Abstract
One of the key aspects in leak-before-break analyses is to predict the maximum load-carrying capacity of a circumferentially cracked pipe. Such analyses require the fracture resistance of the material using the J-integral parameter, typically using small-scale laboratory specimens, such as compact tension, C(T), or three-point bend specimens. To evaluate the similitude between the laboratory specimens and a circumferentially through-wall-cracked pipe, the toughness can be evaluated directly from the pipe using an analysis typically called an η-factor approach. The fracture resistance from the pipe tests can then be compared to laboratory specimen toughness values to assess similitude issues. Additionally, several analysis methods (i.e., LBB.NRC, LBB.ENG, LBB.GE, etc.) that predict maximum load capability of through-wall-cracked pipes have η-factor analyses embedded in them. Hence, the evaluation of the J-R curve accuracy or consistency with small-scale specimens is a verification of one step in such predictive analyses. This paper presents extensions to the earlier η-factor solutions for circumferentially through-wall-cracked pipes where the previous analyses were for cracks in pipes under either pure bending or pure tension. The improvements investigated account for loading under combined bending and tension due to internal pressure. The application of these methods to full-scale pipe tests is presented in Part II (Miura and Wilkowski, 1998) of this paper.