SOME PROBLEMS OF FIBER M E T A L L U R G Y M. Yu. B a l ' s h i n ,
M. K. R y b a l ' c h e n k o ,
O. V. P a d a l k o ,
a n d N, P. E s k i n a
A. A. Baikov Institute of Metallurgy Translated from Poroshkovaya Metallurgiya, No. 3(21), pp. 16-22, May-June, 1964 Original article submitted March 12, 1963
In recent years, substantial progress has been made in a new branch of powder metallurgy-fiber metaIlurgy [1-4]. The starting m a t e r i a l - m e t a l fibers 25-100 /1 (occasionally up to 500 g) in diameLer, with a length: diameter ratio of 50-200 : l - i s produced by cutting wire, machining metals, and, for not too refractory materials, by atomizing and spraying molten metal. Effective methods for preparing fibers from molten metals have been developed by P. K. Oshchepkov and coworkers. Metal fibers have very poor flow characteristics and low bulk density. For this reason, they do not readily lend themselves to the direct fabrication of materials, semiproducts, and components by compacting, roiling, extrusion, or hot pressing. Before metal fibers can be subjected to shaping or sintering, felt must be produced from them. The manufacture of metallic felt is analogous to the production of paper from fibers and, basica!ly, is a variety of the ship casting process. Felt manufacture consists in depositing fibers from a suspension in a viscous liquid on to a filter or gauze by gravitation, vacuum, ,or pressure filtering, or centrifuging. The diagram in Fig. 1 shows the principle o f f e r deposition. The resultant felt, which as a porsity ranging from about 70 to 97%, may be subjected to subsequent working (cold compacting in dies roiling), sintering, hot pressing, infiltration [2, 4], or welding. Thus, the manufacture of materials and componem:s from metal fibers ts somewhat more tedious and time-consuming than the manufacture of parts from powder. However, fiber metals have a number of advantages ove r powder ones, which justify the additional complexity of the process. The possible porosity range for powder metals is from zero to 60%, while for fiber metals it is much wider, extending from zero to 97%. Consequently, the property range for fiber metals is also much wider. It is possible to produce fiber metals with very high capacity, unattainabie in powder metals, for absorbing sound and mechanical vibrations in a wide frequency range. Even at equal porosity, fiber metals have higher strength and ductility than powder metals. The impact strength of fiber metals is frequently higher by a whole order of magnitude than that of powder metals of the same porosity. Similarly, compared with powder metals, fiber materials have higher erosion and thermalfatigue resistance [2, 3, 5].
Fibers
r---
I
Glycerin Metallic felt
f To vacuum pump Fig. 1. Diagram showing preparation of felt by deposition of fibers from viscous medium.
Fig. 2. Kinematic diagram of wire cutting device (for keysee text). 185
Fig. 3. Particle and pore orientation in structure of fiber metah a) in plane perpendicular to direction of shaping force; b) in plane parallel to direction of shaping force. Fiber metallurgy methods are widely employed for the manufacture of filters of high porosity, permeability and strength, parts of de-icing systems, components of mechanisms operating at high impact and vibrational loads, etc. New fiber metals may also be used as active-mass carriers of catalysts, storage batteries, and electrolytic ceils. In addition, the whole problem of heat resistance may possibly be solved by reinforcing heat-resistant, brittle ceramic materials and cermets of low thermal fatigue resistance with a metal-fiber skeleton having high ductility and thermal conductivity. In the present study, use was made of fibers produced by cutting grade MT copper wire of 40, 50, and 100 diameter, Figure 2 shows a kinematic diagram of the cutting device. The roll 1 is the driving element of the device. On this ro11 is mounted the crown 2, the cutting slots of which transmit rotation through the driving roll 3 to the roll 4. The wire 5 is unwound from a coil, passes between the roll 4 and the idle roll 7, and enters into the spinneret of the guide plate 6. On leaving the spinneret, it is cut. TABLE 1. Comparative Compressibility Values for Electrolytic Copper Powder and Copper Fibers Compacting pressure, daN/mm~* 20 25 30
ZO
Porosity, % powder fibers 36 30 26
* 1 daN/mm 2 = 1.02 k g / m m 2
186
o
8,O
[ 30 24 20
~t; 5,o 4.0
aol 0
2~
,b 5 /o ~J +b
Porosity, % Fig. 4. Elastic recovery during sintering of copper fiber specimen as function of porosity and initial fiber diameter. Wire diamter (in p): 1) 100; 2) 40
TABLE 2. Impact Strength a k of Copper Fiber Specimens as Function of Porosity
and Specimen Preparation Conditions (fiber diameter 100 ~) Process
Process P, % 0 5 I0 20 3O 40
ak, k g - m / c m ~ 2,44 2.07 1.71 1.49 1 .O5 0,62
p,% 0 3 4 9 18 32
ak, k g - m / c m 2 2,5 2,3 2.1 2.0 1,3 0.8
Process
p,~, ak, k g - m / c m 2 -i 3 7 13 17
s 2.4 2.1 1,g 1,1
With this arrangement, it is possible to cut either individual wires, each of which is fed through a separate spinneret, or braid twisted from several wires. It is assumed that braid will be less likely to buckle than individual wires. As can be seen from the diagram, the outlet cross section of the spinneret and the point of contact between the slot crown 2 and the driving roll are arranged in such a manner that, at the instant of cutting, the driving roll becomes disengaged from the slot crown. As a result of this, wire feed is stopped during the whole tirne when the slot surface passes by the outlet eross section of the spinneret. This prevents the jamming of the spinneret with crumpled wire, which might take place during continuous feeding. Three length r a n g e s - 1 0 - 1 g , 5-8, and 2-4 r a m - o f fibers 100 p in diameter were tested. The tensile strength o b of specimens from the 10-15 and 5-8 mm long fibers was found to be identical. The tensile strength of specimens from the fibers 2-4 mm long was 10-15% lower. In view of this and the diffieuly of producing small specimens from fibers 10-15 mm long, the 5-8 mm fibers were selected for further study. There appears to be a definite correlation between the optimum fiber length and the blank and component size, so that the optimum length of the starting fibers may be expected to decrease with decreasing blank and component size. In the production of felt blanks for the preparation of specimens, a fiber suspension in glycerin was poured into the die of a press tool and was filtered through gauze according to the arrangement in Fig. 1, with a boosting vacuum pump. After the fibers had settled and the glycerin had been removed, the press tool was lifted from the gauze, and felt was compacted in it by double-action compression. As has already been noted above, one of the characteristics of fiber metals, resulting from the pronounced nonequiaxiality of their structural elements (the length: diameter ratio in our experiments was 50-160), is that the particles have poor flow properties. Because of this the special preshaping of the felt is an essential operation in the technology of fiber metallurgy. Another characteristic of fiber metals, which is also due to the same nonequiaxialiry, is a marked particle and pore structure anisotropy in the direction of the force acting during felt shaping and compacting (the direction of the compacting pressure coincides with the direction of the force responsible for felt settling, the direction of the force of gravity during gravitation settling, and the direction of the pressure drop during the vacuum or pressure filtration of the fiber suspension). For this reason, the structure of such felt, as well as of compacts and sintered fiber metal~ has a characteristic particle and pore orientation, the longer (larger) cross sections of the particles and pores tying in planes at right angles to the direction of the shaping force (Fig. 3a) and their smaller cross sections in planes paratlei to the direction of this force (Fig. 3b). A similar orientation in powder metals has been noted in [6]*. It must be emphasized, however, that, both in powder [6] and fiber metals, the pore anisotropy is only structural , and not qualitative, as is wrongly claimed in [7]. For a microsection area containing a sufficiently large number of fibers, the same proportion of the area is occupied by pores in directions parallel .and perpendicular to the direction of the compacting pressure. Quantitatively, this proportion is equal to the voiume porosity of the fiber metal. * There is also another orientation [6], in which the longer cross sections are parallel to the die walls (and centerline)o Consequently, the final orientation depends to a large extent also on the ratio of the component dimensions determined by the die, and the fiber length.
187
28
t 20. 24', ""
Ifi
In compacting, fibers are characterized by high densification (in spite of the high initial porosity of felt prior to shaping). The excellentcompressibility is due to the high degree of fiber orientation in the starting felt, the smoothness and uniformity of each fiber, and, finally, the low porosity resulting from the ideal stacking of the structural elements without their deformation. Thus, for example, for powders with a spherical particle shape, porosity at idealstacking is about 25%, while for fibers with a cylindrical particle shape, porosity at ideal stacking is 10%. For this reason, as shown in Table 1, copper fibers have better compressibility than copper powder.
~,", 3
D
# 4 I
0
ib
2o 3o
Porosity, % Fig. 5. Tensile strength of fiber copper as function of porosity and conditions of preparation (as described in text).
r
r-4
"~ %
/ /
r-4 O
The pronounced growth of volume and dimensions (particularly height) as a result of elastic recovery during the sintering of fiber metals is due to two factors. On the one hand, a part is played by a substantial springing effect and the resultant large elastic strain of the long fibers; on the other hand, the structural elements of fiber metals are smoother then particles of powder metals, and have a smaller specific surface.
/"
J /
0,40 Fig. 6. Effect of relative density 0 on electrical conductivity of copper fiber material. TABLE 3. Impact Strength a k of Molybdenum Fiber Specimens as Function of Porosity (fiber diameter 50 p) P, % 21 26 31 43
ak,kg-m/cm 2
The third characteristic of fiber metals is that, in comparison with powder metals, they exhibit stronger elastic recovery. This elastic recovery manifests itself to some extent immediately after pressure has been removed and the compact taken out of the die, and leads to an increase in both the compact diameter and height. The effect of elastic recovery manifests itself particularly strongly during sintering.
For this reason, capillary pressure, which acts as the motive force of sintering, is smaller in fiber metals than in powder ones. Consequently, during sintering, it neutralizes elastic recovery to a smaller extent in fiber metals than in powder ones. Capillary pressure decreases with increasing fiber diameter and initiat compact porosity. Thus, elastic recovery due to sintering increases with increasing initial fiber thickness and compact porosity (Fig. 4). Since the magnitude of capillary pressure in fiber metals is is small, they must be sintered at high temperatures. In our experiment s, copper fiber specimens were sintered for 9 h in hydrogen at 980* C (939 of the absolute melting temperature).
1,58 1.46 1,41 1.40
Figure 5 shows the effect of porosity on the tensile strength o b of copper fiber specimens obtained by three processes: 1) sintering of compacted felt; 2) sintering of compacted felt, second compacting at the same pressure as for the first compacting, second sintering under the same conditions as for the first sintering; 3) sintering of compacted felt, second compacting at double the pressure used for the first compacting, second sintering. It follows from Fig. 5 that the strength of porous fiber copper sintered without subsequent working does not differ greatly from the strength of powder copper at the same degree of porosity. The tensile strength of nonporous dense fiber copper after second compacting and second sintering was 31 daN/ram 2. For dense powder copper and cast copper (after working and anI~ealing), o b = 2 2 - 2 4 daN/ram z. Thus, the tensile strength of dense fiber copper is 33-45%higher than that of powder and cast materials.
188
It was also established in this study that, at equal porosity, the strength of copper from thinner fibers (50 g diameter) is much higher than that of copper from thicker fibers (100 g diameter). As in the case of powder m a t e rials [7], the relationship between strength and density for fiber metals is quite accurately described by the formula:
GOp ~ d
= t g m : ( | - - p)m
(1)
where Obp is the tensile strength of porous metal, Obp the tensile strength of dense metal, ~ relative density, P porosity, and m = 2 . 2 - 4 . It can be seen from Eq. (1) that, as in the case of powder metals, the tensile strength of fiber metals is determined by th~ magnitude of the contact section. Experimental data on the impact strength of fiber copper specimens subjected to the same treatments as those employed for tensile test specimens are given in Table 2. As can be seen from Fig. 5, second compacting increased, at equal porosity, the values of strength. On the other hand, at equal porosity, second compacting reduced the value of impact strength (Table 2). For example, without second compacting (process No. 1) at 17% porosity, a k lay between 1.49 and 1.71; with additional compacting (process No. 2) at the same porosity, the value of ak was slightly less than 1.8; with even more intensive additional compacting (process No. 3) at the same porosity, the value of ak was lower still and was only 1.1 k g - m / c m 2. Thus, at equal porosity, the values of ak of fiber copper decreased with increasing strength and, consequently, with the magnitude of contact section. The opposite effect is invariably observed in powder metals: At equal porosity, impact strength always increases with increasing tensile strength and contact section. Additional compacting and sintering always raise the impact strength of powder materials. T h i s somewhat paradoxical behavior, together with the high values of impact strength and the less pronounced effect of porosity on the latter can only be attributed to the fact that, in fiber metals, impact strength is a property which is determined not only by interpartiele contact, but also, to a large extent, by intraparticle characteristics-the high impact strength of the actual structural elements of fiber metal, i . e . , its individual fibers. Table 3 shows results of impact strength tests of molybdenum fiber specimens. What is particularly noticeable is that the effect of porosity on the impact strength of fiber molybdenum is much more pronounced than in the case of copper and that the vaIue of impact strength is very high (several tirnes higher than that of dense bar specimens from powder or cast molybdenum). It can be seen from the table that the impact strength of fiber molybdenum depends on intraparticle properties to an even greater extent than does the impact strength of fiber copper. As is known [7], the electrical conductivity p of powder metals is essentially, if not exclusively, an interpaP ticle contact property and is generally proportional to the contact section. It can be seen from Fig. 6 that t h e e l e c trical conductivity of fiber copper is directly proportional not to the contact section, but to relative density ~, i . e . , to the mean cross section of the material without the pores. Thus, electrical conductivity is governed primarily by the intraparticle, and not contact, properties of fiber metal. From this follows the fourth and most important characteristic of fiber metals, namely, that, compared with powder metals, their properties depend more on intraparticle factors and less on interparticle contact factors. In comparison with uninfiltrated specimens, porous fiber copper specimens infiltrated with Bakelite were found to have higher tensile strength (2-4 d a N / m m 2 higher), lower impact strength (0.1-0.2 k g - m / c m 2 lower), and approimately the same electrical conductivity. The effect of porosity on the air permeability of fiber copper may be described by the formula:
~=k- P
(2)
where a is the coefficient of air permeability, k a coefficient depending on testing conditions, P porosity, and n an exponent. For sintered fibers 100 and 40 /z in diameter, n is 2-4 and 4-8, respectively. SUMMARY Problems of the preparation and sintering of fibers are discussed, and the properties of porous and dense copper fiber materials are studied. Fiber metals differ from powder metals by the following characteristics: pronounced
189
anisotropy, comiderable elastic recovery during sintering and compacting, greater dependence of properties on intraparticte factors, and lesser dependence on interparticle contact factors. LITERATURE
1. 2. 3, 4. 5.
CITED
ASM-SLA Classification, Revised Edition (1957), [Metals Review], 30, 2 (1957), p. 7. R.H. Read, Metal treatment and drop forging, 27,178 (1960). A.G. Metcalfe, C. H. Sump, and W. C. Troy, Metal Progress, 67, 3 (1956), p.p. 81-84. R.H. Read, Material in design engineering, XII, 104 (1952). C . H . Sump and W, Pollack, Proceedings Thirteenth Annual Meeting Metal Powder Association,I, Chicago, 111 (1957). M.Yu. Bal'shin, Powder Metallography [in Russian], Moscow, Metallurgizdat (i948). G, A. Meerson, Collection: Powder Metallurgy [in Russian], 4 (1986).
6. 7.
All abbreviations of periodicals in the above bibliography are ietter-by~letter transliterations of the abbreviations as given in the original Russian journal. Some or all of this periodical l i t e r a t u r e m a y Well b e available in English t r a n s l a t i o n . A Complete l l s t of the c o v e r - t o c o v e r E n g l i s h t r a n s l a t i o n s a p p e a r s at the b a c k of this i s s u e ,
190