SiFe Electrical Steel Research Articles

Deformation Structure and Recrystallization of Ferritic FeSi Steels

The magnetic properties of finally fabricated nonoriented FeSi electrical steels depend critically on microstructure and crystallographic texture components. The resulting microstructure and texture are influenced along the whole processing route of Fe–Si steels: hot rolling, coiling and cooling, hot band annealing before cold rolling (optional), cold rolling, and final thermal treatment. There is an interplay of microstructure and texture between the various processing steps, which is not investigated in any detail. Especially for ferritic FeSi steels are no specific investigations on the influence of microstructure and texture of cold‐rolled material on their evolution at final annealing. In this paper, authors summarize the recent experimental results on the evolution of microstructure and texture at final annealing for cold‐rolled material of ferritic FeSi steels with 2.4 and 3 wt% Si. The evolution of microstructure and texture is analyzed by metallographic studies and EBSD. Special attention is paid to the through‐thickness dependence of microstructure and texture, which is caused by the strain gradients as well as thermal gradients at processing. The obtained experimental results altogether demonstrate clearly the effect of different misoriented substructures after cold rolling due to nonuniform strain on recrystallization behavior for the ferritic FeSi steels.

Disordering induced work softening of Fe–6.5 wt%Si alloy during warm deformation

The application of Fe–Si electrical steel as iron cores of transformers and engines requires it to be cold rolled into thin sheet because the deformation process is a path directly reaching a final dimension with good magnetic and mechanical properties. However, it is impossible for the Fe–6.5wt%Si alloy to be directly cold rolled due to its embrittlement. One alternative way to improve its workability is warm rolling. In this work, warm deformation behavior of Fe–6.5wt%Si alloy is studied. The results show that the flow stress decreases remarkably after 30% compression, accompanying with a reduction of micro-hardness, indicating a work softening behavior of the alloy. The size of the ordered domains and the degree of order decrease gradually with deformation, and a disordered alloy is obtained finally. When deformation reaches a certain value, small dislocation cells appear and finally subgrains are formed indicating the occurrence of dynamic recovery. It is concluded that the deformation-induced disordering and dynamic recovery are the main softening mechanisms.

Electrodeposition and magnetic characterization of iron and iron-silicon alloys from the ionic liquid 1-butyl-1-methylpyrrolidinium trifluoromethylsulfonate.

The electrodeposition of soft magnetic iron and iron-silicon alloys for magnetic measurements is presented. The preparation of these materials in 1-butyl-1-methylpyrrolidinium trifluoromethylsulfonate, [Py1,4]TfO, at 100 °C with FeCl2 and FeCl2 +SiCl4 was studied by using cyclic voltammetry. Constant-potential electrolysis was carried out to deposit either Fe or FeSi, and deposits of approximately 10 μm thicknesses were obtained. By using scanning electron microscopy and X-ray diffraction, the microstructure and crystallinity of the deposits were investigated. Grain sizes in the nanometer regime (50-80 nm) were found and the presence of iron-silicon alloys was verified. Frequency-dependent magnetic polarizations, coercive forces, and power losses of some deposits were determined by using a digital hysteresis recorder. Corresponding to the small grain sizes, the coercive forces are around 950-1150 A m(-1) and the power losses were at 6000 J m(-3), which is much higher than in commercial Fe(3.2 wt %)Si electrical steel. Below a polarization of 1.8 T, the power losses are mainly caused by domain wall movements and, above 1.8 T, by rotation of magnetic moments as well as domain wall annihilation and recreation.

Differential Scanning Calorimetry and Metallographic Analysis of Fe-Si Electrical Steel

The microstructure development in cold rolled electrical steel under dynamic heat treatments was subjected to investigation. Significantly distinguish types of microstructures were obtained in the investigated steels confirming the different character of grain boundary motion. Application of annealing temperature within two phase region (austenite+ferrite) leads to abnormal grain growth in silicon steels. Moreover, in the optimum temperature range, there was a particular temperature leading to the most optimal microstructure and texture[1]. The effect of Si content on the phase transition temperature of the electrical steel (0.6, 1, 2.5, 2.9 % Si) was studied by using differential scanning calorimetry (DSC) analysis. The result indicated that DSC analysis could be used to detect the shift temperature of phase transformation in the electrical steel with different Si addition. DSC have been used in thermochemical studies and as complementary to the study of phase transformation. It can be used as a compliment to optical and electron microscopy.

On the hot working of FeSi ferritic steels

The high temperature deformation behavior of FeSi electrical steels (ferritic at the testing temperature) was studied under hot compression conditions. The flow curves of these steels exhibit a conventional dynamic recovery behavior. The analysis of the hot working data was carried out by applying the classical hyperbolic-sine equation to derive the peak stress in the flow curve by determining the apparent stress exponent, n, and the activation energy, Q. A modified hyperbolic-sine equation showed a better correlation with experimental data. The modified equation assumes that the rate controlling mechanism is the glide and climb of dislocations. Consequently, Q can be taken equal to the iron self-diffusion in ferrite (Q=239kJ/mol) and the creep exponent can be set equal to 5. The whole flow curves were also characterized by the one variable approach which in turn requires deriving characteristic dynamic recovery and work hardening terms at any processing conditions. A direct relationship between the hardening and softening terms and the Si content of the steels was obtained. Finally all the flow curves were discussed and modeled as a function of the processing parameters and chemical composition. Finally, a universal constitutive equation to describe the high temperature deformation of these steels taking account of the chemical composition is proposed.

Effect of Cold Rolling Reduction Rate on Secondary Recrystallized Texture in 3 Pct Si-Fe Steel

The phenomenon of secondary recrystallization in 3 pct Si-Fe electrical steel subjected to relatively high cold rolling reduction rates has been investigated. The texture of the secondary recrystallized sample that has a cold rolling reduction rate of 97.2 pct consists mainly of {110}〈112〉 component, which is quite different from the ideal Goss ({110}〈001〉) texture obtained after lower cold rolling reduction rates. The grain boundary character distribution (GBCD) analysis on the primary recrystallized sample with a cold rolling reduction rate of 97.2 pct indicates that the {110}〈112〉 component has the highest frequency of high energy (HE) boundary with a misorientation angle between 20 and 45 deg, whereas the Goss component in the sample subjected to lower cold rolling reduction rates has the highest frequency of HE boundary. These results indicate that the component with the highest frequency of HE boundary surrounding it after primary recrystallization has the privilege to outgrow other components during secondary recrystallization. However, the GBCD analysis for coincidence site lattice (CSL) boundary points out that the Goss component has the highest frequency of CSL boundaries in the primary recrystallized texture irrespective of the cold rolling reduction applied. These results suggest that the HE model can predict the orientation relationship between the primary and secondary recrystallized textures better than the CSL model.

The Static Characteristic Loop and the External Demagnetizing Factor

In this paper we demonstrate, that shearing is changing only one parameter of the static loop. By using the shearing factor Ns, linked to the widely used, demagnetization coefficient ND, we show the one parameter link between the static unsheared and that of the sheared saturation loop, obtained by a non-toroidal, open circuit hysteresis measurement. The paper illustrates the simple relation between open circuit loop data and measured real static saturation data. The proposed theory is illustrated by using the hyperbolic model. For experimental illustration, tests results are used, which were carried out on two closed and open toroidal samples, made of NO Fe-Si electrical steel sheet, mimicking the demagnetization effect of the open circuit VSM measurement. These are both theoretical and experimental demonstrations, that shearing only changes the inclination of the static hysteresis loop. These test results, presented here, agree very well with the calculated results, based on the proposed method.

Microstructure of hot dip coated Fe–Si steels

Hot dipping is a coating technique pre-eminently used in industry to galvanize machine parts or steel sheets for constructional applications. However, other hot dipping applications have been developed in order to have a positive effect on specific material properties. For instance, in Fe–Si electrical steels, a Si/Al rich top layer is applied and followed by diffusion annealing to increase the electrical resistivity of the material and consequently, lower the power losses. Hot dipped aluminised mild steels have been developed with increased corrosion resistance for high temperature applications by the development of a dense Al 2O 3 layer. Regardless of the type of steel coated and the intended application, after the interaction between the molten Al and the solid material, three constituents are formed: Fe 2Al 5, FeAl 3 and an Al-rich alloy. The structural morphology, which can negatively affect the wear resistance and the thermal stability, also appears to be highly dependent on the chemical composition of the base material. To study thermo-mechanical and compositional effects on the coating behavior after hot dipping, cold rolling with different reductions was performed on different Fe–Si materials. It was demonstrated that hardness differences between the layers caused crack formation inside the Fe 2Al 5 layer during subsequent deformation. The present work reports the results obtained on materials that were hot dipped in a hypo-eutectic Al + 1 wt.% Si bath. The bath was used to coat Fe–Si steel substrates with variable silicon content with dipping times ranging from 1 to 20 s. Before dipping, the samples were heated to 700 °C and subsequently immersed in the liquid bath at temperatures of 710 °C, 720 °C and 740 °C. To further evaluate the interactions between Al, Si and Fe, a diffusion annealing treatment at 1000 °C was performed. The main diffusing elements during this treatment are Al and Fe, although small variations in Si content are also observed. At a certain distance from the surface, voids were observed, which most probably can be related to the Kirkendall effect. A characterization of the formed intermetallics was performed by Scanning Electron Microscopy (SEM), Energy Dispersive Spectroscopy (EDS), X-Ray Diffraction (XRD) and Electron Backscatter diffraction (EBSD).

Total power loss density in a soft magnetic 49% Co–49% Fe–2% V-alloy

The total power loss density of a soft magnetic alloy with wt. % 49 Co–49 Fe–2 V was investigated. Measured loss data of strip material with thickness 0.1 mm were in good agreement with the statistical loss theory that is based on separation of losses into hysteresis, eddy current, and excess contributions. As a standard reference material a low-loss 2.7 wt. % Si–Fe nonoriented electrical steel (with P = 2.25 W/kg at 1.5 T and 50 Hz at thickness 0.35 mm) was also investigated. The results indicate that the amount of excess loss is distinctly higher for the SiFe electrical steel while hysteresis loss is slightly lower. As a consequence the total loss density is lower for the CoFe alloy above a limit value of frequency that is identified for material thicknesses 0.1 mm, 0.2 mm, and 0.35 mm at several induction values. This result makes the material attractive for an increasing number of applications where energy efficiency is important. Kerr-microscopic images of the magnetic domain structure of the two materials show smaller domains of the CoFe alloy as the potential reason for the difference in excess losses.

Cast iron (CI) based soft magnetic BMG Ci88.3Al2Ga1P4.35B4.35

Thermal stability, structure, and magnetic properties of bulk type Ci88.3Al2Ga1P4.35B4.35 alloy in ribbon form have been studied using differential thermal analysis, x-ray diffraction and magnetic measurements. Results reveal that crystallization peak temperature (Tx) and Curie temperature (Tc) of the as-cast alloy are respectively 513 and 370 °C. Crystallization of the specimen starts after annealing at 460 °C and α-Fe is precipitated out. Annealing at temperatures higher than 515 °C, produces apart from α-Fe, hard magnetic precipitants (Fe2B, Fe3B), which deteriorate the soft magnetic properties. Lowest coercive field – 9.8 A/m, highest saturation of induction – 1.55 Tesla and best losses – 0.42 W/kg (at 50 Hz and 0.4 kA/m) were obtained for as-cast specimen. Observed good soft magnetic properties of these low cost cast-iron based alloys suggest perspective applications of these soft magnetic alloys as an alternative to the conventional Fe-Si electrical steel and Mn-Zn ferrites.

Hysteresis loss component under non-sinusoidal flux waveforms

Separation of total energy dissipation per magnetisation cycle into a frequency-dependent dynamic component and a frequency-independent hysteresis component is a common practise in evaluating electromagnetic losses in Si–Fe electrical steel sheet. The assumed frequency-independent hysteresis component is defined by a coefficient C 0 (J/kg). In this work, the value of C 0 was determined using a linear extrapolation method and quasi-static hysteresis energy loss per cycle. The extrapolation method gave a considerable error when applied to non-sinusoidal excitation voltages (pulse width modulation and square) in a frequency range from 25 to 100 Hz. For this reason the coefficient values obtained from the quasi-static measurements at 0.01 Hz were assumed. The product of C 0 and frequency is commonly used to calculate the proportion of hysteresis loss in the total specific loss under sinusoidal magnetisation. However, since this approach is not applicable for non-sinusoidal conditions an equation is proposed, which takes into account the contribution of the higher harmonic components. The hysteresis loss component becomes higher for more distorted waveforms as compared with the value under sinusoidal excitation. Therefore, some variability of the anomalous component contribution is expected from the loss separation.

Microstructure Characterization by EBSD of Hot Rolled High-Silicon Steel

Non oriented electrical steels are soft magnetic materials used in the core of electrical motors. No preferential anisotropy of the electrical texture in the rolling plane is desired. Nowadays these special steels are mainly alloyed with Si, Al and some additives to improve the magnetic properties and to reach a good of formability. For (Si, Al)-concentrations higher than 2 wt.% the α- γ-α phase transformation is suppressed, resulting in a bcc crystalline structure from liquidus to room temperature. These electrical steels, which will be discussed in the paper, exhibit the lowest values of the magnetic losses. Hot rolling of FeSi electrical steels has been found to be one of the fundamental steps in producing these materials with optimum properties. The resulting properties, as well known, are determined by the type of magnetic textures and the structural inhomogeneities. Electron Backscattered Diffraction (EBSD) is a reliable tool for microstructural and texture characterization of different materials. Two compositions of electrical steel are studied by optical microscopy and EBSD, with special attention paid to characterize the grain morphology and its texture through thickness.

Sensitivity analysis for estimation of power losses in magnetic cores using neural network

Abstract Experimental data from a sample of 42 cores made from grain oriented 0.27 mm thick 3% SiFe electrical steel with dimensions ranging from 35 to 160 mm outer diameter, 25–100 mm inner diameter and 10–70 mm strip width and a flux density range 0.2–1.7 T have been obtained at 50 Hz and used as training data to a feed forward neural network. An analytical equation for prediction of power loss as depends on input parameters from the results of sensitivity analysis has been obtained. The calculated power losses with the analytical expression have also been compared with power loss obtained from the Preisach model after it has been applied to toroidal cores. The results show the proposed model can be used for estimation of power losses in the toroidal cores.

Ordering phenomena in Fe-Si alloys with variable Si-content (0.2-11.7 wt.%) and thermomechanical processing

Fe-Si electrical steel around 6.5 wt. % Si is an optimum material for soft magnetic applications. Nevertheless, the brittleness appearing above 3.5 wt. % Si makes the cold processing very difficult to carry out. It is generally believed that the embrittlement is caused by order/disorder phenomena. Mossbauer spectroscopy was used to investigate the ordering processes in several Fe-Si-alloys with up to 11.7 wt. % Si. Type and degree of ordering were determined and results show that the alloys are never random solid solutions and their order degree increases with the silicon content. In addition, the effect of the thermomechanical processing was analysed. It is shown that slow cooling promotes B2 ordering while fast cooling does the opposite. On the other hand, cold rolling tends to cause disordering on the alloys.

Neural Network Using Genetic Algorithm for Magnetic Performance Prediction of Toroidal Wound Cores at 50 Hz

Geometrical and building parameters have a strong influence on magnetic performance of toroidal wound cores made from grain oriented 3% SiFe electrical steel. From a sample of 40 cores with dimensions ranging from 35 to 160 mm outer diameter, 25 to 100 mm inner diameter and JO to 70 mm strip width and a flux density range of 0.1 to 1.7 T have been obtained and used as training data to a generalised feedforward neural network.

Domain structure of chemically thinned non-oriented electrical sheet

The anomaly factor tends to increase steeply when the thickness of non-oriented Si–Fe electrical steel sheets is decreased. To investigate the behavior of the anomaly factor in relation to the sheet thickness, magnetic domain observation was performed, using the Kerr effect. Grains of a thinned sheet exhibit a stripe domain pattern. The grains exhibiting the stripe domains on the surface had closure domain structure in a cross-section of the sheet. It was found that a field applied in the sheet plane induced changes of the magnetization component normal to the surface, which may increase the anomaly factor of thin sheets.

Model to evaluate the AC–DC magnetostriction characteristics of SiFe electrical steels

A simple model, describing the polarization dependence of the amplitude of AC peak-to-peak ( λ (p–p)) and zero-to-peak ( λ (0–p)), and of the DC ( λ DC) magnetostriction of electrical steels, based on some well-known mechanisms governing the domain structure of these materials is proposed. Three main features of the domain structure were considered in relation to the increase of polarization: At low polarization level (Part I of Fig. 1), a slight exponential increase of the magnetostrictive signal occurs according to the increase of volume fraction of sub-domain structure. Further on, owing to the annihilation of main 180° walls, the sub-structure is suppressed and its collapse can be described by a Dirac-type function (Part II). Almost at the same time, rotation of domains start to dominate (Part III), resulting in a sharp increase of the magnetostriction. These relationships can be formulated for the polarization dependence of the λ (p–p), the λ (0–p) and λ DC. Excellent fit of the experimental data was obtained by means of least squares, not only for different grades of grain-oriented but also for non-oriented steels, in wide range of thickness and Si content. In some cases, close relation of the fitted parameters to the materials properties were found.

無方向性珪素鋼板の厚みによる磁区構造変化

The anomaly factor that is the ratio of the measured eddy current loss to the classical calculated value tends to increase steeply when the thickness of non-oriented Si-Fe electrical steel sheets is decreased. To investigate the behavior of the anomaly factor in relation to the sheets’ thickness, magnetic domain observation was performed, using the Kerr effect. The grains of a thinned sheet exhibit a stripe domain pattern. Grains exhibiting a stripe domain on the surface had a closure domain structure in a cross section of a sheet. It was found that a field applied in the sheet plane induced a magnetization component normal to the sheet plane, which may increase the anomaly factor of thin sheets.

Effect of tensile and compressive stress on dynamic loop shapes and power losses of FeSi electrical steels

Magnetic energy losses and hysteresis loops have been determined in FeSi non-oriented laminations as a function of the applied compressive and tensile stress, made to range between − 50 and + 50 MPa. The loss separation analysis has been carried out in association with hysteresis loop calculation by means of the Dynamic Preisach Model. The strong modifications to the hysteresis loop and area introduced in a specific way by compressive stresses can be correctly accounted for by modelling.

Investigation of the limits of statistical power loss model for 3.2% SiFe electrical steels. I. Role of hysteresis loss

Loss separation experiments on non-oriented and grain-oriented 3.2 wt.% SiFe electrical steels have been used to study, according to the statistical model proposed by Bertotti, the polarisation dependence of the hysteresis loss (Phys) and the internal coercive field (Vo). Furthermore quasi-static characterisation was performed by means of a newly developed hysterisograph, improved to assure constant speed of polarisation under controlled duration of measuring period. The measured hysteresis loss on N.O.-0.35mm samples has confirmed the previous results i.e. the limits of the applicability of model (up to 1.6T in this case) and that the polarisation dependence of hysteresis can be interpreted by two different power law functions, instead of one. The polarisation dependence of Vo field, derived from excess loss behaviour, resulted to be more complex. In special cases, the analysis of quasi-static virgin magnetising curve by means of thermodynamic representation of magnetising processes confirmed the presence of different magnetising mechanisms. The good correlation between the break-points of virgin magnetising curves and the polarisation dependence of hysteresis Phys and Vo field, furthermore the strong dependence of the investigated parameters on the microstructure of material suggest to use these models to examine the changes of magnetising mechanisms, both dynamic and quasi-static characteristics, under varied conditions of microstructure.