We investigate the hadron transverse momentum (p T) spectra in Pb–Pb (Pb–Pb, Xe–Xe, p–Pb) collisions at 2.76 (5.02, 5.44, 5.02) TeV in the framework of the Tsallis blast wave model with a linear transverse velocity profile and a constant velocity profile. In this model, the Tsallis temperature (T), the average radial flow velocity (⟨β⟩) and the degree of non-equilibrium (q) of the system are common for all hadrons when a combined fit is performed to the p T spectra of different particles at a given centrality. It is found that the model can describe the particle spectra well up to 3 GeV c −1. For both profiles, the transverse flow velocity decreases from central to peripheral collisions while the non-extensive parameter q exhibits the opposite behavior, indicating a more rapid expansion and reduced off-equilibrium of the system in more central collisions. Moreover, we observe that, in central collisions, ⟨β⟩ and q (T) from the fit with the linear profile are smaller (is slightly larger) than those (that) from the constant profile, while in peripheral collisions ⟨β⟩, T and q from the linear profile are compatible with those from the constant profile. We also derived and discussed the relation between the Tsallis temperature and the thermal temperature. In addition, to check whether a scenario of an early freeze-out of strange particles exists at the Large Hadron Collider, the particle spectra are investigated by grouping them into strange and non-strange hadrons. The combined fit gives an insight on the degree of non-equilibrium, the radial flow and the Tsallis temperature of the system at the kinetic decoupling. It provides a comparison between the results at different energies in the same collision system and the results in different collision systems at the same or similar energy.

We investigate the nuclear modification factor (R AA ) of identified particles as a function of transverse momentum (p T ) in Pb-Pb collisions at √ s NN = 2.76 and 5.02 TeV, as well as p-Pb collision at 5.02 TeV in the framework of Boltzmann transport equation with relaxation time approximation. In this framework, the initial distribution of particles is chosen as the Tsallis distribution and the local equilibrium distribution as the Boltzmann-Gibbs blast-wave distribution. The nonextensive parameter q pp , the Tsallis temperature T pp and the equilibrium temperature T eq are set to be in common for all particles, while the ratio of kinetic freeze-out time to relaxation time t f /τ is different for different particles when we performed a combined fit to the R AA spectra of different particles at a given centrality. We observe that the fitted curves describe the spectra well up to p T ≈ 3 GeV/c. q pp and T eq (t f /τ ) decrease (increases) with centrality nonlinearly, while T pp is almost independent of centrality. The dependence of the rate at which q pp , T eq or t f /τ changes with centrality on the energy and the size of the colliding system is discussed.

We systematically investigate the scaling property of mesons (pions and kaons) and baryons (protons, Λ, Ξ and Ω) transverse momentum (p T ) spectra at different centrality classes (0-5%, 5-10%, 10-20%, 20-40%, 40-60%, 60-80% and 80-100%) in proton-lead collisions with center of mass energy per nucleon pair 5.02 TeV. In the low p T region with p T ≤ 3.9 (3.1, 2.5, 2.7, 2.4 and 2.8) GeV/c, a universal scaling independent of the centrality is observed in the pion (kaon, proton, Λ, Ξ and Ω) spectra when a dilatation, p T → p T /K, is applied. Here K is a scaling parameter depending on the centrality class. We find that the rates at which lnK changes with the logarithmic value of the average value of the number of participating nucleons, ln N part , are stronger for baryons than those for mesons. In the high p T region, there is a deviation from the scaling. The more peripheral the collisions are, the more obvious the violation of the scaling is. In the framework of the colour string percolation (CSP) model, we show that mesons and baryons are generated from the decay of clusters formed by strings overlapping in the transverse plane with the same size dispersion but with different mean size. The mean size of clusters for baryons is smaller than that of mesons. For the same hadrons at different centrality classes, the mean size of clusters decreases with the increase of centrality. The fragmentation functions for cluster decay are different for different hadrons, while they are universal for the same hadrons at different centrality classes. The universal scaling of the meson and baryon spectra in the low p T region can be quantitatively understood with the CSP model at the same time.1 Unless otherwise stated, in the later of the paper, we will follow the same definition of centrality classes.