Nondestructive Safety Evaluation of Thermally Tempered Glass

Abstract

Thermally tempered glass fails into small and blunt pieces upon fracture, and therefore is the product of choice when a degree of safety is required from a monolithic glass panel. The characteristic fracture pattern is the result of residual stresses in the glass that are induced during the tempering process and released during fracture. Previous research sought to correlate residual stress with the size of the fragments. In this paper, the two salient models (Barsom’s and Gulati’s) were described and validated against new experimental data. The experimental data were obtained by inducing fracture in specially prepared thermally tempered glass panels. The glass panels consisted of an unprecedented wide range of residual stress, obtained by a carefully executed thermal relaxation process. The residual stress of the specimens was measured with a scattered light polariscope that was calibrated by means of a specially devised experimental approach. It was shown that Barsom’s model is the more accurate of the two and is appropriate for predicting the fragmentation of glass with a surface residual compression larger than 60 MPa. The accuracy and precision of the scattered light polariscope were found to be ±4.7 MPa and ±3.9 MPa, respectively. The findings from this study provide a basis for a nondestructive safety evaluation of thermally tempered glass.