Conference Agenda

A field-circuit coupled finite element calculation model embedded loss separation theory is proposed, which can effectively deal with the nonlinear core loss of transformer caused by subsynchronous frequency components injection. In order to deal with the increased time period due to the subsynchronous frequency components, and the problem of excessive loss current and too small parallel resistance due to zero-crossing point of voltage, the Parareal algorithm is introduced to solve the model. The algorithm was verified with TEAM Problem 21 A and physical transformer experiments.

During the process of additive manufacturing of permanent magnets, a magnetic field is applied to create an anisotropy in the magnet. To generate a homogenous magnetic field with the least temperature rise, different coil topologies are discussed and compared with regard to the homogeneity and the power loss of the coils. Compared to other coil topologies, the Braunbeck coil is selected as the best solution in this application, because it can generate the magnetic field with the largest homogeneous area, and requires only small currents, which results in a lower power loss and temperature rise at the surface of the coil. Finally, a liquid cooling system is proposed for the Braunbeck coil.

A model order reduction (MOR) of induction motor using the Cauer ladder network is developed in the frequency domain. A multiport frequency transformation between the stator and mover domains is derived by neglecting spatial harmonic interactions. Even after neglecting the harmonic interactions, the reduced model provides reasonably accurate frequency response.

In this paper, we describe an integrodifferential formulation for modeling non-simply connected topologies. It’s an iterative process based on an integral analysis and a finite elements analysis. It has the benefit to consider only conductive areas. Thus, air and dielectrics are not meshed, which allows reducing significantly the size of the solved system. The performance of the suggested method was proofed by measurements and classical finite elements calculation.

The magnetotelluric (MT) method is a geophysical electromagnetic induction technique for imaging the electrical conductivity and structure of the Earth. It exploits the disruption on Earth’s electric (E) and magnetic (H) fields, caused by solar wind, to capture complex conductivity structures. Several inversion methods are routinely applyed to analyze near-surface measures of the electric and magnetic field, leading to conductivity heat maps of Earth structure. This work compares the performance of two metaheuristics against an exact method in solving the 1D MT inverse problem. The selected metaheuristics were particle swarm optimization (PSO) and differential evolution (DE), while the exact method was the generalized inverse technique (GI). The experiments show metaheuristics outperformed GI for the selected 1D instance.

The modeling of the capacitive phenomena including the inductive effects becomes critical, especially in the case of power converter with high switching frequencies supplying an electrical device. To capture the coupled capacitive-inductive effects, the Darwin model is invoked in frequency domain, which only neglects the radiation effects of the full Maxwell system. In this work, we are interested in the comparison of the different solvers, both iterative and direct, for ungauged and gauged finite-element (FE) system for Darwin model. In addition, to handle the huge FE resulting matrix due to the industrial application, a novel formulation is proposed.

For the problem of solving the total electric field near HVDC electrical equipment, this paper proposes a method for updating the charge density near the insulating medium. According to the electrostatic field simulation results, the charge density is determined to update the switch state, and then the upwind finite element method (FEM) is used to update the node surface charge density. Finally, the total electric field of the 500 kV DC Voltage divider is calculated. The results show that the existence of space charges can improve the electric field distribution on the surface of the insulated insulator sheds of the DC voltage divider. In the space far away from the high-voltage electrode, the total electric field strength will increase. The method in this paper provides guidance for the external insulation of the total electric field near the DC divider.

Solar power generation systems are used as representative renewable energy due to low installation cost and simple maintenance. For this reason, many studies are being conducted every year to improve the performance of solar power generation. Solar power generation is affected by the amount of power generated by the surrounding environment, such as partial shade and old objects. In order to improve this, a micro-converter system was developed, but the buck converter used in the existing micro-converter operates in a bootstrap driving method, which causes drive loss and increases the unit cost for circuit configuration. In this paper, to solve this problem, we propose a negative buck converter in which the positions of the switch and inductor are changed. Since the negative buck converter has a MOSFET connected to the ground, switching control is possible through a direct drive method. Accordingly, the circuit is simplified and the driving loss can be reduced. In order to verify this, simulation characteristics through modeling of the existing Buck converter and Negative Buck converter were compared and their validity was verified through experiments.

The air gap flux density of a motor has very important impact on its performance. In order to effectively improve the performance of permanent magnet motor, asetof splicing permanent magnets is introduced. The distribution of the permanent magnet can both improve the magnetic flux density amplitude of the permanent magnet motor at the wave crest and restrain the magnetic density distortion at other positionsSo the proposed technique can effectively improve the electromagnetic density of the motor. In this paper, a 8-pole 12 slot permanent magnet motor is taken as an example, the analytical model is constructed, and the optimal size of splicing permanent magnet is obtained through finite element parametric analysis. The rationality and accuracy of the design by numerical computation has been also verified by comparing the results from using analytical method.

In this paper, a DC modeling approach is proposed to analyze the inductance variation of pancake coils made of 1G HTS tape. The model takes into consideration the non-uniformity of the current density in the tape section using a power minimization criterion. Experimental and numerical studies are combined for a better analysis of the variation of the inductance with the applied current in HTS coils.

For the purpose of local forces calculation, permanent magnet models differ depending on the underlying assumption about the magnetoelastic behaviour. The usually considered models are recalled and a new model, assuming that magnetic moment does not depend on strain, is proposed. For each model the corresponding Maxwell tensor is presented and then used as a source for the resolution of an elasticity problem. An axial flux permanent magnet machine is analysed as an application. The permanent magnet models result in very different localization of forces. It is shown that, at the rotor/bearing interface, where stress concentrates, all models give similar results. In other parts of the structure, with lower stress values, differences can be observed between stress distributions.

A mesh-based magnetic equivalent circuit (MEC) model has been recently developed as part of an existing thermoelectric lumped-parameter model for the transient simulation of High Temperature Superconductor (HTS) field coil synchronous generator. This model allows the inclusion of the magnetic flux density inside the definition of the HTS resistance in addition to the temperature and current. The objective is to gain further accuracy on the thermal and electromagnetic behavior of the superconducting machine during the partial or total loss of the superconducting state of the HTS winding. In the present work, the MEC is firstly cross-checked against a Finite Element Analysis (FEA) using a magnetostatics approximation. Then, the MEC is used for studying the thermoelectric and particularly the magnetic response of a 100 MVA hydroelectric HTS field coil synchronous machine applying a three-phase shortcircuit at its terminals.

High temperature superconductors (HTS) are strong material candidates for the development of future electrical machines due to their high energy density at low AC losses resulting in greater electrical efficiency for lighter and compacter designs than traditional machines. The features are particularly appealing for transports. The present work focuses on the electromagnetic modelling of a superconducting synchronous generator for aeronautic applications with a particular interest in the coupling between Finite Element Model (FEM) and external electrical circuits. The FEM implements different formulations of the Maxwell equations, namely H-A, T-A and the classic A-V , which can also be combined with the H, to find the best suited combination FEM - electrical circuit for computation load reduction (time and memory). The external circuit simulates a power converter interconnecting the machine to the load. The performance of both the machine as a whole and the superconductor is evaluated for different load conditions.

We discuss the relevance of several finite-element formulations for systems containing high-temperature superconductors (HTS) and ferromagnetic materials (FM) in the context of a 3D motor pole model, in terms of their numerical robustness and efficiency. We propose a coupled h-φ-a-formulation as an optimal choice, modeling the problem with an a-formulation in the FM, and an h-φ-formulation in the remaining domain.

In this paper, we present a fast model to calculate the iron losses for thick laminations used in the pole shoes of large synchronous machines. The method is based on approximating the dynamic relationship between the surface magnetic field strength and the average magnetic flux density in the lamination cross-section by an analytical relationship parametrized both in terms of the width and thickness of the lamination. Initially, 2-D finite element simulations were performed in the lamination cross-section by considering different lamination widths. Then, the proposed model was fitted against the finite-element simulation results, and reasonably accurate results were obtained. The proposed model is planned to be used in conjunction with the finite-element analysis, where the field quantities are numerically computed and then losses are calculated in a post-processing manner.

Axial Green function methods(AGMs) are designed for calculating the numerical solution of governing equations in complicated domains. Despite the advantage that AGMs are fit for complicatd domains, unlike the Dirichlet boundary condition, the Neumann boundary condition is not easy to implement unless the boundary for the condition to be imposed on is parallel to an axis. In order to treat with the Neumann boundary condition for the electric potential in two-dimensional complicated domains, finding the way to impose that condition on a line segment that is not parallel to any axis, we apply this approach to electric field computations in complicated domains with the Neumann boundary condition on curved boundaries.