2024 Theses Doctoral

# Three-dimensional magnetic fields: from coils to reconnection

This thesis is a work divided into two parts on aspects of three-dimensional (3D) magnetic fields: (I) magnetic reconnection treated from a strictly 3D viewpoint and (II) the design of coils for producing the 3D magnetic fields of optimized stellarators.

In astrophysical settings, magnetic fields are generically 3D. 3D divergence-free fields have rich topological structures such as magnetic nulls and chaotic field line structures. Standard reconnection literature identifies magnetic nulls as locations of magnetic reconnection, and that intense currents will build up around them. This idea is explored with a key realization that by placing a vanishingly small sphere around the null, boundary conditions on field lines passing through the sphere may be sorted out. The main result here is (1) the dismissal of the notion that nulls are crucial places for magnetic reconnection and current accumulation, instead identifying separatrices of topological type on the boundaries of null-passing field lines to be crucial. Standard reconnection literature dismisses chaotic flows yet 3D fields generically have chaotic flows. An inherent property of chaotic flows is exponentiation. The main result here is (2) the identification of exponentiation as a natural mechanism for magnetic reconnection and that the associated current builds up linearly in time in contradiction to standard results requiring the formation of high-density current sheets.

The magnetic fields of optimized stellarators are intricate, producing complex 3D magnetic surfaces. These fields are conventionally generated by non-planar electromagnetic coils, though these coils are costly to manufacture, slow device assembly, and hinder stellarator maintenance. Part II of this thesis explores methods of stellarator coil simplification that do not involve modular coils. All of this work uses current potentials, which are stream functions of the current sheets that produce magnetic surfaces. We begin with a result found using analytic methods on current potentials that (1) there may be an inherent limitation in the ability of modular coils to produce fields at a distance. This result is not surprising, though further analysis is necessary to work out some complexities of the result.

Next, (2) a novel method to produce localized patches of current potential, representative of patches of current sheets, is developed and used to identify crucial locations of current placement for shaping magnetic surfaces. Most notably, these current sheet patches are able to produce much of the surface shaping while occupying a small fraction of the winding surface, resulting in good open-access stellarator coil configurations. Continuing the trend away from modular coils, (3) helical coils are optimized to support stellarator magnetic fields.

This work agrees with related work on the optimization of helical coils, finding them unsuitable to the precise production of equilibria generated by modular coils. To improve this result, we use coil sets of mixed-type: helical coils with windowpane coils or permanent magnets, to mitigate field error left behind by the helical coils. Finally, (4) the development of a generalized method to cut modular, helical, and windowpane coils out of current potentials and to identify the associated coil currents is developed and used in coil optimization.

## Files

- Elder_columbia_0054D_18183.pdf application/pdf 13.9 MB Download File

## More About This Work

- Academic Units
- Applied Physics and Applied Mathematics
- Thesis Advisors
- Boozer, Allen H.
- Degree
- Ph.D., Columbia University
- Published Here
- November 1, 2023