FILTER

Abstract

A filter comprising a plurality of metal fibers having a non-round cross section, in particular a rectangular, quadric, partial circular or an elliptical cross section, with the cross section comprising a large axis and a small axis, wherein a ratio of the small axis to the large axis lies in the range of 0.99 to 0.05. The invention further relates to a treatment method for metal fibers comprising an elliptical or rectangular cross section, both having a large axis and a small axis, wherein a ratio of a length of the small axis to a length of the large axis is smaller than 1, preferably smaller than 0.5, wherein the treatment method comprises the step of heating the fibers (10) in an oven to a temperature value in C. between 70 and 95% of the melting temperature in C., such that the ratio of the length of the small axis (D2) to the length of the large axis (D1) increases, preferably to the range of 0.05 to 0.99, wherein the metal fibers (10) are at least a part of a filter according to the invention.

Claims

1-40. (canceled)

41. A filter comprising a plurality of metal fibers having a non-round cross section, with the cross section comprising a large axis and a small axis, wherein a ratio of the small axis to the large axis lies in the range of 0.99 to 0.05.

42. The filter according to claim 41, wherein the ratio of the small axis to the large axis is in the range of 0.7 to 0.1.

43. The filter according to claims 41, wherein the ratio of the small axis to the large axis is in the range of 0.5 to 0.1.

44. The filter according to claim 41, wherein the cross section of the metal fibers comprises rounded edges.

45. The filter according to claim 41, wherein the cross section of the metal fibers is elliptical.

46. The filter according to claim 41, wherein the fibers comprise a length of 0.1 mm or more.

47. The filter according to claim 41, wherein a length of the large axis is equal to or smaller than 100 m.

48. The filter according to claim 41, wherein the length of the small axis is 2 m or less.

49. The filter according to claim 41, wherein the fibers form an ordered or an unordered network.

50. The filter according to claim 41, comprising a porosity selected in the range of 93 to 99.9%.

51. The filter according to claim 41, wherein the filter comprises an average mean pore size selected in the range of 0.1 to 300 m.

52. The filter according to claim 51, wherein the average mean pore size is determined according to the bubble point method, in particular as specified in the description.

53. The filter according to claim 41, wherein the filter comprises a fiber volume fraction in the range of 0.01 and 30 vol %.

54. The filter according to claim 41, wherein the thickness of the filter is in a range of 0.1 to 100 mm.

55. The filter according to claim 41, wherein the fibers are held by a frame.

56. The filter according to claim 41, wherein the fibers are sintered to one another.

57. The filter according to claim 41, wherein the fibers are composed of a metal alloy such as CuSn8, CuSi4, AlSil, Ni, stainless steel, Cu, Al or vitrovac.

58. The filter according to claim 41, wherein at least some of the metal fibers of the plurality of metal fibers are sintered or processed by a thermal treatment.

59. The filter according to claim 41, wherein the fibers are obtainable by a melt spinning process.

60. The filter according to claim 41, comprising a fiber density selected in the range of 0.002 to 6.5 g/cm.sup.3.

61. The filter according to claim 41, wherein the cross section of the metal fibers comprises rounded edges, and wherein the filter comprises a porosity selected in the range of 93 to 99% and/or a mean pore size selected in the range of 0.1 to 300 m.

62. The filter according to claim 61, wherein the filter comprises a thickness selected in the range of 6 to 49 mm.

63. Treatment method for metal fibers comprising an elliptical or rectangular cross section, both having a large axis and a small axis, wherein a ratio of a length of the small axis to a length of the large axis is smaller than 1, wherein the treatment method comprises the step of heating the fibers in an oven to a temperature value in C. between 70 and 95% of the melting temperature in C., such that the ratio of the length of the small axis to the length of the large axis increases, wherein the metal fibers are at least a part of a filter, the filter comprising a plurality of said metal fibers having a non-round cross section, with the cross section comprising a large axis and a small axis, wherein a ratio of the small axis to the large axis lies in the range of 0.99 to 0.05.

64. The method of claim 63, wherein inside the oven a protective atmosphere is applied to the fibers.

Description

[0047] The invention will now be described in further detail and by way of example only with reference to the accompanying drawings and pictures as well as by various examples of the network and method of the invention. In the drawings there are shown:

[0048] FIG. 1: a cross section of a schematic fiber with its small and large axis;

[0049] FIG. 2: different cross sections of schematic fibers corresponding to different ratios;

[0050] FIG. 3: a schematic illustration showing the difference of a gas flow around a round fiber and an elliptical fiber;

[0051] FIGS. 4a and 4b: elliptical fiber networks with different aspect ratios;

[0052] FIGS. 4c and 4d: round fiber networks with different fiber dimensions;

[0053] FIG. 5a to 5d: simulations of elliptical fiber networks with different aspect ratios and their corresponding porosities;

[0054] FIG. 6: pictures of fibers before and after a thermal treatment;

[0055] FIG. 7: pictures of the cross sections of the fibers of FIG. 6;

[0056] FIG. 8: fiber dimensions of treated and untreated fibers;

[0057] FIG. 9: fiber cross sections of treated and untreated fibers;

[0058] FIG. 10: pictures of CuSi.sub.4 fibers before and after a thermal treatment; and

[0059] FIG. 11: pictures of AlSi.sub.1 fibers before and after a thermal treatment.

[0060] FIG. 1 shows an ellipsoid cross section of a fiber 10 with its large and small axis D1, D2. As it is commonly known, the large and small axis D1, D2 of an ellipse intersect at a center point C of the ellipse, such that they represent the longest and widest part of the ellipse, respectively. Depending on how big the difference between the lengths of said two axes D1 and D2 is, the more or less the ellipse looks like a circle or can be almost flat. Said difference is represented by the ratio of the length of D2 to the length of D1, which lies between 1 for a perfect circle and 0 for a parabola. Thus, for a circular cross section, the length of both axes D1 and D2 is equal, i.e. a ratio of D2 to D1 is equal to 1.

[0061] Different cross sections of fibers 10 with different ratios are depicted in FIG. 2. As one can see, the smaller the value of said ratio is, the flatter (and longer) is the ellipse.

[0062] The filters according to the invention comprise a plurality of such metal fibers 10 with an elliptical cross section with a large axis D1 and a small axis D2. How big the ratio between D1 and D2 is, can be chosen according to the application. Fibers 10 with rounder cross sections (ratio near 1) comprise a higher mechanical stability than fibers 10 with more elliptical cross sections (ratio well below 1). On the other hand, filters made out of elliptical fibers 10 comprise a lower weight compared to filters made out of round fibers 10 since fewer fibers 10 are needed for a filter of the same size. Hence, it may be chosen according to the application which characteristic is more important.

[0063] Furthermore, the schematic flow behavior of a gas or liquid differs with respect to the geometries of the fibers. Comparing FIGS. 3a and 3b it can be seen that the flow around a round fiber 10 comprises a rather small region of lower pressure right behind (i. e. below in FIG. 3a) the fiber 10 while in FIG. 3b a larger region of lower pressure appears behind (below) the fiber 10. This effect can also be observed at air plane wings. Along this line the filtration efficiency of the filter made out of highly elliptical fibers 10 is strongly enhanced, whilstas mentioned above the volume fraction of the fiber 10 is simultaneously reduced.

[0064] In order to hold said plurality of fibers 10 together, they can either be sintered to one another or be held together by a frame (not shown). Such filters can comprise a thickness between 1 to 100 mm. The precise value can be chosen according to the application of the filter.

[0065] The effects of different fiber shapes on the porosity of a filter have been studied with simulations (see FIG. 5). The experimental results are discussed below.

[0066] For said simulations a fiber volume fraction of 2 vol % has been set as a constant value and only the ratio between the axes D2 and D1 has been varied (see Table 1 below).

TABLE-US-00001 TABLE 1 porosity of filters made out of fibers with different aspect ratios Slightly Highly Round elliptical Elliptical elliptical Large elliptic axis 10 10 10 10 (D1) [m] Small elliptic axis 10 5 2 1 (D2) [m] Ratio 1 0.5 0.2 0.1 (D2/D1) [m] Volume fraction 2 2 2 2 [vol %] Mean Pore size 113 83 53 38 [m]

[0067] One can clearly see that the mean pore size of filters made out of round fibers (column 1) is significantly bigger than the pore size of a filter made out of highly elliptical fibers (fourth column). Smaller mean pore sizes lead to a better filter performance since gases or liquids flowing through such a filter can be cleaned from smaller particles compared to filters with bigger mean pore sizes. Hence, it could be observed that the smaller the aspect ratio of the fibers 10 is, the smaller the mean pore sizes and the better the filter performance will get.

[0068] As a consequence of the smaller mean pore size observed for elliptical fibers 10, two networks (i.e. filters) with different fiber sizes and the same aspect ratios (highly elliptical) have been compared with two filters of different fiber sizes but with the same aspect ratios (round) (FIG. 4. and Table 2). For these comparisons, a standardized flow simulation (ASTM D73704 Standardized Air permeability test) can be used to test the flow performance of these structures.

TABLE-US-00002 TABLE 2 comparison of elliptical fibers with round fibers 5 m 10 m 4 m 8 m MPI-MF MPI-MF Fiber Fiber metal fibers metal fibers Network Network Diameter D1 5 10 4 8 [m] Aspect ratio 0.1 0.1 1 1 Volume fraction 2 2 4 4 [v %] Filter Thickness 0.5 0.5 0.5 0.5 [mm] Mean pore size 20.1 10 37.5 15 35 15 69 37 [m] Flow Rate @ 125 174 260 649 2188 Pa [l/s/m.sup.2]

[0069] Having closer look at Table 2, it becomes apparent that the performance of the metal fiber network does not depend on the fiber size or their diameter, but mostly on the porosity of the fiber network. Hereby, it could be shown that by using a volume fraction of 2 vol % instead of 4 vol % for the elliptical fiber network (filter), smaller porosities and highly decreased flow rates could be achieved. Comparing similar porosities (4 m fiber network vs. 10 m elliptical MPI-MF metal fiber network) the elliptical fibers are able to achieve smaller flow rates due to additional vortexes forming behind the fibers, according to FIG. 3. These vortexes prolong the flight path of molecules, particles and even the air flow in the metal fiber network.

[0070] In order to show the effect of such smaller aspect ratio fibers on heating particles, i.e. water droplets, in a fiber network, a theoretical calculation has been done to evaluate the power output of the filter. It has been assumed that a metal fiber network (i.e. a filter) has a certain temperature and air containing particles or droplets are blown through it. The heat of the network is consequently transmitted to the particles by either heat convection or heat radiation. The Stefan-Boltzmann-Equation for grey bodies has been used to calculate the heat emitted for different metal fiber networks at 190 C. and an emission coefficient of =0.1. The results are displayed in Table 3.

TABLE-US-00003 TABLE 3 heat emission of metal fiber networks at 190 C. made of CuSn8 5 m 10 m 4 m 8 m MPI-MF MPI-MF Fiber Fiber metal fibers metal fibers Network Network Diameter D1 5 10 4 8 [m] Aspect ratio 0.1 0.1 1 1 Volume fraction 2 2 4 4 [v %] Filter Thickness 0.5 0.5 0.5 0.5 [mm] Mean pore size 20.1 10 37.5 15 35 15 69 37 [m] Emitted Power 50.36 20.11 15.4 5.13 [J/s]

[0071] It becomes apparent that structures with significantly larger inner surfaces are able to emit larger amounts of radiation. Especially fibers with an elliptical cross section are able to radiate a large amount of heat. Furthermore, due to the inherently low emission gradient of metals, the radiated heat is scattered and reflected on the inner surfaces of the fibers, leading to a large transmission efficiency.

[0072] In order to obtain fibers 10 with a given shape it is possible to use fibers 10 which have been produced by a melt spinning process. Corresponding apparatuses for melt spinning can be found, for example, in the not yet published international application PCT/EP2020/063026 and from published applications WO2016/020493 A1 and WO2017/042155 A1. With such apparatuses big batches of fibers having given size dimensions and a good quality can be fabricated rather quick and easily.

[0073] Said melt spinning processes are known for producing rather flat ribbons with a rectangular cross section. Hence, their aspect ratios (width to length) are quite small and thus such ribbons would lead to a good filter performance. However, as it has already been mentioned above, the mechanical stability of round fibers can be higher. Hence, a treatment process for rounding said flat ribbons to a certain degree is needed.

[0074] According to the invention the treatment method comprises heating the fibers 10 in an oven (not shown) to a temperature between 70 and 95% of the melting temperature, such that the ratio increases, preferably up to 1.

[0075] By means of such a thermal treatment below the melting temperature of the fibers, flat fibers of any size can be rounded as can be seen from the experimental results below.

[0076] As an example, fibers 10 of a copper alloy (CuSi4; 4% by weight Si and 96% by weight Cu) were thermally treated and the rounding was investigated. FIG. 6 shows untreated fibers 10 (left) and CuSi4 fibers of the same fiber batch that have been aged under argon at 900 C. for 1 hour (right). It is easy to see that the previously very flat fibers get almost perfectly round, and thus their aspect ratio has changed from x: 1 (x>1) to almost 1:1. In-situ tests have shown that the rounding is almost complete after just a few minutes. Here, the more the aspect ratio differs from 1:1 or, in other words, the less round the fibers 10 are, the faster the rounding takes place and the slower the rounding takes place the rounder the fiber becomes (compare equation (1) above) because the surface free energy has already been greatly minimized and is closer to the optimum. Table 4 shows the investigated temperatures for fibers 10 of the copper alloy CuSi4 as well as whether a rounding has taken place or not. Some non-thermally treated and thermally treated fibers 10 were cast in epoxy resin, polished metallographically and then measured by means of optical microscopy. FIG. 7 shows the metallographic cross-section of thermally untreated fibers 10 (left) and thermally treated fibers 10 (right) made out of CuSi4 type of the same batch. It is easy to see that the flat, elliptical, fibers 10 are almost perfectly round after the thermal treatment.

[0077] FIG. 8 shows the theoretical evaluation of the fiber dimensions in width and height for the flat fibers 10 and a diameter for the thermally treated rounded fibers 10. While the thermally untreated fibers 10 have an average aspect ratio of approximately 1:3.5, after the thermal treatment the aspect ratio is 1:1. Thus, said rounded fibers 10 are then in a thermodynamically more favorable state due to the minimized surface.

[0078] In FIG. 9, the area of the fiber cross-section of the thermally untreated and thermally treated fibers 10 was determined. As expected, this value does not change during the thermal treatment and the fibers still have the same statistical distribution (due to production) as the thermally untreated fibers.

TABLE-US-00004 TABLE 4 experimented temperatures for the copper alloy CuSi4 and its respective rounding Temperature 800 C. 850 C. 900 C. 950 C. Rounding flat round round round

[0079] The physical basis of the rounding, i.e. diffusion, can be found in any material. Therefore, this process can be applied to a wide range of metallic fibers, regardless of their size and composition. A specific temperature must be used for each material, usually in the range of 70-95% of its melting temperature. The higher the temperature, the faster the process takes place. The choice of time is also essential, as larger fibers tend to be rounded at a lower temperature and for a longer time in order to avoid undesired breaks or deformations.

[0080] It may also be possible to choose the treatment time and temperature such that the rounding takes place only to a certain degree such that fibers 10 with different aspect ratios can be produced.

[0081] In order to show the independence of the dimensions, in addition to the very fine fibers of FIG. 6 (35 m width, 8 m height), quite coarse fibers 10 (150 m width, 10 m height) were thermally treated at 930 C. for 1 hour under application of argon (FIG. 10). It is easy to see that also here the fibers 10 are almost perfectly round after the thermal treatment.

[0082] FIG. 11 shows fibers made out of the AlSi1 (1% by weight Si; 99% by weight Al). The fibers were thermally treated at 630 C. for 1 hour under the application of argon. Also here it can be seen that the fibers 10 are almost perfectly rounded after the thermal treatment. This shows that the process is not material dependent.