1. Field of the Invention:

This invention relates ^" to a high strength, high toughness and low thermal conductivity ceramic composition, ό More specifically, this invention relates to a ceramic com¬ posite of finely divided Zr0 ₂ -Hfθ2 solid solution particles dispersed in a mixture of Al2θ3*-Cr2θ3 solid solution.

2. Brief Description of the Prior Art:

It is well known in the internal combustion engine 0 art that an increase in temperature in the combustion chamber and a minimization of the associated heat loss during com¬ bustion will theoretically resuli in increased efficiency of the engine. This in turn will lead to improved performance and economy. Thus, the so-called adiabatic (no heat loss) 5 ceramic engine, high temperature. ceramic-based .turbine and ceramic recuperators (heat exchangers) are well known and publicly acknowledged as contemporary research and develop¬ ment objectives (for example, see "The Coming Age of Ceramic Engines", March 1982, Popular Science, p 64). However, at these higher temperatures, the conventional ceramic materials employed in constructing such devices are inadequate in one or more critical properties.

Thus, in the case of ceramic lined diesel engines and similar applications, the ideal ceramic composition used as a lining material should possess and exhibit high strength, high toughness and very low thermal conductivity at ultrahigh combustion temperatures as well as high resistance to thermal shock, wear and corrosion. Although ceramics are generally known for their high temperature strength, heat resistance ^' and high temperature thermal insulation characteristics, they are also known as being extremely brittle. In general, to overstress a ceramic part leads to disintegration of the ceramic composition.

Although contemporary applications involving calcium and yttrium stablized zirconia are reported to result in im¬ proved strength, toughness and thermal conductivity, it has

-2-

also been reported that such partially stabilized zirconia (P _S Z) deteriorates rapidly at temperatures below engine operating temperatures. Furthermore, the use of titanium alloys as found in U. S. patent 3,152,523; the silicon nitride, lithium aluminium silicate, fused silica, silicon carbide, sintered silicon carbide, reaction sintered silicon carbide and reaction bonded silicon nitride ceramics as proposed in U. S. patent 4,242, ^' 948; and the cordierite, beta spodumene-mullite and fused silica-clay of ϋ. S. pat- ent 4,245,611 are felt to be deficient as ceramic compositio for the adiabatic engine in one or more of the above critica properties.

More specifically, it is generally known that zircon dioxide (zirconia) exists in three allotropic forms; mono- clinic, tetragonal and cubic and that there is a large volum expansion during the transition from monoclinic to tetragona Further, it has been historically accepted that because of this disruptive phase transition, the refractory properties of zirconia cannot be used. However, recent developments relating to suppressing or disrupting the deleterious effect of the phase transition have been discovered. For example, in the so-called partially stabilized zirconia (PSZ) the ι addition of metal oxide (e.g. lime stabilized zirconia) is viewed as creating a multiphase material having a fine-scale precipitate of monoclinic zirconia in a stabilized cubic matrix which in turn results in enhanced strength. More recently, an even more powerful strengthening mechanism viewed as involving a dispersion of a metastable tetragonal zirconia in cubic zirconia has been suggested. In this more recent development, the martensitic transformation (fast and diffusionless) between monoclinic and tetragonal phases is partially alleviated by inducing and creating tetragonal zirconia in sintered bodies or domains of a resulting time-stabilized zirconia. Although these transformation stabilizing ceramic mechanisms and their underlying rational may be .questionable and although their respective effects on high temperature properties of the resulting transformati

toughened ceramics are encouraging, the breadth of applic¬ ability of these general principles to the field of ceramics generally and the extent to which the properties can be im¬ proved is still not well defined or understood. Thus, the -5. use of transformation toughened ceramics and ceramic coatings in specific pragmatic applications (e.g. light diesel engines and/or ceramic engines) still remains uncertain.

This invention was made with Government support under DAAG-460-82-C-0080 awarded by the Department of Army. O

The Government has certain rights in this invention.

SUMMARY OF THE INVENTION In view of the deficiencies associated with known ceramic compositions particularly relative to the high tem¬ perature properties required in such applications as the 5 ceramic engine, I have discovered an improved ceramic com¬ position comprising:

(a) a matrix phase selected from the group consisting of solid solutions characterized ^" by the formula Al2θ3 ^« 2Siθ2 + [3Cr2θ3'2Siθ2- where x is the relative mole fraction of 10 Cr2θ ₃ or 3Cr ₂ θ3 ^* 2Si0 ₂ ; and

(b) ^" a disperse phase characterized by the formula ^" Zrθ2*yHfθ2 where y is the relative mole fraction of Hfθ2«

The improved ceramic compositions according to the present invention are viewed as transformation toughened 15 ceramics wherein a fine dispersed Zrθ2~Hfθ2 solid solution is present in either a chromium alumina or chromium mullite solid solution matrix wherein the HfU2 and Cr2θ ₃ content of the respective solid solutions can be selected to optimize the balance of the high temperature properties. According to 20. the present invention, the mole fraction of r^O, is preferabl from about .02 (2 mole %) to about .3 (30 mole %) and the mole fraction of Hfθ2 is preferably from about 0 to about .5(50 mole %) with 20 mole percent Cr2θ3 and 10 to 20 mole percent Hfθ2 representing a particularly preferred combination. 2-S It is an object of the present invention to provide material suitable for application as light diesel engine cylinder and head liner and piston cap. It is a further objec to provide transformation toughened ceramics useful in the super hot adiabatic engine, gas-turbine engine and recuperator 30 ^' heat exchanger applications. Fulfillment of these objects and the presence and fulfillment of additional objects will be apparent on complete reading of the specification and at¬ tached claims when taken in conjunction with the attached drawing. 35.

BRIEF DESCRIPTION OF THE DRAWINGS FIGURE 1 is a typical plot of the indentation diag¬ onal squared, a^, versus the applied load, P, used to evaluate the hardness, H _v , of a ceramic composition accord- ing to the present invention.

FIGURE 2 is a typical plot of the crack length of the three halves power, c ' , versus the applied load, P, used to valuate the fracture toughness, _jC , of a ceramic composition according Jzo the present invention. FIGURE 3 is a typical linear plot of the crack length to the three halves power, c 3'/2, versus the indentation diag¬ onal squared, a^, characteristic of a ceramic composition according to the present invention.

FIGURE 4 is a cross sectional view of a thermal con- ductivity specimen cell used to measure high temperature thermal conductivity of ceramic composition according to the present invetnion.

FIGURES 5. through 12 illustrate experimental data plots using identation test data measured for the (Al ₂ θ * xCr2θ )-15 vol. % (Crθ2* HfO_) ceramic compositions accord¬ ing to the present invention.

FIGURES 13 through 18 illustrate experimental data plots using indentation test data measured for the (3Al2θ3*-2Si0 ₂ + x[3Cr ₂ θ3*2Si0 ₂ J ) -(Zr0 ₂ *yHf0 ) ceramic com- positions according to the present invention.

FIGURES 19 through 23 illustrate experimental data plots using thermal conductivity data measured for composi¬ tions of Figures 5 through 12.

— o —

DESCRIPTION OF THE PREFERRED EMBODIMENTS The novel transformation toughened ceramic composi¬ tions according to the present invention, how they are pre- Dared and the pragmatic significance of their high temperatur properties can perhaps be best explained and understood by reference to a series of compositions characteristic of the alumina system and the mullite system. The particular alumin system of interest is the ceramic compositions having the con¬ tinuous matrix phase of A^O-^'xCr^O- _s solid solution and a ^• dispersed phase within this continuous matrix phase of finely . divided Zrθ2"yHf0 ₂ solid solution particles. The x and y re¬ present the mole fraction or mole percent of Cr^Ov relative to Al_0 ₃ and mole fraction or mole percent of Hfθ2 relative to ZrO- in the respective solid solutions. Similarly, the mullit system of interest is the ceramic compositions having a contin ous matrix phase of 3A1 ₂ 0 ₃ *2Si0 ₂ + x[3Cr ₂ 02 ^* 2Si0 ₂ l and a finel dispersed phase of Z θ2*yHfθ2 where x and y again represent th relative mole fraction or mole percent of the 3Cr2θ ₃ *2Siθ2 and HfO-, respectively. For purposes of this invention, the mole fraction or the mole percent, whether designated by x or y, refers to the relative mole fraction of the second compnent of a two compone system wherein it is to be understood that the mole fraction (mole %) of the first component plus the ^' x or y sums to unity; e.g., Zrθ2 ^* yHfθ2 means (1-y) Zrθ2*yHfθ2.

In order to study and evaluate the high temperature properties of the above ceramic compositions, twenty-two compositions in the alumina system and ten compositions in the mullite system were prepared. In the alumina system the continuous matrix phase, Al ₂ θ *xCr2θ ₃ , was present in 85 percent by volume and the dispersed phase, Zrθ2 ^* Hfθ2, was present in 15 percent by volume; i.e. (Al-0," Cr2θ3) plus 15 vol. % (Zr0 ₂ ^* yHf0 ₂ ) • In the mullite system, the continuous matrix phase, 3Al2θ ₃ -*2Siθ2 + x[3Cr2θ3 ^* Si0 ₂ _ , was present as the major phase and the dispersed phase, Zrθ2 ^* yHfθ2, was present as the minor phase. In both systems, the relative mole fractions of both the chromium oxide com-

ponent and the hafnium oxide component were varied such as to characterize a broad relative range of concentrations of each respective component. Thus, in the case of the alumina system, the twenty-two specimens were distributed as illust- rated in TABLE I wherein two of the samples involved either no Hf0 ₂ or no Zr0 ₂ (controls) and the remaining twenty in¬ volved compositions with all four oxides. Similarly, the x and y (relative mole fractions) of the mullite samples were intentionally distributed over a broad range of com- positions with four additional specimens less one component being included as controls (as explained later in TABLE II) .

TABLE I

COMPOSITIONS PREPARED IN THE ALUMINA SYSTEM: (Al ₂ 0 ₃ -χCr2θ3) + 15 vol. % (Zr0 ₂ 'yHf0 ₂ ) x, mole fraction of Cr ₂ 0 ₃ y, mole fracton of Hfθ2 in Al ₂ θ3 ^* xCr ₂ θ3* in Zrθ2 ^* yHfθ2

0 0.1 0.2 0.3 0.5 1.0

. 0 * * .02 * * * * .05 * * * * .1 * * ** *

2 * * * * _m 3 * ' * * *

** A1 ₂ 0 ₃ - Alcoa XA- 139 Cr2θ ₃ - Reagent grade,. J. T. Baker Chemical Company

Zrθ2 - Zircoa A

Hf0 ₂ - 99.9% Apache Chemicals, Inc.

Specimens representative of both the alumina and mullite composition were prepared by hot pressing of mixtures of pre-solutionized powders. Hardness and fracture toughness were determined by the microhardness indentation method. Thermal conductivity of the respective specimens was deter¬ mined by comparison with known standards. The following examples illustrate the preparation and composition of the specimens employed in the high temperature property evaluation.

EXAMPLE I

Five separate solid solutions of Al ₂ 03~Cr ₂ 02 having the mole fraction of C 2O3 specified in the ordinate of TABLE I and the four separate solid solutions of Zrθ2~Hf0 ₂ having the specified mole fraction of Hf0 ₂ as found in the abscissa of TABLE I were prepared by mixing appropriate amounts of metal oxides in a ball mill and then reacting the mixture at 1350°C for 24 hours. The metal oxides employed were from commercial sources as follows: AI2O3 was Alcoa XA 139; the Cr ₂ 0 ₃ was Reagent Grade from J. T. Baker Chemical Company ^; Zr0 _{2 was zircoa A and the Hfθ2 was} 99.9% Hf0 ₂ from Apache Chemical Inc. The twenty-two compositions specified in TABLE I were prepared by adding 85 parts by volume of the Al ₂ 0 ₃ *xCr ₂ 0 ₃ solutionized powder and 15 parts by volume of the Zr0 ₂ ^* yHfθ2 solutionized powder and then ball milling the mixture for 43 hours. Specimens for microindentation tests were hot pressed at 1600°C for one hour in boron nit- ride coated graphite dyes under a pressure of 30 MN/m . Aft hot pressing, the samples were then oxidized in air at 1350° for two hours. Specimens for thermal conductivity measure¬ ments were hot pressed at 1600°C for one hour under a

2 pprreessssuurree of 15 MN/m . In both cases, "full density was achieved.

EXAMPLE II In order to evaluate the* mullite system, (3Al2θ3«

2Si0 ₂ + [3Cr ₂ θ3 ^* Siθ2l)-(Zrθ2'yHfθ2) , a series of ten com¬ positions and four controls as specified in TABLE II was prepared. Half of the compositions (type B in TABLE II) were prepared by a physical mixing or blending technique ana gous to the process described in EXAMPLE I. The other half of the compositions were prepared by a co-precipitation tech nique (type C in TABLE II) . Appropirate stoichiometric amounts of aluminum hydroxide, silicic acid and chromic acid were used as starting materials for the acid/base neutrali- zation reaction used to prepare the mullite solid solutions. Oxide powders were used to prepare the Zr0 ₂ ~Hf0 ₂ solid solu¬ tions, again in a manner described in EXAMPLE I. Weighed owders were mixed in ro ortions corres ondin to TABLE

and ball milled for seventy-two hours. In the co-precipitated compositions, the slurries corresponding to the compositional properties of TABLE II were dried and hot pressed at 1550°C for 30 minutes at a pressure of 30 MN/m . The hot pressing (HP) and the annealing (Ann.) of the specimens was otherwise s r sQ £±&ά ia TA&LE XI.

TABLE II

Mole % of T. Relative Amount of Phases K ^r Pe IC H, M Cr HfO, HP Ann. Mullite ZrO- (Zr-Hf)0 ₂ (Zr ^« Hf)Si0 ₄ Cadmium (MPa'm) (kg/mm ² )

M-0-0 B

1500°C 1500°C 100 17.6 2.43 1180 30 min 1 hr

M-0-10 B 1600°C 1250°C 100 20.6 1.1 2.08 1207 1 hr 24 hrs

1500°C 1500°C 100 43.0 2.35 1217 30 min 1 hr ¹ M-0-20 B 1600°C 100 23.6 7.0 1.63 1178 1 hr

30 min 1 hr

M-5-10 B 1600°C 1350°C 100 23.6 2.8 1.94 1070 1 hr 24 hrs

1500°C 1500°C 100 30.7 4.1 2.76 1083 30 min 1 hr

M-5-20 B 1600°C 1550°C 100 30.5 4.2 0.7 1.70 1062. 1 hr 2 hrs

1500°C 1500°C 100 30.5 3.2 2.76 1219

With respect to measuring and evaluating the high temperature properties of the ceramic compositions of EXAMPLES I and II, the microindentation technique was used to determine relative fracture toughness and absolute hard- ness. A Tukon microhardness testing machine was used for microhardness and indentation fracture toughening studies. A minimum of five indentations were made at.each of the five or more different loads for each sample. The loads varied from 2 kg to 15.9 kg depending on specimen composition. A Vickers diamond indentor (136°) was used in all studies. The following equation (1) developed by Anstis et al (for further explanation see "A Critical Evaluation of Indentation Techniques for Measuring Fracture Toughness: I. Direct Crack Measurements, " J. Am. Ceraπu Soc. 64 (9) 33-8 (1981.)) was used to calculate these two material properties. constant = κ _χc (H/E) ^l / ² / (?/c ^3/2 ) (1)

The above equation relates a material independent constant to the fracture toughness, hardness, elastic modulus, crack size and applied load. According to this equation, the c 3'/2 versus P plot where c is the crack length and P is the applied load should yield a straight line with a slope equal to: slope = constant (E/H) ^λ ^ ² /K _IC (2)

By rearranging the above equation and solving for fracture toughness, the following is calculated: K _IC -= constant (E/H) ¹ ^ ² /slope (3)

The hardness was determined from the a versus P curve where a is one half of the diagonal of the indentation. From Hucke's work ("Process Development for Silicon Carbide Base Structural Ceramics", Report DAAG 46-80-C-0056-P0004, June, 1982, AMMRC, Watertown, Massachusetts), this hardness, H , is independent of load and is the hardness at large loads. The equation used to calculate this value is:

H _v = k/s' (4) where k is a proportionality constant for a Vickers diamond indentor (136°) and is equal in this case to 4,636. The slope, s', is determined from the a~ versus P curve. The value of the constant in equation (4) is of little signifi¬ cance since a standard (NC203) was used to make all the cal¬ culations.

To determine _χc , the value of the elastic modulus, E, for the composition being measured must be available or determined. For the alumina system, the literature value of the elastic modulus for I2O3 of 60 x 10° psi was used (see Engineering Property Data on Selected Ceramics, Vol.

Ill, Single Oxides, MCIC-HB-07, Metal and Ceramics Informa¬ tion Center, Battelle, Columbus, Ohio) . It was assumed that this value was constant and characteristic of all alumina compositions employed. For the mullite system a value of 25 x 10° psi was used. This value is intermediate between g two values, 21 and 32 x 10 psi, reported in literature.

(See Van Vlack, L.H., Elements of Material Science, Addison-

Wesley, Reading, Massachusetts, pf. 5.4.1-22 and Matdigas i,

K. S., and Brown, L. M. , "Synthesis in Mechanical Properties of Stoiσhiome ric Alumin Silicate (Mullite) , J. Am. Ceram. Soc. 55 (11) 548-552 (1972)).

Plots of a ² versus P and σ ' versus P curves were constructed for all indentation test data generated. The degree of linear fit was excellent, r ² = .99, for a 11 a ² versus P curves. The linear fit was good, for most tests r ² was at least .97, for the c ³ ' versus P. Curves. Figures

1 through 3 of the drawings illustrate typical a ² versus P,

3/2 3/2 2 c ' versus P, and σ ' versus a curves generated by plotting the respective data characteristic of compositions according to the present invention. As clearly indicated in Figures 1 through 3, the experimentally measured data is essentially linear across the entire range of interest consistent with the nature of the above equations.

The thermal conductivity measurements of compositions according to the present invention were performed by a com¬ parative method. As illustrated in Figure 4, the method in¬ volved the use of a commercially available comparative ther¬ mal conductivity instrument manufactured by Dynatech Corpora¬ tion (Model TCFCM) . The comparative measurement involved placing test specimen 10 to be measured within the Model TCFCM 12 such that it is stacked between a top reference standard specimen 14 and a bottom reference standard 16. As further illustrated in Figure 4, the stack of specimens

is sandwiched between an upper surface plate 18 and a lower surface plate 20 which in turn rest on auxiliary heater 22 and is capped by a main heater 24. The top heater 24 in turn is covered by insulation, while the auxiliary heater 5 22 rests on a spacer 28 and heat sink 30. A plurality of thermocouples 32 are stategically positioned within the sample cavity 34 at critical interfaces, such as to make temperature measurements while pressure pad 36 compresses the stack of specimens. 0 The actual measurement and computation of the thermal conductivity is defined by the following equation: q = kA(dT/dx) (5) where q is heat flux, A is the specimen cross section area, T is temperature, x is distance tween two points in the 5 sample and k is the desired thermal conductivity. When the specimens with the same A of two different materials are arranged as shown in Figure 4 then:

T(k/dx) _sara pi _e = ΔT(k/ Δx _refe rence ⁽⁶⁾ Thus, by measuring the temperature difference between two 0 thermocouples at a distance x apart in both reference ma¬ terials and te test sample, the thermal conductivity of the test material can be evaluated.

Figures 5 through 12 graphically present and summarize the results of the indentation tests on the alumina system 5 compositions of EXAMPLE I as specifically set out in TABLE

III. The twenty-two alumina compositions measured contained about 2 to about 30 mole percent C 2O3 in the alumina/chromium solid solution matrix and from 0 to about 50 mole percent Hfθ2 in the zirconia/hafnium dioxide dispersed solid solution par- 0 tides. As illsutrated in Figures 5 through 8 and as pre¬ sented in TABLE III, the measured fracture toughness generally decreased with increasing Hfθ2 content at constant Cr2θ3 con¬ tent. This may be attributed to the decrease in the critical particle size of the tetragonal-monoclinic phase transition 5 with increasing Hfθ2 content. However, it has not been veri¬ fied that the particle size of the dispersed Zrθ2~Hfθ2 solid solution phase in any of the samples was small enough to re¬ tain the -tetragonal phase (as further discussed below) .

TABLE III

SAMPLE THERMAL CONDUCTIVITY

⁽ A1 ₂ 0 ^* xCr ₂ 0.,)- k = q/A- 15 vol % (Zr0 ₂ - (dt/dx)

₂ - yHfo ₂ ) (ca1/cm• °C-sec)

X y 70°C 250°C 400°C

0 - 100 0.0399 0.219 0.0156

0 - 0 0.0273 0.0177 0.0137

2 - 10 0.0378 0.0240 0.182

2 - 20 0.0332 0.018.0 0.0132

2 - 30 0.0348 0.0179 0.0120

2 - 50 0.0292 0.0197 0.0147

5 - 10

5 - 20 0.0274 0.0174 0.0133

5 - 30 0.0277 0.0164 0.0121

5 - 50 0.0274 0.0165 0.0127

- 10 0.0165 0.0125 0.0108

- 20 0.0241 0.0170 0.0138

- 30 0.0229 0.0144 0.0108

- 50 0.0233 0.0152 0.0118

- 10 0.0153 0.0105 0.0082

- 20 0.0179 0.0121 0.0094

- 30 0.0188 0.0137 0.0111

- 50 0.0191 0.0138 0.0110

- 10 0.0168 0.0122 0.0099 - - 20 0.0156 0.0113 0.0093 - - 30 0.0178 0.0139 0.0119

OMPI

Figures 9 through 12 and TABLE III illustrate that for constant Hfθ2 contents of 10, 20 and 50 mole percent there appears to be a shift in the peak of the fracture toughness, KiC versus Cr2θ ₃ content curve. The peaks appear at 10, 20 and 30 mole percent Cr2θ3 content, respectively. The fracture toughness continually decreased with increasing Cr2θ ₃ content with the effect of the Hfθ2 content on the hardness being less discernible. Generally, the hardness stayed fairly constnt with increasing Hf0 ₂ content and con- stant Cr ₂ 0 ₃ content.

The average particle size for the dispersed Zrθ2~ Hfθ2 phase in most of the above specimens was of the order of 5 am. This is felt to be too large to effectively retain the metastable tetragonal phase which, as previously indicated, mechanistically is thought to produce an increase in fracture toughness due to a stress-induced phase transformation of even greater significance than the originally observed phenomena associated with PSZ. In this respect, the time- stabilized zirconia and the associated high energy absorbing tetragonal to monoclinic transition beneficial effects may • not be fully realized (a feature which from a particle size analysis should be kept in mind when evaluating the co- precipitation versus ball milled specimens of EXAMPLE II) . In interpreting the above experimental data and property measurements, it should also kept, in mind that using a constant modulus value, E, for all compositions will intro¬ duce some error in the calculated fracture toughness value. Although present, this error is felt to be small.

In view of the above data, the transformation toughened ceramic alloys involving twenty percent Cr2θ ₃ -10% Hf0 ₂ and/or 20% Cr2θ ₃ -20% Hfθ2 having a relatively high Cr2θ ₃ content should exhibit significantly increased hard¬ ness and modulus. Further, high Hf©2 content (e.g. 50 mole %) should significantly increase the transformation tempera- ture and thus, increase the potential for transformation toughening. For the same reason, the critical particle size for transformation is preferably decreased. To achieve a

reasonable increase in the toughening, and keep the critical particle size within pragmatic limits, intermediate values of Hf0 ₂ (e.g., about 10 to about 20 mole %) in the dispersed solid solution phase are preferred. 5 Figures 13 through 18 graphically display and summar¬ ize the results of the indentation tests on the mullite sys¬ tem of EXAMPLE II as specifically found in the far right columns of data of TABLE II. Again, the figures illustrate and suggest certain general trends in the data. For example,

IX). Figures 13 through 15 suggest that for the co-precipitated samples, keeping Hf0 ₂ constant and increasing the Cr mullite content, results in an increase in the fracture toughness; while, the effect of a change in the Hfθ2 content on fracture toughness is more obscure. Similarly, for mechanically mixed

15 samples, an increase in HfO~. content generally results in a decrease in fracture toughness at constant Cr mullite content. With no Hfθ2 the fracture toughness remains constant while increasing the Cr mullite content. In all cases except one, for the same composition, the fracture toughness of the co-

2-0 precipitated samples was greater than the fracture toughness of the mechanically mixed samples. This may be due to the difference in hot pressing conditions and/or subsequent particle size and distribution.

Figures 16 through 18 indicate that the hardness of

25 the co-precipitated samples increased with increasing Hf0 ₂ content with 0 percent and 5 percent Cr mullite. The hard¬ ness stayed relatively constant with increasing Hf0 content with 10 percent mullite. For the mechanically mixed specimens, the hardness generally decreased with increasing Hf0 ₂ content. 0 For seven of the ten compositions the hardness of the co- precipitated samples were greater than the corresponding values of the mechanically mixed samples. Thus, the co- precipitation method of sample preparation appears to enhance both fracture toughness and hardness simultaneously.

35 The thermal conductivity of the twenty-two compositions in the alumina system, (Al2θ3"XCr2θ ₃ )-(Zr0 ₂ *yHfθ2) , prepared in EXAMPLE I were measured at three different temperatures in the previously described thermal conductivity instrument

illustrated in Figure 4. The specimens for these measure¬ ments were hot pressed cylinders 3 centimeters in diameter and 2 centimeters in height. Both top and bottom surfaces were lapped with 15 im diamond disk. The results of the thermal conductivity.measurements expressed in cal/cm ^* °Csec at the three respective temperatures are given in TABLE III. The sample notations are expressed in terms of the x and y of the above alumina formula and represent the relative mole percent of the Cr ₂ C* ₃ in the matrix phase and the Hf0 ₂ in the dispersed phase. The data are also plotted as thermal con¬ ductivity versus Cr-O- content at various Hfθ2 contents in Figures 19 through 22. As indicated in these figures, the curves can be extrapolated back to zero percent Crθ2 and no Hfθ2» From the extrapolation, it can be concluded that pure ^Zr θ2 dispersed in a pure AI2O matrix should have a thermal conductivity value of about 0.038 cal/cm-°C-sec. Greve et al (see "Thermal Diffusivity/Conductivity of Alumina with Zirconia Disperesed Phase" Am. Ceram. Soc. Bull. 56 (5) 514- 5 (1977)) reported that an Al ₂ 0_ matrix containing fifteen volume percent Zrθ2 had a thermal conductivity value of 0.018 cal/cm-°C* sec. Thus, the 70°C thermal conductivity curves in Figures 19 through 22 can be normalized by multiplying a factor of 0.018/0.038 to become the normalized curve shown in Figure 23. The compositions of the present invention are advan¬ tageously employed as transformation toughened ceramics at high temperatures in application that require a balance of physical, chemical and mechanical properties. In particular, the compositions are useful in that the minor constituents can be altered to take advantage of the specific enhancement of critical properties such as hardness, fracture toughness and low thermal conductivity (as demonstrated in the presented test data) without significantly sacrificing the known corrosion, wear and high temperature stability of the major ceramic constituent. As such, the present invention is to be viewed as being directed broadly to a class of novel ceramic compositions wherein the Cr ₂ 0- content in the matrix phase and the Hfθ2 in the zirconia dispersed phase can

varied to achieve the improvement in the critical properties. Similarly, this novel concept is envisioned as being applic¬ able and beneficial at various relative proportions of dis¬ perse phase to hostmatrix phase. Having thus described and exemplified the preferred embodiments with a certain degree of particularity, it is to be understood that the invention is not limited to the em¬ bodiments set forth herein for purposes of exemplification, but is to be limited only ^" by the scope of the attached claims, including a full range of equivalents to which each element thereof is entitled.