The ductile fracture arrest capability of gas pipelines is seen as one of the most important factors in the future acceptance of new high strength pipeline steels for high pressure applications. It has been acknowledged for some time that the current methods for characterizing and predicting the arrest toughness for ductile fracture propagation in high strength steels are un-conservative. This observation is based on the inability of existing models to predict the required arrest toughness in full-scale ductile fracture propagation tests. While considerable effort is currently being applied to develop more accurate methods for predicting ductile facture arrest, the resulting models are still in a preliminary stage of development and are not immediately amenable for use by the general engineering community. As an interim solution, a number of authors have advocated the empirical adjustment or reformulation of the existing models for use with the newer, high strength pipe grades. While this approach does not address the fundamental issues surrounding the fracture arrest problem, it does provide methods that can be used in the near term for analysis and preliminary design. The desire to use these existing methods, however, is tempered by the uncertainty associated with their applicability in situations involving high pressures and/or high toughness materials. In an attempt to address some of these concerns, a statistical analysis was conducted to assess the accuracy of a number of available fracture arrest models by comparing predictions to actual values determined from full-scale fracture propagation experiments. From the results, correction factors were developed for determining the required toughness levels in high pressure applications that account for the uncertainty in the theoretical prediction methods.
Author: Wolodko, J. and Stephens, M.
Publisher: International Pipeline Conference, September 25-29, Calgary, Alberta
Purchase URL: Purchase this paper