Browsing by Author "Belahcen, Anouar, Prof., Aalto University, Department of Electrical Engineering and Automation, Finland"

Now showing 1 - 7 of 7

Circulating and Eddy-Current Losses in Random-Wound Electrical Machines
(2017) Lehikoinen, Antti
School of Electrical Engineering | Doctoral dissertation (article-based)
This thesis focuses on resistive winding losses in rotating electrical machines, with special attention on circulating currents in random-wound machines. Five methods are developed for computationally efficient analysis, including one analytical circuit model and four 2D finite element models. The circuit model, as well as two of the numerical models, focus on circulating currents, while the remaining two numerical models also include all eddy current effects. Furthermore, a stochastic approach is proposed to model the uncertainty included in random-wound windings. Finally, a large set of measurement data concerning high-speed induction machines is analysed and used to validate the stochastic approach. According to simulations, all of the developed models yield reasonably accurate results at a significantly reduced computational cost. The two numerical methods that also consider eddy current effects are found to be particularly precise. The stochastic approach is also found to be able to predict the variation in measured circulating current losses. The measured machines are observed to exhibit significant circulating current losses. On average, they increase the winding losses by 60 %. Furthermore, the losses are found to vary significantly even between nominally identical machines. Finally, according to simulations, also medium-speed machines with 50 Hz rated supply are found susceptible to high circulating current losses, especially if supplied by an inverter.
Magneto-vibro-acoustic Computational Techniques for Electrical Machines
(2020) Sathyan, Sabin
School of Electrical Engineering | Doctoral dissertation (article-based)
This thesis investigates the electromagnetic, magnetomechanical and vibro-acoustic interactions in an electrical machine, and proposes numerical computational tools and experimental techniques for modeling these phenomena in electrical motors. In an electrical motor, the supplied electrical energy goes through different physical domains in the form of magnetic, mechanical, acoustic and thermal energy. In this thesis, computational methods for magnetic forces, magnetostriction, mechanical deformations, vibrations, and finally the acoustic sound produced by electrical motors are successfully implemented. The study starts with the analysis of magnetic forces in ferromagnetic materials and electrical motors by crafting finite element methods for computing these forces and validating them with experiments. From there, the investigations proceed to structural mechanics zone, where the effects of magnetic forces and magnetostriction on the deformations and vibrations of electrical steel sheets and the stator core of electrical motors are studied using numerical methods and experimental justifications. The causes of vibrations in two induction motors were examined with computational and experimental methodologies. The generation of acoustic noise due to mechanical vibrations is modeled using a numerical method that is based on a combination of the finite-element and boundary-element methods. The results and findings in the thesis from numerical and experimental studies provide efficient tools for analyzing and computing various quantities pertaining to the behavior of ferromagnetic materials and electrical motors. These methods can be applied in practical facets such as the design and condition-monitoring stages of electrical motors. This thesis has succeeded in addressing different domains of energy conversions and their effects in electrical machines by developing reliable and proficient computational tools and experimental methods in the realms of magnetics, mechanics and acoustics.
Material characterization, modeling, and incorporation of the models in the machine simulation of large-diameter synchronous machines
(2023) Gürbüz, Ismet Tuna
School of Electrical Engineering | Doctoral dissertation (article-based)
This dissertation presents a comprehensive methodology for the realistic and computationally efficient simulation of large-diameter synchronous machines accounting for the effect of cutting on iron losses and magnetization. To achieve this, magnetic materials used in the machine parts are experimentally characterized, material models based on the characterization are developed, the developed models are incorporated into finite-element (FE) simulation software, and machine simulation is performed with the incorporated models. Non-oriented punched electrical steel sheets used in the stator laminations are studied experimentally under different uniaxial stress conditions using a modified single-sheet tester. It is shown that the effect of stress on the iron losses of the punched samples differs based on the extent of degradation observed in the samples following the cutting procedure. Subsequently, in the material modeling, the focus is given to the modeling of punching. A continuous material modeling approach with an exponential deterioration profile is used for the magnetization, and iron losses are modeled similarly by modifying the coefficients of Jordan's method. Thick laser-cut steel laminations used in the rotor poles are studied experimentally using a ringcore measurement system. The characterization of the material properties and iron losses is then achieved by a 2-D axisymmetric FE modeling of the lamination cross-section with the inclusion of a continuous local material model using a quadratic deterioration profile. It is shown that the inclusion of the edge effects for the thick laminations is needed, which requires a 2-D modeling. In light of this, a simple 2-D analytical model is developed for eddy-current loss computation. To achieve a computationally efficient and accurate implementation of cutting deterioration into electromagnetic FE simulation, a new methodology for numerical integration is proposed. The validity of this approach is confirmed by comparing it to the analytical solution for a 2-D beam geometry. Subsequently, the method is utilized in the 2-D FE simulation of transformers, resulting in an enhanced computational efficiency when compared to existing methods. Time-stepping simulation of the studied large-diameter synchronous machine is achieved with the incorporated models developed for the stator laminations and rotor poles following the proposed methodology. The effect of cutting on the loss components and machine operating points is analyzed. The results demonstrate that accurate incorporation of the cutting effect in the machine simulation increases the machine's losses by 16.4 kW, necessitating improved cooling capabilities.
Model Order Reduction for Simulation and Control of Synchronous Machines
(2019) Farzam Far, Mehrnaz
School of Electrical Engineering | Doctoral dissertation (article-based)
Model order reduction approaches aim at reducing the computational complexity of numericalmodels. The reduction is achieved by lowering the dimension of the state space or the numberof degrees of freedom in the original high-order model, which in return generates a reducedorder model. This reduced order model is an appropriate substitution for the original model inthe applications with restricted computational resources or a demanding large number of simulations. This thesis proposes a novel method, named the orthogonal interpolation method, to reducethe computational burden of numerical models, which makes the models suitable for real-timeexecution. This method is applied to a 2-D finite element model of a 2.2 kW interior permanentmagnet synchronous motor. According to the simulation results, the resulting reducedmodel accurately imitates the behaviour of the finite element model of the machine, and successfully lowers the computational time and memory requirements. Furthermore, the proposedmethod grants a more significant reduction in the computational complexity comparedto other model order reduction techniques, such as proper orthogonal decomposition coupledwith a discrete empirical interpolation method. As an application, the proposed reduced model is employed in a control system for real-timecontrol of the motor. The high computational efficiency of the reduced model allows direct implementation of the resulting control system in the embedded processor of the drive. Moreover,the simulation and the experimental results show the capability of the developed controlsystem in considering the magnetic cross-coupling and saturation phenomena of the motor,and therefore produces higher torque and output power.
Models of Magnetic Anisotropy for Nonoriented Silicon Steel Laminations of Electrical Machines
(2022) Upadhaya, Brijesh
School of Electrical Engineering | Doctoral dissertation (article-based)
This dissertation deals with magnetic anisotropy and its impact on the flux density and loss distribution for non-oriented silicon steel laminations. Magnetic material models, such as anhysteretic reluctivity, Bergqvist’s vector Jiles-Atherton and energy-based hysteresis models, are considered in this work. The vector hysteresis models mentioned above are extended: first, to account for magnetic anisotropy observed in non-oriented silicon steel, and second, to predict hysteresis losses for the input excitation rotating in the plane of the sheet, as well as alternating in different directions. Non-oriented silicon steel lamination, commonly used in a rotating electrical machine’s magnetic core, shows a significant level of magnetic anisotropy. Numerical analysis tools often assume the laminated core is isotropic. Preliminary investigations reveal that the anisotropy alters the distribution of magnetic flux density in the core. Thus, some core parts get saturated, which may adversely affect the core losses. Hence, determining how the losses predicted by the anisotropic models differ from their isotropic counterparts is paramount. Firstly, an extension of the isotropic anhysteretic reluctivity model is proposed to account for magnetic anisotropy. After preliminary investigations, the ideas from the anisotropic reluctivity model are then incorporated into the advanced hysteresis models. Hysteresis models, such as Bergqvist's vector Jiles-Atherton and energy-based models, rely upon the anhysteretic properties for quantitatively describing the intrinsic anisotropy. Anhysteretic magnetic characteristics are identified from unidirectional alternating measurements in several directions to include intrinsic anisotropy in the models. Moreover, to reasonably predict the hysteresis losses, the model parameters are allowed to depend on the magnitude and polar direction of the flux density vector. Secondly, the anisotropic models are coupled with the finite element method. This coupling is then applied to simulate magnetic fields and associated losses in the commonly used measurement setups such as a toroidal inductor, round rotational single sheet tester, and a transformer-like device. The results of the finite element simulations are briefly compared and discussed. Furthermore, the material level validation of the proposed hysteresis models is in the peer-reviewed articles at the end of this thesis.
Multiaxial Magneto-Mechanical Interactions in Electrical Steel Sheets
(2018) Aydin, Ugur
School of Electrical Engineering | Doctoral dissertation (article-based)
This thesis studies multiaxial magneto-mechanical interactions in electrical steel sheets with particular attention to non-oriented electrical steel sheets. Commonly, during the magneto-elastic analysis of electromagnetic applications only uniaxial stress models are used where the effect of multiaxial stress is often neglected. A new rotational single sheet tester (RSST) that is capable of applying arbitrary in-plane magneto-mechanical loading to steel sheets is designed and manufactured. Experiments on a non-oriented electrical steel sheet are performed for analyzing the effect of multiaxial stress on magnetic properties and iron losses in the material. Performed experiments reveal that the effect of multiaxial stress on magnetic properties and iron losses can be much more significant than that of uniaxial stress. A simplified multiscale (SM) and a macroscopic Helmholtz energy based (HE) model are used to model the multiaxial magneto-mechanical behavior of non-oriented electrical steel sheets and their prediction capabilities are compared when limited measurement data is available for identification. The models were studied for modeling both anhysteretic and hysteretic behavior of three different materials that were characterized by different measurement setups. Comparison of the modeling results to the measured results shows that the SM model is accurate for certain cases, whereas the HE model is successful for all three materials. In order to predict the effect of multiaxial stress on hysteresis and excess losses, a stress dependent iron loss model is developed utilizing statistical loss theory. The model is verified with measurements obtained from the manufactured RSST. For validation purposes, both the HE model and the developed loss model is implemented to a 2D finite element model of a transformer that is under mechanical stress. The applicability of the models is proven by comparing the modeling results to the measurements. It is concluded that if mechanical stresses are present in an application, using conventional methods to calculate the losses can lead to inaccurate results.
Power Balance in the Finite Element Analysis of Electrical Machines
(2017) Silwal, Bishal
School of Electrical Engineering | Doctoral dissertation (article-based)
This dissertation deals with the study of the power balance in the numerical simulations of electrical machines. In the numerical analysis of electrical machines using the finite element method, several methods and techniques to compute torque and force exist, but the existing methods are not free from accuracy issues. In this dissertation, approaches based on the power balance of the machine to obtain torque and force are presented, and the possibility to use the power balance approach to validate the existing methods is shown. Both healthy machines and machines with an eccentric rotor have been considered. The applicability of the power balance approach to study the electromagnetic damping of the mechanical vibrations during eccentricity is also assessed. This dissertation carefully compares a conventional torque and force computation method and the power balance approach, based on the finite element mesh used in the air gap of the machine during simulations. Variations in the air gap mesh is brought by changing the number of layers of elements and the layer used for the rotation of the rotor and for the torque and force computation. Results show that the conventional method suffers from an accuracy issue mainly related to the finite element discretization of the air gap of the machine. Variations in the air gap mesh do not affect torque from the power balance. However, force computation from the power balance is not yet robust. The influence of the rotor eccentricity on the torque of the machine was also studied. Results show that an eccentric machine does not exhibit the same torque as a healthy machine. The harmonic components around the principal slot harmonic are most affected. In this thesis, a measurement set-up to measure the torque harmonics of a machine was designed. The measured results were compared with the simulated results.