Program Speaker: Bruce Bunker: Enviromentally-caused Fracture and Crack Growth in Glass.

Text by Stephen Attaway and Susan Wilson

Bruce Bunker of the Surface and Interface Science Department at Sandia National Laboratories presented an excellent talk for our May meeting on environmentally-caused fracture and crack growth in glass. Bruce investigates crack formation and propagation in glass in order to better understand environmental effects on trans-oceanic optical fiber cables used in telecommunications.

Bruce and his team at Sandia want to be able to predict the time dependence of crack growth, because this has the greatest effect on the viability of the fiber cable structure, aside from sharks using the fiber cables as dental floss. Bruce performed a series of experiments that proved how mechancal strain drastically enhances chemical reactions in silica glasses. His research showed how fracture behavior is drastically modified by water interacting with the glass atomic bonds, creating an atomic rupture mechanism.  Basically, he showed that water molecules distort the bonds between the silica atoms, greatly enhancing the cracking behavior and lessening the time to failure.

The researchers found that the very high strain associated with crack tips enhance the chemical reactions between silica and water. A household advice tip relates that a piece of glass is easier to break if one first scores the glass with a sharp metal point and then licks the crack (you add water). Cracks on car windshields also propagate faster during rainy weather. Bruce described how strain promotes formation of reactive intermediates, explaining how strain can change the reaction rate between silica and water by several orders of maginitude. By using infrared spectroscopy, the relative reaction rates can be detected on selected strained surface defects that had a known silica bond angle. By using computer models of strained atomic structures, he concluded that mechanical strain energy increases the rate of chemical reaction. He was able to predict crack velocity curves for silica under a variety of conditions.

Bruce discussed technical topics like how SI-O bonds rupture, the kinetics of crack tip and surface defect reactions, and the effects of crack tips geometry. He further described attempts to model fracture behavior by atomic bond rupture mechanism, which are summarized in an article he coauthored in Scientific American. Bruce was able to show that the molecule size is also important in understanding crack growth rates. Before water or some other fluid can interact with the strained bonds at a crack tip, it must first be able to diffuse into the crack. He discussed how size constraints allow small molecules, like water, do damage, while large molecules, like oils and alcohols, are not as able to penetrate to the crack tip as fast due to their larger size.