Alloy, magnet core and method for producing a strip from an alloy

Abstract

An alloy of Fe.sub.100-a-b-c-d-x-y-zCu.sub.aNb.sub.bM.sub.cT.sub.dSi.sub.xB.sub.yZ.sub.z and up to 1 atomic % impurities; M is one or more of Mo or Ta, T is one or more of V, Cr, Co or Ni and Z is one or more of C, P or Ge, wherein 0.0 atomic %a<1.5 atomic %, 0.0 atomic %b<3.0 atomic %, 0.2 atomic %c4.0 atomic %, 0.0 atomic %d<5.0 atomic %, 12.0 atomic %<x<18.0 atomic %, 5.0 atomic %<y<12.0 atomic % and 0.0 atomic %z<2.0 atomic %, and wherein 2.0 atomic %(b+c)4.0 atomic %, produced in the form of a strip and having a nanocrystalline structure in which at least 50% by volume of the grains have an average size of less than 100 nm, a remanence ratio J.sub.r/J.sub.s<0.02, J.sub.r being the remanent polarization and J.sub.s being the saturation polarization, and a coercitive field strength H.sub.c which is less than 1% of the anisotropic field strength H.sub.a and/or less than 10 A/m.

Claims

1. An alloy having a composition consisting of Fe.sub.100-a-b-c-d-x-y-zCu.sub.aNb.sub.bM.sub.cT.sub.dSi.sub.xB.sub.yZ.sub.z and up to 1 atomic % impurities, wherein M is Mo and/or Ta, T is one or more of the elements V, Cr, Co or Ni and Z is one or more of the elements C, P or Ge, and wherein 0.0 atomic %a<1.5 atomic %, 0.0 atomic %b<3.0 atomic %, 0.2 atomic %c4.0 atomic %, 0.0 atomic %d<5.0 atomic %, 12.0 atomic %<x<18.0 atomic %, 5.0 atomic %<y<12.0 atomic %, 0.0 atomic %z<2.0 atomic % and 2.0 atomic %(b+c)4.0 atomic %, wherein the alloy is in the form of a strip, wherein the alloy comprises a nanocrystalline structure, at least 50% by volume of the grains having an average size of less than 100 nm, wherein the alloy has a remanence ratio J.sub.r/J.sub.s<0.02, J.sub.r being the remanent polarisation and J.sub.s being the saturation polarisation, wherein the alloy has a coercitive field strength H.sub.c which is less than 1% of the anisotropic field strength H.sub.a, wherein the strip is heat-treated in a continuous process at a annealing temperature between 450 C. and 750 C. under a tension of 5 MPa to 1000 MPa with a dwell time of 2 seconds to 2 minutes, and wherein the remanent polarisation J.sub.r, the saturation polarization J.sub.s, the coercitive field strength H.sub.c and/or the anisotropic field strength H.sub.a or permittivity of the strip are continuously measured as the strip leaves a continuous furnace, and if a deviation from a permitted deviation range of the remanent polarisation J.sub.r, the saturation polarization J.sub.s, the coercitive field strength H.sub.c, and/or the anisotropic field strength H.sub.a or permittivity is detected, the tension applied to the strip is adjusted to bring the remanent polarisation J.sub.r, the saturation polarization J.sub.s, the coercitive field strength H.sub.c, and/or the anisotropic field strength H.sub.a or permittivity measured to be outside the permitted deviation range within the permitted deviation range.

2. The alloy according to claim 1, wherein the remanence ratio J.sub.r/J.sub.s is <0.01.

3. The alloy according to claim 1, wherein the hysteresis loop of the alloy has a nonlinearity factor NL, NL being <0.5%, and
NL=100/2(J.sub.auf+J.sub.ab)/J.sub.s wherein J.sub.auf is the standard deviation of the magnetic polarisation from a regression line through the ascending branch of the hysteresis loop between polarisation values of 75% of the saturation polarisation J.sub.s and J.sub.ab is the standard deviation of the magnetic polarisation from a regression line through the descending branch of the hysteresis loop between polarisation values of 75% of the saturation polarisation J.sub.s.

4. The alloy according to claim 1, wherein the alloy has a permeability between 40 and 10000.

5. The alloy according to claim 1, wherein the alloy has a saturation magnetostriction of less than 1 ppm.

6. The alloy according to claim 1, wherein the alloy has a saturation polarisation J.sub.s that is 1.22 T and the coercitive field strength H.sub.c is 8 A/m.

7. The alloy according to claim 1, wherein 0.0 atomic %b<2.5 atomic %.

8. The alloy according to claim 1, wherein 2.1 atomic %(b+c)3.0 atomic %.

9. The alloy according to claim 1, wherein 0.0 atomic %d<2.0 atomic %.

10. The alloy according to claim 1, wherein 14.0 atomic %<x<17 atomic % and 5.5 atomic %<y<8.0 atomic %.

11. The alloy according to claim 1, wherein the strip is heat-treated in the continuous process under a tension of 10 MPa to 250 MPa with a dwell time of 2 seconds to 2 minutes.

12. The alloy according to claim 1, wherein the strip is heat-treated in the continuous process under a tension of 250 MPa to 1000 MPa with a dwell time of 2 seconds to 2 minutes.

13. A magnet core made from an alloy according to claim 1.

14. The magnet core according to claim 13, having the form of a wound strip.

15. The magnet core according to claim 13, wherein the strip has an oxide layer with a thickness of <0.2 m on its surface.

16. The magnet core according to claim 13, wherein the strip is coated with an additional insulating layer.

17. The alloy according to claim 1, wherein the minimum niobium content is 1.8 atomic % and the minimum Mo content is 0.2 atomic %.

18. The alloy according to claim 1, wherein the alloy does not contain any tantalum, except as a possible impurity.

19. The alloy according to claim 1, wherein M is Mo and 1.8 atomic %b<3.0 atomic %.

20. The alloy according to claim 1, wherein 0.0 atomic %<b<2.5 atomic % and 2.1 atomic %(b+c)<3.0 atomic %.

21. The alloy according to claim 1, wherein the alloy has a permeability in the range of 50 to 200.

22. The alloy according to claim 1, wherein the alloy has a coercitive field strength H.sub.c which is less than 10 A/m.

23. A method for producing a strip, comprising the following: providing a strip from an amorphous alloy with a composition consisting of Fe.sub.100a-b-c-d-x-y-zCu.sub.aNb.sub.bM.sub.cT.sub.dSi.sub.xB.sub.yZ.sub.z and up to 1 atomic % impurities, wherein M is Mo and/or Ta, T is one or more of the elements V, Cr, Co or Ni and Z is one or more of the elements C, P or Ge, and wherein 0.0 atomic %a<1.5 atomic %, 0.0 atomic %b<3.0 atomic %, 0.2 atomic %c4.0 atomic %, 0.0 atomic %d<5.0 atomic %, 12.0 atomic %<x<18.0 atomic %, 5.0 atomic %<y<12.0 atomic %, 0.0 atomic %z<2.0 atomic % and 2.0 atomic %(b+c)4.0 atomic %, wherein the alloy has a remanence ratio J.sub.r/J.sub.s<0.02, J.sub.r being the remanent polarisation and J.sub.s being the saturation polarisation, and the alloy has a coercitive field strength H.sub.c which is less than 1% of the anisotropic field strength H.sub.a, heat treating the strip under a tension of 5 MPa to 1000 MPa with a dwell time of 2 seconds to 2 minutes in a continuous process at an annealing temperature T.sub.a, wherein 450 C.T.sub.a750 C., continuously measuring the remanent polarisation J.sub.r, the saturation polarization J.sub.s, the coercitive field strength H.sub.c and/or the anisotropic field strength H.sub.a or permittivity of the strip as the strip leaves a continuous furnace, and if a deviation from a permitted deviation range of the remanent polarisation J.sub.r, the saturation polarization J.sub.s, the coercitive field strength H.sub.c and/or the anisotropic field strength H.sub.a or permittivity is detected, adjusting the tension applied to the strip to bring the remanent polarisation J.sub.r, the saturation polarization J.sub.s, the coercitive field strength H.sub.c and/or the anisotropic field strength H.sub.a or permittivity measured to be outside the permitted deviation range within the permitted deviation range.

24. The method according to claim 23, wherein the strip is heat-treated in the continuous furnace.

25. The method according to claim 24, wherein the strip is pulled through the continuous furnace with a speed s, so that a dwell time of the strip in a temperature zone of the continuous furnace at the temperature T.sub.a is between 2 seconds and 2 minutes.

26. The method according to claim 23, wherein the strip is heat-treated in the continuous furnace under a tension of 5 MPa to 1000 MPa.

27. The method according to claim 26, wherein the strip is heat-treated in the continuous furnace under a tension of 10 MPa to 250 MPa.

28. The method according to claim 26, wherein the strip is heat-treated in the continuous furnace under a tension of 250 MPa to 1000 MPa.

29. The method according to claim 23, further comprising: predetermining a desired value of the anisotropic field strength H.sub.a or the permeability and/or a maximum value of the remanence ratio J.sub.r/J.sub.s of less than 0.02 and/or a maximum value of the coercitive field strength H.sub.c which is less than 1% of the anisotropic field strength H.sub.a and/or less than 10 A/m, as well as the permitted deviation range for each of these values.

Description

BRIEF DESCRIPTION OF DRAWINGS

(1) Embodiments will now be explained in greater detail with reference to the following examples, tables and drawings.

(2) FIG. 1 shows the hysteresis loops of an alloy according to the invention which is heat-treated under two different tensions,

(3) FIG. 2 shows magnetic properties for alloys according to the invention with various Nb and Mo contents, produced at different annealing temperatures,

(4) FIG. 3 shows magnetic properties for alloys produced at different tensions, and

(5) FIG. 4 is a diagrammatic view of a continuous furnace.

(6) Table 1 lists the magnetic properties for various alloys according to the invention and for comparative examples,

(7) Table 2 shows further alloy examples and their magnetic properties, and

(8) Table 3 lists crystallisation temperatures T.sub.x1 and T.sub.x2 (DSC 10 K/min, peak) and annealing temperatures T for three alloys from Table 1.

DETAILED DESCRIPTION OF SPECIFIC EMBODIMENTS

(9) Various alloys based on Fe.sub.100-a-b-c-d-x-y-zCu.sub.aNb.sub.bM.sub.cT.sub.dSi.sub.xB.sub.yZ.sub.z are produced in the form of an amorphous strip. Typical strips have a width of 6 mm to 10 mm and a thickness of 17 m to 25 m. The amorphous strip can for example be produced in the desired composition by means of a rapid solidification technology. These amorphous strips are then heat-treated to produce a nanocrystalline structure and the desired magnetic properties.

(10) In alloys based on Fe.sub.100-a-b-c-d-x-y-zCu.sub.aNb.sub.bM.sub.cT.sub.dSi.sub.xB.sub.yZ.sub.z, the reduction of the Nb content is desirable in order to reduce raw material costs without at the same time increasing the coercitive field strength too much. Below, it is disclosed that this can be achieved by wholly or partially replacing Nb by Mo or Ta, wherein the total content of the elements from the Nb and/or Mo and/or Ta group(s) is at least 2 atomic % and the niobium content is less than 3 atomic % or less than 2 atomic %.

(11) Table 1 shows the saturation polarisation J.sub.s as measured in the production state and the values for saturation magnetostriction s, nonlinearity NL, remanence ratio J.sub.r/J.sub.s, coercitive field strength H.sub.c, anisotropic field strength H.sub.a and relative permeability as measured after heat treatment under a tension of 5010 MPa for various alloy compositions. Composition data are given in atomic percent.

(12) The heat treatment was performed under a tension of 5010 MPa for a duration of 4 seconds in the case of comparative examples (a) and (i) and for a duration of 6 seconds in the case of comparative examples (ii) and (iii) and in the case of examples 1 to 10 according to the invention at the annealing temperatures T given in the table. Examples 1 to 10 in Table 1 all have a reduced Nb content of less than 2 atomic %.

(13) In the alloy examples 1, 2 and 3, Nb is completely replaced by various Mo contents. For Mo contents from 2 atomic %, the coercitive field strength is less than 8 A/m, decreasing further with increasing Mo contents.

(14) In the alloy examples 4 and 5, Nb is completely replaced by various Ta contents. For Ta contents around 2 atomic %, the coercitive field strength, being H.sub.c=3 A/m, is comparable to the comparative examples, but magnetic saturation polarisation J.sub.s is higher.

(15) One advantage of Ta and Mo over Nb is their better availability on the world market. Ta has the advantage of being more effective in reducing coercitive field strength, in particular compared to Mo. The high raw material costs of Ta are a disadvantage, however. In view of this, attempts were made to replace Nb only partially, if possible with Mo.

(16) In alloy example 6, the major part of Nb was replaced by Mo and Ta. Here, too, coercitive field strength values are comparable to the comparative examples, combined with a higher magnetic saturation polarisation J.sub.s.

(17) In the alloy examples 7 to 10, Nb is partially replaced by Mo. Here, too, coercitive field strength values are markedly less than 10 A/m, if the total content of Nb and Mo is at least 1.9 atomic %. If the composition approaches the lower limit, it is advantageous if the Nb content is slightly higher than the Mo content.

(18) Table 2 shows further alloy examples 11 and 12 and their magnetic properties after a heat treatment of 6 s under a tension of 5010 MPa at the annealing temperature given in the table.

(19) The magnetic properties demonstrate that the addition of Mo and Ta is possible if the Nb content is higher than 2 atomic %. Alloy example 11, for instance, indicates that even a minor addition of 0.2 atomic % Mo combined with a reduction of the Nb content by 0.3 atomic % results in a slight H.sub.c reduction compared to comparative example (a) from Table 1, the saturation polarisation J.sub.s being advantageously increased by about 15%. In alloy example 12, Nb is substituted by Ta, resulting in magnetic properties comparable to those of example (a) from Table 1, if the alloy is heat-treated using a suitable tension at a suitable annealing temperature.

(20) Further embodiments are disclosed in FIGS. 1, 2 and 3.

(21) FIG. 1 shows a typical hysteresis loop which results from heat treatment under tension. FIG. 1 shows the quasi-static hysteresis loop of the alloy Fe.sub.74.7Cu.sub.0.8Nb.sub.1.4Mo.sub.1Si.sub.15.5B.sub.6.6 after a heat treatment of 6 s at 650 C. with two different tensions, wherein .sub.a150 MPa and .sub.a2140 MPa.

(22) FIG. 1 further illustrates the definition of the magnetic saturation polarisation J.sub.s, of the anisotropic field strength H.sub.a, of the coercitive field strength H.sub.c and of the remanent polarisation J.sub.r. For an alloy according to the invention, the coercitive field strength should, at an anisotropic field strength H.sub.a of approximately 1000 A/m, be less than 10 A/m, i.e. less than approximately 1% of H.sub.a. Such low values are difficult to measure at full modulation of the hysteresis loop (measuring accuracy approximately 1/Am) and therefore hardly identifiable with the bare eye in FIG. 1. Nevertheless, if remagnetisation losses are to be minimised, it is advisable to keep to such low values.

(23) A characteristic of the hysteresis loop is its linearity in the centre of the hysteresis loop. A measure for this is a low remanence ratio J.sub.r/J.sub.s.

(24) FIG. 2 shows the saturation magnetostriction s, the anisotropic field strength H.sub.a, the coercitive field strength H.sub.c and the remanence ratio J.sub.r/J.sub.a as a function of the annealing temperature T for Fe.sub.77.1-x-yCu.sub.0.8Nb.sub.xMo.sub.ySi.sub.15.5B.sub.6.6 with two different Nb contents and increasing Mo contents after a heat treatment of approximately 6 seconds under a tension of approximately 50 MPa. The compositions involved are Fe.sub.75.6Cu.sub.0.8Nb.sub.1Mo.sub.0.5Si.sub.15.5B.sub.6.6, Fe.sub.75.1Cu.sub.0.8Nb.sub.1Mo.sub.1Si.sub.15.5B.sub.6.6, Fe.sub.74.6Cu.sub.0.8Nb.sub.1Mo.sub.1.5Si.sub.15.5B.sub.6.6, Fe.sub.75.7Cu.sub.0.8Nb.sub.1.4Si.sub.15.5B.sub.6.6, Fe.sub.75.2Cu.sub.0.8Nb.sub.1.4Mo.sub.0.5Si.sub.15.5B.sub.6.6 and Fe.sub.74.7Cu.sub.0.8Nb.sub.1.4Mo.sub.1Si.sub.15.5B.sub.6.6.

(25) The desired magnetic properties, i.e. a low saturation magnetostriction s, a defined anisotropic field strength H.sub.a, a low coercitive field strength H.sub.c and a low remanence ratio J.sub.r/J.sub.s, are obtained in a specific annealing window which is characteristic for the respective alloy and which is characterised by a minimum annealing temperature T.sub.1 and a maximum annealing temperature T.sub.2. This annealing range can be determined by a standard measurement of the crystallisation temperatures T.sub.x1 and T.sub.x2, for example by means of DSC (differential scanning calorimetry), allowing the annealing temperature T to be defined.

(26) Table 3 shows crystallisation temperatures T.sub.x1 and T.sub.x2 (DSC 10 K/min, peak) and suitable annealing temperatures T for the alloy Fe.sub.75.5-xCu.sub.0.8Nb.sub.1.4Mo.sub.xSi.sub.15.5B.sub.6.6 for annealing times of approximately 6 seconds. The example number corresponds to the alloy composition given in Table 1. Table 3 shows by way of example the context for the annealing time of approximately 6 seconds used here.

(27) The results from FIG. 2 make clear that the saturation magnetostriction and the anisotropic field strength behave approximately in the same way in all examples, while there are noticeable differences in coercitive field strength and remanence ratio.

(28) Complementing Table 1, FIG. 2 discloses that alloys with a total (Nb+Mo) content from approximately 2 atomic % (which includes 1.9 atomic %) have in a wide annealing temperature range a coercitive field strength significantly lower than 10 A/m. Alloys with a total (Nb+Mo) content>2.3 atomic % exhibit with H.sub.c=5 A/m even better values within a large range, which furthermore react less sensitively to the precise annealing temperature. Compared to this, alloys with an (Nb+Mo) content typically have a coercitive field strength between 10 and 20 A/m and correspondingly high hysteresis losses. In addition, H.sub.c changes relatively markedly with the annealing temperature.

(29) The above examples relate to a annealing tension .sub.a of approximately 50 MPa. FIG. 3 shows the effect of this annealing tension on magnetic values.

(30) FIG. 3 shows the relative permeability the anisotropic field strength H.sub.a, the coercitive field strength H.sub.c, the remanence ratio J.sub.r/J.sub.a and the nonlinearity factor of the alloys Fe.sub.75.7-yCu.sub.0.8Nb.sub.1.4Mo.sub.ySi.sub.15.5B.sub.6.6 with y=0.5 atomic % and y=1 atomic % after a heat treatment of 6 seconds at 640 C. for Mo=0.5 atomic % or at 650 C. for Mo=1 atomic % compared to Fe.sub.75.5Cu.sub.1Nb.sub.1Si.sub.15.5B.sub.6.5 at a heat treatment of 4 seconds at 610 C. as a function of the tension .sub.a applied during the heat treatment.

(31) FIG. 3 discloses that the anisotropic field strength H.sub.a increases proportionally with the tension applied during the heat treatment, while the permeability is reduced inversely proportionally to .sub.a. The annealing tension .sub.a is finally selected such that a predefined value for permeability and anisotropic field strength is set. In this respect, all of the alloy examples shown behave in a similar way, while there are noticeable differences in coercitive field strength, i.e. in hysteresis losses. The alloys according to the invention exhibit even better coercitive field strength values at increased annealing tensions. For example, while in an alloy with 1.5 atomic % Nb the coercitive field strength increases noticeably with the tension applied, an Mo addition of only 0.5 atomic % reduces the tension-dependence of H.sub.c, thereby effecting an improvement. This applies correspondingly to an addition of 1 atomic %, which has even better effects. This also applies to lower annealing tensions, which are used if a lower anisotropic field strength and permeability values equal to or higher than 2000 are to be set.

(32) FIG. 4 is a diagrammatic view of an apparatus 1 suitable for producing the alloys with a composition according to any of the above embodiments in the form of a strip. The apparatus 1 comprises a continuous furnace 2 with a temperature zone 3 which is adjusted such that the temperature in the furnace within this zone is within 5 C. of the annealing temperature T. The apparatus 1 further comprises a reel 4 on which the amorphous alloy 5 is wound and a take-up reel 6 which receives the heat-treated strip 7. The strip 7 is pulled by the reel 4 through the continuous furnace 2 to the take-up reel 6 with a speed s. In this process, the strip is subjected to a tension .sub.a in the running direction from the device 9 to the device 10.

(33) The apparatus 1 further comprises a device 8 for continuously measuring the magnetic properties of the strip 6 after it has been heat-treated and pulled out of the continuous furnace 2. In the region of this device 8, the strip 7 is no longer subjected to tension. The measured magnetic properties can be used for adjusting the tension .sub.a under which the strip 7 is pulled through the continuous furnace 2. This is indicated diagrammatically in FIG. 13 by arrows 9 and 10. By this measuring of the magnetic properties and the continuous tension adjustment, the uniformity of the magnetic properties along the length of the strip can be improved.

(34) TABLE-US-00001 TABLE 1 Composition J.sub.s T.sub.a .sub.s NL H.sub.c H.sub.a No. (atomic %) (T) ( C.) (ppm) (%) J.sub.r/J.sub.s (A/m) (A/m) (a) Fe.sub.74Cu.sub.1Nb.sub.3Si.sub.15.5B.sub.6.5 1.21 690 0.1 0.3 0.004 3 850 1130 (i) Fe.sub.75.5Cu.sub.1Nb.sub.1.5Si.sub.15.5B.sub.6.5 1.34 635 0.6 0.6 0.008 13 1180 890 (ii) Fe.sub.75.7Cu.sub.0.8Nb.sub.1.4Si.sub.15.5B.sub.6.6 1.36 625 0.4 0.7 0.011 11 1000 1085 (iii) Fe.sub.75.6Cu.sub.0.8Nb.sub.1Mo.sub.0.5Si.sub.15.5B.sub.6.6 1.37 625 0.5 0.7 0.013 13 1000 1085 1 Fe.sub.75.1Cu.sub.0.8Mo.sub.2Si.sub.15.5B.sub.6.6 1.30 625 0.5 0.2 0.010 7 1170 880 2 Fe.sub.74.1Cu.sub.0.8Mo.sub.3Si.sub.15.5B.sub.6.6 1.23 655 0.06 0.5 0.006 6 1000 980 3 Fe.sub.73.1Cu.sub.0.8Mo.sub.4Si.sub.15.5B.sub.6.6 1.14 640 0.2 0.07 0.003 3 1020 945 4 Fe.sub.75.1Cu.sub.0.8Ta.sub.2Si.sub.15.5B.sub.6.6 1.31 640 0.3 0.10 0.003 3 1080 885 5 Fe.sub.74.1Cu.sub.0.8Ta.sub.3Si.sub.15.5B.sub.6.6 1.23 640 0.4 0.07 0.002 2 1010 950 6 Fe.sub.74.1Cu.sub.0.8Nb.sub.1Mo.sub.1Ta.sub.1Si.sub.15.5B.sub.6.6 1.24 640 0.2 0.09 0.004 4 990 965 7 Fe.sub.74.6Cu.sub.0.8Nb.sub.1Mo.sub.1.5Si.sub.15.5B.sub.6.6 1.27 650 0.4 0.3 0.004 4 930 1095 8 Fe.sub.74.7Cu.sub.0.8Nb.sub.1.4Mo.sub.1Si.sub.15.5B.sub.6.6 1.28 650 0.06 0.1 0.002 2 960 1060 9 Fe.sub.75.2Cu.sub.0.8Nb.sub.1.4Mo.sub.0.5Si.sub.15.5B.sub.6.6 1.32 640 0.4 0.3 0.003 3 1000 1040 10 Fe.sub.75.1Cu.sub.0.8Nb.sub.1Mo.sub.1Si.sub.15.5B.sub.6.6 1.31 625 0.6 0.3 0.005 6 1025 1020 (a) comparative example (i), (ii), (iii) comparative example (1)-(10) examples according to the invention

(35) TABLE-US-00002 TABLE 2 Composition J.sub.s T.sub.a .sub.s NL H.sub.c H.sub.a No. (atomic %) (T) ( C.) (ppm) (%) J.sub.r/J.sub.s (A/m) (A/m) 11 Fe.sub.74.2Cu.sub.0.8Nb.sub.2.7Mo.sub.0.2Si.sub.15.5B.sub.6.6 1.24 640 0.7 0.1 0.002 2 930 1025 12 Fe.sub.74.0Cu.sub.0.8Nb.sub.2.2Ta.sub.0.9Si.sub.15.5B.sub.6.6 1.22 675 0.1 0.1 0.003 4 970 980

(36) TABLE-US-00003 TABLE 3 No. Mo (atomic %) T.sub.x1 ( C.) T.sub.x1 ( C.) Annealing temperature T (ii) 0 488 645 540 C. to 630 C. 9 0.5 498 662 550 C. to 650 C. 8 1.0 505 678 550 C. to 670 C.