Deep neural-network based optimization for the design of a multi-element surface magnet for MRI applications

We present a combination of a CNN-based encoder with an analytical forward map for solving inverse problems. We call it an encoder-analytic (EA) hybrid model. It does not require a dedicated training dataset and can train itself from the connected forward map in a direct learning fashion. A separate regularization term is not required either, since the forward map also acts as a regularizer. As it is not a generalization model it does not suffer from overfitting. We further show that the model can be customized to either finding a specific target solution or one that follows a given heuristic. As an example, we apply this approach to the design of a multi-element surface magnet for low-field magnetic resonance imaging (MRI). We further show that the EA model can outperform the benchmark genetic algorithm model currently used for magnet design in MRI, obtaining almost 10 times better results.

Investigation of transport mechanisms induced by filament-coupling bridges- network in Bi-2212 wires

One of the features unique in Bi-2212/Ag wires is the network of bridges between the filaments formed by grains grown through the Ag matrix during the partial-melt heat treatment process. Although these interconnections favor a redistribution of the current among the filaments allowing high critical current density, they represent a strong electrical coupling between the filaments themself. Such a coupling increases the AC losses, present also in case of charge and discharge of DC magnets, principal applications of this kind of superconductor. In this work, through transport and magnetic measurements and their comparison, we study the behavior of these bridges as a function of applied magnetic field and temperature and the implications they have on the electrical coupling. The experiment has been performed on two multifilamentary wires prepared by GDG-PIT process starting from two commercial Bi-2212 precursor powders: Nexans and Engi-Mat. The reported results provide information on the effective length scale on which the filaments are coupled as a function of the field and temperature and we believe that such findings can be useful in magnet design.

Integrating micromagnets and hybrid nanowires for topological quantum computing

Majorana zero modes are expected to arise in semiconductor-superconductor hybrid systems, with potential topological quantum computing applications. One limitation of this approach is the need for a relatively high external magnetic field that should also change direction at the nanoscale. This proposal considers devices that incorporate micromagnets to address this challenge. We perform numerical simulations of stray magnetic fields from different micromagnet configurations, which are then used to solve for Majorana wavefunctions. Several devices are proposed, starting with the basic four-magnet design to align magnetic field with the nanowire and scaling up to nanowire T-junctions. The feasibility of the approach is assessed by performing magnetic imaging of prototype patterns.

Actively-shielded ultrahigh field MRI/NMR superconducting magnet design

Active shielding technology has been widely applied to the superconducting magnetic resonance imaging (MRI) and nuclear magnetic resonance (NMR) magnets design, revealing excellent performance on the stray field control. For such a highly homogeneous field superconducting magnet design, an appropriate optimization strategy is essential to guarantee the magnetic field homogeneity in the central region and the expected 5 Gauss line range, especially for the ultrahigh field superconducting magnet. Based on the compensating field optimization method, an actively-shielded whole-body 14T MRI magnet and an actively-shielded 1.3GHz NMR magnet were presented, and detailed analyses were conducted to evaluate the feasibility of the designs. The developed magnet design method, coil pattern, wire arrangement, and stress/strain adjustment will be used to guide the corresponding project implementation.

Efficiency Improvement of Permanent Magnet BLDC Motors for Electric Vehicles

A permanent magnet Brushless DC (BLDC) motor has been designed with different rotor configurations based on the arrangement of the permanent magnets. Rotor configurations strongly affect the torque and efficiency performance of permanent magnet electric motors. In this paper, different rotor configurations of the permanent magnet BLDC motor with parallel the Halbach array permanent magnet were compared and evaluated. Many applications of electric drives or air-crafts have recently preferred the surface-mounted permanent magnet design due to its ease of construction and maintenance. The finite element technique has been used for the analysis and comparison of different geometry parameters and rotor magnet configurations to improve efficiency and torque performance. A comprehensive design of a three-phase permanent magnet BLDC 35kW motor is presented and simulations were conducted to evaluate its design. The skewing rotor and Halbach magnet array are applied to the permanent surface-mounted magnet on the BLDC motor for eliminating torque ripples. In order to observe the skewing rotor effect, the rotor lamination layers were skewed with different angles and Halbach sinusoidal arrays. The determined skewing angle, the eliminated theoretically cogging torque, and the back electromotive force harmonics were also analyzed.

Magnetostatic reciprocity for MR magnet design

Electromagnetic reciprocity has long been a staple in magnetic resonance (MR) radio-frequency development, offering geometrical insights and a figure of merit for various resonator designs. In a similar manner, we use magnetostatic reciprocity to compute manufacturable solutions of complex magnet geometries, by establishing a quantitative metric for the placement and subsequent orientation of discrete pieces of permanent magnetic material. Based on magnetostatic theory and non-linear finite element modelling (FEM) simulations, it is shown how assembled permanent magnet setups perform in the embodiment of a variety of designs and how magnetostatic reciprocity is leveraged in the presence of difficulties associated with self-interactions, to fulfil various design objectives, including self-assembled micro-magnets, adjustable magnetic arrays, and an unbounded magnetic field intensity in a small volume, despite realistic saturation field strengths.

Magnetostatic reciprocity for MR magnet design

Electromagnetic reciprocity has long been a staple in MR radio-frequency development, offering geometrical insights and a figure of merit for various resonator designs. In a similar manner, we use magnetostatic reciprocity to compute manufacturable solutions of complex magnet geometries, by establishing a quantitative metric for the placement and subsequent orientation of discrete pieces of permanent magnetic material. Based on magnetostatic theory and nonlinear FEM simulations, it is shown how assembled permanent magnet setups perform in the embodiment of a variety of designs, and how magnetostatic reciprocity is leveraged in the presence of difficulties associated with self-interactions, to fulfil various design objectives, including self-assembled micro magnets, adjustable magnetic arrays, and an unbounded magnetic field intensity in a small volume, despite realistic saturation field strengths.