Evaluation of characteristic fracture toughness for Fitness-for-Service (FFS) assessments of structures containing crack-like flaws

The present study addresses an exploratory application of statistical-based approaches to determine a characteristic value of fracture toughness for use in fitness-for-service (FFS) procedures based on the minimum of three equivalent (MOTE) method and a conventional Weibull statistics to fracture toughness data. The work also focuses on applications of the Master Curve (MC) methodology to estimate lower bound fracture toughness values by direct evaluation of the reference temperature, $T_0$, and by utilizing some well-established $\mathrm{CV}N-T_0$ correlations. Fracture toughness tests using five different types of structural steels, including typical pressure vessel steels and an ultrahigh strength martensitic steel, provide the experimental data upon which the MOTE, the Weibull statistical analyses and the MC methodology are conducted. Overall, these exploratory analyses indicate that, while some existing correlations between Charpy CVN energy and $T_0$ can advantageously be used to estimate fracture toughness in routine assessments and engineering-level applications, they also have some limitations in predictive accuracy imposed by material dependency and their variability.