Effect of stretching during annealing and decarbonizing treatment on the magnetic properties of grain-oriented electrical steels

ABSTRACT

The ability of the electrical steels to amplify an externally applied magnetic field is due to a strong texture induced during secondary recrystallization, the step in which grains with Goss orientation nucleate in the primary matrix. At this stage, normal grain growth is inhibited by a scattering of precipitated particles. The main objective of this research was to investigate the effect of stretching during the annealing and decarbonizing treatment on the magnetic properties of grain-oriented electrical steels with 3% Si. The study subjected grain-oriented electrical steel samples to heat treatment simultaneously to different tensile stresses. After the heat treatment, the samples were characterized by scanning electron microscopy (SEM) and transmission electron microscopy (TEM). The best results, in terms of core losses, were obtained for the sample treated without stretching. The presence and size of MnS and AlN precipitates were also found to significantly influence the magnetic properties. These results focused on the importance of controlling stretching during annealing to optimize the magnetic performance of grain-oriented electrical steels for applications in transformers and generators.

Keywords:
Electrical steels; Precipitation; Stretching; Electron microscopy

1. INTRODUCTION

Iron alloys containing silicon as their principal alloying element, commonly referred to as electrical or silicon steels, were developed to reduce magnetic losses in electrical equipment. Achieving both high energy efficiency and low production costs requires identifying the optimal chemical composition and mechanical processing methods for manufacturing these alloys.

Grain-oriented electrical (GO) steels, with their excellent magnetic properties in the rolling direction, are crucial for a variety of practical applications [¹]. These alloys are widely used in transformers, power reactors, hydrogenerators, and turbogenerators, applications that demand a material capable of easy magnetization and demagnetization [²]. In the case of a single crystal of Fe-Si, which typically adopts a body-centered cubic (BCC) structure, the [100] direction corresponds to the easy magnetization axis. To minimize energy losses in transformer cores, this axis must be aligned with the applied magnetic field [³]. To achieve this ideal alignment on macroscopic scale, the American inventor Norman P. Goss patented [⁴] in 1934 a treatment for silicon steels that produces the desired crystallographic texture, the (110)[001] orientation, now referred to as the Goss texture.

GO steels have an extremely strong crystallographic texture such that most crystals have their {110} planes parallel to the sheet surface, with the [001] direction parallel to the rolling direction [⁵]. Cold rolling is a common way to develop texture in polycrystalline metals by plastic deformation [⁶]. However, the main process for developing the Goss texture in GO steels is secondary recrystallization, which is associated with an abnormal grain growth that favors the evolution of a strong (110)[001] texture [⁴].

Goss-oriented recrystallized grains are developed during the hot rolling and hot band annealing stages [¹]. Therefore, controlling the dispersion of fine particles that precipitate at these stages plays a fundamental role in developing the final texture and magnetic properties [¹].

Precipitate formation depends on the temperature of the processing steps and on the concentration of elements that form nitrides, carbides, and sulfides [⁷]. Manganese sulfide (MnS), copper sulfide (CuS), and aluminum nitride (AlN) are the most common precipitates found in electric steels and can exhibit varying sizes, from 10 nm to 400 nm [⁷]. The precipitation of small inhibitors of the normal growth of the primary grains results in the desirable abnormal growth of the Goss grains during secondary recrystallization [⁸], leading to a higher magnetic flux density [⁹].

The main objective of this investigation was to advance the understanding of the effect of stretching during the annealing and decarbonizing heat treatment on the properties of grain-oriented electrical steels with 3% Si. For this, 3% Si steel samples were decarburized under tensile stress. The influence of stretching was investigated by correlating the characteristics of the second-phase particles observed by SEM and TEM and the evolution of texture with the magnetic properties of the samples.

2. MATERIALS AND METHODS

The material used in this study was a single type of grain-oriented electrical steel, specifically the P989A steel, produced with a low plate reheating temperature and supplied by Aperam South America. The samples analyzed were all from the same batch, which were cold rolled in a single-step reduction of 88.3 %, to 1045 × 700 × 0.27 mm. The sheets were then cut to 600 × 50 × 0.27 mm in dimensions for this study, configuring the as-received material, GO steel. Its chemical composition is presented in Table 1.

Table 1
Chemical composition of the P989A steel (wt.%).

C	Mn	S	Si	Al	Cu	Ti
0.044	0.152	0.005	3.137	0.03	0.191	0.002

The as-received material underwent annealing along with decarburizing and nitriding treatments. During the annealing, decarburizing, and nitriding treatments, a pull roller applied a tensile load to the steel strips of sample 9M, causing them to stretch. Sample 2M did not undergo stretching during these processes. The specific loads applied and their effects on samples 2M and 9M are detailed in Table 2.

Table 2
Load applied to the samples during annealing stretching.

P989A STEEL	LOAD (Kgf)	STRESS (MPa)	STRETCH (%)
2M	5	3.7	0.00
9M	35	25.9	13.2

The 2M and 9M samples were annealed and decarburized in a continuous furnace at 860 °C, with a strip speed of 1.0 m/min in a 50% H₂ atmosphere with a dew point of 63 °C. Subsequently, nitriding treatment was performed in a pilot line furnace at 840 °C with a strip velocity of 2.0 m/min in a controlled atmosphere of 75% H₂ and a dew point of 5 °C. After nitriding, the samples were coated with magnesia (magnesium oxide, MgO) and subjected to final annealing in a box furnace. Nitrogen contents, in ppm, measured after nitriding and the secondary box annealing, are shown in Table 3.

Table 3
Nitrogen content in 2M and 9M samples.

P989A STEEL	N1 (ppm)	N2 (ppm)
2M	407	0.00
9M	435	0.97

• 1 After nitriding.
• 2 After secondary annealing in a box furnace.

The 9M material was stretched by approximately 13.2% through the application of a tensile force, using a 25 MP load applied via the pull roller as the strip passed through the annealing furnace. Load control was achieved using a load cell and digital display connected to the pull roller system, with the roller speed adjustable between 0 and 5 m/min.

The magnetic tests were conducted at the Magnetic Test Laboratory of the Aperam Research Center using a Brockhaus Single Plate Meter (SST Frame 30 × 280) with frequencies of 50 Hz and 60 Hz for 1.5 T and 1.7 T magnetic inductions.

TEM images were acquired using a JEOL JEM-2010 HRTEM transmission electron microscope, operated at 200 kV with a LaB6 source and equipped with an X-ray Energy Dispersion Spectroscopy (EDS) system. A single title holder was used for sample positioning. For TEM analysis, carbon replicas were extracted to identify precipitates and analyze their average size without matrix interference, while thin films were used to identify precipitation sites. Brightfield images and corresponding EDS spectra were obtained for precipitate characterization and identification. The choice of magnification depended on the particle sizes, with lighting conditions adjusted using a condenser aperture and spot size of 2–3; the objective aperture was not used for image acquisition.

For precipitate extraction via carbon replicas, samples were cut, polished with 3 and 1 µm diamond paste, and etched with a 10% Nital solution for 15 seconds. A thin layer of approximately 20 nm amorphous carbon was then deposited using a LEICA EM ACE600 depositor. A 10% Nital solution and a 100 mesh hexagonal Cu grid were used to remove the carbon replica. ImageJ software was employed to measure the average particle size, considering the equivalent average diameter.

For thin film preparation, samples were stamped into discs with a 3 mm diameter, sanded through a sequence of 240, 320, 600, and 1200 mesh, down to a thickness of 60 µm, and polished with 3 and 1 µm diamond pastes. Final polishing to obtain the thin area was performed electrolytically using a StruersTenuPol-5 electrolytic preparation system, with a solution of 95% acetic acid and 5% perchloric acid as the electrolyte. The polishing was carried out at room temperature with a voltage of 20–25 V and a flow rate of about 40 mL/min for approximately 3 minutes.

SEM images were obtained using a Quanta FEG 250 scanning electron microscope to observe the sample microstructure and analyze the dispersion of second-phase particles. EDS was used to identify the chemical elements. The samples were cut to an approximate size of 20 × 20 mm and mounted on an acrylic resin support. They were then sanded using a sequence of 320, 600, and 1200 mesh, polished with 3 and 1 μm diamond paste, and etched with a 10% Nital solution for 15 seconds. Images were captured using a secondary electron detector with 10, 12, and 15 kV voltage, a spot size of 4.5-5, and working distances ranging from 10 to 12 mm and 18 to 24 mm.

Texture analyses were performed using a PANalytical X’PERT PRO MRD X-ray diffractometer with a cobalt anode (λ = 1.79 Å) and an iron filter. X-ray emission was conducted at 40 kV and 45 mA. The samples were prepared by sanding with 320 and 600 mesh and etched with a 95% hydrogen peroxide (H₂O₂) solution and hydrofluoric acid (HF) at room temperature for 1–2 minutes. For primary recrystallization texture analysis, three pole figures–{110}, {200}, and {211}–were measured.

The material textures were analyzed using a crystallographic orientation distribution function (ODF), calculated by PopLA software using the three aforementioned pole figures. The ODF was calculated and presented for sections φ2 = 0° and φ2 = 45° of Euler space, according to Bunge notation [¹].

3. RESULTS AND DISCUSSION

The magnetic properties were evaluated by measuring magnetic loss under the following conditions: 1.5 T/50 Hz (P15/50), 1.7 T/50 Hz (P17/50), 1.5 T/60 Hz (P15/60), and 1.7 T/60 Hz (P17/60), as well as magnetic induction at 800 A/m/60 Hz (B8), as shown in Figure 1. The 2M sample exhibited lower losses and higher magnetic induction compared to the 9M sample.

Figure 1
Magnetic loss results of the samples.

Figure 2 illustrates the hysteretic behavior of the 2M and 9M samples. The areas enclosed by the curves represent the energy density dissipated per cycle. The magnetization and demagnetization processes in alternating current lead to hysteresis, which reflects a delay between the induction and the magnetic field. The energy dissipated during this process is known as the hysteretic portion of magnetic losses [¹⁰].

Figure 2
Comparison of hysteresis curves measured at 60 Hz and 1.7 T, highlighting the energy dissipated during magnetization and demagnetization.

As a result, the sample subjected to the highest tensile stress during annealing and decarburization (9M) exhibited higher hysteresis loss, suggesting that stretching is not recommended for the production of GO steel.

The precipitates present in the as-received materials were observed by TEM, identified via EDS, and analyzed based on their morphology, in accordance with previously reported findings in the literature [6, 7]. Thus, the precipitates observed were classified as:

• CuxS: when the EDS analysis showed the elements copper and sulfur;

• CuxS+AlN: when the EDS analysis showed the elements copper, sulfur, aluminum, and nitrogen;

• MnS: when in the EDS analysis showed the elements manganese and sulfur;

• (Cu, Mn)S: when the EDS analysis showed the elements copper, manganese, and sulfur;

• (Cu, Mn)S+AlN: when the EDS analysis showed the elements copper, manganese, sulfur, aluminum, and nitrogen;

• Cu: when the EDS analysis showed only copper;

• Coprecipitates: when the EDS analysis showed the elements mentioned above and titanium.

Figure 3 presents TEM carbon replica micrographs of the as-received material. The particle with spherical morphology was identified as AlN with an average size of 105 nm; the particles with cubic morphology were identified as CuxS+AlN precipitates with sizes of 162 and 187 nm; a coprecipitate containing Cu, Al, and Ti with a size of 76 nm was also identified.

Figure 3
TEM carbon replica images of precipitates in the as-received material. The particles are identified as (a), (b) CuxS+AlN, (c) AlN, and (d) coprecipitate of Cu, Al and Ti.

CuxS+AlN particles with cubic morphology had previously been observed in electrical steels. ALCÂNTARA et al. [⁷] suggested that CuxS particles precipitate on top of AlN, causing a slight increase in this average particle size.

Figures 4 and 5 show TEM replica micrographs of as-received material, without stretching, 2M. It is possible to observe the presence of elongated and irregularly shaped particles. The elongated particles, with cubic morphology, were identified as MnS precipitates with sizes between 84 and 210 nm, measured by the equivalent diameter, using ImageJ software. The irregular particles were identified as precipitates (Cu, Mn)S, (Cu, Mn)S+AlN, and CuxS with sizes between 67 and 144 nm. A Cu precipitate with a size of 100 nm was also observed.

Figure 4
TEM carbon replica images of the precipitates in sample 2M (no stretching), highlighting (a), (b), (d) irregular (Cu, Mn)S precipitates, and (c) elongated MnS precipitates.

Figure 5
Additional TEM carbon replica images of precipitates in sample 2M, showing the presence of (a), (c) MnS, (b) CuxS, and (d) Cu precipitates.

Figures 6 and 7 present precipitates of MnS, (Cu, Mn)S, and (Cu, Mn)S+AlN obtained by replica extractions of the as-received material, with stretching, 9M. The particle identified as MnS exhibited cubic morphology with a size of 548 nm. The (Cu, Mn)S+AlN particles showed irregular morphology and size of 148 nm and 1.2 μm. (Cu, Mn)S particles had a rod shape indicating elongated cubic morphology with a size of 118 nm.

Figure 6
TEM carbon replica images of the precipitates in sample 9M (with stretching), showing (a), (b) (Cu, Mn)S+AlN, (c) coarser MnS precipitates, and (d) (Cu, Mn)S.

Figure 7
TEM carbon replica images of the precipitates in sample 9M, further illustrating the coarser precipitates resulting from the stretching.

It is essential to highlight here that the precipitates of the 2M sample, processed without stretching, were much smaller than the equivalents of 9M, annealed with stretching.

It is known that manganese sulfide (MnS), copper sulfide (CuS), and aluminum nitride (AlN) have been extensively used as grain growth inhibitors in electrical steels. Copper and manganese sulfide particles and aluminum nitride particles delay grain growth during primary recrystallization, leading to preferential growth of grains with Goss orientation during secondary recrystallization. The sulfides and aluminum nitride particles frequently precipitate during the hot rolling process. As observed in ALCÂNTARA et al. [⁷], the presence of nitride particles was consistently linked to sulfide particles, particularly CuS. These particles tended to precipitate on AlN. AlN precipitates were also found to be associated with (Cu,Mn)S [⁶]. The precipitate sizes are only slightly modified in subsequent heat treatments [¹].

The ideal size of the AlN and MnS particles is less than 100 nm because they can hinder the grain boundary movement by increasing the anchoring force and reducing the driving force for the secondary growth of the grains. Coarse particles larger than 100 nm are less effective at delaying boundary movement; they do not act as grain growth inhibitors [⁸]. Secondary recrystallization begins near the plate surface, and the grains grow by consuming the primary matrix inhibited by the precipitate AlN+MnS [¹¹].

Research on low-temperature annealing provides deeper analysis into the relation between grain growth and processing parameters. CESAR et al. [¹²] investigated the impact of annealing temperatures between 825–845 °C on the grain growth of 3% Si steel, highlighting the role of MnS particles as inhibitors. Their findings demonstrate that even at these lower temperatures, the stability and distribution of MnS particles significantly influence grain growth behavior. This sensitivity to processing conditions, even in the absence of stretching, underscores the potential for stretching to further alter the grain growth kinetics and resulting microstructure, as observed in the present study, stretching at higher temperatures led to coarser grains and larger precipitates.

In this investigation, the precipitates in the 9M sample, which was stretched by approximately 13.2%, were observed to be coarser than those in the 2M sample, as determined using the equivalent diameter via ImageJ software. Coarser precipitates, particularly those larger than 100 nm, are known to be less effective in promoting the development of Goss texture, which is crucial for enhancing the magnetic properties of grain-oriented electrical steels [¹³].

The magnetic properties of the two samples were directly measured, and it was found that the 2M sample, which was not subjected to stretching, exhibited lower magnetic losses and higher magnetic induction compared to the 9M sample. Associating these results with the TEM images of second phase particles, Figures 4–7, and the results of the magnetic tests, one concludes that the 9M sample presented higher magnetic losses and lower induction.

Analyzing this with the TEM images, it was seen aspects of the size, distribution, and morphology of the precipitates in both samples. The 9M sample showed a greater proportion of coarser precipitates compared to the 2M sample. As the literature suggests [6, 11], precipitates larger than 100 nm are less efficient in restricting grain growth during recrystallization, which can negatively influence the development of the desired Goss texture.

Thus, the coarser precipitates in the 9M sample likely hindered the formation of a robust Goss texture, contributing to its inferior magnetic performance. The correlation between the magnetic behavior and the underlying microstructure, as revealed by the TEM analysis, strongly supports that the microstructural differences, particularly in precipitate characteristics, are responsible for the variations in magnetic properties between the 2M and 9M samples. Figure 8 shows TEM images of the 2M sample. In Fig. 8(a), regions of high dislocation densities can be seen with coarse and fine precipitates in the grain boundaries.

Figure 8
TEM thin film image of sample 2M. (a) Shows regions of high dislocation densities with coarse and fine precipitates in the grain boundaries. (b) Shows the morphology and Mn,S precipitate distribution on the surface area.

Figure 8(b) shows the morphology and precipitate distribution on the surface area of sample 2M. It can be observed that the precipitates, composed of Mn and S, are distributed in the matrix and the grain contours and present round or elongated shapes. A region of precipitate clusters can also be noted.

A small volume fraction of medium and/or large precipitates in the grain boundaries may not have a significant thermodynamic effect on recrystallization [¹³]. However, a large volume fraction of small precipitates will impact the grain or sub-grain boundary mobilities [¹³]. Therefore, in this investigation, the magnetic results obtained for the 2M sample were higher than in 9M, consistent with the smaller precipitated particles observed for 2M samples. This has significant implications for the design and production of these steels.

Figure 9 presents the microstructure of the as-received material, before being submitted to the annealing and decarburization treatment. In this stage of Fe-3 % Si steel production, ferrite grains, second-phase regions (SG), and the precipitation of fine carbides can be observed. The image was obtained by SEM with a secondary electron detector and magnifications of 20000x and 15000x.

Figure 9
SEM micrographs of the: (a,b) As-received material, displaying ferritic phase, second-phase (SG) and fine carbides; and after (c) 2M, grain size: 32.78 ± 2.34, (d) 9M, grain size: 41.16 ± 5.26 μm.

Figure 10(a) shows the as-received material, obtained by scanning microscopy with an SE detector and 30000x magnification. The formation of AlN-type precipitates and coprecipitates containing Al and S can be observed. Figure 10(b) shows these elements identified via EDS.

Figure 10
(a) SEM micrograph of the as-received material, where the formation of AlN-type precipitates and coprecipitates was detected; (b) EDS of points P1 and P2.

Figure 11 shows the SEM image (a) and the EDS spectrum (b) of the 2M sample after annealing and decarburization. For this sample, which did not undergo stretching, precipitates of the aluminum, silicon, and manganese (Al, Si, Mn) N-type with cubic morphology were identified. The arrows on the micrograph indicate the precipitates (Al, Si, Mn)N measuring approximately P1 312 nm, P2 210 nm, and P3 102 nm.

Figure 11
(a) SEM micrograph of the 2M sample with (Al, Si, Mn)N precipitates; (b) EDS of points P1, P2 and P3.

The precipitation mechanism of aluminum, silicon, and manganese nitride in grain-oriented electrical steels has already been reported in the literature [¹²]. It was observed that (Al, Si, Mn)N particles precipitated with an average size of 30 nm during primary recrystallization are efficient grain growth inhibitors during secondary recrystallization [⁹].

AlN particles present in sample 9M were also observed by SEM and identified by mapping the chemical elements, as shown in Figure 12.

Figure 12
(a) SEM micrograph of the 9M sample where AlN-type precipitates can be observed; (b) EDS of the larger precipitate; (c) mapping of the chemical elements: N-red; Al-green; Fe-blue.

Coprecipitates and AlN precipitates with an elongated shape and cubic morphology were also identified in this sample, as shown in Figure 13.

Figure 13
(a) and (c) SEM micrographs of sample 9M where coprecipitates and AlN-type precipitates can be observed; (b) and (d) EDS of points P1 and P2.

The crystallographic texture of as-received material was examined in two processing steps: after cold rolling (deformation texture) and after annealing and decarburization (primary recrystallization texture).

The steel textures after cold rolling were observed from the sections of φ2 = 0 and φ2 = 45 of the ODF in Figure 14. The texture intensity was quantified by comparing the measured data to a random distribution (time random intensity). For example, an intensity of 4 would be four times greater than expected for a random or texture-free sample. The main identified texture components are the (001)[110] component, a rotated cube with an intensity greater than 6; the α fiber, (hkl)[110], where <110> is parallel to the rolling direction, also with an intensity greater than 6; and the γ fiber with the {111} plane parallel to the sheet plane, showing a peak at ($\overline{1}11$)[$0\overline{1}1$] with level 4.5. In general, after a higher degree of deformation, the tendency is to reduce the rotated cube intensities and strengthen the α fiber. In this case, the α fiber peaks stayed at the same level as the α fiber and a little higher than the γ one.

Figure 14
Sections (a) φ2 = 0° and (b) φ2 = 45° ODF of the as-received material (cold rolled with 88.3% reduction).

After annealing and decarburization without stretching (Figures 15 and 16), the steel texture presented main components of {001}<$1\overline{3}0$> with intensity 4.5; another peak near {115}<$4\overline{9}1$> with level 5; the component {223}<$3\overline{6}2$> with intensity of 4.5; and (332)[$\overline{11}3$] orientation with level 3.5. The sample subjected to concomitant stretching with the annealing and decarburization treatment showed comparable texture components with dissimilarities mainly in intensities: {001}<$1\overline{3}0$> orientation with intensity 3.5; a component near {115}<$4\overline{9}1$> at a level greater than 5.5; the peak {223}<$3\overline{6}2$> at level 5.5; and another orientation near (332)[$0\overline{2}3$] with intensity 4.5. One can observe that the stretching decreased the intensity of {001}<$1\overline{3}0$> components and increased the others. These differences in texture between the samples with or without stretching during the mentioned heat treatment associated with precipitation distributions can influence the development of secondary recrystallization texture, which is essential for the magnetic properties of the steel [¹⁴].

Figure 15
Sections (a) φ2 = 0° and (b) φ2 = 45° ODF of the 2M sample (annealed and decarburized without stretching).

Figure 16
Sections (a) φ2 = 0° and (b) φ2 = 45° ODF of the 9M sample (annealed and decarburized with 13.2% stretching).

Although the sample subjected to stretching may have had an unfavorable initial orientation, and still developed secondary recrystallization, other factors like the size and distribution of precipitates also played a critical role in facilitating it [7, 9, 11]. As the precipitates are essential in controlling grain growth during primary recrystallization, acting by pinning grain boundaries, which inhibit undesired grain growth and allow for the selective growth of grains with the Goss orientation during secondary recrystallization. This interplay between texture, precipitate characteristics (size, distribution, and type), and the development of the secondary recrystallization texture has been analyzed through TEM, SEM, and texture data from both the stretched and unstretched samples. Figures 3–13 present visual evidence of precipitate size, distribution, and morphology in both sample types. While Figures 14–16 depict texture analysis results, providing the crystallographic orientations of the samples.

The macrographs in Figure 17 show the samples with 0.0% and 13.2% stretching, respectively, after secondary recrystallization. Both samples underwent complete recrystallization, but the stretching influenced the final grain size, with the stretched sample exhibiting larger grains. As reported in other studies, magnetic loss depends on the average grain size after secondary recrystallization [¹⁵], as grain boundaries restrict domain wall movement [¹⁶]. Larger grain sizes increase domain spacing, leading to higher magnetic loss [¹⁷]. Thus, the differences in grain size between the stretched and unstretched samples highlight the role of texture, precipitate control, and grain boundary engineering in optimizing the magnetic properties of as-received material.

Figure 17
Macrographs of the secondary structure of (a) 2M and (b) 9M samples after secondary annealing, illustrating the grain size differences between stretched and unstretched samples.

Measuring the texture of secondary recrystallization is complex due to the large grain sizes, which range from millimeters to centimeters, as shown in Figure 17. However, it is well established that there is a strong correlation between deviations from the ideal Goss orientation {110}<001> and the magnetic properties. In the case of grain-oriented electrical steels, it is common to associate the magnetic properties with the texture, and the texture with the magnetic properties through the disorientation angle of the <001> direction relative to the rolling direction. Typically, a deviation of up to 7° is considered acceptable for Conventional Grain Oriented steels and 3° for High Permeability Grain-Oriented steels [¹⁸]. In this study, this deviation was indirectly estimated by correlating it with the magnetic induction at 800 A/m (B8), as established by CRAIK and MCINTYRE [¹⁹] and LITTMANN [²⁰], and presented in Figure 18. For sample 2M, which had a B8 value of 1.899 T, the average deviation angle from the ideal Goss orientation was approximately 2°. In contrast, for the sample with 13.2% elongation (9M), which exhibited a B8 value of 1.610 T, the estimated deviation angle was around 11°.

Figure 18
Relation between deviation angle from the ideal Goss {110}<001> orientation and magnetic induction at 800 A/m, based on data from CRAIK and MCINTRYE [¹⁹] and LITTMANN [²⁰].

A qualitative summary of the magnetic property measurements, precipitated particle size, and agglomeration site is presented in Tables 4 and 5. It is possible to see that a stretch of 13.2% produced a maximum loss of 3.09 W/Kg under the test conditions of 1.7 T with a frequency of 60 Hz and induction of 1.61 T, which are poor magnetic properties.

Table 4
Qualitative comparison between the magnetic results obtained from the 2M and 9M samples and the desirable benchmark.

MAGNETIC TEST	SAMPLE 2M	SAMPLE 9M	BENCHMARK
Induction	Higher	Lower	Higher
Magnetic Loss	Lower	Higher	Lower
Hysteresis	Lower	Higher	Lower

Table 5
Qualitative comparison between the electron microscopy results obtained from 2M and 9M samples and the desirable benchmark.

MICROSCOPY	SAMPLE 2M	SAMPLE 9M	BENCHMARK
Precipitates size	Lower	Higher	Lower
Particle Agglomeration	Higher	Lower	Higher

The best core loss results were obtained for the sample annealed without stretching (2M), as it showed lower loss and more significant magnetic induction compared to the 9M sample under the same test conditions. This result was a maximum loss of 1.263 W/Kg and an induction of 1.899 T.

4. CONCLUSION

The objective of this investigation was to enhance the understanding of how stretching during annealing and decarburization affects the magnetic properties of grain-oriented electrical steels with 3% Si. To achieve this, 3% Si steel samples were processed concurrently with annealing under tensile stress.

The experimental results demonstrated that applying stress during annealing negatively impacted the steel’s magnetic properties. Stretching during annealing and decarburization primarily influenced the growth of grains during primary recrystallization. Specifically, stretching led to the formation of larger precipitates (greater than 100 nm), which were less effective at pinning grain boundaries. This reduced pinning facilitated more extensive grain growth during primary recrystallization, resulting in coarser grains. These larger grains deviated from the ideal Goss orientation, increasing magnetic losses compared to the non-stretched process.

The best magnetic properties observed in the steel without stretching (2M sample) were likely due to the smaller precipitates of MnS and AlN, which are known to inhibit grain growth more effectively. This more effective pinning contributed to controlled grain growth during primary recrystallization, which may have indirectly resulted in a reduced deviation from Goss orientation in secondary recrystallization.

The average grain size in the non-stretched sample (2M) was smaller than that in the stretched sample (9M), indicating that the smaller precipitates led to a finer grain structure. The 2M sample exhibited a maximum loss of 1.263 W/kg, significantly lower than the 3.09 W/kg loss measured in the 9M sample, highlighting the negative impact of stretching on magnetic performance.

Therefore, it was observed that GO steels exhibit better magnetic properties when not subjected to stretching during processing. This emphasizes the importance of controlling precipitate size during annealing and decarburization to optimize the magnetic properties of grain-oriented electrical steels. By tailoring processing parameters to achieve a fine and uniform distribution of small precipitates, improved energy efficiency in electrical equipment can be achieved.

5. ACKNOWLEDGMENTS

The authors thank Aperam South America for providing the material and CAPES for financial support.

6. BIBLIOGRAPHY

• ^[1] JAIN, V., KUMAR, P., BAGUI, S., et al, “Comprehensive study on the through-process Goss texture evolution in Fe-3.78 wt.%Si grain oriented electrical steel”, Materials Chemistry and Physics, v. 309, pp. 128428, 2023. doi: http://doi.org/10.1016/j.matchemphys.2023.128428.
  » https://doi.org/10.1016/j.matchemphys.2023.128428.
• ^[2] BOHDAL, Ł., PATYK, R., TANDECKA, K., et al, “Influence of shear-slitting parameters on workpiece formation, cut edge quality and selected magnetic properties for grain-oriented silicon steel”, Journal of Manufacturing Processes, v. 56, n. Pt A, pp. 1007–1026, 2020. doi: http://doi.org/10.1016/j.jmapro.2020.05.049.
  » https://doi.org/10.1016/j.jmapro.2020.05.049.
• ^[3] OUYANG, G., CHEN, X., LIANG, Y., et al, “Review of Fe-6.5 wt% Si high silicon steel—A promising soft magnetic material for sub-kHz application”, Journal of Magnetism and Magnetic Materials, v. 481, pp. 234–250, 2019. doi: http://doi.org/10.1016/j.jmmm.2019.02.089.
  » https://doi.org/10.1016/j.jmmm.2019.02.089.
• ^[4] GOSS, N.P., Electrical Sheet and Method for its Manufacture and Test U.S. Patent 1965559, 1934.
• ^[5] SHINGAKI, Y., TAKASHIMA, M., HAYAKAWA, Y., “Influence of initial crystal orientation and carbon content on rolling-induced texture in 3 Mass% Si steel”, ISIJ International, v. 62, n. 11, pp. 2382–2388, 2022. doi: http://doi.org/10.2355/isijinternational.ISIJINT-2022-194.
  » https://doi.org/10.2355/isijinternational.ISIJINT-2022-194.
• ^[6] JAIN, V., MODAK, P., PATRA, S., et al, “Origin of Goss texture in grain oriented electrical steel: role of shear bands”, Materialia, v. 22, pp. 101398, 2022. doi: http://doi.org/10.1016/j.mtla.2022.101398.
  » https://doi.org/10.1016/j.mtla.2022.101398.
• ^[7] ALCÂNTARA, F., BARBOSA, R., CUNHA, M., “Study of aluminum nitride precipitation in Fe-3%Si steel”, Materials Research, v. 16, n. 5, pp. 1039–1044, 2013. doi: http://doi.org/10.1590/S1516-14392013005000090.
  » https://doi.org/10.1590/S1516-14392013005000090.
• ^[8] CASTRUITA ÁVILA, L.G., GARCIA PASTOR, F., CASTRO ROMÁN, M.D., “Evolution of texture and microstructure during thermo-mechanical processing in ultrathin low carbon steels”, Matéria (Rio de Janeiro), v. 20, n. 3, pp. 714–721, 2015. doi: http://doi.org/10.1590/S1517-707620150003.0075.
  » https://doi.org/10.1590/S1517-707620150003.0075.
• ^[9] GAO, Y., XU, G., GUO, X., et al, “Primary recrystallization characteristics and magnetic properties improvement of high permeability grain-oriented silicon steel by trace Cr addition”, Journal of Magnetism and Magnetic Materials, v. 507, pp. 166849, 2020. doi: http://doi.org/10.1016/j.jmmm.2020.166849.
  » https://doi.org/10.1016/j.jmmm.2020.166849.
• ^[10] BIRČÁKOVÁ, Z., KOLLÁR, P., FÜZER, J., et al, “Energy loss and hysteresis of reversible magnetization processes in iron-based soft magnetic composites”, Journal of Magnetism and Magnetic Materials, v. 587, pp. 171291, 2023. doi: http://doi.org/10.1016/j.jmmm.2023.171291.
  » https://doi.org/10.1016/j.jmmm.2023.171291.
• ^[11] BERNIER, N., XHOFFER, C., VAN DE PUTTE, T., et al, “Structure analysis of aluminum silicon manganese nitride precipitates formed in grain oriented electrical steels”, Materials Characterization, v. 86, pp. 116–126, 2013. doi: http://doi.org/10.1016/j.matchar.2013.09.014.
  » https://doi.org/10.1016/j.matchar.2013.09.014.
• ^[12] CESAR, M.D., PAOLINELLI, S.D., ALCÂNTARA, F.L., et al, “Effect of the low temperature annealing on primary and secondary structures and magnetic properties of Fe-3% Si”, Materials Research, v. 15, n. 1, pp. 1–8, 2012. doi: http://doi.org/10.1590/S1516-14392011005000103.
  » https://doi.org/10.1590/S1516-14392011005000103.
• ^[13] JENKINS, K., LINDENMO, M., “Precipitates in electrical steels”, Journal of Magnetism and Magnetic Materials, v. 320, n. 20, pp. 2423–2429, 2008. doi: http://doi.org/10.1016/j.jmmm.2008.03.062.
  » https://doi.org/10.1016/j.jmmm.2008.03.062.
• ^[14] GUBERNATOROV, V.V., SOLOVEI, V.D., GEVRAS’EVA, I.V., et al, “Effect of tensile deformation rate on texture formation during subsequent rolling and recrystallization of electric steel”, The Physics of Metals and Metallography, v. 113, n. 11, pp. 1024–1028, 2012. doi: http://doi.org/10.1134/S0031918X12110099.
  » https://doi.org/10.1134/S0031918X12110099.
• ^[15] ALMEIDA, A.A., RODRIGUES-JR, D.L., PERASSA, L.S.P., et al, “Anomalous loss hysteresis loop”, Materials Research, v. 17, n. 2, pp. 494–497, 2014. doi: http://doi.org/10.1590/S1516-14392014005000020.
  » https://doi.org/10.1590/S1516-14392014005000020.
• ^[16] NADOUM, A., ROBINSON, F., BIROSCA, S., “On the correlation between magnetic domain and crystallographic grain orientation in grain oriented electrical steels”, Journal of Magnetism and Magnetic Materials, v. 494, pp. 165772, 2020. doi: http://doi.org/10.1016/j.jmmm.2019.165772.
  » https://doi.org/10.1016/j.jmmm.2019.165772.
• ^[17] CARDOSO, A.V.M., PAOLINELLI, S.C., SILVEIRA, C.C., et al, “Secondary recrystallization characteristics of 3%Si grain-oriented electrical steel”, Journal of Materials Research and Technology, v. 8, n. 1, pp. 217–221, 2019. doi: http://doi.org/10.1016/j.jmrt.2017.12.002.
  » https://doi.org/10.1016/j.jmrt.2017.12.002.
• ^[18] FROMMERT, M., ZOBRIST, C., LAHN, L., et al, “Texture measurement of grain-oriented electrical steels after secondary recrystallization”, Journal of Magnetism and Magnetic Materials, v. 320, n. 20, pp. e657–e660, 2008. doi: http://doi.org/10.1016/j.jmmm.2008.04.102.
  » https://doi.org/10.1016/j.jmmm.2008.04.102.
• ^[19] CRAIK, D.J., MCINTYRE, D.A., “Magnetostatic effects in grain-oriented silicon iron”, IEEE Transactions on Magnetics, v. 5, n. 3, pp. 378–383, 1969. doi: http://doi.org/10.1109/TMAG.1969.1066460.
  » https://doi.org/10.1109/TMAG.1969.1066460.
• ^[20] LITTMANN, M.F., “Iron and silicon-iron alloys”, IEEE Transactions on Magnetics, v. 7, n. 1, pp. 48–60, 1971. doi: http://doi.org/10.1109/TMAG.1971.1066998.
  » https://doi.org/10.1109/TMAG.1971.1066998.

Publication Dates

• Publication in this collection
  21 Mar 2025
• Date of issue
  2025

History

• Received
  29 Oct 2024
• Accepted
  05 Jan 2025