Spin Hydrodynamics in Biquaternionic Form
Introduction
A fluid whose elements carry spin has a second conserved quantity, and its hydrodynamics is the hydrodynamics of a fluid with two currents: the energy-momentum element and the spin element. The new thermodynamic variable is the spin polarization, whose conjugate intensive quantity is the spin potential $\mu_{\mathrm{spn}}$:
$$ \mu_{\mathrm{spn}}=\mu_{\mathrm{spn}}\left(T,\ \mu_{\mathrm{chm}},\ \omega_{\lambda\nu}\right), $$
and the article's subject is what the fluid does when that variable is not in equilibrium — when the polarization is not the one the local temperature and vorticity would give it.
The spin polarization is an antisymmetric object with six independent components, and the corpus writes it as an element whose components are $\omega_{\lambda\nu}$. Three of the six are spatial rotations and three are boosts; the rotation part is the one a fluid element's spin couples to, and the boost part is the one a spin current transports. The article keeps every index low and labels the direction, following the frame and metric conventions: $\omega_{\lambda\nu}$ is an antisymmetric array of numbers, not a tensor slot.
The article's central caution is stated at the outset and derived later: the naive local relation is incomplete.
$$ \frac{\partial p_{\mathrm{th}}}{\partial \omega_{\lambda\nu}}\neq S_{\Psi,\lambda\nu}\ \text{in general}, $$
so the differential relations cannot close the equations. The equality holds only in global thermal equilibrium, where the polarization is the one the temperature fixes; off equilibrium the pressure is not a function of the polarization alone, and a closure that assumes it is has assumed away the physics. That is the article's most useful content and its warning against the most common shortcut in the spin-hydrodynamics literature.
What the article builds on. The fluid's stress is The Local Thermal State of a Fluid's, the entropy current and its Hermiticity declaration are The Local Entropy Current's, the spin element is the rotor's, which is The Spin-Half Fluid of the Biquaternion Dirac Field's, and the conservation law the spin obeys is Spin Conservation and the Frenkel Condition's. The article adds the polarization as a thermodynamic variable, its diffusion and its relaxation, and the closure caution.
The Spin Polarization and the Spin Potential
The polarization element
The spin polarization is the rotor's generator, read as a thermodynamic variable. With the rotor $\tilde u_{\mathrm r}$ of the polar form and its generator $\tilde\omega_\Psi=\tfrac12\omega_{\lambda\nu}\tilde e_{\lambda\nu}$ — the antisymmetric part of the algebra's basis, with $\lambda<\nu$ — the polarization element is proportional to the generator:
$$ \tilde S_{\Psi,\eta}=\frac{\hbar}{2}\,\tilde u_{\mathrm r}\,\tilde e_{\lambda\nu}\,\tilde u_{\mathrm r}^{\natural}\,\omega_{\lambda\nu}, $$
the rotor conjugating the generator into the fluid's frame and the factor $\hbar/2$ making the components angular momenta. The display is the polar-form article's rotor read as a density: the spin element is what the rotor looks like when it is summed over a volume.
The components, low. Writing the six independent components $\omega_{\lambda\nu}=-\omega_{\nu\lambda}$ and the corresponding components $S_{\Psi,\eta,\lambda\nu}$ of the spin element, the article never raises an index; the pair $(\lambda,\nu)$ labels which rotation, and the array is antisymmetric in it. The spin tensor's components are the angular momentum density in the $\lambda\nu$ plane, and the flux of that component in the $\eta$ direction is written with $\eta$ last, so the current carries three labels: one for the direction of transport and two for the plane.
The density is the temporal slot of the current, and it is the object the thermodynamics uses. Write
$$ S_{\Psi,\lambda\nu}\equiv S_{\Psi,t,\lambda\nu}, $$
so that $S_{\Psi,\lambda\nu}$ is the angular momentum in the $\lambda\nu$ plane per unit proper volume, the $\eta=t$ component of the current. Every thermodynamic relation below uses the two-index density and every conservation or transport equation uses the three-index current, and the distinction is not bookkeeping: $\omega_{\lambda\nu}$ has two indices, so only the two-index object can be contracted with it.
The zero-divisor caution. The rotor is an element of norm one, and the spin element it builds is regular as long as the rotor is; on the zero-divisor cone the quaternionic norm of the spin element vanishes and the polarization has no inverse. The corpus states the restriction because the algebra's degenerate elements are where its structure changes, and a hydrodynamic expansion around a degenerate polarization is not the same expansion.
The two chemical potentials
A spin fluid has two intensive variables conjugate to two conserved densities, and the local thermal state is a function of both.
$$ d w_{\mathrm{fl}}=T\,d s_{\mathrm{ent}}+\mu_{\mathrm{chm}}\,d\rho_{\mathrm{num}}+\frac{1}{2}\omega_{\lambda\nu}\,dS_{\Psi,\lambda\nu}, $$
with $\mu_{\mathrm{chm}}$ the chemical potential and $\omega_{\lambda\nu}$ the angular velocity conjugate to the spin density. The display is the first law for a spin fluid, and its third term is the one a spinless fluid does not have; the factor of one half counts the antisymmetric pairs once.
The spin potential is the intensive variable of the polarization, and it is not the angular velocity. The spin potential $\mu_{\mathrm{spn}}$ is the Legendre transform's variable,
$$ \mu_{\mathrm{spn}}=\left(\frac{\partial w_{\mathrm{fl}}}{\partial \rho_{\mathrm{num,spn}}}\right)_{s_{\mathrm{ent}},\rho_{\mathrm{num}}}, $$
where $\rho_{\mathrm{num,spn}}$ is the number density of polarized elements; it measures how much energy one more polarized element costs, whereas $\omega_{\lambda\nu}$ measures how fast the frame turns. The two are equal in equilibrium and different off it, which is the distinction that the closure caution later turns on.
The local thermal state with spin, written in the corpus's letters. The state is specified by the pointwise temperature, the chemical potential and the spin potential, and the equation of state is The Equation of State in Biquaternion Form's with the third variable added:
$$ p_{\mathrm{eff}}=p_{\mathrm{th}}\left(\rho_{\mathrm{num}},\beta,\mu_{\mathrm{spn}}\right)+p_{\mathrm{qnt}}\left(\rho_{\mathrm{num}}\right), $$
the quantum part unchanged because it depends on the density alone and the thermal part gaining a dependence on the spin potential through the polarization's contribution to the energy.
Rotational Flow and Spin Gradients
The vorticity of the flow and the polarization
The flow's vorticity and the polarization are two different objects, and confusing them is the second commonest error in the subject. The vorticity is
$$ \omega^{\mathrm{vrt}}_{\mu\nu}=\partial_\mu v_{\Phi,\nu}-\partial_\nu v_{\Phi,\mu}, $$
the antisymmetric part of the flow's gradient, built from the flow element's components; the polarization $\omega_{\lambda\nu}$ is a thermodynamic variable describing how the elements are spinning about their own centres. A fluid can rotate without polarized elements and be polarized without rotating, and the two names are distinguished in this article's notation by the label $\mathrm{vrt}$ on the vorticity.
Local equilibrium ties the two together, and the tie is the source of the spin-rotation coupling. In equilibrium a fluid element in a rotating flow is polarized in proportion to the rotation, with the standard result quoted:
$$ \mu_{\mathrm{spn}}^{\mathrm{eq}}=\frac{\hbar\,\omega^{\mathrm{vrt}}_{\mathrm{mag}}}{2}, \qquad P_{\mathrm{pol}}=\tanh\left(\frac{\mu_{\mathrm{spn}}^{\mathrm{eq}}}{k_B T}\right), $$
with $\omega^{\mathrm{vrt}}_{\mathrm{mag}}$ the magnitude of the rotation vector and $P_{\mathrm{pol}}$ the polarization of one element. At $\hbar\omega^{\mathrm{vrt}}_{\mathrm{mag}}=0.019\,75\,\mathrm{GeV}$ and $k_BT=0.15\,\mathrm{GeV}$ the polarization is $0.065\,738$, so a rotation fast by hadronic standards polarizes the fluid only weakly; at $\hbar\omega^{\mathrm{vrt}}_{\mathrm{mag}}=0.065\,82\,\mathrm{GeV}$ the polarization is $0.215\,946$. The relation is the equilibrium one and is used only to identify the equilibrium polarization; off equilibrium the two differ by the polarization's own relaxation.
The coupling's energy, which is the interaction between the two. With the polarization and the vorticity both present, the energy density gains
$$ e_{\mathrm{rot}}=-\,\frac{1}{2}\,\omega^{\mathrm{vrt}}_{\lambda\nu}\,S_{\Psi,\lambda\nu}, $$
the standard spin-rotation coupling: a polarized element aligned with the rotation sits lower in energy, and a misaligned one sits higher. The display is the third term of the first law with the angular velocity replaced by the flow's vorticity, and the replacement is valid only in local equilibrium — which is the same restriction as before, and the article keeps track of it.
Spin gradients and the stress
A gradient in the polarization is a gradient in the spin density, and it contributes to the stress the way any gradient of an intensive variable does. The stress element gains a term
$$ \tilde\Sigma_{\Psi,\mathrm{spn}}=-\left(\partial_\lambda \mu_{\mathrm{spn}}\right)\tilde S_{\Psi,\eta}\ \text{in the first-order form}, $$
whose divergence is the force that a polarization gradient exerts on the flow. The display is schematic in its indices — the spin element is antisymmetric and contracts with the gradient in the $\lambda$ slot — and the article states the structure and does not write the index algebra in full, because the full algebra is bookkeeping without physics.
The structure is worth having, because it is what couples the flow to the spin. A fluid with a polarization gradient is a fluid whose momentum balance feels the spin, so the flow and the spin cannot be solved separately; the two are one system, and that is the sense in which spin hydrodynamics is not ordinary hydrodynamics with an extra field.
The entropy production of the spin channel, written with the inherited Hermiticity. The spin contributes a term to the production of The Local Entropy Current,
$$ \sigma_{\mathrm{ent,spn}}=\frac{1}{T}\,\kappa_{\mathrm{spn}}\left(\partial_\lambda \mu_{\mathrm{spn}}\right)\left(\partial_\lambda \mu_{\mathrm{spn}}\right)\ \ge 0, $$
with $\kappa_{\mathrm{spn}}$ the spin diffusion coefficient. The current remains Hermitian and the inequality remains real, so the declaration made in the entropy article covers this term without a fresh $\mathrm{Re}$; the term is a square, and its sign is the non-negativity of $\kappa_{\mathrm{spn}}$.
Spin Diffusion and Spin Relaxation
Diffusion
A polarization gradient drives a spin current, and the current's law is Fick's law in the corpus's letters.
$$ J_{\Psi,\mathrm{spn},\lambda}=-\,\kappa_{\mathrm{spn}}\,\partial_\lambda\mu_{\mathrm{spn}}, $$
with $\kappa_{\mathrm{spn}}$ the same coefficient as in the production. The two appearances are one fact: a current proportional to a gradient produces entropy proportional to the gradient squared, and the coefficient's sign is fixed by the second law's inequality.
Combining the current with the conservation law gives the diffusion equation, and it is written on the spin potential because that is the scalar the two-index object cannot carry alone. The density $S_{\Psi,\lambda\nu}$ is a function of $\mu_{\mathrm{spn}}$, with the susceptibility $\chi_{\mathrm{spn}}=\partial S_{\Psi}/\partial\mu_{\mathrm{spn}}$, so the balance $\partial_tS_{\Psi,\lambda\nu}=-\partial_\lambda J_{\Psi,\mathrm{spn},\lambda}$ divided by $\chi_{\mathrm{spn}}$ gives
$$ \partial_t\mu_{\mathrm{spn}}=D_{\mathrm{spn}}\,\partial_\lambda\partial_\lambda\mu_{\mathrm{spn}}+\text{relaxation}, \qquad D_{\mathrm{spn}}=\frac{\kappa_{\mathrm{spn}}}{\chi_{\mathrm{spn}}}, $$
the standard diffusion equation with a source, in the variable whose gradient the current follows. Its diffusion length over a time $\tau$ is $\ell_{\mathrm{spn}}=\sqrt{D_{\mathrm{spn}}\tau}$, which is the length over which a polarization disturbance is felt and the number that decides whether a given fluid element is polarized by its neighbours. The write-up of the equation in the density rather than in the potential is where the spin channel's index bookkeeping usually goes wrong: two indices on the left and none on the right is a type error and not a shorthand.
Relaxation
The polarization relaxes to its local equilibrium value on a time scale $\tau_{\mathrm{spn}}$, and the relaxation is the spin channel's answer to the shear's viscosity.
$$ \frac{d}{dt}\left(S_{\Psi,\eta,\lambda\nu}-S_{\Psi,\eta,\lambda\nu}^{\mathrm{eq}}\right)=-\,\frac{1}{\tau_{\mathrm{spn}}}\left(S_{\Psi,\eta,\lambda\nu}-S_{\Psi,\eta,\lambda\nu}^{\mathrm{eq}}\right), $$
the relaxation-time approximation, which is the standard closure and is quoted as such. Its solution is exponential, and its content is one number: how long a polarized element remembers its polarization.
A number, for scale. With $\tau_{\mathrm{spn}}=0.5\,\mathrm{fm}/c=1.667\,820\,476\times10^{-24}\,\mathrm{s}$, the surviving fraction after $0.5\,\mathrm{fm}/c$ is $e^{-1}=0.367\,879$, after $1\,\mathrm{fm}/c$ it is $0.135\,335$, after $2\,\mathrm{fm}/c$ it is $0.018\,316$, and after $5\,\mathrm{fm}/c$ it is $4.540\times10^{-5}$. The polarization is gone after a few relaxation times, which is why spin hydrodynamics is a description of the early stage of a collision and not of its late stage.
The relaxation is dissipative and its production is positive. The relaxation term contributes to $\sigma_{\mathrm{ent}}$ a piece proportional to the departure squared over $\tau_{\mathrm{spn}}T$, so the decay of the polarization heats the fluid, exactly as the decay of a shear does. The two channels — spin relaxation and shear — are separate squares of the same second law, and the article counts them separately.
Why relaxation is not optional. A first-order theory with a diffusive spin current and no relaxation is parabolic and therefore acausal, exactly as the first-order heat conduction is; the relaxation time is what makes the spin channel hyperbolic, and Causal Dissipative Hydrodynamics in Biquaternionic Form is where that structure is written for the fluid as a whole.
The Caution: Why the Differential Relations Do Not Close
The identity and its domain
The relation $\partial p_{\mathrm{th}}/\partial\omega_{\lambda\nu}=S_{\Psi,\lambda\nu}$ is a thermodynamic identity of the equilibrium state, and it is not an identity of the hydrodynamic equations. In equilibrium the pressure is a function of the complete set of state variables, the polarization among them, and the derivative is a genuine partial derivative of a function. Off equilibrium the pressure is not a function of those variables at all — it is a functional of the state's history through the relaxation — so the derivative on the left is not defined as a partial derivative, and the equality has no content.
Written with the corpus's letters, the caution is
$$ \frac{\partial p_{\mathrm{th}}}{\partial \omega_{\lambda\nu}}\neq S_{\Psi,\lambda\nu} \qquad\text{when }\omega_{\lambda\nu}\neq\omega_{\lambda\nu}^{\mathrm{eq}}, $$
and the inequality is the statement that the two sides are different objects rather than two values of one. The left side is the derivative of the equation of state along the equilibrium surface; the right side is the actual spin density, which relaxes towards that surface at its own rate. They agree only on the surface.
Why the closure fails
The differential relations are a system with more unknowns than equations, and the missing equations are the relaxation equations and not the equilibrium identity.
| Quantity | Count |
|---|---|
| Components of the flow element | $4$ |
| Temperature and chemical potential | $2$ |
| Components of the spin density | $6$ |
| The pressure and the energy density | $2$ |
| Unknowns | $14$ |
| Conservation of the energy-momentum element | $4$ |
| Conservation of the spin element | $6$ |
| The equation of state, $p$ and $w_{\mathrm{fl}}$ as functions of $T$ and $\mu$ | $2$ |
| The unit norm of the flow, one algebraic condition | $1$ |
| The second law, one inequality | $0$ (it constrains, it does not determine) |
| Relations | $13$ |
The count is the article's argument in one table: thirteen relations for fourteen unknowns, so one is missing. The flow's four components are counted before its unit norm is imposed, and the norm is listed as the algebraic condition that ties them, so the arithmetic is $14-13=1$. The equation of state is counted twice over, once for each thermodynamic quantity it determines, because $p$ and $w_{\mathrm{fl}}$ are separately listed among the unknowns and neither follows from $T$ and $\mu$ without its own relation; counting it once would overstate the deficit. The missing relation is the constitutive one for the spin: the spin current in terms of the polarization's gradient, which is also what fixes the polarization's relaxation. It cannot be supplied by the equilibrium identity, because the identity would replace six unknowns by the derivative of one function and would leave the relaxation time undetermined.
What happens if the identity is used anyway, which is what the shortcut does. Substituting $S_{\Psi,\lambda\nu}=\partial p_{\mathrm{th}}/\partial\omega_{\lambda\nu}$ adds no equation at all; it removes six unknowns, so the eight that remain are held by thirteen relations and the system is over-determined by five, the six spin conservation laws being converted into six constraints on the flow. A generic flow satisfies none of them, so the "solution" is a flow slaved to its own polarization: parabolic, acausal, and correct only on the equilibrium surface where the constraints degenerate. That is why the shortcut is tempting and why it fails — on the surface it is exact, and off it, where the physics is, it is not even a consistent truncation.
The identity's correct use, which the article keeps. It is a diagnostic: on the equilibrium surface it relates the equation of state to the spin density, so a measurement of one checks the other; and it is the target of the relaxation equation, which drives the spin density towards the value the identity gives. Using it as a definition is the error; using it as the relaxation's fixed point is the method.
What the spin sector needs instead
The missing relation is the spin current's constitutive law, and writing it honestly changes the count. The law is $J_{\Psi,\mathrm{spn},\lambda}=-\kappa_{\mathrm{spn}}\partial_\lambda\mu_{\mathrm{spn}}+\cdots$, and it introduces the current itself as a new field while removing the polarization from among the determined quantities, so the balance of the table is not restored by one line. The standard resolution is the Israel–Stewart one, applied to the spin channel exactly as Causal Dissipative Hydrodynamics in Biquaternionic Form applies it to the heat channel: the current becomes an independent variable, its relaxation equation $\tau_{\mathrm{spn}}\dot J_{\Psi,\mathrm{spn},\lambda}+J_{\Psi,\mathrm{spn},\lambda}=-\kappa_{\mathrm{spn}}\partial_\lambda\mu_{\mathrm{spn}}$ supplies the missing relation, and the count closes because each equation brought in comes with its own unknown.
The balance of the table is restored not by one relation but by a matched pair: every field the relaxation equation introduces arrives with the equation that determines it, so the deficit of the equilibrium count is carried along rather than closed. What finally removes it is a modelling choice and not a dynamical equation — the spin equation of state, or equivalently the pseudo-gauge that fixes which of the two spin tensors is the physical one — and the article says so rather than presenting a closed system it has not got. The transport coefficients $\kappa_{\mathrm{spn}}$, $\tau_{\mathrm{spn}}$ and the spin-rotation coupling remain parameters, constrained in sign by the second law's inequality and in no other way here. That is the standard Israel–Stewart counting applied to the spin sector, and the article's contribution is to say so in the corpus's letters rather than to re-derive the counting.
What Is Established and What Is Interpretation
Established, and standard. That a spin fluid's thermodynamics has two intensive variables, the temperature and the spin potential, and that its first law carries the term $\tfrac12\omega_{\lambda\nu}dS_{\Psi,\lambda\nu}$; that the polarization diffuses with a coefficient whose sign the second law fixes; that it relaxes to its local equilibrium value on a time scale $\tau_{\mathrm{spn}}$; that the relaxation is dissipative and produces entropy; that the equilibrium identity between the pressure's derivative and the spin density holds on the equilibrium surface and does not close the equations off it.
The framework's contribution. The spin element is the rotor's generator conjugated into the fluid's frame, so the polarization is a property of the rotor and not an independent field; the spin's contribution to the stress and to the entropy current is written on elements, with every index low; and the entropy inequality that constrains the spin channel inherits The Local Entropy Current's Hermiticity declaration, so no fresh $\mathrm{Re}$ is written.
Interpretation, and marked as such. That the spin element is $\tfrac{\hbar}{2}$ times the conjugated generator with the components read as angular momenta is the corpus's normalisation and not a derivation. That the relaxation is exponential is the relaxation-time approximation and is quoted, not derived.
Not claimed. That the transport coefficients have been computed — they have not. That the full index algebra of the spin stress has been written — it has not, and the article says so where the display is schematic. That the spin conservation law has been derived from the rotor's equation of motion — that is Spin Conservation and the Frenkel Condition's subject, and this article cites it. That the theory is causal — that is Causal Dissipative Hydrodynamics in Biquaternionic Form's, and the spin channel's hyperbolicity is asserted here on the strength of its relaxation equation rather than proved.
Summary
A spin fluid has two intensive variables. The spin potential $\mu_{\mathrm{spn}}$ is conjugate to the spin polarization, whose element is the rotor's generator conjugated into the fluid's frame, $\tilde S_{\Psi,\eta}=\tfrac{\hbar}{2}\tilde u_{\mathrm r}\tilde e_{\lambda\nu}\tilde u_{\mathrm r}^{\natural}\omega_{\lambda\nu}$, with the six components $\omega_{\lambda\nu}$ antisymmetric and every index low. The first law gains one term, $dw_{\mathrm{fl}}=Tds_{\mathrm{ent}}+\mu_{\mathrm{chm}}d\rho_{\mathrm{num}}+\tfrac12\omega_{\lambda\nu}dS_{\Psi,\eta,\lambda\nu}$, and the equation of state gains one argument.
Rotational flow and spin gradients couple. The vorticity $\omega^{\mathrm{vrt}}_{\mu\nu}=\partial_\mu v_{\Phi,\nu}-\partial_\nu v_{\Phi,\mu}$ is not the polarization, and in equilibrium $\mu_{\mathrm{spn}}^{\mathrm{eq}}=\hbar\omega^{\mathrm{vrt}}_{\mathrm{mag}}/2$ with $P_{\mathrm{pol}}=\tanh(\mu_{\mathrm{spn}}^{\mathrm{eq}}/k_BT)$, which is $0.065\,738$ at $\hbar\omega^{\mathrm{vrt}}_{\mathrm{mag}}=0.019\,75\,\mathrm{GeV}$ and $0.215\,946$ at $0.065\,82\,\mathrm{GeV}$ for $k_BT=0.15\,\mathrm{GeV}$. The spin-rotation energy is $e_{\mathrm{rot}}=-\tfrac12\omega^{\mathrm{vrt}}_{\lambda\nu}S_{\Psi,\eta,\lambda\nu}$ and the spin gradient contributes to the stress.
Spin diffusion is $J_{\Psi,\mathrm{spn},\lambda}=-\kappa_{\mathrm{spn}}\partial_\lambda\mu_{\mathrm{spn}}$, with the production $\sigma_{\mathrm{ent,spn}}=T^{-1}\kappa_{\mathrm{spn}}(\partial_\lambda\mu_{\mathrm{spn}})(\partial_\lambda\mu_{\mathrm{spn}})\ge0$, real by the inherited declaration. Spin relaxation drives the polarization to its equilibrium value on the time scale $\tau_{\mathrm{spn}}$; with $\tau_{\mathrm{spn}}=0.5\,\mathrm{fm}/c$ the surviving fraction is $0.367\,879$ after $0.5\,\mathrm{fm}/c$ and $4.540\times10^{-5}$ after $5\,\mathrm{fm}/c$.
The caution is the article's centre: $\partial p_{\mathrm{th}}/\partial\omega_{\lambda\nu}\neq S_{\Psi,\lambda\nu}$ off the equilibrium surface, so the differential relations do not close — thirteen relations for fourteen unknowns — and the missing relation is the spin's constitutive one, not the equilibrium identity. Using the identity as a closure over-determines the system and turns the spin conservation laws into constraints on the flow; using it as the relaxation's fixed point is the method.
Summary of Notation
| Symbol | Meaning |
|---|---|
| $\mu_{\mathrm{spn}}$ | Spin potential |
| $\mu_{\mathrm{spn}}^{\mathrm{eq}}$ | Its equilibrium value |
| $\mu_{\mathrm{chm}}$ | Chemical potential |
| $\omega_{\lambda\nu}=-\omega_{\nu\lambda}$ | Spin polarization components |
| $\omega^{\mathrm{vrt}}_{\mu\nu}$ | Vorticity of the flow |
| $\omega^{\mathrm{vrt}}_{\mathrm{mag}}$ | Magnitude of the rotation vector |
| $\tilde S_{\Psi,\eta}$ | Spin element |
| $S_{\Psi,\eta,\lambda\nu}$ | Components of the spin current, three labels, all low |
| $S_{\Psi,\lambda\nu}\equiv S_{\Psi,t,\lambda\nu}$ | Spin density, the temporal slot of the current; the conjugate of $\omega_{\lambda\nu}$ |
| $J_{\Psi,\mathrm{spn},\lambda}$ | Spin diffusion current of the $\lambda$ direction, Fick's law with the coefficient $\kappa_{\mathrm{spn}}$ |
| $\chi_{\mathrm{spn}}=\partial s_{\mathrm{ent}}/\partial\mu_{\mathrm{spn}}$ | Spin susceptibility, the entropy density's derivative with respect to the spin potential |
| $D_{\mathrm{spn}}=\kappa_{\mathrm{spn}}/\chi_{\mathrm{spn}}$ | Spin diffusivity, with $\ell_{\mathrm{spn}}=\sqrt{D_{\mathrm{spn}}\tau}$ |
| $S_{\Psi,\eta,\lambda\nu}^{\mathrm{eq}}$ | Its equilibrium value |
| $\tilde u_{\mathrm r}$ | Rotor of the polar form |
| $\tilde e_{\lambda\nu}$ | Antisymmetric basis element |
| $\rho_{\mathrm{num}}=\lVert\tilde\Xi\rVert^{2}$ | Proper density |
| $\rho_{\mathrm{num,spn}}$ | Density of polarized elements |
| $w_{\mathrm{fl}}$, $p_{\mathrm{th}}$, $p_{\mathrm{qnt}}$, $p_{\mathrm{eff}}$ | Energy density and pressures |
| $s_{\mathrm{ent}}$ | Entropy density |
| $T$, $\beta$ | Temperature, imaginary-time circumference |
| $P_{\mathrm{pol}}$ | Polarization of one element |
| $e_{\mathrm{rot}}$ | Spin-rotation energy density |
| $\kappa_{\mathrm{spn}}$ | Spin diffusion coefficient |
| $\tau_{\mathrm{spn}}$ | Spin relaxation time |
| $\ell_{\mathrm{spn}}$ | Spin diffusion length |
| $\sigma_{\mathrm{ent}}$, $\sigma_{\mathrm{ent,spn}}$ | Entropy production, its spin part |
| $\tilde\Sigma_{\Psi,\mathrm{spn}}$ | Spin contribution to the stress |
| $v_{\Phi,\nu}$ | Components of the flow element |
| $\hbar$, $k_B$, $c$ | Planck constant, Boltzmann constant, speed of light |
Further Reading
- F. Becattini, "Covariant statistical mechanics and the stress-energy tensor," Physical Review Letters 108 (2012) 244502, for the spin potential and the equilibrium density operator with angular momentum.
- F. Becattini and L. Tinti, "The ideal relativistic rotating gas as a perfect fluid with spin," Annals of Physics 325 (2010) 1566–1594, for the thermodynamics of a rotating spin fluid.
- W. Florkowski, B. Friman, A. Jaiswal and E. Speranza, "Relativistic fluid dynamics with spin," Physical Review C 97 (2018) 041901, for the spin hydrodynamics equations used here as the standard reference.
- W. Israel and J. M. Stewart, "Transient relativistic thermodynamics and kinetic theory," Annals of Physics 118 (1979) 341–372, for the relaxation-time closure and the counting argument.
- L. D. Landau and E. M. Lifshitz, Statistical Physics, Part 1 (Pergamon, 1980), for the thermodynamics of a system in rotation.
- F. Becattini, L. Bucciantini, E. Grossi and L. Tinti, "Local thermodynamical equilibrium and the $\beta$ frame for a quantum relativistic fluid," European Physical Journal C 75 (2015) 191, for the local-equilibrium density operator and the frame dependence.
- The companion articles: The Local Entropy Current, The Spin-Half Fluid of the Biquaternion Dirac Field, Spin Conservation and the Frenkel Condition, and Causal Dissipative Hydrodynamics in Biquaternionic Form.