Search results

This paper presents various available and new techniques for prediction of the hysteresis loop, no‐load current curve and hysteresis losses. It is shown that linearization is a convenient method to be employed for quick estimation of the hysteresis loop with acceptable accuracy. Although the third and fifth order functions for saturation curve prediction lead to more accurate results, it requires more data and also more complicated equations resulting in longer computation time. Use of various third and fifth order functions for saturation curve are the noticeable advantages of the techniques. The three proposed techniques could predict the hysteresis loop of transformers using simple experiments and iterative computer computations.

Surface heat treatment by induction heating (10-100 kHz) requires precise prediction and control of the depth of the induced phase transformation. This paper aims at identifying common issues with the measurement and modeling of magnetic properties used in induction heating simulations, and it proposes ways to improve the situation.

In particular, it is demonstrated how intrinsic magnetic properties (i.e. the B-H curve) of a sample can change during the magnetic characterization process itself, due to involuntary annealing of the sample. Then, for a B-H curve that is supposed perfectly known, a comparison is performed between multiple models, each one representing the magnetic properties of steel in time-harmonic (TH) finite element method simulations. Finally, a new model called “power-equivalent model” is proposed. This model provides the best possible accuracy for a known nonlinear and hysteretic B-H curve used in TH simulations.

By carefully following the guidelines identified in this paper, reduction of errors in the range of 5-10 per cent can be achieved, both at the experimental and modeling levels. The new “power-equivalent model” proposed is also expected to be more generic than existing models.

This paper highlights common pitfalls in the measurement and modeling of magnetic properties, and suggests ways to improve the situation.

The purpose of this paper is to propose an accurate model for simulation of inrush current in power transformers with taking into account the magnetic core structure and hysteresis phenomenon. Determination of the required model parameters and generalization of the obtained parameters to be used in different conditions with acceptable accuracy is the secondary purpose of this work.

The duality transformation is used to construct the transformer model based on its topology. The inverse Jiles-Atherton hysteresis model is used to represent the magnetic core behavior. Measured inrush waveforms of a laboratory test power transformer are used to calculate a fitness function which is defined by comparing the measured and simulated currents. This fitness function is minimized by particle swarm optimization algorithm which calculates the optimal model parameters.

An analytical and simple approach is proposed to generalize the obtained parameters from one inrush current measurement for simulation of this phenomenon in different situations. The measurement results verify the accuracy of the proposed method. The developed model with the determined parameters can be used for accurate simulation of inrush current transient in power transformers.

A general and flexible topology-based model is developed in PSCAD/EMTDC software to represent the transformer behavior in inrush situation. The hysteresis model parameters which are obtained from one inrush current waveform are generalized using the structure parameters, switching angle, and residual flux for accurate simulation of this phenomenon in different conditions.

A major purpose of vector hysteresis models lies in the prediction of power losses under rotating magnetic fields. The well-known vector Preisach model by Mayergoyz has been shown to well predict such power losses at low amplitudes of the applied field. However, in its original form, it fails to predict the reduction of rotational power losses at high fields. In recent years, two variants of a novel vector Preisach model based on rotational operators have been published and investigated with respect to general accuracy and performance. This paper aims to examine the capabilities of the named vector Preisach models in terms of rotational hysteresis loss calculations.

In a first step, both variants of the novel rotational operator-based vector Preisach model are tested with respect to their overall capability to prescribe rotational hysteresis losses. Hereby, the direct influence of the model-specific parameters onto the computable losses is investigated. Afterward, it is researched whether there exists an optimized set of parameters for these models that allows the matching of measured rotational hysteresis losses.

The theoretical investigations on the influence of the model-specific parameters onto the computable rotational hysteresis losses showed that such losses can be predicted in general and that a variation of these parameters allows to adapt the simulated loss curves in both shape and amplitude. Furthermore, an optimized parameter set for the prediction of the named losses could be retrieved by direct matching of simulated and measured loss curves.

Even though the practical applicability and the efficiency of the novel vector Preisach model based on rotational operators has been proven in previous publications, its capabilities to predict rotational hysteresis losses has not been researched so far. This publication does not only show the general possibility to compute such losses with help of the named vector Preisach models but also in addition provides a routine to derive an optimized parameter set, which allows an accurate modeling of actually measured loss curves.

In the modeling of transducers, especially magnetic transducers, hysteresis may affect performance. Hysteresis models have been improved greatly and are capable of modeling a large variety of rate‐independent phenomena, and are capable of describing minor loops. Of these, the most useful are: the Preisach model, the play model, and the stop model. Coupling these purely magnetic models with other phenomena, such as magnetostriction, enhances the model’s usefulness for transducer applications. This paper will discuss the conditions under which these models may be inverted, and for the invertible media, a technique for inverting them.

On the basis of the Kolmogorov‐Arnold theory, the feedforward type artificial neural networks (NNs) are able to approximate any kind of nonlinear, continuous functions represented by its discrete set of measurements. A NN‐based scalar hysteresis model has been constructed preliminarily on the function approximation ability of NNs. An if‐then type knowledge‐base represents the properties of the hysteresis characteristics. Vectorial generalization to describe isotropic and anisotropic magnetic materials in two and three dimensions with an original identification method has been introduced in this paper.

The Jiles‐Atherton model of magnetization gives a general description of magnetic induction and hysteresis loss against magnetic field and corresponding current. At wider regions of magnetization, however, the differences between calculated and empirical values become noticeable, especially in the case of hysteresis loss. It seems that these inaccuracies are caused to a large degree by the variation of the optimum values of model parameters in the successive regions of magnetization, and in particular by the change of the pinning coefficient. The application of a simple linear dependence of this parameter on the amplitude of magnetization has substantially improved the accuracy of the model.

This work introduces an efficient and accurate technique to solve the eddy current problem in laminated iron cores considering vector hysteresis.

The mixed multiscale finite element method based on the based on the T,Φ-Φ formulation, with the current vector potential T and the magnetic scalar potential Φ allows the laminated core to be modelled as a single homogeneous block. This means that the individual sheets do not have to be resolved, which saves a lot of computing time and reduces the demands on the computer system enormously.

As a representative numerical example, a single-phase transformer with 4, 20 and 184 sheets is simulated with great success. The eddy current losses of the simulation using the standard finite element method and the simulation using the mixed multiscale finite element method agree very well and the required simulation time is tremendously reduced.

The vector Preisach model is used to account for vector hysteresis and is integrated into the mixed multiscale finite element method for the first time.

The purpose of this paper is to reduce computation time of magnetic characteristic analysis considering 2D vector magnetic properties.

The paper proposes a complex E&S modelling with assumption that both flux density and field strength waveforms are sinusoidal. The computation time of the complex E&S modeling becomes 1/10 in comparison with one of the conventional E&S modeling. This modeling is applicable up to 1.4 T of the local magnetic flux density condition in the case of non‐oriented magnetic materials.

In the results of the magnetic field analyses of a linear‐induction motor model core by means of the finite element method taking account of the complex E&S modeling, the distributions of the flux density and the field strength were able to be approximately analyzed and their phase differences in space were represented. The results of the magnetic characteristic analysis of the linear‐induction motor showed that the teeth‐end shape had large influences on the thrust and cogging.

This technique helps to know approximately local vector magnetic properties in core materials. This modeling is very useful for magnetic core design taking account of the simplified 2D vector magnetic properties.

The method presented in this paper enables expression of the simplified 2D vector magnetic properties in magnetic field analyses. The computation time can be considerably reduced in comparison with the conventional method.

The paper presents a method of creating electrical equivalent diagrams of magnetic circuits. The method is based on bond‐graph techniques, using flux derivative as flow variable. Couplings between magnetic and electric part of the system are represented by gyrators. Simple models of magnetic branches, including non‐linear effects due to saturation, hysteresis and eddy currents, are presented. These models can be easily combined into magnetic circuit models, which can be transformed into dual electric equivalent circuit. Various equivalent circuits of transformers are discussed. The proposed models are simple and particularly useful for system‐level simulation of power electronic and motion control systems with magnetic elements. These models can be easily developed if needed. Theoretical considerations are illustrated by examples of digital simulation and experimental results.