A 3-fluid hydrodynamic model for simulating relativistic heavy-ion collisions is introduced. Alongside with two baryon-rich fluids, the new model considers time-delayed evolution of a third, baryonfree (i.e. with zero net baryonic charge) fluid of newly produced particles. Its evolution is delayed due to a formation time τ , during which the baryon-free fluid neither thermalizes nor interacts with the baryon-rich fluids. After the formation it starts to interact with the baryon-rich fluids and quickly gets thermalized. Within this model with pure hadronic equation of state, a systematic analysis of various observables at incident energies between few and about 160A GeV has been done as well as comparison with results of transport models. We have succeeded to reasonably reproduce a great body of experimental data in the incident energy range of E lab ≃ (1-160)A GeV. The list includes proton and pion rapidity distributions, proton transverse-mass spectra, rapidity distributions of Λ andΛ hyperons, elliptic flow of protons and pions (with the exception of proton v2 at 40A GeV), multiplicities of pions, positive kaons, φ mesons, hyperons and antihyperons, including multi-strange particles. This agreement is achieved on the expense of substantial enhancement of the interflow friction as compared to that estimated proceeding from hadronic free cross sections. However, we have also found out certain problems. The calculated yield of K − is approximately by a factor of 1.5 higher than that in the experiment. We have also failed to describe directed transverse flow of protons and pion at E lab ≥ 40A GeV. This failure apparently indicates that the used EoS is too hard and thereby leaves room for a phase transition.

Central collisions of gold nuclei are simulated by several existing models and the central net baryon density rho and the energy density eps are extracted at successive times, for beam kinetic energies of 5-40 GeV per nucleon. The resulting trajectories in the (rho,eps) phase plane are discussed from the perspective of experimentally exploring the expected first-order hadronization phase transition with the planned FAIR at GSI or in a low-energy campaign at RHIC.Comment: 11 pages formatted, 17 eps files for 9 figure

Freeze-out procedure accepted in the model of 3-fluid dynamics (3FD) [1] is analyzed. This procedure is formulated in terms of drain terms in hydrodynamic equations. Dynamics of the freeze-out is illustrated by 1-dimensional simulations. It is demonstrated that the resulting freezeout reveals a nontrivial dynamics depending on initial conditions in the expanding "fireball". The freeze-out front is not defined just "geometrically" on the condition of the freeze-out criterion met but rather is a subject the fluid evolution. It competes with the fluid flow and not always reaches the place where the freeze-out criterion is met. Dynamics of the freeze-out in 3D simulations is analyzed. It is demonstrated that the late stage of central nuclear collisions at top SPS energies is of the form of three (two baryon-rich and one baryon-free) fireballs separated from each other.

Collective transverse flow in heavy-ion collisions at incident energies E lab ≃ (1-160)A GeV is analyzed within the model of 3-fluid dynamics (3FD). Simulations are performed with purely hadronic equation of state (EoS). At the AGS energies the flow turns out to be sensitive to the stopping power of nuclear matter rather than only to the stiffness of the EoS. When the stopping power is fixed to reproduce other observables, the flow data favor more and more soft EoS with the incident energy rise, which can be associated with "a transition from hadronic to string matter" reported in the Hadron-String-Dynamics (HSD) model. Problems, which are met in simultaneous reproduction of directed and elliptic flows within the 3FD, suggest that the transverse flow is very sensitive to the character of the transverse-momentum nonequilibrium at the initial stage of collision. Arguments in favor of "early-stage" nature of the flow observable are put forward. This suggests that the flow (especially the directed one) is determined by early-stage evolution of the collision rather than freeze-out stage.

No abstract