Calibration of the Statistical-Interference Factors for the Design of Tempered Structural Glass

Thermal tempering can induce a permanent residual prestress in glass, which is beneficial because it compensates for the low tensile strength of the material. The characteristic design resistance of tempered glass is usually fixed as the simple sum of the 5% quantiles of the statistical distributions of pristine glass strength and heat-induced surface prestress, but this definition neglects the positive contribution of the statistical interference between the two stochastic variables. Here, a verification formula for structural design is discussed, in which this effect is considered through a statistical-interference factor. Its calibration is now obtained by comparing level III (full probabilistic) and level I (semiprobabilistic) methods of design in paradigmatic case studies, illustrative for a wide range of applications, which differ for what concerns the size effect, the level of prestress, the duration of applied actions, and the class of consequence. The results show that consideration of the interference factor can lead to significant savings in terms of material consumption.