Low-cycle fatigue crack growth considerations in pipe fracture analyses

A recently completed program at Oak Ridge National Lab. (ORNL) for the Dept. of Energy involved conducting a pipe experiment for the New Production Heavy-Water Reactor. This paper presents some results from the ORNL cyclic pipe experiment that demonstrated the magnitude of low-cycle fatigue crack growth from simulated seismic loading. The concern with this result is that current pipe flaw evaluation procedures do not consider the cyclic crack growth during a seismic event, i.e., the peak dynamic load is used in a static analysis. This was the main incentive for the analyses developed in this paper. Additional applications of the methodology developed in this paper would be high stress--low cycle fatigue transients that can occur in service. In this paper the authors showed that the low-cycle fatigue crack growth data from the ORNL pipe test agreed well with the extrapolation of high cycle fatigue crack growth data from small specimens. The next step was to evaluate the accuracy of several different methods to predict the low-cycle fatigue crack growth. In this evaluation it was found that for low-cycle fatigue LEFM analysis would significantly underpredict the crack growth, and EPFM analyses were needed. Three different EPFM analyses were used. The GE/EPRI method was found to be overly conservative, which is consistent with comparison with past pipe fracture data. The LBB.NRC and LBB.ENG2 J-estimation schemes provided good predictions of the experimental crack growth data when the crack closure load was taken into account