Transcript Critical Questions for SiD
Critical Questions for SiD
A Configuration has been proposed for the Silicon Detector.
This serves as a starting point, and suggests the Critical Questions.
Effort must now systematically explore optimization.
Based on “Critical Questions for SiD,” J. Brau, M. Breidenbach, J. Jaros, H. Weerts, Aug 23, 2003 (It is assumed this list is incomplete) J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 1
LC Detector Requirements
Any design must be guided by these goals: a) Two-jet mass resolution comparable to the natural widths of W and Z for an unambiguous identification of the final states.
b) Excellent flavor-tagging efficiency and purity (for both b- and c quarks, and hopefully also for s-quarks). c) Momentum resolution capable of reconstructing the recoil-mass to di-muons in Higgs-strahlung with resolution better than beam energy spread. d) Hermeticity (both crack-less and coverage to very forward angles) to precisely determine the missing momentum . e) Timing resolution capable of separating bunch-crossings to suppress overlapping of events.
J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 2
Architecture arguments
Silicon is expensive, so limit area by limiting radius Get back BR 2 by pushing B (~5T)
This argument may be weak, considering quantitative cost trade offs. (see plots)
Maintain tracking resolution by using silicon strips Buy safety margin for VXD with the 5T B-field.
Keep (?) track finding by using 5 VXD space points to determine track tracker measures sagitta.
J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 4
Cost Trade-offs
Cost Partial R_Trkr
140.0
120.0
100.0
80.0
60.0
40.0
20.0
0.0
0.5
0.75
1
R_Trkr (m )
1.25
1.5
D
$ vs R_Trkr ~1.7M$/cm Cost Partial, Fixed BR^2
10 5 0 0 -5 30 25 20 15 1 2 3
B
4 5 1.85
1.75
1.65
1.55
1.45
1.35
6 1.25
Linear Power Radius
Delta $, Fixed BR
J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004
2 =5x1.25
2
5
Assumptions
Energy flow calorimetry is essential for good jet resolution We need to demonstrate this, and to determine rational major detector parameters that optimize it. Detector cost is constrained. This assumption will not be buttressed by simulation, but is considered reasonable by most. The energy flow demonstration is a simulation and reconstruction strategy issue, as are most of these questions. However, there are a few specific hardware developments that are crucial to determining “rational major detector parameters”.
J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 6
Tracking
Tracking for any modern experiment should be conceived as an integrated system, combined optimization of: the inner tracking (vertex detection) the central tracking the forward tracking the integration of the high granularity EM Calorimeter Pixelated vertex detectors are capable of track reconstruction on their own, as was demonstrated by the 307 Mpixel CCD vertex detector of SLD, and are being developed for the linear collider Track reconstruction in the vertex detector impacts the role of the central and forward tracking system J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 7
Silicon Tracking
• Superb spacepoint precision allows linear collider tracking measurement goals to be achieved in a compact tracking volume • Compact tracker makes the calorimeter smaller and therefore cheaper, permitting more aggressive technical choices (assuming cost constraint) • Robust to spurious, intermittent backgrounds (esp. beam loss) extrapolated from SLC experience • linear collider is not storage ring J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 8
Calorimetry
Current paradigm: Particle/Energy
Flow (unproven)
Jet resolution goal is 30%/ E In jet measurements, use the excellent resolution of tracker, which measures bulk of the energy in a jet Neutral Hadrons
EM
Charged Hadrons
Headroom for confusion Particles in Jet
Charged Photons Neutral Hadrons
Fraction of Visible Energy
~65% ~25% ~10%
Detector
Tracker ECAL ECAL + HCAL
Resolution
< 0.005% p T negligible ~ 15% / E ~ 60% / E
~ 20% /
E
J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 9
Energy/Particle Flow Calorimetry
Identify EM clusters not associated with charged tracks (gammas) Follow charged tracks into calorimeter and associate hadronic showers Remaining showers will be the neutral hadrons J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 10
Calorimeter Questions
For a fixed detector technology, bigger appears to be better. There is general agreement that the numerator of an energy flow figure of merit is BR 2 .
R is the outer radius of the tracker or the inner radius of the EMCal, probably different by about a cm.
If the cost of the calorimeter can be reduced without affecting performance, then BR 2 can be increased.
Therefore, primary questions are 1. Can the number of layers of Si in the EMCal (currently 30) be reduced?
2. Do we need 30 radiation lengths? Is 25 or 20 enough?
3. Would a tungsten based HCal with 2X 0 thick W ease the EMCal requirements?
J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 11
EM Calorimetry
Physics with isolated electron and gamma energy measurements require ~10-15% /
E
1% Particle/Energy Flow requires fine grained EM calorimeter to separate neutral EM clusters from charged tracks entering the calorimeter
Small Moliere radius
Tungsten
Small sampling gaps – so not to spoil R M Separation of charged tracks from jet core helps
Maximize BR 2
material Iron Lead Tungsten Uranium
Natural technology choice – Si/W calorimeters
Good success using Si/W for Luminosity monitors at SLD, OPAL, ALEPH Oregon/SLAC/BNL CALICE
Alternatives – Tile-Fiber (challenge to achieve required granularity) Scintillator/Silicon Hybrid Shaslik Scintillator Strip R M 18.4 mm 16.5 mm 9.5 mm 10.2 mm
J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 12
Energy Flow Issues
Using the latest version of the parametric detector calculator, increasing the tracker radius (from 1.25 m) at fixed B=5T costs about $2.1M/cm. If BR 2 is held fixed at 7.8 Tm 2 costs about $0.6M/cm. , then increasing the tracker radius (from 1.25 m) Note that the baseline design of R =1.25 m and B=5T is a cost minimum if B=5T is considered a maximal field. So assuming an EMCal with gaps of 1 mm and pixels small compared to the Moliere radius, and sampling often in depth, then: 4. Are the EMCal assumptions above realizable (Physical prototype required)?
5. Is BR 2 =7.8 sufficient for the physics benchmark processes?
6. Is the improvement expected from increasing R at fixed BR would this improve things ?
2 justified by the improvement in physics benchmark performance? Why 7. Can a reasonable energy flow figure of merit beyond BR 2 be demonstrated by simulation and reconstruction by early 2005? This should be analogous to understanding the performance variation with B, R, and the calorimeter properties. It is likely that calorimeter means both EMCal and HCal. Are there issues for the z position of the forward calorimeters.
J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 13
Silicon/Tungsten EM Calorimeter
SLAC/Oregon/BNL
Conceptual design for a dense, fine grained silicon tungsten calorimeter well underway First silicon detector prototypes are in hand Testing and electronics design well underway Test bump bonding electronics to detectors by end of ’04/early ‘05 Test Beam in ’05/’06 J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 14
Silicon/Tungsten EM Calorimeter (2)
Pads ~5 mm to match Moliere radius Each six inch wafer read out by one chip < 1% crosstalk Electronics design Single MIP tagging (S/N ~ 7) Timing < 5 nsec/layer Dynamically switchable feedback capacitor scheme (D. Freytag) achieves required dynamic range: 0.1-2500 MIPs Passive cooling – conduction in W to edge Angle subtended by R M GAP J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 15
Hadron Calorimeter
The baseline assumption is that the HCal is inside the coil, and that it is 4 l thick (nominally 34 layers of stainless 20 mm thick with 10 mm gaps).
8. Is this HCal configuration sufficient for the benchmark physics processes?
a. Is this the “right” radiator? How about tungsten?
b. How about more sampling, and is 4 l sufficient?
c. The gaps are expensive because they drive out the coil radius. Could they be reduced?
d. 1 to 2 cm square pixels have been assumed. Is this right, particularly if the HCal density is increased?
e. Can thin, cheap, reliable, good resolution detectors be made?
(Physical Prototype required) (Note that Si is out of the question!!) J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 16
Digital Hadron Calorimetry
1 m 3 prototype planned to test concept
Lateral readout segmentation: 1 cm 2 Longitudinal readout segmentation: layer-by layer Gas Electron Multipliers (GEMs) and Resistive Plate Chambers (RPCs) being evaluated
Objectives
Validate RPC approach (technique and physics) Validate concept of the electronic readout Measure hadronic showers with unprecedented resolution Validate MC simulation of hadronic showers Compare with results from Analog HCAL J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004
Argonne National Laboratory Boston University University of Chicago Fermilab University of Texas at Arlington
17
Superconducting Solenoid
The superconducting solenoid is large, with more than a GJ of stored energy. There are concerns that the hoop stress in a 5T, Rcoil=2.6 m might be excessive. In addition, the coil is a major cost driver, and it thus affects directly what BR 2 might be within the cost constraints.
9. Are there serious technical problems with a (thick) solenoid of these nominal parameters? Does the addition of serpentine correction coils for a crossing angle introduce horrible problems?
10. What is a rational cost parameterization for coils of this scale?
J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 18
Silicon Tracking
The barrel tracking and momentum measurement are baselined as 5 layers of pixellated vertex detector followed by 5 layers of Si strip detectors (in ~10 cm segments) going to 1.25 m. The momentum resolution for found tracks seems excellent.
11. Does it need to become more complicated?
12. Develop a baseline for the Forward direction.
13. Does this system find tracks well? What about machine and physics backgrounds?
14. Are there issues regarding K 0 ’s and Λ’s i.e. can they be detected efficiently ?
15. Demonstrate (if true) the need to minimize tracker material to minimize multiple scattering.
J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 19
Silicon Tracking
With superb position resolution, compact tracker is possible which achieves the linear collider tracking resolution goals Compact tracker makes the calorimeter smaller and therefore cheaper, permitting more aggressive technical choices (assuming cost constraint) Linear Collider backgrounds (esp. beam loss) extrapolated from SLC experience also motivate the study of silicon tracking detector, SiD Silicon tracking layer thickness determines low momentum performance 3 rd dimension may be achieved with segmented silicon strips, or silicon drift detectors (1.5% / layer) (TPC) J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 20
Optimizing the Configuration
Central Tracking (Silicon)
support Cooper, Demarteau, Hrycyk R. Partridge J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 H. Park 21
Silicon Tracking w/ Calorimeter Assist
Primary tracks started with VXD reconstr.
V0 tracks reconstructed from ECAL stubs J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 E. von Toerne 22
Other Important Questions
There are many other important questions that must be studied, but still do not seem to drive the basic design or challenge the fundamental strategy of SiD. For illustration, some of these questions are: 1. What is a rational technology and a more detailed design for the VXD?
2. What is the technology for the muon trackings? Should it be the same as the HCal?
3. What is a design for the very forward calorimetry?
J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 23
Very Forward Instrumentation
• •
Hermiticity depends on excellent coverage in the forward region, and forward system plays several roles
maximum hermiticity precision luminosity shield tracking volume monitor beamstrahlung
High radiation levels must be handled
• 10 MGy/year in very forward detectors J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 24
Machine Detector Interface
A critical area of detector R&D which must be optimized is where the detector meets the collider
Preserve optimal hermiticity Preserve good measurements Control backgrounds Quad stabilization
20 mr crossing angle, silicon detector J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 25
Summary
A systematic investigation of the Silicon Detector is needed soon.
An initial list of Critical Questions has been constructed.
What are the additional Critical Questions which should be added to the high priority list?
J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 26
J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 27
Collider Constraints
Linear Collider Detector R&D has had to consider two different sets of collider constraints: X-Band RF and Superconducting RF designs With the linear collider technology selection, the detector efforts can concentrate on one set of parameters The ILC creates requirements similar to those of the TESLA design #bunch/train #train/sec bunch spacing bunches/sec length of train train spacing crossing angle X-Band GLC/NLC 192 SuperRF TESLA 2820 150/120 1.4 nsec 28800/23040 5 337 nsec 269 nsec 6.6/8.3 msec 7-20 mrad 14100 950
m
sec 199 msec 0-20 mrad
J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 28
Inner Tracking/Vertex Detection
Detector Requirements
Excellent spacepoint precision (
< 4 microns
Superb impact parameter resolution (
5µm
)
10µm/(p sin 3/2
)
) Transparency
( ~0.1% X 0 per layer
) Track reconstruction (
find tracks in VXD alone
)
Concepts under Development for Linear Collider
Charge-Coupled Devices (CCDs) demonstrated in large system at SLD Monolithic Active Pixels – CMOS (MAPs) DEpleted P-channel Field Effect Transistor (DEPFET) Silicon on Insulator (SoI) Image Sensor with In-Situ Storage (ISIS) HAPS (Hybrid Pixel Sensors) J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 29
Inner Tracking/Vertex Detection (CCDs)
Issues
Readout speed and timing Material budget Power consumption Radiation hardness
R&D
Column Parallel Readout ISIS Radiation Damage Studies
SLD VXD3
307 Mpixels 5 MHz 96 channels 0.4% X 0 / layer ~15 watts @ 190 K 3.9 m m point res.
av. - 2 yrs and 307 Mpxl
J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 30
Column Parallel CCD
SLD Vertex Detector designed to read out 800 kpixels/channel at 10 MHz, operated at 5 MHz => readout time = 200 msec/ch Linear Collider demands 250 nsec readout for Superconducting RF time structure Solution: Column Parallel Readout LCFI (Bristol, Glasgow, Lancaster, Liverpool, Oxford, RAL) J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 (Whereas SLD used one readout channel for each 400 columns) 31
Column Parallel CCD (2)
Next Steps for LCFI R&D
Bump bonded assemblies Radiation effects on fast CCDs
High frequency clocking Detector scale CCDs w/ASIC & cluster finding logic; design underway – production this year In-situ Storage Devices
Resistant to RF interference Reduced clocking requirements
J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 32
Image Sensor with In-situ Storage (ISIS)
EMI is a concern (based on SLC experience) which motivates delayed operation of detector for long bunch trains, and consideration of ISIS Robust storage of charge in a buried channel during and just following beam passage (required for long bunch trains) Pioneered by W F Kosonocky et al IEEE SSCC 1996, Digest of Technical Papers, 182 T Goji Etoh et al, IEEE ED 50 (2003) 144; runs up to 1 Mfps.
•
charge collection to photogate from 20 30
m
m silicon, as in a conventional CCD
•
signal charge shifted into stor. register every 50
m
s, providing required time slicing
•
string of signal charges is stored during bunch train in a buried channel, avoiding charge-voltage conversion
•
totally noise-free charge storage, ready for readout in 200 ms of calm conditions between trains for COLD LC design
•
particles which hit the storage register (~30% area) leave a small ‘direct’ signal (~5% MIP) – negligible or easily corrected
J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 33
Radiation Effects in CCDs
Drift of charge over long distance in CCD makes detector very susceptible to effects of radiation: • Transfer inefficiency • Surface defects N. Sinev et al.
Traps can be filled •
neutrons induce damage clusters
•
low energy electrons create point defects – but high energy electrons create clusters – Y. Sugimoto et al.
Hot pixels •
number of effective damage clusters depends on occupation time – some have very long trapping time constants – modelled by K. Stefanov
•
Expect ~1.5x10
11 /cm 2 /yr of ~20 MeV electrons at layer-1
•
Expect ~10 9 /cm 2 /yr 1 MeV-equivalent dose from extracted beamline
J. Brau - ACFA Workshop, Taipei - Critical Questions November 9, 2004 34