AC Loss Computation in Large-Scale Low-Temperature Superconducting Magnets: Multiscale and Semianalytical Procedures

In this article, we introduce two complementary approaches for the prediction of ac losses in large-scale low-temperature superconducting (LTS) magnets subjected to slow ramp rates. These methods account for the temperature rise within the LTS coil and its impact on ac losses. The first approach is multiscale and relies on the coupling between a macroscopic homogenized model of the LTS coil and a mesoscopic model of a single filament for loss prediction. The second approach is semianalytical and is based on analytical approximations for the hysteresis losses, which are validated against the single filament model. The second approach offers a faster computation suitable for initial design considerations, while the multiscale method is shown to take into account more complex phenomena for the ac loss evaluation at the filament scale. We apply both methods to the prediction of ac losses generated in the LTS coil inside the IBA S2C2 synchrocyclotron during its ramp-up procedure. Additionally, we discuss the convergence properties of the multiscale approach and demonstrate the good agreement between the numerical results and experimental data.