The FLARE Facility

Abstract: The Facility for Laboratory Reconnection Experiments (FLARE) has been constructed to study magnetic reconnection in multiple X-line regimes relevant to space, astrophysical, and fusion plasmas. Building upon the successful design of the Magnetic Reconnection Experiment (MRX), FLARE features a larger physical volume, stronger magnetic fields, and an independent ohmic heating drive to significantly extend the accessible parameter space, targeting Lundquist numbers up to S ~ 10^5 and normalized system sizes up to λ~ 10^3. This paper details the facility's core engineering components, including the primary vacuum vessel, internal flux cores, highly segmented external coil systems, modular capacitor banks, and the safety interlock and control architecture. An initial diagnostic suite is presented, comprising high-resolution 2D magnetic probe arrays, triple Langmuir probes, a fully fiber-coupled interferometer, ion Doppler spectroscopy, and fast camera imaging. Initial operations demonstrate the device's experimental flexibility and reliability, successfully executing symmetric push-pull reconnection, spheromak merging, and asymmetric downstream configurations. Currently operating within "Stage 2.5" with S ~ 2,500 and λ~ 60 for anti-parallel reconnection, FLARE provides immediate access to the multiple X-line regimes. Planned hardware upgrades, advanced diagnostic additions, and integration with fully kinetic simulations will further expand its capabilities as it transitions into a collaborative user facility for the broader plasma science community.

• The paper presents FLARE, a purpose-built reconnection facility that expands accessible parameter space to target plasmoid-dominated, multiple X-line regimes relevant to space, astrophysical, and fusion plasmas.
• The facility combines 6.3 MJ capacitor banks, extensive magnetic-probe coverage, high-speed data acquisition, and staged operation toward Lundquist number S ≈ 100,000 and normalized system size λ ≈ 1,000.
• Initial experiments demonstrate symmetric push–pull reconnection, spheromak merging, and asymmetric reconnection, while kinetic VPIC simulations reproduce key magnetic topologies and support integrated experiment–simulation studies.

The Facility for Laboratory Reconnection Experiments (FLARE), described in "The FLARE Facility" (2608.17332), is a purpose-built laboratory device designed to study magnetic reconnection in the multiple X-line regimes that characterize most space, astrophysical, and fusion plasmas. Constructed at Princeton Plasma Physics Laboratory as a scaled successor to the Magnetic Reconnection Experiment (MRX), FLARE extends the accessible dimensionless parameter space by roughly two orders of magnitude in the Lundquist number $S$ and one order of magnitude in the normalized system size $\lambda$, providing laboratory access to plasmoid-dominated reconnection physics that has previously been reachable only through theory and simulation.

Motivation and scientific context

Magnetic reconnection research has progressed through three phases: MHD-scale models, kinetic (non-MHD) mechanisms of fast reconnection, and, most recently, the recognition that Sweet–Parker current sheets become unstable to plasmoid formation above a critical Lundquist number $S_c \approx 10^4$, producing multiple X-lines whose reconnection rate is approximately $R \sim S_c^{-1/2} \approx 0.01$ and independent of $S$. Because kinetic mechanisms operate only on ion or electron scales, a central open question is how MHD-scale dynamics generate thin current sheets and when this occurs. FLARE is designed to answer this question experimentally by spanning the collisional, collisionless, and hybrid multiple X-line regimes identified in the reconnection phase diagram, where $\lambda \equiv L/\rho_s$ (or $L/d_i$ for anti-parallel reconnection) and $S \equiv \mu_0 L V_A / 4\eta_{\text{Spitzer}}$. The paper maps a wide range of natural and fusion plasmas onto this phase diagram and positions FLARE to test whether the theoretically predicted regime boundaries—potentially refined by electron pressure anisotropy and other effects—are accurate. The project also responds directly to community planning documents from 2010, 2016, and 2020.

Design and engineering

FLARE retains the two defining features of MRX: a pair of internal flux cores that generate and control the reconnecting field, and extensive in-situ magnetic probe coverage. The key parameters relative to MRX are summarized below:

Parameter	MRX	FLARE
Device diameter / length	1.5 m / 2 m	3 m / 3.6 m
Flux core separation $L$	0.8 m	1.6 m
Guide field at $R = R_0$	0.1 T	0.5 T
Ohmic drive	none	0.5 V·s
Capacitor bank energy	~30 kJ	~6.3 MJ
$\lambda$0 (guide field)	2,900	100,000
$\lambda$1	140	1,000

The vacuum vessel is a fully welded Type 304 stainless-steel cylinder (2.46 m length, 3.07 m diameter, 9.4 mm wall) with a magnetic diffusion time of approximately 10 ms through the wall. A multi-stage pumping system of turbomolecular and cryogenic pumps, combined with an extensive differential pumping manifold for the flux cores and center stack, has already achieved a base pressure of $\lambda$2 Torr, exceeding the $\lambda$3 Torr target. The center stack houses the ohmic heating (OH) solenoids and 48 water-cooled guide-field (GF) conductors in a composite fiberglass/PVC structure rated for 36,000 ft-lb torque.

Each flux core contains independently powered poloidal-field (PF) and toroidal-field (TF) windings, with peak design currents of 540 kA (PF) and 250 kA (TF). Achieved current rise times are 25–50% longer than design minimums due to stray busbar inductance, a concession the authors note explicitly. Finite-element validation with ANSYS Maxwell shows good agreement between measured vacuum poloidal-field profiles and calculations, confirming that vessel eddy currents are accurately captured for the PF system; by contrast, the externally mounted equilibrium-field (EF) coils exhibit a roughly 10 ms lag between peak coil current and peak internal field, which must be compensated in waveform programming. The GF system currently reaches 0.16 T at midplane, with 0.5 T planned, and shows negligible vessel-eddy-current distortion.

Power is supplied by 11 modular capacitor banks (total 6.3 MJ at Stage 3) with common circuit topologies using NL8900 ignitrons, crowbar protection, and dump resistors. The facility is staged: Stage 1 (first plasma) and Stage 2 are defined, and the device currently operates in Stage 2.5, with 2.82 MJ available, because the drive coils are not yet installed and the GF coils are limited to a single bank by pending structural bracing.

Control, data, and safety

The control system is built on EPICS with pvAccess, Python-based software, and FPGA-implemented microsecond-scale protection logic, replacing legacy LabView infrastructure. Timing resolution is 10 ns across 32 channels. The data acquisition system comprises 26 D-TACQ units providing 1,248 channels at up to 40 MHz, with 1,200 channels using active analog integrators (phase error $\lambda$4 from 200 kHz to 5 MHz), synchronized by CERN White-Rabbit switches and stored in a hierarchical HDF5 scheme. The interlock system employs NRTL-listed trapped-key methodology with monitored egress and redundant power, forming the independent protection layers credited in the facility's LOPA.

Initial diagnostics

The initial suite consists of five instruments: a 2D magnetic probe array, modified triple Langmuir probes, a fiber-coupled interferometer, ion Doppler spectroscopy (IDS), and a fast camera. The magnetic probe array is the most distinctive element: each probe spans 1 m radially with 129 miniature coils in triplets, achieving 0.5 cm resolution over the inner 28 cm and 4 cm spacing at the extremes, with 15 probes at 3 cm spacing in $\lambda$5. This coverage permits direct computation of current density via Ampère's law and the poloidal flux function without separate background probes, and the 40 ns probe response resolves Alfvénic and lower-hybrid timescales. The triple Langmuir probe uses dual floating tips to cancel systematic $\lambda$6 errors from local electric fields up to 1000 V/m, and carries a stated 10–20% uncertainty. The 1550 nm heterodyne Michelson interferometer replaces MRX's vibration-prone CO$\lambda$7 system with a fully fiber-coupled design. IDS uses a high-throughput F/1.8 volume-phase-holographic spectrometer with 80 collection fibers and 1 cm spatial resolution, enabling tomographic inversion of ion temperature and flow under an axisymmetry assumption. A second phase of diagnostics—Mach probes, fluctuation probes, dipole probes, electron energy analyzers, soft x-ray tomography, Thomson scattering, and LIF—is planned to resolve kinetic-scale physics.

Demonstrated operating modes

Initial operations demonstrate three configurations, all in counter-helicity Case-I geometry:

Symmetric push–pull reconnection in deuterium shows a push current sheet (positive $\lambda$8, radially elongated) transitioning to a pull sheet (negative $\lambda$9, axially elongated) with a clear quadrupole $S_c \approx 10^4$0 structure—a two-fluid Hall signature—during the pull phase. OH-driven discharges show an approximately 50% increase in electron temperature, confirming the OH system's effectiveness as an auxiliary drive.

Spheromak merging in argon, produced by delayed PF crowbar and current reversal, yields two pinch-off spheromaks that merge through a push-type current sheet. Notably, the merged configuration retains substantial reversed toroidal field and relaxes toward a reversed-field-pinch-like state rather than the field-reversed configuration observed in the analogous MRX experiment, with total merged poloidal flux exceeding 25 mWb. The paper identifies this distinct relaxation pathway as a subject requiring further investigation.

Asymmetric downstream reconnection in helium, driven by a 10% difference in PF bank charge voltage, produces pronounced left–right asymmetry in current-sheet structure and $S_c \approx 10^4$1 topology, with the weaker-pressure side developing the more evident quadrupole. Whether this asymmetry corresponds to asymmetric ion exhaust remains an open measurement, planned via Mach probes.

In Stage 2.5 the device routinely achieves $S_c \approx 10^4$2 and $S_c \approx 10^4$3 for anti-parallel reconnection ($S_c \approx 10^4$4–$S_c \approx 10^4$5 m$S_c \approx 10^4$6, $S_c \approx 10^4$7–15 eV, $S_c \approx 10^4$8 T, $S_c \approx 10^4$9 m), already providing access to the multiple X-line collisionless regime that is the stated focus of initial research.

Simulation integration

The facility program couples experiments with fully kinetic VPIC simulations run in realistic cylindrical geometry with measured PF and TF current waveforms as inputs. A baseline simulation reproduces the global magnetic topology and resolves a central plasmoid flanked by multiple X-points, demonstrating that first-principles kinetic modeling of FLARE discharges—including Coulomb collisions—is feasible across the collisionless-to-collisional range. This capability is central to the facility's strategy of testing phase-diagram predictions against both experiment and simulation.

Limitations and open questions

Several constraints qualify the present results. The drive coils are designed but not fabricated or installed, limiting the achievable reconnection electric field and hence $R \sim S_c^{-1/2} \approx 0.01$0; the GF structural bracing is incomplete, capping the guide field at 0.16 T rather than the 0.5 T design value and limiting $R \sim S_c^{-1/2} \approx 0.01$1. Global plasma currents were found to affect the TF and PF circuits more strongly than in MRX, so analytic current predictions serve only as guidance. The IDS tomographic inversion assumes axisymmetry, which may fail in the asymmetric configurations FLARE is designed to study. The spheromak merging relaxation pathway differs from MRX and is not yet explained. More broadly, the paper enumerates ten major research problems—spanning multiscale coupling, 3D dynamics, energy partition, boundary conditions, onset, partial ionization, explosive phenomena, flow-driven dynamos, turbulence and shocks, and extreme conditions—and concedes that the last three require platforms beyond FLARE, such as high-energy-density laser or pulsed-power systems, potentially co-located with the device.

Conclusion

The FLARE facility paper documents the construction, commissioning, and initial operation of a laboratory platform that extends reconnection research into the multiple X-line regimes relevant to most natural and fusion plasmas. With demonstrated operational flexibility across push–pull, merging, and asymmetric configurations, a validated engineering basis, a staged path to $R \sim S_c^{-1/2} \approx 0.01$2 and $R \sim S_c^{-1/2} \approx 0.01$3, and tight coupling to first-principles kinetic simulation, FLARE is positioned to test the theoretical phase diagram of magnetic reconnection and to operate as a DOE collaborative user facility. The central scientific question it addresses—whether the predicted division of reconnection regimes is correct, and how plasmoid-mediated dynamics couple MHD scales to kinetic dissipation—remains open and is now experimentally addressable.