Slow crack propagation gives rise to intriguing patterns that encompass a very wide range of length scales from microns to kilometers. What is the underlying mechanism that is responsible for the diversity of scales and yet the similarity of morphology? What are the characteristics of slow fracture--does it happen in a catastrophe, like a first order phase transition, or gradually and smoothly like a continuous one? We attempt to answer these and similar questions by carrying out analytic and extensive numerical studies of simplified models using concepts and methods of statistical physics, along with simple experiments. | |||||||||
|
|||||||||
To address
the pattern formation aspect of quasistatic fracture, we proposed, analyzed,
and simulated a model system which describes the typical situation, namely the quasistatic crack formation in a brittle layer in contact with a substrate[1,2]. We identified the stick-slip action on the frictional substrate and crack propagation on the layer as the two competing, elementary processes, and developed our model based on close experimental observations, including our own. |
|||||||||
|
|||||||||
The model consists of particles interconnected with springs. Particles sit on a frictional substrate. When the net force on a particle exceeds the force threshold, the particle slips to an equilibrium position. When the tension on a spring exceeds the cracking threshold, it breaks irreversibly. Realistic crack patterns are generated upon iteration. | |||||||||
|
|||||||||
We obtained several novel results[1]: 1. The complementary nature of the effects of substrate
and thickness, via a scaling variable; These results suggest why morphologically similar patterns are so ubiquitous over a wide range of scales. In earlier works[2], we also explored the continuous transitions exhibited in quasistatic fracture by means of a coarse-grained spring-block model. On the two-dimensional parameter space spanned by the strain and substrate coupling, a phase transition line was located for the evolution of cracks and fragment-size statistics. The correlation of cracks was shown to crossover from short-range to long-range order across the transition, with power-law decay at the transition, in close analogy to critical phenomena. As similar transitions have been observed in recent fragmentation experiments, the results are not specific to the model we studied. This work served as the basis for later studies. The effect of disorder and limiting stress amplification
were studied analytically using the democratic fiber bundle model[3,4].
We discovered a novel tricritical rupture controlled by disorder, a meeting
place between abrupt and continuous behavior. With the model and simulation
of a model with local interaction, References
Last revised
October 10, 2002
©KtL
|