Systematic research into the phenomenon of propagating ruptures in gas pipelines began in the 1950s, stimulated by long brittle failures that occurred in the field. From a scientific point of view, this problem could be addressed fairly simply, based on an understanding of the fracture mode transitional behaviour of the pipeline material and its relationship to metallurgical variables. Such relationships were coming to be well-understood by the Sixties. The appearance of long ductile fractures shortly afterwards, however, required a much more sophisticated understanding of the interplay between the behaviour of the gas exhausting from the rupturing pipe and the fracture process in the pipe material. Several approaches were developed quite quickly that allowed the material requirements for ductile fracture arrest to be determined, for then-conventional design conditions. Over the last thirty years, as the operating conditions and dimensions of actual and planned gas pipelines have grown more demanding, the adaptation, improvement and even replacement of these ductile fracture models have been a continuing theme. In the future, as even more exacting conditions of pressure, temperature and gas composition are contemplated, together with still-higher levels of matertal strength, further developments will bneeded. This paper briefly reviews the historical evolution of our understanding of ductile fracture propagation control, considers the current state of knowledge, and aims to identify some critical issues for future gas pipeline projects. The most significant of these issues may be that, as operating pressures continue to rise, conventional wisdom regarding fracture arrest criteria is no longer applicable. As a result, approaches to fracture control that are based on empirical adjustments to existing models may no longer be viable.
展开▼