1. High Energy Physics in the LHC Era 2016 6 th International Workshop Vicki Greene for the PHENIX Collaboration Vanderbilt University 7 January 2016 Measurements of directed, elliptic, and triangular flow in Cu+Au collisions at √S NN = 200 GeV using the PHENIX detector at RHIC 1

2. Outline Introduction Detector configuration Directed, elliptic, and triangular flow Charged particles Identified hadrons Scaling behavior Other collision systems Model comparisons Conclusions 2

3. An example of anisotropic flow: Elliptic Flow Elliptic flow: initial spatial anisotropy  pressure gradients  momentum anisotropy 3

4. Anisotropic Flow Harmonics – Event Plane Method 4

5. Anisotropic flow harmonics Reflect properties of initial state and evolution of collision system Probe different length scales Sensitive to Equation of State and viscosity/entropy ratio  s Uncertainties in energy density deposition in initial state are limiting factor in deducing  s Asymmetric collisions probe effect of initial geometry 5

6. v 1 sign conventions used v 1 is defined to be positive at positive  Cu-going) x is positive if spectators flow outwards Measurements use Au spectators, signs are flipped 6

7. 7

8. Au+Au p+p d+Au Cu+Cu U+U Cu+Au He+Au p+Au p+Al Collision Systems at BNL-RHIC Run 12 (2012) 200 GeV 5 weeks 7.6 B events  < 0.35 arXiv:1509.07784 PHENIX data in this analysis 8

9. Event Plane Resolution three sub-event method used to determine the resolution:  1 : SMDS,  2,3 : BBCS+BBSN 9

10. v 1 Magnitude decreases from central to more peripheral events V 2 Magnitude increases from central to more peripheral events Centrality Dependence 10

11. Centrality Dependence v 2 Magnitude increases from central to more peripheral events v 3 Weak centrality dependence – dominated by fluctuations 11

12. v 2 System size dependence: Au+Au, Cu+Au, Cu+Cu Cu+Au v 2 lies between Cu+Cu and Au+Au 12

13. v 2 (  2 scaled)  2 scaling reorders the results by system size 13

14. v 2 (  2 N part 1/3 scaled ) – length scale universal behavior in all centralities and systems: Cu+Cu, Cu+Au, Au+Au 14

15. For the same centrality,  3 is larger in the smaller system due to increased fluctuations 15

16. v 3 system size dependence v 3 Cu+Au > v 3 Au+Au 16

17. v 3 (  3 scaled) Close agreement at low-intermediate p T Within systematic uncertainties at high p T 17

18. V 3 scaled by  3 N part 1/3 Agreement within systematic uncertainties at all p T 18

19. Mass ordering at low p T for v 2 for all centralities Identified particle v 2 19

20. v1v1 v3v3 Mass ordering at low p T for v 1,3 Identified particle v 1 and v 3 20

21. P. Bozek, Phys. Lett. B717 (2012) 287 |  | < 0.35 v 1 comparison to viscous hydrodynamics Indirect comparison shows qualitative agreement, assuming spectators curl outward from the z-vertex |  | < 1.0 21

22. v 2 and v 3 comparison to viscous hydrodynamics v2v2 v3v3 22 For 0-5% centrality, η/s =0.8 better reproduces data For 20-30% centrality, both values of η/s agree with data

23. v2v2 v3v3 Comparison to AMPT 23

24. Conclusions In Cu+Au the magnitude of v 1 decreases from central to peripheral, opposite to v 2 behavior. v 3 is not strongly centrality-dependent System size comparison: v 2,3 in different systems scale with  2,3 N part 1/3. Mass ordering is seen for all harmonics. v 2 and v 3 are consistent with viscous hydrodynamics AMPT with  = 3.0 mb describes v 2 and v 3 for p T < 2 GeV. 24

25. backup 25

26. backup 26

27. backup 27

28. Contributions to systematic uncertainties Event plane resolution correction Event plane using different detectors V n from background tracks Acceptance dependencies PID purity 28

29. PHENIX Run 12 Detector Configuration 29