STATISTICAL-ANALYSIS OF THE ASME K-IC DATABASE

The American Society of Mechanical Engineers (ASME) K-Ic curve is a fu nction of test temperature (T) normalized to a reference nil-ductility temperature, RTNDT, namely, T - RTNDT. It was constructed as the lowe r boundary to the available K-Ic database. Being a lower bound to the unique but limited database, the ASME K-Ic curve concept does not disc uss probability matters. However, a continuing evolution of fracture m echanics advances has led to employment of the Weibull distribution fu nction to model the scatter of fracture toughness values in the transi tion range. The Weibull statistic/master curve approach was applied to analyze the current ASME K-Ic database. It is shown that the Weibull distribution function models the scatter in K-Ic data from different m aterials very well, while the temperature dependence is described by t he master curve. Probabilistic-based tolerance-bound curves are sugges ted to describe lower-bound K-Ic values.