Pre-equilibrium evolution effects on heavy-ion collision observables

Apr 8, 2015
19 pages
Published in:
  • Phys.Rev.C 91 (2015) 6, 064906,
  • Phys.Rev.C 92 (2015) 4, 049904 (erratum)
  • Published: Jun 8, 2015
e-Print:

Citations per year

20142017202020232025051015
Abstract: (APS)
To investigate the importance of pre-equilibrium dynamics on relativistic heavy-ion collision observables, we match a highly nonequilibrium early evolution stage, modeled by free-streaming partons generated from the Monte Carlo Kharzeev-Levin-Nardi (MC-KLN) and Monte Carlo Glauber (MC-Glb) models, to a locally approximately thermalized later evolution stage described by viscous hydrodynamics and study the dependence of final hadronic transverse momentum distributions, in particular their underlying radial and anisotropic flows, on the switching time between these stages. Performing a three-parameter fit of the measured values for the average transverse momenta 〈p⊥〉 for pions, kaons, and protons, as well as the elliptic and triangular flows of charged hadrons v2,3ch, with the switching time τs, the specific shear viscosity η/s during the hydrodynamic stage, and the kinetic decoupling temperature Tdec as free parameters, we find that the preferred “thermalization” times τs depend strongly on the model of the initial conditions. MC-KLN initial conditions require an earlier transition to hydrodynamic behavior (at τs≈0.13 fm/c), followed by hydrodynamic evolution with a larger specific shear viscosity η/s≈0.2, than MC-Glb initial conditions, which prefer switching at a later time (τs≈0.6 fm/c) followed by a less viscous hydrodynamic evolution with η/s≈0.16. These new results including pre-equilibrium evolution are compared to fits without a pre-equilibrium stage where all dynamic evolution before the onset of hydrodynamic behavior is ignored. In each case, the quality of the dynamical descriptions for the optimized parameter sets, as well as the observables which show the strongest constraining power for the thermalization time, are discussed.
  • 25.75.Cj
  • 25.75.Ld
  • 24.10.Nz
  • hydrodynamics: viscosity
  • heavy ion: scattering
  • model: boundary condition
  • transverse momentum: momentum spectrum
  • transverse momentum: hadronic
  • decoupling: temperature
  • decoupling: kinetic
  • plemented MC-KLN [10, 11] and MC-Glauber [12] initial conditions at LHC energies, allowing them to be evolved by free-streaming for a time τs before switching to a viscous hydrodynamic description for the rest of the evolution
    • Our work focuses on the effects brought by freestreaming on both the hydrodynamic initial conditions and the final observables. We mainly focus on the MCKLN model which provides a complete prediction for the initial gluon distribution, not only in space but also in momentum. However, we show that for massless partons moving with the speed of light the shape of the initial momentum distribution is irrelevant as long as it is locally isotropic. This allows to apply our description also to
      • MC-Glb initial conditions although that model makes no prediction per se about the initial parton momentum distribution. Free-streaming evolves the initial conditions from an initial parton formation time τ0 (which is taken to be very close to zero) to the switching time τs when we switch to a near-equilibrium hydrodynamic description
        • The sudden transition to approximate local equilibrium is implemented by applying the Landau matching procedure. By tuning the switching time, we can enforce fast thermalization by setting τs τ0, or slow thermalization by setting τs τ0. The hydrodynamic initial conditions obtained from the Landau matching procedure vary with the switching time, enabling an investigation of the influence of τs on the final observables. The hydrodynamic evolution is performed with the code VISH2+1 [8, 9], without hadron cascade afterburner. For the hydrodynamic evolution, we use a constant specific shear viscosity η/s. Freeze-out is implemented at a fixed kinetic freeze-out temperature Tdec, followed by a Cooper-Frye procedure with full resonance decay cascade to convert the hydrodynamic output into final stable particle spectra
          • In Section II, we outline the free-streaming evolution in the pre-equilibrium stage and describe the Landau matching procedure. Its consequences on the hydrodynamic initial conditions are discussed in Section III
            • Section IV shows how the hydrodynamical evolution responds to initial conditions generated at different switching times. A difficulty related to the conversion of partons to hadrons that arises from a late switching time τs is discussed and resolved in VI and
              • Section V. In Sections
              • VII, the energy flow anisotropy and hadron mean transverse momenta are constructed to illustrate how final observables change with switching time. Finally in Section
                • VIII we introduce a multidimensional parameter search procedure to systematically study the preferred ranges of
                  • τs, η/s and Tdec. For both MC-KLN and MC-Glb initial conditions, both with and without a free-streaming pre-equilibrium stage before the onset of hydrodynamic behavior, we determine the best-fit parameters and their uncertainty ranges, and discuss their relative quality of describing the data. Conclusions are presented in Section
                    • IX
                      • II. FORMULATION OF FREE-STREAMING
                        • AND LANDAU MATCHING
                          • The evolution of partons in the free-streaming model is described by the collisionless Boltzmann equation pµ
                            • ∂µf(x, p) = 0. (1)
                              • We work in Milne coordinates and write f(x, p) = f(x⊥, ηs, τ
                                • p⊥, y), with longitudinal proper time
                                  • τ =
                                    • √ t2-z2, space-time rapidity ηs = 1
                                      • [2]
                                        ln[(t+z)/(t-z)], and rapidity y = 1
                                        • [2]
                                          ln[(E+pz)/(E-pz)]. We assume massless partons for which E = |p| = p2
                                          • ⊥+p2 z. The collisionless Boltzmann equation is easily solved analytically, relating the final parton distribution f(x⊥, ηs, τs
                                            • p⊥, y) to the f(x⊥, ηs, τ0
                                              • p⊥, y) by a spatial coordinate shift, keeping the p⊥ distribution unchanged
                                                • For massless partons one finds: f(x⊥, ηs, τs
                                                  • p⊥, y) = f(x⊥-(τs-τ0)ˆp⊥, ηs, τ0