High Order Free Boundary MHD Equilibria in DESC

Rory Conlin \aff 1 \corresp, Jonathan Schilling \aff 2 Daniel W. Dudt \aff 1, Dario Panici \aff 1, Rogerio Jorge \aff 3, Egemen Kolemen \aff 1 \corresp · 2024

Read High Order Free Boundary MHD Equilibria in DESC by Rory Conlin \aff 1 \corresp, Jonathan Schilling \aff 2 Daniel W. Dudt \aff 1, Dario Panici \aff 1, Rogerio Jorge \aff 3, Egemen Kolemen \aff 1 \corresp (2024) 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-2412-05680.txt.

Introduction

Rory Conlin and co-authors compute free-boundary plasma equilibria at high order in the DESC code. The task asks what shape a hot plasma takes when its edge is not fixed by a wall. Such shapes matter for fusion devices that confine plasma with magnets. Boundary rules split into fixed and free cases, with a sheet current allowed at the edge in some.

The team reviews when that sheet current can exist at the plasma-vacuum seam. A fast, exact method scores how well a trial shape meets the rules. DESC then finds equilibria in vacuum and at set plasma pressure. Runs cover cases with and without the sheet current. The work backs design studies for stellarator fusion devices.

Catalog record

CREATORRory Conlin \aff 1 \corresp, Jonathan Schilling \aff 2 Daniel W. Dudt \aff 1, Dario Panici \aff 1, Rogerio Jorge \aff 3, Egemen Kolemen \aff 1 \corresp
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE24 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-05680

Browse the full library · About the Tower of Babel

Full text

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

High Order Free Boundary MHD Equilibria in DESC Rory Conlin \aff 1 \corresp, Jonathan Schilling \aff 2 Daniel W. Dudt \aff 1, Dario Panici \aff 1, Rogerio Jorge \aff 3, Egemen Kolemen \aff 1 \corresp Abstract In this work we consider the free boundary inverse equilibrium problem for 3D ideal MHD. We review boundary conditions for both fixed and free boundary solutions and under what circumstances a sheet current may exist at the plasma-vacuum interface. We develop an efficient and accurate algorithm for computing the residual of these boundary conditions and use it to compute free boundary equilibria in the DESC code both in vacuum and at finite plasma beta, with and without sheet currents. 1 Introduction Free boundary 1 1 1 Here and throughout we follow the convention in MHD equilibrium and use ”free boundary” to mean the case where the boundary shape of the plasma is allowed to change, though the computational boundary is often fixed. This is in contrast to traditional fluid modelling where free boundary refers to the computational boundary being allowed to move. 3D MHD equilibrium calculations are a key step in many areas of fusion research. For experiments, such calculations are required to reconstruct the internal behavior of the plasma from external measurements Hanson et al. 2009 ; Schilling 2018 , as well as in experimental planning when to determine coil currents. Similarly in design they are an important verification step when designing coils for a new equilibrium. In stellarator optimization, free boundary calculations are generally required when performing "single stage" optimization where both the plasma and coils are optimized at the same time to achieve specific objectives Henneberg et al. 2021 ; Drevlak & Lobsien 2022 . The primary code presently used for free boundary calculations is VMEC Hirshman & Whitson 1983 ; Hirshman et al. 1986 , which uses the NESTOR code Merkel 1986 ; Merkel 1988 ; Merkel & Strumberger 2015 to compute the required vacuum field outside the plasma. NESTOR solves the exterior Neumann problem using a Green’s function method, and a singularity subtraction scheme to solve the resulting singular integrals. This method is only 2nd order accurate in the number of points used to discretize the integral, consistent with the 2nd order accuracy of VMEC’s radial discretization. The DESC code Dudt et al. 2024 improves upon VMEC in a number of ways, and has been shown to be significantly more accurate than VMEC Panici et al. 2023 , though is currently limited to fixed boundary calculations. In extending DESC to free boundary, it is natural to seek a method that will preserve this improved accuracy. Recent work by Malhotra Malhotra et al. 2019 ; Malhotra et al. 2020 has developed a new "partition of unity" method for computing singular boundary integrals arising in MHD equilibrium calculations that has been shown to be much more accurate, achieving 10th to 12th order accuracy Malhotra et al. 2020 , and is the method implemented in DESC and discussed in this work. We note that NESTOR has been ported to Python/JAX and interfaced with DESC and is available for comparisons. We begin by reviewing the boundary conditions for a fixed boundary problem in section 2 and the free boundary problem in section 3 . We then move on to the formulation of these boundary conditions that is solved in DESC in section 4 and the details of the numerical implementation in section 5 . We finally show benchmarks of free boundary DESC vs both VMEC and direct field line tracing in section 6 . 2 Fixed Boundary Conditions Before considering the free boundary problem, it is helpful to first consider the fixed boundary problem, in which the shape of the last closed flux surface ( R_{b},Z_{b} ) is specified, along with the normal field (taken to be zero). When a code such as DESC or VMEC is run with a fixed boundary, the magnetic field it finds is the total magnetic field in the plasma volume due to all sources, not just the currents in the plasma volume. This total field \mathbf{B}_{\textrm{fixed}} can be thought of as being made up of two parts as \mathbf{B}_{\textrm{fixed}}=\mathbf{B}_{\textrm{plasma}}+\mathbf{B}_{\textrm{fake}} , where \mathbf{B}_{\textrm{plasma}} is the field due to the volume currents in the plasma and \mathbf{B}_{\textrm{fake}} is a "fake" external field that is required to make the last closed flux surface coincide with the specified boundary. Because the source of the "fake" field is not inside the plasma volume, within the plasma volume it is curl free: \nabla\times\mathbf{B}_{\textrm{fake}}=0 . A common physical model for such a configuration is that of a plasma surrounded by a rigid superconducting shell. Because the magnetic field \mathbf{B}_{\textrm{fixed}} inside a ideal superconductor is zero, the boundary conditions for Maxwell’s equations in matter give \mathbf{K}_{\textrm{SC}}=-\frac{1}{\mu_{0}}\mathbf{n}\times\mathbf{B}_{\textrm{fixed}} (1) where \mathbf{n} is the surface normal vector and \mathbf{K}_{\textrm{SC}} is a sheet current density that exists on the plasma/superconductor interface, cancelling the field due to currents in the plasma volume such that the total field outside the plasma is zero, and being the source of the "fake" field inside the plasma. The goal of coil design in the second stage of a stellarator design process is to find external coils that replicate the effect of this sheet current inside the plasma volume - effectively to replace the "fake" field in the plasma volume with the field from external coils. With the plasma boundary held fixed and the normal field from the plasma \mathbf{B}_{\textrm{plasma}}\cdot\mathbf{n} specified, the solution inside the plasma volume is unique, and so any set of coils that lie outside the plasma volume and give (\mathbf{B}_{\textrm{plasma}}+\mathbf{B}_{\textrm{coil}})\cdot\mathbf{n}=0 will give the same equilibrium in the plasma volume. The question may then be asked: what happens to the sheet current? If one performs a Biot-Savart integral over the plasma currents in the volume to find \mathbf{B}_{\textrm{plasma}}\cdot\mathbf{n} , the effect of the sheet current will not be included, and so properly designed coils will create the original equilibrium with no sheet current present. The same is generally true when the volume integral is transformed into a surface integral using the virtual casing principle - in fact we see that the virtual casing sheet current \mathbf{K}_{\textrm{VC}}=-\frac{1}{\mu_{0}}\mathbf{n}\times\mathbf{B}_{\textrm{fixed}} is exactly the physical sheet current that exists in a fixed boundary equilibrium. In general, if we first obtain a fixed boundary equilibrium and then design coils for it, the free boundary equilibrium from the given coils will not have any sheet current on the plasma-vacuum interface. 3 Free Boundary Conditions The boundary conditions for a plasma supported by external fields can be found in many standard texts on MHD Freidberg 2014 ; Goedbloed et al. 2019 , and are summarized below \mathbf{B}\cdot\mathbf{n}=0 (2) [[p+B^{2}/2\mu_{0}]]=0 (3) \mathbf{n}\times[[\mathbf{B}]]=\mu_{0}\mathbf{K} (4) where \mathbf{B} is the magnetic field, p is the plasma pressure, \mathbf{K} is a surface current density, \mathbf{n} is the normal vector to the plasma boundary, and [[x]] is the jump in x across the boundary. Because \mathbf{B} lies tangent to the surface, \mathbf{K} will as well. These conditions can be derived intuitively by noting that if the first were not satisfied, particles streaming along magnetic field lines could leave the domain, while if the second were not satisfied the pressure difference would cause the plasma boundary to

Read the complete text