Engineering & Technologyarticle2026-10-12

Numerical Investigation of Crack Driving Force and Ultimate Bearing Capacity of D1,422 mm X80 Pipelines with Composite Groove Girth Weld

Open access0 citations

Abstract

Existing prediction models for the ultimate bearing capacity of circumferential welds are primarily limited to manual or combined automatic welding with single V-grooves, failing to address full-automatic welding with composite grooves, and that is a critical technical gap for large-diameter, thick-walled X80 steel pipelines using automatic welding processes. To fill this gap, this study focused on cracked circumferential welds with composite grooves in full-automatic welded X80 pipelines, establishing a nonlinear finite element model that accurately accounts for the composite groove’s geometric characteristics, three material zones, and canoe-type crack simulation. Spider-web mesh refinement was applied at crack tips to ensure precision, and the model was validated using full-scale wide-plate tensile tests. Using J-integral as the crack driving force parameter, this study systematically analyzed the effects of nine key factors: load parameters, material properties, geometric parameters, and crack dimensions. The results revealed that residual stress had negligible impact on bearing capacity, while internal pressure (0→12 MPa) reduced capacity by 44.726%, misalignment (0→3 mm) by 52.982%, and crack depth (1→3 mm) by 47.451%. In contrast, increasing wall thickness (21.4→30.8 mm) enhanced capacity by 101.72%, and raising the weld strength matching coefficient from 0.8 (low-strength matching) to 1.0 (equal-strength matching) improved capacity by 96.234%. Finally, a ResNet18-based prediction model was built using a 10,935-case database from parallel calculations. This model enables rapid prediction of ultimate bearing capacity and provides a technical basis for the design and serviceability evaluation of such welds.

// Source

View paper (DOI)Open access versionOpenAlexProcessesPublished 2026-10-12

Authors: Yue Feng Yang, Han Zhang, Dong Zhang, Pengchao Chen, Rui Li, Jian Chen, Hong Zhang

Institutions: China University of Petroleum, Beijing