Demands for higher power density in electrical machine are increasing. In the past, magnetic circuit design and thermal design were sometimes performed separately, but as demands become stricter, the importance of multifaceted evaluation that combines magnetic field analysis and thermal analysis is increasing as a means of obtaining a design proposal that simultaneously meets the requirements for magnetic circuits and heat.
The temperature of motors and other components is affected by the contact thermal resistance of each part. To accurately calculate the motor temperature at the magnetic circuit design stage, highly accurate and easy-to-apply contact thermal resistance is required. For this reason, we have developed a database of contact thermal resistance based on actual measurements.
This time, we have expanded the coverage range of the contact state between the coil and core, and measured the contact thermal resistance between the laminated core and the frame. We would appreciate your feedback on the coverage range and other future developments.

To obtain design proposals that satisfy increasingly sophisticated requirements, it is necessary to shift from the design process that has traditionally relied on the experience and expertise of experienced designers to a data-driven approach that explores vast design spaces at high speed. JMAG continues to develop functions to realize data-driven design.
In addition to improving the performance of essential optimization functions, we have also incorporated surrogate models to improve calculation efficiency and increased the speed of large-scale calculations to enable highly accurate narrowing down of options.
This poster will explain the developments that have been carried out over the past year toward data-driven design and future plans.

JMAG is developing to realize automated design of electric machines.
JMAG’s automated design starts with wide-area design exploration, and proceeds with design exploration including multi-disciplinary evaluation while ensuring consistency with the system, and virtual prototyping using analysis data.
Optimization calculations quickly explore the appropriate range. Enhanced size range exploration function improves the success rate of shape generation in multivariable size optimization. Design variables can also be treated as discrete values. In addition, stable and fast optimization can be performed using a surrogate model.
Not only magnetic design but also thermal design functions are made easier to use. Thermal resistance is defined in a script to improve the expression ability of thermal circuits.
Detailed models equivalent to the actual equipment are calculated quickly. Large-scale models using FEM conductors are also calculated quickly with parallel direct method.

JMAG-Express can perform multifaceted evaluations including magnetic, thermal, strength, and insulation.
Although it is possible to evaluate motor performance by selecting from a library of pre-registered geometries, we have received many requests to use geometries that customers are considering.
In order to handle user-defined shapes in JMAG-Express, it is necessary to tie magnets and coils together.
Furthermore, in order to change the shape according to the intent, constraints must be given to the shape to make it fully constrained.
JMAG-Express has a function that guides the user through the necessary steps for registration in an easy-to-understand manner.
In addition, we are developing a function to automatically perform complete restraint setting, which is said to be complicated and difficult.
This poster explains the process of importing CAD data into JMAG-Express and performing optimization calculations.

Motor type, size, and slot combination are considered in the early stages of motor design.
To meet advanced requirements, the design space must be large and explored at high speed.
In recent years, the requirement to drive motors over a wide range has increased, and in these cases, efficiency maps are used to evaluate the entire operating range.
In addition, since temperature, strength, and dielectric strength are often in conflict with efficiency, power, and torque, they must be evaluated simultaneously.
Thus, the design space that must be evaluated in the early stages of design is vast, but the man-hours required to prepare input data are large, and the number of evaluation items as well as design variables is enormous.
JMAG-Express has geometry and evaluation contents as templates, which greatly reduces the amount of time required for preparation.
The evaluation includes not only magnetic properties, but also temperature, stress, and dielectric strength, and simultaneous and multifaceted evaluation reduces rework.
By using simulation to obtain multi-case results, motor selection can be performed at high speed.
In this poster, we will show how designs that satisfy motor performance requirements, temperature, strength, and dielectric strength can be explored from a vast design space.

As electrical machine becomes smaller and more powerful, thermal prediction at the early design stages is becoming increasingly important.
In the thermal design of electrical machine, an accurate understanding of losses that are the source of heat is essential, and is closely related to magnetic circuit design.
Meanwhile, in magnetic circuit design, accurate temperature distribution predictions are necessary to consider thermal demagnetization of magnets in PM motors and cooling methods.
In other words, it is important to simultaneously evaluate magnetic circuits and heat from multiple angles in electrical machine design, but thermal evaluation is not necessarily the specialty of magnetic circuit designers, and they can even be unsure of how to perform an analysis in the first place.
Therefore, in this seminar, we will introduce a thermal analysis process starting with JMAG-Express.
By using JMAG-Express, even beginners to thermal evaluation can intuitively build thermal equivalent circuits and cooling models and perform thermal analysis without hesitation.

In the initial stages of motor design, multi-faceted evaluations such as magnetic, thermal, structural, and insulation assessments are crucial for reducing design rework and controlling development costs.
Among these evaluations, insulation assessment is particularly important as it contributes to the long-term performance and safety of the motor.
However, conducting insulation evaluations requires separate modeling of the wire, which can be time-consuming.
With JMAG-Express, wire modeling can be easily accomplished by simply inputting specifications into the wire template.
Additionally, since models for magnetic, thermal, structural, and insulation evaluations are integrated, any changes to a single parameter are reflected across all evaluations.
This integration facilitates easy multi-faceted evaluations and optimization.
This poster introduces how JMAG-Express allows for easy insulation evaluation by simply inputting motor specifications.
It also introduces optimization in multi-faceted evaluations, considering insulation in addition to existing magnetic, thermal, and structural assessments.

In motor-driven systems like EV powertrains, the demand for higher efficiency and output power is essential, as is the need to reduce noise and vibration.
The electromagnetic force of the embedded motor is widely acknowledged as a major source of noise and vibration.
However, the magnetic design and NVH design are often performed independently, so the magnetic design may require revisions if resonance from electromagnetic forces in the machine fails to meet NVH specifications.”,
When the motor design is revised, the NVH needs to be reevaluated in a costly iterative process.
Moreover, system-wide vibration analysis demands significant computational resources, creating a barrier to its integration into the new design process.
Therefore, to solve these problems practically, it is important to be able to perform “rapid” magnetic design with “NVH integrated from the early design stage”.
In this seminar, we will discuss the vibration generation mechanism associated with motor electromagnetic forces and how to express it in FEA, We will also propose a high-speed system-level NVH design flow for electrical machine designers, with the introduction of transfer functions.

When using a JMAG-RT plant model in control design, a large number of circuit simulations must be run.
Therefore, reducing the circuit simulation time is of great significance in shortening the development period.
Since the processing time of the plant model accounts for a non-negligible proportion of the total circuit simulation time, we set out to speed up the processing time in the plant model. In this poster, the current status and future plans will be introduced.
In addition, the time required to generate plant models has been increasing due to the use of 3D models for higher accuracy and the inceased degrees of freedom of currents in open-end windings, compared to conventional star-connected windings.
The use of HPC as an effective method to reduce generation time and its effects will also be presented.

JMAG-RT is a useful plant model that can be used in a wide range of applications, from system design to ECU verification with HILS.
Due to the wide variety of uses of JMAG-RT, many people may not be able to make full use of it.
This poster aims to provide a deeper understanding of how JMAG-RT can be used by including introductions to characteristic examples of use in addition to explanations of JMAG-RT functions.
For example, the JMAG-RT Viewer, which evaluates the characteristics of RTT files as a map, will be explained as a functional explanation, and a case study of a dual inverter system will be discussed as an application example.
We hope you will consider how JMAG-RT can contribute to your design work.

In motor design, it is generally known that it is important to understand physical quantities such as losses and excitation forces that depend on harmonic components of the current at an early stage.
However, the reality is that the use of current waveforms that do not include harmonic components is still the norm, especially in the early stages of electromagnetic design.
In response to this situation, JMAG is equipped with a function called “control macro elements,” which allows the creation of motor control circuits in JMAG alone by simply entering design variables.
This poster proposes a new design method that takes into account the harmonic component of the current due to control from the initial stage of electromagnetics design, and presents the merits of incorporating this method with specific examples.
If you are looking for a better initial design proposal in electromagnetic design, or for a better initial electromagnetic design mechanism, please stop by.

In the development of motors with increasingly high power density, it has become necessaryto include not only magnetic design but also thermal design from the initial stage.
The JMAG-RT model used in control simulations can handle magnet and coil temperature increases due to copper and iron losses, and allows control studies to take into account changes in characteristics due to these increases.
In addition, the JMAG-RT loss map model can output various losses according to the motor drive state, making it possible to evaluate temperatures when combined with the cooling system in a vehicle system driving simulation.
In this poster, examples of the application of these JMAG-RT models to thermal management systems will be presented.

Multipoint search, such as multi-objective genetic algorithms, is used as a method to obtain design alternatives that satisfy increasingly sophisticated requirements.
In addition, as equipment performance requirements become more stringent, models that are faithful to the actual equipment, such as 3D models, are required.
However, optimization calculations using 3D models take an enormous amount of time, so in order to obtain a solution within practical time it is necessary to accelerate the optimization calculation.
Recently, surrogate models have been attracting attention because they can be evaluated faster than FEA.
JMAG proposes the use of surrogate models in optimization calculations to speed up the process.
In this poster, an example of optimization of an axial gap motor will be presented.
The flow of setting up a surrogate model is also introduced for induction heating and other rotating machines examples.

Why not utilise JMAG optimization for electrical equipment designs that meet high performance requirements?
JMAG provides a parameter optimization function that can handle a large number of design variables.
For example, if the dimensions of the equipment geometry are a design variable, RangeFinder can automatically avoid geometry breakdowns due to interference from multiple dimensional parameters.
In addition, multi-variable optimization requires a huge number of calculations to be performed. The challenges of computation time and storage consumption are also solved by high-performance distributed execution and the ability to delete files during the optimization calculation.
This poster introduces JMAG solutions for parameter optimization.

Automated design by data-driven approach, such as optimization calculations, is attracting attention in the design of electric machines.
Topology optimization, which has a high degree of freedom in shape, is being used in conceptual design.
On the other hand, topology optimization often results in shapes that are difficult to manufacture due to the high degree of freedom in shape.
In such cases, by not subjecting the entire design domain to topology optimization but a partial area of it to parametric optimization, it is possible to maintain a manufacturable shape.
In addition, the convergence of optimization calculations tends to depend on the quality of initial cases at the first generation.
In JMAG V23.2, the specified individuals can be set to initial cases for topology + parametric optimization and it can improve the convergence of optimization.
In this poster, we will introduce the topology + parametric optimization functions and discuss requests for these functions.

The optimization techniques are being employed for motor designing to get higher torque and efficiency than ever.
Especially, the dimensional parameter optimization using the continuous optimization method has been widely tackled.
However, the continuous optimization cannot consider some constraints in actual designing. For example, manufacturing tolerance and the standard wire gauge.
The number of turns of the coil is also an essential design variable but it is not suitable to the continuous optimization.
We introduce the discrete optimization feature in JMAG V24, and show that the constraints can be satisfied.
The results of not only a bulk coil model but also a wire model are shown.
We continue to develop the discrete optimization feature and welcome your requests.

In multi-case calculations such as optimization calculations, it is necessary to increase the number of
simultaneous executions of jobs to shorten processing times.
In recent years, the number of many-core workstations has been increasing.
It is becoming possible to execute dozens of jobs simultaneously on a PC, even without a cluster,
which is expensive to purchase and manage.
JMAG has previously improved performance for Linux clusters and cloud environments.
These technologies are also useful for Windows PCs.
In this poster, we will show examples of optimization using simultaneous execution on a Windows PC.
We will also introduce performance improvements that we have begun to prepare for PCs.

The efficiency map is a critical tool in the design and performance assessment of electrical machines.
However, the non-negligible impact of thermal dependency highlights the need for more advanced modeling techniques.
We will explore JMAG’s vision for thermal-dependent machine design, transitioning from the current thermal modeling capabilities to the future tools needed for the accurate assessment of electrical machine performance.
Emphasizing the plans of JMAG to integrate a ‘true’ continuous-use NT curve accounting for thermal limits as well as a thermal 2-way drive cycle performance evaluation.

In recent years, the JMAG usage environment has become complicated, such as centralized management of licenses and use at multiple locations.
With the license usage log analysis function added in JMAG V24, it is possible to check and analyze which functions are being used, when, by whom, and to what extent.
In this document, we will introduce usage log output methods and analysis examples to lead to appropriate license operation.

Data-driven approaches that use large-scale data for accurate and quick decision making are beginning to be applied in a variety of fields.
In the field of electrical machine design, JMAG-Designer’s distributed calculation capabilities and high-speed solvers make it possible to collect large amounts of data.
Furthermore, JMAG-Designer has been equipped with machine learning modeling functions since Ver. 23.2 to support not only the collection of data, but also its effective use.
This poster explains the features of JMAG’s machine learning model creation function.
Specifically, it will be shown that the function for removing anomalies in training data and the neural network creation function (to be introduced in Ver. 24.0) are useful for creating high-quality models based on large-scale data.
Examples of motor sizing script construction using this function will be included.
Future improvements needed to make machine learning models even more reliable will also be discussed.

To realize virtual prototyping, the analysis requires accuracy equivalent to that of the actual machine.
In that case, a 3D model is necessary, and the number of elements will be more than one million. Then it will need long calculation times.
This makes it essential to speed up the process using a parallel solver.
For example, by using a parallel solver, it is now possible to perform the analysis of an axial flux motor with one million elements in less than one minute, which is comparable to a 2D model.
This poster will show the performance of the speedup achieved by a parallel solver.
And also introduce the latest trends, such as the combination of a direct method and an MPP solver, a parallel solver using a GPU, and speedup using a parallel solver for structural analysis.

When the power supply frequency is high or harmonics are superimposed, eddy currents flow in the windings and AC copper loss cannot be ignored.
Calculating AC copper loss requires modeling of the eddy currents in the windings, which requires large-scale calculations.
JMAG provides the FEM conductor as a function for calculating eddy currents in windings.
However, FEM conductors with long winding lengths can degrade the convergence of ICCG, which can lead to increased calculation times.
An effective solution to this problem is to use the direct solver instead of ICCG.
The direct solver does not require convergence calculations, so it can calculate quickly for problems with poor convergence.
Furthermore, by using the MPP direct solver in V24.0, it is possible to calculate large-scale winding models in a short time.
This poster introduces parallel calculation using the direct solver for a large-scale AC copper loss analysis model.

JMAG’s calculation speed continues to improve through three new versions each year, so you can expect to see shorter calculation times by simply using the latest version.
For example, in optimization, which extracts the optimal design from a wide design space, calculations are performed for many cases.
To shorten this calculation time, the latest JMAG includes technologies that make efficient use of many-core and cluster cloud computing, which have become popular in recent years.
Improvements to the program have also made it possible to speed up calculation processing per case.
This poster will introduce the improvements in distributed processing efficiency and speed-up of calculation processing over the past two years.

Achieving reliable design proposals that meet demanding performance requirements and minimize rework requires a sufficiently broad design space and high computational accuracy.
This implies an increase in both the number of design cases and the computational time per case.
With the MJR feature (job system) that supports MPP calculations since JMAG-Designer Ver.23.2, distributed multi-case runs and highly parallel MPP calculations can be combined easily to achieve high performance design proposals within severe time constraints. In this poster, the functions
and benefit of the MJR are introduced through application examples which include a massive calculation using more than 10,000 cores.

Magnets are used in many electrical devices and have a significant effect on the characteristics of those devices, so magnet design is essential to further improve the efficiency of electrical devices.
For example, when magnetizing multiple times by changing the current value during built-in magnetization, the magnetization distribution of the magnet can be evaluated with higher accuracy by taking magnetic hysteresis into account and modeling the demagnetization state during re-magnetization.
In addition, if the surface magnetic flux density of the magnet after magnetization does not match the actual measurement, it is possible to obtain magnetization characteristics with small deviation from the actual measurement by identifying the magnetization direction and coercive force through optimization.
This poster will focus on magnetization and will explain multi-stage magnetization modeling that takes magnetic hysteresis into account, as well as optimization of magnetization characteristics, with examples included.

In the development and design of high-efficiency motors, high accuracy is required to enable verification and performance evaluation through simulation with the aim of virtual prototyping.
For example, an efficiency error of 1% cannot be ignored in areas where the efficiency exceeds 90%. To keep the efficiency error within 1% for a motor with an efficiency of 90%, the loss error must be kept within approximately 11%.
The factors that affect the accuracy of loss calculations are known to be building factors such as time harmonics, AC copper loss, stress, and processing strain.
Taking these into account in the analysis is expected to improve the accuracy of losses, but the cost of the analysis will be high, so we believe that being able to predict which loss factors should be calculated with emphasis before designing (and conversely, how much difference will occur if they are not taken into account) will serve as a guideline for analysis.
In this poster, we will show with examples what motors and operating conditions have the greatest impact on the above factors that affect losses.

Demands for higher power density in electrical equipment are increasing. Conventionally, magnetic circuit design and thermal design were performed separately, but as the demands become stricter, the importance of multifaceted evaluation that couples magnetic field analysis and thermal analysis is increasing as a means of obtaining a design proposal that simultaneously meets the requirements for magnetic circuit and heat.
To accurately predict the temperature, it is necessary to estimate the heat transfer of cooling (air cooling, water cooling, oil cooling, etc.). The heat transfer coefficient can be obtained by thermal fluid analysis (CFD), but the calculations are generally large-scale, and a highly accurate and easy-to-apply cooling model is required to perform it simultaneously with the magnetic circuit design.
For this reason, we have developed cooling models based on the results of CFD in advance and empirical rules. However, new cooling methods are being adopted in the development of motors with increasingly high power density, and we believe that previous cooling models are insufficient.
In this poster, we review currently available cooling models and present a plan for future development. We would appreciate your feedback on the coverage, priority, etc.

In the development of induction heating coils, the optimum coil design is required depending on the parts to be heated.
It is also important to consider the cooling process to suppress deformation of the parts being heated.
In order to efficiently develop such heating coils, simulation is used for design.
This poster presents an example of design exploration that takes into account the deformation of the heating coil in order to maintain its durability.
We will also introduce an example of induction heating calculation that takes into account the cooling process and how to set it up in JMAG.
We will also show the performance of distributed and parallel computing.

In order to calculate AC losses and to take into account magnetic flux from coil ends, it is necessary to accurately model the coil end geometry and winding arrangement.
However, creating the coil end geometry is a time-consuming task as it can be difficult to draw due to the high density and parts can interfere with each other.
In these situations, we recommend using coil template function in JMAG.
Just by specifying parameters, you can create the winding geometry of a rotating machine while avoiding these problems.
It supports round and square wire as well as coil shapes of any cross section, and it is possible to add coil ends not only to geometry created in JMAG but also to geometry from CAD data.
In addition, since it supports parametric analysis and optimization calculations, it is also suitable for considering multiple cases.
In this poster, we will introduce examples of coil end creation for radial gap type motors and axial gap type motors, as well as examples of parametric analysis.
If you feel the need to consider coil ends, please stop by.

In the finite element method, the analysis accuracy and analysis time vary greatly depending on the mesh used in the calculation.
Therefore, it becomes necessary to reduce the number of mesh elements by providing appropriate density.
In radial gap motors, it is possible to reduce the number of elements by dividing the motor finely in the circumferential direction and roughly in the axial direction, and by using a symmetrical mesh for cogging torque calculations.
Not only for motors but also for induction heating calculations, it is possible to reduce the number of elements by changing the density in each direction depending on the model shape.
JMAG has an automatic mesh generation function that can create meshes like the one above.
Here we take radial gap motors and induction heating as examples.
We will explain how to create a sufficient mesh to obtain motor torque accuracy, and how to reduce the number of meshes while maintaining loss accuracy in induction heating problems which tend to become large-scale.

Electromagnetic force calculations are essential when considering vibration, noise, deformation, and displacement in electrical equipment design.
However, have you ever been troubled by not knowing which electromagnetic force calculation method to choose and what the calculation range should be?
Also, have you ever felt that it is difficult to make accurate calculations because the calculations do not match actual measurements?
This poster will introduce JMAG’s functions and modeling guidelines for improving calculation accuracy, along with examples of motors and magnetic shields.

Placing the resolver close to the motor saves space for the entire system, but makes it more susceptible to flux leakage from the motor.
If magnetic flux leakage enters the resolver, the output signal may be distorted, reducing the accuracy of angle detection.
Although the installation of magnetic shields can be expected to reduce the effects of flux leakage, the cost and weight of the magnetic shields themselves are disadvantageous.
This poster presents a case study of optimization calculations of magnetic shield geometry to suppress output signal distortion and reduce volume.

Synchronous reluctance motors excel in environmental resistance and maintainability because they do not have windings or permanent magnets.
Additionally, these motors are not limited to rotary machines but are also practical for linear motion applications, such as in railway systems.
The main evaluation criteria include thrust required for speed and positioning accuracy, as well as the reduction of cogging force, which is also an important evaluation item.
However, the characteristics of synchronous motors vary greatly depending on the width and number of iron parts and magnetic barriers, requiring fine adjustments.
JMAG allows for practical time optimization calculations through distributed parallel execution to explore optimal shapes.
This poster introduces a case study of exploring rotor shapes that maximize thrust and minimize cogging force using dimensional optimizatio

Axial flux motors are capable of producing high torque even with a thin shape, and are attracting attention as motors for use in vehicles where space is limited.
JMAG-Express is ideal for easily exploring a wide design space, and calculations for axial flux motors can also be performed with JMAG-Express.
As an example, this poster will introduce the results of evaluating an efficiency map using JMAG-Express.
In order to conduct wider design exploration by changing the number of poles and slots, additional functions is required.
Therefore, in v24.0 we have added a mesh generation function, making calculations easier.
We plan to continue adding new functions to enable a wider range of design exploration, and also have plans to support JMAG-Express scenarios.
We will show you what types of shapes and patterns we are considering supporting, and hope to be able to discuss on the day whether they are sufficient.

Magnetic resonance type wireless power transfer devices are used for long-distance power transfer, and are being considered as a method of charging electrical devices such as electric vehicles.
The receiving coil installed in the vehicle must be lightweight from the perspective of vehicle weight, and must have high power transmission efficiency even when the receiving and transmitting coils are misaligned.
Reducing the weight of the receiving side reduces power transmission efficiency, resulting in a mutually contradictory design. Multi-objective optimization is effective for problems like this.
Furthermore, because misalignment must be taken into account, three-dimensional consideration is required.
Even for this type of three-dimensional optimization, JMAG can obtain optimal solutions in a short amount of time through distributed execution combined with parallel computing.
This poster will introduce the optimization of wireless power transfer devices that take misalignment into account.

Reactors and transformers for converters are required to have improved electrical performance, such as DC superposition characteristics and low loss. However, problems such as noise due to electromagnetic excitation force and rise in component temperature due to losses in the windings and core can occur.
In conventional design flows, evaluations of magnetics, vibration, and heat are often performed sequentially, and even if a design proposal that satisfies electrical requirements is obtained, if the vibration and heat requirements are not met, the design must be redone from the magnetic design.
Simultaneous and multifaceted evaluation of magnetics, vibration, and heat reduces reversals in the design and development process, leading to improved performance and quality.
This poster uses the example of a reactor modeled on a power supply circuit to introduce an example of parameter optimization through multi-disciplinary evaluation of magnetics, structure, and heat.

In recent years, motors have become faster and higher frequency due to inverters.
In addition, as the frequency increases, electromagnetic interference (EMI) caused by power converters in control circuits has become more serious.
Electronic devices such as EMI filters are installed to mitigate this.
The frequency characteristics of these high-frequency motors and electronic devices are often measured, and there is a need to clarify them through analysis.
In analysis, a mesh is required to model the wires and capture the skin effect. As a result, a 3D model with more than one million elements is required.
And many cases calculations are required to investigate the frequency characteristics, which increases the calculation time.
By utilizing the parallel calculation and distributed simultaneous execution of JMAG, frequency characteristics can be evaluated in a short time.
This poster will introduce analysis examples of motors and electronic devices with their analysis time.

JMAG’s efficiency map function for conceptual design (early design stage) works quickly in few seconds, making it suitable for design exploration.
This poster explains the optimization of PMSM/EESM (permanent magnet / externally excited synchronous machine) incorporating the efficiency map into the objective function.
We will use an EESM example to confirm that different design proposals can be obtained depending on the efficiency map requirements,
and demonstrate that efficiency map optimization is useful for narrowing down design proposals in conceptual design.
We will also introduce the performance improvements that have been made in the latest version when using HPC (cluster computers).
Whether you are already working on efficiency map optimization or not,
please come and see for yourself whether JMAG’s efficiency map optimization function can help you create the design you really want.

Induction motors are cost-effective and easy to maintain because they do not use permanent magnets, and are widely used as drive motors.
In the case of automotive applications, the driving range is wide, so efficiency maps are useful for evaluating characteristics.
To accurately calculate the efficiency map of an induction motor, calculations must be performed using a model that captures the equipment characteristics and harmonic components.
With JMAG, even highly detailed efficiency maps can be calculated in a reasonable amount of time using parallel processing.
This poster will introduce the different efficiency map calculation methods to use depending on the design stage of an induction motor, and examples of efficiency map optimization.

In motor design, it is necessary to consider not only magnetic property evaluation but also stress evaluation.
Designs that anticipate a large safety factor in terms of stress will limit the design space, so there is a possibility that only design proposals that do not satisfy the requirements will be obtained.
On the other hand, it is not easy to obtain a design proposal that satisfies conflicting magnetic and stress properties.
In such considerations, multi-objective optimization that considers magnetic and stress is effective.
In this poster, we will explain how to handle nonlinear contact modeling between the rotor magnet and core and nonlinear material modeling at high speed rotation through the example of optimizing an IPM motor.
We would like to discuss evaluation methods for achieving both magnetic properties and strength properties at a high level in motor design.

With the increasing efficiency of electromagnetic devices, high accuracy is also required for iron loss analysis.
The iron loss calculation that combines the play model and 1D method is a promising method because it can calculate iron loss for arbitrary excitation waveforms, and the cost of measuring physical property data is the same as that of conventional iron loss analysis methods.
On the other hand, the combination of the play model and the 1D method does not take into account anomalous eddy current loss.
Therefore, a modeling of abnormal eddy current loss is necessary for practical use.
In JMAG, we have been working to develop an anomalous eddy current loss model that can be applied to arbitrary excitation waveforms without frequency dependence.
This poster presents an anomalous eddy current loss model that focuses on the magnetic susceptibility, which has shown relatively good performance.

By using JMAG, it is possible to perform detailed analysis of superconducting phenomena.
Specifically, in superconducting analysis, it is possible to calculate the current density, temperature, and stress of the wire, considering the important nonlinear resistance characteristics.
These characteristics are realized through the modeling of superconducting resistance using the IV model, cooling analysis considering thermal transfer boundaries, and stress calculation through structural analysis. This poster introduces the functions of superconducting analysis using the example of a solenoid coil for nuclear fusion.

Traditionally, various methods have been used to evaluate the characteristics of induction machines, one of which is the T-type equivalent circuit.
This poster will explain the T-type equivalent circuit, which has constant equipment constants such as leakage inductance and excitation inductance, in relation to FEA models.
However, constant equipment constants have the problem that slip dependence and current dependence cannot be taken into account.
This poster will demonstrate how the power of expression can be improved by taking into account slip dependence, current dependence, and spatial harmonics.
In addition, we will introduce a case study where actual measurements were compared to verify the validity of the equipment constants of the T-type equivalent circuit obtained by analysis.

In order to effectively utilize electromagnetic field analysis tools in design, it is necessary to ensure the reliability of the analysis.
As a means of ensuring reliability, there are two methods: checking the accuracy of the analysis settings (verification) and checking the validity of the analysis using actual measurements as a reference (validation).
During this verification process, we confirmed the torque sensitivity and accuracy of the analysis setting parameters. As a result, we found that each compensation value in the torque control model is highly sensitive to torque.
Additionally, during the validation process, we discovered that the delay in the control system of the actual machine and the delay in the encoder affected the target torque value.
This poster shows these contents.

Perform a comparative analysis between Finite Element Analysis (FEA) predictions and experimental measurements for a Hybrid Excited Flux Switching Machine (HEFSM), an electric motor that utilizes permanent magnets and electromagnetic excitation to generate variable magnetic flux, optimizing efficiency and performance.
This analysis will focus on identifying accuracy limitations in FEA predictions by considering various uncertainty sources, including the physical state of the machine and potential measurement inaccuracies. By examining both model-based and measurement-related uncertainties, this assessment aims to delineate the reliability boundaries of FEA results and identify factors contributing to discrepancies between predicted and observed results. Our findings highlight that achieving a 1% accuracy target, even with 3D simulations, is challenging and demands extensive understanding of the machine’s state along with exacting measurement precision.

HPC is increasingly being used to obtain design alternatives that meet high performance requirements and complex constraints in a short time.
In an HPC environment, including cloud computing, multiple computers are running.
As a result, the probability of encountering problems caused by the environment, such as analysis errors, increases compared to a single PC.
In addition, identifying the cause of a problem requires the construction of a similar environment, which takes time to resolve.
We are investigating a system that can simulate the environment of an HPC environment in software for the purpose of early problem solving and proactive risk detection.
This poster will present the details of this initiative using examples of environment-dependent problems that have been encountered.
We invite you to send us issues you have encountered in order to improve the accuracy of the simulation of this system.