Summary Detector Working Group Neutrino Factory International Design Study Meeting 17 January 2008 Paul Soler

Summary Detector Working Group Summary Detector Working Group Summary Detector Working Group Summary Detector Working Group

Neutrino Factory International Design Study

Meeting 17 January 2008

Paul Soler

Neutrino Factory International Design Study Meeting RAL 17 January 2008

2

ContentsContents

MIND summary and tasks – Synergy with TASD: develop common

software (TASD performance covered by Walter for low E Neutrino Factory)

Near Detector summary and tasks Note: no discussion of silver detectors


3

MIND MIND

iron (4 cm) scintillators/RPCs (1cm)

beam

100 m

14 m

14 m

B=1 T1cm transverse resolution

M~100 KTon

Easy to detect muons in iron by rangeEasy to discriminate against hadron showers

Based in known technology: ~MINOSCan be very massiveCost is not prohibitive: 300-400 M$

Easy to magnetise iron

cannot detect electrons or tausthe energy threshold is high


4

Simulation and resolutionSimulation and resolution

Including QE

Essential to measure the oscillation pattern

Crucial to solve degeneracies

Fully contained muons by rangeScaping muons by curvatureHadron shower: E

Ehad

Detector effects not simulated Perfect pattern recognition Reconstruction based on parameterisation Dipole field instead of toroidal field


5

Kinematic cutsKinematic cuts


6

Charge identificationCharge identificationSimple exercise. Assumptions:

No border effects

Non-gaussian scatters can be identified via local χ2 criterion with a Kalman Filter

Assume gaussian MS

Gluckstern formula + MS term

BFe =1.25 T

1.7 T 2T

Event simulationRealistic flux

Non-gaussian MS

Border effects

LSQ fit

L>150

cm

L>150

cm

L>75 cmL>75 cm


7

Wrong charge assignmentsWrong charge assignments


8

BackgroundsBackgrounds


9

Signal efficiencySignal efficiency

Old analysis II: P>5 GeV, Qt

> 0.7 GeV

Old analysis I: P>7.5 GeV, Qt

> 1 GeV

CC signal

Efficiency plateau between 5 and 8 GeV depending on Lμcut

L> 75 cmL>150 cmL>200 cm

baseline: Lμ > 150 cmEnsures charge mis-ID

below 10-3


10

Aims of full Aims of full simulation/reconstructionsimulation/reconstructionDemonstrate that for E < 10 GeV

Backgrounds are below 10-3

The efficiency can be increased with respect to fast analysis

Compute:

Signal and backgrounds efficiency as a function of energy

Energy resolution as a function of energy

Identify crucial parameters to be optimised to maximise

the sensitivity to the osc. parameters

Optimise segmentation and B field based on the above parameters

and taking into account feasibility and cost


11

Hadron showerHadron shower

The hadron shower energy and angle are smeared according to

MINOS proposal + MINOS CalDet + Monolith testbeam

Hadrons are stopped when they decay or undergo a nuclear interaction

We then record their energy and momenta: p1, p2, ..., E1,E2, ...

Their length is also recorded: L1, L2, ...

Fast analysis


12

MuonMuon

Muon is followed until it stops, decays or escapes the detector

The position of all hits is recorded

And also its 3-momentum

Muon hits are smeared with 1cm transverse resolution

A track fit gives its charge

For the kinematical analysis the muon momentum is smeared

according to Gluckstern formula + MS term

Fast analysis


13

In real lifeIn real lifeThe muon is not isolated: pattern recognition

2 independent views XZ and YZ that should be matched

The event sense can be computed from timing (?)


14

Muon reconstructionMuon reconstruction

1.Reconstruct the vertex from event topology

2.Cellular automaton or Hough transform for planes with small activity

3.Match X and Y views in planes with small activity

4.Find approximate muon parameters based on these planes and vertex

5.Incremental Kalman Filter from the end of the track towards vertex

•Multiple scattering, energy loss and B field map

Reconvertex

Cellular automatonKalman filter


15

StepsSteps


16

Event generationEvent generationOnly DIS interactions as coming from LEPTO has

been generated so far

Including QE and RES should have a big impact at

low neutrino energies:

No hadron shower:

Easy pattern recognition

Better neutrino energy resolution

Help in improving the threshold energy and reduce

backgrounds

Generators: Nuance, Neut, Neugen, Genie


17

Synergies with TASDSynergies with TASDScintillator bars, PD and

electronics are the same. This is

the most difficult part

B field production is different and

more difficult

A common framework for simulation and reconstruction

(M. Ellis)

15 m

15

m

100 m


18

Detector optimisation: Longitudinal Detector optimisation: Longitudinal segmentationsegmentation


19

Detector optimisation: Transverse Detector optimisation: Transverse segmentationsegmentation

Assuming perfect pattern recognition 1 cm transverse resolution

is enough for charge and Qt measurements

Pattern recognition:

better segmentation should improve it

which resolution saturates the patter recognition performance ?

Lines: 1, 1.5 and 2 GeV/c muon momentum

BFe=1.25 TeslaFe thickness = 4 cm

BFe=1.25 TeslaFe thickness = 2.5 cm

BFe=2 TeslaFe thickness = 2.5 cm


20

Detector optimisation:Detector optimisation: magnetic fieldmagnetic fieldEven if we are able to isolate a 1 GeV/c muon, the ratio curvature/MS

is not sufficient. ~5% charge mis-ID

The magnetic field strength is the crucial parameter

Going from 1.25 to 1.7 Tesla average is feasible (J. Nelson, Golden07)

> 1 o.o.m improvement at 1 GeV/c. 10-3 level

1 GeV/c

2 GeV/c

1.5 GeV/c

MINOS MIND


21

Conclusions MIND Conclusions MIND

Fast simulation/reconstruction was very

useful until now

But it’s time to move forward with a full

simulation/reconstruction

What are the main backgrounds at low energies ?

What is the background level ?

Where is the efficiency plateau ?

What are the parameters to be optimised ?

Prototyping program should go in parallel


22

Sim/Rec/Analysis task listSim/Rec/Analysis task listEvent simulation (NUANCE)--> bHEP1

converter between NUANCE and bHEP format

Event transport (GEANT4) --> bHEP2 Geometry and bHEP interface

Digitisation --> bHEP3hits:

2D points, pulse height, timelink to true particle

Dummy digitisation with MIND fast simulation Reconstruction --> root fileBuild the framework:

Define bHEP formatRead dst (bHEP)Event likelihoodCellular automaton (import from T2K)Kalman filter (RecPack)

Identified manpower for these tasks

In Valencia/Brunel/Glasgow

+ EuroNu manpower

Tasks to be done in parallel


23

Beam Diagnostics and Near Detector Beam Diagnostics and Near Detector aimsaims

Beam diagnostics (needed for flux measurement)– Number of muon decays

– Measurement of divergence

– Measurement of Muon polarization Near detector measurements needed for neutrino oscillation systematics:

– Flux control for the long baseline search.

– Measurement of charm background

– Cross-section measurements: DIS, QES, RES scattering Other near detector neutrino physics (electroweak and QCD):

– sin2W - sin2W ~ 0.0001

– Unpolarised Parton Distribution Functions, nuclear effects

– Polarised Parton Distribution Functions – polarised target

– Lambda () polarisation S from xF3 - S~0.003 _

– Charm production: |Vcd| and |Vcs|, CP violation from D0/ D0 mixing

– Beyond SM searches

– …


24

Beam DiagnosticsBeam Diagnostics Beam Current Transformer (BCT) to be included at entrance of

straight section: large diameter, with accuracy ~10-3.

Beam Cherenkov for divergence measurement? Could affect quality of beam.

storage ring

shielding

the leptonic detector

the charm and DIS detector

Polarimeter

Cherenkov

BCT


25

Beam DiagnosticsBeam Diagnostics Muon polarization:

Build prototype of polarimeter

Fourier transform of muon energy spectrum

amplitude=> polarization

frequency => energy

decay => energy spread.


26

Flux Measurement at Near Detector Flux Measurement at Near Detector Best possibility: Inverse Muon Decay: scattering off electrons in the

near detector. Known cross-sections μ+νe+ν μe

μ+e+ e


27

Near Detector used to extract PNear Detector used to extract Pee Use matrix method with Near Detector data (even if spectrum not identical in

near and far detector!) to extract oscillation probability:

Where: M1=matrix relating event rate and flux of e at ND

M2=matrix relating event rate and flux of at FD

M=matrix relating measured ND e rate and FD rate

MnOsc=matrix relating expected e flux from ND to FD

Method works well

but need to extract

syst errors of method:

P e M2

1MM1MnOsc 1

Probability of oscillation determined by matrix method under “simplistic” conditions. Need to give more realism to detector and matter effects.


28

Charm measurement Charm measurement Motivation: measure charm cross-section to

validate size of charm background in wrong-sign muon signature

Charm cross-section and branching fractions poorly known

Semiconductor vertex detector only viable option in high intensity environment (emulsion too slow!)


29

Cross section measurementsCross section measurements

Measure of cross sections in DIS, QE and RES. Coherent Different nuclear targets: H2, D2

Nuclear effects, nuclear shadowing, reinteractions

At NUFACT, with modest

size targets can obtain very

large statistics, but is <1%

error achievable?

What is expected cross-

section errors from

MiniBoone, SciBoone,

T2K, Minerva, before

NUFACT?


30

Other physics: Parton Distribution FunctionsOther physics: Parton Distribution Functions

Unpolarised and Polarised Parton Distribution Functions

S from xF3 - S~0.003 Sum rules: e.g. Gross-Llewelyn

Smith polarization: spin transfer from

quarks to — NOMAD best data— Neutrino factory 100 times

more data


31

Near Detector DesignNear Detector Design

Muon chambers

EM calorimeter

HadronicCalorimeter

Overall design of near detector(s):– Near Detector could be a number of specialised detectors to perform

different functions (ie. lepton and flux measurement, charm measurement, PDFs, etc.) or larger General Purpose Detector


32

Near Detector ConclusionsNear Detector Conclusions Near Detector considerations: optimisation design

– Vertex detector: Choice of Pixels; eg. Hybrid pixels, Monolithic Active Pixels (MAPS), DEPFET; or silicon strips

– Tracker: scintillating fibres, gaseous trackers (TPC, Drift chambers, …)– Other sub-detectors: PID, muon ID, calorimeter, …

Tasks:– Simulation of near detector and optimisation of layout: could benefit

from common software framework for Far Detector– Flux determination with inverse muon decays, etc.– Analysis of charm using near detector– Determination of systematic error from near/far extrapolation– Expectation of cross-section measurements– Test beam activities to validate technology (eg. vertex detectors)– Construction of beam diagnostic prototypes– Other physics studies: PDFs, etc. (engage with theory community for

interesting measurements)

Documents

Summary Detector Working Group Neutrino Factory International Design Study Meeting 17 January 2008 Paul Soler