---
title: 3D Magneto-Micropolar Equations
url: https://www.emergentmind.com/topics/3d-magneto-micropolar-equations
type: topic
---

# 3D Magneto-Micropolar Equations

The 3D magneto-micropolar equations are coupled nonlinear partial differential equations for electrically conducting micropolar fluids in three spatial dimensions. In their standard incompressible form, the unknowns are the velocity \(u\), the micro-rotation or angular-velocity field \(\omega\), the magnetic field \(b\), and the pressure \(p\). Relative to classical magnetohydrodynamics, the system adds an internal rotational degree of freedom and an antisymmetric \(u\)–\(\omega\) coupling, together with linear damping in the spin equation. Across recent work, the subject includes whole-space Cauchy problems, periodic and strip-domain formulations, stationary Liouville theory, partial regularity, pressure-based continuation criteria, energy-stable numerical schemes, compressible extensions, and application-driven models with explicit magnetization dynamics [2006.14427].

## 1. Canonical equations and principal variants

A common incompressible formulation on \(\mathbb R^3\) is
\[
\begin{cases}
\partial_t u + (u\cdot\nabla)u + \nabla p
= (\mu+\chi)\,\Delta u + \chi\,\nabla\times\omega + (b\cdot\nabla)b,\\[4pt]
\partial_t \omega + (u\cdot\nabla)\omega
= \gamma\,\Delta\omega + \nabla(\nabla\!\cdot\omega)
+ \chi\,\nabla\times u - 2\chi\,\omega,\\[4pt]
\partial_t b + (u\cdot\nabla)b
= \nu\,\Delta b + (b\cdot\nabla)u,\\[4pt]
\nabla\!\cdot u=\nabla\!\cdot b=0,
\end{cases}
\]
where \(\mu>0\) is the kinematic viscosity, \(\gamma>0\) the micro-rotation viscosity, \(\nu>0\) the magnetic diffusivity, and \(\chi>0\) the vortex-viscosity. The term \(-2\chi\,\omega\) is a linear damping term and plays a decisive role in decay and stability theory [2006.14427].

The stationary equations are obtained by removing the time derivatives. In the normalization used by Zhang–Zu, one seeks smooth fields \((u,\omega,b,\pi)\) on \(\mathbb R^3\) satisfying
\[
\begin{cases}
-\Delta u + (u\cdot\nabla)u + \nabla\pi = \chi\,\mathrm{curl}\,\omega + (b\cdot\nabla)b,\\
-\Delta\omega + (u\cdot\nabla)\omega - \nabla(\mathrm{div}\,\omega) + 2\chi\,\omega = \chi\,\mathrm{curl}\,u,\\
-\Delta b + (u\cdot\nabla)b - (b\cdot\nabla)u = 0,\\
\mathrm{div}\,u=\mathrm{div}\,b=0,
\end{cases}
\]
with \(\chi>0\) [2508.02449].

Generalized models replace the Laplacians in the \(u\), \(b\), and \(\omega\) equations by fractional powers \(\Lambda^\alpha\) and \(\Lambda^\beta\), while retaining the divergence-free constraints and the damping term \(-2\chi w\). Kim’s pressure-criterion study treats
\[
\partial_t u + (u\cdot\nabla)u +\nabla\pi
= -(\mu+\chi)\Lambda^\alpha u + \chi\nabla\times w + (b\cdot\nabla)b,
\]
with corresponding fractional equations for \(w\) and \(b\), for \(\alpha,\beta\in(0,2]\) [2310.00217].

Other normalizations differ by constant factors but preserve the same structural triad: incompressible transport, magnetic stretching, and micro-rotation coupling. On \(\mathbb T^3\), for example, Shang studies
\[
\partial_t u + u\cdot\nabla u -(\mu+\chi)\Delta u
= -\nabla p + 2\chi\,\mathrm{curl}\,\omega + b\cdot\nabla b,
\]
\[
\partial_t \omega + u\cdot\nabla \omega + 4\chi\,\omega - \kappa\nabla(\nabla\cdot\omega)-\eta\Delta\omega
=2\chi\,\mathrm{curl}\,u,
\]
\[
\partial_t b + u\cdot\nabla b - \nu\Delta b = b\cdot\nabla u,
\]
again with \(\nabla\cdot u=\nabla\cdot b=0\) [2603.02868]. In strip-domain work without resistivity and spin viscosity, the \(w\)-equation may reduce to an ODE-type damping-coupling law \(w_t+(u\cdot\nabla)w+4\chi w=2\chi\nabla\times u\) [2605.21927].

A standard structural observation is that suppressing the micro-rotation field recovers a classical MHD-type system. Several recent papers use this comparison to isolate what the microstructure contributes analytically: faster decay, effective dissipation, and in some regimes a stabilizing mechanism [2603.02868].

## 2. Geometric settings and solution frameworks

The theory is distributed across several domains and function-space regimes. For the whole-space nonstationary problem on \(\mathbb R^3\), Niche–Perusato work with Leray-type weak solutions satisfying
\[
(u,b)\in L^\infty(0,\infty;L^2_\sigma(\mathbb R^3))\cap L^2(0,\infty;\dot H^1(\mathbb R^3)),
\quad
\omega\in L^\infty(0,\infty;L^2(\mathbb R^3))\cap L^2(0,\infty;H^1(\mathbb R^3)),
\]
with initial data \((u_0,b_0)\in L^2_\sigma(\mathbb R^3)\), \(\omega_0\in L^2(\mathbb R^3)\) [2006.14427].

For stationary problems on \(\mathbb R^3\), the usual hypothesis is full smoothness, \((u,\omega,b,\pi)\in C^\infty(\mathbb R^3)\), with decay or growth encoded by Lebesgue norms over expanding annuli \(A_R=B_{2R}\setminus B_R\). Zhang–Zu introduce the scale-invariant quantities
\[
X_{\alpha,p,R}=R^{-\alpha}\|u\|_{L^p(A_R)},\quad
Y_{\beta,q,R}=R^{-\beta}\|\omega\|_{L^q(A_R)},\quad
Z_{\gamma,r,R}=R^{-\gamma}\|b\|_{L^r(A_R)},
\]
which organize the Liouville theory in a form sensitive to scaling and annular growth [2508.02449].

Periodic problems on \(\mathbb T^3\) are typically posed for zero-mean Sobolev data. Shang assumes \((u_0,\omega_0,b_0)\in H^3(\mathbb T^3)\) or \(H^N(\mathbb T^3)\), with \(\nabla\cdot u_0=\nabla\cdot b_0=0\), and analyzes global strong solutions near the zero state or near a constant background magnetic field \(\alpha\) [2603.02868].

Bounded-domain formulations split into two major classes. In numerical analysis, Qiu considers a bounded Lipschitz domain \(\Omega\subset\mathbb R^3\), homogeneous Dirichlet conditions for \(u\) and \(w\), and perfect-conductor conditions \(B\cdot n=0\), \((\nabla\times B)\times n=0\) on \(\partial\Omega\) [2303.06000]. In the compressible Coulomb-force problem of Liu–Zhong, \(\Omega\) is a bounded simply connected \(C^3\) domain, and slip conditions are imposed:
\[
u\cdot n=0,\ (\mathrm{curl}\,u)\times n=0,\ 
w\cdot n=0,\ (\mathrm{curl}\,w)\times n=0,\ 
b\cdot n=0,\ (\mathrm{curl}\,b)\times n=0,\ 
\partial_n\Phi=0
\]
on \(\partial\Omega\) [2203.06658].

The strip domain \(\Omega=\mathbb R^2\times(0,1)\) has become central in the non-resistive, non-spin-viscous theory. There the velocity satisfies the no-slip condition \(u|_{x_3=0,1}=0\), while the magnetic and micro-rotation equations are handled through Lagrangian reformulations, div–curl structure, and refined trace estimates rather than standard dissipative boundary conditions [2605.21927, 2607.05166].

## 3. Sharp decay and asymptotic behavior

The sharp \(L^2\)-decay theory for the 3D whole-space incompressible system is organized by the decay character \(r^*(v_0)\), which measures the lowest-frequency behavior of the initial datum through
\[
P_r(v_0)=\lim_{\rho\to0}\rho^{-2r-n}\int_{|\xi|<\rho}|\widehat{v_0}(\xi)|^2\,d\xi,
\qquad
r^*(v_0):\ 0<P_{r^*}(v_0)<\infty.
\]
For \(z=(u,b,\omega)\), Niche–Perusato prove that if \(r^*=r^*(z_0)\in(-3/2,\infty)\) and \(32\,\chi(\mu+\chi+\gamma)>1\), then
\[
\|u(t)\|_{L^2}^2+\|b(t)\|_{L^2}^2
=\|z(t)\|_{L^2}^2
\le C(1+t)^{-\min\{\frac32+r^*,\,\frac52\}},
\]
and for \(-3/2<r^*\le1\) there is a matching lower bound
\[
\|z(t)\|_{L^2}^2\ge C(1+t)^{-(\frac32+r^*)}.
\]
The micro-rotation field decays faster:
\[
\|\omega(t)\|_{L^2}^2
\le C(1+t)^{-\min\{\frac52+r^*,\,\frac72\}},
\]
so one always has at least \((1+t)^{-1/2}\)-decay for \(\omega\), regardless of \(r^*\) [2006.14427].

That faster decay is not a secondary refinement but a direct consequence of the linear damping. Niche–Perusato also compare the nonlinear solution with the linearized flow
\[
\bar u_t=(\mu+\chi)\Delta\bar u+\chi\nabla\times\bar\omega,\qquad
\bar\omega_t=\gamma\Delta\bar\omega+\nabla(\nabla\cdot\bar\omega)+\chi\nabla\times\bar u-2\chi\bar\omega,\qquad
\bar b_t=\nu\Delta\bar b,
\]
showing
\[
\|\bar z(t)\|_{L^2}^2\approx (1+t)^{-(\frac32+r^*)},
\qquad
\|\bar\omega(t)\|_{L^2}^2\approx (1+t)^{-(\frac52+r^*)},
\]
and establishing first-order asymptotic estimates for \(z(t)-\bar z(t)\) and \(\omega(t)-\bar\omega(t)\). Their proof combines decay character, Fourier splitting, and Duhamel estimates exploiting the factor \(e^{-2\chi(t-s)}\) [2006.14427].

Anisotropic dissipation produces a different but related decay theory. For the 3D system with horizontal dissipation, Zhang and collaborators establish global well-posedness when the initial data and their vertical derivatives are sufficiently small in \(L^2\). Under the additional assumption of anisotropic negative regularity \((u_0,B_0,w_0,\partial_3u_0,\partial_3B_0,\partial_3w_0)\in\dot H_h^{-\sigma}\), with \((k-1)/(2(k-2))<\sigma<1\), the global solution satisfies
\[
\|(u,B,w,\partial_3u,\partial_3B,\partial_3w)(t)\|_{L^2}^2\le C(1+t)^{-\sigma},
\]
\[
\|(\nabla_hu,\nabla_hB,\nabla_hw)(t)\|_{L^2}^2\le C(1+t)^{-(1+\sigma)},
\]
which are stated as the optimal large-time decay rates for the \(H^1\)-norm of the solution [2509.19015].

## 4. Stability with partial or degenerate dissipation

Recent work has emphasized regimes in which classical viscous or magnetic smoothing is missing, and where the microstructure compensates for that loss. On \(\mathbb T^3\), Shang studies two such cases. In the first, \(\mu=0\) but \(\chi>0\), \(\eta>0\), and \(\nu>0\); for small \(H^3\) data, there exists a unique global strong solution with
\[
\sup_{t>0}\|(u,\omega,b)(t)\|_{H^3}\le C\varepsilon,
\qquad
\|(u,\omega,b)(t)\|_{H^3}\le C\varepsilon e^{-C_0t}.
\]
In the second, \(\mu=0\) and \(\nu=0\), one perturbs around a constant background magnetic field \(\alpha\), assumes \(|\alpha|^2<\chi<2\), a Diophantine condition
\[
\forall\,k\in\mathbb Z^3\setminus\{0\},\qquad |\alpha\cdot k|\ge c|k|^{-r},
\]
and obtains a unique global solution with algebraic decay in lower norms, for example
\[
\|(u,\omega,B)(t)\|_{H^{r+5}}\le C(1+t)^{-3/2}.
\]
The central conclusion is that the microstructure has the effect of enhancing dissipation and contributes to stabilize the fluid [2603.02868].

This point is especially striking because the corresponding MHD limits remain difficult. Shang explicitly notes that when \(\chi=0\), the first regime becomes the 3D inviscid and resistive MHD equations, whose stability problem is still a challenging open problem, and the second becomes 3D ideal MHD, for which weighted spaces are usually required in global results [2603.02868]. A plausible implication is that the \(u\)–\(\omega\) coupling acts as an effective dissipation mechanism even when direct velocity viscosity or magnetic diffusion is absent.

The strip-domain theory develops the same theme with boundary effects. Zhao studies the incompressible system in \(\Omega=\mathbb R^2\times(0,1)\) without resistivity and spin viscosity,
\[
u_t+(u\cdot\nabla)u-(\mu+\chi)\Delta u+\nabla p=(B\cdot\nabla)B+2\chi\nabla\times w,
\]
\[
w_t+(u\cdot\nabla)w+4\chi w=2\chi\nabla\times u,
\qquad
B_t+(u\cdot\nabla)B-(B\cdot\nabla)u=0,
\]
and proves global well-posedness of small classical solutions by a two-layer energy method adapted from Guo–Tice. The resulting lower-order energy and dissipation satisfy
\[
\mathcal E_L(t)\lesssim (1+t)^{-2},
\qquad
\mathcal D_L(t)\lesssim (1+t)^{-2},
\]
while higher-order norms remain uniformly controlled [2605.21927].

Luo and collaborators analyze a related strip problem with zero magnetic and angular viscosities near the steady state
\[
u=0,\qquad w=0,\qquad b={\bf e}_3=(0,0,1).
\]
For sufficiently small perturbations in high Sobolev norms, they obtain a unique global classical solution and establish decay
\[
\sum_{l=0}^1 \|\partial_t^l u(t)\|_{H^{2(N+2)-2l}}^2
+\sum_{l=0}^1 \|\partial_t^l w(t)\|_{H^{2(N+2)-2l}}^2
+\|\eta(t)\|_{H^{2(N+2)}}^2
\lesssim \mathcal E_{2N}(0)(1+t)^{4-2N}.
\]
For \(N\gg1\), the decay is faster than any polynomial and is described as “almost exponential” [2607.05166].

## 5. Stationary equations and Liouville theory

The stationary 3D magneto-micropolar equations on \(\mathbb R^3\) support a substantial Liouville theory, whose aim is to identify conditions under which every smooth stationary solution must be trivial. In the 2022 study of Cho, Neustupa, and Yang, the system is considered with \(\gamma=\nu=\chi=1\), and three distinct criteria are proved. First, if there exist smooth matrix-valued potentials \(\Phi,\Psi,\Upsilon\) with
\[
\nabla\cdot\Phi=u,\qquad \nabla\cdot\Psi=b,\qquad \nabla\cdot\Upsilon=\omega,
\]
and their mean oscillations satisfy a BMO-type growth bound for some \(3<s\le6\), then any smooth solution is trivial. Second, if \(u,b,\omega\in L^p(\mathbb R^3)\) for some \(2\le p<9/2\), then \(u\equiv b\equiv\omega\equiv0\). Third, if \(1\le p<9/4\), \(u,b,\omega\in L^p(\mathbb R^3)\), and \(u,b,\omega\to0\) as \(|x|\to\infty\), then again the solution is trivial [2204.05759].

These theorems are based on a cut-off energy inequality, an auxiliary divergence-correction field \(W_R\) solving a Galdi-type divergence problem, and estimates of the form
\[
\int_{B_R}\bigl(|\nabla u|^2+|\nabla b|^2+|\nabla\omega|^2\bigr)
+\int_{B_R}|\omega|^2
\lesssim
\frac1{R^2}\int_{B_{2R}\setminus B_R}(|u|^2+|b|^2+|\omega|^2)
+\frac1R\int_{B_{2R}\setminus B_R}(|u|^3+|b|^3+|\omega|^3),
\]
followed by an iteration lemma [2204.05759].

Zhang–Zu substantially sharpen this picture. Assuming \(\chi\in(0,2)\), they prove Liouville theorems under annular growth conditions encoded by \(X_{\alpha,p,R}\), \(Y_{\beta,q,R}\), and \(Z_{\gamma,r,R}\). In the “large-\(p\)” and “small-\(p\)” regimes they obtain triviality under explicit algebraic relations among \(\alpha\), \(\gamma\), \(p\), and \(r\), and in the Lebesgue-space corollary they show \(u=\omega=b=0\) whenever \((u,\omega,b)\in L^p\times L^q\times L^r\) with one of the index ranges
\[
p\in[3,9/2],\ q\in[1,\infty],\ r\in[1,2p'],
\]
or
\[
p\in[3,9/2],\ q\in[1,\infty],\ r\in[2p',6],\ 1/p+2/r\ge2/3,
\]
or
\[
p\in(3/2,3),\ q\in[1,\infty],\ r\in[1,6].
\]
They also prove purely micropolar analogues for \(b\equiv0\), with the sharp threshold \(\kappa>1/4\) [2508.02449].

The most distinctive feature of Zhang–Zu’s result is the treatment of the angular velocity. Compared with the velocity and magnetic fields, they raise the most relaxed restriction for the angular velocity, allowing the \(L^q\)-norm on annuli to grow polynomially at any degree. Their proof combines a Bogovskii-map correction, an iterative energy-cutoff inequality, Gagliardo–Nirenberg interpolation for \(u\) and \(b\), and a novel two-case interpolation for \(\omega\), followed by Giaquinta’s lemma [2508.02449]. This directly contradicts the common intuition that \(\omega\) must satisfy decay assumptions comparable to those imposed on \(u\) and \(b\).

## 6. Regularity theory and numerical approximation

Local regularity theory for the nonstationary 3D equations has been developed in the suitable-weak-solution framework. In the perturbed system studied by Campanato-type methods, one introduces Elsässer variables \(u=U+B\), \(b=U-B\), a divergence-free perturbation field \(a\), external forces \(f,g\), and a local energy inequality for
\[
E(t)=\int_{\mathbb R^3}\tfrac12\bigl(|u|^2+|b|^2+|\omega|^2\bigr)\phi(\cdot,t)\,dx.
\]
There exists an absolute \(\varepsilon^\ast>0\) such that if
\[
\sup_{0<r<R}\frac1{r^2}\iint_{Q_r(x_0,t_0)}
\bigl(|\nabla u|^2+|\nabla b|^2+|\nabla\omega|^2\bigr)\,dx\,dt
<\varepsilon^\ast,
\]
then \(u\), \(b\), and \(\omega\) belong to \(C^{\alpha,\alpha/2}(Q_{R/2}(x_0,t_0))\) for some \(\alpha\in(0,1)\). An equivalent cubic criterion using
\[
r^{-2}\iint_{Q_r(x_0,t_0)}\bigl(|u|^3+|b|^3+|\omega|^3\bigr)\,dx\,dt
\]
is also available [2111.06653].

Kim’s pressure-based regularity criteria place continuation conditions directly on the pressure in Lorentz spaces. For Leray–Hopf weak solutions of the generalized fractional system, if \(\alpha=\beta\in[1,5]\) and either
\[
\pi\in L^{q,\infty}(0,T;L^{p,\infty}(\mathbb R^3)),
\qquad
\frac3p+\frac2q=2(2\alpha-1),
\]
with the norm small, or
\[
\nabla\pi\in L^1(0,T;L^r(\mathbb R^3)),
\qquad
2\alpha<r\le\infty,
\]
then the weak solution is smooth on \((0,T]\). In the MHD-type case \(\alpha=1\), an additional criterion involving small Lorentz norms of \(\nabla\pi\) and \(b\) also yields smoothness up to time \(T\) [2310.00217].

The numerical analysis of the full 3D nonstationary equations has concentrated on preserving the underlying energy law. In a bounded Lipschitz domain, Qiu derives the continuous identity
\[
\frac{d}{dt}\Bigl(\tfrac12\|u\|_{L^2}^2+\tfrac S2\|B\|_{L^2}^2+\tfrac12\|w\|_{L^2}^2\Bigr)
+( \nu+\nu_r)\|\nabla u\|_{L^2}^2
+S\mu\|\nabla B\|_{L^2}^2
+(c_a+c_d)\|\nabla w\|_{L^2}^2
+(c_0+c_d-c_a)\|\nabla\cdot w\|_{L^2}^2
+2\nu_r\|w\|_{L^2}^2
=(f,u)+(g,w),
\]
and constructs two families of unconditionally energy stable schemes: a first-order semi-implicit Euler discretization and a second-order Crank–Nicolson scheme with extrapolated nonlinear terms preserving skew-symmetry [2303.06000].

Under standard regularity assumptions, the semi-implicit Euler method yields first-order accuracy \(O(h+\Delta t)\) for \(u\), \(B\), \(w\), and the pressure in the stated norms, while the Crank–Nicolson–extrapolation scheme improves the time order to \(O(\Delta t^2)\) with spatial order \(O(h)\). Qiu also proposes fully decoupled first-order schemes based on Gauss–Seidel-type splitting, and reports numerical tests with manufactured solutions, monotone decay of the discrete energy \(E_h^n\), and a 3D lid-driven cavity computation at high Reynolds number [2303.06000].

## 7. Compressible and application-driven extensions

The compressible theory couples magneto-micropolar dynamics with density evolution and, in some models, electrostatic effects. Jia–Tan–Zhou study the compressible magneto-micropolar Navier–Stokes–Poisson system on \(\mathbb R^3\), with unknowns \((\rho,u,w,b,\Phi)\), and show global existence near a constant state under the smallness condition \(\mathcal E_3(0)\le\delta\ll1\), even when higher Sobolev norms may be large. They further prove optimal decay rates in negative Sobolev or Besov scales,
\[
\|\nabla^k(\nabla\Phi,\varrho,u,w,b)(t)\|_{L^2}
\le C_0(1+t)^{-(k+s)/2},
\]
and obtain the usual \(L^p\)-\(L^2\) decay without requiring the \(L^p\) norm of the initial data to be small [1906.06848].

Liu–Zhong treat a bounded-domain compressible isentropic system with Coulomb force and slip boundary conditions. If the initial total energy
\[
E_0=\int_\Omega\Bigl(\tfrac12\rho_0|u_0|^2+\tfrac12\rho_0|w_0|^2+\tfrac12|b_0|^2+G(\rho_0)\Bigr)\,dx
\]
is sufficiently small, then there exists a unique global classical solution satisfying
\[
0<\rho\le2\overline\rho,\quad
\rho-\tilde\rho\in C([0,T];W^{2,q}),\quad
(u,w)\in C([0,T];H^2)\cap L^2(0,T;H^3),\quad
b\in C([0,T];H^2)\cap L^2(0,T;H^4),
\]
for every \(T>0\). Their proof relies on piecewise estimates, slip-boundary identities, and effective viscous fluxes involving both velocity and micro-rotational velocity [2203.06658].

A separate application-driven line introduces explicit magnetization and micromagnetorotation (MMR). Sabeel and Hameed formulate a 3D incompressible magneto-micropolar model with unknowns \((\boldsymbol U,\boldsymbol W,p,\boldsymbol H,\boldsymbol B,\boldsymbol M)\), including the MMR torque
\[
\boldsymbol M\times\boldsymbol H
\]
in the angular-momentum equation and an MMR constitutive law with new nondimensional parameters
\[
\alpha=\frac{\tau U_0}{L},\qquad \beta=\frac{M_0}{\mu_0H_0}.
\]
They then derive a boundary-layer reduction and study its ODE form numerically [2308.08457].

Aslani and collaborators use a related micropolar MHD blood-flow model with magnetization law
\[
M_i=\frac{M_{eq}}{|H|}\bigl[H_i-\tau\,\epsilon_{ijk}H_j\omega_k\bigr]
\]
and micromagnetic torque \(\epsilon_{ijk}M_jH_k\), and solve the 3D stenosed-artery problem numerically with the OpenFOAM solvers `epotMicropolarFoam` and `epotMMRFoam`. Their computations indicate that when MMR is disregarded, the magnetic field does not significantly alter blood flow regardless of its intensity, whereas with MMR the flow exhibits reductions of up to \(30\%\) in velocity and vorticity and up to \(99.9\%\) in microrotation, with damping of vortices and disturbances [2504.13678].

Taken together, these extensions show that the 3D magneto-micropolar framework is not a single equation set but a family of closely related models. The core analytic features—antisymmetric velocity–spin coupling, magnetic stretching, and often a damping term in the micro-rotation equation—persist across incompressible, compressible, periodic, bounded, strip-domain, and magnetization-augmented formulations. The current literature suggests that these features are not merely constitutive embellishments: they materially alter decay, Liouville rigidity, regularity thresholds, and global stability mechanisms [2006.14427, 2603.02868].

Source: https://www.emergentmind.com/topics/3d-magneto-micropolar-equations