Tackling Next Generation Motor Development using JMAG

The evolution of technology for motors and control systems has been tremendous. The progress of development is driven around an environmental axis which has provided motors that can cut the world’s power consumption in half and brought us electric vehicles. Electromagnetic field analysis software is becoming a tool indispensable in designing and evaluating the performance of these types of products. Prof. Akatsu uses the electromagnetic field analysis software, “JMAG,” for his M&E Energy Conservation Lab. at Shibaura Institute of Technology which is known for its achievements reducing torque ripple in permanent magnet motors and he is an executive member of several technical committees in The Institute of Electrical Engineers of Japan. In this interview, Prof. Akatsu discusses motor development and electromagnetic field analysis software.

Prof. Akatsu
Motors are thought of as a very mature technology, but their technological innovation continues today. Currently, motor and control systems continue to advance, especially to address environmental issues. Although this statistic is not very well known, 57% of the power in Japan is consumed by electric motors. This includes all of the motors built into compressors, pumps, and various other equipment. Motors that have power usage efficiency of less than 30%, or “power wasting motors” make up more than a hand full of the motors that are used. Motors controlled by inverters are said to have a usage efficiency of roughly 16% to 17%.
It is estimated that the world’s C02 emissions could be reduced by 7% if motors just in Japan were replaced by the most efficient products in the market today.
This means that highly efficient, small, low cost, and high performance motor systems need to be adopted around the world when considering the environmental burden.
Interest in electric vehicles is growing and the technological progress of motor systems, which is at the heart of these vehicles, is advancing rapidly. For example, motors for automobiles require high torque at low speeds, such as when starting a car on a hill. However, motors only capable of producing large torque would be too powerful and pointless under normal driving conditions. Additionally, because these motors need to be installed inside the vehicle, small yet powerful motors are desired.
These challenges drive a new type of research in motor electronics that needs to unify the development of the materials and structure of motors with the development of the control.

Prof. Akatsu
JMAG is used as a design tool as well as a tool to examine our theories. As an example, we investigate the effects of nonlinear characteristics using simulation after creating a linear design. We also examine whether the theories we are considering for a new motor are actually valid, which relates to the numerical models we create JMAG is also used to create control models by deriving the control parameters of the motor.
Linear models are no longer used in the design of control systems. Therefore, results from magnetic field and circuit simulations using JMAG-RT are implemented in real-time control simulations.
JMAG is a vital tool that is taught to the students working in my lab. The students address specific production challenges of motors, which deepens their understanding as they attempt the inductance calculations*1, iron loss calculations*2, and output curve calculations*3 used in the manufacturing process.

*1 Inductance = A property caused by variations of current becoming induced electromotive force in coils, etc. Inductance is also referred to as induction coefficient.
*2 Iron loss calculation = The amount of electric energy lost when current is applied to the core of a motor wound with coil and magnetized. The efficiency of motors worsens as iron loss increases.
*3 Output curve calculation = Variations of output (power) for motors.

Evaluating High Torque & Reducing Torque Ripple with JMAG

Prof. Akatsu
When aiming to gain higher torque which provides more powerful motors, the trade-offs related to the torque ripple*4 always need to be considered. To solve this problem, development that is “collaborative between the motor design and control” is required to increase the average torque in the motor design and limit the torque ripple in the control. We use JMAG in this type of development.
More specifically, we measured the torque ripple instantly by deriving numerical models encompassing the torque ripple. Then, we are able to reduce the torque ripple by applying a current using a current command to eliminate the torque ripple that we obtained.
Formulating the equation for the torque ripple was extremely difficult, but we were able to create a model matching the calculation results in JMAG and the actual results that we measured. Analyses allowed us to further our research because we know that the torque ripple results obtained in JMAG match those measured in the prototype.

Prof. Akatsu
JMAG is suited to generate these types of numerical models. Knowing the results in JMAG match the actual phenomena allows us more freedom in designs at the design stage.
In the past, the person making the motor and the person controlling the motor were different. There were many people developing control methods that believed “the motor works just like the numerical model.” However, in the real world, this is not true.
Collaboration between the people making the motors and those controlling the motor is made possible by the highly accurate torque ripple models generated in JMAG.

*4 Torque ripple = Pulses of torque (force) produced when the motor rotates. The torque ripple that occurs causes noise and vibrations and may worsen controllability.

Prof. Akatsu
There are two directions I plan to expand my research. First, we have been able to control the torque ripple and the vibrations caused by the torque ripple in testing, but I would like to apply the same techniques to a larger motor with nonlinear characteristics, such as a switched reluctance motor (SRM)*5, which is a motor that does not use permanent magnets.I also plan to utilize JMAG in the actual controller. I want to use JMAG to provided feedback to the controller in real time by being able to fully trust the torque ripple calculated for the numerical model derived in JMAG. I am confident that this will happen within the next 10 years.

*5 Switched reluctance motor (SRM) = A motor that rotates using the difference in electrical resistance between the stator and the rotor. Even though switched reluctance motors have superior heat resistance due to a simple structure compared to rotating magnetic field type motors that have permanent magnets embedded into the rotor, there application into electric automobiles has been deemed difficult because the usage efficiency of torque and energy is low making miniaturization difficult. However, there has been some success in small, high output SR motor development.

Appearance of the equipments

Torque ripple control (simulation results)

Torque ripple control (actual results)

Evolving Motor Technology and Analysis Software

Prof. Akatsu
It is no coincidence that the introduction of permanent magnet motors in the early 1990’s and viable analyses using the finite element method coincide.
It could be said that the needs and innovation have created a cycled development. Each progressing mutually from the others’ success.
There was a time when development was all about solving equations for the control. Motors with new structures or new methods of control can’t be discovered by repetitively solving equations, and if they could, it would take a vast amount of time. The “Prius” hybrid*6 and other electric vehicles would not be around today without analysis software using the finite element method.
There is bilateral development where this type of software promotes advances in motor technology which in turn filters back to the software’s own innovation. This cycle of innovation will continue into the future.

Prof. Akatsu
JMAG has supported development engineers since its release. Engineers needed a tool like JMAG and they were waiting patiently for its release.
I started using JMAG in 2000 for a research center of an automotive manufacturer when I was asked to compare the magnetic field analysis results of the various software that were available. Any misgivings about the analysis accuracy were due to a lack of understanding about how to use the software as a tool as well as the limited number of analysis examples that combined data comparing the analysis results to actual data. However, the more analysis experience that was accumulated, the more trust JMAG gained.
Another reason that JMAG earned as much confidence as it did was the support provided by JSOL. The user friendly technical perspective they offered builds a strong foundation of trust. The support team is very fast in responding to any trouble or questions that we run into. They are on top of the problem right after we contact them. I am very grateful for this support, and I don’t think it is going to far to say it is unacceptable that other companies are not keeping pace with this kind of support.

*6 Prius is a registered trademark of the Toyota Motor Corporation.

Understanding the “Logic” of Results Obtained in JMAG Brings Technological Reform Reducing Post-processing

Prof. Akatsu
I would like them to consider the results they obtain until they are convinced of the accuracy. They need to know intuitively that the analysis results are a simulation. An analysis, no matter how advanced the software, will not provide any answers if the analysis is not setup correctly. They need to use the software with this always in the back of their minds.
Even in “the 5 rules” I have put in place at my lab., I have included “always consider the fundamental principles.” They need to measure the results and then investigate whether or not they have obtained the right answer. This type of approach is important.

Prof. Akatsu
First I would like to see more ways to utilize the analysis results effectively.
For example, Professor Katsumi Yamazaki’s (Chiba Institute of Technology) method for isolating the magnetic flux produced by the magnets and the magnetic flux produced by the armature windings suggests a new way of using analysis results. This means that the “analysis results” are not just taken as is, but rather evaluated by “breaking down” and analyzing those results. If the contents and logic behind the results is understood, the areas that need to be focused on, including problems that may occur at the next stage, can be grasped. Maybe it should even be called a “post-processing breakdown.”
The plots to display the magnetic flux density distribution by frequency are very advantageous. I am sure this type of freedom in post-processing will determine the value of analysis tools in the future. I would also like an established method to simulate the charge and spin of magnetic fields (spintronics) to capture in what way electromagnetic fields are produced, something that is still not fully understood.
Nevertheless, JMAG is indispensable for companies in Japan and the rest of the world to develop motors viable in extremely competitive industrial markets. Permanent magnets are wonderful, but we relay on China for the rare-earth materials. This is where the concept of a high-performance SRM comes from. SRM do not require magnets and allow us to conserve natural resources.
Even further motor development expanding added value can be expected, from motors that cannot be reverse engineered to motors exhibiting two opposing characteristics, or motors with built-in inverters. JMAG is also in the background supporting these advancements by promoting a more collaborative development.

Freshman Noriya Nakao
Electrical Engineering and Computer Science Graduate Program
Shibaura Institute of Technology

Student’s Perspective
I feel that JMAG is a very advanced analysis tool, but, at the same time, I feel it takes a certain amount of engineering experience. For example, to set windings, someone with no experience would not be able to visualize how to specify the various settings. Often, the user doesn’t realize they have made a mistake until after the results have been obtained.
This problem might stem from the difficulty of switching from thinking about things 2-dimensionally to thinking about them 3-dimensionally. In other words, the real “power” of JMAG is its ability to grow as a tool as the user gains more experience.

Interviewee

Associate Professor Kan Akatsu
Department of Electrical Engineering Shibaura Institute of Technology

Laboratory name:

Associate Professor Kan Akatsu

School name:

Department of Electrical Engineering Shibaura Institute of Technology

Research overview:

The M&E Energy Conservation Lab. was established in 2009 at the Toyosu campus of the Shibaura Institute of Technology. Their research focuses on the primary challenges for higher efficiency and performance of motors and generators. The M&A Energy Conservation Lab.’s mission is to “contribute to reducing CO2 emissions 50% through engineering by 2050.”
Associate Professor Kan Akatsu received his Ph.D in electrical engineering from Yokohama National University before joining the Nissan Research Center. In 2003, he joined the department of Electrical and Electric Engineering at Tokyo University of Agriculture and Technology from 2003 to March 2009 until accepting a position as an associate professor in the Department of Electrical Engineering at Shibaura Institute of Technology. Kan Akatsu specializes in electromechanical energy conservation, power electronics, and control engineering. He is also a member of the Institute of Electrical Engineers Japan and various other societies.

URL:

http://www.sic.shibaura-it.ac.jp/~akatsu/english/

[JMAG Newsletter January, 2011]