|Ph.D Student||Gleizer Anna|
|Subject||The Fundamental of Crack Initiation and Propagation in|
Brittle Single Crystals at the Low Energy
|Department||Department of Materials Science and Engineering||Supervisor||Dr. Dov Sherman|
The way cracks initiate and propagate in brittle crystals is still not well understood. Experiments and atomistic computations over the last decades have yielded crack initiation energy that suffers from large scatter. Atomistic simulations of cracked bodies predict crack initiation at energies much higher than the Griffith barrier, twice the surface energy, and invalidate crack propagation at low speeds; this is known as the lattice trapping effect. Environmental effects, although being taught, are not well identified due to the absence of accurate measurements of the fracture energy.
We studied, experimentally, the fundamentals of crack initiation and low speed propagation in brittle crystals under pure Mode I. Silicon crystal was used as a model material, and specimens cut from silicon wafer were cleaved using a recently developed coefficient of the thermal expansion mismatch method. The method is aiming at manipulating the energy flow to the crack tip. Extensive finite element analysis was employed to evaluate the quasi-static strain energy release rate. Confocal and Atomic Force Microscopes were used to characterize the fracture surfaces and to evaluate crack speed by the Wallner-lines method.
Cleavage experiments performed on the and low energy cleavage systems of silicon at room environment have yielded, respectively, 2.2±0.1 J/m2 and 2.7±0.3J/m2 cleavage energy at initiation, the lowest ever obtained for silicon. However, at atmospheric pressure of pure Argon (reduced oxygen) environment the obtained values were 2.9±0.3J/m2 and 3.5±0.2J/m2, respectively, in excellent agreement with density functional theory calculations. The reduced fracture energy value is the first direct evidence for stress corrosion cracking phenomenon in silicon crystal.
Our crack speed measurements and energy calculations showed that the Freund equation of motion well describes crack propagation at the low energy regime and suggests that twice the free surface energy, in brittle crystals, is the lower bound for the cleavage energy with no evidence for the large amount of the lattice trapping effect. The agreement between the experiments and the Freund equation of motion also suggests that surface features at low speed are consuming negligible energy, or alternatively, have minimal effect on the crack speed. Finally, we have managed to show that when the energy flow to the crack tip is comparable to 2γs, cracks in brittle crystals can propagate at extremely low speed. This is in contrast to the lattice trapping effect that was obtained in atomistic calculations, which stated that cracks cannot stably propagate at low speed.