Longitudinal correlation in heavy-ion collisions

Transcription

1 Longitudinal correlation in heavy-ion collisions Jiangyong Jia Stony Brook University & Brookhaven National Laboratory QCD Chirality Workshop 2016 Feb 22-26, 2016, UCLA 1

2 2 Nature of sources seeding these long-range collective ridges? Frozen PDF fluctuation Frozen PDF fluctuation vn ε n! 1 nsources How many such sources, their sizes & transverse distribution?

3 3 Nature of sources seeding these long-range collective ridges? Particles (entropy) are produced early in collision p+p Pb+Pb B N part N F part F B Underlying Event Underlying Event MPI: nf nb Forward/backward multiplicity/flow correlations provide a handle

4 Importance of sub-nucleonic sources! Multi-parton interactions (MPI) required to described N ch distribution 4! N MPI N ch

5 Three types of longitudinal correlations Fluctuation of participants in two nuclei " size and transverse-shape different! N F part! ε F e inψ F n n! N B part! ε B e inψ B n n 5 Consequences:! ε F B ε 2 2! N F B (a) N part part (b) (c)! Ψ F B Ψ 2 2! d η direction η direction η direction Asymmetry in multiplicity Asymmetry in flow magnitude Torque/twist of flow plane

6 Glauber model estimation 6 FB asymmetry width of A Npart =(N F B -N part part )/(N Pb+Pb F B +N part part ) FB asymmetry width of A F B F B 2 =( 2-2 )/( ) width of A F- B)/( F 3 =( B ) Twist F B width of 2( Φ 2 * -Φ 2 * ) F B width of 3( Φ 3 * -Φ 3 * ) N part N part N part! N part -asymmetry large in peripheral.! 2 nd order: ε-asymmetry and twist largest in central! 3 rd order: ε-asymmetry and twist ~ independent of centrality

7 Two-particle correlation observables 7! Most general 2PC: C(η 1, η 2, Δϕ) C(η 1, η 2, Δφ)= C N (η 1, η 2 ) 1+2! n V nδ (η 1, η 2 )cosnδφ FB Multiplicity fluc. N( η 1 )N η 2 C N (η 1, η 2 )= N η! 1 ( ) ( ) N( η 2 ) Driven by N part asymmetry FB flow fluc. V nδ (η 1, η 2 ) = v n (η 1 )v n (η 2 )cosn[ Φ n (η 1 ) Φ n (η 2 )] Driven by ε n twist & asymmetry! C N ~ (1+ a 1 η 1 )(1+ a 1 η 2 ) =1+ a 12 η 1 η 2 r(η,η ref )= V ( η,η ) nδ ref V! nδ (η,η ref ) a 1 N F part F N part B N part B + N part CMS observables Large gap between η ref and η,-η

8 2 nd -order flow 8! e! Decrease toward mid-central collisions, then increase toward peripheral collisions

9 3 rd -order flow 9! Slight increase toward peripheral collisions " Importance of e.g. subleading flow, subnucleon dof?

10 Compare to 3+1D hydrodynamics 10 Pang, arxiv: !! Good agreement except for 0-5%!! Much stronger effect at RHIC energy (investigate in BES)!

11 r 2 (η a,η b ) in high-multiplicity ppb 11 The rise toward smaller system consistent with importance of subnucleonic dof.

12 0 th -order: FB multiplicity asymmetry 12

13 Observable! 2-D pseudorapidity correlation function C =! N( η 1 )N η 2 ( ) ( ) N( η 2 ) = R η S 1 N η 1 Mixed events R S ( )R S η 2 ( ) events ( η) N(η) N(η) Single particle distribution η <Y= Traditional observables N b N f b = N f N b N f N b σ N f σ Nb CF disentangles statistical fluctuation from dynamical fluctuation

14 Forward/backward multiplicity correlation 14! dn/dη shape reflects asymmetry in num. of forward/backward sources! Seen directly in p+pb collisions. Pb matter p arxiv: FB asymmetry is expected in Pb+Pb or pp collisions on event-by-event bases!

15 Property of the multiplicity correlation 15 Short-range correlation Δη=η 1 -η 2 ~0 = Data-driven method to separate SRC from LRC Long range correlation large Δη =! SRC reflects correlations in the same source! LRC reflects FB-asymmetry of number of sources, e.g.

16 Quantifying the SRC and LRC 16 SRC Quantify by average amplitude: η <Y=2.4 LRC Shape approximate by: Implication: deviation from average is linear in η! R S ( η) N(η) N(η) evts 1+ a 1 η! C = R η S ( 1)R S ( η 2 ) 1+ a 2 1 η 1 η 2 R S Event 1 Event 2 1 η

17 Dependence on N ch and collision systems 17 both N ch and system size only by N ch SRC LRC Compare: SRC/LRC control by N ch or transverse geom. size?

18 Dependence on N ch and collision systems 18 both N ch and system size only by N ch SRC LRC SRC depends on num. of sources LRC depends on FB asymmetry of sources independent source picture :! Fit with # LRC: num. of sources, n, controlled by N ch, think in terms of partons! # SRC: pp vs PbPb at same N ch " n is similar but pairs/source is larger?

19 Features of dn/dη distribution 19 Double hump" FB asymmetry? p+a collision suggest it is the latter "p+p FB fluctuation is even larger!

20 20 LRC SRC Asymmetry entirely due to SRC "larger on proton side!

21 21 LRC SRC Asymmetry entirely due to SRC "larger on proton side!! SRC at given η scale as nsourcem2 1 = 1 2 dn dη (nsourcem) nsource!! " Strength of SRC vs η+ reflects fluctuation of dn/dη shape " Amplitude increase on p-side " Width unchanged

22 Compare pp to ppb, PbPb at same Nch 22

23 Compare pp to ppb, PbPb at same Nch 23 +! High-multi. pp has same η correlation as symmetrized p+pb given Nch " Ebye asym. of dn/dη in high-multi. pp is as large as that in ppb!! " Pb+Pb collision more symmetric.

24 Summary! Longitudinal correlation provide unique information on initial conditions in pp, pa and AA collisions 24! Longitudinal flow correlations! consistent with rapidity dependent mixing of initial condition controlled by projectile and target participants! Longitudinal multiplicity correlations! LRC depends only on total multiplicity of the event! SRC depends strongly on collision system and charge combination! Both follows power-law of N ch with an index close to 0.5 " information on the number of sources for particle production?! FB asymmetry in pp is as strong as ppb in same multiplicity! High multiplicity pp collision is highly asymmetric system! Similar longitudinal initial condition in high multiplicity ppb and pp?