Prioritizing the Correct Crack-Like Feature for Remediation Requires Burst and Fatigue Calculations

Abstract Prioritization of crack-like features for remediation requires a burst pressure assessment as well as a fatigue life assessment to ensure all features are managed accurately. As the ability of in-line inspection technologies to provide complex feature profiles becomes more commonplace, a process for estimating the criticality of crack-like features with these profiles becomes more imperative. One approach is to apply an effective area method (e.g., similar to RSTRENG) to estimate criticality. This method evaluates all contiguous combinations of feature dimensions that can be defined via the detailed profile. The effective areas from these combinations are converted into semi-ellipses, or other idealized shapes, which can then be analyzed using fracture mechanics models. The lowest predicted burst pressure of these combinations is considered the burst pressure of the feature. As an extension, the fatigue life of each combination can also be estimated such that the effective dimensions resulting in the shortest predicted life define the overall fatigue life of the feature. While the effective area approach is mostly applicable to metal loss rather than crack-like features, this approach has been used in the past and its application can yield illustrative results. For some profiles, the effective dimensions corresponding to the lowest burst pressure are not the same as those corresponding to the shortest fatigue life. This points to a larger implication that can be applied to any set of features, whether they have simple or complex profiles. Specifically, the most critical feature based on burst pressure may be different from the feature with the shortest fatigue life. Different features could be prioritized for remediation differently depending on the failure condition being evaluated. A full understanding of which features to prioritize for remediation efforts therefore requires both a burst pressure assessment and a fatigue life assessment. This paper seeks to explore this implication in more detail by examining the ranges of assumptions and feature parameters under which this phenomenon occurs as well as the extent of its influence when using the MAT-8, Raju-Newman, and CorLAS™ fracture mechanics models.