Effect of Al content on magnetic properties of Fe-Al Non-oriented electrical steel

Effect of Al content on magnetic properties of Fe-Al Non-oriented electrical steel

Accepted Manuscript Effect of Al content on Magnetic Properties of Fe-Al Non-oriented Electrical Steel Jaewan Hong, Hyunseo Choi, Seil Lee, Jae Kwan K...

984KB Sizes 2 Downloads 18 Views

Accepted Manuscript Effect of Al content on Magnetic Properties of Fe-Al Non-oriented Electrical Steel Jaewan Hong, Hyunseo Choi, Seil Lee, Jae Kwan Kim, Yangmo Koo PII: DOI: Reference:

S0304-8853(16)32743-3 http://dx.doi.org/10.1016/j.jmmm.2017.03.082 MAGMA 62711

To appear in:

Journal of Magnetism and Magnetic Materials

Received Date: Revised Date: Accepted Date:

24 October 2016 7 December 2016 6 March 2017

Please cite this article as: J. Hong, H. Choi, S. Lee, J.K. Kim, Y. Koo, Effect of Al content on Magnetic Properties of Fe-Al Non-oriented Electrical Steel, Journal of Magnetism and Magnetic Materials (2017), doi: http://dx.doi.org/ 10.1016/j.jmmm.2017.03.082

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Effect of Al content on Magnetic Properties of Fe-Al Non-oriented Electrical Steel Jaewan Honga, Hyunseo Choia, Seil Leeb, Jae Kwan Kima, Yangmo Kooa*1 a) Graduate Institute of Ferrous Technology, Pohang University of Science and Technology (POSTECH), Cheongam-ro 77, Nam-gu, Pohang, Gyeongbuk, Korea 37673 b) POSCO Pohang Technical Laboratories Steel Product Ⅱ Research Group, Donghaean-ro 6261, Namgu, Pohang, Gyeongbuk, Korea 37673

Abstract The effects of Al content (0.53 ≤ Al ≤ 9.65 wt%) on microstructure, texture, magnetic flux density, permeability, core loss and magnetic domain structure of Fe-Al based electrical steel were measured or observed. Average grain size decreased as Al content increased, but Al contents had no severe effects on texture. Magnetic flux density and permeability tends to decrease as Al content increased. Total core loss Ptot was separated into hysteresis loss Ph, eddy-current loss Pe and anomalous loss Pa. As Al increased, Ph increased, but Pe and Pa decreased, so the optimal grain size increased. To reduce core loss of electrical steel with high resistivity, annealing should be conducted at high temperature and for a long time to increase grain size.

Keywords: Electrical steel; Al composition; Magnetic induction; Core loss; Loss separation

1.

Introduction

Electrical steel is one of the most important soft magnetic materials for rotating machinery and transformers; it is classified as either non-oriented (NO) or grain-oriented (GO) electrical steel. Research on improving NO *

Corresponding author. Tel.: +82 54 279 9019; fax: +82 54 279 9299. E-mail address: [email protected] (Yang Mo Koo) 1

electrical steels focusses especially on obtaining high magnetic flux density and low core losses. High magnetic flux density can be achieved by manipulating crystallographic orientation to easily-magnetized direction such as <001>{0uv} by addition of segregation element [1-4], cross-rolling [5], α →γ→α phase transformation [6,7] and some other methods. Methods to reduce core loss include increasing the content of high-resistivity alloying elements such as silicon (Si) and aluminum (Al), optimizing grain size [8], and reducing thickness [9]. The content of Si and Al in electrical steel composition is one of the most important topics in research on electrical steels. Addition of Si or Al or other high-resistivity alloying elements increases the resistivity of the steel, so core loss can dramatically decrease. However, Si and Al addition reduce the cold workability of conventional NO electrical steel, so its greatest Si content is limited to 3.5 wt%, and its greatest total Si + Al content is limited to 4.5 wt% [10]. To overcome these limits, methods such as spray-forming [11], warm-rolling [12], diffusionannealing-induced Si/Al addition [13-15] and strip casting method [16, 17] have been introduced. Although studies for increasing high resistivity alloying element have been well established, the pure effects of Al and Si on the magnetic properties have not been well studied. This paper presents the effects of Al content on texture development, grain growth, magnetic flux density, permeability and total core loss. Based on measured properties, optimal grain sizes for Al contents were calculated. The results provide guidelines for designing new electrical steels.

2.

Experimental

Fe-Al binary phase electrical steel ingots of 30 kg containing 0.53 ≤ Al ≤ 9.65 wt. % were prepared by vacuum melting. The as-cast ingots were reheated at 1100 °C for 2 h for homogenization, then abruptly hotrolled to a thickness of 2.0 mm. Hot bands were annealed at 1000 °C for 5 min until elongated grains were not observed and grain structure on the surface and at the center became homogeneous at all Al contents. The oxidized layers that formed during hot rolling and hot-band annealing were removed using 35 wt% HCl etchant, then the samples were cold-rolled to a thickness of 0.35 mm in a laboratory mill. Cold-rolled sheets were annealed at tube furnace of radius 100 mm with heating rate of 30 °C/s. Annealing of the sheets was conducted at 1000 °C for 1 min, 3 min, 5 min, 10 min, 30 min or 60 min in H2 atmosphere to induce recrystallization and grain growth and cooled automatically at 55°C/s by using a pull-in and pull-out system. The final sheets were 2

prepared by annealing at 750 °C for 2 h in Ar atmosphere to relieve stress. The chemical compositions (Table 1) of the cold-rolled sheets were obtained using wet chemical analysis. Specimens of annealed hot band and final sheets were observed using an optical microscope on the cross section parallel to rolling direction, then mechanically polished and etched using 2 wt% nital at an elevated temperature. Grain size of the final sheets was measured using the intercept method on observations an area with dimensions 350 µm × 3 mm. Textures of cold-rolled sheets and final sheets were identified using CoKα radiation on a Bruker D8-discover X-ray diffractometer equipped with a texture goniometer. Orientation distribution functions (ODFs) were calculated from pole figures obtained using X-ray diffraction (XRD). From these data, textures of final sheets were evaluated using the texture coefficient  =

∑   ,

∑    ,

(1)

,

where Nhkl, is the multiplicity factor of (hkl) planes, Ihkl is the X-ray intensity of the (hkl) plane, and IR,hkl is the X-ray intensity of randomly-oriented grains; i.e., Phkl represents the ratio of the surface area covered by the (hkl) plane to the surface area covered by randomly-oriented grains in a sample [18]. Magnetic properties were measured by using a Brockhaus Instrument Model MPG 100, coil system of 50 mm × 50mm single strip tester. Measured magnetic flux density and core loss of longitudinal and transverse directions were averaged. Magnetic domain structure of the final sheets with Al contents of 1.85 wt% and 6.54 wt% were measured by magneto-optic Kerr microscopy using a ZEISS Axio Imager D1m optical microscope. For observation of magnetic domains, specimens were mechanically polished, then electro-polished to remove the stress that was induced by mechanical polishing.

3.

Results and Discussions

3.1. Microstructure, Grain size and Texture Because grain boundaries interrupt motion of magnetic domain walls, the microstructure of the final sheet should be investigated. [19] Microstructure and grain size of the sheets before stress relief annealing (SRA) was 3

also measured. However as recrystallization temperature was 250 °C higher than SRA temperature, its microstructure change or further grain growth were not observed. Microstructures of Fe-Al specimens with TD section (Figure 1) were uniform from the surface to the center at annealing time of 5 mins, except Al 6.54 wt% and Al 9.65 wt%. At high Al contents, grains at the surface tends to smaller than the grains at the center of the sheets. That might be due to strong surface pinning effect [20] at high Al contents. At all Al contents, measured average grain size of final sheets (Figure 2) was ~ 25 µm at 1 min of annealing. Based on grain growth kinetics, grain growth depends on annealing time t as t1/n [21]. In Fe-Al composition, grain growth exponent n changes from 45.5 at Al 0.53 wt% to 19.4 at Al 9.65 wt%. Therefore, as t increases, grain growth tend to decrease at high Al contents, and finally at 60 min annealing, reached 80 µm for Al contents of 1.85 wt% and 50 µm for Al contents of 9.65wt%. In case of Al contents of 0.53 wt% and 1.85 wt%, their carbon and nitrogen composition is quite higher than other compositions. The elements are well-known for suppressing the grain growth in electrical steel by precipitation pinning. But at the given annealing condition of hot bands and cold-rolled sheets, carbon and nitrogen were kept in solid solution before cold rolling. Thus their effects were quite limited on grain growth as in figure 1 and 2. Because grain size significantly affects magnetic properties, final sheets with 25 µm and 50 µm average grain size were chosen for texture measurement, magnetic property measurement, magnetic domain observation and loss analysis. However, at Al contents of 0.53wt%, the composition undergoes α →γ→α phase transformation at the annealing composition [22]; the existence of this loop means that the composition is not stable single-phase ferrite at the annealing condition; therefore the texture and magnetic properties analysis and comparison at this composition were not conducted in this study. The area fractions of {110}, {200}, {111} planes were calculated by Phkl XRD measurements of final sheets (figure 3). High-index planes such as {310} and {321} were not described. The {110} and {200} planes, which have strong positive effects on magnetic properties, each composed only ~ 5% of the area tested. The {111} plane, which has negative effects on magnetic properties, occupied ~35% of the total area. As t increased, the intensity of the γ-fiber increased, and the proportional {110} and {200} texture decreased. The difference between the fractions of each texture was less than 5% at similar grain size, but such texture change at similar grain size does not have any significant effects on change in magnetic flux density or core loss. This assertion is 4

supported by analysis of the texture parameter [23]. The texture parameters of Fe-Al final sheets ranged from 36.3 to 38.2; considering that the average texture parameter of 100% cubic texture is 22.5, such texture difference will not show any large difference on magnetic properties. The texture difference decreased when average grain size increased to 50 µm, so magnetic flux density might not be affected by texture difference caused by Al contents in the final sheet. The relationship between texture change and magnetic flux density is discussed later.

3.2. Magnetic flux density and Permeability Magnetic flux densities of final sheets with d = 25 µm and d = 50 µm were measured under 5000 mA/m of ) (B50). Using chemical compositions of the final sheets, the saturation magnetic flux densities magnetic field ( (Bs) were calculated as [24]  ) = 2.20 − 0.049 ∙ ! #$%.

(2)

B50 decreased as Al content increased (Figure 4a), but B50/Bs was almost constant at ~0.84 (figure 4b). Such remarkable decrease in magnetic flux density at higher Al contents is due to decreased saturation magnetic flux density at higher Al contents. Because texture is one of most important factors that affects magnetic flux density, these results agree well with texture results (figure 3) and a previous result [18]. Thus Al contents does not have any significant effects on the magnetic flux density of final sheets, except decreased magnetic flux density induced by reduced saturation magnetic flux density at higher Al contents. Thus, without any texture manipulation, it is hard to expect high magnetic flux density at high Al composition.  of 50 mA/m to 7500 mA/m. To determine maximum permeability (µmax), final sheets were tested under   required to induce µmax was constant until Al contents of 4.68 wt% µmax decreased as Al contents increased, but  (figure 5). Decrease of µmax at high Al contents is due to decrease in saturation induction and permeability by  required to induce adding Si or Al to reduce core loss causes [25]. When Al contents higher than 4.68 wt%,  µmax increased and µmax drastically decreased.

5

3.3. Core loss 3.3.1. Loss separation method Theoretically, Ptot steel can be assigned to three effects: hysteresis loss Ph , eddy-current loss Pe and anomalous Pa loss: &'& =  + ) + * .

(3)

Ph is affected by the area of hysteresis, grain size, and frequency [26]. Pe in a thin sheet is described by classical eddy-current loss [27]. Pa is the unexplained loss that remains after subtracting Ph and Pe from Ptot [8]. Based on precedent study, each loss can be described using grain size d, thickness of the sheet t, working frequency f, resistivity ρ, density D and some other factors as +

 = ∮ ./, , ) =

1 01 2345 & 1 +1

* = 9:

678

(4)

.

(5)

√, < < $ =*> ? @/< , 7

(6)

Among the above equations, f is common factor in these equations, so Ptot can be described as a function of f: &'& = B ? + B) ? < + B* [email protected]/< ,

(7)

where the kh, ke and ka is hysteresis, anomalous and eddy-current loss parameter for each. Because Pe is the only factor that is affected by geometry or chemical composition of specimens, it can determined without further calculation. After excluding ke by calculation, kh, and ka can be estimated by regression vs. f. Therefore, core losses of the materials were measured at 20 Hz ≤ f ≤ 200 Hz while inducing magnetic flux density (J) of 1.5 T for loss separation.

3.3.2. Hysteresis and Anomalous loss in Fe-Al electrical steel

6

Pe was subtracted from measured Ptot, then kh (Figure 6) and ka (Figure 7) were determined. Increase of Al content in the final sheet increased kh (Figure 6a); this result can be understood by magnetic anisotropy K1, magnetostriction λ100, and the relationship between magnetic domain wall energy and domain wall thickness [28]. The magnetic domain size is mainly determined by the balance between magneto-static energy and magnetic domain wall energy. At here the magnetic domain wall energy is proportional to the square root of magnetic anisotropy constant K1. In Fe-Al alloy, K1 decreases and λ100 increases as Al content increases [29, 30]. For example, low Al alloys have high K1, their domain walls are thin and have large energy. Thus the reduced domain wall energy causes domain wall area to increase as Al contents increase in order to reduce magnetostatic energy. Besides, as domain wall area increases, the domain structure becomes more complex with 180o and 90o walls. Considering higher Al content causes lower K1 and higher magneto-static energy, increase in Al content causes complex magnetic domain structure as demonstrated in figure 6 (b) and (c), which is magnetic domain structure of demagnetized stated of Fe-Al 1.85 wt% and 6.54 wt%. Since {100} in BCC iron have cubic anisotropy on the surface, degree of freedom in domain structures increases as the domain size decreases, thus hysteresis loss, kh, increases. This sensitivity of kh to Al content increases as average grain size increases, because the difference between the domain sizes of low Al content and high Al content increases as average grain size increases. This result is well accordance with previous report, although Al contents increased by PVD method [15]. Change in ka can be understood by examining equation (6). Equation (6) indicates that Pa increases as grain size increases, therefore ka became higher when grain size became 50 µm. As Al content increased, resistivity increased, so ka decreased (figure 7). However, the 90°-domain structure in the magnetic domain increased as Al content increased (figure 6b, c), so ka could not decrease further. Till now, for improving eddy-current and anomalous loss, increasing the composition of high resistivity alloying element to the electrical steel was one of the important issues. However, magneto-static energy and anisotropy constant change causes the complexity, the increased degree of freedom and the increased domain wall area in magnetic domain structure of the final sheets [29, 30]. Therefore the material shows limited Al effects on anomalous loss than theoretically expected at high Al contents as in figure 7.

7

3.4. Optimal grain sizes for minimizing core loss According to Campos et al., a mathematical model for optimal grain size dsOp in electrical steel is: F7


.CD = E21GH& 1+I/1J

,

(8)

which requires experimentally-determined parameters such as c [8]. Although estimations of optimal grain size in this section did not use the above parameters directly, calculated core loss contains the parameter in the form of kh, ke and ka . Based on kh, ke and ka, Ptot at 25 µm grain size at 50 Hz working frequency and inducing magnetic field 1.5 T, Ptot was calculated using equation (8) for 10 ≤ d ≤ 150 µm. Experimental data and calculated Ptot matched well (figure 8a). Optimal grain size to minimize Ptot at different Al content were obtained from calculated Ptot. As Al content increased, optimal grain size to minimize core loss increased (figure 8b). This tendency is due to increased resistivity at high Al content, and agrees well with Campos’ model [8]. A similar trend in electrical steel with different electrical steel is observed in Fe-Si electrical steel, in which increase in Si content increases grain size that minimizes core loss [31]. Considering these results of optimal grain size change, to reduce Ptot of electrical steel with high resistivity, annealing should be conducted at high temperature and for a long time to increase average grain size.

4.

Conclusions

Microstructure, texture and magnetic flux density, permeability (µmax) and total core loss Ptot were quantified in Fe-Al based electrical steels with 0.53 wt% ≤ Al ≤ 9.65 wt%. Total core loss was separated into hysteresis loss, eddy-current loss and anomalous loss and based on three major losses, optimal grain sizes for Fe-Al based electrical steel was calculated. The exponent that determines grain growth decreased as Al content increased, and the change suppressed grain growth at high Al content. However, Al content and annealing time did not cause drastic texture differences that can affect magnetic induction. 8

Higher Al content causes reduced magnetic flux density (B50), but B50/Bs was constant at ~0.84 for all Al content, so such decrease in B50 is due to the reduced saturation magnetic (Bs) flux density at high Al contents. Maximum permeability also decreased as Al content increased. At Al content over 6.54 wt%, maximum permeability appeared under higher external magnetic field than at lower Al content. Ptot of final sheets was decomposed into hysteresis loss Ph, eddy-current loss Pe, and anomalous loss Pa . As Al content increased Ph increased due to complexity of magnetic domain structure, but Pa decreased due to increase in resistivity. However, the complexity of magnetic domain structure at high Al content affects Pa, so it did not follow resistivity change exactly. Based on loss parameters, Ptot was calculated; it agree well with experimentally measured Ptot. According to calculated core loss, increasing Al content allows increase in the grain size that minimizes Ptot.

Acknowledgments The authors thank Prof. Jae Sang Lee and Prof. Se Kyun Kwon for their review and comments on this research. The authors also gratefully acknowledge the facilities and financial support provided by POSCO for experimental work. This work was supported by POSCO.

References [1] I. Tanaka, H. Yashiki, J. Magn. Magn. Mater. 304 (2006) e611. [2] W.S. Ko, J.Y. Park, J.Y. Byun, J.K. Lee, N.J. Kim, B.J. Lee, Scr. Mater. 68 (2013) 329. [3] N.H. Heo, Scr. Mater. 51 (2004) 339. [4] M. Godec, M. Jenko, H.J. Grabke, R. Mast, ISIJ Int. 39 (1999) 742. [5] S. Taguchi, A. Sakakura, Kinzoky Butsuri (in Japanese), 7 (1968) 221. [6] O. Hashimoto, S. Satoh, T. Tanaka, Trans. ISIJ, 23 (1983) 1028. [7] T. Tomida, T. Tanaka, ISIJ Int. 35 (1995), 538. [8] M.F. de Campos, J.C. Teixeira, F.J.G. Landgraf, J. Magn. Magn. Mater. 301 (2006) 94. 9

[9] M. Komatsubara, K. Sadahiro, O. Kondo, T. Takamiya, A. Honda, J. Magn. Magn. Mater. 242-245 (2002) 212. [10] K. Narita, N. Teshima, Y. Mori, M. Enokizono, IEEE Trans. Magn. 17 (1981) 2857. [11] C. Bolfarini, M.C.A. Silva, A.M. Jorge Jr., C.S. Kiminami, W.J. Botta, J. Magn. Magn. Mater. 320 (2008) e653. [12] K.N. Kim, L.M. Pan, J.P. Lin, Y.L. Wang, Z. Lin, G.L. Chen, J. Magn. Magn. Mater. 277 (2004) 331. [13] Y. Takada, M. Abe, S. Masuda, J. Inagaki, J. Appl. Phys. 64 (1988) 5367. [14] J. Barros, T. Ros-Yanez, J. Schneider, O. Fischer, Y. Houbaert, K. Magn. Magn. Mater. 304 (2006) e614. [15] S.M. Park, J.S. Lee, J.S. Kim, K.S. Han, Y.M. Koo, Met. Mater. Int. 3 (2015) 593. [16] H.T. Liu, Z.Y. Liu, Y. Sun, F. Gao, G.D. Wang, Mater. Lett. 91 (2013) 150. [17] H.T. Liu, Z.Y. Liu, Y.Q. Qiu, Y.Sun, G.D. Wang, J. Mater. Process. Tech. 212 (2012) 1941-1945. [18] R.M.S.B. Horta, W.T. Roberts, D.V. Wilson, Trans. Met. Soc. AIME 245 (1969) 2535. [19] S. Lee, B.C.D. Cooman, ISIJ Int. 52 (2012) 1162. [20] A.H. King, Scripta Mater. 62 (2010) 889. [21] J.E. Burke, D. Turnbull, Progress in Metal Physics, Vol. 3, Pergamon Press, London (1952) 220. [22] M. Palm, Intermetallics 13 (2005) 1286. [23] L. Kestens, S. Jacobs, Texture Stress Microstruct. 173083 (2008) 1. [24] J.B. Lorenzo, Production of High Si-steel by hot dipping and diffusion annealing, Ph.D. thesis Ghent University, 2006. [25] A. Honda, B. Fukuda, I. Ohyama, Y. Mine, J. Mater. Eng. 12 (1990) 41. [26] T.D. Yensen, N.A. Ziegler, Trans. ASM 23 (1935) 536. [27] J.J. Thomson, Electrician 28 (1892) 599. [28] B.D. Culity, Introduction to Magnetic Materials, Addison-Wesley, Reading, Ma, 1972. [29] S. Chikazumi, K. Suzuku, H. Iwata, J. Phys. Soc. Japan 15 (1960) 250. [30] R.C. Hall, J. Appl. Phys. 28 (1957) 711. [31] H. Shimanaka, Y. Ito, K. Matsumura, B. Fukuda, J. Magn. Magn. Mater. 26 (1982) 57.

10

Table 1. Chemical composition of steels. Al

Si

C

S

N

P

Ti

Fe

(wt. %)

(wt. %)

(ppm)

(ppm)

(ppm)

(ppm)

(ppm)

(wt. %)

0.53

0.0028

280

15

140

<10

<10

bal.

1.85

0.0043

220

16

11

<10

<10

bal.

2.64

0.014

20

10

2

<10

<10

bal.

4.68

0.011

44

9

1

<10

<10

bal.

6.54

0.0069

17

9

22

<10

<10

bal.

9.65

0.0040

20

9

1

<10

<10

bal.

11

RD

(a)

(b)

(c)

(d)

(e)

(f)

ND

100µm Figure 1. TD section microstructure of Fe-Al specimens (a) 0.53 wt. % (b) 1.86 wt. % (c) 2.64 wt. % (d) 4.68 wt. % (e) 6.54 wt. % (f) 9.65 wt. % with annealing time of 5 min.

12

160

0.53wt.% 1.85wt.% 2.64wt.% 4.68wt.% 6.54wt.% 9.65wt.%

140

Grain Size (µm)

120

n=45.5

100

n=29.8 n=28.1 n=29.0 n=24.2 n=19.4

80

60

40

20 1

10

100

log(t) min Figure 2. Average grain sizes of Fe-Al final sheets with different annealing times t at 1000 °C. The average grain sizes were measured by the interception method of TD sections of final sheets.

13

60

(110) (200) (222) (110) (200) (222)

55 50

Area Fraction (%)

45 40 35 30 25 10 5 0 2

4

6

8

10

Al Contents (wt.%) Figure 3. (110), (200) and (222) surface area fraction of Fe-Al specimens with grain size of 25 µm (■, ●, ▲) and 50 um (□, ○, △). The surface area fraction was estimated from the texture parameter Phkl achieved by XRD.

14

1.8

1.00

B50 B50/Bs

0.95

1.7

1.6

0.85

B50/Bs

B50 (T)

0.90

0.80 1.5 0.75

1.4

0.70 2

4

6

8

10

Al Contents (wt.%) Figure 4. Magnetic induction B50 and ratio of saturation induction Bs to B50 in Fe-Al specimen with 50 µm grain size.

15

8000

1.85wt% 2.64wt% 4.68wt% 6.54wt% 9.65wt%

7000

Permeability

6000 5000 4000 3000 2000 1000 0 100

1000

logH (A/m) Figure 5. Permeability change of Fe-Al electrical steels of grain sizes of 50 µm by external AC magnetic field.

16

Figure 6. (a) Hysteresis loss parameter at J = 1.5 T and f = 50 Hz with grain sizes of 25 µm and 50 µm. (b) Magnetic domain structure of Al content 1.85 wt. % and (c) 6.54 wt. % at demagnetized state.

17

0.00025

Anomalous loss parameter (25µm) Anomalous loss parameter (50µm) Eddy-current loss parameter

0.0035

0.00020

0.0030

0.00015

0.0025

0.00010

0.0020

0.00005

0.0015

Eddy-current loss parameter

Anomalous loss parameter (ka)

0.0040

0.00000

2

4

6

8

10

Al contents (wt.%) Figure 7. Anomalous loss parameter of Fe-Al specimens with grain size of 25 µm and 50 µm and eddy-current loss parameter.

18

(a)

20

Calculated core loss Experimental data

Core loss (W/kg)

15

10

5

0 0

50

100

150

Grain size (µm)

(b)

140

Optimal grain size (µm)

120

100

80

60

40 2

4

6

8

10

Al contents (wt%) Figure 8. Calculated core loss under by using each loss parameter and experimental core loss of Al 1.85 wt% (a)

19

and optimal grain sizes for core loss as a function of Al content (b). Experimental data and conditions for calculation is average grain size = 25 µm, J = 1.5 T and f = 50 Hz.

20

Research Highlights

► Based on Fe-Al based electrical steel, microstructure, texture, magnetic properties and magnetic domain structure were studied. ► According to Al content, grain size tends to decrease at elevated Al content, but there was no significant effects on texture. ► As Al content increase, the steel sheets had decreased magnetic flux density and permeability tends to decrease. ► The Al contents has negative effects on hysteresis loss while eddy-current and anomalous loss was improved. ► Based on the hysteresis, eddy-current and anomalous loss, optimal grain sizes for Fe-Al based electrical steel was calculated.

21