betaFeedback

WorksPhysics and Astronomy

Spin dynamics with realistic hydrodynamic background for relativistic heavy-ion collisions

Abstract

The equations of perfect spin hydrodynamics are solved for the first time using a realistic (3+1)-dimensional hydrodynamic background, calibrated to reproduce a comprehensive set of hadronic observables, including rapidity distributions, transverse momentum spectra, and elliptic flow coefficients for Au+Au collisions at the beam energy of sNN=200\sqrt{s_{\rm NN}} = 200 GeV. The spin dynamics is governed by the conservation of the spin tensor, describing spin-12\frac{1}{2} particles, with particle mass in the spin tensor treated as an effective parameter. We investigate several scenarios, varying both the effective mass and the initial evolution time for the spin polarization tensor. The model predictions are then compared with experimental measurements of global and longitudinal spin polarization of Lambda hyperons. Our results indicate that a successful description of the data requires a delayed initial evolution time for the perfect spin hydrodynamics of about 4 fm/cc (in contrast to the standard initial time of 1 fm/cc used for the hydrodynamic background). This delay marks a transition from the phase where spin-orbit interaction is significant to the regime where spin-conserving processes dominate. Our findings suggest that the spin-orbit dissipative interaction plays a significant role only in the very early stages of the system's evolution.

Introduction

The measurements of the spin polarization of the Λ\Lambda hyperons and vector mesons has opened completely new possibilities for studying the properties of hot and dense matter produced in heavy-ion collisions [1, 2, 3, 4, 5, 6, 7, 8] (see Ref. [9] for a theory review). However, despite the increasing availability of data and plans for further measurements [10, 11], there remains a lack of consensus on the theoretical side regarding potential explanations for the observed polarization effects. The most widely adopted approach originates from quantum kinetic theory, resulting in a framework where polarization effects are determined solely by gradients of standard hydrodynamic variables at freeze-out. Notably, at least two significantly different implementations of this scheme yield satisfactory descriptions of the Λ\Lambda polarization data [12, 13, 14, 15, 16], leaving uncertainty about the true physical mechanism responsible for the observed polarization phenomena.

An alternative approach to those mentioned above is the framework of spin hydrodynamics, which has been developed over several years by independent research groups worldwide, employing a variety of physical concepts [12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88]. Thus far, efforts have primarily focused on establishing the foundational framework of spin hydrodynamics, assessing its stability and causality properties [58, 59, 69, 70, 71, 72], and deriving quasi-analytic solutions for simple expansion geometries, particularly Bjorken flow [61, 63, 69, 89].

The application of spin hydrodynamics to describe the spin polarization data appears natural, given the remarkable success of conventional relativistic hydrodynamics in explaining hadronic observables, particularly flow coefficients (see, for example, Ref. [90]). The ongoing development of spin hydrodynamics aims to incorporate spin polarization effects within the well-established framework of relativistic hydrodynamics.

In this work, we employ the formulation of perfect spin hydrodynamics as outlined in Refs. [25, 26, 28]. Key features of this approach include the use of a symmetric energy-momentum tensor together with a conserved spin tensor to describe spin-1/2 particles. For small values of the spin polarization tensor (a dimensionless quantity in natural units), the spin dynamics decouples from the background [27], i.e. the spin dynamics can be treated as externally determined by the spinless hydrodynamic background — specifically, a solution of the standard equations of relativistic dissipative hydrodynamics. Here, for the first time, we solve the equations of perfect spin hydrodynamics using a realistic (3+1)-dimensional [(3+1)D] hydrodynamic background. This background is calibrated to reproduce bulk hadronic observables, including rapidity distributions, transverse momentum spectra, and elliptic flow coefficients.

We examine several scenarios by varying the effective mass and initial evolution time of the spin polarization tensor and compare the model predictions with experimental data. Our results indicate that a successful description of the data requires a moderately delayed evolution time for spin hydrodynamics of about 4 fm/c, compared to the standard initial timescale of 1 fm/c. This timescale marks the transition from the phase where the spin-orbit interaction is significant to the regime dominated by spin-conserving processes.

The structure of our paper is as follows: In Sec. II, we introduce the hydrodynamic equations for both the background and the spin tensor. The numerical methods used to solve these equations are described in § 2. Our results are presented in § 3, where we compare them with experimental data and other theoretical models. Finally, we summarize and conclude in § 3.3. Technical details are provided in five appendices.

Conventions: We work in natural units where ℏ=kB=c=1\hbar= k_{\mathrm{B}} = c = 1. The metric is defined as gμν=diag⁡(1,−1,−1,−1)g_{\mu\nu} = \operatorname{diag}(1,-1,-1,-1). We use Milne coordinates (τ,x,y,ηs)(\tau,x,y,\eta_{\mathrm{s}}), where the longitudinal proper time τ\tau and spacetime rapidity ηs\eta_{\mathrm{s}} are defined as τ=t2−z2\tau= \sqrt{t^{2}-z^{2}} and ηs=12log⁡(t+zt−z)\eta_{\mathrm{s}} = \frac{1}{2}\log\left(\frac{t+z}{t-z}\right), respectively. For the Levi-Civita tensor ϵμναβ\epsilon^{\mu\nu\alpha\beta}, we follow the sign convention ϵ0123=−ϵ0123=+1\epsilon^{0123}=-\epsilon_{0123}=+1. We also adopt the notation A[μBν]=(AμBν−AνBμ)/2A^{[\mu}B^{\nu]}=(A^{\mu}B^{\nu}-A^{\nu}B^{\mu})/2 and A(μBν)=(AμBν+AνBμ)/2A^{(\mu}B^{\nu)}=(A^{\mu}B^{\nu}+A^{\nu}B^{\mu})/2 for the antisymmetric and symmetric parts of tensors.

Theoretical framework

Background hydrodynamics

The equations of relativistic viscous hydrodynamics arise from the conservation laws for the energy-momentum tensor, TαβT^{\alpha\beta}, and the net baryon current, NαN^{\alpha},

DαTαβ(x)=0,(1)D_{\alpha}T^{\alpha\beta}(x)=0, \tag*{(1)}
DαNα(x)=0,(2)D_{\alpha}N^{\alpha}(x)=0, \tag*{(2)}

where DαD_{\alpha} denotes the covariant derivative.4 We define the flow velocity uαu^{\alpha} using Landau’s convention, which specifies that in the local rest frame of the fluid cell, the energy flux should vanish

Tαβuβ=ϵuα.(3)T^{\alpha\beta}u_{\beta}=\epsilon u^{\alpha}. \tag*{(3)}

In this case, the constitutive relation for the energy-momentum tensor takes the form

Tαβ=ϵuαuβ−(Peq+Π)Δαβ+παβ,(4)T^{\alpha\beta}=\epsilon u^{\alpha}u^{\beta}-(P_{\mathrm{eq}}+\Pi)\Delta^{\alpha\beta}+\pi^{\alpha\beta}, \tag*{(4)}

where ϵ\epsilon is the energy density, PeqP_{\mathrm{eq}} is the corresponding equilibrium pressure, παβ\pi^{\alpha\beta} is the shear stress tensor, Π\Pi is the bulk pressure, and Δαβ=gαβ−uαuβ\Delta^{\alpha\beta}=g^{\alpha\beta}-u^{\alpha}u^{\beta} is the projector onto the space orthogonal to the flow velocity.

Spin hydrodynamics

The baryon current is expressed as

Nα=nuα+nα,(5)N^{\alpha}=n u^{\alpha}+n^{\alpha}, \tag*{(5)}

where nn represents the baryon number density. For this analysis, we assume that dissipative corrections from charge diffusion vanish, setting nα=0n^{\alpha}=0. At second order in spacetime gradients, the time evolution of the shear stress tensor παβ\pi^{\alpha\beta} and bulk pressure Π\Pi follows from the relaxation-type equations derived in the Denicol-Niemi-Molnar-Rischke (DNMR) framework [91, 92]:

Π˙=ΠNS−ΠτΠ−δΠΠτΠΠθ+λΠπτΠπαβσαβ,(6)\dot{\Pi}=\frac{\Pi_{\mathrm{NS}}-\Pi}{\tau_{\Pi}}-\frac{\delta_{\Pi\Pi}}{\tau_{\Pi}}\Pi\theta+\frac{\lambda_{\Pi\pi}}{\tau_{\Pi}}\pi^{\alpha\beta}\sigma_{\alpha\beta}, \tag*{(6)}
π˙⟨αβ⟩=πNSαβ−παβτπ−δππτππαβθ+λπΠτπΠσαβ−τππτππγ⟨ασβ⟩γ+ϕ7τππγ⟨απβ⟩γ.(7)\dot{\pi}^{\langle\alpha\beta\rangle}=\frac{\pi_{\mathrm{NS}}^{\alpha\beta}-\pi^{\alpha\beta}}{\tau_{\pi}}-\frac{\delta_{\pi\pi}}{\tau_{\pi}}\pi^{\alpha\beta}\theta+\frac{\lambda_{\pi\Pi}}{\tau_{\pi}}\Pi\sigma^{\alpha\beta}-\frac{\tau_{\pi\pi}}{\tau_{\pi}}\pi_{\gamma}^{\langle\alpha}\sigma^{\beta\rangle\gamma}+\frac{\phi_{7}}{\tau_{\pi}}\pi_{\gamma}^{\langle\alpha}\pi^{\beta\rangle\gamma}. \tag*{(7)}

Here, ( ˙)≡u⋅D(\dot{\ })\equiv u\cdot D denotes the comoving derivative, θ≡D⋅u\theta\equiv D\cdot u is the expansion scalar, and σαβ≡ΔαγΔβδDγuδ\sigma^{\alpha\beta}\equiv\Delta^{\alpha\gamma}\Delta^{\beta\delta}D_{\gamma}u_{\delta} is the shear flow tensor. The projector Δγδαβ≡12[ΔγαΔδβ+ΔδαΔγβ−(2/3)ΔαβΔγδ]\Delta_{\gamma\delta}^{\alpha\beta}\equiv\frac{1}{2}\left[\Delta_{\gamma}^{\alpha}\Delta_{\delta}^{\beta}+\Delta_{\delta}^{\alpha}\Delta_{\gamma}^{\beta}-(2/3)\Delta^{\alpha\beta}\Delta_{\gamma\delta}\right] isolates the symmetric, traceless, and orthogonal part of a rank-2 tensor relative to uαu^{\alpha}.

The Navier-Stokes forms of the dissipative currents in Eqs. (6)–(7) are ΠNS=−ζθ\Pi_{\mathrm{NS}}=-\zeta\theta and πNSαβ=2ησαβ\pi_{\mathrm{NS}}^{\alpha\beta}=2\eta\sigma^{\alpha\beta}, where the bulk and shear viscosities are chosen as in [92, 93],

η=Cηϵeq+PeqT,ζ=75η(13−cs2)2(8)\eta=C_{\eta}\frac{\epsilon_{\mathrm{eq}}+P_{\mathrm{eq}}}{T},\qquad\zeta=75\eta\left(\frac{1}{3}-c_{\mathrm{s}}^{2}\right)^{2} \tag*{(8)}

with

cs2=∂Peq∂ϵeq∣neq+neqϵeq+Peq∂Peq∂neq∣ϵeq(9)c_{\mathrm{s}}^{2}=\left.\frac{\partial P_{\mathrm{eq}}}{\partial\epsilon_{\mathrm{eq}}}\right|_{n_{\mathrm{eq}}}+\frac{n_{\mathrm{eq}}}{\epsilon_{\mathrm{eq}}+P_{\mathrm{eq}}}\left.\frac{\partial P_{\mathrm{eq}}}{\partial n_{\mathrm{eq}}}\right|_{\epsilon_{\mathrm{eq}}} \tag*{(9)}

representing the speed of sound squared. The coefficient CηC_{\eta} is set equal to 0.12. The relaxation times for bulk and shear corrections as well as the second-order transport coefficients, derived using the 14-moment approximation in massless limit, are given by [92, 94]:

τπ=5ηϵeq+Peq,τΠ=ζ15(13−cs2)2(ϵeq+Peq)(10)\tau_{\pi}=\frac{5\eta}{\epsilon_{\mathrm{eq}}+P_{\mathrm{eq}}},\qquad\tau_{\Pi}=\frac{\zeta}{15\left(\frac{1}{3}-c_{\mathrm{s}}^{2}\right)^{2}(\epsilon_{\mathrm{eq}}+P_{\mathrm{eq}})} \tag*{(10)}
δΠΠτΠ=23,λΠπτΠ=85(13−cs2),δππτπ=43(11)\frac{\delta_{\Pi\Pi}}{\tau_{\Pi}}=\frac{2}{3},\qquad\frac{\lambda_{\Pi\pi}}{\tau_{\Pi}}=\frac{8}{5}\left(\frac{1}{3}-c_{\mathrm{s}}^{2}\right),\qquad\frac{\delta_{\pi\pi}}{\tau_{\pi}}=\frac{4}{3} \tag*{(11)}
λπΠτπ=65,τππτπ=107,ϕ7τπ=970Peqτπ.(12)\frac{\lambda_{\pi\Pi}}{\tau_{\pi}}=\frac{6}{5},\qquad\frac{\tau_{\pi\pi}}{\tau_{\pi}}=\frac{10}{7},\qquad\frac{\phi_{7}}{\tau_{\pi}}=\frac{9}{70P_{\mathrm{eq}}\tau_{\pi}}. \tag*{(12)}

The choice of η\eta and ζ\zeta in Eq. (8) implies τπ=τΠ=5CηT\tau_{\pi}=\tau_{\Pi}=\frac{5C_{\eta}}{T}.

Spin hydrodynamics

For a symmetric energy-momentum tensor, such as the one considered in Eq. (4), the conservation law for total angular momentum, DαJα,βγ=0D_{\alpha}J^{\alpha,\beta\gamma}=0, implies the conservation of the spin current

DαSα,βγ(x)=0.(13)D_{\alpha}S^{\alpha,\beta\gamma}(x)=0. \tag*{(13)}

The spin current is a rank-3 tensor that is antisymmetric in its last two indices. In the de Groot-van Leeuwen-van Weert (GLW) pseudogauge, the spin current is expressed as [27, 95]

Sα,βγ=A1uαωβγ+A2uαu[βωγ]δuδ+A3(u[βωγ]α+gα[βωγ]δuδ).(14)S^{\alpha,\beta\gamma}=\mathcal{A}_1u^\alpha\omega^{\beta\gamma}+\mathcal{A}_2u^\alpha u^{[\beta}\omega^{\gamma]\delta}u_\delta+\mathcal{A}_3\left(u^{[\beta}\omega^{\gamma]\alpha}+g^{\alpha[\beta}\omega^{\gamma]\delta}u_\delta\right). \tag*{(14)}

where ωβγ\omega^{\beta\gamma} is the spin polarization tensor, which is antisymmetric: ωβγ=−ωγβ\omega^{\beta\gamma}=-\omega^{\gamma\beta}. The coefficient functions A\mathcal{A} are given by the following expressions:

A1=cosh⁡(μBT)T3π2[(4+z22)K2(z)+zK1(z)],\mathcal{A}_1=\cosh\left(\frac{\mu_B}{T}\right)\frac{T^3}{\pi^2}\left[\left(4+\frac{z^2}{2}\right)K_2(z)+zK_1(z)\right],
A2=2cosh⁡(μBT)T3π2[(12+z22)K2(z)+3zK1(z)],\mathcal{A}_2=2\cosh\left(\frac{\mu_B}{T}\right)\frac{T^3}{\pi^2}\left[\left(12+\frac{z^2}{2}\right)K_2(z)+3zK_1(z)\right],
A3=12(A1−A22).(15)\mathcal{A}_3=\frac{1}{2}\left(\mathcal{A}_1-\frac{\mathcal{A}_2}{2}\right). \tag*{(15)}

where μB\mu_B is the baryon chemical potential, z≡m/Tz\equiv m/T is the ratio of the particle mass mm to the temperature TT, and Kn(z)K_n(z) denotes the modified Bessel functions of the second kind of order nn. Even though the spin degrees of freedom evolve independently, the spin-polarized particles interact with the medium. As a result, their in-medium effective mass can differ from the vacuum one. Therefore, we treat mm as an effective parameter.

By substituting Eq. (14) into Eq. (13), one obtains a system of six evolution equations for the components of the spin polarization tensor ωβγ\omega^{\beta\gamma}, provided that the temperature TT, baryon chemical potential μB\mu_B, and four-velocity uαu^\alpha are known.

Numerical framework

Background hydrodynamics

Equations (1), (2), (6), and (7), supplemented by an equation of state relating pressure to energy density and baryon density, Peq=Peq(ε,n)P_{\rm eq}=P_{\rm eq}(\varepsilon,n), form a closed set of eleven differential equations for eleven unknown hydrodynamic fields: ε\varepsilon, nn, uμu^\mu, Π\Pi, and πμν\pi^{\mu\nu}.5 It is important to note that, within the model, we consider the limit of small polarization [27], where Eqs. (1), (2), (6), and (7) do not receive feedback from spin equations of motion via ω\omega. Therefore, we solve these equations numerically in (3+1)(3+1) dimensions in Milne coordinates (τ,x,y,ηs)(\tau,x,y,\eta_s) using the code developed in Ref. [96], which employs the Godunov-type relativistic Harten-Lax-van Leer-Einfeldt (HLLE) approximate Riemann solver, as described in Ref. [97]. Additionally, we extend the code from Ref. [96] to include the cross terms in Eqs. (6)–(7), associated with the non-vanishing transport coefficients λΠπ\lambda_{\Pi\pi}, λπΠ\lambda_{\pi\Pi}, τππ\tau_{\pi\pi}, and φ7\varphi_7.

In contrast to Ref. [96], in the present study, we use the lattice-QCD-based equation of state (EOS) at finite net baryon density, NEOS-BQS [98, 99], which exhibits a crossover phase transition across the entire parametric space of the phase diagram.

Herein, we focus on the Au+Au collisions at the top RHIC energy of sNN=200\sqrt{s_{\rm NN}}=200 GeV. We initialize the background evolution at the proper time τ0=1\tau_0=1 fm. The initial energy density and baryon density profiles are determined according to the model based on the Glauber collision geometry with local energy-momentum conservation [100, 101]. However, unlike in Refs. [100, 101], to compute the thickness functions and wounded nucleon densities we use the optical limit of the Glauber model. The initial transverse flow components are chosen to be zero whereas the longitudinal flow is numerically obtained from the initial energy-momentum tensor components. For simplicity, we also assume that the dissipative corrections initially vanish, i.e., πμν(τ0)=0\pi^{\mu\nu}(\tau_0)=0 and Π(τ0)=0\Pi(\tau_0)=0. Further details of the initial state model used for the energy-momentum and baryon currents are provided in Appendix A.

The initial conditions obtained within the optical limit of the Glauber model involve a fixed impact parameter. Its value, corresponding to a given centrality range, is determined using the two-component Glauber model. The procedure for determining the impact parameter is detailed in Appendix B, and the values obtained for the centrality classes considered in this work are listed in Table III.

Centrality (%)bb (fm)
5-103.8
10-205.6
20-307.2
30-408.6
40-509.7
20-508.4

Table III. The values of mean impact parameter for the different centrality classes considered.

We evolve the hydrodynamic equations until the energy density in the system decreases below the threshold value εsw=0.5 GeV/fm3\varepsilon_{\rm sw}=0.5\ {\rm GeV}/{\rm fm}^3. The condition ε(T,μB)=εsw\varepsilon(T,\mu_B)=\varepsilon_{\rm sw} defines the switching hypersurface Σ\Sigma, which is extracted using the CORNELIUS code [102]. The hydrodynamic fields on the hypersurface are passed to a hadron sampler [103, 104, 105], which generates particles from fluid elements. The resulting particle set serves as input to the SMASH transport model [106, 107], which describes subsequent hadron interactions and decays.

Spin hydrodynamics

Using the hydrodynamic background obtained from Eqs. (1), (2), (6), and (7) (as described in the previous section), we determine the effective temperature T(x)T(x) and baryon chemical potential μB(x)\mu_B(x). These, together with the four-velocity field, allow us, through Eq. (14), to compute the dynamics of the spin polarization tensor ω\omega from Eq. (13). For this purpose, we further extend the code to incorporate the spin conservation equation (13). The respective numerical implementation, using also HLLE algorithm, is explained in detail in Appendix C.

Motivated by Refs. [12, 14, 111, 112], we use the following initial condition for the spin polarization tensor

ωμν(τ0s)=ϖμνiso+4τ^[μξν]ρisouρ,(16)\omega_{\mu\nu}(\tau_0^s)=\varpi^{\mathrm{iso}}_{\mu\nu}+4\hat{\tau}_{[\mu}\xi^{\mathrm{iso}}_{\nu]\rho}u^\rho, \tag*{(16)}

where the right-hand side is understood to be evaluated at the initial time of the spin evolution τ0s\tau_0^s, which we treat as a free parameter of the model to be extracted from the experimental data. In Eq. (16), ϖμνiso=∂[νuμ]\varpi^{\mathrm{iso}}_{\mu\nu}=\partial_{[\nu}u_{\mu]} is the isothermal part of thermal vorticity ϖμν=∂[νβμ]\varpi_{\mu\nu}=\partial_{[\nu}\beta_{\mu]}, namely, ϖμν=ϖμνiso+ϖμνT\varpi_{\mu\nu}=\varpi^{\mathrm{iso}}_{\mu\nu}+\varpi^T_{\mu\nu},6 where ϖμνT=1Tu[μ∂ν]ln⁡T\varpi^T_{\mu\nu}=\frac{1}{T}u_{[\mu}\partial_{\nu]}\ln T and βμ\beta^\mu is the ratio of the flow velocity uμu^\mu to local temperature TT, βμ=uμ/T\beta^\mu=u^\mu/T. Similarly, we introduce ξμνiso=∂(νuμ)\xi^{\mathrm{iso}}_{\mu\nu}=\partial_{(\nu}u_{\mu)}, which represents the isothermal part of the thermal shear tensor ξμν=∂(νβμ)\xi_{\mu\nu}=\partial_{(\nu}\beta_{\mu)}. The part ϖμνT\varpi^T_{\mu\nu} (as well as the respective part of the thermal shear tensor) vanishes when temperature gradients are absent. Note that τ^μ=(1,0,0,0)\hat{\tau}^\mu=(1,0,0,0) is a fixed timelike vector which in Milne coordinates is normal to the constant-τ\tau hypersurface.

As an alternative, we also consider the analog of the initial condition (16) which includes all temperature gradients present in the system

ωμν(τ0s)=ϖμν+4τ^[μξν]ρuρ.(17)\omega_{\mu\nu}(\tau_0^s) = \varpi_{\mu\nu} + 4\hat{\tau}_{[\mu}\xi_{\nu]\rho}u^\rho. \tag*{(17)}

We note that by using Eqs. (16) or (17), we intend to account for equilibration of spin degrees of freedom resulting from strong spin-orbit interactions occurring in the early stages of heavy-ion collisions, before the system reaches the regime where perfect spin hydrodynamics is applicable.

Spin polarization of Lambdas

In this work, we focus on polarization observables for Λ\Lambda hyperons. After determining the spacetime dynamics of the spin polarization tensor ω(x)\omega(x), we calculate the components of the polarization vector Sμ(p)S^\mu(p) for Λ\Lambda particles with momenta pp. For the GLW pseudogauge used in the present study, the polarization vector is given by the formula [111]

Sμ(p)=−18mΛϵμνρσpσ∫dΣ⋅p nF(1−nF)ωνρ∫dΣ⋅p nF(18)S^\mu(p) = -\frac{1}{8m_\Lambda}\epsilon^{\mu\nu\rho\sigma}p_\sigma\frac{\int d\Sigma\cdot p\ n_F(1-n_F)\omega_{\nu\rho}}{\int d\Sigma\cdot p\ n_F} \tag*{(18)}

where nF=nF(T,μB,p⋅u;mΛ)n_F=n_F(T,\mu_B,p\cdot u;m_\Lambda) is the Fermi-Dirac distribution function, which depends on the background hydrodynamic variables and the Λ\Lambda mass, mΛm_\Lambda. The integral is performed over the switching hypersurface Σ\Sigma determined as described in § 2 A.

Our results are compared with those obtained using the spin polarization formula obtained at first order of thermodynamic gradients [14, 111],

Sμ(p)=Sϖμ(p)+Sξμ(p),(19)S^\mu(p)=S^\mu_\varpi(p)+S^\mu_\xi(p), \tag*{(19)}

where the term containing contribution from thermal vorticity, Sϖμ(p)S^\mu_\varpi(p), is given by [114]

Sϖμ(p)=−18mΛϵμνρσpσ∫dΣ⋅p nF(1−nF)ϖνρ∫dΣ⋅p nF.(20)S^\mu_\varpi(p)=-\frac{1}{8m_\Lambda}\epsilon^{\mu\nu\rho\sigma}p_\sigma\frac{\int d\Sigma\cdot p\ n_F(1-n_F)\varpi_{\nu\rho}}{\int d\Sigma\cdot p\ n_F}. \tag*{(20)}

For the contribution from thermal shear, there are currently two prescriptions available in the literature: one proposed in Ref. [14] (henceforth denoted as BBP), given by

Sξ,BBPμ(p)=−ϵμνρσ4mΛpσpλp⋅t^∫dΣ⋅p nF(1−nF)t^νξλρ∫dΣ⋅p nF(21)S_{\xi,\mathrm{BBP}}^\mu(p)=-\frac{\epsilon^{\mu\nu\rho\sigma}}{4m_\Lambda}\frac{p_\sigma p^\lambda}{p\cdot\hat{t}}\frac{\int d\Sigma\cdot p\,n_F(1-n_F)\hat{t}_\nu\xi_{\lambda\rho}}{\int d\Sigma\cdot p\,n_F} \tag*{(21)}

and another proposed in Ref. [12] (henceforth denoted as LY), given by

Sξ,LYμ(p)=−ϵμνρσ4mΛpσpλ∫dΣ⋅p nF(1−nF)βuνp⋅u∂ρuλ∫dΣ⋅p nF=−ϵμνρσ4mΛpσ∫dΣ⋅p nF(1−nF)pλuνp⋅uξλρ∫dΣ⋅p nF(22)\begin{aligned} S_{\xi,\mathrm{LY}}^\mu(p)=-\frac{\epsilon^{\mu\nu\rho\sigma}}{4m_\Lambda}p_\sigma p^\lambda\frac{\int d\Sigma\cdot p\,n_F(1-n_F)\frac{\beta u_\nu}{p\cdot u}\partial_\rho u_\lambda}{\int d\Sigma\cdot p\,n_F} \\ &=-\frac{\epsilon^{\mu\nu\rho\sigma}}{4m_\Lambda}p_\sigma\frac{\int d\Sigma\cdot p\,n_F(1-n_F)\frac{p^\lambda u_\nu}{p\cdot u}\xi_{\lambda\rho}}{\int d\Sigma\cdot p\,n_F} \tag*{(22)} \end{aligned}

where ∂⊥μ≡Δνμ∂ν\partial^\mu_\perp\equiv\Delta^{\nu\mu}\partial_\nu and p⊥μ≡Δνμpνp^\mu_\perp\equiv\Delta^{\nu\mu}p_\nu.

Following recent works, we use Eqs. (20)–(22) in two ways: with and without temperature gradients included. In the latter case, we replace ϖ\varpi and ξ\xi by ϖiso\varpi^{\mathrm{iso}} and ξiso\xi^{\mathrm{iso}}, respectively. The approach with neglected temperature gradients has been shown to be crucial for correctly reproducing the longitudinal component of the spin vector in Refs. [15, 16, 24, 101].

It is important to note that since we perform calculations on the switching hypersurface Σ\Sigma defined by a constant energy density, our hypersurface does not exactly correspond to a constant value of TT. However, as shown in Figure 1, the temperature is nearly constant on Σ\Sigma. On the other hand, the ratio μB/T\mu_B/T exhibits a rather broad distribution.

Area-weighted distributions of temperature $T$ and scaled baryon chemical potential $\mu_B/T$ on the switching hypersurface $\Sigma$

Figure 1. Area-weighted distributions of temperature TT and scaled baryon chemical potential μB/T\mu_B/T on the switching hypersurface Σ\Sigma.

Results

Background hydrodynamics

Our numerical results describing charged hadron observables are presented in Figure 2. The upper panel shows the transverse-momentum spectra, the middle panel displays the pseudorapidity distributions, and the lower panel illustrates the elliptic flow. These results are compared with experimental data from Au+Au collisions at sNN=200\sqrt{s_{\mathrm{NN}}}=200 GeV across various centrality bins. Overall, we observe good agreement between the model calculations and the experimental data.

Bulk observables for Au+Au collisions at $\sqrt{s_{NN}}=200$ GeV

Figure 2. Bulk observables for Au+Au collisions at sNN=200\sqrt{s_{NN}=200} GeV. The solid lines represent our numerical results, while the different symbols correspond to experimental data taken from Ref. [108] in the top panel, from Ref. [109] in the middle panel, and from Ref. [110] in the bottom panel.

Spin hydrodynamics

Figures 3 – 10 present a collection of our numerical results for the spin polarization of Λ\Lambda hyperons, obtained using various model parameters and initial conditions. We varied the initial time for spin hydrodynamics, τ0s\tau_0^s, the effective mass appearing in the spin tensor, mm, and the form of the initial conditions by switching between Eqs. (16) and (17).

Figures 3 and 4 show our results (solid red lines) for the global and longitudinal polarizations, respectively, using τ0s=4\tau_0^s=4 fm and m=mΛm=m_\Lambda. Experimental data are represented by teal dots with vertical lines marking the experimental errors. For comparison, we also show the results obtained within the frameworks proposed in Ref. [14] (blue dashed lines, denoted as BBP) and Ref. [12] (green dotted lines, denoted as LY), using Eqs. (21) and (22), respectively. In the case of the global polarization, Figure 3, our prediction for the pTp_T dependence of PJP_J agrees with the experimental data for pTp_T smaller than 1 GeV. On the other hand, our results for the rapidity dependence of PJP_J are very close to the data in the whole measured range. Similar trends can be observed for the longitudinal polarization shown in Figure 4. The results presented in Figure 3 and Figure 4 indicate that the LY and BBP frameworks describe the data better, especially the LY approach, which accurately reproduces all the experimental points, including those at large pTp_T for PJP_J.

Component of $\Lambda$ polarization along the orbital angular momentum direction

Figure 3. Component of Λ\Lambda polarization along the orbital angular momentum direction, plotted as a function of transverse momentum, pTp_T (top), and pseudorapidity, η\eta (bottom). Experimental data is taken from Ref. [3] with updated decay parameter (αΛ=0.732\alpha_\Lambda= 0.732). Solid lines represent our results from spin hydrodynamics with the initial condition model given by Eq. (16), using the parameters, τ0s=4 fm\tau_0^s = 4\ \mathrm{fm} and m=mΛm = m_\Lambda.

Longitudinal $\Lambda$ polarization

Figure 4. Longitudinal Λ\Lambda polarization, plotted as a function of transverse momentum, pTp_T (top), and azimuthal angle, ϕ\phi (bottom). Experimental data are taken from Ref. [4] with updated decay parameter (αΛ=0.732\alpha_\Lambda= 0.732). Solid lines represent our results from spin hydrodynamics with the initial condition model given by Eq. (16), using the parameters τ0s=4 fm\tau_0^s = 4\ \mathrm{fm} and m=mΛm = m_\Lambda.

To examine the effect of reducing the effective mass, we performed calculations within our framework using m=300m=300 MeV. We stress that we use the reduced mass in the calculations that determine the spacetime evolution of the spin tensor only. To calculate the spin polarization we use the physical Λ\Lambda mass. Our results with the reduced mass (with initial time and form of the initial condition for the spin polarization tensor unchanged) are shown in Figure 5 and Figure 6. Here, we observe a clear improvement of the agreement with the data compared to the case where the value m=mΛm = m_\Lambda is used throughout the calculations. In this case, our results closely align with the BBP predictions.

Global and longitudinal polarization as functions of transverse momentum and pseudorapidity

Figure 5. Same as Figure 3 but with m=300m=300 MeV.

Four panels showing longitudinal and transverse polarization observables as functions of transverse momentum, azimuthal angle, and pseudorapidity

Figure 6. Same as Figure 4 but with m=300m = 300 MeV.

To further explore the sensitivity of our results to the form of the initial conditions, we have replaced the form (16) by (17), thereby including temperature gradients in the initial conditions. Again, with the reduced mass but unchanged initial time for spin hydrodynamics we obtain the results shown in Figure 7 and Figure 8. While the magnitude of PJP_J is only slightly changed, the longitudinal polarization changes sign — an effect known from early attempts to describe PzP_z with thermal vorticity, Eq. (20), only and dubbed the “sign problem”.

Four panels showing polarization observables for the initial condition (17)

Figure 7. Same as Figure 3 but with initial condition (17) and parameters τ0s=4\tau_0^s = 4 fm and m=300m = 300 MeV.

Global and longitudinal spin polarization results for initial condition (17)

Figure 8. Same as Figure 4 but with initial condition (17) and parameters τ0s=4\tau_0^s = 4 fm and m=300m = 300 MeV.

The results presented so far are consistent with earlier findings suggesting that omitting temperature gradients or reducing the mass improves agreement with data. Our framework offers, however, new insights into the spin evolution of matter produced in heavy-ion collisions. Since our framework is based on the equation that describes conservation of the spin part of the total angular momentum, its ability to reproduce the experimental data suggests that the last stages of the evolution of matter are dominated by the processes that do not exchange the spin and orbital parts of the angular momentum.

To determine when the initial stage with a strong spin-orbit coupling transitions into a stage where this coupling is weak, we initiated our spin evolution at different times, specifically τ0s=2,3,\tau_0^s = 2, 3, and 44 fm. The results of these calculations (using m=mΛm = m_\Lambda) are shown in Figure 9 and Figure 10. Inspection of these figures indicates that the initialization time for spin hydrodynamics cannot be set too early. The acceptable agreement with the data is only achieved with τ0s=4\tau_0^s = 4 fm. In a way, the exclusion of early times from our analysis of spin dynamics seems reasonable, since the formalism of spin hydrodynamics used in this work refers to well-defined particles (quasiparticles) with spin 1/21/2. In the very early stages of heavy-ion collisions, the system is very strongly interacting, the concept of quasiparticles can be questioned, and the currently developed methods of spin hydrodynamics might be not applicable.

Numerical results for the component of $\Lambda$ polarization along the orbital angular momentum direction

Figure 9. Our numerical results for the component of Λ\Lambda polarization along the orbital angular momentum direction for different initial time of spin evolution τ0s\tau_0^s keeping mass fixed as mΛm_\Lambda. Experimental data is taken from Ref. [3] with updated decay parameter (αΛ=0.732\alpha_\Lambda= 0.732).

Numerical results for longitudinal Λ polarization

Figure 10. Our numerical results for longitudinal Λ\Lambda polarization for different initial time of spin evolution τ0s\tau_0^s keeping mass fixed as mΛm_\Lambda. Experimental data are taken from Ref. [4] with updated decay parameter (αΛ=0.732\alpha_\Lambda= 0.732).

Summary and outlook

In this work, we solved the equations of perfect spin hydrodynamics for the first time using a realistic 3+1D hydrodynamic background, calibrated to reproduce a wide range of hadronic observables measured in Au+Au collisions at the beam energy of sNN=200\sqrt{s_{\mathrm{NN}}} = 200 GeV. The spin dynamics of spin 1/21/2 particles has been determined from the conservation law for the spin part of the total angular momentum. The mass appearing in the spin tensor has been treated as an effective parameter. We have analyzed several scenarios by varying the effective mass, the initial evolution time, and the form of the initial spin polarization tensor. Our model predictions have been compared with the data on global and longitudinal spin polarization of Λ\Lambda hyperons. The results indicate that a successful description of the data requires a delayed initial evolution time for the perfect spin hydrodynamics of about 4 fm (compared to the standard initial time of 1 fm used for the hydrodynamic background). This delay signals a transition from the phase where the spin-orbit interaction is significant to the regime where spin-conserving processes dominate. Our results suggest that the spin-orbit dissipative interaction plays a significant role only in the very early stages of the system evolution.

Acknowledgments

The authors (S.K.S and W.F.) thank ECT* for support allowing participation in the workshop “Spin and quantum features of the QCD plasma” where some part of this work was done. This research was supported in part by the Polish National Science Center (NCN) Grants No. 2018/30/E/ST2/00432 (R.R., S.K.S.) and 2022/47/B/ST2/01372 (W.F.). S.K.S is supported by ICSC - Centro Nazionale di Ricerca in High Performance

Computing, Big Data and Quantum Computing, funded by the European Union - NextGenerationEU

Initial condition

We use initial conditions defined in Refs. [100, 101], which are based on the Glauber collision geometry with local energy-momentum conservation. However, in contrast to these references, we apply the optical Glauber model to compute thickness functions and wounded nucleon densities. We denote the wounded nucleon densities per unit area in the transverse plane for nuclei AA and BB as nAn_A and nBn_B, respectively. If the impact parameter is bb, then nAn_A and nBn_B are given by:

nA(x,y)=TA(x−b2,y)[1−(1−σNNTB(x+b/2,y)NB)NB],(23)n_A(x,y) = T_A\left(x-\frac{b}{2},y\right)\left[1-\left(1-\frac{\sigma_{NN}T_B(x+b/2,y)}{N_B}\right)^{N_B}\right], \tag*{(23)}
nB(x,y)=TB(x+b2,y)[1−(1−σNNTA(x−b/2,y)NA)NA],(24)n_B(x,y) = T_B\left(x+\frac{b}{2},y\right)\left[1-\left(1-\frac{\sigma_{NN}T_A(x-b/2,y)}{N_A}\right)^{N_A}\right], \tag*{(24)}

where σNN\sigma_{NN} denotes the nucleon-nucleon inelastic cross-section, and NIN_I denotes the total number of nucleons in nucleus I∈{A,B}I \in\{A,B\}. The nuclear thickness functions TI(x,y)T_I(x,y) are calculated by integrating the nuclear density ϱI(x,y,z)\varrho_I(x,y,z) along the longitudinal zz-axis (the beam direction in the collision)

TI(x,y)=∫−∞∞ϱI(x,y,z′) dz′.(25)T_I(x,y) = \int_{-\infty}^{\infty}\varrho_I(x,y,z')\,\mathrm{d}z'. \tag*{(25)}

The nuclear density is assumed to have the Woods-Saxon form

ϱI(x,y,z)=ϱ01+e(x2+y2+z2−R)/δ,(26)\varrho_I(x,y,z) = \frac{\varrho_0}{1+e^{\left(\sqrt{x^2+y^2+z^2}-R\right)/\delta}}, \tag*{(26)}

where RR is the nuclear radius and δ\delta is the surface thickness. For Au+Au collisions considered in this work, the values used are R=6.37R=6.37 fm and δ=0.535\delta=0.535 fm. The normalization constant ϱ0\varrho_0 is set such that

∫ϱI(r⃗) d3r⃗=NI.(27)\int\varrho_I(\vec{r})\,\mathrm{d}^3\vec{r}=N_I. \tag*{(27)}

If ybeam=cosh⁡−1(sNN2mN)y_{\mathrm{beam}}=\cosh^{-1}\left(\frac{\sqrt{s_{NN}}}{2m_N}\right) denotes the beam rapidity, then the energy and longitudinal momentum associated with a nucleon are mNcosh⁡(ybeam)m_N\cosh(y_{\mathrm{beam}}) and mNsinh⁡(ybeam)m_N\sinh(y_{\mathrm{beam}}), respectively. Assuming that nuclei AA and BB move along zz-axis in positive and negative directions, respectively, the net energy and net longitudinal momentum deposited per unit area in the transverse plane are:

d2Edxdy=[nA(x,y)+nB(x,y)]mNcosh⁡(ybeam)≡M(x,y)cosh⁡(Y(x,y)),\frac{\mathrm{d}^2\mathcal{E}}{\mathrm{d}x\mathrm{d}y}=[n_A(x,y)+n_B(x,y)]m_N\cosh(y_{\mathrm{beam}})\equiv M(x,y)\cosh(Y(x,y)),
d2Pzdxdy=[nA(x,y)−nB(x,y)]mNcosh⁡(ybeam)≡M(x,y)sinh⁡(Y(x,y)).\frac{\mathrm{d}^2P_z}{\mathrm{d}x\mathrm{d}y}=[n_A(x,y)-n_B(x,y)]m_N\cosh(y_{\mathrm{beam}})\equiv M(x,y)\sinh(Y(x,y)).

The local invariant mass, MM, and the local center-of-mass rapidity, YY, are defined in terms of nAn_A and nBn_B as:

M(x,y)=mNnA2+nB2+2nAnBcosh⁡(2ybeam),Y(x,y)=tanh⁡−1[nA−nBnA+nBtanh⁡(ybeam)].(28)M(x,y)=m_N\sqrt{n_A^2+n_B^2+2n_An_B\cosh(2y_{\mathrm{beam}})},\qquad Y(x,y)=\tanh^{-1}\left[\frac{n_A-n_B}{n_A+n_B}\tanh(y_{\mathrm{beam}})\right]. \tag*{(28)}

To match the deposited energy and momentum with those obtained from the energy-momentum tensor, we follow the ansatz in Ref. [101] for the energy-momentum tensor at the initial proper time τ=τ0\tau=\tau_0 hypersurface:

Tττ(τ0;x,y,ηs)=ε(τ0;x,y,ηs)cosh⁡(fY),(29)T^{\tau\tau}(\tau_0;x,y,\eta_s)=\varepsilon(\tau_0;x,y,\eta_s)\cosh(fY), \tag*{(29)}
Tτx(τ0;x,y,ηs)=Tτy(τ0;x,y,ηs)=0,(30)T^{\tau x}(\tau_0;x,y,\eta_s)=T^{\tau y}(\tau_0;x,y,\eta_s)=0, \tag*{(30)}
Tτηs(τ0;x,y,ηs)=1τ0ε(τ0;x,y,ηs)sinh⁡(fY),(31)T^{\tau\eta_s}(\tau_0;x,y,\eta_s)=\frac{1}{\tau_0}\varepsilon(\tau_0;x,y,\eta_s)\sinh(fY), \tag*{(31)}

where ff is a parameter controlling the fraction of longitudinal momentum converted into fluid velocity. The parameter ff is set equal to 0.15. For symmetric collisions, the initial energy density is given by

ε(τ0;x,y,ηs)=ε⊥(x,y)F(ηs),(32)\varepsilon(\tau_0;x,y,\eta_s)=\varepsilon_\perp(x,y)F(\eta_s), \tag*{(32)}

where F(ηs)F(\eta_s) is defined as [100]

F(ηs)=exp⁡[−(∣ηs−(1−f)Y∣−η0)22ση2Θ(∣ηs−(1−f)Y∣−η0)](33)F(\eta_s) = \exp\left[-\frac{\left(\left|\eta_s-(1-f)Y\right|-\eta_0\right)^2}{2\sigma_\eta^2}\Theta\left(\left|\eta_s-(1-f)Y\right|-\eta_0\right)\right] \tag*{(33)}

with Θ(x)\Theta(x) denoting the Heaviside function. By substituting the above expressions into the matching condition at the initial time, we obtain ε⊥(x,y)=NεM(x,y)\varepsilon_\perp(x,y)=\mathcal{N}_\varepsilon\mathcal{M}(x,y), where Nε\mathcal{N}_\varepsilon is a normalization constant expressed in terms of η0\eta_0 and ση\sigma_\eta. The choice of initial energy-momentum tensor results in zero transverse velocity, and the ηs\eta_s-component of the velocity is determined numerically by solving the implicit equation

vηs(τ0;x,y,ηs)=Tτηs(τ0;x,y,ηs)Tττ(τ0;x,y,ηs)+Peq(ε,n)v^{\eta_s}(\tau_0;x,y,\eta_s)=\frac{T^{\tau\eta_s}(\tau_0;x,y,\eta_s)}{T^{\tau\tau}(\tau_0;x,y,\eta_s)+P_{\mathrm{eq}}(\varepsilon,n)}

For the initial baryon density, we consider the following profile

n(τ0;x,y,ηs)=Nn[gA(ηs)nA(x,y)+gB(ηs)nB(x,y)](34)n(\tau_0;x,y,\eta_s)=\mathcal{N}_n\left[g_A(\eta_s)n_A(x,y)+g_B(\eta_s)n_B(x,y)\right] \tag*{(34)}

where gA(ηs)g_A(\eta_s) and gB(ηs)g_B(\eta_s) are given by [93]:

gA(ηs)=Θ(ηs−ηB,0)exp⁡[−(ηs−ηB,0)22σB,out2]+Θ(ηB,0−ηs)exp⁡[−(ηs−ηB,0)22σB,in2],(35)g_A(\eta_s)=\Theta(\eta_s-\eta_{B,0})\exp\left[-\frac{(\eta_s-\eta_{B,0})^2}{2\sigma_{B,\mathrm{out}}^2}\right]+\Theta(\eta_{B,0}-\eta_s)\exp\left[-\frac{(\eta_s-\eta_{B,0})^2}{2\sigma_{B,\mathrm{in}}^2}\right], \tag*{(35)}
gB(ηs)=Θ(ηs+ηB,0)exp⁡[−(ηs+ηB,0)22σB,in2]+Θ(−ηB,0−ηs)exp⁡[−(ηs+ηB,0)22σB,out2].(36)g_B(\eta_s)=\Theta(\eta_s+\eta_{B,0})\exp\left[-\frac{(\eta_s+\eta_{B,0})^2}{2\sigma_{B,\mathrm{in}}^2}\right]+\Theta(-\eta_{B,0}-\eta_s)\exp\left[-\frac{(\eta_s+\eta_{B,0})^2}{2\sigma_{B,\mathrm{out}}^2}\right]. \tag*{(36)}

The normalization constant Nn\mathcal{N}_n is chosen by demanding that

∫τ0 dx dy dηs n(τ0;x,y,ηs)=Npart⇒∫dηs n(τ0;x,y,ηs)=1τ0[nA(x,y)+nB(x,y)],(37)\int\tau_0\, dx\,dy\,d\eta_s\,n(\tau_0;x,y,\eta_s)=N_{\mathrm{part}}\quad\Rightarrow\quad\int d\eta_s\,n(\tau_0;x,y,\eta_s)=\frac{1}{\tau_0}\left[n_A(x,y)+n_B(x,y)\right], \tag*{(37)}

where NpartN_{\mathrm{part}} denotes the number of participants. The above condition gives

Nn=1τ02π1σB,in+σB,out.(38)\mathcal{N}_n=\frac{1}{\tau_0}\sqrt{\frac{2}{\pi}}\frac{1}{\sigma_{B,\mathrm{in}}+\sigma_{B,\mathrm{out}}}. \tag*{(38)}

The parameter values used in the initial condition model are provided in Table I.

sNN\sqrt{s_{NN}} (GeV)τ0\tau_0 (fm)ση\sigma_\etaσB,out\sigma_{B,\mathrm{out}}σB,in\sigma_{B,\mathrm{in}}η0\eta_0ηB,0\eta_{B,0}
2001.00.60.12.02.53.5

Table I. Parameters for the initial condition used in this work. The values are taken from [100].

Mean impact parameter determination

To estimate the mean impact parameter across various centralities, we simulate 7×1047\times10^4 nuclear collisions within an impact parameter range of 0–15 fm, following the procedure in [115]. First, we sample the positions (x,y,z)(x,y,z) of NAN_A nucleons for nucleus AA from a Woods-Saxon distribution using acceptance-rejection sampling. We compute the center of mass coordinates as follows:

xˉ=1NA∑i=1NAxi,yˉ=1NA∑i=1NAyi.\bar{x}=\frac{1}{N_A}\sum_{i=1}^{N_A}x_i,\qquad\bar{y}=\frac{1}{N_A}\sum_{i=1}^{N_A}y_i.

Next, we adjust the coordinates of the nucleons such that the center of mass coincides with the coordinate system’s origin. We repeat this procedure for nucleus BB. Afterward, we sample the impact parameter from a linear distribution,

P(b)∝bP(b) \propto b. We then shift the xx-coordinates of the nucleons in nucleus AA by −b/2-b/2 and those in nucleus BB by b/2b/2 to the right, so that the distance between the centers of the nuclei equals bb.

To determine binary collisions, we loop over the nucleons in nucleus AA and, for each, check whether there is a nucleon in nucleus BB within a transverse separation of σNN/π\sqrt{\sigma_{\mathrm{NN}}/\pi}. We count all such nucleons from BB for each nucleon in AA, which yields the number of binary collisions associated with each nucleon in AA. If there is at least one binary collision, the nucleon from AA is classified as “wounded”. We repeat this process for nucleus BB, then compute the total number of binary collisions (NcollN_{\mathrm{coll}}) and participants (NpartN_{\mathrm{part}}) by summing the contributions from both AA and BB.

In the two component Monte-Carlo Glauber model, the number of particles produced at mid-rapidity is given by

dNchdη∣η=0=npp[αhNcoll+(1−αh)Npart2],\left.\frac{dN_{\mathrm{ch}}}{d\eta}\right|_{\eta=0}=n_{pp}\left[\alpha_h N_{\mathrm{coll}}+(1-\alpha_h)\frac{N_{\mathrm{part}}}{2}\right],

where nppn_{pp} is the mean multiplicity per participant pair, and αh\alpha_h represents the fraction of particle production due to hard scattering processes. To reproduce the tail of the multiplicity distribution, we sample M=αhNcoll+(1−αh)Npart2M=\alpha_h N_{\mathrm{coll}}+(1-\alpha_h)\frac{N_{\mathrm{part}}}{2} random variables from a negative binomial distribution (NBD) with mean nppn_{pp} and variance ∼1k\sim\frac{1}{k} [117]. The sampled values from the NBD are summed to obtain the model prediction for charged particle multiplicity. We tune the parameters to fit the distribution from the experiment, as shown in Figure 11, with the parameter values provided in Table II.

Comparison of the charged particle multiplicity distribution from the two-component Monte-Carlo Glauber model

Figure 11. Comparison of the charged particle multiplicity distribution from the two-component Monte-Carlo Glauber model (solid red line) with STAR data [116] (green symbols).

sNN\sqrt{s_{\mathrm{NN}}}σNN\sigma_{\mathrm{NN}} (mb)nppn_{\mathrm{pp}}αh\alpha_hkk
20042.01.800.115.0

Table II. Parameters for impact parameter determination using the two-component Monte-Carlo Glauber model

We denote the charged particle multiplicity distribution (normalized to 1) as P(dNch/dη)P(dN_{\mathrm{ch}}/d\eta). If (dNch/dη)max⁡(dN_{\mathrm{ch}}/d\eta)_{\max} denotes the maximum multiplicity, then the centrality class 0−C1%0-C_1\% corresponds to the multiplicity range (dNch/dη)1(dN_{\mathrm{ch}}/d\eta)_1–(dNch/dη)max⁡(dN_{\mathrm{ch}}/d\eta)_{\max} such that

∫(dNch/dη)1(dNch/dη)max⁡P(N′) dN′=C1100.\int_{(dN_{\mathrm{ch}}/d\eta)_1}^{(dN_{\mathrm{ch}}/d\eta)_{\max}} P(N')\,dN'=\frac{C_1}{100}.

Similarly, multiplicity range (dNch/dη)2(dN_{\mathrm{ch}}/d\eta)_2–(dNch/dη)max⁡(dN_{\mathrm{ch}}/d\eta)_{\max} for the centrality class 0−C2%0-C_2\%, for C2>C1C_2>C_1, is

∫(dNch/dη)2(dNch/dη)max⁡P(N′) dN′=C2100,\int_{(dN_{\mathrm{ch}}/d\eta)_2}^{(dN_{\mathrm{ch}}/d\eta)_{\max}} P(N')\,dN'=\frac{C_2}{100},

so that the multiplicity range corresponding to C1−C2%C_1-C_2\% class is (dNch/dη)2(dN_{\mathrm{ch}}/d\eta)_2–(dNch/dη)1(dN_{\mathrm{ch}}/d\eta)_1 (note that (dNch/dη)1>(dNch/dη)2(dN_{\mathrm{ch}}/d\eta)_1>(dN_{\mathrm{ch}}/d\eta)_2). The mean impact parameter in C1−C2%C_1-C_2\% centrality class is computed by averaging the bb values of events with multiplicities between (dNch/dη)2(dN_{\mathrm{ch}}/d\eta)_2–(dNch/dη)1(dN_{\mathrm{ch}}/d\eta)_1. This exercise is repeated across all centrality classes, ultimately yielding the mean impact parameter as a function of centrality. The values of mean impact parameter for the different centrality classes considered in this work are listed in Table III.

Numerical implementation of spin equation of motion

The spin equations of motion in arbitrary coordinates can be expressed as:

DαSα,βγ=0,(39)D_\alpha S^{\alpha,\beta\gamma}=0, \tag*{(39)}

where DαD_\alpha represents the covariant derivative.

Expanding the covariant derivative using Christoffel symbols, the above equation becomes

∂αSα,βγ+ΓρααSρ,βγ+ΓαδβSα,δγ+ΓαωγSα,βω=0.(40)\partial_\alpha S^{\alpha,\beta\gamma}+\Gamma^\alpha_{\rho\alpha}S^{\rho,\beta\gamma}+\Gamma^\beta_{\alpha\delta}S^{\alpha,\delta\gamma}+\Gamma^\gamma_{\alpha\omega}S^{\alpha,\beta\omega}=0. \tag*{(40)}

Since the hydrodynamic code is written in the Milne coordinates (τ,x,y,ηs)(\tau,x,y,\eta_s), the only non-vanishing Christoffel symbols are Γτηsηs=Γηsτηs=1τ\Gamma^{\eta_s}_{\tau\eta_s}=\Gamma^{\eta_s}_{\eta_s\tau}=\frac{1}{\tau} and Γηsηsτ=τ\Gamma^\tau_{\eta_s\eta_s}=\tau. Hence, expressing the Eqs. (40) component-wise, we obtain:

∂τ(τSτ,τx)+∂x(τSx,τx)+∂y(τSy,τx)+∂ηs(τSηs,τx)+τ2Sηs,ηsx=0,(41)\partial_\tau(\tau S^{\tau,\tau x})+\partial_x(\tau S^{x,\tau x})+\partial_y(\tau S^{y,\tau x})+\partial_{\eta_s}(\tau S^{\eta_s,\tau x})+\tau^2 S^{\eta_s,\eta_s x}=0, \tag*{(41)}
∂τ(τSτ,τy)+∂x(τSx,τy)+∂y(τSy,τy)+∂ηs(τSηs,τy)+τ2Sηs,ηsy=0,(42)\partial_\tau(\tau S^{\tau,\tau y})+\partial_x(\tau S^{x,\tau y})+\partial_y(\tau S^{y,\tau y})+\partial_{\eta_s}(\tau S^{\eta_s,\tau y})+\tau^2 S^{\eta_s,\eta_s y}=0, \tag*{(42)}
∂τ(τSτ,τηs)+∂x(τSx,τηs)+∂y(τSy,τηs)+∂ηs(τSηs,τηs)+Sτ,τηs=0,(43)\partial_\tau(\tau S^{\tau,\tau\eta_s})+\partial_x(\tau S^{x,\tau\eta_s})+\partial_y(\tau S^{y,\tau\eta_s})+\partial_{\eta_s}(\tau S^{\eta_s,\tau\eta_s})+S^{\tau,\tau\eta_s}=0, \tag*{(43)}
∂τ(τSτ,xy)+∂x(τSx,xy)+∂y(τSy,xy)+∂ηs(τSηs,xy)=0,(44)\partial_\tau(\tau S^{\tau,xy})+\partial_x(\tau S^{x,xy})+\partial_y(\tau S^{y,xy})+\partial_{\eta_s}(\tau S^{\eta_s,xy})=0, \tag*{(44)}
∂τ(τSτ,xηs)+∂x(τSx,xηs)+∂y(τSy,xηs)+∂ηs(τSηs,xηs)+Sτ,xηs+Sηs,xτ=0,(45)\partial_\tau(\tau S^{\tau,x\eta_s})+\partial_x(\tau S^{x,x\eta_s})+\partial_y(\tau S^{y,x\eta_s})+\partial_{\eta_s}(\tau S^{\eta_s,x\eta_s})+S^{\tau,x\eta_s}+S^{\eta_s,x\tau}=0, \tag*{(45)}
∂τ(τSτ,yηs)+∂x(τSx,yηs)+∂y(τSy,yηs)+∂ηs(τSηs,yηs)+Sτ,yηs+Sηs,yτ=0.(46)\partial_\tau(\tau S^{\tau,y\eta_s})+\partial_x(\tau S^{x,y\eta_s})+\partial_y(\tau S^{y,y\eta_s})+\partial_{\eta_s}(\tau S^{\eta_s,y\eta_s})+S^{\tau,y\eta_s}+S^{\eta_s,y\tau}=0. \tag*{(46)}

Following methodology of Ref. [97], we define scaled spin current as

S~α,βγ=Sα,βγforα,β,γ≠ηs,\widetilde{S}^{\alpha,\beta\gamma}=S^{\alpha,\beta\gamma}\qquad\text{for}\qquad\alpha,\beta,\gamma\ne\eta_s,
S~α,βγ=τSα,βγif any index out of α,β,γ is ηs,\widetilde{S}^{\alpha,\beta\gamma}=\tau S^{\alpha,\beta\gamma}\qquad\text{if any index out of }\alpha,\beta,\gamma\text{ is }\eta_s,
S~α,βγ=τ2Sα,βγif any two indices out of α,β,γ is ηs.\widetilde{S}^{\alpha,\beta\gamma}=\tau^2 S^{\alpha,\beta\gamma}\qquad\text{if any two indices out of }\alpha,\beta,\gamma\text{ is }\eta_s.

Defining ∂~ν={∂∂τ,∂∂x,∂∂y,1τ∂∂ηs}\widetilde{\partial}_\nu=\left\{\frac{\partial}{\partial\tau},\frac{\partial}{\partial x},\frac{\partial}{\partial y},\frac{1}{\tau}\frac{\partial}{\partial\eta_s}\right\}, we may write Eqs. (41)–(46) as

∂~τ(τS~τ,τx)+∂~x(τS~x,τx)+∂~y(τS~y,τx)+∂~ηs(τS~ηs,τx)=S~ηs,xηs,(47)\widetilde{\partial}_\tau(\tau\widetilde{S}^{\tau,\tau x})+\widetilde{\partial}_x(\tau\widetilde{S}^{x,\tau x})+\widetilde{\partial}_y(\tau\widetilde{S}^{y,\tau x})+\widetilde{\partial}_{\eta_s}(\tau\widetilde{S}^{\eta_s,\tau x})=\widetilde{S}^{\eta_s,x\eta_s}, \tag*{(47)}
∂~τ(τS~τ,τy)+∂~x(τS~x,τy)+∂~y(τS~y,τy)+∂~ηs(τS~ηs,τy)=S~ηs,yηs,(48)\widetilde{\partial}_\tau(\tau\widetilde{S}^{\tau,\tau y})+\widetilde{\partial}_x(\tau\widetilde{S}^{x,\tau y})+\widetilde{\partial}_y(\tau\widetilde{S}^{y,\tau y})+\widetilde{\partial}_{\eta_s}(\tau\widetilde{S}^{\eta_s,\tau y})=\widetilde{S}^{\eta_s,y\eta_s}, \tag*{(48)}
∂~τ(τS~τ,τηs)+∂~x(τS~x,τηs)+∂~y(τS~y,τηs)+∂~ηs(τS~ηs,τηs)=0,(49)\widetilde{\partial}_\tau(\tau\widetilde{S}^{\tau,\tau\eta_s})+\widetilde{\partial}_x(\tau\widetilde{S}^{x,\tau\eta_s})+\widetilde{\partial}_y(\tau\widetilde{S}^{y,\tau\eta_s})+\widetilde{\partial}_{\eta_s}(\tau\widetilde{S}^{\eta_s,\tau\eta_s})=0, \tag*{(49)}
∂~τ(τS~τ,xy)+∂~x(τS~x,xy)+∂~y(τS~y,xy)+∂~ηs(τS~ηs,xy)=0,(50)\widetilde{\partial}_\tau(\tau\widetilde{S}^{\tau,xy})+\widetilde{\partial}_x(\tau\widetilde{S}^{x,xy})+\widetilde{\partial}_y(\tau\widetilde{S}^{y,xy})+\widetilde{\partial}_{\eta_s}(\tau\widetilde{S}^{\eta_s,xy})=0, \tag*{(50)}
∂~τ(τS~τ,xηs)+∂~x(τS~x,xηs)+∂~y(τS~y,xηs)+∂~ηs(τS~ηs,xηs)=S~ηs,τx,(51)\widetilde{\partial}_\tau(\tau\widetilde{S}^{\tau,x\eta_s})+\widetilde{\partial}_x(\tau\widetilde{S}^{x,x\eta_s})+\widetilde{\partial}_y(\tau\widetilde{S}^{y,x\eta_s})+\widetilde{\partial}_{\eta_s}(\tau\widetilde{S}^{\eta_s,x\eta_s})=\widetilde{S}^{\eta_s,\tau x}, \tag*{(51)}
∂~τ(τS~τ,yηs)+∂~x(τS~x,yηs)+∂~y(τS~y,yηs)+∂~ηs(τS~ηs,yηs)=S~ηs,τy.(52)\widetilde{\partial}_\tau(\tau\widetilde{S}^{\tau,y\eta_s})+\widetilde{\partial}_x(\tau\widetilde{S}^{x,y\eta_s})+\widetilde{\partial}_y(\tau\widetilde{S}^{y,y\eta_s})+\widetilde{\partial}_{\eta_s}(\tau\widetilde{S}^{\eta_s,y\eta_s})=\widetilde{S}^{\eta_s,\tau y}. \tag*{(52)}

It should be noted that, apart from the source terms, the set of equations (C9)–(C14) is in the conservative form

∂~τW+∂~xF+∂~yG+∂~ηsH=0.(53)\widetilde{\partial}_\tau W+\widetilde{\partial}_x F+\widetilde{\partial}_y G+\widetilde{\partial}_{\eta_s}H=0. \tag*{(53)}

One can employ any algorithm from the class of Godunov-type algorithms which are designed to solve such equations. Following Ref. [97], we choose the relativistic Harten-Lax-van Leer-Einfeldt (HLLE) approximate Riemann solver [118, 119]. The numerical solution of Eqs. (47)–(52) proceeds with the following steps:

  1. We initialize the background hydrodynamics using the initial condition model and the equation of state described in § 2 A and Appendix A. If the initial time for spin evolution, τ0s\tau_0^s, is chosen to coincide with the initial time for background hydrodynamics, τ0\tau_0, then the background hydrodynamics is evolved separately for three timesteps, and the data for βμ=uμT\beta^\mu= \frac{u^\mu}{T} is stored in a data file. These data are then used to compute the rank-2 tensor DμβνD_\mu\beta_\nu, from which the thermal vorticity ϖμν\varpi_{\mu\nu} and thermal shear ξμν\xi_{\mu\nu} are calculated. We use a forward finite difference scheme for time derivatives and a central finite difference scheme for spatial derivatives.

  1. If τ0s\tau_0^s is different from the initial time for background hydrodynamics, we evolve the background and spin hydrodynamics simultaneously, with initial spin current as 00. Due to the ideal nature of spin equations, the spin current will be zero during this evolution. At τ=τ0s\tau= \tau_0^s, we compute ϖμν\varpi_{\mu\nu} and ξμν\xi_{\mu\nu} and put the spin polarization tensor, ω\omega, to either thermal vorticity or one of the initial condition models in Eqs. (16) and (17).

  1. We solve the equations on a three dimensional grid Nx×Ny×Nηs=201×201×121N_x \times N_y \times N_{\eta_s} = 201 \times201 \times121 in the following range:

(xmin⁡,xmax⁡)=(−20,20),(ymin⁡,ymax⁡)=(−20,20),(ηs,min⁡,ηs,max⁡)=(−12,12).(x_{\min},x_{\max}) = (-20,20), \quad(y_{\min},y_{\max}) = (-20,20), \quad(\eta_{s,\min},\eta_{s,\max}) = (-12,12).

Each grid point with coordinates (xi,yj,zk)(x_i,y_j,z_k) is treated as a fluid cell identified as (i,j,k)(i,j,k). The interface between fluid cells at (i,j,k)(i,j,k) and (i+1,j,k)(i+1,j,k) is denoted by (i+12,j,k)(i+\frac{1}{2},j,k). We choose Δτ=0.05\Delta\tau= 0.05, which satisfies the Courant-Friedrichs-Lewy (CFL) criterion for the stability of the algorithm.

  1. The conserved quantity of a fluid cell is updated by computing the net flux entering the cell independently from each of the three directions. For example, this translates to the following prescription for Eq. (53):

Wijkn+1=Wijkn+ΔtΔx(Fi−12,jkn−Fi+12,jkn)+ΔtΔy(Gi,j−12,kn−Gi,j+12,kn)+ΔtτΔηs(Hij,k−12n−Hij,k+12n).(54)W_{ijk}^{n+1} = W_{ijk}^{n} + \frac{\Delta t}{\Delta x}\left(F_{i-\frac{1}{2},jk}^{n} - F_{i+\frac{1}{2},jk}^{n}\right) + \frac{\Delta t}{\Delta y}\left(G_{i,j-\frac{1}{2},k}^{n} - G_{i,j+\frac{1}{2},k}^{n}\right) + \frac{\Delta t}{\tau\Delta\eta_s}\left(H_{ij,k-\frac{1}{2}}^{n} - H_{ij,k+\frac{1}{2}}^{n}\right). \tag*{(54)}

In case of spin hydrodynamics, the conserved quantities are S=(Sτ,τ,Sτ,τx,Sτ,τy,Sτ,τηs,Sτ,xy,Sτ,xηs,Sτ,yηs)\mathbf{S} = (S^{\tau,\tau}, S^{\tau,\tau x}, S^{\tau,\tau y}, S^{\tau,\tau\eta_s}, S^{\tau,xy}, S^{\tau,x\eta_s}, S^{\tau,y\eta_s}).

  1. The numerical fluxes in the previous step are computed from the approximate solution of the local Riemann problem, according to HLLE prescription, at the interface between two fluid cells in each direction. The Riemann problem arises due to the reconstruction of the conserved quantities at the left and right boundaries of a fluid cell using the minmod slope limiter. For example, along the xx-direction after reconstruction, if Wi+12,jkW_{i+\frac{1}{2},jk} denotes the value at the right boundary of the fluid cell at (i,j,k)(i,j,k), and W(i+1)−12,jkW_{(i+1)-\frac{1}{2},jk} denotes the value at the left boundary of the fluid cell at (i+1,j,k)(i+1,j,k), then the flux that leaves the fluid cell at (i,j,k)(i,j,k) and enters the fluid cell at (i+1,j,k)(i+1,j,k) is given by

Fi+12,jk=brF(Wi+12,jk)−blF(W(i+1)−12,jk)+blbr(W(i+1)−12,jk−Wi+12,jk)br−bl,F_{i+\frac{1}{2},jk} = \frac{b_rF(W_{i+\frac{1}{2},jk}) - b_lF(W_{(i+1)-\frac{1}{2},jk}) + b_lb_r\left(W_{(i+1)-\frac{1}{2},jk} - W_{i+\frac{1}{2},jk}\right)}{b_r-b_l},

where blb_l and brb_r are signal velocities, to be described in Appendix D.

  1. The computation of fluxes and source terms at each timestep requires a transformation from conservative quantities (Sτ,μν)(S^{\tau,\mu\nu}) to primitive quantities (ωμν)(\omega^{\mu\nu}). Due to the choice of the spin current, its components can be mapped linearly to the primitive quantities Ω=(ωτx,ωτy,ωτηs,ωxy,ωxηs,ωyηs)\boldsymbol{\Omega} = (\omega^{\tau x},\omega^{\tau y},\omega^{\tau\eta_s},\omega^{xy},\omega^{x\eta_s},\omega^{y\eta_s}), for example, the transformation between Sτ,μνS^{\tau,\mu\nu} and ωμν\omega^{\mu\nu} can be written as

S=AΩ,\mathbf{S} = \mathbf{A}\boldsymbol{\Omega},

where A\mathbf{A} is a 6×66 \times6 matrix given by (after omitting explicit τ\tau)

A16=A25=A34=A43=A52=A61=0,A11=(A1−A3)uτ+A22uτ[(ux)2−(uτ)2],A22=(A1−A3)uτ+A22uτ[(uy)2−(uτ)2],A33=(A1−A3)uτ+A22uτ[(uηs)2−(uτ)2],A44=A1uτ+A22uτ[(ux)2+(uy)2],A55=A1uτ+A22uτ[(ux)2+(uηs)2],A66=A1uτ+A22uτ[(uy)2+(uηs)2],A12=A21=A56=A65=A22uτuxuy,A13=A31=−A46=−A64=A22uτuxuηs,A14=−A41=A63=−A36=−A32uy−A22(uτ)2uy,A15=−A51=A26=−A62=−A32uηs−A22(uτ)2uηs,A23=−A32=A45=−A54=A22uτuyuηs,A24=−A42=A35=−A53=A32ux+A22(uτ)2ux.\begin{aligned} &A_{16}=A_{25}=A_{34}=A_{43}=A_{52}=A_{61}=0,\qquad A_{11}=(\mathcal{A}_1-\mathcal{A}_3)u^\tau+\frac{\mathcal{A}_2}{2}u^\tau\left[(u^x)^2-(u^\tau)^2\right],\\ &A_{22}=(\mathcal{A}_1-\mathcal{A}_3)u^\tau+\frac{\mathcal{A}_2}{2}u^\tau\left[(u^y)^2-(u^\tau)^2\right],\qquad A_{33}=(\mathcal{A}_1-\mathcal{A}_3)u^\tau+\frac{\mathcal{A}_2}{2}u^\tau\left[(u^{\eta_s})^2-(u^\tau)^2\right],\\ &A_{44}=\mathcal{A}_1u^\tau+\frac{\mathcal{A}_2}{2}u^\tau\left[(u^x)^2+(u^y)^2\right],\qquad A_{55}=\mathcal{A}_1u^\tau+\frac{\mathcal{A}_2}{2}u^\tau\left[(u^x)^2+(u^{\eta_s})^2\right],\\ &A_{66}=\mathcal{A}_1u^\tau+\frac{\mathcal{A}_2}{2}u^\tau\left[(u^y)^2+(u^{\eta_s})^2\right],\\ &A_{12}=A_{21}=A_{56}=A_{65}=\frac{\mathcal{A}_2}{2}u^\tau u^x u^y,\qquad A_{13}=A_{31}=-A_{46}=-A_{64}=\frac{\mathcal{A}_2}{2}u^\tau u^x u^{\eta_s},\\ &A_{14}=-A_{41}=A_{63}=-A_{36}=-\frac{\mathcal{A}_3}{2}u^y-\frac{\mathcal{A}_2}{2}(u^\tau)^2u^y,\qquad A_{15}=-A_{51}=A_{26}=-A_{62}=-\frac{\mathcal{A}_3}{2}u^{\eta_s}-\frac{\mathcal{A}_2}{2}(u^\tau)^2u^{\eta_s},\\ &A_{23}=-A_{32}=A_{45}=-A_{54}=\frac{\mathcal{A}_2}{2}u^\tau u^y u^{\eta_s},\qquad A_{24}=-A_{42}=A_{35}=-A_{53}=\frac{\mathcal{A}_3}{2}u^x+\frac{\mathcal{A}_2}{2}(u^\tau)^2u^x. \end{aligned}
  1. The primitive quantities are then obtained by inverting the matrix equation in the above step i.e.

Ω=A−1S.\Omega=\mathbf{A}^{-1}\mathbf{S}.

The values of primitive quantities get very large for fluid cells towards the boundaries due to small temperatures. To keep this under check, we compute Sτ,μνωμν\sqrt{S^{\tau,\mu\nu}\omega_{\mu\nu}} after transformation from conserved quantities and whenever its value increases a certain threshold, we re-scale the primitive quantities. We have checked that our results are insensitive towards such a re-scaling, as the fraction of cells whose values gets larger is very small.

  1. Consider the equation

∂~τW+∂~xF+∂~yG+∂~ηsH=X.\widetilde{\partial}_{\tau}W+\widetilde{\partial}_{x}F+\widetilde{\partial}_{y}G+\widetilde{\partial}_{\eta_s}H=X.

To handle the source terms, we first evolve WnW^n according to Eq. (54) to obtain W∗n+1W^{*n+1}. The final value of WW is obtained as

Wijkn+1=Wijk∗n+1+ΔτXijkn.W_{ijk}^{n+1}=W_{ijk}^{*n+1}+\Delta\tau X_{ijk}^n.

Signal velocity

One of the inputs needed to solve the equations of relativistic hydrodynamics using the relativistic HLLE algorithm is the signal velocity, which gives the speed at which information propagates from a point. Signal velocities are often estimated from the eigenvalues of the Jacobian matrix of the flux function [120] at the interface between two fluid cells. If uμ=(γ,γvx,γvy,γvz)u^\mu= (\gamma,\gamma v_x,\gamma v_y,\gamma v_z) denotes the four-velocity of a relativistic uncharged fluid with γ=11−vx2−vy2−vz2\gamma= \frac{1}{\sqrt{1-v_x^2-v_y^2-v_z^2}} being the Lorentz factor, then the state vector is

w⃗=(ϵ,vx,vy,vz),\vec{w} = (\epsilon,v_x,v_y,v_z),

and the fluxes are:

F0=[(ϵ+Peq)γ2−Peq,(ϵ+Peq)γ2vx,(ϵ+Peq)γ2vy,(ϵ+Peq)γ2vz],F^0 = \left[(\epsilon+P_{\rm eq})\gamma^2-P_{\rm eq},(\epsilon+P_{\rm eq})\gamma^2v_x,(\epsilon+P_{\rm eq})\gamma^2v_y,(\epsilon+P_{\rm eq})\gamma^2v_z\right],
F1=[(ϵ+Peq)γ2vx,(ϵ+Peq)γ2vx2+Peq,(ϵ+Peq)γ2vxvy,(ϵ+Peq)γ2vxvz],F^1 = \left[(\epsilon+P_{\rm eq})\gamma^2v_x,(\epsilon+P_{\rm eq})\gamma^2v_x^2+P_{\rm eq},(\epsilon+P_{\rm eq})\gamma^2v_xv_y,(\epsilon+P_{\rm eq})\gamma^2v_xv_z\right],
F2=[(ϵ+Peq)γ2vy,(ϵ+Peq)γ2vxvy,(ϵ+Peq)γ2vy2+Peq,(ϵ+Peq)γ2vyvz],F^2 = \left[(\epsilon+P_{\rm eq})\gamma^2v_y,(\epsilon+P_{\rm eq})\gamma^2v_xv_y,(\epsilon+P_{\rm eq})\gamma^2v_y^2+P_{\rm eq},(\epsilon+P_{\rm eq})\gamma^2v_yv_z\right],
F3=[(ϵ+Peq)γ2vz,(ϵ+Peq)γ2vxvz,(ϵ+Peq)γ2vyvz,(ϵ+Peq)γ2vz2+Peq].F^3 = \left[(\epsilon+P_{\rm eq})\gamma^2v_z,(\epsilon+P_{\rm eq})\gamma^2v_xv_z,(\epsilon+P_{\rm eq})\gamma^2v_yv_z,(\epsilon+P_{\rm eq})\gamma^2v_z^2+P_{\rm eq}\right].

The Jacobian matrix, JαJ^\alpha, is defined as

Jα=∂Fα∂w⃗.J^\alpha= \frac{\partial F^\alpha}{\partial\vec{w}}.

The hyperbolic condition for conservation equations is given by the following expression

det⁡(J1−λJ0)=0.\det(J^1-\lambda J^0)=0.

Here λ\lambda denotes the characteristic speeds of different waves. For a fluid flowing along xx-direction, i.e., vy=vz=0v_y=v_z=0 and vx=v≠0v_x=v\ne0, the above equation gives

λ=v,v,v+cs1+vcs,v−cs1−vcs.(55)\lambda=v,v,\frac{v+c_s}{1+vc_s},\frac{v-c_s}{1-vc_s}. \tag*{(55)}

In the numerical solution of background hydrodynamics, a convenient choice for the right (brb_r) and left (blb_l) signal velocities at an interface is

br=max⁡(0,vˉ+cˉs1+vˉcˉs,v+cs1+vcs),bl=min⁡(0,vˉ−cˉs1−vˉcˉs,v−cs1−vcs),(56)b_r=\max\left(0,\frac{\bar{v}+\bar{c}_s}{1+\bar{v}\bar{c}_s},\frac{v+c_s}{1+vc_s}\right),\quad b_l=\min\left(0,\frac{\bar{v}-\bar{c}_s}{1-\bar{v}\bar{c}_s},\frac{v-c_s}{1-vc_s}\right), \tag*{(56)}

where vˉ\bar{v} is energy-weighted average of the left and right velocities at the interface, and similarly for cˉs\bar{c}_s [121]. As signal velocities are estimates for speeds of wave propagation, one can also consider constant cs=13c_s=\frac{1}{\sqrt{3}} for simplicity [97].

A similar analysis is possible for spin hydrodynamics. The state vector is

w⃗=(ωtx,ωty,ωtz,ωxy,ωxz,ωyz)\vec{w}=(\omega^{tx},\omega^{ty},\omega^{tz},\omega^{xy},\omega^{xz},\omega^{yz})

and the fluxes are

Fα=(Sα,tx,Sα,ty,Sα,tz,Sα,xy,Sα,xz,Sα,yz).F^\alpha=\left(S^{\alpha,tx},S^{\alpha,ty},S^{\alpha,tz},S^{\alpha,xy},S^{\alpha,xz},S^{\alpha,yz}\right).

We compute the matrices JαJ^\alpha and solve the equation

det⁡(J1−λJ0)=0,\det(J^1-\lambda J^0)=0,

to obtain characteristic speeds of waves for the case of a fluid flowing along the xx-direction as

λ=v,v,v(1−α~2)+α~(1−v2)1−v2α~2,v(1−α~2)−α~(1−v2)1−v2α~2≡λ1,λ2,λ3,λ4,(57)\lambda=v,v,\frac{v(1-\tilde{\alpha}^2)+\tilde{\alpha}(1-v^2)}{1-v^2\tilde{\alpha}^2},\frac{v(1-\tilde{\alpha}^2)-\tilde{\alpha}(1-v^2)}{1-v^2\tilde{\alpha}^2}\equiv\lambda_1,\lambda_2,\lambda_3,\lambda_4, \tag*{(57)}

where α~\tilde{\alpha} is given by

α~=∣A3∣2∣S1∣S2.\tilde{\alpha}=\frac{\lvert A_3\rvert}{2\sqrt{\lvert S_1\rvert S_2}}.

In the above expression, S1=A1−A22−A3S_1=A_1-\frac{A_2}{2}-A_3 and S2=A1S_2=A_1. The characteristic speeds depend not only on the fluid velocity but also on the temperature and mass of the particles. We plot the third eigenvalue, λ3\lambda_3, of Eq. (57) in Figure 12 for T=200T=200 MeV and compare it with the corresponding quantity from background hydrodynamics for constant speed of sound, cs2=13c_s^2=\frac{1}{3}. Since the characteristic speed for spin hydrodynamics is always smaller than that of background hydrodynamics in Eq. (56), irrespective of temperature, mass and fluid velocity, we use the same signal velocity as in background hydrodynamics. This choice is made not only for simplicity but also because we use the same grid for solving both spin and background hydrodynamics. The signal velocities in Eq. (56), along with the CFL criterion, already ensure that waves from two neighboring fluid cells do not overlap, thus maintaining algorithm stability.

Third eigenvalue for spin hydrodynamics

Figure 12. Third eigenvalue for spin hydrodynamics (see text for description) plotted as a function of fluid velocity for different masses (solid red and dash-dotted blue lines) when temperature TT is fixed at 200 MeV. The corresponding speed of background hydrodynamics (green dashed line) for constant cs=13c_s=\frac{1}{\sqrt{3}} is also shown for comparison.

Test of the code

Since no analytical solution is available for relativistic spin hydrodynamics, we perform an indirect test of the code. Assuming boost invariance, we have uτ=1u^\tau= 1, ux=uy=uηs=0u^x = u^y = u^{\eta_s} = 0. Under these conditions, Eqs. (47)–(52) reduce to a simplified form:

τS1ω~˙τx+(S1+τS1′T˙+A32)ω~τx=0,τS1ω~˙τy+(S1+τS1′T˙+A32)ω~τy=0,τS1ω~˙τηs+(S1+τS1′T˙)ω~τηs=0,τA1ω~˙xy+(A1+τA1′T˙)ω~xy=0,τA1ω~˙xηs+(A1+τA1′T˙−A32)ω~xηs=0,τA1ω~˙yηs+(A1+τA1′T˙−A32)ω~yηs=0.\begin{aligned} \tau S_1 \dot{\tilde{\omega}}^{\tau x}+\left(S_1+\tau S'_1\dot{T}+\frac{\mathcal{A}_3}{2}\right)\tilde{\omega}^{\tau x}&=0,\\ \tau S_1 \dot{\tilde{\omega}}^{\tau y}+\left(S_1+\tau S'_1\dot{T}+\frac{\mathcal{A}_3}{2}\right)\tilde{\omega}^{\tau y}&=0,\\ \tau S_1 \dot{\tilde{\omega}}^{\tau\eta_s}+\left(S_1+\tau S'_1\dot{T}\right)\tilde{\omega}^{\tau\eta_s}&=0,\\ \tau\mathcal{A}_1\dot{\tilde{\omega}}^{xy}+\left(\mathcal{A}_1+\tau\mathcal{A}'_1\dot{T}\right)\tilde{\omega}^{xy}&=0,\\ \tau\mathcal{A}_1\dot{\tilde{\omega}}^{x\eta_s}+\left(\mathcal{A}_1+\tau\mathcal{A}'_1\dot{T}-\frac{\mathcal{A}_3}{2}\right)\tilde{\omega}^{x\eta_s}&=0,\\ \tau\mathcal{A}_1\dot{\tilde{\omega}}^{y\eta_s}+\left(\mathcal{A}_1+\tau\mathcal{A}'_1\dot{T}-\frac{\mathcal{A}_3}{2}\right)\tilde{\omega}^{y\eta_s}&=0. \end{aligned}

The primes in above equations denote derivative with respect to TT. If we further assume Bjorken expansion for the background, the temperature evolution is given by T˙=−T3τ\dot{T}=-\frac{T}{3\tau}. These equations can be solved using MATHEMATICA’s NDSolve function. We compare the numerical solution from our code with MATHEMATICA solution in Figure 13, and the two solutions show good agreement.

Comparison of code output with MATHEMATICA solution for different components of the spin polarization tensor

Figure 13. Comparison of code output with MATHEMATICA solution for different components of the spin polarization tensor.

References

  1. [1]G. Agakishiev et al. (HADES), Eur. Phys. J. A 50, 81 (2014), arXiv:1404.3014 [nucl-ex].DOI
  2. [2]L. Adamczyk et al. (STAR), Nature 548, 62 (2017), arXiv:1701.06657 [nucl-ex].arXiv
  3. [3]J. Adam et al. (STAR), Phys. Rev. C 98, 014910 (2018), arXiv:1805.04400 [nucl-ex].DOI
  4. [4]J. Adam et al. (STAR), Phys. Rev. Lett. 123, 132301 (2019), arXiv:1905.11917 [nucl-ex].DOI
  5. [5]S. Acharya et al. (ALICE), Phys. Rev. Lett. 125, 012301 (2020), arXiv:1910.14408 [nucl-ex].arXiv
  6. [6]M. S. Abdallah et al. (STAR), Phys. Rev. C 104, L061901 (2021), arXiv:2108.00044 [nucl-ex].DOI
  7. [7]S. Acharya et al. (ALICE), Phys. Rev. C 101, 044611 (2020), [Erratum: Phys.Rev.C 105, 029902 (2022)], arXiv:1909.01281 [nucl-ex].arXiv
  8. [8]T. Niida and S. A. Voloshin, Int. J. Mod. Phys. E 33, 2430010 (2024), arXiv:2404.11042 [nucl-ex].DOI
  9. [9]F. Becattini, M. Buzzegoli, T. Niida, S. Pu, A.-H. Tang, and Q. Wang, International Journal of Modern Physics E 33, 2430006 (2024), arXiv:2402.04540 [nucl-th].arxiv.org/abs/2402.04540
  10. [10]Y. Bondar and W. Florkowski, Acta Phys. Polon. B 55, 9 (2024), arXiv:2408.11050 [hep-ph].DOI
  11. [11]A. Sorensen et al., Prog. Part. Nucl. Phys. 134, 104080 (2024), arXiv:2301.13253 [nucl-th].DOI
  12. [12]S. Y. F. Liu and Y. Yin, JHEP 07, 188, arXiv:2103.09200 [hep-ph].DOI
  13. [13]B. Fu, S. Y. F. Liu, L. Pang, H. Song, and Y. Yin, Phys. Rev. Lett. 127, 142301 (2021), arXiv:2103.10403 [hep-ph].DOI
  14. [14]F. Becattini, M. Buzzegoli, and A. Palermo, Phys. Lett. B 820, 136519 (2021), arXiv:2103.10917 [nucl-th].DOI
  15. [15]F. Becattini, M. Buzzegoli, G. Inghirami, I. Karpenko, and A. Palermo, Phys. Rev. Lett. 127, 272302 (2021), arXiv:2103.14621 [nucl-th].arXiv
  16. [16]S. Alzhrani, S. Ryu, and C. Shen, Phys. Rev. C 106, 014905 (2022), arXiv:2203.15718 [nucl-th].DOI
  17. [17]F. Becattini and L. Tinti, Annals Phys. 325, 1566 (2010), arXiv:0911.0864 [gr-qc].DOI
  18. [18]F. Becattini, Phys. Part. Nucl. Lett. 8, 801 (2011).
  19. [19]F. Becattini, V. Chandra, L. Del Zanna, and E. Grossi, Annals Phys. 338, 32 (2013), arXiv:1303.3431 [nucl-th].DOI
  20. [20]F. Becattini, L. Csernai, and D. J. Wang, Phys. Rev. C 88, 034905 (2013), [Erratum: Phys.Rev.C 93, 069901 (2016)], arXiv:1304.4427 [nucl-th].DOI
  21. [21]I. Karpenko and F. Becattini, Eur. Phys. J. C 77, 213 (2017), arXiv:1610.04717 [nucl-th].DOI
  22. [22]F. Becattini, I. Karpenko, M. Lisa, I. Upsal, and S. Voloshin, Phys. Rev. C 95, 054902 (2017), arXiv:1610.02506 [nucl-th].DOI
  23. [23]F. Becattini and I. Karpenko, Phys. Rev. Lett. 120, 012302 (2018), arXiv:1707.07984 [nucl-th].DOI
  24. [24]A. Palermo, E. Grossi, I. Karpenko, and F. Becattini, Eur. Phys. J. C 84, 920 (2024), arXiv:2404.14295 [nucl-th].arXiv
  25. [25]W. Florkowski, B. Friman, A. Jaiswal, and E. Speranza, Phys. Rev. C 97, 041901 (2018), arXiv:1705.00587 [nucl-th].DOI
  26. [26]W. Florkowski, B. Friman, A. Jaiswal, R. Ryblewski, and E. Speranza, Phys. Rev. D 97, 116017 (2018), arXiv:1712.07676 [nucl-th].DOI
  27. [27]W. Florkowski, A. Kumar, and R. Ryblewski, Phys. Rev. C 98, 044906 (2018), arXiv:1806.02616 [hep-ph].DOI
  28. [28]W. Florkowski, A. Kumar, and R. Ryblewski, Prog. Part. Nucl. Phys. 108, 103709 (2019), arXiv:1811.04409 [nucl-th].DOI
  29. [29]D. Montenegro and G. Torrieri, Phys. Rev. D 100, 056011 (2019), arXiv:1807.02796 [hep-th].DOI
  30. [30]A. Kumar, in 13th Workshop on Particle Correlations and Femtoscopy (2018) arXiv:1810.05364 [nucl-th].DOI
  31. [31]A. Kumar, PoS Confinement2018, 281 (2018), arXiv:1812.00217 [nucl-th].DOI
  32. [32]F. Becattini, W. Florkowski, and E. Speranza, Phys. Lett. B 789, 419 (2019), arXiv:1807.10994 [hep-th].DOI
  33. [33]W. Florkowski, A. Kumar, R. Ryblewski, and R. Singh, Phys. Rev. C 99, 044910 (2019), arXiv:1901.09655 [hep-ph].DOI
  34. [34]D. Montenegro and G. Torrieri, Phys. Rev. D 102, 036007 (2020), arXiv:2004.10195 [hep-th].DOI
  35. [35]M. Garbiso and M. Kaminski, JHEP 12, 112, arXiv:2007.04345 [hep-th].DOI
  36. [36]S. Bhadury, W. Florkowski, A. Jaiswal, A. Kumar, and R. Ryblewski, Phys. Lett. B 814, 136096 (2021), arXiv:2002.03937 [hep-ph].DOI
  37. [37]S. Bhadury, W. Florkowski, A. Jaiswal, A. Kumar, and R. Ryblewski, Phys. Rev. D 103, 014030 (2021), arXiv:2008.10976 [nucl-th].DOI
  38. [38]S. Bhadury, W. Florkowski, A. Jaiswal, A. Kumar, and R. Ryblewski, Phys. Rev. Lett. 129, 192301 (2022), arXiv:2204.01357 [nucl-th].DOI
  39. [39]S. Bhadury, A. Das, W. Florkowski, G. K. K., and R. Ryblewski, Phys. Lett. B 849, 138464 (2024), arXiv:2307.12436 [hep-ph].
  40. [40]R. Singh, M. Shokri, and S. M. A. T. Mehr, Nucl. Phys. A 1035, 122656 (2023), arXiv:2211.15047 [hep-ph].
  41. [41]N. Weickgenannt, X.-L. Sheng, E. Speranza, Q. Wang, and D. H. Rischke, Phys. Rev. D 100, 056018 (2019), arXiv:1902.06513 [hep-ph].DOI
  42. [42]N. Weickgenannt, E. Speranza, X.-L. Sheng, Q. Wang, and D. H. Rischke, Phys. Rev. D 104, 016022 (2021), arXiv:2103.04896 [nucl-th].DOI
  43. [43]N. Weickgenannt, E. Speranza, X.-L. Sheng, Q. Wang, and D. H. Rischke, Phys. Rev. Lett. 127, 052301 (2021),
  44. [44]D. Wagner, N. Weickgenannt, and D. H. Rischke, Phys. Rev. D 106, 116021 (2022), arXiv:2210.06187 [nucl-th].DOI
  45. [45]N. Weickgenannt, D. Wagner, E. Speranza, and D. H. Rischke, Phys. Rev. D 106, 096014 (2022), arXiv:2203.04766 [nucl-th].DOI
  46. [46]N. Weickgenannt and J.-P. Blaizot, Phys. Rev. D 109, 056012 (2024), arXiv:2311.15817 [hep-ph].arXiv
  47. [47]N. Weickgenannt and J.-P. Blaizot, Phys. Rev. D 109, 056019 (2024), arXiv:2312.05917 [hep-ph].arXiv
  48. [48]N. Weickgenannt and J.-P. Blaizot, (2024), arXiv:2409.11045 [hep-ph].arXiv
  49. [49]D. Wagner, M. Shokri, and D. H. Rischke, Phys. Rev. Res. 6, 043103 (2024), arXiv:2405.00533 [nucl-th].DOI
  50. [50]H.-H. Peng, J.-J. Zhang, X.-L. Sheng, and Q. Wang, Chin. Phys. Lett. 38, 116701 (2021), arXiv:2107.00448 [hep-th].DOI
  51. [51]C. Cartwright, M. G. Amano, M. Kaminski, J. Noronha, and E. Speranza, Phys. Rev. D 108, 046014 (2023), arXiv:2112.10781 [hep-th].arXiv
  52. [52]J. Hu, Phys. Rev. D 105, 076009 (2022), arXiv:2111.03571 [hep-ph].DOI
  53. [53]S. Li, M. A. Stephanov, and H.-U. Yee, Phys. Rev. Lett. 127, 082302 (2021), arXiv:2101.12318 [hep-th].
  54. [54]S. Shi, C. Gale, and S. Jeon, Phys. Rev. C 103, 044906 (2021), arXiv:2008.08618 [nucl-th].DOI
  55. [55]K. Hattori, M. Hongo, X.-G. Huang, M. Matsuo, and H. Taya, Phys. Lett. B 795, 100 (2019), arXiv:1901.06615 [hep-th].DOI
  56. [56]K. Fukushima and S. Pu, Phys. Lett. B 817, 136346 (2021), arXiv:2010.01608 [hep-ph].DOI
  57. [57]A. Daher, A. Das, W. Florkowski, and R. Ryblewski, Phys. Rev. C 108, 024902 (2023), arXiv:2202.12609 [nucl-th].DOI
  58. [58]A. Daher, A. Das, and R. Ryblewski, Phys. Rev. D 107, 054043 (2023), arXiv:2209.10460 [nucl-th].DOI
  59. [59]G. Sarwar, M. Hasanuzzaman, J. R. Bhatt, H. Mishra, and J.-e. Alam, Phys. Rev. D 107, 054031 (2023), arXiv:2209.08652 [nucl-th].
  60. [60]C. Yi, S. Pu, J.-H. Gao, and D.-L. Yang, Phys. Rev. C 105, 044911 (2022), arXiv:2112.15531 [hep-ph].DOI
  61. [61]D.-L. Wang, S. Fang, and S. Pu, Phys. Rev. D 104, 114043 (2021), arXiv:2127.11276 [nucl-th].
  62. [62]D.-L. Wang, X.-Q. Xie, S. Fang, and S. Pu, Phys. Rev. D 105, 114050 (2022), arXiv:2112.15535 [nucl-th].DOI
  63. [63]R. Biswas, A. Daher, A. Das, W. Florkowski, and R. Ryblewski, Phys. Rev. D 107, 094022 (2023), arXiv:2211.02934 [nucl-th].DOI
  64. [64]Z. Cao, K. Hattori, M. Hongo, X.-G. Huang, and H. Taya, PTEP 2022, 071D01 (2022), arXiv:2205.08051 [hep-th].DOI
  65. [65]N. Weickgenannt, D. Wagner, and E. Speranza, Phys. Rev. D 105, 116026 (2022), arXiv:2204.01797 [nucl-th].DOI
  66. [66]N. Weickgenannt, D. Wagner, E. Speranza, and D. H. Rischke, Phys. Rev. D 106, L091901 (2022), arXiv:2208.01955 [nucl-th].DOI
  67. [67]N. Weickgenannt, Phys. Rev. D 108, 076011 (2023), arXiv:2307.13561 [nucl-th].arXiv
  68. [68]R. Biswas, A. Daher, A. Das, W. Florkowski, and R. Ryblewski, Phys. Rev. D 108, 014024 (2023), arXiv:2304.01009 [nucl-th].DOI
  69. [69]X.-Q. Xie, D.-L. Wang, C. Yang, and S. Pu, Phys. Rev. D 108, 094031 (2023), arXiv:2306.13880 [hep-ph].arXiv
  70. [70]A. Daher, W. Florkowski, and R. Ryblewski, Phys. Rev. D 110, 034029 (2024), arXiv:2401.07608 [hep-ph].arxiv.org/abs/2401.07608
  71. [71]X. Ren, C. Yang, D.-L. Wang, and S. Pu, Phys. Rev. D 110, 034010 (2024), arXiv:2405.03105 [nucl-th].arxiv.org/abs/2405.03105
  72. [72]A. Daher, W. Florkowski, R. Ryblewski, and F. Taghinavaz, Phys. Rev. D 109, 114001 (2024), arXiv:2403.04711 [hep-ph].arxiv.org/abs/2403.04711
  73. [73]Z. Fang, K. Hattori, and J. Hu, (2024), arXiv:2409.07096 [hep-ph].DOI
  74. [74]S. Dey and A. Das, (2024), arXiv:2410.04141 [nucl-th].arxiv.org/abs/2410.04141
  75. [75]D. She, Y.-W. Qiu, and D. Hou, (2024), arXiv:2410.15142 [nucl-th].DOI
  76. [76]A. D. Gallegos, U. Gürsoy, and A. Yarom, SciPost Phys. 11, 041 (2021), arXiv:2101.04759 [hep-th].DOI
  77. [77]A. D. Gallegos, U. Gursoy, and A. Yarom, JHEP 05, 139, arXiv:2203.05044 [hep-th].DOI
  78. [78]M. Hongo, X.-G. Huang, M. Kaminski, M. Stephanov, and H.-U. Yee, JHEP 11, 150, arXiv:2107.14231 [hep-th].DOI
  79. [79]A. Kumar, D.-L. Yang, and P. Gubler, Phys. Rev. D 109, 054038 (2024), arXiv:2312.16900 [nucl-th].DOI
  80. [80]D. She, A. Huang, D. Hou, and J. Liao, Sci. Bull. 67, 2265 (2022), arXiv:2105.04060 [nucl-th].DOI
  81. [81]D. Montenegro, L. Tinti, and G. Torrieri, Phys. Rev. D 96, 056012 (2017), [Addendum: Phys.Rev.D 96, 079901 (2017)], arXiv:1701.08263 [hep-th].DOI
  82. [82]G. Torrieri and D. Montenegro, Phys. Rev. D 107, 076010 (2023), arXiv:2207.00537 [hep-th].DOI
  83. [83]W. Florkowski and M. Hontarenko, (2024), arXiv:2405.03263 [hep-ph].DOI
  84. [84]F. Becattini, A. Daher, and X.-L. Sheng, Phys. Lett. B 850, 138533 (2024), arXiv:2309.05789 [nucl-th].arxiv.org/abs/2309.05789
  85. [85]M. Kiamari, N. Sadooghi, and M. S. Jafari, Phys. Rev. D 109, 036024 (2024), arXiv:2310.01874 [nucl-th].arxiv.org/abs/2310.01874
  86. [86]A. Tiwari and B. K. Patra, (2024), arXiv:2408.11514 [hep-th].DOI
  87. [87]Z. Drogosz, W. Florkowski, and M. Hontarenko, Phys. Rev. D 110, 096018 (2024), arXiv:2408.03106 [hep-ph].DOI
  88. [88]Z. Drogosz, W. Florkowski, M. Hontarenko, and R. Ryblewski, (2024), arXiv:2411.06249 [hep-ph].DOI
  89. [89]Z. Drogosz, W. Florkowski, N. Lygan, and R. Ryblewski, (2024), arXiv:2411.06154 [hep-ph].DOI
  90. [90]B. Schenke, Rept. Prog. Phys. 84, 082301 (2021), arXiv:2102.11189 [nucl-th].DOI
  91. [91]G. S. Denicol, H. Niemi, E. Molnar, and D. H. Rischke, Phys. Rev. D 85, 114047 (2012), [Erratum: Phys.Rev.D 91, 039902 (2015)], arXiv:1202.4551 [nucl-th].DOI
  92. [92]G. S. Denicol, S. Jeon, and C. Gale, Phys. Rev. C 90, 024912 (2014).
  93. [93]G. S. Denicol, C. Gale, S. Jeon, A. Monnai, B. Schenke, and C. Shen, Phys. Rev. C 98, 034916 (2018), arXiv:1804.10557 [nucl-th].DOI
  94. [94]E. Molnár, H. Niemi, G. S. Denicol, and D. H. Rischke, Phys. Rev. D 89, 074010 (2014).
  95. [95]W. Florkowski, R. Ryblewski, R. Singh, and G. Sophy, Phys. Rev. D 105, 054007 (2022), arXiv:2112.01856 [hep-ph].arXiv
  96. [96]S. K. Singh and J. Alam, The European Physical Journal C 83, 585 (2023).
  97. [97]I. Karpenko, P. Huovinen, and M. Bleicher, Comput. Phys. Commun. 185, 3016 (2014), arXiv:1312.4160 [nucl-th].DOI
  98. [98]A. Monnai, B. Schenke, and C. Shen, Phys. Rev. C 100, 024907 (2019).
  99. [99]https://sites.google.com/view/qcdneos/.sites.google.com/view/qcdneos
  100. [100]C. Shen and S. Alzhrani, Phys. Rev. C 102, 014909 (2020), arXiv:2003.05852 [nucl-th].DOI
  101. [101]S. Ryu, V. Jupic, and C. Shen, Phys. Rev. C 104, 054908 (2021), arXiv:2106.08125 [nucl-th].DOI
  102. [102]P. Huovinen and H. Petersen, The European Physical Journal A 48, 171 (2012).
  103. [103]I. A. Karpenko, P. Huovinen, H. Petersen, and M. Bleicher, Phys. Rev. C 91, 064901 (2015), arXiv:1502.01978 [nucl-th].DOI
  104. [104]A. Schäfer, I. Karpenko, X.-Y. Wu, J. Hammelmann, and H. Elfner, The European Physical Journal A 58, 230 (2022).
  105. [105]https://github.com/smash-transport/smash-hadron-sampler.github.com/smash-transport/smash-hadron-sampler
  106. [106]J. Wei et al., Phys. Rev. C 94, 054905 (2016).
  107. [107]Wergieluk, A., & al., . et . (2024). Smash-3.1.
  108. [108]J. Adams et al. (STAR Collaboration), Phys. Rev. Lett. 91, 172302 (2003).
  109. [109]I. G. Bearden et al. (BRAHMS Collaboration), Phys. Rev. Lett. 88, 202301 (2002).
  110. [110]J. Adams et al. (STAR Collaboration and STAR-RICH Collaboration), Phys. Rev. C 72, 014904 (2005).
  111. [111]M. Buzzegoli, Phys. Rev. C 105, 044907 (2022).
  112. [112]D. Wagner, (2024), arXiv:2409.07438 [nucl-th].arXiv
  113. [113]H.-Z. Wu, L.-G. Pang, X.-G. Huang, and Q. Wang, Phys. Rev. Research. 1, 033058 (2019), arXiv:1906.09385 [nucl-th].DOI
  114. [114]F. Becattini, G. Inghirami, V. Rolando, A. Beraudo, L. Del Zanna, A. De Pace, M. Nardi, G. Pagliara, and V. Chandra, Eur. Phys. J. C 75, 406 (2015), [Erratum: Eur.Phys.J.C 78, 354 (2018)], arXiv:1501.04468 [nucl-th].DOI
  115. [115]S. Chatterjee, S. K. Singh, S. Ghosh, M. Hasanuzzaman, J. Alam, and S. Sarkar, Physics Letters B 758, 269 (2016).
  116. [116]B. I. Abelev et al. (STAR Collaboration), Phys. Rev. C 79, 034909 (2009).
  117. [117]P. Bożek and W. Bronowski, Phys. Rev. C 88, 014903 (2013).
  118. [118]A. Harten, P. D. Lax, and B. v. Leer, SIAM Review 25, 35 (1983), https://doi.org/10.1137/1025002.
  119. [119]V. Schneider, U. Katscher, D. Rischke, B. Waldhauser, J. Maruhn, and C.-D. Munz, Journal of Computational Physics 105, 92 (1993).
  120. [120]J. A. Font, J. M. Ibáñez, A. Marquina, and J. M. Martí, Astronomy and Astrophysics 282, 304 (1994).
  121. [121]D. H. Rischke, S. Bernard, and J. A. Maruhn, Nuclear Physics A 595, 346 (1995).

Paper details

vlnc:W_xdb5w5a3
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents