Towards a new standardized configuration for the coaxial double test for float glass

The macroscopic properties of float glass are governed by the opening of surface cracks in mode I. To bypass the influence of crack orientation and defectiveness of the borders (due to the cutting process), in the experimental measurement of the material strength the panacea would be to induce an equibiaxial state of stress in the core of the specimen. The Coaxial Double Ring (CDR) test achieves this ideal condition when geometric non-linearities are of minor importance. To compensate for second-order spurious components, ASTM C1499-09 indicates a CDR configuration with variable geometry according to the specimen thickness, whereas EN 1288-2 proposes the application of an additional overpressure. An analytic theoretical study of the non-linear effects in a CDR test is here presented. Assuming a Weibull statistical distribution of defects, for the CDR configuration with no overpressure, we obtain expressions in closed form for the effective area, a parameter that allows the re-scaling of the experimental data to a reference condition (equibiaxial stress on a unitary area) according to a criterion of equal failure probability. This method is used to propose a new standardized CDR testing method, with fixed geometry and no overpressure. Since the procedure of EN 1288-2 is proved to be ineffective because the induced stress state is not uniform and equibiaxial, a supplementary experimental campaign is reputed necessary, considering that the reference strength of glass in product standards has been determined with this testing method.