Several different criteria have been proposed over the years to predict the minimum toughness for arrest of an axial propagating crack for natural gas pipelines. The initial ones were empirically based. The Battelle Two-Curve Method (TCM) was subsequently developed and was somewhat less empirical. The TCM is still used frequently today. Nevertheless, all of these criteria use the Charpy energy as a measure of the material’s ductile fracture resistance. As higher-grade steels have been developed, it has been found from full-scale tests that a multiplier was needed on the predicted minimum Charpy arrest energy value as calculated from the original TCM. Several researchers have also suggested that a correction factor was needed on the Charpy energy as the Charpy energy value increased above a certain level. This was a nonlinear correction factor that essentially showed that as the Charpy energy value surpassed a certain level, the effective energy for ductile fracture arrest is less than the total energy from the Charpy test. This paper presents background information on several of these toughness correction factors, as well as statistical analyses of full-scale pipe burst tests on 186 lengths of X60 to X100 grade pipes using these methods. The results show the effects of grade level on not only the original TCM predictions, but also several other modifications for high Charpy energy levels. Additionally, a method has also been developed where the DWTT energy was used instead of the Charpy energy in the Battelle TCM. The results of the statistical analyses showed that all the Charpy-energy-based criteria required an increasing correction factor as the grade level increased. The one DWTT energy criterion was statistically constant with grade level. This difference between the Charpy criteria and the DWTT criterion was traced back to a changing relationship between the Charpy and DWTT energy values as the grade of the steel increases.

This content is only available via PDF.
You do not currently have access to this content.