A Novel Simulation Model of Shielding Performance Based on the Anisotropic Magnetic Property of Magnetic Shields

To achieve a near-zero magnetic field environment, the use of permalloy sheets with high-performance magnetic properties is essential. However, mainstream welding processes for magnetically shielded rooms (MSRs), such as argon arc welding and laser welding, can degrade the magnetic properties of the material. Additionally, neglecting the anisotropy of permalloy sheets can introduce unpredictable errors in the evaluation of MSR performance. To address this issue, this paper proposes a modified model for calculating the shielding factor (SF) of MSRs that incorporates the anisotropic magnetic characteristics of permalloy sheets. These characteristics were measured using a two-dimensional single sheet tester (2D-SST). A high-precision measurement system was developed, comprising a 2D-SST (to generate two-dimensional magnetic fields and sense the induced B and H signals) and a control system (to apply in-phase 2D excitation signals and amplify, filter, and record the B and H data). Hysteresis loops were tested at low frequencies (0.1-9 Hz) and under different magnetization states (0.1-0.6 T) in two orientations-parallel and perpendicular to the annealing magnetic field-to verify anisotropy under varying conditions. Initial permeability, near-saturation magnetization, and basic magnetization curves (BM curves) were measured across different directions to provide parameters for simulations and theoretical calculations. Based on these measurements and finite element simulations, a mathematical model was developed to adjust the empirical coefficient λ used in theoretical SF calculations. The results revealed that the ratio of empirical coefficients in different directions is inversely proportional to the ratio of magnetic permeability in the corresponding directions. A verification group was established to compare the original model and the modified model. The mean squared error (MSE) between the original model and the finite element simulation was 49.97, while the MSE between the improved model and the finite element simulation was reduced to 0.13. This indicates a substantial improvement in the computational accuracy of the modified model.