Task%20C%20Activities%20in%20the%20CMS%20Experiment

1
Task C Activities in the CMS Experiment

• Greg Landsberg
• DOE Site Visit
• September 28, 2006

2
Brown and CMS

• Next energy frontier and a natural way for our
  group to evolve is LHC
• While have realized this a while ago, it would
  have been a hard decision to move part of our
  DOE-supported group away from DØ w/o stretching
  thin our major responsibilities there
• The NSF CAREER Award won by GL in June 2003
  allowed us to expand in the new direction w/o
  compromising our DØ strength
• Why CMS?
• Better detector
• Strong Fermilab presence allowing for an
  adiabatic transition
• December 2003 formal application to join CMS
• March 2004 presentation of our application at
  the CMS Week
• June 2004 Brown was voted in and became a full
  member of CMS
• At the same time, Ryan Hooper, a new RA to work
  on DØ and CMS has been hired using the NSF Grant
• Based on our strong ties with Fermilabs SiDet,
  we have joined Tracker Outer Barrel construction
  project
• Where do we stand now, two years since weve
  joined?

3
CMS Soon to be Lowered!
September 2006 CMS Week
4
CMS Explained
Jos Engelen, SLAC 2006 Summer School
5
CMS All-Silicon Tracker
Brown group involvement
Outer Barrel (TOB)
Pixels
End Caps (TEC 12)
Inner Barrel Disks (TIB TID)
2,4 m
5.4 m
volume 24.4 m3 running temperature 20 0C
6
TOB Construction

• Major hardware project accomplished via joint
  effort of Brown, Fermilab, KU, Rochester, UCSB,
  UIC, and UCR
• Two mirror production facilities Fermilab and
  UCSB
• Some TOB statistics
• Total number of modules 5208 5 spares
• Total number of rods 688 5 spares
• Total silicon area 1000 ft2
• Total silicon weight 100 lbs
• Brown involvement QA, rod assembly, single-rod
  test stand
• People GL, Ryan Hooper, Hai Duong Nguyen,
  Patrick Tsang, Scott Wolin, several more SS
• Rod production took place in FY2006
  (November-July)

• Uncovered and fixed many problems since mid 2005
• Common mode noise in ST sensors replaced w/
  Hamamatsu
• Wire bond breakage due to vibration module
  encapsulation
• I2C communication problems redesigned cards
• Degradation of the Ag-based epoxy bias connection
  switched to wire-bonding
• Sensor damage due to discharges between the
  ground and the silicon surface encapsulated
  wire bonds and modified HV power supplies
• Despite the unexpected delays, the project has
  been successfully finished
• All the rods have been assembled, tested, and
  delivered to CERN on time!

7
Rods Single-Rod Test Stand
Single-Rod Test Stand (SRT),and the Brown Box
for reading rod temperature.Over 700 rods went
through SRT
An assembled rod and its readout end
8
Typical SRT Rod Failure
SRT Opto Scan and Noise
Saturated AOH laser channel
9
TOB at CERN Rods Being Installed
10
Actual Rod Installation
100 mm axial precisionover the length of therod
(3ft)! 0.5 mil/sensor!
11
LPC Activities

• LHC Physics Center (LPC) at Fermilab has been
  founded in 2004 as a Mortar and Brick
  foundation to accommodate critical mass of
  experts and support LHC physics efforts by the US
  community
• Brown group was one of the first groups to join
  (GL served on the Advisory Board in 2004-2005
  Yuri Gershtein, than a Brown RA, became a
  convener of the e/g group)
• Many successes, from thorough work on development
  and optimization of ID algorithms within the new
  CMS software framework, to support of MC physics
  studies by students and software tutorials
• Helped our group to move to CMS adiabatically,
  without compromising the DØ connection

• LPC grew significantly over the last two years
  still remains to be seen if the critical mass of
  experts can be achieved during CMS running
• Could serve as a good alternative to maintaining
  the core of the group at CERN (COLA, remoteness,
  need to relocate several people at once)
• 2007 LHC pilot run will serve as a proof of
  principle
• In 2005-2006 GL took a year-long sabbatical
  supported by LPC and has been appointed as the
  LPC Trigger Group convener
• Meenakshi Narain just started as an LPC b-tagging
  Group convener
• Maintain LPC office space to provide students,
  postdocs, and commuting faculty with excellent
  working conditions

12
Triggering at Hadron Colliders

• ee- colliders low total cross section, low
  rates
• Trigger pretty much on everything, perhaps with
  the exception of very forward processes
  (low-angle Bhabha)
• Hadron colliders enormous cross section,
  unattainable rates
• Trigger is very selective
• Only small fraction of collisions is written to
  tape
• Additional complications due to pile-up
• LHC
• stot 110 mb, sin 70 mb
• L 1034 cm-2s-1 10 nb-1s-1
• 25 ns bunch crossing
• Total rate 109 s-1 or 20/crossing
• An order of magnitude more complex triggering
  than that at the Tevatron

PDG
13
Trigger Architecture

• Must reduce 2.5-40 MHz of input interactions to
  50-100 Hz
• Do it in steps/successive approximations
  Trigger Levels

x400 rejection
x1000 rejection
14
CMS Level-1 Trigger Scheme
Muons
Electrons, Photons, Jets, MET
3lt?lt5 ?lt3 ?lt3
?lt2.1 0.9lt?lt2.4 ?lt1.2
15
LPC Trigger Group

• Charge in concert with the international CMS
  Online Selection PRS group
• Simulate, develop, and maintain L1 and High-Level
  Trigger algorithms
• Work on creating realistic trigger table for the
  startup pilot run, as well as for low and high
  luminosity physics running
• Maintain and operate Remote Operations Center
  (ROC) at Fermilab a mirror control room for the
  experiment
• Participate in the MTCC (Magnet Test Cosmic
  Challenge) shifts and data monitoring and
  analysis
• LPC Trigger Group Statistics
• About 25 active members
• Regular biweekly meetings (one later this
  afternoon!)
• Two successful workshops (L1 HLT) in the past
  year
• Deep connection with Online Selection PRS group

16
LPC Trigger Group Activities

• Contribution to vol.2 of the CMS Physics TDR
  (Brown, Fermilab, Wisconsin)
• Trigger tables and rates (Brown, Fermilab,
  Wisconsin)
• Optimization of the L1 Calorimeter trigger
  (Brown)
• L1 calorimeter calibration and jet energy
  corrections (Brown, Wisconsin)
• Implementation of the L1 calorimeter emulator in
  the CMSSW framework (TAMU, Wisconsin)
• HLT Jet/MET trigger suite implementation and
  timing studies (Brown, UIC)
• Remote Operations and Control (Fermilab,
  Maryland Brown is getting involved)
• Test beam and cosmics data analysis (Fermilab,
  UIC)
• Calorimeter noise studies (Brown, Rochester, UIC)
• Minbias trigger design (UIC)

17
L1 Calorimeter Trigger Optimization

• Pursued by GL and Sara Vanini
• Focus on the basics
• Systematic studies of noise
• Adjusting trigger thresholds to match noise
• Optimizing trigger algorithms
• Improving turn-on and resolution
• Improving MET resolution
• In the process uncovered and fixed many trigger
  simulation bugs and cleaned up ORCA code and
  propagate fixes into CMSSW
• Example
• greatly improved single jet trigger turn-on an
  important tool for multiobject triggers (ejets,
  mjets, etc.)

Original defaultthresholds
ETgen, GeV
Optimized thresholds
ETgen, GeV
18
Hardware Implementation

• Proposed optimized thresholds have been endorsed
  and approved by the CMS HCAL Electronics and
  Trigger groups
• They are currently being implemented in firmware
  and the new L1 emulator code
• Brown group spearheads this effort and will redo
  simulations with the new software
• Also work on implementing and simulating dynamic
  baseline subtraction scheme
• Energy is determined as a weighted sum E ?fiEi
  over several time-slices

TB04
5 70 23 2

• Automatic baseline subtraction ?fi 0
• Being implemented for HCAL fi -1.0, -1.0, 1.0,
  1.0
• Will require slight timing realignment (1 ns)
• Robust against coherent noise and hot cells

19
L1 Calorimeter Trigger Calibration

• Problem CMS calorimeter is
• Non-linear
• Non-compensated
• Has h-dependent response
• Goal correct for these deficiency via proper
  calibration either at the jet or the trigger
  tower level
• New L1 JES corrections
• Fit (or take the mean) of ETgen ETL1
  distributions in bins of ET, h

• Effect on MET in Wjets
• Left ME resolution
• Right ME pull
• Corrections remove pull ?, but worsen the
  resolution by similar amount ?
• Next step introduce new dimension in tower
  calibration e/h

GL, Scott Wolin
GL, Sara Vanini
20
Future Plans

• CMS Hardware Commissioning
• Our current CMS hardware project has been a
  success, but it came to an end
• Important to maintain hardware experience in the
  group, especially given new students and postdocs
  joining the group
• Want to maintain close ties with the US Silicon
  Consortium work well together, common goal in
  bringing tracker online
• Will participate in TOB commissioning via ROC in
  MTCC-II (later this year) and underground SX5
  cosmic running (2007)
• Consider manning test beam and commissioning
  shifts at CERN via short-term visits

• SLHC Upgrade
• A lot of opportunities for new major hardware
  efforts
• Participate in the SLHC Steering Group meetings
  and SLHC workshops
• CMS SLHC LoI will go to the funding agencies
  later this year
• Hope for enthusiastic support from you guys!
• Discussed participation in the following areas
• Construction of a part of the CMS silicon tracker
  replacement
• L1 tracking trigger design and construction
  (hardware firmware)
• More in Meenakshis presentation

21
The LHC Schedule Winter 2006

• Physics running 140 days/year
• ATLAS/CMS running 100 days/year
• Typical efficiency for physics 40
• Effective ATLAS/CMS running time/year 1000
  hours 4 x 106 s 4 x 1038 cm-2 4 x 1014 b-1
  400 pb-1 _at_ 1032cm-2s-1
• Note that the schedule below R. Bailey, Aspen
  2006 is all goes well scenario

2010, 25ns L1x1034 cm-2s-1?Ldt 40 fb-1
Pilot run, 75ns L5x1030 cm-2s-1?Ldt 20 pb-1
2008, 75/25ns L3x1032 cm-2s-1?Ldt 1.2 fb-1
2009, 25ns L1x1033 cm-2s-1?Ldt 4 fb-1
22
Revised LHC Schedule (Summer 2006

• Pilot 2007 Run
• Starts 2 months later than previously expected
• Accelerator runs at injection energy (450 x 450
  GeV pp)
• Expected luminosity 1029 cm-2s-1
• Collision data sometimes in November 2007 50
  nb-1
• Three-month shutdown following with 14 TeV data
  in Spring 2008
• The goal to deliver a few fb-1 by the end of the
  run
• The rest of the schedule stays the same

23
CMS Physics

• All what we have been doing is really driven by
  our desire to become major players in the CMS
  physics, in the same way we maintained physics
  leadership within DØ over the last 15 years
• Currently focus on low-level task optimization,
  validation, and certification of trigger,
  reconstruction and particle ID
• See Tulikas talk on this
• Actively contributed to the Volume 2 of the CMS
  Physics TDR
• Getting involved in the Pilot 2007 Run data
  analysis as a preparation to the 2008 physics run
• Picking/negotiating physics topics to work on
  from rediscovering SM with the first data to
  searches for black holes at the LHC
• Use our two-decade experience and leading role in
  the Tevatron physics as a guiding star

24
(Wo)manpower on CMS Doubling!

• FY2006
• Greg Landsberg PI (90)
• Tulika Bose RA (50)
• Ryan Hooper RA, NSF (75)
• Sara Vanini VRA (75)
• Zongru Wan VRA (20)
• Hai Duong Nguyen GS, MO (100)
• Patrick Tsang SS (25)
• Scott Wolin SS, UTRA (20)
• Total 4 FTE

• FY2007
• Dave Cutts PI (50)
• Greg Landsberg PI (80)
• Meenakshi Narain PI (50)
• Tulika Bose RA (75)
• Leonard Christofek RA (50)
• Selda Esen RA, NSF (100)
• Aram Avetisyan GS (75)
• Paul Huwe GS (40)
• Hai Duong Nguyen GS (100)
• Patrick Tsang GS (100)
• 2-3 SS (50-75)
• Total 8 FTE

• FY2005
• Greg Landsberg PI (30)
• Ryan Hooper RA, NSF (100)
• Hai Duong Nguyen SS, MO (25)
• 3 Temps SS (75)
• Total 2 FTE

25
Our Newest Addition to the Group

• Selda Esen, who is graduating from Turkish
  University of Cukurova this week and will start
  with us on October 1
• CMS thesis based on jet studies and HCAL
  test-beam data analysis
• Will work 100 on CMS and replace Ryan Hooper
• Supported by the NSF grant

26
Conclusions

• Our 2004 bid for LHC, made possible by the NSF
  support, has paid up
• Our group is now well-plugged and recognized
  within the CMS community and has contributed to
  several aspects of hardware, firmware, and
  software
• Our adiabatic transition from DØ to CMS has been
  successful we still maintain important
  responsibilities in DØ (Luminosity Monitoring,
  Level 3/DAQ, B-physics, Top, Single Top, New
  Physics Searches), while ramping up our CMS
  effort
• Our CMS involvement doubled every year since we
  have joined CMS while such a growth cant be
  sustained much longer, its up to you to ensure
  that it will continue
• LHC is imminent three years from now we very
  well may be reporting a major discovery made
  possible by the Brown group efforts!