- Atlas
- Engineering
- Fusion Engineering
Fusion Engineering
Fusion engineering is the branch of nuclear engineering that tries to build a machine that can reproduce, on Earth and under control, the reaction that powers the Sun.
11 frameworks, from 1951 to today · Nuclear Engineering · Engineering
- Magnetic confinement fusion 1951-Present
- Stellarator 1951-Present
- Lawson criterion framework 1957-Present
- Inertial confinement fusion 1960-Present
- Tokamak 1968-Present
- Fusion nuclear materials engineering 1980-Present
- Tritium breeding blanket engineering 1980-Present
- Remote handling and maintenance engineering 1990-Present
- Fusion power plant systems integration 2000-Present
- High-temperature superconducting magnet engineering 2015-Present
- Private compact fusion development 2015-Present
Overview
Fusion engineering is the branch of nuclear engineering that tries to build a machine that can reproduce, on Earth and under control, the reaction that powers the Sun12. In that reaction, light atomic nuclei—most often two forms of hydrogen, deuterium and tritium—fuse into a heavier nucleus, releasing enormous energy1. The fuel is abundant, the reaction cannot run away the way a fission chain reaction can, and its main waste is short-lived. What is at stake is a power source that could supply large amounts of carbon-free energy for a very long time2. The difficulty is that fusion only happens at temperatures of over a hundred million degrees, so the fuel becomes a plasma—an ionized gas of charged particles—that no material wall can touch13. Fusion engineering is the long, unfinished effort to hold that plasma, heat it, fuel it, extract its heat, and maintain the machine.
The idea and its first machines, 1920s–1950s
The astrophysicist Arthur Eddington proposed in the 1920s that stars are powered by the fusion of hydrogen into helium. In 1934, Ernest Rutherford, Mark Oliphant, and Paul Harteck fused deuterium nuclei in a laboratory, confirming that the reaction releases energy. The physicist George Gamow and others worked out the rates at which such reactions proceed at given temperatures, and by the early 1950s the conditions for a practical reactor were roughly understood: a plasma hot enough, dense enough, and confined long enough. The product of density and confinement time became the standard yardstick, later called the Lawson criterion after John Lawson, who published a systematic assessment in 1957.
The problem was confinement. Two routes were pursued. In magnetic confinement, strong magnetic fields hold the charged plasma away from the walls1. In inertial confinement, a tiny fuel pellet is compressed and heated so violently that it fuses before it can fly apart1. Both were explored in the 1950s, much of it in secret3. In the United States, Lyman Spitzer built the first stellarator at Princeton in 1951; a stellarator is a torus—a doughnut-shaped vessel—whose confining field is produced entirely by external coils twisted into complex shapes14. In the Soviet Union, Andrei Sakharov and Igor Tamm proposed the tokamak in the early 1950s: a torus in which a large current driven through the plasma itself generates most of the confining field. In Britain, the ZETA machine at Harwell, a torus using a powerful current pulse, was announced in 1958 as a possible fusion success; the claim did not survive scrutiny, and the episode taught the field how hard plasma measurement would be.
The tokamak becomes the standard, 1960s–1980s
Through the 1960s the stellarator struggled with plasma losses, while the tokamak's simpler, more symmetric field proved far better at holding heat4. The decisive moment came in 1968, when Soviet researchers reported plasma temperatures of several million degrees in their T-3 tokamak15. A British team led by Derek Robinson confirmed the result with their own laser measurement in 1969. From then on the tokamak dominated, and laboratories worldwide converted or built machines around it5.
The 1970s and 1980s were the era of scaling up. The Princeton Large Torus reached temperatures around 60 million degrees in the late 1970s, and neutral-beam heating—firing energetic neutral atoms into the plasma—became a standard tool. The Tokamak Fusion Test Reactor at Princeton, operating from 1982, was the first machine designed to use deuterium–tritium fuel; in the mid-1990s it produced several megawatts of fusion power. In Europe, the Joint European Torus (JET), completed in 1983, was built as a coordinated European facility and in 1991 carried out the first controlled fusion experiment to produce significant power from deuterium–tritium fuel. In 1997 JET reached about 16 megawatts of fusion power, a record that stood for decades. The Soviet T-15 and later machines pushed superconducting magnets, which use coils cooled to very low temperature so that they carry current with no resistance and can run continuously.
Stellarators, inertial fusion, and the materials problem
The stellarator never disappeared. Its advantage is that it can run steadily without the large current that a tokamak needs, and it does not suffer the tokamak's current-driven instabilities. Its disadvantage is that the twisted coils are hard to design and build. The German Wendelstein 7-X, which began operation in 2015, was designed with computer-optimized coil shapes to control plasma behavior precisely; it demonstrated long, stable discharges and showed that optimization could make stellarators competitive again.
Inertial confinement developed in parallel, driven largely by nuclear-weapons research because the compression physics overlaps with weapons physics6. The idea is to focus lasers or other intense beams onto a millimeter-scale pellet, imploding it to enormous density and temperature. The Nova laser at Lawrence Livermore National Laboratory, operating in the 1980s, and its successor the National Ignition Facility, completed in 2009, pursued this. In December 2022 the National Ignition Facility achieved a fusion reaction that released more energy than the laser energy delivered to the target—a scientific milestone, though the facility's total electricity draw is far larger than the fusion yield16. Inertial fusion energy as a power source remains a separate engineering challenge: targets must be cheap, repetition rates high, and the chamber able to survive.
Meanwhile, the materials problem became central. A fusion reactor's inner walls and blanket face intense neutron bombardment, and the neutrons carry most of the fusion energy1. Fusion nuclear materials engineering studies how structural materials—reduced-activation steels, vanadium alloys, silicon carbide composites—swell, embrittle, and change under that flux. Because no fusion machine yet produces the full neutron environment, much of this work uses fission reactors and ion beams as stand-ins, and the planned International Fusion Materials Irradiation Facility would provide a dedicated neutron source.
Breeding the fuel: tritium and the blanket
Deuterium is plentiful in seawater, but tritium is scarce and radioactive, with a half-life of about twelve years. A power plant must therefore breed its own tritium by surrounding the plasma with a blanket containing lithium, where fusion neutrons convert lithium into tritium1. Tritium breeding blanket engineering designs that blanket: which lithium compound to use (solid ceramics or liquid metals), how to extract the tritium, and how to recover the heat. The blanket must also shield the magnets and structure, so it is simultaneously a heat exchanger, a fuel factory, and a shield. No fusion device has yet demonstrated a closed tritium fuel cycle at power-plant scale, and breeding blankets are among the least-tested major components.
Remote handling and the integrated plant
A fusion plant will be radioactive, so maintenance must be done by machine. Remote handling and maintenance engineering designs the manipulators, transporters, and viewing systems that can replace blanket modules and divertor components inside a sealed vessel. JET pioneered this with a remote handling system used for major in-vessel work, and ITER's maintenance concept depends on it entirely.
As the machines grew, fusion power plant systems integration became a discipline of its own: fitting the plasma-facing components, magnets, heating systems, vacuum pumping, tritium plant, and heat conversion into one plant that can run reliably. The ITER project, agreed in 2006 by seven partners and under construction in France, is the largest expression of this integration effort2. It is designed to demonstrate sustained fusion power at the scale of hundreds of megawatts and to test breeding blanket modules, but it is not a electricity-producing plant.
Superconducting magnets and the private era
High-temperature superconducting magnet engineering transformed the field's prospects. Conventional superconductors such as niobium-tin must be cooled to near absolute zero with liquid helium; high-temperature superconductors, particularly rare-earth barium copper oxide tapes, can operate at higher temperatures and, crucially, carry much stronger fields in compact coils. Stronger fields allow a smaller, cheaper tokamak for the same performance. The MIT-led SPARC design, announced in the 2010s, was built around this idea, and in 2021 a large high-temperature superconducting magnet built by the MIT–Commonwealth Fusion Systems collaboration reached a field of about 20 tesla, a landmark for the approach.
This fed a wave of private compact fusion development. Commonwealth Fusion Systems, Tokamak Energy, General Fusion, TAE Technologies, and others raised substantial private capital from the 2010s onward, pursuing smaller machines and faster timelines than government projects. Their designs span compact tokamaks, spherical tokamaks, stellarators, and alternative concepts. The claims are ambitious and unproven; what is new is that private capital now funds engineering at a scale once reserved for states.
Where the field stands
Today the tokamak remains the most developed concept, with ITER as its flagship and a generation of superconducting machines behind it. Stellarators have returned as a serious alternative, aided by optimized coil design4. Inertial fusion has proved ignition is possible and now faces the engineering of repetition-rate targets and chambers. The materials and tritium-breeding problems remain the least mature parts of the chain, and no machine has yet produced net electricity. The open questions are concrete: which confinement concept scales best, how to build blankets that breed tritium reliably, how to make plasma-facing components survive years of neutron damage, and whether compact high-field machines can deliver on their promise. The history above is the record of a field that has repeatedly turned hard physics into engineering problems—and has not yet finished the last one.
References
- Fusion power (introduction, Mechanism, Magnetic confinement, Inertial confinement, First tokamak, 2020s, Deuterium, tritium, Energy capture), Wikipedia
- ITER (Background, introduction), Wikipedia
- Magnetic confinement fusion, Wikipedia
- Stellarator, Wikipedia
- Joint European Torus (Background), Wikipedia
- Inertial confinement fusion, Wikipedia
The frameworks of Fusion Engineering
Each framework is a way of working the field has used: when it held, and the question it answers.
- 1951-Present
- 1951-Present
- 1957-Present
- 1960-Present
- 1968-Present
- 1980-Present
- 1980-Present
- 1990-Present
- 2000-Present
- 2015-Present
- 2015-Present

Fusion Engineering is a planet in Nousfarer
Its city has a tower for each of these 11 frameworks. Land, walk in, and climb one concept at a time.
Land on Fusion EngineeringRelated fields
More in Nuclear Engineering
- Nuclear Engineering
- Nuclear Fuel Cycle And Waste
- Nuclear Safety And Radiation Protection
- Reactor Physics And Design
See where Fusion Engineering sits on the Map of Human Knowledge