Structure of proton

1
What do we do with pp data?
Partially based on WSU group analysis plans

• New possibilities with ALICE
• Higher energy -gt Higher particle density
• Access to clean high pt physics, in
  particularwith EMCal addition (triggering and
  jet reconstruction)
• High statistics (including high multiplicity
  trigger) pp data sample

• ALICE during the first years before EMCal data to
  come
• correlation analysis of pp data
• correlation analysis of PbPb data
• participation in preparation of EMCal data
  analysis tools
• The first two items are essential in order to
  understand ALICE analysis infrastructure, be
• noticeable to the Collaboration, be ready for
  analysis and paper writing of EMCal results

2
Proton-proton collisions unprecedented physics
reach
Enormous reach in multiplicity and
transverse momentum.

• Physical question to pursue
• Jets structure, in particular difference in gluon
  and quark fragmentations
• Hard parton spatial distribution inside proton
• Physics of hadronization, role of coalescence
• Collectivity on pp collisions
• Nature of KNO scaling (multiple Pomeron
  exchange?)
• Reference measurements for future AA analyses

ALICE PPRv2, J.Phys.G32 (2006)1295

• Measurements/analyses/(papers ?)
• 2- and 3-particle high pt correlations
  (A,B,C,D,F)
• 2-particle azimuthal, rapidity, and pt
  correlations (D,E,F)
• Identified particle spectra (A,B,C,D, F)

Emphasis is on the correlation measurements,where
the we have the leading expertise, to
betterunderstand the nature of phenomena. We
start with non-identified 2-particlecorrelations,
then proceed with identified spectra and
3-particle correlations
3
How well we can measure correlations? Pt-Nch
A.Asryan, G.Feofilov, A.Ivanov, A.Grebenyuk,
R.Kolevatov, P.Naumenko, V.Vechernin SAINT-PETERSB
URG STATE UNIVERSITY G.Feofilov , PWG2 , 03.06.
2008 (SPb-CERN, EVO),
Long-range pt-n correlations, pp 900GeV. Pythia
ESD We assume 2-3 x 105 to be minimum
statistics required for Pt-N correlations
estimation. This corresponds to -1..0, 0..1
rapidity windows. In a case of smaller windows of
0.2 units of rapidity the minimal required
statistics will be 1 - 2 x 106 .
Long-range Pt-Nch correlations. Forward rapidity
window 1..0 , Backward rapidity window 0..-1.
Data from CAF PDC06 Run 600 601 Link
http//aliceinfo.cern.ch/export/sites/default/Offl
ine/Analysis/CAF/v4-04-Rev-07/ESD600_601_900GeV_v2
.txt
4
How well we can measure correlations? Nch-Nch
Long-range n-n correlation, pp 900GeV. Pythia
ESD We assume 2-3 x 105 to be minimum
statistics required for Pt-N correlations
estimation. This minimum corresponds to -1..0,
0..1 rapidity windows. In a case of decreasing
window with to 0.2 units of rapidity minimal
statistics will be increased to 1 - 2 x 106 .
N-N correlation saturation study will require
statistics of about 107.
Long-range Nch-Nch correlations. Forward rapidity
window 1..0 , Backward rapidity window 0..-1.
Data from CAF PDC06 Run 600 601 Link
http//aliceinfo.cern.ch/export/sites/default/Offl
ine/Analysis/CAF/v4-04-Rev-07/ESD600_601_900GeV_v2
.txt
5
Elementary NN-collision. Inclusive correlation
functions.
Inclusive
Distribution of correlated pairs
Distribution of associatedparticles (2) per
triggerparticle (1), Balance function
Correlations are due to local charge(s)
conservation, resonances, fluctuations in number
of produced strings, e.g. number of qq-collisions
in const. quark approach
Probability to find a correlated pair
Semi- inclusive (topological)
might use probability density correlation
functions
Production via Nc clusters e.g. independent NN
collisions
At midrapidity, the probability to find a
particle is about 60 larger if one particle has
been already detected.
6
Inclusive vs semi-inclusive
Which do we need to compare with AA data? It is
likely continuous change from semi-inclusive to
inclusive as we move from peripheral to
central. We need to measure both, including
charge dependence, and ID, whenever possible
7
Level of collectivity in high multiplicity pp
events
Observations
Proposed measurements/analyses
two step mean pt corr. dependenceon
multiplicity Multistring events?
Two particle multiplicityand pt correlations. pt
spectra.
Measure azimuthal correlations in pplook for
possible collective effects also, as a reference
for AA.
pure non-flow orelliptic flow in pp? Is it
collective?Contribute to AA?
radial flow? Momentum conservation?
Non- and identified particle spectra asfunction
of multiplicity and blast wave fits.
8
Multistring events
Multireggeon exchanges
Multipomeron exchanges
Multiple constituent quark interaction
The possibility to have one, two, or three
string interactions (e.g. collisions of pairs
of constituent quarks)
9
Nature of multistring collisions
Observations
Proposed measurements/analyses
Mean pt dependence on multiplicity looks like
two step dependence
Two-particle eta, pt, and azimuthal
correlationsas function of multiplicity
Emphasis on understanding the space time
picture of multi-particle production in pp
including initial condition. Check the 3
constituent quark structure of nucleon. We
anticipate the dilution of eta and pt
correlationsin 2 collision events compared with
singlecollision events and at the same time
increase inazimuthal correlations (analogous to
elliptic flow) The understanding of the
structure of pp multi-particle processes will
improve drastically our ability to describe
initial condition in AA as well as hadronization
stage.
Two particle pt correlationsas function of
multiplicity looks like two step dependence
Note clear separation of differentevent classes
at high energies. Expect modification of the
shapeof the correlation function.
The structure of the correlation function best
interpreted as a superposition of events with
different number of strings.
10
Mean pt, two-particle pt and azimuthal
correlations
Azimuthal correlations will we be able to
observe rescatterings of particles from two qq
collisions? ? Strong dependence on
multiplicity. Anisotropic source by HBT?
11
Elliptic flow in pp? How much is it relevant for
AA?
Using PYTHIA (plots from Subhasis student VECC)
Density gradients select specificconfigurations
of fluctuations might lead to multiparticle
azimuthal correlations event anisotropy.
12
Generalized parton distributions spatial
distribution of fast partons
Proposed measurements/analyses
Idea/question
Study frequency of double hard collision events
vs single hard collision events
What is the spatial distribution of fast and
slow partons inside proton?
With additional assumptions, one be also able to
look into internal structure of protons as an
object made of 3 constituent quarks
The results of this study will be important not
only for understanding of structure of the
proton, but would clarify the nature of the
saturation phenomena in AA collisions.
13
Parton spatial distribution How much of black vs
gray?
We can study this by looking at double hard
collisions!
14
Even more fun with identified particles
Current observations
Proposed measurements/analyses
Identified particle spectra and 2-particle
correlationsas function of multiplicity
Breakdown of mt-scaling due to quark vs gluon
jet production
Charged particle spectra in quark vs. gluon jets
15
Summary
Study of plain multiparticle production can
bring exciting physics.
16
Higher energies ? cleaner separation of multiple
collisions
17
What defines the probabilities of multiple hard
collisions?
The values of the background larger than defined
by inclusive single particle spectrum meansthat
the probability of two hard collisions is larger
than the square of probabilities ofsingle hard
collision.
If hard collisions happen only in the overlap
region of area A, then
For two overlapping discsand average over bdb,
this ratio is rather large, 4 - 64/(3?2) ?1.84
, but still lower than the (STAR) experimental
value,of about 2.5
18
Scales of hadronic physics
The first scale 0.6-1.0 fm ?-1QCD typical
hadron radius (radius of confinement).
- Radius of small size instantons sets const.
quark radius.
Additive quark model (hadron spectroscopy,
magnetic moments, etc.)
Primordial kT in lepton pair and large pt
production 1-2 GeV ?

• Kovner - Wiedemann (PRD 66034031, 2002)
• Total cross section grows for two reasons
• Increase in parton density (darkness)
  described by hard Pomeron.
• Increase in radius (soft Pomeron)
  asymptotically saturates Froissart bound (?).

19
Toward measuring constituent quark size...
Consider a case when more energetic partons are
distributedin a region occupying only a fraction
q of the total areaover which softer partons
are distributed. Then
??