Efficient single-stage optimization of islands in finite-β stellarator equilibria

We present single-stage optimization of islands in finite-β stellarator equilibria. Stellarator optimization is traditionally performed as a two-stage process; in the first stage, an optimal equilibrium is calculated that balances a set of competing constraints, and in the second stage, a set of coils is found that supports the said equilibrium. Stage one is generally performed using a representation for the equilibrium that assumes nestedness of flux surfaces, even though this is not warranted and occasionally undesired. The second stage optimization of coils is never perfect, and the mismatch leads to worse performing equilibria and further deteriorates if additional constraints, such as force minimization, coil torsion, or port access, are included. The higher fidelity of single-stage optimization is especially important for the optimization of islands as these are incredibly sensitive to changes in the field. In this paper, we demonstrate an optimization scheme capable of optimizing islands in finite β stellarator equilibria directly from coils. We furthermore develop and demonstrate a method to reduce the dimensionality of the single-stage optimization problem to that of the first stage in the two-stage approach.