We use leading-order anisotropic hydrodynamics to study an azimuthally symmetric boost-invariant quark-gluon plasma. We impose a realistic lattice-based equation of state and perform self-consistent anisotropic freeze-out to hadronic degrees of freedom. We then compare our results for the full spatiotemporal evolution of the quark-gluon plasma and its subsequent freeze-out to results obtained using 1+1D Israel-Stewart second-order viscous hydrodynamics. We find that for small shear viscosities, $4\pi \eta /s\sim 1$, the two methods agree well for nucleus-nucleus collisions; however, for large-shear-viscosity-to-entropy-density ratios or proton-nucleus collisions we find important corrections to the Israel-Stewart results for the final particle spectra and the total number of charged particles. Finally, we demonstrate that the total number of charged particles produced is a monotonically increasing function of $4\pi \eta /s$ in Israel-Stewart viscous hydrodynamics, whereas in anisotropic hydrodynamics it has a maximum at $4\pi \eta /s\sim 10$. For all $4\pi \eta /s>0$, we find that for Pb-Pb collisions Israel-Stewart viscous hydrodynamics predicts more dissipative particle production than anisotropic hydrodynamics.

• Received 26 June 2015

DOI:https://doi.org/10.1103/PhysRevC.92.044912