Fast and accurate 3D FEM model for electromagnetic simulations of no-insulation HTS coils based on polygon-anisotropic-resistivity

Abstract For no-insulation (NI) high-temperature superconducting (HTS) coils, a 3D electromagnetic model, which is fast and accurate, conducive to establish, and straightforward to multi-physics coupling, is still required. This paper introduces a polygon-anisotropic-resistivity (PAR) method for 3D FEM electromagnetic simulations of NI HTS coils. This model avoids dividing each tape into the specific HTS-tape layer and turn-to-turn contact layer, which yields: (1) a reduced computational burden; (2) improved convergence due to smaller element aspect ratios. The significance of the PAR method lies in its indispensable role in achieving a 3D anisotropic-resistivity model with high computing speed and accuracy. The proposed PAR model is rigorously evaluated through three types of simulations: (1) charge and discharge tests; (2) AC losses of the NI coil subjected to AC fields with a DC current supply; (3) heat-triggered quench and recovery scenarios. For these simulations, the PAR model is validated by comparisons with the full-element model, namely, the 3D FEM model that explicitly incorporates each specific HTS-tape layer and turn-to-turn contact layer in the H -formulation model, and is also validated by previous experimental data for AC losses. Good consistency is observed. The computing speed of the PAR model is tested to be 12–38 times that of the full-element model with the same accuracy. The PAR model achieves a 40% reduction in degrees of freedom compared to the full-element model, with the same mesh density along the HTS tape width and length, facilitating more precise and larger scale coil simulations within the same computational memory limits. Additionally, the PAR model entirely eliminates the inherent inaccuracies found in the conventional-anisotropic-resistivity 3D model, which stem from discrepancies between the arranged anisotropic-resistivity and the actual computed coil meshes. The proposed PAR model will enhance the prevalence of 3D electromagnetic analyses of NI HTS coils.