ZrB2 - 2
ZrB2 - 2
www.elsevier.com/locate/jeurceramsoc
Abstract
Two zirconium diboride-base composites were produced and characterised. The chosen starting compositions were: 55 wt.%
ZrB2+41 wt.%TiB2+4 wt.% Ni and 83 wt.% ZrB2+13 wt.% B4C+4 wt.% Ni. The microstructure and properties of these com-
posites were compared to those of a monolithic ZrB2+4 wt.% Ni material. In all cases, metallic Ni as the sintering aid promoted
the formation of the liquid phase which improved mass transfer mechanisms during sintering. From the powder mixture
ZrB2+TiB2, two solid solutions of Zr–Ti–B were obtained. In the case of the other mixture, B4C particles were dispersed in the
ZrB2 matrix. The composite materials have better mechanical properties than those of the monolithic ZrB2 ceramic; in particular
the fracture toughness and the flexural strength were almost doubled at room temperature. Long term oxidation tests indicated that
the ZrB2-based composites, particularly the composite containing B4C as the second phase, were more resistant to oxidation than
the monolithic ZrB2 due to the formation of surface oxide products which were protective against the complete degradation by
oxidation observed for the ZrB2 matrix material. # 2001 Elsevier Science Ltd. All rights reserved.
Keywords: B4C; Composites; Electrical properties; Mechanical properties; Thermal properties; TiB2; ZrB2
of electroconductive monolithic or composite ZrB2-based shown in Table 2. A Ni powder was selected as a sin-
ceramics, the possibility of machining them by electrical tering aid for the monolithic and composite materials.
discharge offers a powerful tool for the manufacture of The powder batches were homogenised by wet mixing
complex shaped components.2629 Another critical aspect in ethanol using zirconia balls. The hot pressing cycles
of refractory borides is their low thermal stability in air at were carried out under vacuum using a BN-lined gra-
high temperature, although several studies have shown the phite die.
formation of a surface protective scale of borosilicate The microstructure was analysed by scanning electron
glasses.3 Again, attempts were made to improve the oxi- microscopy and EDX analysis on polished and fracture
dation resistance of ZrB2-based materials, through the surfaces. The crystalline phases were identified by X-ray
addition of appropriate additives.9,20,3032 diffraction analyses.
This work aims to highlight the microstructure and The following mechanical properties were measured:
properties of two ZrB2-based ceramic composites elastic modulus (E) by the resonance frequency method
obtained from fine commercial ZrB2 powders and the on 2880.8 mm3 samples, 4-pt flexural strength () on
addition of TiB2 or B4C as secondary phases. Micro- 252.52 (lengthwidththickness) mm3 test bars,
structure, mechanical properties, electrical resistivity using an outer span of 40 mm and an inner span of 20
and oxidation resistance were evaluated and compared mm with a crosshead speed of 0.5 mm/min, from room
to those of monolithic ZrB2.33 temperature up to 1400 C; microhardness (HV1.0) by a
Vickers indenter with an applied load 9.81 N; fracture
toughness (KIC ) using two methods: (i) direct crack
2. Experiments measurement (DCM)34 with a load of 98.1 N, (ii) chev-
ron notched beam method (CNB)35 on 2522.5
ZrB2, TiB2 and B4C powders produced by H. C. (lengthwidththickness) mm3 bars with a machine
Starck (Germany) were selected as raw materials. crosshead speed of 0.05 mm/min.
Their main properties are summarised in Table 1: the The thermal expansion behaviour was tested in air up
chemical composition was supplied by the producer to 1400 C.
except the oxygen amounts which were determined by The electrical resistivity measurements were carried
LECO. The specific surface areas were measured by the out by a four probe DC method at room temperature,
BET method while grain size and morphology were inducing a current in bar specimens of 22.525 mm3.
evaluated by SEM analyses. The particle shape of all the The current and the voltage readings were detected at
powders was irregular and mainly acicular, with a large the same time in two different digital high-resolution
grain size distribution. multimeters. The resistivity values were determined
Compositions of the two ZrB2-based composites, of from the electrical resistance measurement taking
the monolithic material and the adopted parameters account of the test lead distance and cross-section area
necessary to obtain high densities by hot pressing are of the samples.
Table 1
Characteristics of the raw powders (from H. C. Starck)
Table 2
Composition, hot pressing parameters and densities of the hot pressed samples
Long term oxidation behaviour at 1000 C was studied analysis revealed areas containing metallic Ni with tra-
in a laboratory kiln with interruptions in the tests in order ces of oxygen as marked in Figs. 1b and 2a. Metallic
to measure weight change at fixed times. Microstructure nickel could not be detected on X-ray diffractograms, as
of the surface and polished cross-sections of the oxidised its main peak is superimposed by a peak of zirconium
samples were analysed by X-ray diffraction analyses, diboride.
scanning electron microscopy and EDX analyses. SIMS analyses, reported in a previous work,36
revealed the presence of phases such as Zr–B–O, Ni–B,
Zr–Ni–O in this system. These phases show some degree
3. Results and discussion of reactivity between the different elements during hot
pressing. Ni reacts with the O2 present as an impurity in
3.1. Microstructure of the hot pressed materials the starting ZrB2 powders; subsequently, the exchange
reactions with ZrB2 give Ni2B and ZrO2. As confirma-
Material A reached a final density of 6.05 g/cm3 tion, small ZrO2 particles were detected, mostly within
which is about 98% of the theoretical density. The ZrB2 grains (Fig. 1b). Moreover, (Zr-O) compounds,
shape and size of ZrB2 grains turned out to be quite probably Zr2O5,36 were also present in the spherical
homogeneous; Ni-rich grain boundary phases were inclusions distributed mainly at grain boundaries (Fig. 2a
located mainly at triple points (Fig. 1a and b). It has to and b). The internal part of these inclusions consisted of
be pointed out that in the back scattered electron ima- an aggregation of small zirconium oxide particles
ges (an example of which is shown in Fig. 1b), the con- embedded in a boron oxide-based glassy phase where
trast arises from the channelling of the accelerated zirconium, oxygen, and low amounts of impurities like
electrons as a consequence of the different orientation of Al and Si segregated, as revealed by EDX analyses. It
the ZrB2 crystals and not from the change of the mean can be hypothesised that the boron oxide-liquid phase
atomic number of the phases present, i.e. variations in formed at high temperature as a consequence of the
composition and stoichiometry. From X-ray diffraction
analyses, traces of Ni2B were identified and micro-
Fig. 1. (a) fracture surface of material A: ZrB2+4 wt.%Ni; (b) back Fig. 2. Back scattered electron images of a polished section of material
scattered electron image of a polished surface. The following details A. In picture a: (1) Ni-rich grain boundary phases; (2) ZrB2 grains; (3)
are marked: (1) Ni-rich grain boundary phase; (2) ZrB2 grains; (3) spherical glassy inclusions, containing clusters of precipitated ZrO2
spherical glassy inclusions; (4) ZrO2 particles. nanoparticles. The microstructural features are highlighted in picture (b).
282 F. Monteverde et al. / Journal of the European Ceramic Society 22 (2002) 279–288
concentration of oxygen and other impurities and their with different volume fraction and stoichiometry (Fig. 3).
reaction with boron and that the glass formed during On the basis of the cell parameters, two solid solutions
cooling from the hot pressing temperature. Zirconium were identified, one rich in titanium, (Ti0.8Zr0.2)B2, and
oxide nanoparticles later precipitated in this glass. SIMS the other rich in zirconium, (Zr0.8Ti0.2)B2, in amounts of
analyses36 brought out two important points: (i) the 44 and 37 by vol.% respectively. Of the starting borides,
presence of several BxOy and ZrBxOy species, indicating an amount of about 13% of ZrB2 was left, while TiB2
boron’s tendency to form clusters and, (ii) the presence disappeared almost completely indicating that under the
of other contaminants (Na, Ca, K, Cr, Ti) in addition to adopted hot pressing conditions, the dissolution and
those indicated by the powder supplier. segregation of zirconium in TiB2 was favoured in com-
The above reported results indicate that during hot parison to the solubility of titanium in ZrB2. However, a
pressing another melt formed, besides the Ni-based complete solution did not occur as the processing tem-
melt: a liquid glassy phase formed by B2O3 and metal perature of 1600 C was well below 2000 C. Additional
cations, such as Zr and impurity elements. It is likely factors, which could have influenced the rate of forma-
that a liquid film of B2O3 formed at grain boundaries tion of solid solutions and their stoichiometry, were: the
due to the oxygen released from ZrB2 powder particles. difference in atomic radii between Ti and Zr (Zr being
B2O3 melts at 723 C and vaporises at temperatures about 10% larger than Ti38), the heating atmosphere,
above about 800 C, but due to the applied pressure the cooling rate after hot pressing (slow cooling may
during sintering and to the fact that the liquid and induce decompositions15), the relative amounts of the
vapour boron oxide-based phases were contained within starting phases, the impurities present in the starting
closed interstitial spaces among solid particles, B2O3 powders, and the addition of metallic sintering aids
could not volatilise and remained in a liquid state. which promote the formation of metal borides as sec-
Probably dewetting took place, as the liquid assumed ond phases.14,15,38 In the present case a low amount of
approximately the shape of a sphere. This means that Ni2B was actually found. Oxygen impurity leads to the
once formed, the liquid did not wet the zirconium formation of zirconia particles randomly distributed
diboride particles (depleted of the amount of the surface (Fig. 4a and b) at grain boundaries. The various phases
oxygen) and the interfacial area was reduced due to a and relative compositions are marked in the micro-
strong driving force for shape change towards the geo- graphs of Fig. 4a and in the EDX spectra of Fig. 4c and
metrical configuration with the lowest surface energy. d. During sintering, two phenomena affect the forma-
The microstructural features (grain size and shape tion and distribution of solid solutions:38 (i) a composi-
and grain boundary phases) confirmed also that the tional gradient in the solid solution grains, which should
presence of Ni promoted the formation of a liquid phase be limited to the grain surface for equilibrium reasons;
at high temperature, similar to the phenomena observed (ii) a graded concentration of solute at triple points and
during the sintering of TiB26 and in agreement with within ZrB2 grains, since segregation is enhanced by the
previous results on the effects of the addition of Fe to different diffusion coefficients. Both these features were
ZrB2.7 The Ni-based liquid phase has two main effects: observed in the micrographs of Fig. 4a and b. Increased
(i) it favours powder particle rearrangement, and (ii) solution concentration caused lattice dilatation. These
enhances mass transfer mechanisms such as activated microstructural features affect the material properties.
volume diffusion, surface diffusion and formation of a In particular, as lattice anisotropy and elastic strain
solid solution.37 The temperature of this liquid phase
formation, as indicated by the densification curve, is
about 1260 C. Evaporation phenomena of ZrB2 was
also retarded due to the presence of liquid phase and
enhanced grain growth was avoided. Being dependent
upon the wetting behaviour, which is influenced by the
surface oxidation of ZrB2 particles, some residual pores
which were not completely infiltrated by the liquid
phase remained at the triple junctions.
Composite B, produced by the mixture ZrB2+
TiB2+Ni, reached a density of 5.65 g/cm3 which is
higher than the theoretical one calculated on the basis of
the starting components. This was due to the formation
of (Ti, Zr)B2 solid solutions as the two starting diborides
are mutually and fully soluble at T>2000 C.16,38 The
X-ray diffraction spectra of the crystalline phases and
their relative amounts pointed out that the main part of Fig. 3. X-ray diffractograms of sample B (composite ZrB2/TiB2), evi-
the starting phases formed two solid solutions (Ti, Zr)B2 dencing the formation of two different Zr–Ti–B solid solutions.
F. Monteverde et al. / Journal of the European Ceramic Society 22 (2002) 279–288 283
Fig. 4. (a) Back scattered electon image of a polished surface of composite B. The following phases are indicated: (1) ZrB2; (2) TiB2; (3) (Zr,Ti)B2;
(4) (Ti, Zr)B2; (5) ZrO2. (b) Secondary electron image of the same area represented in (a). (c) Windowless EDX spectra from the two phases indi-
cated as 1 and 2 in picture (a). (d) Windowless EDX spectra of the two solid solutions indicated as 3 and 4 in picture (a).
energy increase in the solid solutions,38 they determine concurrent factors: (i) the relatively low dissociation
the grain boundary cohesion in polycrystalline materi- energy for B4C that favoured the formation of volatile
als. Grain size was not homogeneous, and the material carbon oxide species and in boron (boron oxide) spe-
contained some defects and microcracks (Fig. 4a and b). cies, preventing also the formation of zirconia particles,
It was not possible to confirm the amount of each phase (ii) the reaction between boron oxide and nickel (prob-
by image analysis because the correspondence grey ably nickel oxide) that resulted in liquid phases (above
level-phase was disturbed by the channelling effect of about 1300 C).40 These liquids either could have been
the electrons. squeezed out by the applied pressure due to their low
Composite C, produced with the addition of 13 wt.% viscosity or could have originated volatile species, being
of B4C to the ZrB2+4 wt.%Ni matrix material, had a the processing temperature (1870 C) very high.
final density of 5.16 g/cm3 which is about the 99.6% of
the theoretical density. The microstructure (Fig. 5a–c) 3.2. Mechanical properties
showed ZrB2 grains of 5–15 mm mixed with B4C grains
of smaller dimensions (1–10 mm). The only crystalline The mechanical properties of the three tested materi-
phases revealed by X-ray diffraction analyses were ZrB2 als are summarised in Table 3.
and B4C. Very low amounts of Ni-based phases were
only randomly found within B4C particle aggregates 3.2.1. Young’s modulus
(Fig. 5c). Therefore, the main part of nickel introduced This property was clearly dependent on the composi-
as sintering aids was lost during hot pressing due to two tion. The value measured in material A is higher than
284 F. Monteverde et al. / Journal of the European Ceramic Society 22 (2002) 279–288
values found in literature.1,14,15 The addition of TiB2, and phase. In this respect, material B was tougher than
the consequent formation of (Ti,Zr)B2 solid solutions, material A and material C was tougher than material B.
decreased its value in agreement with literature data.1416 While the toughness values measured by DCM and
CNB were in a very close agreement for the composite
3.2.2. Hardness materials, the CNB value measured on material A was
The sample with the highest value resulted the com- higher than the DCM value. As the crack pattern gen-
posite containing B4C, however all the measured values erated by indentation in this material (Fig. 6) is defi-
were in the range reported in literature.1,4,16 nitely far from any reasonable approximation, scarce
reliability should be attributed to the data collected with
3.2.3. Fracture toughness this method. A direct comparison of our toughness
The comparison of the fracture toughness values values with those reported in literature is blurred by the
revealed the toughening effect associated with the addi- different compositions and densities of the tested mate-
tion of a second reinforcing phase or to a solid solution rials as well as the numerous models, equations and
experimental parameters used to calculate the fracture
toughness by DCM. The authors are not aware of any
fracture toughness estimation by CNB of this kind of
material.
Table 3
Thermo-mechanical properties and electrical resistivity of the ZrB2-based materials
p
Sample (m.cm) E (GPa) l (106 / C) HV (GPa) KIC (MPa m) (MPa)
A 7.42 496 7.49 7.70 7.51 14.40.8 2.80.3 3.380.42 37124 616 8 62437 2373
B 15.55 439 7.79 8.25 8.63 17.70.5 4.30.3 4.090.14 599167 — 58228 25014
C 16.08 448 7.20 7.72 8.25 19.21.1 4.5.3 4.530.24 64386 — 57399 14810
F. Monteverde et al. / Journal of the European Ceramic Society 22 (2002) 279–288 285
4. Conclusions
Ni (4 wt.%) powder promoted the formation of a liquid 8. Øvrebø, D. N. and Riley F. L., Densification of zirconium
phase which improved mass transfer mechanisms during diboride. In Sixth-EcerS Conference & Exhibition’’, Extended
Abstract, Vol. 2. IOM Communications, University Press, Cam-
sintering.
bridge, UK, 1999, pp. 19–20.
From the powder mixture ZrB2+TiB2, solid solutions 9. Ogata, T., Mori, T., Nakamura, K., Kobaiashi, K. and Kuwa-
Zr–Ti–B with two different stoichiometries were jima, H., Mechanical properties of particulate dispersed and SiC
obtained. In the case of the other mixture, B4C particles whiskers reinforced boride composite materials by hot pressing.
are dispersed in the ZrB2 matrix. Microstructural fea- In Advanced Structural Inorganic Composites, ed. P. Vincenzini.
tures are discussed in relationship with the processing Elsevier Science, Amsterdam, 1991, pp. 235–243.
10. Sorrell, C. C., Stubican, V. S. and Bradt, R. C., Mechanical
parameters and starting compositions. properties of ZrC/ZrB2 and ZrC/TiB2 directionally solidified
From the mechanical point of view the composite eutectics. J. Am. Ceram. Soc., 1986, 69, 317–321.
materials generally performed better than the mono- 11. Sheppard, L. M., Advances in Composites processing. Ceram.
lithic ZrB2 ceramic: at room temperature, the fracture Ind., 1996, 146, 72–76.
toughness and the flexural strength were almost dou- 12. Cheng, Y. M. and Gadalla, A. M., Synthesis and analysis of
ZrB2-based composites. Mater. Manuf. Processes, 1996, 11, 575–
bled. The high temperature strength behaviour was dif- 587.
ferent for the monolithic and the composite materials 13. Schuldies, J. J. and Branch, J. A., Ceramic composites: emerging
due to the different grain boundary phase characteristics. processes, applications. Ceram. Ind., 1992, 138, 43–46.
Oxidation tests indicated that the ZrB2-based composites, 14. Mroz, C., Processing and properties of microcomposite TiZrC
particularly composites containing B4C as a second phase, and TiZrB2 materials. Ceram Eng. Sci. Proc., 1993, 14(9-10),
725–735.
were more resistant to oxidation than the matrix material 15. Mroz, C., Processing TiZrC and TiZrB2. Am. Ceram. Soc. Bull.,
due the formation of surface oxide products which were 1994, 73, 78–81.
protective against the complete degradation by oxidation 16. Moryiama, M., Aoki, H. and Kobayashi, Y., Fabrication and
as observed for the ZrB2 matrix material. mechanical properties of hot pressed TiB2–ZrB2 ceramic system.
From the technological point of view, the ability of the J. Ceram. Soc. Jap., 1998, 106(12), 1196–1200.
17. Jin, Z., Zhang, G., Bao, Y., Bai, C., Yue, X. and Zhao, H., Size
diborides to form solid solutions provides an opportunity effect of strengthened particles of ceramic Composites. J. Chin.
to manufacture high density materials at relatively low Ceram. Soc., 1995, 23(6), 610–617.
temperature, in fact composition B reached full density at 18. Jimbou, R., Takahashi, K., Matsushita, Y. and Kosugi, T., SiC–
1600 C, about 250 C lower than in the case of the other ZrB2 electroconductive ceramic composite. Adv. Ceram. Mat.,
1986, 1, 341–346.
two compositions A and C.
19. Woo, S. K., Kim, C. H. and Kang, E.S., Fabrication and micro-
structural evaluation of ZrB2/ZrC/Zr compoisites by liquid infil-
tration. J. Mater. Sci., 1994, 29, 5309–5315.
Acknowledgements 20. Bundschuh, K., Schuze, M., Muller, C., Greil, P. and Heider, W.,
Selection of materials for use at temperatures above 1500 C in
This project was supported by ASI (Agenzia Spaziale oxidizing atmospheres. J. Eur. Ceram. Soc., 1998, 18, 2389–2391.
21. Janeway, P. A., Reinforcing tomorrow’s technology. Ceram Ind.,
Italiana), Project NEIN700257. 1992, 138, 42–44.
22. Breval, E. and Johnson, W. B., Microstructure of platelet rein-
forced ceramics prepared by the direct reaction of zirconium with
References boron carbide. J. Am. Ceram. Soc., 1992, 75, 2139–2145.
23. Champion, Y. and Hagege, S., Study of composite interfaces in
1. Morz, C., Annual mineral review. Zirconium diboride. Am. the Zr–ZrB2 system. J. Mater. Sci. Lett., 1992, 11, 290–293.
Ceram. Soc. Bull., 1995, 74(6), 165–166. 24. Jiang, D., Li, Y., Wang, J, Ma, L. and Yang, X., Study of
2. Low, I. M. and McPherson, R., Fabrication of new zirconium strengthening of SiC–ZrB2 multiphase ceramics. J. Chin. Ceram.
boride ceramics. J. Mater. Sci. Lett., 1989, 8, 1281–1285. Soc., 1990, 18, 123–129.
3. Upadhya, K., Yang, J. M. and Hoffmann, W.P., Materials for 25. Rahman, M., Wang, C. C., Chen, W., Akbar, S. A. and Mroz,
ultrahigh temperature structural applications. Am. Ceram. Soc. C., Electrical resistivity of titanium diboride and zirconium
Bull., 1997, 58, 51–56. diboride. J. Am. Ceram. Soc., 1995, 78, 1380–1382.
4. Van de Goor, G., Sagesser, P. and Berroth, K., Electrically con- 26. Jimbou, R., Suzuki, Y. and Takahashi, K., Effect of dispersion of
ductive ceramic composites. In Advanced Multilayered and Fibre- ZrB2 on resistivity of SiC-ZrB2 electroconductive ceramic com-
Reinforced Composites, ed. Y. M. Haddad. Kluwer Academic, posites. J. Ceram. Soc. Jap., 1990, 98, 225–230.
The Netherlands, 1998, pp. 311–322. 27. Jimbou, M. R., Suzuki, Y. and Takahashi, K., Effect of the dis-
5. Hayami, R., Iwasa, M. and Kinoshita, M., Effects of applied persion of ZrB2 on resistivity of Si-ZrB2 electroconductive cera-
pressure on hot pressing of ZrB2. Yogyo-Kyokai-Shi, 1987, 88, mic composites. J. Ceram. Soc. Jpn. Inter. Ed., 1990, 98, 234–239.
352–358. 28. Nakamura, M., Shigematsu, I., Kanayama, K. and Hirai, Y.,
6. Bellosi, A. and Monteverde, F., Microstructure and properties of Surface damage in ZrB2-based composite ceramics induced by
titanium nitride and titanium boride-based ceramics. In Engineering electro-discharge machining. J. Mater. Sci., 1991, 26, 6078–6082.
Ceramics: Multifuncional Properties-New Perspectives, Key Engi- 29. Cheng, Y. M., Eubank, P. T. and Gadalla, A. M., Electrical dis-
neering Materials, Vols. 175–176, ed. P. Sajgalik and Z. Lences. charge machning of ZrB2-based ceramics. Mater. Manuf. Pro-
Trans Tech Publications, Switzerland, 2000, pp. 130–140. cesses, 1996, 11, 565–574.
7. Woo, S.-K., Han, I.-S., Kang, H.-S., Yang, J.-H. and Kim, C.- 30. Pankov, G. A., Fomina, G. A., Ivanov, D. A. and Val’yano,
H., Sintering of zirconium diboride through Fe-based liquids G. E., Strength and scaling resistance of a composite based on
phases. J. Korean Ceram. Soc., 1966, 33, 259–262. zirconium diboride. Refractories, 1994, 35(9-10), 298–300.
288 F. Monteverde et al. / Journal of the European Ceramic Society 22 (2002) 279–288
31. Kobayashi, K., Sano, H., Maeda, K. and Uchiyama, Y., Oxidation 36. Daolio, S., Fabrizio, M., Piccirillo, C., Muolo, M. L., Passerone,
behaviour of graphite-B4C/SiC/ZrB2 composite in dry and moist A. and Bellosi, A., Secondary ion mass spectrometry in the char-
atmospheres. J. Ceram. Soc. Jap. Int. Ed., 1992, 100, 407–411. acterisation of boron-based ceramics. Rapid Commun. Mass
32. Zhong, X. and Zhao, H., High temperature properties of refrac- Spectrom., 2001, 15, 1–7.
tory composites. The Am. Ceram. Soc. Bull., 1999, 78, 98–101. 37. Pastor, H., Metallic borides, preparation of solid bodies-sintering
33. Bellosi, A., Monteverde, F., Dalle Fabbriche, D. and Melandri, methods and properties of solid bodies. In Boron and Refractory
C., Microstructure and properties of ZrB2-based ceramics. Sub- Borides, ed. V. I. Matkovich. Springer, New York, 1977, pp. 457–
mitted to Materials Processing and Manufacturing Science. 493.
34. Anstis, G. R., Chantikul, P., Lawn, B. R. and Marshall, D. B., A 38. Zdaniewsky, W. A., Solid solubility effect on properties of tita-
critical evaluation of indentation techniques for measuring frac- nium diboride. J. Am. Ceram Soc., 1987, 70(11), 793–797.
ture toughness: I, direct crack measurements. J. Am. Ceram. Soc., 39. Lebugle, A. and Monte, G., Etude comparee de l’oxidation des
1981, 64, 533–538. diborured de zirconium, d’hafnium at de titane. Rev. Int. Htes
35. Munz, D. G., Shannon, J. L. and Bubsey, R. T., Fracture Temp. Rt Refract., 1974, II, 231–244.
toughness calculation from maximum load in four point bend 40. Berkes, J. S. and White, W. B., Phase relations in the sys-
tests of chevron notch specimens. Int. J. of Fracture, 1980, 16, tem Li2O.B2O3–B2O3-NiO. J. Am. Ceram. Soc., 1969, 52,
R137–R141. 481–484.