Article: Paper Introduction


Issue 6 Lesson on Advanced Iron Loss Analysis

In this series, I would like to introduce various papers that present ways of using JMAG while performing electromagnetic field simulation. In this 6th edition, I will describe nine papers that focus on the impact of iron loss and enlighten about the necessary knowledge and modeling technologies for advanced iron loss evaluation.


Mr. Kazuyuki Narita from JSOL is in charge of introducing the papers covered in this issue.Issue 5 described initiatives related to JMAG loss analysis. The artiucle talked about how advanced evaluation of iron loss in electromagnetic field simulation required taking a variety of steps to counter the effects of iron loss and then introduced many initiatives from JMAG. In the July 2013 next-generation seminar that we held loss analysis was also an issue brought up, as were numerous countermeasures against loss. This article will introduce papers that describe the various causes of the different types of losses.
Images in this text are all self-created and have not been taken from papers or edited.

Various Factors that Affect Loss

[1] Kaido. "Kaitenki ni okeru Tesson Kyodo ni tsuite (Iron loss behavior in rotating machinery)," Electrical Engineering Rotating Machiner Research Materials, RM-00-119, 2000.

Magnetic objects like magnetic steel sheets or ferrite, in which loss occurs, are affected by a number of factors that outnumber the ideal of having no impact. That is what makes evaluation of loss in electromagnetc field simulation difficult. This paper looks at the five major causes of iron loss in rotating machinery (magnetic flux distribution, rotational fields, stress strain, time harmonics and spatial harmonics) and broadly summarizes their mechanisms and effects. It also touches how to model electromagnetic steel sheet in the equivalent magnetic circuits to understand magnetic steel sheet properties and evaluate rotating machine performance.
I introduce this book first because it contains a large variety of causes and presents them in an easily understandable manner.

Factors in the Production Process

To produce rotating machines or transformers, magnetic steel sheets are punched out of a press and laminated. After lamination, the magnetic steel sheets are caulked, welded or bolted to fix them into place. Finally, the lamintated iron core is pressed into a frame and undergoes a process called hardening. Magnetic steel sheets are exposed to all sorts of effects during this production process. I would like to describe here for you a paper that touches on these factors within the production process.

[2] Kaido, Mogi, Fujikura, Yamasaki: "Punching Deterioration Mechanism of Magnetic Properties of Cores", IEEJ Trans. FM, vol.128, no.8, 2008

Pressing magnetic steel sheets during the production process is known to have an effect on the magnetic properties of the magnetic steel sheets and iron loss properties. But the mechanism of those effects is less well-known. This paper tries to clarify the mechanism of the impact on properties that punching has on magnetic steel sheets by observing a cross-section of a magnetic area's structure taken from a magnetic steel sheet. This paper tries to clarify the mechanism of the impact on properties that punching has on magnetic steel sheets by observing a cross-section of a magnetic area's structure taken from a magnetic steel sheet. The breadth of the regions subjected to plastic deformation appear to variate due to factors such as clearance, plate thicknes or Si content. The breadth of the regions subjected to plastic deformation appear to variate due to factors such as clearance, plate thicknes or Si content.
It is also a good reference due to the rich amount of data it contains regarding coercive force through punching, B-H curves and hysteresis loss degradation. We will show you how we have done this, using examples.

[3] Kashiwara, Fujimura, Okamura, Imamura, Yashiki. "Denjikoban no Uchinuki ni yoru Jiki Tokusei Rekkaryo ni Oyobasu Uchiunuki Joken no Eikyo no Kaiseki (Analysis of the Impact of Punching Conditions on Degradation of Magnetic Properties due to Punching of Magnetic Steel Sheets)," The Institute of Electrical Engineers of Japan (IEEJ) Magnetics Technical Meeting. MAG-08-74, 2008.

A modeling method is to measure the properties in a thin magnetic steel sheet punched out as a test and use those properties in a magnetic field analysis of the impact of the punching process on regions (This is also in the paper mentioned in [2], but generally a punched item 2-to-3 times thicker than the insulation is called a region) and give those material properties. Even if understanding that punching casuses degradation in properties, how do you know what punching process influence should you model in an analysis? This is a practical method, but the test piece must be punched under exactly the same conditions as for actual machinery and the impacted region width needs to be estimated, both of which can be problem areas.
This paper describes a method for modeling a combination of a deformation analysis (Abaqus from SIMULIA) and an electromagnetic field analysis (JMAG) to determine the impact of the punching process. Firstly, conduct the deformation analysis to obtain the warping or remnant stress distribution caused by punching the magnetic steel sheet. Then use a table to allow for the impact of this warping and remnant stress distribution on magnetic properties and loss and map these as a distribution amoun in the material properties for an electromagnetic field analysis. Performing these steps will, in principle, enable modeling of the impact of the punching process. It is also valid because it enables understanding of the property degradation mechanism by changing the punching conditions through a deformation analysis.
This paper essentially examines only a single magnetic steel sheet, but it could be expanded to practial application in items such as rotating machinery by eflecting the warping or remnant stress distribution material properties to an electromagnetic field analysis using the homogenization method.

Fig. 1 Impact of Punching on Loss
Fig. 1 Impact of Punching on Loss

[4] Nakano, Fujino, Tani, Daikoku, Tsude, Yamaguchi, Arita, Yoshioka. "Tesshin Naibu no Ouryoku Bunpu wo Koryo shita Koseido Tesson Kaiseki Hoho (Method of Advanced Analysis of Iron Loss Allowing for Stress Distribution in the Inner Core," IEEJ Journal, vol.129, No.11, pp.1060-1067, 2009.

To fix in place a laminated core in a rotating machine, use press fitting or hardening of the core in the case. This will ensure the core undergoes some stress and change these to magnetic properties or loss. Normally, you would use principal stress (compressive or tensile) to allow for the influence of stress, but when the stress and magnetic fluzx are in different directions, the expression is insufficient.
This paper isolates the magnetic flux density from the core parts in an electromagnetic field analysis in the principal stress in a structural analysis to allow for the angle formed by stress and magnetic flux. As a result, it shows that conventional methods not allowing for direction and treating stress as having a major impact and the methods used here obtain the results closest to those obtained from actual measurements.

Fig. 2 Impact Stress has on Loss
Fig. 2 Impact Stress has on Loss

[5] Miyagi, Aoki, Nakano, Takahashi: "Effect of Compressive Stress in Thickness Direction on Iron Losses of Nonoriented Electrical Steel Sheet", IEEE Transactions on Magnetics, vol.46, No.6, pp2040-2043, 2010

In rotating machines or transformers, magnetic steel sheets can be caulked, welded, or fixed into place with bolts above and below a clamped or framed laminated core. In these cases, stress is placed on the lamination thickness direction of the core. Apart from our paper there are many other examples of papers describing changing properties if the stress is exerted through something like pressing toward the core's in-plane direction, but few report on the property changes stress brings about in a lamination thickness direction.
Our paper will show that even by exerting a comparatively minor compressed stress of 0.5MPa in the lamination thickness direction can increase hysteresis loss by as much as a maximum of 12%. We will also show how this increases abnormal eddy current loss.
Our paper contains one of the few documented measurement data of lamination thickness direction stress, which makes it an important contribution.

Magnetomotive Force Waveform Factors

The most commonly used iron loss evaluation methods are either Steinmetz's empirical law or another method based on that. Steinmetz's empirical law has a coefficient that expresses hysteresis loss and eddy current loss, but this is calculated from loss values obtained through Epstein's law or applying sinusoidal alternating flux in a single plate test method. But magnetic flux in actual rotating machines or transformers fluctuates in a much more complicated manner. For example, with a rotating machine, slot structures include items such as spatial harmonics or carrier time harmonics caused by PWM power conversions. When a transformer comes in contact with the rectification circuit, the core moves into a state of direct current bias magnetism. This paper gives an example of knowledge and modeling technologies regarding loss that does not come from sinusoidal alternating flux.

[6] Kaido, Yabumoto, Lee, Miyata: "Minor-loop Magnetic Properties of Non-oriented Electrical Steel Sheets" The Papers of Joint Technical Meeting on Static Apparatus and Rotating Machinery, IEE Japan SA-05-35ARM-05-35A2005

As stated above, hwen core magnetic flux density incudes harmonics or in direct current bias magnetism, it is possible to render a minor loop trajectory on the B-H curve. Measuring minor loop properties is an onerous task, so the data on them is important.
This paper is a vital reference as it displays numerous types of hysteresis loop properties. A must-read, in particular, is data about how even though the fluctuation quantity of magnetic flux density (AC components) stays the same, the iron loss greatly increases depending on the amount of direct current bias magnetism (DC components) (iron loss grows threefold with 1.2T of direct current bias magnetism).
In recent years, it seems increasingly likely that hysteresis modeling will be possible on even OTC software, but whether solutions it provides are correct and how it came about getting those results and what sort of hysteresis properties exist still need to be understood. This paper helps understand hysteresis properties.

Fig. 3 Hysteresis Properties in 50A1300 (Lower curve)
Fig. 3 Hysteresis Properties in 50A1300 (Lower curve)

[7] Matsuo, Shimode, Terada, Shimasaki: "Application of Stop and Play Models to the Representation of Magnetic Characteristics of Silicon Steel Sheet"CIEEE Transactions on Magnetics, vol.39, No.3, pp.1361-1364, 2003

In an electromagnetic field analyses, play modeling has recently been under the spotlight as a form of practically modeling magnetic hysteresis. The authors have continued research into play modeling, including methods of applying it to electromagnetic field analyses, methods to identify necessary functions for the play model from the measured data and expanding interest to AC hysteresis.
This paper includes an early paper by one of thie authors about the possibilities of practical application of play modeling in electromagnetic field analysis. This work is reccomended for those interested in play modeling.

Fig. 4 Lay Modeling an SR Motor Loss Analysis
Fig. 4 Lay Modeling an SR Motor Loss Analysis

[8] Bottauscio, Chiampi, Chiarabaglio: "Advanced Model of Laminated Magnetic Cores for Two-Dimensional Field Analysis", IEEE Transactions on Magnetics, vol.36, No.3, pp.561-573, 2000

As magnetic steel sheets have nonlinear magnetic properties, flux variations in the sheet are rapid and when the surface skin effect is strong, permeability is distributed in a complicated manner in the direction of the plate thickness. If that distribution includes harmonics or direct current bias magnetism it will have different conditions to those of the sunsoidal alternating magnetic flux. Consequently, it becomes difficult to predict the eddy current loss due based on Steinmetz's empirical law with assumptions about sinusoidal waveform alternating magnetic flux.
To precisely calculate eddy current loss under the complicated manner stated above requires a 3D transient response analysis modeling the sheet's thickness direction. But 3D transient response analyses are large and take a lot of time for calculations, both of which are problems.
To counter this, use a 2D analysis or magnetic density boundary condition obtained from a treated 2D analysis and conduct a 1D analysis to understand the eddy current distribution in what is a complicated and difficult proposal. Due to making an allowance for non-linear magnetic properties, it becomes possible to evaluate with high precision eddy current distribution when magnetic flux variations are complicated and the authors could report of successfully applying the application to items such as rotating machines. This paper presents a case study from the initial stage. This paper also describes the Preisach method to model hysteresis (Mathematically equivalent to the play model described in [7].

Fig. 5 1D Analysis of the Magnetic Steel Plate Thickness Direction
Fig. 5 1D Analysis of the Magnetic Steel Plate Thickness Direction

[9] Narita, Sakashita, Yamada, Akatsu: "Iron Loss Calculation of PM Motor by Coupling Analysis between Magnetic Field Simulator and Control Simulator(Second Report)," ICEMS2009, 2009

I'm sorry to boast, but this is a paper I co-wrote. Permanent magnet synchronous motors (PMSM) or variable speed induction motors as their power adjusted by the PWM inverter, but when doing so the PWM carrier generates time harmonics in the rotating machine and these harmonics cause an increase in iron loss. This paper obtains current including a PWM carrier in an operating state and calculates iron loss in an electromagnetic field analysis by incorporating PMSM behavior models created in an electromagnetic field analysis into the control and circuit system simulator. The result was that the tendencies in iron loss depending on the variations in items like carrier frequency or DC voltage matched well with the tendencies of actual measurements.
This kind of method can be used for optimal designing for the entire drive system losses including controllers and inverters.

Fig. 6 Impact on Harmonics Loss
Fig. 6 Impact on Harmonics Loss


I described the impact and causes of iron loss and informed about the necessary knowledge and modeling technologies for advanced iron loss evaluation.

(Katsuyuki Narita)