Description

[0001] This invention was made under a CRADA NFE-07-00881 between Buswell Energy LLC, and Bekaert Corporation / NV Bekaert SA and UT-Battelle, LLC operating and management Contractor of the Oak Ridge National Laboratory for the United States Department of Energy. The Government has certain rights in this invention.

Technical Field

[0002] The invention relates to a steel wire adapted for a magnetic flux path such as the core of a transformer. The invention also relates to a magnetic flux path comprising such a steel wire and to a method of manufacturing such a steel wire.

Background Art

[0003] In 1831 Faraday made the important discovery of magnetic induction. An embodiment, made by Faraday himself, shows an iron ring where two separate windings or coils of wire were wound around. This embodiment can be considered as one of the first transformers ever made. The first coil of wire forms the primary coil and was brought in connection with a battery. Upon closing the windings of the primary coil a magnetic flux was induced in the iron core. This magnetic flux induced an electrical current in the windings of the secondary coil.

[0004] The function of the core of the transformer is to transfer power between primary and secondary windings by acting as a flux magnifier. The extent to which a transformer core is capable of magnifying the flux will be determined amongst others by the losses. There are two main types of losses, namely the eddy current losses and the hysteresis losses. Eddy currents are generated when the magnetic flux in the core material changes with time, generating an electromotive force in the plane at right angles to the direction in which the flux is changing. The eddy current losses are proportional to the square of the frequency. The hysteresis losses are a consequence of the non-linear relationship between the applied magnetic field H and the resulting magnetic induction B. The smaller the surface of the well-known BH curve, the smaller the hysteresis losses. The hysteresis losses are proportional to the frequency.

[0005] The history of the transformer has known two basic types of transformer cores: a core with sheet material and a core with wire material. However, from the beginning of the twentieth century, the embodiment with the wire core has been left, and all development has been focused on the embodiment with the sheet material until the development of so-called electrical steel sheet in 1933. Since then, advances have been made to further enhance material properties through thinner gages, alloying modifications, laser scribing, but the basic processes have remained more or less the same. Today, the transformer with the sheet material is the standard.

[0006] Despite the widely accepted use and more than hundred years of

development, a magnetic core of sheet material has a number of drawbacks.

[0007] One major drawback is that further reduction of the magnetic losses, and particularly the reduction of the Eddy current losses (proportional to the square of the thickness of the sheets), would necessitate a further reduction in the thickness of the sheet. While this can be done technically, it is no longer justified for economic reasons.

[0008] A second drawback is that sheets inherently lack flexibility. Indeed it is difficult to bend sheets in order to match a particular geometry of magnetic cores. In other words, some geometries of magnetic cores either demand special preparation of the sheets or are not possible to realize with sheets. [0009] Another drawback is that sheets are not fit for magnetic cores in case the magnetic flux has a three-dimensional pattern, since the sheets do not follow the direction of the flux here.

[0010] Since the year 1990 there has been a renewed attention for making

magnetic cores of wire. WO-A1 -91/09442 discloses a magnetic flux path for an electrical device, where the magnetic flux path is formed from discrete magnetic material wire wrapped in the desired return path configuration. This use of discrete magnetic material wire greatly reduces the Eddy current losses because of the smaller cross-section of the wire material in comparison with the sheet material. Other documents which also disclose the use of wire as magnetic core are WO-A2-00/44006, WO- A-02/059914, WO-A-02/059916, WO-A-02/059918, WO-A-2007/084963.

[001 1] Despite this renewed attention for wire material for magnetic cores, the market has not seen the reappearance of wire material on a commercial scale. It is recognized that one of the reasons blocking this reintroduction is the feature that - in contrast to eddy current losses - hysteresis losses remain high. Indeed, the normal crystallographic texture and crystal structure obtained by drawing and / or by rolling a steel wire is not favourable for hysteresis losses.

Disclosure of Invention

[0012] It is an object of the present invention to avoid the drawbacks of the prior art.

[0013] It is a main object of the present invention to provide a steel wire which is suitable for use in a magnetic flux path.

[0014] It is another object of the present invention to reduce the hysteresis losses in a steel wire. [0015] According to a first aspect of the present invention, there is provided a sheared steel wire with a width of the cross-section ranging from 0.4 mm to 6.0 mm, preferably ranging from 0.5 mm to 5.0 mm. The thickness of the cross-section ranges from 0.05 mm to 1.0 mm, preferably from 0.10 mm to 0.80 mm. The width to thickness ratio ranging from 1.05 to 30, preferably from 1.10 to 25.

The smaller cross-section of a steel wire in comparison with a sheet decreases the eddy current losses.

Due to the shearing the steel wire now has obtained a <100> fiber orientation over at least part of its cross-section, and preferably over whole its cross-section. This <100> orientation along the wire axis greatly increases the permeability and reduces the hysteresis losses. To the knowledge of the inventors, this is the first time a <100> orientation along the wire axis was obtained for work hardened steel wire with a body centered cubic (BCC) crystal structure.

[0016] The hysteresis loss of this sheared steel is less than 10 Watt per kilogram wire measured at 50 Hz and for a magnetic induction of 1.4T, e.g. less than 5 Watt per kilogram wire measured at 50 Hz, e.g. less than 4 Watt per kilogram wire measured at 50 Hz.

[0017] A suitable steel composition for a steel wire according to the invention is as follows:

- maximum carbon content of 0.10 weight %, e.g. of 0.08 weight %, e.g. of 0.06 weight %;

- silicon content ranging from 0.0 weight % until 8.0 weight %, e.g. from 0.0 weight % to 5.0 weight %;

- maximum manganese content of 3.0 weight %, e.g. of 2.5 weight %, e.g. of 1.5 weight %;

- maximum aluminium content of 2.0 weight %, e.g. of 0.5 weight %;

- maximum copper content of 2.0 weight %, e.g. of 0.5 weight %;

- maximum chromium content of 5.0 weight %, e.g. of 1.0 weight %;

- maximum nickel content of 5.0 weight %, e.g. of 1.0 weight %; - maximum contents of sulphur, phosphorus, oxygen, and nitrogen of 0.0230 weight %, e.g. of 0.010 weight % (individual maximum, not cumulated).

The remainder of the steel composition is iron.

[0018] In a particular embodiment of the steel composition, the silicon content ranges from 1.5 weight % to 6.0 weight %, e.g. from 1.5 weight % to 5.0 weight %, e.g. from 2.0 weight % to 4.0 weight %. The higher the silicon and aluminium content, the higher the electrical resistance, and the lower the eddy current losses.

[0019] In another particular embodiment of the steel composition, the silicon

content is limited and ranges from 0.0 weight % to 0.8 weight %, e.g. from 0.0 weight % to 0.5 weight %. The silicon content may even be limited to the presence of unavoidable impurities. A suitable steel composition has a maximum carbon content of 0.05 weight %, a manganese content ranging from 0.25 weight % to 0.40 weight %, a maximum sulphur content of 0.05 weight % and a maximum phosphorus content of 0.05 weight %, the remainder being iron and unavoidable impurities. This steel composition is cheaper than the one with the higher silicon content. Although such a steel wire has a lower electrical resistance, the disadvantageous effect hereof is limited for applications at high frequencies. In addition, this low carbon steel composition has the advantage that it has a higher magnetic saturation limit B _s . Another advantage is that the lower silicon content or even the absence of silicon is favourable for workability.

[0020] The steel wire according to the invention is preferably work hardened, e.g. work hardened by drawing followed by rolling, e.g. cold drawing, e.g. cold dry drawing followed by cold rolling, for reaching the final dimensions. Thereafter the steel wire can be annealed.

[0021] The steel wire is provided with an electrically insulating flexible coating such as an oxide, e.g. MgO or an iron oxide, in order to limit the transverse fluxes and to reduce the eddy current losses. From a general point of view the coating must not only provide an electrical insulation but must also be heat resistant to an annealing treatment. From a general point of view also the thickness of the coating may range from 0.05 μηη to 10 μηη, e.g. from 0.2 μηη to 8 μηη.

[0022] According to a second aspect of the present invention, there is provided a magnetic flux path comprising a number of windings of a sheared steel wire according to the first aspect of the invention. This number of windings forms one or more layers in the magnetic flux path.

[0023] According to a third aspect of the present invention, there is provided a method of manufacturing a sheared steel wire according to the first aspect. This method comprises the following steps:

a) providing a steel wire rod;

b) subject the wire rod to a heat treatment;

c) drawing the steel wire rod until a drawn wire at an intermediate diameter, e.g. by cold drawing, e.g. by cold dry drawing;

d) annealing the steel wire

e) optionally rolling the drawn wire until a rolled wire with a flattened cross- section;

f) shearing the rolled wire to a sheared wire with at least partially a <100> fiber texture over its cross-section;

g) optionally annealing the sheared wire.

[0024] The shearing step f) can be done at room temperature or at an elevated temperature ranging from 100 °C to 750 °C, e.g. from 150 °C to 350 °C.

[0025] Shearing can be done by means of a rolling operation wherein the two rolls operate at different linear speeds. These different linear speeds may be obtained either by rolls having an equal diameter but different rotational speeds or by using rolls having a different diameter or by a combination of both.

[0026] The shearing can also be done by means of a rolling operation wherein the two rolls have a different surface roughness.

[0027] The shearing operation can be done in one single step or in several

subsequent steps.

[0028] The shearing treatment is followed by a step of coating the steel wire with a thin, electrically insulating flexible coating.

[0029] The coated wire is then wound in a form to be used as a magnetic flux path in a magnetic device such as a transformer or an electrical motor. A number of windings will form a first layer. There may be one or more layers, one above the other.

[0030] Before being used as a magnetic flux path, the thus wound wire structure can be subjected to an annealing treatment, i.e. a stress-relieve annealing treatment. This is the reason why the coating on the wire must have heat resistant properties.

Brief Description of Figures in the Drawings

[0031] Figures 1 (a), 1 (b) and 1 (c) show <1 10> pole figures obtained from

simulations of wire drawing for BCC polycrystals.

[0032] Figures 2(a), 2(b) and 2(c) show <100> pole figures obtained from

simulations of a combination of compression and shear on BCC polycrystals staring from a wire drawing texture.

[0033] Figures 3(a), 3(b) and 3(c) show other <100> pole figures obtained from simulations of a different combination of compression and shear on BCC polycrystals starting from a wire drawing texture. [0034] Figures 4(a) shows a simulated <1 10> pole figure and Figure 4(b) shows an experimental <1 10> pole figure.

[0035] Figures 5(a) and 5(b) show cross-sections of a flattened steel wire.

[0036] Figure 6a shows the longitudinal section of a non-sheared steel wire while Figure 6b shows the longitudinal section of a sheared steel wire.

[0037] Figure 7 illustrates the <100> fiber orientation.

[0038] Figure 8 shows a set-up to measure the magnetic losses in a magnetic core composed of wires.

[0039] Figure 9a shows the magnetic induction B as a function of the applied magnetic field H for several wires and a reference material.

Figure 9b shows the total magnetic losses as a function of the magnetic induction B for several wires and a reference material.

[0040] Figure 10 shows a first embodiment to obtain a sheared steel wire.

[0041] Figure 1 1 shows a second embodiment to obtain a sheared steel wire.

[0042] Figure 12 shows a third embodiment to obtain a sheared steel wire.

[0043] Figure 13 illustrates a magnetic core used in another test set-up.

[0044] Figure 14 illustrates a power loss measurement circuit.

[0045] Figure 15 illustrates a measured BH loop of a magnetic core.

[0046] Figure 16 illustrates a comparison of magnetic core power losses. Mode(s) for Carrying Out the Invention

[0047] A novel inductive design based on a magnetic component in the form of a wire requires that the crystallographic texture in the wire should have a strong <100> fiber texture in the direction parallel to the wire axis in a non- rotating magnetic field. A wire made up of an iron - silicon alloy or a plain low-carbon steel wire has a body centered cubic (BCC) crystal structure. When subjected to conventional wire drawing and wire rolling operations, the crystallographic texture typically obtained results in the <1 10> direction parallel to the wire axis. This direction has inferior magnetic properties.

[0048] A modelling effort has been carried out to obtain the deformation and

annealing conditions under which a <100> fiber, the desirable wire texture, could be achieved. For this purpose a microstructural deformation modelling technique based on a crystal plasticity approach has been used. The deformation texture was further evolved through annealing

simulations using a Monte Carlo approach that incorporated a new nucleation model for recrystallization. The deformation and annealing models were validated using available experimental data for wire drawing of FCC (face centered cubic) and BCC polycrystals.

[0049] In FCC materials, the experimental wire drawing deformation texture has a mixture of <1 10> and <100> fibers. Annealing experiments indicate that the <100> fiber strengthens at the expense of the <1 10> fiber.

[0050] In contrast with FCC materials, experimental results for BCC materials show that the as drawn texture is a <1 10> fiber, which is retained after annealing. Figures 1 (a), 1 (b) and 1 (c) all show <1 10> pole figures obtained from simulations for BCC materials. Figure 1 (a) shows the pole figure of the initial texture. Figure 1 (b) shows the pole figure of the wire drawing texture. Figure 1 (c) shows the pole figure of the annealed texture all showing a strong <1 10> fiber along the wire axis. [0051] The next step was to use the validated simulation tools to identify the deformation conditions under which a strong <100> fiber texture could be obtained in a drawn and annealed iron wire. The initial texture for these simulations was the conventional wire drawing texture for BCC

polycrystals. The deformation path was varied through the externally imposed velocity gradient. Using the simulations it was possible to identify an industrially viable deformation path that would result in the optimum texture in annealed iron wire. The evolution of the deformation texture through two such deformation paths resulting in a final cube texture is shown in Figures 2 and 3. The deformation paths are based on a combination of compression and shear loads.

[0052] Figures 2(a), 2(b) and 2(c) show simulation results for a deformation path based upon a combination of compression and shear that finally leads to a cube texture. In the case of Figures 2 the shear was obtained using the compression and shear conditions seen in equal channel angular extrusion.

[0053] Figures 3(a), 3(b) and 3(c) show simulation results for a deformation path based upon a combination of compression and shear that finally leads to a cube texture. In the case of Figures 3 the shear was obtained using compression and shear conditions seen in asymmetric rolling.

[0054] A next step was to check the results obtained from simulation with

experimental results.

[0055] A first thing was to optimize the conditions under which rolling is carried out in order to transmit through thickness shear between the steel strip and the rolls. A series of rolling experiments were performed with sheet samples 1 mm in thickness in which pins of 1.65 mm in diameter were embedded. The sheet had a composition of iron and 2.6 weight % Si, the wire had a composition of iron and 3.0 weight % Si. The initial tests showed that room temperature deformation introduced very little shear in the pin, while deformation at approximately 300 °C introduced a significant amount of shear deformation in the through thickness direction. The reason for the significantly greater shear during warm rolling is not clear. It could be related to the formation of a surface oxide or to the fact that the material is softer or to the fact that there is more friction between the wire and the rolls. Later experiments, however, showed that heating up is not always required.

[0056] Figure 4a shows the simulation results for warm shear rolling a 0.4 x 0.8 mm thick strip. Figure 4b shows the experimental texture obtained along the axis of a 0.4 x 0.8 strip that was warm shear rolled at Bekaert using asymmetric rolling. It is clear that a strong <100> orientation was obtained in both cases.

[0057] Figure 5(a) shows a flattened steel wire 150 obtained after rolling with two rolls. The steel wire 150 has a thickness T and a width W. Examples of steel wire 150 according to the invention are:

- 0.40 mm x 0.80 mm;

- 0.60 mm x 1.20 mm;

- 0.1 1 mm x 2.0 mm;

- 0.18 mm x 2.64 mm;

- 0.1 1 mm x 2.27 mm;

- 0.38 mm x 4.08 mm;

- 0.25 mm x 2.29 mm.

The steel wire shown in Figure 5(a) is a flat steel wire with so-called round natural edges.

[0058] As shown in Figure 5(b), the steel wire 152 can also have a rectangular cross-section, with four straight sides, two pairs of parallel sides. Such a steel wire may be obtained by means of so-called Turks heads. [0059] Figure 6a and Figure 6b illustrate schematically the difference between a normal, non-sheared wire and a sheared wire by means of pins which are integrated in the wire.

[0060] Figure 6a shows a longitudinal section of a normal, non-sheared wire 160.

Pins 162 run more or less perpendicular to the rolling direction shown by the arrow or only exhibit a relatively small angle deviating from this perpendicular direction.

[0061] Figure 6b shows a longitudinal section of a sheared wire 164. The pins 166 are running more horizontal in comparison with the pins 162. The direction of the pins 166 deviate to a large extent from the direction of the pins 162. This different orientation and different texture is due to the shearing treatment, i.e. a treatment which is different on the upper surface than on the lower surface of the wire.

[0062] Figure 7 illustrates the <100> fiber orientation. An iron-silicon steel wire or a plain low carbon steel wire typically has a body centered cubic crystal structure (BCC structure).

Conventional wire drawing or rolling gives a <1 10> orientation to the steel wire, which is unfavourable from a magnetic point of view. Indeed the <1 10> orientation results in a low magnetic permeability and in high losses. It requires a magnetic field strength H more than three times as high to obtain the same magnetic flux density than compared with sheet material. Similarly, the losses are more than three times higher when using a conventionally drawn wire in comparison with grain oriented sheet material.

[0063] Referring to Figure 7, steel wire 170 is rolled in the rolling direction RD.

TD refers to the transverse direction, and ND refers to the normal direction. The cubic crystal 172 illustrates the better <100> fiber orientation, which is favourable from a magnetic point of view. With this orientation, the applied magnetic field H can be reduced a number of times and still the same amount of magnetic flux obtained. With the <100> orientation the losses can be reduced to less than 50% in comparison with non grain oriented wire.

[0064] Figure 8 illustrates a first test set-up 180 for measuring the losses in a magnetic flux path 182. The magnetic flux path 182 can be composed of the material to be measured. This can either be wires of different dimensions or sheet material that is used as a reference. A first number of electrical windings 184 is wound around the magnetic flux path 182 and forms the primary winding. A second number of electrical windings 186 is wound around the magnetic flux path 182 and forms the secondary winding. An input voltage is applied at the primary winding and the induced voltage at the secondary winding is measured. This is being done for different frequencies. As the eddy current losses are proportional to the square of the frequency and the hysteresis losses are proportional to the frequency, a distinction can be made between both types of losses in the following way: The total losses divided by the frequency are plotted as a function of the frequency. The hysteresis losses divided by the frequency intercepts with the vertical axis and the eddy current losses divided by the frequency is the gradient.

[0065] Figure 9a shows the magnetic induction B; expressed in Tesla, as a

function of the applied magnetic field H, expressed in A m at a frequency of 50 Hz.

In general, the higher the value of the magnetic induction B for a given applied magnetic field, the higher the magnetic permeability μ and the smaller the resistance to magnetic fluxes.

Curve 191 represents the measurement values relating to a magnetic flux path of round iron silicon wire (2.57weight % Si) with a diameter of 0.5 mm, drawn in a conventional way, without any shearing treatment whatsoever.

Curve 192 shows the measurements relating to a magnetic flux path of an unannealed sheared iron silicon wire, with 2.57 weight % of Si and with dimensions 0.30 mm x 2.33 mm.

Curve 193 shows the measurements relating to a magnetic flux path of an annealed sheared iron silicon wire, with 2.57 weight % of Si and with dimensions 0.23 mm x 3.07 mm.

Curve 194 shows the measurements relating to a magnetic flux path of an annealed sheared iron silicon wire, with 2.57 weight % of Si and with dimensions 0.30 mm x 2.33 mm.

Curve 195 represents the measurement values relating to magnetic flux path of commercially available grain oriented electrical sheet material with a thickness of 0.23 mm.

Curves 193 and 194 of a magnetic flux path of an annealed sheared iron silicon wire are closest to the reference of a commercially available electrical sheet material.

Even if not annealed, the magnetic properties become better because of the shearing, see curve 192.

Figure 9b shows the total magnetic losses, i.e. mainly hysteresis losses and eddy current losses, expressed in W/kg, as a function of the magnetic induction B, expresses in Tesla, at a frequency of 50 Hz. The lower the losses the better.

Curve 191 ' represents the measurement values relating to a magnetic flux path of round iron silicon wire (2.57 weight % Si) with a diameter of 0.5 mm, drawn in a conventional way, without any shearing treatment whatsoever.

Curve 192' shows the measurements relating to a magnetic flux path of an unannealed sheared iron silicon wire, with 2.57 weight % of Si and with dimensions 0.30 mm x 2.33 mm.

Curve 193' shows the measurements relating to a magnetic flux path of an annealed sheared iron silicon wire, with 2.57 weight % of Si and with dimensions 0.23 mm x 3.07 mm.

Curve 194' shows the measurements relating to a magnetic flux path of an annealed sheared iron silicon wire, with 2.57 weight % of Si and with dimensions 0.30 mm x 2.33 mm. Curve 195' represents the measurement values relating to magnetic flux path of commercially available grain oriented electrical sheet material. Curves 193' and 194' of a magnetic flux path of an annealed sheared iron silicon wire are closest to the reference of a commercially available electrical sheet material.

Even if not annealed, the magnetic properties become better because of the shearing, see curve 192' in comparison with curve 191 '.

[0067] Table 1 hereunder mentions the total losses in W/kg of the various tested magnetic flux paths for a magnetic induction of 1.2 Tesla at various frequencies.

[0068] Table 1 : Total losses at 1.2 Tesla

[0069] Table 2 hereunder mentions the values of the hysteresis losses in W/kg at 1.2 Tesla of the various tested magnetic flux paths. These values are derived from the total losses according to the method as outlined above. [0070] Table 2: Hysteresis Losses at 1.2 Tesla

[0071] Table 3 hereunder mentions the total losses in W/kg of the various tested magnetic flux paths for a magnetic induction of 1.4 Tesla at various frequencies.

[0072] Table 3: Total Losses at 1.4 Tesla

[0073] Table 4 hereunder mentions the values of the hysteresis losses at 1.4

Tesla in W/kg of the various tested magnetic flux paths. These values are derived from the total losses according to the method as outlined above.

[0074] Table 4: Hysteresis Losses at 1.4 Tesla

[0075] A sheared wire according to the first aspect of the present invention can be made as follows.

[0076] Starting steel composition is e.g. 0.05 weight % C, 0.30 weight % Mn, 0.03 weight % P and 0.03 weight % S, the remainder being Fe and unavoidable impurities.

[0077] A wire rod with this composition is first subjected to a heat treatment.

[0078] Thereafter the wire rod is cold drawn from the wire rod diameter 5.5 mm to a diameter 1.75 mm in a number of subsequent dry drawing steps.

[0079] The thus drawn steel wire can be annealed.

[0080] The thus annealed steel wire is then subjected to one or more subsequent cold rolling steps.

[0081] The rolled steel wire is then sheared between two rolls as will be explained hereinafter. [0082] Figure 10 illustrates a first way and equipment 200 to apply a shearing treatment to a steel wire 202. The top roll 204 and the bottom roll 206 rotate at the same rotational speed. Top roll 204 has a larger diameter than bottom roll 206, and thus has a higher linear speed, which results in the upper surface of the rolled steel wire 208 to be deformed heavier than the under surface. Possible ratios of roll diameters are 2 to 1 and 2.8 to 1.0.

[0083] Figure 1 1 illustrates a second way and equipment 210 to apply a shearing treatment to a steel wire 212. The difference between Figure 1 1 and Figure 10 is that in Figure 1 1 the rolls 214 and 216 have the same roll diameter but rotate at a different rotational speed so that here again a different linear speed occurs at the contact with the steel wire 218.

[0084] Figure 12 illustrates yet another and third way and equipment 220 to apply a shearing treatment to a steel wire 222. Here both rolls 224 and 226 have the same roll diameter and rotate at the same rotational speed so resulting in the same linear speed at the contact with the steel wire 228. One of the rolls, however, in this case the upper roll 224, has an outer surface that is provided with roughnesses which are greater than the roughnesses present on the surface of the under roll 226 and so resulting in a different treatment at the upper surface of the steel wire 228.

[0085] In contrast to the first experiments which showed that shearing had to be carried out by means of hot rolling, good results have been obtained in the meantime by means of cold rolling, in other words by means of rolling at room temperature. A possible explanation is that the heating up may be favourable by creating an increased friction between the rolls and the wire. However, this increased friction may be obtained by other purely mechanical means as well.

[0086] Other methods to manufacture a sheared wire are applying a heating on only one side of the wire and using one roll at one side and a flat surface at the other side.

[0087] In addition to the shearing treatment, a post heat treatment can be applied after the shearing to further improve the magnetic properties, such as a further reduction of the hysteresis losses.

[0088] An flat iron silicon wire (2.56 weight % Si) with dimensions 0.25 mm x 2.29 mm and a thin oxide coating was subjected to another series of tests to measure the BH loop and the core power losses.

[0089] Referring to Figure 13 a ring core 230 was made by winding the flat wire on a coil former with an inner diameter d of 3 cm and an outer diameter of 5 cm (i.e. t = 1 cm). The number of windings is 240 and the height of the coil former is 4 cm so that 16 windings of flat wire in one layer are wound. The cross-section of the magnetic core is 159 mm ² .

[0090] A digitizing method to measure the core power losses according to the International Standard IEC 62044-1 [2], IEC 62044-3 is applied.

[0091 ] The test set-up 240 is illustrated in Figure 14. A sinusoidal waveform is generated by means of a function wave generator 242 Agilent 33250A and is amplified by power amplifier 243 KEPCO. A DC capacitor 244 (100 iF or 1000 [iF) is used to stop DC current in the circuit. The thus generated and amplified waveform is applied to the inductor 245 as formed in Figure 13. The alternating current in the primary winding excites a voltage in the secondary winding. An oscilloscope 246 YOKOGAWA DL 1740EL with a bandwidth up to 500 MHz is used to digitize and record current through inductor and secondary winding voltage. The digitized data were transferred from the oscilloscope 246 to a PC for further data processing. The average of the product of the current and voltage of the inductor derived from measurement over one cycle was used to calculate the core power loss. The obtained results were compared with the results obtained by the calorimetric method. This calorimetric method is based upon the principle that if a component that dissipates power is put inside an air-stop box, then the temperature inside the box will rise and this temperature increase is proportional to the power dissipated inside the box. It has been found that the power loss differences between the calorimetric method and the digitizing method are smaller than 4% so that the results obtained by the digitizing method can be considered as reliable.

[0092] Measurements were done for following frequencies (all in Hz): 50, 100,

200, 400, 500, 1000, 2000, 4000, 5000 and for following flux densities (all in Tesla): 0.075, 0.125, 0.15, 0.2, 0.25, 0.4, 0.5.

[0093] Figure 15 illustrates the obtained BH loop 250.

[0094] Figure 16 compares the obtained values of the power losses of the core with the wire material with conventional silicon steel sheet materials.

Curve 260 relates to M800-65A silicon sheet, curve 262 relates to M400- 50A silicon sheet, curve 264 relates to M250-50A silicon sheet and curve 266 relates to the 0.25 mm x 2.29 mm wire core. It can be concluded that this wire material has a slightly larger power loss density than conventional silicon steel sheet materials.