Determining the Validity of Tokamak Perturbed Equilibrium Modeling Using Nonlinear Equilibria

J. Halpern, N.C. Logan, E. Paul, C. Paz-Soldan · 2025

Read Determining the Validity of Tokamak Perturbed Equilibrium Modeling Using Nonlinear Equilibria by J. Halpern, N.C. Logan, E. Paul, C. Paz-Soldan (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-2506-14938.txt.

Introduction

Halpern and co-authors test in 2025 which reference frame makes tokamak perturbed equilibrium models valid, by checking them against full nonlinear equilibria. Small 3D fields disturb these plasma devices, and model answers hinge on the assumed symmetry frame, coil centroid or toroidal field coil. Fully 3D equilibria from the VMEC code supply the ground truth.

For a SPARC case with coil asymmetry, the toroidal field coil centroid approximated the right frame well. For NSTX-U, the radial spot of the inboard coil legs at midplane worked. Axis shift in the true equilibria picks the frame. Linear theory held for current coil tolerances, but the margin is thin for future devices.

Catalog record

CREATORJ. Halpern, N.C. Logan, E. Paul, C. Paz-Soldan
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE42 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2506-14938

Browse the full library · About the Tower of Babel

Full text

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

Determining the Validity of Tokamak Perturbed Equilibrium Modeling Using Nonlinear Equilibria J. Halpern, N.C. Logan, E. Paul, C. Paz-Soldan Abstract The prediction of perturbed equilibrium models for tokamaks with small non-axisymmetric fields is strongly dependent on which reference frame for axisymmetry is assumed. This assumption directly affects the applied field spectrum on the plasma, with the incorrect choice resulting in an incorrect prediction of the plasma response. We use fully 3D equilibria generated by VMEC to determine the correct reference frame when calculating error fields in perturbed equilibrium codes subject to n = 1 n=1 misalignments in several coil sets. We analyze a case of independently offset toroidal field coils and central solenoid/poloidal field coils in the SPARC tokamak and find that the appropriate reference frame can be well approximated by the centroid of the toroidal field coil set. We also consider the case of NSTX-U with an independent centerpost consisting of the inner legs of the toroidal field coils and central solenoid, and find that the reference frame can be well approximated by the radial location of the toroidal field coil inboard legs at the midplane. We determine the correct frame by analyzing the shifted magnetic axis in the nonlinear equilibria, and use our findings to generalize to 3D fields from misalignments which modify the reference frame and those which impact the plasma response. We also analyze the magnetic field line displacement to identify where the linearized MHD theory begins to break down, and compare that to typical coil tolerances in existing tokamaks. We find that linear theory is valid for existing tolerances, validating the use of perturbative codes to set these tolerances, but with a sufficiently small margin to be of note for future devices with relative tolerances larger than approximately 1 % 1\% of the minor radius. This study enables engineers to confidently use 3D perturbative models for determining assembly tolerances by providing insight into the correct applications of the theory. August 8, 2026 \ioptwocol 1 Motivation Tokamaks are designed to be nominally axisymmetric devices. However, unavoidable non-axisymmetric “error fields” occur due to construction errors, current feeds into the electromagnetic coils, and ferritic materials in the vicinity of the device [ 1 , 2 , 3 ] . Error fields on the order of ten thousand times weaker than the axisymmetric field can reduce plasma rotation through neoclassical torque [ 4 , 5 , 6 ] , modify the heat flux distribution on the divertor [ 7 , 8 ] , and create magnetic islands [ 9 , 10 ] which can reduce confinement and lead to a disruption [ 11 , 12 , 13 ] . These features make identifying and correcting error fields of the utmost importance in tokamak operation and design, and ultimately set the allowed tolerances for device construction [ 14 , 15 , 16 , 17 ] . The effects of error fields on the plasma are commonly modeled using 3D perturbed equilibria, which describe how a nominally axisymmetric equilibrium responds to externally applied 3D magnetic fields. This linear formulation can be solved more simply than fully nonlinear equilibria, and several perturbative codes have been developed [ 18 , 19 , 20 , 21 , 22 , 23 ] and used to predict plasma responses to both error fields resonating in the plasma core and resonant magnetic perturbations in the edge [ 24 , 25 ] . In this study, we will focus on two main limitations faced when using perturbative models: the choice of reference frame and the validity of the linearized MHD model. Reference Frame A subtlety that arises when using a perturbed equilibrium code is the correct choice of the reference axis for defining axisymmetry, hereafter referred to as the reference frame. As the equilibrium position is fixed within pertubative codes, this frame is dictated by the input coil positions and varied by rigid shifts of the entire coil set. To illustrate this impact, we consider an example configuration for the SPARC tokamak [ 26 ] , with the cross section in the x,z plane visualized in fig. 1 (a). In this example, the entire toroidal field (TF) coil set in blue has been rigidly shifted in the -\hat{x} direction during assembly, with the poloidal field (PF) coils and central solenoid (CS) in red installed as specified in the lab frame, the intended axis of symmetry of the device. When specifying the input 3D coil set, the user assumes that the computed equilibrium will obey the constraints of the linearized theory, which has a singular point at the magnetic axis and forces the axis displacements to go to zero. The perturbed equilibrium must therefore retain the same magnetic axis as the initial axisymmetric equilibrium, requiring the user to specify the coils in the frame that preserves the magnetic axis. In other words, the shifted coil set should be specified relative to the true magnetic axis location. Figure 1 : (a) Sample coil set with toroidal field coils in blue rigidly shifted in the +\hat{x} direction. The cross section of the unshifted poloidal field coils and central solenoid is shown in red. (b) Energy normalized applied flux spectrum calculated in GPEC, \Phi^{x}_{m} , versus poloidal mode number in Hamada coordinates, m . The red dashed spectrum is the applied flux on the red surface in (a) from the shifted TF coils, and the blue spectrum is the applied flux on the blue surface from the PF/CS coils. In our example, the user must assume if the plasma axis is fixed with respect to the PF/CS coils, TF coils, or somewhere in between. For the first case, which we denote the PF/CS frame, we assume that the plasma retains the original axis of symmetry (red dashed line in fig. 1 (a)), and the input equilibrium in red is perturbed due to the shifted TF coils. As the PF/CS coils are unshifted, the PF/CS frame is equivalent to the lab frame in this case. In the TF frame, the user assumes that the plasma shifts with the TF coils, resulting in an input equilibrium that is axisymmetric with respect to the TF centroid (blue dashed line in fig. 1 (a)) with the same plasma boundary as the original equilibrium; the 3D coil set is then rigidly shifted relative to the plasma, and the nonaxisymmetric field becomes the PF/CS coils shifted in the +\hat{x} direction. For all frames in between, the correct choice of input coils would include both the TF and PF/CS coils shifted in opposite directions, with a relative shift between them equal to the original TF shift. This assumption of which reference frame is correct will result in differing applied 3D field spectra, as shown in fig. 1 (b) where we plot the normalized applied flux on the plasma surface, \Phi^{x}_{mn}=\frac{1}{(2\pi)^{2}}\iint e^{i(n\phi-m\theta)}\frac{\delta\vec{B}\cdot\hat{n}}{\vec{B}\cdot\nabla\theta}d\theta d\phi, (1) in both the TF frame and PF/CS frame. Here, \vec{B} is the magnetic field, \phi/\theta the toroidal/poloidal angles, n/m the toroidal/poloidal mode numbers, and \hat{n} the surface unit normal. The applied field spectrum determines the plasma response [ 23 ] , indicating that the response differs in these two frames despite being the same physical coil set translated in space. The magnitude of the response difference

Read the complete text