US3663356A

US3663356A - Reinforced metal-matrix composites

Info

Publication number: US3663356A
Application number: US769938A
Authority: US
Inventors: Chou H Li
Original assignee: CHOU H LI
Current assignee: CHOU H LI
Priority date: 1968-10-23
Filing date: 1968-10-23
Publication date: 1972-05-16
Anticipated expiration: 1989-05-16

Abstract

The invention discloses the structure of, and method and apparatus for making, improved metal-matrix composites. Each of the composites comprises a strain-hardenable metal matrix, and a plurality of high-strength, high-modulus, elongated reinforcing members arranged in spatial relationship therein. Each of the members is effectively bonded to the surrounding matrix at discrete regions along its length, so as to locally strain the matrix around the discretely bonded regions upon application of load on the composite. Being thus locally strained, the matrix is spatially selectively and differentially strain-hardened and strengthened near the bonded regions. The spheres of influence of the straining and strain-hardening in the matrix near the bonded regions on one member are designed to significantly overlap similar spheres of influence near the neighboring members. This composite structure substantially increases mechanical interaction and load transfer ability between the members through the intervening hardened and strengthened matrix.

Description

waited States Eatent Li [15] 3,663,356 [4 1 May 16, 1972 [54] REINFORCED METAL-MATRIX COMPOSITES Chou H. Li, 379 Elm Drive, Roslyn, NY. 1 1576 [22] Filed: Oct. 23, 1968 [21] Appl.No.: 769,938

[72] Inventor:

Primary Examiner-John T. Goolkasian Assistant Examiner-C. B. Cosby ABSTRACT The invention discloses the structure of, and method and apparatus for making, improved metal-matrix composites. Each of the composites comprises a strain-hardenable metal matrix, and a plurality of high-strength, high-modulus, elongated reinforcing members arranged in spatial relationship therein. Each of the members is effectively bonded to the surrounding matrix at discrete regions along its length, so as to locally strain the matrix around the discretely bonded regions upon application of load on the composite. Being thus locally strained, the matrix is spatially selectively and differentially strain-hardened and strengthened near the bonded regions. The spheres of influence of the straining and strain-hardening in the matrix near the bonded regions on one member are designed to significantly overlap similar spheres of influence near the neighboring members. This composite structure substantially increases mechanical interaction and load transfer ability between the members through the intervening hardened and strengthened matrix.

27 Claims, 4 Drawing Figures REINFORCED METAL-MATRIX COMPOSITES The invention relates to reinforced metal-matrix composites and their methods of manufacture, and more particularly to composites comprised of strain-hardenable metal-matrix with elongated reinforcing members dispersed therein and held in position thereby but selectively and differentially bonded thereto.

Laboratory tests have shown that metal-matrix composites, compared with metals, glasses, ceramics, plastics, etc., offer weight savings in aerospace or other structures of up to 45 percent or more. Composites with reinforcing members made of graphite, ceramics, or other refractory compounds promise to have high strengths, high modulii, and long-life high-temperature capabilities. Material scientists have now agreed that composites are not only here to stay, but will become ever more important, their uses expanding as rapidly as technologies permit. A list of matrix metals already tested include aluminum, magnesium, titanium, beryllium, iron, copper, silver,

- molybdenum, and nickel. The materials of the reinforcing members also have varied from boron, beryllium, graphite, sapphire, silicon carbide, boron carbide, to stainless steel.

At least dozens of laboratories are intensively working on these metal-matrix composites. One laboratory has even made titanium composite samples having ultimate tensile strengths of over 1,47 ksi, compressive strengths of 300 ksi, but not much ductility. Unfortunately, the field of metal-matrix composites is still not very far along. The advances being made so far have been largely empirically based. Different production methods have been tried. Such methods include vapor deposition, diffusion or roll bonding, liquid metal infiltration, unidirectional melt growth, powder metallurgy, electroforming, draw cladding, and plasma spraying. But all are beset by major manufacturing problems. The magnitude of these problems has been almost universally considered as insummountable for the next few years.

Foremost amount these problems is the difficulty of developing optimum bonding between the matrix and fibers or the reinforcing members. The bonding interfaces apparently play a very significant part, if not the key role, in the determination of the resultant composite properties. Thus, observed composite failures mostly occur at or near the matrixfiber interfaces. It is generally considered that a uniformly strong bonding is essential to the reinforcing mechanism. Too strong bonding, however, appears to render the composite severly limited by the matrix. It is also thought that the fibers must completely wet the matrix; yet chemical reactions, interdiffusions, or intermetallic compound formations often occur when they do wet, resulting in severe chemical degradations or even destructions of the reinforcing fibers.

Another problem with most present composite manufacturing processes is improper and inconsistent fiber alignment. Yet alignment of discontinuous fibers is very important in achieving high tensile strengths, particularly at high temperatures. Misalignment of even only 3 causes significant decreases in tensile strengths, particularly when short fibers are used.

A third serious problem is the physical or chemical degradation of the fibers. It is not unusual'for the extracted fibers to lose over 50 percent of their tensile strengths. Because the fibers are very fine in diameters, they have very little crosssectional areas and are, therefore, easily broken under transverse shears, such as are necessarily present in powder metallurgy, rolling, or other mechanical operations. Also, the diffusion rates, which govern many pertinent chemical reactions or brittle intermetallic compound formations, are often large compared to the fiber diameters, so that even within the processing time severe, locallized or general penetrations may result. Examples of such chemical degradations have been repeatedly observed. Slight interactions on sapphire fibers, for example, have shown to significantly reduce their strengths. Also, silicon carbide fibers are easily and completely dissolved in nickel matrices, though not in aluminum. In other cases, intermetallic phases form and grow at the fiber-matrix interfaces. The ductilities of the composites are then markedly reduced by the brittle compounds and surface irregularities on the fibers that result from such reactions. The composite strength is, of course, directly related to the unreacted fiber volume portion.

manufacturing, reinforced metal-matrix composites. The new composite comprises a strain-hardenable metal matrix, and a plurality of high-strength, high-modulus, elongated reinforcing members dispersed therein and held in position thereby. The dispersion of the members in the matrix is in a preselected spaced arrangement and at least a major portion of the members is each bonded efiectively .to the surrounding matrix differentially along its length. Differentially bonded regions on each such member selectively though nonunifonnly strain and strain-harden the surrounding matrix upon the application of external load on the composite, so as to substantially increase the efficiency of physical interaction and load transfer between the members through the intervening hardened and strengthened matrix. The method comprises preparing the requisite elongated reinforcing members, dispersing the members in a strain-hardenable metal matrix according to a preselected spaced arrangement, and selectively bonding the members to the matrix.

The objects of the invention are:

1. To overcome the aforementioned manufacturing problems and disadvantages of prior-art composites comprised of the usual high-strength, high-modulus reinforcing members embedded in the usual strain-hardenable metal matrices;

2. To provide improved reinforced, metal-matrix composites;

3. To achieve reproducible alignment and distribution of the reinforcing members in the composites;

4. To obtain optimum combination and structure of physical and chemical bonding between the matrix and fibers, for controlled load-transfer characteristics between the fibers and for improved overall performance of the composites;

5. To provide means to control the chemical bonding, reaction, or diffusion between the fibers and the matrix so as to reliably achieve high composite strength;

6. To provide unique, transverse protrusions or ridges on the reinforcing members to achieve novel, and useful physical and/or chemical results;

7. To control and relieve mismatch stresses and strains between the matrix and the fibers;

8. To provide means to increase the resistance of the fibers to transverse and dynamic stresses and strains;

9. To improve the life and reliability of high-temperature metal-matrix composites; and

10. To disclose novel processes and equipments to manufacture composites having the above-described desirable properties.

Further objects and advantages of my invention will appear as the specification proceeds.

The preferred form of my invention is illustrated in the accompanying drawing in which:

FIG. 1 is a front view showing a portion of the inside of a composite-making apparatus according to the invention;

FIG. 2 is a top view of the above apparatus taken along the cross-sectional line 2-2 of FIG. 1;

FIG. 3 shows the presence of protrusions or ridges, and their various shapes, on the reinforcing members; and

FIG. 4 shows the arrangements of the reinforcing members with their associated protrusions thereon for obtaining optimum reinforcing effects in the composites.

With reference to the drawing, there is shown in FIG. 1 one form of the apparatus for producing the metal-matrix composites according to the method of the invention. Here, a number of parallel, fine tungsten, tantalum, or other supporting wires 21 are arranged by means of guides 22 in a substantially uniform and equispaced, triangular or hexagonal pattern, such as is seen in FIG. 2. Non-metallic supporting wires are often equally useful. Metallic wires, however, are generally easier to handle and, furthermore, allow the passage of electrical currents therein for heating or electroplating. These wires are pulled from their respective spools (not shown) at their lower ends vertically upward at fairly uniform speeds to pass through a number of processing zones, i.e., Zones I to V, as shown. The preferred vertical movements eliminate gravity effects and result in symmetrical and uniform deposits or surface treatments to be discussed.

Zone I' is for wires preparation and may include such processing steps as mechanical cleaning by brushes or scrubbers 23, chemical cleaning including degreasing by organics 24 and oxide removal by acids 25, and vacuum cleaning such as vacuum heating by heaters 26 and/or ion bombardment. It is evident that suitable dust removal means or ducts 27, liquid seals 28, vacuum locks 29, fluid inlets and outlets, instruments, heating and pressurizingmeans, etc., must be provided for these operations.

Zone II is for depositing the reinforcing metallic or nonmetallic materials 32 The depositing method used here is chemical or vacuum vapor deposition from suitable sources 30, although it may also be electroplating or another method such asv pulling the supporting wires through a bath of suitable molten material. For a specified reinforcing material 32, skilled persons can readily select the appropriate deposition methods and estimate the optimum processing conditions. Preferably, each wire has its own deposition chamber, such as is outlined by partitions 31, so that the deposition thereon can be adjusted and controlled independently of the deposition on the other wires. According to one aspect of the invention, the depositing conditions are periodically changed, such as by varying the degree of energization on the sources 30, or, in other cases, the plating or heating currents in the supporting wires, or the deposition temperature and chemicals concentrations in the deposition chambers. These periodic changes in deposition conditions cause periodic variations in the deposition rates resulting, at constant supporting wires traveling speed, in transverse protrusions 33 being regularly and purposedly formed on the wires. The shape, size, and position of these protrusions on each wire can be exactly and independently programmed and controlled in a predetermined manner, account being taken of the shape, size, and position of similar protrusions on the neighboring wires. For most uses, these protrusions should be equal in size and shape but zigzag from one wire toward its neighbors along a longitudinal direction of the wires, as is shown in FIG. 1. After passing through the deposition zone, or Zone II, the wires become reinforcing, threadlike fibers, being covered thereon with the required amount of reinforcing materials 32.

In Zone III, the reinforcing fibers receive their post-deposition treatments, which may be simply a rinsing treatment after electroplating, or vacuum and/or thermal degassing, or, as shown, heating by heaters 41 to promote diffusional bonds between the supporting wires 21 and deposited reinforcing materials 32. The fibers may also receive other special treatments in Zone III. For example, in one such treatment, the reinforcing fibers are passed through sponges of proper sizes and elasticity and in proper spaced arrangement and, furthermore, wetted with a solution of chemical bond-promoting materials to improve the chemical bonding between the matrix and the reinforcing materials 32. A metallizing solution of molybdenium or tungsten oxide, for example, molybdenum trioxide, is particularly useful for promoting chemical bonding to metal matrices by many ceramic reinforcing fibers such as sapphire, beryllia, and graphite, or by many metallic reinforcing fibers having non-metallic oxides or other refractory materials on the surfaces. It is to be noted that these chemical bond-promoting materials are applied only to the protruded portions of the reinforcing members by the sponge technique described above. On fibers without protrusions, these chemical bond-promoters may be applied to selected portions on the fibers by, e.g., periodically compressing the wet sponges against the fibers, or by using suitable sources 42 of the bondpromoting materials together with masks, shutters, or programmed periodic source energizers. Greases, oils, or other special materials contained in these same sponges or sources may also be applied as masks or shields only to selected portions so that only portions between these selected portions are now susceptable to a subsequent treatment. This treatment may comprise applying a surface layer by chemical reaction or electrodeposition to prevent wetting and to reduce chemical bonding between the matrix and the non-protruded portions. Another post-deposition treatment in Zone III may be to apply, either selectively or uniformly, a soft, plastic, metallic cushioning surface layer of substantial thickness to increase the resistance of the reinforcing members to transverse and dynamic stresses and strains.

The reinforcing fibers prepared in the above manner may be used as such according to conventional fiber-matrix integration techniques, such as pressing, liquid infiltration, extrusion, and the like. However, according to another aspect of the invention, the reinforcing members or fibers coming out of Zone III are passed directly into Zone IV for the application, as a part of a continuous process, of the relatively soft and plastic but strain-hardenable metal matrix 34 thereon, by liquid penetration, electroplating, vapor or vacuum deposition, such as from the sources 43 as shown. Special shields, and liquid, vapor, or vacuum locks, matrix metal sources, etc., may be required here, but these are also easily designed by persons skilled in the art.

By using special matrix metal sources, masks or shields, or deposition control means, the matrix-coated reinforcing fibers may come out of Zone IV as smooth or fairly smooth-surfaced rods. However, according to another aspect of the invention, these rods are purposedly to have enlarged nodules corresponding to the protrusions 33 on the reinforcing fibers. These enlarged nodules achieve the following results. Firstly, they facilitate zigzagging the protrusions from one reinforcing member to its neighbors when these fibers or members are densified. Secondly, they allow regulated pre-straining of the metal matrix between the protrusions or nodules during the composite compaction process. Thirdly, the nodules achieve much greater strain-hardening in the matrix close to them than otherwise possible, while still retaining softness or ductility in other portions of the matrix distant from the nodules. By regulating the shape, size, and spacing of the protrusions or nodules on the reinforcing members, and the thickness of the matrix metal thereon, the strain-hardening effect can be maximized.

The last zone, i.e., Zone V, is for final preparation, which may include such steps as mechanical or chemical cleaning of the matrix-deposited reinforcing members, densification by extrusion or pressing, thermal treatments (for degassing-and controlling the grain-size of the matrix or fiber material, or for promoting bonding only at selected areas between the matrix and reinforcing fibers), conditioning the composite including pre-strain-hardening selective portions of the matrix, packaging, and the like. In FIG. 1, the reinforcing fibers with the matrix metal vacuum-deposited thereon from the deposition sources 43 are seen to be guided by guides 44 to pass through a heating zone heated by heaters 45 into densifying die 46. Finished composite 47 then comes out of the exit end of the densifying die 46. This densifying die may be simply an extrusion die.

In summary, in the preferred mode described above, metallic or non-metallic supporting wires 21 are prepared in Zone I of the equipment of FIG. I by mechanical, chemical, and

vacuum-thermal means. The prepared wires then enter into Zone II for deposition of the reinforcing material 32, and preferably also for formation of the spaced protrusions 33 layer by a sponge technique. Zone IV is for the deposition of 5 the matrix metal 34 on the reinforcing member. The final preparation" Zone V may be designed for cleaning, densifying, heating, conditioning, packaging, and/or the like.

It can be seen that the above scheme completely solves the persistent problems of improper fiber alignment and distribution. Each reinforcing fiber now has exactly six closest neighbors in the above scheme at substantially the same distances apart. This results in substantially balanced transverse forces on each fiber from its neighboring fibers, thereby avoiding fiber breakage. It also avoids, on the one hand, too much spacing between reinforcing fibers so that the efficiency of mechanical interaction or load transfer between them is impaired or nullified and, on the other hand, too closely-spaced fibers so that the full strain-hardening effect of the matrix surrounding the fibers is not achieved thereby resulting in failure of the intervening matrix because of localized high stresses and strains. Full strain-hardening of the matrix, according to this invention, is achieved when there are present therein large, controlled and regularly-spaced, and preferably uniformly-shaped strain-hardened zones of continually increasing straining intensity toward the chemically bonded portions on the fibers.

The protrusions 33 which are purposedly introduced in the above case, or the chemically-bonded portions selectively arranged on the protruded or non-protruded fibers, enable each such fiber to be bonded effectively to the surrounding matrix differentially along its length to selectively and differentially strain and strain-harden the surrounding matrix upon the application of external load on the composite, so as to substantially increase the efficiency of mechanical interaction and load transfer between the fiber and its neighbors through the intervening hardened and strengthened matrix. The less strained matrix between these strain-hardened portions remain relatively soft and plastic to allow therein some stress and strain adjustments or energy absorption under the load thereby preventing microcracks from forming or growing and hence reducing composite failures.

It is to be particularly noted that in the above example, the bonding between each fiber and its surrounding matrix varies in a systematic, predetermined manner periodically along the length of the fiber, the periodic distance being determined by the distance between two neighboring, similarly bonded portions on the fiber or, in the case of protruded fibers, the distance between two neighboring protrusions on the same fiber.

Preferably, eachof the fibers is at least partly chemically bonded to the surrounding matrix to prevent complete boundary separation between the fibers and matrix.'Yet complete chemical bonding, though specified, taught, or implied in prior-art composites, is neither necessary nor desirable. This is not only because efiective load transfer between fibers can be achieved without complete chemical bonding, but because the matrix and fibers can never be exactly matched, both statically and dynamically, in thermophysical properties. Complete chemical bonding between the matrix and fibers, therefore, inevitably results in build-up of substantial mismatched stresses and strains, and often leads to premature failures during services, particularly under dynamic, mechanical or thermal loading conditions. On the other hand, imperfect chemical bonding at well-chosen or selected portions, such as on the non-protruded portions of protruded fibers, though against ordinary use and teaching, relieves such substantial mismatch stresses and strains and actually improves the overall performance and reliability of the resultant composites. Mere physical bonding is often enough between the chemically bonded discrete regions so that in this case the bonding between each fiber and its surrounding matrix is alternately chemical and physical in character.

As indicated above, a well-known problem in present composite making is fiber degradation through physical breakages and/or chemical reaction or diffusion between, the matrix and fibers. The soft plastic, metallic cushioning layers on the fibers previously described are specifically designed to minimize physical breakages. According to yet another aspect of the invention, on selected portions of the fiber where perfect chemical bonding is not required, an inert surface layer is purposely applied, as mentioned previously. An example of such an inert surface layer is titanium carbide on titanium, to prevent the diffusion of iron thereinto.

Composites according to the above invention can be achieved with reinforcing fibers having or having no transverse protrusions thereon. When no protrusions are needed, the same manufacturing procedures described previously in connection with FIG. 1 may be employed, except that there are then no periodic variations in the deposition conditions for the reinforcing materials 32 and that the fibers come out of Zone II roughly as smooth-surfaced threadlike reinforcing fibers.

However, the protrusions achieve many novel and important purposes: Firstly, they effectively strain and strain-harden the surrounding matrix, thereby raising the load that can be transferred between the fibers through the intervening strengthened matrix. This achieves a new and useful result, which is particularly evident in the matrix located among oppositely compressing or shearing protrusions on neighboring fibers (See FIG. 4). Secondly, the protrusions facilitate and simplify the regulation and control of differential bonding between the matrix and fibers, described above as an essential feature of my new mode or principle of composite reinforcing mechanism. Thirdly, the protrusions prevent the reinforcing members from being pulled out of the composite, particularly at high temperatures. Fourthly, the protrusions provide bearing areas so that the load transfer between fibers are not limited to interfacial shears that tend to break the fiber-matrix bonds, but also at leastly partly taken up by compressions, to which the matrix has much greater resistance to failure than shear. Fifthly, according to another aspect of the invention, only surfaces of the protrusions on the fibers are chemically bonded to the matrix, so that the non-protruded portions on the fibers are prevented from chemically reacting or diffusing with the matrix. In general, such chemical reactions or diffusions follow well-known diffusion laws so that for a given system under a specified service (temperature) conditions, the time to penetrate a given depth is roughly proportional to the square of the diameter. Since the protruded portions have diameters of, typically 2 to 3 times the fiber diameter, the time to penetrate and fail at the protruded portions is about 4 to 9 times greater than at the non-protruded portions. That is, by having protrusions on the reinforcing fibers and limiting chemical bonding only to the protruded portions, the hightemperature life of the composite will often be increased manyfold.

The protrusions should generally have no sharp edges or comers that may act as stress raisers. Each of the protrusions may have a shape symmetrical with respect to a plane normal to the fiber, or have no such symmetrical shape with respect to every such plane. in some cases, protrusions having non-symmetrical shapes are desirable so that the load transfer efficiency from one fiber to its neighbors in one of the longitudinal directions is substantially higher than that in the other direction. It is also desirable to have the direction of high load transfer efficiency alternate from one fiber to its neighbors, such as is shown in FIG. 4b with the tensile forces indicated thereon. These arrangements achieve the oppositely compressing and shearing, and the resultant great strain-hardening, effects on the matrix.

The protrusions should not be too large in sizes because then the thin fibers are frequently subjected to large bending stresses from loads applied on the flanges of the protrusions. Neither should the protrusions be too small, or the beneficial effect of protrusions is not fully exploited. For optimum results, these protrusions should have diameters of about 1.5 to 4 times those of the fibers. Since each protrusion must optimally interact not only with protrusions on neighboring fibers but also with those on the same fiber, the spacing of the protrusions on the fiber should be less than 10 or times the fiber diameters, otherwise such interactions become negligible or insignificant. On the other hand, too closely spaced protrusions on the fiber, such as less than 3 times the fiber diameter, result in too small and insufficiently developed strainhardened zones in the matrix. For similar reasons, the spacing between reinforcing fibers in the composite should roughly be between 3 to times the fiber diameter for optimal results, being also somewhat dependent on the height of the protrusions relative to the fiber diameters. The best size of protrusions, or spacing between protrusions and fibers, varies from instance to instance, being also dependent on the strainhardening behavior of the matrix.

The optimum volume percentage of reinforcing fibers can be determined according to the following rule to achieve optimal strength by insuring equal failure rates in the fibers and in the surrounding strain-hardened matrix. Thus, the effective area of the hardened and strengthened matrix in the strainhardened zone surrounding the fiber (at the chemically bonded or protruded portions) multiplied by the ultimate tensile strength of the strain-hardened matrix should equal the cross-sectional area of the reinforcing fibers, or at the nonprotruded portions thereon if protrusions are present, multiplied by the ultimate tensile strength of the fibers.

Each of the reinforcing fibers in the above example comprises a metallic or non-metallic wire core 21 for support, and an outer layer of the reinforcing material 32. This material may have the same composition as the core so that the boundary between the core and the outer layer may be physically or chemically indistinguishable. If, in addition, the core material is single-crystalline, under suitable conditions, the outer layer can also be made single-crystalline and crystallographically indistinguishable from the core, thereby resulting in single-crystalline fibers. Evidently, even commercially available fibers or other reinforcing members can be fed directly into Zone IV or V of the equipment of FIG. 1, thereby simplifying the equipment by eliminating the Zones I to III or IV.

Single-crystalline fibers or whiskers can also be grown without the use of core wires at all. This has been done by unidirectional eutectic melt growth. Another method comprises nucleating with, e.g., a special catalytic impurity vacuum-deposited on a substrate according to a hexagonal pattern; growing the whiskers by chemical or vapor depositions, from chemical solutions, by electroplating, or otherwise such as by the vapor-liquid-solid growth mechanism (See J. Metals 16, 761, 1964); and possibly even keeping the whiskers aligned by mechanical pulling or by electrostatic means.

Although only parallel or non-parallel threadlike fibers are almost exclusively described above and used as the reinforcing members; sheets, with or without protrusions or ridges thereon but preferably nearly parallel and equally thick and spaced for obvious strength reasons, can also be used as reinforcing members according to the principle of reinforcement by differential strain-hardening of the matrix as described above. The equipment for making sheet-reinforced composites may be very similar to the one shown in FIG. 1. While parallel fibers-reinforced composites are excellent to resist tension in the direction of the fibers, parallel sheet-reinforced composites admirably withstand tensions in any direction in addition to bending, provided the stresses are in planes parallel to the reinforcing sheets. The sizes and spacings of the protrusions be very similar to the one shown in FIG. 1. While parallel fibers-reinforced composites are excellent to resist tension in the direction of the fibers, parallel sheet-reinforced composites admirably withstand tensions in any direction in addition to bending, provided the stresses are in planes parallel to the reinforcing sheets. The sizes and spacings of the protrusions or ridges, and of the sheets, can be similarly determined for optimum results.

These reinforcing sheets and fibers can even be in the form of a eutectic compound grown from a suitable melt. In melt eutectic growth, the periodic transverse protrusions or ridges can be obtained by periodically varying the melt growth conditions, such as by changing the melt temperature by, say, 10 to 20 C, or temperature gradient, or local impurity concentrations in the melt. Periodically varying the magnetic fields transversely of the direction of growth over the melt changes the intensity of electromagnetic convection damping effects, and also causes variations in transverse cross-sections on the grown fibers or sheets. One can even control differential chemical bonding to such melt-grown reinforcing eutectic members (fibers or sheets) by adding to the original melt a chemical bond-promoting or bond-'mhibiting melt-segregatable impurity and allow the impurity to segregate on the protrusions or other selected portions of the fibers or sheets, care being taken to adjust the growth conditions in relation to eutectic solidification and impurity segregation.

FIG. 3 shows three basic forms or shapes of the protrusions or ridges on the reinforcing members. These forms are: square wave (3a), triangular wave (3b), and elliptical wave (30). These forms can also be distorted, e. g., the distorted partial ellipse (3d) and distorted square (3e). The last form (3e) can also be considered as a combination of the square and triangular forms. Other combination forms include two oppositely tapering or triangular forms (3]), tapering with superposed el lipse (3g and 3h), and tapering with superposed square (3i).

FIG. 4 shows two arrangements whereby neighboring reinforcing members of the protruded type are arranged to achieve oppositely compressing and/or shearing effects on the intervening matrix. When a reinforced composite of this invention is loaded, some pairs of neighboring reinforcing members must be oppositely tensioned, and the situation depicted in or similar to FIG. 4 always results. FIG. 4a is a useful arrangement in which the load-transfer efficiency is indepen dent of the direction of the tensile forces, i.e., whether as shown or oppositely directed. On the other hand, in the arrangement of FIG. 4b, the same efficiency depends on the direction of tensile forces on the reinforcing members. Notice that here these members have combination protrusions, i.e., triangular protrusions on tapering bases. Further, for each member the direction of easy load transfer on the triangular protrusion is directly opposite to that for the tapering base. Specifically, the triangular protrusions cause the top member of FIG. 4b to have high load transfer efficiency if this member is pulled to the right as is indicated, but the tapering base has low load transfer efficiency for the same pulling force on the same top member of FIG. 4b.

When a composite with protruded reinforcing members dispersed therein is subjected to mechanical working, such as rolling or extrusion, differential bonding between the members and matrix occurs. Further, such mechanical operations can be regulated so that the character and degree of bonding, and the shape, size, and spacing of the bonded portions can be exactly controlled and programed. Also, not only is the bonding better and more chemical in the protruded portions than in the non-protruded portions but, on the same protrusion, the side first coming out of the rolling or extrusion die is better bonded than the other side. When the mechanical working or area reduction is sufficiently severe, such as between 40 to percent of the allowable straining and strain-hardening without failing the matrix; and when the straining is sufficiently rapid, i.e., if the rolling or extruding speed is comparable with or exceeds the thermal diffusion rate in the matrix, then substantial chemical bonding results from the almost adiabatic heating by heat converted from the plastic deformation work. The composite with protruded reinforcing members dispersed therein and coming out of the extrusion die 46 of FIG. 1 can, therefore, be selectively and differentially bonded, either physically or chemically, in respect to the bonding between the members and matrix, if suitable processing procedures are adopted. However, if the reinforcing material 32 or matrix metal 34 is easily or heavily oxidized, then special chemical bond-promoters are desirable to insure positive bonding. Even with the extruded or rolled composites containing nodulated members, differential bonding is further intensified upon application of load on the composite. Rolling and extruding do strain-harden and strengthen the matrix metal around the nodules, but according to a different type of force system and resulting in a different spatial straining pattern. This is'because the forces in rolling and extrusion are predominantly lateral, rather than longitudinal as usually occurs during service of the composites. That is, the extruded or rolled composites containing nodulated reinforcing members go through additional processes of matrix differential straining and strain-hardening, upon load application on the composites.

It is to be reemphasized that composites according to the present invention achieve high load-transfer efficiencies between the reinforcing members and matrix through controlled, selective and differential straining and strain-hardening of the matrix, a natural result of regulated, selective and differential bonding between the matrix and the reinforcing members dispersed therein. Such a reinforcing mechanism is completely different from those involved in fiberglasses or reinforced concretes. In these materials, the matrix is not strain-hardenable and, in contrast to the composites herein disclosed, merely acts as glue in fiberglasses while carries practically all the load in reinforced concretes. In both these cases, the load transfer between the reinforcing members is limited by the low shear strengths of the unchanging, relatively weak matrices. Even prior-art metal-matrix composites greatly differ from my composites in structure, result, and mode of operation, because they lack controlled and optimum strengthening of the matrix, protrusions on the reinforcing members, regulated differential bondings, alternate chemical and physical bondings, intentionally imperfect chemical bondings at periodic intervals to relieve mismatch stresses and strains, and the like.

It is to be particularly noted that in my composites the load transfer is not directly from one reinforcing member to another, but indirectly through the intervening strainhardened and strengthened matrix. Further, the strain-hardening and strengthening in the matrix is not to be abrupt (such as is obtained by many other methods of composite manufacture including soldering, brazing, and welding fibers together), but gradually varies in intensity, both laterally and longitudinally, reaching maximum values at near the centers of the chemically bonded matrix-member interfaces. Also, the matrix metal is alternately hard and soft along the length of each discretely bonded member, hard where strength is needed to insure efficient load transfer from member to member while soft where some ductility is required to allow stress and strain relief.

The invention is not to be construed as limited to the particular forms disclosed herein, since these are to be regarded as illustrative rather than restrictive. Thus, while it is possible to achieve most or all of the 10 objectives of the invention with the equipment and process described herein, merely applying the differential bonding principle of composite reinforcement allows one to achieve a few of the stated objectives (e.g., high load transfer, great strength, high temperature capability), simultaneously and with either modulated or nonnodulated fibers together with conventional production equipment but with modified bonding techniques. Also, composites may be made partly according to the present method, e.g., having a substantial or major portion of the reinforcing members differentially bonded to the matrix, to achieve partially benefial results.

What is claimed is:

1. A reinforced metal-matrix composite comprising a strainhardenable metal matrix, and a plurality of high-strength, high-modulus, elongated reinforcing members arranged in spatial relationship therein, at least a substantial portion of the members being each effectively bonded to the surrounding matrix at discrete regions along its length so as to locally strain the matrix near the discretely bonded regions upon application of external load on the composite in the general direction of the members whereby the matrix in the space between the members is selectively though nonuniforrnly strain-hardened and strengthened, the spheres of influence because of the straining and strain-hardening in the matrix within the space near the bonded regions on one member of the portion significantly overlapping similar spheres of influence near the neighboring members of the same portion so as to substantially increase mechanical interaction and load transfer ability between the one member and the neighboring members through the intervening hardened and strengthened matrix.

2. The composite of claim 1 wherein the portion is a major portion that approaches percent.

3. The composite of claim 1 wherein the portion is a major portion and the effectively bonded discrete regions are substantially similarly-shaped and equally spaced along the length of each member of the portion.

4. The composite of claim 3 wherein the ratio of the effective area of the hardened matrix in the strain-hardened space to the minimum effective cross-sectional area of each member of the major portion is substantially equal to the ratio of the ultimate tensile strength of the member to that of the strainhardened surrounding matrix.

5. The composite of claim 3, wherein the members of the major portion are substantially parallel and oriented generally in the direction of the intended principal tensile loading on the composite when in use.

6. The composite of claim 1 wherein the portion is a major portion and wherein the bonding is at least partly due to chemical reaction and is also strong enough to prevent complete boundary separation between each member of the major portion and its surrounding matrix.

7. The composite of claim 6 wherein the bonding along each member of the major portion to its surrounding matrix is alternately chemical and physical in character.

8. The composite of claim 6 wherein the matrix around each member of the major portion includes therein regularlyspaced strain-hardened zones of gradually increasing straining intensity toward the chemically bonded discrete regions on each such member.

9. The composite of claim 6 wherein each member of the major portion has a chemical bond-promoting layer in each of the discretely bonded regions.

10. The composite of claim 9 wherein each member of the major portion has an outer surface consisting essentially of a non-metallic, refractory material and the bond-promoting layer consists essentially of a bond-promoting oxide of a metal selected from the group consisting of tungsten and molybdenum.

11. The composite of claim 6 wherein the transverse section thickness of each of the effectively and discretely bonded members varies periodically along the length of the member.

12. The composite of claim 11 wherein each member of the major portion has thereon substantially equal-sized transverse protrusions periodically along its length.

13. The composite of claim 12 wherein the protrusions are spaced apart on each protruded member at a distance of from 3 to 15 times the minimum thickness of the protruded members, the maximum lateral dimension of the protrusions is between 1.5 to 4 times the minimum thickness, and the spacing between neighboring protruded members is between 3 to 20 times the minimum thickness.

14. The composite of claim 12 wherein the chemical bonding occurs exclusively on the surfaces of the protrusions.

15. The composite of claim 12 wherein each protruded member has on the non-protruded portions thereof surface areas wherein the bonding to the matrix is imperfect to the extent to allow absorption therein of thermoplastic mismatch stresses and strains between these members and the matrix.

16. The composite of claim 12 wherein the protruded members are nearly parallel to one another and the transverse protrusions on these members zigzag from one member toward its neighbors along a longitudinal direction of these members.

17. The composite of claim 12 wherein each of the protruded members has on each non-protruded portion thereof a protective inert layer thereon to prevent the nonprotruded portions from chemically bonding to and reacting with the surrounding matrix.

18. The composite of claim 12 wherein each of the protrusions has a symmetrical shape with respect to a plane normal to the parent protruded member.

19. The composite of claim 11 wherein the periodic transverse thickness variation is in the form of two tapers oppositely directed longitudinally of each said member toward a common transverse plane having therein an extreme value in the transverse section thickness.

20. The composite of claim 11 wherein the periodic transverse thickness variation comprises a uniform taper.

21. The composite of claim 11 wherein the periodic transverse thickness variation includes within each period a transverse plane of substantially constant but extreme section thickness, and wherein each of the effectively and discretely bonded regions is centered on the plane.

22. The composite of claim 1 wherein the reinforcing members are in the form of fibers.

23. The composite of claim 1 wherein the reinforcing members are in the form of sheets.

24. The composite of claim 23 wherein each sheet of the substantial portion has thereon substantially equally spaced transverse protrusions.

25. The composite of claim 23 wherein each sheet of the substantial portion has thereon transversely protruding ridges.

26. A reinforced metal-matrix composite comprising a strain-hardenable metal matrix, and a plurality of highstrength, high-modulus, elongated reinforcing members arranged in spatial relationship therein, at least a substantial portion of the members forming selected members each of which is effectively bonded chemically to the surrounding matrix at discrete regions along its length, the matrix in the space between the selected members being selectively though non-uniformly strain-hardened and strengthened near the discrete regions, the spheres of influence because of the strainhardening in the matrix within the space near one selected member significantly overlapping similar spheres of influence near the neighboring selected members so as to substantially increase mechanical interaction and load transfer ability between the selected members through the intervening hardened and strengthened matrix.

27. The composite of claim 26 wherein each selected member has along its length substantially equally spaced transverse protrusions generally zigzagging from the one selected member toward its neighborhood selected members.

Claims

2. The composite of claim 1 wherein the portion is a major portion that approaches 100 percent.
3. The composite of claim 1 wherein the portion is a major portion and the effectively bonded discrete regions are substantially similarly-shaped and equally spaced along the length of each member of the portion.
4. The composite of claim 3 wherein the ratio of the effective area of the hardened matrix in the strain-hardened space to the minimum effective cross-sectional area of each member of the major portion is substantially equal to the ratio of the ultimate tensile strength of the member to that of the strain-hardened surrounding matrix.
5. The composite of claim 3, wherein the members of the major portion are substantially parallel and oriented generally in the direction of the intended principal tensile loading on the composite when in use.
6. The composite of claim 1 wherein the portion is a major portion and wherein the bonding is at least partly due to chemical reaction and is also strong enough to prevent complete boundary separation between each member of the major portion and its surrounding matrix.
7. The composite of claim 6 wherein the bonding along each member of the major portion to its surrounding matrix is alternately chemical and physical in character.
8. The composite of claim 6 wherein the matrix around each member of the major portion includes therein regularly-spaced strain-hardened zones of gradually increasing straining intensity toward the chemically bonded discrete regions on each such member.
9. The composite of claim 6 wherein each member of the major portion has a chemical bond-promoting layer in each of the discretely bonded regions.
10. The composite of claim 9 wherein each member of the major portion has an outer surface consisting essentially of a non-metallic, refractory material and the bond-promoting layer consists essentially of a bond-promoting oxide of a metal selected from the group consisting of tungsten and molybdenum.
11. The composite of claim 6 wherein the transverse section thickness of each of the effectively and discretely bonded members varies periodically along the length of the member.
12. The composite of claim 11 wherein each member of the major portion has thereon substantially equal-sized transverse protrusions periodically along its length.
13. The composite of claim 12 wherein the protrusions are spaced apart on each protruded member at a distance of from 3 to 15 times the minimum thickness of the protruded members, the maximum lateral dimension of the protrusions is between 1.5 to 4 times the minimum thickness, and the spacing between neighboring protruded members is between 3 to 20 times the minimum thickness.
14. The composite of claim 12 wherein the chemical bonding occurs exclusively on the surfaces of the protrusions.
15. The composite of claim 12 wherein each protruded member has on the non-protruded portions thereof surface areas wherein the bonding to the matrix is imperfect to the extent to allow absorption therein of thermoplastic mismatch stresses and strains between these members and the matrix.
16. The composite of claim 12 wherein the protruded members are nearly parallel to one another and the transverse protrusions on these members zigzag from one member toward its neighbors along a longitudinal direction of these members.
17. The composite of claim 12 wherein each of the protruded members has on each non-protruded portion thereof a protective inert layer thereon to prevent the non-protruded portions from chemically bonding to and reacting with the surrounding matrix.
18. The composite of claim 12 wherein each of the protrusions has a symmetrical shape with respect to a plane normal to the parent protruded member.
19. The composite of claim 11 wherein the periodic transverse thickness variation is in the form of two tapers oppositely directed longitudinally of each said member toward a common transverse plane having therein an extreme value in the transverse section thickness.
20. The composite of claim 11 wherein the periodic transverse thickness variation comprises a uniform taper.
21. The composite of claim 11 wherein the periodic transverse thickness variation includes within each period a transverse plane of substantially constant but extreme section thickness, and wherein each of the effectively and discretely bonded regions is centered on the plane.
22. The composite of claim 1 wherein the reinforcing members are in the form of fibers.
23. The composite of claim 1 wherein the reinforcing members are in the form of sheets.
24. The composite of claim 23 wherein each sheet of the substantial portion has thereon substantially equally spaced transverse protrusions.
25. The composite of claim 23 wherein each sheet of the substantial portion has thereon transversely protruding ridges.
26. A reinforced metal-matrix composite comprising a strain-hardenable metal matrix, and a plurality of high-strength, high-modulus, elongated reinforcing members arranged in spatial relationship therein, at least a substantial portion of the members forming selected members each of which is effectively bonded chemically to the surrounding matrix at discrete regions along its length, the matrix in the space between the selected members being selectively though non-uniformly strain-hardened and strengthened near the discrete regions, the spheres of influence because of the strain-hardening in the matrix within the space near one selected member significantly overlapping similar spheres of influence near the neighboring selected members so as to substantially increase mechanical interaction and load transfer ability between the selected members through the intervening hardened and strengthened matrix.
27. The composite of claim 26 wherein each selected member has along its length substantially equally spaced transverse protrusions generally zigzagging from the one selected member toward its neighborhood selected members.