Many properties of rapid fracture may profitably be studied in atomic scale computer simulations involving relatively small numbers of atoms. A first result of such a study is that qualitative properties of Mode III fracture change little when one explores various shapes of the interparticle potential, introduction of randomness, and elevated temperatures. A second result is that Mode I fracture is considerably more susceptible to instability than had previously been understood, and that to obtain stable Mode I fracture may require non-central forces between atoms.

One of the most interesting questions in the dynamics of brittle fracture is how a running brittle crack responds to an overload, i.e. to a mechanical energy release rate larger than that due to the increase in surface energy of the two cleavage surfaces. To address this question, dynamically running cracks in different crystal lattices are modelled atomistically under the condition of constant energy release rate. Stable crack propagation as well as the onset of crack tip instabilities are studied.

It will be shown that small overloads lead to stable crack propagation with steady state velocities which quickly reach the terminal velocity of about 0.4 of the Rayleigh wave speed upon increasing the overload. Further increasing the overload does not change the steady state velocity but significantly changes the energy dissipation process towards shock wave emission at the breaking of every single atomic bond. Eventually the perfectly brittle crack becomes unstable, which then leads to dislocation generation and to the production of cleavage steps. The onset of the instability as well as the terminal velocity are related to the non-linearity of the interatomic interaction.

Implementing molecular dynamics on the IBM SP2 parallel computer, we have studied the fracture of two-dimensional notched solids under tension using million atom systems. Brittle materials are modelled through the choice of interatomic potential function and the speed of the failure process, and our interest is to learn about the dynamics of crack propagation in ideal materials. Recent laboratory findings occur in our simulation experiments, one of the most intriguing is the dynamic instability of the crack tip as it approaches a fraction of the sound speed. A detailed comparison between laboratory and computer experiments is presented, and microscopic processes are identified. In particular, an explanation for the limiting velocity of the crack being significantly less than the theoretical limit is provided.

We present a theory for the morphology of the fracture surface left behind by slowly propagating cracks in linear, isotropic and homogeneous three dimensional solids. Our results are based on first order perturbation theory of the equations of elasticity for cracks whose shape is slightly perturbed from planar. For cracks propagating under pure type I loading we find that all perturbation modes are linearly stable, from which we can predict the roughness of the fracture surface induced by fluctuations in the material. We compare our results with the classical results for cracks propagating in two dimensional systems, and discuss the effects in the three dimensional analysis which result from taking into account contributions from non-singular terms of the stress field, as well as the effects arising from finite speeds of crack propagation.

We modeled the initiation of fracture in vitreous silica at various strain rates using molecular dynamics simulations. We avoided biasing the location for fracture initiation within the sample so that we could study the effects of dynamics and structure on determining the path to fracture, defined as the particular bonds that break during the course of fracture. We sought to show that the path to fracture would be primarily determined by the local variations in the structure of the vitreous phase at low strain rates, with diminished sensitivity on structural variations at higher strain rates. However, the results of our model indicate that the path to fracture is dependent not only on the initial structure of the system and the applied strain rate, but also on the initial phase of the thermal vibrations. This underscores the importance of atomic dynamics in determining the path to fracture in brittle materials and provides a justification for extending the analysis of fracture surfaces to the near-atomic scale.

The existing models for the “classical” Portevin-Le-Chatelier effect have been analyzed, and the non-linear dynamical model has been proposed in order to quantify the nature of temporal instabilities in fatigued metallic alloys. The model employs the concept of a positive feed-back among the populations of mobile, immobile and Cottrell-type dislocations with atmospheres of point defects. Three major types of loading have been numerically simulated: pure sinusoidal, creep fatigue (“the Lorenzo-Laird bursts”) and ramp loading (“the Neumann bursts”, when the amplitude of otherwise cyclic loading grows linearly with time). Computer movies of the temporal evolution of stress and dislocation densities have been prepared as an aide for analysis and illustration. The model successfully reproduces stress serrations in terms of the underlying dislocation mechanisms and thus for the first time establishes a fundamental link between the micro-and macromechanics of cyclic deformation.

The morphology of fracture surfaces in complex metallic alloys is analysed. The simultaneous use of Atomic Force Microscopy (AFM) and Scanning Electron Microscopy (SEM) allows the measurement of the universal roughness exponent ζ┴ = 0.78 over five decades of lengthscales (0.5 nm - 0.5 mm). Furthermore, a small lengthscales regime (lnm - 1 μm) is shown to be characterised by a roughness index ζ┴QS ≃ 0.5.

On the other hand, cracks fronts stopped during their propagation at pinning microstructural obstacles in two different metallic alloys is analysed, and their “in-plane” roughness index is determined for the first time. In the case of the 8090 Al-Li alloy, which has a very anisotropic microstructure, the roughness of the tensile crack front propagating along the grains length is equal to ζ ≃ 0.60. On the other hand, the roughness of the purely three-dimensional fatigue crack front in the Superα2 Ti3Al-based alloy is equal to ζ ≃ 0.54.

It has been recently observed that synthetic materials subjected to an external elastic stress give rise to scaling phenomena in the acoustic emission signal. Motivated by this experimental finding we develop a mesoscopic model in order to clarify the nature of this phenomenon. We model the synthetic material by an array of resistors with random failure thresholds. The failure of a resistor produces an decrease in the conductivity and a redistribution of the disorder. By increasing the applied voltage the system organizes itself in a stationary state. The acoustic emission signal is associated with the failure events. We find scaling behavior in the amplitude of these events and in the times between different events. The model allows us to study the geometrical and topological properties of the micro-fracturing process that drives the system to the self-organized stationary state.

An efficient image processing method is developed to determine the fractal dimension from a high resolution 3-D reconstruction of fracture surface. The third dimension of the fracture surface is obtained from a stereo pair. The elevation is locally determined through a recursive window matching algorithm centered on the cross-correlation operation. The method allows to reach high vertical and horizontal resolutions of the elevation map.

The fractal dimension is obtained by an implementation of the slit-island method. A series of contour lines is extracted from the reconstructed surface. The images of the contour lines are progressively reduced in size and at every step the curve length is estimated. The result is then normalized according to the scale factor to build a Kolmogorov plot.

The performance of this method is demonstrated on a synthetic image generated by a fracture surface mathematical model.

We present a nanometer scale description of the fracture surface of soda-lime glass. This is achieved by the use of Atomic Force Microscopy. The mirror zone is shown to be built with elementary entities, the density of which increases continuously while the mist and hackle zones are approached. Moreover, the overall picture leads to some kind of self-similarity, in the sense that small regions of the hackle zone exhibit the full set of mirror, mist and hackle areas.

The paper presents a review of recent results on the problem of size effect (or the scaling problem) in nonlinear fracture mechanics of quasibrittle materials and on the validity or recent claims that the observed size effect may be caused by the fractal nature of crack surfaces. The problem of scaling is approached through dimensional analysis and asymptotic matching. Large-size and small-size asymptotic expansions of the size effect on the nominal strength of structures are presented, considering not only specimens with large notches (or traction-free cracks) but also structures with no notches. Simple size effect formulas matching the required asymptotic properties are given. Regarding the fractal nature of crack surfaces, it is concluded that it cannot be the cause of the observed size effect.

We study a 3D random fuse network model with computer simulations. The breaking thresholds are distributed randomly, corresponding to quenched disorder. We find for the roughness exponent of the final fracture surface ζ = 0.47 ± 0.19, which is close both the minimum energy surface value and the directed percolation depinning model value in 2+1 dimensions. It is also similar to results from measurements of fracture surfaces at nanometer scale, and from experiments in which the fracture process occurs slowly as in fatique. The traditional measure of damage, the number of broken bonds grows faster than the area effect (nb ˜ L2.28), with no signs of a trivally brittle regime.

It is shown that a small ridge exists on the surface of the detachment fold or Schallamach's wave (1971). Such waves take place in the contact area between a moving transparent hemispherical asperity and the smooth, flat surface of a soft elastomer sample. The space-time evolution diagram, composed by juxtaposition of contact area cross-sections, recorded one after the other, proves that there is a correlation between the appearance of these ridges and microdynamic shifting of forward and backward detachment fold limits. A rolling of a rubber band, compressed between two rigid plates, enables us to reproduce the ridges in the macroscale approach.

It is suggested that perfluoropolyether lubricating oil coatings applied to the carbon overcoat film of magnetic recording layers become decomposed by electrons emitted from frictional surfaces. However, no work has at yet been reported as to triboemission of electrons from frictional carbon films.

This paper describes the behavior of triboemission of electrons and the friction coefficient during wear of sputtered hydrogenated carbon films (with various hydrogen contents on the glass substrate). The triboemission of electrons, together with friction coeficient, was measured in a frictional system of Al2O3 sliding on carbon films in a reduced dry air atmosphere. The worn surfaces of the carbon films were then observed using both a SEM and an AFM. The results showed that intense triboemission of electrons were observed during wear of hydrogenated carbon films. The electron emission intensity and friction coefficient transit from low to high with hydrogen content in the film. These results are discussed including physical properties of the carbon films such as internal stress and surface wettability.

Peeling delamination and image-layer fracture of a multilayer imaging film is analyzed by a simple beam approximation. Criteria for propagation of the delamination cracks and tensile failure of the imaging layer are established. The critical role of substrate strain in controlling tensile strain in the imaging layer is explored. For the case of a thin imaging layer with a fracture strain below a critical level, image layer fracture occurs while the leading-edge delaminations length is comparable to the layer thickness. Experimental Helios film image quality is optimized at a small level of substrate tensile strain where delamination length at leading and trailing edges of small holes is balanced and small and where hole size is minimized.