UPGRADE 2 TRACKING SYSTEM - TRACKING AND SIMULATION STUDIES (UPSTREAM AND DOWNSTREAM) - CERN Indico
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
1
UPGRADE 2
TRACKING SYSTEM
TRACKING AND SIMULATION STUDIES (UPSTREAM AND DOWNSTREAM)
Lucia Grillo
(for the Mighty Tracker group)
5th Workshop on LHCb Upgrade II, Barcelona, 30/03/2020
Apologies for links to LHCb internal documents, those are just for completeness for LHCb colleagues
INTRODUCTION 2
Mighty Tracker
TRACKING SYSTEM
UT
~21% of long tracks *
~50% of long tracks *
~29% of long tracks *
* At first layer of T1, inclusive-b,
Upgrade 2 luminosity
T-STATIONS: INNER, MIDDLE TRACKER, SCIFI 3
LHCb-INT-2019-007
MIGHTY TRACKER Inner Tracker (LS3)
20cm
20cm
Middle Tracker (LS4) ~54 cm (one Sci-Fi module) 20 cm
▸ HV-CMOS: Inner Tracker (IT), Upgrade
1b + Middle Tracker (MT), Upgrade 2
▸ Design choices that impact track
reconstruction performance (track
finding efficiency, fake rate,
momentum resolution): IT
Area per layer = 6 lots of 20x54 cm = 0.7 m2 (minus beam hole)
Total Area = 6 layers of 0.7 m2 = 3.9 m2 (minus beam hole)
IT+MT
Area per layer = 28 lots of 20x54 cm = 3.0 m2 (minus beam hole)
Total Area = 6 layers of 0.7 m2 = 18.1 m2 (minus beam hole)
L = 1.5 ⇥ 1034 cm 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
2
s 1
L = 2 ⇥ 1033 cm 2 s 1
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
2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 …
LS2 Upgrade Ia LS3 Upgrade Ib LS4 UII
SciFi Only SciFi + Si Inner Tracker SciFi +
Inner +
U Ib TDR UII TDR Middle
UII Framework TDR U Ib Install U II Install Tracker
T-STATIONS 4
MIGHTY TRACKER - UPGRADE 1B 4 layers of hybrid + 8 fibre
modules. How many ‘old’
▸ Inner Tracker (IT), Upgrade 1b (new ideas)
SciFi modules we can afford
for tracking performance?
536 mm
Fibremat top
Inner /
4835 mm
2 mm Middle
Tracker
Dead region
Fibremat bottom
2 mm 2 mm y-size initially chosen for reduction of
Standard Hybrid
igure 3.6: The dimensions of a module as described in the simulation and the definition of
SciFi Module Module
Shortened
SciFi Module
fake rate (assuming fake rate negligible in
he stereo angles. The size of the dead material is increased to be visible. The simulation
escribes one full fibre mat in each module whereas the final module will be constructed using
ight separate 13.5 cm fibre mats placed side by side.
the silicon region). Larger y-size justified
for specific physics cases?
ime (less than a few ns). Single dopant fibres that emit in the blue-region (such as PMP,
r p-terphenyl (PT) dye) typically have a shorter attenuation length of ⇠1 m, due to the
elf-absorption of the light by the dye. Dyes that exhibit a larger separation between their
bsorption and emission spectra (Stokes’ Shift), such as 3HF which emits in the green,
ave longer attenuation lengths (> 2 m). A second approach is to use two scintillating
yes. The primary dye has a high quantum efficiency to absorb the energy from the base
material and the second dye is a wavelength shifter (⇠0.05% by weight). It absorbs the
mission of the primary dye, via radiative or non-radiative transfer, and fluoresces at a
onger wavelength where re-absorption in the fibre is less likely to occur [43].
For the time being, all experimental work has been concentrated on the multi-clad
lue emitting fibre of type SCSF-78MJ4 from Kuraray5 . It uses p-terphenyl (PT) as a
rimary dye, plus tetraphenyl-butadiene (TPB)6 as a wavelength shifter [44–47] and was
hosen as the baseline because of previous experience and knowledge gained from using
he scintillating fibres in other experiments. Co-operation with a second fibre supplier,
4
▸ Additional options?
The M indicates multi-clad. The J indicates a high purity distillation process was used. This results
n an extended attenuation length.
5
Kuraray Co., Ltd., Ote Center Building, 1-1-3, Otemachi, Chiyoda-ku, Tokyo 100-8115, Japan.
6
The dyes have not been confirmed directly by Kuraray, but are based on private communications in
T-STATIONS: NEEDS OF ION PHYSICS 5
Benjamin Audurier 23.10.2019
LHCb performance in PbPb Collisions LHCb-INT-2020-004
/ Centrality = f(eCal energy)
▸ Higher centrality = higher
energy density = creation of
Quark-Gluon Plasma rA rB
▸ PbPb track reconstruction up to
30% centrality with the upgrade
detector, SciFi occupancy limiting
factor
Maximum occupancy of the SciFi
LHCb Unofficial
▸ IT reduces occupancy in the SciFi: which IT geometry would allow to reach 0%
centrality?
T-STATIONS: NEEDS OF ION PHYSICS 6
Benjamin Audurier 04.03.2020
LHCb performance in PbPb Collisions LHCb-INT-2020-004
▸ Method: use ALICE Phys. Lett. B 772 (2017) 457 to estimate charged particle
multiplicity in LHCb
multiplicity x 3.5
multiplicity
x 2.2
▸ IT extended to 2nd to most
central modules is needed to
reach 0% centrality
▸ Additional technical complications
and/or costs
▸ “Original” IT allows to reach
10%-20% centrality
T-STATIONS: INNER AND MIDDLE TRACKER 7
Inner Tracker (LS3)
LHCb-INT-2019-00720cm
MIGHTY TRACKER: BASELINE DESIGN UPGRADE 2 Middle Tracker (LS4)
20cm
~54 cm (one Sci-Fi module) 20 cm
▸ Extension of IT/MT/SciFi regions 1060mm
200mm
▸ Geometrical constraints; Occupancy
per SciFi fibre < 2% (MT region)
▸ IT/MT # layers and arrangement
IT
▸ Pixels: track finding efficiency close to 100%
Area periflayer
dead= 6 lotsarea
of 20x54iscm < 5%
= 0.7 withbeam
m2 (minus 6 layers
hole)
Total Area = 6 layers of 0.7 m2 = 3.9 m2 (minus beam hole)
▸ IT/MT pixel/strip size IT+MT
Area per layer = 28 lots of 20x54 cm = 3.0 m2 (minus beam hole)
Total Area = 6 layers of 0.7 m2 = 18.1 m2 (minus beam hole)
▸ Occupancy: strips shorter than 100mm needed in IT Upgrade 2
▸ Pixel size in x to have momentum resolution as good as Upgrade 1: 100μm
▸ Pixel size in y to maximise track angle (wrt then beam axis) resolution.
Multiple scattering dominates for pixel size larger than (100-400)μm for
tracks of (1.5-40)GeV, with 40mm spacing between layers
T-STATIONS: INNER AND MIDDLE TRACKER 8
MightyPix - pixel size
▸ Baseline: 100μm x 300μm other sizes up to 50μm x 150μm
0.9 ×10−p3 = 15000.00[MeV] d = 40.00[mm]
δ p / p [%]
σ track y-angle [rad]
0.8 0.2
0.7 0.18
0.16
0.6
0.14
0.5
0.12
0.4 0.1
0.3 0.08
Upgrade I
0.2 Upgrade Ib, 100µm
Upgrade Ib, 200µm
0.06
hit resolution
0.1 Upgrade II, 100µm 0.04 MS
Upgrade II, 200µm MS + hit resolution
0 0.02
0 50 100 0.5 1
Momentum [GeV] pitch[mm]
▸ Further studies on going to refine this number
104
Abs(tan(αx))
98% of clusters
▸ Cluster size~ 1.02 pixel from d · tan(↵) 103
contain one hit
Entries 548037
geometrical considerations
Mean 0.518 ± 0.00113
102 Std Dev 0.832 ± 0.000798
(slope between exit and
10
entry points) other effects
may dominate 1
0 2 4 6 8 10
FINDING TRACKS 9
“Forward”
FINDING “LONG: TRACKS - TODAY
▸ Long (+downstream) tracks:
what everyone needs
“Matching”
UT
▸ You need to match VELO(UT) track and either hits on T-stations or T-track
▸ Following studies use Bs→φφ MC, generated with Upgrade 2 luminosity and
Upgrade 1 geometry. MCHits are used (+ smearing to add hit resolution effect)
FINDING TRACKS - “VELO-UT” SEGMENT 10
ra ck
T
EXTENDING VELO TRACK SEGMENT IN UT
t line
straigh
VELO
x
z
UT, layer 1
▸ Find UT hit to extend the VELO track segment by
opening a search window in x and y
relative to total integral (normalised to 1)
1
▸ 14mm x 9mm search window ensures the 0.8
search window in X
roughly 98% in x
correct hit is within the search window more search window in Y
than 98% of the cases; average 7-11 hits 0.6 roughly 98% in Y
within search window 0.4
▸ Next steps: include region dependence and 0.2
try to reduce fake rate (timing?)
0
0 5 10 15 20
Luke Mitchell, Edinburgh summer student project search window size [mm]FINDING TRACKS - “FORWARD” TRACKING 11
Vadym Denysenko 12.02.2020
FORWARD TRACKING
“kink” plane
1 2
search window in x (200mm) and y (75mm) Use search window in y and x information
3 4
Repeat for all hits within initial search window, excluding
Use search window in y and x information
downstream segments incompatible for tracks slopes in x and y
▸ Assuming VeloUT segment
▸ Find tracks passing basic criteria: not electrons, no secondary interactions between
last upstream hit and first downstream hit, minimum number of hits and momentumFINDING TRACKS - “FORWARD” TRACKING 12
Vadym Denysenko 12.02.2020
FORWARD TRACKING
“kink” plane
1 2
search window in x (200mm) and y (75mm) Use search window in y and x information
3 4
Repeat for all hits within initial search window, excluding
Use search window in y and x information
downstream segments incompatible for tracks slopes in x and y
At least 3 downstream hits (1 per station belonging to the correct particle)
~95% ▸ First study. Next steps:
investigate strategies to
~33% reduce the fake rateFINDING TRACKS - TRACK “MATCHING” 13
TRACK MATCHING
▸ Assuming stand-alone track
reconstruction in then T-
stations is possible,
▸ Match the VELO-UT
segments to the T-station
segments, by minimising:
2 (xpred xmeas ) 2
(ypred ymeas ) 2 (ty pred ty meas )2
match = 2
+ 2
+ 2
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
x y ty
Irene Cortinovis, Zurich student project Alessandro Scarabotto, CERN summer student projectFINDING TRACKS - TRACK “MATCHING” 14
TRACK MATCHING
If for every VELO-UT track a
match is found.
Discarding VELO-UT with too
high Χ^2:
Upgrade 1 Upgrade 2
98% Matched VELO-UT, 16% ghost rate
Efficiency 0.968 ± 0.005 0.932 ± 0.003
96% Matched VELO-UT, 15% ghost rate
Ghost rate 0.127 ± 0.008 0.173 ± 0.004
Upgrade 1 luminosity,
using same algorithm
▸ Dominant contribution to these ghost rates is from
tracks not reaching the T-stations (cut applied for
following studies)
Matched VELO-UT tracks [%]
Irene Cortinovis, Zurich student project Alessandro Scarabotto, CERN summer student projectFINDING TRACKS - TRACK “MATCHING” Zakariya Aliouche 25.03.2020 15
TRACK MATCHING - ADDING MOMENTUM & (PV) TIMING
2 (xpred xmeas )2 (ypred ymeas )2 (ty pred ty meas )2 (pV eloU T pT stat )2
match = 2
+ 2
+ 2 + 2 2
x y ty p,V eloU T + p,T stat
▸ VELO-UT momentum
resolution ~15%
▸ VELO-T-stations momentum
resolution ~2%
2 (xpred xmeas ) 2
(ypred ymeas ) 2 (ty pred ty meas )2 (tP V
velo tP V
T stat ) 2
match = 2
+ 2
+ 2 + 2 2
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
x y ty velo + T stat
▸ time resolution
assumed to be 30ps AAACBXicdVC7SgNBFJ2NrxhfUUstBoNgIWE3bkjSBW0sI+QFyRJmJ5NkyOyDmbuBsGxj46/YWChi6z/Y+TdOXqCiBy6cOede5t7jhoIrMM1PI7W2vrG5ld7O7Ozu7R9kD4+aKogkZQ0aiEC2XaKY4D5rAAfB2qFkxHMFa7njm5nfmjCpeODXYRoyxyNDnw84JaClXva0q/jQI714wkSQXK5edQUEkl42Z+Zt27QrZaxJsVgqWQtilirYyptz5NAStV72o9sPaOQxH6ggSnUsMwQnJhI4FSzJdCPFQkLHZMg6mvrEY8qJ51ck+FwrfTwIpC4f8Fz9PhETT6mp5+pOj8BI/fZm4l9eJ4JB2Ym5H0bAfLr4aBAJDAGeRYL7XDIKYqoJoZLrXTEdEUko6OAyOoTVpfh/0izkrat84c7OVa+XcaTRCTpDF8hCJVRFt6iGGoiie/SIntGL8WA8Ga/G26I1ZSxnjtEPGO9fEKmZmQ==
velo , T stat
PV PV
tvelo = ttrue + velo
tP V
T stat = t PV
i + T stat
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FINDING TRACKS - TRACK “MATCHING” Zakariya Aliouche 25.03.2020 16
TRACK MATCHING - MOMENTUM & TIMING
2 (xpred xmeas )2 (ypred ymeas )2 (ty pred ty meas )2 (pV eloU T pT stat )2 (tP V
V elo tP V
T stat )
2
match = 2
+ 2
+ 2 + 2 2 + 2 2
x y ty p,V eloU T + p,T stat t,V elo + t,T stat
▸ Only 10% of the wrong
matches come from the
same PV
▸ ~60% of the wrong
matches are electrons
▸ Momentum more
beneficial than (PV) timingFINDING TRACKS - T-TRACKS 17 T-TRACK FINDING ▸ distance between pair of layers is current x1-x2 distance, baseline pixels: good hit resolution in x and y ▸ start from all the hits in the highest z layer and open a search window in y ▸ for each pair of hits search third hit within y search window ▸ for each set of 3 hits used linear and parabolic extrapolation to centre a x-y search window to find the hits in the following layers Jiazhen Tang, CERN summer student project
FINDING TRACKS - T-TRACKS 18
Possible tracks from one point in y component
280
Real track
T-TRACK FINDING 210
y position/mm
▸ Multiple track candidates for the same 140
initial hit: for p>5GeV the less curved
candidate (in x) is the real track in most 70
of the cases, for now choosing 0
7800 8225 8650 9075 9500
candidate including hits closer to the z position/mm
Possible tracks from one point in x component
predictions 450
0
x position/mm
-450
Real track
-900
-1350
tight search loose search -1800
7800 8225 8650 9075 9500
window window z position/mm
Efficiency 89.0% 93.9%
▸ Very first study. several
Ghost rate 12.9% 15.7% improvements to make it more
performing and more realistic
Jiazhen Tang, CERN summer student projectDETECTOR GEOMETRY DESCRIPTION 19
Tai-Hua Lin 04.03.2020
DETECTOR DESCRIPTION
▸ Full geometry of Mighty Tracker implemented in DD4hep
▸ Passive material not (decided and) implemented yet
▸ Some details like gap sizes needs to be confirmed
▸ Will allow for more realistic track reconstruction studies!SUMMARY AND OUTLOOK 20
OUTLOOK
▸ Open design questions/optimization
▸ Pixel size, module design (material, support, structure), strategy for Upgrade 1b
▸ First track reconstruction performance studies (using MCHits)
▸ First pattern recognition algorithm for MT implemented - few more ideas to be tried
and hits in the fibre region to be added: tool to optimise pixel size/layer spacing
▸ Extension of VELO track segment with UT hits and downstream hits to be better
investigated (might be profit from timing?)
▸ Upstream/downstream track segment matching looks promising, momentum looks
more beneficial than timing
▸ Progress in DD4hep detector geometry description
▸ Geometry implemented in DD4hep, next steps: material implementation and
realistic tracking studies
▸ Good progress so far, but a lot still to do21 BACKUP
UT: FIRST STUDIES 22
Jianchun Wang, Quan Zou, Yiming Li, 18.03.2020
UT OCCUPANCY - UPGRADE 1
▸ Minimum bias (MB) events used to estimate
the strip occupancy and the number of hits to
be recorded
▸ At inner sensors the average occupancy is
~1%. Some ASICs are higher ~2%UT: FIRST STUDIES 23
Luke Mitchell, Edinburgh summer student project
UT OCCUPANCY - UPGRADE 2
▸ Could imagine equipping using HVCMOS of IT and/or outer part with silicon strips
▸ Occupancy / radiation requirements an order of magnitude higher than in Mighty
Tracker
Strip Occupancy 1D Hist in x
Occupancy [%]
35
0-100mm in Y
30
100-200mm in Y
25 200-300mm in Y
20
15
10
5
1% 100x500μm pixels
0
−800 −600 −400 −200 0 200 400 600 800
x [mm]
▸ From occupancy considerations, strips in external region (2/3 of the area) feasible,
pixels internally (1/3 of the area)REGIONS OF COVERAGE 24
OCCUPANCY STUDIES
▸ Hybrid modules: same width for IT/MT and SciFi modules
Upgrade 2
Upgrade 2
Low occupancy per silicon
sensor preserves track
reconstruction (e.g. allowing
stand alone track
reconstruction in T-stations)
Constraint on occupancy of the SciFi to Upgrade 1
occupancy (to preserve tracking performance and contain
radiation damage of the fibres) sets the MT dimension
▸ Number of hits obtained from Geant4 simulation (MCHits) - safety factors needed
for: pp collisions, tracking threshold, material assumes the full SciFiREGIONS OF COVERAGE 25
OCCUPANCY STUDIES - SCIFI
▸ Choice of MT size: maximum occupancy in Upgrade 2 in SciFi does not exceed the
occupancy of SciFi in Upgrade 1
▸ Maximum integrated occupancies along the fibres
Occupancy per fibre averaged over 40 fibres in top half (y>0) station 1, layer 1
Occupancy [1 / fiber event]
0.025
Upgrade 1b SciFi Occupancy [1 / fiber event] 0.16 Upgrade 2 SciFi
0.14
2% SciFi with IT SciFi with MT
0.02 0.12
0.1
0.015
0.08
0.01 0.06
0.04
0.005 2%
0.02
0 0
−2000 0 2000 −2000 0 2000
x [mm] x [mm]REGIONS OF COVERAGE 26
OCCUPANCY STUDIES - IT & MT
▸ Could we use silicon strips?
▸ Estimated the occupancy per pixel/strip in IT & MT
Occupancy per bin (0.1mm x100 mm - strip size) per event
Upgrade 1b Upgrade 2
▸ All acceptable values for Upgrade 1b, likely shorter strips needed for Upgrade 2UPGRADE 1B 27
INNER TRACKER: FIRST ESTIMATE OF FAKE RATE REDUCTION
▸ While the horizontal dimension of IT is constrained to be 1060mm by the width of
the two (central) SciFi module dimension, there is room for optimisation of the
vertical size
▸ Hits in different IT region are removed to estimate the improvement in fake rate
(in the SciFI)
Full SciFi, fake rate 18% Upgrade 1b
Nominal IT area, fake rate 3%
▸ VELO-like track reconstruction in the IT assumedNEW PROPOSAL FOR UPGRADE 1B 28
UPGRADE 1B - QUICK CHECK
▸ Hits in different IT region are removed, standard (in the SciFI - VELO-like track
reconstruction is assumed)
▸ B+->J/psiK+ sample, L = 2 ⇥ 1033 cm
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
2
s 1
, O(1k) events digitised & reconstructed,
▸ Attenuation maps considered for the fibres https://gitlab.cern.ch/lhcb-conddb/
SIMCOND/tree/upgrade/master/Conditions/FT/Calibration (documented here:
http://cdsweb.cern.ch/record/2673602/files/LHCb-PUB-2019-007.pdf),
Full SciFi
x6 x4
1 Eff(long>5GeV) 95.4% 96.0% 95.9%
25fb Ghost rate 14.6% 5.5% 8.8%
1 Eff(long>5GeV) 95.2% 95.4%
50fb Ghost rate 13.8% 8.3%
1 Eff(long>5GeV) 93.5% 94.2%
100fb 14.2% 8.2%
Ghost rateFINDING TRACKS - TRACK “MATCHING” Zakariya Aliouche 25.03.2020 29
TRACK MATCHING - MOMENTUM & TIMING
2 (xpred xmeas )2 (ypred ymeas )2 (ty pred ty meas )2 (pV eloU T pT stat )2 (tP V
V elo tP V
T stat )
2
match = 2
+ 2
+ 2 + 2 2 + 2 2
x y ty p,V eloU T + p,T stat t,V elo + t,T stat
▸ Only 10% of the wrong
matches come from the
same PV
▸ ~60% of the wrong
matches are electrons
▸ Momentum more
beneficial than (PV) timing
▸ VELO-T-stations momentum resolution ~1%NUMBERS OF LAYERS 30
PIXELS - TRACKING IN Y - 6 LAYERS
▸ Pattern recognition not only in the bending plane ?
▸ Pixels: high resolution in x and y: for pixel based IT/MT the number of layers can be
reduced, achieving similar or better resolution
▸ With 100% sensor active area 100
∈rec %
(possible if DMAPS pixel chips are
98
arranged to overlap in each layer),
96
6 layers would obtain the effective
94
measurements of the current IT Track finding efficiency
92
close to 100% if dead area
90 is < 5% with 6 layers
88 >3 hits
86
>4 hits toy simulation
0 5 10
dead fraction %PIXEL SIZE (X) 31
MOMENTUM RESOLUTION
▸ Momentum resolution studied on D→Kππ Upgrade 1 sample, described
accounting for Multiple Scattering (MS) and hit resolution (res)
Upgrade 1 toy simulation to extrapolate to Upgrade 2
0.9 0.9
δ p / p [%]
δ p / p [%]
0.8 0.8
0.7 0.7
0.6 0.6
0.5 0.5
0.4 ⇣ p ⌘2 0.4
0.3 = A2MS + (p ⇥ Bres )2 0.3
Upgrade I
0.2 p
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
0.2 Upgrade Ib, 100µm
Upgrade Ib, 200µm
Upgrade II, 100µm
0.1 0.1 Upgrade II, 200µm
0 0
0 50 100 0 50 100
Momentum [GeV] Momentum [GeV]
▸ Pixel size 200μm results in a slightly worse momentum resolution, pixel size 100μm
results in an improvement (up to 20% at high momentum)PIXEL SIZE (Y) 32
TRACK ANGLE RESOLUTION
▸ Track toy (linear segment)
obtained from a truth angle,
and hits from smearing due to < ✓M S >=
13.6MeV
z
r
X 1
p
hit resolution and Multiple
cp X0 2
X/X0 = 0.565% (3.9mm spacing)
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
Scattering (MS)
▸ A “per-track” quantity allows to account for the pixel size and position of the
tracker layers at the same time: track angle resolution (in y)
▸ Pixel size in y of (100-400)μm optimal for tracks of (1.5-40)GeV
×10−p3 = 15000.00[MeV] d = 40.00[mm] ×10−p3 = 40000.00[MeV] d = 40.00[mm]
0.24
σ track y-angle [rad]
0.2 σ track y-angle [rad] 0.22
0.18 0.2
0.16 0.18
0.14 0.16
0.12 0.14
0.12
0.1
0.1
0.08 0.08
0.06 0.06
hit resolution hit resolution
0.04 MS 0.04 MS
MS + hit resolution MS + hit resolution
0.02 0.02
0.5 1 0.5 1
pitch[mm] pitch[mm]PIXEL SIZE (Y) AND ARRANGEMENT OF LAYERS 33
SPACING BETWEEN LAYERS
▸ Detector material considered for a double layer design (3.9mm spacing)
▸ When the layers in each doublet are too close, the hits are not distinguishable, and
cannot provide a direction useful for the pattern recognition (larger search windows
needed, extra random hits…)
▸ Which fraction of tracks has the same hit position in y?
p = 15000.00[MeV] pitch = 0.30[mm]
fractionmerged hits layers 1&2 and/or 3&4
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
10 20
d[mm]NUMBER AND ARRANGEMENT OF LAYERS 34
Kapton flex-circuit
Si Sensors
ARRANGEMENT OF LAYERS CF skin
Core with
Integrated cooling
▸ Minimise the material: 3 double-sided
layers with the sensor layers separated by
(3.9-40)mm - constraints: support, cooling
structure needed
▸ Distance between layers in each
doublet should be good enough for
pattern recognition to form starting
seeds (2 hits, direction information)
▸ Spacing between layers to be studied together with pixel size
▸ Alternative to 3 double-sided layers, giving larger spacing but increasing
material budget is to locate the 6 layers in the x layers of the x-u-v-x of SciFi
▸ Alternative to the 6 layers is 8 layers, arranged in 2:2:4 for T stations 1:2:3, to
minimise the material seen by the SciFiPIXEL SIZE (Y) AND LAYERS POSITIONS 35
IF THE LAYERS OF THE DOUBLETS ARE TOO CLOSE
▸ Detector material considered for a double layer design (3.9mm spacing)
▸ When the layers in each doublet are too close, the hits are not distinguishable,
and cannot provide a direction useful for the pattern recognition (larger
search windows needed, extra random hits…)
▸ Which fraction of tracks has the same hit position in y?
p = 15000.00[MeV] d = 40.00[mm] p = 15000.00[MeV] d = 3.90[mm]
fractionmerged hits layers 1&2 and/or 3&4
fractionmerged hits layers 1&2 and/or 3&4
0.14 0.9
0.12 0.8
0.7
0.1
0.6
0.08 0.5
0.06 0.4
0.3
0.04
0.2
0.02 0.1
0 0
0.5 1 0.5 1
pitch[mm] pitch[mm]
▸ spacings like 40mm looks preferred
▸ To-do: evaluate the impact of thisTRACK MOMENTUM 36
MOMENTUM DISTRIBUTION IN IT, MT, SCIFI REGIONS
5000
Middle Tracker tracks
Inner Tracker tracks
Entries 11679 Entries 20578
800 Mean 2.51e+04 Mean 7.78e+03
700 Std Dev 1.97e+04 4000 Std Dev 7e+03
600
500 3000
400
2000
300
200 1000
100
0 ×103 0 ×103
0 50 100 150 200 0 50 100 150 200
p [MeV] p [MeV]
SciFi tracks
4500 Entries 8516
▸ Inclusive-b MC 4000 Mean
Std Dev
2.94e+03
1.91e+03
sample, 1.5x10^34, 3500
3000
Long tracks, at first 2500
layer of first station 2000
1500
1000
500
0 ×103
0 50 100 150 200
p [MeV]PIXEL SIZE (Y) AND LAYERS POSITIONS 37
DIFFERENT CONFIGURATIONS [15GEV- MEAN P IN MT ACCEPTANCE]
×10−p3 = 15000.00[MeV] d = 3.90[mm] ×10−p3 = 15000.00[MeV] d = 40.00[mm]
0.22
σ track y-angle [rad]
σ track y-angle [rad]
0.2 0.2
0.18 0.18
0.16 0.16
0.14 0.14
0.12 0.12
0.1 0.1
0.08 0.08
0.06 0.06
hit resolution hit resolution
0.04 MS 0.04 MS
MS + hit resolution MS + hit resolution
0.02 0.02
0.5 1 0.5 1
pitch[mm] pitch[mm]
×10−p3 = 15000.00[MeV] d = 184.80[mm]
0.25
σ track y-angle [rad]
silicon on x1 and x2 layers of T-stations
0.2 ▸ Small dependence on the
0.15
layers positions. For
doublets (3.9-40mm) no
0.1
gain for < 200μm
0.05 hit resolution
MS
MS + hit resolution
0.5 1
pitch[mm]PIXEL SIZE (Y) AND LAYERS POSITIONS 38
DIFFERENT CONFIGURATIONS [5GEV]
×10−d3 = 5.00[mm] ×10−d3 = 40.00[mm]
0.25
σ track y-angle [rad]
σ track y-angle [rad]
0.25
0.2
0.2
0.15
0.15
0.1 0.1
0.05 geometry 0.05 geometry
MS MS
0.5 1 0.5 1
pitch[mm] pitch[mm]
−3
0.24 ×10 d = 15.00[mm] ×10−d3 = 184.80[mm]
σ track y-angle [rad]
0.3
σ track y-angle [rad]
0.22 silicon on x1 and x2 layers of T-stations
0.2 0.25
0.18
0.16 0.2
0.14
0.12 0.15
0.1
0.08 0.1
0.06
geometry
0.04 0.05 geometry
MS MS
0.02
0.5 1 0.5 1
pitch[mm] pitch[mm]CHIP SPECIFICATION DOCUMENT 39
SIZE OF THE CLUSTER (IN #PIXEL)?
z
d* tan(Beta) Track, 4um radius of charge deposition
Thickness d
Beta (LHCb z dirn)
track angle to
vertical
x(y)
Example entry at back and exit
X, y view of pixels at front points of trackCHIP SPECIFICATION DOCUMENT 40
SIZE OF THE CLUSTER (IN #PIXEL)?
▸ MCHits in the T-stations (i.e. secondaries etc included)
▸ Slope between exit and entry points of the MCHits along x and y directions (in
plots: average over the events of the sample)
104
Abs(tan(αx)) 104 Abs(tan(αy))
103 Entries 548037 Entries 548037
103
Mean 0.518 ± 0.00113 Mean 0.245 ± 0.000889
102 Std Dev 0.832 ± 0.000798
102 Std Dev 0.657 ± 0.000629
10 10
1 1
0 2 4 6 8 10 0 2 4 6 8 10
▸ d = 30microns, baseline pixels 100microns x 300microns
▸ 97.5% of the hits has cluster size 1, 2.5% of the hits has cluster size 2… Cluster
size should be ~1.02 pixelsFINDING TRACKS - TIMING 41
TIME RESOLUTION AND TRACK MATCHING/PV ASSOCIATION
▸ Example: use TORCH time
measurement (~15ps per-track
resolution) for downstream and
UT (layer or sensors - 25-200ps
per-track resolution)
▸ To be studied impact on full
matching upstream/
LHC collision time
sigma ~ 200 ps
downstream and rate of mis-
σUT = 50.00[ps] σMT = 15.00[ps] σUT = 200.00[ps] σMT = 15.00[ps] association (fakes)
18 Entries 250 Entries 250
16 Mean 12.6 35 Mean −4.21 σMT = 15.00[ps]
Std Dev 200
downstream-t downstream )
Std Dev 102
30
from upstream
14
12 1
25
10 0.8
20
8
15 0.6
6
Relative σ(tmeas
4 10
0.4
2 5
0 0.2
−500 0 500 0
−500 0 500
t 0[ps] t meas from upstream 0 50 100 150 200
downstream downstream [ps]
-t Upstream per-track resolution[ps]You can also read
