Surface Current Optimization and Coil-Cutting Algorithms for Stage-Two Stellarator Optimization

D. Panici \aff 1 \corresp, R. Conlin \aff 2, R. Gaur \aff 5, D. W. Dudt \aff 4, Y. G. Elmacioglu \aff 1, M. Landreman \aff 2, T. Elder \aff 2, N. Snir \aff 3, I. Gissis \aff 3, Y. Nikulshin \aff 3, E. Kolemen \aff 1 \corresp · 2025

Read Surface Current Optimization and Coil-Cutting Algorithms for Stage-Two Stellarator Optimization by D. Panici \aff 1 \corresp, R. Conlin \aff 2, R. Gaur \aff 5, D. W. Dudt \aff 4, Y. G. Elmacioglu \aff 1, M. Landreman \aff 2, T. Elder \aff 2, N. Snir \aff 3, I. Gissis \aff 3, Y. Nikulshin \aff 3, E. Kolemen \aff 1 \corresp (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-09321.txt.

Introduction

Dario Panici and co-authors detailed in 2025 the second stage of stellarator design, where coils are shaped to create a chosen magnetic field. A stellarator confines fusion plasma with twisted external coils. In stage one, the plasma boundary is tuned for physics goals; in stage two, the boundary is fixed and coils must recreate it. Past papers gave few details on how coils are cut from a surface current.

The team presents basic physical quantities of that surface current and how they relate to the terms of the current potential, backed by explicit derivations for the first time, in their account. They also detail how to include an external field in the accounting. The catalog creator field carries raw LaTeX markers; the head byline governs. Readers get the missing manual for a key design step.

Catalog record

CREATORD. Panici \aff 1 \corresp, R. Conlin \aff 2, R. Gaur \aff 5, D. W. Dudt \aff 4, Y. G. Elmacioglu \aff 1, M. Landreman \aff 2, T. Elder \aff 2, N. Snir \aff 3, I. Gissis \aff 3, Y. Nikulshin \aff 3, E. Kolemen \aff 1 \corresp
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE50 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2508-09321

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Surface Current Optimization and Coil-Cutting Algorithms for Stage-Two Stellarator Optimization D. Panici \aff 1 \corresp, R. Conlin \aff 2, R. Gaur \aff 5, D. W. Dudt \aff 4, Y. G. Elmacioglu \aff 1, M. Landreman \aff 2, T. Elder \aff 2, N. Snir \aff 3, I. Gissis \aff 3, Y. Nikulshin \aff 3, E. Kolemen \aff 1 \corresp Abstract Stellarator optimization often takes a two-stage approach, where in the first stage the boundary is varied in order to optimize for some physics metrics, while in the second stage the boundary is kept fixed and coils are sought to generate a magnetic field that can recreate the desired stellarator. Past literature dealing with this stage lacks details on the coil cutting procedure and the mathematical and physical properties of the surface current potential which dictates it. In this work, some basic physical quantities of the surface current and how they relate to the parameters in the current potential are presented, and supported for the first time by explicit mathematical derivations. Additionally, the details of how to account for the presence of an external field in the surface current algorithm are explicitly presented. These relations underpin the procedure of discretizing the surface current into coils. Finally, the conventionally-used algorithm for discretizing the surface current into coils is detailed, along with an example coil optimization for both a modular and a helical coilset. The algorithm is implemented in the DESC code, with both modular and helical coil capabilities, where it is available for use in stellarator coil design. 1 Introduction Stellarator coil design comprises the second stage of the conventional two-stage stellarator optimization approach. Stage-one of the optimization utilizes a three-dimensional ideal magnetohydrodynamics (MHD) equilibrium code running in fixed-boundary mode to optimize the shape of the plasma boundary (with the profiles either prescribed or part of the optimization) in order to minimize given physics-based objectives, such as particle confinement (through a proxy like quasisymmetry) and avoidance of rational surfaces in the rotational transform profile. Once a satisfactory fixed boundary equilibrium (as defined by the profiles, enclosed toroidal flux, and the boundary shape) is found, the second stage of the optimization involves finding an external coilset that can provide the necessary enclosed toroidal flux, while also ensuring the plasma boundary is indeed a flux surface (i.e. the total normal magnetic field on that surface is zero) ( Boozer 2000b ) . It has been observed by Merkel that this second requirement cannot be exactly satisfied ( Merkel 1987 ) , so instead a minimum of the integral of the normal field squared is sought, i.e. seeking a coilset which minimizes \chi^{2}_{B}\coloneqq\int_{S_{plasma}}(\bm{B}\cdot\bm{n})^{2}dA, (1) where S_{plasma} is the plasma boundary, \bm{n} is the unit surface normal vector and \bm{B} is the total magnetic field (with contributions from the coils external to the plasma and from the plasma field itself). There also has been recent work done on the concept of "single-stage" optimization, where both the plasma boundary and the external coils are optimized simultaneously ( Henneberg et al. 2021 ; Lee et al. 2023 ; Giuliani et al. 2022 ; Jorge et al. 2023 ; Jorge et al. 2024 ) , but this paper will not discuss the single-stage approach. Stage-two optimization of the coilset is typically approached one of two ways. The first assumes a fixed winding surface upon which a surface current lies, and minimizing the above metric with respect to this current, which we will call the surface current formalism ( Merkel 1987 ) . The second is to parameterize a discrete coilset, either filamentary or finite-build, and minimize \chi^{2}_{B} with respect to the degrees of freedom of this coilset. The first approach leads to a linear least-squares problem that can be efficiently solved for the global optimum, but is difficult to add other coil objectives to, while the second is a nonlinear minimization problem which can be augmented with other coil complexity objectives and sample a wider range of configurations not restricted to a pre-defined winding surface. Among the codes that perform this second approach are ONSET ( Drevlak 1998 ) , COILOPT (though the coils in this code were still restricted to a winding surface) ( Strickler et al. 2002 ) , COILOPT++ ( Gates et al. 2017 ) , and FOCUS ( Zhu et al. 2018 ) . Furthermore, the surface current formalism has been shown to be related to the problem of minimizing normal fields using permanent magnet (PM) arrays ( Helander et al. 2020 ) , and a number of methods and codes have been created to address the PM approach in both continuous and discrete formulations ( Kaptanoglu et al. 2023b ; Landreman & Zhu 2021 ; Hammond & Kaptanoglu 2024 ; Kaptanoglu et al. 2023a ; Elder & Boozer 2024 ) . While the prior works on the surface current formalism all provide detail on the general algorithm, the details of the coil cutting procedure are not elaborated upon by many of the references, nor are the properties of the surface current potential that inform the coil cutting procedures. This work intends to explain these in a clear way, for ease of introduction to this useful tool of stage-two stellarator optimization. In § 2 we summarize the existing literature on the surface current formalism for coil optimization. We then present the surface current formalism in § 3 . Section § 4 will present some basic physical quantities in the surface current and how it relates to the parameters in the current potential, supported for the first time by some basic mathematical derivations underpinning the procedure of discretizing the surface current into coils. Then, § 6 will detail the algorithm for discretizing the surface current into coils that is conventionally used in the community, both for modular and helical coils. Example stage-two coil optimizations with modular and helical coils are then presented with these methods. 2 Literature Review The surface current formulation of the stage-two coil optimization problem was first introduced by Rehker and Wobig ( S. Rehker & H. Wobig 1973 ) . Merkel ( Merkel 1987 ) extended this formulation to winding surfaces placed arbitrarily far away from the plasma surface with the NESCOIL code. This code has been used to obtain coilsets for optimized stellarators given a fixed winding surface ( Miner et al. 2001 ; Ku & Boozer 2010a ; Ku & Boozer 2010b ) , used in-situ along with the stage one optimization to inform coil complexity objectives ( Zarnstorff et al. 2001 ; Drevlak et al. 2018 ) , or used to create coilsets as starting points for further refinement with nonlinear coil optimization codes ( Drevlak et al. 2013 ; Zarnstorff et al. 2001 ) . Later, NESCOIL was modified to include a truncated singular value decomposition (TSVD) in a code titled NESVD which attempts to better regularize the problem ( Pomphrey et al. 2001 ) . More recently, the algorithm was modified to include a Tikhinov-like regularization based off of the surface current density magnitude ( Landreman 2017 ) which was shown to yield much better behaved results than the earlier formulations. This algorithm was implemented in FORTRAN in the original publication, and as part of this work has

Read the complete text