Thermally adaptive magnetic excitation architecture for high-stability magnetorheological finishing in constrained geometries

• Planar air-gap geometry achieves >180 mT peak flux with a 4.5 mm uniform zone. • Hollow copper tube excitation remains stable below 45°C at high current. • Finite element–genetic algorithm optimizes yoke gradients for uniform fields. This investigation demonstrates a thermally adaptive magnetic excitation architecture for magnetorheological finishing (MRF), achieving previously unattainable magnetic flux stability within spatially constrained polishing environments. Three excitation configurations—multi-turn enameled coil, low-turn high-current cable, and square-section hollow copper tube—were evaluated under identical ampere-turn conditions. Experimental characterization and finite element analysis using ANSYS Maxwell revealed that only the hollow copper tube, with a 4 mm planar air gap, consistently delivered a peak magnetic flux density over 180 mT, while maintaining a thermal rise below 45°C. To enhance field uniformity and energy efficiency, a Python-coupled genetic algorithm optimized the angular geometry of the magnetic yoke, revealing a nonlinear flux regulation mechanism governed by structural gradients. The optimized design ensures a stable shear zone in both space and temperature, with simulation-experiment deviation <5 mT and material removal fluctuation under 10% in repeated K9 glass polishing. These results establish a quantitative framework for thermal-magnetic coupling under geometric and thermal constraints, proposing a new design paradigm for modular, field-tunable excitation systems. This approach offers a scalable solution to fundamental limitations in high-efficiency, atomic-level surface processing.