HTS coils for a magnetic mirror are a pair of compact 20-tesla REBCO magnets that 📝Commonwealth Fusion Systems designed and built for an axial mirror device rather than a tokamak — published magnet work on a non-tokamak confinement concept.
Two papers in IEEE Transactions on Applied Superconductivity, two years apart, mark the distance between a drawing and a delivered magnet. 📝Alexey Radovinsky and eight co-authors published the preliminary design in 2023 (33(5), 10.1109/TASC.2023.3240377): a conduction-cooled 📝REBCO magnet under two tonnes reaching roughly 20 📝tesla on the tape and about 17 tesla at the centre bore, built from eight single 📝pancake windings in series — each one interchangeable, dry-wound and partially insulated. Once charged it holds constant current while operating in the presence of other coils. Field-induced stress in the winding pack is carried by "partitions" keyed into structural plates, and the pack is supported and thermally isolated inside the 📝cryostat by high-strength, low-conductivity brackets rated for over 60 tonnes of axial load, with cryocoolers holding it at 20 kelvin. 📝Brian LaBombard of the 📝MIT Plasma Science and Fusion Center and 📝Dan Brunner are among the co-authors.
The 2025 paper is the same magnets after they were made, and the byline order reverses to match: Grant Kristofek is first author, Radovinsky second (35(5), 10.1109/TASC.2025.3542351). It reports design, construction, assembly and full-field dual-magnet factory acceptance testing completed on both coils, at 20 tesla on tape and 17 tesla in the warm bore, with 📝Bob Mumgaard among the twelve authors. What was measured is the magnets, in a factory. Neither paper reports the mirror device itself operating.
The two coils are the high-field end coils for "An HTS Axisymmetric Magnetic Mirror on a Faster Path to Lower Cost Fusion Energy," a project under the 📝ARPA-E fusion programs BETHE programme on which CFS is a subrecipient alongside the University of Wisconsin–Madison. A 📝magnetic mirror confines plasma axially between two high-field regions instead of in a closed torus, so the magnet set is a pair of solenoids rather than a 📝toroidal field coil cage. What carries across from 📝SPARC is not the geometry but the conductor and the winding practice — the same tape, partial insulation and conduction cooling documented in 📝The Production Magnet Papers.
These are not SPARC coils, but they are SPARC's conductor and SPARC's winding practice aimed at a different machine. We built two, tested them at full field, and they became the high-field ends of someone else's device.