Yan Lu for STAR collaboration Department of Modern Physics

System Size Dependence of Strange Hadron Elliptic Flow from 200 GeV Au +

Au and Cu + Cu collisions

Yan Lu

for STAR collaboration

Department of Modern Physics University of Science and Technology of China

Quark Matter 2008The 20th International Conference on Ultra-relativistic Nucleus-Nucleus CollisionsFebruary 4–10, Jaipur, India

Yan Lu Quark Matter 2008, Feb. 4 – 10, Jaipur, India 2

Outline

Motivation

Data analysis

Results and Discussion

Summary and Outlook

Yan Lu Quark Matter 2008, Feb. 4 – 10, Jaipur, India 3

Thermalization?

Data: S. Voloshin, nucl-ex/ 0701038Hydro: Kolb, Sollfrank, Heinz, PRC 62 (2000) 054909

21 dNS

vdy

2 2S x y

Low density limit:

2

In case of thermalization:

v2 22

22

yx

xy

part is for the participant zone.

Possible thermalization only for the most central collisions.

‘We argue that the centrality dependence of elliptic flow should be a good indicator of the degree of equilibration reached in the action. ’ S. Voloshin, A. Poskanzer, PLB 474, 27, 2000.

Yan Lu Quark Matter 2008, Feb. 4 – 10, Jaipur, India 4

Partonic Collectivity at RHIC

‘They are made via coalescence of seemingly thermalized quarks in central Au+Au collisions, the observations imply hot and dense matter with partonic collectivity has been formed at RHIC’ STAR: Phys. Rev. Lett., 99, 112301(07), nucl-ex/0703033

minimum bias

Yan Lu Quark Matter 2008, Feb. 4 – 10, Jaipur, India 5

Why System Size and Centrality Dependence?

versus

- Systematic study for identified hadrons will gain information on interplay of collectivity, NQ scaling and thermalization as a function of collision system size and centrality.- Determine parameters for partonic EoS.

Collectivity Local Thermalization nq scaling

System SizeCentrality

Yan Lu Quark Matter 2008, Feb. 4 – 10, Jaipur, India 6

Dataset and Cutsmultiplicity collision centrality

• System: Au+Au collision • Energy : sNN = 200 GeV• Event sample: ~22 M events• Centrality: 0 – 80 %, 0 – 10%,

10 – 40%, 40 – 80%

• System: Cu+Cu collision • Energy : sNN = 200 GeV• Event sample: ~21 M events• Centrality: 0 – 60 %, 0 – 20%, 20

– 60%

• Event cuts: |Vertex z| < 30cm• FTPC track cuts: 6 =< nFitHits < 12, nHitsFit/nHitsPoss > 0.52 pT > 0.1, 2.5 < || < 4.0, Dca < 2.0 cm

Yan Lu Quark Matter 2008, Feb. 4 – 10, Jaipur, India 7

PID and v2 Methods

0S

() p (p )

)()( )()(

KK

v2 vs. minv method (background subtraction technique): - used for K0S, Λ , Ξ, , and analysis and for Au+Au and Cu+Cu datasets flow method - Event Plane method used for Au+Au and Cu+Cu datasets - non-flow systematic: Lee-Yang Zero method for Au+Au and Event Plane method with FTPC reaction plane for Cu+Cu

Lee-Yang Zero method: R. S. Bhalerao, N. Borghini, J.-Y.Ollitrault, Nucl. Phy. A 727 (2003) 373-426Event Plane method: A. M. Poskanzer, S. A. Voloshin, Phys. Rev. C58, 1671 (1998)

Yan Lu Quark Matter 2008, Feb. 4 – 10, Jaipur, India 8

Centrality Dependence ofIdentified v2

All centralities: mass ordering at low pT mT-m scaling at low mT-m. All centralities: baryon v2 > meson v2 at intermediate pT or mT-m.

STAR Preliminary Transverse kinetic energy: mT – m = (pT)2 + m2 - mTransverse momentum

Hydro: P. Huovinen, private communications, 2007

200 GeV Au+Au

Yan Lu Quark Matter 2008, Feb. 4 – 10, Jaipur, India 9

Centrality Dependence of Number-of-Quark Scaling

nq scaling is observed at all centrality bins.

200 GeV Au+Au

STAR Preliminarynq: number of quark within a hadron

Yan Lu Quark Matter 2008, Feb. 4 – 10, Jaipur, India 10

Eccentricity Scaling(?)

STAR Preliminary

Not observe the part scaling claimed by PHENIX.

Divided v2 by <v2>ch instead of p

art, it appears that the scaling works better.

Scale by part to remove initial geometry effect.

Larger v2/part indicates stronger flow in more central collisions.

PHENIX: Phys. Rev. Lett 98, 162301 (2007)Phys. Lett. B 503, 58 (2001)

200 GeV Au+Au200 GeV Au+Au

(c)

Yan Lu Quark Matter 2008, Feb. 4 – 10, Jaipur, India 11

pT-integrated v2/part

STAR Preliminary

Charge particle data: STAR, Phys. Rev. C 72, 014904, 2005.

v2/part versus Npart data: increasing trend indicates stronger flow in more central collisions. hydro: little sensitivity to the collision centrality as expected in equilibrium scenario.

v2/part for a given centrality hydro: a clear hadron mass dependence. data: not clear due to large errors.

Above Npart ~ 170, integrated v2 consistent with hydro prediction indicates local thermalization.

200 GeV Au+Au

p

Yan Lu Quark Matter 2008, Feb. 4 – 10, Jaipur, India 12

System Size DependenceSTAR Preliminary

A given colliding system: clear scaling with nq. System size dependence: - No part scaling claimed by PHENIX - v2 seems to fall at lower pT in Cu+Cu than in Au+Au

PHENIX: Phys. Rev. Lett 98, 162301 (2007)See S. S. Shi’s poster

Yan Lu Quark Matter 2008, Feb. 4 – 10, Jaipur, India 13

Summary We report azimuthal anisotropy parameter v2 of strange and multi-strange hadrons from Au+Au and Cu+Cu collisions at 200 GeV. Detailed centrality dependence is presented up to pT ~ 6 GeV/c.

At low pT region, collectivity is observed at all centrality bins under study. it follows a mT-m scaling.

Not observe the part scaling. larger v2/part indicates stronger flow in more central collisions and larger system.

v2/part increase with Npart, reaching hydro prediction in central Au+Au collisions. indicate thermalization in central Au+Au collisions.

Future:

1) More systematic measurements for , : partonic EoS parameters

2) Heavy quark collectivity: early thermalization of light quarks

Yan Lu Quark Matter 2008, Feb. 4 – 10, Jaipur, India 15

Backup slides!

Yan Lu Quark Matter 2008, Feb. 4 – 10, Jaipur, India 16

Elliptic flow

xz

y

px

pyy

x

y 2 x 2y 2 x 2

v2 cos2 , tan 1(pypx)

Reaction plane: z-x plane

impact parameter

beam

Yan Lu Quark Matter 2008, Feb. 4 – 10, Jaipur, India 17

v2 vs minv method• analysis method: v2 vs. minv • motivated by

Borghini et al. [nucl-th/0407041]

• advantages over standard method:– only one fit per pT bin- smaller systematic uncertainties

• method used for K0S, Λ , Ξ, , and analysis• standard method and v2 vs. minv method give consistent results

SIG + BG

BG

BG(SIG+BG)

SIG (SIG+BG)

v2TOT(minv)

)()()(

)(2cos()(

22

2

invinvBG

invSIG

invTOT

mBGSIG

BGmvmBGSIG

SIGv

mv

Yan Lu Quark Matter 2008, Feb. 4 – 10, Jaipur, India 18

Lee-Yang Zero method: - measure flow using genuine correlations among all particles. - flow determined from first minimum r0

of generating function.

Systematic Errors Evaluation Standard method (Event Plane method): - flow vector - event plane angle

iiixQ )2cos(2

i

iiyQ )2sin(2 2/tan

2

212

x

y

QQ

Sensitive to non-flow. FTPC Reaction Plane: - Similar to Event Plane method - Correlate particles in TPC region with FTPC reaction plane.

TPCFTPC FTPC

A. M. Poskanzer, S. A. Voloshin, Phys. Rev. C58, 1671 (1998)

R. S. Bhalerao, N. Borghini, J.-Y.Ollitrault, Nucl. Phy. A 727 (2003) 373-426

Suppress non-flow.

all particles

arbitrary variable

Yan Lu Quark Matter 2008, Feb. 4 – 10, Jaipur, India 19

Non-flow systematics

200 GeV Au+Au 10-40%. h± (yellow bands): decreasing trend. PID: no clear pT and PID dependence. v2{LYZ} < v2{EP}: ~10% effect.

200 GeV Cu+Cu 0-60%. v2{EP, FTPC} < v2{EP, TPC}. larger effect at higher pT. v2 from two methods consistent for within large errors.

Yan Lu Quark Matter 2008, Feb. 4 – 10, Jaipur, India 20

System Size and Centrality DependenceSTAR Preliminary PHENIX: Phys. Rev. Lett 98, 162301 (2007)

A given centrality of a colliding system: clear scaling with nq. A given system: stronger flow in more central collisions. Not observe the part scaling claimed by PHENIX, which is independent on colliding system .