An ESPI experimental study on the phenomenon of fracture in glass. Is it brittle or plastic?

The crack opening displacement (COD) in annealed soda-lime (float) glass has been measured with an electronic speckle pattern interferometry (ESPI) apparatus using coherent laser light. Specimens, naturally pre-cracked with a particular technique, were loaded under strain-driven bending until crack propagated; at regular intervals loading was paused to let the crack reach subcritical equilibrium and the COD measured. By using a post-processing algorithm comparing four images lighted with phase-shifted laser beams, surface displacements could be measured at a resolution of View the MathML source. Glass transparency has allowed to see through that the propagating crack front is not sharp but curved, jagged and merged in an opaque neighborhood. Numerical simulations show that the measured CODs cannot be reproduced if cohesive surface forces à la Barenblatt–Dugdale bridge the crack lips; instead a plastic-like region must form in a bulk neighborhood of the tip, where inelastic strains are associated with volume increase rather than deviatoric distortion. For this, a Gurson–Tvergaard model of porous plasticity, accounting for the formation of microvoids/microcracks, has been found more efficient than classical von Mises plasticity. This study confirms the formation at the crack tip of a process zone, whose occurrence in brittle materials like glass is still a subject of controversy.