High-Temperature Failure of Polycrystalline Alumina: I, Crack Nucleation

A combination of techniques has been used to compare the compositional and microstructural characteristics of two commercially hot-pressed aluminas. Second-phase material containing Ni was observed in one material, mainly as small intergranular particles. Thermal treatment in air caused Ni to concentrate at the surfaces, in small precipitates. Large-grained heterogeneities, with Ti at the core, were identified in the same material. Hot-pressing flaws were observed in the second material. The large-scale heterogeneities act as crack nucleation sites during creep. Stress concentrations associated with these heterogeneities are considered to contribute to the premature crack nucleation, in conjunction with the presence of localized regions of amorphous second phase.     

High-Temperature Failure of Polycrystalline Alumina: III, Failure Times

High-temperature failure data have been obtained for two polycrystalline aluminas, revealing a strongly stress-dependent failure time. The observed stress dependence has been compared with the dependence predicted by models of the nucleation, growth, and coalescence stages of failure. Only models of the coalescence phase indicate sufficient nonlinearity to be a plausible explanation of the data. Further observations and measurements of coalescence, involving continuous crack nucleation and shear-band formation, are identified as requirements for further understanding of the rupture process.     

High-Temperature Creep of Polycrystalline Magnesia: II,Effects of Additives *

The creep of magnesia doped with 0.035 to 2.26 cation % of nine other oxides and three binary mixtures thereof and of three seawater products (about 96, 98, and 99.5y0 MgO) was evaluated in transverse bending at 1200° to 1500°C, with strain rates of about 10−²%/hr, and average grain sizes of 5 to 50p. The results obtained were compared with those for pure magnesia. Most additives accentuated the plastic (diffusion-controlled) nature of the creep process presumably by pinning dislocations and/or slowing grain growth. In most cases the rate-determining diffusing species seemed to be the cation, Mg, but in two cases it was suspected that oxygen boundary diffusion was controlling. Porosities above ˜10% appear to increase the temperature dependence of creep, probably by introducing boundary sliding. The agreement of the creep data with those of other diffusion-controlled processes (electrical conductivity, sintering, and grain growth) is demonstrated.     

High-Temperature Creep of Polycrystalline Magnesia: I,Effect of Simultaneous Grain Growth *

The creep of pure magnesia (99.9 +% MgO) was tested in transverse bending at temperatures from 1200° to 1500°C, strain rates near 10⁻²%/hr, and grain sizes of 4 to 50μ. In most cases, grain growth during the test affected the apparent creep behavior more than all the other variables combined. An analytical graphical method was used to separate the grain growth effect from other effects and to obtain more meaningful creep data. Creep occurred primarily by a viscous mechanism (Nabarro-Herring type, cation-lattice-diffusion controlling) with a minor amount of plastic creep (dislocation climb). The agreement with previous creep data was good.     

Analysis of Concurrent Grain Growth During Creep of Polycrystalline Alumina

During low-stress creep experiments in a fine-grained alumina, the strain rates decreased continuously with time due to the occurrence of concurrent grain growth. The grain-growth kinetics were independent of the applied stress and depended only on the time of exposure to the elevated test temperature. An analysis of the effect of concurrent grain growth was consistent with experimental observations.     

High-Temperature Fatigue Behavior of Polycrystalline Alumina

The strength and fatigue behavior of a 99.5% polycrystalline alumina were measured as a function of temperature. Both the strength and fatigue behavior remained essentially constant up to 500°C; from 800° to 1100°C the strength and fatigue resistance decreased markedly and at >1100°C macroscopic creep was observed. It is believed that the decrease in strength and fatigue resistance is caused by a grain-boundary glassy phase enhancing subcritical crack growth. Proof-testing at room temperature was effective in improving the strength distributions at both room temperature and 1000°C; however, at 1000°C it was not effective, due to crack growth during the proof test. The good agreement between proof-test results and fracture-mechanics theory indicates that the same flaws control the strength at room temperature and at high temperatures.     

High-Temperature Compressive Creep of Polycrystalline Mullite

Aluminosilicates of three compositions with mullite as the major phase were synthesized by a sol-gel process and characterized with bulk and microchemical analyses and microstructural observation. An apparatus for measuring the compressive creep up to 1900 K with a sensitivity of ±1 μm was constructed and used to measure the creep of singlephase mullite, mullite with second-phase glass, and mullite with second-phase corundum. Measurements in air at stresses of 15 to 100 MPa and temperatures of 1471 to 1724 K determined that samples with second-phase glass crept more rapidly than single-phase mullite or mullite with secondphase corundum. The apparent creep activation energies determined at 100 MPa were 742 kJ/mol for the mullite containing glass, 819 kJ/mol for the single-phase mullite, and 769 kJ/mol for the mullite with second-phase corundum. The stress exponents determined at 1724 K were 1.6 for the mullite plus glass, 1.5 for the single-phase mullite, and 1.2 for the mullite with α-Al₂O₃. The creep behavior of the aluminosilicates containing glass were consistent with rate control by the viscous flow of the glass and the measured creep rates were in good agreement with creep rates calculated from a model by Dryden. The creep behavior of the completely crystalline aluminosilicates was consistent with rate control by diffusional creep.     

High-Temperature Creep and Cavitation of Polycrystalline Aluminum Nitride

Dense, polycrystalline AlN samples of grain size between 1.8 and 19 μm were fabricated by hot-pressing. Bar-shaped samples were subjected to creep in four-point bending under static loads in nitrogen atmosphere. The outer fiber stress was varied between 20 and 140 MPa and the temperature between 1650 and 1940 K. Steady-state creep rate, dɛ/d t was proportional to σ ⁿ d ^{− m} where the stress exponent, n , was between 1.27 and 1.43 and grain-size exponent, m , between ∼ 2.2 and ∼ 2.4. The activation energy for creep ranged between 529 and 625 kJ/mol. Both round (r type) and wedgeshaped (w type) cavities were observed in electron micrographs of the deformed samples. No noticeable change in the dislocation density was observed. Contribution of cavitation to the creep rate was estimated using an unconstrained cavity model. Based on this study it is concluded that the dominant mechanism of creep in polycrystalline AlN is diffusional.     

Diffusional Crack Growth and Creep Rupture of Silicon Carbide Doped with Alumina

This study deals with tensile creep and crack growth behavior of silicon carbide doped with alumina at 1400°C. Excellent creep resistance was observed for stresses from 150 MPa to 200 MPa. From the creep exponent of 1.4 and the activation energy of 320 kj/mol, the principal creep mechanism was Coble creep. The creep failure was caused by slow crack growth from a preexisting flaw. The crack was found to grow subcritically along grain boundaries almost in isolation. The relation between the time–to–failure and the applied stress was well treated by a diffusive crack growth model, and the threshold stress of this material at 1400°C was estimated at 165 MPa.     

Duality in the Creep Rupture of a Polycrystalline Alumina

Creep rupture experiments performed on a polycrystalline alumina have revealed a duality in fracture behavior associated with a crack blunting threshold. At high stresses, or in the presence of large preexisting flaws, fracture is dictated by the slow creep growth of flaws and exhibits substantial statistical variability. At lower stresses, preexisting flaws blunt, and failure is delayed. In this regime, failure occurs by the coalescence of creep damage, manifested as shear bands, and the failure strain becomes inversely dependent on the applied stress.     

High-Temperature Creep and Dislocation Etch Pits in Polycrystalline Beryllium Oxide

Compressive creep of high-density polycrystalline beryllium oxide was investigated in the range 1850° to 2050°C. Creep rate was dependent on the applied stress to the 2.5 power, and the apparent activation energy for creep was 145 kcal/mole. Etch pit studies showed that the dislocation density in tested specimens was two orders of magnitude greater than that in assintered material. The diffusion process controlling creep was ascribed to volume diffusion of the anion. The deformation behavior was governed by dislocation motion.     

High-Temperature Creep Deformation of Alumina — SiC-Whisker Composites

The creep resistance at temperatures between 1200° and 1300°C in air of alumina—SiC-whisker composites was investigated via four-point flexure to examine (1) the effect of whisker content and (2) the influence of densification additives (i.e., Y₂O₃ (plus MgO)). The creep resistance of polycrystalline alumina is greatly improved with the addition of ≤ 20 vol% SiC whiskers. The interlocking/pinning of grains by whiskers which limits grain-boundary sliding contributes to the improvement in creep resistance. However, the creep rates of alumina composites in air increase at whisker contents ≥ 30 vol%. Electron microscopy observations suggested that the degradation in creep resistance for whisker content ≥ 30 vol% originated from (1) the promotion of creep cavitation and subsequent microcrack generation from the higher number density of nucleation sites and (2) more extensive formation of grain-boundary amorphous phase(s) associated with an observed increased oxidation rate. Along this one, the excellent creep resistance of alumina composites containing 20 vol% SiC whiskers was significantly degraded by the presence of the intergranular amorphous phases introduced by the addition of the Y₂O₃ densification additive.     

Damage and Fracture Mechanisms During High-Temperature Creep in Hot-Pressed Alumina

The mechanisms responsible for creep damage accumulation and fracture have been examined in two commercial hot-pressed aluminas. Differences between the two materials can be ascribed to minor compositional variations. Three damage regimes have been identified, depending on stress. However, in all three regimes, failure is controlled by crack propagation. At high stress, a single crack, nucleated at a processing flaw, controls failure. These cracks grow in a linear elastic stress field. At intermediate stresses, crack tip stresses relax, and many microcracks are nucleated. They grow and link under strain control. The details of this process differ under tension and bending, thus invalidating the flexure test as a means of establishing creep life, even in simple, single-phase materials. At the lowest stress, extensive cavitation, with relatively little microcrack development, is observed. However, failure continues to be dominated by the growth of cracks. The material is damage tolerant and can be thought of as superplastic. We find that processing flaws (primarily large grains) control the creep life at all stresses. These should therefore be carefully controlled in materials aimed at high-temperature structural applications.     

Crack Propagation Behavior of Polycrystalline Alumina under Static and Cyclic Load

Various causes for cyclic-loading fatigue in ceramics have been proposed. Degradation in the grain-bridging effect is the most important cause for cyclic-loading fatigue in nontransforming ceramics. Cyclic- and static-loading crack propagation behavior in terms of crack propagation rate, load—strain curve, and R -curve in compact tension specimens of polycrysalline aluminas with two types of average grain size is reported. Significant bridging is observed in coarse-grained alumina. The results are consistent with the proposal that grain bridging increases with grain size and that degradation in grain bridging is the most important cause for cyclic-loading fatigue in alumina ceramics.     

High-Temperature Creep Behavior of Polycrystalline SrZrO3

The high-temperature creep behavior of sintered polycrystalline SrZrO₃ containing 1.35 wt% Fe₂O₃ was investigated as a function of temperature, stress, grain size, and strain level over the ranges 1160° to 1275°C, 780 to 3110 psi, 0.45 to 2.04 μm, and 0.0014 to 0.014, respectively. A constant-load 4-point (pure bending) method was used to load the specimens. The creep rate is time-dependent, decreasing exponentially with strain, i.e.     , where the decay constant (β=118, measured at the 1560 psi stress level over the strain range 0.0014 to 0.014) is independent of temperature and grain size. No significant grain growth occurred during creep. The activation energy of 169±10 kcal/mol obtained for creep is relatively independent of temperature, stress, grain size, and strain level over the ranges investigated. The creep rate is directly proportional to the cube of the stress and the reciprocal of the grain size; this result is consistent with nonviscous creep theories based on dislocation generation and climb as the rate-controlling deformation mechanism.     

Crack Growth Resistance of Textured Alumina

The crack growth resistance of a textured, extruded alumina body was compared with that of an isotropic, isopressed body of similar grain size, density, and chemistry. R -curve levels reflected the preferred orientation; however, R -curve slopes ( dK _IR/ d Δ a ) were the same in all instances, implying a similar crack growth resistive mechanism. Three orthongonal orientations of crack growth in the two structures exhibited similar forms of K _IR versus Δ a curves, for which a schematic diagram for polycrystalline ceramics is proposed.     

Crack Blunting of High-Silica Glass

The mechanical strength of abraded high-silica glass was measured after immersion in water and silicic acid solution at room temperature and 88°C. The strength increase was observed at 88°C. This phenomenon is usually explained by crack blunting. From an observation of the dissolution rate of a porous glass with a similar composition and a comparison of the strength increase in water and silicic acid, it was concluded that the dissolution and subsequent precipitation of the high–silica glass is the mechanism of the crack blunting.     

Comparison of Slow Crack Growth Behavior in Alumina and SiC-Whisker-Reinforced Alumina

Results of dynamic fatigue experiments in water at room temperature on an indented Al₂O₃/25 wt% SiC whisker composite material have shown that this composite has a high resistance to slow crack growth. Aging tests in water revealed that the residual stress due to the indentation does not play an important role, and interrupted fatigue tests showed that the crack starts to grow at very low stress intensities, but the velocity is very low. Detailed fractography indicated that the slow crack growth path is intergranular in the whisker composite. The slow crack growth behavior in the whisker composite is discussed in association with the indentation residual stress, the change of the crack shape during the bending test. These results are compared with a bio-grade Al₂O₃ with lower resistance to slow crack growth, and important differences are pointed out.