Glass ceramics: an innovative material for structural applications.

Glass-ceramic (GC) is mostly produced in two steps. First, a glass is formed by a glass manufacturing process; then, the glass is cooled down and further reheated in a second step. In this heat treatment glass partly crystallizes, coagulating into grains within the amorphous matrix. This underlying microstructure provides a fracture toughness one order of magnitude higher than that of glass, because crack opening is constrained by the bridging effect of the grains. Moreover, the heat treatments increase the performance at high temperature: whereas float glass begins to mollify at about 600°C, GC can withstand up to 900-950°C with no sensible decrease of stiffness. GCs can be made transparent or translucent, with an aesthetic finishing comparable to that of glass, but mechanical properties are better: the higher fracture toughness mitigates the brittleness, whereas the excellent response at high temperature increases resistance to fire. Here, we summarize results obtained in recent experimental campaigns for the characterization of a commercial type of GC for structural applications in buildings. In order to investigate the toughening mechanism, naturally pre-cracked GC specimens have been tested under 3-P bending to let the crack slowly propagate, while crack opening displacement (COD) was measured with the ESPI apparatus described in [1]. The ESPI arrangement, through the technique of phase shifting, is sensible to displacements of the order of 10-2 m, with a precision of 10-1 m. Processing the data, one finds that the COD profile has analogies with the cusp-like trend of a cohesive crack à la Barenblatt, confirming that cohesive forces, attributable to the bridging action of the crystalline-phase grains in GC, do act at the crack tip. Similar experiments have evidenced that this toughening mechanism is not present in annealed glass. In a successive investigation for the statistical characterization of the bending strength, the experimental campaign was composed of: (a) Ring on Ring biaxial bending tests (UNI-EN 1288-5) and (b) 4PB tests (UNI-EN 1288-5). Two batches of samples of thicknesses 6 and 8 mm were tested, each one composed of 50 and 20 specimens for test type (a) and (b), respectively (140 specimens in total). Classical Weibull probability distribution is shown to be able to accurately interpret the data. The strength of GC is comparable with that of tempered glass but, remarkably, it breaks into large pieces like annealed glass (this property is important for the post-breakage performance of laminated panels). The data dispersion is much lower than glass. To evaluate the effects of static fatigue (decrease of strength with time under dead load), following ASTM C1368 additional 4PB tests were performed at four different load rates (0.02, 0.2, 2, 20 MPa/s). A phenomenological model of equivalent-crack growth has been applied in order to characterize the initial equivalent defects of the material and make predictions of the decay of strength with loading time, calculating the corresponding nominal reduction of strength. As expected, GC results much less sensitive than glass to static fatigue. Higher strength with lower dispersion, higher fracture toughness, better resistance to high temperature, fragmentation into larger pieces: these are all properties that render GC much better than glass for what structural application is concerned. The cost of GC is in general higher than that of glass, but the improvement of production technology and the increase of market request may render, in a close future, its use in the building industry competitive with that of glass.