Upload
dafydd
View
18
Download
0
Embed Size (px)
DESCRIPTION
Standard Model. Lesson #1 Measurement of the Z e W bosons properties. 3. 2. 4. 1. e -. a. a. e -. . e +. b. e +. b. Introduzione. s,t,u – le variabili invarianti di Mandelstam. p 1 = [ E , p ,0,0]; p 2 = [ E ,-p,0,0]; p 3 = [ E , p cos , p sin ,0]; - PowerPoint PPT Presentation
Citation preview
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
Lesson #1
Measurement of the Z e W bosons properties
Standard Model
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
in approssimazione di massa nulla per tutte le particelle di stato iniziale e finale (m 0, E |p| )
p1 = [E, p, 0, 0]; p2 = [E, -p, 0, 0]; p3 = [E, p cos, p sin, 0]; p4 = [E, -p cos, -p sin, 0];
s = ( p1 + p2 )2 = ( p3 + p4 )2 = 4E2; t = ( p1 - p3 )2 = ( p4 - p2 )2 = - ½ s (1 - cos); u = ( p1 - p4 )2 = ( p2 - p3 )2 = - ½ s (1 + cos);
s + t + u = m12 + m2
2 + m32 + m4
2 0 (2 variabili indipendenti).
s,t,u – le variabili invarianti di Mandelstam
e-
b
a
e+
e-
b
ae+
1
2
3
4
Introduzione
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
e++
e-
-canale “s”
e+
e+
e+
e+
canale “t”
• si chiamano processi di “canale s” quelli, come e+e- +-, in cui la particella emessa e riassorbita ( in questo caso) ha come quadrato del quadri-momento il valore s, la variabile di Mandelstam che caratterizza il processo;
• viceversa, si chiamano processi di “canale t” quelli, come e+e+ e+e+, in cui la particella scambiata ( anche in questo caso) ha come quadrato del quadri-momento il valore t;
• talvolta, il processo (ex. e+e- e+e-) è descritto da più diagrammi di Feynman, di tipo s e t; in tal caso si parla di somma di “diagrammi di tipo s” o di “tipo t”
(+ interferenza).
Canale “s” e canale “t”
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
Luminosity definition and measurement
dt
dNtL eventsproduced _)(
dttLN eventsobserved )(_
Integrated luminosity
dt
dNtL eventsobserved _)(
Efficiency (trigger + reconstruction +selection)
collisionxy
ee
fArea
NNL
[cm -2 sec -1 ]
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
Based on low angle Bhabha scattering event counting
e- e+
e+
e-
e+e- e+e-
Dominated by the photon exchange “t cannel”:
(deg)
d
d
45. 90.Region used by the luminometer: 1-10 deg
e+
e-
“s channel”
e+
e-
Luminosity measurement LEP
Bhabha Homi Jehangir, indian theoretica physicist (Bombay 1909 – monte Bianco 1966)
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
cos1
)cos1(2
)cos1(
4)cos1(2)cos1(
4
2
2
22
2
sd
d QED
e+
e-
t) s)
e+
e-
(s)-(t)
21cos 4
2 14
sd
d QED
(t)-(t)(s)-(s)
Z(s)-(s) Z(s)-(t) Z(t)-(s) Z(t)-(t) Z(s)-Z(s) Z(s)-Z(t) Z(t)-Z(t)
Z(s)e+
e-
e+
e-
Zt)
(non polarized emectrons)
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
• Coasting beams with crossing angle and beam currents
x
y
x
z
yo
Luminosity (rate) insensitive to offsets in x and z, but sensitive to offsets in y:
See S. Van der Meer, ISR-PO/68-31, June 18th, 1968
Now the trick: scan the offset while measuring the rate, over the whole non-zero range, then integrate the result:
Rate vs offset, taken from Potter in Yellow report CERN 94-01 v1
Van der Meer scan
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
The standard Model is our particle interaction theory
It’s based on the two non –abelian gauge groups:
QCD (Quantum CromoDynamics) : color symmetry group SU(3)
QEWD (Quantum ElectroweakDynamics) : symmetry group SU(2)xU(1)
We have one theory but many Monte Carlo programs because the cross section for every possible process is not a trivial calculation also starting from the “right” Lagrangian.
Standard ModelStandard Model
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
Standard Model
The QEWD Lagrangian (cfr. Halzen, Martin, “Quarks & leptons”, cap.13 - 15):
LQEWD = Lgauge + Lfermioni + LHiggs + LYukawa
=> Fermions – Vector bosons interaction term
)()(2
1 2 VDLHiggs
BBWWLgauge 4
1
4
1
2
1
LRRLGL TiYukawa
*
Lfermioni = Llept+ Lquark
=> Fermions – Scalar boson interaction term
BY
igWg
iD2
'2
42 baV
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
=(1,2,3) : Pauli matrixes
g g’ a b
v = |
Gi
2
vGm e
e
,,
___
2'
2'
2),(
elRR
L
Llept BY
igBY
igWg
iL
bsddtcuu
RR
L
Lquark uBY
igud
uB
YigW
giduL
,,,,
___
2'
2'
2),(
amH 2
b
a
2
Parameters of the model
Removing the mass of the fermions and the Higgs mass
we have only 3 residual free parameters: g g’ v
RR dB
Yigd
2'
_
is the minimum of the Higgs potential
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
2
21
WWW
2
21
WWW
22
3
'
'
gg
gBWgA
22
3
'
'
gg
BggWZ
vgMW 2
10AM
22 '2
1ggvM Z
,, 21 Small oscillation around the vacuum.
For
we neglect terms at order > 2nd
42 baV 22221
2
1
After the symmetry breaking :
2 complex Higgs fields 1 real Higgs field - 3 DOF
4 massless vector bosons 1 massless + 3 massive vector bosons + 3 DOF
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
W Weinberg angle
g g’ v 22 '
'
gg
gge
22 '
'sin
gg
gW
2
12v
GF
e Millikan experiment (ionized oil drops)
GF lifetime W Gargamelle (1973) asymmetry from scatteringp p
WF
WG
eM
2
22
sin24
W
WZ
MM
cos
electron charge Fermi constant
sin2W 0.23 ( error 10 % )
MW 80 GeV MZ 92 GeV
UA1 pp √s = 540GeV (1993) MW = 82.4±1.1 GeV MZ = 93.1±1.8 GeV
Before the W and Z discovery we had strong mass constraints:
Historical measurements:
The mass of the vector bosons are related to the parameters:
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
The Z° boson can decay in the following 5 channels with different probabilities:
p=0,20 (invisible)e- e+ p=0,0337 pv= 0,0421
Z° - + p=0,0337 pv= 0,0421- + p=0,0337 pv= 0,0421qq p=0,699 pv= 0,8738
The differences can be partially explained with the different number of quantum states:
includes the 3 different flavors: e , , qq includes the 5 different flavors: uu , dd , ss , cc , bb
for every flavor there’re 3 color states ( tt is excluded because mt >MZ)
“partially explained” : 0,20 > 3 × 0,0337 and 0,699 > 15 × 0,0337
we will see later they are depending also to the charge and the isotopic spin
Z0 boson decay (channels and branching ratios)
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
Z0 boson decay (selection criteria)
e+e- Z0 e+e-
Efficiency 96 %
Contamination 0.3 % ( + - events)
e+e- Z0 hadrons
• Charged track multiplicity 3
• E1ECAL > 30 GeV E2
ECAL > 25 GeV
• < 10 o
Efficiency 98 %
Contamination 1.0 % (+ - events)
Nucl. Physics B 367 (1991) 511-574
150.000 events(hadronic and leptonic)
collected betweenAugust 1989 - August 1990
NchFraction of √s
a) Charged track multiplicity 5
b) Energy of the event > 12 % s
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
e+e- Z0 +-
e+e- Z0 +-
• Charged track multiplicity 6
• Etot > 8 GeV , pTmissing > 0.4 GeV
• > 0.5 o
• etc.
• Charged track multiplicity = 2
• p1 e p2 > 15 GeV
• IPZ < 4.5 cm , IPR < 1.5 cm
• < 10 o
• Association tracker- muon detector
• EHCAL < 10 GeV (consistent with MIP)
• EECAL < 1 GeV (consistent with MIP)
Efficiency 99 %
Contamination: ~1.9 % (+ - events)
~1.5 % (cosmic rays)
Efficiency 70 %
Contamination: ~0.5 % (+ - events)
~0.8 % (e+e- events)
~0.5 % (qq events)
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
19 input variables:
• P and Pt of the most energetic muon
• Sum of the impact parameters
• Sphericity , Invariant mass for different jets
3 output variables:
• Probability for uds quarks
• Probability for c quarks
• Probability for b quarks
Quark flavor separation
Classification using neural network
MC uds MC c
MC b Real data
h
bbR
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
s)e+
e-
(s)-(s)
Z(s)e+
e-
Z(s)-Z(s)weak
ffff
em
ff
EW
ff
d
d
d
d
d
d
d
d
int
Z0 line shape
The line shape is the cross section (s) e+e- ff with √s values arround MZ
Can be observed for one fermion or several together (i.e. all the quarks)
_
f
ffTOT
)cos1(4
222
s
NQ
d
dCf
em
ff
f
f
f
f_ _
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
gVf = I3f - 2 Qf sin2W gAf = I3f
s
NQ
s
MsI
MMs
ss CfZ
ZZZ
ZEW
3
4
)()(
222
02222
2
Resonance term (Breit – Wigner) ( I QeQf )
)(26
223
AfVfZF
Cf ggMG
N 220
12
ZZ
fe
M
f
fZ
( terms proportional to ( mf / Mz )2 have been neglected )
The expected line shape from the theory is a Breit-Wigner function characterized by 3 parameters:
Mass (MZ) – Width (Z) – Peak cross section (0)
With unpolarized beams we must consider the mean value of the 4 different helicity
Branching ratios f / Z are related to Qf and I3f
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
We can take the values of the parameters reported in the PDG
and calculate the expected value of the partial widths f
MeVMG ZF 01.088.165)4/14/1(
26
3
3 families
MeVMG ZF
l 005.0419.83)4/10014.0(26
3
GeVGF510)00001.016637.1( GeVM Z 0021.01876.91
3 families
MeVMG ZF
cu 015.0446.285)4/10368.0(26
33
.
)(26
223
AfVfZF
Cf ggMG
N
2 families
gVf = I3f - 2 Qf sin2W gAf = I3f
MeVMG ZF
bsd 02.095.367)4/11197.0(26
33
,.
3 families
MeVZ 07.064.2422 MeVsperZ 3.22.2495
00013.023116.0sin2 W
%3
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
(s) e+e- hadrons
Born(s)
(s) experimental cross section0
Branching ratios
e+ e-
Process ff / Z (%)
B.R. exp. (%)
Neutrinos 20.54 20.00±0.06
Leptons 10.33 10.10±0.02
Hadrons 69.13 69.91±0.06
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
Radiative corrections
The radiative corrections modify the expected values at the tree level:
Z*,
Relevant impact: • decreases the peak cross section of ~30%• shift √s of the peak of ~100 MeV
QED corrections
s
Born dsssGszss0
'),'()'()(
1-z = k2/s fraction of the photon’s momentum
(1) Initial state radiation (QED ISR)
(2) Final state radiation (QED FSR)
f
fQED
fh )1(0
Z*,
4
3 2ff
QED
Q
G(s’,s) = function of the initial state radiation
Initial state radiation correction:
~ 0.17 %
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
(4) Propagator correction (vacuum polarization)
f+ n loop
2
22
22
log3
)(1
)()(
Q
Q 22fmQ
)(1)(
202
137
10
128
1)( ZM
610)88()95( GeVGeV
064.0)( ZM
s
Born dsssssHszss0
')',()',()'()(
(3) Interference between initial state radiation and final state radiation
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
EW corrections
Z/f+ n loop
)(1)(
ss
)(1
)(sincossin)(sinsin 222
s
ss Z
WWWWW
2
)(ZM
ss
)(1)(
s
GsGG
Z
FFF
(1) Propagator corrections
Photon exchange
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
QCD corrections
f
QCDf
QEDf
h )1)(1(0
%17.04
3 2
ff
QED
Q%8.3
)( 2
ZsfQCD
M
(1) Final state radiations (QCD FSR)
(2) Vertex corrections
W/fZ*, WffftVfVf QImsgg 2
3 sin2),(
fftAfAf Imsgg 3),(
Contributions from virtual top terms
Z*, g
2
2
1Z
tZf m
mM
Z*, W/f
~ 1 + 0.9 %
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
Z
ZZZ
Z
MsMs
ss
22222
2
0 )/'()'(
')'('
s
dsssssHszss0
')',()',()'(')(
Line shape
220
12
ZZ
fe
f M
)1)(1)((26
223
QCDf
QEDfAeVe
ZFCf gg
MGN
Interference EW- QED
WfffVf
ffAf
QIg
Ig
2
3
3
sin2
MZ Z 0
QED ISR QED Interference IS-IF
Photon exchange
QED FSR QCD FSR
EW vertex
Breit-Wigner modified by EW loops
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
Padova 4 Aprile 2011
Ezio Torassa
MZ
Z
0h , 0
e , 0 , 0
MZ
Z
eh , ee , e , e
%06.000.20/
%06.091.69/
%008.0367.3/
%007.0366.3/
%004.0363.3/
0023.04952.2
0021.01876.91
Z
Zh
Z
Z
Ze
Z
Z
GeV
GeVM
PDG
2010
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
Padova 4 Aprile 2011
Ezio Torassa
inv = Z – had - 3lept - 3
The number of the neutrino familiescan be added in the fit. The result is compatible only with N=3
We can also extract the width for new physics (emerging from Z decay):Z , had , lept can be measured can be estimated from the SMThe result is compatible with zero orvery small widths.
Number of neutrino families
We can suppose to have a 4th generation family with all the new fermions heavier than the Z mass (except the new neutrino) . How to check this hypothesis ?
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
Gamma-gamma interactions
The gamma-gamma interactions produce two leptons with small energy (W << Ebeam) and small angles, most of them are not detected.
At the Z0 peak the gamma-gamma cross section is about 150 nb. After the selection (pT , cuts)is reduced to a non resonant 6 nb background.
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
• QED was assumed for the ISR function H(s,s’) and the interference IF-FS function (s,s’)• QEWD was assumed for the interference QED-EW function Z(s)• QCD was assumed for the QCD FSR correction
With the cross section measurements at higher energies (s= 130-200 GeV), the interference term can estimate from the data.
Residual dependence from the model
s
MsI
MsMs
s Z
ZZZ
Z2
022222
2
)/()(
s
NQ Cf
3
4 22
)(sEW
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
LEP luminosities LEP2
From 1990 to 1994 about 170 pb-1
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
qq()
WW
ZZ
s
Center of mass energy after the initial state radiation
Relevant ISR contributions (radiative return to the Z0 peak)
Searches beyond the Standard Model have more sensitivity to the non radiative events: s /s > 0.85
ff() production at LEP2
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
Jet
Jet
|| 1p || 2p
|| p
Identification of ISR photon and s´ estimation (SPRIME)
Search of ISR candidates:
Sarch of signal inside calorimeters, luminometer included, with E>10 GeV (not associated to charged tracks)
Jet 1 and Jet 2 reconstruction
Photon inside the beam pipe hypothesys
Energy and momentum conservation:
None ISR photon detected
1221
12
sinsinsin
sin
sinsinsin
sin
s
CBA
Asp
2222 )()(' EEspEss
ISR photons are emitted at low angle, they mainly induce polar angle unbalance (R unbalance is neglected)
R
z
A
Bpp
sin
sin|||| 1
A
Cpp
sin
sin|||| 2
|||||| 21 ppps
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
One ISR photon detected
If the photon is coplanar with the two jets ( > 345o ) his direction is used.
Considering the low resolution of the calorimeters the energy momentum balance is used to estimate the energy of the photon
otherwise a second undetected photon inside the
beam pipe is considered.
1212
12
sinsinsin
sin
sp
22 )()(' hh ppEEss
ISR photon spectrum s = 130 GeV
22)(' pEss
The photon emission at the energy of s – 91 GeV increases because the process is traced back to the Z0 (cross section peak).
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
2/NDoF =160/180 for the mediated ff data at LEPII
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
ZZ production at LEP2
ZZ(s)
Test at 5% precision
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
Produzione dei bosoni W+W- e misura di Mw a LEP II
MZ e sin2W misurati a LEPI MW permette la definizione di vincoli piu’ stringenti.W+
W -
e+
e-
W+
W -
+Z*
W+
W -
+ + rad.corr.
I vertici ZWW previsti come conseguenza della strutta non abeliana della teoria di gauge SU(2)L×U(1)Y esistono
Per ricavare WW occorre distinguere il segnale WW dal fondo: ff() ff(gg)
-
Le cancellazioni previste dalla teoria di gauge sono state verificate al livello dell’ 1 %.
-
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
Decadimenti adronici
La caratteristica è la ricostruzione di 4 jets. Talvolta anche gli eventi ff possono fornire 4 jets. I jets dovuti a radiazione di gluoni sono caratterizzati da un piccolo angolo e da una bassa energia. La variabile D è in grado di discriminare permettendo la riduzione del fondo:
-
Decadimenti semileptonici
La caratteristica è la ricostruzione di 2 jets ed un leptone energetico ed isolato.
Decadimenti totalmente leptonici
La caratteristicha è la ricostruzione di 2 leptoni energetici isolati di carica opposta. L’esempio in figura riporta solo 2 tracce cariche ma il leptone puo’ anche essere un Una variabile discriminante è la direzione del momento mancante che per il fondo ff ha piccoli angoli
)( minmax
min
max
min
EEE
ED
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
MW a LEP II
MW si può considerare una grandezza derivata da GF , il cui valore è conosciuto con precisione. La relazione è semplice a livello (r=0) mentre con le correzioni di
polarizzazione del vuoto risulta dipendente da mt ed MH.
Una precisa misura della massa del W permette una ulteriore verifica del modello
ed allo stesso tempo fornisce dei limiti per le masse del top e dell’Higgs.
All’inizio di LEP II le misure dirette di mt al Tevatron (18012 GeV) avevano ancora
errori piuttosto grandi.
La massa del W può essere ottenuta:
• dall’andamento WW(MW) che in prossimità della soglia s ~160 GeV varia rapidamente
• mediante la ricostruzione della massa invariante.
)),(1(
1
)/1(2 222HtZWW
F MmrMMMG
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
Determinazione della massa da WW
W+
W -
e+
e-
W+
W -
Z*W+
W -
Diagrammi CC03
Lo stato finale in 4 fermioni, oltre ai contributi dei diagrammi CC03, può ricevere contributi da altri diagrammi elettrodeboli. Tali contributi sono stati considerati con dei termini di correzione della sezione d’urto (v. Tabella). La sezione d’urto così corretta è stata confrontata con l’andamento previsto in funzione della massa MW
Dati 1996 a 161 GeV L = 10 pb -1
WW decay mode Corr (CC03)
qqqq eqq ()qq ll
0.996 1.087 1.006 1.045
pbtotWW 19.067.3 97.0
85.0
2/09.044.040.80 cGeVmW DELPHI
Contributi, interferenza inclusa, di altri diagrammi che generano 4 fermioni mediante il coinvolgimento di 0,1 o 2 bosoni vettori massivi
Vengono selezionati 29 eventi (15 + 12 + 2) da cu si ricava:
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
Ricostruzione diretta della massa
Per s > 2 MW è possibile ricostruire direttamente la massa dai decadimenti.
La ricostruzione ottenuta solo con le tracce osservate non ha una sufficiente risoluzione. Applicando dei vincoli quali la conservazione dell’energia e dell’impulso del centro di massa la risoluzione migliora significativamente. La ricostruzione si effettua solo per i decadimenti adronici e semileptonici, per quelli totalmente leptonici non si dispone di sufficienti vincoli (rispetto ai semileptonici manca l’asse del jet che determina la direzione del W opposto).
DELPHI Dati 1998 a 189 GeV L = 150 pb -1
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
2/)(035.0)(017.0)(034.0)(087.0387.80 cGeVFSILEPsysstatmW
FSI: Final State Interaction
I due W decadono a distanze di frazioni di fm
nel caso adronico contribuiscono le interazioni dei partoni nello stato finale
FSI
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
L’FSI riduce il peso del canale adronico rispetto al canale semileptonico
Risultato combinato LEP II
2/042.0412.80 cGeVMW
Contributi agli errori statistici e sistematici
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
MZ/MZ 2.3 10-5MW/MW 5.2 10-4
Measurements since 1989
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
Discrepancy observed in 1995 not confirmed after more precise Rb measurements
hadb /
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
EWK physics at LHC
LEP 170 pb-1 at the Z0 peak e+e-~ 3 104 pb → 5 M Z0 / experiment
LHC 25 fb-1 pp~ 3 104 pb → 750 M Z0 / experiment
XXVIII Ph.D in Physics Ezio TorassaPadova, April 22th 2013
Quarks & Leptons – Francis Halzen / Alan D. Martin – Wiley International Edition
The Experimental Foundation of Particle Physics – Robert N. Cahn / Gerson Goldhaber
Cambridge University Press
Determination of Z resonance parameters and coupling from its hadronic and leptonic decays - Nucl. Physics B 367 (1991) 511-574
Z Physics at LEP I CERN 89-08 Vol 1 – Radiative corrections (p. 7) Z Line Shape (p. 89)
Measurement of the lineshape of the Z and determination of electroweak parameters from its hadronic decays - Nuclear Physics B 417 (1994) 3-57
Measurement and interpretation of the W-pair cross-section in e+e- interaction at 161 GeV Phys. Lett. B 397 (1997) 158-170
Measurement of the mass and width of the W boson in e+e- collision at s =189 GeV Phys. Lett. B 511 (2001) 159-177
http://www.pd.infn.it/~torassa/dottorato/2013/