This paper presents a method to extract the homogenized transversal permittivity of carbon fiber reinforced plastics which contains coated carbon fibers. The applied unit cell based EMQS model requires only geometrical and basic material parameters as input data. Results obtained from the method are compared to experimental/numerical data presented by other studies. It is also demonstrated that—up to a certain filament density—instead of modeling the unknown random inner fiber distribution of the composite, a regular and periodic model is suitable for the extraction of the homogenized permittivity.
We developed a scanning magnetoresistance microscope tailored for magnetic tape technology research. With initial results, we demonstrate the spatial and temporal resolution of the microscope by imaging recorded servo and data patterns on a perpendicularly oriented tape media sample, and the static and dynamic fields produced by a tape write transducer.
Corona discharges in air are relevant for many engineering applications, among which, ionic air-breathing propulsive systems have recently attracted much interest. In designing such systems, geometry is a key design feature as the electrode configuration effectively determines the ionization characteristic and power vs thrust performance. Motivated by this, in the framework of the IPROP project, we aimed at developing efficient numerical modeling techniques to accurately predict the performance of many complex electrode geometries in a design–evaluate–modify loop in order to enable use of numerical simulations as an effective practical design tool. In the present contribution we describe the modeling assumptions and numerical algorithms that form the basis of our simulation code, and we assess its accuracy and efficiency in evaluating the behavior of large and complex 2D and 3D thruster geometries.
This contribution presents an analytical formulation to derive the magnetic field spatial distribution in spherical induction actuators. The system is modeled as an ironless spherical structure, with a spherical conductive rotor and excitation coils fed with sinusoidal currents placed on a constant-radius surface. By introducing the second-order vector potential, the Helmholtz equation is solved to derive the spatial distribution of the magnetic flux density produced by a three-phase winding, considering the relative motion between the rotor and the stator magnetic field. Moreover, the algorithm estimates the developed torque directly from the spatial harmonics amplitude, avoiding the creation of a mesh grid to discretize the domain and thereby reducing computational time compared with FE simulations. The model is used to analyze an actuator with a rotor radius of about 36 mm and a stator radius of about 40 mm. Comparison with full-3D FE simulations highlighted the approach’s accuracy and computational efficiency.
Spatially nonuniform magnetic fields generate magnetic body forces in ferrofluids, causing internal pressure redistribution and altering the effective buoyancy of immersed objects. This effect is often described as an apparent-density field, yet direct quantification remains difficult in opaque ferrofluids under strong field gradients. Here, we establish a quantitative relationship between magnetic additional pressure and apparent density. Internal pressure is directly measured using an immersed pressure sensor, and the apparent-density distribution is reconstructed accordingly. The model is validated by experiments and multiphysics simulations incorporating measured magnetization curves rather than assuming saturation. Infrared thermography of density-calibrated spheres is used to identify equilibrium levitation positions and independently assess the predicted field. The measured pressure agrees closely with simulations. The inferred apparent density matches levitation positions with a mean relative error of 3.9%, while reconstructed profiles agree with simulations, with errors below 10% for water-based ferrofluid and below 5% for oil-based ferrofluid.
To reduce the meshing difficulty and computational cost caused by laterally continuous thin nanocrystalline laminated shields in wireless power transfer (WPT) systems, a layered transition boundary condition (LTBC) method is proposed for efficient leakage magnetic field computation. The physical laminated shield is replaced by equivalent zero thickness boundaries, while the skin effect dependent electromagnetic transmission behavior of each conductive magnetic layer is retained through surface and transfer impedances. Two formulations are developed, namely LTBCs for laminates with identical layer properties and LTBCd for laminates with different material properties or thicknesses. The proposed models are compared with an anisotropic homogenized model under different lamination numbers and nanocrystalline material parameters. The results show that the LTBC methods maintain magnetic-field prediction accuracy while reducing the degrees of freedom and computation time by up to 86.5% and 96.6%, respectively. Experimental measurements further validate the accuracy of the proposed method for multilayer thin laminated shields.
This paper proposes an efficient framework for rapid uncertainty analysis of the global electromagnetic performance of cross-stacked grain-oriented electrical steel (GOES) cores. To reduce the computational cost associated with detailed laminated finite-element models, an equivalent modeling approach is developed by combining an orientation distribution function (ODF) model with a multiscale homogenization technique. The ODF model is first employed to reconstruct the B-H characteristics corresponding to different cutting angles. Based on the reconstructed curves, we adopt the energy-based finite element method with a multi-scale homogenization procedure to establish an equivalent anisotropic model of the cross-stacked laminations, this approach enablies accurate modelling while significantly improving computational efficiency. Using the equivalent model, we propose a model that combines neural network predictions and Sobol index analysis to quantify the contributions of uncertain parameters. This scheme enables rapid analysis of tolerances in manufacturing.
To eliminate the need for narrow-track readers in high-track-density bit-patterned magnetic recording (BPMR) systems, this work investigates encoding schemes and dual-track simultaneous reading using a wide-track reader. To mitigate two-dimensional (2-D) interference, two enhanced signal processing methods for single-reader/dual-track reading are introduced: 1) a simple modulation code at a rate of 3/4, to avoid ambiguous dual-track bit patterns and generate distinguishable readback signal levels, and 2) a modified Viterbi detection scheme that employs an optimized trellis structure specifically designed for dual-track sequence estimation. Since data from two tracks is read simultaneously, the transfer speed is one and a half times that of single-track reading. Moreover, the simulation results show that our proposed BPMR system performs better than single-track reading in terms of bit-error rate at the same user densities (UDs), even at UDs exceeding 3.0 Tb/in2 under severe media noise and 2-D interference conditions.
The rapid transition from hard disk drives (HDDs) to solid-state drives (SSDs) results in massive HDD decommissioning, raising data security and environmental challenges. Before recycling, data sanitization through magnetic degaussing is required; however, HDDs' composite aluminum-steel casing significantly attenuates the applied magnetic field. This paper presents a three-dimensional (3D) finite element method (FEM) study investigating frequency-dependent magnetic flux density attenuation through aluminum-steel HDD cases and establishing mesh granularity guidelines, while quantifying the frequency ranges in which local attenuation predictions achieve convergence. The study employed an open-source FEM toolchain (NETGEN, ELMER FEM, GNU Octave) and first-order (linear) Whitney/Nedelec H(curl)-conforming edge elements to analyze frequency-dependent attenuation for different mesh resolutions and field orientations. Results reveal that coarse meshing leads to underestimated shielding effects, especially above 1 kHz, potentially causing false predictions in degaussing efficiency. This paper formalizes frequency-aware meshing guidelines linking the maximum element size to the electromagnetic skin depth, verifies global energy convergence using energy-norm indicators, and identifies the frequency ranges in which local attenuation predictions remain mesh-sensitive. The framework supports the design of secure HDD sanitization systems, which can be reproduced with free software.
This paper proposes a conical axial-flux permanent-magnet motor (AFPMM) integrated with a layer-wise machined maximum slot occupancy (MSO) coil to maximize torque density within a limited envelope volume. While the conical geometry alleviates radial core saturation by introducing a radially varying axial structure, it inherently creates tapered stator slots that reduce the fill factor of conventional windings. To overcome this limitation, the proposed MSO coil features layer-wise varying conductor heights to closely follow the tapered boundary, effectively maximizing the usable slot area. To evaluate this coupled topology efficiently, a variable cross-sectional area phase-resistance model was developed and combined with quasi-3D (Q3D) finite element analysis (FEA). Based on this framework, a parametric design sweep was conducted to optimize the cone angle, rotor thickness, and magnet-thickness ratio under identical total magnet-volume conditions. The optimal configuration was rigorously validated via full 3D FEA, demonstrating a significant torque density increase from 45.98 Nm/L to 54.63 Nm/L, which corresponds to an 18.8% performance improvement under the same copper-loss condition.
In this paper, we present a computational model that evaluates the electromagnetic scattering from reconfigurable metasurfaces of finite dimensions. Due to its Fourier-domain representation, the proposed model admits an inversion, which can be used to estimate the metasurface configuration that produces a desired scattering pattern. The methodology described is based on physical optics, a framework that can analytically compute the scattered field from arbitrarily profiled and illuminated finite apertures, accounting for near- and far-field diffraction. The forward model is compared with established literature models, while the performance of the inverse process is evaluated through numerical simulations. These results validate the framework, highlighting its relevance in forward and inverse scattering calculations.
This paper proposes a dual-layer bit-patterned magnetic recording (DL-BPMR) system with unequal recording-layer thicknesses to improve bottom-layer detection under severe inter-layer interference (ILI). The top recording layer is designed to be thicker than the bottom layer, resulting in different readback signal amplitudes that enable reliable top-layer detection with a simple threshold detector. The top-layer decisions are then used to perform min-max normalization, rescaling the four-peak amplitude distribution of the mixed readback signal into a two-peak distribution, substantially reducing ILI ambiguity before bottom-layer detection. The rescaled signal is subsequently processed by a multi-layer perceptron (MLP) detector using a sliding window input construction that combines rescaled signal samples and estimated top-layer bits to capture nonlinear signal characteristics that persist after rescaling. At an areal density of 4.0 Tb/in2, the proposed system achieves a bit-error rate (BER) of around 10−5 at a signal-to-noise of 20 dB and outperforms bottom-layer rescaling with one-dimensional partial-response maximum-likelihood detection by approximately 2 dB at a BER of 10−4. These results demonstrate the effectiveness of combining unequal recording-layer thicknesses, readback signal rescaling, and MLP-based detection for high-density DL-BPMR systems.
The consequent-pole permanent magnet bearingless motor (CPPMBM) is of significant interest in industrial applications due to its reduced coupling between suspension control and torque control. This paper presents an analytical model for the CPPMBM considering stator slotting and rotor eccentricity. First, fractional linear transformation is applied to address the magnetic field distortion caused by rotor eccentricity of the motor. The relative permeance functions in the x and y directions are calculated. Next, under concentric conditions, the analytical domain of the CPPMBM is divided into three subregions. Laplace’s and Poisson’s equations are solved for each subregion using the corresponding boundary conditions, and the concentric no-load air-gap magnetic flux density is derived through the vector magnetic potential. Then, the concentric no-load air-gap magnetic flux density of CPPMBM is corrected using the relative permeance function to obtain the no-load air-gap field under eccentric conditions. Finally, the analytical results for the no-load air-gap magnetic flux density, cogging torque, and unbalanced magnetic pull are compared with the finite element results. The air-gap magnetic field is evaluated by the regression evaluation indexes. The analytical results are consistent with the finite element results, which verifies the validity and accuracy of the proposed analytical model.
Doping of heavy rare earth elements in Sm2Co17 permanent magnets demonstrates a significant potential for reducing remanence temperature coefficient, thereby improving the precision of magnetic instrument with temperature variation. However, such alloying additions tend to compromise magnetic performance. This study aims to improve the magnetic properties by ball milling process optimization in Sm2Co17 permanent magnets doped by Dy, Gd and Er. Prolonging the drying time of ball-milled powders will increase the content of (Sm, Gd)2O3, therefore reducing the content of Sm in matrix and a higher proportion of cell interiors. The increased cell interiors helps to the remanence improvement, which is increased from 8.612 kG to 8.798 kG at room temperature. On the other hand, enriched (Sm, Gd)2O3 precipitates may be strong domain pinning sites and responsible for the coercivity enhancement, which is increased from 15.26 kOe to >26.00 kOe at room temperature. Meanwhile, the thermal stability is also improved by drying time prolonging. The remanence temperature coefficient between 20-200 °C is slightly reduced from -0.0118%/°C to -0.0037%/°C. It is attributed to the increase in the proportion of heavy rare earth elements in matrix, which is caused by more Sm consumed by oxidation when compared with heavy rare earth. This work provides a viable method for magnetic performance enhancement in Sm2Co17 permanent magnets with heavy rare earth elements doping.
Heat-assisted interlaced magnetic recording (HIMR) increases areal density capability (ADC) of hard disk drives by alternating narrow top tracks with wide bottom tracks. The wide bottom tracks require elevated laser current, which degrades reliability. This work maps HIMR ADC as a function of laser and writer current and shows the ADC gain saturates at 6~10% increase of laser current relative to conventional magnetic recording. For one head design, the ADC gain reaches 10.0% as the bottom track laser current is increased by 10% and climbs slightly to 10.8% as the laser current increase is doubled to 20%. For the other head design, the ADC gain reaches 12.0% at 6% laser current increase. The ADC saturation is attributed to the plateauing of spatial signal-to-noise ratio of the bottom track confined between two top tracks. The ADC saturation with laser current can be enhanced by increasing writer current for the bottom tracks, enabling significant HIMR ADC gain with minimal reliability penalty.
Modern computers increasingly favor low-precision arithmetic, and sparse iterative linear solvers are typically memory bandwidth-bound. Hence, mixed-precision techniques have become a promising approach to improving the performance of iterative linear solvers. In this study, we evaluate our previously proposed mixed-precision sparse linear solver F3R, which can effectively utilize IEEE 754 half-precision (FP16) arithmetic, in quasi-static electromagnetic field analysis for a switched reluctance motor. We combine F3R with a reordering-based ILU(0) preconditioner and compare it with conventional preconditioned Krylov subspace methods, including BiCGStab and GMRES, as well as a previous F3R variant coupled with an approximate inverse preconditioner. Numerical experiments on both CPU and GPU nodes show that F3R uses FP16 arithmetic efficiently and has performance advantages over the conventional solvers and the previous F3R variant, demonstrating its applicability to electromagnetic analysis.
This paper presents an accelerated, non-intrusive multivariate sensitivity-adaptive (MVSA) polynomial chaos expansion (PCE) framework for uncertainty quantification in time-varying electromagnetic structures. In essence, the method enriches the active polynomial basis with multiple dominant admissible multi-indices per iteration, reducing the adaptive construction cost relative to the standard MVSA procedure. The proposed formulation is coupled with a Floquet modal solver and successfully validated through a time-periodic dielectric slab with both geometrical and material uncertainties.
Accurate current-rating assessment is essential for the reliable operation of power cables under high loading conditions. Although IEC 60287 provides standardized analytical methods, it relies on simplifying assumptions that may not fully capture cable electromagnetic and thermal behavior. To evaluate their impact, finite element method (FEM) models are developed for a benchmark 132 kV cable system in trefoil and flat formations, considering two alternative sheath-modeling approaches. FEM-based ratings are compared with IEC 60287 results. The findings indicate that IEC 60287 may overestimate the permissible current rating, highlighting the value of detailed numerical modeling.
This paper presents an asymmetric rotor shape designed to reduce the torque ripple of a wound-field synchronous motor (WFSM) without sacrificing the average torque. A symmetric model and two asymmetric models with oppositely extended pole shoes were first analyzed using finite element analysis (FEA), and their torque waveforms were shown to exhibit a near-inverse phase relationship in the dominant ripple harmonic. Based on this observation, a single-piece rotor combining both asymmetric pole shoe geometries on alternating poles was proposed, and fast Fourier transform (FFT) analysis confirmed that the dominant ripple harmonic was effectively suppressed while the average torque was preserved. Two shape variables were then defined to further optimize the proposed rotor, and the torque and torque ripple characteristics were analyzed according to their variation. An optimal design was derived from the resulting electromagnetic characteristic maps, and its torque, torque ripple, and cogging torque were compared with those of the symmetric and asymmetric models. Compared with the Symmetric Model, the optimal proposed model reduced the torque ripple from 27.16% to 14.12% and the cogging torque from 5.68 Nm to 3.78 Nm, while the average torque remained nearly unchanged, demonstrating that the proposed asymmetric rotor shape is an effective and manufacturable solution for torque ripple reduction in WFSMs.
Mesoporous silica-coated Fe3O4-based nanoparticles were synthesized using CTAB and PEG as structure-directing agents. TEM imaging and diffraction revealed multicore core-shell morphologies composed of crystalline iron oxide nanodomains embedded in silica. Nitrogen adsorption-desorption analysis showed that CTAB templating produced a higher surface area mesoporous structure, whereas PEG templating resulted in lower surface area with mixed micro-mesoporosity. XRD and electron diffraction showed reflections consistent with a cubic Fe3O4-type iron oxide structure together with amorphous silica. SQUID magnetometry confirmed superparamagnetic behavior at 300 K. Modified Langevin fitting gave magnetic core diameters of approximately 8-10 nm, in agreement with TEM. The decrease in mass-specific magnetization after silica coating is attributed to dilution by nonmagnetic silica. The results show that template selection controls pore architecture while retaining magnetic responsiveness.
