Event by Event Flow in ATLAS and CMS

Transcription

1 Event by Event Flow in ALAS and CMS Gregor Herten Universität Freiburg, Germany LHCP 2015 St. Petersburg,

2 Some basic heavy-ion physics terminology 2

3 Centrality and Glauber Model Centrality in A+A: impact parameter cannot be directly measured and has to be estimated from measurements of Nch, E, ZDC,.. Centrality is typically expressed as a % fraction of the total geometrical cross section: central is 0% centrality. Glauber Model: connects centrality to the number of binary collisions (Ncoll) and nucleon participants (Npart) 3

4 Modelling Primordial Fluctuations MC-Glauber: MC-KLN: IP-Glasma: DIPSY: Event-by-event fluctuations due to sampling of nucleon positions. Soft particle production proportional to the density of participating nucleons. Initial entropy is proportional to number of participating nucleons and number of binary collisions. Entropy production is determined by initial gluon production, calculated from structure function or participating nucleons. his model builds on the IP-Sat (impact parameter dependent saturation) model to generate finite, deformed, fluctuating initial gluon field configurations in the transverse plane (longitudinal fluctuations are not yet included). MC event generator, based on gluon radiation from colored dipoles (via dipole splitting), that uses BFKL evolution. hese initial fluctuations are then evolved through nonlinear viscous hydrodynamics into the final-state particle flow. 4

5 Anisotropic Flow dn d Li Yan, Stony Brook 1+2 X n v n cos[n( n)], Fourier transform of azimuthal angle distribution Φ. Fourier coefficients vn Event plane angles Ψn Event plane angles Ψn characterize the direction of maximum particle density in the event. v2 elliptic flow - due to initial asymmetry v3 and higher orders - due to initial fluctuations 5

6 Flow methods General Prob. distribution p( n, m,..., n, m,...)= 1 N evts One measures projections of this general prob-distribution: dn evts d n,d m,...,d n,d m,... Event plane method, Scalar product method: dn d 1+2 X n v n cos[n( EP,n)] Multi-particle correlations: X 2-PC 4-PC Lee-Yang Zero (All-Particle Correlation) e.g. 2 PC dn pair d 1+2 X n V n cos(n ) 6

7 X Flow methods Correlations of 2k particles: hh2kii = hcorr n {2k}i = hhe in( k k+1... {2k}) ii = hv n {2k} 2k i Cumulant method: he idea of using 2k-particle cumulants is to suppress the non-flow contribution by eliminating the correlations which act between fewer than 2k particles. 7

8 Observables J. Jia 2014 J. Phys. G: Nucl. Part. Phys

9 Single flow harmonics vn v n 2 1/ a v 2 v 3 v 4 v 5 ALICE data v n {2}, p > 0.2 GeV η/s = 0.2 v n 2 1/ b ALAS 20 30%, EP Narrow: η/s() Wide: η/s = Centrality percentile p (GeV) Heinz, Snelling, Ann.Rev.Nucl.Part.Sci. 63 (2013) Good agreement is found between data and model calculations based on viscous hydrodynamical calculations (IP-Glasma), which include gluon fluctuations and gluon saturation. 9

10 Flow Distribution p(vn) ALAS, JHEP 11 (2013) 183 Probability distribution of EbyE vn for several centrality bins. he shaded bands indicate the uncertainty on the vn-shape. Solid lines: Bessel-Gaussian function based on measurement of <vn> for the fluctuation-only scenario.

11 Flow Distribution p(vn) ) p(v ) p(v centrality: 2-3% ALAS Pb+Pb s NN =2.76 ev = 7 µb L int p >0.5 GeV, η < centrality: 0.2 5% 1 centrality: 3-4% centrality: % centrality: 4-5% centrality: % centrality: 50% centrality: % Deviations for v2 from Bessel- Gaussian function at mid-central and peripheral collisions centrality: % centrality: % centrality: % centrality: % ALAS JHEP 11(2013) 183 ) 2 1 p(v centrality: % centrality: % centrality: % centrality: % ) 1 p(v v v v v 2 Figure 14. he probability density distributions of v for p > 0.5 GeV in several centrality 11

12 Comparison of different vn measurements {8} to v 2 {4} as a function of the average s, N part,forellipticflowcoefficients hod (left) and calculated from the measured p(v 2 ) distribution [17] (right). he error bars denote statistical and systematic errors added in quadrature. he ratio symbols are shifted horizontally with respect to each other for clarity ALAS, Eur. Phys. J. C (2014) 74:3157 Comparisons of vn measurements using different methods. endency: vn{2} > vn{ep} > vn{ebye} > vn{4} art dependence of the v 2 (top left), v 3 (top right) and v 4 (bottom) harmonics measured with different methods, with lue ofgregor the corresponding Herten, LHCP p(v2015 n ).heerror bars denote statistical and systematic uncertainties added in quadrature 12

13 Comparison PbPb and ppb CMS uses multi-particle correlation techniques to measure flow and event-by-event fluctuations. 0. CMS PbPb s NN = 2.76 ev 0.3 < p < 3.0 GeV/c; η < 2.4 CMS ppb s NN = 5.02 ev 0.3 < p < 3.0 GeV/c; η < 2.4 v v 2 {2, η >2} v 2 {4} v 2 {6} v 2 {8} v 2 {LYZ} offline N trk offline N trk Figure 2: he v values as a function of N offline. Open data points are published two- and fourv2 signal also in ppb: v2{2} > v2{4} v2{6} v2{8} v2{lyz} ± 2% (PbPb) ± % (ppb) CMS, PRL 115 (2015)

14 FCal q 2 2 (a) Centrality 0% 2 1/N evt dn evt /dq Flow amplitude correlations p(vn, vm) ALAS s NN Pb+Pb = 2.76 ev = 7 µb FCal q 2 IV. DAA ANALYSIS Measurement of flow vector q2 A. Event-shape selection (shape parameter) in the forward calorimeter for the 1% most-central collisions L int v 5 2 v ALAS Pb+Pb Pb+Pb NN 2.76 ev s NN = 2.76 ev 0.01 L int = 7 µb 0.15 = 7 µb < p < 2 GeV 2< η <5 0.1 Centrality Centrality 0-5% 0-5% % 5% 25-30% 25-30% 40-45% 0 0 ALAS Pb+Pb < p < GeV 2< η < = 2.76 ev FCal q 0.04 s NN = 7 µb L int r triangularity in each event is characterized by the so-called flow vector calculated from Centrality the E % ) deposited in the FCal [14, 39]: 5% 25-30% 40-45% q m = q m e im obs w m j e im j = hq 0 w m i evts,m= 2 or 3 (7) j ALAS Pb+Pb 0.5 < p < 2 GeV 2< η <5 v Centrality % 5% % ALAS, arxiv % (7) in the 1% most central collisions. he vertical lines indicate the boundaries L int of the fifteen q m ranges, Correlation between q2 and v2 in four centrality bins. s NN = 2.76 ev 0.03 j in the FCal. Subtraction = 7 µb of the event-averaged FCal q ine) he distributions of the magnitude of the flow vector, q 2 (left panel) and q 3 (right panel), calculated action of events as indicated. ever, the e ciency varies more strongly with and event multiplicity [31]. For p > 0.8 GeV, it t 0 to 57% for > 2 in peripheral collisions, while it ranges from 72% at 0 to about 42% al collisions. j is the E of the j th tower at azimuthal angle L where int wj is the E of the j in Eq. (7) removes biases due to th tower at azimuthal angle φj in the FCAL. obs detector e ects [40]. he angles m are the observed event uate around the true event planes m due to the finite number of particles in an event. A standard 0.02 ed togregor remove Herten, the LHCP small2015 obs residual nonuniformities in the distribution of m. hese procedures are 2 14

15 Flow amplitude correlations p(vn, vm) ALAS, arxiv < 4 GeV} {3 < p 0.2 (a) ALAS Pb+Pb s NN L int = 2.76 ev = 7 µb Centrality 0-70% 2< η <5 < 4 GeV} {3 < p 0.1 (b) ALAS Pb+Pb s NN L int = 2.76 ev = 7 µb Centrality 0-70% 2< η <5 v 2 Peripheral v 3 Central Central v 2 {0.5 < p < 2 GeV} Peripheral v 3 {0.5 < p < 2 GeV} Correlation of v2 and v3 for two p bins. Values are calculated in fourteen 5% centrality bins in the range 0-70%. 15

16 Flow amplitude correlations p(vn, vm) ALAS, arxiv v Centrality 0-70%, no shape selection (a) 0.5 < p < 2 GeV 2< η <5 ALAS Pb+Pb Centrality interval with q selection: 0-5% 2 5% (b) 20-25% 30-35% 40-45% 50-55% 60-65% s NN = 2.76 ev 0.03 L int = 7 µb Central Peripheral v 2 ALAS Pb+Pb s NN = 2.76 ev L int = 7 µb v 2 Fourteen 5% centrality bins. No shape selection. Fifteen q2 intervals in seven centrality ranges 16

17 Flow amplitude correlations p(vn, vm) v (a) 0.5 < p < 2 GeV 2< η <5 ALAS s NN L int Pb+Pb = 2.76 ev = 7 µb (b) ALAS s NN L int Pb+Pb = 2.76 ev = 7 µb 0.02 Peripheral 0.01 Central Centrality 0-65%, no shape selection Centrality interval with q 2 selection: % 5% 20-25% 30-35% 40-45% 50-55% 60-65% v 2 v 2 hirteen 5% centrality bins. No shape selection. Fifteen q2 intervals in seven centrality ranges 17

18 Event plane correlations p(φn, Φm,...) ALAS, Phys. Rev. C 90, (2014) PHYSICAL REVIEW C 90, (2014) Solid line: scalar product method Dashed line: event plane method FIG.. (Color online) Comparison of six two-plane correlators, cos( ), with = jk( n m ), with results from the AMP model calculated via the SP method (solid lines) and the EP method (dashed lines) from Ref. [37]. he error bars on the lines represent the statistical uncertainties in the calculation. v 5 signal [5,44]: combines the initial-state geometry fluctuations of the Glauber model and final-state interactions through a parton and hadron i5 5 i5 i2 i3 18

19 Factorization breaking in correlations CMS uses multi-particle correlations to study factorization breaking effects which are due to initial state fluctuations. hese measurements provide information about event-plane fluctuations. 19

20 r2 in PbPb CMS, arxiv: p dependent factorization ratio as function of p a -p b in bins of p a for different centrality ranges in PbPb. Comparison with MC-Glauber : dashed line MC-KLN: solid green line Factorization breaking at high p a and high p a -p b 20

21 r3 in PbPb CMS, arxiv: p dependent factorization ratio as function of p a -p b in bins of p a for different centrality ranges in PbPb. Comparison with MC-Glauber : dashed line MC-KLN: solid green line Factorization breaking at high p a and high p a -p b 21

22 rn vs. centrality CMS, arxiv: he p-dependenct factorization ratios as function of event multiplicity. Breakdown of factorization observed in r2 for centrality < 5%. For r3 factorization holds at the 2-3% level. No MC calculation can describe data over full centrality range. 22

23 CMS: Pseudo-Rapidity Factorization Breakdown Study of longitudinal fluctuations 23

24 CMS: r2(η a, η b ) in PbPb Factorization breaking effects below 5% 24

25 CMS: r3(η a, η b ) in PbPb r3 is more sensitive to longitudinal fluctuations than r2 25

26 % % CMS: r4(η a, η b ) in PbPb η a % η a Figure : Similar distributions as shown in Fig. 9, but for the factorization ratio r CMS PbPb s NN = 2.76 ev r 4(η a,η b ) < η b < < η b < 4.0 Exponential fits 0-0.2% centrality r 4(η a,η b ) η a < p p b 0-20% a > 0 GeV/c < 3.0 GeV/c r 4(η a,η b ) η a % η a Figure 11: Similar distributions as shown in Fig. 9, but for the factorization ratio r 4 in fewer centrality ranges. data. By plugging Eq. (11) into Eq. (9), the r n can be expressed as r n (h a, h b ) e 2Fh n h a, (12) which is independent of h b, consistent with the results in Figs According to Eqs. (11) and (12), the r n (h a, h b ) also corresponds to a measurement of event plane fluctuations between a a r4 also is more sensitive to longitudinal fluctuations than r2 26

27 Conclusions High precision measurements on azimuthal anisotropy in PbPb and ppb by ALAS and CMS. Large variety of (new) methods, e.g. vn-vm, EP correlations, show promising potential for further insight in HI collisions. ALAS and CMS results on event-plane fluctuations Collective flow also established in ppb collisions Good description of data by viscous hydrodynamic models with fluctuating initial-state conditions. 27

28 Backup Slides 28

29 Single flow harmonics vn: ultra-central collisions Ultra-central Pb+Pb collisions are sensitive to EbyE fluctuations J. Jia, J. Phys. G: Nucl. Part. Phys. 41 (2014) CMS, JHEP02(2014)088 Luzum, Ollitrault, Nuclear Physics A (2013) Comparison with hydrodynamic calculation at various initial conditions show discrepancies, mainly in the relative strength of v2 and v3. 29

30 Flow amplitude correlations p(vn, vm) ALAS, arxiv < 4 GeV} {3 < p v (a) Centrality with q 2 0-5% 5% 20-25% 30-35% 40-45% 50-55% 60-65% selection: 2< η <5 < 4 GeV} {3 < p v (b) 3 Centrality with q 0-5% 5% 20-25% 30-35% 40-45% 50-55% selection: 2< η <5 0.1 ALAS Pb+Pb s NN = 2.76 ev 0.05 ALAS Pb+Pb s NN = 2.76 ev L int = 7 µb Centrality 0-70%, no shape sele v 2 {0.5 < p < 2 GeV} L int = 7 µb Centrality 0-70%, no shape sele v 3 {0.5 < p < 2 GeV} Correlation of v2 and v3 for two p intervals for various centrality bins. Data points are calculated in each centrality bin for several intervals in the shape parameter qm. hey increase monotonically with qm. 30

31 Flow amplitude correlations p(vn, vm) 31