Critical Review of Strain Measures for Characterisation of Fatigue Damage in ASME Section III Fatigue Assessments

Rules for fatigue evaluation of nuclear pressure vessels and piping components are provided in Subsection NB of Section III of the ASME code. The code prescribed fatigue procedure requires the comparison of an alternating stress amplitude with fatigue allowables (design fatigue curves), usually derived through uniaxial specimen testing. For elastic assessments of multiaxial loading, typical from thermal shocks, a Tresca stress is used to characterise the stress field into a single effective stress measure for comparison with ASME fatigue allowables. For nonlinear elastic-plastic assessments, Appendix XIII-3440(b) of Section III specifies that “the numerically maximum principal total strain range” (interpreted as Maximum Total Principal (MTP) strain range) should be used for comparison with fatigue allowables. Two alternative methods for the characterisation of multiaxial strain fields are presented in the ASME code. Section VIII Division 2 provides alternative rules for the construction of pressure vessels, with Part 5 specifying the use of a Von Mises based Effective Strain Range (ESR) for elastic-plastic analysis. Section III Division 5 Subsection NBB provides rules for the assessment of components at elevated temperatures, also specifying the use of a Von Mises based Equivalent Total Strain Range (ETSR) measure. The two alternative strain measures are differentiated by their treatment of the elastic strain contribution. In the ESR method an equivalent elastic strain is calculated and summated with the plastic strain component. In the ETSR method the total strain (elastic plus plastic) is used thus evaluating the elastic and plastic contributions simultaneously. More complex critical plane approaches have also been proposed in recent years to better characterise multiaxial loading conditions. This paper presents a comparison of the various ASME specified strain measures and simplified critical plane approaches for fatigue evaluation of complex multiaxial loading. In support of this comparison, predictions of initiation lives to 0.254 mm defect in the stepped pipe specimen reported in PVP2004-2748 are provided to quantify the additional conservatism contained in elastic-plastic fatigue assessments of nuclear components. Predictions use the methodology presented in the companion paper PVP2019-93847 for the generation of short crack fatigue curves and the associated modification to environmental enhancement factors. It is concluded that use of the ASME specified strain measures, in conjunction with lower bound stress-strain data, conservatively underestimate the initiation life to a 0.254 mm defect by a factor of four for the example considered. However, use of more complex critical plane strain measures were observed to provide significant improvement in prediction accuracy of elastic-plastic fatigue evaluations.