JP2002156361A - Magnetic characteristics distribution estimating method for magnetic material and quality evaluation method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide magnetic characteristics distribution estimation method of a magnetic material, capable of quantitatively expressing microscopic magnetic characteristics distribution, and to provide a quality evaluation method using it. SOLUTION: In an excitation process of a magnetic circuit and a magnetic material excited by a drive coil, the magnetic characteristics of respective image points are estimated by two or more visualized images made to correspond to the excited state, so that two-dimensional or three-dimensional magnetic characteristics distribution of the magnetic material is estimated. When an image average magnetic flux density is known in the two or more visualized images of the excited state, it is desirable that the magnetic flux density be determined, using the known magnetic flux density by the image processing and that the Helmholtz equation be used for the estimation of the magnetic characteristics.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for estimating a magnetic property distribution of a magnetic material and a quality evaluation method using the same.

[0002]

2. Description of the Related Art Magnetic materials include soft magnetic materials such as electromagnetic steel sheets, soft ferrite, and permalloy, and hard magnetic materials such as permanent magnets. The quality of these magnetic materials is evaluated by a magnetization curve and a hysteresis loop. In particular, a soft magnetic material such as an electromagnetic steel sheet is evaluated by an Epstein test or a ring test, and a magnetization curve, a hysteresis loop, a magnetic permeability, an iron loss, and the like are measured as an average value of the entire sample.

[0003] Since higher performance is required for these magnetic materials, the magnetic properties have been improved by macroscopic evaluation based on the average value of the entire sample as described above. However, general magnetic materials are polycrystalline, and the distribution of magnetic characteristics is caused by individual crystal grains inside the material, which greatly affects the magnetic characteristics of the entire sample. It is also important to evaluate the magnetic properties of local parts. For this reason, a local magnetic flux density or a magnetic field is detected by a search coil or the like, and the magnetic characteristic distribution is measured. In the method using a search coil, the minimum area that can be detected is often larger than a crystal grain, so that a magnetic domain observation is performed for further minute analysis.

The magnetic domain observation method includes a method of detecting a magnetization direction based on a difference in the state of reflection or scattering of electrons incident on a sample by an SEM to obtain a magnetic domain image, and a method of utilizing a difference in light reflection depending on the magnetization direction by the Kerr effect. And a method of detecting magnetic flux leakage from the sample surface with magnetic powder to know the magnetization state. By observing these magnetic domains, the motion of the domain wall and the magnetization distribution are estimated.

[0005]

However, since the results obtained from these images are not quantitative, it is difficult to compare the images, it is an empirical evaluation, and individual differences are likely to appear. Further, evaluation of the entire image (quantitative evaluation of two-dimensional distribution such as magnetization direction, size of magnetic domain, ease of movement of domain wall, etc.) has not been performed much. On the other hand, in the magnetic domain observation by the Kerr effect, the whole reflected light is taken in and the magnetization curve and the like are measured, but the reflection area is relatively large compared to the size of the crystal grain, and the area is narrowed down to one crystal grain or the like. , The sensitivity is lowered, and the accuracy of the obtained magnetization curve is not obtained.

SUMMARY OF THE INVENTION The present invention has been made in view of the above-mentioned problems of the prior art, and a magnetic property distribution estimating method capable of quantitatively expressing a microscopic magnetic property distribution of a magnetic material and a magnetic material distribution estimating method. It is an object of the present invention to provide a method capable of performing quality evaluation with high accuracy.

[0007]

The gist of the present invention is as follows. (1) In the excitation process of the magnetic material, two-dimensional or three-dimensional magnetic characteristic distribution of the magnetic material is estimated by estimating the magnetic characteristics of each point of the image from the visualized images corresponding to two or more excitation states. A method for estimating a magnetic property distribution of a magnetic material. (2) When the average magnetic flux density of the visualized images in two or more excited states is known, the magnetic flux density by image processing is determined using the known magnetic flux density. A method for estimating magnetic property distribution of a magnetic material as described in the above. (3) Using the magnetic property distribution estimation method of the magnetic material described in (1) or (2) above, in a specified area, average the magnetic properties of each point in the area to estimate an average magnetic property. A method for estimating magnetic properties. (4) The method for estimating magnetic properties of a magnetic material according to any one of the above (1) to (3), wherein the Helmholtz equation is used for estimating the magnetic properties. (5) A quality evaluation method for a magnetic material, wherein the quality of a material or a member is evaluated using the method according to any one of the above (1) to (4).

[0008]

BEST MODE FOR CARRYING OUT THE INVENTION The magnetic material of the present invention is a soft magnetic material such as a permalloy such as an electromagnetic steel sheet, an amorphous magnetic material, a ferrite, or a hard magnetic material of a permanent magnet, and may be a semi-hard magnetic material. The visualized image corresponding to the excitation state is a magnetic domain observation image by the SEM method, the Kerr effect method, a magnetic domain image by the Bitter method, or a magnetic domain image by another method, and the microscopic magnetization state is indicated by analog or digital. Any method and method may be used for an image.

In the magnetic domain image obtained by the SEM, the magnetization direction can be detected based on the difference in the reflection (or scattering) state of the sample incident electrons, and the magnetization state at a certain depth from the surface of the magnetic material is detected, thereby obtaining a magnetic domain image. By changing the acceleration voltage (incident energy), the detected depth can be changed. The magnetic domain image by the Kerr effect is a method of obtaining a magnetic domain image by utilizing a difference in light reflection depending on a magnetization direction, and it is possible to know a magnetization state near a surface of a magnetic material. In the image by the bitter method, the magnetic state can be detected by detecting the magnetic flux leakage from the magnetic material surface with the magnetic powder.

If a transmission electron microscope can be applied to obtain a magnetic domain image, the average magnetization state of the entire magnetic material in the depth direction can be known. Therefore, in one excitation state, an SEM magnetic domain image in which the acceleration voltage is changed,
By combining the two or more types of magnetic domain images,
Since a three-dimensional magnetic domain image can also be estimated, not only two-dimensional processing of magnetic characteristics but also three-dimensional data processing including the depth direction can be performed.

The image data at each point in the visualized image obtained by the above method is an analog or digital amount corresponding to the amount of electrons, the amount of light, and the amount of magnetic flux leakage (or the amount of magnetic powder) detected corresponding to the magnetization. In the SEM method, the Kerr effect method, etc., the image data corresponds to the magnitude of the magnetization direction with respect to the excitation direction or the magnitude of the cosine component with respect to the excitation direction, and the bitter method corresponds to the magnitude of the magnetic flux leakage. Corresponding to the magnetic flux density of

In the present invention, a two-dimensional visualized image corresponding to two or more excitation states or a three-dimensional image estimated by a combination of images can be used to arbitrarily determine the surface of the magnetic material or an arbitrary region (point) near the surface. , Lines, planes, and volumes). If there are two or more excitation states, it is possible to correspond to two or more points in the magnetization curve. Therefore, if the magnetic flux density is estimated from image data in an arbitrary area of the image, the whole or a part of the magnetization curve or the hysteresis loop can be obtained. You can ask.

The magnetic field (or the magnitude of the magnetic field) H corresponding to the excitation state is obtained from the current excited from the outside, the induced electromotive force of the magnetic field detecting coil, and the voltage generated by a Hall element, a magnetoresistive element, or the like. When the exciting current is used, the average magnetic field in the sample is obtained, and when the coil or element for detecting the magnetic field is used, the average magnetic field in a region corresponding to the size is obtained. If the relationship between the domain wall motion and the local magnetic field is known in advance, a magnetic field that can be estimated from the domain wall motion may be used. It goes without saying that these types of magnetic fields need to be selected according to the purpose.

The magnetic flux density B corresponding to the data U at an arbitrary point on the image may be obtained by directly converting the image data U to the magnetic flux density B if the correspondence between U and B is clear. When the average magnetic flux density of the entire image in the visualized image of the state is known, the magnetic flux density by image processing may be determined from the known average magnetic flux density and the average value of the entire image of U.

As the excitation state, a demagnetization state (magnetic flux density =
0), magnetic saturation state (magnitude is saturation magnetization)
Or a state in which the magnetic flux density is measured for a certain magnetic field. The image data and the magnetic flux density exist for the excitation point, and the excitation point alone may be sufficient if the object is achieved, but if necessary,
The interpolation between the excitation points may be performed by various methods. For example, a method of linearly approximating two excitation points, or three excitation points H ₁ , H ₂ , H
₃ and the corresponding image data U ₁ (H ₁ ), U ₂ (H
₂ ), U ₃ (H ₃ ), constants a, Λ, and b of the following equation are determined, and image data U between or near these magnetic fields is determined.
(H) may be obtained, and the magnetic flux density may be obtained. U (H) = a · exp (−ΛH) + b

This method is based on the following Helmholt method.
z). ▽ ² U + ε∂U / ∂H = −σ where U is scalar data at an arbitrary point in the image, is a function of the magnetic field H, the first term on the left side represents spatial spread, and the second term is based on the magnetic field or the like. The right side is the source density of the image of the final image. If the final image data is U _Final , ▽ ² U _Final = −σ. Now, assuming that the initial values are H _Start and U _Start ,
In general, U (H) = (U _Start −U _Final ) exp ｛−Λ (H−H
_Start )｝ + U _Final , which can be obtained, so this method may be used.

As described above, if the magnetic flux density is determined corresponding to the image data, the magnetic characteristic is determined as a two-dimensional distribution or a three-dimensional distribution corresponding to the visualized image. Here, the magnetic characteristics are a magnetization curve and a hysteresis loop, and are a residual magnetization, a coercive force, a magnetic permeability, an iron loss and the like obtained from these.

According to the present invention, not only the two-dimensional distribution and the three-dimensional distribution of the magnetic characteristics are obtained, but also the magnetic characteristics at an arbitrary point of the image can be obtained instantaneously from the magnetic characteristics distribution. In addition, it is also possible to specify a necessary area and obtain an average magnetic characteristic in the specified area. The designated region is a line, a plane, or a space (volume) other than a point, a crystal grain boundary, a crystal grain, a region divided by a certain size, and the whole or a part of a measurement range.

Using such a method, the quality of a magnetic material or a member made of a magnetic material can be evaluated. Magnetic properties are affected not only by the crystal grains (grain boundaries, size) and texture of the material, but also by strain, stress, and defects in the material. Surface properties (surface shape, film properties, etc.)
Is also affected. Therefore, as a material, the size of crystal grains, grain boundaries, and texture, precipitates and inclusions in the material,
Crystal defects such as dislocations can be detected, and from these results, nondestructive inspection of members using a magnetic material becomes possible, and information relating to shape defects, cracks and fractures can be obtained.

[0020]

EXAMPLES Example 1 Magnetic domain images of a test piece of grain-oriented electrical steel sheet from which the insulating film had been removed were sequentially excited from a demagnetized state to 324 A / m, and measured by SEM. Table 1 shows the results of the measurement of the magnetic flux density in each excitation state at the same time.
Further, among the SEM images, magnetic domain photographs of image numbers 1 to 10 are shown in FIGS.

U _k , U _{k + 1} , U _{k + 2 of} magnetic domain image data
Was used to determine Λ _k from the following equation. Λ _k = − (1 / ΔH) ln ｛(U _{k + 1} −U _{k + 2} ) / (U _k
−U _{k + 2} )｝ ΔH = H _{k + 1} −H _{k Using} this Λ _k , U = (U _k −U _{k + 2} ) exp ｛−Λ _k (H−H _k )｝ + U
_{As k + 2} , image data at an arbitrary magnetic field H in H _k <H <H _{k + 1} was obtained. Conversion to magnetic flux density is performed by using images k and k
The average values of the image planes of the image data U _k and U _{k + 1} of +1 are defined as the magnetic flux densities B _k and B _{k + 1 in} the excited state of the images k and k + 1, and linear interpolation using these two points is performed.

In this manner, the magnetic field is from 0 A / m to 324 A / m
The magnetic domain image at ~ -11 A / m was continuously estimated with respect to the magnetic field, and the entire image was averaged to estimate the magnetization curve of the grain-oriented electrical steel sheet. FIGS. 11 to 17 show the estimated magnetic domain images and magnetization curves. FIGS. 18, 19, and 20 show estimated magnetization curves at respective points in FIG. The designated points 1 and 2 are the magnetization curves of the 180 ° magnetic domain portion, but are easily magnetized (FIG. 18), and are hardly magnetized in the lancet domain portions 3 and 4 and the strained portions 5 and 6 (FIG. 18). 19 and FIG.
0) is known. In addition, it can be seen that the hysteresis of the strained portion is increased even in a high magnetic field.

[0023]

[Table 1]

Example 2 An image corresponding to an instantaneous iron loss distribution was obtained from the magnetic domain image of Example 1. The imaginary part of the lambda _k denotes the delay of the magnetic domain changes to the magnetic field, relate to the loss i.e. iron loss. The results are shown in FIGS. 21 and 22. FIG. 21 shows an iron loss distribution at a magnetic flux density of 1.7 T when the magnetic flux rises.
2 corresponds to the iron loss distribution at 1.9 T when the magnetic flux rises. In FIG. 22 with a higher magnetic flux density, the occurrence of iron loss is large, and it can be seen that the iron loss occurs on the entire surface of the observed steel sheet.

[0025]

As described above, the magnetic material distribution estimation method of the present invention estimates a two-dimensional or three-dimensional magnetic characteristic distribution, and quantitatively expresses a microscopic magnetic characteristic distribution. can do. Also, a very microscopic expression can be obtained as compared with a conventional method using a search coil, a Hall element, or the like. Therefore, the effect of the magnetic material on the magnetic characteristics of the magnetic material can be microscopically investigated.

For example, since the magnetic characteristics of each crystal grain can be obtained individually, it is possible to examine the difference in the magnetic characteristics between crystal grains and to quantify the magnetic influence of the crystal grain boundary.
It can also be used for subboundary effects and quantification at grain boundaries,
There is a possibility that minute distortion of the crystal structure can be detected.
In addition, there is a possibility that it is possible to quantify the improvements and precipitates in the material, and to quantify the size of the area where the improvements and precipitates magnetically influence, etc. Can also be used for quantification of When this method is applied to a member using a magnetic material, it can be applied to nondestructive inspection, and it is considered that defect locations can be specified and quantified.
It can be used for nondestructive inspection and quality evaluation of magnetic members in all fields such as manufacturing equipment and nuclear power equipment.

[Brief description of the drawings]

FIG. 1 is a view showing an actually measured SEM magnetic domain image (image number 1).

FIG. 2 is a view showing an actually measured SEM magnetic domain image (image number 2).

FIG. 3 is a view showing an actually measured SEM magnetic domain image (image number 3).

FIG. 4 is a view showing an actually measured SEM magnetic domain image (image number 4).

FIG. 5 is a view showing an actually measured SEM magnetic domain image (image number 5).

FIG. 6 is a view showing an actually measured SEM magnetic domain image (image number 6).

FIG. 7 is a view showing an actually measured SEM magnetic domain image (image number 7).

FIG. 8 is a view showing an actually measured SEM magnetic domain image (image number 8).

FIG. 9 is a view showing an actually measured SEM magnetic domain image (image number 9).

FIG. 10 is a view showing an actually measured SEM magnetic domain image (image number 10).

FIG. 11 shows an estimated magnetic domain image and a magnetization curve (0.06
T).

FIG. 12 shows an estimated magnetic domain image and a magnetization curve (0.69).
T).

FIG. 13 shows an estimated magnetic domain image and a magnetization curve (1.78).
T).

FIG. 14 shows an estimated magnetic domain image and a magnetization curve (1.95).
T).

FIG. 15 shows an estimated magnetic domain image and a magnetization curve (1.77).
T).

FIG. 16 shows an estimated magnetic domain image and a magnetization curve (0.61).
T).

FIG. 17 shows an estimated magnetic domain image and a magnetization curve (−1.55).
T).

FIG. 18 is an estimated magnetization curve of a designated point (designated points 1 and 2).

FIG. 19 is an estimated magnetization curve of a designated point (designated points 3 and 4).

FIG. 20 shows estimated magnetization curves of designated points (designated points 5 and 6).

FIG. 21 is a graph corresponding to iron loss distribution (1.7T).

FIG. 22 is a graph corresponding to iron loss distribution (1.9 T).

[Explanation of symbols]

 1, 2: 180 ° magnetic domain portion (FIG. 18) 3, 4: Lancet magnetic domain portion (FIG. 19) 5, 6: Strain-containing portion (FIG. 20)

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Choko Saito 3-7-2 Kajino-cho, Koganei-shi, Tokyo Faculty of Engineering, Hosei University (72) Inventor Hisashi Endo 3-7-2, Kajino-cho, Koganei-shi, Tokyo Hosei F-term within the Faculty of Engineering (reference) CB13 CC03 DB02 DB09 DC16

Claims (5)

[Claims] In the exciting process of a magnetic material, a magnetic property of each point of an image is estimated from a visualized image corresponding to two or more exciting states, thereby obtaining a two-dimensional or three-dimensional magnetic material.
A method for estimating a magnetic property distribution of a magnetic material, comprising estimating a two-dimensional magnetic property distribution. 2. The method according to claim 1, wherein when the average magnetic flux density is known in two or more excited visualized images, the known magnetic flux density is used to determine the magnetic flux density by image processing. 2. A method for estimating a magnetic property distribution of a magnetic material according to the first aspect. 3. A method for estimating a magnetic property distribution of a magnetic material according to claim 1 or 2, wherein a magnetic property of each point in the designated area is averaged to estimate an average magnetic property. A method for estimating magnetic properties. 4. The method for estimating magnetic properties of a magnetic material according to claim 1, wherein a Helmholtz equation is used for estimating the magnetic properties. 5. A quality evaluation method for a magnetic material, wherein the quality of a material or a member is evaluated using the method according to any one of claims 1 to 4.