Computationally-efficient optimization of the remanence angles of permanent magnet circuits for magnetic refrigeration

Permanent magnets are used in many engineering fields, but their high cost and restricted availability demand optimization techniques to generate high magnetic fields over large areas with minimal material. This paper advances an optimization technique for magnetic circuits composed of NdFeB and iron wedges. The remanence angles of each magnetized block are optimized to generate a desired magnetic waveform in the air gap, a critical parameter of the magnetocaloric cooling response in magnetic refrigerators. An advantage of the proposed method is that the number of finite element simulations is twice the number of magnetized blocks, often capped due to cost reasons. Therefore, the method can fully optimize a given geometry in a few minutes. Two-dimensional and three-dimensional models are developed and compared, and the full three-dimensional model is used to design the magnetic circuit for an actual magnetic refrigerator, capable of achieving 1.1T in the high field region. An encouraging agreement is observed between the modeling results and the local experimental magnetic field measured within the air gap. The most significant deviations between simulated and measured values, never higher than 40%, are observed in the region where the magnetic field gradient is the steepest and are likely caused by manufacturing imperfections that are hard to control, such as magnetization magnitude and direction deviations and localized self-demagnetization. Deviations between experimental data and numerical results smaller than 20% are observed in the high field region.