Videos by Johannes J H Paulides
#aegroup #aeuk #aeelectricdrives #aesuspensiom #aeindia #aemarine #electricisfun
13 views
Papers by Johannes J H Paulides
—In comparison with common combustion motor driven vehicles future electric driven vehicles vary ... more —In comparison with common combustion motor driven vehicles future electric driven vehicles vary more in their reachable range because of the high drive train efficiency and the different energy demand requirements in different drive conditions. One of the main issues to improve the overall range is the minimization of energy consumption. On the one hand an adapted energy management strategy is needed to achieve a more comfortable and acceptable equalization in the range. On the other hand it is important to recuperate mechanical, electrical and thermal energy, both inside and from the vehicles wherever possible. For example, it is thinkable to use the kinetic energy of braking vehicles outside the vehicle to power traffic lights. If a combustion process is used, a lot of waste thermal energy can be re-used from the tailpipe. Or solar energy may be used. If a vehicle has a varying duty-cycle it makes sense to adapt the powertrain to suit the power need. Whatever the available energy, it has to go hand in hand with minimizing the energy consumption. This paper is based on the knowledge and combination of future technologies [2], which are useful to collect and convert energy coming from different sources within a vehicle.
In the last decades the magnetic charge method has proven to be an excellent tool to analyze and ... more In the last decades the magnetic charge method has proven to be an excellent tool to analyze and design permanent magnet based systems. Recent research has shown that inclusion of the relative permeability of magnetic materials is also possible using the extended charge method. Contrary to popular believe, it is also possible to include currents into the equations for the magnetic field using a fictitious magnetization. In this paper the fictitious magnetization is calculated and used to determine the field of a rectangular coil. Compared to the current method to calculate the field, on the same computational hardware and software, a large reduction in calculation time (81 vs 11 seconds) is observed.
Variable flux reluctance machines (VFRM) are an interesting candidate to substitute permanent-mag... more Variable flux reluctance machines (VFRM) are an interesting candidate to substitute permanent-magnet synchronous machines in many applications, mainly owing to the absence of rare-earth permanent magnets and improved field weakening capability. However, the torque ripple is intrinsically large. In this paper, various rotor skewing strategies are researched to minimize the torque ripple of a 12/10 (stator/rotor pole ratio) VFRM, including stepped rotor skewing with 2 modules [1], non-uniformly distributed rotor teeth [2] and conventional rotor skewing. The result shown in this digest is based on finite element simulations, although that also analytic results will be incorporated in the final paper. Specifically, for 12/10 VFRM, the 5 th back-emf harmonic and 6 th self-inductance harmonic of the field winding largely influence the torque ripple. For the method of stepped rotor skewing, the geometry is shown in [1]. The rotor is divided into two skewed segments. This method generates opposing torque ripple contributions, such that the sum of the two reaches a minimum. The skewing angle selection is strongly related to the torque ripple harmonic components. When the torque ripple, induced by the variation of the field winding inductance, is dominant, a skewing angle close to 3 mechanical degree (half of the period of the 6 th harmonic in field winding self-inductance) is the most effective way to minimize the torque ripple. When the torque ripple, created by the harmonics in the back-emf, is dominant, a skewing angle close to 3.6 degree (half of the period of the 5 th back-emf harmonic) is preferred. The relationship between the skewing angle and the reduction of the torque ripple is shown in . The non-uniformly distributed rotor teeth topology is shown in , in which every other tooth is skewed and the skewing angle is the same along the stack length. This topology prevents the identity of the harmonics in coils, such that the superposition of the harmonics is minimized. An example is given in , which shows the 5 th harmonic of the induced emf in different coils. The effectiveness of this topology is shown in , which illustrates the relationship between the skewing angle and the reduction of torque ripple. Both methods can significantly reduce the torque ripple. In the full paper, the aforementioned methods will be explained in detail. Further, the average torque reduction, influence of saturation and ease of manufacture will be discussed.
Permanent Magnets (PMs) are ever more used in high-performance applications such as electrical ma... more Permanent Magnets (PMs) are ever more used in high-performance applications such as electrical machines, linear servo actuators, contactless magnetic bearings, etc. This paper focuses on PM modeling for a contactless lithographic vibration isolation device with an integrated PM-based gravity compensator. This device needs to position a high mass (tons) with micrometer accuracy, with permanent magnets providing zero-power gravity compensation, and active devices ensuring stability and accurate positioning. To minimize the energy consumption of the active devices due to modeling errors it is essential that the PM modeling accuracy is maximized. Further, low calculation time is required for quick and efficient topology optimization. The 3D Finite Element Method (FEM), although accurate, consumes significant amounts of time, especially for models with small stroke compared to overall device dimensions. The 3D analytical charge model, on the other hand, provides field expressions which are exact for ironless structures, i.e. they are not approximations and, in contrary to FEM, remain very time inexpensive. Furthermore , these analytical field solutions exhibit significantly lower noise (150dB) compared to FEM solutions, and consequently provide an increased accuracy, especially at large magnetic field gradients (e.g. the magnet edges). Analytical charge modeling has been commonly used in literature to analytically describe the 3D magnetic fields induced by PMs, e.g. in [1,2] to analyze the field of and the interaction between cuboidal permanent magnets. It has been illustrated in [3] that the PM topology of the gravity compensator is of significant influence on the interaction force. More specifically, topologies which focus the magnetic flux to a small volume, such as the Halbach topology, provide the highest force density. In [3], solely cuboidal PMs were used to investigate the achievable force density for bidirectional arrays and Halbach arrays. Differently shaped PMs can contribute to increased flux focusing properties, hence, increase force density. This paper specifically presents novel analytical field expressions of the pyramidal frustum, or 3D trapezoidal PM (Fig. 1a), for the gravity compensator. It is defined such, that the bottom and top surface are rectangular, parallel, and share the vertical symmetry axis. By reducing the horizontal dimensions of the top surface towards zero, a pyramid is easily obtained. Fourteen rectangular and triangular shaped surfaces are distinguished on the contours of the PM. A charge is applied to each surface, and the respective magnetic fields are obtained analytically as described in [1] and [4], respectively. A summation of these fields results in the exact analytical expressions for the magnetic field exhibited by the 3D trapezoidal PM. The derivation of all surfaces, rotations, field expressions and the summations which result in the total field are elaborated on in the paper. The paper discusses anomalies in the mathematical field description, and methods to overcome those, which results in a smooth field solution. Comparison with 3D FEM solutions (Fig. 2b) demonstrates the high accuracy of the analytical solution (Fig. 2a) for the pyramidal frustum. Besides the pyramidal frustum, additional PM shapes, such as the hexagonal or octagonal prism (Fig. 1b), may contribute to an increased force density. The analytical model of such PMs is obtained by superposition of cuboidal PMs and triangular PMs, described in [1] and [4], respectively. In the final paper, a comparison with FEM shows the high accuracy for such PM shapes. Using the analytical field expressions obtained for exotic-shaped PMs, novel array topologies are investigated. Due to space considerations, this paper presents force density for a selection of possible topologies to be incorporated in the gravity compensator. [1] J.P. Yonnet and G. Akoun, " 3D analytical calculation of the forces exerted between two cuboidal magnets, " (a) pyramidal frustum and (b) top view of hexagonal and octagonal prism showing the cuboidal and triangular shapes (a): analytically obtained absolute value of the magnetic flux density, parallel to one of the side surfaces (projected in the figure). (b): difference with results obtained by FEM.
A semianalytical framework is presented for the analysis of synchronous reluctance motors. The mo... more A semianalytical framework is presented for the analysis of synchronous reluctance motors. The model is based on a recently refined definition of the harmonic modeling technique for electromagnetic actuators in polar coordinates, and includes the magnetic field solution in the soft-magnetic motor parts. In order to implement the model, a polar representation of the motor geometry is required. Therefore, a method to define the polar geometry representation is proposed and applied to model a benchmark motor topology. The calculation results of the benchmark motor are then validated against finite-element analysis predictions acquired for the actual geometry of the motor. This comparison shows that good agreement is obtained.
—This paper concerns an initial simplified steady-state method to start to evaluate the behavior ... more —This paper concerns an initial simplified steady-state method to start to evaluate the behavior of a dc-excited flux-switching motor at a steady-state, or slow dynamic, vehicle level. Starting from a simplified finite-element motor representation, a two-dimensional co-energy lookup table model is extracted. Then, a model reduction is performed to create steady-state loss maps of the system parts at different torque-speed points, considering the simplified losses. These loss maps are integrated into an electric vehicle model, where a discussion is introduced to illustrate the limitation of using efficiency maps. Finally, to determine an average efficiency of the electrical powertrain during drive cycles, a fast simplified steady-state calculation is performed.
A widely used method for the magnetic field calculation in permanent magnet (PM) machines is the ... more A widely used method for the magnetic field calculation in permanent magnet (PM) machines is the harmonic modeling (HM) method. Despite its many advantages, the application of this method is limited to machines in which armature currents, as a source of the magnetic field, are known. Since most of the PM machines are supplied with three-phase voltages, any variation of parameters in the armature electric circuit could lead to the uncertainty in knowing the armature currents, and therefore, limit use of the HM method. In this paper, an approach for overcoming this limitation is presented, which enables the calculation of the magnetic field in the PM machines without a priori knowledge of the armature currents. Index Terms— Harmonic modeling (HM), high-speed machines, permanent magnet (PM) machines, rotor eddy currents.
Recent advances have been made into modeling the soft-magnetic materials using the surface charge... more Recent advances have been made into modeling the soft-magnetic materials using the surface charge model. Until now, the model has only been used to model the field outside the soft-magnetic material. To include the saturation in the future, the magnetic field inside the material should be known. In this paper, the model is used to calculate the field inside the material when three different shapes of boundaries are considered. Comparing the results with Finite Element Method shows that the differences in the air region are <1 mT and those in the soft-magnetic region are <4 mT. Larger discrepancies can, however, occur near the corner of the material.
High-speed permanent magnet machines, supplied by pulsewidth modulation (PWM) voltage source inve... more High-speed permanent magnet machines, supplied by pulsewidth modulation (PWM) voltage source inverters, operate with distorted stator currents. Harmonics present in these stator currents deteriorate the machine performance by generating losses. Mostly, these losses are following the machine design using a transient finite-element model. The precise measurement of these rotor eddy-current losses is extremely difficult, hence, only a few papers provide convincing comparisons between predictions and measurements. This paper presents a fast and precise analytical approach, verified with measurements, to consider rotor losses of machines, supplied by PWM voltages, already during the design procedure. Index Terms— High-speed permanent magnet (PM) machines, pulsewidth modulation (PWM) losses, rotor eddy-current losses.
With the ever advancement and increased force density of slotted brushless linear permanent magne... more With the ever advancement and increased force density of slotted brushless linear permanent magnet (PM) actuators the attraction forces become more dominating, where in the early years this attraction force was advantageous it now becomes a problem due to the reduced bearing lifetime.
Hybrid or full electric propulsions for inland ships are becoming more popular. In these applicat... more Hybrid or full electric propulsions for inland ships are becoming more popular. In these application, direct-drive PM propulsion motors are a preferred machine configuration. This paper discusses the challenges to determine the losses, as estimated with simulations, during the testing procedures of a 350kW at 300rpm, respectively. The full-load testing of the drive system is performed by mechanically coupling two identical machines, of which one operates as a motor and the other as a generator, or " back-to-back " testing configuration. Two Direct-Drive PM machines have been manufactured to validate key findings from the modelling, particularly in terms of loss predictions, thermal modelling and influence of the design features such as magnet segmentation. A back-to-back setup is created for testing these machines with a speed range of 0-450 rpm. Before the measurement commended, tests were carried out in accordance with IEC60034-1, IEC60034-15, IEC60085-1, IEEE43, IEEE118 and Lloyd's register. These tests included: surge, resistance, winding symmetry, high voltage test, insulation resistance and polarization index. All these tests were successfully completed and agreed with the analysis as described before. Following the motors have been installed in an inland ship hybrid propulsion.
—This paper provides a comparative analysis of three types of electrical drives with rare-earth-f... more —This paper provides a comparative analysis of three types of electrical drives with rare-earth-free reluctance motors for next-generation 48V mild hybrid automotive applications. The drives with switched reluctance motors (SRM), variable flux reluctance motors (VFRM) and synchronous reluctance motors (SynRM) are compared considering aspects like complexity of control algorithms and the structure of the power electronic circuits, construction of the motors and their performance characteristics, such as torque density, power factor, acoustic noise, and efficiency. The analysis is based on the results of simulation using 2D FEM and Simulink together with published research outcomes. The analyzed drives are compared to the internal permanent magnet motor (IPM) drive, selected as a benchmark. Index Terms—Rare-earth-free reluctance motor drive; automotive ; mild hybrid powertrain.
Highlights • We have developed an harmonic modeling technique for the analysis of slotted inducti... more Highlights • We have developed an harmonic modeling technique for the analysis of slotted induction motor structures. • We have examined two different methods to represent the slots and slot opening in the model. • The results of the model have been compared to 2D FEA simulations for a set of four test motors. • The obtained magnetic field solution is in relatively good agreement with the magnetic field predicted by linear 2D FEA simulations. • The electromagnetic torque calculated from the harmonic model is in very good agreement with linear 2D FEA simulation results. Abstract A fast, semi-analytical model for induction motors (IMs) is presented. In comparison to traditional analytical models for IMs, such as lumped parameter, magnetic equivalent circuit and anisotropic layer models, the presented model calculates a continuous distribution of the magnetic flux density in the air gap and both the stator and rotor slots of the IM. Due to its semi-analytical nature, the computation time of the model is short in comparison to finite element analysis (FEA) simulations. However, the curved shapes of typical IM slots require a suitable representation of each slot in polar coordinates. Two different polar slot representations are investigated and it is shown that a double region representation is favorable to a single region representation. Finally, the model is implemented for four different benchmark motors. The calculation results are shown to be in good agreement with linear 2D FEA simulation results. This makes the model a suitable candidate for fast design tools.
In this study, six different modelling methods for permanent magnet electric machines are compare... more In this study, six different modelling methods for permanent magnet electric machines are compared in terms of their computational complexity and accuracy. The methods are based primarily on conformal mapping, mode matching, and harmonic modelling. In the case of conformal mapping, slotted air gap of a complex machine geometry is transformed to a smooth slotless air gap where analytical expression for field solution is available. The solution in the canonical domain is then mapped back to the original slotted air-gap domain. Mode matching or subdomain method, as it is called in different sources, is using a solution of Laplace's equation to model the slotted air gap. In harmonic modelling, the machine cross-section is divided into homogeneous regions that are represented using Fourier series and coupled with each other using boundary conditions. The boundary value problems in both the mode matching and harmonic models are solved to obtain the field solutions. The performance of the modelling methods are evaluated by comparing the global parameters such as cogging torque, electromagnetic torque, back-emf as well as the simulation time with the results of finite-element transient analysis.
—Recently, series-hybrid drivetrains are given more emphasis over parallel-hybrid drivetrains due... more —Recently, series-hybrid drivetrains are given more emphasis over parallel-hybrid drivetrains due to their simplicity, freedom in implementation and higher efficiency. However, the modeling phase of such machines takes either long calculation time with numerical methods such as Finite Element Analysis (FEA) or produces low accurate results with Magnetic Equivalent Circuit (MEC) models. In this paper a method is used to build a reluctance network where high accuracy is obtained yet still with acceptable calculation time.
Purpose – The space-mapping (SM) optimization technique, with its input, implicit or output mappi... more Purpose – The space-mapping (SM) optimization technique, with its input, implicit or output mapping-based implementations, provides a basis for computationally efficient engineering optimization. Various algorithms and design optimization problems, related to microwave devices, antennas and electronic circuits, are presented in numerous publications. However, a new application area for SM optimization is currently expanding, i.e. the design of electromechanical actuators. The purpose of this paper is to present an overview of the recent developments. Design/methodology/approach – New algorithm variants and their application to design problems in electromechanics and related fields are briefly summarized. Findings – The paper finds that SM optimization offers a significant speed-up of the optimization procedures for the design of electromechanical actuators. Its true potential in the area of magnetic systems and actuator design is still rather unexplored. Originality/value – This overview is complementary to the previous published reviews and shows that the application of SM optimization has also extended to the design of electromechanical devices.
Linear motors find their utilization in an increasing number of industrial applications. Permanen... more Linear motors find their utilization in an increasing number of industrial applications. Permanent magnet linear synchronous motors PMLSMs are favorable in many applications due to their servo characteristics, robustness, and high force density. The major disadvantage of moving coil type PMLSMs is the cable slab that energizes the coils from the fixed world to the moving parts of the machine. These cable slabs introduce extra wear and dynamical distortions. Moreover, in precision application the cable slab is supported by additional linear drives. These disadvantages can be eliminated if the coils could be powered wirelessly. In this paper two topologies are proposed that are capable of transferring 1 kW of power at most. The transformer part of the CET is characterized by means of two dimensional finite element analysis, and the influence of using additional capacitors to boost the output power is examined. Furthermore, an analysis of the core losses is conducted. Conclusions are drawn from the results.
Uploads
Videos by Johannes J H Paulides
Papers by Johannes J H Paulides