Longitudinal decorrelation of anisotropic flows in heavy-ion collisions at the CERN Large Hadron Collider

Fluctuations in the initial transverse energy-density distribution lead to anisotropic flows as observed in central high-energy heavy-ion collisions. Studies of longitudinal fluctuations of the anisotropic flows can shed further light on the initial conditions and dynamical evolution of the hot quark-gluon matter in these collisions. Correlations between anisotropic flows with varying pseudorapidity gaps in $\mathrm{Pb}+\mathrm{Pb}$ collisions at the CERN Large Hadron Collider are investigated using both an event-by-event $\left(3+1\right)$-dimensional ideal hydrodynamical model with fluctuating initial conditions and a multiphase transport (AMPT) Monte Carlo model for high-energy heavy-ion collisions. Anisotropic flows at different pseudorapidities are found to become significantly decorrelated with increasing pseudorapidity gaps due to longitudinal fluctuations in the initial states of heavy-ion collisions. The longitudinal correlation of the elliptic flow shows a strong centrality dependence while the correlation of the triangular flow is independent of the centrality. Longitudinal fluctuations as a source of the decorrelation are further shown to consist of a twist or gradual rotation in flow angles between the forward and backward direction and additional fluctuations on top of the twist. Within the AMPT model, longitudinal correlations of anisotropic flows are also found to depend on the value of partonic cross sections. The implications of constraining the initial conditions and shear viscosity to entropy density ratio of the partonic matter in high-energy heavy-ion collisions are also discussed.

Figure 1

The longitudinal correlation functions $C_2\left(\Delta \eta \right)$ (a) and $C_3\left(\Delta \eta \right)$ (b) in 20%–25% (squares), 5%–10% (triangles), and 0%–5% (circles) central $\mathrm{Pb}+\mathrm{Pb}$ collisions at $\sqrt{s_{NN}}=2.76$ TeV from ideal hydrodynamic model simulations.

Figure 2

The dependence of the longitudinal correlation functions $C_2\left(\Delta \eta \right)$ (a) and $C_3\left(\Delta \eta \right)$ (b) on the smearing parameter ${\sigma }_{{\eta }_s}$ in 0%–5% central $\mathrm{Pb}+\mathrm{Pb}$ collisions at $\sqrt{s_{NN}}=2.76$ TeV from ideal hydrodynamic model simulations.

Figure 3

The same as Fig. 1 except from the AMPT model simulations with a partonic cross section $\sigma =20$ mb.

Figure 4

Correlation functions $C_2\left(\Delta \eta \right)$ (circles and triangles) and $C_3\left(\Delta \eta \right)$ (squares and rhombuses) calculated with 100% (circles and squares) and 70% of charged hadron multiplicity (triangles and rhombus) in 20%–25% $\mathrm{Pb}+\mathrm{Pb}$ collisions at $\sqrt{s_{NN}}=2.76$ TeV from the AMPT model simulations with a partonic cross section $\sigma =20$ mb.

Figure 5

Correlation functions $C_2\left(\Delta \eta \right)$ (a) and $C_3\left(\Delta \eta \right)$ (b) calculated with all charged hadrons (circles), all charged pions (squares), and pions with the same signs of charge (triangles) in 20%–25% $\mathrm{Pb}+\mathrm{Pb}$ collisions at $\sqrt{s_{NN}}=2.76$ TeV from the AMPT model simulations with a partonic cross section $\sigma =20$ mb.

Figure 6

The correlation function $C_2\left(\Delta \eta \right)$ from ideal hydrodynamic (squares) and the AMPT model simulations of 0%–5% (a), 5%–10% (b), and 20%–25% (c) central $\mathrm{Pb}+\mathrm{Pb}$ collisions at $\sqrt{s_{NN}}=2.76$ TeV with two different values of the partonic cross section, $\sigma =1.5$ mb (circles) and 20 mb (triangles).

Figure 7

The same as Fig. 6 except for the correlation function $C_3\left(\Delta \eta \right)$.

Figure 8

Correlation functions $C_2\left(\Delta \eta \right)$ (a) and $C_3\left(\Delta \eta \right)$ (b) for different hadronic evolution times ($t=10$, 20, and 30 fm). The partonic cross section in the AMPT model is taken to be $\sigma =1.5$ mb.

Figure 9

Event distributions in the twist angles $|\Delta {\Phi }_2^{\mathrm{FB}}|$ (a) and $|\Delta {\Phi }_3^{\mathrm{FB}}|$ (b) from hydrodynamic simulations of $\mathrm{Pb}+\mathrm{Pb}$ collisions at $\sqrt{s_{NN}}=2.76$ TeV with different centralities.

Figure 10

Correlation functions $C_2\left(\Delta \eta \right)$ for different twist angles $\Delta {\Phi }_2^{\mathrm{FB}}$ from hydrodynamic simulations of 0%–5% (a) and 20%–25% (b) central $\mathrm{Pb}+\mathrm{Pb}$ collisions at $\sqrt{s_{NN}}=2.76$ TeV.

Figure 11

Correlation functions $C_3\left(\Delta \eta \right)$ for different twist angles $\Delta {\Phi }_3^{\mathrm{FB}}$ from hydrodynamic simulations of 0%–5% (a) and 20%–25% (b) central $\mathrm{Pb}+\mathrm{Pb}$ collisions at $\sqrt{s_{NN}}=2.76$ TeV.