CMS-UK RAL 20/4/06 R M Brown - CCLRC 1 CMS Submission to PPGC UK Deliverables Status of CMS UK Progress & Commissioning Plans Computing UK Physics Programme

CMS-UK RAL 20/4/06 R M Brown - CCLRC 1

CMS Submission to PPGC

• UK Deliverables

• Status of CMS

• UK Progress & Commissioning Plans

• Computing

• UK Physics Programme

• Summary


The CMS Collaboration

April 2006

846

1061

462

528

2051

64

69

49182

UK Global Co-ordination RolesDeputy Spokesperson

T. VirdeeCMS Electronic Co-ord.

J. NashTK Project Manager

P.SharpTK Deputy PM

G. HallECAL IB Chair

R.BrownECAL Endcap Co-ord.

D.CockerillECAL Commiss.Co-ord.

K.BellTK Offline S/W Co-ord.

I.Tomalin‘e-’ Group Co-ord.

C. SeezHiggs Group Co-ord.

A. NikitenkoComputing Com. Chair

D.Newbold

M&O Authors 1081UK (4.0%) 43

Scientific Authors 2051UK 106Countries 38Institutes 182


Compact Muon Solenoid

ECAL

Tracker

4T solenoid

Muon chamber

s

HCAL

Iron yoke

Total weight: 12,500 tOverall diameter: 15 mOverall length: 21.6 mMagnetic field: 4 T

Person!


The UK in CMS

Bristol Imperial College

Brunel

Tracker Readout System Global Calorimeter TriggerElectromagnetic Calorimeter Endcaps

Rutherford Appleton Laboratory

Source Card

Leaf Card

Computing - Software - Simulation - Physics


LHC Intersection Point 5CMS is pre-assembled on the surface and lowered in 17 sections weighing between 300 and 2000 tons

Surface Assembly Hall

Service Cavern(Readout, Trigger, P/S)

Experiment Cavern


Surface Assembly

Magnet complete, coil cold

Yoke Endcap (YE+1)installation ~ done

Muon Barrel Drift-Tube/RPCinstallation well advanced

10 sectors in wheels 0,+1,+2.2 sectors in wheels -1,-2.

HCAL being inserted into solenoid

300

0

T OK

Time Days

Gas out

Gas in


Magnet closed: : May 06Magnet test/cosmic ray challenge: : June-Aug 06ECAL 1st half-barrel (EB+) installation : Aug 06 SX5Service Cavern ready for crates: : April-May 06HF lowering: : July 06Experiment Cavern ready for crates : Jul 06First connection to Service Cavern : Aug-Sep 06Central Yoke Section+Coil (YB0) lowering : Nov 06ECAL 2nd half-barrel (EB-) installation : Oct 06 SX5+(Jan-May 07 UXC) Tracker installation : Jan-Feb 07 ECAL/Tracker cabling complete : Apr-May 07Heavy lowering complete : Feb-Mar 07Initial CMS ready for beam : 30 Jun 07ECAL Endcaps, Pixels installation : Dec 07 – March 08

Schedule is tight, with no master contingencyOptions to recover ~2 months after Tracker installation under consideration

CMS Schedule: Key Dates


CMS ECAL Endcap (EE)UK has led EE project from inception

- Monte Carlo simulations - Performance optimisation- Vacuum Photo-Triode (VPT) R&D- Prototype beam tests at CERN

Deliverables:- Overall mechanical design- On-det. electronic design(+CERN)- VPTs (pay 50%, test 100%)- Mechanical structures(+Russia)- Assembly (with Russia, CERN)

Vacuum phototriodes:- Single stage pmt- Gain 8 -10 at B = 4 T- Radiation resistant (UV glass window)- Q.E. ~ 20% at 420 nm

Presampler

Environmental Shield (Russia)

Supercrystals

PositionalSpacers

Back plate

Ring flange

Moderators(Russia)

HadCal

ElectronicsMechatronic

s

3.5 m

1"


EE Status

Electronic noise performance verified with a

detector module (25 channels)

mounted on a Dee(~3800 e/channel)

Alignment test of 2 Dees and Ring Flange mounted

on HCAL

Mechanics:All major components deliveredCMS ‘Gold Award’ to TM Engineers (UK)

13800/15500 delivered (89%)

VPT Production & Testing


EE Future Work2006 On detector electronics production

Assembly jigs and tooling completed Launch EE crystal production:SIC (China) Assembly of Dee1 from July

2007 Launch EE crystal production:BTCP (RF) Assembly of Dees Installation of Dees 1 & 2

2008 Installation of Dees 3 & 4

Delivery of EE crystals is on the critical pathAssembly schedule tight: Parallel work to minimise assembly latency

Long term UK commitments - Maintaining and operating the EE- Maintaining the associated databases- Understanding and calibrating the detector - Estimated EID effort: 0.25 FTE in 08/09 & 09/10

EE Commissioning will only START in earnest AFTER end of Ring Fence


TOBTOB

TIDTIDTIBTIB

TECTEC

PDPD

CMS Tracker• Two main sub-systems: - Silicon Strip Tracker and Pixel Detector• Silicon Strip Tracker comprises: - Outer Barrel (TOB) - Inner Barrel and Disks (TIB-TID) - End-Cap (TEC)

~ 107 Channels, 220m2 Silicon

TIB Layer 2, after burn-in

Cosmics in TEC+


Tracker Electronic System

Input Rate ~ 1.2 TBytes/sec1.2 TBytes/secOutput rate ~ 50 GBytes/sec50 GBytes/sec

• Analogue readout• No on-detector zero suppression • Optical analogue data transfer

UK Deliverables:• Development of readout & control system• Delivery of APV & APVMUX ASICs ( ~100,000 die, inc spares)• Procurement of Front End Driver (~500 modules)

Status:• APV Project Complete On time, within budget• FED Procurement underway(eXception received a Gold Award)


Tracker Future Work2006: Final assembly and integration in TIF • Four large systems to be assembled and operated• Test complete tracker in up to 25% units• Delivery within budget requires staff from institutes

2007: Commissioning at Point 5• Lower, insert, cable, power,…• Commission off-detector electronics & first running

2008: Operations• Calibrate and tune performance

Optimise S/N, cluster finding in firmware• APV, APV emulator, FED, DAQ, Online Software

are key elements for Tracker to deliver good data• All require UK expertise• System of ~500 complex boards 1 FTE EID needed long term for maintenance


Global Calorimeter TriggerVital Role in CMS Trigger: Electron/ triggers Jet Triggers Transverse Energy Triggers RCT Readout System Luminosity Monitoring

Original GCT design was technically challenging - required very high speed (~3GB/s)

data transfer between many boards

major revision Jan 2006

- Large, powerful FPGAs concentratealgorithm processing in few locations

- Optical links bring large number ofinput signals to processing boards

New design to be delivered by July 07- UK led with CERN engineering effort


The New GCT Design

Software: Simulation (efficiencies)Testing and CommissioningOnline MonitoringOnline Control

There are four different types of card:

Source Card (54 needed):Translates electrical signals from RCT to optical

Leaf Card (8 +1() needed):Main algorithm card of the GCTElectron and jet finding and sorting, all ET calculationsProven design from a satellite project (LANL)

Concentrator Card (1+1() needed):Communicates with rest of Trigger and upstream DAQ

Wheel Card (2 needed):Supports Leaf Cards, performs final jet-finder calcs


GCT Future Work

2006: Development and integration phaseAug Integration of Source Card + Leaf Card

Oct Integration Source + Leaf + Concentrator

2007: Commissioning phaseJan Commissioning Electron TriggersJul Commissioning Jet Triggers

2008: Operations phaseRespond to actual CMS conditionsCalibrate, evaluate & tune trigger performance

Trigger must operate with 100% availability 1 FTE EID needed long term for maintenance

New design can be scaled and commissioned in stages

Electron data from ½ RCT



Leaf Cards

Wheel Cards

Concentrator Card


Computing status Major milestones in last two years:

– CMS Computing Model fully defined, documented in Computing TDR– Operational prototypes of all ‘infrastructural’ software in place

Major UK contributions to data and workflow management systems– Integration with Grid systems (WLCG) now well advanced– Decision taken to adopt new application framework in May 2006

Upcoming milestones before 2008– CSA2006 (computing, software, analysis): 25% scale test of full

computing system, October 2006– Full computing system readiness by April 2007– Factor-of-two final resource ramp up before April 2008

CMSUK Deliverables 2006 - 8:– Delivery and support of software components for CMS– Ramp-up and operation of Tier-1 and Tier-2 computing centres– Maintenance and support of computing system for UK analysis


UK Computing Project Project organisation

– Organised as single coherent project, covering GridPP, e-science, centre operations, service challenge effort

– Project currently limited by available manpower Further physicist / e-science effort will become available in 2007 Vital to retain our expert GridPP / e-science personnel

– Project plan is fully tied to CMS global milestones Work areas:

– ‘Infrastructural’ software deliverables (key for CMS) [2 FTE] PhEDEx: manages all offline data; BOSS: user interface to the Grid

– Computing system ramp-up and support [1.75 FTE] Software installation and maintainance; DB management Data challenge activities; integration with new RAL CASTOR system

– ‘Physics ops’: data handling, reconstruction, MC prod [2 FTE]– Documentation, user support for analysis [1 FTE]


Computing Plans & Challenges Major UK milestones 2006 – 8:

– Jul ‘06: All UK centres take part in new framework MC production;new user support resources (web-based) in place

– Sep – Oct ‘06: Full UK participation in CSA2006 at 25% scale– Dec ‘06: RAL CASTOR system fully integrated with CMS system– Apr ‘07: UK systems ready to take data; final versions of PhEDEx, BOSS;

funded plan for ramp-up to full system scale– Apr ‘08: Ready for full scale data taking

Key challenges:– Retention of expert staff: systems are highly complex, will require constant

‘tuning’ in 2007 – 8– Scale of Tier-1 hardware resources

CMS extremely short of Tier-1 storage and CPU; UK is the smallest contributor, despite high profile on CMS

CMS Computing Model may be compromised in light of the shortfall Critical factor for effective physics output in UK and Europe


UK Physics Strategy Build on existing expertise

- Tracker- ECAL- Test beam analyses- Simulation studies

Plan smooth transition from detector/reconstruction studies to physics

Emphasise preparations for leading involvement in early analyses

But balance with developing tools to achieve longer term physics goals

Develop close working collaboration with theorists to ‘gain an edge’

Work cohesively with the wider world of CMS

UK provides Convenors for 3 of the 8 CMS ‘PRS’ Groups:(Physics, Reconstruction, Software)

ECAL- e/ Tracker- b/ Higgs

Organisation:2 ‘Physics Co-ordinators’

(One at CERN, One in UK)Regular programme of meetings


CMS Physics TDRCMS physics activities are focused on Physics TDR

Volume 1: Published December 2005- Validation of detector simulation- Development of reconstruction algorithms- Techniques for Calibration, alignment, synchronisation…

Volume 2: To be published May 2006- Complete analyses (backgrounds, misalignment,

miscalibration)- Demonstration of physics capability with 10 fb-1, 30 fb-1

(low luminosity L=2×1033)

Volume 3: Due January 2007- Commissioning- Start-up procedures- First physics- Timing of High Level Trigger processing

Detector Performance and Software

Physics Technical Design Report, Volume I

Physics PerformancePhysics Technical Design Report, Volume II


UK Contributions to Physics TDR Defining content and work, final editing, detailed contributions:

Physics TDR Volume I: Detector Performance and Software

ECAL Intercalibration using We

106 Fully simulated

events

ECAL Test Beam analysis

Beam

0 50 100 150 200 GeV

/E%

1.4

1.2

1.0

0.8

0.6

0.4

0.2

0.0

Super Module


UK Contributions to Physics TDRPhysics TDR Volume II: Physics Performance

qqH, H ℓ + jet WW, ZZ

Fusion

jet2

jet1

b jet1b jet2

jet2

jet1

b jet1b jet2

MSSM: bbH/A,

H/A 2 jet


First Physics Exciting results possible with ~fb-1 of data

Concentrate initial analysis efforts on early discoveries

Exploit connections between detector understanding,reconstruction & selection algorithms and physics

Focus on e, , energy-flow, and then eg: Study W e, Z ee, leading to:

W' e, Z' ee, (and eventually, H ZZ* eeℓℓ

eg: Study W + jets, Z + jets, Z , leading to:b-jet tagging, H searches


Higgs Physics As luminosity increases, concentrate progressively on Higgs searches

Strong UK interest demonstrated by range of studies already underway:

)h MSSMor HSM(ττ,qq 0

ττ/AH,/AHbb 0000

bbhandZ,ZhA 000 model)Sundrum-(Randalγγhandbbh,hhRadion 0000

class) thisindecays other (plusνχorχqqχ,bχt~,bbh,ht~t~ 01

0111

00

scenario) CPX (inν Wand bbH W,HH,bHt 11

tt/AH 00 decay)invisible(i.e.χχh,hqq 0

101

00 ττH,Htt 00


Direct ,Triple Gauge Couplings UK leads studies of gluon structure function of

through prompt production Useful results with 1-3 fb-1

Analysis will provide valuable input for development of H searches

UK also leads TGC study for Physics TDR

Responsible for W+, Z+ channels & backgrounds

TGC sensitivity requires high L If no ‘smoking gun’: Anomalous TGC could provide vital signature of BSM

Wq

q

ℓ

sensitive to

non-SM Physics W+jet background to W+


Summary of CMS RequestRequest (£k) 2006/2007 2007/08 2008/09 2009/10

Maintenance & Operating (M&O) Costs

165.0 170.0 171.0 171.0

Common Fund (M&O) Costs 120.0 157.0 180.0 196.0

Travel - Short Term (UK + Abroad) 0.0 0.0 170.2 170.2

Travel - Long Term 0.0 0.0 254.5 254.5

TOTAL 285 327 776 792

RAL EFFORT (SY)

ED Effort 0 0.00 0.10 0.10

ID Effort 0 2.00 2.15 2.15

Other (specify) 0 0.00 0.00 0.00

TOTAL 0 2.00 2.25 2.25


SummaryCMS assembly is proceeding without significant technical problems

Solenoid is cold: Magnet test and ‘Cosmic Challenge’ start in June

Schedule is tight for ‘Ready-for-Beam’ in July 07

UK Tracker Readout deliverables on schedule

ECAL Endcap completion limited by crystal delivery: Install Spring 08

Revised GCT design is advancing rapidly

UK-CMS has provided key tools for CMS Computing (PhEDEX, BOSS)

There are concerns about the CMS Tier-1 capacity

Continuation of posts currently supported by GridPP/e-Science is vital

UK is making strong contributions to CMS Physics TDR

A coherent UK physics programme is underway


Back-ups


Tier-1 Issue Role of Tier-1 centres:

– Long-term storage (‘curation’) of raw data– Bulk reprocessing of data for reco, skimming, calibration, analysis– Serving data to Tier-2 centres (all data comes through Tier-1)– Stores full copy of AOD

Resource situation for CMS:– European Tier-1 contributors do not have large CMS fractions– WLCG ‘pledge system’ results in major shortfall in Tier-1 resources

This is emergent behaviour: funding ‘formulae’ may locally make sense in each country, but do not address the global problem

– Individual Tier-1 centres are well-run, but far below critical mass Only US, Italian centres approach ‘nominal’ Tier-1 size

Strategy:– Funding system must be examined - LHCC asks for CMS input– Must look at new ways to make maximally efficient use of resources

Potential danger if redundancy & robustness traded against resources– The CMS computing model may be compromised (in Europe)

Urgent need for adequate and realistic Tier-1 resources in UK


WLCG Computing Resources (2008)


Deployment of effort

CMS Bid 200608/09 09/10

Table 1 Uni M&O: Grants-supported 8.0 7.0

Table 2 Uni M&O: Academics 2.7 2.5

Table 3 PPD M&O 6.0 6.0

Table 5 Physics Analysis Effort: Academics 9.9 9.5Physics Analysis Effort: Rolling Grant 7.6 7.5Physics Analysis Effort: Other 0.0 0.0Physics Analysis Effort: TOTAL 17.5 17.0


Exclusive diff. Higgs prod. pp p H p

Hgap gap

b

b

-jet

p p

420m

High luminosity:Full 3D detectorsand pixel readout

A. De Roeck, V. A. Khoze, A. D. Martin, R. Orava and M. G. Ryskin, Eur. Phys. J. C 25 (2002) 391

V. A. Khoze, A. D. Martin and M. G. Ryskin, Eur. Phys. J. C 34 (2004) 327

Alberta, Antwerp, Cambridge, Helsinki, FNAL, Leuven, Torino,UCL..

R&D with participants from Atlas and CMS/Totem

Documents

CMS-UK RAL 20/4/06 R M Brown - CCLRC 1 CMS Submission to PPGC UK Deliverables Status of CMS UK Progress & Commissioning Plans Computing UK Physics Programme