Design of passive and structural conductors for tokamaks using thin-wall eddy current modeling is the 2024 Nuclear Fusion paper introducing ThinCurr, the three-dimensional electromagnetic code used to design the conducting structures 📝SPARC and 📝DIII-D depend on during a 📝disruption.
When a tokamak plasma disrupts, its current collapses in milliseconds and the magnetic energy it was carrying has to go somewhere. Part drives induced 📝eddy current loops through every conducting surface nearby; part flows as 📝halo current through plasma that has reached the wall and returns through the structure. The metal absorbing those loads is the same metal holding the machine together, which is why the disruption forces 📝Ryan Sweeney projects for SPARC are an electromagnetic design problem before they are a mechanical one.
ThinCurr, built on the existing PSI-Tet finite-element code, solves those currents in the thin-wall limit — surfaces rather than solids, which is what makes a full three-dimensional machine tractable. A. F. Battey and seven co-authors, among them 📝Carlos Paz-Soldan, Sweeney, 📝Alex Tinguely and 📝Alex Creely, built models of both SPARC and DIII-D including the 📝vacuum vessel and other structural elements at realistic material resistivities.
The application driving the paper is the passive 📝runaway electron mitigation coil: a non-axisymmetric coil with no trigger and no power supply, driven by the disruption itself to apply large three-dimensional fields that destroy 📝magnetic flux surfaces and deconfine the seed population from which 📝runaway electrons grow. Sweeping coil resistivity, current quench duration and plasma vertical position produced two conclusions: current designs should apply substantial 3D fields at the plasma surface during a disruption, and the coil conductors must sit away from the machine midplane to remain robust in off-normal disruption scenarios. Both are modeled results for machines whose disruptions have not yet been measured against them.
The runaway electron mitigation coil is a part we intend never to need. Modeling like this is how we satisfy ourselves it will work the one time it has to, on a tokamak that has not yet disrupted.