METAL STRIP, METHOD FOR PRODUCING AN AMORPHOUS METAL STRIP AND METHOD FOR PRODUCING A NANOCRYSTALLINE METAL STRIP

Abstract

A metal strip is provided having a casting-wheel side that has been solidified on an outer surface of a heat sink, an opposing, air side and a microstructure. The microstructure is at least 80 vol. % amorphous or has at least 80 vol. % nanocrystalline grains and a residual amorphous matrix in which at least 80% of the nanocrystalline grains have an average grain size of less than 50 nm and a random orientation. The air side of the metal strip has a surface crystallisation proportion of less than 23%.

Claims

1. A metal strip, comprising: a casting-wheel side that has solidified on an outer surface of a heat sink; an air side opposing the casting-wheel side, and a microstructure that is at least 80 vol. % amorphous or has at least 80 vol. % nanocrystalline grains and a residual amorphous matrix in which at least 80% of the nanocrystalline grains have an average grain size of less than 50 nm and a random orientation, the air side having a surface crystallisation proportion of less than 23%.

2. A metal strip according to claim 1, wherein the air side comprises a surface crystallisation proportion of less than 5%.

3. A metal strip according to claim 1, wherein the casting-wheel side comprises a surface crystallisation proportion of less than 23%.

4. A metal strip according to claim 1, wherein the metal strip comprises a surface layer that forms between 0.01% and 5% of the total volume and contains crystalline grains that form the surface crystallisation.

5. A metal strip according to claim 3, wherein 80 vol. % of the crystalline grains of the surface crystallisation have an average grain size of greater than 100 nm.

6. A metal strip according to claim 4, wherein the crystalline grains of the surface crystallisation have a crystallographic texture.

7. A metal strip according to claim 1, wherein the metal strip comprises a width of 2 mm to 300 mm, and/or a thickness of less than 50 m.

8. A metal strip according to claim 1, wherein the metal strip comprises a titanium content of less than 0.25 at. %, an aluminium content of less than 0.4 at. %, a manganese content of less than 0.4 at. % and a sulphur content of less than 0.35 at. %.

9. A metal strip according to claim 1, wherein the metal strip consists of (Fe,T).sub.aM.sub.b and up to 1 at. % impurities, where 70 at. %a90 at. % and 10 at. %b30 at. %, T being one or more of the elements Co, Ni, Cu, Cr, Zn, Sn and V, and M being one or more of the elements Nb, Mo, Zr Ta, B, Si, C and P.

10. A metal strip according to claim 1, wherein the metal strip consists of F.sub.100-a-b-w-x-y-z T.sub.a M.sub.b Si.sub.w B.sub.x P.sub.y C.sub.z (in at. %) and up to 1 at. % impurities, T being one or more of the group consisting of Co, Ni, Cu, Cr, Zn, Sn and V, and M being one or more of the group consisting of Nb, Mo, Zr and Ta, where 0a80, 0b10, 0w25, 3x20, 0y7, and 0z2.

11. A metal strip according to claim 9, wherein the metal strip has .sub.dyn>100000.

12. A metal strip according to claim 9, wherein the casting-wheel side of the metal strip having a surface roughness with an arithmetic mean, Ra, of less than 0.8 m, preferably less than 0.7 m.

13. A method for the production of an amorphous metal strip using a rapid solidification technology, the method comprising: providing a molten mass of an alloy; continuously pressing a rolling device against the outer surface of the heat sink as the molten mass is poured onto the moving outer surface of the heat sink with a pressure sufficient to smooth the outer surface of the heat sink, pouring the molten mass onto a moving outer surface of a moving heat sink, the molten mass solidifying on the outer surface and an amorphous metal strip being formed, the amorphous metal strip having a casting-wheel side that has solidified on the outer surface of a heat sink, an opposing, air side and a microstructure that is at least 80 vol. % amorphous, the air side having a surface crystallisation proportion of less than 23%.

14. A method according to claim 13, wherein the casting-wheel side has a surface crystallisation proportion of less than 23%.

15. A method according to claim 13, wherein the amorphous metal strip has a width of 2 mm to 300 mm and/or a thickness of less than 50 m.

16. A method according to claim 13, wherein the metal strip has a surface layer that comprises between 0.01% and 5% of the total volume and contains crystalline grains that form the surface crystallisation.

17. A method according to claim 16, wherein 80 vol. % of the crystalline grains in the surface crystallisation have an average grain size of greater than 100 nm.

18. A method according to claim 13, wherein the rolling device is pressed against the outer surface of the heat sink such that it continuously reduces the roughness of the outer surface of the heat sink as the molten mass is poured onto the outer surface of the heat sink.

19. A method according to claim 13, wherein a rotatable roller is provided as the rolling device and the surface of the rotating roller is pressed against the outer surface of the rotating heat sink with a pressure that the outer surface of the heat sink is reshaped.

20. A method according to claim 13, wherein a rotatable roller is provided as the rolling device and the roller is driven in a first direction of rotation and the heat sink is driven in a second direction of rotation, the first direction of rotation being opposite the second direction of rotation.

21. A method according to claim 13, wherein the roller is moved over the outer surface of the heat sink parallel to the second axis of rotation of the heat sink such that contact with the outer surface of the heat sink is spiral-shaped.

22. A method according to claim 13, wherein the outer surface of the heat sink is protected by organic material at least at the point at which the molten mass hits the outer surface as the molten mass is poured onto the moving outer surface of the moving heat sink.

23. A method according to claim 13, wherein the heat sink is formed from a material with a thermal conductivity of greater than 200 W/mK.

24. A method according to claim 13, wherein the metal strip consists of (Fe,T).sub.aM.sub.b and up to 1 at. % impurities, where 70 at. %a90 at. % and 10 at. %b30 at. %, T being one or more of the elements Co, Ni, Cu, Cr, Zn, Sn and V and M being one or more of the elements Nb, Mo, Zr, Ta B, Si, C and P.

25. A method according to claim 13, wherein the metal strip consists of Fe.sub.100-a-b-w-x-y-z T.sub.a M.sub.b Si.sub.w B.sub.x P.sub.y C.sub.z (in at %) and up to 1 at. % impurities, T being one or more of the group consisting of Co, Ni, Cu, Cr, Zn, Sn and V, and M being one or more of the group consisting of Nb, Mo, Zr and Ta, where 0a80, 0b10, 0w25, 3x20, 0y7, and 0z2.

26. A method for the production of a nanocrystalline foil, comprising: heat treating an amorphous foil produced using a method according to claim 11 at a temperature Ta, where 400 C.Ta750 C., in order to generate a nanocrystalline structure in the foil in which at least 80 vol. % of the grains have an average size of less than 50 nm and a random orientation.

27. A method according to claim 26, wherein the strip is heat treated in a continuous furnace.

28. A method according to claim 27, wherein the strip is drawn through the continuous furnace at a speed s such that a strip dwell time in a temperature zone of the continuous furnace with a temperature Ta is between two seconds and two minutes.

29. A method according to claim 26, wherein the strip is continuously heat treated under a tensile stress of 5 MPa to 1000 MPa.

30. A method according to claim 26, wherein a desired anisotropy field strength value, Ha, or a desired permeability value and optionally a maximum remanence ratio value, J.sub.r/J.sub.s, of less than 0.02, and/or a maximum coercive field strength value, H.sub.c that is less than 1% of the anisotropy field strength, H.sub.a, and/or less than 10 A/m, and a permitted deviation range of each of these values being predetermined, and the magnetic properties of the strip is measured continuously as the strip leaves the continuous furnace, and if deviations from the permitted deviation ranges of the magnetic properties are detected, the tensile stress on the strip is adjusted in order to bring the measured values of the magnetic properties back within the permitted deviation ranges.

31. A soft magnetic core or an inductive component including the metal strip according to claim 1.

32. An antennae or sensor including the soft magnetic core of claim 31.

33. A blade, an amorphous spring or a knife blade including the metal strip according to claim 1.

34. A brazing foil including the metal strip according to claim 1.

35. A layer of a laminate including the metal strip according to claim 1.

36. A shielding foil in an object with parts for wireless charging including the metal strip according to claim 1.

37. A shielding foil in an object with parts to be shielded including the metal strip according to claim 1.

38. A shielding foil according to claim 37, the parts to be shielded being one or more of the group consisting of electronic components, cables, sensor ranges and cavities.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0110] The invention is explained in greater detail below with reference to the drawings and examples.

[0111] FIG. 1 shows a schematic time-temperature transformation diagram.

[0112] FIG. 2a shows a dark-field TEM image of an amorphous sample with surface crystallisation.

[0113] FIG. 2b shows a dark-field TEM image of the same sample in the heat-treated, nanocrystalline state.

[0114] FIG. 3 shows a diffractogram of an amorphous sample.

[0115] FIG. 4 shows a diffractogram of a nanocrystalline sample.

[0116] FIG. 5a shows a diffractogram of an amorphous sample with low surface crystallisation.

[0117] FIG. 5b shows a diffractogram of an amorphous sample with very marked surface crystallisation.

[0118] FIG. 6 shows a diffractogram of a nanocrystalline sample with marked surface crystallisation.

[0119] FIG. 7 shows a diffractogram of an amorphous sample with surface crystallisation that is not textured.

[0120] FIG. 8 shows diagrams comparing the extent of surface crystallisation on the free side of a metal strip and the side of a metal strip facing the casting wheel.

[0121] FIG. 9 shows a graph of coercive field strength as a function of grain size.

[0122] FIG. 10 shows graphs of AC permeability with two-way sine modulation at 1.5 A/m and 50 Hz Dyn (a) and with modulation at 0.3 A/m and 100 kHz 100 kHz (b) as a function of surface crystallisation.

[0123] FIG. 11 shows a graph of the surface crystallisation as a function of average strip thickness.

DETAILED DESCRIPTION OF EXAMPLE EMBODIMENTS

[0124] One example of a rapid solidification technology is the melt-spinning process. During production using melt-spinning a glass-forming metal alloy is melted in a crucible that is typically made substantially of oxide ceramic (e.g. aluminium oxide) or graphite. Depending on the reactivity of the molten mass, the melting process may take place in air, in a vacuum or in an inert gas such as argon or nitrogen, for example.

[0125] Once the alloy has been melted down at temperatures well above the liquidus point, the molten mass is transported to a casting tundish and injected through a casting nozzle, which generally has a slit-shaped outlet opening, onto a rotating heat sink, e.g. a roller or wheel made of a copper alloy. To this end, the casting nozzle is brought very close to the surface of the rotating copper cylinder at a distance of approx. 50 m to 500 m from it. The molten mass passes through the nozzle outlet and hits the moving surface of the heat sink where it solidifies at cooling rates of approx. 10.sup.4 K/min to 10.sup.6 K/min. The rotational movement of the roller carries the solidified molten mass away as a continuous foil strip, detaches it from the cool roller and winds it onto a winding device.

[0126] Amorphous metal strips produced using the melt-spinning process have thicknesses of between approx. 10 m and approx. 30 m. Standard widths range from 5 mm to 200 mm.

[0127] In addition to the aforementioned high cooling rates, the production of amorphous metal strips also requires a non-metallic, so-called metalloid or glass-forming content in the alloy in order to disrupt the formation of a crystalline microstructure, Elements commonly used as metalloids in this context are boron, silicon, phosphorus and niobium, and the total content is usually between 10 at.-% and 30 at.-%. Iron, nickel and cobalt, in particular, but also copper, are used as the metals owing to their ferromagnetic order at room temperature. Typical alloys and their production are described in EP 0 271 657 A, for example.

[0128] It has been found that both the magnetic properties in the case of magnetic alloys and the workability of foils are significantly influenced, even impaired, by surface crystallisation caused during production. Crystallites with a typical grain size of a few hundred nanometres form directly under the surface. In the case of iron-based alloys, they may for example, be cubic, body-centred -Fe or Fe3Si grains. One possible explanation for this observation is explained with reference to FIG. 1.

[0129] The present invention is based on the new finding that partial crystallinity in the proximity of the surface of rapidly solidified strips generally occurs when heat cannot be dissipated sufficiently quickly during production of the strip to generate a completely amorphous microstructure, and that the cooling rate may vary due to various factors.

[0130] The necessary cooling rate is determined amongst other factors by the metalloid content of the alloy. The larger the content of structure-disrupting atoms, i.e. the larger the metalloid content, the lower the cooling rates needed for complete amorphous solidification.

[0131] FIG. 1 shows a schematic representation of a time-temperature transformation diagram for metallic glasses on which possible cooling curves are plotted. Depending on the cooling speed, the material coming from the molten mass or melt reaches either the glass or the crystalline state with crystallisation starting locally at so-called crystallisation nuclei such as structural defects in the material, for example.

[0132] Curve A with the highest cooling rate very clearly generates an amorphous microstructure. Curve B with its middle cooling rate still generates an amorphous microstructure and can easily be transformed into a partially crystalline microstructure due to crystallisation nuclei. The lowest cooling rate C clearly results in a crystalline state. The presence and the number of crystallisation nuclei therefore have a major influence on the crystallisation of the undercooled molten mass.

[0133] According to the invention crystallisation nuclei should therefore be avoided. The following measures are examined and adjusted as ways of keeping the cooling rate high and so better preventing the crystallisation of the undercooled molten mass.

[0134] In the melt-spinning process selected, surface crystallinity can occur because a surface always represents a disruption in microstructure and so acts as a crystallisation nucleus. In addition, heat dissipation takes place via a casting wheel made of copper on which the strip lies and the surface opposite the casting wheel, the air side of the metal strip, therefore has lower heat dissipation. This effect can be amplified further if exothermic oxidation with the oxygen in the surrounding air takes place on this free surface because the additional oxidation heat must also be dissipated.

[0135] However, surface crystallisation is not observed exclusively on the free surface. Less frequently, the side facing the casting wheel, i.e. the casting-wheel side of the metal strap, can also exhibit surface crystallisation because crystalline copper can act as a very good crystallisation nucleus. For this reason, in what follows below both sides, i.e. the free surface and the surface facing the casting wheel, of the rapidly solidified metallic strip are always examined. As a result it was possible to examine all four cases: surface crystallisation on the air side only, on the casting-wheel side only, on both sides or on neither side.

[0136] In order to maintain the outstanding soft magnetic properties and the good mechanical workability of rapidly solidified metallic strips even in the face of the increasing demands for larger widths, higher saturation inductions, lower raw material costs and reduced thicknesses described above, a strip that is as free as possible of surface crystallisation is cast.

[0137] Impurities in the molten mass are minimised firstly because they can act as crystallisation nuclei, Secondly, impurities of this type, in particular, that tend to strong exothermic oxide formation in the air, such as aluminium or titanium, are damaging because they promote surface crystallisation. This may be due to the exothermic nature of the oxide formation leading to a locally reduced cooling rate. However, it may also be due to heterogeneous nucleation where these elements are already present in oxide form.

[0138] The influence of the casting wheel on surface crystallisation is also examined since it affects surface crystallisation. This includes both the thermal conductivity of the copper alloy and the geometric dimensions of the casting wheel and the casting wheel surface. Wear on this surface during the casting process results in the occurrence of cavities that transport process gas under the molten metal droplet and cause contact difficulties between the molten mass or strip and the roller. This significantly reduces the cooling rate at least locally. In order to minimise casting wheel wear, a high-strength casting-wheel material is selected. Generally, however, the properties of strength and thermal conductivity tend to act in opposite directions in the copper materials usually used in melting metallurgy.

[0139] A further parameter that can be used to influence cooling rate is the material of the casting wheel surface.

[0140] It has been observed that a significantly higher percentage of strips produced with a casting wheel surface made of a copper-nickel-silicon alloy with a high strength of 200 HV (HV30) exhibit surface crystallisation than the average of approx. 50% of strips produced with a casting wheel surface made of a CuBe alloy. This is due to the alloy's low thermal conductivity of only 150 W/mK.

[0141] The tendency to form surface crystallinity significantly improve when beryllium-alloyed copper materials with a thermal conductivity of over 200 W/mK were used. With these materials thermal conductivity increases as beryllium content falls. The best results by a distance were achieved with a material that had a thermal conductivity of 330 W/mK. Low beryllium contents are also advantageous in terms of occupational health and safety since beryllium dusts are poisonous. For this reason the addition of beryllium should be limited to 2 wt. % preferably 1 wt. %.

[0142] The beryllium-alloyed copper materials have strengths of between 130 HV and 250 HV (HV30) depending on the beryllium content. However, the material with the lowest hardness of 130 HV is also that with the highest thermal conductivity.

[0143] In order to achieve good, uniform heat contact between the molten mass or strip and the roller with these soft materials and to be able to use them in the casting of amorphous strips in the long term it is advantageous to ensure even processing of the contact surface during the production process itself. The roughness of the casting wheel surface should therefore be kept uniform.

[0144] However, material-removing processes such as polishing and brushing, for example, can result in local contact problems or the development of gases due to processing residues (e.g. polishing agents, dust, brushes) on the casting wheel with the aforementioned negative effects on local cooling rates, and in increased levels of surface crystallisation. To avoid surface crystallisation the invention therefore discloses the use of a reshaping process like that described in DE 10 2010 036 401 A1 for processing the surface of the heat sink, in particular the surface of the casting track onto which the molten mass is poured.

[0145] To summarise, an amorphous metal strip with a low level of surface crystallisation even over long continuous lengths is provided that provides good, uniform mechanical properties, in particular ductility and elongation at fracture. In the case of magnetic alloys, the amorphous metal strip also has good, uniform magnetic properties, where appropriate depending on the composition following nanocrystallisation.

[0146] If the amorphous metal strip is heat treated in order to produce the nano crystalline metal strip, this nanocrystalline metal strip has good, uniform mechanical properties. In the case of a magnetic alloy, it also has good, uniform magnetic properties.

[0147] Examples of metal strips according to the invention are detailed in Table 1. Table 1 indicates the width B (mm), thickness D (m), composition and thermal conductivity of the casing-wheel material (W/mK), the level of surface crystallisation determined (%) and the .sub.Dyn of various examples of metal strips measured. The proportion of surface crystallisation is determined by means of powder diffractometry.

[0148] The strips are produced using rapid solidification technology, the surface of the casting track on the casting wheel being reshaped and smoothed by a roller during the casting process.

[0149] Table 1 also shows the relationship between the presence of marked surface crystallisation and the deterioration of soft magnetic properties by means of dynamic AC permeability Dyn (<100000).

TABLE-US-00001 TABLE 1 Thermal conductivity of casting Proportion wheel of surface B D Fe Nb Ti Al material crystallisation [mm] [m] [Wt.] [Wt.] [Wt.] [Wt.] [W/mK] [%] .sub.Dyn 1* 46 17.98 82.95 5.47 0.007 0.006 150 76 78353 2* 58 19.25 85.53 2.87 0.007 0.004 290 93 3* 58 20.12 86.02 4.69 0.004 0.002 230 65 4* 58 20.29 82.98 5.42 0.006 0.007 290 53 81982 5 25 18.63 82.97 5.47 0.004 0.002 290 1 193549 6 46 18.10 82.97 5.46 0.002 0.004 290 6 115467 7 58 18.22 83.06 5.42 0.005 0.003 290 12 152960 8 46 18.69 83.39 4.21 0.007 0.003 290 4 9 58 22.90 85.94 4.35 0.005 0.002 330 0 10 58 17.40 86.01 4.34 0.005 0.002 330 1 11 108 82.98 5.43 0.005 0.002 290 0 138708 *Comparative examples Wt. denotes weight percent

[0150] Examples 1 to 4 in Table 1 are comparative examples and show that surface crystallinity increases in wide (>50 mm), thick (>19 m) strips with an increased iron content (>85 wt. %) and reduced niobium content (<5 wt. %) supported by titanium or aluminium impurities and when the thermal conductivity of the casting-wheel material is poor (<200 W/mK).

[0151] Examples 5 and 6 according to the invention show the opposite behaviour to examples 1 to 4. They involve a thin, narrow strip with a low iron content, a high niobium content and a very small amount of titanium and aluminium impurities. They show low levels of less than 10% surface crystallisation even at average casting-wheel material thermal conductivity.

[0152] Examples 7 and 8 according to the invention show that it is also possible to produce wide strips and strips with a reduced niobium content using an appropriate production process and appropriate processing of the casting wheel surface during production.

[0153] Examples 9 to 11 according to the invention illustrate that it is also possible to produce wide, thick strips with an increased iron content and reduced niobium content at the highest casting-wheel material heat conductivities and with a sufficiently low level of impurities.

[0154] As has already been mentioned above, surface crystallisation is characterised by the formation of crystalline grains that are different in the amorphous and the nanocrystalline microstructure of the core of the metal strip. These crystalline grains can be detected using transmission electron microscopy, for example.

[0155] FIG. 2a shows a a dark-field transmission electron microscope image of a sample of an amorphous metal strip with surface crystallisation. The region of the image below the blue line, labelled Pt, is merely a platinum layer applied for the purposes of sample preparation. The upper part of the image shows the amorphous microstructure of the sample by means of a uniform grey coloration. Below it, at the surface of the sample, two individual crystalline grains embedded in the amorphous microstructure with a grain size of approx. 140 nm can be seen. The arrows represent the crystalline orientation of the grain as determined by means of electron diffraction measurements.

[0156] Readily visible directly below the surface are two crystalline grains with a grain size of approx. 140 nm.

[0157] FIG. 2b shows a dark-field TEM image of the same sample in the heat-treated nanocrystalline state. The amorphous sample is subjected to an appropriate heat treatment in order that the nanocrystalline microstructure is formed, that is clearly visible in FIG. 2b. Here the grains at the surface have grown into a continuous layer approx. 150 nm (one grain) deep with grain sizes of up to 300 nm. The upper portion of the image shows the nanocrystalline microstructure, each dot representing a grain with a grain size of approx. 15 nm. It should be noted that the two images have different length scales. A continuous crystalline layer can now been seen on the sample surface. It is exactly one grain length deep and so projects approx. 150 nm into the sample. The grain size of this layer is approx. 300 nm.

[0158] Though it is possible to detect surface crystallisation by means of transmission electron microscopy, this requires a long and complex sample preparation process. Here it is, therefore, powder diffractometry that is used as the detection method for measuring and evaluating surface crystallisation in metal strips,

[0159] A polycrystalline material in the individual grains exhibits statistically distributed crystal orientations similar to a powder, and can therefore be treated as such.

[0160] The measurements are carried out using K radiation from a copper anode in a Bragg-Brentano arrangement and an angular range of 2=20 to 2=125. The measuring spot is approx. 10 mm in diameter and measurements are taken directly on the untreated surface of the strip. Copper K radiation with a wavelength of 1.54 has substantially less energy than molybdenum K radiation with a wavelength of 0.71 , for example. This ensures that the penetration depth is not too great for the measurement of surface crystallisation.

[0161] Evaluation of the penetration depth of the copper K radiation for Fe3Si gives d(1/e)=6.29 m, i.e. radiation falls to a faction of 1/e after approx. 6 m. For molybdenum, in contrast, the penetration depth is d(1/e)=46.22 m. With a foil thickness of approx. 20 m, it is therefore possible to observe the entire sample volume in the exposed region of approximately one square centimetre with this radiation. However, since the crystalline surface layer represents no more than approx. 1.5% of the total volume (20.15 m), it cannot be resolved in a measurement of the whole sample volume. This is different for copper K, where only the volume close to the surface contributes to the material response. The fact that surface crystallisation can only be detected using powder diffractometry when appropriate radiation is used may explain the lack of testing of this important material property.

[0162] The diffractogram of a complete amorphous sample shows no sharp reflexes, merely blurred reflexes that correspond to the associated crystalline microstructure in what are roughly the strongest reflex layers. FIG. 3 shows a diffractogram of this type for a sample of VITROPERM 800 (nominal composition in at. %: Fe.sub.73.9Cu.sub.1Nb.sub.3Si.sub.15.5B.sub.6.6)

[0163] The diffractogram in FIG. 3 shows the blurred, amorphous halo at a scattering angle of 2=44.7. The second, clearly smaller halo is also visible at roughly 2=82. The so-called amorphous halo for a typical nanocrystalline iron-based alloy has a scattering angle of 2=447

[0164] Completely crystalline, cubic, body-centred iron shows the strongest reflex, belonging to the (110) planes, at 2=44.674 (100%), followed by the reflex of the (211) planes at 2=82.335 (30.0%) and the (200) planes at 2=65.023 (20.0%). Fe3Si, also cubic and body-centred, shows the strongest reflex of the (220) planes at 2=45.237 (100%), followed by the reflex of the (422) planes at 2=83.536 (21.3%) and the (400) planes at 2=65.902 (128%).

[0165] FIG. 4 shows a typical diffractogram for a completely nanocrystalline sample in which approx. 80% of the volume is nanocristallites with static orientation. The amorphous portion is visible only as a slight widening in the trailing flanks of the reflex. This amorphous portion can be separated using the methods described in EP 1260812, for example.

[0166] The diffractogram in FIG. 4 shows the sharp reflexes typical of a polycrystalline, cubic, body-centred crystal structure with statistical grain distribution. The amorphous portion is visible only as a slight widening in the trailing flanks of the reflexes. The strongest reflex layers for Fe3Si occur at:

(220) plane: 2=45.237 (100%),
(422) plane: 2=83.536 (21.3%) and
(400) plane: 2=65.902 (12.8%).

[0167] FIG. 5a shows a diffractogram of an amorphous sample with little surface crystallisation and FIG. 5b shows a diffractogram of an amorphous sample with very marked (b) surface crystallisation.

[0168] In addition to the amorphous halo already seen in FIG. 3, FIG. 5a also shows a strong crystalline reflex at 2=65.9. This corresponds to cubic, body-centred crystallites completely oriented in the (100) direction. This is referred to as the texture or crystallographic texture of a material.

[0169] FIG. 5b shows an amorphous sample with surface crystallisation. This diffractogram shows only the (400) reflex in addition to the amorphous microstructure. The surface crystallites are therefore very strongly textured in the (100) direction. This is confirmed by electron diffraction measurements. For this reason it is easy to separate the crystalline portion and the amorphous portion.

[0170] All the alloys in these examples contain silicon. The same considerations can, however, also be applied to silicon-free alloys.

[0171] The ratio between the surface portion of the crystalline reflex and the total surface portion of the amorphous halo and the crystalline reflex is regarded as a measure of surface crystallinity, with only the host halo of the amorphous microstructure being taken into account. Though the result corresponds in theory to the content by volume, it is strongly affected by the exposed volume and is consequently only comparable with an identical measurement setup.

[0172] The estimation of surface crystallisation in the nanocrystalline strip is somewhat less precise. FIG. 6 shows a diffractogram of a nanocrystalline sample with marked surface crystallisation. This diffractogram corresponds to the one shown in FIG. 4, but with a considerably higher (400) reflex. This is due to the textured crystalline portion in the volume considered, which is caused by surface crystallisation. The reflexes to the left of the (220) reflex and the (400) reflex are artefacts of the measurement (due to the inadequate filtering of K radiation).

[0173] The significantly increased (400) reflex is clearly visible and can also be used to define a measure of surface crystallinity by determining the ratio between the area under the (400) reflex less 20% and the area under the (220) reflex and the total (400) reflex. This measure is comparable with the measure defined for the amorphous state, though less precise. This is firstly because only the strongest reflexes are used to calculate the total surface, while the portion of other reflexes in the diffractogram is ignored. Secondly, strictly speaking the deduction of 20% is only applicable to silicon-free samples. In samples containing silicon, the portion of surface crystallinity is underestimated to the extent of the silicon content of the crystalline grains formed.

[0174] FIG. 7 shows the diffractogram of an amorphous sample with surface crystallisation. The amorphous sample has the composition Fe.sub.81.1Co.sub.4Cu.sub.0.8Si.sub.0.5B.sub.9.54P.sub.3.94C.sub.0.12 (figures in at. %) with a sulphur impurity of 0.005 at. %. This diffractogram is realised in the same manner as the one shown in FIG. 4a, but the surface crystallisation has no texture. For this reason the strongest crystalline reflex in the surface crystallisation is the (220) reflex, coinciding with the amorphous halo as would be expected with a statistical orientation distribution. The (422) and (400) reflexes are suggested. Here, too, the extent of surface crystallisation determined by calculating the area under the (220) peak and the amorphous halo. To adjust the curves, the fold in the two overlapping peaks is taken into account. It should be noted that it would be impossible to determine the level of surface crystallisation in the nanocrystalline state here since it would be no different in crystallographic terms from the surface crystallisation, which is untextured in this case.

[0175] This powder diffractometry method is used to examine a plurality of samples in order to examine the influence of various production parameters on surface crystallisation.

[0176] FIG. 8 shows a graph of the extent of surface crystallisation comparing the free side and the side facing the casting wheel. The top diagram shows the entire graph, the bottom diagram a section of it providing better resolution for the region under 0.15. FIG. 8 shows the standardised number of over 400 samples examined that have a corresponding proportion of surface crystallisation. The compositions of these 400 samples range from amorphous soft magnetic alloys via nanocrystalline compositions to brazing alloys. All compositions are within the range indicated above.

[0177] It shows that no surface crystallisation proportions above 35% are measured for the side facing the casting wheel, while levels of up to 100% are found for the air-facing side. At 76.2%, the number of samples without surface crystallisation on the side facing the casting wheel is somewhat larger than that on the air side, which stands at 69.5%. In addition, no proportion of surface crystallisation of higher than 35% (fraction of 0.35) are measured on the casting-wheel side, while surface crystallisation proportions of up to 100% (fraction of 1) are observed on the air side. Overall, surface crystallisation occurred on approximately half of the samples examined (46.8%).

[0178] In the amorphous state, rapidly solidified, metallic strips exhibit very good mechanical properties. In a two-point bend test, for example, they exhibit elongations at fracture of up to 100% and hardnesses of approx. 10 CPa measured using a nanoindenter with a three-sided diamond Syntori Berkovich tip. It has been found that the elongation at fracture drops by up to two orders of magnitude when crystallisation occurs in the material. This embrittlement can also be observed in the presence of surface crystallisation in otherwise amorphous material, though to a less marked extent. Surface crystallisation should therefore be avoided in order to achieve the problem-free mechanical further processing of the strips.

[0179] Amorphous and nanocrystalline materials have a small coercive field, as is required for good soft magnetic materials, as long as the grain sizes of possible crystalline regions are below the exchange interaction length and the orientation of the crystallites is statistical distributed. In such cases, crystal anisotropy averages out and has no disrupting, macroscopic influence.

[0180] FIG. 9 shows a diagram of the relationship between coercive field strength and grain size taken from the Handbook of Magnetic Materials, G. Herzer, Vol. 10, Chapter 3. It shows that the highest coercive field strengths are measured for grain sizes of between 100 nm and 300 nm. This corresponds to the grain size of a Typical surface crystallisation, which in this case is also generally strongly textured.

[0181] This makes it possible to reliably provide good magnetic properties when surface crystallisation is limited and surface crystallisation is preferably avoided as far as is possible.

[0182] FIG. 10 shows AC permeability measured with two-way sine modulation at 1.5 A/m and 50 Hz, so-called dynamic permeability Dyn measured at a sine modulation of 0.3 A/m and 100 kHz as a function of surface crystallisation, the maximum air and casting-wheel side values being applied. Although both the measurement of surface crystallinity and the determination of dynamic permeability are subject to significant measurement inaccuracies, the negative influence of surface crystallinity on the magnetic characteristics of the material is clearly visible.

[0183] The top image shows that it is still entirely possible to achieve a dynamic permeability of greater than 150000 with a surface crystallisation proportion of 23%.

[0184] FIG. 9 reveals that a surface crystallisation proportion of up to 23% is tolerable without a significant reduction in soft magnetic performance. Brittleness tests confirm this result, the figure of 23% relating to the measuring method described above using X-ray powder diffractometry with copper K radiation.

[0185] FIG. 11 shows a graph of surface crystallisation as a function of average strip thickness. It shows that surface crystallisation increases with rising strip thickness. Owing to its low metalloid content in general and its low niobium content in particular, the present alloy tends towards higher levels of surface crystallisation. It was chosen in order to make the connection between thickness and surface crystallisation visible even at standard thicknesses of approx. 20 m. The two green dots correspond to embodiments 9 and 10 and show that it is also possible, at high casting-wheel material thermal conductivity and optimum production, to produce this alloy free of surface crystallisation in the standard thickness.

[0186] To summarise, an amorphous or nanocrystalline metal strip is provided that has surface crystallisation proportion of less than 23%, preferably less than 5%, preferably 0%, the proportion of surface crystallisation being determined using the powder diffractometry process described above. An amorphous strip of this type is produced using a rapid solidification technology. A nanocrystalline strip can be formed from the amorphous strip by heat treating the amorphous strip.

[0187] When producing the amorphous metal strip, the cooling rate of the solidification of the molten mass or melt is adjusted and maintained over the duration of the casting process such that the formation of crystallisation at the surface of the metal strip, and in particular in a surface layer on the air side of the metal strip, is prevented as far as possible over the length and width of the metal strip. In this way, longer amorphous metal strips can be reliably provided on an industrial-scale and with reliable mechanical and, depending on composition, magnetic properties.