Magnetic Field Design in a Cylindrical High-Permeability Shield: The Combination of Simple Building Blocks and a Genetic Algorithm

TL;DR

This paper introduces a novel design method combining analytic coil modeling and genetic algorithms to optimize magnetic fields inside cylindrical high-permeability shields, improving field uniformity and control for compact quantum devices.

Contribution

It develops a new analytic approach for coil design within magnetic shields and applies genetic algorithms to optimize coil configurations, enhancing magnetic field control in confined spaces.

Findings

Achieved a sevenfold increase in the volume with accurate axial gradient fields.

Realized a threefold increase in the volume with accurate transverse bias fields.

Demonstrated the method's potential for low-cost, compact magnetic field shaping systems.

Abstract

Magnetically-sensitive experiments and newly-developed quantum technologies with integrated high-permeability magnetic shields require increasing control of their magnetic field environment and reductions in size, weight, power and cost. However, magnetic fields generated by active components are distorted by high-permeability magnetic shielding, particularly when they are close to the shield's surface. Here, we present an efficient design methodology for creating desired static magnetic field profiles by using discrete coils electromagnetically-coupled to a cylindrical passive magnetic shield. We utilize a modified Green's function solution that accounts for the interior boundary conditions on a closed finite-length high-permeability cylindrical magnetic shield, and determine simplified expressions when a cylindrical coil approaches the interior surface of the shield. We use an analytic formulation of simple discrete building blocks to provide a complete discrete coil basis to generate any physically-attainable magnetic field inside the shield. We then use a genetic algorithm to find optimized discrete coil structures composed of this basis. We use our methodology to generate an improved linear axial gradient field, $\mathrm{d}B_z/\mathrm{d}z$, and transverse bias field, $B_x$. These optimized structures increase, by a factor of seven and three compared to the standard configurations, the volume in which the desired and achieved fields agree within $1\%$ accuracy, respectively. This coil design method can be used to optimize active--passive magnetic field shaping systems that are compact and simple to manufacture, enabling accurate magnetic field control in spatially-confined experiments at low cost.