Recent Progress in Development of Ductile Fracture Arrest Methodology Based on CTOA: Test Standard, Transferability and Methodology

Abstract

Abstract Significant progress has been made in development of a new fracture arrest methodology based on a toughness parameter designed to characterize propagation — the crack-tip opening angle (CTOA). A CTOA test procedure using lab-scale DWTT-type specimens has been standardized by ASTM, and recently published experimental work has demonstrated transferability of CTOA from DWTT to full-scale pipe. This paper will present the basic methodology for determination of CTOA using DWTT-type specimens (i.e., ASTM E3039) and other specimens such as modified double-cantilever-beam (MDCB). Recent numerical studies using cohesive zone models (CZM) and others based on damage mechanics will be discussed, including models of full-scale pipe fracture. The effects on CTOA of loading rate, specimen flattening and constraint (bending vs. tension) will be reviewed. The effect on CTOA of loading rate between quasi-static and impact (covering five orders of magnitude) is small or negligible, being within experimental scatter. Observed differences between surface and mid-thickness CTOA values will be discussed. Models of DWTT specimens using damage mechanics have shown that the CTOA for tensile loading is the same at the surface and mid-thickness and equal to the mid-thickness value for bend loading, but that the surface CTOA is significantly larger than the mid-thickness CTOA in bending. Model calculations have revealed the dependence of crack velocity on stress for a given CTOA, enabling construction of fracture resistance curves (pressure required to propagate fracture as a function of crack velocity). These first-principles curves based on CTOA can then be used in the Battelle two-curve model (BTCM) to replace empirical resistance curves based on Charpy absorbed energy (Cv). It has been known for some time that Cv over-represents the propagation resistance for high-strength high-toughness steels, requiring empirical “correction factors” to Cv in the BTCM. Experiments have shown that there is a non-linear correlation between Cv and CTOA, explaining the need for correction factors to Cv and supporting the use of CTOA as a more appropriate propagation toughness.