Numerical Failure Pressure Prediction of Crack-in-Corrosion Defects in Natural Gas Transmission Pipelines
MetadataShow full item record
The aim of this study was to use the finite element method to model crack, corrosion, and Crack-in-Corrosion defects in a pipeline. The pipe material under investigation for this study was API 5L X60, 508 mm diameter with a wall thickness of 5.7 mm. The pipe material was evaluated using Tensile, Charpy, and J testing in order to model the defects and to establish the numerical failure criteria. Corrosion defects were modeled as flat-bottomed grooves. The collapse pressure was predicted when the deepest point in the bottom of the defect reached a critical stress. Based on this criterion, the FE corrosion failure pressure predictions were conservative compared to the experimental failure pressures, conducted by Hosseini , with an average error of 10.13%. For crack modeling, the failure criteria were established considering the plastic collapse limit and the fracture limit. Both the Von Mises stress in the crack ligament and the J-integral values around the crack were monitored to predict the failure pressure of the model. The crack modeling was done based on two approaches, the uniform depth profile and the semi-elliptical profile. The crack with uniform depth profile was done because the uniform shape is the logical equivalent shape for a colony of cracks. The crack with the semi-elliptical profile was done to have a less conservative results and because the experiments were done with semi-elliptical cracks. The FE crack modeling results were conservative compared to the experimental collapse pressure with an average error of 19.64% for the uniform depth profile and 5.35% for the semi-elliptical profile. In crack-in-corrosion (CIC) defect modeling, the crack was modeled with uniform depth because it was very difficult to model the semi-elliptical crack profile when the crack defect is coincident with a corrosion defect. The results were conservative compared to the experimental results with an average error of 22.18%. In general, the FE modeling provides the least conservative failure pressure prediction over the existing analytical solutions for pipe with longitudinal corrosion, crack, and CIC defects.