Project Matterhorn

Lyman Spitzer with the figure-eight-shaped Model A stellarator.

The stellarator concept predates the tokamak. Lyman Spitzer proposed it in 1951 as a way to confine plasma using external magnetic fields. His first stellarator, Model A, was built in 1953 as part of Project Matterhorn at Princeton. It consisted of a borosilicate-glass tube bent into a figure-eight shape and surrounded by magnetic windings, providing an early experimental test of stellarator plasma confinement.

The subsequent Model B series remained relatively small and was used to investigate plasma heating and confinement in more detail. Ohmic heating in Model B-1 produced plasma temperatures of about 100,000 K. Later Model B devices explored improved magnetic configurations and plasma exhaust concepts, including early divertor arrangements.

After the Model B series, the much larger Model C stellarator began operation in 1962. It became an important test bed for plasma transport and auxiliary heating techniques, including neutral beam heating and ion cyclotron resonance heating. These methods later became standard tools in magnetic-confinement fusion research. By the late 1960s, Model C had achieved plasma temperatures of several million kelvin.

Diffusion Problems

Early stellarators suffered from unexpectedly large particle and energy losses, with measured transport far exceeding classical predictions. These confinement problems cast doubt on the practical potential of the concept. Following the striking success of Soviet tokamaks, PPPL decided in 1969 to convert Model C into the Symmetric Tokamak (ST).

Despite these difficulties, stellarator research continued in several countries during the 1960s and 1970s. New configurations such as torsatrons and heliotrons were developed, using helical windings and additional shaping coils to improve the magnetic field. Experiments including the CLEO devices at Culham demonstrated that stellarator confinement could be substantially improved. However, further progress increasingly depended on better plasma theory, numerical modelling and more precise magnetic-field optimisation.

Modular Stellarator

First hydrogen plasma in the Wendelstein 7-X stellarator on 3 February 2016.

A major step came in 1988 with Wendelstein 7-AS in Garching, Germany, the first of a new generation of Advanced Stellarators. Its magnetic field was produced by 45 specially shaped non-planar modular coils calculated using numerical optimisation. The experiment demonstrated that such complex coil systems could be manufactured accurately and could produce the intended improved magnetic configuration. Wendelstein 7-AS also became the first stellarator to operate with a divertor, providing important experience for its successor, Wendelstein 7-X.

Second Generation

A new generation of large and optimised stellarators emerged in the late 1990s. Japan’s Large Helical Device (LHD), based on the heliotron concept, began operation in 1998. Around the same time, the Helically Symmetric eXperiment (HSX) was developed at the University of Wisconsin-Madison to test quasi-helical symmetry, a magnetic configuration designed to reduce neoclassical particle losses.

LHD subsequently demonstrated one of the key advantages of the stellarator concept: very long plasma operation, with discharges lasting for tens of minutes and several gigajoules of injected heating energy during long-pulse experiments.

Wendelstein 7-X in Greifswald, Germany, produced its first plasma in December 2015. The fully optimised modular stellarator was built to test reactor-relevant confinement and long-pulse operation. Subsequent upgrades enabled progressively longer and higher-energy plasma discharges.

Achievements

Wendelstein 7-X progressively extended both plasma performance and pulse duration. In 2017 it achieved a stellarator record for the fusion triple product. In 2023 it maintained a plasma for eight minutes with an energy turnover of 1.3 GJ. In May 2025 it reached 1.8 GJ during a 360-second plasma and set a new record for the triple product in long-duration discharges, bringing stellarator performance into the range previously achieved only by tokamaks for such long pulses.

LHD began deuterium plasma experiments in 2017 and achieved ion temperatures of about 120 million kelvin. Long-pulse operation later reached 2,859 seconds, or almost 48 minutes, with total injected heating energy of 3.36 GJ. After more than 27 years of operation, the LHD experimental programme was completed in December 2025.