Triggering bulk flaws in glass

Abstract

Ongoing research at TU Delft focuses on recycling low-quality glass by casting it into volumetric elements, under the assumption that bulk flaws, and thus compatible contamination, have little influence on the volumetric component's strength. However, to validate the structural behavior of volumetric glass with significant bulk flaws, a corresponding uniform tensile testing method is needed that is not governed by the surface quality of glass and thus, can trigger bulk flaws. This research explores the applicability of the Theta-specimen (Durelli et al. 1962), as a method for measuring tensile strength of volumetric cast glass, while avoiding the main drawbacks of a direct uniaxial tensile test. Three theta-sample geometries are evaluated using FEA, and by mechanical testing of 4-5 CNC waterjet cut float glass specimens per geometry. Polarized light is used to visualize the development of stresses within each sample. Post-fracture fractographic analysis is performed to identify the origins of fracture, and to estimate failure stresses based of the fracture mirror radius using Orr's formula. The photo-elastic patterns closely match the FEA prediction, confirming that the largest tensile stresses occur within the intended central test strip. Polarized light reveals a sensitivity to eccentric loading for the two newly proposed designs, which can be minimized by introducing a neoprene interlayer between the sample and testing machine. All samples failed at multiple points; stress estimations based on fracture mirror size indicate that the lowest failure stresses consistently occurred within the test strip, further confirming that failure initiated within this area. It is concluded that the Theta-sample has potential as a uni-axial tensile testing method for brittle materials such as glass, though further research is required to fully confirm its reliability.