The object of the invention is to prepare components using a powder-metallurgic manufacturing technique and startin from metal powder and refractory ceramic powder with high re¬ quirements as to the binding between the components involved.

The powder-metallurgic manufacturing technique is cha¬ racterized by production in long series of parts having high dimensional accuracy. The chain of manufactures starts by mixi a metallic powder, for example iron powder, optionally having added pulverulent alloying agents thereto with a lubricant to facilitate the subsequent compacting operation. In this opera¬ tion the powder mixture is compacted to a moulding blank the shape of which closely or completely corresponds to the shape of the final part. The blank is heated and is then maintained at a temperature at which it by sintering obtains its final properties, such as strength, toughness etc.

Serial" manufacture of dry-compacted refractory ceramic shows several similarities with the powder-metallurgic manufac ring technique. In iron powder based manufacture sintering nor mally takes place within the temperature range 1000-1300 C. Fo more low-melting metals and alloys, such as for example alumi¬ nium and bronze, the sintering temperature usually lies betwee 500-900 C. When manufacturing powder-compacted refractory cera mics the sintering, however, takes place within the temperatur range 1400-1700°C.

Refractory materials are in combination with metals, for example as heat shielding, of quite a great interest. The difficulty of adjoining the refractory material with the metal in a durable manner lies in the fact that the materials invol¬ ved normally have different coefficients of expansion. Joining may then be accomplished using for example countersunk bolts which, however, when used under mechanical load i.a. can result in unsuitably high weight to obtain the necessary strength. It is also possible to perform joining using a glue, but then it is a requirement that heating cannot take place beyond the heat resistance of the glue. If inorganic glue, for example water- glass, is used in such cases when the ceramics is expected to

be subjected to a relatively high temperature there is an ob¬ vious risk that different coefficients of expansion will cau¬ se failure of the ceramics. Moreover, the multitude of opera¬ tional steps in this manufacture results in high costs. Other known methods are for example soldering on ceramics. This is a laborious method requiring high accuracy and at least seven different operational steps. Moreover, it is required that the grain size of the soldering powder shall be less than 1 micron. According to the invention it has, however, been found that it is possible by powder-pressing and subsequent sintering to obtain in a simple manner suprisingly good binding between ceramics and metal. The simplicity consists i.a. in the fact that it has been found possible in one step to press together the ceramic powder and the metal powder (for example iron pow- der) . The pressing can take place at such low pressures as abou

2 0.5 ton/cm , and high strength can be obtained also in the cera mic part at such low temperature as 800 C.

There are two partly overlapping problems which the in¬ ventors have been obliged to solve. The dilatation curves for metals and ceramics usually differ strongly from each other.

It has been found that the damaging effect of this can be eli¬ minated by using as a binding phase between a ceramic layer and ^* a metal layer one or several layers of a mixture of the compo¬ nents involved. The fraction of metal or ceramics to be present in the mixture is wholly dependant on the area of use and whe¬ ther the mixture shall be bound to ceramics and/or metal. It is obvious that sintered bodies also can be manufactured in accor¬ dance with the invention in only one layer and where the amount of the components involved can be controlled by the area of us At the temperature which can be endured by the metal part of th sintered body the refractory ceramics sinters only poorly. It h however, been found possible to modify the ceramic material inv ved in such a manner that the necessary strength will be obtain during sintering. This can be achieved in two ways, namely eith by adding alumina or aluminum hydrate and phosphoric acid or dry phosphoric, acid and a suitable quantity of water, or by adding monoaluminium phosphate solution or monoaluminium phosphate and water, such added material forming an effective binder between metal and ceramics.

In the following the invention will be exemplified and experiments will be presented which have been made using the technique according to the invention and the unexpected re¬ sults hereby obtained. Example 1

A mixture, designated A, was prepared in the following manner. 600 g of alumina powder having a maximum particle size of 400 microns and 350 g alumina powder having a maximum partic le size of 45 microns were thouroughly mixed together with 50 g of 80 phosphoric acid.

500 g of the mixture prepared according to the above was a mixed with 500 g of ironsponge powder having a maximum particle size of 147 microns. The mixture thus prepared is in the follo¬ wing designated B. A mixture C consisting of the following components was prepared: 990 g of ironsponge powder having a maximum particle size of 147 microns, 10 g of zinc stearate powder having a ma¬ ximum particle size of 147 microns.

Of each of mixtures A-C there were prepared test bodies having the dimensions 0 25 mm, h 25 mm, by pressing at a pres¬ sure of 200 MPa, heat treatment in evaporation oven at 300°C for a period of time of 15 min. in a neutral atmosphere, and the bodies were then sintered at a temperature of 1050 C for 30 min. in a non-oxidizing atmosphere. The dilatation process was determined for the test bo¬ dies thus prepared within the temperature range 20-600°C. The result which is presented in diagram 1 surprisingly shows that the dilatation process for the test body manufactured from mix¬ ture B corresponded considerably closer to the dilatation pro- cess registered for the test body manufactured from mixture A than to the process for the test body from mixture C in spite of the fact that mixture B is constituted by the same parts by weight of mixture A and C. Example 2 Four powder mixtures, D, E, F and G having a composition according to the table below were prepared:

Mixture D: 90 % alumina powder having a maximum particle size of 150 microns 5.0 monoaluminum phosphate solution

5.0 carbon-forming organic powder. Mixture E: 50.0% of mixture D.

50.0 of an ironsponge powdeτ having a maximum par¬ ticle size of 210 microns. Mixture F: 95.0% alumina powder having a maximum particle size of 150 microns.

5.0% carbon-forming organic powder. Mixture G: 50.0% of mixture F.

50.0% of an ironsponge powder having a maximum par- tide size of 210 microns.

Testbodies having the dimensions 0 25mm and h 25mm were pressed in two layers at a pressure of 90MPa. The composition of the testbodies is clear from the table below. Testbody I: Layer 1: consisting of powder from mixture D. Layer 2: consisting of powder from mixture E. Testbody II:

Layer 1: consisting of powder from mixture F. Layer 2: consisting of powder from mixture G. The testbodies were sintered at 900°C for 20 min. in a neutral protective gas atmosphere.

The testbodies thus prepared were investigated with re¬ gard to compressive strength. For testbody II there was measure a total load at break of 190 N, whereas testbody I was still co- herent at a load of 5500 N, the investigation being discontinue Moreover, the testbodies were examined under microscope making clear that the significantly better strength measured for testb dy I is a result of the fact that the addition of monoaluminum phosphate provides for sintering bonds between the ceramicr par tides but also, which is a wholly surprising effect, provides for sintering bonds between the ceramic particles and the meta particles. The results of the experiments described herein show that the binding between ceramics and metal can be considerably improved by using the process according to the invention. Example 3

Multi-layer testbodies having the composition according to the table below were prepared. Testbody III: Layer 1: consisting of 92.0% alumina powder and 8.0% dry

phosphoric acid + a suitable quantity of water for forming monoaluminum phosphate. Layer 2: consisting of 46.0% of alumina powder and 4.0% of dry phosphoric acid + water according to the above and 50,0% iron powder having a maximum particle si ze of 400 microns. Layer 3: consisting of 96.0% of iron powder having a maximu particle size of 400 microns and 3.5% carbon-formi organic powder having a particle size of less than 400 microns and 0.5% of zinc stearate powder.

Testbody IV:

Layer 1: consisting of 95,0% alumina powder and 5.0% mono- aluminum phosphate solution. Layer 2: consisting of 71.0% alumina powder and 5.0% mono- aluminum phosphate solution and 24.0% of iron pow¬ der. Layer 3: consisting of 70.0% of iron powder and 25.0% of a- lu ina powder and 5.0% of monoaluminum phosphate solution. Layer 4: consisting of 99.3% of iron powder and 0.7% of zin stearate powder. The testbodies were pressed at 500 MPa, heat treated i evaporation oven and sintered at 1120 C for 30 min. in protect gas atmosphere. The dilatation curves of the testbodies were me sured. The result which is presented in diagram 2 shows that t dil4tation process for multi-layered bodies prepared according the invention corresponds very closely to the dilatation process for pressed and sintered ceramics.

Moreover, the testbodies were examined under microscope and it could be observed that good binding between ceramics and metal was obtained and creation of cracks between ceramic parti les and metal particles could not be observed.

As a preliminary test for endurance to heat chock the testbodies were heated with a welding torch on the metal layer to glowing red and were then quenched in water. The test was re peated until all testbodies had disintegrated. None of the test bodies showed increased tendency for formation of cracks at the layer boundaries.

In the examples presented herein for the purpose of exemplifying the invention the inventors have used alumina and iron powder. However, experiments have shown that the ceramics may equally well be aluminum silicates, magnesium-aluminum sili- cates, zirconias , zirconium silicates, orto silicates, silicon carbides and slags. Nor is the invention limited to iron powder and alloys thereof but also other metals, such as aluminum, coppe nickel, chromium, molybdenium, manganese and alloys thereof can be used. It has also been found in continued experiments that it is possible for certain applications to use as a binder also o- ther acid phosphate salts of 2- or 3-valent metals.

The wording used in claim 1 : "or in several layers havin different mixing ratios between the ceramics and metal used" is intended to cover also a composite body wherein the said ratio varies continuously over a cross-section of the body, for example from one side of it to the other. In other words, this embodiment is constituted by infinitesimally thin layers. In an extreme case the body may thus on one side thereof consist of metal whereas the opposite side consists of ceramics.

With regard to the hinder used the presence of water is important. Thus at least one of the components of the binder may be added in the form of a solution. In the alternative dry mate¬ rial may be supplemented with external water. The amount of binder (based on solids thereof) is not particularly critical, but is suitably up to about 10% by weight, based on the weigh of the composite body.

The lower limit is determined by the quantity necessary for effective binding. An amount of about 1%. is usually necessar and at least about 2 or 3% by weight is usually employed.