Pub Date : 2022-10-01DOI: 10.1109/CEFC-LONG57069.2022.10107547
Hendrik Hensel, Marvin-Lucas Henkel, N. Haussmann, C. Jörgens, S. Stroka, M. Clemens
High voltage alternating current gas insulated systems are utilized in narrow spaces. In operation electric field peaks can occur and may impair the functioning of the system. The knowledge of the electric field distribution in gas insulated lines is essential to ensure a reliable operation of the system. This can be determined by numerical simulations and requires a fast linear solver. Utilizing the AmgX library on NVIDIA graphic processing units allows a fast solution of a Poisson system of linear equations with millions of degrees of freedom within fractions of a second.
{"title":"GPU-Accelerated Field Simulation of HVAC Gas Insulated Lines","authors":"Hendrik Hensel, Marvin-Lucas Henkel, N. Haussmann, C. Jörgens, S. Stroka, M. Clemens","doi":"10.1109/CEFC-LONG57069.2022.10107547","DOIUrl":"https://doi.org/10.1109/CEFC-LONG57069.2022.10107547","url":null,"abstract":"High voltage alternating current gas insulated systems are utilized in narrow spaces. In operation electric field peaks can occur and may impair the functioning of the system. The knowledge of the electric field distribution in gas insulated lines is essential to ensure a reliable operation of the system. This can be determined by numerical simulations and requires a fast linear solver. Utilizing the AmgX library on NVIDIA graphic processing units allows a fast solution of a Poisson system of linear equations with millions of degrees of freedom within fractions of a second.","PeriodicalId":301005,"journal":{"name":"2022 IEEE 20th Biennial Conference on Electromagnetic Field Computation-Long papers (CEFC-LONG)","volume":"791 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117051504","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-01DOI: 10.1109/CEFC-LONG57069.2022.10107524
Ye Liu, Kang Ma, Zhuoran Zhang, Ziqiang Wei
In this paper, a high-speed permanent magnet synchronous motor (PMSM) is designed and manufactured for hydrogen recirculation pumps (HRPs). The motor rotor is exposed to a high-pressure mixture of steam and hydrogen, which makes hydrogen embrittlement occur in permanent magnets (PMs). A protective coating process is necessary for the PMs in high-pressure hydrogen. However, in the process of sleeve interference installation, the protective coating of the PMs is easily damaged. This paper aims at the high-strength rotor optimization of high-speed PMSM for HRPs in terms of electromagnetic and mechanical performance. A new sleeve structure is proposed to strengthen the motor rotor considering the manufacturing process. The electromagnetic performance is optimized by adjusting the ratio of PMs with different magnetization directions with the different sleeve structures. A 25000r/min, 800W prototype has been manufactured and the effectiveness of analysis methods has been validated by the experimental results.
{"title":"Comparative Analysis of Mechanical Characteristics of Different Rotor Structures of High-Speed Permanent Magnet Synchronous Motor for Hydrogen Recirculation Pumps","authors":"Ye Liu, Kang Ma, Zhuoran Zhang, Ziqiang Wei","doi":"10.1109/CEFC-LONG57069.2022.10107524","DOIUrl":"https://doi.org/10.1109/CEFC-LONG57069.2022.10107524","url":null,"abstract":"In this paper, a high-speed permanent magnet synchronous motor (PMSM) is designed and manufactured for hydrogen recirculation pumps (HRPs). The motor rotor is exposed to a high-pressure mixture of steam and hydrogen, which makes hydrogen embrittlement occur in permanent magnets (PMs). A protective coating process is necessary for the PMs in high-pressure hydrogen. However, in the process of sleeve interference installation, the protective coating of the PMs is easily damaged. This paper aims at the high-strength rotor optimization of high-speed PMSM for HRPs in terms of electromagnetic and mechanical performance. A new sleeve structure is proposed to strengthen the motor rotor considering the manufacturing process. The electromagnetic performance is optimized by adjusting the ratio of PMs with different magnetization directions with the different sleeve structures. A 25000r/min, 800W prototype has been manufactured and the effectiveness of analysis methods has been validated by the experimental results.","PeriodicalId":301005,"journal":{"name":"2022 IEEE 20th Biennial Conference on Electromagnetic Field Computation-Long papers (CEFC-LONG)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116000296","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-01DOI: 10.1109/CEFC-LONG57069.2022.10107519
Jaeho Ryu, I. Hahn
The eddy current is one of the unavoidable disadvantages of electrical machines, and the derivation of its calculation and results is not easily achieved. In particular, the eddy current by the magnetic field induces the second magnetic field, and it results in the induced eddy current according to the excitation frequency. In order to study the induced eddy current, this paper uses the Finite Difference Method (FDM) as the numerical method by applying the finite-difference stencil of the curl, div, and discrete Laplace operator. This method is a simple, accurate technique for solving partial differential equations. However, the FDM requires a square mesh, and this requirement has limitations for solving complex shapes, and it increases computation time by making all shapes into fine meshes for calculating the induced eddy current density. To solve this problem, this paper proposes the FDM with a non-uniform mesh to evaluate the eddy current generated at the edges of a structure requiring a fine mesh, which also contributes to the reduction of computing cost. In addition, to ensure the continuity between different size meshes in the non-uniform mesh, the multi-linear interpolation is applied. The numerical results from the non-uniform mesh are compared with a commercial FEM tool and analytical method to demonstrate the accuracy of the results.
{"title":"Numerical modeling for induced eddy current calculation according to non-uniform mesh","authors":"Jaeho Ryu, I. Hahn","doi":"10.1109/CEFC-LONG57069.2022.10107519","DOIUrl":"https://doi.org/10.1109/CEFC-LONG57069.2022.10107519","url":null,"abstract":"The eddy current is one of the unavoidable disadvantages of electrical machines, and the derivation of its calculation and results is not easily achieved. In particular, the eddy current by the magnetic field induces the second magnetic field, and it results in the induced eddy current according to the excitation frequency. In order to study the induced eddy current, this paper uses the Finite Difference Method (FDM) as the numerical method by applying the finite-difference stencil of the curl, div, and discrete Laplace operator. This method is a simple, accurate technique for solving partial differential equations. However, the FDM requires a square mesh, and this requirement has limitations for solving complex shapes, and it increases computation time by making all shapes into fine meshes for calculating the induced eddy current density. To solve this problem, this paper proposes the FDM with a non-uniform mesh to evaluate the eddy current generated at the edges of a structure requiring a fine mesh, which also contributes to the reduction of computing cost. In addition, to ensure the continuity between different size meshes in the non-uniform mesh, the multi-linear interpolation is applied. The numerical results from the non-uniform mesh are compared with a commercial FEM tool and analytical method to demonstrate the accuracy of the results.","PeriodicalId":301005,"journal":{"name":"2022 IEEE 20th Biennial Conference on Electromagnetic Field Computation-Long papers (CEFC-LONG)","volume":"65 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126485567","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-01DOI: 10.1109/CEFC-LONG57069.2022.10107520
Arbaaz Khan, Saikou Ceesay, Yuan-Po Teng, Ruoli Wang, Haipeng Yue, D. Lowther
Over the past couple of decades, dependence on magnetic materials has increased tremendously. Since every kind of magnetic material behaves differently, it is very difficult to create a universal model of these behaviors. While there exist models and simulation tools that can represent the behavior of these materials, it may take days or even an entire week to compute when embedded in an analysis system, which is highly inefficient. With a good machine learning (ML) model, the computation time can be reduced by a significant amount with minimal error. To determine performance parameters such as the efficiency of an electrical machine, the hysteretic behavior of the material is crucial, and the representation used can impact the performance of a simulation tool. The goal of this work is to explore several, different deep learning (DL) network architectures which might both provide a generalized representation methodology for hysteresis problems and reduce the computational effort needed to include the impact of hysteresis in a finite element-based simulation over existing approaches.
{"title":"Generalizable Deep Neural Network Based Multi-Material Hysteresis Modeling","authors":"Arbaaz Khan, Saikou Ceesay, Yuan-Po Teng, Ruoli Wang, Haipeng Yue, D. Lowther","doi":"10.1109/CEFC-LONG57069.2022.10107520","DOIUrl":"https://doi.org/10.1109/CEFC-LONG57069.2022.10107520","url":null,"abstract":"Over the past couple of decades, dependence on magnetic materials has increased tremendously. Since every kind of magnetic material behaves differently, it is very difficult to create a universal model of these behaviors. While there exist models and simulation tools that can represent the behavior of these materials, it may take days or even an entire week to compute when embedded in an analysis system, which is highly inefficient. With a good machine learning (ML) model, the computation time can be reduced by a significant amount with minimal error. To determine performance parameters such as the efficiency of an electrical machine, the hysteretic behavior of the material is crucial, and the representation used can impact the performance of a simulation tool. The goal of this work is to explore several, different deep learning (DL) network architectures which might both provide a generalized representation methodology for hysteresis problems and reduce the computational effort needed to include the impact of hysteresis in a finite element-based simulation over existing approaches.","PeriodicalId":301005,"journal":{"name":"2022 IEEE 20th Biennial Conference on Electromagnetic Field Computation-Long papers (CEFC-LONG)","volume":"103 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131303825","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-01DOI: 10.1109/CEFC-LONG57069.2022.10107526
Issah Ibrahim, D. Lowther
Acoustic noise and torque ripple are some of the undesirable performance indices in the design of electric motors. In fact, knowing the relationship between the two quantities can be very crucial, especially in the optimization process. For instance, where an increase in torque ripple is known to exacerbate the noise and vibrations, the two quantities could be considered as non-conflicting objectives. What this means is that minimizing one quantity could result in the optimization of the other without necessarily treating the two as mutually exclusive objectives. This paper, therefore, explores the correlation, if any exists, between torque ripple and the acoustic noise performance of the electric motor. The procedure involves exploring the design space of a 10 pole 12 slot surface-mounted permanent magnet synchronous motor to compute the torque ripple and the corresponding acoustic noise related to thousands of different motor configurations. Then, a statistical analysis is performed to establish the correlation strength between torque ripple and acoustic noise.
{"title":"A Study of the Relationship between Acoustic Noise and Torque Pulsation in Permanent Magnet Synchronous Motors","authors":"Issah Ibrahim, D. Lowther","doi":"10.1109/CEFC-LONG57069.2022.10107526","DOIUrl":"https://doi.org/10.1109/CEFC-LONG57069.2022.10107526","url":null,"abstract":"Acoustic noise and torque ripple are some of the undesirable performance indices in the design of electric motors. In fact, knowing the relationship between the two quantities can be very crucial, especially in the optimization process. For instance, where an increase in torque ripple is known to exacerbate the noise and vibrations, the two quantities could be considered as non-conflicting objectives. What this means is that minimizing one quantity could result in the optimization of the other without necessarily treating the two as mutually exclusive objectives. This paper, therefore, explores the correlation, if any exists, between torque ripple and the acoustic noise performance of the electric motor. The procedure involves exploring the design space of a 10 pole 12 slot surface-mounted permanent magnet synchronous motor to compute the torque ripple and the corresponding acoustic noise related to thousands of different motor configurations. Then, a statistical analysis is performed to establish the correlation strength between torque ripple and acoustic noise.","PeriodicalId":301005,"journal":{"name":"2022 IEEE 20th Biennial Conference on Electromagnetic Field Computation-Long papers (CEFC-LONG)","volume":"35 6","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132570618","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-01DOI: 10.1109/CEFC-LONG57069.2022.10107548
Samuel Kvasnicka, K. Roppert, C. Riener, T. Bauernfeind, M. Kaltenbacher
The generation of reduced order models describing the input-output behavior of linear quasi-stationary electromagnetic field problems is of particular interest in numerous applications, for example, in the field of wireless power transfer. Especially when rigorous optimization strategies shall be applied which require a large number of function calls. This contribution demonstrates the construction of a differential-algebraic equation system to describe a given single-input single-output behavior of a two and three dimensional eddy current model which is essential for a broad range of model order reduction techniques. The description of the problem in terms of a differential-algebraic equation system furthermore allows an ease coupling of electric circuits.
{"title":"Model order reduction techniques for linear differential-algebraic equations applied to semi-discretized eddy current model","authors":"Samuel Kvasnicka, K. Roppert, C. Riener, T. Bauernfeind, M. Kaltenbacher","doi":"10.1109/CEFC-LONG57069.2022.10107548","DOIUrl":"https://doi.org/10.1109/CEFC-LONG57069.2022.10107548","url":null,"abstract":"The generation of reduced order models describing the input-output behavior of linear quasi-stationary electromagnetic field problems is of particular interest in numerous applications, for example, in the field of wireless power transfer. Especially when rigorous optimization strategies shall be applied which require a large number of function calls. This contribution demonstrates the construction of a differential-algebraic equation system to describe a given single-input single-output behavior of a two and three dimensional eddy current model which is essential for a broad range of model order reduction techniques. The description of the problem in terms of a differential-algebraic equation system furthermore allows an ease coupling of electric circuits.","PeriodicalId":301005,"journal":{"name":"2022 IEEE 20th Biennial Conference on Electromagnetic Field Computation-Long papers (CEFC-LONG)","volume":"4 4","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120983677","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-01DOI: 10.1109/CEFC-LONG57069.2022.10107550
J. Smajic, M. Schneider, Raphael Baumeler, M. Schueller
Linear induction motors (LIMs) represent an elegant modern solution for future transportation, motion control, and linear automation systems. As opposed to their rotational counterparts, simulation based design and optimization of LIMs is still an open research subject due to their specific geometrical-and electromagnetic features. The time-domain FEM eddy current analysis is a widely used tool for analyzing LIMs capable of taking into account a dynamically moving rail and a coupling of the motor with its power supplying electric circuit. To replace the computationally intensive time-domain FEM analysis of LIMs the paper suggests a frequency-domain FEM analysis and a concept of an effective rail conductivity for taking into account the rail movement. The numerical results of the proposed method are compared against the classical time-domain approach and measurements and a high level of agreement has been found.
{"title":"Accelerated Frequency-domain Field Simulations of Linear Induction Motors","authors":"J. Smajic, M. Schneider, Raphael Baumeler, M. Schueller","doi":"10.1109/CEFC-LONG57069.2022.10107550","DOIUrl":"https://doi.org/10.1109/CEFC-LONG57069.2022.10107550","url":null,"abstract":"Linear induction motors (LIMs) represent an elegant modern solution for future transportation, motion control, and linear automation systems. As opposed to their rotational counterparts, simulation based design and optimization of LIMs is still an open research subject due to their specific geometrical-and electromagnetic features. The time-domain FEM eddy current analysis is a widely used tool for analyzing LIMs capable of taking into account a dynamically moving rail and a coupling of the motor with its power supplying electric circuit. To replace the computationally intensive time-domain FEM analysis of LIMs the paper suggests a frequency-domain FEM analysis and a concept of an effective rail conductivity for taking into account the rail movement. The numerical results of the proposed method are compared against the classical time-domain approach and measurements and a high level of agreement has been found.","PeriodicalId":301005,"journal":{"name":"2022 IEEE 20th Biennial Conference on Electromagnetic Field Computation-Long papers (CEFC-LONG)","volume":"90 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121156537","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-01DOI: 10.1109/CEFC-LONG57069.2022.10107522
A. Gschwentner, K. Reppert, M. Kaltenbacher
In this work, the influence of classical mortaring and Nitsche-type mortaring on the torque, calculated via Maxwell's stress tensor, is determined. First, the influence of the element distortion of quadrilateral and triangular elements on the torque using a conforming mesh is investigated. Subsequently, a convergence study is performed to gain insight into the influence of the Nitsche factor used in the non-conforming interface technique and to determine optimal parameters. Based on these results, the torque of a four pole squirrel cage induction motor is calculated for selected rotation speeds and compared with torque values calculated with a harmonic approach using a conforming mesh.
{"title":"Computation of Torque using different non-conforming Interface Techniques","authors":"A. Gschwentner, K. Reppert, M. Kaltenbacher","doi":"10.1109/CEFC-LONG57069.2022.10107522","DOIUrl":"https://doi.org/10.1109/CEFC-LONG57069.2022.10107522","url":null,"abstract":"In this work, the influence of classical mortaring and Nitsche-type mortaring on the torque, calculated via Maxwell's stress tensor, is determined. First, the influence of the element distortion of quadrilateral and triangular elements on the torque using a conforming mesh is investigated. Subsequently, a convergence study is performed to gain insight into the influence of the Nitsche factor used in the non-conforming interface technique and to determine optimal parameters. Based on these results, the torque of a four pole squirrel cage induction motor is calculated for selected rotation speeds and compared with torque values calculated with a harmonic approach using a conforming mesh.","PeriodicalId":301005,"journal":{"name":"2022 IEEE 20th Biennial Conference on Electromagnetic Field Computation-Long papers (CEFC-LONG)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133098851","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-01DOI: 10.1109/CEFC-LONG57069.2022.10107549
Yapeng Jiang, Yunjun Guo, Qingyu Yang, Junquan Chen, Kai Bo, Dong Wang
For high-speed electrical machines, the AC copper loss caused by high electrical frequency is very considerable. This paper presents a computationally efficient method for evaluating the circulating current between parallel strands in electrical machines with stranded windings. Firstly, the lumped parameter circuit model is introduced, and then the key parameters, namely the mutual slot-leakage inductance matrix, are calculated by the proposed non-strand finite element model. Compared with the traditional parameter extraction method, the time consumption of the proposed model is greatly reduced. Finally, the accuracy of the method is validated by FE simulations.
{"title":"A Computationally Efficient Method for Modelling of Circulating Current Loss in Electrical Machines with Stranded Windings","authors":"Yapeng Jiang, Yunjun Guo, Qingyu Yang, Junquan Chen, Kai Bo, Dong Wang","doi":"10.1109/CEFC-LONG57069.2022.10107549","DOIUrl":"https://doi.org/10.1109/CEFC-LONG57069.2022.10107549","url":null,"abstract":"For high-speed electrical machines, the AC copper loss caused by high electrical frequency is very considerable. This paper presents a computationally efficient method for evaluating the circulating current between parallel strands in electrical machines with stranded windings. Firstly, the lumped parameter circuit model is introduced, and then the key parameters, namely the mutual slot-leakage inductance matrix, are calculated by the proposed non-strand finite element model. Compared with the traditional parameter extraction method, the time consumption of the proposed model is greatly reduced. Finally, the accuracy of the method is validated by FE simulations.","PeriodicalId":301005,"journal":{"name":"2022 IEEE 20th Biennial Conference on Electromagnetic Field Computation-Long papers (CEFC-LONG)","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124686331","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：