Method for fabricating a functionally-graded monolithic sintered working component for magnetic heat exchange and an article for magnetic heat exchange

Abstract

An article for magnetic heat exchange includes a functionally-graded monolithic sintered working component including La.sub.1-aR.sub.a(Fe.sub.1-x-yT.sub.yM.sub.x).sub.13H.sub.zC.sub.b with a NaZn.sub.13-type structure. M is one or more of the elements from the group consisting of Si and Al, T is one or more of the elements from the group consisting of Mn, Co, Ni, Ti, V and Cr and R is one or more of the elements from the group consisting of Ce, Nd, Y and Pr. A content of the one or more elements T and R, if present, a C content, if present, and a content of M varies in a working direction of the working component and provides a functionally-graded Curie temperature. The functionally-graded Curie temperature monotonically decreases or monotonically increases in the working direction of the working component.

Claims

1. An article for magnetic heat exchange, comprising: a monolithic sintered working component comprising La.sub.1-aR.sub.a(Fe.sub.1-x-yT.sub.yM.sub.x).sub.13H.sub.zC.sub.b with a NaZn.sub.13 structure, wherein M is one or more element selected from the group consisting of Si and Al, T is one or more element selected from the group consisting of Mn, Co, Ni, Ti, V and Cr, R is one or more element selected from the group consisting of Ce, Nd, Y, and Pr, 0a0.5, 0b1.5, 0.05x0.2, 0y0.2, and 0z3, a content of at least one of the elements selected from the group consisting of T, R, C, and M varying in a working direction of the working component, a Curie temperature monotonically decreasing or monotonically increasing in the working direction of the working component, and the working component comprising an average gradient of the Curie temperature of 5 C./mm to 0.5 C./mm over 80% of a length of the working component.

2. The article according to claim 1, wherein the Curie temperature increases or decreases with a gradient that lies within 50% of a linear function over 80% of a length of the working component, and the linear function is defined as the difference between the Curie temperature in degrees Celsius at one end of the working component and the Curie temperature at an opposing end of the working component divided by a distance in millimeters between the two ends.

3. The article according to claim 2, wherein the Curie temperature decreases or increases with a gradient that lies within 5% of the linear function.

4. The article according to claim 2, wherein the Curie temperature decreases or increases with a gradient that lies within 10% of the linear function over 90% of the length of the working component.

5. The article according to claim 1, wherein the working component has a density d in a defined portion of 5 vol % to 10 vol % of a total volume of the working component and wherein the density d lies within a range of 5% of an average total density d.sub.av of the working component.

6. The article according to claim 1, wherein the working component comprises an increasing M content in the working direction of the working component or a decreasing M content in the working direction of the working component for increasing amounts of one or more of the elements T and R in the working direction of the working component.

7. The article according to claim 1, wherein T of the working component comprises Co and/or Mn, M comprises Si, and a Si content, Si.sub.act, lying within 5% of Si.sub.m, wherein Si.sub.m=3.850.0573Co.sub.m0.045Mn.sub.m.sup.2+0.2965Mn.sub.m, wherein Si.sub.m is a metallic weight fraction of Si, Mn.sub.m is a metallic weight fraction of Mn, and Co.sub.m is a metallic weight fraction of Co.

8. The article according to claim 1, wherein nowhere along a length of the working component is there a gradient of the Curie temperature that exceeds 10 C./0.5 mm of the length.

9. An article for magnetic heat exchange, comprising: a monolithic sintered working component comprising La.sub.1-aR.sub.a(Fe.sub.1-x-yT.sub.yM.sub.x).sub.13H.sub.zC.sub.b with a NaZn.sub.13 structure, wherein M is one or more element selected from the group consisting of Si and Al, T is one or more element selected from the group consisting of Mn, Co, Ni, Ti, V and Cr, R is one or more element selected from the group consisting of Ce, Nd, Y, and Pr, 0a0.5, 0b1.5, 0.05x0.2, 0y0.2, and 0z3, a content of at least one of the elements selected from the group consisting of T, R, C, and M varying in a working direction of the working component, and a Curie temperature monotonically decreasing or monotonically increasing in the working direction of the working component, wherein nowhere along a length of the working component is there a gradient of the Curie temperature that exceeds 10 C./0.5 mm of the length.

10. The article according to claim 9, wherein the Curie temperature increases or decreases with a gradient that lies within 50% of a linear function over 80% of a length of the working component, and the linear function is defined as the difference between the Curie temperature in degrees Celsius at one end of the working component and the Curie temperature at an opposing end of the working component divided by a distance in millimeters between the two ends.

11. The article according to claim 10, wherein the Curie temperature decreases or increases with a gradient that lies within 5% of the linear function.

12. The article according to claim 10, wherein the Curie temperature decreases or increases with a gradient that lies within 10% of the linear function over 90% of the length of the working component.

13. The article according to claim 9, wherein the working component has a density d in a defined portion of 5 vol % to 10 vol % of a total volume of the working component and wherein the density d lies within a range of 5% of an average total density day of the working component.

14. The article according to claim 9, wherein the working component comprises an increasing M content in the working direction of the working component or a decreasing M content in the working direction of the working component for increasing amounts of one or more of the elements T and R in the working direction of the working component.

15. The article according to claim 9, wherein T of the working component comprises Co and/or Mn, M comprises Si, and a Si content, Si.sub.act, lying within 5% of Si.sub.m, wherein Si.sub.m=3.850.0573Co.sub.m0.045Mn.sub.m.sup.2+0.2965Mn.sub.m, wherein Si.sub.m is a metallic weight fraction of Si, Mn.sub.m is a metallic weight fraction of Mn, and Co.sub.m is a metallic weight fraction of Co.

16. The article according to claim 15, wherein Si.sub.act lies within 2% of Si.sub.m.

17. The article according to claim 9, wherein 0.05b0.5.

18. The article according to claim 9, wherein 1.4<z3.

19. The article according to claim 9, wherein T is selected from one or more of the group consisting of Co and Ni and wherein z=0.

20. The article according to claim 9, wherein T is selected from one or more of a group consisting of Mn, Ti, V and Cr, R is selected from one or more of a group consisting of Ce, Nd and Pr, and wherein 1.4<z3.

Description

(1) Embodiments and specific examples will now be described in connection with the following figures and Tables.

(2) FIG. 1 illustrates a green body according to a first embodiment.

(3) FIG. 2 illustrates a working component for magnetic heat exchange fabricated from the green body of FIG. 1.

(4) FIG. 3 illustrates a typical curve of temperature change as a function of temperature for a single slice of a working component.

(5) FIGS. 4A-4C illustrate graphs of Curie temperature, maximum temperature change and full width at half maximum value as a function of the position of the slice in a first working component.

(6) FIGS. 5A-5C illustrate graphs of Curie temperature, maximum temperature change and full width at half maximum value as a function of the position of the slice in a second working component.

(7) FIGS. 6A-6C illustrate graphs of Curie temperature, maximum temperature change and full width at half maximum value as a function of the position of the slice in a third working component.

(8) FIGS. 7A-7C illustrate graphs of Curie temperature, maximum temperature change and full width at half maximum value as a function of the position of the slice in a fourth working component.

(9) FIG. 8 illustrates a definition of two different Curie temperature gradients and of the diffusion zone used to establish the values given in Table 4.

(10) FIG. 9 illustrates a schematic diagram of the concentration profile over the working component at t=0.

(11) FIG. 10 illustrates the measured Curie temperature as a function of position in a first working component.

(12) FIG. 11 illustrates a graph of measured Curie temperature as a function of position for a second working component.

(13) FIG. 12 illustrates a graph of Curie temperature as a function of position for a third working component.

(14) FIG. 13 illustrates a graph of entropy change as a function of temperature for differing carbon contents.

(15) FIGS. 14A-14C illustrate graphs of Curie temperature, entropy change as a function of carbon content and entropy change as a function of Curie temperature.

(16) FIGS. 15A and 15B illustrate measured Curie temperatures as a function of silicon content and carbon content.

(17) FIGS. 16A and 16B illustrate maximum entropy change as a function of silicon content and carbon content.

(18) FIG. 17 illustrates a graph of sinter density in dependence of sinter temperature for samples including varying Si and Al contents.

(19) FIG. 18 illustrates a graph of sinter density in dependence of sinter temperature for samples with varying aluminium contents.

(20) FIG. 19 includes Table 1 which illustrates the composition of three starting powders with a cobalt composition selected to give differing Curie temperatures.

(21) FIG. 20 includes Table 2 which illustrates the Curie temperatures of four multilayer green bodies.

(22) FIG. 21 includes Table 3 which illustrates sinter and diffusion heat treatments.

(23) FIG. 22 includes Table 4 which summarizes the measured diffusion zones and Curie temperature gradient.

(24) FIG. 23 includes Table 5 which summarizes the calculated diffusion coefficient of cobalt in the working components.

(25) FIG. 24 includes Table 6 which summarizes the compositions of the three starting powders of the second set of examples.

(26) FIG. 25 includes Table 7 which summarizes the composition of multi-layer green bodies were fabricated from layers of the powders of Table 6.

(27) FIG. 26 includes Table 8 which summarizes the density of green bodies heat treated at 1120 C. for 4 hours, 16 hours or 64 hours.

(28) FIG. 27 includes Table 9 which summarizes the compositions of examples of differing carbon content.

(29) FIG. 28 includes Table 10 which summarizes the sinter temperatures and the density of the samples of Table 9 after heat treatment at different temperatures.

(30) FIG. 29 includes Table 11 which summarizes the compositions of a set of examples in which carbon is substituted for silicon.

(31) FIG. 30 includes Table 12 which summarizes the compositions of six samples of differing carbon and silicon content.

(32) FIG. 31 includes Table 13 which summarizes the sinter temperature and density measured for the samples of Table 12.

(33) FIG. 32 includes Table 14 which summarizes the compositions of six samples of varying Si and Al content.

(34) FIG. 33 includes Table 15 which summarizes the density of the samples of Table 14 after heat treatment at different temperatures.

(35) FIG. 34 includes Table 16 which summarizes the compositions of six samples of varying Al content.

(36) FIG. 35 includes Table 17 which summarizes the density of the samples of table 16 after heat treatment at different temperatures.

(37) FIG. 1 illustrates a green body 1 according to a first embodiment which may be heat treated to form a working component for a magnetic heat exchanger which includes a functionally-graded Curie temperature.

(38) The green body 1 includes three portions 2, 3, 4. Each portion includes elements having a stoichiometry suitable to form a La.sub.1-aR.sub.a(Fe.sub.1-x-yT.sub.yM.sub.x).sub.13 phase. The composition of each portion differs from that of the others and is selected to give a preselected Curie temperature, T.sub.c. In one particular example, the Curie temperature of the first portion 2 is 60 C., the Curie temperature of the central portion 3 is 30 C. and the Curie temperature of the third portion 4 is 15 C. The green body 1 therefore has three separate portions with a composition selected to produce a Curie temperature which decreases along the length of the green body 1 in a step like fashion as is illustrated in FIG. 1.

(39) The Curie temperature of the three portions can be preselected by selecting the composition of the cobalt content of the La.sub.1-aR.sub.a(Fe.sub.1-x-yT.sub.ySi.sub.x).sub.13 phase according to the following equation:
T.sub.c=14.82Co.sub.m87.1(1)
wherein Co.sub.m is the metallic weight fraction of cobalt.

(40) Additionally, the silicon content of the three portions 2, 3, 4 of the green body 1 is adjusted so that each of the portions has a sinter activity that is similar. The silicon content is selected according to the following equation:
Si.sub.m=3.850.0573Co.sub.m(2)
wherein Si.sub.m is the metallic weight fraction of silicon and Co.sub.m is the metallic weight fraction of cobalt.

(41) The similar sinter activity enables the three portions 2, 3, 4 of the green body 1 to be formed as integral parts of a single monolithic sintered working component after heat treatment of the green body 1. A monolithic working component having a plurality of Curie temperatures which decrease over the length of the working component is provided.

(42) In order to form a monolithic working component from the green body 1, the green body 1 is sintered. The sintered body 1 is subsequently heat treated at a temperature T.sub.diff which lies within the range of 900 C. to 1200 C. for a time within a range of 1 hour to 100 hours. The temperature and the time of the heat treatment are selected so as to allow diffusion of one or more elements along the length of the green body 1 and form a working component 7.

(43) In particular, a diffusion zone 8, 9 is created at the interfaces 5, 6 between adjacent portions 2, 3 and 3, 4 respectively as a result of the heat treatment, as is illustrated in FIG. 2. The Curie temperature of the working component 7 when viewed as a function of the length no longer has a stepped structure as in the green body illustrated in FIG. 1 but has a smooth undulating decrease which is generally linear, as is illustrated in FIG. 2. The Curie temperature monotonically decreases from left to right in the view of FIG. 2.

(44) The heat treatment conditions may be selected so that this decrease in the Curie temperature is as linear as possible. The heat treatment temperature and the time may be selected depending on the diffusion coefficients of the particular elements included in the basic LaFe.sub.13 phase which is or are used to adjust the Curie temperature and depending on the thickness of the portions of differing composition included in the green body.

(45) As a Curie temperature gradient is desirable across the length of the working component, the diffusion time should not be selected to be too long as eventually the entire working component would have the same composition and a uniform Curie temperature across its length.

(46) By providing a monolithic sintered working component 7 with a functionally-graded Curie temperature which decreases in a generally linear fashion from one end 10 of the working component 7 to the other end 11 of the working component 7, the efficiency of the heat exchange can be increased over that possible by a stepped decrease in the Curie temperature as, for example, illustrated in FIG. 1.

(47) In a first set of embodiments, the Curie temperature of the working component was adjusted by adjusting the cobalt content. Table 1 illustrates the composition of three starting powders with a cobalt composition selected to give a Curie temperature of 10 C., 35 C. and 60 C. The silicon composition was also selected according to equation (2) above.

(48) Four green bodies were formed including two or three of these powders in a layered structure. The Curie temperatures of the three green bodies are summarised in Table 2. The first green body is fabricated using powders having a composition selected to provide a Curie temperature of 10 C. and 60 C. The second green body is fabricated from the powders having a composition selected to have a Curie temperature of 10 C. and 35 C. The third green body is fabricated using three powders having the differing Curie temperatures of 10 C., 35 C. and 60 C. The fourth green body is fabricated using powders having compositions for the two Curie temperatures of 35 C. and 60 C.

(49) In each case, the different powders were placed in layers in a press. The powder with the lowest Curie temperature was placed in the press first and pressed with a pressure of around 0.15 tonne/cm.sup.2. A flat surface was produced on which the second powder was introduced without mixing of the two powders. After the layers were built, the green body was then pressed with a pressure of 2 tonne/cm.sup.2.

(50) The green bodies were given a sintering and diffusion heat treatment at 1100 C. for 4 hours, 16 hours or 64 hours as summarized in Table 3. These times were chosen so that the diffusion length is expected to be double that of the previous time as the diffusion distance/length is generally expected to be proportional to the square root of the diffusion time. In particular, the green bodies were heated to 1100 C. for 3 hours in a vacuum and the remaining dwell time in argon before furnace cooling to 800 C. where the temperature was held for a further 8 hours before fast cooling to room temperature.

(51) The annealing at 800 C. was performed in order to decompose the La(Fe, Co, Si).sub.13 phase in order to allow machining of the working component without forming undesirable cracks. This method uses the teaching of the published application WO2010/038099 A1 which is hereby incorporated by reference in its entirety.

(52) In order to investigate the diffusion between the different portions of the working components, the working components were cut into slices with a spark erosion technique. The working components were cut so that some slices extend over the length of the working component. Others were cut in a perpendicular direction so that slices along the length of the working component could be tested to establish the Curie temperature of this particular slice. The individual slices were around 1 mm thick and the total length of the working component was around 25 mm.

(53) To recombine the magnetocalorically active La.sub.1-aR.sub.a(Fe.sub.1-x-yT.sub.ySi.sub.x).sub.13 phase, the slices were heat treated at 1050 C. for 4 hours and quickly cooled.

(54) The Curie temperature of the slices was measured with an infrared monitor whilst the test sample rotated in a magnet system so that it was subjected to an alternating external field of 1.6 T and 0 T. FIG. 3 illustrates a typical curve of temperature change as a function of temperature for a single slice.

(55) The peak temperature, which may also be described as the Curie temperature, the maximum temperature change and the full width at half maximum value of the curves is illustrated as a function of the position of the slice in the working component for the working components 1, 2, 3 and 4 in FIGS. 4, 5, 6 and 7, respectively.

(56) Diffusion between the layers of differing composition is seen for each of the working components as the form of the decrease in the Curie temperature becomes smoother and more linear for increasing heat treatment time. The plateaus become shorter and less sharply defined for increasing diffusion times.

(57) Table 4 includes a summary of the measured diffusion zones and Curie temperature gradient of these samples.

(58) Two different ways of establishing the Curie temperature gradient are used. FIG. 8 illustrates the two different Curie temperature gradients and the definition of the diffusion zone use to establish the values given in Table 4.

(59) The gradient illustrated with the solid line in FIG. 8 is established by measuring the gradient of the central portion of the diffusion zone by establishing the best fit of this portion of the curve.

(60) The mean gradient, illustrated in FIG. 8 by the dashed line, is the slope of the line connecting the two points in the end zones at which the peak temperature starts to deviate from a constant value. These two endpoints were also used to establish the diffusion zone.

(61) The results may be used to calculate the diffusion coefficient of cobalt in these working components. The calculated values are summarised in Table 5.

(62) The following equation was used which is applicable only for diffusion coefficients which are independent of the concentration of the diffusing element. This factor was assumed in this case. It was assumed that the concentration on the left is c.sub.1 and that on the right is c.sub.2 and the concentration profile over the working component is known at start point of t=0 as illustrated in FIG. 9 and summarised in the following equation:

(63) $\begin{array}{cc}\frac{c}{t}=D\frac{c^2}{x^2}& \left(2\right)\end{array}$

(64) The above differential equation can be solved as follows:

(65) $\begin{array}{cc}c\left(x,t\right)=\frac{c_1+c_2}{2}+\frac{c_1-c_2}{2}.Math.\mathrm{erf}\frac{\left(x\right)}{\sqrt[2]{\mathrm{Dt}}}& \left(3\right)\end{array}$
where erf is an error function, t the diffusion time and D is the interdiffusion coefficient.

(66) Table 5 illustrates that the calculated interdiffusion coefficients vary only slightly over the nine samples. The average value is 2.310.sup.11 m.sup.2/s.

(67) In a further set of embodiment, a varying carbon content was used to adjust the Curie temperature.

(68) The compositions of the three starting powders are illustrated in Table 6. As in the first set of embodiments, the composition of the three powders was selected so as to produce a Curie temperature of 10 C., 35 C. or 60 C. The carbon content increases from 0.06 weight percent, 0.36 weight percent and 0.68 weight percent for the three powders, respectively.

(69) Green bodies were fabricated from layers of two or three of these powders. The compositions are summarised in Table 7. The fifth green body has a layer with a Curie temperature of 10 C. and a second layer with the Curie temperature of 35 C.

(70) The sixth green body has a first layer having a Curie temperature of 10 C. and a second layer having a Curie temperature of 60 C.

(71) The seventh green body has three layers having the Curie temperatures 10 C., 35 C. and 60 C., respectively.

(72) As in the first set of embodiments, the powder with the lowest Curie temperature was placed in the press first, pressed to form a flat surface before the second powder was placed into the press. In case of the seventh green body with three different differing Curie temperatures, the second layer was pressed to form a flat surface before the powder forming a third layer was placed into the press.

(73) These green bodies were given a sintering and diffusion heat treatment at 1120 C. for 4 hours, 16 hours or 64 hours. The density of the working components after heat treatment of the green bodies is summarised in Table 8.

(74) In this embodiment, the sinter activity was not adjusted for the differing carbon contents by adjusting the silicon content. The working component fabricated from the second sort of powder and heated for 64 hours split into two pieces. It is though that this is a result of the sinter activity being sufficiently different for the two portions of varying carbon content. These results in combination with the sinter density achieved for other carbon and silicon contents, which are summarized in Table 10 and for examples 1 to 3 of Table 13, indicates that the silicon content should be reduced for increased carbon contents.

(75) If aluminium is used instead of silicon, it is expected that the aluminium content also has to be reduced for increasing carbon content as can be seen from the results summarized in Table 16 and Table 17.

(76) FIG. 10 illustrates the measured Curie temperature as a function of position for the fifth working component heat treated at 4 hours, 16 hours and 64 hours. FIG. 10 illustrates that the gradient of the Curie temperature (Curie temperature per unit length) becomes more linear with a less pronounced curve across the interface between the two portions of the green body with increasing diffusion time.

(77) FIG. 11 illustrates a graph of measured Curie temperature as a function of position for the working component fabricated from the sixth green body which was heated at 1120 C. for 16 hours.

(78) FIG. 12 illustrates a graph of Curie temperature as a function of position for a working component made from the seventh green body which was heated for 16 hours and 64 hours.

(79) FIGS. 10 to 12 illustrate that diffusion between the layers having different carbon compositions has occurred to give a functionally-graded Curie temperature.

(80) The following set of examples was carried out in an attempt to establish whether carbon is accommodated in the NaZn.sub.13 structure in interstitial positions or at lattice positions.

(81) In the first set of examples, a composition including 16.7 weight percent lanthanum, 3.48 weight percent silicon, 6.55 weight percent cobalt and 73.250% iron was fabricated. This composition should have a Curie temperature of 10 C. In this composition all of the lattice sites should be occupied by these elements.

(82) 0.4 weight percent carbon was introduced into a portion of this powder in the form of graphite powder and mixed for 30 minutes with steel balls. This graphite containing powder was mixed in varying proportions with non-graphite containing powder to produce examples of differing carbon content. The compositions are summarised in Table 9.

(83) The samples were heated at one or more of five heat treatment conditions at varying temperatures for 3 hours in vacuum and 1 hour in argon before fast cooling. The sinter temperatures and the density of the samples heated at different temperatures are summarised in Table 10. In order to achieve the highest density, the sinter temperature may be increased for increasing carbon contents.

(84) The entropy change as a function of temperature for differing carbon contents is summarised in FIG. 13. As the carbon content increases, the peak temperature, which corresponds to the Curie temperature, increases. The relationship between the Curie temperature, entropy change and carbon content and the relationship between entropy change and Curie temperature are summarised in FIG. 14.

(85) In a further set of examples, carbon was substituted for silicon in a La.sub.1-aR.sub.a(Fe.sub.1-x-yT.sub.ySi.sub.x).sub.13-based composition. The compositions are summarised in Table 11 and varying portions of the powders were mixed to form six samples of differing carbon and silicon content. The compositions of the six samples are summarised in Table 12. In this set of embodiments, it was attempted to establish whether carbon could replace silicon in the NaZn.sub.13 structure.

(86) Samples were heated at various sintering temperatures for 3 hours in vacuum followed by 1 hour in argon before fast cooling. The sinter temperature and density measured are summarised in Table 13.

(87) FIG. 15 illustrates the measured Curie temperatures as a function of silicon content and carbon content.

(88) FIG. 16 illustrates the maximum entropy change in an external field of 1.6 T for the samples as a function of silicon content and carbon content.

(89) Samples 1 to 3 illustrate a dependence of the Curie temperature on the silicon and carbon content which can be summarised 30 by the following equation:
T.sub.c=97C87.1(4)
wherein C is the carbon content in weight percent.

(90) This is similar to that found in the first set of embodiments. However, samples 4 to 6 are found to exhibit different behaviour and a decrease in Curie temperature. This indicates that carbon does not take the same sites as Si, but is incorporated only on interstitial sites, as is the case for hydrogen. The results also indicate that at least 3.5 weight percent silicon is required to stabilize the NaZn.sub.13 crystal structure for the entire cobalt content of 3.9 weight percent.

(91) These sets of embodiments illustrate that the sinter activity also depends on the carbon content as well as on the silicon content. Consequently, the silicon content should be adjusted for a given carbon content required to provide a given T.sub.c as well as to provide a uniform sinter activity and uniform resulting density of the working component. This hinders the cracking and/or delamination of portions of the working component. A similar result is expected by adjusting the aluminium content in dependence of the carbon content.

(92) In a further group of embodiments, the effect of varying the silicon and aluminium contention the sinter density was investigated.

(93) Table 14 summarizes the compositions of six samples of varying Si and Al content. The total content of silicon and aluminium remains largely similar, but the proportion of silicon to aluminium is varied so that sample 1 includes only aluminium and no silicon and sample 6 includes only silicon and no aluminium.

(94) Table 15 summarizes the density of the samples of Table 14 after heat treatment at different temperatures in the range of 960 C. to 1110 C. These results are also illustrated in the draft of FIG. 17.

(95) These results indicate that the aluminium content can also be adjusted to adjust the density of the working component. As the silicon content increases and the aluminium content decreases, the sinter temperature should be increased to achieve a high sinter density.

(96) In a further group of embodiments, the effect of aluminium on the sinter density for samples with no silicon, i.e. silicon-free samples, was investigated.

(97) Table 16 summarizes the compositions of six samples of varying Al content. The samples were heat treated at temperatures between 940 C. and 1040 C. The sinter density of these samples is summarized in Table 17.

(98) FIG. 18 illustrates a graph of sinter density in dependence of sinter temperature for samples with varying aluminium contents.

(99) The results indicate that for increasing Al content, the sinter temperature has to be increased in order to achieve a high sinter density. This trend is comparable to the embodiments in which the samples only contain silicon and no aluminium.