Detection and tracking of muons in the ATLAS experiment at LHC: study for an online Z?

1
Detection and tracking of muons in the ATLAS
experiment at LHC study for an online Z?µµ
selection

• Physics program at the Large Hadron Collider
• The ATLAS experiment at the LHC
• The muon spectrometer
• MDT Operating principles
• MDT Chambers
• Tracking in the experiment
• Conclusions

• production and test
• tracking, autocalibratiom, resolution

• fast tracking and momentum measurement
• Z?µµ selection and luminosity measurement

1
2
Physics program at the Large Hadron Collider
(LHC)
The Standard Model describes accurately present
data, but

• The Higgs mechanism of electroweak symmetry
  breaking (particle masses) has to be observed
  experimentally.
• Search for Higgs boson in the mass range 114
  GeV lt mH lt 1 TeV.
• Lower limit set by direct search in previous
  experiments, upper limit set by the
• stability of the theory. Present data suggest
  mH lt 200 GeV.
• Experimental behaviour of the coupling constants
  suggest a possible unification (GUT) at an energy
  scale ?GUT 1014 1016 GeV.
• Higgs mass diverges quadratically with ?
  (naturalness problem).
• ? supersymmetric theories (MSSM)
• Search supersymmetric particles (Msusy gt 100
  GeV) and in particular study the Higgs sector in
  the MSSM

pp collider CM energy 14 TeV luminosity
1034cm-2s-1 bunch crossing period 25 ns.
The ATLAS detector has been planned to fully
exploit LHC potential.
Fabrizio Petrucci Dipartimento di Fisica
E.Amaldi - Università Roma TRE
2
3

• Higgs boson search
• Low mass range (mH lt 130 GeV)
• H ? bb BR 100
  
• b-jet tagging and invariant mass
  resolution
• H ? gg BR 10-3
  
• g energy and direction measurement
• High mass range (mH gt 130 GeV)
• H ? WW() , ZZ()
• (Z ? ee, mm, jet -
  jet )
• (W ? en, mn, jet - jet )
• m and e p , E measurement
• leptonic decay to detect signal

H(130Gev)?ZZ ? 4e
?
Higgs sector in the MSSM 5 bosons (h, A, H0,
H) A, H ? tt h, H ? bb , gg H ? ZZ ? 4l
Supersymmetric particles Unknown masses, decay
chain to the LSP Missing energy W e Z boson
production excess.

• General requirements
• Particle identification e/g jets m
  missing energy
• Leptonic decays and high transverse momentum
  particles to detect signal above background
• p , E measurement

?
Fabrizio Petrucci Dottorando XV ciclo
Università Roma TRE
3
4
ATLAS detector
4
Fabrizio Petrucci Dipartimento di Fisica
E.Amaldi - Università Roma TRE
5
Muon Spectrometer

• Requirements
• Good momentum resolution in the range 6GeV-1TeV

2) h coverage up to h2.7 3) Trigger capability
on single or double muons with programmable pt
thresholds.
Dedicated trigger chambers
Solutions Monitored Drift Tube (MDT) Cathode
Stip Chamber (CSC) precision chambers Resistive
Plate Chamber (RPC) Thin Gap Chamber (TGC)
trigger chambers
5
Fabrizio Petrucci Dipartimento di Fisica
E.Amaldi - Università Roma TRE
6
RPC
MDT
Monitored Drift Tube (MDT) Proportional drift
tubes of 3cm diameter and of variable length
(1.8-5.2 m). Assembled in 2 multilayers of 3 or 4
tubes. Internal laser alignement system. Single
point resolution 80 mm. Maximum drift time
700 ns.
Resistive Plate Chamber (RPC) ionizanition
chambers built with two resistive plates and
read-out in both coordinates with cathodic
strips. Space resolution 1 cm. Time
resolution 2 ns.
6
Fabrizio Petrucci Dipartimento di Fisica
E.Amaldi - Università Roma TRE
7
MDT (Monitored Drift Tube)
100 ep/cm produced
Aluminium tube, diameter3cm, thickness400 mm
tungsten wire, 50 mm
t dc
electrons drift time
pressurized ArCO2 gas mixture
Working conditions
Gas Mixture Argon (93, high primary ionization
density) - CO2(7) Pressure 3 bar (High
pressure reduces diffusion effects) Gas gain
2104 (HV3080V) Discriminator threshold 20
primary e (3mV/e ? 60mV)
7
Fabrizio Petrucci Dipartimento di Fisica
E.Amaldi - Università Roma TRE
8
MDT Chamber test site
Tubes are individually tested and assembled
before arriving in Roma Tre.
Cosmic-ray hodoscope in Roma TRE
Chambers equipped with gas system, HV connection,
read-out electronics and tested with cosmics
before shipping to CERN.
RPC planes
BIL chamber

• 4 tubes per multilayer
• 2144 288 tubes per chamber (270 cm)
• Total volume
• 2275 l 550 l

8
Fabrizio Petrucci Dipartimento di Fisica
E.Amaldi - Università Roma TRE
9
MDT Chamber tests
Chambers have to fulfil specific requirements
concerning mechanical precision, gas tightness,
electrical properties, noise level and uniformity
of response.

• Assembly and test sequence
• Gas distribution system
• assembly and test
• 2) Gas distribution mounted
• on the chamber
• 3) Test for gas tightness
• 4) High Voltage distribution boards
• 5) Test of the electrical properties (current
  drawn by the chamber)
• 6) Read-out electronics
• 7) Tube maps and noise level
• 8) Cosmic data analysis
• 9) Chamber response check

9
Fabrizio Petrucci Dipartimento di Fisica
E.Amaldi - Università Roma TRE
10
MDT Chamber test beam
2001 H8 test beam set-up
Muon beam at the CERN SPS p 10-180 GeV

• Systems test and systems integration
• Reduced multiple scattering
• High events rate ? large data sample in the same
  working conditions

2002 H8 test beam set-up
10
Fabrizio Petrucci Dipartimento di Fisica
E.Amaldi - Università Roma TRE
11
Tracking
1)

• 1) list of hit tubes in the event tube
  identifiers (position) and drift time (tdc
  measurement).
• 2) group aligned tubes in a multilayer to form a
  candidate track (only geometrical informations).
• 3) drift time to drift distance using the proper
  r-t relation.
• fit a line to the drift circles and eventually
  drop hits with an high contribution to the ?2.
• track points definition and track parameters
  calculation.
• Track can be extended to two multilayers

2)
3)
track segment
track point
11
Fabrizio Petrucci Dipartimento di Fisica
E.Amaldi - Università Roma TRE
12
Autocalibration finding r-t relation

• Iterative procedure.
• Straight line computed fitting drift circles
  obtained with a seed r-t relation.
• Residuals are computed.
• The mean value of residuals distribution is
  computed in different drift time slices.
• It is used as the correction to the r-t relation.

12
Fabrizio Petrucci Dipartimento di Fisica
E.Amaldi - Università Roma TRE
13
Effects due to variations of temperature,
pressure and gas composition change the r-t
relation. Different chamber can have different
r-t relations.
RM07 ml 12 RM07 ml 11 RM01 ml 12 RM01 ml 11
RM07 ml 12 RM07 ml 11 RM01 ml 12 RM01 ml 11
Time (ns)
Systematic uncertainty in r-t relation are of the
order of 10 µm
Time (ns)
13
14
Tube Resolution

• Selection of good events (single track, 8 hits,
  good c2).
• Residual for each tube and its extrapolation
  error
• are computed with the track obtained with n-1
  points.
• Residuals distribution width is given by

s(r) ? Resolution(r)2 ltextrapolation
errorgt(r)2
Resolution(r) ? s(r)2 - ltextrapolation
errorgt(r)2
14
Fabrizio Petrucci Dipartimento di Fisica
E.Amaldi - Università Roma TRE
15
Run 2011 - QBIL 6? Nominal conditions
The resolution on different layers
4 layers average resolution
resolution (mm)
resolution (mm)
Signed radius (mm)
Signed radius (mm)
15
Fabrizio Petrucci Dipartimento di Fisica
E.Amaldi - Università Roma TRE
16
Fast tracking in the spectrometer

• A fast tracking procedure in the spectrometer is
  needed for calibration purpose and detector
  response monitoring.
• Montecarlo simulation has been used
• - physic processes included multiple
  scattering, energy loss, d-ray production
• - detailed geometry, material and magnetic
  field description
• - tube response is simulated using realistic
  r-t relation, resolution and efficiency.

• MDT measure only in the banding plane (R-f
  plane) second coordinate from RPC hits to
  properly account for the magnetic field.
• Track fit in each chamber parameters of the
  segment, track points.
• Comparison in both projections of segment
  parameters to form a track.
• Fast tracking assume circular trajectory
• Look for the circle best fit to all track
  points.
• Radius of curvature and error matrix computed
• analitically.
• Fast computation (150 µs).

Outer station
Middle station
P(GeV)0.3B(Tesla)Rcurv(m)
Inner station
16
Fabrizio Petrucci Dipartimento di Fisica
E.Amaldi - Università Roma TRE
17
Fast Tracking performance
Fabrizio Petrucci Università Roma TRE e INFN
17
18
Z?µµ

• Z boson production and decay in muons is a clean
  and unambiguous signal.
• Can be used for the calibration of the detector
  response and for luminosity measurement.
• s pp?Z Bz? ll 1.8 nb
• d(s pp?Z ) 5 at the LHC energy
• (aS, parton distribution
  functions, normalization of data sets)
• Bz? ll very well known

Physics event Montecarlo generator and detector
simulation
0.1 events with both muons in the barrel
muons in the barrel
all muons
18
Fabrizio Petrucci Dipartimento di Fisica
E.Amaldi - Università Roma TRE
19
Z?µµ reconstruction

• Only muon spectrometer used
• Muon pair invariant mass to select events

19
Fabrizio Petrucci Dipartimento di Fisica
E.Amaldi - Università Roma TRE
20
Background
Muon pairs with an invariant mass close to that
of the Z boson. Main sources heavy quarks
semileptonic
decays pp?qqX?µµX (qc,b,t)
20
Fabrizio Petrucci Dipartimento di Fisica
E.Amaldi - Università Roma TRE
21
Z?µµ selection and luminosity measurement
Range around the Z peak 10 GeV (15
GeV) Selection efficiency 84 (91) ? 156 pb
(169 pb) Background contamination 1.4 pb (2.2
pb)
Ls/N Luminosity can be measured using a
process with a small theoretical error on the
production cross section. d(s pp?Z ) 5
To keep statistical uncertainty below theoretical
uncertainty at least 103 Z needed s pp?Z 160 pb
? 103 Z 6 pb-1 integrated luminosity ? 20
minutes (3 h) of data taking at nominal (low)
luminosity.
21
Fabrizio Petrucci Dipartimento di Fisica
E.Amaldi - Università Roma TRE
22
Conclusions

• MDT BIL chambers construction and test
• Setting up of the cosmic-ray hodoscope.
• Definition of the procedures for chambers
  assembly and test.
• The read-out software has been written and the
  prototype electronics has been
• exploited.
• 9 chambers produced and tested.
• Chambers performance tested both at the test
  site and at the test-beam
• showing the desired construction quality.
• - Single point resolution from 250 µm close
  to the wire down to 60 µm at the
• maximum drift distance.
• - Average single tube efficiency gt97 over
  the full drift path.
• - Autocalibration r-t relation systematics
  lower than 10 µm.

22
Fabrizio Petrucci Dipartimento di Fisica
E.Amaldi - Università Roma TRE
23
Conclusions (2)

• Fast tracking and momentum measurement method in
  the barrel spectrometer
• Mean resolution varies from 3.5 at 25 GeV to
  10 at 1 TeV.
• No bias in the momentum measurement up to 200
  GeV.
• Processing time is less than 10 ms on a 600 MHz
  processor.
• Reconstruction and selection of Z?µµ events
• About 10 of pp ? Z X ?µµ X events with
  both muons in the barrel.
• Resolution of 3 in Z mass measurement.
• Background due to heavy quarks semileptonic
  decay has been studied and
• accounts for less than 2 in Z counting.
• A statistical uncertainty of 3 can be obtained
  in 20 min. (3 h) at nominal
• (low) luminosity.

23
Fabrizio Petrucci Dipartimento di Fisica
E.Amaldi - Università Roma TRE
24
Backup slides
25
LHC parametri e condizioni di misura
Parametri di LHC Parametri di LHC
Luminosita 1034cm-2s-1
Energia nel CDM vs14 TeV
Periodo di incrocio dei fasci 25 ns
protoni per bunch 1011
numero dei bunch 3600
stot(pp) 70mb ? 109 eventi/s (25
eventi ogni incrocio dei fasci) sH 10 pb ? 10-1
eventi/s il fondo e 10 ordini di grandezza
maggiore ? fondamentale la selezione (trigger) in
impulso trasverso delle particelle
25
Fabrizio Petrucci Dipartimento di Fisica
E.Amaldi - Università Roma TRE
26
ATLAS il tracciatore interno

• Misura dellimpulso delle particelle cariche ed
  identificazione di vertici secondari
• Capacita di tracciamento fino a ?lt2.5
• Risoluzione ?pT/pT lt30 (50) per ?lt2
• 
  (2lt?lt2.5)
• Efficienza e gt 95 su tutto O per pT gt 5 GeV

MSGC (Micro Strip Gas Chamber) camere a
guadagno moderato con elettrodi di lettura
segmentati a strisce s35 µm
6 punti di precisione 36 negli straw tubes
TRT (Transition Radiation Tracker) straw tubes
con s170 µm (identificazione degli elettroni
tramite i ? generati)
SCT (SemiConductor Tracker) rivelatore al
silicio (pixel strisce) ulteriore strato
vicino al vertice per la misura di vertici
secondari. Risoluzione sul singolo punto s13 µm.
26
Fabrizio Petrucci Dipartimento di Fisica
E.Amaldi - Università Roma TRE
27
ATLAS il calorimetro
Calorimetro elettromagnetico geometria
accordion, piombo e Argon liquido (2.5 mm, 4 mm)
sE/E10/v(E(GeV))1
Calorimetro adronico a campionamento ferro e
scintillatore nel barrel (TILE)
sE/E50/v(E(GeV))3
Identificare e misurare elettroni, fotoni, getti
adronici e energia mancante (copertura fino a
?4.5, profondita 10?)
Calorimetri in avanti ad Argon liquido
Calorimetro adronico a campionamento rame e
Argon liquido nelle zone in avanti
sE/E100/v(E(GeV))10
27
Fabrizio Petrucci Dipartimento di Fisica
E.Amaldi - Università Roma TRE
28
MDT (Monitored Drift Tube)
100 ep/cm produced
Aluminium tube, diameter3cm,thickness400 mm
thick
tungsten wire, 50 mm
t dc
pressurized ArCO2 gas mixture
electrons drift time
Gas mixture Argon (high primary ionization
density) CO2 High
pressure (reduced diffusion effects)
Good resolution on single point measurement
Limits on gas gain
Small signals to the read-out electronics
Gas Mixture Argon (93) - CO2(7) Pressure 3
bar Gas gain 2104 (HV3080V) Discriminator
threshold 20 primary e (3mV/e ? 60mV)
Working conditions
28
Fabrizio Petrucci Dipartimento di Fisica
E.Amaldi - Università Roma TRE
29
DAQ and read-out electronics
Si utilizzano prototipi dellelettronica finale
per lesperimento. Il software per il DAQ e
stato sviluppato a Roma Tre.
Chamber Service Module (CSM) raccoglie dati da
18 mezzanini tramite un adattatore ed e letto da
una CPU via un bus VME.
Trigger esterno (ad esempio dal telescopio)
Mezzanini schede di front-end per la lettura di
64 tubi. Contengono un chip ASD (Amplificatore,
Shaper, Discriminatore) e un TDC
29
Fabrizio Petrucci Dipartimento di Fisica
E.Amaldi - Università Roma TRE
30
New hardware setup
data link
final mezzanine (AMT2)
jtag in
Adapter
Final mezzanine 10 K test site electronics.
jtag out
C P U
VME
One more adapterino is needed (noise source)
Fabrizio Petrucci Università Roma TRE e INFN
30
31
Drift time distribution

• Two effects take place when temperature
• increases at constant pressure and interplay
• Gas is less dense ? less charge per unit path AND
  Chamber GAIN modifications
• Drift velocity is larger

31
32
Efficiency
1) Tracks that cross the tube under analysis are
fitted excluding that tube. 2) Check the hit in
the tube - Hit not present - High
contribution to the c2
tube not efficient
Hits due to d rays can hide track hits. Effect
grows with radius.
Residuals (mm)
high C2 hits
Radius (mm)
32
Fabrizio Petrucci Dipartimento di Fisica
E.Amaldi - Università Roma TRE
33
Radius of curvature
We look for the circle which better fits all the
track points. ?2 minimization with respect to R2
instead of R. ?2 S( f(xc,yc) - Ri2 ) 2 / s2R
f(xc,yc)
(x-xc)2(y-yc)2 Impose that the first track point
(x1,y1) belongs to the track (x1-xc)2(y1-yc)2-Rc
2 0 () Use (x1,y1) as origin for other
points Xi xi - x1 Yi yi - y1 f(xc,yc) -
Ri2 Xi2 Yi2 2Xi (x1-xc) 2Yi (y1-yc)
(Ri2 Rc2) Its possible to find the point
(xc,yc) which minimize the ?2 analitically. Also
the error matrix is computable exactely. The
curvature radius is the obtained from () The
computation is fast (150 µs).
33
34
Fast tracking

• G4 spectrometer simulation
• Track segments in the single chambers.
• Second coordinate from RPC hits with a proper
  smearing (digitization not ready)
• Comparison of fitted tracks parameters to match
  tracks.
• Fast tracking circular trajectories (radius of
  curvature computation ?)

2 track segments
3 track segments
34
35
Momentum measurement
P(GeV)0.3Bl(Tesla)Rcurv(m)
Large sector
Small sector
25 GeV muons
corrections needed
Approximations not accurate expecially in small
sectors
4 1 parameters needed (no ? and no momentum
dependence)
f
35
Fabrizio Petrucci Dipartimento di Fisica
E.Amaldi - Università Roma TRE
36
Performance (II)
Fabrizio Petrucci Università Roma TRE e INFN
36
37
Resolution effects
Fabrizio Petrucci Università Roma TRE e INFN
37
38
R-t relation effect (I)
Tubes with different r-t relation. Example from
H8 test beam analysis triplet of tubes in the
same multilayer with different max drift time.
Effect simulated in digitization. Events
reconstructed using a mean r-t relation (the same
for all tubes).
Fabrizio Petrucci Università Roma TRE e INFN
38
39
R-t relation effect (II)
Fabrizio Petrucci Università Roma TRE e INFN
39
40
Trigger
Sezione durto differenziale di produzione di m
3 livelli di trigger in cascata, riduzione della
rate del fondo ed elevata efficienza per eventi
di segnale.
Criteri utilizzati Tagli in impulso trasverso,
richiesta di isolamento
40
Fabrizio Petrucci Dipartimento di Fisica
E.Amaldi - Università Roma TRE
41
Trigger µ
Schema del 1o lvl di trigger m
Calcolo dellimpulso al 2o lvl di trigger
41
Fabrizio Petrucci Dipartimento di Fisica
E.Amaldi - Università Roma TRE