Multi-Region Relaxed MHD (MRxMHD)

• Multi-Region Relaxed MHD is a variational framework that divides plasma into discrete relaxation regions where magnetic fields relax to linear Beltrami states while preserving key invariants.
• The framework employs a constrained energy functional and Euler–Lagrange equations to enforce ideal-MHD constraints and maintain pressure continuity across interfaces.
• MRxMHD underpins numerical codes like SPEC and facilitates modeling of 3D toroidal equilibria featuring islands, chaotic regions, and rotational flows.

Multi-Region Relaxed Magnetohydrodynamics (MRxMHD) is a variational MHD framework in which a plasma is subdivided into a finite number of discrete relaxation subregions, each separated by perfectly conducting ideal interfaces. Within each subregion, the magnetic field is fully relaxed, typically to a linear Beltrami or generalized Taylor–Woltjer state, while preservation of key fluxes, helicities, and other global invariants is enforced inside each volume. At each ideal interface, the field-line mapping, magnetic flux, and thermodynamic quantities satisfy ideal-MHD-like constraints, most importantly the continuity of total pressure. This structure has proven foundational for the study of three-dimensional toroidal equilibria with partial chaos, islands, and profiles with discontinuous pressure gradients, and forms the basis for advanced equilibrium codes such as SPEC. Extensions include flow, pressure anisotropy, Hall effects, and dynamic formulations. MRxMHD recovers classical ideal MHD in the continuum-interface limit and Taylor relaxation in the single-volume limit, enabling controlled interpolation between these extremes (Dennis et al., 2014, Hudson et al., 2012, Dennis et al., 2012, Hudson et al., 2011, Qu et al., 2020, Dewar et al., 2015).

1. Variational Principle and MRxMHD Energy Functional

The central construction in MRxMHD is a constrained energy functional extremized over the plasma domain subdivided into $N$ nested, toroidal regions $V_i$, bounded by ideal transport barriers $I_i$. Within each $V_i$, the constrained energy (for the isotropic, zero-flow case) reads

$$W = \sum_{i=1}^N E_i - \frac{1}{2}\sum_{i=1}^N \mu_i (K_i - K_i^0) - \sum_{i=1}^N \nu_i (M_i - M_i^0)$$

where

• $$E_i = \int_{V_i} \left[ \frac{1}{2}|B|^2 + \frac{p}{\gamma - 1} \right] d^3x$$,
• $K_i = \int_{V_i} A \cdot B\, d^3x$ is the magnetic helicity,
• $M_i$ is an entropy or "mass" integral, e.g., $\int p^{1/\gamma} d^3x$ for isentropic closure,
• $\mu_i,\nu_i$ are Lagrange multipliers.

Extending to MRxMHD with flow introduces additional constraints and multipliers to enforce conservation of flow-helicity $C_i = \int_{V_i} B \cdot u\, d^3x$ and optionally toroidal angular momentum $L_i$; the full functional then includes terms $-\lambda_i (C_i - C_i^0) - \Omega_i (L_i - L_i^0)$, and the kinetic energy $\frac{1}{2}\rho |u|^2$ inside $E_i$ (Dennis et al., 2014, Qu et al., 2020).

Key properties:

• Each $V_i$ allows full Taylor relaxation subject only to global invariants.
• At the interfaces, field lines are frozen-in and fluxes are constrained.
• The pressure is constant within each region, leading to a staircase profile, with discontinuities localized at the interfaces.

2. Euler–Lagrange Equations and Interface Conditions

The stationarity of the MRxMHD functional yields, in each subregion $V_i$,

• A (generalized) Beltrami equation:

$$\nabla \times B = \mu_i B + \lambda_i \nabla \times u$$

In the absence of flow ($\lambda_i = 0$), this reduces to $\nabla \times B = \mu_i B$.

• Flow–magnetic field relation:

$\rho u = \lambda_i B + \rho \Omega_i R e_\phi$

• Bernoulli/Energy relation:

$$\nu_i = \frac{1}{2}|u|^2 + \frac{\gamma}{\gamma-1} \sigma_i \rho^{\gamma-1} - \Omega_i R u_\phi$$

• In the anisotropic extension additional closure relations relate $p_\parallel$, $p_\perp$ to density and invariants (Dennis et al., 2014).

On each ideal interface, the principal interface conditions are:

• Continuity of total (thermal plus magnetic) pressure:

$[[p + \frac{1}{2}|B|^2]] = 0$

where $[[\cdot]]$ denotes the jump across the interface.

• Continuity of the normal magnetic field component: $B \cdot n = 0$.
• For interfaces with flow, continuity of normal velocity: $u \cdot n = 0$ (Dennis et al., 2014, Qu et al., 2020, Hudson et al., 2011).

3. Limiting Cases and Relationship to Classical MHD

MRxMHD interpolates between well-known MHD limits:

• Single-volume Taylor relaxation: When the entire plasma is treated as a single relaxation region, the result is the global force-free Taylor state ($\nabla \times B = \mu B$), with no pressure gradients sustained.
• Ideal MHD limit ($N \to \infty$): As the number of regions increases and interfaces are distributed densely, the MRxMHD functional converges to ideal MHD. In this limit, the piecewise-constant $\mu_i$, $p_i$ profiles approach continuous functions, and the force balance becomes $J \times B = \nabla p$ everywhere, with all local ideal-MHD constraints restored (Dennis et al., 2012, Dennis et al., 2014).
• Inclusion of general flow: In the infinite-interface limit, MRxMHD with flow reduces to ideal MHD with flow or, with axisymmetry only at the edge, ideal MHD in a rotating frame; in the absence of angular momentum constraint, the flow must become field-aligned (Dennis et al., 2014).

Extensions such as MRxMHD with anisotropy (Dennis et al., 2014) or Hall effects (Lingam et al., 2016) yield partial relaxation analogues of classical two-temperature or double-Beltrami states, respectively.

4. Numerical Implementation: SPEC and Algorithmic Details

The Stepped Pressure Equilibrium Code (SPEC) is the principal computational realization of MRxMHD (Hudson et al., 2011, Hudson et al., 2012). Its implementation centers on:

• A spectral representation of interface geometries, typically Fourier poloidal–toroidal expansions for interface surfaces, and finite-element or spectral representations for the magnetic potential.
• Mixed finite-element/Fourier discretization for the magnetic field within each region, subject to prescribed poloidal/toroidal fluxes and boundary conditions.
• Iterative solution strategy: For fixed interface shapes and target invariants, the Beltrami equation is solved within each region to obtain the magnetic field. The interface shapes are then iteratively adjusted (using Newton or conjugate-gradient descent) to enforce total-pressure continuity and prescribed fluxes.
• For flow or anisotropy, the field and flow equations are solved iteratively, alternating between updating density, velocity, and magnetic field (Qu et al., 2020).
• Prescribed “noble-irrational” rotational transforms for interfaces are used for robustness against resonant perturbations.

Convergence studies demonstrate rapid spectral and radial convergence, agreement with known axisymmetric equilibria (e.g., VMEC), and proper reproduction of nonlinear features such as Shafranov shift, islands, and ergodic layers.

5. Applications and Physical Insights

MRxMHD, especially with flow, enables the construction and analysis of equilibria with non-axisymmetric structure, islands, and chaotic regions, which are inaccessible to conventional ideal-MHD solvers. Key application domains include:

• Rotating 3D equilibria: MRxMHD with flow predicts minimum-energy states that are rigidly rotating in the laboratory frame, relevant to “snakes” and long-lived modes in tokamak plasmas. The rotation frequency emerges from the self-consistent angular momentum constraint, not as an imposed parameter (Dennis et al., 2014).
• Reversed Field Pinch (RFP) plasmas: MRxMHD with flow, as implemented in SPEC, successfully reproduces observed features such as parallel flow flattening following sawtooth crashes in MST experiments, and generates self-consistent equilibria with helical core islands (Qu et al., 2020).
• Tearing instability and reconnection: MRxMHD provides a variational framework for modeling the stepwise reconnection process in, e.g., sawtooth crashes, including the sequence of “partial relaxation” states and the quantification of magnetic energy release (Qu et al., 24 Jan 2025).
• Resonant magnetic perturbations: The formation of current sheets, shielding, and half-island structure under applied boundary ripple can be quantitatively described by MRxMHD, with analytic agreement to singular limit theory (Dewar et al., 2016, Huang et al., 2021).

6. Extensions: Anisotropy, Hall Effects, and Dynamics

MRxMHD has been generalized to include key physics beyond isotropic, static MHD:

• Pressure anisotropy and general flows: The anisotropic MRxMHD model (Dennis et al., 2014) incorporates separate $p_\parallel$, $p_\perp$ and general flow, with additional entropy and magnetic-moment constraints, and demonstrates convergence to anisotropic ideal MHD in the continuum limit.
• Hall MHD: By constraining canonical helicity (not just magnetic helicity), MRxHMHD (Lingam et al., 2016) recovers double-Beltrami states and allows for non-field-aligned flows and two-fluid relaxation equilibria.
• Dynamical MRxMHD: A fully dynamical RxMHD/MRxMHD variational principle has been constructed (Dewar et al., 2020), enabling the study of time-dependent relaxed dynamics, linear waves, and interface motion, with cross-helicity conservation and pressure-balance interface conditions derived self-consistently. This formalism clarifies issues regarding the applicability of the ideal Ohm's law, especially for dynamic, turbulent, or relaxation processes.
• Near-ideal RxMHD, augmented Lagrangian methods: Recent developments (Tavassoli et al., 2024) employ augmented Lagrangian approaches to enforce the (possibly weak) ideal Ohm law, allowing region boundaries and relaxation structure to self-organize during optimization, further reducing the need for ad hoc interface specification.

7. Physical and Computational Significance

MRxMHD provides a mathematically well-posed, computationally tractable framework for modeling realistic three-dimensional toroidal equilibria, capturing essential features of island formation, partial chaos, stepped-pressure profiles, and rotating structures. By interpolating smoothly between the completely constrained ideal MHD and the globally relaxed Taylor state, MRxMHD enables:

• Robust handling of partially relaxed, partially ordered systems in parameter regimes where traditional MHD models fail.
• Inclusion of experimental constraints (such as those from measured fluxes, helicities, or rotation) in equilibrium reconstruction.
• Direct connection to linear and nonlinear stability analysis, including tearing modes and inner-layer theory, with accurate reproduction of classical results (Kumar et al., 2022, Dewar et al., 2016).
• Systematic and numerically controlled convergence to the ideal-MHD limit as interface count increases, validated through benchmark studies and theoretical proofs (Dennis et al., 2012, Hudson et al., 2011).

MRxMHD thus forms a foundational theoretical and computational apparatus for advanced equilibrium analysis in fusion plasma physics and related fields.