ILC VTX workshop at Ringberg, 29 th May 2006Sonja Hillert (Oxford)p. 0 The Vertex Detector in the LDC Concept Sonja Hillert (Oxford) on behalf of the LCFI

ILC VTX workshop at Ringberg, 29th May 2006 Sonja Hillert (Oxford) p. 1

The Vertex Detector in the LDC Concept

Sonja Hillert (Oxford)

on behalf of the LCFI collaboration

A Vertex Detector for the ILC

Workshop at Ringberg Castle ~ 28 – 31 May 2006

Vertex detector requirements and differences in the 3 concepts

Vertex charge reconstruction for performance evaluation

Comparison of different detector designs

Effects not yet included in simulation

The LCFI Vertex Package – Plans and Status

Summary


At the ILC, the most stringent requirements on the vertex detector are common

to all the concepts.

In part they are given by the machine:

• readout at least at effectively 50 s intervals

• excellent point resolution / small pixel size

• adequate radiation hardness

Other requirements result from physics needs:

• hermeticity / good angular coverage (to cos = 0.96)

• proximity to the IP / small radius of innermost detector layer

• minimal material amount, as far as possible evenly distributed over layer

• low power consumption to permit gas cooling

Requirements for the ILC vertex detector

337 ns

2820x

0.2 s

0.95 ms Beam bunch structure at ILC

Multiple collisions


Detector concepts differ in overall size and therefore in B-field and

inner radius of the vertex detector.

Other differences are not necessarily related to global detector design:

• SiD: short barrel and forward disks

• LDC: long barrel and tracking from additional silicon detectors in forward region

Differences between vertex detectors of the 3 concepts

250 mm

15 mm60 mm

100 mm

LDC vertex detector

SiD vertex detector

design concept

(Norman Graf)


Vertex charge reconstruction

in the 40% of cases where b quark hadronises to charged B-hadron quark sign can

be determined by vertex charge – good performance indicator for vertex detector

probability of mis-reconstructing vertex charge is small for both charged and neutral cases

neutral vertices require ‘charge dipole’ procedure from SLD still to be developed for ILC

need to find all stable tracks from

B decay chain:

• define seed axis

• cut on L/D (normalised distance

between IP and projection of track POCA

onto seed axis)

• tracks that form vertices other than IP

are assigned regardless of their L/D


Energy and polar angle dependence

impurity is given by probability 0

of reconstructing neutral B hadron

as charged

degradation at large cos :

multiple scattering of oblique

tracks in detector material,

loss of tracks at detector edge

effects more strongly seen at

lower jet energy

(jet cone is broader and

more low momentum tracks

multiple scattering more severe)

results obtained with SGV fast MC


Comparing detector with different inner layer radii250 mm

15 mm

60

mm

100 mm

8 mm25 mm

e+e- BEAM

importance of small beam pipe radius demonstrated

by recent study of probability 0 for reconstructing

neutral hadron as charged (Snowmass 2005)

can be translated into ‘luminosity factor’: factor

by which luminosity would have to be varied to

reach same statistical significance as for standard

detector (plot: processes requiring vtx chg for 2 jets)


Comparison with SiD & GLD vertex detectors: Results

material at the end of SiD short barrel staves compromises performance at large cos

GLD performance affected by larger beam pipe radius compared to LDC detector

50

(August 05) (August 05)


Sensitivity to other parameters

since vertex charge performance is

sensitive to multiple scattering need to

keep layer thickness small (target 0.1 % X0)

also strong dependence on momentum cut

(track selection) – this depends critically

on tracking performance:

• track finding capability

• background rates

• linking across subdetector boundaries

should push all these parameters to their limits, as all these effects will eventually

add up in the real detector


A future optimised full MC simulation and reconstruction will need to reliably model

• pair background

• cluster shapes (different for signal and pair background: also transverse component)

• time / position correlation (different for different

sensor technologies)

• dependence on any non-uniformities in

material distribution, effect of conical beam-pipe

• pattern recognition, leading to detailed understanding

of tracking precision, track finding inefficiency,

incorrect hit association, fake tracks

• non-Gaussian tails on track errors

• track linking between sub-detectors

• reconstruction of Ks and

ILC community will need to look at all these items in detail to arrive at reliable conclusions

regarding detector performance.

Effects not yet included

T Woolliscroft

(Liverpool U)


The LCFI Vertex Package

The LCFI collaboration is working hard to provide a C++ based Vertex Package comprising

the vertex finder ZVTOP, NN-based flavour tagging and vertex charge reconstruction.

To make our code available to the community and to profit from future improvements of

the ILC software frameworks, this package will be provided in the form of MARLIN processors

with input / output in LCIO format.

For part of the output, we have proposed the introduction of a new Vertex class into LCIO:

currently under discussion in the LCIO developers group, for more information see the

ILC forum at http://forum.linearcollider.org/index.php

Performance will be checked using SGV. Parts for which FORTRAN code exists will be

cross-checked against the FORTRAN version. Improvements of the code are guided by

comparison to MC information.

Coding / validation is making good progress – release envisaged this summer


interface SGV to

internal format

interface LCIO to

internal formatinput to LCFI Vertex Package

output of LCFI Vertex Packageinterface internal

format to SGV

interface internal

format to LCIO

ZVRES ZVKIN

ZVTOP:

vertex information

track attachment

assuming c jet

track attachment

assuming b jet

track attachment

for flavour tag

find vertex-

dependent

flavour tag

inputs

find vertex charge

find vertex-

independent

flavour tag

inputs

neural net flavour tag


The ZVTOP vertex finder

two branches: ZVRES and ZVKIN (also known as ghost track algorithm)

The ZVRES algorithm:

tracks approximated as Gaussian ´probability tubes´

from these, a ´vertex function´ is obtained:

3D-space searched for maxima in the vertex function that satisfy

resolubility criterion; track can be contained in > 1 candidate vertex

iterative cuts on 2 of vertex fit and maximisation of vertex

function results in unambiguous assignment of tracks to vertices

has been shown to work in various environments differing in

energy range, detectors used and physics extracted

very general algorithm that can cope with arbitrary multi-prong decay topologies

D. Jackson,

NIM A 388 (1997) 247


The ZVKIN (ghost track) algorithm

more specialised algorithm to extend coverage to b-jets in which one or both

secondary and tertiary vertex are 1-pronged and / or in which the B is very

short-lived;

algorithm relies on the fact that IP, B- and D-decay vertex lie on an approximately

straight line due to the boost of the B hadron

should improve flavour tagging capabilities

ZVRES

GHOST

SLD VXD3 bb-MC


tanh (MPt / 5 GeV)

joint probability

uds

c

b

Flavour tag Vertex package will provide flavour tag procedure developed by R. Hawkings et al

(LC-PHSM-2000-021) and recently used by K. Desch / Th. Kuhl as default

NN-input variables used:

• if secondary vertex found: MPt , momentum

of secondary vertex, and its decay length and

decay length significance

• if only primary vertex found: momentum and

impact parameter significance in R- and z for the

two most-significant tracks in the jet

• in both cases: joint probability in R- and z (estimator of

probability for all tracks to originate from primary vertex)

will be flexible enough to permit user further tuning of the input variables for the neural net,

and of the NN-architecture (number and type of nodes) and training algorithm

b

c


Status of C++ ZVTOP development ZVRES branch: coding completed, validation ongoing

left: comparison of decay length reconstructed by C++ to the FORTRAN value

right: comparison of C++ reconstructed to true track origin (iso = isolated tracks from ZVTOP)

Ben Jeffery (Oxford U) MC track origin

2 vertices 3 vertices

pri sec iso pri sec ter iso

Primary 98.6 0.4 1.1 98.1 1.1 0.0 0.8

B decay 8.8 75.6 15.6 2.3 90.1 3.7 3.9

D decay 2.1 80.5 17.4 0.5 17.8 77.4 4.4

Mark Grimes (Bristol U)

coding of ZVKIN branch ongoing, determination of ghost track direction complete


Towards completion of the Vertex Package

b

c

c (b bkgr)c

c (b bkgr)

b

Z peak ECM = 500 GeV

test of full chain of C++ ZVTOP with FORTRAN flavour tag and vertex charge imminent

pure FORTRAN results (from SGV) show below:

C++ code for calculation of inputs for flavour tag being written

Vertex charge reconstruction for c-jets under development


Summary and outlook

Stringent requirements from the accelerator on the one hand and from ILC physics on the

other hand make the design of an ILC vertex detector very demanding.

Careful modelling of all relevant aspects is necessary to arrive at reliable conclusions

regarding requirements and performance achieved.

Relaxing design targets prematurely could result in performance being severely compromised –

targets for inner layer radius, layer thickness and momentum cutoff should be maintained.

Realistic assessment of vertex detector performance and comparison of designs will be helped

by the LCFI vertex package, providing vertex finding, flavour tag and vertex charge

reconstruction – due to be released this summer.

Output of our code crucially relies on the quality of the input – precise reconstruction of

low momentum tracks particularly challenging, likely to influence design of FTD system.

ILC community still has a long way to go for many aspects of the reconstruction.


Additional Material


Position of vertices wrt detector layers: percentages

CM energy (GeV) 50 100 200 500

inside beam pipe 99.99 % 99.49 % 94.69 % 74.49 %

between layer 1 & layer 2 0.01 % 0.48 % 4.48 % 14.95 %

between layer 2 & layer 3 -- 0.02 % 0.67 % 5.78 %

between layer 3 & layer 4 -- -- 0.11 % 2.49 %

between layer 4 & layer 5 -- -- 0.04 % 1.11 %

outside vertex detector -- -- 0.01 % 1.19 %

CM energy (GeV) 50 100 200 500

inside beam pipe 95.39 % 93.97 % 85.89 % 57.64 %





outside vertex detector 3.00 % 3.82 % 4.18 % 6.08 %

MC-level

secondary

vertex

MC-level

tertiary

vertex


quark b

hadron

B- B0

vtx charge

X-- X- X0 X- X0 X+

Interpretation

b

0.4(1-pm)+0.30+0.30

01-0pm pm1-2pm

0

b-bar

B0-bar B+

X- X0 X+ X0 X+ X++

b-bar

0.4(1-pm)+0.30+0.30

b quark sign selection


Vertex charge reconstruction


Increasing beam pipe thickness for standard detector

difference in layer thickness

between standard detector

and 4-layer detector

only small effect

main difference between the

detectors due to difference

in inner layer radius


Suggestion for a new Vertex class in LCIOa new vertex class could look like this:

+~Vertex()

+getMomentum() : const float*

+getMass() : float

+getCharge() : float

+getPosition() : const FloatVec&

+getCovMatrix() : const FloatVec&

+getChi2() : float

+getProbability() : float

+getDistanceToPreviousVertex() : float

+getErrorDistanceToPreviousVertex() : float

+getParameters() : const LCParameters&

+getTracks() : const TrackVec&

+addTrack() : void

+getPreviousVertex() : Vertex&

Documents

ILC VTX workshop at Ringberg, 29 th May 2006Sonja Hillert (Oxford)p. 0 The Vertex Detector in the LDC Concept Sonja Hillert (Oxford) on behalf of the LCFI