Top Banner
Italo-Hellenic School of Physics 2005 – Martignano June 2005 C.Roda University and INFN Pisa 1 Calorimetry at LHC Why should I want a calorimeter ? Interaction relevant for Electromagnetic Calorimeters Calorimeter characteristics: linearity and resolution The ATLAS and CMS em calorimeters: different choices Hadronic interactions and issues relevant to hadron calorimeters ATLAS and CMS hadronic calorimeters C.Roda INFN & Universita` di Pisa
110

Calorimetry at LHC

Jan 09, 2016

Download

Documents

kuniko

Why should I want a calorimeter ? Interaction relevant for Electromagnetic Calorimeters Calorimeter characteristics : linearity and resolution The ATLAS and CMS em calorimeters: different choices Hadronic interactions and issues relevant to hadron calorimeters - PowerPoint PPT Presentation
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

1

Calorimetry at LHCCalorimetry at LHC

Why should I want a calorimeter ?

Interaction relevant for Electromagnetic Calorimeters

Calorimeter characteristics: linearity and resolution

The ATLAS and CMS em calorimeters: different choices

Hadronic interactions and issues relevant to hadron calorimeters

ATLAS and CMS hadronic calorimeters

C.Roda

INFN & Universita` di Pisa

Page 2: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

2

ReferencesReferences

R. Wigmans, “Calorimetry, Energy Measurements in Particle Physics” all figures without other cited source are from this book

Priscilla B.Cushman, “Electromagnetic and Hadronic Calorimeters”

D.Prieur, “Etalonnage du calorimetre electromagnetique du detector ATLAS”, PhD Thesis

M.Diemoz, “Calorimetri elettromagnetici a cristalli per la fisica delle alte energie” Lezioni Villa Gualino 3.2.2005

U.Amaldi, “Fluctuations in Calorimetry measurements” 1981 Phys.Scr.23 409

C.W.Fabjan and F.Gianotti, “Calorimetry for particle physics”, Reviews of Modern Physics, Vol.75, October 2003

R.Wigmans et al., “On the energy measurement of hadron jets”

ATLAS & CMS TDRs

Page 3: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

3

What is a Calorimeter ?What is a Calorimeter ?

The Calorimeter concepts originates from thermodynamics: thermally isolated box containing a substance under study of which we want to measure the temperature. “Our” calorimeters also measure temperature as an energy measurement. The very basic concept is thus taken from thermodynamics but the sensitivity we need is much higher, the effect of 1 TeV (1 eV = 10-19 J) in 1 liter of water (cwater = 4.19 J g-1K-1) at 20o is:

T = E / cwater M = 1.6 10-7 / 103 4.19 = 3.9 10-7 K

the sensitivity that we need is much higher.…also calorimeter in particle and nuclear physics are invasive devices:

Calorimeters are detectors able to measure the particle energy through total absorption. The first idea to use calorimeter was originated by the need to measure not only charged particles (bending magnetic field) but also neutral particles: 0 …there were born around the 1970 …

Page 4: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

4

Wigm

ans - C

alorim

etry

Page 5: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

5

Key role in the pastKey role in the past

UA2 measurement of W → jj invariant mass before and after background subtraction, the wider is the peak the more difficult it is to see the signal on the QCD background. The sigma of the signal peak is 8 GeV of which 5 GeV are attributed to calorimeter resolution. Here the resolution is not enough to separate the W and Z peaks.

Page 6: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

6

General charcteristicsGeneral charcteristics

• Sensitivity both to neutral and charged particles;• Energy measurement precision (more or less) with E

spectrometer measurement precision with p;• Do not need magnetic field (infact it is easier without);• Shower length ln(E) thus dimension are compact;• Particle identification;• Not only E but also spatial measurement through

segmentation;• They can have a fast response, useable at high rate and

for trigger signals.

CALOE

Page 7: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

7

Key role in LHCKey role in LHC

Higgs discovery: H → H → ZZ → 4 e (particle identification against jets)

SUSY discovery: easiest event signature is given by excess of events high ET miss e high pT jets;

top mass measurement tt → WWbb → ljjbb, W → jj; precise ET miss measurement requires precise and hermitic

calorimetries. forward jet tagging …

I hope I have convinced you that there are numerous reasons why we need a calorimeter…Also we will see how some of the mentioned physics channels will be used to define the design requirements of the LHC calorimeters.

Page 8: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

8

A few concepts on Electromagnetic interactions

A few concepts on Electromagnetic interactions

Page 9: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

9

How does a EM shower forms ?How does a EM shower forms ?

e interaction with matter First issure is to understand the mean energy

deposit/interactionPhoton Electrons and Positrons

PDG 2004 PDG 2004

Page 10: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

10

What do we need to understand …What do we need to understand …

Since the cross section of these processes depends on the particle energy the relevance of each process changes as the shower develops. The cross section depends on Z of the material thus the characteristics of the signal depends strongly on the type of material we use to build the calorimeter. Now we try to better understand the relevant points of this processes for what concerns shower formation.

Page 11: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

11

Electron and positron BremsstralungElectron and positron Bremsstralung

Radiation of real photons in Coloumb field nuclei.

QED

0/0

0

Xx

Brem

eEEX

E

dx

dE 0/

00

Xx

Brem

eEEX

E

dx

dE

• mean energy loss per unit length (per gr-1cm2) proportional to energy of the particle;

• Scaling factor for high energy ele in one X0 the particle reduces its energy by 63%.

• X0 can be multiplied by the density to measure in cm

E > 100 MeV it is the most important process for energy loss for e+e-

][180 2

20 cmgr

Z

AX

Page 12: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

12

Ionization lossIonization loss

Interactions of electrons with the atoms characterized by many interactions with a small release of energy.

92.0

71024.1

610

Z

MeVE

Z

MeVE

c

c

cioniz

cBrem

c EEdx

dEE

dx

dE )()(

Material Z X0/cm Ec/MeV

Liquid Ar 18 14 37

Fe 26 1.8 22

Lead 82 0.56 7.4

Sol

Liq

Two regimes of energy loss → the border is set by the critical energy EC:

Page 13: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

13

Electrons vs photonsElectrons vs photons

There is a main difference between the interactions of electrons (and positrons) and photons with matter at high energy. Electrons loose energy but they do not disappear, photons as they interact they are destroyed.

dx

dE )(E

electrons photons

Page 14: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

14

Pair productionPair production

Interaction of photons with the field of the nucleus (or of the electrons):

nucleus → nucleus + e+ e-

Threshold: E ≥ 2 me

pairx

pairAApair

eIxI

N

A

XN

A

/0

09

7

High energy approssimation, E independent

Reduction of photon beam intensity.

Photons scaling factor is of the same order of electrons: pair = 9/7 X0 1.3 X0

Page 15: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

15

Photoelectric and Compton interactionPhotoelectric and Compton interaction

3

5

E

ZicPhotelectr

A → A+ e- e → ’ e’

E

EZCompton

ln

In both interactions secondaries do not follow the direction of the incident electron, almost no reminder of initial particle direction.

shows strong dependence on Eshell

Page 16: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

16

Direction of particles that release energy ?Direction of particles that release energy ?

Page 17: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

17

Photon cross sectionsPhoton cross sections

3

5

E

Zricphotoelect

E

EZCompton

ln

CarbonZ = 6

LeadZ = 82

ZlungBremsstrah P

article

Da

ta B

oo

k 20

04

Page 18: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

18

Page 19: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

19

In summary how the shower is formedIn summary how the shower is formed

The shower is formed through a process of particle multiplication that degrades the particle energy;

Interplay between different interaction processes depends on Z of material;

As the energy of the particles reaches very low energies eV,KeV, electrons and positrons are absorbed by the material which is “heated” by the released energy.

Page 20: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

20

Scaling of shower profile with E and X0Scaling of shower profile with E and X0

The position of the shower maximum XMaximum is approximatly described as a function of X0 – since both gamma pair and brem scale with it - and the particle initial energy by the simple formula:

00 ln t

Ec

EXX Maximum

to= - 0.5 for electrons

0.5 for photons

Page 21: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

21

Electron longitudinal shower profileElectron longitudinal shower profile

Electron longitudinal shower profile in copper

Shower maximum moves with energy as log(E)

[Wigm

ans – Text B

ook]

Page 22: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

22

Photon/electron differencePhoton/electron difference

Few photons do not interact at all

Almost no electrons do not release

[Wigm

ans – Text B

ook]

Page 23: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

23

X0 scaling is approximateX0 scaling is approximate

Shower Mx is deeper in Lead than in Aluminium: multiplication continues for longer since critical energy is lower in Lead than in Aluminum (7.4 MeV vs 43 MeV).

[Wigm

ans – Text B

ook] 10 GeV e-

Page 24: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

24

X0 scaling is approximateX0 scaling is approximate

Shower “decade” slowlier in lead than in aluminum since the total number of particle created is 3 times higher than in Aluminium.

[Wigm

ans – Text B

ook] 10 GeV e-EGS4

Page 25: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

25

Consequence on longitidunal Shower containment

Consequence on longitidunal Shower containment

Pe

rce

ntu

al s

how

er

con

tain

me

nt

More radiation lengths of U than of Al needed to absorb 95% of em showers.

Page 26: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

26

Calorimeter dimensionCalorimeter dimension

6.908.0%95 ZXX Max

Calorimeters of 25X0 allows to contain electron showers at 1% up to 300 GeV.

25X0 25-50 cm

Material needed for 95% shower containment:

Page 27: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

27

Transverse profileTransverse profile

The lateral shower development is dominated by two effects:

• multiple scattering at the early phase of the shower;

• long free path for low energy photons in Compton energy range.

The measurement of the transverse size, integrated over the full longitudinal range, is given by the Molière radius (same units as X0):

)(21 0

MeVEc

XMeVRM

On average 90% of the shower is contained in 1 RM.

Page 28: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

28

Transverse profileTransverse profile

10 GeV e- in copper

Transverse profile at various depths.

Two regimes: multiples scattering and Compton photons travelling away from the axis.

Material Z X0/cm Ec/MeV RM/cm

LAr 18 14 37 8

Fe 26 1.8 22 1.7

Lead 82 0.56 7.4 1.6

Page 29: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

29

What are the particle that deposit energyWhat are the particle that deposit energy

Fraction of energy deposited to the material by a 10 GeV electron:

The low energy particles are responsible for most of the energy deposition.

Page 30: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

30

What is the range of the particle that release energy

What is the range of the particle that release energy

Page 31: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

31

From energy deposit to signal From energy deposit to signal

From the energy deposit we have to generate the signal. Two calorimeter design possibilities:

Homogeneous: the calorimeter consists of a single material which acts both as absorber and active device that transform all e+ e- energy deposit in signal.

Sampling: absorber and active device are made of different materials and signal is generated from a sample of the total e+ e- energy deposit.

Page 32: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

32

Signal generationSignal generation

The most used techniques to generate the signal in calorimeters are:

• Cerenkov radiation from e+ e-• Scintillation signals • Ionization of the detection medium

All these tecniques are characterized by a threshold energy which is the minimum detectable energy Es.

Light collection

Charge collection

Page 33: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

33

What I need from the calorimeter What I need from the calorimeter

Linearity in a given energy range: Signal = a E

The larger the range the more difficult it is for example range @ LHC [MIP → TeV]

Signal/Energy: pC/GeV, ADC count/MeV …

The request might seem easy but many different source might spoil the calorimeter response (a).

a =

Sig

nal/E

nerg

y

Page 34: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

34

Factors that affect the linearityFactors that affect the linearitySpoiling the linearity:• Saturation effects: of electronics, of energy deposition … • Leakage (transverse or lateral)• noise, this at low side Example: PMT saturation.

Injected charge (a.u.) Injected charge (a.u.)

PM

T s

ign

al (

a.u

.)

PM

T s

ign

al/i

nje

cted

ch

arg

e

Linearity within 2%1

1.02

0.98

Page 35: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

35

What else I need from a calorimeterWhat else I need from a calorimeterResponse to monochromatic

source of energy E

Calorimeter signal

background

H good resolution

Signal = constant

integrated B →

S/B 1/

… but = f(calo)

(calo) defines the energy resolution for energy E.

m

H bad resolution

Perfect good bad

Page 36: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

36

What affects the resolutionWhat affects the resolution

Up to this moment we have described the <mean> behaviour of the calorimeter, fluctuations around this value are the sources of the calorimeter resolution.

The sources of fluctuations are various:

• Signal quantum fluctuations (i.e.: photoelectric statistics …)

• Sampling fluctuations

• Shower leakage

• Instrumental effects (i.e.: structural non-uniformity, electronic noise, light attenuation, …)

Usually in each calorimeter, and in each energy range, one of these sources dominates.

Page 37: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

37

Simple model: a particle of energy E will produce N signal quanta:

N E/Ec

N is the number of e+ e- that realese energy by ionization and excitation. The signal S is proportional to the total track length (T) :

T X0 E/Ec

The measured energy EM is proportional to the particle energy E:

EM= k T

Resolution and signal quanta fluctuationResolution and signal quanta fluctuation

Page 38: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

38

E

t

TTTkkT

kT

E

E TTk

M

M cos1)()(22

Resolution and signal quanta fluctuationResolution and signal quanta fluctuation

Stochastic term

Assuming (for the moment) that k = 0

EM= k T

Fluctuation of number of track segments is poissonian →

gaussian for large number of track segment

Page 39: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

39

Resolution and signal quanta fluctuationsResolution and signal quanta fluctuationsThe intrisic limit to the energy resolution is given by the maximum detectable track length which depends on the signal threshold energy:

Tdetectable = fs T fs X0 E/Ec

fs fraction of N particles over energy threshold Es. Thus:

EM = k Tdetectable k fs X0 E/Ec

sectable

ectable

M

M

fETkT

kT

E

E

ectable

111)()(

detdet

det

Low energy threshold for detecting → high fs

Page 40: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

40

Crystal calorimeters have best intrinsic limit on energy resolution

Crystal calorimeters have best intrinsic limit on energy resolution

Compare processes with different energy threshold

Scintillating crystals

/)%31(~/ GeVEE

eV~EE gaps

MeV/1010 42

Cherenkov radiators

MeV7.0~En

1s

/)%510(~/ GeVEE

GeV/2000600

/)%52(~/ GeVEE /)%3.003.0(~/ GeVEE

Real Resolution with all contributions:

Page 41: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

41

Resolution in sampling calorimeters Resolution in sampling calorimeters In sampling calorimeters there is a further contribution to

fluctuations which is due to the sampling procedure and usually dominates other stochastic fluctuations:

Absorber plates

Active mean

Electron shower in a cloud chamber with lead absorber

Rossi gave a semi-emipirical expression for the sampling fluctuations considering the fluctuation of the number of particles crossing “a set of active layers equally spaced at distance x”:

Emip

EN

Emip = energy lost by a mip on a

sampling layer (Active + absorber)

Page 42: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

42

Resolution in sampling calorimeters Resolution in sampling calorimeters

EEmip

E

E

M

M 1)(

The higher the number of planes the smaller the Emip → the better the energy resolution

However this is clearly only part of the story …

Page 43: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

43

Sampling fluctuationsSampling fluctuationsThe previous formula however fails to describe the correct dependence of the resolution with the active layer thickness … it goes in the opposite direction.

Page 44: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

44

Sampling fluctuationsSampling fluctuations

We have seen that the calorimeter signal is given by many low energetic ( MeV) e+ and e-:

• e+ e- created in active layers

• e+e- created in absorber that reach the active layers

• the pathlength of particles with E 1MeV is fraction of the distance between active layers thus increasing the number of boundary surfaces between layers increases the contribution to the signal

The fluctuations depend on:

Sampling fraction

Sampling frequency

)()(

)(

absorberEactiveE

activeEf

mipmip

mipsamp

dd/2

Page 45: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

45

Sampling fluctuationsSampling fluctuations

Ef

da

E

E

sampM

M 1)(

fsamp ↓ resolution

↓d ↓ resolution

Page 46: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

46

Other contribution to energy resolutionOther contribution to energy resolution

Many other sources affect the energy resolution which can be parametrized as the sum of three terms added in quadrature assuming independent sources:

c

E

b

E

a

E

E

a = stochastic term, fluctuations in signal quanta

b = noise term (Stot = Sparticle + Snoise): electronic noise but also contribution from pile-up

c = smearing of the calorimeter response due any structure non uniformity that cause variation in the signal generation, non hermetic coverage (cracks)

Page 47: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

47

Resolution constant termResolution constant term

It is the leading term at high energies. It is affected by non uniform response of the detector as a function of the impact point position (equalization), temperature… It is mainly related to the precision and stability of setting working conditions … EM = kTdetectable where now we are considering the variation of k:

...)(

kE

E k

M

M

Very hard work to have a low constant term in order not to spoil resolution at high energy expecially if the stochastic term is low….

Page 48: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

48

Resolution and shower leakageResolution and shower leakage These fluctuations are non poissonian since are due to

fluctuations in numbers of interactions in first calo layer …and increase with ln(E) lateral shower leakage much less fluctuating the

longitudinal one Usefull parametrization for longitudinal fraction energy

lost f < 10%:

)f50f41(EE

2

L

i.e. for f = 5% → 13% degradation in energy resolution

Page 49: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

49

Resolution and shower leakageResolution and shower leakageC

HA

RM

Co

llabo

ratio

nN

IM 1

980

17

8,2

7

Percentual energy loss

Longitudinal dominated by first interaction, lateral by fluctuations of many low energy particles

pair = 9/7X0

Page 50: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

50

Which is the source I should take care of …Which is the source I should take care of …

Es.: ATLAS EM barrel Calorimeter

0.7%

Page 51: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

51

A Calorimeter for the Large Hadron ColliderA Calorimeter for the Large Hadron Collider

1034

1033

<1032

Lumi

cm-2s-1

100

10

0.3

Int. Lumi/y

fb-1

14 LHC(high lumi)

14 LHC(low lumi)

1.8TeVatron Run I

ECM

TeV

Page 52: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

52

Good and Bad at LHCGood and Bad at LHC

1110

Page 53: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

53

Minimum bias e pileup per bunchMinimum bias e pileup per bunch

tot(pp)100 mb

10 102 103 104

Centre-of-mass energy (GeV)

tot (pp) and inel = tot- el - diff

@ LHC inel 70 mb

Pileup:<n> = inel x L x t = 70 mb x 1034 cm-2s-1 x 25 ns 20 interactions/BC

Big change with respect to previous machines:LEP: t = 22 s <n> << 1SppS: t = 3.3 s <n> 3HERA: t = 96 ns <n> << 1Tevatron : t = 3.5 s <n> << 1Tev RunII: t = 0.4 s <n> 2

Page 54: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

54

Minimum bias characteristics Minimum bias characteristics

<pT> ~ 500 MeV

pp inelastic events at s = 14 TeV Roughly speaking at high Roughly speaking at high luminosity:luminosity:

dncharged/d dnneutral/d 7.5 in = 1

+- <pT> 0.6 GeV

da 0 <pT> 0.3 GeV

pseudo rapidity

Calorimeter acceptance –5 < < 5 ( = 0.8o):

most energy is lost down the beam pipe

~ 1100 GeV transverse energy (~ 3000 particles) in the calorimeters every 25 ns

Nch

/

= 1

E(T

eV)\

= 1

Page 55: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

55

A calorimeter for this environmentA calorimeter for this environment

ATLAS and CMS have been designed to:Minimize the pile-up in:Minimize the pile-up in: time: fast detector with a time response compatible with the bunch crossing distance 25/50 ns space: high granularity thus high number of channels Radiation resistance: Radiation resistance: appropriate tecnique for each rapidity range. Measurement of neutrinos: Measurement of neutrinos: high ermeticity Use of performance on important channels Use of performance on important channels to define the reuirement on calorimeter performance. For EM calorimeters H→ …

Page 56: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

56

Performance for Em calorimeters: H→Performance for Em calorimeters: H→

2/2

1

2

2

1

1

tgEEmm

c

E

b

E

a

E

E

Natural width: for MH 100 GeV → H /MH ≤ 10-3

Experimental width of m = 2 E1 E2 (1 - cos ) :

E

mrad 50)(

Same for ATLAS and CMS …

Page 57: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

57

ATLAS and CMS calorimeter systems are completely different

ATLAS and CMS calorimeter systems are completely different

Solenoidal inner section +

Toroidal outer section

Solenoidal field up to muon spectrometerPictures approximately to scale …CMS requires

a compact detector

Page 58: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

58

Very similar requirements but …Very similar requirements but …

ATLAS and CMS makes different choices: ATLAS require segmented calorimeter to have redudant mesurement of angle

CMS relies on vertex reconstruction from tracking and point to homogenous calorimeter with very low stochastic term aiming for excellent energy resolution.

Page 59: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

59

ATLAS and CMS electromagnetic calorimetersATLAS and CMS electromagnetic calorimeters

• Compact • Excellent energy resolution• Fast • High granularity• Radiation resistance• E range MIP → TeV

Homogeneous calorimeter made of 75000 PbW04

scintillating crystals

• good energy resolution• Fast • High granularity• Longitudinally segmented• Radiation resistance• E range MIP → TeV

Sampling LAr-Pb, 3 Longitudinal layers + PS

Page 60: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

60

CMS choice: crystal calorimeterCMS choice: crystal calorimeter

Compact Transverse segmentation

Material X0/cm Ec/MeV RM/cm

Fe 1.8 22 1.7

Lead 0.56 7.4 1.6

PbWO4 0.89 2.2

Crystal dimensions:

longitudinal 25 X0 = 22.2 cm

Transverse 1 RM = 2.2 cm

95% of the shower contained in 2 RM

Module type 2 - Rome

Page 61: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

61

CMS electromagnetic calorimeter: fastCMS electromagnetic calorimeter: fast

Conduction band

valence band

bandgap

E, T

Slow component is induced by defects and impurities.

In high quality crystals 80% is emitted in 25 ns.

200 300 400 500 600 700

inte

nsity

(a.u

.)

wavelength (nm)

Stokes shift

Page 62: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

62

Photon detectors for CMSPhoton detectors for CMS

Example of problematics due to the readout device and to the experimental environment.

NIM

A37

8 (1

996)

410

-426

Very sensitive to magnetic field (4T) … not impossible but very much care should be taken to correct for all the effects and loss of amplification.

One of the drawback with PbWO4 is the low light yield 100 /MeV thus photon detector should amplify the light: first choice PMTs.

Page 63: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

63

Photon detectors for CMSPhoton detectors for CMS

Si-PhotoDiode

OK for B field, however … no moltiplication → very large tails from electrons going through the silicon.

Si-Avalanche PhotoDiode

OK for B field, however … x25 moltiplication and good resolution.

NIM

A37

8 (1

996)

410

-426

Page 64: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

64

The constant term in the resolutionThe constant term in the resolution

Many sources to be kept under control:• Longitudinal uniformity of light collection• Strong light yield variation with temperature (-

2.3%/0C)• APD gain variation with applied tension

(-3%/Volt) and termperature (-2.3%/0C)• Light collection uniformity• light transmission due to radiation damageOther terms:Leakage front and rear

Page 65: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

65

CMS EM ResolutionCMS EM Resolution

E

MeV

E

129%40.0

%93.2

Resolution as a function of energy from test on beam:• final prototype matrix

Page 66: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

66

A slice of ATLAS electromagnetic calorimeterA slice of ATLAS electromagnetic calorimeter

Sampling: accordion lead structure filled with LAr

47 cm

• Longitudinal dimension:

25 X0 = 47 cm (CMS 22 cm)

• 3 longitudinal layers

4 X0 0 rejections separation of 2 photons very fine grain in

16 X0 for shower core

2 X0 evaluation of late started showers

• Total channels 170000

Particles from collisions

Page 67: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

67

Signal formation in LArSignal formation in LAr

GerbeEM

e-

e-

e+

Plo

mb

E ~ 1kV/mm

Argonliquide

Ele

ctro

de

ions

e-

HTIphys

E

Signal is given from collection of released electrons

Drift velocity depends on electron mobility and applied field. In ATLAS :

Lar gap 2 mm

V = 2kV

HV

400 ns 16 LHC BC

LAr

Page 68: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

68

Good and bad with LArGood and bad with LAr

High number of electron-ion pair producedNo amplification neeeded of signal, low fluctuationsLiquid → Very uniform response (purification)Stability with timeMain fluctuations are due to sampling fluctuationsIntrinsically radiation hardcheap slow time response 400 ns boling temperature 87K → criogeny neededTemperature sensitivity 2% signal drop for T=1

Page 69: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

69

LAr Calorimeter in the LHC environment …LAr Calorimeter in the LHC environment …

In order to cope with pileup background the time response is shaped with a short resolving time pulse with 0 time integral:

Each “” is a bunch crossing

Signal used is only a fraction thus I need good S/N

but shaping time has 0 time integral → mean value of pileup is cancelled.

Page 70: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

70

LAr Calorimeter at LHC rates …LAr Calorimeter at LHC rates …

The “perpendicular” geometry allows to have low detector capacitance (series of electrodes) and signal close to preamplifier (low inductance) this give high S/N.

The accordion geometry allows to have this feature without very large variation of sampling fraction for perpendicular crossing particles.

Series

Parallel

Page 71: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

71

Signal in ATLAS electromagnetic calorimeterSignal in ATLAS electromagnetic calorimeter

The accordion geometry makes also the calorimeter particularly hermetic, much easier to get the signal out but also “solve” the time response problem

Signals on copper electrodes due to current induced by electrons:

2 mm gap in 500 ns

Page 72: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

72

LAr barrel EM calorimeter after insertion into the cryostat

The cryostate for the ATLAS electromagnetic calorimeter

The cryostate for the ATLAS electromagnetic calorimeter

Page 73: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

73

ATLAS EM calorimeterATLAS EM calorimeter

The constant term in the resolution is dominated by: the equalization of the electronic readout. The equalization procedure requires to know the shaping function of each cell at few percent level → equalization with an electronic control signal the non uniformity in the electric field and in the sampling fraction introduced by the accordion structure.

Page 74: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

74

Contribution to constan term in resolution ATLAS em

Contribution to constan term in resolution ATLAS em

90 GeV e-

Some contribution also come from variation of sampling fraction due to the accordion structure

Page 75: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

75

Which uniformity do I get ?Which uniformity do I get ?

Scan on a complete module with monoenergetic electrons

Scan in

Const. term 0.57% over ~ 500 spots

Page 76: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

76

Comparing the design resolution Comparing the design resolution

E (GeV)

(E

)/E

Many of us are working to make this resolutions become reality for the whole calorimeter

CMS vs ATLAS

Page 77: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

77

Comparing the resolution on prototypes modules

Comparing the resolution on prototypes modules

E (GeV)

(E

)/E

E (GeV)(

E)/

E

%7.0)(

25.0

)(

%10)(

GeVEGeVEE

E%5.0

)(

2.0

)(

%7.2)(

GeVEGeVEE

E

Page 78: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

78

Comparison of performances on H→Comparison of performances on H→

Resolution on invariant mass:

CMS 0.7 GeV

ATLAS 1.2 GeV

CMS better resolution requires a very precise control of constant term

ATLAS has better power to measure the direction → potentially a higher efficiency

… is a question of point of views …

Page 79: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

79

Two calorimeters two points of viewTwo calorimeters two points of view

CMS goes for excellent energy resolution thus points on a technique which has a very small stochastic term, also requires a very compact calorimeter for B field choice

ATLAS points to a moderate energy resolution but to a tecnique where the intrinsic uniformity is almost for free and requires also angle reconstruction and more powerful capabilities for particle id.

Page 80: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

80

Hadronic interactionsHadronic interactions

Page 81: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

81

An extra complication…An extra complication…

231

int 35 gcmAeraction

The extra complication of strong and nuclear interation makes hadron calorimeters more difficult to optimize. The performance that one can expect from an hadron calorimeter at the moment are resolution of the order of 50% - 100%/E and linearity with a few percent 5% (recall 3%-10 %/E and <1% for EM calorimeters.

The first hadron interaction is governed by:

X0,

I [

cm]

Material Z I/cm X0/cm

Fe 26 16.8 1.8

Cu 29 15.1 1.4

Page 82: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

82

Hadron interactionsHadron interactions

When a hadron interact the are two components:Electromagnetic fraction: , 0 → s and develop and electromagnetic shower same as the one we already discussed.Hadron fraction: neutrons, protons, pions. As an example in lead this energy deposit is divided in:

56% ionizing particles (2/3 spallation protons)10% neutrons (very low energy neutrons)34% invisible (mainly nuclear binding energy, few pion decays …)Average deposit from simple model (Wigman).

Page 83: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

83

The electromagnetic fractionThe electromagnetic fraction

The average electromagnetic energy fem depends on the energy of the incident particle E0:

1

0

1

k

em E

Ef

k 0.8 and E0 = average energy needed for 0 production 1 – 2 GeV. This is obtained assuming that 30% of energy goes to 0 at each interaction, the value of k is related to the track multiplicity.

Gabriel et al.

E (GeV)

Fe E0 = 0.7 GeV k = 0.8

Pb E0 = 1.3 GeV k = 0.8

Page 84: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

84

Global view of mean energy depositGlobal view of mean energy deposit

Lead

f0

MIP

Rev.Mod.Physics Vol.75 Oct.2003

Invisible

Page 85: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

85

What are the implication of this type of showerWhat are the implication of this type of shower

We have seen that there are two types of shower components: electromagnetic Ee() and hadronic Eh().

The calorimeter reponse to the two components is

different: e, h in general e > h.

Calorimeter response to hadrons: Rh = eEe+ h Eh

Ee >> EhEe <<Eh

Rh = eEe+ h Ehe > h

Page 86: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

86

What are the implication of this type of showerWhat are the implication of this type of shower

We have seen that there are two types of shower components: electromagnetic Ee() and hadronic Eh().

The calorimeter reponse to the two components is

different: e, h in general e > h.

Calorimeter response: R = eEe+ h Eh

Ee >> EhEe <<Eh

Rh = eEe+ h Ehe = h

R

Page 87: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

87

What are the implication of this type of showerWhat are the implication of this type of shower

Calorimeter response to hadrons and electrons:

h

e

h

e

e

h

Ef

R

R

e

11 0

Often indicated as e/h

eee

eehhh

ER

EER

Page 88: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

88

Non linearity due to e/hNon linearity due to e/h

Page 89: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

89

Longitudinal Hadronic shower shapeLongitudinal Hadronic shower shape

95% containment 300 GeV 8 i.e.: 85 cm U.

For electron containment of same energy 10 cm U.

Page 90: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

90

Longitudinal leakageLongitudinal leakage

Position of shower maximum: tmax ln(E)

For 98% containment of 10 GeV = 2, for 100 GeV 7

Page 91: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

91

Mean is a rough concept for hadrons …four different longitudianl profiles

Mean is a rough concept for hadrons …four different longitudianl profiles

Page 92: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

92

Transverse shower shapeTransverse shower shape

Two components: em core + non-em halo mainly non relativistic particles

95% containment 80 GeV 1.5 / 32 cm

For electrons 95% containment of same energy 3.5 cm

A factor 9 for in both directions.

Page 93: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

93

Lateral leakageLateral leakage

The em components increases and the shower gets sharper.

150 GeV

10 GeV

Page 94: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

94

Energy resolutionEnergy resolution

The energy resolution is dominated by fluctuations in: • visible energy - ultimate limit for hadronic fluctuations• em component – this is the dominant factor in calorimeter with e/h 1 as is the case for non compensated calorimeter

em component

150 GeV -

As we have seen this fluctuation induces a worse resolution as e/h is different from 1.

Page 95: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

95

Effect of e/h on hadron lineshape Effect of e/h on hadron lineshape

Page 96: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

96

Dealing with e/hDealing with e/h

The calorimeters with e/h 1 are said to be non-compensated. In order to recover linearity and to improve the hadronic resolution two possible strategies:Hardware compensation: tuning the sampling fraction, sampling frequency and the type of materials used in sampling calorimeters it is possible to enhance the response to the hadronic part of the shower thus reaching e/h = 1Software compensation exploting the longitudinal and transverse segmentation of the calorimeter it is possible to correct event by event the reconstructed energy by weighting differently em-like deposits and hadron-like deposits.ATLAS and CMS have chosen the software compensation.

Page 97: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

97

An hadronic calorimeter for LHCAn hadronic calorimeter for LHC

The same consideration made for EM requirements in LHC environment (radiation hardness, rates …) are true for Hadronic calorimeters. Calorimeters will have the main impact on performances for:

Jets : collimated particles with different energies produced by parton hadronization

ET miss: Jets are the worst reconstructed objects thus have impact on Et miss resolution

Single hadrons i.e. from tau or W decaysRequest on resolution and linearity set with benchmark channels: W → jj, top mass, sensitivity to compositness

Page 98: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

98

Response to jetsResponse to jets

Jets are composed by many low energetic particles. Very semplified model for jet composition.

100 GeV jet

R.W

igm

an

s et a

l. “On

the

en

erg

y m

ea

sure

me

nt o

f ha

dro

n je

ts”

It is very important to understand the behaviour of the calorimeter up to 1 GeV hadron to understand the performances to jets.

Page 99: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

99

The Hadronic calorimeter The Hadronic calorimeter

The hadronic calorimeter is composed by:

- EM calorimeter section (about 1 , 25X0)

- Hadronic calorimeter section

We will see that the performances on hadrons are due to both these sections.

Page 100: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

100

LAr/Cu 1.7 <|| < 3.2

4 longitudinal sections

ATLAS Hadronic sectionATLAS Hadronic section

Both hadronic and em

LAr/Cu or W 3.2 <|| < 4.9

3 longitudinal sections

Tile Calorimeter || < 1.7

Fe / Scintillator

3 longitudinal sections

Longitudinal depth about 8 -10

Different technologies to cope with higher radiation at higher eta.

Page 101: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

101

Atlas central hadronic sectionAtlas central hadronic section

Barrel

Ext. Barrel

Principle of operation of TILE

Page 102: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

102

TileCal - Sezione centraleTileCal - Sezione centrale

Tile Preassembly

TileCal modules

Page 103: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

103

CMS Hadronic sectionCMS Hadronic section

Central Hadronic || < 1.7 :

Brass/Scintillator + WLS

2 + 1 (HO) Longitudinal section

5.9 + 3.9 (|| =0)

Endcap Hadronic 1.3< || < 3 :

Brass/Scintillator + WLS

2/3 Longitudinal sections

Forward calorimeter 2.85 < < 5.19:

Ferro/fibre di quarzo

Brass has been chosen since it is a non magnetic material

COIL

Page 104: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

104

CMS Hadron CalorimeterCMS Hadron Calorimeter

Page 105: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

105

CMS and ATLASCMS and ATLAS

• The choices made for the hadronic central section by ATLAS and CMS are similar: sampling calorimeters with scintillator as active material. • In both case the dominant factor on resolution and linearity is the e/h 1

ATLAS: e/hhad 1.4 e/hem 1.5

CMS: e/hhad 1.4 e/hem 1.6

• ATLAS higher segmentation and better stochastic term gives better total resolution

Page 106: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

106

e/h at work…e/h at work…

15%

E/

p

e/e/

Ebeam (GeV)Ebeam (GeV)

Page 107: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

107

CMS Energy resolution on pionCMS Energy resolution on pion

interacting in HCAL

In HCAL or ECAL

no weigthing

o passive weighting

dynamic weighting

Effect of different e/h + no longitudinal sampling in EM

%4%101

E

E

%5%122

E

E

Page 108: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

108

ATLAS energy resolution on pionsATLAS energy resolution on pions

EE

8.1%8.1

%9.41

E

Linearita < 2%

Shown for ATLAS but

similar for CMS

NIM

A44

9(20

00)

461-

477

EM scale

Corrected with longitudinal samples

Page 109: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

109

CMS and ATLAS CalorimeterCMS and ATLAS Calorimeter

ATLAS has much higher longitudinal segmentation thus correct the hadron signal with sw compensation.

CMS has chosen a non segmented em calorimeter and a less segmented hadron calorimeter thus is more difficult to obtain sw compensation

Also the presence of the coil and calorimeter design of CMS starts with a much higher stochastic term

It should be noticed that when considering jets there are many effects related to jet reconstruction (out of cone correction, parton to jet calibration…) that affect the resolution.

Again we see a different point of view … a different bet

Page 110: Calorimetry at LHC

Italo-Hellenic School of Physics 2005 – Martignano June 2005

C.Roda University and INFN Pisa

110

ConclusionsConclusions

I hope I have given you an idea of what are the reasons behind the design and the choices of the calorimeters. I have skipped many items that are important for calorimeters:• Calorimeter calibration, how do I set the E scale …• Calorimeter monitoring, how do I keep the E scale • Effect of detector integration on calo performances• Particle ID with calorimeters• position measurement with calorimeters• Jet reconstruction and calibration• …..you can find much more than I said in the references that I listed at the beginning of the presentation.

Good luck and I wish to all of us a great discovery times …