Contribution to the size effect on the strength of flat glass

Abstract

The American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE) has pubhshed a diagram for the required thicknesses of rectangular glass plates with simply supported edges under uniform load. This diagram is based on experimental fracture tests performed with samples of large glass plates homogeneously loaded. The average fracture load of a sample was divided by the safety coefficient 2.5 and the load so defined was regarded as permissible and safe. In the present paper this diagram is analyzed and the maximum principal tensile stresses are calculated for two selected load levels. For large plates (16 m2) of annealed glass the maximum permissible tensile stress was 20 N/mm2, for small plates (0.54 m2) 37 N/mm2. The difference between the two values can be explained quantitatively through the "size effect" on the strength of glass; however, the stress distribution being nonuniform these calculations must not be performed for the real surface areas of the plates, but for "effective surface areas" which can be defined through Weibull statistics. For tempered glass, too, the permissible maximum tensile stress depends on the size of the plate, but, as the permanent compressive surface stresses of the tempered glass plates tested are unknown, corresponding analysis of the results represented in the ASHRAE diagram [1] was not possible. The maximum principal tensile stress which was regarded as safe is approximately 55 N/mm2 for the largest plates of tempered glass used in practice.