Similitude experiments on projectile-induced fracture of monolithic glass

Abstract

Experiments were used to identify the II-groups that best scaled the fracture of monolithic plate glass under central impact by a projectile mass. Threshold impact velocities for glass fracture were measured for three groups of geometrically similar, laboratory-scale, rectangular specimens, for two types of edge constraints: (i) all edges clamped, and (ii) short edges clamped and long edges free. Different sets of Π-groups were needed for the two types of edge constraints, if the threshold impact velocities measured for the smallest specimens are to be used to predict those for the larger ones. Furthermore, edge flaws on the specimen face, caused by the diamond cutter used to size the specimens, and the orientation of the cut face to the oncoming projectile, affected the threshold impact velocities significantly. Extensive drop tests showed that, for other parameters remaining constant, the threshold projectile impact velocity is higher for a window glass whose previously cut surface faces toward, rather than away from, the oncoming projectile, regardless of the edge constraints.