Method of reproducibly predamaging float glass as a basis to determine the bending strength

Strength measurements on commercial float glass result in a wide range of randomly distributed values. Such a random statistical distribution leads, if extrapolated to low fracture probabilities, to unrealistically low strength values which cannot be taken as a basis for design. We introduce a method of reproducibly predamaging float glass. The glass surfaces which are under tensile stress are predamaged to an extent which is not achieved in normal practical applications. This predamaging reduces the bending strength and narrows the random range of measured values. The strength tests with the following statistical evaluation of the treated float glass specimens will more easily come up to the fracture behaviour of glass in practical use and can thus be adopted as a basis for the design of float glass according to a probabilistic approach