Optimizing stellarators for better equilibria through the addition of phyiscs objectives to DESC

Stellarators require optimizing relevant quantities to design magnetic field configurations that admit desirable equilibria. Because complicated physics governs these quantities, more efficient algorithms and proxy functions prove essential for covering larger search spaces to find good equilibria. This work presents the addition of physics objectives and optimization parameters, such as the magnetic well, current profile constraints, and other particle confinement metrics to the DESC stellarator code suite [1-4]. These parameters measure equilibrium stability and plasma confinement, rendering them good figures of merit alongside symmetries for designing future stellarators. Implementing them in DESC – a code with automatic differentiation and fast gpu portability – improves our understanding of equilibria quality and allows optimizing for new equilibria which may better satisfy a larger set of objectives.