EP1268105A1

EP1268105A1 - Method for manufacturing metal parts

Info

Publication number: EP1268105A1
Application number: EP01929442A
Authority: EP
Inventors: Wolfgang Kochanek
Original assignee: Endrich Manfred
Current assignee: Endrich Manfred
Priority date: 2000-03-24
Filing date: 2001-03-22
Publication date: 2003-01-02
Anticipated expiration: 2021-03-22
Also published as: EP1268105B1; CA2424733A1; US20040067152A1; JP2003528979A; ATE267655T1; CA2424733C; DK1268105T3; US6939509B2; WO2001072456A1; DE50102410D1; ES2222991T3; AU2001256212A1; DE10014403A1

Abstract

The invention relates to a method for producing metal bodies by mixing metal compound particles with a binding agent and compacting them to form shaped parts, whereby the binding agent is subsequently removed and the metal compound is reduced to metal by gassing the same with a reductive gas at high temperatures. The reduction is carried out at temperatures lower than the sintering temperature of the reduced metal compound so that the formed reduced shaped bodies, while maintaining their original dimensions to a large extent, have a density less than that of the metal compound used. The metallic matrix obtained in such a manner can be subjected to a post compaction by applying mechanical forces. High-strength steel parts having very low residual porosities and a high geometric tolerance are obtained by sintering the matrix at usual sintering temperatures after compaction.

Description

Process for manufacturing metal parts

The present invention relates to a powder metallurgical process for the production of metal parts.

Metal parts manufactured using powder metallurgy are used to a considerable extent in the automotive, electrical appliance and lock industry. There are two main production processes: the classic press sintering technique (PM) with the special case of sinter forging and the metal powder injection molding process (MIM).

The parts accessible by the classic PM process are characterized by simple geometries, which are unidirectionally pressed from relatively coarse powders and avoid thin bars, narrow bores, as well as bevels and undercuts. Typical part weights range from a few grams (e.g. locking nuts in the lock industry) to around one kilogram in the automotive sector (e.g. oil pump rotor, sprockets, ABS sensors). The manufacturing costs of such parts are low. In addition to the shape restriction mentioned, the low mechanical strength of classic PM parts is particularly disadvantageous. They generally have densities below 7 g / cm ³ and thus have a considerable pore volume. This leads to a strong notch sensitivity, which does not allow the use of classic PM parts in applications subject to alternating loads (e.g. high-speed gearwheels in gearboxes). Although it is possible to increase the densities to values in the range from 7 up to 7.2 g / cm ³ by techniques of double-press sintering, an almost pore-free matrix with material densities above 7.4 g / cm ³ can only be achieved by the complex sinter forging , In order to increase the unsatisfactory material density of classic PM parts, attempts have also been made to add finely divided metal powders (eg carbonyl iron powder) to the coarse PM powders to improve the sintering activity. In addition to the high raw material costs and problems of segregation, these approaches have so far failed due to the fact that the penetration of very fine powder particles into the gap between the punch and the die means that the pressing tools are subject to high wear.

The principle of the metal powder injection molding (MIM) process shows a fundamental way out, which has become increasingly important in industrial production for the serial production of geometrically complex metal parts in the last 10 years. Despite material densities above 7.4 g / cm ³ and the associated good mechanical strength, the use of these parts has so far been limited. The reasons for this limitation are initially the high raw material costs for the fine-particle metal powder required, which limits the economic limit with regard to competing manufacturing processes to part weights below approx. 50 g. In addition, MIM parts shrink considerably during the manufacturing process, resulting in a maximum controllable part size. Taking normal tolerance specifications into account, this is a diameter of approx. 50 mm. For the reasons mentioned, a typical MIM part has a weight of approx. 2 to 20 g and is significantly above the price level of classic pressed sintered parts in terms of manufacturing costs.

In the process of metal powder injection molding, finely divided metal powders (particle diameter typically <22μ, 90% point) are kneaded with a binder to a homogeneous mass (feedstock). The proportion of binder is Dependence on particle density and morphology in general at 5 to 15 wt.%. The binder, which no longer occurs in the actual end product (sintered steel part), takes on the task of encasing the metal particles in the process sequence and making the mass homogeneously flowable without any noteworthy segregation. Most of the binders used industrially are based on the interaction of the following three components, as explained below: removable component (K1), polymer (K2) and surface-active auxiliary (K3).

This feedstock, which has the flow properties of a filled thermoplastic, is processed to shaped bodies (green compacts) on conventional injection molding machines, this sub-step of the process corresponding to the molding principles of plastic injection molding and thus allowing the production of geometrically complex shaped bodies. In a subsequent process step with component K1, the predominant part of the binder is removed from the green compact of the process, a porous part being obtained (brown compact), the external geometry of which is practically identical to that of the green compact and the shape of which is formed by a polymer (component K2 ) is held together. By removing component K1, pores are created which, during the subsequent pyrolysis of the polymer skeleton, allow the pyrolysis gases that are formed to escape to the outside without building up internal pressure (i.e. without damaging the component through bubbles and cracks). Both binder mixtures are known in which K2 and K1 are homogeneously soluble in one another, and those in which these form two discrete phases after cooling / Lit5 /.

Component K1 can be removed thermally, chemically, microbially or solvent-based. For example, processes are described in which component K2 comprises polymers from the following classes: polyolefins, polystyrene, polyamides, acrylates, cellulose acetate, polyacetals. In order to largely suppress segregation between binder and particle phase when the feedstock is sprayed, a further component K3 is often added to the binder as a flow improver which has surface-active properties and thus allows the surface of the metal particles to be wetted as homogeneously as possible.

The brown product obtained is then sintered in the presence of H2, H2 / N2 mixtures or in vacuo at temperatures below the melting point of the alloy. The components K2 and K3 decompose and the Braunling shrinks under internal compression by the original volume fraction of the binder. The shrinkage in the x, y, z direction is approximately isotropic and, depending on the binder content and composition, is approximately 13-20%. For a given geometry of the sintered part, the green part is accordingly to be designed in x, y, z with a length extension of SF = 1.13 to 1, 20.

The origins of this concept go back to K. Schwarzwalder, who described the manufacture of ceramic spark plugs in 1937, as well as work in the 1940s in the manufacture of components for uranium enrichment plants based on the shaping of nickel powder using organic binders / Lit2 /. However, this concept was only taken up and implemented industrially at the end of the 1970s with the patents of Rivers / Lit3 / and Wiech IUXΛI. Since the basic patents for the manufacturing concept of molding injection-molded metal or ceramic powders using organic binders have expired or their concepts have been described in previous work, the patent situation can be regarded as free. Accordingly, commercially available binder mixtures are now offered on the market by various providers.

As an example of the multitude of industrial processes, three concepts are mentioned here that outline the possibilities. EP-PS 125 912 describes a method in which K2 one of the above-mentioned thermoplastics is processed with a wax K1.

EP 0465 940 B1 is a thermoplastic composition in which component K2 is a polyolefin and K1 is a polyoxymethylene, K1 being removed by acid catalysis and K2 then being expelled pyrolytically.

D-OS 38 08 123 describes a process in which the binder consists of K2 = polyethylene and K1 = oleic acid decyl ester, an ethylene-acrylic acid copolymer being added as an additive for promoting adhesion between the metal particles and the binder. K1 is made from the Braunling by solvents e.g. Alcohols and or chlorinated hydrocarbons are extracted.

The pyrolysis of carbon-containing binder components (K1 or K3) often results in an uncontrolled incorporation of carbon into the matrix of the metal particles, particularly with larger wall thicknesses. Since in iron-based alloys C contents above approximately 0.9% by weight significantly deteriorate the mechanical properties, OZ 0050/40736 describes a special "process for dewaxing and improving the properties of injection molded metal parts". Here the addition of 2-30 % By weight, preferably 4 to 10% by weight of a high surface area carbonyl iron oxide with a specific surface area of 10 to 120 m ² / g, preferably 70 to 110 m ² / g, which is added to the binder by intensive grinding with the metal powder this reduces the incorporation of carbon into the metallic matrix by reducing the amount of carbon formed by binder pyrolysis.

Although the process of metal powder injection molding opens up a variety of technical possibilities, the comparatively high raw material costs with increasing part weight increasingly limit its economic efficiency compared to competing manufacturing processes. For example, parts with a weight of approx. 20 g or more are generally accessible more cheaply via the investment casting, since its raw material costs (costs of the molten metal) are significantly lower than the high costs of the finely divided powders required in the MIM process. This disadvantage becomes even more serious if one takes into account that with the MIM method, the system-inherent shrinkage, in particular in the case of larger parts, leads to unsatisfactory statistics of the final dimensions, which means that an increased scrap must be taken into account.

US 4,445,936 and US 4,404,166 describe one possibility of increasing the accuracy of MIM parts by compressing them after inserting into a press die after sintering at 2150 ° F (1177 ° C) with plastic deformation of the metallic matrix formed during sintering , In the described calibration process, according to patent information, high accuracies are achieved without the formation of cracks, the density increasing minimally compared to the sintered parts. If oxides are used as a component of the feedstock, then when sintering under hydrogen at approx. 1200 ° C (2150 ° F), parts that can be deformed with the hammer are obtained, the volume of which has shrunk considerably compared to the originally shaped starting geometry. Stressed and described only molded parts are first sintered and then calibrated, ie the end product is the calibrated sintered body whose geometry corresponds exactly to the geometry of the calibration mold.

Due to the high raw material costs of the MIM process, there has been no lack of attempts to use inexpensive metal powders, in particular water-atomized and mechanically ground metal powders as the raw material base / Lit6 /. However, since the MIM process requires very finely divided powders with good rheology (sprayability of the feedstock) and high sintering activity (high final density) due to the process steps involved, the inexpensive powders mentioned, on the other hand, are coarse-grained (> 40 μ) and also have an irregular structure the addition of coarse-grained powders always leads to a deterioration in the mechanical properties compared to comparable parts that were produced using fine-grained powders. These relationships have been sufficiently examined in the literature. / Lit3 /

It would be conceivable to synthesize very finely divided metal powders by reducing powders of the corresponding metal compounds (in particular their inexpensive oxides) in a preceding process step. The disadvantage here, however, is that, for thermodynamic reasons, an almost complete conversion of these oxides requires temperatures at which the metal powders produced in this way already have considerable sintering activity. This high sintering activity - which, on the other hand, is one of the reasons for using such extremely fine powders in the MIM process - means that the primary grains at the grain boundaries already frit together during the reduction to form irregularly shaped aggregates. Because of this morphology, the rheological properties of a feedstock manufactured via the preceding reduction of corresponding metal compounds are unsatisfactory, so that it can only be sprayed on with the addition of unacceptably high amounts of binder. However, this high proportion of binder has many disadvantages and leads, among other things. to segregation in the green compact, which in the sintered end part leads to sink marks, flow seams and sealing inhomogeneities. Although the frying of the primary particles can be minimized by lowering the temperature during the reduction process, an alternating mixture of metal powder and starting compound is obtained in this case instead of a defined metal powder, which leads to an undefined shrinkage behavior when used further in the context of the MIM process Parts during sintering.

In order to avoid this problem, it would be conceivable to add compounds which are mixed in a low concentration with the metal compounds before their reduction in order to maintain the temperature required for the conversion. temperatures (approx. 550-750 ° C) to suppress frying of the primary grains. If compounds which are thermally or chemically destroyed above the reduction temperature are used here, their effect is limited to the process step of powder production without interfering with the subsequent sintering process of the MIM process which takes place at a higher temperature. However, this upstream process step for the production of the powder requires additional investments, so that here, too, the raw material costs are significantly lower, but nevertheless not to be neglected. In addition, the handling of these powders on an industrial scale requires extensive safety precautions, since the powders so produced tend to self-ignite in air at room temperature due to their high specific surface area.

The present invention has therefore set itself the task of creating a possibility to significantly expand the technically / economically conditioned component upper limit for MIM parts by using cheaper starting products instead of the expensive, finely divided metal powders previously required and at the same time significantly reducing the shrinkage during sintering.

The solution to this problem is achieved by incorporating the corresponding unreduced metal compounds (for example, as inexpensive oxides) into the binder instead of the finely divided metal powders and only after the shaping of the green body while largely maintaining the initial geometry with a reducing gas at elevated temperature, definitely reduced to metal below the sintering temperature. The minimum temperature required depends on the redox potential of the cation under consideration and increases with the increasingly base character of the metal, e.g. from Cu (approx. 270 ° C) to Ni (approx. 650 ° C) to Fe (approx. 700 ° C).

The reduced moldings have a high, precisely controllable porosity and a correspondingly low density. They can be manufactured inexpensively using simple principles within narrow geometric tolerances. In principle, any reducible metal cations can be used in free or complex-bound form with any inorganic or organic anions, such as oxides; which form volatile or non-disruptive end products under the reduction conditions; Hydroxides, sulfides, nitrates, carbonates, formates, oxalates, acetates or metallates (eg paratungstate) and mixtures of such compounds are used. For economic and ecological reasons, oxides or mixtures of different oxides and ammonium metallates are preferably used, especially since their metal content based on the initial volume of the metal compound used is comparatively high.

Apart from the fact that the binder components should not undergo any uncontrolled chemical reaction with the metal compound particles at processing temperature, the composition of the binder is not subject to any technical restriction. It is therefore possible to use commercially available binder systems, and in particular the principle well known from MIM technology can be used to construct the binder from an easily removable (e.g. extractable) component in combination with a crackable polymer. Since the metal particles are incorporated into the binder in oxidized form, water-extractable binder systems can be used without corrosion problems.

The removal of the binder can be carried out in any manner known per se, with the dimensional stability of the shaped bodies formed by reduction, in contrast to the conventional MIM process, being removed by removing the crackable binder components under oxidizing conditions in air or a water vapor-containing atmosphere at temperatures between about 400 and 950 ° C has proven to be advantageous. With this measure, partial sintering of the highly porous matrix, which is under volume shrinkage, as well as its expansion, is avoided. continuous carburization avoided, so that the porous matrix formed by reduction provides dimensionally stable parts.

In order to set close geometrical tolerances of the porous shaped bodies, it has proven to be advantageous to abort the reduction in the vicinity of the equivalent point in order to prevent an uncontrolled subsequent sintering of the highly porous matrix formed which is caused by volume shrinkage. Furthermore, in order to achieve dimensionally stable parts, it has proven advantageous to clamp the metal matrix on the surface at the beginning of the reduction by incorporating foreign atoms in order to prevent uncontrolled sintering of the reduced metal matrix as the reduction progresses. This inhibition of sintering can be made possible in a simple manner by using a carbon-containing gas, the temperature to avoid soot formation on the freshly formed metal surface above the Bouduard decay but below the sintering temperature. In order to reduce an uncontrolled carburization of the metal surface, it has proven advantageous to add ammonia to the carbon-containing reducing gas.

The two-stage reduction can be made possible in a simple manner by adding a low-molecular-weight organic compound to the reactor in the first sub-step (reduction under a carbon-containing atmosphere), e.g. a lower alcohol is added with the addition of aqueous ammonia solution and is only reduced with hydrogen after a certain (component-dependent) partial conversion has been achieved.

When examining the properties of the highly porous metal matrix produced by reduction, it was surprisingly found that this shows a ductile flow behavior transverse to the pressing direction in the unsintered state when exposed to mechanical forces. This unusual behavior makes it possible to force compression with simultaneous shaping by applying pressure even without a split ram, within the matrix thus compacted An almost homogeneous density distribution occurs in the component even with complicated component geometry. Since the materials produced in this way have excellent mechanical characteristics after subsequent sintering, the process considerably expands the shaping possibilities of powder metallurgy without complex additional steps.

The porous shaped bodies formed by reduction can

• due to their low density, they can either be used directly as open-pore metal foams (catalysts, shock absorbers),

• are converted into steel parts with reduced porosity and completely new material properties due to the continuous pore structure by infiltration or CVD process while maintaining their x, y, z geometry.

Due to their ductile flow behavior while maintaining their x-y geometry, are pressed in a subsequent step before sintering in the Z direction and then sintered to final density,

• are sintered due to their high sintering activity while shrinking in the x, y, z direction in analogy to the conventional MIM process.

The present invention thus circumvents the disadvantages of the prior art and describes a process which reduces the raw material costs of the MIM process to a negligible proportion and only requires minor additional investments. This is achieved by using the unreduced metal compounds (eg as inexpensive oxides) as the basic component of the feedstock instead of expensive, finely divided metal powder these are reduced to metal only after the shaping of the green compact.

This method is not restricted to special binders, and is described below using an example of a commercially available binder composition (model binder). The possibility of using water-soluble binders is advantageous, since the oxidic particle matrix does not cause any corrosion problems here.

Example 1 :

A commercially available mixture (company TEKON: Marktheidenfeld) consisting of the binder components common in the MIM process: polymer (K2 = polyamide), removable component (K1 = long-chain ester) and auxiliary agents (K3 = fatty acid) is mixed with commercially available iron oxide (ground and floated Fe3O4 -Ore with a purity of 99.5% and an average particle diameter of 6-8 my) processed in a discontinuous double Z kneader at 175 ° C by kneading to feedstock. 5.92% by weight of carbonyl nickel powder (INCO 123) (based on the Fe3O4 + Ni) is also added to the binder. The proportion of binder required for processing is 9.3% by weight based on the total mass.

Green parts with a part weight of 10.49 g are injected from this feedstock on a conventional injection molding machine. After removing component K1 by extracting component K1 in acetone for 12 hours, the brown body is kept in the presence of hydrogen or hydrogen-containing gases for several hours at temperatures between 550 and 1250 ° C. and thus converted into a porous matrix.

Due to the difference in density between the original metal oxide (5.1 g / cm ³ ) and the iron formed by reduction (7.86 g / cm ³ ), additional volume is formed inside the part when the oxide matrix is converted. If the conversion were to be carried out while maintaining the external dimensions, a theoretically a pore fraction of approx. 65 vol.%, Which results from the 33 vol.% Pores of the binder removed, plus the approx. 32 vol.% Pore volume, would be expected come from the reduction of the oxide. However, since the transformation of the oxide matrix is already superimposed by partial sintering of the highly reactive metal particles at temperatures above approx. 650 ° C., this maximum proportion of pores is not reached. The extent of this shrinking process essentially depends on the reduction temperature, the duration of the reduction, the gas composition and the specific gas feed (I H2 / h / Kg Braunling). Typical values for the total shrinkage expressed as SF value lie between SF = 1.03 (transformation temperature Tmax below 600 ° C) and SF = 1.20 (Tmax = 800 ° C). In the following, the quotient between the currently considered length and the associated output length in the green compact is referred to as the SF value.

If the transition temperature is kept below 600 ° C, the pre-sintered parts are mechanically very susceptible, since due to the low surface diffusion at these temperatures, hardly any sintering processes take place. The three-dimensional network of metal particles that forms is therefore held together only by very weak forces.

The temperature profile during the reduction of the metal connections has to be adapted to the part geometry, whereby high wall thicknesses rather require a slow rise in temperature in order to achieve a conversion that is as uniform as possible within the matrix. If the temperature is increased too quickly, the reaction rate in the outer areas is very high, while the comparatively slow diffusion of the hydrogen into the interior of the part and the diffusion of the water vapor formed in the opposite direction mean that the almost completely reduced areas near the wall are still largely faces the original output matrix inside the part. Especially at higher temperatures (> 900 ° C), at which the three-dimensional particle network begins to shrink due to the onset of sintering processes, the different density between the starting and the end product leads to stresses in the part, which either show as cracks or as faults in the pre-sintering. For parts with the geometries and wall thicknesses common in the MIM process, a temperature profile has proven itself, in which the temperature is increased from 550 ° C to 800 ° C within 3 to 8 hours. Since the reduction of the metal oxide is an equilibrium reaction, it makes sense to drive with an excess of hydrogen during the conversion and to discharge the water formed in the reaction in a circuit. To achieve a complete conversion of the oxide used, the highest possible final temperature should be selected.

The pre-sintered product obtained in the manner described above can now be sintered into the end product analogously to the classic MIM process either in a separate process step or directly by further increasing the temperature. In particular for thick-walled parts, preference is given to final sintering under hydrogen, since complete conversion of the oxide is achieved here at the high temperature.

In the present case, the brown compact, which was reduced at 850 ° C., was sintered in a vacuum at a temperature of 1280 ° C. for 30 minutes. The final density achieved at 7.55 g / cm ³ corresponded to that customary in the MIM process.

In contrast to the conventional MIM process, in which the comparatively low shrinkage of approx. SF = 1.13 to 1.20 is a known problem, particularly in the case of larger parts, the conversion shrinkage also occurs in the process according to the invention. Overall, the parts from Example 1 therefore theoretically have a very high total shrinkage of approximately SF = 1.5. This high total shrinkage is uncontrollable, especially in the case of parts with different wall thicknesses, especially if one takes into account that the attack of the hydrogen takes place from the outside and thus internal tension in the part is pre-programmed. The problems associated with the high shrinkage of the warpage and the high dispersion of the target dimensions can be greatly minimized if the sequence of the process steps is reversed, as explained in Example 2.

Example 2:

A sintered body (hereinafter referred to as invert sintered body) is first produced from the chemically precisely defined Braunling of Example 1 without the addition of reducing gases. For this purpose, the Fe3O4 brown body is heated to 800 to 1360 ° C under nitrogen or vacuum (30 min holding time at maximum temperature). Following the thermal decomposition of the binder constituents, a gas evolution is observed in the temperature range of approx. 350-500 ° C, above approx. 750 ° C by reaction of the cracked residual polymer from the Braunling with the Fe3O4. This reaction leads to a weight loss which is due to the fact that the cracked residual polymer reduces part of the Fe304 to FeO / Fe. The conversion due to this reaction depends on the invert sintering temperature and the gas atmosphere set and is e.g. in vacuum depending on the maximum temperature approx. 4% (850 ° C) and 28% (1360 ° C). If you use an inert gas (e.g. N2), the sales are slightly lower.

Accordingly, the invert sintering essentially consists of the sintered starting product (here Fe3O4 with Ni), which, depending on the maximum temperature of the invert sintering, has a residual porosity of approx. 8% by volume (1360 ° C) to approx. 32% by volume (850 ° C) ) has. The invert sinterling is mechanically extremely stable, especially at higher sintering temperatures (from approx. 900 ° C) and shows no deformation or cracks despite the relatively high wall thickness.

Depending on the sintering temperature, its SF value is between 1.01 (800 ° C) and 1.15 (1360 ° C) (see Fig. 3). The statistical distribution of dimensions for different parts of the same series is strikingly narrow with a maximum of +/- 0.4% of the mean. The micro density of the open-pore structure increases due to the parallel partial reduction of Fe3O4 with increasing invert sintering temperature from 5.2 g / cm ³ (invert sintering temperature 700 ° C) to 5.5 g / cm ³ (inverting sintering temperature 1360 ° C), the macro density increases in the same Direction from 3.6 to 5.1 g / cm ³ .

In a subsequent step, the invert sinterling is reduced to iron in analogy to example 1. The conversion at approx. 900 ° C in the H2 / N2 mixture has proven to be favorable. The required reaction time depends on the wall thickness of the parts and is usually around 3 to 7 hours.

In contrast to the procedure described in Example 1, the shrinkage which occurs when the invert sinterling is reduced to the outside is comparatively low at temperatures below 1000 ° C. The SF value between invert sinterling and Braunling is only approx. 1.005 to approx. 1.030 depending on the maximum temperature passed. This is due to the fact that during the previous sintering of the unconverted Fe304 brown body (invert sintering) a mechanically stable skeletal structure with an internal residual porosity of approximately 8-32 vol.% Depending on the temperature used is formed. In contrast to Example 1, the shrinkage resulting from the conversion of the oxide therefore does not manifest itself to the outside, but, while maintaining the outer shape, leads to an increase in the internal porosity of approximately 32% by volume and thus (depending on the previous invert sintering temperature) reduction is 43 to 65% by volume. In contrast to the directly reduced brown compacts from Example 1, the converted invert sinterings are largely free of cracks and warpage due to the skeleton structure mentioned, even at a comparatively low transition temperature.

The macro density of the reduced invert sinterlings was around 2.6 to 4.2 g / cm ³ , depending on the conversion conditions. The micro density was independent of the Invert sintering temperature of approx. 7.5 to 7.7 g / cm ^{3 is} approximately the theoretically maximum possible value.

The tensile strength of the reduced (converted) invert sintered blocks corresponds approximately to that of plastics, but the fracture behavior is without elastic components. The tensile strength increases with increasing invert sintering temperature and reaches a typical value of approx. 70 N / mm2 at 1345 ° C (invert sintering temperature) after reduction in the H ₂ stream (900 ° C; 3 hours). In cases where plastic has the necessary mechanical rigidity, but due to the low heat resistance and low thermal conductivity, it creates constructional problems, these parts can already take on design tasks as an independent family of parts despite their high porosity. The tensile strength can be increased slightly if the porous molded body is infiltrated with polymerizable monomers, for example a mixture of isocyanates and polyols, with the formation of polyurethanes

If the converted invert sintering is sintered in a subsequent step at a higher temperature (e.g. in a vacuum at 1320 ° C for 1 h), the strength of the parts increases to approx. 300 N / mm2 with a macro density of approx. 5.3 g / cm , The residual porosity of these parts is around 25%.

Example 3:

In contrast to Example 2, the temporal and spatial separation of invert sintering and reduction is eliminated in this concept, which means that, due to the lack of intermediate cooling, comparatively low invert sintering and transition temperatures remain manageable without cracks.

For this purpose, in analogy to example 2, a batch of the brown pieces from example 1 consisting of 150 parts is run into the hot belt furnace in a stream of N2. From the technical dimensions of the furnace, the set temperature of the 5th Heating zones (300/600/900/900/900 ° C) and the belt speed are calculated as a heating rate of approx. 20 ° C / min. After reaching the 4th heating zone (900 ° C), the belt advance was put out, kept under N2 for 30 minutes and then the oxide components of the brown body were reduced to iron within 2 hours at this temperature by adding 1.5 Nm3 H2 / h. A mixture of hydrogen and nitrogen with a circulatory procedure with simultaneous removal of the water formed has proven to be advantageous.

The parts obtained, hereinafter referred to as DI sintered parts (direct inversion), show approximately the same geometric dimensions as the brown parts used at a temperature of 900 ° C, whereby the SF values can be controlled within certain limits without cracks by appropriate gas and process control , The optimal process control depends on the geometry of the parts (especially their specific surface), the specific loading of the furnace and the water vapor concentration set in the furnace, which in turn results from various other process parameters such as e.g. Gas throughput and furnace volume results. With a correspondingly high loading density, surprisingly even DI sintered parts are obtained which are larger than the brown parts used (values up to SF = 0.89 were found).

The pore volume resulting from the volumes of the former binder components and the conversion shrinkage (reduction of Fe3O4 to Fe) is in the range from 60 to 70% by volume, i.e. the removal of the binder and the conversion take place with appropriate process control while largely maintaining the outer geometry while simultaneously building up a high, homogeneously distributed inner porosity.

As expected, the Dl sintered products produced in this way have a low tensile force of approx. 10 to 20 N / mm2, but are promising candidates in such applications in view of the low macro density of approx. 2.6 g / cm ³ where metal foams (e.g. hot gas filters; crash absorbers) are discussed. These metal foams have so far not been made of steel but, due to the process, only accessible from alloys that have comparatively low melting points. (e.g. decomposition of TiHx in Al and Zn melts)

Example 4:

The DI sinterings produced according to Example 3 were sintered at high temperatures (for example 1320 ° C. for 1 hour in a vacuum). As expected, the parts shrank and the macro density increased to approx. 7 g / cm ³ . At the same time, the tensile strength that can be achieved increases to approx. 400 N / mm2.

Surprisingly, the tolerances of the final sintered parts can be set within relatively narrow limits despite the shrinkage factors above 1.3. Thus, the statistics of the dimensions of +/- 0.7% are not worse than those of the usual MIM method, despite the significantly higher shrinkage.

Example 5:

A tablet made from the feedstock of Example I with a diameter of 27 mm and a height of 25 mm was debinded and the brown product thus obtained was converted into a highly porous DI sintered product under N2 / H2 as described in Example 3. (Response time 5 hours; T = 900 ° C). The DI sintered compact (density 2.74 g / cm ³ ) thus obtained had practically not shrunk compared to the green compact and had a diameter of 26.85 mm and a height of 25.0. It was placed in a press die (diameter 27 mm and pressed with upper and lower punches at a given press pressure. The molded article obtained, hereinafter referred to as PDI (Pressed after Direct Inversion), showed an increasing density with increasing press pressure. This PDI was then sintered in vacuo (10 ° C / min; 1320 ° C over 1 hour; vacuum).

The evaluation of the sinterings obtained in this way shows that the sintered density increases with the density of the PDI and thus with the pressing pressure. Taking into account the pressures of max. 6 t / cm ² , there is a compression density of the PDI of approximately 6.4 g / cm ^3, which leads to a final density of 7.5 g / cm ³ during sintering. When using high pressures (15 t / cm ² ) a density in the PDI of 7.14 was obtained, which led to a sintered density of 7.62. The fact that the experimentally observed density, even when using high compression pressures of 7.62 g / cm ^{3, is} below the theoretical density of the FeNiδ alloy formed (approx. 7.9 g / cm ³ ) is due to the slight contamination of the mined starting material (Fe3O4 content> 99.5%) is justified. This gait contained in the ore is visible in the cut of the sintered stone as impurities and is identified in the microsensor as phosphates and silicates. Since the diameter of these homogeneously distributed inclusions is very small (usually approx. 1 my, in exceptional cases up to approx. 10 my), it does not affect the material properties. Tensile strengths of approx. 650 to 720 N / mm ² with HB values of> 200 were determined on material samples. This is remarkable because, according to its history, the material contains practically no carbon. The metallographic examination of the parts shows that the metallic matrix of the material is extremely fine-grained, absolutely homogeneous and pore-free.

If the sintering is then hardened and carburized, the hardness increases to 52 HRC with a simultaneous increase in tensile strength to> 1000 N / mm ² .

As expected, an inclusion-free metallic matrix was obtained with a synthetic precipitated iron oxide (commercially available Bayferrox ®). In view of the already very high mechanical strength, which is based on the mined oxide obtained, the use of high-purity starting materials can generally be dispensed with in the process according to the invention for cost reasons.

The toughness and notched impact strength of the materials produced according to Example 5 are high. Even if the pressure applied to the PDI is only 2.6 t / cm ² and the sintered density is only around 6.95 g / cm ³ , the tensile strength is over 500 N / mm2. In contrast to conventional PM materials with a comparable density, these parts are surprising because of the significantly higher tensile strength and the significantly lower sensitivity to notch impact, which is due to the extremely fine-grained structure. The method according to the invention can thus be used to achieve material properties at comparatively low pressures, which are clearly superior to the conventional PM parts produced using comparable pressures. For a given press output, significantly larger parts can be produced using the method according to the invention than is possible with the classic PM.

To protect the press die, it has proven to be useful to at least partially soak the highly porous DI with a commercially available oil before pressing. This low-viscosity oil emerges from the die during subsequent pressing and leads to a more homogeneous density distribution in the pressed body. Since, in contrast to the classic press sintering technique, no powder, but an impregnable porous molded body is involved in the pressing process of the method according to the invention, the service life of the pressing tools can be increased considerably by this simple measure without the formation of lubricating pastes during the subsequent filling process of the mold.

With increasing density of the PDI (ie with increasing pressing pressures) the porosity of the DI sintered body is increasingly eliminated by compression in the Z direction. The porosity remaining in the molded body after pressing disappears during the ters to the final density and leads to a sintering shrinkage that occurs evenly in the first approximation in the x, y and z directions. Taking into account technically manageable pressing forces (approx. 6 t / cm ² ), pre-compaction to approx. 6.4 g / cm ^{3 is} achieved so that the remaining sintering shrinkage in the x and y direction is approx. 5.5% (SF = 1 , 055) is located. This value is significantly below the SF value which must be taken into account for the same material class in the MIM process (approx. SF (MIM) = 1, 175). Within the given manufacturing tolerances, significantly larger component geometries are therefore technically manageable using the process according to the invention. The accuracy of the process depends on the geometry of the component and its pre-compression before sintering. With compression to approx. 6.4 g / cm ³ , tolerances of less than 0.3% based on the target dimensions can be reliably controlled. Additional calibration steps can therefore generally be dispensed with.

Example 6:

In classic press sintering, the compression of a "single-height" molded body, ie a geometry that has only one height in the Z direction (for example the tablet of Example 5), is simple. On the other hand, if a part is to be manufactured with different heights, each height theoretically requires own stamping in order to enable an individual compression in Z-range without density gradients in the X, Y range.This problem in the classic PM with multi-height parts leads to the necessity of a large number of individually controllable stamps. The underlying presses and tools are accordingly very complex and It becomes particularly critical if, instead of two or three-height molded parts, geometries with a continuous change in height (eg an oblique edge) or undercuts in the pressing direction are to be produced. manoeuvrable tools or mechanical reworking of corresponding blanks.

In theory, this problem would also be expected when compressing the DI from Example 5. Surprisingly, however, it was found that its highly porous matrix has ductile properties when pressed and has the ability to compensate for density differences within certain limits in the x and y direction to move material transversely to the pressing direction. Taking advantage of this ductile behavior, a helical bevel gear with module 0.76 and a diameter of D = 53 mm was produced with a simple 2-part mold from the upper and lower punches at a pressure of 6 to / cm ² . Despite different heights (2 mm on the outer edge, 6 mm in the center), a density of 7.48 g / cm ^{3 was} obtained over the entire cross-section of the sintered gear (1320 ° C; 1h; vacuum). The surface hardness was uniformly 209 to 212 HB 187 / 2.5. The scatter of the diameter in the sintered gear was very small at +/- 0.06 mm.

Example 7

This flow behavior in combination with the fact that, in contrast to the classic press sintering technique in the present process, not a pile of powder, but a homogeneous, geometrically narrowly defined shaped body is inserted into the press mold, allows the mold parting plane of the pressing tool to be within certain limits with respect to the outer edge of the component reset. If you consider a circumferentially toothed gear, in classic sintering technology the mold separation inevitably runs on the outer edge of the component, with the result that this gear as a PM part often has an impermissible sharp burr formation on the running flank, which also impermissibly high point forces into the component initiates. In the present method, this can be done in can easily be avoided by the fact that the separation of the press ram does not run on the tooth edge, but is drawn into the component by a few 1/10 mm. The tooth edge itself can now be rounded. Despite the resulting undercut in the pressing direction, the partial contour lying in the undercut also condenses as a result of the ductile cross flow and, when correctly designed, there are no notable differences in density and strength in the sintered component. The mold is removed from the mold via a second parting plane in the press die, which is closed during the pressing process.

Example 8

The ductile flow behavior mentioned in Example 7 makes it possible, within certain limits, to fill in the press die such material contours whose material is not pre-formed in the porous molded body inserted into the mold. That the porous molded body need not necessarily represent the shape of the pressed body expanded in the pressing direction.

Compared to the classic press sintering technology, this results in significantly expanded shaping options. Using this flow behavior, form-fitting connections between two workpieces can be produced in the simplest way.

For this purpose, a PDI (insert A, compression density 6.4 g / cm ³ ) produced in analogy to example 5 is produced and inserted as an insert into a press die. A porous DI manufactured according to Example 3 is then placed on this PDI and, using the ductile flow behavior, pressed in a form-fitting manner with insert A transversely to the pressing direction. The two parts connected by pressing technology are now sintered together. Since both part geometries (assuming the same pre-compression) shrink identically during sintering, the part combination is rigid, with the original separation point healing due to the high pre-compression and the high sintering activity of the finely divided powders during sintering. This significantly extends the range of technical possibilities, since in classic press sintering technology the high press forces that are inevitably required for larger components limit the accessible part size to approx. 100 cm ² (projected press area). A sequential pressing is not possible in the classic PM, because pressed surfaces do not form a sinter bond worth mentioning and a homogeneous filling of the pressing tool with powder particles around inserted contours is generally not possible

It is advantageous to be able to build up parts with a high space filling or parts with openings in mutually perpendicular planes so that the individual geometries to be pressed are each optimally constructed in terms of press technology. The concept also opens up uncomplicated access to composite parts made of different materials, provided that a regime overlapping in the sintering parameters can be found for both materials.

Example 9

As shown in Example 8, PDI and porous molded body DI are not necessarily geometrically similar bodies that are only compressed in the z-axis. Since the ductile flow properties of the DI allow material to be moved transversely to the pressing direction, it is possible to produce geometrically complex shapes with flowing transitions (multi-height parts) in simply constructed pressing dies in that a porous molded body identical in basic geometry is shaped like a stem in the rear shaft is extended by the additional volume required for pressing. The additional volume must be designed so that it stores the material of the target geometry required for compression. Assuming undisturbed compression (ie here 6.4 g / cm ³ at 6 to / cm ² ) this would be the case in the DI described in Example 3 (2.54 g / cm ³ ) theoretically about 2.52 times the target volume to be compressed.

This procedure leads to relatively simple press molds if, in addition to the geometrically easy-to-compress (e.g. single-height) sub-geometries, there are also filigree multi-height contours in a component under consideration.

The application of this principle enables a high degree of flexibility for the production of geometrically similar parts, since in a simple way a still geometrically undifferentiated mass part (e.g. a key blank manufactured according to example 3) can be converted into a part-specific final contour in a subsequent pressing step using a coded adjustable pressing tool can.

Although in the case of more complex geometric structures the density achieved by transverse compression in the PDI does not naturally achieve the homogeneity that can be achieved by compressing a geometrically simple tablet, due to the excellent material properties, even in those areas that are not fully compressed during pressing after sintering good tensile strengths and notched bar impact values were observed. In a critical area to be compacted, the comparatively low pre-compression to approx. 5 g / cm ³ is sufficient to achieve a density of 6.9 g in this area after the sintering of the FeNi8 material used in this example. After sintering, this leads to a tensile strength of approximately 500 N / mm ² .

Overall, the process offers a clear competitive advantage with comparable material properties compared to conventional metal powder injection molding. The low raw material costs and the significantly lower sintering shrinkage allow the production of highly cantilevered geometries that cannot be mastered when sintering classic MIM parts. So it is possible via a Relining with support geometries to stabilize these projecting partial structures during sintering and to subsequently mechanically remove these support structures. This concept is generally uneconomical in the classic MIM process due to the prohibitively high raw material costs.

As stated in Examples 5 ff, the porous moldings must be placed in a mold for pressing. To minimize the cost of this additional step, cycle times of a few seconds are required. The pressing process itself is generally in the range of less than 1 sec and does not require a holding time. The speed-determining step therefore lies in feeding the parts to the press mold, which should be automated to minimize costs. Due to the high stability of the porous molded bodies, this is possible without any problems, provided that they are manufactured within relatively narrow tolerances and can therefore be inserted into the specified press cavity without tolerance problems.

With simple component geometries, this is possible without any problems, since when an undersized molded body is inserted, a gap remains between the compression mold and the molded body, but this is filled during the pressing by transverse displacement of the material. A certain amount of oversize is also permitted if the press mold has an insertion bevel, thus avoiding blunt shearing of material at the edge of the mold.

It becomes more critical, however, if the internal dimensions have to be adhered to when inserting the shaped bodies, since the highly porous material can withstand compressive forces only to a very small extent with approximately 10 to 20 N / mm ² , without problems. If, for example, a molded body in the shape of an "8" is to be pressed, it must be placed on two standing pins when it is inserted into the mold. If the deviation from the nominal size is such that tensile stress is exerted on the part when it is being threaded (considered The case would be the porous "8" too turned out small), the part tears, whereby the break gates formed separate from each other. If there is enough material in the vicinity of these break gates, these break gaps heal when pressed, since material flows from the other parts of the part due to the ductile flow behavior. Due to the high sintering activity, the material is homogeneous after sintering and also has the same high tensile strength as the rest of the matrix at the healed fracture gate.

However, if the material cross-section at the "8" node is small, the ductile flow behavior is not sufficient to transport the material missing between the fracture ports through the adjacent small flow cross-sections perpendicular to the actual pressure gradient. In this case, component damage remains even after sintering.

For the reasons mentioned, it is necessary for most components in x / y to adhere to relatively narrow tolerances of the porous molded body. A deviation of +/- 1.5% can be used as an order of magnitude. Since the reduction of the oxide component matrix requires a diffusion of the reducing gases (e.g. H2) into the component depth, when using pure hydrogen this leads to the fact that filigree part geometries have already been completely converted, while part geometries with high wall thicknesses still contain significant oxide components in the middle , This diffusion front between metal oxide and reduced metallic matrix can often be clearly seen with the naked eye. As stated in Examples 1 and 2, the completely reduced material already has a certain sintering activity even below the usual sintering temperatures. This means that filigree part geometries tend to shrink considerably, while the material in more compact part geometries does not yet show complete conversion. Extensive tests have shown that it is difficult for critical components, ie those with changing wall thicknesses and internal dimensions to be maintained, using pure hydrogen, Finding process parameters that guarantee satisfactory process statistics of the dimensions, ie, uniform, reproducible SF values over the entire component. The problem is exacerbated if, in the case of technically relevant batch sizes, inhomogeneities in the gas routing of technical apparatus come to the problems mentioned, so that there is also. result in different SF values over the total batch considered.

By adding coarse atomized powder to the feedstock mixture, the sintering activity can often be reduced here so that the thin-walled geometries mentioned do not shrink inadmissible, but this is generally accompanied by a deterioration in the material properties and an increase in the pressing pressure.

Example 10

29 parts of a brown body made from the feedstock of Example I with a green body weight of 10.5 g and a diameter DX = 25.42; DY = 25.42 and height H = 12.96 were heated in a gas-tight furnace with gas circulation and exhaust gas torch lying on a perforated plate at 20 ° C / min to 900 ° C and then flowed through from below at this temperature with 0.6 Nm3 H2 / h reduced.

After a reaction time of 2 hours, the weight of the parts had been reduced to 7.1 g by reducing the oxide. The parts had a light gray metallic appearance and were in the area of the thin-walled outer arches (wall thickness 1.1 ^* 0.9 mm) by a factor of 1.05 to 1.09; in the thick-walled center, on the other hand, only shrunk by SF = 0.98 to 1.015. When these parts were manually inserted into the mold, the thin outer arches broke. Hair cracks were clearly visible in the components pressed in this way, which led to component failure in the functional test of the subsequently sintered parts. In an identical arrangement, the 29 parts were flowed through in a second experiment instead of H2 / N2 with 0.6 Nm3 / h of a gas mixture of CO / H2 / CH4 (30/65/5 vol.%). After a reaction time of 2 hours, the weight of the parts had been reduced to 7.2 to 7.4 g by reducing the oxide. The parts had a dark gray metallic appearance and had shrunk uniformly in X and Y by a factor of 0.985 to 1.015%. Carbon deposits were visible on the surface of some parts, particularly in the area of edges and thin-walled geometries, which can be attributed to a Bouduard decay of the CO on the freshly formed iron surface. At the same time, these thin-walled geometries swelled from 25.42 mm to 26.4 mm.

Example 11:

The experiment from Example 10 was repeated to suppress Bouduard decay with the addition of 5% by volume of NH3, water being added to the system at the same time to increase the O: C content. The parts showed no carbon deposits at an SF value of 0.975 to 1.02

Example 12:

150 parts of the brown product described in Example 7 were heated to 900 ° C. in a gas-tight convection oven while feeding 20 l of N 2 / min. 500 g / h of an oniacal ethanol solution were then metered into this oven for 2 hours (870 g of 96 ethanol with 130 g of 25% strength aqueous NH 3 solution). The escaping gases were flared. After 2 hours, the mixture was cooled in a stream of N2. The parts were metallic gray and had a uniform weight of 7.15 to 7.35 g. The parts had no visible surface Carbon deposits. The shrinkage of the parts was uniform over the entire part at SF = 0.97 to 1.02 with a reject rate of 2.7%. The dimensions of the parts thus corresponded to those of the within a spread of + / 0.4 -0.2 mm originally used Braunlings. The parts could be fed to the press mold automatically. After sintering these parts at 1280 ° C in vacuo, an inadmissibly high C content was found in some parts, which led to partial melting.

Example 13:

300 parts of the brown body described in Example 7 were heated to 900 ° C. in a gas-tight convection oven while feeding 20 l of N 2 / min. 1.1 kg / h of an ammoniacal ethanol solution were then metered into this oven for 1 hour (870 g of 96% ethanol with 130 g of 25% strength aqueous NH 3 solution). The escaping gases were flared. After 1 hour, the mixture was reduced under hydrogen (2 m ³ / h for a further 2 hours. The mixture was then cooled in a stream of N 2. The parts were metallic gray and had a uniform weight of 7.12 g. They had a C content of approx. 0.75% and had a uniform shrinkage of 0.99 to 1.01 over the entire part.

The porous parts were impregnated with a commercially available mineral oil and then pressed in a mold using a pressure of 28 tons. The compacts obtained had a macro density of 6.3 to 6.4 g / cm ³ with a micro density of 7.55 g / cm ³ . These parts were sintered at 1280 ° C under hydrogen (7.5 ° C / min; 1 hour holding time at maximum temperature). After sintering, the parts had a uniform weight of 6.98 g, a macro density of 7.5 g / cm ³ and were ductile deformable under the action of impact. After hardening and tempering of the parts at 940 ° C (HRC = 52), they had a tensile strength of 2.2 kN under defined test conditions Computationally corresponded to a tensile strength of approx. 1100 N / mm2. The scatter of the dimensions in diameter and height was within 24.2 +/- 0.08 within a narrow tolerance.

Example 14:

The parts were produced analogously to Example 13, but after the conversion, the porous moldings were infiltrated with a concentrated ammoniacal Cu [(NH ₃ )] ₄ ^{2 "} solution and under hydrogen at 900 ° C. in a belt furnace (total residence time 1.5 ^' h) reduced to Cu °. The parts showed a copper color on a metallic gray matrix, which continued homogeneously into the part. These parts were pressed, sintered, hardened and tempered as described in Example 13. In comparison, the tensile strength was about 10 % higher than that of the comparison parts from example 12 without Cu infiltration.

Claims

claims

Process for the production of a plastically deformable metal body of defined geometry by mixing metal compound particles with a binder and pressing into shaped parts, after which the binder is removed and the metal compound is reduced to metal by gassing with a reductive gas at higher temperatures, characterized in that the reduction is carried out at Performs temperatures below the sintering temperature of the reduced metal compound.

A method according to claim 1, characterized in that the metal compounds are metal oxides or mixtures thereof, in particular iron oxides (e.g. magnetite or Ruthner oxides) and / or nickel oxide and / or molybdenum oxide.

Process according to Claim 1 or 2, characterized in that the reduction is carried out at temperatures of between approximately 550 and 1050 ° C.

Method according to one of claims 1 to 3, characterized in that a binder mixture of a removable, for example soluble and a stable, for example insoluble component is used, the removable component is removed, for example with the aid of a solvent, and then the shaped body in an oxidizing atmosphere, for example in air and / or water vapor at a temperature of between about 550 and 950 ° C., largely converting the stable binder fraction into gaseous degradation products, for example CO / CO ₂ and H ₂ / H ₂ O, and removing them from the matrix of the shaped body. Process according to one of Claims 1-4, characterized in that the shaped body is pre-reduced with a low molecular weight organic compound, for example a lower alcohol, preferably in the presence of ammonia at temperatures above the Bouduard decay.

Process according to one of claims 1-5, characterized in that the pre-reduced shaped body is reduced with hydrogen at temperatures above about 550 ° C.

Method according to one of claims 1-6, characterized in that during the reduction with hydrogen this is terminated at the end of the water formation which occurs in this case.

Method according to one of claims 1-7, characterized in that the reduced shaped bodies are pressed to end products.

Method according to one of claims 1-8, characterized in that the pressed molded body is heated to sintering temperature.

Process according to Claims 1 and 2, characterized in that the reduction of the metal compound takes place only after sintering of the as yet unconverted metal compound in the brown body.

Process according to Claims 1 to 3, characterized in that the porous shaped body produced by reduction is sintered after the metal compounds have been reduced. A method according to claim 4, characterized in that the porous molded body produced by reduction of the metal component is compressed by mechanical force before sintering.

Process according to Claims 1 to 5, characterized in that the mixture of metal oxides used as a secondary component of metal or alloy powder e.g. Contains Cr, CrNi steels or ferromanganese.

Process according to Claims 1 to 3, characterized in that the metal compounds are tungsten compounds.

Process according to Claims 1 to 6, characterized in that a gas, preferably a gas containing H ₂ and / or CO or NH ₃ or a mixture of these gases, is used as the reducing agent.

Process according to Claim 9, characterized in that the gas used as the reducing agent is formed by thermal decomposition of a starting compound, in particular by decomposition of lower alcohols and / or hydrocarbons, if appropriate with the addition of water and / or NH ₃ .

A method according to claims 9 and 10, characterized in that the reduction of the metal compounds is carried out first by a carbon-containing gas and in a subsequent step by a hydrogen-containing gas. A method according to claim 5, characterized in that the body to be pressed does not necessarily correspond to the target geometry of the compact to be sintered corrected for the volume contraction of the pressing process, in that the volume compression takes place by ductile flow of the porous matrix transverse to the pressing direction.

Method according to Claim 11, characterized in that, in combination with a second molded part, positive component assemblies are produced from the same or different materials during the pressing.

Process according to claims 5 and 11, characterized in that a lubricant (e.g. mineral oil or stearates) is added during the pressing process or the porous molded body is at least partially soaked with such a lubricant before pressing.

Process according to Claims 1 to 3, characterized in that graphite and / or metal carbides are added to the feedstock.

Process according to Claims 1 to 3, characterized in that the porous intermediate stage produced by reduction of the metal component or its associated brown body is infiltrated with the cation of a reducible metal compound, for example a Cu [(NH ₃ )] 4 ^{2 "} solution.

Process according to Claims 1 to 3, characterized in that the porous intermediate stage produced by reduction of the metal component or its associated brown body is enriched with a metal by means of CVD principles (e.g. by thermal decomposition of metal carbonyls).

Shaped body obtainable according to one of claims 1-9.