Excessive scatter in fracture toughness data for ferrous materials tested in the ductile-to-brittle transition region makes it difficult to develop conservative estimates of fracture toughness for use in structural analysis and design. Several procedures are developed to predict fracture toughness from scattered data sets. These procedures either rely solely on statistical analysis or else combine a correction for excess plasticity with statistical analysis.
A round robin testing program including 150 fracture toughness tests of SA508C13 in the ductile-to-brittle transition region has been conducted by the Materials Property Council (MPC) in conjunction with the Japan Society for the Promotion of Science (JSPS). This test program, which uses IT compact fracture specimens tested at three transition region temperatures, provides a large database for comparing procedures designed for dealing with data scatter. In this work, the Anderson-Dodds constraint correction procedure is compared to the available energy analysis procedure based upon their suitability for correcting for excess plasticity in scattered transition region fracture toughness data sets such as the MPC/JSPS round robin results. Various forms of two- and three-parameter Weibull statistics are reviewed for the ability to fit uncorrected and plasticity corrected data sets.