Fallback Plans

1
Fallback Plans

• John Womersley

2
How did we get here?

• December 1999
• letter from Directorate to DØ asking us to
  investigate fallback options
• January 2000
• established a Contingency Planning Panel within
  DØ to advise the spokesmen, solicit input from
  the collaboration, and review the documents
• Ken Johns (chair), Alice Bean, John Ellison, Paul
  Grannis, Nick Hadley (physics coordinator),
  Sijbrand de Jong, Greg Landsberg, Darien Wood
• Weekly open meetings to present studies, discuss
  issues

3
Documents

• January 2000
• Looked at Options
• DØ Note 3724 Fallback Options for the DØ
  Upgrade
• March/April 2000
• Formed a plan
• DØ Note 3741 DØ Schedule and Contingency Plan
• Physics Justification
• Physics Impact of Tracking Detector Contingency
  Plans
• these three documents are part of the materials
  made available to the committee

4
Our goals

• We fully intend to complete the whole DØ detector
  for the start of Run II on March 1, 2001
• In exploring fallbacks and contingency plans, we
  have not learned of anything that would prevent
  this from being accomplished
• We have taken the view that developing these
  fallback strategies will help us accomplish this
  goal

5
Where are we exposed?

• In general the critical areas are
• Silicon Tracker
• Electronics
• Installation (time constraints)
• The following are either complete or schedule is
  not an issue
• central muon system
• calorimeters
• solenoid
• central preshower
• forward preshower and ICD
• fiber tracker detector assembly
• forward muon detector assembly
• online system
• Not an issue, despite some concerns in January
• availability of sufficient SVX chips

6
DØ central tracker
7
Physics goals for tracking

• The Run II Physics menu is a broad one
• QCD, B-physics, precision EW, top, SUSY, Higgs .
  . .
• High pT processes
• typically require good performance at moderate to
  high momenta and central rapidities (?lt 2)
• SUSY multilepton signals
• demand pT 5 GeV and coverage in ? up to the
  limits of electron ID (?lt 2.5)
• B-physics
• requires even lower pT 1 GeV and maximal
  coverage in ?
• The environment
• vertex ? 30 cm in z
• radiation 400 krad per fb-1 at the inner layer

8
DØ Silicon Tracker
Six Barrels Four layers barrels 2-5 two 2º, two
90º stereo layers barrels 1-6 two 2º stereo,
two axial layers
12
11
10
9
6
5
4
4
3
2
3
1
2
1
8
7
6
5
Twelve F-disks each with 12 double-sided wedges
15º stereo
Four H-disks each with 24 single-sided wedges
glued back-to-back
9
SMT acceptance

• Barrels
• Provide acceptance at relatively low rapidity and
  for central zvtx
• Disks
• Primary tracking system at moderate to large
  rapidity (2 lt?lt 3)
• Cover gaps between barrel detectors for ?lt 1.5

10
Physics studies

• Evaluated existing design studies
• Heinson, Heintz, 1993-98
• To assess the impact of descope options, new
  studies were carried out
• Barberis, Demina, Heinson, Jesik, Kulik, Khanov,
  Whiteson
• MCFAST
• GEANT DØ C tracking code
• GEANT silicon groups standalone tracking code
• Have investigated single tracks, top events, and
  B ? ? KS

11
B ? ? KS
MCFAST, R. Jesik
12
Barrels

• Small angle stereo ladders (used in all barrels)
• crucial for pattern recognition no viable
  fallback except to use lower quality devices if
  necessary
• layer 2 detectors suffer significant radiation
  damage use higher depletion voltage devices in
  this layer
• Single-sided ladders (barrels 1 and 6)
• only concern is non-planarity of some ladders
• impact on silicon track trigger has been studied
• procedure developed to improve planarity

13
Barrels 90 ladders

• 90 stereo ladders (barrels 2-5)
• several concerns at the start of this process
• p-stop shorts (manufacturing defects)
• fears about ability to bias at sufficiently high
  voltage to withstand the radiation dose
• high occupancy raised questions about usefulness
  of 90 strip information to the pattern
  recognition
• it was therefore suggested that we resort to
  using single-sided axial detectors everywhere

14
Barrels 90 ladders

• Studies showed
• p-stop defects cause only 1-2 loss of strips
• radiation damage studies show layer 1 probably OK
  to 3-4 fb-1
• subsequent studies of B-physics observables show
  that the 90 strips are crucial
• omission of 90 strip information worsens the
  proper time resolution from 95 fs to 150 fs
• reduces effective flavor tagging efficiency by
  10 (worsens sin2? error by 10)
• We therefore have no plan to discard the 90
  double sided ladders

15
Close up the barrels?

• There is a loss of acceptance between the barrel
  modules
• gaps in z are required for the F-disks
• if the barrels were pushed closer together, a
  3-5 gain in acceptance for high pT tracks could
  be realized (for barrel hits alone)
• BUT
• this would still leave gaps due to inactive
  detector area
• in principle all the tracks passing through these
  gaps can be recovered using F-disk hits
• On April 26, we committed to cut the openings in
  the support cylinder in such a way that the
  barrels are positioned as designed
• accommodate full complement of central F-disks
• precludes possibility of closing the gaps between
  the barrels
• this has now been done

16
F-disks

• Concerns
• Component deliveries
• Assembly
• these are the most complex combination of
  sensors, HDI and SVX chips
• Physics impact
• central disks
• acceptance of central tracks without these
  disks, we would lose 15 of fully reconstructed B
  ? ? K0S events ( 3-4 ? single track
  inefficiency)
• expect similar problems whenever three or four
  tracks are required e.g. secondary vertex
  tagging of high pT b-jets
• end disks
• in B ? ? K0S, primarily influence tagging
  efficiency and proper time resolution
• Disks 1 and 12 carry radiation sensors, and disks
  2 and 11 carry kinematic mounts

17
H-disks

• Affect acceptance and resolutions for ?? 2
• Production is well advanced with little schedule
  risk
• Modules can be installed very late, if necessary

18
Fallback Strategy

• Small angle stereo ladders
• install detectors with imperfections if necessary
• we have already done this in barrel 1
• 90 stereo ladders
• install detectors with one p-stop defect (1-2
  bad channels) if necessary
• Single sided ladders
• no issue
• F-disks
• Decision of the Contingency Planning Panel
  (April 19)if there is a need to reduce the
  complement of F-disks installed, sacrifice disks
  3 10 leaving a triplet of disks at each end

19
First Barrel Assembly
Grade B (electrical)
Grade C (electrical)
20
Reducing F-disk complement
Remove disks 3 and 10

• IF DONE, THIS WOULD BE A NON-RECOVERABLE DESCOPE
• The first decision point for this is July 1

21
Decision Triggers
22
June 1

• Schedule calls for
• 2 barrels, 4 disks complete
• Actual status was
• 1.5 barrels, 2.5 disks
• Actions required
• Review likelihood of SMT-S installation by
  8/1/00. Consider early reduction of F disk
  complement if 3 disks are not complete
• What we did
• Project managers and spokesmen discussed progress
  with silicon management
• all parts are in hand for first six disks (first
  half SMT)
• support rings were a holdup, but now proceeding
• F-disk managers believe that all six disks can be
  completed by August 1
• we will revisit the status on July 1, as called
  for in our plan

23
July 1

• Schedule calls for
• 3 barrels, 5 disks complete
• Actions required
• If two barrels are not complete
• Delay SMT-S installation date to 10/1/00.
• If 3 F disks are not complete
• reduce F disk complement by two to 10.
• Silicon managers anticipate that the second
  barrel and two more F-disks should be complete
  by this date, so we do not presently anticipate
  triggering the descope.

24
Installation

• Over the last six months a great deal of effort
  has gone into understanding and tuning the
  installation schedule
• We have been able to gain significant schedule
  contingency through logistical changes and by
  splitting the SMT into two halves
• Nonetheless, the present schedule does not have a
  lot of slack left
• Exposure in
• possible late delivery of silicon
• hookup time for CFT and SMT
• forward muon installation
• engineering design and fabrication still required
  on supports
• delayed availability of electronics and VLPC
  cassettes
• mainly impacts commissioning activities, and
  complicates hookup

25
Silicon Installation

• Baseline dates for detector completion (ready to
  move to DØ)
• SMT-S August 10, SMT-N October 27
• silicon group internal goal is completion of
  SMT-S by August 1
• Baseline dates for start of installation in DØ
• SMT-S September 21, SMT-N November 2
• Latest possible dates for installation in DØ
  consistent with roll-in
• SMT-S November 1, SMT-N January 1
• SMT installation depends on completion of the CFT
  waveguide installation two shift operation was
  assumed, could go to three if necessary, or work
  on N and S simultaneously
• if SMT-N comes this late, hookup will need to be
  done in the collision hall
• Ultimate option
• install SMT-N after roll-in

26
Forward Muon Installation

• We have exercised one fallback option already
• June 1 decision NOT to install the south muon
  truss in the hall during August
• it was felt better to use resources elsewhere
• We have prioritized and largely serialized the
  installation sequence
• A, C, B
• B layer occupies the last two months 11/15-1/15
• ultimately we have the option of installing layer
  B once the detector is rolled into the collision
  hall
• some additional engineering will be required

27
Electronics

• We expect all readout electronics to be in place
  for March 1 in any case, delays do not impact
  roll-in
• We have developed alternative strategies for
  hookup and commissioning
• Tracking detectors
• CFT commissioning plan assumes no production
  electronics are available use stereo boards
  (cosmic ray test electronics)
• SMT will use prototype interface boards
• Muon system
• extensively tested before being moved to DØ
• will commission with limited number of prototype
  boards
• Calorimeter
• 5000 channels (one quadrant) will be available in
  September
• Ultimate fallback
• start running with a partially instrumented
  detector

28
Conclusion

• We have developed a set of fallback options and
  contingency plans and have started to implement
  them.
• The major decisions remaining are in the silicon
  tracker and in the installation.
• We have learned a great deal about the detector
  and we have also made significant progress in
  production since December 1999 when this process
  was begun.
• In light of this progress, we do not believe that
  we will need to exercise any of these options
  nonetheless, we now know what to do if we have
  to.
• The DØ detector will be installed and ready for
  Run II on March 1, 2001