle
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 4 / 49
Introduction
Classical MechanicsIf we know initial positions and velocities off all constituents of the system(eg. Solar system objects), we can foresee the future state of the system(as well as determine its history).
Quantum mechanicsMotion of a particle described as a propagation of probability wave(according to quantum wave equations)
Probability of detecting particle at given point depends on the amplitudeof its wave function.
We can not be sure where the particle is until we make a measurement.Only the dedicated measurement can “probe” the wave function and giveits actual position. Without the measurement, we can only guess...
Also, we are not able to measure particle state with infinite precision -uncertainty principle
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 4 / 49
Introduction
Classical MechanicsIf we know initial positions and velocities off all constituents of the system(eg. Solar system objects), we can foresee the future state of the system(as well as determine its history).
Quantum mechanicsMotion of a particle described as a propagation of probability wave(according to quantum wave equations)
Probability of detecting particle at given point depends on the amplitudeof its wave function.
We can not be sure where the particle is until we make a measurement.Only the dedicated measurement can “probe” the wave function and giveits actual position. Without the measurement, we can only guess...
Also, we are not able to measure particle state with infinite precision -uncertainty principle
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 4 / 49
Introduction
Classical MechanicsIf we know initial positions and velocities off all constituents of the system(eg. Solar system objects), we can foresee the future state of the system(as well as determine its history).
Quantum mechanicsMotion of a particle described as a propagation of probability wave(according to quantum wave equations)
Probability of detecting particle at given point depends on the amplitudeof its wave function.
We can not be sure where the particle is until we make a measurement.Only the dedicated measurement can “probe” the wave function and giveits actual position. Without the measurement, we can only guess...
Also, we are not able to measure particle state with infinite precision -uncertainty principle
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 4 / 49
Introduction
Particle detection
In the macroscopic world, we are ableto make observations which do notinterfere with the process under study
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 5 / 49
Introduction
Particle detection
In the macroscopic world, we are ableto make observations which do notinterfere with the process under study
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 5 / 49
Introduction
Particle detection
In the macroscopic world, we are ableto make observations which do notinterfere with the process under study
In the particle world, each measurement is related to some interaction.We are not able to “see” particles without changing their state!
We can not observe particles which do not interact
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 5 / 49
Introduction
Particle detection
In the macroscopic world, we are ableto make observations which do notinterfere with the process under study
In the particle world, each measurement is related to some interaction.We are not able to “see” particles without changing their state!
We can not observe particles which do not interact
Main processes used for particle detection:
ionization and scintillation
photoelectric effect
Cherenkov radiation
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 5 / 49
Ionization
Structure of matterProperties of different materials depend on the strength of valenceelectrons bonding with atomic cores.
InsulatorAll electrons tightly bonded with atoms
ConductorValence electrons can move freely
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 6 / 49
Ionization
IonizationIs the phenomena used in most of the particle detectors
Charged particle passing through the insulator interacts with valenceelectrons and passes part of its energy to them, sufficient to “liberate”them from their atoms. Free charge carriers are created
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 7 / 49
Gas detectors
Ionization in gases is very small, of the order of 100 e/cm.Measurement of the corresponding current is not possible, unless we applyelectric field strong enough to create an electron avalanche.
In the uniform field we are not ableto obtain large multiplication factors.
Strong electric field resulting in largemultiplication can be easily obtainedaround the thin anod wire.
Detected charge is still very low, but can be measured with sensitiveelectronics. Charge multiplication is crucial...
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 8 / 49
Gas detectors
Ionization in gases is very small, of the order of 100 e/cm.Measurement of the corresponding current is not possible, unless we applyelectric field strong enough to create an electron avalanche.
In the uniform field we are not ableto obtain large multiplication factors.
Strong electric field resulting in largemultiplication can be easily obtainedaround the thin anod wire.
Detected charge is still very low, but can be measured with sensitiveelectronics. Charge multiplication is crucial...
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 8 / 49
Gas detectors
Geiger-Muller counterElectrons accelerated in strong electric field (resulting from high voltageapplied), can create secondary ionization when scattering off atoms. Athighest voltages charge multiplication can lead to almost full ionization ofthe gas near the wire surface (Geiger-Muller mode).
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 9 / 49
Gas detectors
Multiwire proportional chamber (MWPC)
Georges Charpak 1970(Nobel 1992)
Cheap!Electronic readout possible!electronics+computers⇒ revolution
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 10 / 49
Gas detectors
TPCTimeProjectionChamber
Heavy Ioncollision event
STAR detectorat RHIC (BNL)
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 11 / 49
Semiconductor detectors
Silicon is much denser than gas.About 100 electron-hole pairs are created in 1µm.Charge multiplication not needed, direct charge measurement possible.
Large semiconductor crystalscan be used for energymeasurement
Very precise position measurement possiblewith proper sensor segmentation
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 12 / 49
Semiconductor detectors
Silicon is much denser than gas.About 100 electron-hole pairs are created in 1µm.Charge multiplication not needed, direct charge measurement possible.
Large semiconductor crystalscan be used for energymeasurement
Very precise position measurement possiblewith proper sensor segmentation
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 12 / 49
Silicon detectors
Pixel detectors most precise position measurementMany different technologies used, including CCD sensors(as used for digital photography)
Each CCD camera is a particle detector!
Image from astronomic CCD camera: Enlarged section:
It’s not UFO. It’s a particle...M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 13 / 49
Silicon detectors
Rapid developmentSilicon detectors give very high measurement precision.They are still relatively expensive, but prices decrease fast with technologydevelopment.
⇒ more and more widely used Single pixel sensor
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 14 / 49
Scintillation detectors
ScintillationCharged particle passing the mediumcan ionize or excite the atom.
Return of the atom to its groundstate can be accompanied by thephoton emission - scintillation
PhotonsPhotons can also interact withelectrons in atom.They can transfer all their energy tosingle electron (photoelectric effect)or only part of it (Compton effect)
In both cases, electron is “released”from atom.
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 15 / 49
Scintillation detectors
Scintillation
In some materials, atoms excited byionizing particle emit photons.
Light produced in scintillator can bemeasured with photomultiplier.Classical set-up:
No position measuremenVery good time measurement
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 16 / 49
Scintillation detectors
PhotomultiplierSingle electrons can release single electrons from photocatode.We multiply this charge by applying high voltage between subsequentdynodes. When hitting dynodes accelerated electrons produce secondaryelectrons - charge avalanche.
Single photon, if it produces the first electron (photoelectric effect) resultsin macroscopic charge.
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 17 / 49
Scintillation detectors
Pixelized Photon Detector (PPD)
Also called the Silicon Photomultiplier (SiPM).Large number (∼ 103) of avalanche photo-diodes on small (∼ 1mm2)surface - possibility of counting single photons
Parameters similar to PMT: 105 – 106 gain, response time ∼ 1nsMuch smaller! Low voltage operation! ⇒ more and more popular
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 18 / 49
Scintillation detectors
Modern detectors
Classical scintillation detectorsnot used frequently
New solution:⇐ scintillating fibres,
pixelized photon detectors ⇓
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 19 / 49
Cherenkov detectors
Cherenkov radiation
Emitted by a charged particle travelingwith speed larger than the speed of light ingiven medium.
Light emitted in a cone.When passing a thin mediumlayer distinctive annulus shapeis obtained on the screen.
Observed in water, ice, air...Very cheap technology for very large detectors!
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 20 / 49
Cherenkov detectors
Cherenkov radiation
If the particle path in medium is longer, we can use special mirrors tofocus produced Cherenkov light
Scheme Image in the detector
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 21 / 49
Cherenkov detectors
Cherenkov radiation
If the particle path in medium is longer, we can use special mirrors tofocus produced Cherenkov light
Scheme Image in the detector
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 21 / 49
Calorimeters
All detectors described so far were designed to measure position of thecharged particle ⇒ tracking detectors
To measure the particle energy, we have to force it to transfer its fullenergy to the detector ⇒ calorimeters.
Electromagnetic cascadeHigh energy electronsloose energy in bremsstrahlung
e−
γ
High energy photonsconvert to e+ e− pairs
γ
e
e
−
+
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 22 / 49
Calorimeters
Electromagnetic calorimetersHigh energy electron or photon entering the detector creates a cascade ofsecondaries, with N ∼ E particles
By counting secondaries or measuring their total track length in thedetector (total ionization) we can determine the energy of primary particle.
Uniform calorimeter
eg. scintillating cristal
Sampling calorimeter
detector layers interleaved with denseabsorber
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 23 / 49
Calorimeters
Hadronic calorimeters
Simulation of the hadroniccascade (proton energymeasurement)
Hadronic cascade muchlonger than theelectromagnetic one
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 24 / 49
Collider experiments
Layer structureModern universal detectors at particle colliders are build from manydifferent sub-components.
We arrange detectors in such a way as to obtain best measurement for allpossible particles, as well as their (partial) identification.
detectors which interact least with produced particles are placed closest tothe interaction point - gas detectors, thin silicon sensors
detectors which absorb particles are placed at largest distances frominteraction point - calorimeters, muon detectors
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 25 / 49
Collider experiments
Layer structureModern universal detectors at particle colliders are build from manydifferent sub-components.
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 25 / 49
Collider experiments
Universal detectorLayout describing most of recent and present-day experiments at colliders(LEP, HERA, Tevatron, LHC, ILC):
Starting from the center of the detector:vertex detectoras close to the beam line as possible, used to measure the exactposition of the interaction point, allows for identification of short-livedparticles (secondary vertexes)silicon pixel detectors
tracking detectorsmeasure tracks of charged particles, allows to determine theirmomentum from bending in magnetic fieldgas detectors or silicon strip detectors
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 26 / 49
Collider experiments
Universal detectorLayout describing most of recent and present-day experiments at colliders(LEP, HERA, Tevatron, LHC, ILC):
Starting from the center of the detector:vertex detectoras close to the beam line as possible, used to measure the exactposition of the interaction point, allows for identification of short-livedparticles (secondary vertexes)silicon pixel detectors
tracking detectorsmeasure tracks of charged particles, allows to determine theirmomentum from bending in magnetic fieldgas detectors or silicon strip detectors
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 26 / 49
Collider experiments
Universal detector
electromagnetic calorimeterelectron and photon energy measurementdense material absorbing EM cascade(copper, lead, tungsten)
hadron calorimeterhadron energy measurement (protons, neutrons, pions, kaons)dense material absorbing hadronic cascade; hadronic cascade is manytimes longer than EM one
muon detectorsidentify muons - only charged particles which can pass bothcalorimeters with small energy losses
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 27 / 49
Collider experiments
Universal detector
electromagnetic calorimeterelectron and photon energy measurementdense material absorbing EM cascade(copper, lead, tungsten)
hadron calorimeterhadron energy measurement (protons, neutrons, pions, kaons)dense material absorbing hadronic cascade; hadronic cascade is manytimes longer than EM one
muon detectorsidentify muons - only charged particles which can pass bothcalorimeters with small energy losses
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 27 / 49
Collider experiments
Universal detector
electromagnetic calorimeterelectron and photon energy measurementdense material absorbing EM cascade(copper, lead, tungsten)
hadron calorimeterhadron energy measurement (protons, neutrons, pions, kaons)dense material absorbing hadronic cascade; hadronic cascade is manytimes longer than EM one
muon detectorsidentify muons - only charged particles which can pass bothcalorimeters with small energy losses
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 27 / 49
Collider experiments
+−
+−
+−
γ
ν
ο
calorimetersTPCVTX e.−m. hadronic
muon
µ
, p...
n, K...
π
e
det.tracking det.
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 28 / 49
ATLAS experiment
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 29 / 49
CMS experiment
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 30 / 49
Collider experiments
Compact Muon Solenoid - CMS
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 31 / 49
Reconstructed events
CMS Schematic detector view
https://www.i2u2.org/elab/cms/event-display/
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 32 / 49
Reconstructed events
CMS Single muon event
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 32 / 49
Reconstructed events
CMS Single muon event
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 32 / 49
Reconstructed events
CMS Single muon event
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 32 / 49
Reconstructed events
CMS Single muon event
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 32 / 49
Reconstructed events
CMS Event with two muons
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 33 / 49
Reconstructed events
CMS Event with two muons
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 33 / 49
Reconstructed events
CMS Event with electron production
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 34 / 49
Reconstructed events
CMS Event with electron production
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 34 / 49
Reconstructed events
CMS Event with electron production
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 34 / 49
Reconstructed events
CMS Event with two electrons
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 35 / 49
Reconstructed events
CMS Event with two photon production (H → γγ candidate)
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 36 / 49
Reconstructed events
CMS Event with two photon production (H → γγ candidate)
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 36 / 49
Reconstructed events
CMS Four lepton event (H → ZZ → e+e−µ+µ− candidate)
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 37 / 49
Reconstructed events
CMS Four lepton event (H → ZZ → e+e−µ+µ− candidate)
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 37 / 49
Reconstructed events
CMS How do we interpret events like this?
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 38 / 49
Reconstructed events
CMS How do we interpret events like this?
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 38 / 49
Reconstructed events
CMS How do we interpret events like this?
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 38 / 49
Hadronization
When a pair of quarks isproduced in the collision:gg → qq
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 39 / 49
Hadronization
When a pair of quarks isproduced in the collision:gg → qq
Color field increases betweenquarks moving apart
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 39 / 49
Hadronization
When a pair of quarks isproduced in the collision:gg → qq
Color field increases betweenquarks moving apart
Gluons are emitted
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 39 / 49
Hadronization
When a pair of quarks isproduced in the collision:gg → qq
Color field increases betweenquarks moving apart
Gluons are emitted
Gluons convert toquark-antiquark pairs
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 39 / 49
Hadronization
When a pair of quarks isproduced in the collision:gg → qq
Color field increases betweenquarks moving apart
Gluons are emitted
Gluons convert toquark-antiquark pairs
Quarks and antiquarks form“white” hadrons
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 39 / 49
Hadronization
When a pair of quarks isproduced in the collision:gg → qq
Color field increases betweenquarks moving apart
Gluons are emitted
Gluons convert toquark-antiquark pairs
Quarks and antiquarks form“white” hadrons
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 39 / 49
Reconstructed events
CMS Event with six jet production
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 40 / 49
Collider experiments
CMS Unbalanced transverse momentum - signature of missing particle
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 41 / 49
Reconstructed events
ATLAS Event with W+ boson production (W+ → e+νe)
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 42 / 49
Event selection
Up to 50 proton pairs collide at LHCat each bunch crossing
New particles are produced at almosteach collision
About billion of interactions persecond!
We have no possibilities to storemore than about 100 events persecond!
How to select the interesting ones?
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 43 / 49
Event selection
Trigger systemSignals from the detectorcomponents are “checked”after each bunch crossing bythe dedicated electronics, socalled trigger.
Only “interesting” signals areread from the detector.
These events are then passedthrough subsequent “filters” -dedicated programs rejectingall “rubbish”
Only events looking interestingare stored to disk!
particle mass (GeV)
σ rate ev/yearLHC √s=14TeV L=1034
cm-2
s-1
barn
mb
µb
nb
pb
fb
50 100 200 500 1000 2000 5000
GHz
MHz
kHz
Hz
mHz
µHz
1
10
102
103
104
105
106
107
108
109
1010
1011
1012
1013
1014
1015
1016
LV1 input
max LV2 inputmax LV1 output
max LV2 output
σ inelastic
bb–
tt–
W
W→lνZ
Z→l+l-
ZSM
→3γ
gg→HSM
qq–→qq
–H
SM
HSM
→ZZ(*
)→4l
HSM
→γγh→γγ
tanβ=2-50Z
ARL→l
+l-
Zη→l+l-
scalar LQ
SUSY q~q~+q
~g~+g
~g~
tanβ=2, µ=mg~=m
q~
tanβ=2, µ=mg~=m
q~/2
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 44 / 49
Event selection
Trigger system
Final event selection require their very detailedanalysis. But no computer could process 40milion events per second!
Solution: multilevel trigger system!
Level 1: very fast (dedicated electronics),rejects 99.9% of rubbish
Level 2: checks basic event parameters,selects 1% for final analysis
Level 3: full even analysis and final decision
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 45 / 49
Event selection
Trigger system scheme
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 46 / 49
Simulation
We use very complicated detector systems to measure particle interactions.We use trigger systems to select events of interest.We use dedicated algorithms to reconstruct the particle properties.
How do we know how to interpret the measurements?How to determine the efficiency of event selection?How to estimate precision of energy or mass measurement?How to verify that our results are consistent with expectations?
Monte Carlo methodWe use simulation methods to model all aspects of the experiment:beam particle collision, interactions inside the detector, detector response,trigger system decision.
We produce event samples which are equivalent to the actual data.We can reconstruct them in the same way and compare⇒ verify our knowledge about physics and detector performance
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 47 / 49
Simulation
We use very complicated detector systems to measure particle interactions.We use trigger systems to select events of interest.We use dedicated algorithms to reconstruct the particle properties.
How do we know how to interpret the measurements?How to determine the efficiency of event selection?How to estimate precision of energy or mass measurement?How to verify that our results are consistent with expectations?
Monte Carlo methodWe use simulation methods to model all aspects of the experiment:beam particle collision, interactions inside the detector, detector response,trigger system decision.
We produce event samples which are equivalent to the actual data.We can reconstruct them in the same way and compare⇒ verify our knowledge about physics and detector performance
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 47 / 49
Simulation
Monte Carlo methodSimple example: how to calculate the area of an irregular figure?Defined eg. by a complicated set of mathematical formulaThere are two approaches:
We can divide the space into a largenumber of unit elements and countthe elements belonging to the figure
We can generate random points andcount those inside the figure⇒ much more efficientin large number of dimensions
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 48 / 49
Simulation
Monte Carlo methodSimple example: how to calculate the area of an irregular figure?Defined eg. by a complicated set of mathematical formulaThere are two approaches:
We can divide the space into a largenumber of unit elements and countthe elements belonging to the figure
We can generate random points andcount those inside the figure⇒ much more efficientin large number of dimensions
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 48 / 49
Simulation
Monte Carlo simulation is just an efficient way to “integrate” all ourknowledge on the observed phenomena. It allows to predict theexperimental result with arbitrary precision.
But it is not a “magic box” - predictions can be calculated only for theprocesses which are fully understood!
Examples of Monte Carlo simulation results compared with data:
[GeV]4l
m100 150 200 250
Eve
nts
/5 G
eV
0
5
10
15
20
25
30
35
40
1Ldt = 4.6 fb∫ = 7 TeV s1Ldt = 20.7 fb∫ = 8 TeV s
4l→ZZ*→H
Data 2011+ 2012
SM Higgs Boson
=124.3 GeV (fit)H m
Background Z, ZZ*
tBackground Z+jets, t
Syst.Unc.
ATLAS
M.Krawczyk, A.F.Zarnecki Particles and Universe 5 April 12, 2016 49 / 49