1. Steve McMahon RAL Towards Phase II tracking at ATLAS
Mini-Workshop on ATLAS Upgrades for High Luminosity
May 20th 2014
Prague
S. McMahon : Prague May 2014

2. The Current ATLAS Inner Detector (ID)
Initial Specifications & Limitations
3 layers of Pixels (PIX), 4 layers of Si micro-strips (SCT), straw tracker (TRT)
Additional PIX-layer (IBL) to be inserted in Long Shut-Down 1 (LS )
ID designed for 10 years of operation, peak luminosity of 1 x 1034 cm-2s-1
The assumed pile up (m) was 23, bunch crossing frequency 25ns
and the L1 trigger rate of 100kHz. 2 Tesla Magnetic field
Current
Inner Detector
PIX: designed to withstand MeV neq/cm2 or about 400fb-1
IBL: designed to about 850fb-1
SCT: designed to withstand 2x1014 1MeV neq/cm2 or about 700fb-1
Radiation Damage
Both SCT and PIX apply zero suppression to accommodate <m> up to 50
This is being expanded in LS1 to get bveyond 75 (3 x ns)
However, there are “hard wired” limitations which mean one cannot go beyond this. PIX-Limits pixels per 25ns. For SCT data rate between FE chip and ReadOut Driver (ROD) limit us to 3 x 1034 cm-2s-1
Bandwidth Limitation
With the occupancy expected at the HL-LHC the SCT would not be able to resolve close-by tracks in the core of High PT jets
The TRT will approach 100% occupancy and tracking efficiency will suffer accordingly
Occupancy limits
S. McMahon : Prague May 2014

3. The Current ATLAS Inner Detector (ID)
Initial Specifications & Limitations
3 layers of Pixels (PIX), 4 layers of Si micro-strips (SCT), straw tracker (TRT)
Additional PIX-layer (IBL) to be inserted in Long Shut-Down 1 (LS )
ID designed for 10 years of operation, peak luminosity of 1 x 1034 cm-2s-1
The assumed pile up (m) was 23, bunch crossing frequency 25ns
and the L1 trigger rate of 100kHz. 2 Tesla Magnetic field
Current
Inner Detector
PIX: designed to withstand MeV neq/cm2 or about 400fb-1
IBL: designed to about 850fb-1
SCT: designed to withstand 2x1014 1MeV neq/cm2 or about 700fb-1
Radiation Damage
Both SCT and PIX apply zero suppression to accommodate <m> up to 50
This is being expanded in LS1 to get bveyond 75 (3 x ns)
However, there are “hard wired” limitations which mean one cannot go beyond this. PIX-Limits pixels per 25ns. For SCT data rate between FE chip and ReadOut Driver (ROD) limit us to 3 x 1034 cm-2s-1
Bandwidth Limitation
With the occupancy expected at the HL-LHC the SCT would not be able to resolve close-by tracks in the core of High PT jets
The TRT will approach 100% occupancy and tracking efficiency will suffer accordingly
Occupancy limits
S. McMahon : Prague May 2014

4. The Current ATLAS Inner Detector (ID)
Initial Specifications & Limitations
3 layers of Pixels (PIX), 4 layers of Si micro-strips (SCT), straw tracker (TRT)
Additional PIX-layer (IBL) to be inserted in Long Shut-Down 1 (LS )
ID designed for 10 years of operation, peak luminosity of 1 x 1034 cm-2s-1
The assumed pile up (m) was 23, bunch crossing frequency 25ns
and the L1 trigger rate of 100kHz. 2 Tesla Magnetic field
Current
Inner Detector
PIX: designed to withstand MeV neq/cm2 or about 400fb-1
IBL: designed to about 850fb-1
SCT: designed to withstand 2x1014 1MeV neq/cm2 or about 700fb-1
Radiation Damage
Both SCT and PIX apply zero suppression to accommodate <m> up to 50
This is being expanded in LS1 to get bveyond 75 (3 x ns)
However, there are “hard wired” limitations which mean one cannot go beyond this. PIX-Limits pixels per 25ns. For SCT data rate between FE chip and ReadOut Driver (ROD) limit us to 3 x 1034 cm-2s-1
Bandwidth Limitation
With the occupancy expected at the HL-LHC the SCT would not be able to resolve close-by tracks in the core of High PT jets
The TRT will approach 100% occupancy and tracking efficiency will suffer accordingly
Occupancy limits
S. McMahon : Prague May 2014

5. The Current ATLAS Inner Detector (ID)
Initial Specifications & Limitations
3 layers of Pixels (PIX), 4 layers of Si micro-strips (SCT), straw tracker (TRT)
Additional PIX-layer (IBL) to be inserted in Long Shut-Down 1 (LS )
ID designed for 10 years of operation, peak luminosity of 1 x 1034 cm-2s-1
The assumed pile up (m) was 23, bunch crossing frequency 25ns
and the L1 trigger rate of 100kHz. 2 Tesla Magnetic field
Current
Inner Detector
PIX: designed to withstand MeV neq/cm2 or about 400fb-1
IBL: designed to about 850fb-1
SCT: designed to withstand 2x1014 1MeV neq/cm2 or about 700fb-1
Radiation Damage
Both SCT and PIX apply zero suppression to accommodate <m> up to 50
This is being expanded in LS1 to get bveyond 75 (3 x ns)
However, there are “hard wired” limitations which mean one cannot go beyond this. PIX-Limits pixels per 25ns. For SCT data rate between FE chip and ReadOut Driver (ROD) limit us to 3 x 1034 cm-2s-1
Bandwidth Limitation
With the occupancy expected at the HL-LHC the SCT would not be able to resolve close-by tracks in the core of High PT jets
The TRT will approach 100% occupancy and tracking efficiency will suffer accordingly
Occupancy limits
S. McMahon : Prague May 2014

6. The Current ATLAS Inner Detector A few selected silicon highlights (with personal bias)
S. McMahon : Prague May 2014

7. The Current ATLAS Inner Detector A few selected silicon highlights (with personal bias)
S. McMahon : Prague May 2014

8. The Current ATLAS Inner Detector A few selected silicon highlights (with personal bias)
S. McMahon : Prague May 2014

9. The Current ATLAS Inner Detector A few selected silicon highlights (with personal bias)
S. McMahon : Prague May 2014

10. The Current ATLAS Inner Detector A few selected highlights (with personal bias)
S. McMahon : Prague May 2014

11. IBL : Phase-0 upgrades (LS-1)
Insertable B Layer (IBL)
Inserted at small radius (see insert)
Closest measurement 3.3cm from IP
Significantly improves light jet rejection
3→ 4 layer pixel device
n-in-n planar & 3D sensors
Installed in May 2014
Uses the 2cm x 2cm FEI4 rad hard chip produced in 130nm CMOS
IBL stays in place to Phase II
NOW
With IBL
Just one slide with a picture of IBL and one plot of performance improvement due to IBL
S. McMahon : Prague May 2014

12. ID Pixel : Phase-0 consolidation (LS-1)
Improvements to the
Existing pixel services
The so –called NSQP
Extraction April 2013
Return December 3rd 2013
Surface Clean Room
S. McMahon : Prague May 2014

13. The LHC roadmap (updated December 2013)
S. McMahon : Prague May 2014

14. The Future Phase II ATLAS Tracker General ITk Requirements
Guided by the physics requirements (see Phil’s talk)
Maintain and improve performance of existing ATLAS - ID to 3,000fb-1
Tracking to <m> ~ , expect 1000 tracks per unit of rapidity
Measure PT and direction of isolated particles (e and m),
reconstruct vertices of pile-up events
identify vertex of hard scatter
Identify secondary vertices in b-jets with high efficiency and purity
Measure tracks in the core of dense jets with good resolution
Identify the decays of t leptons, including impact parameter resolution
Reconstruct tracks associated with converted photons
The tracking must be robust against minor losses of acceptance
Tracking at first level of triggering
General ITk Requirements
Pixel
ITk Pixels.
Fluences up to 2x1016 1MeV neq/cm2
Low mass to reduce multiple scattering (<1.5%Xo Inner, < 2.0%Xo Outer)
Inner layers removable/replaceable in a short LHC stop (clam shell)
Approximately 10m2 of pixels
ITk Strips.
Fluences up to 2x1016 1MeV neq/cm2
Low mass to reduce multiple scattering (<2.5%Xo Inner, < 2.0%Xo Outer)
Approximately 200m2 of strips
Strip Detector
S. McMahon : Prague May 2014

15. The Future Phase II ATLAS Tracker General ITk Requirements
Guided by the physics requirements (see talk by Takanori Kono Monday)
Maintain and improve performance of existing ATLAS - ID to 3,000fb-1
Tracking to <m> ~ , expect 1000 tracks per unit of rapidity
Measure PT and direction of isolated particles (e and m),
reconstruct vertices of pile-up events
identify vertex of hard scatter
Identify secondary vertices in b-jets with high efficiency and purity
Measure tracks in the core of dense jets with good resolution
Identify the decays of t leptons, including impact parameter resolution
Reconstruct tracks associated with converted photons
The tracking must be robust against minor losses of acceptance
General ITk Requirements
ITk Pixels.
Fluences up to 2x1016 1MeV neq/cm2
Low mass to reduce multiple scattering (<1.5%Xo Inner, < 2.0%Xo Outer)
Inner layers removable/replaceable in a short LHC stop (clam shell)
Approximately 10m2 of pixels , good bandwidth required, low cost
Pixel
ITk Strips.
Fluences up to 2x1016 1MeV neq/cm2
Low mass to reduce multiple scattering (<2.5%Xo Inner, < 2.0%Xo Outer)
Approximately 200m2 of strips
Strip Detector
S. McMahon : Prague May 2014

16. The Future Phase II ATLAS Tracker General ITk Requirements
Guided by the physics requirements (see talk by Takanori Kono Monday)
Maintain and improve performance of existing ATLAS - ID to 3,000fb-1
Tracking to <m> ~ , expect 1000 tracks per unit of rapidity
Measure PT and direction of isolated particles (e and m),
reconstruct vertices of pile-up events
identify vertex of hard scatter
Identify secondary vertices in b-jets with high efficiency and purity
Measure tracks in the core of dense jets with good resolution
Identify the decays of t leptons, including impact parameter resolution
Reconstruct tracks associated with converted photons
The tracking must be robust against minor losses of acceptance
General ITk Requirements
ITk Pixels.
Fluences up to 2x1016 1MeV neq/cm2
Low mass to reduce multiple scattering (<1.5%Xo Inner, < 2.0%Xo Outer)
Inner layers removable/replaceable in a short LHC stop (clam shell)
Approximately 10m2 of pixels
Pixel
ITk Strips.
Fluences up to ~1x1015 1MeV neq/cm2
Low mass to reduce multiple scattering (<2.5%Xo Inner, < 2.0%Xo Outer)
Approximately 200m2 of strips , optimal cost choices
Strip Detector
S. McMahon : Prague May 2014

17. Baseline (LoI) Layout http://cds.cern.ch/record/1502664?ln=en
ITk
S. McMahon : Prague May 2014

18. Baseline (LoI) Layout http://cds.cern.ch/record/1502664?ln=en
ITk
Polyethane Moderator
Pixel Barrel x 4
Beam Pipe 33mm
Pixel Discs x 6
S. McMahon : Prague May 2014

19. Baseline (LoI) Layout http://cds.cern.ch/record/1502664?ln=en
PIXELS
8.2 m2
638x106 channels
Baseline (LoI) Layout
STRIPS
193 m2
74x106 channels
ITk
Stub
Layer x 1
Disc
Strips x 7
Long (23.8mm)
Strips x 2
Short (47.8mm)
Strips x 3
Pixel Barrel x 4
Pixel Discs x 6
S. McMahon : Prague May 2014

20. ITk - Radiation Fluences : 1 MeV neq.cm-2
Small radius dominated by flux from IP, larger radius significant contributions from cascades in
Calorimeter. Factor of flux at large radius with and without polymoderator
Simulations with FLUKA, for integrated luminositiy to 3,000fb-1
S. McMahon : Prague May 2014

21. ITk - Radiation Fluences : 1 MeV neq.cm-2
216kGy , 5.3x1014n cm-2
63kGy , 2.9x1014n cm-2
288kGy , 8.1x1014n cm-2
7.7MGy , 1.4x1016n cm-2
0.9MGy , 1.7x1015n cm-2
0.9MGy , 1.8x1015n cm-2
Simulations with FLUKA to 3,000 fb-1
S. McMahon : Prague May 2014

22. Performance of the ATLAS Phase II Tracker Compared with the existing ID + IBL
The expected performance of the ATLAS Phase II tracker compared to the performance
of the existing ID with the addition of the IBL.
sx(∞) refers to sx for PT → ∞ which allows one to remove the effect of material.
Performance parameters extracted from full ATLAS simulation including realistic service routing
and appropriate engineering considerations.
S. McMahon : Prague May 2014

23. Performance of the ATLAS Phase II Tracker Number of hits on Track
fakes
Sample of tt events
S. McMahon : Prague May 2014

24. Performance of the ATLAS Phase II Tracker Number of hits on Track
fakes
Sample of tt events
S. McMahon : Prague May 2014

25. Performance of the ATLAS Phase II Tracker Number of hits on Track
2.7 ?
Pattern Rec
Number of hits on muon tracks with PT > 5GeV as a function of pseudo-rapidity (Z=0,150mm)
Optimization in progress
S. McMahon : Prague May 2014

26. Construction of the ATLAS Phase II Tracker Detector Occupancies
Hit occupancies in different sub-detectors in the presence of 200 pile up events (peak around 0.9)
S. McMahon : Prague May 2014

27. Construction of the ATLAS Phase II Tracker ITk Material
Particular attention put into reducing the material in the tracker volume
Note that everywhere less than 0.7Xo while in ID (+IBL) ~1.2%
Achieved through carful choice of materials, service routing etc
S. McMahon : Prague May 2014

28. Alternate Layouts of Phase II Pixels
Conical (Barrel area 5.1 → 4.6 m2)
Fifth Pixel (optimized pattern rec)
Alpine (reduced area ∟ modules)
Work also ongoing with coverage out to a pseudorapidity of 4
S. McMahon : Prague May 2014

29. Alternate Layouts of Phase II Pixels
Conical (Barrel area 5.1 → 4.6 m2)
Fifth Pixel (optimized pattern rec)
Alpine (reduced area ∟ modules)
Work also ongoing with coverage out to a pseudorapidity of 4
S. McMahon : Prague May 2014

30. Alternate Layouts of Phase II Pixels
Conical (Barrel area 5.1 → 4.6 m2)
Fifth Pixel (optimized pattern rec)
Alpine (reduced area ∟ modules)
Work also ongoing with coverage out to a pseudorapidity of 4
S. McMahon : Prague May 2014

31. Alternate Layouts of Phase II Pixels
Conical (Barrel area 5.1 → 4.6 m2)
Fifth Pixel (optimized pattern rec)
Alpine Stave
Alpine (reduced area ∟ modules)
Work also ongoing with coverage out to a pseudorapidity of 4
S. McMahon : Prague May 2014

32. ATLAS Phase II Pixels Candidate Sensor Technologies
Choices largely driven by radiation tolerance requirements.
Standard Planar Sensors
Thin (150mm) High Resistivity Silicon [n-in-n (inner layers) or n-in-p (outer)]
Demonstrated to HL-LHC fluxes, Lots of experience and known vendors, mass production understood.
3D sensors
Used in the IBL, Low depletion voltage after irradiation, radiation tolerant
Diamond Sensors
Good radiation tolerance , Low capacitance (noise), no cooling, BUT expensive
Used in ATLAS beam monitoring, current and replacement in 2014
CMOS
Emergent technology (used on STAR and baseline for ALICE ITS), low power, varying degrees of sensor-hybrid integration (move away from standard implementation) to a more MAPS like approach, Cheap, Need to demonstrate radiation tolerance and readout speed.
Developments & Choices
Extensive R&D in progress, wait for appropriate time to make decisions
S. McMahon : Prague May 2014

33. ATLAS Phase II Pixels Electronics Technologies & Challenges
Specifications and Challenges.
Radiation tolerance, Low power consumption per channel, Low noise
High density of channels -> interconnects (TSV, bump-bonding=cost)
Move towards deep(er) sub-micron technologies 65nm. CERN - RD53 now established and operational.
On detector power conversion (DC-DC (baseline) and Serial Power under consideration)
FEI4
Currently being used for IBL, requirement well matched to outer layers of ITk pixel detector. To limit cooling requirements chip plus sensor power should not exceed 450mW/cm2.
RD65 to develop the next generation of chips
Choices
Note that we do not necessarily need the same technologies at large and small r
Extensive R&D in progress, wait for appropriate time to make decisions, cost is a driver
S. McMahon : Prague May 2014

34. ATLAS Phase II : Changes to trigger architecture
New design in time for Phase II
2-level system at Phase II
Phase-I L1 becomes Phase-II L0, new L1 includes tracking
Make use of improvements made in Phase 1 (NSW, L1Calo) in L0
Introduce precision muon and inner tracking information in L1
Better muon pT resolution
Track matching for electrons,…
Requires changes to detector FE electronics feeding trigger system
Drives ITk bandwidth requirements.
~3,200 optical links running at 5.6Gb/s
New (post Q2-2014) Requirements
Level-0
Rate ~ 1,000 kHz, Lat. ~10 ms
Level-1
Rate ~400 kHz, Lat. ~30 ms
S. McMahon : Prague May 2014

35. Ongoing Evolution of ITK Organisation
ITK chair
S. McMahon
Pixel
P. Morettini
M.G.Sciveres
Strips
I.Gregor
T. Affolder
Collaboration
Institute
Matters
Upgrade
Coordinator
P. Allport
Software
M. Elsing
Performance
& Layout
H.Hayward
A. Korn
Mechanics
A. Catinaccio
G. Viehhauser
Costing and
Finance
C. Buttar + SJM
Technical
Coordinator
B. Di Girolomo
Readout
A. Grillo
Rap-5 Task Force
K. Einsweiler
A. Henriques
Upgrade-Physics
Groups
P. Wells
Production Prep
& Test Facilities
Test Beams
Strip and Pixel coordinators, nominated by the community and
selected by 2 search committees by November 2013 AUW.
In December they were endorsed by the ITK community
S. McMahon : Prague May 2014

36. ITk: Draft Timetable : V1.0
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023 CB
Endorse
Pixel TDR
~ 2019
Preproduction
2-3 Years Pixel Production
ITk-IDR
APR-OCT
ITk Ready for Installation
PIXEL
UCG, LHCC,
Pixel-MOU, RRB
Begin
Integration
UCG, LHCC,
Strip-MOU, RRB
STRIP
ITk Commission
on Surface
ITk-Kick off
EB Approval
Strips TDR
~ 2018
Preproduction
3-4 Year Strip Production
YR=1 2 3 4 5 6 7 8 9 10
CB= collaboration board, EB=executive board, IMOU=interim memorandum of understanding, UCG=upgrade cost group, RRB= Resources review board, IDR=initial design review (internal), TDR=technical design report (external)
S. McMahon : Prague May 2014

37. Some Open Questions
Optimisation of the baseline layout for performance & cost
Can we reduce the material in the baseline?
Fewer layers, less material per layer?
Can we extend the existing layout to higher rapidities?
Effect of magnetic field, material and track length
Can we readout the tracker in a more efficient way
Now planned to readout the Pixels at L0
Does it make sense to readout the strips at L0
What are the penalties in terms of performance and cost?
What about alternate technologies
Can we exploit emergent detector technologies… CMOS?
What about new chip technologies 65nm
Can we source sensors (perhaps larger) from new vendors
S. McMahon : Prague May 2014

38. ATLAS Phase II : Other Challenges
Low mass “high” pressure ON detector cooling
Thermal and environmental management
local and global supports and mechanics
Alignment (with tracks and possibly lasers)
Integration, decommissioning, installation commissioning and possibilities for access maintenance and repair
Power supplies, cables, RAD hard fibres, possible reuse of existing servies.
Multiplexed powering (HV and LV)
DAQ, ATCA, Fast routing of data to L1-Track.
DCS, now becomes part of the detector-data-stream
Long term reliability issues...electronics, mechanics, infrastructure
Can we exploit 3D printing
Logistics ….
S. McMahon : Prague May 2014

39. Some Open Questions How are we going to build the new detector?
Time to develop and qualify processes and institutes
Which institutes will participate with what
Unique production challenges in the scale (strips = 200m2)
Difficult sociological challenges (see plot)
Have we got all of the bases covered at the appropriate level …
How would we respond if/when
The Funding Agencies ask “can you do the same job with 10% less….”
The LHC changes the machine parameters
Instantaneous luminosity increases
Changes in the shape of the luminous region
The interaction region changes and safety factors are decreased
We need to be able to answer these questions efficiently and accuratly
S. McMahon : Prague May 2014

40. The Open Questions Risks : How would we respond to
Instabilities in vendors (ref IBM)
The ongoing needs to support the existing ATLAS detector. …
All of the above takes an enormous amount of effort
Simulation is absolutely critical … still under resourced
Software is critical … current priority is for RUN-2
Having a balanced, appropriate R&D program, covering all of the bases at the appropriate level, on the appropriate time scale, and that starts NOW.
S. McMahon : Prague May 2014

41. ATLAS Phase II : Conclusions
The HL-LHC physics program will make very stringent demands of the ATLAS tracker beyond phase II
It will be necessary to replace the existing tracker to meet all of these requirements.
A baseline option (ITk) has been developed that builds on the performance and experience gained from the existing detector (ID) and our understanding of the harsh tracking environment at HL-LHC.
We are now moving towards a TDR (2016) and construction in 2017 and we are continuing to develop
There are a large number of challenges ahead of us.
S. McMahon : Prague May 2014

42. ITk – Institutions : 81 in total

43. The End
S. McMahon : Prague May 2014

44. ITk Mailing Lists : Please sign up
ITk general mailing list called:
atlas-ITk-general
This will be used for ITk collaboration wide mails.
Suggest everybody on ITk signs up to this (currently 18 69, 140 members)
open with approval of owner/administrators
Posting is restricted to the e-group members.
ITk TWIKI
S. McMahon : Prague May 2014

45. ITk - LOI Costing : Total 131 MCHF

46. Known (to me) Institutional Affiliations
ITk – Institutional : Affiliation
Known (to me) Institutional Affiliations
40 Strip
51 Pixel
17 Both
10 Not yet known.