Abstract
Engineering Critical Assessment (ECA) is commonly undertaken to derive the acceptance criteria for girth weld flaws in rigid pipelines deployed subsea by low-strain installation methods, such as S-Lay or J-Lay, or high-strain installation methods, such as Reel-Lay. The ECA generally considers the whole load history seen by the pipeline from fabrication to the end of service, and involves fracture and fatigue assessments. Fracture, which is the main focus of this paper, is deemed to have initiated when either (i) the crack driving force, expressed in terms of the J-integral or the Crack Tip Opening Displacement (CTOD), δ, is greater than the materials resistance, or (ii) the applied load exceeds the bearing capacity of the ligament of a cracked structure, also referred to as the plastic collapse or limit load. The robustness of the ECA procedure relies on the accuracy of the assessment solutions.
Most flaws in pipeline girth welds are embedded. Unlike surface breaking flaws, embedded flaws are typically not directly assessed in a high-strain fracture ECA because the available assessment solutions are too conservative. A work-around approach is often followed, where the maximum acceptable surface breaking flaw sizes are also considered acceptable below the surface if the embedment depth is equal to or greater than half of the flaw height. Otherwise, an embedded flaw must be reclassified as a surface breaking flaw with a height equal to the sum of the embedded flaw height and embedment depth.
To enable the direct fracture assessment of embedded flaws, the authors undertook in a previous work a parametric finite-element (FE) study on the effect of the embedment depth, the crack height and the crack length on the plastic collapse load of the shorter ligament of embedded flaws. Subsequently, a new limit load solution was proposed for the fracture assessment of embedded flaws in evenmatch pipeline girth welds subjected to tension and/or bending. This closed-form solution was shown to be significantly more accurate for estimating the crack driving force and the ligament plastic collapse load than other solutions available in the literature. For some geometries, however, the predicted limit load still needs to be significantly adjusted (increased) to correctly evaluate the J-integral, in a combined tearing and collapse assessment. This suggests that further enhancement of the solution is possible.
This paper describes small-scale fracture tests which were undertaken to determine the load required to collapse a smaller ligament of embedded flaws in a modified middle crack tension (MMCT) specimen. A closed-form solution, which can also be used as a flaw reclassification criterion, is fitted to the test results and then compared to the FE-based solution. Finally, recommendations are made for the direct fracture assessment of embedded flaws in evenmatch pipeline girth welds subjected to load or displacement-controlled conditions.
