Our objective is to develop atomic force microscopy-based methods to measure the structure of crack tips, the rates of crack growth, and the roughness of fracture surfaces at the nanometer scale in glasses. These measurements will enable more accurate predictions of the lifetimes of glasses used in optical, structural, and electronic components and devices, thereby enhancing the reliability of these components and devices.
Our approach is to fracture glass samples under controlled conditions in order to elucidate fracture mechanisms. Specimen geometries with well-defined fracture mechanics behavior are used to propagate cracks at controlled velocities ranging from 10-11 m s-1 to 102 m s-1. Fracture surfaces are then examined by atomic force microscopy (AFM). By comparing opposing areas on a given fracture surface, we can evaluate many aspects of the fracture process and crack growth: cavity formation at crack tips; sharpness of cracks that form; rate of corrosion of fracture surfaces nanometers away from arrested crack tips; and nature of the residue left on the fracture surface as a consequence of corrosion.
Impact and Customers:
Recent work on subcritical crack growth has concentrated on determining the behavior of cracks at very slow crack growth rates using AFM. We have studied crack tips near the fatigue limit (arrested crack motion) in soda-lime silicate glasses and found that: (1) ion exchange at crack tips forms very basic solutions that corrode the crack surfaces near the crack tip; (2) ion exchange sets up compressive stresses around the crack tip that assist in retarding crack growth; and (3) despite the corrosion, the crack tips remain sharp.
One of the common mechanisms of crack growth in metals and polymers is the formation and growth of cavities near the crack tip. Linkage of the cavities with the crack tip determines the rate-limiting step for crack growth. Recently, it has been suggested that this mechanism of crack growth also occurs in glass. In a study of crack growth at slow velocities, opposing fracture surfaces that formed during the crack growth process were compared to see if residual damage due to cavity formation could be observed on the fracture surfaces. We found that the opposing fracture surfaces matched to within 0.3 nm, suggesting that cavities did not form at crack tips in the glass.
The roughness of fracture surfaces was studied as a function of crack velocity. We observed that the roughness depended on the rate of crack growth, becoming smoother as the crack accelerated. Silica glass and soda-lime silicate glass behaved very differently. The dependence of the root mean square (RMS) roughness of silica glass on crack growth rate was much smaller than that of soda-lime silicate glass. In a theory developed to explain the data, the roughness was attributed to the basic structure of the glass, and consequent local fluctuations in structure and chemical composition. Silica glass, which has a much more uniform structure than soda-lime silicate glass, is therefore much smoother. A quantitative theory developed to explain the experimental results was based on elastic interactions between the crack tip and the local fluctuations with the glass. The theory was consistent with the data. Work is underway to determine the effect of glass microstructure on RMS roughness of glass fracture surfaces.