Tracking and Alignment in LHCb Florin MACIUC on behalf of LHCb - - PowerPoint PPT Presentation

▶
tracking and alignment in lhcb
SMART_READER_LITE
LIVE PREVIEW

Tracking and Alignment in LHCb Florin MACIUC on behalf of LHCb - - PowerPoint PPT Presentation

Mar 11, 2024 5 likes •325 views

Tracking and Alignment in LHCb Florin MACIUC on behalf of LHCb collaboration florin.maciuc@mpi-hd.mpg.de Max-Planck Institute for Nuclear Physics Heidelberg Physics at LHC 2010 Hamburg p. 1/22 LHCb and B-physics LHCb - Large Hadron

slide-1
SLIDE 1

Tracking and Alignment in LHCb

Florin MACIUC on behalf of LHCb collaboration

florin.maciuc@mpi-hd.mpg.de

Max-Planck Institute for Nuclear Physics Heidelberg

Physics at LHC 2010 Hamburg – p. 1/22

slide-2
SLIDE 2

LHCb and B-physics

  • LHCb - Large Hadron Collider beauty detector.
  • LHCb aims lay primary in the B-physics sector.
  • Nominal luminosity of about 2 × 1032 cm−2s−1 =

⇒ 1012 b¯ b per

year.

  • The dominant channel behavior explains the single-arm forward

spectrometer geometry chosen for LHCb.

Gluon fusion before fragmentation forward beaming of b¯ b in the LHCb frame

Physics at LHC 2010 Hamburg – p. 2/22

slide-3
SLIDE 3

LHCb Detector

M1 M3 M2 M4 M5 RICH2 HCAL ECAL SPD/PS Magnet z 5m y 5m 10m 15m 20m TT T1 T2 T3 Vertex Locator

z y X x Primary Vertex (PV)

Physics at LHC 2010 Hamburg – p. 3/22

slide-4
SLIDE 4

LHCb Detector

M1 M3 M2 M4 M5 RICH2 HCAL ECAL SPD/PS Magnet z 5m y 5m 10m 15m 20m TT T1 T2 T3 Vertex Locator

VErtex LOcator (VELO): Silicon Detector Inner Tracker (IT): Silicon Detector Tracker Turicensis (TT): Silicon Detector pitch 183 µm, depth 500 µm pitch 198 µm, depth 320-410 µm pitch 38-102 µm, depth 300 µm

Physics at LHC 2010 Hamburg – p. 3/22

slide-5
SLIDE 5

LHCb Detector

M1 M3 M2 M4 M5 RICH2 HCAL ECAL SPD/PS Magnet z 5m y 5m 10m 15m 20m TT T1 T2 T3 Vertex Locator

Outer Tracker (OT): Straw Tube Detector radius 2.45 mm hit resolution 200 µm

Physics at LHC 2010 Hamburg – p. 3/22

slide-6
SLIDE 6

LHCb Detector

M1 M3 M2 M4 M5 RICH2 HCAL ECAL SPD/PS Magnet z 5m y 5m 10m 15m 20m TT T1 T2 T3 Vertex Locator

Warm Magnet : integrated magnetic field of 4 T · m

Physics at LHC 2010 Hamburg – p. 3/22

slide-7
SLIDE 7

VErtex LOcator

  • Primary Vertex (PV) is inside VELO, towards middle;
  • VELO is a retractable detector, 2 VELO sides:

⋆ To protect from damage, VELO is in Open position before the

beam is stable, and closed afterward.

⋆ Open VELO: sensors 30 mm further from the beam, ⋆ Closed VELO: sensors are about 8 mm from the beam line,

VELO double-sensor modules: R+φ Schematic: one side of VELO

Physics at LHC 2010 Hamburg – p. 4/22

slide-8
SLIDE 8

VErtex LOcator

  • Primary Vertex (PV) is inside VELO, towards middle;
  • VELO is a retractable detector, 2 VELO sides:

⋆ To protect from damage, VELO is in Open position before the

beam is stable, and closed afterward.

⋆ Open VELO: sensors 30 mm further from the beam, ⋆ Closed VELO: sensors are about 8 mm from the beam line,

Schematic VELO sensors in Open and Closed positions

Physics at LHC 2010 Hamburg – p. 4/22

slide-9
SLIDE 9

Primary Vertex Resolution

  • Primary Vertex (PV) is determined with VELO tracks.
  • Method: randomly split event track container in two, and

reconstruct PV.

  • Results close to expected,

⋆ A residual ≈ 40 % difference - e.g. when using 25 tracks. ⋆ Improving.

PV resolution vs track used, real data PV resolution vs track used, MC MC Data ∆x(µm) 11.5 15.8 ∆y(µm) 11.3 15.2 ∆z(µm) 57 91

Physics at LHC 2010 Hamburg – p. 5/22

slide-10
SLIDE 10

Impact Parameter Resolution

  • Impact parameter (IP) - Closest approach to PV of a track.
  • IP resolution is determined primarily by:

⋆ random scattering in VELO material, VELO misalignments and

hit resolutions.

  • IP resolution for MC and data given.

Impact Parameter resolution in X Impact Parameter resolution in Y

(c/GeV)

T

1/p

0.5 1 1.5 2 2.5 3

mm

0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 VELO Closed

2010 Data m Foil µ

  • Sim. 300

m Foil µ

  • Sim. 250

LHCb VELO Preliminary m µ

T

2010 Data: 16.2 + 24.6/p m µ

T

m Foil: 11.2 + 21.0/p µ

  • Sim. 300

m µ

T

m Foil: 11.2 + 19.9/p µ

  • Sim. 250

T

Resolution Vs 1/p

X

IP

(c/GeV)

T

1/p

0.5 1 1.5 2 2.5 3

mm

0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 VELO Closed

2010 Data m Foil µ

  • Sim. 300

m Foil µ

  • Sim. 250

LHCb VELO Preliminary m µ

T

2010 Data: 15.7 + 24.4/p m µ

T

m Foil: 11.5 + 20.6/p µ

  • Sim. 300

m µ

T

m Foil: 11.9 + 19.3/p µ

  • Sim. 250

T

Resolution Vs 1/p

Y

IP

Physics at LHC 2010 Hamburg – p. 6/22

slide-11
SLIDE 11

Impact Parameter Resolution

  • Impact parameter (IP) - Closest approach to PV of a track.
  • IP resolution is determined primarily by:

⋆ random scattering in VELO material, VELO misalignments and

hit resolutions.

  • IP resolution for MC and data given.
  • 15-40 % difference between MC and data.
  • Accounted for already.

⋆ Some disagreement in material description of MC. ⋆ Misalignment between VELO sides.

  • Remaining:
  • residual misalignments of sensors, 4.4µm,
  • too optimistic hit resolution in MC,
  • charge sharing.

Physics at LHC 2010 Hamburg – p. 6/22

slide-12
SLIDE 12

Alignment Status of Subdetectors

  • Optical alignment of VELO, OT, IT, TT : Survey.
  • Updated software alignment Aligned.
  • Monte Carlo results: black histograms.
  • Rtrack − Rhit, measurement residual distribution gauges the

alignment quality.

VELO R-sensor residuals OT residuals

Physics at LHC 2010 Hamburg – p. 7/22

slide-13
SLIDE 13

Alignment Status of Subdetectors

  • Optical alignment of VELO, OT, IT, TT : Survey.
  • Updated software alignment Aligned.
  • Monte Carlo results: black histograms.
  • Rtrack − Rhit, measurement residual distribution gauges the

alignment quality.

IT residuals TT residuals

Physics at LHC 2010 Hamburg – p. 7/22

slide-14
SLIDE 14

Silicon Trackers: Hit Resolution

  • 40-50% difference between Monte Carlo and Data for IT and TT.
  • IT and TT are single-sided silicon strip detectors.
  • One source of disagreement was found in the charge sharing

between neighboring strips.

⋆ This effect was overestimated in MC. ⋆ After correction: an increase from 40 µm to 50 µm for IT hit

resolution.

  • We expect residual misalignments to account for the rest.

charge sharing between two strips larger cluster of strips improve measurement resolution

Physics at LHC 2010 Hamburg – p. 8/22

slide-15
SLIDE 15

Long Track Efficiency

  • Long track efficiency obtainable from KS candidates.
  • Method:

⋆ Finds VELO segment and the associated CALO cluster, ⋆ Gets Long tracks from reconstruction, ⋆ KS Candidates 1: VELO+CALO track and a Long track, ⋆ KS Candidates 2: 2 Long tracks.

  • The method supplies IT/OT/TT efficiency in tracking.
  • Results close to 100%, with MC and data agreement.

Long-Long KS candidates, mass plot Efficiency as a function pT

[MeV]

π π

m

400 450 500 550 600

candidates / 2 MeV

5000 10000 15000 20000 25000

longtrack + velo-calo track signal component longtrack + ( longtrack & velo-calo track ) signal component

= 7 TeV s LHCb preliminary

[MeV]

T

p

200 400 600 800 1000

efficiency

0.2 0.4 0.6 0.8 1

Data Monte Carlo

= 7 TeV s LHCb preliminary

Physics at LHC 2010 Hamburg – p. 9/22

slide-16
SLIDE 16

An Other Method for Track Efficiency

  • Method, phase 1:

⋆ For all VELO segments, finds a corresponding CALO

cluster in the bending plane (x,z)

⋆ Checks in the non-bending (z, y) plane, ⋆ Fits track VELO+CALO,

  • Phase 2:

⋆ IT/OT/TT segments are matched to the found track. ⋆ the previous segments are provided by the various Pattern-Recognition algorithms.

Physics at LHC 2010 Hamburg – p. 10/22

slide-17
SLIDE 17

An Other Method for Track Efficiency

  • Method, phase 1:

⋆ For all VELO segments, finds a corresponding CALO

cluster in the bending plane (x,z)

⋆ Checks in the non-bending (z, y) plane, ⋆ Fits track VELO+CALO,

  • Phase 2:

⋆ IT/OT/TT segments are matched to the found track. ⋆ the previous segments are provided by the various Pattern-Recognition algorithms.

Difference in y for the track and CALO cluster includes only VELO+CALO tracks, which position, includes all VELO+CALO tracks have an associated Downstream segment

ǫeff = n2

n1

Physics at LHC 2010 Hamburg – p. 10/22

slide-18
SLIDE 18

Particle Zoo

  • Mass values of several detected particle agree with the PDG values

to per mil level.

  • Small signal widths , e.g. 2.8 MeV for Λ, 2.7 MeV Ξ−, 8.5 MeV

D0, 2.5 MeV Ω, etc. KS Λ J/ψ Ξ−

)

2

(MeV/c

π K

m

1800 1850 1900

)

2

Entries / (3 MeV/c

50 100 150 200 250

)

2

(MeV/c

π K

m

1800 1850 1900

)

2

Entries / (3 MeV/c

50 100 150 200 250

46 ± = 1539 signal N 2 0.27 MeV/c ± = 1863.38 µ Mass 2 0.24 MeV/c ± = 8.69 σ Mass = 7 TeV Data s

Preliminary LHCb

) 2 mass (MeV/c π S K m 1800 1850 1900 1950 Events / ( 5 ) 5 10 15 20 25 ) 2 mass (MeV/c π S K m 1800 1850 1900 1950 Events / ( 5 ) 5 10 15 20 25 = 7 TeV Data s

Preliminary LHCb

Gauss σ m Signal N = = = 8.48 1869.9 68.8 1.2 MeV ± 1.5 MeV ± 11 ±

)

2

mass (MeV/c

π pK

m

2200 2250 2300 2350

)

2

Events / ( 4.75 MeV/c

5 10 15 20 25 30 35

)

2

mass (MeV/c

π pK

m

2200 2250 2300 2350

)

2

Events / ( 4.75 MeV/c

5 10 15 20 25 30 35 = 7 TeV Data s

Preliminary LHCb

Gauss σ m Signal N = = = 3.73 2286.1 51.9 0.79 MeV ± 0.72 MeV ± 9.2 ±

D0 D+ Ω Λc

Plus many more other ....

Physics at LHC 2010 Hamburg – p. 11/22

slide-19
SLIDE 19

Summary and Conclusions

  • Already more than 100 Million 7 TeV Collisions in the 2010 LHCb

data.

  • Main conclusion: Alignment and tracking are in good shape for

physics analysis.

  • Monitoring of alignment and tracking quality in progress.
  • Already done gradual improvements in:
  • Detector description,
  • Tracking tools,
  • Alignment.
  • As result, MC and data reconstruction give a better agreement.
  • More to do ... but“Terra Nova”/“Terra Incognita”in sight, as we

reconstruct particles from 7 TeV pp collisions with high precision.

Physics at LHC 2010 Hamburg – p. 12/22

slide-20
SLIDE 20

Backup Slides

Physics at LHC 2010 Hamburg – p. 13/22

slide-21
SLIDE 21

Impact Parameter (IP)

  • 2010 data , VELO Closed

Impact Parameter resolution in X

Impact Parameter resolution in Y

(c/GeV)

T

1/p

0.5 1 1.5 2 2.5 3

mm

0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 VELO Closed

2010 Data Simulation LHCb VELO Preliminary m µ

T

2010 Data: 16.2 + 24.6/p m µ

T

Simulation: 11.2 + 19.9/p

T

Resolution Vs 1/p

X

IP

(c/GeV)

T

1/p

0.5 1 1.5 2 2.5 3

mm

0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 VELO Closed

2010 Data Simulation LHCb VELO Preliminary m µ

T

2010 Data: 15.7 + 24.4/p m µ

T

Simulation: 11.9 + 19.3/p

T

Resolution Vs 1/p

Y

IP

Physics at LHC 2010 Hamburg – p. 14/22

slide-22
SLIDE 22

VELO Sensor Alignment

Sensor alignment correction for 88 sensors 168 DoF in X and Y

m) µ Misalign (

  • 50
  • 40
  • 30
  • 20
  • 10

10 20 30 40 50 Entries 10 20 30 40 50 60 70 80

Overview of misalignments Sensor X and Y m µ Alignment: 4.4

LHCb Preliminary

Physics at LHC 2010 Hamburg – p. 15/22

slide-23
SLIDE 23

Residual Distributions for IT

IT pull plots single-sided silicon strip sensor

σ residual/

  • 5
  • 4
  • 3
  • 2
  • 1

1 2 3 4 5

#

0.01 0.02 0.03 0.04 0.05 0.06 0.07

Physics at LHC 2010 Hamburg – p. 16/22

slide-24
SLIDE 24

Primary Vertex Z

PV resolution vs track used PV resolution vs track used real data MC

Physics at LHC 2010 Hamburg – p. 17/22

slide-25
SLIDE 25

VELO Stability, Sensor Alignment

  • VELO retractable: Left/Right sides.

⋆ VELO is closed after stable beam conditions fulfilled.

  • Primary Vertex reconstruction with tracks from separate sides.

⋆ Difference gives an estimate of misalignment between VELO

sides.

∆X difference of PV (µm)

Run Number 69500 70000 70500 71000 71500 72000 m] µ PVx(Left-Right) [ ∆ Mean

  • 10
  • 5

5 10 15

LHCb Preliminary

X misalignment

Physics at LHC 2010 Hamburg – p. 18/22

slide-26
SLIDE 26

VELO Stability, Sensor Alignment

  • VELO retractable: Left/Right sides.

⋆ VELO is closed after stable beam conditions fulfilled.

  • Primary Vertex reconstruction with tracks from separate sides.

⋆ Difference gives an estimate of misalignment between VELO

sides.

∆Y difference of PV (µm)

Run Number 69500 70000 70500 71000 71500 72000 m] µ PVy(Left-Right) [ ∆ Mean

  • 6
  • 4
  • 2

2 4 6

LHCb Preliminary

Y misalignment

Physics at LHC 2010 Hamburg – p. 18/22

slide-27
SLIDE 27

VELO Stability, Sensor Alignment

  • VELO retractable: Left/Right sides.

⋆ VELO is closed after stable beam conditions fulfilled.

  • Primary Vertex reconstruction with tracks from separate sides.

⋆ Difference gives an estimate of misalignment between VELO

sides.

∆Z difference of PV (µm)

Run Number 69500 70000 70500 71000 71500 72000 m] µ PVz(Left-Right) [ ∆ Mean

  • 40
  • 30
  • 20
  • 10

10 20

LHCb Preliminary

Z misalignment

Physics at LHC 2010 Hamburg – p. 18/22

slide-28
SLIDE 28

Tracking Methods and Alignment

  • Reconstruction phase:

⋆ various pattern recognition algorithm + Kalman-Filter tracking.

  • Runge-Kutta extrapolator to deal with highly inhomogeneous field

in the tracking stations.

Physics at LHC 2010 Hamburg – p. 19/22

slide-29
SLIDE 29

Tracking Methods and Alignment

  • Reconstruction phase:

⋆ various pattern recognition algorithm + Kalman-Filter tracking.

  • Runge-Kutta extrapolator to deal with highly inhomogeneous field

in the tracking stations.

  • “Closed-form”alignment methods used:

⋆ Alignment with track model based on Kalman-Filter, ⋆ An alignment based on Millepede method, with parametrized

trajectory - Volker Blobel,

  • Equivalent methods, χ2 minimization over alignment and track

parameters simultaneously.

Physics at LHC 2010 Hamburg – p. 19/22

slide-30
SLIDE 30

Impact Parameter and Material

RF-foil divides Sides of VELO and prevents outgasing.

Physics at LHC 2010 Hamburg – p. 20/22

slide-31
SLIDE 31

Silicon Trackers: Hit Resolution

  • The charge sharing depends relatively strongly on the track slope.
  • Note for the experts: previous fact is detrimental to some of the

alignment parameters which couple strongly to the track slope.

Charge sharing vs. on track slope

Physics at LHC 2010 Hamburg – p. 21/22

slide-32
SLIDE 32

Downstream Tracks, Mass Resolutions

  • The best physics candidates are made from Long tracks.
  • Long lived particles: e.g., KS and Λ may decay outside VELO.

Down-Downstream tracks for KS Mass resolution vs. zdecay for KS

[MeV]

−

π

+

π

m 400 420 440 460 480 500 520 540 560 580 candidates/2 MeV 5000 10000 15000 20000 25000 30000

LHCb 2010 data, preliminary

−

0.01 MeV ± = 8.68 σ

decay vertex [mm]

s

z position of K 500 1000 1500 2000 2500 mass resolution [MeV] 1 2 3 4 5 6 7 8 9 10

Monte Carlo data

= 7 TeV, preliminary s LHCb 2010 data,

  • Hence, some physics studies are possible even without VELO...

Physics at LHC 2010 Hamburg – p. 22/22