One of the many aspects of natural gas pipeline design and material selection is the consideration of propagation and arrest of high-speed axial ductile fracture in the line pipes. Understanding the material ductile fracture behavior is essential for establishing an integrated fracture control plan. This is particularly important for pipelines of high design pressures utilizing large-diameter and high-grade line pipes. The procedure of Battelle Two-Curve Method (TCM) has been most commonly used in ductile fracture analysis in the prediction of fracture speed and minimum arrest toughness for axially running cracks. In the past decades, discussions and research have been in that the TCM approach, among with others, could not accurately predict either fracture speed or minimum arrest fracture toughness for high-grade pipes, and with pipe grade increasing the prediction errors are getting larger. Recent research work at TransCanada indicates that for a better prediction of pipeline ductile fracture, understanding the basic material mechanical behavior and its fundamental fracture mechanism is essential. One of the important findings of the work is that pipe material fracture toughness is not a constant as being commonly treated, rather the fracture toughness, in terms of both steady-state CTOA and steady-state DWTT fracture energy is fracture speed dependent, being decreasing with increasing fracture speed. Corresponding modifications have been made to the traditional TCM by introducing speed-dependent fracture toughness. The improved model gives much better predictions in both fracture speed and toughness for high grade pipes. This paper presents recent work at TransCanada, together with its industry partner Engineering Mechanics Corporation of Columbus (EMCC), on high-speed pipe-material fracture testing technique (using the modified back-slot DWTT specimen) and high-grade material testing data. The test data supports the predictions of early published work on speed-dependent fracture toughness. The fracture speeds obtained from the modified back-slot DWTT specimens were very close to actual full-scale pipeline ductile fracture speeds and this in turn enhanced the applicability of the modified TCM model.

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