W. H. Duckworth

Mixed-mode fracture in biaxial stress state: Application of the diametral-compression (Brazilian disk) test

Abstract Mixed-mode fracture of soda-lime glass was studied using a diametral-compression test that features disk specimens with symmetric through-cracks. The test enables one to study fracture under pure mode I loading, pure mode II loading, or any combination of mode I and mode II loading by a simple alignment of the crack relative to the diameter of compression loading. The disk specimens were precracked with the aid of both chevron notches and water-assisted subcritical crack growth. The directions of noncoplanar crack extensions and the relative magnitudes of mode I and mode II stress-intensity factors for mixed-mode fracture under inert conditions were compared to the predictions of three different mixed-mode fracture theories. None of the theories was completely adequate to explain the experimental observations, but a maximum hoop stress criterion modified to include second order, nonsingular term in the series solution for the crack-tip region stress gave reasonable agreement with the experimental results.

Analysis of R-curve behavior of non-phase-transforming ceramics

Abstract An analysis is presented of the phenomenon of increased crack resistance with crack growth ( R- curve behavior) exhibited by ceramics such as alumina, which do not undergo phase-transformation in the process and by ceramic-matrix composites. These materials contain a wake of unbroken ligaments and crack bridges, and these are modeled in this paper to account for R- curve behavior. The analysis is based on a modified Dugdale strip yield model. Various sizes and geometries of specimens are analysed, and a small-scale yielding estimation is also presented. Good correlation is obtained between predicted and experimental R- curve behavior. The model is also used to predict the dependence of saturation fracture toughness on grain size.

Effects of moisture-assisted slow crack growth on ceramic strength

An investigation was made of strength degradation caused by moisture-assisted slow crack growth of surface flaws in two dense polycrystalline ceramics. Specimen sizes were varied. The observed strength degradation of each ceramic was predictable usingKIC as a material property in fracture-mechanics relations, suggesting that the only material variable involved was critical flaw size. The strength of one of the ceramics in water decreased significantly with increased specimen size, but its Weibull modulus was essentially unaffected by specimen size and slow crack growth.

Estimating damage laws from bend-test data

Bend-test curves were calculated for materials obeying three very different damage laws. One result of the calculations is that the neutral axis migrates towards the compression face as the load-point displacement increases. For the three damage laws investigated, there is a simple relationship between the location of the neutral axis and the secant modulus derived from the bend-test record; this relationship is quite insensitive to the exact form of the damage law. The stress at a point 20 per cent of the beam depth from the tensile outer fibre is also fairly insensitive to the form of the damage law. Combining these two observations, it is possible to use bend-test data to make a good estimate of the tensile stress-strain curve of a material subject to damage.

EVALUATION OF THE STATISTICAL DISTRIBUTION OF FRACTURE STRENGTH: DISCUSSION OF SOME RECENT CONTRIBUTIONS.

Several recent statistical analyses of ceramic strength have concentrated on the form of the distribution function associated with a set of uniform specimens. It is pointed out that prediction of the effect of specimen size is an integral part of current statistical strength theory, and can be used to discriminate among alternative distribution functions.

Correlation of uniaxial and biaxial fracture strengths of ceramics

The increased utilization of brittle materials in structural applications has necessitated a need for better understanding of factors governing fracture strengths in multiaxial stress states. The research on which this paper is based examines the problem of relating fracture strengths of ceramics in biaxial stress states to strengths in uniaxial stress states. The problem is divided into two parts: in the first part fracture criteria for failure of brittle materials from single flaws are evaluated through controlled-flaw experiments in different stress states and fracture mechanics theory. The derived fracture criteria are then employed in a statistical fracture theory framework to predict strengths of ceramics failing from surface flaws.