IntroductionIntroduction toto CalorimetryCalorimetry SteveSteve BoydBoyd UniversityUniversity ofof WarwickWarwick OutlineOutline ● A short overview ● Particle shower basics ● Calorimeters Sampling Calorimeters Homogeneous Calorimeters ● Readout and DAQ ● Example systems ● Advanced Technologies Particle Flow Dual-Readout 2 CalorimetryCalorimetry -What-What isis it?it? ● A calorimeter measures the energy of an incoming particle Stopping the particle Converting the energy into something detectable Basic mechanism: electromagnetic/hadronic showers The measured output is linear to the incoming energy ● It measures the location of the energy deposit Allows “tracking” of neutrals, e.g. photons and neutrons ● A hermetic calorimetry is essential to measure “missing energy” From all particles escaping detection Neutrinos, Neutralinos and all that 3 CalorimetryCalorimetry && ParticlesParticles ● Only ~ 13 Particles actually seen by a detector Everything else is too short-lived ● Charged Hadrons π±,p±, K± Generate hadronic Showers ● Electrons & photons Generate Electromagnetic showers ● Neutral Hadrons n,KL Generate hadronic Showers ● Muons Usually only a track through the calorimeters 4 AsAs donedone inin aa collidercollider (CMS)(CMS) 5 andand fixedfixed targettarget (NuTeV)(NuTeV) 6 ParticleParticle ShowersShowers ● Calorimeters stop particles by generating particle showers ● Two basic types Electromagnetic showers Hadronic showers ● Electromagnetic Showers Driven by QED Clean and simple ● Hadronic showers Nuclear interactions and EM component Quite complicated Very difficult to model 7 EMEM InteractionInteraction withwith MatterMatter ● Electrons and photons as the main components ● Above ~ 1 GeV Electrons:Bremsstrahlung radiating off photons Photons: Pair production Increase of particles ● Below a critical energy EC Ionization dominates Shower slowly dies out ● Material dependent Density ρ Number of Protons (Z) and nucleons (A) 8 EMEM showershower basicsbasics From T. Virdee 9 EMEM DefinitionsDefinitions ● Radiation length (X ) 0 716.4A 1 X 0= ⋅ When the energy has been Z (Z+1)⋅ln(287/√Z ) ρ reduced to 1/e Characterizes the shower depth ● 610 MeV Critical Energy (EC) E = C , solid/liquid Z +1.24 Energy, where Ionization takes over 710 MeV EC , gas= ● Z+0.92 Moliere Radius (rMoliere) Radius which contains 90 % of the shower X 0 Characterizes the width of the shower r Moliere=21.2 MeV EC ● Shower Max(imum) The peak of the shower E Incoming S max=ln( ) EC 10 MaterialMaterial DependenceDependence ZZ ρ (g/cm3) XX (cm)(cm) λλ (cm)(cm) ρ (g/cm ) 0 Int C 6 2.2 19 38.1 Al 13 2.7 8.9 39.4 Fe 26 7.87 1.76 16.8 Cu 29 8.96 1.43 15.1 W 74 19.3 0.35 9.6 Pb 82 11.35 0.56 17.1 U 92 18.7 0.32 10.5 11 ShowerShower regionsregions 50 GeV electron SiW Longitudinal development Shower decays through ionisation scales with X up to Shower 0 and Compton scattering - the tail does not scale with X Max since and 0 sbrem spair are roughly energy independent 12 ShowerShower ShapesShapes 10 GeV electron 50 GeV electron SiW SiW Layer Layer 25 cm 13 EMEM ShowersShowers PicturesPictures W Fe 20 GeV electrons longitudinal shower profile 14 ShortShort SummarySummary ● EM showers dependent on density and Z ● As Z increases shower maximum shifts to greater depth Slower decay after the Shower maximum ● The typical scale of EM showers is mm A EM Calorimeter is not a very thick object ● Location of Shower max scales with ln(E) ● Allows to build compact calorimeters ● Electrons interact right away. Photons may travel into the calorimeter for some distance before showering. 15 HadronicHadronic ShowersShowers ● Hadronic showers are much more complex ● Incoming particle hits nucleus → secondaries electromagnetic component (from π0) + strong interaction component (from n,p, π ) ● fission ... ● knock-off ... ● Delayed photons ● Hadronic Showers are much broader extend deeper in the calorimeter have significant event-by-event fluctuations 16 HadronicHadronic ShowerShower basicsbasics From T. Virdee 1. Incoming particle hits target nucleus in a hard scatter 2. Generates a shower of hadrons - p +/- , p0 and kaons - Multiplicity scales with E and particle type - p0 gg generates an electromagnetic core - leading particle effects (protons produce less p0 than pions) 3. Hadrons now leave nucleus and cascade through calorimeter - Fast hadrons knockout lower energy protons and neutrons - Multiplicity depends on target type and energy - Low energy hadrons can generate low energy photons and neutrons from nuclear de-excitation17 HadronicHadronic showershower definitionsdefinitions ● Basic quantity is the nuclear interaction length A Analog to the radiation length λ I = N A⋅σ Total 1 Order of magnitude larger than X 0 3 λ I ∼A ● Only approximations for Shower max S max (λI )∼0.2⋅ln( E)+0.7 ● Shower fractions f as electromagnetic fraction arising EM (k−1) (largely) from p0 produced in shower. E f em=1− (E0 ) Large event-to-event fluctuations f for the strong interaction fraction E depends on material and had 0 incoming particle, k ~ 0.8 Generally fEM increases with energy Neutron gas as well 18 HadronicHadronic ShowerShower ShapesShapes 10 GeV neutron 50 GeV neutron Fe+Scintillator Fe+Scintillator Layer Layer 275 cm 19 LongitudinalLongitudinal profileprofile Long shower (x10 of EM shower) For 200 GeV hadron need about 10 interaction lengths to contain 99% of the shower energy The maximum at low depth values is due to the EM component in the shower that develops more readily due to the X0 dependence on Z compared to λI. Energy deposition in copper as a function of calorimeter depth 20 LateralLateral ProfileProfile Core of shower comes from EM component Pion shower profile ATLAS Testbeam 21 IndividualIndividual ShowersShowers 50 GeV neutrons on Fe-Scintillator Stack Large event-to-event fluctuations 22 SelectingSelecting HCALHCAL materialmaterial Z ρ (g/cm3) X (cm) λ (cm) ρ (g/cm ) 0 Int C 6 2.2 19 38.1 Al 13 2.7 8.9 39.4 Fe 26 7.87 1.76 16.8 Cu 29 8.96 1.43 15.1 W 74 19.3 0.35 9.6 Pb 82 11.35 0.56 17.1 U 92 18.7 0.32 10.5 23 MaterialsMaterials againagain 24 CalorimeterCalorimeter responseresponse Response of a calorimeter to the hadronic part of a hadron-induced shower is usually different from the response to the EM part of the shower. In general the EM response (e) is linear, but the hadronic response (h) is not e/h 1 Calorimeters for which e/h= 1 are called “compensating” calorimeters. Best response come from compensating calorimeters. e e/mip = >1 h f rel rel/mip+f p p/mip+f n n/mip+f inv inv /mip relativistic proton neutron invisible energy ionisation response response response Can tune e/h by playing around with weighting factors 25 CompensationCompensation ● As already stated, hadronic showers have electromagnetic component (e) strong interaction component (h) e/h ≠1 ● EM fraction increases with energy Non-linearities ● Event by Event fluctuations tend to be non-gaussian Affect the resolution ● What can be done ? Compensating calorimeters to achieve e/h=1 26 e/e/pp inin leadlead 27 HowHow toto compensatecompensate ?? ● Software-based Try to reweight on a shower-by-shower basis difficult ● Reduce EM-Component High Z material for filtering out photo-electrons ● Boost hadronic response mainly the neutron component ● Use of Organic (hydrogen-rich) materials have a large neutron cross- section Uranium (Nuclear fission triggered by neutrons) Dual readout calorimeters? 28 TheThe UraniumUranium questionquestion ● Depleted Uranium was en vogue for a while as absorber Several Calorimeters, e.g. ZEUS, D0 ● But compensation mainly due to EM suppression Boosting hadronic response ● The fission fragments carried lots of energy But to slow to matter ● Uranium is a nasty material Radioactive Very reactive ( grinds catch fire) Mechanical properties ● These disadvantages made it unpopular 29 ShortShort summarysummary ● Hadronic Showers are very complex ● They have two components electromagnetic strong-interaction ● Electromagnetic fraction increases with energy leads to non-linearity ● Compensation trying to achieve e/h=1 30 CalorimeterCalorimeter typestypes ● Basically there are two classes Homogeneous Calorimeters Sampling Calorimeters Photo-Detector ● Either type is extensively used for ECALs ● HCALs are almost exclusively sampling calorimeters ● Decision for either depends on application Absorber Readout 31 HomogeneousHomogeneous CalorimetersCalorimeters ● Three ways to make one Scintillating crystals lead glass (Cerenkov light) Noble gas liquids ● Either offers very good resolution ● Disadvantages no direct longitudinal shower information Crystals are expensive very non-linear for hadrons 32 Read-outRead-out ● Mostly light-based tends to be blue ● Classical Photomultiplier ● Advanced Avalanche Photo-Diodes Silicon-Photo-multipliers ECAL HCAL ● Caveat Readout electronics always at the end highly non-linear for hadrons Electronics 33 TheThe CMSCMS ECALECAL 34 TargetTarget ApplicationsApplications ● ECAL only systems ● ECAL+ HCAL e.g. B-Factories If ultimate ECAL resolution is needed generally medium energy machines e.g. H→γγ ● Good ECAL is essential ● Necessary compromise on HCAL performance no Jet physics at all ● ● Examples Examples BaBar, Belle CMS KTeV L3 35 CalibrationCalibration Crystal calorimeters are quite stable and can be largely calibrated pre-experiment Monitoring in-beam Intercalibration of crystals at the same pseudo-rapidity p0 and h invariant mass Electrons from Z->ee decays