T. Bowcock1 Can CP save life? LHCb and its applications…

T. Bowcock 1

Can CP save life?

“LHCb and its applications…”

T. Bowcock 2

Overview

• CP physics and the LHCb experiment

• LHCb vertex detector technology• Optimization• MAP project• Applications• Summary

T. Bowcock 3

CPz

x

yP: )( rr C:(particle antiparticle)

CP violated only rarely … compensated by T violation (CPT)

e.g KLe±

±

T. Bowcock 4

Why study CP violation?

• Origin of CP (T) violation? – Matter/AntiMatter

Asymmetry• Sakharov

• CP violation– Only observed in

K system. – CP effects in B

system in SM

CosmologyParticle Physics

T. Bowcock 5

CP studies • 1st Generation B -sector studies

<2005. Two possibilities:– either: there will be already a sign of

new physics– or: measurements will look “consistent”

with the Standard model.

• A dedicated experiment at LHC is needed. A “CP-interferometer”. About 10x more sensitive.

T. Bowcock 6

CP studies

Bd+-

Bds

BsDsK

B

s mixing

V ub/V

cb

Bs

First generation of experiments will study Vub/Vcb and BdKs.

To study all angles/decay modes need a specialized detector.

Detailed Bs studies out of reach of Generation 1 experiments.

Bd JS Bs JBs DSK (PID)Bd DK (PID,Trigger)Bd D* (PID)Bd (PID)Bd K (PID)Bd Bs KBs Kll Bs 75 38Bs

LHCb ATLAS/CMS

T. Bowcock 7

B-hadron productionB-hadron production

323222

CP effects 10K events

About 11012 BB produced/year (108 Gen. 1)

T. Bowcock 8

LHCb Detector

T. Bowcock 9

Vertex Detector

• Precision tracking that: – identification of B vertices– measurement of lifetime (40fs)

Bs Ds K

T. Bowcock 10

Geometry

10cm

Detectors separated 6cm during injection

series of 17 1/2 disks

smalloverlap

Positioning and movement to 5m

Precision detectors: Very high radiation doses

£3M project

T. Bowcock 11

Vertex Detector

T. Bowcock 12

Radiation Environment

cm

1013

1014

1 M

eV e

quiv

alen

t neu

tron

s/cm

2

1 2 3 4 5 6

station 6

Dose after 1yr • Including effects of walls, vessel

• High doses at tips– (1/r2)

T. Bowcock 13

150 micron thin n on n detectors with double metal layer (multiple scattering, radiation)

r phi z geometry (reflects forward symmetry of the events, optimizes resolution / channel numbers, ease L1 trigger)

Baseline Solution

r-measuring detector-measuring detector

5 ° “stereo” tilt resolves ambiguities

T. Bowcock 14

Detectors fabricated on 100mm wafer

-measuringdetectors

r-measuringdetectors

inner radius 10mm

readout tracksspaced 50m

UK prototypes

T. Bowcock 15

Testbeam

Testbeam

T. Bowcock 16

Resolution

T. Bowcock 17

Noise

• S/N with VA2 (slow) electronics

• Noise – readout line

length– strip length

T. Bowcock 18

Module Design

0 50 100 150 200 250 300

W/mm2

0

-4

-8Tem

pera

ture

at T

ip (

°C)

Thermal Model: hold cooling at -10°C

Thermal Runaway

LHCbthick detectors

Single Sided r and module

LHCbUK

T. Bowcock 19

Irradiation

• 4 Detectors Irradiated 11/98– Heidelberg 20MeV

p

• Lab test• Awaiting test

beam– stored at -25C

T. Bowcock 20

First Test with a Fast Readout Chip (SCTA 128)

Main goal: Noise Study Over Spill

19.10.1998:first output from one SCTA chip

Now: 4 SCTA/hybrid uniform pedestals Noise: 600 ENC 26.4.1999:

Next testbeam

T. Bowcock 21

Summary 1998

• Prototype in test beam• LHCb like events produced with a

target• Alignment issues

– design modifications

• Noise Studies• Resolution

T. Bowcock 22

Physics Performance

Bs decays

B->

Level 1 Trigger

z positions+radii

#planes

tilt angle

material effects

pitch vs. r

break positions

stereo angle

Use LINUX farm at Liverpool + VX fast simulation

Optimization

T. Bowcock 23

Some ideas for improvement need to be investigated:

smaller pitches -> p on n technology improve resolution, (price )

vary pitch as function of radius optimize occupancy / number of channels / resolution

move closer to beam axis

reduce outer radius to fit on one wafer reduces number of individual detector modules, ease alignment

Improvements?

T. Bowcock 24

Simulation

• Radiation damage• Signal

– ISE(3D)/Kurata– full 3D diode

• Thermal analysis

T. Bowcock 25

New p-strip design

• reduced inner radius: 8 mm

• reduced outer radius: 40 mm requires a few extra stations

• 182o modules with 45.5o sectors (r-detector)

• inner strip pitch 25 (26) micron, strip pitch increases with r outer strip pitch 99 (131) micron

r-detector2048 strips

phi-detector2048 strips

Try different thickness 300, 200 and 140 micron

40.0mm

21.2mm 17.9mm

8mm

T. Bowcock 26

1999

• Tests of irradiated n-strips in Lab• Tests of irradiated detectors (with

SCT128A) in test beam

T. Bowcock 27

1999 Irradiation

T. Bowcock 28

Irradiated Detector

Irradiated (200V) unirradiated

(V. Prelim) Irradiation at 3*1014

T. Bowcock 29

Module

Precision Mount to platformdesigned

Hybrid Design

L’pool +IDE

T. Bowcock 30

Mechanics

• Can adjust to <10microns in space.

T. Bowcock 31

Thermal Model

• Heat Conduction Away from module– separate cooling and mechanical path– no cooling pipes?

T. Bowcock 32

Biases?

T. Bowcock 33

VD Milestones

• Technology Choice 9/00• Technical Design Report 01/01?

T. Bowcock 34

Current Work

• SCTa electronics• Cosmic Ray & Laser setups• Evaluation of Irradiated Detectors

– comparison with testbeam

T. Bowcock 35

Computing

• In operation 1PByte stored/year– LCB has set up MONARC

• Short term– data sets of about 108 events

• optimize detector design• background suppression

• “Monte-Carlo Array Processor

T. Bowcock 36

MAP operation• Run for several weeks 106-107 events

– data will reside on disk!

• Model for use has been developed– “commodity disk” servers -1TByte each`– 8 TBytes at MAP– 5 mirror sites with 0.5TBytes each

• Minimum required for UK LHCb program– RICH and VD design and optimization

T. Bowcock 37

MAP+analysis

Institutes

Liverpool

T. Bowcock 38

Summary• Responsibility for LHCb Si• Extensive programme of R&D• Computing for LHCb-UK

– solve immediate optimization problems

T. Bowcock 39

Need for Monte Carlo

• At LHCb about 1 interaction /25ns !– 4*1014/year– if you want to do physics you need to

know the backgrounds• generating just the signals doesn’t work

– need to generate large MC samples•O(107) to O(108) events.

T. Bowcock 40

Summary

• LHCb recommended for funding by PPESP !

• Intense programme of work to TDR until 2001

T. Bowcock 41

MAP cont’d

T. Bowcock 42

MAP status

• Method– Linux– Batch System– Broadcast

• point to point transfers for “broken sockets”

– complex code– handling failed/dropped packets.

T. Bowcock 43

MAP hardware

• 300 processors– 400MHz PII– 128 Mbytes memory– 3 Gbytes disk– 100BaseT ethernet +hubs– commercial units BUT

• custom boxes for packing and cooling

T. Bowcock 44

MAP use

• Prepare a job • Submit to Batch Queue

– at the moment suppress o/p

• Histograms/Ntuples transmitted back at end of job

• Random Numbers handled automatically

T. Bowcock 45

MAP

• Performance– About 120s/event GEANT +

reconstruction

• Produce about 200,000 events/day• Results

– move into production

T. Bowcock 46

MAP capabilites

• Can be used in “throwaway” mode– just keep Ntuples

• MAP possesses 1Tbyte internal storage – 3 Gbytes/machine– events stored locally (1million events)– repeatedly analyse QUICKLY

T. Bowcock 47

MAP+COMPASS

T. Bowcock 48

COMPASS

• Meant to give large storage capability for use of all UK institutes– RICH/VELO studies– PHYSICS

• DELL prototype in place– 1TByte + High End Server (600MHz)

• transfer rates• reliability

T. Bowcock 49

COMPASS

T. Bowcock 50

COMPASS

• Purpose– Will show this in place and working

with MAP– Model for LHC analysis

• store events on disks (cheap!)• move JOB to the DATA• NO HSM

T. Bowcock 51

BACKUP SLIDES

BACKUP slides

T. Bowcock 52

Technology Options for LHCb Vertex Detector

pro con

p strip Single sided processing Rapidly falling efficiencyReduced cost (30%)and ease of handling below partial depletion

Minimum pitch 12 m High field region on opposite surface to readout

Thin detectors advantageous for Needs to be thinner for givenmultiple scattering operating voltage.(Lower signal)

Handling(cost) of thin detectors.

n strip High efficiency at partial depletion gives Lithographic processing of backlower operating voltage and lower power (Cost and handling)High field region (after irradiation) at Minimum pitch 40m.readout strips.

Operating partially depleted at tip still allows full depletion (high CCE) elsewhere.

both Material Difficult to handleCharge Correlation Offset voltages on one side of

detector for electronics.Thermal contact - sensitive face?

Proto

type

1998

T. Bowcock 53

RF shieldprimary ->

secondary Vacuum2100 micron / detector station

Best case: 100 micron once

Worst case: 2.4100 micron/detector station for low angle tracks (but high p)

T. Bowcock 54

Noise Study

T. Bowcock 55

Noise Study

T. Bowcock 56

pictures of rf shield

T. Bowcock 57

Expected Resolution

LHCb 97-020 TRAC

6 micron for theta = 0o

LHCb 97-020 TRAC

6 micron for theta = 0o