Magnetic Exhaust Effect: Mechanisms & Applications

• Magnetic Exhuast Effect is the process by which magnetic topology channels mass, momentum, and energy into structured outflows, observed in diverse plasma environments.
• Distinct mechanisms include plasmoid ejection in Kerr black holes, density-depleted reconnection cavities in collisionless plasmas, engineered flux expansion in fusion divertors, and nozzle-like acceleration in propulsion.
• Studies reveal that magnetic geometry governs energy conversion and exhaust efficiency, impacting plasma confinement, jet formation, and thrust generation across astrophysical and laboratory settings.

Searching arXiv for recent and relevant papers on “magnetic exhaust effect” and closely related usage across black-hole accretion, reconnection exhausts, fusion divertors, and magnetic nozzles. The magnetic exhaust effect is not a single universally fixed term; rather, it denotes a family of magnetically mediated exhaust phenomena in which magnetic topology, magnetic stresses, or reconnection-driven structures redirect mass, momentum, energy, or particles into an outflow channel. In the literature, the term is used in at least four technically distinct senses: equatorial plasma ejection from vertically magnetized Kerr black holes (Karas et al., 2020); density-depleted reconnection exhausts in strong-guide-field collisionless plasmas that support electron energization and ECMI emission (Treumann et al., 2017, McCubbin et al., 2021); magnetic shaping of heat and particle exhaust in fusion divertors and quasi-continuous edge exhaust regimes (Gorno et al., 2022, Harrer et al., 2021, Verhaegh et al., 2023, Frerichs et al., 13 Apr 2026); and electromagnetic-to-kinetic energy conversion in magnetic nozzles, where azimuthal magnetic and rotational energy are converted into directed axial flow (Smolyakov et al., 24 Jun 2025). Across these domains, the common structure is the same: magnetic fields do not merely confine plasma, but organize a preferred exhaust geometry and mediate conversion of stored magnetic or rotational free energy into transported mass, momentum, radiation, or heat.

1. Terminological scope and unifying idea

The phrase “magnetic exhaust effect” appears in multiple research programs with different local meanings. In black-hole GRMHD, it denotes an equatorial magnetic venting process in which accreted vertical flux builds magnetic pressure near the horizon, forms a midplane current sheet, reconnects, and ejects plasmoids and magnetic flux outward, thereby reducing accretion (Karas et al., 2020). In strong-guide-field reconnection, it refers to the drifting, density-depleted reconnection exhaust itself: an electron quasi-hole with strongly deformed velocity distributions that can drive the electron cyclotron maser instability (Treumann et al., 2017). In gyrokinetic analyses of guide-field reconnection, the exhaust is the downstream separatrix-bounded region in which electrons are energized primarily by $j_{\parallel,e}E_{\parallel}$ and exhibit distinctive field–particle correlation signatures (McCubbin et al., 2021). In fusion research, the term is used more broadly for magnetic-topology control of power and particle exhaust through divertor shaping, baffling, and pedestal-edge transport regimes (Gorno et al., 2022, Harrer et al., 2021, Verhaegh et al., 2023, Frerichs et al., 13 Apr 2026). In plasma propulsion, it denotes acceleration in a magnetic nozzle by conversion of azimuthal electromagnetic and rotational energy into axial kinetic energy (Smolyakov et al., 24 Jun 2025).

This multiplicity is not merely semantic. A plausible implication is that “magnetic exhaust effect” is best understood as a cross-disciplinary descriptor for magnetically structured exhaust formation rather than a single mechanism. In every usage, the exhaust is enabled by magnetic geometry: equatorial current sheets near Kerr horizons, guide-field reconnection cavities, divertor leg expansion and baffling, or converging–diverging flux tubes.

2. Magnetically driven exhaust near black holes

In the black-hole context, the term is explicitly developed for a Kerr black hole immersed in a large-scale vertical Wald field after an initially spherical Bondi inflow has reached steady state (Karas et al., 2020). The spacetime is Kerr in Boyer–Lindquist coordinates, with line element

$${\rm d}s^{2} = -\frac{\Delta\Sigma}{A}\,{\rm d}t^{2} +\frac{\Sigma}{\Delta}\,{\rm d}r^{2} +\Sigma\,{\rm d}\theta^{2} +\frac{A\sin^2\theta}{\Sigma} \left({\rm d}\phi-\omega\,{\rm d}t\right)^{2},$$

where

$$\Delta=r^2-2r+a^2,\qquad \Sigma=r^2+a^2\cos^2\theta,\qquad A=(r^2+a^2)^2-\Delta a^2\sin^2\theta,\qquad \omega=\frac{2ar}{A},$$

and the horizon radius is

$R_+=1+\sqrt{1-a^2}.$

The imposed vertical field is the Wald solution, with nonzero four-potential components

$$A_t = Ba\Big[r\Sigma^{-1}\left(1+\cos^2\theta\right)-1\Big],$$

$A_\phi = B\Big[{\textstyle\frac{1}{2}(r^2+a^2)-a^2r\Sigma^{-1}(1+\cos^2\theta)\Big]\sin^2\theta.$

The numerical setup first establishes a Bondi inflow satisfying

$4\pi r^2\rho(r)v(r)=-\dot M,$

with Euler equation

$$v\frac{dv}{dr}+\frac{1}{\rho}\frac{dP}{dr}+\frac{GM}{r^2}=0,$$

or, using $dP=c_s^2 d\rho$,

$$\frac{1}{2}\left(1-\frac{c_s^2}{v^2}\right)\frac{d\,v^2}{dr} = -\frac{GM}{r^2}\left[1-\frac{2c_s^2r}{GM}\right], \qquad r_s=\frac{GM}{2c_s^2(r_s)}.$$

After the Bondi state is established, the vertical field is switched on and evolved with axisymmetric HARM in the ideal-MHD/force-free regime, enforcing

$${\rm d}s^{2} = -\frac{\Delta\Sigma}{A}\,{\rm d}t^{2} +\frac{\Sigma}{\Delta}\,{\rm d}r^{2} +\Sigma\,{\rm d}\theta^{2} +\frac{A\sin^2\theta}{\Sigma} \left({\rm d}\phi-\omega\,{\rm d}t\right)^{2},$$0

The grid is $${\rm d}s^{2} = -\frac{\Delta\Sigma}{A}\,{\rm d}t^{2} +\frac{\Sigma}{\Delta}\,{\rm d}r^{2} +\Sigma\,{\rm d}\theta^{2} +\frac{A\sin^2\theta}{\Sigma} \left({\rm d}\phi-\omega\,{\rm d}t\right)^{2},$$1, with outer boundary at $${\rm d}s^{2} = -\frac{\Delta\Sigma}{A}\,{\rm d}t^{2} +\frac{\Sigma}{\Delta}\,{\rm d}r^{2} +\Sigma\,{\rm d}\theta^{2} +\frac{A\sin^2\theta}{\Sigma} \left({\rm d}\phi-\omega\,{\rm d}t\right)^{2},$$2, inner boundary at $${\rm d}s^{2} = -\frac{\Delta\Sigma}{A}\,{\rm d}t^{2} +\frac{\Sigma}{\Delta}\,{\rm d}r^{2} +\Sigma\,{\rm d}\theta^{2} +\frac{A\sin^2\theta}{\Sigma} \left({\rm d}\phi-\omega\,{\rm d}t\right)^{2},$$3, and polytropic index $${\rm d}s^{2} = -\frac{\Delta\Sigma}{A}\,{\rm d}t^{2} +\frac{\Sigma}{\Delta}\,{\rm d}r^{2} +\Sigma\,{\rm d}\theta^{2} +\frac{A\sin^2\theta}{\Sigma} \left({\rm d}\phi-\omega\,{\rm d}t\right)^{2},$$4 (Karas et al., 2020).

The mechanism is magnetic arrest and midplane venting. The inflow advects vertical magnetic flux toward the horizon; field lines bend into a near-radial, split-monopole-like configuration; a current sheet forms in the equatorial plane; reconnection intermittently ejects plasmoids radially outward; and magnetic pressure and tension gradients continue to drive equatorial mass loss (Karas et al., 2020). The relevant competition is between magnetic pressure,

$${\rm d}s^{2} = -\frac{\Delta\Sigma}{A}\,{\rm d}t^{2} +\frac{\Sigma}{\Delta}\,{\rm d}r^{2} +\Sigma\,{\rm d}\theta^{2} +\frac{A\sin^2\theta}{\Sigma} \left({\rm d}\phi-\omega\,{\rm d}t\right)^{2},$$5

and gas or ram pressure. The reported criterion is

$${\rm d}s^{2} = -\frac{\Delta\Sigma}{A}\,{\rm d}t^{2} +\frac{\Sigma}{\Delta}\,{\rm d}r^{2} +\Sigma\,{\rm d}\theta^{2} +\frac{A\sin^2\theta}{\Sigma} \left({\rm d}\phi-\omega\,{\rm d}t\right)^{2},$$6

The paper characterizes magnetization using $${\rm d}s^{2} = -\frac{\Delta\Sigma}{A}\,{\rm d}t^{2} +\frac{\Sigma}{\Delta}\,{\rm d}r^{2} +\Sigma\,{\rm d}\theta^{2} +\frac{A\sin^2\theta}{\Sigma} \left({\rm d}\phi-\omega\,{\rm d}t\right)^{2},$$7 and emphasizes that when the inner flow approaches equipartition, $${\rm d}s^{2} = -\frac{\Delta\Sigma}{A}\,{\rm d}t^{2} +\frac{\Sigma}{\Delta}\,{\rm d}r^{2} +\Sigma\,{\rm d}\theta^{2} +\frac{A\sin^2\theta}{\Sigma} \left({\rm d}\phi-\omega\,{\rm d}t\right)^{2},$$8, inflow is partially diverted into outflow and the accretion rate is reduced (Karas et al., 2020).

This exhaust is related to MAD phenomenology but is not a Blandford–Znajek polar jet. It is equatorial, intermittent, reconnection-mediated, and tied to a midplane current sheet rather than a polar low-density funnel (Karas et al., 2020). The same broader topic connects to the Meissner problem for rapidly spinning isolated black holes: for extremal uncharged weakly isolated horizons under local stationarity and axisymmetry, the normal magnetic flux density on the horizon vanishes, implying magnetic-flux expulsion in the extremal limit (Gürlebeck et al., 2017). That result concerns horizon-threading flux in electrovac equilibrium, whereas the equatorial exhaust of vertically magnetized accretion flows arises once conducting plasma advects flux inward and removes vacuum Meissner expulsion (Karas et al., 2020, Gürlebeck et al., 2017).

3. Reconnection exhausts in space and astrophysical plasmas

In collisionless reconnection, “exhaust” denotes the downstream outflow region bounded by separatrices or compound discontinuities. In strong current-aligned guide fields, reconnection proceeds in small antiparallel components while the guide field remains largely passive; electrons stay magnetized because their gyroradii remain much smaller than exhaust scales (Treumann et al., 2017). The resulting magnetic exhaust effect is the formation of a drifting, density-depleted electron cavity at the X point, elongated along the reconnecting component and thin across it. This cavity acts as an electron quasi-hole and develops hot dilute interior populations, boundary beams, ring-like or shell-like anisotropy in $${\rm d}s^{2} = -\frac{\Delta\Sigma}{A}\,{\rm d}t^{2} +\frac{\Sigma}{\Delta}\,{\rm d}r^{2} +\Sigma\,{\rm d}\theta^{2} +\frac{A\sin^2\theta}{\Sigma} \left({\rm d}\phi-\omega\,{\rm d}t\right)^{2},$$9, and positive perpendicular gradients $$\Delta=r^2-2r+a^2,\qquad \Sigma=r^2+a^2\cos^2\theta,\qquad A=(r^2+a^2)^2-\Delta a^2\sin^2\theta,\qquad \omega=\frac{2ar}{A},$$0 that favor ECMI (Treumann et al., 2017).

The basic ECMI resonance condition is

$$\Delta=r^2-2r+a^2,\qquad \Sigma=r^2+a^2\cos^2\theta,\qquad A=(r^2+a^2)^2-\Delta a^2\sin^2\theta,\qquad \omega=\frac{2ar}{A},$$1

with

$$\Delta=r^2-2r+a^2,\qquad \Sigma=r^2+a^2\cos^2\theta,\qquad A=(r^2+a^2)^2-\Delta a^2\sin^2\theta,\qquad \omega=\frac{2ar}{A},$$2

For X-mode emission, the paper emphasizes the combination of a positive perpendicular gradient and low density such that $$\Delta=r^2-2r+a^2,\qquad \Sigma=r^2+a^2\cos^2\theta,\qquad A=(r^2+a^2)^2-\Delta a^2\sin^2\theta,\qquad \omega=\frac{2ar}{A},$$3 (Treumann et al., 2017). In the exhaust frame, the frequency maps the local field via

$$\Delta=r^2-2r+a^2,\qquad \Sigma=r^2+a^2\cos^2\theta,\qquad A=(r^2+a^2)^2-\Delta a^2\sin^2\theta,\qquad \omega=\frac{2ar}{A},$$4

Remote observation is limited by the X-mode stop band between the cold-plasma cutoffs $$\Delta=r^2-2r+a^2,\qquad \Sigma=r^2+a^2\cos^2\theta,\qquad A=(r^2+a^2)^2-\Delta a^2\sin^2\theta,\qquad \omega=\frac{2ar}{A},$$5 and $$\Delta=r^2-2r+a^2,\qquad \Sigma=r^2+a^2\cos^2\theta,\qquad A=(r^2+a^2)^2-\Delta a^2\sin^2\theta,\qquad \omega=\frac{2ar}{A},$$6, so source motion to weaker $$\Delta=r^2-2r+a^2,\qquad \Sigma=r^2+a^2\cos^2\theta,\qquad A=(r^2+a^2)^2-\Delta a^2\sin^2\theta,\qquad \omega=\frac{2ar}{A},$$7 and/or lower density is important for escape (Treumann et al., 2017).

A distinct kinetic use of the term appears in gyrokinetic reconnection studies of large-guide-field exhausts (McCubbin et al., 2021). There, the exhaust is the downstream region inside the ion diffusion region, and the dominant electron energization mechanism is non-resonant bulk acceleration of out-of-plane electron flow by $$\Delta=r^2-2r+a^2,\qquad \Sigma=r^2+a^2\cos^2\theta,\qquad A=(r^2+a^2)^2-\Delta a^2\sin^2\theta,\qquad \omega=\frac{2ar}{A},$$8. The electron energy transfer rate is

$$\Delta=r^2-2r+a^2,\qquad \Sigma=r^2+a^2\cos^2\theta,\qquad A=(r^2+a^2)^2-\Delta a^2\sin^2\theta,\qquad \omega=\frac{2ar}{A},$$9

with $R_+=1+\sqrt{1-a^2}.$0 in the strong-guide-field gyrokinetic limit, so energization is dominated by $R_+=1+\sqrt{1-a^2}.$1 (McCubbin et al., 2021). The phase-space formulation uses

$R_+=1+\sqrt{1-a^2}.$2

The field–particle correlation diagnostic is

$R_+=1+\sqrt{1-a^2}.$3

with reduced form

$R_+=1+\sqrt{1-a^2}.$4

and

$R_+=1+\sqrt{1-a^2}.$5

At the X point and exhaust midplane, the signatures are broad and symmetric in $R_+=1+\sqrt{1-a^2}.$6, consistent with non-resonant bulk acceleration; near separatrices they become mirrored asymmetric patterns due to the combination of bulk flow and quadrupolar density perturbations (McCubbin et al., 2021). The proposed single-point diagnostic for identifying a reconnection exhaust uses precisely this pattern: strong localized $R_+=1+\sqrt{1-a^2}.$7, broad symmetric energization at the X point and midplane, and mirrored asymmetry across the separatrices (McCubbin et al., 2021).

Collisionless anti-parallel reconnection introduces another exhaust-boundary effect. Pressure anisotropy from counterstreaming ions lowers the firehose parameter

$R_+=1+\sqrt{1-a^2}.$8

or equivalently, in SI notation in the source paper,

$R_+=1+\sqrt{1-a^2}.$9

depending on convention (Haggerty et al., 2018, Liu et al., 2011). As $$A_t = Ba\Big[r\Sigma^{-1}\left(1+\cos^2\theta\right)-1\Big],$$0, magnetic tension weakens and outflow jets become sub-Alfvénic (Haggerty et al., 2018). Across the exhaust boundary, classical Petschek switch-off slow shocks do not form when anisotropy drives $$A_t = Ba\Big[r\Sigma^{-1}\left(1+\cos^2\theta\right)-1\Big],$$1 sufficiently low. Instead, the structure self-organizes into an anomalous slow shock, a degenerate slow shock around an $$A_t = Ba\Big[r\Sigma^{-1}\left(1+\cos^2\theta\right)-1\Big],$$2 plateau, and a rotational discontinuity (Liu et al., 2011). This suppresses complete magnetic switch-off and reduces the outflow speed below the Walén prediction.

The empirical outflow scaling reported for nearly anti-parallel reconnection is

$$A_t = Ba\Big[r\Sigma^{-1}\left(1+\cos^2\theta\right)-1\Big],$$3

where

$$A_t = Ba\Big[r\Sigma^{-1}\left(1+\cos^2\theta\right)-1\Big],$$4

For sufficiently strong guide field, roughly

$$A_t = Ba\Big[r\Sigma^{-1}\left(1+\cos^2\theta\right)-1\Big],$$5

firehose suppression restores $$A_t = Ba\Big[r\Sigma^{-1}\left(1+\cos^2\theta\right)-1\Big],$$6 (Haggerty et al., 2018).

Near the Sun, Parker Solar Probe has now observed reconnection exhausts with compound boundaries rather than pure pairs of slow shocks (Sun et al., 28 Apr 2026). At $$A_t = Ba\Big[r\Sigma^{-1}\left(1+\cos^2\theta\right)-1\Big],$$7, each boundary consisted of a rapidly evolving inner slow shock and an outer gradual compound structure comprising a slow shock and a rotational discontinuity; deep within the exhaust, high $$A_t = Ba\Big[r\Sigma^{-1}\left(1+\cos^2\theta\right)-1\Big],$$8 and high $$A_t = Ba\Big[r\Sigma^{-1}\left(1+\cos^2\theta\right)-1\Big],$$9 triggered mirror instability and mirror-mode structures (Sun et al., 28 Apr 2026). PSP also observed an extended HCS exhaust at $A_\phi = B\Big[{\textstyle\frac{1}{2}(r^2+a^2)-a^2r\Sigma^{-1}(1+\cos^2\theta)\Big]\sin^2\theta.$0 containing sunward Alfvénic jets, embedded merging islands, and stably trapped protons up to $A_\phi = B\Big[{\textstyle\frac{1}{2}(r^2+a^2)-a^2r\Sigma^{-1}(1+\cos^2\theta)\Big]\sin^2\theta.$1 with a power-law differential spectrum of index $A_\phi = B\Big[{\textstyle\frac{1}{2}(r^2+a^2)-a^2r\Sigma^{-1}(1+\cos^2\theta)\Big]\sin^2\theta.$2 (Desai et al., 2024). That work interprets the exhaust as an acceleration environment in which merging magnetic islands with guide field $A_\phi = B\Big[{\textstyle\frac{1}{2}(r^2+a^2)-a^2r\Sigma^{-1}(1+\cos^2\theta)\Big]\sin^2\theta.$3–$A_\phi = B\Big[{\textstyle\frac{1}{2}(r^2+a^2)-a^2r\Sigma^{-1}(1+\cos^2\theta)\Big]\sin^2\theta.$4 enable Fermi and betatron energization (Desai et al., 2024).

At 1 AU, magnetic-field turbulence within reconnection exhausts has been characterized using higher-order structure functions and the Jensen–Shannon complexity–entropy plane (Miranda et al., 2021). For all four analyzed events, the outflow-aligned $A_\phi = B\Big[{\textstyle\frac{1}{2}(r^2+a^2)-a^2r\Sigma^{-1}(1+\cos^2\theta)\Big]\sin^2\theta.$5 component had lower entropy and higher complexity than $A_\phi = B\Big[{\textstyle\frac{1}{2}(r^2+a^2)-a^2r\Sigma^{-1}(1+\cos^2\theta)\Big]\sin^2\theta.$6 and $A_\phi = B\Big[{\textstyle\frac{1}{2}(r^2+a^2)-a^2r\Sigma^{-1}(1+\cos^2\theta)\Big]\sin^2\theta.$7, implying that coherent structures preferentially organize turbulence along the exhaust direction (Miranda et al., 2021). This suggests that the exhaust is not merely a passive outflow but an intermittency-rich channel where reconnection and turbulence remain coupled.

4. Magnetic exhaust in fusion plasmas

In fusion research, the phrase refers to magnetically engineered exhaust of heat and particles from the confined plasma to the divertor. The physical variables are no longer X-point plasmoids or reconnection cavities but divertor leg length, flux expansion, radiation localization, neutral trapping, and the stability of detached states.

A central geometric quantity is flux expansion. In the MAST Upgrade divertor-shaping study, the total expansion is decomposed as

$A_\phi = B\Big[{\textstyle\frac{1}{2}(r^2+a^2)-a^2r\Sigma^{-1}(1+\cos^2\theta)\Big]\sin^2\theta.$8

with a poloidal metric

$A_\phi = B\Big[{\textstyle\frac{1}{2}(r^2+a^2)-a^2r\Sigma^{-1}(1+\cos^2\theta)\Big]\sin^2\theta.$9

and a field-gradient measure

$4\pi r^2\rho(r)v(r)=-\dot M,$0

used to parameterize total expansion between X point and target (Verhaegh et al., 2023). Three configurations were compared: Conventional Divertor, Elongated Divertor, and Super-X Divertor, with progressively larger target radius, larger flux expansion, and longer $4\pi r^2\rho(r)v(r)=-\dot M,$1 and $4\pi r^2\rho(r)v(r)=-\dot M,$2 (Verhaegh et al., 2023). Under $4\pi r^2\rho(r)v(r)=-\dot M,$3 L-mode conditions, detachment onset in the conventional divertor occurred near $4\pi r^2\rho(r)v(r)=-\dot M,$4, whereas the elongated and Super-X configurations were already detached below $4\pi r^2\rho(r)v(r)=-\dot M,$5 and exhibited much flatter detachment-front sensitivity to density (Verhaegh et al., 2023). Measured peak perpendicular heat-flux reductions were about $4\pi r^2\rho(r)v(r)=-\dot M,$6 for Super-X and $4\pi r^2\rho(r)v(r)=-\dot M,$7 for the elongated divertor relative to the conventional case, exceeding pure geometric expectations and indicating strong additional volumetric losses (Verhaegh et al., 2023).

The target power estimate used there is

$4\pi r^2\rho(r)v(r)=-\dot M,$8

with $4\pi r^2\rho(r)v(r)=-\dot M,$9 and $$v\frac{dv}{dr}+\frac{1}{\rho}\frac{dP}{dr}+\frac{GM}{r^2}=0,$$0 (Verhaegh et al., 2023). A reduced detachment criterion is written in terms of a control parameter

$$v\frac{dv}{dr}+\frac{1}{\rho}\frac{dP}{dr}+\frac{GM}{r^2}=0,$$1

with threshold

$$v\frac{dv}{dr}+\frac{1}{\rho}\frac{dP}{dr}+\frac{GM}{r^2}=0,$$2

Larger flux expansion and longer connection length therefore lower detachment thresholds (Verhaegh et al., 2023).

Neutral baffling is equally central. In TCV, a baffled Snowflake Minus Low-Field Side divertor was optimized to maximize SOL–baffle distance while maintaining heat-flux sharing between the two outer strike points (Gorno et al., 2022). The optimization doubled divertor neutral pressure relative to unbaffled Snowflake cases and reduced ion flux to baffle-facing walls to levels comparable to baffled single-null operation (Gorno et al., 2022). Without nitrogen seeding, the baffled SF-LFS reduced peak outer target heat flux by up to $$v\frac{dv}{dr}+\frac{1}{\rho}\frac{dP}{dr}+\frac{GM}{r^2}=0,$$3 relative to baffled single null, depending on $$v\frac{dv}{dr}+\frac{1}{\rho}\frac{dP}{dr}+\frac{GM}{r^2}=0,$$4 and diagnostic, and by about $$v\frac{dv}{dr}+\frac{1}{\rho}\frac{dP}{dr}+\frac{GM}{r^2}=0,$$5 at SP2 and $$v\frac{dv}{dr}+\frac{1}{\rho}\frac{dP}{dr}+\frac{GM}{r^2}=0,$$6 at SP4 relative to the unbaffled SF-LFS (Gorno et al., 2022). The LP-based parallel heat-flux decomposition used in that work is

$$v\frac{dv}{dr}+\frac{1}{\rho}\frac{dP}{dr}+\frac{GM}{r^2}=0,$$7

with

$$v\frac{dv}{dr}+\frac{1}{\rho}\frac{dP}{dr}+\frac{GM}{r^2}=0,$$8

and

$$v\frac{dv}{dr}+\frac{1}{\rho}\frac{dP}{dr}+\frac{GM}{r^2}=0,$$9

where $dP=c_s^2 d\rho$0 (Gorno et al., 2022). Yet the paper also reports a trade-off: despite an inter-null radiator farther from the core, the SF-LFS showed higher $dP=c_s^2 d\rho$1 than baffled single null in those L-mode conditions, suggesting greater core impurity penetration rather than improved core compatibility (Gorno et al., 2022).

A different fusion use of “exhaust” is the quasi-continuous exhaust scenario, in which small ELMs provide edge transport without large type-I transients (Harrer et al., 2021). There the relevant magnetic geometry is near the LCFS, governed by connection length between good- and bad-curvature regions, local magnetic shear, and $dP=c_s^2 d\rho$2 shear. Ideal ballooning stability is described through

$dP=c_s^2 d\rho$3

with local shear

$dP=c_s^2 d\rho$4

The marginality indicator is

$dP=c_s^2 d\rho$5

with $dP=c_s^2 d\rho$6 denoting ideal ballooning instability (Harrer et al., 2021). In this picture, the exhaust is “quasi-continuous” because ballooning-like fluctuations localized near the pedestal foot provide continuous heat and particle release while the steep-gradient region remains in second ballooning stability, preserving confinement (Harrer et al., 2021).

Recent divertor-design work extends the concept from diagnostic interpretation to rapid engineering evaluation. FIREFLY models heat and particle exhaust from reconstructed field-line geometry, solves a simplified transport equation,

$dP=c_s^2 d\rho$7

and tracks neutralized particles with EIRENE to estimate exhaust efficiency (Frerichs et al., 13 Apr 2026). It defines a heat-load proxy

$dP=c_s^2 d\rho$8

and a particle exhaust efficiency

$dP=c_s^2 d\rho$9

The W7-X example demonstrates how divertor geometry and pump-gap placement can be optimized under heat-load constraints to improve particle removal (Frerichs et al., 13 Apr 2026). This is a more explicitly engineering interpretation of magnetic exhaust: field-line geometry determines where power and neutrals go, and geometry can be numerically optimized for better exhaust performance.

5. Magnetic nozzles and directed plasma acceleration

In plasma propulsion and open mirror applications, the magnetic exhaust effect has a narrower and more literal meaning: acceleration of plasma along a converging–diverging magnetic field through conversion of thermal, azimuthal magnetic, and rotational energy into axial kinetic energy (Smolyakov et al., 24 Jun 2025). The model is one-fluid ideal MHD with continuity, momentum, induction, and a polytropic closure,

$$\frac{1}{2}\left(1-\frac{c_s^2}{v^2}\right)\frac{d\,v^2}{dr} = -\frac{GM}{r^2}\left[1-\frac{2c_s^2r}{GM}\right], \qquad r_s=\frac{GM}{2c_s^2(r_s)}.$$0

$$\frac{1}{2}\left(1-\frac{c_s^2}{v^2}\right)\frac{d\,v^2}{dr} = -\frac{GM}{r^2}\left[1-\frac{2c_s^2r}{GM}\right], \qquad r_s=\frac{GM}{2c_s^2(r_s)}.$$1

$$\frac{1}{2}\left(1-\frac{c_s^2}{v^2}\right)\frac{d\,v^2}{dr} = -\frac{GM}{r^2}\left[1-\frac{2c_s^2r}{GM}\right], \qquad r_s=\frac{GM}{2c_s^2(r_s)}.$$2

$$\frac{1}{2}\left(1-\frac{c_s^2}{v^2}\right)\frac{d\,v^2}{dr} = -\frac{GM}{r^2}\left[1-\frac{2c_s^2r}{GM}\right], \qquad r_s=\frac{GM}{2c_s^2(r_s)}.$$3

In paraxial axisymmetry, the field and velocity decompose into parallel and azimuthal parts, and mass-flux conservation along a flux tube gives

$$\frac{1}{2}\left(1-\frac{c_s^2}{v^2}\right)\frac{d\,v^2}{dr} = -\frac{GM}{r^2}\left[1-\frac{2c_s^2r}{GM}\right], \qquad r_s=\frac{GM}{2c_s^2(r_s)}.$$4

The axial momentum balance contains the terms responsible for nozzle acceleration:

$$\frac{1}{2}\left(1-\frac{c_s^2}{v^2}\right)\frac{d\,v^2}{dr} = -\frac{GM}{r^2}\left[1-\frac{2c_s^2r}{GM}\right], \qquad r_s=\frac{GM}{2c_s^2(r_s)}.$$5

In a diverging section, the field gradient couples both centrifugal and magnetic-pressure/ponderomotive effects into axial acceleration (Smolyakov et al., 24 Jun 2025). The corresponding Bernoulli-like invariant is

$$\frac{1}{2}\left(1-\frac{c_s^2}{v^2}\right)\frac{d\,v^2}{dr} = -\frac{GM}{r^2}\left[1-\frac{2c_s^2r}{GM}\right], \qquad r_s=\frac{GM}{2c_s^2(r_s)}.$$6

The last term makes the physical interpretation explicit: as $$\frac{1}{2}\left(1-\frac{c_s^2}{v^2}\right)\frac{d\,v^2}{dr} = -\frac{GM}{r^2}\left[1-\frac{2c_s^2r}{GM}\right], \qquad r_s=\frac{GM}{2c_s^2(r_s)}.$$7 and/or $$\frac{1}{2}\left(1-\frac{c_s^2}{v^2}\right)\frac{d\,v^2}{dr} = -\frac{GM}{r^2}\left[1-\frac{2c_s^2r}{GM}\right], \qquad r_s=\frac{GM}{2c_s^2(r_s)}.$$8 decrease along the diverging field, conservation of $$\frac{1}{2}\left(1-\frac{c_s^2}{v^2}\right)\frac{d\,v^2}{dr} = -\frac{GM}{r^2}\left[1-\frac{2c_s^2r}{GM}\right], \qquad r_s=\frac{GM}{2c_s^2(r_s)}.$$9 requires $${\rm d}s^{2} = -\frac{\Delta\Sigma}{A}\,{\rm d}t^{2} +\frac{\Sigma}{\Delta}\,{\rm d}r^{2} +\Sigma\,{\rm d}\theta^{2} +\frac{A\sin^2\theta}{\Sigma} \left({\rm d}\phi-\omega\,{\rm d}t\right)^{2},$$00 to increase (Smolyakov et al., 24 Jun 2025).

The paper also derives auxiliary invariants, including

$${\rm d}s^{2} = -\frac{\Delta\Sigma}{A}\,{\rm d}t^{2} +\frac{\Sigma}{\Delta}\,{\rm d}r^{2} +\Sigma\,{\rm d}\theta^{2} +\frac{A\sin^2\theta}{\Sigma} \left({\rm d}\phi-\omega\,{\rm d}t\right)^{2},$$01

$${\rm d}s^{2} = -\frac{\Delta\Sigma}{A}\,{\rm d}t^{2} +\frac{\Sigma}{\Delta}\,{\rm d}r^{2} +\Sigma\,{\rm d}\theta^{2} +\frac{A\sin^2\theta}{\Sigma} \left({\rm d}\phi-\omega\,{\rm d}t\right)^{2},$$02

and identifies slow, Alfvén, and fast critical points where the regular transonic/trans-Alfvénic solution must pass smoothly (Smolyakov et al., 24 Jun 2025). A central result is that for sufficiently large field expansion, the flow can become trans-Alfvénic, and time-dependent simulations converge to the unique regular stationary branch (Smolyakov et al., 24 Jun 2025).

This usage is conceptually the clearest example of “magnetic exhaust effect” as energy conversion into an exhaust beam. It also differs from the reconnection and black-hole cases by lacking current-sheet-mediated release; the acceleration is continuous and nozzle-like, though the same ingredients—magnetic geometry, azimuthal field, and magnetic stresses—remain central.

6. Common structure, contrasts, and misconceptions

The strongest commonality across the literature is that the exhaust is a magnetically selected outflow channel. Magnetic geometry determines both the existence and the efficiency of the outflow. In Kerr accretion, vertical flux accumulation and midplane reconnection create an equatorial escape route (Karas et al., 2020). In guide-field reconnection, the exhaust is a density-depleted kinetic structure whose velocity-space anisotropies and $${\rm d}s^{2} = -\frac{\Delta\Sigma}{A}\,{\rm d}t^{2} +\frac{\Sigma}{\Delta}\,{\rm d}r^{2} +\Sigma\,{\rm d}\theta^{2} +\frac{A\sin^2\theta}{\Sigma} \left({\rm d}\phi-\omega\,{\rm d}t\right)^{2},$$03 energization are controlled by the field configuration (Treumann et al., 2017, McCubbin et al., 2021). In divertors, flux expansion, connection length, and neutral trapping determine where power and particles leave the confined plasma (Gorno et al., 2022, Verhaegh et al., 2023, Frerichs et al., 13 Apr 2026). In magnetic nozzles, a diverging flux tube and the coupling of $${\rm d}s^{2} = -\frac{\Delta\Sigma}{A}\,{\rm d}t^{2} +\frac{\Sigma}{\Delta}\,{\rm d}r^{2} +\Sigma\,{\rm d}\theta^{2} +\frac{A\sin^2\theta}{\Sigma} \left({\rm d}\phi-\omega\,{\rm d}t\right)^{2},$$04 and $${\rm d}s^{2} = -\frac{\Delta\Sigma}{A}\,{\rm d}t^{2} +\frac{\Sigma}{\Delta}\,{\rm d}r^{2} +\Sigma\,{\rm d}\theta^{2} +\frac{A\sin^2\theta}{\Sigma} \left({\rm d}\phi-\omega\,{\rm d}t\right)^{2},$$05 define the conversion of stored electromagnetic energy into directed thrust (Smolyakov et al., 24 Jun 2025).

A frequent misconception is to equate all magnetic exhaust phenomena with jets. The black-hole study explicitly distinguishes its equatorial exhaust from a Blandford–Znajek polar jet (Karas et al., 2020). Fusion exhaust is not a jet at all, but a controlled sink of power and particles to surfaces or pumps (Gorno et al., 2022, Verhaegh et al., 2023, Frerichs et al., 13 Apr 2026). Reconnection exhausts are sometimes treated as purely Alfvénic outflows bounded by simple slow shocks, but both simulation and observation show more complicated compound structures, anisotropy-controlled transitions, and sub-Alfvénic jets when firehose physics is important (Liu et al., 2011, Haggerty et al., 2018, Sun et al., 28 Apr 2026).

Another misconception is that magnetic exhaust is always a steady process. In fact, intermittency is common. The Kerr equatorial exhaust is plasmoid-mediated and variable (Karas et al., 2020). Reconnection exhausts show evolving shock normals, mirrored field–particle signatures, mirror modes, and coherent turbulent structures (McCubbin et al., 2021, Miranda et al., 2021, Sun et al., 28 Apr 2026). Fusion exhaust control seeks to stabilize detachment or make edge release quasi-continuous precisely because unsteady exhaust is operationally problematic (Harrer et al., 2021, Verhaegh et al., 2023).

7. Open problems and broader significance

Several unresolved themes recur across fields. One is dimensionality. The Kerr exhaust calculation is axisymmetric and ideal; 3D GRMHD with non-axisymmetric instabilities and non-ideal reconnection remains necessary to determine whether the equatorial exhaust remains robust in more realistic accretion flows (Karas et al., 2020). Strong-guide-field exhaust analyses identify compelling kinetic signatures, but the dependence of ECMI growth, escape, and saturation on realistic 3D exhaust distributions remains open (Treumann et al., 2017). In space plasmas, the compound nature of reconnection exhaust boundaries, the role of turbulence, and the relation between island trapping and energetic-particle spectra all require broader statistical characterization (Miranda et al., 2021, Sun et al., 28 Apr 2026, Desai et al., 2024).

A second theme is closure physics. Fusion exhaust predictions depend sensitively on neutral trapping, cross-field transport, impurity behavior, and radiative loss channels; simplified models such as FIREFLY are useful precisely because full coupled simulations remain expensive (Frerichs et al., 13 Apr 2026). Yet the TCV and MAST-U results also show that geometry alone is insufficient: neutral closure can either unlock or suppress the benefits of magnetic shaping (Gorno et al., 2022, Verhaegh et al., 2023). In propulsion, ideal-MHD nozzle theory produces robust stationary branches, but detachment, resistivity, turbulence, and two-fluid effects remain critical for practical thruster design (Smolyakov et al., 24 Jun 2025).

A third theme is energy partition. In every magnetic exhaust problem, the core question is not simply whether an exhaust forms, but which reservoir feeds it: magnetic pressure and tension near a black hole (Karas et al., 2020); $${\rm d}s^{2} = -\frac{\Delta\Sigma}{A}\,{\rm d}t^{2} +\frac{\Sigma}{\Delta}\,{\rm d}r^{2} +\Sigma\,{\rm d}\theta^{2} +\frac{A\sin^2\theta}{\Sigma} \left({\rm d}\phi-\omega\,{\rm d}t\right)^{2},$$06 and kinetic anisotropy in reconnection (McCubbin et al., 2021); radiative, molecular, and neutral-mediated losses in divertors (Gorno et al., 2022, Verhaegh et al., 2023); or azimuthal magnetic and swirl energy in a nozzle (Smolyakov et al., 24 Jun 2025). This suggests that the term’s greatest value is comparative: it foregrounds how magnetic structures organize exhaust pathways and redistribute energy across radically different plasma environments.

In that sense, the magnetic exhaust effect is less a single phenomenon than a recurring plasma-physical motif. Whenever magnetic topology creates a privileged outlet through which matter, energy, heat, or radiation escapes—and whenever magnetic stresses or magnetically controlled microphysics determine that outlet’s efficiency—the literature now repeatedly reaches for the same language.