betaFeedback

WorksPhysics and Astronomy

Application range of perfect spin hydrodynamics

Abstract

The application range of perfect spin hydrodynamics is studied in two cases: one based on the classical spin description and the other using a quantum spin density matrix (Wigner function). Different forms of the conditions connecting the components of the spin polarization tensor, particle mass, temperature, and hydrodynamic flow are introduced, and their mutual relations are explained. The results obtained are important for practical applications of spin hydrodynamics to model heavy-ion collisions.

Introduction

Spin-polarization measurements of various hadrons produced in relativistic heavy-ion collisions (for example, of the Λ\Lambda hyperons and vector mesons) have recently attracted much interest [1, 2, 3, 4, 5, 6]. For a current experimental review of this broad topic, we refer the reader to [7]. The experimental data naturally inspired a wide range of theoretical studies [8, 9, 10, 11, 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], which in particular have led to the concept of spin hydrodynamics [12]. The latter framework implements spin degrees of freedom in the formalism of “ordinary” relativistic hydrodynamics that is known as one of the most effective tools used to interpret the heavy-ion data [45, 46, 47, 48, 49]. Although the formalism of spin hydrodynamics has been developed for several years now, fully 3D numerical codes where the spin degrees of freedom follow their own dynamic equations have been formulated only very recently [50, 51]. Moreover, the construction of the spin hydrodynamics formalism itself is not yet complete, as several formal issues still require clarification. One of such problems is the applicability range of perfect spin hydrodynamics, which is thoroughly analyzed in this paper.

The main object of interest in spin hydrodynamics is the spin polarization tensor ωαβ\omega_{\alpha\beta} that can be interpreted as the dimensionless ratio of the spin chemical potential Ωαβ\Omega_{\alpha\beta} to the temperature TT, ωαβ=Ωαβ/T\omega_{\alpha\beta} = \Omega_{\alpha\beta}/T. Formal studies of spin hydrodynamics and its practical applications are mostly restricted to the case where the components of ωαβ\omega_{\alpha\beta} are small (we note that the expansion in ωαβ\omega_{\alpha\beta} is meaningful since the components ωαβ\omega_{\alpha\beta} are dimensionless). The main exceptions are: the review article [52], where the classical description of spin was formulated, and the three more recent publications [53, 54, 55], where the generalized thermodynamic identities were introduced and a novel form of the equilibrium Wigner function was proposed.

In Ref. [55], a criterion was given for the applicability of perfect spin hydrodynamics (describing a relativistic gas of spin-1/21/2 particles), which imposes an upper bound for the magnitude of the components of the spin polarization tensor,

12ωmax⁡(2+2)1+v1−v<mT.(1)\frac{1}{2}\omega_{\max}(2+\sqrt{2})\sqrt{\frac{1+v}{1-v}} < \frac{m}{T}. \tag*{(1)}

Here ωmax⁡\omega_{\max} is the largest absolute value of the components of the spin polarization tensor ωαβ\omega_{\alpha\beta} defined in the laboratory (hereafter LAB) frame, mm is the mass of the particles and vv denotes the magnitude of the flow three-vector (defined again in the LAB). It was also shown in [55] that Eq. (1) is naturally fulfilled by the physical conditions encountered in heavy-ion collisions.

In this work, we elucidate the origin of the condition (1) and derive finer constraints that involve all components of the spin polarization tensor. Our description refers to the classical spin approach [52] and the quantum framework based on the Wigner function [55]. We also discuss in parallel the conditions formulated in the fluid rest frame (LRF) and in an arbitrary frame that we identify with LAB. The results presented herein will be very useful in future applications of spin hydrodynamics to model heavy-ion collisions.

Before we discuss our methods and results, it is important to emphasize two points. First, by the applicability range of perfect spin hydrodynamics, we mean the internal consistency of this framework. We do not address here the issue of including dissipation. Second, by a perfect spin fluid we understand a system in local equilibrium defined by the conservation laws of the baryon number, energy, linear momentum, spin part of the total angular momentum, and entropy. The effects of the spin-orbit interaction leading to a transfer between the spin and orbital parts of the total angular momentum can be added by the dissipative terms, as explained in [53, 54]. They are not related to the problems discussed in this work.

Notation and conventions – Throughout the text, we use natural units ℏ=c=kB=1\hbar= c = k_{\mathrm{B}} = 1. For the Levi-Civita symbol ϵμναβ\epsilon^{\mu\nu\alpha\beta}, we follow the convention ϵ0123=−ϵ0123=+1\epsilon^{0123} = -\epsilon_{0123} = +1. We use the mostly negative metric convention, gμν=diag⁡(+1,−1,−1,−1)g_{\mu\nu} = \operatorname{diag}(+1,-1,-1,-1). The trace over the spinor indices is denoted by tr⁡4\operatorname{tr}_4. The scalar product of two four-vectors aa and bb reads a⋅b=a0b0−a⋅ba \cdot b = a^0b^0 - a \cdot b, where the three-vectors are indicated in bold. The tilde diacritic is used to denote dual tensors, which are obtained from rank two antisymmetric tensors aμνa_{\mu\nu} by contraction with the Levi-Civita symbol and division by a factor of two. For example, a~μν=(1/2)ϵμναβaαβ\tilde{a}_{\mu\nu} = (1/2)\epsilon_{\mu\nu\alpha\beta}a^{\alpha\beta}. The inverse transformation is aρσ=−(1/2)ϵρσμνa~μνa^{\rho\sigma} = -(1/2)\epsilon^{\rho\sigma\mu\nu}\tilde{a}_{\mu\nu}.

Classical spin description

Extended phase space

In the classical-spin approach we use distribution functions f(x,p,s)f(x,p,s) in the extended phase space that includes the spin four-vector sμs^\mu along with the spacetime coordinates xμ=(t,x)x^\mu=(t,\boldsymbol{x}) and the four-momentum pμ=(Ep,p)p^\mu=(E_p,\boldsymbol{p}). Throughout this work, we assume that particles are on the mass shell, hence Ep=m2+p2E_p=\sqrt{m^2+\boldsymbol{p}^2}.

The spin four-vector sμs^\mu is orthogonal to the particle four-momentum, p⋅s=0p\cdot s=0, and satisfies the normalization condition s2=−ℓ2s^2=-\ell^2, where ℓ2=3/4\ell^2=3/4 is the eigenvalue of the Casimir operator for the SU(2) group. Following the works by Mathisson [56, 57], we define the internal angular momentum of a particle by the equation sαβ=(1/m)ϵαβγδpγsδs^{\alpha\beta}=(1/m)\epsilon^{\alpha\beta\gamma\delta}p_\gamma s_\delta. In the particle rest frame (PRF), where pμ=(m,0,0,0)p^\mu=(m,0,0,0), the four-vector sαs^\alpha has only space components, sα=(0,s∗)s^\alpha=(0,\boldsymbol{s}_*), with the normalization ∣s∗∣=ℓ|\boldsymbol{s}_*|=\ell.

The allowed spin configurations can be included by using a covariant measure [52]

∫dS…=mπℓ∫d4s δ(s⋅s+ℓ2)δ(p⋅s)…(2)\int dS\ldots=\frac{m}{\pi\ell}\int d^4s\,\delta(s\cdot s+\ell^2)\delta(p\cdot s)\ldots \tag*{(2)}

The two delta functions in Eq. (2) take care of the normalization and orthogonality conditions, while the prefactor m/(πℓ)m/(\pi\ell) is chosen to obtain the normalization that accounts for the spin degeneracy factor of spin-1/2 particles,

∫dS=mπℓ∫d4s δ(s⋅s+ℓ2)δ(p⋅s)=2.(3)\int dS=\frac{m}{\pi\ell}\int d^4s\,\delta(s\cdot s+\ell^2)\delta(p\cdot s)=2. \tag*{(3)}

The measure in the momentum space, denoted by dPdP, is defined as

∫dP…=∫d3p(2π)3Ep….(4)\int dP\ldots=\int\frac{d^3p}{(2\pi)^3E_p}\ldots. \tag*{(4)}

We note that both dPdP and dSdS are Lorentz invariant.

Local equilibrium function

The local equilibrium distribution functions for particles (+) and antiparticles (−) in the extended phase space have the form

f±(x,p,s)=exp⁡(±ξ−p⋅β+12ωαβsαβ),(5)f^\pm(x,p,s)=\exp\left(\pm\xi-p\cdot\beta+\frac{1}{2}\omega_{\alpha\beta}s^{\alpha\beta}\right), \tag*{(5)}

where βμ=uμ/T\beta^\mu=u^\mu/T, uμ=γ(1,v)u^\mu=\gamma(1,\boldsymbol{v}) is the fluid four-velocity, and ξ=μ/T\xi=\mu/T, where μ\mu is the baryon chemical potential. Equation (5) is valid for the Boltzmann statistics. In this case, the fugacity factor exp⁡(±ξ)\exp(\pm\xi) is irrelevant to our analysis; therefore, we set ξ=μ/T=0\xi=\mu/T=0 and consider only particles in the following discussion.

The spin-polarization tensor ωαβ\omega_{\alpha\beta} appearing in (5) can be expressed in terms of the electric- and magneticlike three-vectors, e=(e1,e2,e3)\boldsymbol{e}=(e^1,e^2,e^3) and b=(b1,b2,b3)\boldsymbol{b}=(b^1,b^2,b^3), as

ωμν=[0e1e2e3−e10−b3b2−e2b30−b1−e3−b2b10],(6)\omega_{\mu\nu}= \begin{bmatrix} 0 & e^1 & e^2 & e^3\\ -e^1 & 0 & -b^3 & b^2\\ -e^2 & b^3 & 0 & -b^1\\ -e^3 & -b^2 & b^1 & 0 \end{bmatrix}, \tag*{(6)}

where the sign conventions follow those used in the Faraday tensor in [58].

Scalar density – classical version

An object of central importance for our analysis is the scalar density

ncl=∫dP∫dSexp⁡(−p⋅β+12ωαβsαβ)(7)n_{\mathrm{cl}}=\int dP\int dS\exp\left(-p\cdot\beta+\frac{1}{2}\omega_{\alpha\beta}s^{\alpha\beta}\right) \tag*{(7)}
=∫dPexp⁡(−p⋅β)∫dSexp⁡(12ωαβsαβ).(8)=\int dP\exp(-p\cdot\beta)\int dS\exp\left(\frac{1}{2}\omega_{\alpha\beta}s^{\alpha\beta}\right). \tag*{(8)}

All other tensors of interest can be obtained by taking derivatives of ncln_{\mathrm{cl}} with respect to βμ\beta^\mu and/or ωαβ\omega_{\alpha\beta}. The integral over the spin space is analytic and equals [52]

∫dSexp⁡(12ωαβsαβ)=2sinh⁡(ℓ∣b∗∣)ℓ∣b∗∣,(9)\int dS\exp\left(\frac{1}{2}\omega_{\alpha\beta}s^{\alpha\beta}\right)=\frac{2\sinh(\ell|\boldsymbol{b}_*|)}{\ell|\boldsymbol{b}_*|}, \tag*{(9)}

where b∗\boldsymbol{b}_* is the magneticlike part of the spin polarization tensor boosted to the particle rest frame

b∗=1m(Epb−p×e−p⋅bEp+mp).(10)\boldsymbol{b}_*=\frac{1}{m}\left(E_p\boldsymbol{b}-\boldsymbol{p}\times\boldsymbol{e}-\frac{\boldsymbol{p}\cdot\boldsymbol{b}}{E_p+m}\boldsymbol{p}\right). \tag*{(10)}

Since ncln_{\mathrm{cl}} is a Lorentz scalar, it can be calculated in the local rest frame of the fluid element (LRF), where u′μ=(1,0,0,0)u'^\mu=(1,0,0,0). Denoting quantities in LRF by primes, we obtain

ncl=∫dP′exp⁡(−Ep′T)∫dS′exp⁡(12ωαβ′s′αβ)(11)n_{\mathrm{cl}}=\int dP'\exp\left(-\frac{E_{p'}}{T}\right)\int dS'\exp\left(\frac{1}{2}\omega'_{\alpha\beta}s'^{\alpha\beta}\right) \tag*{(11)}
=2∫dP′exp⁡(−Ep′T)sinh⁡(ℓ∣b∗′∣)ℓ∣b∗′∣,(12)=2\int dP'\exp\left(-\frac{E_{p'}}{T}\right)\frac{\sinh(\ell|\boldsymbol{b}'_*|)}{\ell|\boldsymbol{b}'_*|}, \tag*{(12)}

where we used Eq. (9) with

b∗′=1m[Ep′b′−p′×e′−p′⋅b′Ep′+mp′].(13)\boldsymbol{b}'_*=\frac{1}{m}\left[E_{p'}\boldsymbol{b}'-\boldsymbol{p}'\times\boldsymbol{e}'-\frac{\boldsymbol{p}'\cdot\boldsymbol{b}'}{E_{p'}+m}\boldsymbol{p}'\right]. \tag*{(13)}

Here b′\mathbf{b}' and e′\mathbf{e}' are the b\mathbf{b} and e\mathbf{e} fields transformed to LRF [58],

e′=γ(e+v×b)−γ2γ+1v(v⋅e),(14)\mathbf{e}' = \gamma(\mathbf{e}+\mathbf{v}\times\mathbf{b})-\frac{\gamma^2}{\gamma+1}\mathbf{v}(\mathbf{v}\cdot\mathbf{e}), \tag*{(14)}
b′=γ(b−v×e)−γ2γ+1v(v⋅b).(15)\mathbf{b}' = \gamma(\mathbf{b}-\mathbf{v}\times\mathbf{e})-\frac{\gamma^2}{\gamma+1}\mathbf{v}(\mathbf{v}\cdot\mathbf{b}). \tag*{(15)}

Constraining the spin polarization components

From Eq. (10) it is easy to conclude that the existence of the integral defining ncln_{\rm cl} is controlled by the behavior of the difference

τ≡−Ep′T+gm∣b′∣(16)\tau\equiv-\frac{E'_{\mathbf{p}}}{T}+\frac{\mathfrak{g}}{m}|\mathbf{b}'| \tag*{(16)}

at large values of the three-momentum ∣p′∣|\mathbf{p}'|. For large ∣p′∣|\mathbf{p}'|, Eqs. (13) and (16) yield

τ=−∣p′∣T+g∣p′∣m∣b′−p^′×e′−(p^′⋅b′)p^′∣,(17)\tau=-\frac{|\mathbf{p}'|}{T}+\frac{\mathfrak{g}|\mathbf{p}'|}{m}\left|\mathbf{b}'-\hat{\mathbf{p}}'\times\mathbf{e}'-(\hat{\mathbf{p}}'\cdot\mathbf{b}')\hat{\mathbf{p}}'\right|, \tag*{(17)}

with hats denoting unit vectors. The integral (10) converges if τ<0\tau<0 for every choice of p^′\hat{\mathbf{p}}'. Therefore, to find whether it converges for given mm, TT, b′\mathbf{b}', and e′\mathbf{e}', we have to maximize the quantity

Q′≡∣b′−p^′×e′−(p^′⋅b′)p^′∣(18)Q'\equiv\left|\mathbf{b}'-\hat{\mathbf{p}}'\times\mathbf{e}'-(\hat{\mathbf{p}}'\cdot\mathbf{b}')\hat{\mathbf{p}}'\right| \tag*{(18)}

with respect to p^′\hat{\mathbf{p}}'. Squaring both sides and making use of elementary trigonometric identities as well as of the cyclic property of the mixed product A⋅(B×C)\mathbf{A}\cdot(\mathbf{B}\times\mathbf{C}) leads to

Q′2=b′2sin⁡2θbp+e′2sin⁡2θep−2p^′⋅(e′×b′),(19)Q'^{2}=\mathbf{b}'^{2}\sin^{2}\theta_{bp}+\mathbf{e}'^{2}\sin^{2}\theta_{ep}-2\hat{\mathbf{p}}'\cdot(\mathbf{e}'\times\mathbf{b}'), \tag*{(19)}

where θbp\theta_{bp} and θep\theta_{ep} denote the angles between p^′\hat{\mathbf{p}}' and the vectors b′\mathbf{b}' and e′\mathbf{e}', respectively. Now, it can be readily seen that to maximize this expression, p^′\hat{\mathbf{p}}' should be chosen as the unit vector antiparallel to e′×b′\mathbf{e}'\times\mathbf{b}'.4 Then, the sines are equal to ±1\pm1, and

Q′2=b′2+e′2+2∣e′×b′∣.(20)Q'^{2}=\mathbf{b}'^{2}+\mathbf{e}'^{2}+2|\mathbf{e}'\times\mathbf{b}'|. \tag*{(20)}

Hence, the criterion for convergence of (10), max⁡(τ)<0\max(\tau)<0, takes the form

gb′2+e′2+2∣e′×b′∣<mT.(21)\mathfrak{g}\sqrt{\mathbf{b}'^{2}+\mathbf{e}'^{2}+2|\mathbf{e}'\times\mathbf{b}'|}<\frac{m}{T}. \tag*{(21)}

Equation (21) is our first important result. We note that it differs from the criterion presented in [59] by the appearance of the cross product and by its “primed” form, meaning that it can be applied in any frame if the primed (LRF) components are expressed by those in LAB.

Maximum-magnitude criteria

If the components eie^{i} (i=1,2,3i=1,2,3) are restricted to lie in the interval −ωmax⁡≤ei≤ωmax⁡-\omega_{\max}\le e^{i}\le\omega_{\max} for a certain positive ωmax⁡\omega_{\max}, and the same holds for the components bib^{i} (i=1,2,3i=1,2,3), then one can prove (see Appendix A) that b2+e2+2∣e×b∣\sqrt{\mathbf{b}^{2}+\mathbf{e}^{2}+2|\mathbf{e}\times\mathbf{b}|} is bounded from above by ωmax⁡(2+2)\omega_{\max}(2+\sqrt{2}). Therefore, based on (21), we can formulate a maximum-magnitude criterion for ωmax⁡′\omega'_{\max}, namely

g ωmax⁡′(2+2)<mT.(22)\mathfrak{g}\,\omega'_{\max}(2+\sqrt{2})<\frac{m}{T}. \tag*{(22)}

We note that (22) is more restrictive than (21), i.e. (22) implies (21). However, it may be simpler to use, especially as it can easily be transformed into a condition that is valid in LAB.

The upper bound on ωmax⁡′\omega'_{\max} given ωmax⁡\omega_{\max} and v\mathbf{v} can be found as follows. Let us assume that ∣e∣≥∣b∣|\mathbf{e}|\ge|\mathbf{b}|. We rotate the coordinate frame so that e=(ωmax⁡,0,0)\mathbf{e}=(\omega_{\max},0,0) and seek the choice of b\mathbf{b} and v\mathbf{v} that maximizes e′1=ωmax⁡′e'^{1}=\omega'_{\max}. We find that the projection of the last term in (14) on e\mathbf{e} is nonpositive regardless of the choice of v\mathbf{v},

−γ2γ+1(v⋅e)2≤0.(23)-\frac{\gamma^{2}}{\gamma+1}(\mathbf{v}\cdot\mathbf{e})^{2}\le0. \tag*{(23)}

Therefore, in the extremal case, this term should be zero, whereas the projection of the second term, γ v×b\gamma\,\mathbf{v}\times\mathbf{b}, on e\mathbf{e} should be as large as possible. It is easy to verify that the choice that maximizes e′1e'^{1} is b⊥e\mathbf{b}\perp\mathbf{e}, v⊥e\mathbf{v}\perp\mathbf{e}, v⊥b\mathbf{v}\perp\mathbf{b}, ∣b∣=∣e∣|\mathbf{b}|=|\mathbf{e}|, with v×b\mathbf{v}\times\mathbf{b} parallel to e\mathbf{e}. Then,

e′=γ(1+v)e=1+v1−v e,(24)\mathbf{e}'=\gamma(1+v)\mathbf{e}=\sqrt{\frac{1+v}{1-v}}\,\mathbf{e}, \tag*{(24)}

and the criterion (22) takes the form5

gωmax⁡(2+2)1+v1−v<mT.(25)\mathfrak{g}\omega_{\max}(2+\sqrt{2})\sqrt{\frac{1+v}{1-v}}<\frac{m}{T}. \tag*{(25)}

This criterion is coarser (more restrictive) than that derived in the previous subsection, as it corresponds to the worst-case arrangement of the vectors e\mathbf{e}, b\mathbf{b}, and v\mathbf{v}, but it may be easier to apply in practical calculations, as it uses only two variables, ωmax⁡\omega_{\max} and vv, instead of the nine components of the considered vectors.

Wigner function approach

Equation (25) has exactly the same form as Eq. (1) provided that the factor g\mathfrak{g} is replaced by 1/21/2. Since Eq. (1) was derived within the framework based on the Wigner function, we first recall the main elements of this approach.

Local equilibrium

A very recent paper [55] introduced a local equilibrium Wigner function defined by the expression

X(x,p)=\exp\left[-\beta_\mu(x)p^\mu+\gamma_5\slashed{a}\right], \tag*{(26)}

where

aμ(x,p)=−12mω~μν(x)pν.(27)a_\mu(x,p)=-\frac{1}{2m}\widetilde{\omega}_{\mu\nu}(x)p^\nu. \tag*{(27)}

The tensor ω~μν\widetilde{\omega}_{\mu\nu} is the dual polarization tensor

ω~μν=(0b1b2b3−b10e3−e2−b2−e30e1−b3e2−e10).(28)\widetilde{\omega}_{\mu\nu}= \begin{pmatrix} 0 & b^1 & b^2 & b^3 \\ -b^1 & 0 & e^3 & -e^2 \\ -b^2 & -e^3 & 0 & e^1 \\ -b^3 & e^2 & -e^1 & 0 \end{pmatrix}. \tag*{(28)}

We note that the following identities hold

12ωμνωμν=−12ω~μνω~μν=b⋅b−e⋅e,\frac{1}{2}\omega_{\mu\nu}\omega^{\mu\nu} =-\frac{1}{2}\widetilde{\omega}_{\mu\nu}\widetilde{\omega}^{\mu\nu} =\boldsymbol{b}\cdot\boldsymbol{b}-\boldsymbol{e}\cdot\boldsymbol{e},
ω~μνωμν=−4e⋅b.(29)\widetilde{\omega}_{\mu\nu}\omega^{\mu\nu}=-4\boldsymbol{e}\cdot\boldsymbol{b}. \tag*{(29)}

The orthogonality condition aμpμa_\mu p^\mu may be treated as a version of the Frenkel condition [60] that is frequently used in spin hydrodynamics in different formulations.

Scalar density – quantum version

The results for conserved tensors (including baryon current, energy-momentum tensor, spin tensor, and entropy current) obtained with classical and quantum descriptions have been shown to agree up to the second order in ω\omega. Interestingly, in this case, the integrals over allowed classical spin configurations are equivalent to the calculation of the traces over the spinor space.

The scalar density calculated with the quantum treatment of spin is given by the formula

n_{\rm qt}=\frac{1}{2}\operatorname{tr}_4\int dP\exp\left(-p\cdot\beta+\gamma_5\slashed{a}\right), \tag*{(30)}

where the prefactor 1/21/2 guarantees that the quantum formula agrees with the classical one for the spinless case. To proceed further, it is convenient to make use of the identity

\exp\left(\gamma_5\slashed{a}\right)=\cosh\sqrt{-a^2}\left[1+\frac{\gamma_5\slashed{a}}{\sqrt{-a^2}}\tanh\sqrt{-a^2}\right]. \tag*{(31)}

that yields

nqt=2∫dPexp⁡(−p⋅β)cosh⁡−a2.(32)n_{\rm qt}=2\int dP\exp(-p\cdot\beta)\cosh\sqrt{-a^2}. \tag*{(32)}

Constraining the spin polarization components

Since the integral (32) is a scalar, we can write it down in the LRF,

nqt=2∫dP′exp⁡(−p′⋅β)cosh⁡−a′2.(33)n_{\rm qt}=2\int dP'\exp(-p'\cdot\beta)\cosh\sqrt{-a'^2}. \tag*{(33)}

This integral converges if for large ∣p′∣|p'| the expression

−p′T+−a′2(34)-\frac{p'}{T}+\sqrt{-a'^2} \tag*{(34)}

is negative for every direction of p′p'. This looks similar to (16) and, in fact, it can be easily verified that

−a′2=∣p′∣24m2[b′2−(p^′⋅b′)2+(p^′×e′)2−2p^′⋅(e′×b′)]=∣p′∣24m2[b′2sin⁡2θbp+e′2sin⁡2θep−2p^′⋅(e′×b′)].(35)\begin{aligned} -a'^2=\frac{|p'|^2}{4m^2}\left[b'^2-(\hat{p}'\cdot b')^2+(\hat{p}'\times e')^2-2\hat{p}'\cdot(e'\times b')\right] \\ &=\frac{|p'|^2}{4m^2}\left[b'^2\sin^2\theta_{bp}+e'^2\sin^2\theta_{ep}-2\hat{p}'\cdot(e'\times b')\right]. \tag*{(35)} \end{aligned}

with θbp\theta_{bp} and θep\theta_{ep} defined analogously as in Eq. (19). Then, the same argument as below Eq. (19) leads to the convergence criterion

12b′2+e′2+2∣e′×b′∣<mT.(36)\frac{1}{2}\sqrt{b'^2+e'^2+2|e'\times b'|}<\frac{m}{T}. \tag*{(36)}

Maximum-magnitude criteria

Since the inequalities (21) and (36) differ only by a multiplicative constant, the entirety of Sec. II E is valid here as well. In the Wigner case, the maximum-magnitude criteria analogous to (22) and (25) take the form

12ωmax⁡′(2+2)<mT,(37)\frac{1}{2}\omega'_{\max}(2+\sqrt{2})<\frac{m}{T}, \tag*{(37)}
12ωmax⁡(2+2)1+v1−v<mT.(38)\frac{1}{2}\omega_{\max}(2+\sqrt{2})\sqrt{\frac{1+v}{1-v}}<\frac{m}{T}. \tag*{(38)}

Summary and conclusions

In this work, we have constructed several criteria that constrain the application of perfect spin hydrodynamics in the formulations defined in Ref. [52, 53, 54, 55]. These criteria are a general restriction on the magnitude of the allowed values of the spin-polarization tensor components. The maximum values allowed depend on the mass-to-temperature ratio, m/Tm/T, and the value of the hydrodynamic flow vv. We have verified all the conditions by performing numerical calculations/integrations.

The specific form of the criteria depends on the set of variables given. A full collection of our results is provided in Table I. If we have complete data on the vectors ee, bb and vv (in the laboratory frame), we can apply the most exact version (third row in Table I), which uses the Lorentz-transformed fields; see Eqs. (14) and (15). If only the magnitude of vv is given but not its direction relative to bb and ee, we can notice that the left-hand side is bounded from above by an expression that uses non-transformed fields multiplied by a Doppler effect-like factor (second row in Table I). Finally, given only vv and ωmax⁡\omega_{\max} (the largest component of the vectors ee and bb), we can use the criterion that involves only these two variables and is the upper bound of the two finer criteria (first row in Table I).

VariablesClassicalWigner
ωmax⁡,v\omega_{\max},vsωmax⁡(2+2)1+v1−v<mT\mathfrak{s}\omega_{\max}(2+\sqrt{2})\sqrt{\frac{1+v}{1-v}}<\frac{m}{T}12ωmax⁡(2+2)1+v1−v<mT\frac{1}{2}\omega_{\max}(2+\sqrt{2})\sqrt{\frac{1+v}{1-v}}<\frac{m}{T}
e,b,ve,b,vsb2+e2+2∣e×b∣1+v1−v<mT\mathfrak{s}\sqrt{b^2+e^2+2\lvert e\times b\rvert}\sqrt{\frac{1+v}{1-v}}<\frac{m}{T}12b2+e2+2∣e×b∣1+v1−v<mT\frac{1}{2}\sqrt{b^2+e^2+2\lvert e\times b\rvert}\sqrt{\frac{1+v}{1-v}}<\frac{m}{T}
e,b,ve,b,vsb′2+e′2+2∣e′×b′∣<mT\mathfrak{s}\sqrt{b^{\prime2}+e^{\prime2}+2\lvert e'\times b'\rvert}<\frac{m}{T}12b′2+e′2+2∣e′×b′∣<mT\frac{1}{2}\sqrt{b^{\prime2}+e^{\prime2}+2\lvert e'\times b'\rvert}<\frac{m}{T}

Table 1. Applicability criteria for different sets of available variables and for two different formulations of perfect spin hydrodynamics (the second column – the approach based on the classical spin description, the third column – the approach using the Wigner function; ωmax⁡=max⁡(∣ei∣,∣bi∣)\omega_{\max}=\max(\lvert e^i\rvert,\lvert b^i\rvert), i=1,2,3i=1,2,3).

We have derived the applicability criteria for two different formulations of spin hydrodynamics. The first is based on the classical treatment of spin [52], while the second uses the equilibrium Wigner function (the spin is described quantum mechanically with the help of the spin density matrix [55]). Interestingly, the classical and Wigner cases have the same functional form, differing only by a constant factor. The quantum approach is obtained by replacing the classical spin normalization s=3/4\mathfrak{s}=\sqrt{3/4} by 1/21/2. We thus see how naturally the spin value appears in the quantum framework.

We stress that our conditions include only hydrodynamic variables and differ from those used in the past, that involved both hydrodynamic and particle properties. One of the previously used conditions had the form b∗(x,p)≪1b_*(x,p)\ll1 and could always be violated by particles with very large energies.

The similarity of the results obtained for the classical and quantum cases follows from the fact that in the two calculations one has to check the large momentum behavior of the integrands that have a very similar asymptotics in this limit (both the hyperbolic sine appearing in the classical description and the hyperbolic cosine appearing in the quantum case approach the exponential function for large arguments). Because of this behavior, we expect our results to hold for Fermi-Dirac statistics as well, although this still needs to be worked out in the quantum formulation (the Fermi-Dirac case with classical spin treatment is discussed in [54, 61]).

Acknowledgments — We thank Rajeev Singh for bringing our attention to a poster presented at the Quark Matter 2025 conference. This work was supported in part by the Polish National Science Centre (NCN) Grant No. 2022/47/B/ST2/01372.

Proof of the upper bound on QQ

Here we prove that if all the components (i=1,2,3i=1,2,3) of vectors bb and ee are constrained to lie in intervals −ωmax⁡≤bi≤ωmax⁡-\omega_{\max}\le b^i\le\omega_{\max} and −ωmax⁡≤ei≤ωmax⁡-\omega_{\max}\le e^i\le\omega_{\max} for a certain positive ωmax⁡\omega_{\max}, then the following inequality holds

Q=b2+e2+2∣e×b∣≤(2+2)ωmax⁡.(39)Q=\sqrt{b^2+e^2+2\lvert e\times b\rvert}\le(2+\sqrt{2})\omega_{\max}. \tag*{(39)}

We square both sides,

Q2=b2+e2+2∣e×b∣≤(6+42)ωmax⁡2,(40)Q^2=b^2+e^2+2\lvert e\times b\rvert\le(6+4\sqrt{2})\omega_{\max}^2, \tag*{(40)}

and calculate the second partial derivative with respect to bib^i,

∂2∂(bi)2Q2=2+2e2((e×b)i)2∣e×b∣3≥2,(41)\frac{\partial^2}{\partial(b^i)^2}Q^2=2+\frac{2e^2\left((e\times b)^i\right)^2}{\lvert e\times b\rvert^3}\ge2, \tag*{(41)}

where (e×b)i(e\times b)^i is the ii-th component of the vector product. Analogously, the second partial derivative with respect to eie^i is

∂2∂(ei)2Q2=2+2b2((e×b)i)2∣e×b∣3≥2.(42)\frac{\partial^2}{\partial(e^i)^2}Q^2=2+\frac{2b^2\left((e\times b)^i\right)^2}{\lvert e\times b\rvert^3}\ge2. \tag*{(42)}

Therefore, if we keep five of the components constant and consider Q2Q^2 a one-variable function of the sixth one, this function is convex, and we can use Bauer’s maximum principle to state that it attains its extremum at an end of the interval.

Repeating this argument for each of the components, we arrive at the conclusion that the global maximum must be reached at a corner of the six-dimensional cube [−ωmax⁡,ωmax⁡]6[-\omega_{\max},\omega_{\max}]^{6}. At a corner, b=ωmax⁡(±1,±1,±1)\mathbf{b}=\omega_{\max}(\pm1,\pm1,\pm1) and e=ωmax⁡(±1,±1,±1)\mathbf{e}=\omega_{\max}(\pm1,\pm1,\pm1), from which follows b2=e2=3ωmax⁡2\mathbf{b}^{2}=\mathbf{e}^{2}=3\omega_{\max}^{2}, and there are only two possible values of ∣e×b∣|\mathbf{e}\times\mathbf{b}|, namely 00 and 22ωmax⁡22\sqrt{2}\omega_{\max}^{2}. Since 0<22ωmax⁡20<2\sqrt{2}\omega_{\max}^{2}, it is the latter choice that corresponds to the global maximum of Q=3+3+42ωmax⁡2=(2+2)ωmax⁡\mathcal{Q}=\sqrt{3+3+4\sqrt{2}\omega_{\max}^{2}}=(2+\sqrt{2})\omega_{\max}.

Note that this simple argument, which did not require testing for joint convexity of Q2\mathcal{Q}^{2} (e.g., by testing for the positive semidefiniteness of the Hessian matrix) but only for convexity in each variable separately, could be applied because the domain of Q2\mathcal{Q}^{2} was a Cartesian product of intervals; it would have been insufficient if the domain had a more complex shape.

References

References

  1. [1]Z.-T. Liang and X.-N. Wang, Phys. Rev. Lett. 94, 102301 (2005), Erratum: Phys.Rev.Lett. 96, 039901 (2006)], arXiv:nucl-th/0410079.DOI
  2. [2]Z.-T. Liang and X.-N. Wang, Phys. Lett. B 629, 20 (2005), arXiv:nucl-th/0411101.DOI
  3. [3]L. Adamczyk et al. (STAR), Nature 548, 62 (2017), arXiv:1701.06657 [nucl-ex].
  4. [4]J. Adam et al. (STAR), Phys. Rev. C 98, 014910 (2018), arXiv:1805.04400 [nucl-ex].DOI
  5. [5]J. Adam et al. (STAR), Phys. Rev. Lett. 123, 132301 (2019), arXiv:1905.11917 [nucl-ex].DOI
  6. [6]S. Acharya et al. (ALICE), Phys. Rev. Lett. 125, 012301 (2020), arXiv:1910.14408 [nucl-ex].
  7. [7]T. Niida and S. A. Voloshin, Int. J. Mod. Phys. E 33, 2430010 (2024), arXiv:2404.11042 [nucl-ex].DOI
  8. [8]F. Becattini and L. Tinti, Annals Phys. 325, 1566 (2010), arXiv:0911.0864 [gr-qc].DOI
  9. [9]F. Becattini, Phys. Part. Nucl. Lett. 8, 801 (2011).
  10. [10]F. Becattini, V. Chandra, L. Del Zanna, and E. Grossi, Annals Phys. 338, 32 (2013), arXiv:1303.3431 [nucl-th].DOI
  11. [11]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
  12. [12]W. Florkowski, B. Friman, A. Jaiswal, and E. Speranza, Phys. Rev. C 97, 041901 (2018), arXiv:1705.00587 [nucl-th].DOI
  13. [13]W. Florkowski, B. Friman, A. Jaiswal, R. Ryblewski, and E. Speranza, Phys. Rev. D 97, 116017 (2018), arXiv:1712.07676 [nucl-th].DOI
  14. [14]W. Florkowski, A. Kumar, and R. Ryblewski, Phys. Rev. C 98, 044906 (2018), arXiv:1806.02616 [hep-ph].DOI
  15. [15]K. Hattori, M. Hongo, X.-G. Huang, M. Matsuo, and H. Taya, Phys. Lett. B 795, 100 (2019), arXiv:1901.06615 [hep-th].DOI
  16. [16]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
  17. [17]W. Florkowski, A. Kumar, R. Ryblewski, and R. Singh, Phys. Rev. C 99, 044910 (2019), arXiv:1901.09655 [hep-ph].DOI
  18. [18]S. Bhadury, W. Florkowski, A. Jaiswal, A. Kumar, and R. Ryblewski, Phys. Lett. B 814, 136096 (2021), arXiv:2002.03937 [hep-ph].DOI
  19. [19]D. Montenegro and G. Torrieri, Phys. Rev. D 102, 036007 (2020), arXiv:2004.10195 [hep-th].DOI
  20. [20]N. Weickgenannt, E. Speranza, X.-L. Sheng, Q. Wang, and D. H. Rischke, Phys. Rev. Lett. 127, 052301 (2021), arXiv:2005.01506 [hep-ph].DOI
  21. [21]S. Shi, C. Gale, and S. Jeon, Phys. Rev. C 103, 044906 (2021), arXiv:2008.08618 [nucl-th].DOI
  22. [22]S. Bhadury, W. Florkowski, A. Jaiswal, A. Kumar, and R. Ryblewski, Phys. Rev. D 103, 014030 (2021), arXiv:2008.10976 [nucl-th].DOI
  23. [23]K. Fukushima and S. Pu, Phys. Lett. B 817, 136346 (2021), arXiv:2010.01608 [hep-th].DOI
  24. [24]S. Li, M. A. Stephanov, and H.-U. Yee, Phys. Rev. Lett. 127, 082302 (2021), arXiv:2011.12138 [hep-th].
  25. [25]R. Singh, G. Sophys, and R. Ryblewski, Phys. Rev. D 103, 074024 (2021), arXiv:2011.14907 [hep-ph].DOI
  26. [26]A. D. Gallegos, U. Gürsoy, and A. Yarom, SciPost Phys. 11, 041 (2021), arXiv:2101.04759 [hep-th].DOI
  27. [27]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
  28. [28]D. She, A. Huang, D. Hou, and J. Liao, Sci. Bull. 67, 2265 (2022), arXiv:2105.04060 [nucl-th].DOI
  29. [29]M. Hongo, X.-G. Huang, M. Kaminski, M. Stephanov, and H.-U. Yee, JHEP 11, 150 (2021), arXiv:2107.14231 [hep-th].DOI
  30. [30]J. Hu, Phys. Rev. D 105, 076009 (2022), arXiv:2111.03571 [hep-ph].DOI
  31. [31]R. Singh, M. Shokri, and S. M. A. T. Mehr, Nucl. Phys. A 1035, 122656 (2023), arXiv:2202.11504 [hep-ph].DOI
  32. [32]N. Weickgenannt, D. Wagner, E. Speranza, and D. H. Rischke, Phys. Rev. D 106, 096014 (2022), arXiv:2203.04766 [nucl-th].DOI
  33. [33]D. Wagner, N. Weickgenannt, and D. H. Rischke, Phys. Rev. D 106, 116021 (2022), arXiv:2210.06187 [nucl-th].DOI
  34. [34]S. Dey, W. Florkowski, A. Jaiswal, and R. Ryblewski, Phys. Lett. B 843, 137994 (2023), arXiv:2303.05271 [hep-th].DOI
  35. [35]N. Weickgenannt and J.-P. Blaizot, Phys. Rev. D 109, 056012 (2024), arXiv:2311.15817 [hep-ph].arxiv.org/abs/2311.15817
  36. [36]A. Kumar, D.-L. Yang, and P. Gubler, Phys. Rev. D 109, 054038 (2024), arXiv:2312.16900 [nucl-th].DOI
  37. [37]D. Wagner, M. Shokri, and D. H. Rischke, Phys. Rev. Res. 6, 043103 (2024), arXiv:2405.00533 [nucl-th].DOI
  38. [38]D. Wagner, Phys. Rev. D 111, 016008 (2025), arXiv:2409.07143 [nucl-th].DOI
  39. [39]S. Dey and A. Das, Phys. Rev. D 111, 074037 (2025), arXiv:2410.04141 [nucl-th].arxiv.org/abs/2410.04141
  40. [40]D. She, W. Qiu, and D. Hou, Phys. Rev. D 111, 036027 (2025), arXiv:2410.15142 [nucl-th].DOI
  41. [41]X.-G. Huang, (2024), arXiv:2411.11753 [nucl-th].DOI
  42. [42]S. Bhadury, (2025), arXiv:2503.08428 [hep-ph].arxiv.org/abs/2503.08428
  43. [43]S. Dey, (2025), arXiv:2504.18838 [hep-th].
  44. [44]N. Weickgenannt and J.-P. Blaizot, Phys. Rev. D 111, 056006 (2025), arXiv:2409.11045 [hep-ph].arxiv.org/abs/2409.11045
  45. [45]J.-Y. Ollitrault, Eur. J. Phys. 29, 275 (2008), arXiv:0708.2433 [nucl-th].DOI
  46. [46]P. Romatschke, Int. J. Mod. Phys. E 19, 1 (2010), arXiv:0902.3663 [hep-ph].DOI
  47. [47]W. Florkowski, Phenomenology of Ultra-Relativistic Heavy-Ion Collisions (Singapore: World Scientific, 2010).DOI
  48. [48]C. Gale, S. Jeon, and B. Schenke, Int. J. Mod. Phys. A 28, 1340011 (2013), arXiv:1301.5893 [nucl-th].DOI
  49. [49]A. Jaiswal and V. Roy, Adv. High Energy Phys. 2016, 9623034 (2016), arXiv:1605.08694 [nucl-th].DOI
  50. [50]S. K. Singh, R. Rylewbski, and W. Florkowski, Phys. Rev. C 111, 024907 (2025), arXiv:2411.08223 [hep-ph].DOI
  51. [51]Sapna, S. K. Singh, and D. Wagner, (2025), arXiv:2503.22552 [hep-ph].DOI
  52. [52]W. Florkowski, A. Kumar, and R. Rylewbski, Prog. Part. Nucl. Phys. 108, 103709 (2019), arXiv:1811.04409 [nucl-th].DOI
  53. [53]W. Florkowski and M. Hontarenko, Phys. Rev. Lett. 134, 082302 (2025), arXiv:2405.03263 [hep-ph].DOI
  54. [54]Z. Drogosz, W. Florkowski, and M. Hontarenko, Phys. Rev. D 110, 096018 (2024), arXiv:2408.03106 [hep-ph].DOI
  55. [55]S. Bhadury, Z. Drogosz, W. Florkowski, S. K. Kar, and V. Mykhaylova, (2025), arXiv:2505.02657 [hep-ph].arxiv.org/abs/2505.02657
  56. [56]M. Mathisson, Acta Phys. Polon. 6, 163 (1937).
  57. [57]M. Mathisson, General Relativity and Gravitation 42, 1011 (2010).
  58. [58]J. D. Jackson, Classical Electrodynamics (Wiley, 1998).
  59. [59]S. Rajeev, private communication, see also https://indico.cern.ch/event/1334113/contributions/6291310/attachments/3044936/5379962/QM2025_Abboud_Poster.pdf.
  60. [60]J. Frenkel, Z. Phys. 37, 243 (1926).
  61. [61]Z. Drogosz, (2025), arXiv:2504.03396 [hep-ph].DOI

Paper details

vlnc:W_rzqs2gfy
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