1

Sparsh Navin – University of Birmingham

for the ALICE collaboration

Trigger Efficiencies in ALICE

andDiffraction in PYTHIA

Diffractive and electromagnetic processes at the LHC, Trento 2010

2

Outline

ALICE and pp trigger efficiencies• ALICE detectors for first physics with protons

• Process Types

• First physics trigger efficiencies

Diffraction and PYTHIA• Kinematic comparison – PYTHIA 6.4 and PHOJET

• Diffraction in PYTHIA – then and now

• Comparison of kinematic plots – PYHTIA 8.310

• Summary

3

A Large Ion Collider Experiment • Primary purpose – heavy ion detector for Pb-Pb collisions at 5.5 TeV

• First physics programme is pp at 0.9, 2.36, 7 (maybe 10) and 14 TeV

- Time and space alignments

- pp bench mark to study genuine Pb-Pb effects

• Special features for pp collisions:

- Low magnetic field => low pT cut-off, tracking resolution from 100 MeV/c

- Excellent PID over broad range of momenta

- Primary vertex resolution (100μm for pp, 10 μm for Pb-Pb)

• pp Physics contributions:

- Physics cross checks with previous results

- Good description of underlying event and high multiplicity collisions

• 1st measurements: , multiplicity and momentum distributions

and

chdN

dch

T

dN

dp

vs T chp N

4

A Large Ion Collider Experiment - 2

52 m underground 16 m high 26 m long

10,000 tonnes 0.5 T magnetic field

V0A z = 3.3m

ITS (Inner

Tracking System)

TPC(Time

Projection Chamber)

• SPD – Inner layers - 3.9cm 7.6cm radii - better than 100μm resolution - provide the SPD trigger signal

• Main gas-filled tracking detector

2.8 5.1

3.7 1.7

• Provide the V0A and V0C trigger signals• Time resolution better than 1ns

V0C z=-0.9m

2 1.4

ZDC at 116m

0.9

8.7 ZN 8.4 ZP

ZEM at 7 m 4.8 5.7

55

Process Types

•True cross sections at LHC energies are not known • Scaling of cross sections with energy is model dependent

PHOJETDefault

fractionsPYTHIA

0.134 SD 0.187

0.063 DD 0.127

Figure from Torbjörn Sjöstrand, MCnet school, 2008.

666

V0A V0C

Outgoing undiffracted p beam

Diffractive system

Single Diffraction (SD)0

Pseudorapidity gap

777

V0A V0C

Diffractive system (2)

Diffractive system (1)

Double Diffraction (DD)0

Pseudorapidity gap

8888

V0A V0C

Non Diffractive (ND)

No pseudorapidity gap

99

First Physics Triggers

Minim

um B

ias Triggers

MB1 SPD or (V0A or V0C)

MB3 SPD and (V0A and V0C)

•Different efficiencies for different

Triggers (MB1, MB3)

Type of process (SD, DD, ND)

Event generator (Pythia and Phojet)

Global Fast Or (GFO) is the trigger from the

Silicon Pixel Detector (SPD)

Actual trigger used – at least 2 pixels in coincidence with beams

10

Trigger Efficiencies and corrections

processprocesstotaltriggered efNN /Efficiency

• Need to know the fraction (f) and the efficiency (e) for each process. • Efficiency is process, trigger and generator dependent

Eg: MB1 = SPD or V0A or V0C

Process SD DD ND

Fraction (f) 0.187 0.127 0.686

Efficiency (e) 0.714 0.864 0.999

NDNDDDDDSDSD efefef

Process SD DD ND

Fraction (f) 0.134 0.063 0.803

Efficiency (e) 0.767 0.938 0.999

MB1 efficiencies:

Pythia: 92.9%

Phojet: 96.4%

Reason for difference: f – uncertainty in

fractionse – uncertainty in

kinematics

Major difference is in diffractive events

~2-4% effect each

11

Kinematics – eta ND 7 TeV

12

Kinematics – eta SD 7 TeV

13

Kinematics – pT ND 7 TeV

14

Kinematics – pT SD 7 TeV

15

Kinematics – multiplicity ND 7 TeV

16

Kinematics – multiplicity SD 7 TeV

17

Diffraction in PYTHIA – “old”

Event Generation:

• Diffractive cross sections given by model by Schuler and Sjöstrand (Phys. Rev. D 49, 2257 (1994))

• Diffractive mass ( ) and momentum transfer (t) generated according to:

2

2

1~ b|t|

2X X

d se

dtdM M

Particle Production:

• above mass of incoming particles => isotropic decay into 2-body state

• More massive system treated as a string with quantum numbers of the original hadron

21 XM GeVc

XM

18

Pomeron couples to valence quarks

Dominates at small

Diffraction in “old” PYTHIA – stretching the string

Version 6.214 (Fortran)

• q and g contributions are set by a user-defined fixed ratio

Version 8.1 (C++)

• Slope p and normalisation N set by user

• mass ( ) dependence

• Gluonic domination at large

Has a gluon (g) and quark (q) contribution

XM

P( )

P( ) pX

q N

g M

XM

XM

Pomeron couples to gluon

Dominates at large

XM

19

Diffraction in PYTHIA 8.130 – “new”Event Generation:• Cross sections - same way as before

• Diffractive mass ( ) and momentum transfer (t) picked by Pomeron flux model

Particle Production: • Pomeron-p collisions

• Pomeron PDF with - dependence from H1 data

- H1 2007 DPDF Fit Jets and H1 2006 Fits A and B

- Pion PDF also available

• Standard PYTHIA machinery for multiple interactions, parton showers, hadronization

Mass separation:• For non-perturbative description (as before)

- longitudinally stretched strings

• For perturbative

XM

2Q

1.2 < M 10 XGeV GeV

M 10 X GeV

Work done with Torbjörn Sjöstrand - MCnet

20

• Energy dependent

• Only SS provides separate t spectrum for DD

Pomeron Flux factor

21

PYTHIA “old” vs “new” - eta 7 TeV

22

PYTHIA “old” vs “new” - pT 7 TeV

23

PYTHIA “old” vs “new” - multiplicity7 TeV

24

Comments

PP

proton

Ex

E g

gP

Ex

E

2X PM x s

Diffractive hard scattering cross section:

2/ ( )/ ( )2 2

Pomeron flux Pomeron PDF

ˆ( ) ( , ) P proton P q g P q g

d df x f x Q

dQ dQ

In the massless limit1p1p

2p

Px

gx

X

Not known from first principles

Multiple interactions => screening of diffractive rates

25

• Does cut-off depend on the diffractive mass?

• Introduce a screening factor to go from ep to pp collisions

• Momentum sum of PDFs

• H1 2006 DPDF LO fit

• Tuning to data

Future Plans - PYTHIA

0Tp

• Include Central Diffraction

1p1p

2p2p

X

LRG

LRG

26

Summary

• The ALICE detectors and trigger for the first pp physics programme were discussed

• Trigger efficiencies are model dependent - PYTHIA and PHOJET

• Earlier versions of PYTHIA had a primitive description of diffraction

• No hard diffraction – caused the difference in pT and multiplicity tails compared to PHOJET

• New version (8.130) has Pomeron description of diffraction

• Hard collisions can be simulated

• Better agreement with PHOJET

27

• ALICE first paper

• Extraction of fractions from data - ALICE

• Diffraction in Phojet

Back up slides

28

ALICE first paper - arXiv: 0911.5430 [hep-ex]

• Trigger – atleast 2 fired chips in coincidence with bunches – 284 events

• Magnetic field off; SPD used for analysis; SDD, SSD and V0 used for cross checks and background ID and removal.

• In the region

• Trigger efficiency obtained from MC PYTHIA 6.141 D6T and PHOJET 1.12:

SD – 48% to 58%, DD – 53% to 76%, ND – 98% to 99%

• Fractions taken from UA5:

SD – , DD - , ND –

• Trigger efficiencies: Inelastic - 87% to 91%, NSD – 94% to 97%

• Results obtained with PYTHIA, difference between PYTHIA and PHOJET used to estimate systematic uncertainty.

0.5 3.10 0.13( ) 0.22( )chdN

stat systd

3.51 0.15( ) 0.25( )chdNstat syst

d

-Inelastic

-NSD

0.153 0.031 0.08 0.05 0.767 0.059

29

Extraction of fractions• Trigger on bunch crossing• Define 8 uncorrelated trigger types using SPD, V0A and V0C• Meausre

NININDNDDDDDSDSDgentrig

gen

NIgen

NIgen

NItrig

gen

NDgen

NDgen

NDtrig

gen

DDgen

DDgen

DDtrig

gen

SDgen

SDgen

SDtrig

gentrig

NItrig

NDtrig

DDtrig

SDtrigtrig

efefefefNN

N

N

N

N

N

N

N

N

N

N

N

N

N

N

N

NNN

NNNNN

calcmeasured

2

2

))((Error

)()(

trig measuredtrig

measuredtrigcalctrig

iN

iNiN

• Program works out combinations of fractions to generate so as to minimise

2 calctrigN

trigN

- Z.Matthews 20/03/09, ALICE first physics meeting

30

Central elements of Phojet- R. Engel workshop on soft diffraction at LHC 26/6/09

Two component Pomeron

Only one pomeron with soft and hard contributions

Topological identification of different terms (Dual parton model)

Soft and hard partons differ in impact parameter distribution

Application of existing parton density parametrisation

Initial and final state radiation (leading logQ^2 parton showers)