Properties of Iron Core Fabricated from Flaky-Shaped and Annealed Pure Iron Powder

Abstract

To improve the properties of powder magnetic core for motors, such as iron loss, resistivity and strength, pure iron powder was deformed into flaky-shaped and then annealed to decrease strain by deformation. It was confirmed that the strength of iron core increased with increasing the aspect ratio of powder. However, the resistivity extremely decreased. The iron core showed properties as follows; density d = 7.66 Mg·m⁻³, resistivity ρ = 2.5 × 10⁴ µΩ·m, iron loss P_c (the maximum magnetic flux density, B_max = 1 T, frequency, f = 400 Hz) = 30 W·kg⁻¹, iron loss P_c (B_max = 1 T, f = 800 Hz) = 63 W·kg⁻¹, radial crushing strength σ = 95 MPa. Especially, eddy current loss coefficient K_e was 0.008 mWs²·kg⁻¹, extremely lower than electromagnetic steel sheet and powder magnetic core reported ever.

 

This Paper was Originally Published in Japanese in J. Jpn. Soc. Powder Powder Metallurgy 64 (2017) 638–645. To show the shape of the powder more clearly, Fig. 2 was replaced.

1. Introduction

The powder magnetic core is a bulk body formed into a predetermined shape after subjecting the surface of the soft magnetic powder to binder or insulating coating treatment. The important characteristics are iron loss, strength, magnetic flux density, and magnetic permeability. Iron loss is the sum of hysteresis loss and eddy current loss. Coercive force and resistivity are closely related to those losses respectively. As the operating frequency becomes higher, the eddy current loss becomes dominant.

Comparing the powder magnetic core with the core by electromagnetic steel sheet, there are the following advantages and disadvantages. The advantages are (1) since it has isotropic magnetic characteristics, it is possible to design the flow of magnetic flux three-dimensionally, and it is possible to reduce the size and weight of parts, (2) since the particle size (50 to 150 µm) of the raw material powder is usually smaller than the sheet thickness (0.25 to 0.35 mm) of the electromagnetic steel sheet, the eddy current hardly flows and the iron loss is small in the high frequency region, (3) since powder is formed near net shape, material yield is higher than that of punching of electromagnetic steel sheet, and so on. Considering the advantages and disadvantages as described above, the powder magnetic core has been used in a high frequency region where the eddy current loss is dominant, and practical application to a part like a reactor whose required strength is relatively small, has been promoted.^1–8) Recently, application to motors has also been actively studied.^9–33) In order to apply the powder magnetic core to the motor, it is necessary to reduce the iron loss in the low frequency range and improve the strength. Examining raw material powders sold for powder magnetic core for motors, there is a powder showing an iron loss P_c (maximum magnetic flux density B_max = 1 T, frequency f = 400 Hz) of 29 W·kg⁻¹ and a transverse rupture strength TRS of 90 MPa.^34,35) These properties are good values as a powder magnetic core. However, they are still insufficient as compared with a general non-oriented electromagnetic steel sheet with a thickness of 0.35 mm. On the other hand, in the powder magnetic core using the powder atomized from electrolytic high purity iron and with low impurity content of 120 ppm, the coercive force, that is, the hysteresis loss is further reduced and the iron loss P_c (B_max = 1 T, f = 400 Hz) is decreased to 23 W·kg⁻¹.³⁶⁾ However, its transverse rupture strength TRS is as low as 50 MPa, making it difficult to balance low iron loss and high strength.

Based on the above situation, in this research, high purity spherical gas atomized powder which should have low hysteresis loss is used as a starting material, and by flattening it, the effect of entangling of powder is given, a compatibility of low iron loss and high strength was studied. Since the flattening reduces the demagnetizing field coefficient and increases the effective magnetic field, it is also expected to improve the magnetic flux density in the low magnetic field.^37–39)

The target characteristics were set as follows. In consideration of high revolution and high frequency accompanying the reduction in size and weight of the motor, we focused on low iron loss near 1 kHz. For that purpose, it was assumed that a resistivity ρ of the order of 10⁴ µΩ·m was required. The strength was targeted at 100 MPa, considering the value necessary for the motor.

2. Experimental Procedure

2.1 Preparation of flaky-shaped powder

Ar gas atomized pure iron powder made by Sanyo Special Steel Co., Ltd. with a grain size of −106 µm was processed into a flaky-shape by a wet attritor. Three types of flaky-shaped powder were prepared mainly by changing processing time.

The Powders were annealed in order to remove flattening strain and reduce coercive force by coarsening of crystal grains. In order to prevent sintering during annealing, SiO₂ powders (particle size about 2 µm) made by Admatechs Co., Ltd. were added at 4 to 6 mass%. Annealing was performed under the conditions of 1073 K and 3.6 ks in a H₂ and Ar mixed gas flow (H₂ concentration: 75 vol%). Thereafter, powder was immersed in pure water and ultrasonically cleaned to remove the SiO₂ powder floating on the water surface. After repeating cleaning several times, powder was washed with ethanol and dried.

The coercivity H_c and the saturation magnetization M_s of the powder were measured with a vibrating sample magnetometer VSM-3S-15 manufactured by Toei Industry Co., Ltd.

2.2 Fabrication and evaluation of powder magnetic core

For each powder, Sr–B–P–O phosphate coating,^40–42) 0.24 mass% Al–Si based alkoxide coating^43,44) and 0.24 mass% silicone resin coating⁴⁵⁾ were applied.

The coated powder was molded into a ring shape with an outer diameter of 39 mm, an inner diameter of 30 mm, and a thickness of 4 mm by warm compaction with die-wall lubrication.^46–49) No internal lubricant was added and the lithium stearate suspension as a mold lubricant was sprayed onto a die wall preheated to about 400 K. The molding pressure was 1176 MPa. This green body was annealed under the conditions of 873 K and 3.6 ks in an N₂ gas stream to obtain a powder magnetic core sample.

The obtained sample was evaluated as follows. The density was determined by measuring the weight and shape. The resistivity was measured by a four-terminal method using a digital multimeter. In order to evaluate the magnetic characteristics, 80 turns as a detection coil and 240 turns as an excitation coil of coated copper wire with a diameter of 0.5 mm were wound around the sample. The static magnetic field characteristics were measured using a DC self-recording magnetometer TRF-5A manufactured by Toei Industry Co., Ltd. In the text, B_Xk indicates the magnetic flux density measured at X kA·m⁻¹. The AC magnetic field characteristics were measured using a high frequency B-H curve tracer SY-8232 manufactured by Iwatsu Electric Co., Ltd. The measurement conditions of the iron loss were the maximum magnetic flux density B_max as 1 T and the measurement frequency f as 200, 400, 800, 1000 Hz. The transverse rupture strength TRS was measured according to JIS Z2507. It was separately confirmed that the radial crushing strength σ was equivalent to the TRS.

3. Results and Discussions

3.1 Preparation of flaky-shaped powder

The SEM image of the raw material powder and the prepared flaky-shaped powder are shown in Fig. 1, and the optical microscopic images of the cross section are shown in Fig. 2, respectively. The raw material powder shows a spherical shape peculiar to the gas atomized iron powder. On the other hand, about 20 flaky-shaped powders were randomly selected from the optical microscopic image of the cross section shown in Fig. 2, the long side length and the short side length were measured, and the aspect ratio of the powder (length of long side/length of short side) was calculated. The average aspect ratio of each powder was about 1.0, 3.8, 4.5, and 11.3, respectively.

Fig. 1

SEM images of powders with aspect ratio: a) 1.0, b) 3.8, c) 4.5, d) 11.3.

Fig. 2

Sectional micrographs of powders with aspect ratio: a) 1.0, b) 3.8, c) 4.5, d) 11.3.

The coercivity H_c of the flaky-shaped powder is shown in Fig. 3. The H_c increases as the aspect ratio of the powder increases, but by annealing at 1073 K, the H_c can be reduced to about 200 A·m⁻¹.

Fig. 3

Relationships between aspect ratio of powder and coercivity of powder H_c before and after anneal.

What we note here is the prevention of sintering between powders during annealing. If powder was sintered, pulverization treatment is necessary, and strain is added again, so that the coercivity increases. Thus, how to prevent sintering between powders is important. We have adopted a method using SiO₂ powder as an inhibitor from the experience so far. Sintering can be prevented by adding SiO₂ powder with an appropriate particle diameter at a certain ratio or more. As a method for separating and removing SiO₂ powder from the annealed iron powder, ultrasonic cleaning was examined. Figure 4(a) shows the polarity of the liquid used for ultrasonic cleaning and the saturation magnetization M_s of powder after cleaning with M_s of pure iron powder as a reference. As the polarity of the liquid is higher, it is found that M_s of powder after cleaning approaches M_s of pure iron powder. Defining the SiO₂ powder removal ratio as in the formula (1), the relationship between the polarity of the liquid and the SiO₂ powder removal ratio is as shown in Fig. 4(b).   

\begin{align} &\text{SiO$_{2}$ powder removal ratio (%)}\\ &\quad = \text{($M_{\text{s}}$ after removal of SiO$_{2}$ $-$ $M_{\text{s}}$ before removal of SiO$_{2}$)}\\ &\quad\quad/\text{($M_{\text{s}}$ of pure iron powder $-$ $M_{\text{s}}$ before removal of SiO$_{2}$)}\times 100 \end{align} (1)

Fig. 4

Relationships between relative polarity of liquid in ultrasonic cleaning and a) saturation magnetization M_s after removal of SiO₂, b) SiO₂ removal ratio. (Aspect ratio of powder; 3.8).

That is, it can be understood that pure water with high polarity is most suitable as a liquid for separating and removing SiO₂ powder from iron powder in ultrasonic cleaning. From the above investigation, the strain relief annealing condition of the flaky-shaped powder was optimized by combination of adding SiO₂ powder, removal of SiO₂ powder by ultrasonic cleaning with pure water and drying after alcohol substitution.

Figure 5 shows the powder appearance before and after SiO₂ powder removal. The small black spot on the flaky-shaped powder surface before removal is SiO₂ powder (Fig. 5(a)). Almost no SiO₂ is observed after removal (Fig. 5(b)). SiO₂ powder can be almost completely removed by ultrasonic cleaning considering the polarity of liquid.

Fig. 5

SEM BSE images of powders a) before and b) after SiO₂ removal. (Aspect ratio of powder; 3.8 liquid; water).

Next, a coating treatment for insulating property and strength of the powder magnetic core was carried out. In this experiment, coating conditions were not adjusted according to the change of the aspect ratio, and coating was carried out under same conditions. Figure 6 shows the powder appearance after coating. The dark part is the coating and the bright part is where the coating is insufficient. As the aspect ratio increased, coating became difficult and a missing part of the coating was seen. Particularly, flaky-shaped powder with an aspect ratio of 11.3 was hardly coated. The main reason for this would be that the flat surfaces overlap each other and the coating component becomes difficult to contact with the surface.

Fig. 6

SEM BSE images of coated powders with aspect ratio: a) 1.0, b) 3.8, c) 4.5, d) 11.3.

3.2 Influence of powder aspect ratio on properties of powder magnetic core

Figure 7 shows the optical microscope image of the cross section of the powder magnetic core. And Fig. 8 shows the SEM image of the fractured surface after the radial crushing strength test. In the powder magnetic core using the flaky-shaped powder, the flat powder which is remarkably deformed is entangled so as to be crimped in contrast to the powder magnetic core using the spherical powder.

Fig. 7

Sectional micrographs of iron core by powder with aspect ratio: a) 1.0 and b) 11.3.

Fig. 8

SEM images of fracture of iron core by powder with aspect ratio: a) 1.0 and b) 11.3.

Figure 9 shows the relationship between the powder aspect ratio and the density and resistivity of the powder magnetic core. Density was lower as the aspect ratio was larger. This is considered to be due to deterioration of powder filling properties and compressibility. The specific resistance also decreased as the aspect ratio was larger. This is probably because insulation coating became insufficient as shown in the previous section.

Fig. 9

Relationships between aspect ratio of powder and a) density and b) resistivity of powder core.

Figure 10 shows the relationship between powder aspect ratio and magnetic flux density B_5k, iron loss P_c (B_max = 1 T, f = 800 Hz) of the powder magnetic core. The larger the aspect ratio, the higher the B_5k despite the lower density as mentioned above. This is considered to be an effect that the demagnetizing field coefficient in the circumferential direction becomes smaller and the effective magnetic field becomes larger. However, the larger the aspect ratio, the larger the iron loss. This would be due to an increase in eddy current loss as the resistivity decreases due to insufficient insulation coating.

Fig. 10

a) Relationships between aspect ratio of powder and Magnetic density, B_5k and Iron loss (B_max = 1 T, f = 800 Hz), b) schematic view of ring-shaped powder core specimen.

Figure 11 shows the relationship between the powder aspect ratio and the radial crushing strength of the powder magnetic core. The higher the aspect ratio, the higher the strength. That is considered to be the effect of entanglement of flaky-shaped powder. It was found that to obtain the target strength of 100 MPa, it is sufficient to use a powder with an aspect ratio of about 4. However, as described above, if the aspect ratio is large, the iron loss increases. Figure 12 shows the relationship between the radial crushing strength and the iron loss of the powder magnetic core. The higher the strength, the larger the iron loss. The P_c (B_max = 1 T, f = 800 Hz) is 100 to 150 W·kg⁻¹ when the strength is about 100 MPa.

Fig. 11

Relationship between aspect ratio of powder and radial crushing strength of powder core.

Fig. 12

Relationship between radial crushing strength and iron loss (B_max = 1 T, f = 800 Hz) of powder core.

From the above results, it was found that the larger the powder aspect ratio, the better the magnetic flux density and the radial crushing strength of the powder magnetic core, but the density and resistivity decreased and the iron loss increased. In order to realize the both high strength and low iron loss, the appropriate powder aspect ratio would be 2 to 4.

3.3 Properties of powder magnetic core using flaky-shaped powder

As described above, in order to achieve both high strength and low iron loss, we focused on powder with aspect ratio 2 to 4, and reexamined process conditions such as insulation coating. The conditions were changed systematically to prepare a ring-shaped powder magnetic core and evaluate the properties of the specimens, and then finally the following conditions were determined. First, fine powder of less than 45 µm is removed and the annealing temperature of the powder is raised from 1073 K to 1173 K to reduce the hysteresis loss. Next, by reducing the resin coating amount from 0.24 mass% to 0.10 mass%, the decrease in magnetic flux density was minimized while ensuring specific resistance and strength. Furthermore, by changing the annealing condition from 873 K and 3.6 ks to 838 K and 4.2 ks, both high specific resistance and low hysteresis loss was achieved simultaneously. Figure 13 shows the BH curves of specimen in this research and Table 1 shows various properties of specimen in this research together with comparative material 1³⁴⁾ and 2.³⁶⁾ Comparing with the material 1 and 2, the iron loss of specimen in this research is slightly larger, but the strength is 1.3 to 1.9 times.

Fig. 13

Magnetization curve.

Table 1 Comparison of properties of powder cores.

The iron losses of the powder magnetic core and the non-oriented electrical steel sheet are compared below. It is known that eq. (2) holds about losses.   

\begin{equation} P_{\text{c}} = P_{\text{h}} + P_{\text{e}} = K_{\text{h}}\cdot f + K_{\text{e}}\cdot f^{2} \end{equation} (2)

Here, P_c: iron loss, P_h: hysteresis loss, P_e: eddy current loss, K_h: hysteresis loss coefficient, K_e: eddy current loss coefficient, f: frequency. Equation (3) is obtained from eq. (2).   

\begin{equation} P_{\text{c}}/f = K_{\text{h}} + K_{\text{e}}\cdot f \end{equation} (3)

Figure 14 shows the relationship between frequency and iron loss/frequency value (P_c/f) for several materials. From eq. (3), the intercept of the approximate straight line of this plot is the hysteresis loss coefficient K_h, and the slope is the eddy current loss coefficient K_e. The lower the both coefficient, the lower the iron loss. From Fig. 14, comparing the specimen in this research with other materials, the hysteresis loss coefficient K_h is high but the eddy current loss coefficient K_e is very low. The eddy current loss coefficient K_e was 0.039 mWs²·kg⁻¹ for the electromagnetic steel sheet and 0.024 mWs²·kg⁻¹ for the comparative material 2, whereas 0.008 mWs²·kg⁻¹ for specimen in this research. K_e for specimen in this research is about 1/5 of electromagnetic steel sheet and about 1/3 of comparative material 2. This means that the characteristic of suppressing the eddy current loss is excellent, and thought to be due to high specific resistance of the order of 10⁴ µΩ·m. It is predicted from Fig. 14 that the iron loss of specimen in this research would be lower than those of electromagnetic steel sheet and comparative material 2 at a frequency of about 1800 Hz or more.

Fig. 14

Relationships between frequency and iron loss (hysteresis and eddy current loss).

4. Conclusion

For applying powder magnetic core to motors in the future, we focused on low iron loss near 1 kHz and strength of 100 MPa, and investigated the effect of flaky-shaped pure iron powder then obtained the following findings.

1. (1)    As the powder aspect ratio increases, the coercivity increases, but by annealing at 1073 K, the coercivity can be reduced to about 200 A·m⁻¹. Sintering of iron powder can be prevented by mixing SiO₂ powder during annealing. After annealing, SiO₂ can be almost completely removed by ultrasonic cleaning using pure water.
2. (2)    The larger the aspect ratio of the powder, the lower the demagnetizing field coefficient in the direction parallel to the flat surface and the higher the magnetization. In addition, the radial crushing strength is also improved. However, insulation coating became insufficient, and specific resistance and density decreased.
3. (3)    The powder magnetic core prepared by using flaky-shaped powder having an aspect ratio of 2 to 4 and selected proper process conditions shows a radial crushing strength of 95 MPa and an iron loss (B_max = 1 T, f = 800 Hz) of 63 W·kg⁻¹. The eddy current loss coefficient was 0.008 mWs²·kg⁻¹, which was about 1/5 of the electromagnetic steel sheet and about 1/3 of the comparative material of the powder magnetic core. This is considered because the specimen in this research has a high resistivity of the order of 10⁴ µΩ·m.

Acknowledgments

We greatly appreciate Mr. Mikio Kondo, Dr. Atsuto Okamoto, Mr. Shoji Hotta, and Mr. Takanori Nakagaki who cooperated in carrying out this research.

REFERENCES

• 1)  T. Hattori, S. Tajima, M. Kondoh, M. Sugiyama, T. Yamaguchi, H. Kishimoto, S. Ohkouchi, T. Saito, S. Takemoto and Y. Fujita: Abstracts of Autumn Meeting JSPM, (2009) p. 99.
• 2)  S. Takemoto, Y. Fujita, T. Saito, T. Hattori, S. Tajima, M. Kondoh, M. Sugiyama, T. Yamaguchi, H. Kishimoto and S. Ohkouchi: Abstracts of Autumn Meeting JSPM, (2009) p. 100.
• 3)  K. Hidefumi, S. Ohkouchi, M. Sugiyama, T. Yamaguchi, T. Hattori, M. Tani, T. Suzuki, T. Saito, S. Takemoto and Y. Fujita: Abstracts of Autumn Meeting JSPM, (2009) p. 101.
• 4)  T. Hattori, M. Sugiyama, H. Kishimoto and T. Saito: Abstracts of Autumn Meeting JSPM, (2013) p. 80.
• 5)   S.  Yamamoto,  K.  Kusawake,  J.  Ido,  H.  Kawaguchi,  A.  Ito and  M.  Kato: SEI Tech. Rev. 185 (2014) 29–33.
• 6)   N.  Igarashi,  M.  Uozumi,  T.  Kosuge,  A.  Sato,  K.  Kusawake and  K.  Yamaguchi: SEI Tech. Rev. 186 (2015) 92–97.
• 7)   H.  Hashimoto,  K.  Zaitsu,  S.  Hayashi,  H.  Mitani and  K.  Inoue: Kobe Steel Eng. Rep. 65 (2015) 16–20.
• 8)  T. Inagaki, C. Ishihara and N. Nakayama: Abstracts of Autumn Meeting JSPM, (2016) p. 191.
• 9)  Y. Kageyama, S. Kosaka, N. Hirose and J. Asami: Abstracts of Autumn Meeting JSPM, (2003) p. 118.
• 10)  C. Ishihara, K. Asaka, Y. Enomoto and M. Ito: Abstracts of Spring Meeting JSPM, (2005) p. 201.
• 11)  M. Ueta, S. Uenosono, N. Nakamura and J. Asami: Abstracts of Spring Meeting JSPM, (2005) p. 202.
• 12)  K. Mori, K. Kanagawa and R. Nakayama: Abstracts of Autumn Meeting JSPM, (2005) p. 138.
• 13)  K. Kanagawa, K. Mori and R. Nakayama: Abstracts of Autumn Meeting JSPM, (2005) p. 139.
• 14)  L. Pennander, G. Nord, K. Maezawa, M. Saito and K. Kanno: Abstracts of Autumn Meeting JSPM, (2005) p. 140.
• 15)  K. Kanagawa and K. Mori: Abstracts of Spring Meeting JSPM, (2006) p. 134.
• 16)  H. Oyama, T. Sonoda, T. Nishioka, T. Sawai and Y. Takeda: Abstracts of Spring Meeting JSPM, (2006) p. 136.
• 17)  M. Yokoi, N. Kamiya and T. Nakano: Abstracts of Autumn Meeting JSPM, (2006) p. 136.
• 18)  K. Hiramoto, S. Tajima, T. Hattori and Y. Inaguma: Abstracts of Autumn Meeting JSPM, (2006) p. 197.
• 19)  C. Ishihara, K. Asaka, Y. Enomoto and M. Ito: Abstracts of Autumn Meeting JSPM, (2006) p. 198.
• 20)  K. Kanagawa: Abstracts of Autumn Meeting JSPM, (2006) p. 199.
• 21)  H. Oyama, Y. Yamada, T. Nishioka, T. Sawai and S. Morimoto: Abstracts of Autumn Meeting JSPM, (2006) p. 200.
• 22)   H.  Oyama,  K.  Kitajima,  H.  Toyoda,  T.  Nishioka,  K.  Kimura,  T.  Sawai and  T.  Sonoda: SEI Tech. Rev. 169 (2006) 79–81.
• 23)   M.  Ueta,  S.  Uenosono,  N.  Nakamura and  J.  Asami: J. Jpn. Soc. Powder Powder Metall. 53 (2006) 297–303.
• 24)   Y.  Enomoto,  M.  Ito,  R.  Masaki and  K.  Asaka: IEE J Trans. Ind. Appl. 126 (2006) 1585–1590.
• 25)  C. Ishihara, T. Inagaki, K. Asaka, Y. Enomoto and M. Ito: Abstracts of Spring Meeting JSPM, (2007) p. 100.
• 26)   Y.  Yamada,  S.  Ohashi,  T.  Ariyoshi,  T.  Sawai and  S.  Morimoto: J. Jpn. Soc. Powder Powder Metall. 54 (2007) 428–431.
• 27)  T. Inagaki, C. Ishihara, Y. Enomoto and M. Ito: Abstracts of Autumn Meeting JSPM, (2009) p. 105.
• 28)  S. Kohnushi, Y. Okada and H. Dohmeki: Abstracts of Spring Meeting JSPM, (2010) p. 217.
• 29)  T. Miyamura, H. Mitani, H. Hojo and T. Kamijo: Abstracts of Autumn Meeting JSPM, (2011) p. 241.
• 30)  K. Inoue, H. Hashimoto, H. Hojo and S. Kasai: Abstracts of Spring Meeting JSPM, (2012) p. 148.
• 31)   M.  Namba,  K.  Hiramoto and  H.  Nakai: Inst. Electr. Eng. Japan 135 (2015) 1085–1090.
• 32)  T. Ueno, A. Watanabe, K. Yamada, R. Tsunata and M. Takemoto: Abstracts of Autumn Meeting JSPM, (2016) p. 95.
• 33)  J. Shima and C. Ishihara: Abstracts of Spring Meeting JSPM, (2017) p. 144.
• 34)  Kobe Steel Website, http://www.kobelco.co.jp/products/powder/magmel.html.
• 35)   H.  Hojo,  N.  Akagi,  T.  Sawayama and  H.  Mitani: Kobe Steel Eng. Rep. 60 (2010) 79–83.
• 36)   T.  Maeda,  H.  Toyoda,  N.  Igarashi,  K.  Hirose,  K.  Mimura,  T.  Nishioka and  A.  Ikegaya: SEI Tech. Rev. 166 (2005) 1–6.
• 37)   N.  Nakamura,  Y.  Ozaki and  K.  Higuchi: Kawasaki Steel Giho 33 (2001) 184–187.
• 38)  N. Nakamura, Y. Ozaki, M. Ueta and S. Uenosono: Abstracts of Spring Meeting JSPM, (2001) p. 121.
• 39)  M. Ueta, Y. Ozaki and S. Uenosono: Abstracts of Autumn Meeting JSPM, (2001) p. 140.
• 40)   S.  Tajima,  T.  Hattori,  M.  Kondoh,  H.  Kishimoto,  M.  Sugiyama and  T.  Kikko: IEEE Trans. Magn. 41 (2005) 3280–3282.
• 41)   S.  Tajima,  T.  Hattori,  M.  Kondoh,  H.  Kishimoto,  M.  Sugiyama and  T.  Kikko: J. Jpn. Soc. Powder Powder Metall. 52 (2005) 164–170.
• 42)  Japanese Patent No.4060101.
• 43)  Japanese Patent No.4589374.
• 44)  Japanese Patent No.4641299.
• 45)  Japanese Patent No.4837700.
• 46)  M. Kondoh and H. Okajima: Advances in Powder Metallurgy & Particulate Materials, (2002) 3-47–3-54.
• 47)   S.  Tajima,  T.  Hattori,  S.  Hotta,  M.  Kondoh,  O.  Hiroshi,  K.  Higashiyama and  H.  Kishimoto: J. Jpn. Soc. Powder Powder Metall. 50 (2003) 577–583.
• 48)   M.  Kondoh: J. Jpn. Soc. Powder Powder Metall. 54 (2007) 506–512.
• 49)   M.  Kondoh and  N.  Matsumoto: J. Jpn. Soc. Technol. Plast. 58 (2017) 558–562.